Read ricketsnelder.pdf text version

JOURNAL OF VECTOR ECOLOGY

Volume 30 DECEMBER, 2005 Number 2

CONTENTS

Board of Directors ........................................................................................................................................... ii Guidelines for Contributors ........................................................................................................................... iii Advertising ....................................................................................................................................................... iv

Submitted Papers

Seasonal and geographical distribution of adult Ixodes scapularis Say (Acari: Ixodidae) in Louisiana Andrew Mackay and Lane Foil ............................................................................................................... 168 Tick infestations of the eastern cottontail rabbit (Sylvilagus floridanus) and small rodentia in northwest Alabama and implications for disease transmission Joseph C. Cooney, Willy Burgdorfer, Martin K. Painter, and Cynthia L. Russell ........................... 171 Field trial on the spatial repellency of metofluthrin-impregnated plastic strips for mosquitoes in shelters without walls (beruga) in Lombok, Indonesia Hitoshi Kawada, Yoshihide Maekawa, and Masahiro Takagi ............................................................ 181 Seasonal dynamics of four potential West Nile vector species in north-central Texas Bethany G. Bolling, James H. Kennedy, and Earl G. Zimmerman ..................................................... 186 Mosquito larvicidal activity of aqueous extracts of long pepper (Piper retrofractum Vahl) from Thailand Uruyakorn Chansang, Nayer S. Zahiri, Jaree Bansiddhi, Thidarat Boonruad, Pratom Thongsrirak, Jiranuch Mingmuang, Nipa Benjapong, and Mir S. Mulla .......................................... 195 Host plant selection of two Mansonia Blanchard species (Diptera: Culicidae) in a heterogeneous habitat of Buenos Aires City, Argentina Pablo R. Mulieri, Juan P. Torretta, and Nicolás Schweigmann .......................................................... 201 Variability in natural populations of Anopheles sacharovi (Diptera: Culicidae) from southeast Anatolia, revealed by morphometric and allozymic analyses H. Yurttas, B. Alten, and A. M. Aytekin ................................................................................................ 206 Evidence to support two conspecific cytological races of Anopheles aconitus in Thailand Anuluck Junkum, Narumon Komalamisra, Atchariya Jitpakdi, Narissara Jariyapan, Gi-Sik Min, Mi-Hyun Park, Kang-Hyun Cho, Pradya Somboon, Paul A Bates, and Wej Choochote ........ 213 Geographical distribution of Anopheles minimus species A and C in western Thailand Ampornpan Kengluecha, Pornpimol Rongnoparut, Soamrutai Boonsuepsakul, Ratana Sithiprasasna, Prinyada Rodpradit, and Visut Baimai ........................................................................ 225 Repellent effect of plant essential oils against Aedes albopictus Pin Yang and Yajun Ma .......................................................................................................................... 231 Evaluation of habitat management strategies for the reduction of malaria vectors in northern Belize John P. Grieco, Roy C. Vogtsberger, Nicole L. Achee, Errol Vanzie, Richard G. Andre, Donald R. Roberts, and Eliska Rejmankova ......................................................................................... 235 Methods for monitoring outdoor populations of house flies, Musca domestica L. (Diptera: Muscidae) Christopher J. Geden ............................................................................................................................... 244 Bionomics and distribution of species of Hystrichopsylla in Arizona and New Mexico, with a description of Hystrichopsylla dippiei obliqua, n. ssp. (Siphonaptera: Hystrichopsyllidae) Michael W. Hastriter and Glenn E. Haas .............................................................................................. 251 Abundance and diversity of human-biting flies (Diptera: Ceratopogonidae, Culicidae, Tabanidae, Simuliidae) around a nickel-copper smelter at Monchegorsk, northwestern Russia M. V. Kozlov, N. K. Brodskaya, A. Haarto, K. Kuusela, M. Schäfer, and V. Zverev ....................... 263 Contents continued on inside back cover

CONTENTS (continued)

An annotated checklist of the Anopheles mosquitoes (Diptera: Culicidae) in Iran

Mohammad Mehdi Sedaghat and Ralph E. Harbach ........................................................................... 272 Evaluation of the present dengue situation and control strategies against Aedes aegypti in Cebu City, Philippines Milagros M. Mahilum, Mario Ludwig, Minoo B. Madon, and Norbert Becker ................................ 277 Synergistic efficacy of botanical blends with and without synthetic insecticides against Aedes aegypti and Culex annulirostris mosquitoes Essam Abdel-Salam Shaalan, Deon Vahid Canyon, Mohamed Wagdy Faried Younes, Hoda AbdelWahab, and Abdel-Hamid Mansour ...................................................................................................... 284 Patterns of insecticide resistance in larval Culex pipiens populations in Israel: dynamics and trends Laor Orshan, Maria Kelbert, and Hedva Pener ................................................................................... 289 Effects of sub-lethal concentrations of synthetic insecticides and Callitris glaucophylla extracts on the development of Aedes aegypti Essam Abdel-Salam Shaalan, Deon Vahid Canyon, Mohamed Wagdy Faried Younes, Hoda Abdel-Wahab, and Abdel-Hamid Mansour ........................................................................................... 295 Chemical detection of the predator Notonecta irrorata by ovipositing Culex mosquitoes Leon Blaustein, Jonathan Blaustein, and Jonathan Chase .................................................................. 299 Chemical composition and anti-mosquito potential of rhizome extract and volatile oil derived from Curcuma aromatica against Aedes aegypti (Diptera: Culicidae) Wej Choochote, Dana Chaiyasit, Duangta Kanjanapothi, Eumporn Rattanachanpichai, Atchariya Jitpakdi, Benjawan Tuetun, and Benjawan Pitasawat ......................................................................... 302 Bartonella and Rickettsia in fleas and lice from mammals in South Carolina, U.S.A. Will K. Reeves, Mark P. Nelder, and James A. Korecki ...................................................................... 310 Seasonal abundance of horse flies (Diptera: Tabanidae) from two locations in eastern Croatia Stjepan Kr...mar ........................................................................................................................................ 316 ... Relative abundance of mosquito species (Diptera: Culicidae) on Big Pine Key, Florida, U.S.A. Lawrence J. Hribar .................................................................................................................................. 322 Laboratory estimation of degree-day developmental requirements of Phlebotomus papatasi (Diptera: Psychodidae) Ozge Erisoz Kasap and Bulent Alten ..................................................................................................... 328

Scientific Notes

Vertical distribution of adult mosquitoes in native forest in Auckland, New Zealand

José G. B. Derraik, Amy E. Snell, and David Slaney ............................................................................ 334

Outbreak of dengue in National Capital Territory of Delhi, India during 2003

R.S. Sharma, P.L. Joshi, K.N. Tiwari, Rakesh Katyal, and Kuldip Singh Gill .................................. 337 Molecular evidence for novel Bartonella species in Trichobius major (Diptera: Streblidae) and Cimex adjunctus (Hemiptera: Cimicidae) from two southeastern bat caves, U.S.A. Will K. Reeves, Amanda D. Loftis, Jeffery A. Gore, and Gregory A. Dasch ....................................... 339 A method for determining the sex of larval Aedes aegypti mosquitoes Gail M. Chambers .................................................................................................................................... 342 Culicidae), a primary vector of dengue in Thailand Theeraphap Chareonviriyaphap, Wannapa Suwonkerd, Piti Mongkalangoon, Nicole Achee, John Grieco, Bob Farlow, and Donald Roberts .............................................................................................. 344

The use of an experimental hut for evaluating the entering and exiting behavior of Aedes aegypti (Diptera:

First occurrence of Ochlerotatus japonicus in Missouri

Stephanie Gallitano, Leon Blaustein, and James Vonesh ..................................................................... 347

310

Journal of Vector Ecology

December 2005

Bartonella and Rickettsia in fleas and lice from mammals in South Carolina, U.S.A.

Will K. Reeves1 , Mark P. Nelder2, and James A. Korecki2

Centers for Disease Control and Prevention, Viral and Rickettsial Zoonoses Branch, Mailstop G-13, 1600 Clifton Rd. NE, Atlanta, GA 30333, U.S.A. 2 Clemson University, Department of Entomology, Soils, and Plant Sciences, 114 Long Hall, Clemson, SC 29634, U.S.A. Received 25 March 2005; Accepted 1 August 2005 ABSTRACT: Species in the genera Bartonella and Rickettsia are vector-borne pathogens of humans and domestic animals. The natural reservoirs and enzootic transmission cycles of these bacteria are poorly known in South Carolina. Thirteen species of lice and fleas were collected from urban animals and screened for the presence of Bartonella and Rickettsia by PCR amplification using genus-specific primers. Bartonella henselae was present in cat fleas (Ctenocephalides felis) from Virginia opossums (Didelphis virginiana) and a novel genotype of Bartonella was detected in Orchopeas howardi from an eastern gray squirrel (Sciurus carolinensis). We detected R. typhi and three novel genotypes Rickettsia in other species of fleas and lice. Rickettsia typhi, the causative agent of murine typhus, was detected in two pools of lice (Enderleinellus marmotae) from the woodchuck (Marmota monax). Cat fleas harbored one of two novel genotypes of Rickettsia. A third novel Rickettsia was detected in Orchopeas howardi from an eastern gray squirrel. Journal of Vector Ecology 30 (2): 310315. 2005. Keyword Index: Rickettsia, Bartonella, murine typhus, lice, fleas.

1

INTRODUCTION Lice (Phthiraptera) and fleas (Siphonaptera) are obligate, blood-feeding, ectoparasites of mammals and birds on all continents and most oceanic islands (e.g. Kim et al. 1986, Price et al. 2003). Flea-and louse-borne diseases, such as murine typhus, epidemic typhus, trench fever, and plague, have had a significant influence on human history. Modern antibiotics and vector control methods have dramatically reduced the morbidity and mortality associated with these diseases, but the etiologic agents of louse- and flea-borne diseases remain in the U.S.A. Recent cases of urban trench fever in American cities indicate that active transmission of Bartonella quintana still occurs (Spach et al. 1995). Sporadic cases of epidemic typhus, caused by Rickettsia prowazekii, occur in the United States when humans are exposed to infected southern flying squirrels, Glaucomys volans (Linnaeus), or their ectoparasites (Sonenshine et al. 1978). Emerging infectious diseases caused by previously unknown pathogens, such as the flea-borne pathogen Rickettsia felis, have been associated with typhus and dengue-like illness in humans worldwide (Williams et al. 1992, Azad et al. 1997). Other pathogens, such as Bartonella henselae, cause cat scratch disease in thousands of Americans annually (Kaplan et al. 2002). Vector-borne diseases have played a monumental role in the history of South Carolina by killing or debilitating thousands of citizens (Adler and Wills 2003). Species in the genera Rickettsia and Bartonella are primarily zoonotic pathogens with nonhuman, vertebrate reservoirs and arthropod vectors. Ectoparasites must be screened continuously to detect

new and emerging infectious diseases. We examined 13 species of fleas and lice from urban mammals of upstate South Carolina to determine if these ectoparasites harbored rickettsial agents. Previous studies have not focused on rickettsial agents in ectoparasites from this region of South Carolina. MATERIALS AND METHODS We collected fleas and lice from 24 animals and four animal nests in upstate South Carolina. The majority of the animals were killed by automobiles and collection techniques were reported by Nelder and Reeves (2005). Additional specimens were obtained from the Clemson University Arthropod Collection (CUAC). Collections were limited to the Piedmont and Foothills ecoregions of South Carolina in Aiken, Anderson, Cherokee, Greenville, Oconee, and Pickens counties (Table 1). Each ectoparasite was screened for DNA from Bartonella and Rickettsia by polymerase chain reaction (PCR) amplification. Individual fleas and lice or pools of lice (three per pool) (Table 1) were frozen in liquid nitrogen and crushed with a sterile Teflon pestle. Total DNA was extracted from the pulverized remains with an IsoQuick Nucleic Acid Extraction Kit (ORCA Research Inc., Bothell, WA) and resuspended in nuclease-free water. We detected DNA from Bartonella and Rickettsia spp. by PCR amplification using the BARTON-1 (52 -TAACCGATATTGGTTGTGTT GAAG-32) and BARTON-2 (52-TAAAGCTAGA AAGTCTGGCAACATAACG-32) and primer-1 (5'GCTCTTGCAACTTCTATGTT-3') and primer-2

Vol. 30, no. 2

Journal of Vector Ecology RESULTS AND DISCUSSION

311

(CATTGTTCGTCAGGTTGGCG-3') primers to amplify a fragment of the riboflavin synthase gene (ribC) of Bartonella and the 17 kD antigenic gene of Rickettsia, as described by Johnson et al. (2003) and Webb et al. (1990). We amplified the citrate synthase (gltA) and 16s rDNA genes from specimens with unidentified Rickettsia spp., using primers described by Roux et al. (1997) and the Rick16sF1 (3' GTATGCTTAACACATGCAAGTCGAAC 5') and Rick16sR4 (3' TCCGCGATTACTAGCGATTCC 5') primers. The PCR amplification cycle for the Rick16sF1 and Rick16sR4 primers consisted of an initial DNA degradation at 95o C for 5 min, followed by 30 s denaturation at 95o C, primer annealing at 55o C for 1 min, and extension at 72o C for 2 min. Samples were amplified for 45 consecutive cycles with a 10-min final extension at 72o C. All stock PCR and sequencing primers were at concentrations of 20 Mol. Each PCR tube contained 12.5 l of Taq PCR Master Mix Kit (Qiagen, Valencia, CA), 7.5 l of nuclease-free water, 1.25 l of each primer, and 2.5 l of DNA extract in water. PCR products were separated by 2% agarose gel electrophoresis and visualized under ultraviolet light with ethidium bromide. Positive and negative controls were used in all screens and consisted of genomic DNA of B. henselae, R. rickettsii, or distilled water. Products were purified with a QIAquick PCR Purification Kit (Qiagen, Valencia, CA). Duplicate sequencing reactions were performed with a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA) using PCR primers, and excess dye was removed with a DyeEx 2.0 column (Qiagen, Valencia, CA). Sequences were determined using an ABI 3100 capillary sequencer (Applied Biosystems, Foster City, CA). Primer sequences were removed and sequences assembled with Seqmerge (Accelrys, San Diego, CA). Assembled sequences were compared to those in GenBank using the BLAST 2.0 program (NCBI, Bethesda, MD). Identification of bacterial species was based on sequence similarity to known species. Novel sequences were determined to represent unique taxa. PCR products from the 17kD antigenic gene were digested with AluI for a restriction fragment length polymorphism (RFLP) pattern. Samples (5l) of amplified PCR products from Rickettsia sp. Rf2125 genotype, Rickettsia sp. Rf31 genotype, R. typhi, R. rickettsii, and R. felis were subjected to the restriction endonuclease AluI (New England Biolabs, Beverly, MA) digestion at 65o C for 6 h in 10 l volumes, as recommended by the manufacturer. Digested samples were separated by 4.5% agarose gel electrophoresis and visualized under ultraviolet light with ethidium bromide. Voucher specimens of each flea and louse species are deposited in the CUAC. Accession numbers for the DNA sequences are deposited in GenBank, with the following accession numbers ribC for Bartonella sp. Oh6 (AY953283), B. henselae (AY953284), 17 kD antigenic gene for R. typhi (AY867871), Rickettsia sp. Rf31 (AY953285), Rickettsia sp. Rf2125 (AY953286), and Rickettsia sp. Oh16 (AY953287), gltA from Rickettsia sp. Rf31 (AY953288) and Rickettsia sp. Rf2125 (AY953289), and the 16s rDNA from Rickettsia sp. Oh16 (AY953290).

We collected and screened a total of 42 lice and 38 fleas. DNA from Bartonella and Rickettsia were detected in 23 individual fleas and two pools of lice (Table 1). The prevalence of rickettsial agents ranged from 100% for three Rickettsia spp. in Ctenocephalides felis (Bouché) and pools of Enderleinellus marmotae Ferris, to no detectable rickettsial DNA in the remaining species of lice. PCR products from positive controls were present during all reactions, and PCR with the negative control amplified nothing. Species in the genus Bartonella are Gram-negative bacteria that infect erythrocytes of vertebrates and are putatively transmitted by hematophagous arthropods. There are at least 16 species or subspecies of Bartonella and at least nine of these are associated with disease in humans (Ciervo and Ciceroni 2004, Dehio et al. 2004). The arthropod vectors and transmission cycles of most Bartonella spp. are unknown. We detected two Bartonella spp. in fleas and none in lice. Bartonella henselae, the causative agent of cat scratch disease, has a worldwide distribution and is associated with domestic cats, Felis silvestris Schreber, and the cat flea, C. felis (Chomel et al. 1996). Experimental evidence implicates C. felis as the enzootic vector of B. henselae to cats (Chomel et al. 1996), but additional routes of transmission and vectors probably exist. Cat scratch disease is rarely fatal but more than 22,000 cases occur in the United States annually (Kaplan et al. 2002), making it one of the most common arthropodassociated diseases in the U.S.A. DNA from B. henselae was detected in 2/19 C. felis from Virginia opossums, Didelphis virginiana Kerr. The presence of B. henselae in wild-caught cat fleas implicates both C. felis and opossums in maintaining this pathogen. If cat fleas are competent vectors of B. henselae, they could serve as reservoirs and potential bridge vectors between wild mammals and pets or humans. Fleas collected from pets were negative for the ribC of Bartonella and were determined to be uninfected (Table 1). However, domestic animals such as pet cats and dogs might be exposed to fleas from peridomestic mammals such as the Virginia opossum. Orchopeas howardi (Baker) is mainly an ectoparasite of tree squirrels but over 53 arboreal animals including raccoons, opossums, squirrels, and eight species of birds have been recorded as hosts of this flea (Lewis 2000). The eastern gray squirrel, Sciurus carolinensis Gmelin, is naturally infected with Bartonella spp. and the vectors of these bacteria are suspected to be ectoparasitic arthropods (Bown et al. 2002). Kosoy et al. (2003) reported that some Bartonella of rodents are infectious to humans, but there is no evidence that the squirrel agents are pathogenic to humans or squirrels. Bartonella sp. Oh6 genotype was detected in a squirrel flea, O. howardi. The sequence of the PCR amplicon from this bacterium was 97% similar to Bartonella birtlesii. Bartonella birtlesii naturally infects mice, Apodemus spp., (Bermond et al. 2000) and might be closely related to the Bartonella spp. of squirrels. Durden et al. (2004) reported two novel Bartonella spp. from O. howardi collected from squirrels in southern Georgia but sequenced the citrate synthase gene

312

Journal of Vector Ecology

December 2005

Table 1. Fleas and lice screened for Rickettsia and Bartonella species in South Carolina, U.S.A., 1931-2004.

Number of arthropods examined 3 (1 pool) 4 Rickettsia spp. detected (prevalence) No No Bartonella spp. detected (prevalence) No No

Arthropod species Bovicola caprae (Gurlt) Ceratophyllus celsus Jordan

Ctenocephalides felis (Bouché)

1 4 5 1 1 4 3 6 (2 pools) 1 1 1 3 (1 pool) 3 (1 pool) 6 (2 pools) 1 1 11 3 3 (1 pool) 3 (1 pool) 3 (1 pool) 1 6 (2 pools)

Ctenocephalides felis

Collection data Pickens Co., Sixmile, 8 February 2003, ex Capra hircus Linnaeus Oconee Co., Seneca River Bridge, 12 July 2004, ex Petrochelidon pyrrhonota (Viellot) nest Pickens Co., Clemson, 9 June 2004, ex Didelphis virginiana Kerr Pickens Co., Clemson, 9 June 2004, ex D. virginiana Pickens Co., Clemson, 22 June 2004, ex D. virginiana Pickens Co., Clemson, 13 June 2004, ex Felis silvestris Schreber (domestic cat) Cherokee Co., Gaffney, 7 November 1993, ex Canis lupus Linnaeus (domestic dog) Pickens Co., Central, 30 November 2004, ex F. silvestris Pickens Co., Clemson, 22 December 2004, ex C. lupus Pickens Co., Clemson, 11 May 2004, ex Marmota monax (Linnaeus) Aiken Co., White Pond, 12 November 1963, ex Sus scrofa Linnaeus Pickens Co., Clemson, 18 April 1968, ex S. scrofa Pickens Co., Clemson, 10 January 1984, ex S. scrofa Pickens Co., Sixmile, 8 February 2003, ex C. hircus Pickens Co., Clemson, 1 January 1931, ex Bos taurus Linnaeus Anderson Co., Fants Grove, 10 May 2003, ex Sciurus carolinensis Gmelin Anderson Co., Pendelton, 15 May 2004, ex S. carolinensis Pickens Co., Clemson, 16 June 2004, ex S. carolinensis Greenville Co., Greenville, 21 September 2004, ex S. carolinensis Anderson Co., Anderson, 30 June 2004, ex S. carolinensis Greenville Co., Greenville, 12 September 1994, ex Homo sapiens Linnaeus Pickens Co., Sixmile, 20 May 1996, ex H. sapiens Greenville Co., Ninty Six, 1 January 1975, ex H. sapiens Oconee Co., Seneca, 19 February 2004, ex Mus musculus Linnaeus Pickens Co., Clemson, 5 July 2004, ex Procyon lotor (Linnaeus)

Ctenocephalides felis

Ctenocephalides felis

Ctenocephalides felis

Rickettsia sp. Rf2125 genotype (1/1) Rickettsia sp. Rf2125 genotype (4/4) Rickettsia sp. Rf2125 genotype (5/5) Rickettsia sp. Rf2125 genotype (1/1) Rickettsia sp. Rf31 genotype (1/1) Rickettsia sp. Rf2125 genotype (4/4) Rickettsia sp. Rf2125 genotype (3/3) Rickettsia typhi (2/2) No No No No No No No No No Rickettsia sp. Oh16 (1/3) No No No No No

Bartonella henselae (1/1) No

Bartonella henselae (1/5) No No No No No No No No No No No No No Bartonella sp. Oh6 near B. birtlesii (1/11) No No No No No No

Ctenocephalides felis

Ctenocephalides felis Enderleinellus marmotae Ferris Haematopinus suis (Linnaeus)

Haematopinus suis Haematopinus suis Linognathus africanus Kellogg and Paine Linognathus vituli (Linnaeus)

Neohaematopinus sciuri (Osborn) Neohaematopinus sciuri Orchopeas howardi (Baker) Orchopeas howardi

Orchopeas howardi Pediculus humanus capitus De Geer Pediculus humanus capitus Pediculus humanus humanus Linnaeus Polyplax serrata Burmeister

Stachiella octomaculatus (Paine)

Vol. 30, no. 2

Journal of Vector Ecology

313

rather than the riboflavin synthase genes. The flea harboring the Bartonella sp. Oh6 genotype was collected from a dead squirrel at a zoo and might have fed previously on other hosts. This novel agent was detected in 1/15 O. howardi and the role of fleas and squirrels in the natural transmission cycle of this agent is unknown. Four species in the genus Rickettsia, including R. typhi, were detected in lice and fleas. Rickettsia typhi, the causative agent of murine or endemic typhus, has a worldwide distribution and is endemic in the United States. Murine typhus was historically a widespread disease in the United States and epidemics of murine typhus in 1944 produced more than 5,338 cases (Love and Smith 1960, Mohr et al. 1953). Endemic foci, such as those in southern Texas, currently produce an average of about 50 cases per year (Boostrom et al. 2002). The oriental rat flea, Xenopsylla cheopis Rothschild, was considered the primary vector of R. typhi and maintains this pathogen by both vertical and horizontal transmission (Farhang-Azad et al. 1985). Lice are not considered significant vectors of R. typhi, but the human body louse, Pediculus humanus (Linnaeus), is an experimental vector in the laboratory (Houhamdi et al. 2003) and Hoplopleura pacifica Ewing might transmit the pathogen to rats (Traub et al. 1978). Neither woodchucks, Marmota monax (Linnaeus), nor their lice, E. marmotae, have been associated with R. typhi. Serologic data from M. monax in the northeastern United States were negative for R. typhi antibodies (Nicholson et al. 2003). DNA from R. typhi was detected by PCR in two pools of adult E. marmotae from a road-killed woodchuck. Based on a BLAST search, the 412 bp sequence for the 17 kD antigenic gene was a 100% match to that of the Wilmington strain of R. typhi (GenBank Accession# AE017197). The presence of DNA from R. typhi in lice does not prove that they are vectors

of this agent. The lice were either feeding on an infected host or acquired the pathogen transovarialy. Enderleinellus marmotae feeds exclusively on M. monax and does not bite humans. Several species of fleas, including Pulex irritans Baker, O. howardi, and Oropsylla arctomys (Baker), feed on M. monax (Palmer and Wingo 1972, Whitaker and Schmeltz 1973), but their roles in the transmission of R. typhi are unknown. Fleas were not reported from M. monax in the most recent checklist of fleas from South Carolina (Durden et al. 1999). Our discovery of R. typhi in lice demonstrates that this pathogen is present in urban animals in South Carolina. Three novel genotypes of Rickettsia were detected in fleas from South Carolina (Table 1). DNA from a novel Rickettsia sp. Oh16 genotype was detected in 1/15 O. howardi. Sequences of both the amplicons of the 16s rDNA and 17 kD antigenic genes were unique. The 17 kD gene amplicons were 97% similar to the Rickettsia "ARANHA" genotype (GenBank Accession# AY360215). A second genotype of Rickettsia was detected in a specimen of C. felis from Gaffney (Table 1). The gltA amplicon sequence from this Rickettsia was identical to the Rickettsia sp. Rf31 genotype (GenBank Accession# AF516331) reported from the Thailand-Myanmar border (Parola et al. 2003). All other C. felis in our study harbored an undescribed Rickettsia sp. with a gltA sequence that was identical to that of the Rickettsia sp. Rf2125 genotype (GenBank Accession# AF516333) from Ctenocephalides spp. collected at the Thailand-Myanmar border and reported by Parola et al. (2003). These Rickettsia spp. are not known to cause disease in humans. We are the first to report these bacteria from fleas outside of Asia, but the Rickettsia sp. Rf2125 genotype was also detected in fleas from Egypt (unpublished data). Azad et al. (1997) reported R. felis from fleas in California, Georgia, Tennessee, and Texas and based their identifications on the RFLP patterns produced by AluI

Figure 1. Restriction fragment length polymorphism pattern from the 17kD antigenic gene amplicons digested with AluI at 65oC for 6 h, separated by 5% agarose gel electrophoresis, and visualized under ultraviolet light with ethidium bromide. MWM: molecular weight marker with 300, 200, and 100 base pair bands labeled, 1: Rickettsia rickettsii, 2: R. typhi, 3:R. felis, 4: Rickettsia sp. Rf31 genotype, and 5: Rickettsia sp. Rf2125 genotype.

314

Journal of Vector Ecology

December 2005

digestion of the 17 kD antigenic gene amplicons. The RFLP patterns produced by AluI digestion of the 17 kD antigenic gene amplicons of Rickettsia sp. Rf2125 genotype, Rickettsia sp. Rf31 genotype, and R. felis are nearly identical and essentially indistinguishable (Figure 1). In addition, the Rickettsia detected in C. felis might be antigenically similar to R. felis and cross react when analyzed by an indirect fluorescent antibody test (IFA). We sequenced the PCR amplicons from the fleas in our study and determined that R. felis was not present. All of the cat fleas harbored novel Rickettsia. Contamination of the samples was unlikely because fleas from different collections were tested on separate dates, negative controls were never contaminated, and fleas from other locations in SC and GA were not infected with this agent. The public health implications of these new Rickettsia spp. are unknown. Ctenocephalides felis from pets harbored Rickettsia spp. and pet owners in South Carolina are therefore exposed to these agents. If these genotypes of Rickettsia are not pathogenic and exclude the establishment of R. typhi or R. felis in fleas, then the recent paucity of typhus-like illnesses in South Carolina could result from competitive interactions between species of Rickettsia. Rickettsial DNA was not detected in all of the specimens examined (Table 1), however some material from the CUAC harbored Rickettsial DNA indicating that they were appropriately preserved. Some ectoparasites might be refractory to rickettsial infection, could have fed on uninfected hosts, or were poorly preserved and any rickettsial DNA degraded prior to our study. Further research should clarify the roles of fleas and lice in the natural history of rickettsial pathogens in South Carolina. Acknowledgments We thank C.E. Beard, K.D. Cobb, M.A. MacCarroll, S.K. Kim, and S. Miller for their assistance with field collections of ectoparasites; P.H. Adler, G.A. Dasch, and J.C. Morse for allowing us to use laboratory space and equipment; A.D. Loftis for assisting with PCR primer development; and L.A. Durden for reviewing this manuscript and verifying identifications of E. marmotae. REFERENCES CITED Adler, P. H. and W. Wills. 2003. Legacy of death: the history of arthropod-borne human diseases in South Carolina. Am. Entomol. 49: 216-228. Azad, A.F., S. Radulovic, J.A. Higgins, B.H. Noden, and J.M. Troyer. 1997. Flea-borne rickettsioses: ecological considerations. Emerg. Infect. Dis. 3: 319-327. Bermond, D., R. Heller, F. Barrat, G. Delacour, C. Dehio, A. Alliot, H. Monteil, B. Chomel, H.J. Boulouis, and Y. Piemont. 2000. Bartonella birtlesii sp. nov., isolated from small mammals (Apodemus spp.). Int. J. Syst. Evol. Microbiol. 50: 1973-1979. Boostrom, A, M.S. Beier, J.A. Macaluso, K.R. Macaluso, D. Sprenger, J. Hayes, S. Radulovic, and A.F. Azad. 2002. Geographic association of Rickettsia felis-infected

Virginia opossums with human murine typhus, Texas. Emerg. Infect. Dis. 8: 549-554. Bown K.J., B.A. Ellis, R.J. Birtles, L.A. Durden, J. Lello, M. Begon, and M. Bennett. 2002. New world origins for haemoparasites infecting United Kingdom grey squirrels (Sciurus carolinensis), as revealed by phylogenetic analysis of Bartonella infecting squirrel populations in England and the United States. Epidemiol. Infect. 129: 647-653. Chomel, B.B., R.W. Kasten, K. Floyd-Hawkins, B. Chi, K. Yamamoto, J. Roberts-Wilson, A.N. Gurfield, R.C. Abbott, N.C. Pedersen, and J.E. Koehler. 1996. Experimental transmission of Bartonella henselae by the cat flea. J. Clin. Microbiol. 34: 1952-1965. Ciervo, A. and L. Ciceroni. 2004. Rapid detection and differentiation of Bartonella spp. by a single-run real time PCR. Mol. Cell. Probes 18: 307-312. Dehio, C., U. Sauder, and R. Hiestand. 2004. Isolation of Bartonella schoenbuchensis from Lipoptena cervi, a blood-sucking arthropod causing deer ked dermatitis. J. Clin. Microbiol. 42: 5320-5323. Durden, L.A., W. Wills, and K.L. Clark. 1999. The fleas (Siphonaptera) of South Carolina with an assessment of their vectorial importance. J. Vector Ecol. 24: 171-181. Durden, L.A., B.A. Ellis, C.W. Bank, J.D. Crowe, and J.H. Oliver, Jr. 2004. Ectoparasites of gray squirrels in two different habitats and screening of selected ectoparasites for bartonellae. J. Parasitol. 90: 485-489. Farhang-Azad, A., R. Traub, and S. Baqar. 1985. Transovarial transmission of murine typhus rickettsiae in Xenopsylla cheopis fleas. Science 227: 543-545. Houhamdi, L., Fournier P.E., Fang R., and Raoult D. 2003. An experimental model of human body louse infection with Rickettsia typhi. Ann. New York Acad. Sci. 990: 617-627. Johnson G., M. Ayers, S.C. McClure, S.E. Richardson, and R. Tellier. 2003. Detection and identification of Bartonella species pathogenic for humans by PCR amplification targeting the riboflavin synthase gene (ribC). J. Clin. Microbiol. 41: 1069-1072. Kaplan, S., J. Rawlings, C. Paddock, J. Childs, R. Regnery, and M. Reynolds. 2002. Cat-scratch disease in children - Texas, September 2000­August 2001. MMWR 51: 2647-2649. Kim, K.C., H.D. Pratt, and C.J. Stojanovich. 1986. The sucking lice of North America. Pennsylvania State University Press. University Park, Pennsylvania. U.S.A. 241 pp. Kosoy, M., M. Murray, R.D. Gilmore, Jr., Y. Bai, and K.L. Gage. 2003. Bartonella strains from ground squirrels are identical to Bartonella washoensis isolated from a human patient. J. Clin. Microbiol. 41: 645-650. Lewis, R.E. 2000. A taxanomic review of the North American genus Orchopeas Jordan, 1933 (Siphonaptera: Ceratophyllidae: Ceratophyllinae). J. Vector Ecol. 25: 164-189. Love, G.J. and W.W. Smith. 1960. Murine typhus investigations in southwestern Georgia. Publ. Hlth. Rep.

Vol. 30, no. 2

Journal of Vector Ecology

315

75: 109-120. Mohr, C.O., N.E. Good, and J.H. Schubert. 1953. Status of murine typhus infection in domestic rats in the United States, 1952, and relation of infection by oriental rat fleas. Am. J. Public Health 43: 1514-1522. Nelder, M.P. and W.K. Reeves. 2005. Ectoparasites of roadkilled vertebrates in northwestern South Carolina, U.S.A. Vet. Parasitol. 129: 313-322. Nicholson, W.L., D.J. Kuhar, J.G. Humphreys, and J.A. Childs. 2003. Serologic evidence for a novel Ehrlichia species in woodchucks (Marmota monax) from Pennsylvania, USA. Ann. New York Acad. Sci. 990: 90-93. Palmer, D.B. and C.W. Wingo. 1972. Siphonaptera occurring on Missouri mammals. Trans. Missouri Acad. Sci. 6: 4355. Parola, P., O.Y. Sanogo, K. Lerdthusnee, Z. Zeaiter, G. Chauvancy, J.P. Gonzalez, R.S. Miller, S.R. Telford III, C. Wongsrichanalai, and D. Raoult. 2003. Identification of Rickettsia spp. and Bartonella spp. in fleas from the Thai-Myanmar border. Ann. New York Acad. Sci. 990: 173-181. Price, R. D., R. A. Hellenthal, R. L. Palma, K. P. Johnson, and D. H. Clayton. 2003. The chewing lice: World checklist and biological overview. Illinois Natural History Survey Special Publication 24. 501 pp. Roux, V., E. Rydkina, M. Eremeeva, and D. Raoult. 1997. Citrate synthase gene comparison, a new tool for

phylogenetic analysis, and its application for the rickettsiae. Int. J. Syst. Evol. Microbio. 47: 252-61. Sonenshine, D.E., F.M. Bozeman, M.S. Williams, S.A. Masiello, D.P. Chadwick, N.I. Stocks, D.M. Lauer, and B.L. Elisberg. 1978. Epizootiology of epidemic typhus (Rickettsia prowazekii) in flying squirrels. Am. J. Trop. Med. Hyg. 27: 339-349. Spach, D.H., A.S. Kanter, M.J. Dougherty, A.M. Larson, M.B. Coyle, D.J. Brenner, B. Swaminathan, G.M. Mater, D.F. Welch, R.K. Root, and W.E. Stamm. 1995. Bartonella (Rochalimaea) quintana bacteremia in inner-city patients with chronic alcoholism. N. Engl. J. Med. 332: 424-428. Traub, R., C.L. Wisseman, and A. Farhang-Azad. 1978. The ecology of murine typhus- a critical review. Trop. Dis. Bull. 75: 237-317. Webb, L., M. Carl, D.C. Malloy, G.A. Dasch, and A.F. Azad. 1990. Detection of murine typhus infection in fleas by using the polymerase chain reaction. J. Clin. Microbiol. 28: 530-534. Whitaker, J.O., Jr. and L.L. Schmeltz. 1973. External parasites of the woodchuck, Marmota monax, in Indiana. Entomol. News 84: 69-72. Williams, S.G., J.B. Sacci, Jr., M.E. Schriefer, E.M. Andersen, K.K. Fujioka, F.J. Sorvillo, A.R. Barr, and A.F. Azad. 1992. Typhus and typhuslike rickettsiae associated with opossums and their fleas in Los Angeles County, California. J. Clin. Microbiol. 30: 1758-1762.

316

Journal of Vector Ecology

December 2005

Seasonal abundance of horse flies (Diptera: Tabanidae) from two locations in eastern Croatia

Stjepan Kr...mar

Department of Biology, Faculty of Philosophy, J.J. Strossmayer University, L. Jägera 9, HR-31000 Osijek, Croatia Received 12 May 2005; Accepted 31 July 2005 ABSTRACT: A total of 10,539 tabanid horse flies from 22 species and five genera was collected in the Tikves forest within the Kopa...ki rit Nature Park in eastern Croatia. Seasonal abundance was analyzed for the six most abundant species. Tabanus maculicornis, Tabanus tergestinus, and Haematopota pluvialis reached their highest peak abundance in the fourth week of June. Atylotus loewianus and Tabanus bromius reached their highest peak of abundance in the first week of August, whereas Tabanus sudeticus reached its maximum abundance in the third week of July. Horse flies also were collected once a week on the pasture at Petrijevci from mid-May to mid-September during 1993. Paired collections were made from a Malaise trap and from a horse by using a sweep net. A total of 2,867 tabanids belonging to 26 species was collected. The number of tabanids collected on horses was much higher than the total captured with Malaise traps. On their natural host (horse), 2.6 times more tabanids were collected than in the traps. Seasonal abundance was analyzed only for the eight most abundant species. Chrysops paralellogrammus, Tabanus autumnalis, Tabanus bromius, Tabanus tergestinus, Haematopota pluvialis, and Haematopota subcylindrica all reached their highest peak of abundance in the second week of July, whereas Tabanus maculicornis reached the maximal peak of abundance in the third week of June. Seasonal meteorological variability that occurs periodically from one year to another has a significant influence on the maximal peaks of tabanid abundance. Journal of Vector Ecology 30 (2): 316-321. 2005. Keyword Index: Diptera, Tabanidae, seasonal abundance, Croatia.

INTRODUCTION Horse flies (Diptera: Tabanidae) are known world-wide as important mechanical vectors of viruses, bacteria, protozoans, and helminths that cause diseases in wild and domestic animals (Krinsky 1976, Foil 1989). Emergence of the first specimens depends on the geographical latitude of the area (Chvála et al. 1972). In central Europe, the first species emerge in the second half of May, and they usually belong to the genus Hybomitra. The later appearance of tabanids in these regions is caused by the low temperature of the soil during spring months, which induces late pupation of larvae (Chvála et al. 1972). Because of this, there are often seasonal fluctuations between different populations of tabanids during the year. In Croatia, tabanids are most abundant during the summer season from the second half of June through the end of August (Chvála et al. 1972, Kr...mar 1999). Some of the most economically important species, such as Tabanus bromius and most of the Haematopota species, maintain their maximum populations to the end of August and individuals are often collected throughout the second half of September (Chvála et al. 1972, Kr...mar 2004). Despite the fact that the Kopa...ki rit Nature Park is home to wading birds and a multitude of insects, there are relatively little data about its Diptera fauna. Due to this lack of data, faunistic and ecological studies were initiated to contribute to the knowledge of biodiversity of the insect fauna in this part of Europe. The purpose of this study was to investigate the ecology of the most common tabanid species in the Kopa...ki rit Nature

Park and compare this area with a pasture of Petrijevci in eastern Croatia. The efficiency of traps and natural hosts in collecting the tabanids was also compared. MATERIALS AND METHODS Collections were carried out at two locations in eastern Croatia. The first collection site was the pasture in the village of Petrijevci (N 45° 37' 74" E 018° 29'59"), approximately 12 km NW of the town of Osijek. The pasture is situated between the Karasica and Drava rivers and comprises over 5 km2 . The pasture is in the form of a rectangle and is mostly covered with grasses except the edges which are overgrown with shrubby vegetation. Tabanids were continuously sampled by three homemade Malaise traps constructed according to the design of Townes (1962), and by a sweep net on horses. Malaise traps were baited with 2 kg of dry ice as an attractant. The attractant was always put under the linen cover of the trap next to the entrance into the collection cap. The first Malaise trap was set near a horse pound, and other two traps were placed on the open pasture adjacent to the river Karasica. Distance between traps was 200 m. A total of 22 samples was made over the period of May to September during 1993. Collections were conducted continually from 7 a.m. to 7 p.m. The other sampling site was located at Tikves forest (N 45° 41' 46" E 018° 49' 41") within the Kopa...ki rit Nature Park. Kopa...ki rit is situated in northeast Croatia in the corner made by the confluence of the Drava and Danube rivers. The Kopa...ki rit spreads over some 100 km2; it is oval-shaped, 8-

Vol. 30, no. 2

Journal of Vector Ecology

317

10 km long, and 7-9 km wide. The largest part of Kopa...ki rit lies between 81.5 and 82.5 m above sea level. In spite of its monotonous lowlands, the structure of Kopa...ki rit is very complex. Shallow oval or crescent-formed recesses and channels directly connect Kopa...ki rit with the Drava and Danube allowing flood waters to penetrate the marsh. The central part of the marsh is occupied by a large permanent water body, the Kopa...ko Lake (ca. 200-250 ha). The banks of the lake gradually become marshy meadows. In temporarily flooded areas, vast sections of the marsh are overgrown with associations of common reed and high sedges. The largest parts of the marsh are covered with white willow and black poplar forests. The Tikves forest is on higher terrain, with its largest parts covered in common oak with dyer's green weed, and a smaller part overgrown with common oak and hornbeam forests. Three times a week, tabanid collections were made continually from 7 a.m. to 7 p.m., from mid-May to midSeptember. Tabanids were collected by four gray linen Malaise traps constructed according to the design of Townes (1962) and also by four homemade black and white canopy traps, which were constructed according to the design of Hribar et al. (1991). These eight traps were placed in pairs (one Malaise and one canopy trap) along the lightly-shaded edges of the forest. Each pair was about 400 m apart from the next closest pair. The traps were baited either with aged horse urine, 1octen-3-ol, or a combination of aged horse urine and acetone. Each bait was used in two traps per day, which made a subtotal of six baited traps. Two unbaited traps served as control traps, making a total of eight traps per day. The baits were rotated among the traps daily, so that every bait was used at every trap site. Traps were baited separately with 4 ml 1-octen-3ol, 40 ml aged horse urine, and a combination of aged horse urine and acetone (10 ml aged horse urine to 30 ml acetone). A total of 60 samplings was made during the study. At both locations tabanids were killed in potassium cyanide bottles or in a small bottle charged with diethyl ether. All flies were preserved in ethanol. Identification and nomenclature followed that of Chvála et al. (1972) and Chvála (1988). During each sampling field trip, data for air temperature, relative humidity, and wind velocity were measured by means of a Kestrel 3000 environmental meter. The collection data from the pasture of Petrijevci were evaluated with a t test and by Kendall's "tau" coefficient (SPSS, Inc. 1998). RESULTS Effectiveness of traps and a natural host for the collection of tabanids A total of 2,867 tabanids belonging to six genera and 26 species was collected in the pasture at Petrijevci (Table 1). The total number of tabanids collected on horses in this study (2,069 specimens) was much higher than those captured with Malaise traps (798 specimens) baited with dry ice (Table 1). The results clearly express the difference between the use of traps and a natural host for sampling tabanids (t = - 2. 331, df = 25, P = 0.028). Also, the number of species was different for both methods (Table 1). The rank order of the eight most

common tabanid species captured in traps and on horses is significantly different ( = 0.327). Seasonal abundance of tabanids on the pasture in Petrijevci The seasonal abundance of the eight most abundant species is presented because these eight species made up 93.9% of the tabanid fauna on the pasture in Petrijevci. The other 18 species collected on the pasture represented only 6.1% of the total. The longest flight period from mid-May to early September was determined for the species Haematopota subcylindrica Pandellé 1883. Also, Haematopota subcylindrica was the most abundant with 42.4% followed by Tabanus bromius L., 1758 (16.6%), Chrysops paralellogrammus Zeller 1842 (8.2%), Haematopota pluvialis (L., 1758) (8.5%), Tabanus autumnalis L., 1761 (5.5%), Hybomitra ciureai (Séguy 1937) (5.3%), Tabanus maculicornis Zetterstedt 1842 (4.0%), and Tabanus tergestinus Egger 1859 (3.4%). Specimens of Chrysops paralellogrammus, Hybomitra ciureai, Tabanus autumnalis, Tabanus bromius, Haematopota subcylindrica, and Haematopota pluvialis reached their first peak of abundance in the fourth week of May, the second in the third week of June, the third peak in the second week of July, and a fourth smaller peak in the second week of August (Table 2). Tabanus maculicornis showed only one peak of abundance in the third week of June, whereas Tabanus tergestinus showed two peaks of abundance the first in the third week of June and another in the second week of July (Table 2). The largest number of specimens on the pasture in Petrijevci was collected during June and July (Table 2). Seasonal abundance of tabanids in the NP Kopa...ki rit ... A total of 10,539 tabanids, classified in 22 species were collected during the one-year study at the Kopa...ki rit Nature Park (Table 3). The four canopy traps collected 10,376 specimens; the other 163 specimens were collected by means of the four Malaise traps. Analyses of the trapping data for canopy traps showed that each of the attractants (1-octen-3ol, aged horse urine, and a combination of aged horse urine and acetone) significantly increased the number of tabanids collected, in comparison to the number collected in unbaited traps (Kr...mar et al. 2005). The most successful attractant was 1-octen-3-ol, followed by aged horse urine, and a combination of acetone and aged horse urine (Kr...mar et al. 2005). Tabanus was the most represented genus with six species and 9,157 specimens, followed by the genera Hybomitra with six species and 131 specimens, Haematopota with five species and 403 specimens, Chrysops with three species and 21 specimens, and Atylotus with two species and 827 specimens. Sixteen species had a relative abundance lower than 1%, which when combined, represented only 2.4% of the total catch (Table 3). However, the number of specimens belonging to the six most abundant species represented as much as 97.6% of the total of collected tabanids. The most abundant species were Tabanus bromius (53.4%), Tabanus sudeticus Zeller 1842 (21.6%), Atylotus loewianus (Villeneuve 1920) (7.8%), Tabanus tergestinus (5.7%),

318

Journal of Vector Ecology

December 2005

Table 1. Number of collected tabanids on a pasture of Petrijevci during 1993.

Species Haematopota subcylindrica Pandellé, 1883 Tabanus bromius L., 1758 Haematopota pluvialis (L., 1758) Chrysops parallelogrammus Zeller, 1842 Tabanus autumnalis L., 1761 Hybomitra ciureai (Séguy, 1937) Tabanus maculicornis Zetterstedt, 1842 Tabanus tergestinus Egger, 1859 Haematopota bigoti Gobert, 1880 Tabanus sudeticus Zeller, 1842 Chrysops viduatus (Fabricius, 1794) Tabanus bovinus L., 1758 Heptatoma pellucens (Fabricius, 1776) Atylotus loewianus (Villeneuve, 1920) Hybomitra ukrainica (Olsufjev, 1952) Chrysops caecutiens (L., 1758) Hybomitra muehlfeldi (Brauer, 1880) Chrysops relictus Meigen, 1820 Hybomitra pilosa (Loew, 1858) Atylotus rusticus (L., 1767) Hybomitra bimaculata (Macquart, 1826) Hybomitra solstitialis (Meigen, 1820) Haematopota italica Meigen, 1804 Hybomitra acuminata (Loew, 1858) Hybomitra n.confiformis Chvála et Moucha, 1971 Haematopota scutellata (Olsufjev, Moucha et Chvála, 1964) 26

Malaise traps 374 95 95 118 0 28 31 17 7 0 14 0 1 6 2 4 2 2 0 0 1 1 0 0 0 0 798

Horse 842 385 149 118 156 123 83 81 31 23 5 19 15 5 9 4 6 3 3 2 1 1 2 1 1 1 2,069

Table 2. Seasonal abundance of the most abundant tabanids on the pasture in Petrijevci during 1993.

Species/Month Week H. subcylindrica T. bromius H. pluvialis Ch.parallelogrammus T. autumnalis Hy. ciureai T.maculicornis T. tergestinus 3 48 1 3 1 -

May 4 157 36 25 25 27 31 16 1 110 31 22 22 7 31 36 2

June 2 7 1 1 1 1 1 3 114 94 44 22 19 28 45 24 4 64 17 22 8 18 2 10 3 1 120 13 12 11 1 4 1 2

July 2 469 183 79 85 43 22 5 61 4 37 64 14 44 19 11 2 1 5 1 2 1 1 3 -

August 2 77 40 21 17 17 18 3

September 1 8 3 1 -

Vol. 30, no. 2

Journal of Vector Ecology

319

Table 3. Number of collected tabanids in the NP Kopaèki rit during 2004.

Species Tabanus bromius L., 1758 Tabanus sudeticus Zeller, 1842 Atylotus loewianus (Villeneuve, 1920) Tabanus tergestinus Egger, 1859 Tabanus maculicornis Zetterstedt, 1842 Haematopota pluvialis (L., 1758) Hybomitra ciureai (Séguy, 1937) Haematopota subcylindrica Pandellé, 1883 Tabanus autumnalis L., 1761 Tabanus bovinus L., 1758 Hybomitra bimaculata (Macquart, 1826) Chrysops viduatus (Fabricius, 1794) Atylotus rusticus (L., 1767) Chrysops relictus Meigen, 1820 Haematopota italica Meigen, 1804 Hybomitra acuminata (Loew, 1858) Hybomitra ukrainica (Olsufjev, 1952) Haematopota pandazisi Kröber, 1936 Hybomitra muehlfeldi (Brauer, 1880) Chrysops caecutiens (L., 1758) Hybomitra solstitialis (Meigen, 1820) Haematopota ocelligera (Kröber, 1922) 22

Specimens 5627 2273 819 598 597 359 100 33 32 30 19 12 8 8 6 4 4 4 3 1 1 1 10,539

Relative abundance % 53.39 21.56 7.77 5.67 5.66 3.40 0.94 0.31 0.30 0.28 0.18 0.11 0.07 0.07 0.05 0.03 0.03 0.03 0.02 0.01 0.01 0.01 100

Table 4. Seasonal abundance of the most abundant tabanids in NP Kopaèki rit during 2004.

Month Species/Week T.bromius T.sudeticus A.loewianus T.tergestinus T.maculicornis H.pluvialis 1 1 4 2 74 6 30 4 June 3 322 4 69 123 31 4 468 102 147 269 130 1 231 134 3 35 68 36 July 2 468 333 75 115 86 37 3 842 838 135 116 15 50 4 289 210 46 15 6 7 1 1331 320 233 62 44 2 874 229 144 22 12 August 3 346 61 82 5 3 4 289 37 79 6 1 September 1 84 2 22 2 8 3 -

320

Journal of Vector Ecology

December 2005

Table 5. Recorded values of average monthly air temperatures (°C) for the study area.

Year/Month 1993 2004 Mean for the period 1992-2004

March 4.8 5.8 6.5

April 11.9 11.7 11.7

May 19.0 14.6 17.5

June 20.0 19.2 20.6

July 21.1 21.5 21.8

August 21.1 21.1 21.9

Tabanus maculicornis (5.7%), and Haematopota pluvialis (3.4%). Seasonal abundance was analyzed only for the six most abundant species (Table 4). The longest flight period was observed only for the Tabanus bromius, which extended from early June to mid-September, while other species had a shorter flight period (Table 4). The populations of these six most abundant species mostly fluctuated during the summer season. Specimens of Tabanus bromius, Tabanus tergestinus, and Haematopota pluvialis reached their first peak of abundance in the fourth week of June, the second in the third week of July, and third peak in the first week of August. Also, Atylotus loewianus and Tabanus sudeticus reached their peak of abundance in the same period of the year, the first peak of abundance in the third week of July and second in the first week of August. Females of Tabanus maculicornis reached their first peak of abundance in the fourth week of June and their second in the second week of July. The largest number of tabanids, 81.1%, was sampled during July and August (Table 4). Tabanids are not collected in May because of unfavorable climatic conditions (Table 5). The average monthly air temperature in May 2004 was the lowest in the last twelve years (Table 5). While the average monthly temperature in May 1993 was the second highest in the last twelve years (Table 5). Only in May 2003 was a higher average monthly air temperature recorded. In the weeks not mentioned in Tables 2 and 4, tabanids were not collected because of unfavorable weather conditions (shorter or longer rainy period). DISCUSSION Numerous studies have investigated the effects of meteorological factors on seasonal abundance of tabanids all over the world (Roberts 1971, Strickman and Hagan 1986, Leprince et al. 1991, Strickler and Walker 1993, McElligot and Lewis 1998, Kr...mar 1999, Barros 2001, Hribar et al. 2003). The studies of the seasonal abundance of tabanids are very important from the standpoint of medical and veterinary parasitology, because some species of tabanids participate in the transmission of agents of various diseases (Foil 1989, Vazzeille-Falcoz et al. 1997, Thomson and Connor 2000). Comparison of the seasonal abundance of the most abundant tabanids from two locations in eastern Croatia showed that meteorological factors are very important for the beginning of flight activity of tabanids (Table 5). The average monthly air temperature in May 2004 was 2.9°C lower than

the twelve-year mean for the study area and because of this, tabanids were not collected during this period. Unlike the last year, during 1993 the average monthly air temperature in May was 1.5°C higher than the twelve-year mean. This resulted in the first peak of abundance in May for the six most abundant tabanids. The peak of abundance for the tabanids during 1993 occurred one week earlier than during 2004. Generally, on the pasture in Petrijevci, the species Chrysops paralellogrammus, Tabanus autumnalis, Tabanus bromius, Tabanus tergestinus, Haematopota pluvialis, and Haematopota subcylindrica reached their highest peak of abundance in the second week of July. The same was observed for Tabanus bromius and Haematopota pluvialis in Poland and Russia (Olsufjev 1977, Trojan 1958, 1979, Trojan and Wojciechowska 1966). Also, it was established that Malaise traps baited with dry ice collected 2.59 times fewer tabanids than were collected by sampling nets on horses on the pasture in Petrijevci. The use of dry ice as a bait increases the catch of tabanids in traps. However, these results clearly express the difference between the use of traps and a horse for sampling tabanids. In the Tikves forest within the NP Kopa...ki rit, the highest peak of abundance was determined to be the fourth week of June for Tabanus maculicornis, Tabanus tergestinus, and Haematopota pluvialis. Atylotus loewianus and Tabanus bromius reached their highest peak of abundance in first week of August. These data for Tabanus bromius are identical with the seasonal abundance determined during 2002 in the Monjoros forest (Kr...mar 2004). Only Tabanus sudeticus reached its highest peak of abundance in the third week of July. The results of the analysis of seasonal dynamics of Tabanus maculicornis correspond with the data obtained from the European part of Russia (Olsufjev 1977). It appears that seasonal meteorological variability that occurs periodically from one year to another has a significant influence on the duration of tabanid flight activity, on their emergence, and on their peaks of abundance. For instance, when temperature conditions were more favorable during spring (May), higher numbers of tabanids were caught in the pasture habitat than in the wooded habitat (Tables 2, 4, and 5). Acknowledgments I thank Dr. Lawrence J. Hribar for his suggestions to improve this manuscript.

Vol. 30, no. 2 REFERENCES CITED

Journal of Vector Ecology

321

Barros, A.T.M. 2001. Seasonality and relative abundance of Tabanidae (Diptera) captured on horses in the Pantanal, Brazil. Mem. Inst. Oswaldo Cruz 96: 917-923. Chvála, M. 1988. Family Tabanidae. In: Á. Soós, (ed.), Catalogue of Palaearctic Diptera. Athericidae-Asilidae. Akadémiai kiadó, Budapest, 5: 97-171. Chvála, M., L. Lyneborg, and J. Moucha. 1972. The horse flies of Europe (Diptera, Tabanidae). Entomological Society of Copenhagen, Copenhagen, 499 pp. Foil, L.D. 1989. Tabanids as vectors of disease agents. Parasitol. Today 5: 88-96. Hribar, L.J., D.J. Leprince, and L.D. Foil. 1991. Design for a canopy trap for collecting horse flies (Diptera: Tabanidae). J. Am. Mosq. Contr. Assoc. 7: 657-659. Hribar, L.J., M.N. Hribar and D.J.Demay. 2003. Seasonal abundance of Diachlorus ferrugatus (Diptera: Tabanidae) in Monroe county, Florida. Fla. Sci. 66: 52-54. Kr...mar, S. 1999. Seasonal dynamics of horse flies in Eastern Croatia as a part of the Pannonian Plain (Diptera: Tabanidae). Period. Biol. 101: 221-228. Kr...mar, S. 2004. Ecological notes on Tabanus bromius L., and Haematopota pluvialis (L.), (Diptera: Tabanidae) of some flood areas in Croatian sections of the river Danube. J. Vector Ecol. 29: 376-378. Kr...mar, S., L.J. Hribar and M. Kopi. 2005. Response of Tabanidae (Diptera) to natural and synthetic olfactory attractants. J. Vector Ecol. 30: 133-136. Krinsky, W.L. 1976. Animal disease agents transmited by horse flies (Diptera: Tabanidae). J. Med. Entomol. 13: 225-275. Leprince, D.J., L.J. Hribar, R.T. Bessin and L.D. Foil. 1991. Seasonal patterns of abundance of horse flies (Diptera: Tabanidae) from two locations in southern Louisiana. Proc. La. Acad. Sci. 54: 10-18. McElligot, P.E.K. and D.J. Lewis. 1998. Seasonal changes in

abundance and gonotrophic age of host-seeking Tabanidae (Diptera) from a subarctic Labrador peatland. J. Med. Entomol. 35: 763-770. Olsufjev, N.G. 1977. Fauna CCCP. Nasekome dvukrilie, Slepni-Tabanidae. Akademia Nauk CCCP, Leningrad, 435 pp. Roberts, R.H. 1971. The seasonal appearance of Tabanidae as determined by Malaise trap collections. Mosquito News 31: 509-512. SPSS, Inc. 1998. Systat, Version 8. SPSS, Inc., Chicago. Strickler, J.D. and E.D. Walker. 1993. Seasonal abundance and species diversity of adult Tabanidae (Diptera) at Lake Lansing Park-North, Michigan. Great Lakes Entomol. 26: 107-112. Strickman, D. and D.V. Hagan. 1986. Seasonal and meteorological effects on activity of Chrysops variegatus (Diptera: Tabanidae) in Paraguay. J. Am. Mosq. Contr. Assoc. 2: 213-216. Thomson, M.C. and S. J. Connor. 2000. Environmental information systems for the control of arthropod vectors of disease. Med. Vet. Entomol. 3: 227-244. Townes, H. 1962. Design for a Malaise trap. Proc. Entomol. Soc. Wash. 64: 253-262. Trojan, P. 1958. The Ecological Niches of Certain Species of Horse Flies (Diptera, Tabanidae) in the Kampinos Forest near Warsaw. Ekologia Polska Seria A, Warszawa, 129 pp. Trojan, P. 1979. Tabanidae Slepaki (Insecta: Diptera). Polska Akademia Nauk, Instytut Zoologii, Warszawa, 308 pp. Trojan, P. and B. Wojciechowska. 1966. The specific distinction of Chrysozona pluvialis (L) and Ch. hispanica (Szil.) (Diptera, Tabanidae) in Poland. Ann. Zool. 23: 525-534. Vazzeille-Falcoz, M., C. Helias, F. Le Goff, F. Rodhain, and C. Chastel. 1997. Three spiroplasmas isolated from Haematopota sp. (Diptera: Tabanidae) in France. J. Med. Entomol. 34: 238-241.

322

Journal of Vector Ecology

December 2005

Relative abundance of mosquito species (Diptera: Culicidae) on Big Pine Key, Florida, U.S.A.

Lawrence J. Hribar

Florida Keys Mosquito Control District, 506 106th Street, Marathon, FL 33050, U.S.A. Received 19 May 2005; Accepted 26 July 2005 ABSTRACT: Species diversity of mosquito (Diptera: Culicidae) collections in carbon dioxide-baited light traps was investigated on Big Pine Key, Florida, from 2000 ­ 2004. Twenty species of mosquitoes were collected during this study, the most commonly collected being Anopheles atropos, Culex bahamensis, Deinocerites cancer, and Ochlerotatus taeniorhynchus. Ochlerotatus taeniorhynchus is the dominant species at this location. For most of the mosquito species, more individuals were collected during the "high season" months (June to September) than in the "low season" months (January to March). During most years more mosquito species were collected in the high season than in the low season. Rank order of species collected was significantly correlated between low and high season. Species composition at this site appears to be relatively stable from year to year and from season to season. Journal of Vector Ecology 30 (2): 322-327. 2005. Keyword Index: Mosquitoes, species composition, relative abundance, seasonal distribution.

INTRODUCTION Big Pine Key, one of the Florida Keys, is an island that lies south of the southernmost tip of the Florida peninsula and north of Cuba, between the Straits of Florida and the Gulf of Mexico. It is composed mainly of Miami limestone, with a small peninsula of Key Largo limestone on the southeastern part of the island (Snyder et al. 1990). The flora of the Florida Keys is derived from the West Indies (Stern and Brizicky 1957). Almost 300 species of higher plants have been recorded from Big Pine Key (Dickson et al. 1953). The Florida Keys Mosquito Control District has conducted mosquito control operations on this island since the late 1950s. In 1998, surveillance for adult mosquitoes via carbon dioxide-baited light traps became a significant component of these operations. Twenty-eight species of mosquitoes have been collected from Big Pine Key (Basham 1948, Darsie 2003, Darsie et al. 2002, Darsie and Shroyer 2004, Hribar 2002, Hribar et al. 2004, Pritchard et al. 1949). Only one of these studies (Hribar 2002) included any details of seasonal abundance of mosquito species on Big Pine Key. However, only the most abundant mosquito species were analyzed, viz., Ochlerotatus taeniorhynchus (Wiedemann), Deinocerites cancer Theobald, and Anopheles atropos Dyar and Knab. Although abundance of these three species changed during the course of the year, it is unknown whether the number of mosquito species changes throughout the year. This paper describes an attempt to answer that question. MATERIALS AND METHODS The study site is located at the northern terminus of Key Deer Boulevard on Big Pine Key, Florida, U.S.A. Trapping of mosquitoes has occurred almost continuously since late 1998. Dominant vegetation includes black mangroves (Avicennia germinans L.), red mangroves (Rhizophora

mangle L.), white mangroves (Laguncularia racemosa (L.) Gaertn. f.), buttonwood (Conocarpus erecta L.), saltwort (Batis maritima L.), and other succulents, halophytes, and salt tolerant grasses (Hribar 2002, 2005). Mosquitoes were collected by means of a carbon dioxidebaited light trap (American Biophysics Company, Jamestown, Rhode Island, U.S.A.). A trap, baited with dry ice, was set near the northern shoreline of Big Pine Key, in an area known to support large mosquito populations. A mosquito trap has been placed in the same site and hung from the same tree since late 1998. Distance from the ground to the light source was about 1.5 m. The trap was placed in the field at least once per week in the late afternoon and retrieved the following morning. All mosquitoes were transported to the laboratory, killed by freezing, and identified to species. In order to detect seasonal differences in mosquito species composition, data were analyzed as follows. Data from five years, 2000-2004, were used in the analyses. These years were the most complete in terms of continuous weekly collections of mosquitoes. Two periods of 13 consecutive weeks in each year were selected for analysis. The first 13 weeks of each year, when mosquito populations generally are low (Hribar 2002), was called the "low season." Another 13-week period in the middle of the year, weeks 24-36 (early June to early September) was called the "high season." This period was chosen because the largest number of mosquitoes is collected in traps during that time (Hribar 2002). One week's data, in the middle of the low season, were missing from the dataset; these were replaced by interpolation. One other week's data, at the beginning of one high season, also were missing from the dataset. In this case, the adjacent week's data were used. The maximum number of specimens per species collected each week was used in the analyses. Total number of mosquitoes collected and total number of species collected in low season and high season during each year were tabulated. The repeated measures ANOVA

Vol. 30, no. 2

Journal of Vector Ecology

323

test was used to detect differences between low season and high season mosquito numbers. Prior to performing this test, count data were transformed by X' = (X + 3/8)1/2 (Anscombe 1948) to stabilize variance and obtain a normal distribution. Untransformed means ± SEM are reported. Differences in rank order of species collected during low and high season were investigated via Kendall's statistic, which may be used as an index of community similarity (Ghent 1963). This statistic was calculated twice, once with all species included in the analysis, and again with "rare" species deleted from the analysis (Lounibos 1981). "Rare" species were considered to be those represented by less than 100 total specimens during the five-year collection period. Significance values for Kendall's statistic were obtained from Rohlf and Sokal (1981). The number of mosquitoes and the number of mosquito species collected during low season and high season in each year were plotted against year. The dominance index, d (Berger and Parker 1970), and Hill's second diversity number, N2 (Hill 1973), were calculated for low season and high season within years and also were plotted against year. Differences in total number of species collected and in d between low season and high season were investigated via use of a paired t-test. Statistics were calculated according to Ludwig and Reynolds (1988) or with the aid of SYSTAT computer software (SPSS, Inc. 1998). RESULTS Twenty species of mosquitoes were collected during this study, almost all of which were more abundant during the high season than during the low season. Those species that were collected more often during the low season, viz., Anopheles crucians Wiedemann, Culex erraticus Dyar and Knab, Cx. iolambdis Dyar, Cx. peccator Dyar and Knab, and Uranotaenia lowii (Theobald), were collected in very low numbers (Table 1). During the five-year study period, the mean number (± SE) of mosquitoes collected during high season (37,868.6 ± 10,221.7) was significantly greater than during low season (629.8 ± 197.0) (F = 34.169; df = 1,4; P = 0.004). Larger numbers were collected in the high season during all five years of the study (Figure 1). The mean number (± SE) of species collected was not significantly different between low season (7.4 ± 0.7) and high season (8.8 ± 0.9) (t = 1.1611, df = 4, P = 0.183) (Figure 2). When all years' data were combined and examined, the rank order of species collected in low season and high season was significantly different ( = 0.46, n = 20, P < 0.005, two-tailed test). After rare species were deleted from the calculation and only the most commonly collected species considered, rank order of species was still significantly discordant between low season and high season ( = 0.72, n = 9, P < 0.025), albeit less so than with all species included. This illustrates the disproportionate influence that rare species may have on calculations of community composition (Bullock 1971). The number of very abundant species collected did not change from low season to high season for most years (Figure 3). When N2 was recalculated using all years' data combined,

both low season and high season collections contained one very abundant species (low season N2 = 1.51; high season N2 = 1.07). Deinocerites cancer, Anopheles atropos, and Culex bahamensis Dyar and Knab also are among the most commonly collected mosquito species during both low season and high season. The increase in the value for N2 in 2004 is accounted for by the fact that more Cx. bahamensis and An. atropos were collected during low season in 2004 than in previous years, and more were collected in low season than in high season during 2004. Hill's second diversity number, N2, is an index of the number of very abundant species, and it approaches 1 as diversity decreases (Hill 1973). The increased number of individuals of Cx. bahamensis and An. atropos caused an increase in the value of N2 for low season in 2004. Bias due to sample size exists in diversity indices (Peet 1975), and given the unequal number of individual mosquitoes collected each year, comparisons among years may be suspect. However, N2 does appear to be stable over a wide range of sample sizes (Hill 1973), allaying these concerns somewhat. In all years and seasons, Oc. taeniorhynchus was the dominant species with its mean (± SE) low season d-value (0.79 ± 0.09) not differing significantly from high season (0.96 ± 0.02) (t = -1.927, df = 4, P = 0.126) (Figure 4). The dominance index (d) is independent of the number of species present (May 1975) and provides an indication of how large a proportion of individuals in a community are accounted for by the most numerous species. Even in low season 2004, when more Cx. bahamensis and An. atropos were collected than in previous years, Oc. taeniorhynchus was the most commonly collected mosquito species in all seasons and all years. DISCUSSION The removal of rare species from the calculation of Kendall's revealed that the rank order of the most commonly collected mosquito species at this site is more stable throughout the year than might be suspected when all species present are considered. However, even when rare species were removed from the analysis, there still remained a significant discord between low season and high season rank order of species. Lounibos (1981), in his study of African treehole mosquitoes, found that removal of rare species from calculations of Kendall's had one of two effects on the value of the statistic; either increasing the value of without increasing its significance level, or decreasing the value of . According to Bullock (1971), deleting the rare species has the effect of removing "chance vagrants" and when these are eliminated, patterns may begin to emerge among the more common species. This appears to be the case among the nine most commonly collected species on Big Pine Key, Florida. The data suggest that a few very commonly collected species dominate the mosquito fauna at this trap site in both low and high season, their relative abundances changing during the year, other species of minor importance coming and going as the year progresses. It should be borne in mind, though, that although traps are useful tools for surveying the mosquito species present at a collection site, they have their limitations. For example, Lounibos (1981) found that bamboo

324

Journal of Vector Ecology

December 2005

Table 1. Total mosquitoes collected by species for low season and high season, 2000-2004. Species marked with an asterisk were used for the second calculation of Kendall's statistic.

Species Anopheles atropos Dyar and Knab* An. crucians Wiedemann Culex bahamensis Dyar and Knab* Cx. declarator (Dyar and Knab) Cx. erraticus Dyar and Knab Cx. iolambdis Dyar Cx. nigripalpus Theobald* Cx. peccator Dyar and Knab Cx. quinquefasciatus Say Deinocerites cancer Theobald* Ochlerotatus atlanticus (Dyar and Knab) Oc. condolescens (Dyar and Knab)* Oc. infirmatus (Dyar and Knab)* Oc. sollicitans (Walker) Oc. taeniorhynchus (Wiedemann)* Oc. tortilis (Theobald)* Psorophora ciliata (Fabricius)* Ps. columbiae (Dyar and Knab) Ps. johnstonii (Grabham) Uranotaenia lowii (Theobald) Total

Low Season 193 7 207 0 4 3 8 7 0 158 0 11 0 0 2,545 6 1 0 0 1 3,152

High Season 1,006 6 1,067 1 1 0 237 0 2 1,788 1 1,426 131 82 183,229 253 207 6 4 0 189,445

Vol. 30, no. 2

Total Mosquitoes Collected

Journal of Vector Ecology

325

100000 10000 1000 100 10 1 2000 2001 2002 2003 2004

Year

Low Season High Season

Figure 1. Total individual mosquitoes collected on Big Pine Key in low season and high season, 2000-2004.

12

Number of Species

10 8 6 4 2 0 2000 2001 2002 2003 2004

Year Low Season High Season

Figure 2. Total mosquito species collected on Big Pine Key in low season and high season, 2000-2004.

Hill's N2

3.5 3 2.5 2 1.5 1 0.5 0 2000 2001 2002 2003 2004

Year

Low Season High Season

Figure 3. Number of very abundant mosquito species (Hill's N2) collected on Big Pine Key in low season and high season, 2000-2004.

326

1.2

Journal of Vector Ecology

December 2005

Dominance (d)

1 0.8 0.6 0.4 0.2 0 2000 2001 2002 2003 2004

Year

Low Season High Season

Figure 4. Dominance index (d) for Ochlerotatus taeniorhynchus in low season and high season, 2000-2004.

traps might not give an accurate indication of the relative abundance of mosquito species at a trap site. Use of dry ice as a bait may also be a concern. Service (1993) cautions that it is difficult to control the release rate of gas sublimating from dry ice. One possible result is a deterrent effect on mosquitoes, depending on concentration of carbon dioxide gas and air temperature near the trap. On the other hand, consistent use of traps over a period of years may allow for detection of changes in relative abundance of those species collected by traps (Milby and Reeves 1986). This information is valuable for tracking abundance of disease vectors and for making operational control decisions. By whatever measure is used, be it raw number of individuals per species, the dominance index (d), or Hill's N2, Oc. taeniorhynchus is the most abundant species at this collection site. Although other species may occasionally account for a greater proportion of the total number of mosquitoes during low season, during high season Oc. taeniorhynchus is far and away the most abundant mosquito species on Big Pine Key. Acknowledgments I thank L. P. Lounibos, University of Florida, for his invaluable comments on earlier drafts of this manuscript. REFERENCES CITED Anscombe, F.J. 1948. The transformation of Poisson, binomial, and negative binomial data. Biometrika 35: 246-254. Basham, E.H. 1948. Culex (Melanoconion) mulrennani, a new species from Florida (Diptera: Culicidae). Ann. Entomol. Soc. Am. 41: 1-7. Berger, W.H. and F.L. Parker. 1970. Diversity of planktonic Foraminifera in deep-sea sediments. Science 168: 13451347. Bullock, J.A. 1971. The investigation of samples containing many species. II. Sample comparison. Biol. J. Linn. Soc. 3: 23-56. Darsie, R.F., Jr. 2003. First report of Ochlerotatus

condolescens (Dyar and Knab) (Diptera: Culicidae) in the United States. Proc. Entomol. Soc. Wash. 105: 10671068. Darsie, R.F., Jr., J.J. Vlach, and E.M. Fussell. 2002. New addition to the mosquito fauna of the United States, Anopheles grabhamii (Diptera: Culicidae). J. Med. Entomol. 39: 430-431. Darsie, R.F., Jr. and D.A. Shroyer. 2004. Culex (Culex) declarator, a mosquito species new to Florida. J. Am. Mosq. Contr. Assoc. 20: 224-227. Dickson, J.D., III, R.O. Woodbury, and T.R. Alexander. 1953. Check list of flora of Big Pine Key, Florida and surrounding keys. Q. J. Florida Acad. Sci. 16: 181-197. Ghent, A.W. 1963. Kendall's "tau" coefficient as an indicator of similarity in comparisons of plant and animal communities. Can. Entomol. 95: 568-575. Hill, M.O. 1973. Diversity and evenness: a unifying notation and its consequences. Ecology 54: 427-432. Hribar, L.J. 2002. Mosquito (Diptera: Culicidae) collections in the Florida Keys, Monroe County, Florida, USA. Studia Dipterol. 9: 679-691. Hribar, L.J. 2005. Records and observations for some Diptera in the Florida Keys. Florida Sci. 68: 109-113. Hribar, L.J., J.J. Vlach, D.J. DeMay, S.S. James, J.S. Fahey, and E.M. Fussell. 2004. Mosquito larvae (Culicidae) and other Diptera associated with containers, storm drains, and sewage treatment plants in the Florida Keys, Monroe County, Florida. Fla. Entomol. 87: 199-203. Lounibos, L.P. 1981. Habitat segregation among African treehole mosquitoes. Ecol. Entomol. 6: 129-154. Ludwig, J.A. and J.F. Reynolds. 1988. Statistical ecology: A primer on methods and computing. Wiley-Interscience, N.Y. 337 pp. May, R.M. 1975. Patterns of species abundance and diversity. In: M.L. Cody and J. Diamond (eds.). Ecology and Evolution of Communities, pp. 81-120. Belknap Press, Cambridge, Massachussets. Milby, M.M. and W.C. Reeves. 1986. Changes in the relative abundance of Aedes nigromaculis, Aedes melanimon and Culex tarsalis in the central valley of California. Proc.

Vol. 30, no. 2

Journal of Vector Ecology

327

Calif. Mosq. Vector Contr. Assoc. 54: 96-100. Peet, R.K. 1975. Relative diversity indices. Ecology 56: 496498. Pritchard, A.E., E.L. Seabrook, and J.A. Mulrennan. 1949. The mosquitoes of the Florida Keys. Florida Entomol. 30: 8-14. Rohlf, F.J. and R.R. Sokal. 1981. Statistical tables. W.H. Freeman, San Francisco. 253 pp. Service, M.W. 1993. Mosquito ecology: Field sampling

methods. Elsevier Applied Science, London. 988 pp. Snyder, J.R., A. Herndon, and W.B. Robertson, Jr. 1990. South Florida Rockland. In: R.L. Myers and J.J. Ewel (eds.). Ecosystems of Florida, pp. 230-277. Univ. Centr. Fla. Press, Orlando. SPSS, Inc. 1998. SYSTAT Version 8.0. SPSS Inc., Chicago. Stern, W.L. and G.K. Brizicky. 1957. The woods and flora of the Florida Keys. Introduction. Tropical Woods 107: 3665.

Vol. 30, no. 2

Journal of Vector Ecology

181

Field trial on the spatial repellency of metofluthrin-impregnated plastic strips for mosquitoes in shelters without walls (beruga) in Lombok, Indonesia

Hitoshi Kawada, Yoshihide Maekawa, and Masahiro Takagi

Department of Vector Ecology and Environment, Institute of Tropical Medicine, Nagasaki University, Nagasaki, Nagasaki 852-8523, Japan Received 3 November 2005; Accepted 5 April 2005 ABSTRACT: Field trials on the spatial repellency of metofluthrin-impregnated plastic strips for mosquitoes present in shelters without walls (beruga) were carried out in Lombok, Indonesia. A major reduction in the incidence of human biting by Culex quinquefasciatus was achieved, and the use of two strips per beruga repelled >60% of the mosquitoes for at least 11 wk while four strips repelled >60% of the mosquitoes for more than 15 weeks. The technique was found to be a practical long-term solution for the prevention of mosquito bites without using electricity or heat to evaporate the metofluthrin. Journal of Vector Ecology 30 (2): 181-185. 2005. Keyword Index: Metofluthrin, plastic strip, spatial repellency, Culex quinquefasciatus. INTRODUCTION Metofluthrin, (2,3,5,6-tetrafluoro-4-methoxymethylbenzy1 (E,Z)(1R,3R)-2,2-dimethyl-3-(prop-1-enyl) cyclopropanecarboxylate) (S-1264) is a newly-synthesized pyrethroid produced by Sumitomo Chemical Co. Ltd., Osaka, Japan (Ujihara et al. 2004). Metofluthrin belongs to the relatively safe pyrethroid group (Shono et al. 2004) and has already been registered in several Asian countries such as Singapore, Indonesia, Myanmar, and Vietnam. We reported in our previous papers that metofluthrin impregnated into multilayer paper strips is a promising spatial repellent for mosquitoes under laboratory and field conditions (Kawada et al. 2004a, 2004b). It was proven that mosquitoes were repelled by airborne metofluthrin vapor for 4 weeks in simulated outdoor conditions (Kawada et al. 2004a). The field tests also indicated the effectiveness of metofluthrin against mosquito bites in shelters without walls (beruga), which people in Lombok Island, Indonesia, use for resting, praying, and evening conversations with neighbors. During such occasions there is a high risk of transmission of malaria at a beruga (Kawada et al. 2004b). The above preliminary investigations were carried out using the prototype paper devices, which were prepared in the laboratory. A more advanced technology is required for mass production of these devices, which should be economical, easy to handle, and long-lasting so that their use can be feasible. Therefore, a plastic strip impregnated with metofluthrin was produced on a trial basis. In this paper, we report the residual spatial repellency of this formulation for mosquitoes in the field conditions existing in Lombok, Indonesia. MATERIALS AND METHODS Formulation of metofluthrin-impregnated strips Metofluthrin-impregnated multilayer paper strips (similar to the devices used by Kawada et al. (2004a, 2004b)) and plastic strips were supplied by Sumitomo Chemical Co., Ltd. (Takatsukasa 4-2-1, Takarazuka, Hyogo, 665-8555 Japan). Metofluthrin, 200 mg, diluted with acetone was uniformly applied to the paper strip (tissue paper, 25g/m2 in density, 0.06-0.07mm in thickness), which is multilayered and foldable, and the acetone was allowed to evaporate. The plastic material had a folded cylindrical-shaped network structure (11.5 cm in width, 3-4 mesh) composed of polyethylene impregnated with 5% (w/w) metofluthrin. The plastic material was cut into a strip weighing 20 g (11.5 by 37 cm) (Figure 1). Field evaluation of repellency The field evaluation was carried out at Presak located in the coastal area ca. 7.5 km northwest of Mataram. The dominant mosquito species in the area, which were collected by human baited collection, were Culex quinquefasciatus Say. The mean numbers of outdoor mosquito bites were >300 per man per night (Kawada et al. 2004b). Four beruga (floor areas, 4.4 to 6.2 m2) were chosen as the test sites. The beruga is a simple hut made of wood and palm leaves and has no walls or may have simple screens on one or two sides. The people on Lombok Island use the beruga for resting, praying, and during evening conversations with neighbors. The trials for metofluthrin-impregnated paper strips and plastic strips were carried out from September 29 to October 3, 2003 and from December 8, 2003 to March 17, 2004, respectively. The former period was the relatively dry season and latter was the rainy season. Two treatment regimes were rotated on a daily basis (four replications for the four different beruga) among the beruga; a) four-paper strip treatment regimes and b) untreated control for the paper strip trial and three treatment

182

Journal of Vector Ecology

December 2005

were maintained, as mentioned earlier, outside the beruga throughout the trial and the same evaluation was carried out at one to four-week intervals. Percent repellency was calculated based on the average number of mosquitoes per hour collected at a strip-treated beruga and at an untreated beruga. Ambient temperature during the tests ranged from 24oC to 28oC. RESULTS AND DISCUSSION Spatial repellency of metofluthrin-impregnated paper strips and plastic strips for mosquitoes at beruga are shown in Table 1 and 2. The dominant mosquito species was Cx. quinquefasciatus, followed by a small number of Anopheles sundaicus (Rodenwaldt). The repellency of the strips was evaluated based on the effect on the total number of mosquitoes of both species. The number of mosquitoes collected at the untreated control beruga fluctuated, ranging from 19.9 to 43.5 per h, but it was maintained at a constantly high value throughout the trials (Figure 3). High spatial repellent effect was observed in the beruga that was treated with four metofluthrin-impregnated paper strips on the day of treatment, as previously reported by Kawada et al. (2004b). The repellency, however, decreased rapidly in less than one week (Table 1). This rapid reduction in efficacy is attributable to the rapid decrease in the active ingredient in the paper strip due to degradation and the rapid loss by vaporization since no protective measure against photo degradation and oxidation or against excess loss by vaporization was incorporated into the device. In the case of metofluthrin-impregnated plastic strips, on the other hand, high spatial repellency was

Figure 1. Metofluthrin-impregnated plastic strip used in the study.

regimes were rotated on a daily basis (three to four replications for the four different beruga); a) four plastic strips, b) two plastic strips and c) untreated control for the plastic strip trial, respectively. Two humans (males aged 20 to 30, 50 to 60 kg weight) laid in a bed net (ca 2 by 2 by 2 m), which was hung in each beruga, during the test as human bait, and mosquitoes were collected outside the net by a person with an aspirator. This person stayed outside the bed net throughout the trial. Strips were hung below the ceiling of the beruga outside the bed net (Figure 2). The distance between the strips and the humans inside the bed net was 1 to 1.5 m. Mosquitoes were collected from 1800 h to 2400 h, as described above. Strips

Figure 2. Field test with four plastic strips in a beruga. Arrows indicate the plastic strips treated.

Vol. 30, no. 2

Table 1. Numbers of mosquitoes collected in the metofluthrin-impregnated paper strips at treated and untreated collection sites.

Sample 4 0 1 2 4 4 4 0 0.17 0.04 2.7 16.0 17.4 2.7a (1.8) 16.2a (9.3) 17.5a (11.2) 89.0 40.7 12.2

No. of devices per beruga Weeks after treatment Replications % Repellency4

Average number collected/h Anopheles Culex sundaicus quinquefasciatus Total2 (SD3)

Metofluthrinimpregnated paper strip1 -

0 4 0.13 24.5 24.6b (8.3) 0 1 4 0.21 27.1 27.3a (9.8) 0 2 4 0.29 19.6 19.9a (4.9) 0 1 200 mg metofluthrin was impregnated in a device. 2 Same letters in the same week indicate that there was no significant difference between the number of collection (P > 0.05, Tukey's HSD Test). 3 Standard deviation. 4 % Repellency = (1 - No. treated/No. untreated) x 100.

Untreated

Journal of Vector Ecology

Numbers of mosquitoes collected per h

Figure 3. Changes in the total number of mosquitoes collected per h during the trial for plastic strips. Bars indicate the standard deviations.

183

184

Table 2. Number of mosquitoes collected in the metofluthrin-impregnated plastic strips treated and untreated collection sites.

Sample

No. of devices per beruga Weeks after treatment Replications % Repellency4

Average number collected/h Anopheles Culex sundaicus quinquefasciatus Total2 (SD3)

Journal of Vector Ecology

1 4 0.04 1.7 1.8a (0.2) 94.2 2 4 0 2.0 2.0a (1.5) 91.9 4 4 0.04 4.6 4.6a (3.1) 83.7 2 6 3 0.06 6.3 6.4a (1.8) 80.2 9 3 0 12.8 12.8a (7.2) 70.6 11 3 0 9.6 9.6a (4.2) 63.2 Metofluthrin15 3 0.11 17.9 18.0b (8.7) 38.2 impregnated 1 4 0 0.4 0.4a (0.4) 98.6 plastic strip1 2 4 0 1.3 1.3a (1.1) 94.7 4 4 0.04 1.3 1.4a (0.3) 95.2 4 6 3 0 1.6 1.6a (0.2) 95.0 9 3 0.06 3.6 3.7a (2.0) 91.6 11 3 0 5.9 5.9a (2.0) 77.2 15 3 0 9.0 9.0a (8.1) 69.2 1 4 0.13 30.0 30.1b (9.8) 0 2 4 0.21 24.6 24.8b (8.9) 0 4 4 0.04 28.3 28.3b (14.7) 0 Untreated 6 3 0 32.2 32.2b (9.5) 0 9 3 0.17 43.5 43.7b (12.4) 0 11 3 0.06 26.1 26.1b (6.5) 0 15 3 0 29.1 29.1b (5.0) 0 1 1000 mg metofluthrin was impregnated in a device. 2 Same letters in the same week indicate that there was no significant difference between the number of collection (P > 0.05, Tukey's HSD Test). 3 Standard deviation. 4 % Repellency = (1 - No. treated/No. untreated) x 100.

December 2005

Vol. 30, no. 2

Journal of Vector Ecology REFERENCES CITED

185

maintained much longer than that for the paper strips. A significantly higher (>60%) spatial repellency (P < 0.05; Tukey's HSD test) than that observed in the untreated control, lasted for at least 11 weeks with the two-strip treatment and for >15 weeks with the four-strip treatment (Table 2, Figure 3). We could confirm the effective, long-lasting efficacy of the repellent against mosquitoes by using metofluthrinimpregnated plastic strips in outdoor conditions. However, the effective duration in the present result appears to be insufficient with regard to the practical use of the devices, from the viewpoint of cost of the treatment. The plastic strips examined in the present study were only a prototype that requires further improvement. The long-term effectiveness might be achieved by the further formulation studies that work towards optimizing the release speed of the active ingredient, such as the investigation for the optimum density of plastic polymer. The more improved measures for manufacturing plastic will provide an improved formulation that result in a longer effective duration period of the repellent. Acknowledgments We express our deep gratitude to K. Tusta and K. Shimabukuro for their assistance in the study.

Kawada, H., Y. Maekawa, Y. Tsuda, and M. Takagi. 2004a. Laboratory and field evaluation of spatial repellency with metofluthrin impregnated paper strip against mosquitoes in Lombok Island, Indonesia. J. Am. Mosq. Contr. Assoc. 20: 292-298. Kawada, H., Y. Maekawa, Y. Tsuda, and M. Takagi. 2004b. Trial of spatial repellency of metofluthrin-impregnated paper strip against Anopheles and Culex in shelters without walls in Lombok, Indonesia. J. Am. Mosq. Contr. Assoc. 20: 434-437 . Shono, Y., S. Kubota, M. Sugano, H.H. Yap, and K. Tsushima. 2004. Field evaluation of paper strips and mosquito coil formulation impregnated metofluthrin for mosquito control in Malaysia. p. 40. In: 70th Annual Meeting of American Mosquito Control Association Abstracts, February 21-26, Savannah, GA. American Mosquito Control Association. 57 pp. Ujihara, K., T. Mori, T. Iwasaki, M. Sugano, Y. Shono, and N. Matsuo. 2004. Metofluthrin: A potent new synthetic pyrethroid with high vapor activity against mosquitoes. Biosci. Biotechnol. Biochem. 68: 170-174.

186

Journal of Vector Ecology

December 2005

Seasonal dynamics of four potential West Nile vector species in north-central Texas

Bethany G. Bolling , James H. Kennedy1, and Earl G. Zimmerman1

Department of Microbiology, Immunology, and Pathology, Arthropod-Borne and Infectious Diseases Laboratory, Colorado State University, Fort Collins, CO 80523, U.S.A. 1 Department of Biological Sciences, Environmental Science Division, University of North Texas, P.O. Box 310559, Denton, TX 76203-0559, U.S.A. Received: 18 November 2004; Accepted 2 May 2005 ABSTRACT: A population survey was conducted from April through September 2002 on mosquito species occurring on the Ray Roberts Greenbelt, a riparian corridor used for public recreation on the Elm Fork of the Trinity River, in Denton County, TX. Geographic information system software was used to set up a stratified random sampling design based on habitat parameters. Mosquitoes were collected using light traps, gravid traps, and resting boxes. A total of 29 species was collected during this study belonging to the following genera: Aedes, Anopheles, Coquillettidia, Culex, Mansonia, Ochlerotatus, Orthopodomyia, Psorophora, and Uranotaenia. The four most common species collected during this study were Aedes vexans, Culex erraticus, Culex salinarius, and Psorophora columbiae. West Nile virus (WNV) has been detected in these species in the United States, and they may serve as important vector species in Denton County. Collections were analyzed by the Texas Department of Health for arboviruses. One pool consisting of both Cx. erraticus and Cx. salinarius, collected in August 2002, tested positive for WNV, making it the first mosquito pool to test positive for WNV in Denton County. Canonical correspondence analysis was performed using abundance data of dominant species with selected weather variables and habitat parameters. Important factors for determining dominant species abundance were temperature, precipitation, dew point, and canopy coverage. Spatial and temporal patterns of these species are discussed. Journal of Vector Ecology 30 (2): 186-194. 2005. Keyword Index: Mosquito ecology, canonical correspondence analysis, West Nile virus.

INTRODUCTION Accurate information on the ecology of mosquito fauna in a region is necessary for the development of efficient mosquito control programs (Alten et al. 2000). Currently, there is very little published information on mosquito species occurring in north-central Texas. Spatial and temporal patterns of medically important species are important in determining where infectious agents can be spread (Meade and Earickson 2000). The recent introduction of West Nile virus (WNV) in Texas in 2002, has created an urgent need for more efficient approaches to mosquito surveillance and control efforts in the state. West Nile virus has been found in 60 species of mosquitoes in the United States since 1999 (CDC 2005). The main objective of this study was to determine the mosquito species present in Denton County, TX, and to describe the seasonal population dynamics. This study was planned prior to the anticipated arrival of WNV in Texas. MATERIALS AND METHODS The study site of Denton County is located in northcentral Texas, near the Dallas-Fort Worth Metroplex. The climate is humid subtropical, with moderate rainfall (890 mm/ year) and periodic drought (USDA 1980, Bailey 1995). The study area is located in the Ray Roberts Greenbelt, a recreational area that follows the Elm Fork of the Trinity River.

The Greenbelt is approximately 2,000 ha and is managed by Texas Parks and Wildlife. The dominant vegetation of the forest is cedar elm (Ulmus crassifolia), hackberry (Celtis reticulata and C. laevigata), and green ash (Fraxinus pennsylvanica), with some bur oak (Quercus macrocarpa), Shumard oak (Quercus shumardii), pecan (Carya illinoensis), and eastern cottonwood (Populus deltoides) (Barry and Kroll 1999). The land surrounding the Ray Roberts Greenbelt is primarily used for agricultural purposes. A habitat classification of the study area was derived from digital aerial photos taken in August of 2000. The study area was classified into four habitat classes: bottomland forestold growth, bottomland forest-new growth, open field, and upland forest. The resulting theme was created by drawing polygons over the digital photos using ESRI® ArcView® 3.2 software (Environmental Systems Research Institute, Inc., Redlands, CA 2002). Ground-truthing with a Garmin II GPS unit (Garmin International Inc., Olathe, KS) was incorporated to increase the accuracy of the land cover theme. Population abundance data were obtained through biweekly trapping from April through September 2002. Ten new sites were randomly selected each week of collection and stratified by the percent cover of each habitat type. A random number generator was used to select sampling points in each habitat. A GPS unit was used to locate sampling sites in the field. Adult mosquitoes were collected using Centers for

Vol. 30, no. 2

Journal of Vector Ecology

187

Disease Control (CDC) miniature light traps and gravid traps (John W. Hock Company, Gainesville, FL). Light traps were baited with 1 kg dry ice, powered with 6-V lantern batteries, and hung 1 m above the ground. CDC gravid traps were baited with hay-infused water and powered with 6-V lantern batteries. These traps were placed in the field before dusk and retrieved the following morning. Collection bags were brought to the lab where mosquitoes were aspirated and deposited in shipping containers. Adults collected from CDC traps and gravid traps were shipped to the Texas Department of Health, in Austin, TX, for identification and virus testing. Resting boxes were also used to sample adult mosquitoes. Boxes were made of five 930-cm2 pieces of plywood, painted black on the outside and red on the inside (Service 1993). Boxes were placed in the field before dusk and collected in the afternoon after three days. In order to collect the mosquitoes from the resting box, a mesh screen (with a slit down the middle) was placed over the open end, and mosquitoes were collected using a handheld aspirator. Mosquitoes were placed in plastic containers and returned to the lab for species identification. Weather data were obtained from an online weather database (www.wunderground.com), which compiles information from airport weather stations throughout the country. Environmental data for this study were collected at Denton Municipal Airport, located approximately 11 km from the study area. Variables of interest were temperature, dew point, precipitation, and wind speed. Weekly averages were calculated for these variables. Other variables incorporated to account for possible time lag were number of days since last rain event (greater than 1 cm), precipitation occurring over the 7-day period, and the 14-day period prior to sampling date. Canopy coverage was measured with a spherical densiometer at each collection site (Mills and Stevenson 1999). Statistical analyses were performed on abundance data of the four dominant species collected during the study to determine which environmental elements and habitat variables were most influential on species distributions in the study area. Correlation analysis was used to examine possible relationships between mosquito abundance and environmental variables. Since some of the data sets did not meet normality assumptions, nonparametric Spearman correlation coefficients were calculated for the dominant species abundance with weather variables using SPSS® 11.0.1 (© SPSS Inc., Chicago, IL. 2001). Canonical correspondence analysis (CCA) (ter Braak 1986) was performed to further explore the relationships between species abundance with environmental variables. CCA is a multivariate direct gradient analysis technique, where species composition is directly related to a set of environmental variables. CCA was included in this study because it performs well with skewed species distributions, high noise levels, and complex sampling designs (Palmer 1993). This analysis was carried out using CANOCO 4 for Windows (ter Braak and Smilauer 1998). Cluster analysis was used to compare species composition between habitat types. An agglomerative, hierarchical cluster analysis was conducted on abundance data with SIGTREE 3.4 (© International

Statistics and Research Corp., Government of Canada. 1991), which uses the Bray-Curtis similarity coefficient and the bootstrap method to determine statistical significance. An unweighted average linkage method was used to calculate clusters. RESULTS Mosquito species A total of 36,166 mosquitoes was collected during this study belonging to 29 species and the following nine genera: Aedes, Anopheles, Coquillettidia, Culex, Mansonia, Ochlerotatus, Orthopodomyia, Psorophora, and Uranotaenia. The largest number of species (n=28) and over 96% of the adult specimens were collected with the CDC miniature light trap (Table 1). Thirteen species were collected with the gravid traps and six species with the resting boxes. Table 1 shows the relative abundance of each species collected in the study area expressed as a percentage of the total number of adult mosquitoes collected. Culex erraticus was the most abundant species, comprising more than 51% of the total collections, followed by Aedes vexans (26%), Culex salinarius (10%), and Psorophora columbiae (7%). The remainder of the species combined contributed to less than 6% of the adult collections. The seasonal distributions of the dominant species reveal population fluctuations (Figure 1). Ae. vexans was the primary species collected in April and May. Although the population of this species decreased greatly in June, it was collected in small numbers throughout the sampling period. Ps. columbiae reached its highest population density in the month of June, with a smaller peak occurring at the end of July. Present from May through the end of September, Cx. erraticus was by far the most abundant species collected. The seasonal population distribution of this species exhibited a major peak in midJuly and then a larger one in early August. Abundance of Cx. salinarius followed the same general trend as that for Cx. erraticus but with smaller numbers. West Nile virus was detected in one pool consisting of 43 Cx. erraticus and seven Cx. salinarius, collected on 6 August 2002 from a CDC miniature light trap located in upland forest habitat. This was the first detection of WNV in mosquitoes in Denton County. Environmental data Weather variables used for statistical analyses included weekly averages of temperature, dew point, and wind speeds (Table 2). One-week precipitation refers to rain accumulation 7 days prior to collection date. Two-week precipitation refers to rain accumulation 14 days prior to collection date. Precipitation-days refers to the number of days since the last rain event (> 1 cm). Over the sampling period, temperatures ranged from 12.7°C to 36.4°C. Dew point temperatures ranged from 12.3°C to 23°C. The largest rain event occurred on May 10, with an accumulation of 6.25 cm. The total amount of rain occurring throughout the study period was 41 cm. Average wind speeds ranged from 7.6 to 32.8 km/hr.

188

Journal of Vector Ecology

December 2005

Table 1. Species counts and relative abundance of adult mosquitoes collected April through September 2002 in the Ray Roberts Greenbelt.

Species

Aedes albopictus Aedes vexans Anopheles atropos Anopheles crucians Anopheles punctipennis Anopheles quadrimaculatus Coquillettidia perturbans Culex erraticus Culex coronator Culex quinquefasciatus Culex restuans Culex salinarius Culex tarsalis Culex territans Mansonia titillans Ochlerotatus epacticus Ochlerotatus sollicitans Ochlerotatus sticticus Ochlerotatus taeniorynchus Ochlerotatus triseriatus Ochlerotatus trivittatus Ochlerotatus zoosophus Orthopodomyia signifera Psorophora ciliata Psorophora columbiae Psorophora cyanescens Psorophora discolor Psorophora ferox Uranotaenia sapphirina

Light Trap Gravid Trap Resting Box Total

14 9,383 2 68 80 142 111 18,103 16 0 50 3,341 172 10 27 1 15 5 210 39 164 6 3 9 2,448 26 1 19 566 35,031 96.862 40 9 0 0 0 7 0 49 0 61 219 103 3 0 0 2 0 0 0 10 3 0 14 0 9 0 0 0 0 529 1.463 0 0 0 8 12 25 0 553 0 0 0 7 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 606 1.676 54 9,392 2 76 92 174 111 18,705 16 61 269 3,451 175 10 27 3 15 5 210 49 167 6 17 9 2,457 26 1 19 567 36,166 100

% Total

0.149 25.969 0.006 0.210 0.254 0.481 0.307 51.720 0.044 0.169 0.744 9.542 0.484 0.028 0.075 0.008 0.041 0.014 0.581 0.135 0.462 0.017 0.047 0.025 6.794 0.072 0.003 0.053 1.568 100

Total % Total

Table 2. Weather data retrieved from www.wunderground.com (© The Weather Underground, Inc.) and averaged for each week of sampling.

Mean Temp. (ºC) 20.70 23.30 19.00 23.30 26.90 26.90 28.50 30.00 28.40 27.30 26.30 22.10 Dew Point (ºC) 14.70 18.40 13.90 18.80 20.90 20.70 23.00 22.00 20.70 15.90 19.60 12.30 One-Week Two- Week Precip. Precip. (cm) (cm) 0.05 2.46 0.05 0.10 6.66 10.12 1.65 3.43 3.25 3.28 0.00 3.83 0.08 0.13 0.21 5.17 1.78 1.88 0.00 0.03 2.27 2.73 1.91 1.91 Wind Speed (km/h) 21.60 15.10 19.70 9.80 17.70 10.00 7.60 13.80 9.50 11.80 10.70 12.10

Date 4/23/02 4/30/02 5/14/02 5/29/02 6/11/02 6/26/02 7/10/02 7/22/02 8/5/02 8/26/02 9/9/02 9/24/02

Precip. Days 10 17 4 2 6 10 24 10 6 16 30 5

Vol. 30, no. 2

Journal of Vector Ecology

189

Table 3. Significance values produced from Monte Carlo permutation test.

Variable Mean Temperature Wind Speed Canopy Coverage Open Field Dew Point Rain Accumulation 1 Week Prior to Sampling Rain Accumulation 2 Weeks Prior to Sampling Number of Days Since Last Rain Event (> 1cm) Upland Forest Bottomland Forest-Old Growth Bottomland Forest-New Growth F 86.33 23.22 15.39 35.56 9.51 9.09 9.42 5.5 2.53 0.08 0.44 P-value 0.005 0.005 0.01 0.005 0.005 0.005 0.005 0.005 0.07 0.965 0.43

Correlation analysis Nonparametric correlation analysis was performed on abundance data of the four most abundant species. There was a significant negative correlation between Ae. vexans abundance and maximum temperature (Spearman Rank Order R = -0.345, P = 0.02). Ae. vexans abundance was also significantly correlated with wind speed (R = 0.362, P = 0.01). Cx. erraticus abundance was strongly correlated with two variables, mean temperature (R = 0.816, P < 0.01) and dew point (R = 0.631, P < 0.01). Cx. erraticus abundance had a significant negative correlation with wind speed (R = -0.647, P < 0.01). Cx. salinarius abundance was positively, significantly correlated with mean temperature (R = 0.464, p < 0.01), dew point (R = 0.544, P < 0.01), and two-week rain accumulation (R = 0.317, P = 0.03), but was negatively correlated with wind speed (R = -0.334, P = 0.02). Significant positive correlations existed between abundance of Ps. columbiae with mean temperature (R = 0.460, P < 0.01) and dew point (R = 0.462, P < 0.01). Canonical correspondence analysis The ordination diagram produced by CCA (Figure 2) shows the relationships between species abundance and environmental variables. The arrows represent environmental variables and the circles are species abundance. The length and direction of the arrow indicates the importance of the variable and how it correlates with the species composition axes. The angle between arrows shows correlations among the environmental variables. The location of the circles (species) reveals the environmental preferences of each species (Palmer 1993). The most important variables determining Ae. vexans abundance were temperature measurements and dew point (Figure 2). These were negative associations. Based on this diagram, it does not prefer one habitat to another. Ps. columbiae abundance was positively related to two-week precipitation accumulation, and negatively correlated with average canopy coverage and number of days since the last rain event. This species also displays a strong association with upland and field habitats. Counts of Cx. erraticus and Cx. salinarius were driven by the same weather

variables. Temperature measurements and dew point had the strongest associations with these species. The habitat preferences for these species are unclear from this analysis. The CCA produced four canonical axes with the following eigenvalues: 0.771, 0.405, 0.070, and 0.234. The first two axes were the most important in separating the species. The significance levels produced from the Monte Carlo permutation test (Table 3) show the importance of each of the environmental variables. Mean temperature had the highest F-value at 86.33 (P < 0.01). Cluster analysis A cluster analysis was used to describe species composition within the different habitat types for each week of sampling. The Bray-Curtis similarity coefficient and bootstrap analysis assessed statistical significance of the clusters. As expected, bottomland forest-old growth and bottomland forest-new growth exhibited the most similar patterns. For eight of the 12 sampling weeks (67%), these two habitats clustered together with similarity levels ranging from 0.567 to 0.892 but were not significantly associated (P = 0.280-0.778). Interestingly, bottomland forest-old growth and upland forest clustered together three different times, in three separate months, with statistically significant probabilities (May: P = 0.998, June: P = 0.994, and August: P = 0.958) and similarity levels ranging from 0.648 to 0.727. Bottomland forest-old growth and open field clustered together only once, in April, with low similarity and probability values (similarity = 0.491, P = 0.384). No significant differences were found between species composition among habitat types. DISCUSSION Weather patterns affect adult female mosquito abundance by altering the quality and quantity of larval habitats. This relationship between climate variables and mosquito abundance can provide important information to determine virus activity levels and, therefore, disease risk (Wegbreit and Reisen 2000). Statistical analyses focused on the four most

190 Journal of Vector Ecology

December 2005

Figure 1. Seasonal distribution of the dominant mosquito species found on the Ray Roberts Greenbelt in 2002. Counts are standardized by number of traps operated per week.

Vol. 30, no. 2

Journal of Vector Ecology

191

Figure 2. Ordination diagram produced from canonical correspondence analysis. (1WKPRECIP and 2WKPRECIP = rain accumulation occurring in weeks prior to sampling week, WINDSPD = wind speed, PRECPDAY = number of days since last rain event greater than 1cm, CANCOV = average canopy coverage, MEANTEMP = mean temperature, and DEWPOINT = dew point, UPL = upland habitat, FLD = open field habitat, NBL = bottomland forest-new growth, OBL = bottomland forest-old growth, AED VEX = Aedes vexans, CUL ERA = Culex erraticus, CUL SAL = Culex salinarius, PSO COL = Psorophora columbiae ).

common species collected: Ae. vexans, Cx. erraticus, Cx. salinarius, and Ps. columbiae. West Nile virus has been detected in field-collected samples of these species (CDC 2005), but other factors must be considered to implicate them as important vectors, including vector competence and host preference (Turell et al. 2005, Sardelis et al. 2001). Vector competence refers to infection susceptibility, permissiveness for pathogen replication, duration of extrinsic incubation period, and transmission capability (Higgs and Beaty 2005). West Nile virus is maintained in an enzootic transmission cycle between birds and mosquitoes, with humans and horses considered to be incidental hosts. An effective enzootic vector would be a mosquito that feeds primarily on birds, while a bridge vector would be a more catholic feeder, transmitting virus to non-avian hosts as well (Turell et al. 2001). Incorporating this information with mosquito abundance data can help determine which species may serve as important vectors. Results of correlation and canonical correspondence analyses indicated that maximum temperature was the most influential climatic variable determining Ae. vexans abundance. This association was negative, and it appeared

that as summer temperatures increased, populations of this species declined. Ae. vexans is known as a floodwater species, and immature stages are often found in temporary rain-filled pools (Carpenter and LaCasse 1955), which suggests seasonal occurrence of this species is also correlated with precipitation. In Canada, Helson et al. (1979) found that Ae. vexans peaks of abundance were associated with extensive rainfall 2-4 weeks prior. A lag time between rainfall and peak collections of Ae. vexans ranged from 1-2 weeks during the summer to 3 weeks during the fall months in Nebraska (Janousek and Kramer 1999). In this study, Ae. vexans abundance was highest during the first month of collections and then rapidly decreased as temperatures increased, so we were unable to see a relationship with precipitation patterns. More extensive sampling, beginning earlier in the season, is needed to determine the annual fluctuations of this species. Laboratory studies have shown Ae. vexans to have moderate infection and transmission rates with WNV (Turell et al. 2005, Goddard et al. 2002). This species primarily feeds on mammals (Nasci 1984) and could potentially serve as a bridge vector for WNV. Since Ae. vexans abundance peaked during April in our study and WNV activity generally peaks in late summer, it is unlikely

192

Journal of Vector Ecology

December 2005

that this species plays a major role in epidemic virus transmission. Both correlation and canonical correspondence analyses showed mean temperature to be the most important variable of interest affecting abundance of Cx. erraticus. An increase in temperature, within a species thermal tolerance, will increase metabolic and development rates (Gullan and Cranston 2004) resulting in population growth. Culex erraticus abundance had two peaks during the sampling period in July and August coinciding with the some of the highest temperatures of the summer in 2002. McNelly and Crans (1989) observed a similar distribution in New Jersey, with adults appearing in resting box collections starting in July and reaching peak populations during August and September. Patterns of seasonal abundance of Cx. erraticus in North Carolina also exhibited peaks in summer and early fall (Robertson et al. 1993). West Nile virus has been detected in this species (CDC 2005), but there have been no laboratory studies done to determine vector competence. In our study, WNV was detected in a mixed pool of Cx. erraticus and Cx. salinarius, so it is unclear which species was positive for the virus. Culex erraticus mainly feeds on birds (Tempelis 1975, Edman 1979), but has also been shown to feed on mammals, reptiles, and amphibians (Robertson et al. 1993). With opportunistic feeding habits and high population levels occurring in late summer, this species could serve as an enzootic vector or bridge vector depending on host availability. Based on results from correlation and canonical correspondence analyses, mean temperature was also found to be an important factor associated with abundance of Cx. salinarius. The seasonal distribution of Cx. salinarius on the Greenbelt displayed a small peak beginning in late July and extending into early August, which tends to be the hottest part of the summer. This association was seen in New Jersey, where Cx. salinarius collected with light traps were highest during the summer when temperatures were greatest (Slaff and Crans 1981). Slaff and Haefner (1985) observed the abundance of Cx. salinarius to be highest from March through June, before peak temperatures were reached in central Florida. They suggested that seasonal fluctuations of Cx. salinarius seen in Florida were due to a competitive interaction with Cx. nigripalpus. In contrast, abundance of Cx. salinarius was found to be negatively correlated with monthly temperature in southeastern Louisiana (Samui et al. 2003). Other factors not measured in these studies could also be influencing the population dynamics. Numerous isolations of WNV have been made from this species in other field collections (Turell et al. 2005) and it has been shown to be a competent vector in the laboratory (Sardelis et al. 2001). Blood meal studies indicate that Cx. salinarius is opportunistic in its host preference, feeding on both birds and mammals (Crans 1964). Considering its feeding behavior, vector competence for WNV, and seasonal overlap with WNV activity, Cx. salinarius may serve as an important bridge vector for human transmission (Kulasekera et al. 2001, Sardelis et al. 2001, Andreadis et al. 2004). It is also thought that Cx. salinarius serves as a bridge vector for Eastern equine encephalitis virus

(Vaidyanathan et al.1997). Results of canonical correspondence analysis showed that rain accumulation 2 weeks prior to a sampling period was an important factor in determining Ps. columbiae abundance. Samui et al. (2003) made the same observation in Louisiana, with collections of Ps. columbiae positively correlating with rainfall accumulated 15 days prior. Welch et al. (1986) also determined that precipitation patterns were influential for this species in Texas riceland systems and numerous other studies have concluded that precipitation is a key determinant in adult abundance (Sulaiman and Jeffery 1986, Alto and Juliano 2001). This mosquito is a floodwater species that deposits its eggs in areas subject to temporary inundation (Horsfall 1955). Ps. columbiae is associated with rice fields and irrigated cropland, where the females deposit their eggs on moist soil (Meisch 1994). Several studies have shown that soil moisture is a major factor affecting the abundance of this species (Olson and Meek 1977, Focks et al. 1988). However, measurements of soil moisture were not included in this study. Canonical correspondence analysis also indicated that average canopy coverage was negatively associated and open field was positively associated with abundance of Ps. columbiae, indicating its preference for open fields. Previous studies suggest that Ps. columbiae prefer mammals, primarily cattle, for host feeding (Kuntz et al. 1982). West Nile virus has been detected in field collections (CDC 2005), but vector competence has not been determined for this species. Psorophora columbiae are known to be extremely aggressive biters and may play a role as a bridge vector for WNV transmission, depending on host availability. Another species that should be considered as an important vector for WNV in Denton County is Culex quinquefasciatus. This species is associated with more urban habitats, so very few specimens were collected in this study and abundance data were not statistically analyzed. After WNV was detected in Texas, mosquito surveillance activities were initiated in numerous counties. Since the end of our study, the majority of mosquito pools testing positive for WNV in Denton and surrounding counties have been from Cx. quinquefasciatus (TDH 2004). This species shows a preference for avian blood but will readily feed on mammals when these hosts are abundant (Tempelis 1975, Savage and Miller 1995). Vector competence studies (Sardelis et al. 2001, Goddard et al. 2002) indicate that Cx. quinquefasciatus is susceptible to infection and capable of transmitting WNV by bite. Further studies characterizing the seasonal fluctuations of this species are needed to clarify its role in the WNV transmission cycle. Cluster analyses were performed to explore species composition within the different habitat types but results were inconclusive. The lack of a high frequency of significant habitat associations might be expected, since all four dominant species were typically found in all habitats. Because of the limited size of the study area, juxtaposition of habitat types, and various flight ranges of mosquitoes, significant patterns between habitats were not detected. The description of mosquito population dynamics on the Ray Roberts Greenbelt serves as a tool to aid local health departments in arbovirus surveillance activities. Detailed

Vol. 30, no. 2

Journal of Vector Ecology

193

information about medically important species can aid in control efforts and also provide early warning signs for disease transmission risk. Results from this study indicated that temporal variables were more important in determining vector abundance than were the selected spatial parameters. Since most of the medically important species collected during the sampling period occurred in all habitat types, spatial risk assessments could not be determined for the Greenbelt. Weather variables, specifically mean temperature, were the most important factors determining the abundance of dominant mosquito species in the study area. Acknowledgments We would like to thank Glenna Teltow at the Texas Department of Health for providing advice and processing our specimens. We are grateful to Sonny Solis, the Greenbelt manager at the Ray Roberts State Park Complex and David H. Riskind of the Texas Parks and Wildlife Department for allowing us to collect mosquitoes on the Ray Roberts Greenbelt (Permit No: 75-01). Special thanks to Dr. Chet Moore for reviewing this manuscript and providing helpful comments. REFERENCES CITED Alten, B., R. Bellini, S.S. Caglar, F.M. Simsek, and S. Kaynas. 2000. Species composition and seasonal dynamics of mosquitoes in the Belek region of Turkey. J. Vector Ecol. 25: 146-154. Alto, B.W. and S.A. Juliano. 2001. Precipitation and temperature effects on populations of Aedes albopictus (Diptera: Culicidae): Implications for range expansion. J. Med. Entomol. 38: 646-656. Andreadis, T.G., J.F. Anderson, C.R. Vossbrinck, and A.J. Main. 2004. Epidemiology of West Nile virus in Connecticut: a five-year analysis of mosquito data 19992003. Vector-borne Zoonotic Dis. 4: 360-378. Bailey, R.G. 1995. Description of the Ecoregions of the United States, 2nd ed. USDA-Forest Service Miscellaneous Publication 1391, Washington. Barry, D. and A.J. Kroll. 1999. A phytosociological description of a remnant bottomland hardwood forest in Denton County, Texas. Texas J. Sci. 51: 309-316. Carpenter, S.J. and W.J. LaCasse. 1955. Mosquitoes of North America, North of Mexico. University of California Press, Berkeley and Los Angeles. Centers for Disease Control and Prevention. 2005. West Nile Virus Web Page. http://www.cdc.gov/ncidod/dvbid/ westnile/mosquitoSpecies.htm. Crans, W.J. 1964. Continued host preference studies with New Jersey mosquitoes, 1963. Proc. 51st Ann. Meeting of N. J. Mosq. Extermin. Assoc. 50-58. Edman, J.D. 1979. Host-feeding patterns of Florida mosquitoes (Diptera: Culicidae) VI. Culex (Melanoconion). J. Med. Entomol. 15: 521-525. Focks, D.A., R.E. McLaughlin, and B.M. Smith. 1988. A

dynamic life table model of Psorophora columbiae in the southern Louisiana rice agroecosystem with supporting hydrological submodel. Part I. Analysis of literature and model development. J. Am. Mosq. Contr. Assoc. 4: 266-281. Goddard, L.B., A.E. Roth, W.K. Reisen, and T.W. Scott. 2002. Vector competence of California mosquitoes for West Nile virus. Emerg. Infect. Dis. 8: 1385-1391. Gullan, P.J. and P.S. Cranston. 2004. The Insects, An Outline of Entomology, 3rd ed. Blackwell Publishing, Oxford, U.K. Helson, B.V., G.A. Surgeoner, and R.E. Wright. 1979. Mosquitoes of southwestern Ontario, their seasonal distribution, prevalence and new records. Ontario Minist. Health. 182-198. Higgs, S. and B.J. Beaty. 2005. Natural cycles of vector-borne pathogens. In: Biology of Disease Vectors, 2nd Ed. W.C. Marquardt, W.C. Black, J.E. Freier, H.H. Hagedorn, J. Hemingway, S. Higgs, A.A. James, B. Kondratieff, and C.G. Moore (eds.). Elsevier Academic Press, New York. Horsfall, W.R. 1955. Mosquitoes: Their Bionomics and Relation to Disease. The Ronald Press Company, New York. Janousek, T.E. and W.L. Kramer. 1999. Seasonal incidence and geographical variation of Nebraska mosquitoes, 1994-95. J. Am. Mosq. Contr. Assoc. 15: 253-262. Kulasekera, V.L., L. Kramer, R.S. Nasci, F. Mostashari, B. Cherry, S.C. Trock, C. Glaser, and J.R. Miller. 2001. West Nile virus infection in mosquitoes, birds, horses, and humans, Staten Island, New York, 2000. Emerg. Infect. Dis. 7: 722-725. Kuntz, K.J., J.K. Olson, and B.J. Rade. 1982. Role of domestic animals as hosts for blood-seeking females of Psophora columbiae and other mosquito species in Texas ricelands. Mosq. News. 42: 202-210. McNelly, J.R. and W.J. Crans. 1989. The larval habitat of Culex erraticus in southern New Jersey. Proc. N. J. Mosq. Contr. Assoc. 63-64. Meade, M.S. and R.J. Earickson. 2000. Medical Geography. The Guilford Press, New York. Meisch, M.V. 1994. The dark ricefield mosquito Psorophora columbiae. Wing Beats 5: 8. Mills, K.E. and N.J. Stevenson. 1999. Riparian vegetation. In: Aquatic habitat assessment: common methods. M.B. Bain and N.J. Stevenson (eds.) American Fisheries Society, Bethesda, Maryland. Nasci, R.S. 1984. Variations in the blood-feeding patterns of Aedes vexans and Aedes trivittatus (Diptera: Culicidae). J. Med. Entomol. 21: 95-99. Olson, J.K. and C.L. Meek. 1977. Soil moisture conditions that are most attractive to ovipositing females of Psorophora columbiae in Texas ricelands. Mosq. News 37: 19-26. Palmer, M.W. 1993. Putting things in even better order: The advantages of canonical correspondence analysis. Ecology 74: 2215-2230. Robertson, L.C., S. Prior, C.S. Apperson, and W.S. Irby. 1993. Bionomics of Anopheles quadrimaculatus and Culex

194

Journal of Vector Ecology

December 2005

erraticus (Diptera: Culicidae) in the Falls Lake Basin, North Carolina: seasonal changes in abundance and gonotrophic status, and host-feeding patterns. J. Med. Entomol. 30: 689-698. Samui, K.L., R.M. Gleiser, M.E. Hugh-Jones, and C.T. Palmisano. 2003. Mosquitoes captured in a horse-baited stable trap in southeast Louisiana. J. Am. Mosq. Contr. Assoc. 19: 139-147. Sardelis, M.R., M.J. Turell, D.J. Dohm, and M.L. O'Guinn. 2001. Vector competence of selected North American Culex and Coquillettidia mosquitoes for West Nile virus. Emerg. Infect. Dis. 7: 1018-1022. Savage, H.M. and B.R. Miller. 1995. House mosquitoes of the U.S.A., Culex pipiens complex. Wing Beats. 6: 8-9. Service, M.W. 1993. Mosquito Ecology. Field Sampling Methods 2nd ed. Chapman & Hall, London. Slaff, M. and W.J. Crans. 1981. The activity and physiological status of pre- and posthibernating Culex salinarius (Diptera: Culicidae) populations. J. Med. Entomol. 18: 65-68. Slaff, M. and J.D. Haefner. 1985. The impact of phosphate mining on Culex nigripalpus and Culex salinarius (Diptera: Culicidae) populations in central Florida. Fla. Entomol. 68: 444-450. Sulaiman, S. and J. Jeffery. 1986. The ecology of Aedes albopictus (Skuse) (Diptera: Culicidae) in a rubber estate in Malaysia. Bull. Entomol. Res. 76: 553-557. Tempelis, C.H. 1975. Host-feeding patterns of mosquitoes, with a review of advances in analysis of blood meals by serology. J. Med. Entomol. 11: 635-653. ter Braak, C.J.F. 1986. Canonical correspondence analysis: A new eigenvector technique for multivariate direct gradient analysis. Ecology 67: 1167-1179.

ter Braak, C..J.F. and P. Smilauer. 1998. CANOCO reference manual and user's guide to Canoco for Windows: software for canonical community ordination (version 4). Microcomputer Power, New York. Texas Department of State Health Services. 2004. West Nile Virus Web Page. http://www.tdh.state.tx.us/zoonosis/ diseases/Arboviral/westNile/ Turell, M.J., M.L. O'Guinn, D.J. Dohm, and J.W. Jones. 2001. Vector competence of North American mosquitoes (Diptera: Culicidae) for West Nile virus. J. Med. Entomol. 38: 130-134. Turell, M.J., D.J. Dohm, M.R. Sardelis, M.L. O'Guinn, T.G. Andreadis, and J.A. Blow. 2005. An update on the potential of North American mosquitoes (Diptera: Culicidae) to transmit West Nile virus. J. Med. Entomol. 42: 57-62. Vaidyanathan, R., J.D. Edman, L.A. Cooper, and T.W. Scott. 1997. Vector competence of mosquitoes (Diptera: Culicidae) from Massachusetts for a sympatric isolate of eastern equine encephalomyelitis virus. J. Med. Entomol. 34: 346-352. Wegbreit, J. and W.K. Reisen. 2000. Relationships among weather, mosquito abundance, and encephalitis virus activity in California: Kern County 1990-98. J. Am. Mosq. Contr. Assoc. 16: 22-27. Welch, J.B., J.K. Olson, and M.M. Yates. 1986. Occurrence of Psorophora columbiae eggs in different field types comprising a Texas riceland agroecosystem. J. Am. Mosq. Contr. Assoc. 2: 52-56. USDA. 1980. Soil Survey of Denton County, Texas. Natural Resources Conservation Service, U.S. Department of Agriculture.

Vol. 30, no. 2

Journal of Vector Ecology

201

Host plant selection of two Mansonia Blanchard species (Diptera: Culicidae) in a heterogeneous habitat of Buenos Aires City, Argentina

Pablo R. Mulieri1, Juan P. Torretta2, and Nicolás Schweigmann1,3

Departamento de Ecología, Genética y Evolución, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Pabellón II, C1428 EHA Buenos Aires, Argentina 2 Cátedra de Botánica Agrícola, Facultad de Agronomía, Universidad de Buenos Aires. Av. San Martín 4453. C1417DSQ Buenos Aires 3 Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Argentina Received 7 December 2004; Accepted 4 March 2005 ABSTRACT: Larvae and pupae of the genus Mansonia Blanchard attach to the roots of aquatic plants by means of modified structures to obtain oxygen. A study of the association of larval Ma. indubitans and Ma. titillans with floating macrophytes was conducted at Macáes Pond, Argentina. Fifty-four sampling units were taken from January to May 2003. Three genera of host plants were considered: Pistia, Limnobium, and Salvinia. A total of 402 immatures of Ma. indubitans and 217 of Ma. titillans were captured and associations between Mansonia immatures and roots of each genera were assesed. Significant association was noted between Ma. indubitans and certain host plant species (K-W H=42.74, df=2, p<0.001). The same result was observed for Ma. titillans (K-W H=23.42, df=2, p<0.001). Both Mansonia species utilized roots of P. stratiotes in significantly higher proportions than expected by random selection. Both species showed significant negative association with Salvinia spp., while no clear relationships were detected with L. laevigatum. Journal of Vector Ecology 30 (2): 201205. 2005. Keyword Index: Mansonia, habitat selection, mosquitoes, host plant.

1

INTRODUCTION Animals perceive their environment as a heterogeneous mosaic of habitats (Sutherland and Poppy 1997) and display selective adaptations for occupying those sites with favorable conditions for their breeding, growth, and survival (Greene and Stamps 2001). Several culicid species are adapted to aquatic habitats with the presence of floating plants (Poi de Neiff and Neiff 1980, Balseiro 1986, Lounibos and Dewald 1989, Lounibos and Escher 1985). Among them, larval instars and pupae of species of the genus Mansonia Blanchard attach to roots of aquatic plants by means of modified structures to obtain oxygen (Ronderos and Bachmann 1964). The genus Mansonia is represented by twenty-three species, some of which are vectors of filariasis (Forattini 1965). In spite of this fact, very few studies have been reported regarding Mansonia immatures and host plant associations. Ma. indubitans Dyar and Shannon and Ma. titillans Walker share a wide geographic range in America, from the southern United States down to Buenos Aires Province in Argentina (Forattini 1965). In this area, they associate with a diverse range of host plant species (Slaff and Haefner 1985, Almiron and Brewer 1996, Ferreira et al. 2003). Ferreira (1999) and Ferreira et al. (2003) reported a differential association of Ma. humeralis, Ma. indubitans, Ma. titillans and Ma. amazonensis to Eichhornia crassipes and Pistia stratiotes in the central Amazon region. Roots of Eichhornia crassipes had higher larval numbers of Ma. indubitans and

Ma. titillans, compared to Ceratopteris sp. (Ferreira et al. 2003). In Argentina, García et al. (1995) reported the only work concerning immature seasonality and abundance of Mansonia in pure stands of P. stratiotes in Punta Lara, Province of Buenos Aires. The aim of the present study was to determine host-plant selection and to survey the spatial distribution of Ma. indubitans and Ma. titillans in a heterogeneous habitat of floating macrophytes. MATERIALS AND METHODS Field studies were conducted at the Macáes Pond, located in Costanera Sur Reserve (S 34°36'26.7" O 58°20'54.4"), Buenos Aires City, Argentina. This freshwater habitat was overgrown with floating plants, mainly Salvinia spp. (S. herzogii de la Sota and S. rotundifolia Wlld.), Pistia stratiotes L., and Limnobium laevigatum (H. and B.) Heine, representing a heterogeneous environment for immature instars of Mansonia. Fifty-four quadrangular sampling units of 0.14 m2 were randomly taken from January to May 2003. The species and number of plants per sampling unit were registered. Because of the difficult identification of species level in the field, individuals of Salvinia were considered at generic level. A total of three genera of host plants were considered: P. stratiotes, L. laevigatum, and Salvinia spp. Each plant with a single differentiated root was considered as an individual. Individual plants joined to each other with a stolon were

202

Journal of Vector Ecology

December 2005

separated before larval collection. The number of larvae and pupae was counted for each plant. Immatures were obtained by shaking plants vigorously in a plastic container with water and collecting them with plastic pipettes. They were fixed in situ with 80% ethanol. Fourth-instar larvae and pupae were identified to specific level according to Ronderos and Bachmann (1963). Early-instar larvae were not taken into account in this study. The monthly abundance of host plants and immatures (larvae and pupae) per plant between plant species was compared with the Kruskal-Wallis test, followed by a posteriori comparisons using the Mann-Whitney test. The frequency of both species of Mansonia on different plant species was tested by the standard Chi-Square test. In addition, we calculated the overlap in resource use of both Mansonia species with Hurlbert´s Index (Hurlbert 1978), L=(pij*pik/ai); where pij is the proportion of plant i of the total plants utilized by Mansonia immatures of species j; pik is the proportion of plant i of the total plants utilized by Mansonia immatures of species k, and ai is the proportional amount of plant i in the environment. This index takes a value of 0 if both species share no resources, it is 1 when the two species utilize each resource in proportion to its abundance, and >1 when both species utilize certain resources in selective ways and the preference of the two species tends to coincide (Krebs 1999). The abundance of Ma. indubitans and Ma. titillans was correlated (Spearman rank correlation) with the number of plants for each species. RESULTS During the five months of sampling, the proportion of host plant species was constant and the heterogeneous habitat showed the same pattern of abundance of vegetal species. The mean number of host plant species was compared for each month, but no significant differences were detected for P. stratiotes (K-W H=1.29, df=4, P=0.862), Salvinia spp. (KW H=2.05, df=4, P=0.725), and L. laevigatum (K-W H=8.03, df=4, P=0.09) (Figure 1). A total of 402 immatures of Ma. indubitans and 217 of Ma. titillans were captured. Significant association was noted between M. indubitans and the host plant species (K-W H=42.74, df=2, P<0.001). The same result was observed for M. titillans (K-W H=23.42, df=2, P<0.001). Immatures of M. indubitans were significantly more abundant on P. strtatiotes, and no differences were found between P. stratiotes and L. laevigatum for M. titillans. Both Mansonia were very scarce on Salvinia spp (Table 1). No signifficant differences of abundance per plant were noted between M. indubitans and M. titillans for any host plant species (K-W H=0.43, df=2, P=0.507). The frequency of both culicids on roots of each type of plant showed significantly higher frequencies of Ma. indubitans than Ma. titillans on roots of P. stratiotes (2=75.95, df=1, P<0.001), but no difference was registered on roots of L. laevigatum (2=2.47, df=1, P= 0.116). The calculated Hurlbert´s Index, L= 4.909, suggested that both

Mansonia species utilized roots of P. stratiotes in significantly major proportions rather than selecting resources at random. The total amount of Ma. indubitans immatures was significantly correlated to the plant number of P. stratiotes (Rs=0.716, P<0.001) per samplig unit. On the other hand, a negative correlation was registered between Ma. indubitans immatures and plant number of Salvinia spp. (Rs=-0.637, P<0.001). No significant relationship between Ma. indubitans and L. laevigatum was registered (Rs=-0.228, P=0.096). Similar results were obtained from Ma. titillans. This species showed a significant positive relationship with P. stratiotes (Rs=0.489, P<0.001), as well as significant negative association with Salvinia spp. (Rs=-0.579, P<0.001), and no significant correlation with L. laevigatum (Rs=-0.001, P=0.994). DISCUSSION A strong association between Mansonia immatures and Pistia stratiotes as host plants in subtropical areas of South America is documented (Neiff and Poi de Neiff 1978, Poi de Neiff 1983, García et al. 1995). However, previous works did not focus on the relationship between the abundance of immature instars of Mansonia and the availability of different species of host plants. Our results show a selective Pistia stratiotes pattern of root utilization by Mansonia indubitans and Ma. titillans in Macáes Pond. Compared with results obtained in Florida by Slaff and Haefner (1985) and Ferreira (1999) for the Amazon region, the abundance of Mansonia on P. stratiotes reached lower levels in Buenos Aires. The study area is near the southern extreme of the distribution of P. stratiotes (Morrone and Zuloaga 1996) and this plant undergoes seasonal growth in accordance with a report by Dewald and Lounibos (1990) in North America. As a result, the sampling was restricted to the period in which the coverage of each host plant was constant. Plant abundance and association patterns of both Mansonia showed similar trends, but in the case of M. titillans, analysis of abundance did not show differences between P. stratiotes and L. laevigatum. Conversely, frequency relationships between Ma. indubitans and Ma. titillans were 2:1 in the heterogeneous environment of Macáes Pond. A similar ratio concerning the pure stands of P. stratiotes was extremely biased to Ma. indubitans (García et al. 1995). However, it is interesting to point out that in Macáes Pond the frequency ratio of Ma. indubitans and Ma. titillans is equal on roots of L. laevigatum. These observations, in addition to our results, suggest that M. indubitans exhibits a closer relation with P. stratiotes, while Ma. titillans is associated with a broad range of plant species. Similar trends were detected for Ma. dyari compared to Ma. titillans in North America (Lounibos and Escher 1985, Slaff and Haefner 1985). Differential utilization of the roots of P. stratiotes, as compared to other species, could be related to diverse factors such as the oviposition behavior of female adults or characteristics of the different kind of roots used as habitat. Oviposition on P. stratiotes by adult Mansonia females (Mattingly 1972, Lounibos and Linley 1987, Lounibos and

Vol. 30, no. 2

Journal of Vector Ecology

203

Pistia stratiotes

40 35 30 25 20 15 10 5 0 Jan Feb Mar Apr May

Pistia stratiotes (n)

Limnobium laevigatum (n)

Salvinia spp. (n)

Figure 1. Mean number of host plant species ± SD obtained over five months, from January to May 2003, at Macáes Pond, Buenos Aires City, Argentina.

( ) Pi ti t

Limnobium laevigatum

60 50 40 30 20 10 0 Jan Feb Mar Apr May

( ) Li

Salvinia spp.

80

40

( )S l i i

0 Jan Feb Mar Apr May

204

Journal of Vector Ecology

December 2005

Table 1. Mean number of immatures ± SE (sampling units) and total immature frequency of Ma. indubitans and Ma. titillans collected from aquatic vegetation during January to May, 2003 at Macáes Pond, Buenos Aires City, Argentina.

Ma. indubitans Species Mean ± SE (sampling units) 0.783 ± 0.290 a* Total Ma. titillans Mean ± SE (sampling units) 0.273 ± 0.077 a

Total

Pistia stratiotes (n=446)

354 (88%)

157 (72%)

Limnobium laevigatum (n=617) Salvinia spp. (n= 2280) Total

0.088 ± 0.027 b

38 (10%)

0.113 ± 0.028 a

53 (25%)

0.007 ± 0.002 b

10 (2%)

0.007 ± 0.003 b

7 (3%)

402

217

* Means followed by different letters are significantly different. Kruskal-Wallis test for 3-way comparisons (P<0.01), Mann-Whitney for 2-way comparisons (P<0.05).

Dewald 1989, Ferreira and Nunes de Mello 1999) resulted in high densities of larval and pupal instars. On the other hand, roots of floating plants could have different characteristics, such as aerenchyma size or hardness of its tissues, that could be determinants for positive or negative selection by Mansonia immatures. In this sense, laboratory assays showed selective attachments of 4th instar Mansonia larvae on healthy plants (McDonald and Lu 1973). In our study, both Mansonia species appeared to avoid Salvinia spp. "roots" although this genus was dominant in Macáes Pond. The sizes of the host plant species were not measured. This issue is important because the absence of data on plant size precludes inferences on host plant selection proportional to macrophyte biomass. On the other hand, the particular relationship between niche overlap and competition is complex (Holt 1987). For this reason, the measures of Hurlbert´s Index obtained in our study may not imply interspecific competition between Mansonia species. Acknowledgments We thank A. Bachmann and S. Fischer for reviewing earlier drafts of this manuscript. We also thank the personnel of Reserva Ecológica Costanera Sur for logistical support in field work and Patricia Rodriguez, Valeria Sander, Luciano Patitucci, and Victoria Botazzi for their valuable cooperation in the field work. We would also like to thank the anonymous reviewer for his or her constructive suggestions.

REFERENCES CITED Almirón, W.R. and M.E. Brewer. 1996. Classification of immature stage habitats of Culicidae (Diptera) collected in Córdoba, Argentina. Mem. Inst. Oswaldo Cruz 91: 19. Balseiro, E.G. 1986. Análisis de la estructura de edades y disposición espacial en una población local de Aedeomyia squamipennis (Lynch Arribalzaga) (Diptera: Culicidae). Rev. Soc. Entomol. Arg. 44: 41-46. Dewald L.B. and L.P. Lounibos. 1990. Seasonal growth of Pistia stratiotes L. in south Florida. Aquat. Bot. 36: 263275. Ferreira, R.L.M. 1999. Densidade de oviposição, e quantificação de larvas e pupas de Mansonia Blanchard, 1901(Diptera: Culicidae), em Eichhornia crassipes Solms. e Pistia stratiotes Linn. na Ilha da Marchantaria, Amazonia Central. Acta Amazonica 29: 123-134. Ferreira, R.L.M., E. Silva Pereira, N.T. Ferreira Har, and N. Hamada. 2003. Mansonia spp. (Diptera: Culicidae) associated with two species of macrophytes in a Varzea lake, Amazonas, Brazil. Entomotrópica 18: 21-25. Ferreira, R. L. M. and J. A. S. Nunes de Mello. 1999. Aspectos biológicos de três espécies de Mansonia (Mansonia) Blanchard, 1901 (Diptera, Culicidae) em laboratório. Rev. Bras. Entomol. 43: 29-34. Forattini, O. P. 1965. Entomologia Médica. 3º Volume. Culicini: Haemagogus, Mansonia, Culiseta. Sabethini. Toxorhynchitini. Arboviroses. Filariose bancroftiana, Genética. Universidade de São Paulo. 416 pp.

Vol. 30, no. 2

Journal of Vector Ecology

205

García, J.J., R.E. Campos, and A. Macia. 1995. Observaciones ecológicas sobre Mansonia indubitans y Ma. titillans (Diptera: Culicidae) y sus enemigos naturales en Punta Lara, Argentina. Rev. Soc. Entomol. Arg. 54: 43-50. Greene, C.M. and J.A. Stamps. 2001. Habitat selection at low population densities. Ecology 82: 2091-2100. Holt, R.D. 1987. On the relation between niche overlap and competition: the effect of incommensurable niche dimensions. Oikos 48: 110-114. Hurlbert, S.H. 1978. The measurement of niche overlap and some relatives. Ecology 59: 67-77. Krebs, C.J. 1999. Ecological Methodology. 2nd Edition. University of Columbia. Lounibos, L.P. and L.B. Dewald. 1989. Oviposition site selection by Mansonia mosquitoes on water lettuce. Ecol. Entomol. 14: 413-422. Lounibos, L.P. and R.L. Escher. 1985. Mosquitoes associated with water lettuce (Pistia stratiotes) in Southeastern Florida. Florida Entomologist 68: 169-178. Lounibos, L.P. and J.R. Linley. 1987. A quantitative analysis of underwater oviposition by the mosquito Mansonia titillans. Physiol. Entomol. 12: 435-443. Mattingly, P.F. 1972. Mosquito Eggs XIX. Genus Mansonia (Subgenus Mansonioides Theobald). Mosq. Syst. 4: 5059. McDonald, J.L. and L.C. Lu. 1973. Preference of Mansonia uniformis (Theob.) for specific water hyacinth plants. Mosq. News 33: 466-467. Morrone, O. and F. Zuloaga. 1996. Araceae. In: Catálogo de Plantas Vasculares de la República Argentina. I.

Pteridophyta, Gymnospermae y Angiospermae (Monocotyledoneae). F. Zuloaga and O. Morrone (eds.). St. Louis. Missouri Botanical Garden Press. pp. 100-104. Neiff, J.J. and A. Poi de Neiff. 1978. Estudios sucesionales en los camalotales chaqueños y su fauna asociada. I. Etapa seral Pistia stratiotes-Eichhornia crassipes. Physis (B) 37: 29-39. Poi de Neiff, A. 1983. Observaciones comparativas de la mesofauna asociada a Pistia stratiotes L. (Araceae) en algunos ambientes acuáticos permanentes y temporarios (Chaco, Argentina). Physis (B) 41: 95-102. Poi de Neiff, A. and J.J. Neiff. 1980. Los camalotes de Eichhornia crassipes en aguas lóticas del Paraná y su fauna asociada. Ecosur 7: 185-199. Ronderos, R.A. and A.O. Bachmann. 1963. Mansoniini Neotropicales. I (Diptera-Culicidae). Rev. Soc. Entomol. Arg. 26: 57-65. Ronderos, R.A. and A.O. Bachmann. 1964. Mansoniini Neotropicales. II (Diptera-Culicidae). Com. Invest. Cient. 2: 4-9. Slaff, M. and J.D. Haefner. 1985. Seasonal and spatial distribution of Mansonia dyari, Mansonia titillans, and Coquillettidia perturbans (Diptera: Culicidae) in the Central Florida, USA, phosphate region. J. Med. Entomol. 22: 624-629. Sutherland, J.P. and G.M. Poppy. 1997. Spatial and temporal distribution of phidophagous syrphids (Diptera: Syrphidae) in sown wild flower patches in a winter barley crop. In: Species dispersal and land use processes. Cooper, A. and J. Power (eds.). Proc. Sixth Annu. Conf. IALE, Ulster, UK. pp. 135-142.

206

Journal of Vector Ecology

December 2005

Variability in natural populations of Anopheles sacharovi (Diptera: Culicidae) from southeast Anatolia, revealed by morphometric and allozymic analyses

H. Yurttas, B. Alten , and A. M. Aytekin

Hacettepe University, Faculty of Science, Department of Biology, Ecology Section, 06532 Beytepe, Ankara, Turkey Received 23 December 2004; Accepted 15 March 2005 ABSTRACT: Four populations of Anopheles sacharovi Favre occurring in different ecological subregions at altitudes between 353-1,126 m in the Sanliurfa Province of southeast Turkey were compared using morphometric and allozyme analyses. Four allozyme loci were assessed for genetic differentiation among samples from four localities. The similarity phenogram obtained from the allozyme data showed that populations at Birecik and Sandi branched as a separate group from the Pamuklu and Gedik populations. The Gedik population at the highest altitude (1,126 m) was clustered as a separate branch when linear measurements of 63 morphological characteristics were examined. The UPGMA phenogram also showed that Pamuklu and Sandi formed a cluster while Birecik and Gedik formed separate groups. Journal of Vector Ecology 30 (2): 206-212. 2005. Keyword Index: Anopheles sacharovi, geographical variation, traditional morphometrics, allozyme electrophoresis.

INTRODUCTION Turkey is situated on the edge of the subtropical zone in which certain vector-borne diseases are prevalent at endemic and occasionally epidemic proportions (Ramsdale and Haas 1978). Historically, malaria has been one of the most important vector-borne diseases in Anatolia (Merdivenci 1984, Akdur 1999). Malaria is not evenly distributed in the country and is endemic mainly in the provinces of the Southeastern Anatolia Irrigation Project (GAP) and a portion of eastern Anatolia (Kasap et al. 2000). At present, 23% of the total population of Turkey still lives in those provinces where malaria is endemic. In 2001, 84% of malaria cases that occurred in Turkey were recorded from the GAP area (Anonymous 2002). In southeastern Anatolia, malaria is focused in Sanliurfa province. The geographical, ecological, and socio-economical characteristics of the province play important roles in the distribution and epidemics of malaria (Alten et al. 2003). This disease has become a greater threat in recent years because the economic opportunities in the provinces of the GAP have attracted human populations to Sanliurfa. Malaria has been controlled and/or supressed in Sanliurfa between 2001 and 2003, yet based on a study of 1,306 individuals between 1999 and 2002, the rate of prevalance of malaria in Sanliurfa was between 1.90 to 8.29 % of the total population of 1.3 million in Sanliurfa (Alten et al. 2003). Previous studies have shown that there are thirteen Anopheles species recorded in Turkey (Ramsdale et al. 2001). Among them, Anopheles sacharovi Favre is the most common malaria vector followed by Anopheles maculipennis Meigen, Anopheles claviger Meigen, and Anopheles superpictus Grassi (Kasap et al. 1987, Kasap 1990, Özer et al. 2001). In the past, three indigenous parasite species have been reported in Turkey: Plasmodium malaria, P. falciparum, and P. vivax. Currently, although occasionally imported cases of P. falciparum are observed, all indigenous cases of malaria are

P. vivax (Anonymous 2002, Alten et al. 2003). Turkey has a variety of geographical, climatic, geological, and ecological conditions giving rise to a proliferation of different mosquito species. Furthermore, these differentiations may have also caused inter/intraspecific variations that are revealed by morphometric and allozymic analyses as in other insect species (Belen et al. 2004). The main objective of the current study was to investigate the effects of altitude as an important geographical factor on local populations of An. sacharovi. In light of previous studies done in the region (Alten et al. 2003, Belen et al. 2004), we have focused on four different eco-regions that are classified gradually in this very important endemic malaria site of Turkey. It was hypothesized that habitat differences associated with altitude may affect the morphological characteristics in a multivariate way relating to allozyme variability. MATERIALS AND METHODS The study area The study was carried out in Sanliurfa Province (37o 09' N; 38o 47' E), SE Anatolia, Turkey from May 2001 to October 2002. Sanliurfa Province, which is 18,500 km2, is located in the western part of the southeastern Anatolia region near the border of Syria. The city of Sanliurfa lies at 550 m and has a semi-arid climate with four distinct seasons: a very hot and dry summer (June-August; 31.5°C, 3.20 mm rainfall), a warm, wet autumn (September-October; 20.7°C, 14.14 mm rainfall), a moderately warm and rainy winter (December-February; 7.5°C, 71.33 mm rainfall), and a warm and rainy spring (March-May; 17.0°C, 37.03 mm rainfall). Based on the results of a previous study (Alten et al. 2003), four localities at different elevations were selected. The Birecik (BRC) region (37o01'N, 37o57'S) is the lowest region with an altitude of 353 m located in the west of the province, and the mean temperature is higher than that of the

Vol. 30, no. 2

Journal of Vector Ecology

207

north. The Sandi (SND) region (37o19'N, 39o34'S) has an altitude of 682 m. Pamuklu (PMK) region (37o36'N, 39o19'S) is at 743 m and Gedik (GDK) (37o45'N, 39o40'S) is the highest mountainous and rural area with an altitude of 1,126 m. The incidence of malaria ranged from 1.6 % to 9.8 % among these locations (Alten et al. 2003). The major plant species found in the region are: Astragalus aleppicus Boiss., Astragalus aduncus Willd., Onobrychis crista-galli (L.) Lam., Adonis dentata Del., Medicago orbicularis (L.) Bart. Eryngium criticum Lam., Trifolium pillulare Boiss., Linum mucronotum Bertol., and Salvia brachyntha (Brodz.) Pobed. The main agricultural products of the Sanliurfa region are; Pistacia vera L., Vitis sylvestris Gmelin, Triticum aestivum L., Oryza sativa L., and Gossipium spp. Apart from agriculture, livestock breeding is the major source of income in the area. Sheep are the most common livestock but cattle and chickens can also be found in some areas. Mosquito sampling An. sacharovi adults were collected by aspirating from houses and barns and from animal-bait traps made from polyester netting and baited with cows that were placed near houses of current or past malaria patients in each locality. Collections were also made using CDC miniature light - traps (John W. Hock Co. Florida, U.S.A.) during the summers of 2001 and 2002 (August-September). On each trapping night, four to six light traps were placed in each of the sampling localities. Houses and barns used as sampling stations varied from two-story cement block enclosures to simple brick, stone, or cement houses with basements, cellars, caves, or barns for keeping poultry or livestock. Field-collected live females (fed or gravid) were transported to the laboratory in polyethylene containers that were kept on ice for the later morphometric and allozyme studies (Urbanelli et al. 2000). Taxonomic identification was made using the keys and descriptions of Ross and Roberts (1943), Dubose and Curtin (1965), Glick (1992), Darsie and Samanidou (1997) and Schaffner et al. (2001). Identifications were reconfirmed using voucher specimens of An. sacharovi from Sanliurfa, Turkey. After identification, An. sacharovi females were separated from the other live flies and introduced into rearing cages (50x50x50 cm). About 400 females were collected from study areas, and 92 of them were randomly chosen for morphometric analyses and 250 for allozyme analyses. Laboratory studies Electrophoresis. The thoraces of females were ground up and homogenates were kept at -80º C until they were used for electrophoresis. Four enzyme systems, MDH (malate dehydrogenase, EC 1.1.1.37), PGI (phosphoglucoisomerase, EC 5.3.1.9), EST (esterase, EC 3.1.1), and HK (hexokinase, EC 2.7.1.1), were studied by horizontal starch-gel electrophoresis. Two enzyme systems (EST and PGI) were studied using the Tris-citrate, pH 7.0 buffer system (Shaw and Prasad 1970, Hillis and Moritz 1990). MDH was studied using the Tris-HCl, pH 8.6 buffer system (Shaw and Prasad 1970), while HK was studied using the Tris-malate-EDTA, pH 7.4 buffer system (Shaw and Prasad 1970). Sample and

gel preparation and experimental conditions were similar to those of Kandemir and Kence (1995). Statistical analysis of electrophoretic data was performed using the computer program BIOSYS-1 (Swofford and Selander 1981). Morphometric Analysis. All the specimens were screened for the presence of known parasites in order to prevent possible traumatic variations affecting the morphometric data (Mayr and Ashlock 1991). The body parts were then removed from each specimen with forceps and mounted on slides. All slides were photographed using a Leica MZ-7.5 stereoscopic zoom dissection microscope with a DC-300 digital camera system, digitized and archived. A total of 63 characters were measured using TPSdig (Rohlf 2003) software as follows: 1) length of wing (a-o), 2) width of wing, 3) length of subcosta (b-d), 4) distance between f-e points on r2 vein, 5) length of R-M (gh), 6) distance between n-k points on the Cu2, 7) distance between k-j points on the Cu2 vein, 8) distance between k-l points, 9) distance between h-m points on the m3 vein, 10) length of M-Cu1 (i-j) (Figure 1), 11) length of haltere, 12) length of the apical part of haltere, 13-36) [I) length of femur, II) Width of femur, III) length of tibia, IV) length of 1st segment of tarsus, V) length of 2nd segment of tarsus, VI) length of 3rd segment of tarsus, VII) length of 4th segment of tarsus, VIII) length of 5th segment of tarsus (fore, mid and hind legs 3 x 8 = 24 characters)], 37) length of head, 38) width of head, 39) length of compound eyes, 40) length of clypeus, 41) length of palpus 42) length of 1st and (plus) 2nd segment of palpus, 43) length of 3rd segment of palpus, 44) length of 4th segment of palpus, 45) length of 5th segment of palpus, 46) length of proboscis, 47) length of label, 48-62) length of total antenna, [ I) length of antennal segments (total 14 characters) ], 63) length of cercus. For paired organs, those on the right side were measured (Aytekin and Cagatay 2002). The collected data were tested for allometry by Huxley's model and transformed into natural logarithmic form. For statistical analysis of morphology in terms of size morphometry, the data were transformed into natural logarithmic form (Debat et al. 2003). The data were discriminated using a discriminant multigroup function analysis (Canonical variate analysis CANOVAR) by Syn-tax 2000 (Podani 2001) package (Exeter-Software, U.S.A.). The arithmetic means of the 63 morphometric measurements were clustered by Ward's method with PAST 2004 software package (Hammer et al. 2004). The significant differences among characters were also tested by one way ANOVA (Sokal and Rohlf 1981). RESULTS Electrophoresis results Allelic frequencies and genetic differentiation of all populations for the four loci are given together with their standard errors in Table 1. The similarity phenogram obtained from these data are presented in Figure 2. From Table 1 it appears that although the differentiation is not exact, the BRC and SND constitute one group while PMK and GDK another. Morphometric results Individual distribution and clusters obtained from the

208

Journal of Vector Ecology

December 2005

Table 1. Values of the genetic differentiation based on allozymes for four different population of Anopheles sacharovi collected from four different altitudes (Birecik (BRC) 353 m, Sandi (BRC) 682 m, Pamuklu (PMK) 743 m, and Gedik (GDK) 1,126 m), Sanliurfa, southeastern Turkey (May 2001-October 2002).

Allele frequencies Hk Mdh-1 Est-3 Pgi-1 Mean Mean number of sample size alleles per locus per locus 2.5 (.7) 2.5 (.7) 2.5 (.7) 2.5 (.7) Percentage of loci polymorphic* 50.0 50.0 50.0 50.0 Observed Expected (Hardy-Weinberg**) 303 (.189) 347 (.157) 307 (.184) 338 (.153) 298 (.184) 310 (.142) 323 (.192) 351 (.158) Mean heterozygosity

Population BRC SND PMK GDK

A

B

C

D

E

A

A 0.500 0.443 0.538 0.409

B

C

D

A

B

C

0.188 0.250 0.188 0.208 0.167 1.000 0.241 0.172 0.345 0.172 0.069 1.000 0.065 0.391 0.152 0.304 0.087 1.000 0.100 0.250 0.200 0.325 0.125 1.000

0.300 0.114 0.086 0.500 0.338 0.162 19.5 (6.5) 0.414 0.114 0.029 0.500 0.233 0.267 25.7 (8.1) 0.308 0.128 0.026 0.500 0.486 0.014 22.5 (7.2) 0.318 0.227 0.045 0.500 0.297 0.203 16.0 (5.6)

* A locus is considered polymorphic if the frequency of the most common allele does not exceed 0.95. ** Unbiased estimates (see Nei 1978). morphometric analysis are shown in Figure 3. For morphometrics, although the personal measurement error (æ ) p (Arnqvist and Mårtensson 1998) as well as the standardization and transformation procedure errors are kept at a minimum, the unexpected intravariation is still seen in four characters (length of hind leg segment, wing length, length of cerci, and length of haltere). The errors in accuracy and precision of the data are mostly due to ambiguity of the reference points of these characters, resulting in shifts during measurement and/ or the difficult procedure in the preparation and processing of these body parts. Errors due to two-dimensional viewing of a three-dimensional object can possibly be another reason. Analysis conducted taking the measurement errors into consideration clearly show that GDK (1,126 m) forms a distinct group from the other three regions (BRC, 353 m; SAND, 682 m; PMK, 743 m) for the 63 characters measured along the first two canonical axes (Figure 3). The differences between the local populations of An. sacharovi are clear when all populations are considered. SND and PMK populations are the closest ones along the first two axes. When an UPGMA tree is constructed, Gedik and Birecik populations show distinct group patterns independently from altitude and related factors (Figure 4). Means with standart deviation are also calculated for each character and some of them show significant differences (Table 2). DISCUSSION The distribution of An. sacharovi is from Italy to the former U.S.S.R. and China, and from Jordan to Israel, Syria, Iraq, Iran, Cyprus, and Turkey (Merdivenci 1984). Its range in Turkey includes the western, southern, and northern coastal plains and extends to the central plateau (Postiglione et al. 1973, Merdivenci 1984, Ramsdale et al. 2001). Increasing altitude and longitude may influence climatic factors such as precipitation and temperature that are important determinants relative to the abundance, life-history, and morphology of mosquitoes (Gleiser et al. 2000). The results presented here indicate that there are significant differences among populations. When allozyme data and phenograms are examined it was seen that BRC and SND branched as a separate group from the PMK and GDK populations (Figure 2, Table 1). It is difficult to identify the possible mechanisms of the genetic variability among these populations with these data. Urbanelli et al. (2000) showed that although there was genetic homogenity on a large scale for Aedes albopictus, genetic differentiation between closely situated sites (10-20 km) resulted from genetic drift, in line with low dispersal and founder effects, thus showing that isolation by distance could be effective even at close distances for mosquitoes. However, since no clear pattern of isolation by distance was observed (for example SND and PMK branched separately althought they were almost at the same altitude) human effects such as effective insecticide application appears to be more important. When analyzed by one-way ANOVA, the populations were found to be significantly different for 45 morphological characters (Table 2). The differences among the four populations primarily occurred in the wing, leg, head and haltere characters. This may be associated with factors such as climatic, ecological, or socio-biological effects. When whole size differences were analyzed by UPGMA, GDK and BRC grouped separately and PMK and SND clustered together (Figure 4). Although BRC and GDK populations live at different altitudes and climates, they have a similar morphology. Therefore, size similarity and dissimilarity does not seem to have an obvious environmental explanation like climatic differences brought on by altitude. This situation again hints towards some ideas like differential insecticide applications which might result in bottlenecks as a possible explanation for size variation. Size differences showed somewhat different results when analyzed by CANOVAR. This may indicate a directional selection or genetic drift, especially on the GDK population. Only the GDK population was found to be significantly different from the other populations (Figure 3). Because GDK is situated at the highest altitude, climatic factors could explain this difference. Wing

Figure 1. Characters used in morphometric measurements of wing.

Vol. 30, no. 2

Journal of Vector Ecology

209

Table 2. Arithmetic mean (± standard deviation) for the 45 morphological characters that show significant differences p < 0.05 among four populations of Anopheles sacharovi collected from four altitudes (Birecik (BRC) 353 m, Sandi (SND) 682 m, Pamuklu (PMK) 743 m, and Gedik (GDK) 1,126 m ), Sanliurfa, southeastern Turkey (May 2001- October 2002).

Characters (10-1 mm) 1. Length of wing 2. Width of wing 3. Length of subcosta 4. Distance between f-e points on r2 vein 5. Distance between n-k points on Cu2 vein 6. Distance between k-j points on Cu2 vein 7. Distance between k-l points 8. Distance between h-m points on m3 vein 9. Length of haltere 10. Length of the apical part of halter 11. Length of fore femur 12. Width of fore femur 13. Length of fore tibia 14. Length of 1st segment of fore tarsus 15. Length of 2nd segment of fore tarsus 16. Length of 3rd segment of fore tarsus 17. Length of 4th segment of fore tarsus 18. Length of 5th segment of fore tarsus 19. Length of mid femur 20. Width of mid femur 21. Length of mid tibia 22. Length of 1st segment of mid tarsus 23. Length of 4th segment of mid tarsus 24. Length of 5th segment of mid tarsus 25. Length of hind femur 26. Width of hind femur 27. Length of hind tibia 28. Length of 1st segment of hind tarsus 29. Length of 2nd segment of hind tarsus 30. Length of 3rd segment of hind tarsus 31. Length of 4th segment of hind tarsus 32. Length of 5th segment of hind tarsus 33. Length of head 34. Width of head 35. Length of compound eyes 36. Length of clypeus 37. Length of palpus 38. Length of 3rd segment of palpus 39. Length of 5th segment of palpus 40. Length of proboscis 41. Length of labellum 42. Length of 1st antennal segment 43. Length of 12th antennal segment 44. Length of 14 and 15th antennal segment 45. Length of cercus BRC SND PMK 9.69 ± 0.53 6.10 ± 0.49 5.53 ± 0.35 6.00 ± 0.45 6.60 ± 0.44 3.98 ± 0.27 GDK p

39.78 ± 3.38 38.07 ± 2.86 37.55 ± 2.29 41.85 ± 2.10 0.000 10.27 ± 0.74 9.83 ± 0.77 6.24 ± 0.56 5.82 ± 0.54 6.28 ± 0.51 6.10 ± 0.57 4.16 ± 0.31 6.12 ± 0.63 5.47 ± 0.41 6.00 ± 0.41 6.40 ± 0.54 3.94 ± 0.30 10.27 ± 0.69 0.000 6.61 ± 0.33 6.19 ± 0.47 6.62 ± 0.43 7.02 ± 0.36 4.24 ± 0.22 0.003 0.000 0.000 0.002 0.001 22.97 ± 2.15 21.71 ± 1.96 21.31 ± 1.86 23.39 ± 1.43 0.000 16.37 ± 1.25 16.00 ± 1.10 15.8 ± 0.83 17.15 ± 0.86 0.000

1.95 ± 0.24 1.83 ± 0.19 1.77 ± 0.14 2.02 ± 0.18 0.000 21.01 ± 1.29 19.94 ± 1.40 19.87 ± 1.23 21.37 ± 1.42 0.000 0.99 ± 0.15 0.83 ± 0.17 0.86 ± 0.15 0.89 ± 0.12 0.003 24.43 ± 1.54 23.50 ± 1.85 23.17 ± 1.16 25.28 ± 1.72 0.000 18.61 ± 1.16 17.93 ± 1.37 17.79 ± 0.99 18.92 ± 1.10 0.004 6.85 ± 0.51 6.67 ± 0.55 6.67 ± 0.40 7.12 ± 0.56 0.014 4.71 ± 0.41 2.82 ± 0.24 4.59 ± 0.54 2.65 ± 0.26 4.42 ± 0.27 2.66 ± 0.16 4.90 ± 0.38 2.87 ± 0.23 0.002 0.002

2.00 ± 0.12 1.90 ± 0.17 1.87 ± 0.14 1.97 ± 0.13 0.005 25.20 ± 1.63 24.08 ± 0.83 23.61 ± 1.57 25.44 ± 1.69 0.001 0.76 ± 0.11 0.66 ± 0.09 0.70 ± 0.10 0.74 ± 0.09 0.004 25.89 ± 1.82 25.25 ± 2.05 24.36 ± 1.53 26.61 ± 1.53 0.000 5.77 ± 0.38 3.60 ± 0.28 5.58 ± 0.42 3.43 ± 0.26 5.51 ± 0.34 3.40 ± 0.26 6.04 ± 0.57 3.58 ± 0.23 0.000 0.014

2.20 ± 0.14 2.05 ± 0.16 2.04 ± 0.15 2.10 ± 0.12 0.001 25.74 ± 1.62 24.54 ± 2.00 24.19 ± 1.36 26.29 ± 1.80 0.000 0.66 ± 0.09 0.60 ± 0.10 0.62 ± 0.08 0.67 ± 0.06 0.033 27.30 ± 2.06 26.39 ± 2.24 25.71 ± 1.44 28.18 ± 1.54 0.000 31.96 ± 2.34 30.63 ± 2.74 30.60 ± 2.11 32.74 ± 2.12 0.005 13.03 ± 0.94 12.33 ± 1.02 1.47 ± 0.81 13.32 ± 0.93 0.001 9.61 ± 0.83 5.79 ± 0.38 2.87 ± 0.18 7.72 ± 0.42 7.35 ± 0.34 4.69 ± 0.19 9.17 ± 1.17 5.27 ± 0.65 2.64 ± 0.23 7.48 ± 0.45 7.03 ± 0.36 4.56 ± 0.20 9.13 ± 0.64 5.30 ± 0.38 2.64 ± 0.22 7.39 ± 0.28 6.92 ± 0.35 4.52 ± 0.15 9.95 ± 0.69 5.56 ± 0.58 2.78 ± 0.39 7.59 ± 0.45 7.04 ± 0.38 4.62 ± 0.29 0.005 0.002 0.010 0.044 0.001 0.011

3.12 ± 0.29 2.83 ± 0.26 2.88 ± 0.20 2.86 ± 0.28 0.001 21.15 ± 1.67 20.47 ± 1.55 20.25 ± 1.55 21.74 ± 1.39 0.007 7.94 ± 0.65 2.16 ± 0.22 7.57 ± 0.62 2.11 ± 0.16 7.44 ± 0.65 2.12 ± 0.20 7.98 ± 0.68 2.30 ± 0.19 0.013 0.003

22.50 ± 1.16 21.63 ± 1.32 21.22 ± 1.20 22.65 ± 1.07 0.000 2.02 ± 0.11 1.93 ± 0.14 1.87 ± 0.11 1.96 ± 0.13 0.001 0.93 ± 0.42 1.20 ± 0.11 1.24 ± 0.11 1.97 ± 0.17 0.76 ± 0.08 1.13 ± 0.09 1.18 ± 0.10 1.76 ± 0.15 0.81 ± 0.12 1.12 ± 0.08 1.16 ± 0.10 1.70 ± 0.18 0.77 ± 0.11 1.17 ± 0.06 1.20 ± 0.10 1.73 ± 0.11 0.040 0.011 0.032 0.000

210

Journal of Vector Ecology

December 2005

Figure 2. Similarity phenogram obtained from the data of allozymes for four different populations of Anopheles sacharovi collected from four altitudes (Birecik (BRC) 353 m, Sandi (SND) 682 m, Pamuklu (PMK) 743 m, and Gedik (GDK) 1,126 m), Sanliurfa, southeastern Turkey (May 2001-October 2002).

length is a determinant of body size because it is a relatively fixed character and is easily measured. Mean wing length value is the highest in GDK populations. An explanation for larger body size might be the influence of developmental temperature on growth. When mosquito larvae develop at lower temperatures, they grow more slowly and complete development at a larger size (Tun-Lin et al. 2000, Alto and Juliano 2001). Gleiser et al. (2000) found that in the floodwater mosquito, Aedes albifasciatus, the distribution of wing length varied seasonally and was correlated with both rainfall and breeding site volume. GDK has the highest rainfall and this might explain the observed pattern. We did not find any significant correlation between size variation and allozymes with the incidence of malaria in the region. However, it was interesting to observe the higher vector density in BRC for the future studies.

In conclusion, these morphometric and genetic approaches indicate that geographic variations exist among local populations of A. sacharovi in Sanliurfa. The morphometrical studies conducted here are based only on size differences. Future studies could consider size and shape differences using geometric modeling. With increased pressure to use biological control methods, especially those which rely upon altering the genetic structure of a population, greater attention must be given to understanding the life history characteristics of mosquito populations in the field.

Figure 3. Population distribution of Anopheles sacharovi collected at four different altitudes (Birecik (BRC) 353 m, Sandi (SND) 682 m, Pamuklu (PMK) 743 m, and Gedik (GDK) 1,126 m), Sanliurfa, southeastern Turkey (May 2001October 2002), along the first two principle axes (CANOVAR) based on 63 morphological characters. Numbers within the scatters indicate identification numbers of each individual. 1-23 GDK, 24-46 BRC, 47-69 PMK, and 70-92 SND.

Vol. 30, no. 2

Journal of Vector Ecology

211

Figure 4. The cluster based on the arithmetic means of 63 morphological characters measured from Anopheles sacharovi collected at four different altitudes (Birecik (BRC) 353 m, Sandi (SND) 682 m, Pamuklu (PMK) 743 m, and Gedik (GDK) 1,126 m), Sanliurfa, southeastern Turkey (May 2001-October 2002), (Ward's method).

REFERENCES CITED Akdur, R. 1999. Malaria. In: Ali Ozcel (ed.) The epidemiology of malaria. Turkiye Parasitoloji Dernegi (Turkish). 295 pp. Alten, B., S. Ça--lar, S. Kaynas, and F. Simsek. 2003. Effect of insecticide-treated bednets for malaria control in Southeast Anatolia, Turkey. J. Vector Ecol. 28: 97-107. Alto, B.W. and S.A. Juliano. 2001. Precipitation and temperature effects on populations of Aedes albopictus (Diptera: Culicidae): Implications for range expansion. J. Med. Entomol. 38: 646-656. Anonymous. 2002. Report on the informal consultation on the elimination of residual malaria foci and prevention of re-introduction of malaria, Rabat, Morocco. WHOEM/MAL/285/E/L, Cairo, Egypt. 24 pp. Arnqvist, G. and T. Mårtensson. 1998. Measurement error in geometric morphometrics: emprical strategies to assess and reduce its impact on measures of shape. Acta Zool. 44: 73-96. Aytekin A.M. and N. Cagatay. 2002. A phenetic approach to the subgenera of bumblebees (Apidae: Hymenoptera). Mellifera. 2: 60-64. Belen, A., B. Alten, and A.M. Aytekin. 2004. Altitudinal variation in morphometric and molecular characteristics of Phlebotomus (Phlebotomus) papatasi populations. Med. Vet. Entomol. 18: 343-350. Darsie, R.F. and A. Samanidou-Voyadjoglou. 1997. Keys for the identification of the mosquitoes of Greece. J. Am. Mosq. Contr. Assoc. 13: 231-241. Debat, V., M. Bégin, H. Legout, and J.R. David. 2003. Allometric and nonallometric components of Drosophila wing shape respond differently to developmental temperature. Evolution 57: 2773-2784. Dubose, P. and T. Curtin. 1965. Identification keys to the adult and larval mosquitoes of the Mediterranean area. J. Med.

Entomol. 1: 349-355. Gleiser, R.M., J. Urrutia, and D.E. Gorla. 2000. Body size variation of the floodwater mosquito Aedes albifasciatus in Central Argentina. Med. Vet. Entomol. 14: 38-43. Glick, J.I. 1992. Illustrated key to the female Anopheles of Southwestern Asia and Egypt (Diptera: Culicidae). Mosq. Syst. 24: 125-153. Hammer, Ø., D.A.T. Harper, and P.D. Ryan. 2004. Paleontological Statistics, ver 1.25. Hillis, D. M. and C. Moritz. 1990. Molecular systematics. Sinauer, Sunderland, MA. U.S.A. 588 pp. Kandemir, I. and A. Kence. 1995. Allozyme variation in a central Anatolian honeybee (Apis mellifera L.) population. Apidologie. 26: 503-510. Kasap, H., M. Kasap, O. Demirhan, and D. Alptekin. 1987. Development of Plasmodium vivax in Anopheles superpictus under experimental conditions. Am. J. Trop. Med. Hyg. 27: 241-245. Kasap, H. 1990. Comparison of experimental infectivity and development of Plasmodium vivax in Anopheles sacharovi and Anopheles superpictus in Turkey. Am. J. Trop. Med. Hyg. 42: 111-117. Kasap, H., M. Kasap, D. Alptekin, U. Luleyap, and P.R.J. Herath. 2000. Insecticide resistance in Anopheles sacharovi Favre in southern Turkey. Bull. Wld. Hlth. Org. 78: 687-692. Mayr, E. and P.D. Ashlock. 1991. Principles of Systematic Zoology, 2nd ed. McGraw-Hill Inc., London. Merdivenci, A. 1984. Mosquitoes of Turkey. Istanbul Universitesi, Cerrahpasa Tip Fakultesi, Publ. No: 3215, Istanbul (Turkish) 354 pp. Nei, M. 1978. Genetic distance between populations. Am. Nat. 106: 283-292. Özer, N., B. Alten, and S.S. Ça--lar. 2001. Distribution of malaria vectors in Turkey. Proceedings of The Firsth Balkan Conference, Malaria and Mosquito Control.

212

Journal of Vector Ecology

December 2005

Serres-Greece. 56-60. Podani, J. 2001. SYN-TAX 2000 computer programs for data analysis ecology and systematics. + software users manual, Budapest. Postiglione, M., S. Tabanli, and C.D. Ramsdale. 1973. The Anopheles of Turkey. Riv. Parasitol. 34: 127-159. Ramsdale, C.D. and E. Haas. 1978. Some aspects of epidemiology of resurgent malaria in Turkey. Trans. R. Soc. Trop. Med. Hyg. 72: 570-580. Ramsdale, C.D., B. Alten, S.S. Ça--lar and N. Ozer. 2001. A revised annoted checklist of mosquitoes (Diptera: Culicidae) of Turkey. J. Eur. Mosq. Bull. 9: 18-28. Rohlf, F.J. 2003. TPSdig. Software. Version 1.38. Ecology and Evolution, SUNY Stony Brook. http:// life.Bio.SUNYSB.edu/morph/morph.html. Ross, E.S. and H.R. Roberts. 1943. Mosquito atlas part II. Eighteen old world Anophelines important to malaria. Am. Entomol. Soc. Acad. Nat. Sci. 39 pp. Schaffner, F., G. Angel, B. Geoffroy, J.-P. Hervy, A. Rhaiem,

and J. Brunhes. 2001. Les moustiques d'Europe, The mosquitos of Europe. CD-ROM. Institut de Recherche pour Development /EID Mediterranée. Shaw, C.R. and R. Prasad. 1970. Starch gel electrophoresis-a compilation of recipes. Biochem. Genet. 4: 297-320. Sokal, R.R and F.J. Rohlf. 1981. Biometry. The principles and practice of statistics in biological research. W.H. Freeman. 859 pp. Swofford, D.L. and R.B. Selander. 1981. Biosys-1: a FORTRAN program for the comprehensive analysis of electrophoretic data in population genetics and systematics. J. Hered. 72: 281-283. Tun-Lin, W., T.R. Burkot, and B.H. Kay. 2000. Effects of temperature and larval diet on development rates and survival of the dengue vector Aedes aegypti in north Queensland, Australia. Med. Vet. Entomol. 14: 31-37. Urbanelli, S., R. Bellini, M. Carrieri, P. Sallicandro and G. Celli. 2000. Population structure of Aedes albopictus (Skuse): the mosquito which is colonizing Mediterranean countries. Heredity. 84: 331-337.

Vol. 30, no. 2

Journal of Vector Ecology

213

Evidence to support two conspecific cytological races of Anopheles aconitus in Thailand

Anuluck Junkum1, Narumon Komalamisra2, Atchariya Jitpakdi1, Narissara Jariyapan1, Gi-Sik Min3, Mi-Hyun Park3, Kang-Hyun Cho3, Pradya Somboon1, Paul A Bates4, and Wej Choochote1

2

Department of Parasitology, Faculty of Medicine, Chiang Mai University, Chiang Mai 50200, Thailand Department of Medical Entomology, Faculty of Tropical Medicine, Mahidol University, Bangkok 10400, Thailand 3 Department of Biological Sciences, Inha University, Incheon 402-751, Korea 4 Molecular and Biochemical Parasitology Group, Liverpool School of Tropical Medicine, University of Liverpool, Liverpool, United Kingdom

1

Received 17 December 2004; Accepted 18 April 2005 ABSTRACT: Iso-female lines (isolines) of Anopheles aconitus collected from Mae Hong Son, Phet Buri, and Chiang Mai Provinces were successfully identified to karyotypic forms. The results of identification revealed that An. aconitus Form B (X1, X2, Y2) was obtained from four and 48 isolines in Phet Buri and Chiang Mai Provinces, respectively, and Form C (X1, X2, Y3) was recovered from three and 41 isolines in Mae Hong Son and Chiang Mai Provinces, respectively. When comparing band to band on the same arm of ovarian nurse cell polytene chromosomes of An. aconitus Form B (Phet Buri: four isolines) and C (Mae Hong Son: three isolines, Chiang Mai: 20 isolines) to the standard chromosome mapping of An. aconitus Form B (Chiang Mai: 20 isolines), no major chromosomal rearrangements that related to the karyotype variations were demonstrated. The investigations on allelic frequencies of 4th stage larvae and adult females of three (Form C: Mae Hong Son), four (Form B: Phet Buri), 41 (Form C: Chiang Mai) and 48 (Form B: Chiang Mai) isolines suggested that An. aconitus Form B and C of all strains have similar allelic frequencies. This was observed at 10 isoenzymes 16 loci in 4th stage larvae, and 11 isoenzymes 13 loci in adult females. Hybridization tests among the four laboratory-raised isolines of An. aconitus Form B (Chiang Mai and Phet Buri) and C (Chiang Mai and Mae Hong Son) were employed by induced copulation. The results of crosses indicated that they were genetically compatible, providing viable progeny and completely synaptic salivary gland polytene chromosomes. The complete sequences of rDNA internal-transcribed spacer two (ITS2) and partial sequences of mitochondrial cytochrome c oxidase subunit I and II (COI and COII) from genomic DNA of 12 isolines of An. aconitus Form B and C were identified. Total sequence lengths (ITS2+COI+COII) of An. aconitus isolines varied from 1550bp to 1556bp. Conspecific relationships between the two An. aconitus forms were well supported by low values of intraspecific distances (ranged from 0.1% to 1.0%) and genetic differentiation (dxy: 0.01322) between the two forms. Based on evidence of no pre- and post-mating isolations, and nearly identical of DNA sequences of ITS2, COI and COII regions between An. aconitus Form B and C, we conclude that they are conspecific cytological races in the Thai population. Journal of Vector Ecology 30 (2): 213-224. 2005. Keyword Index: Anopheles aconitus, isoenzyme, hybridization, internal transcribed spacer 2, cytochrome c oxidase subunit I and II.

INTRODUCTION The Myzomyia series in Thailand consists of at least six Anopheles (Cellia) species, i.e., An. aconitus Donitz, An. culicifacies Giles, An. jeyporiensis James, An. minimus Theobald, An. pampanai Buttiker and Beales, and An. varuna Iyengar (Harrison 1980). Among these species, An. minimus and An. aconitus are considered as respective primary and secondary vectors of malaria in Thailand (Scanlon et al. 1968, Harrison 1980). The primary vector belongs to a species complex comprising two sibling species, A and C. The former is found throughout the country, while the latter is limited to Kanchanaburi Province (Sucharit et al. 1988, Green et al. 1990, Baimai et al. 1996a). Based on metaphase karyotype studies, at least three karyotypic forms of An. aconitus [Form A (X1, X2, Y1), B (X1, X2, Y2), and C (X1, X2, Y3)] have been

reported sympatrically from Maetang District, Chiang Mai Province, northern Thailand, whereas Form D (X3, X4, Y4) has been found from only Java, Indonesia (Baimai et al. 1996a). In vector studies of Thailand, significant progress has been made in the population genetics of the primary vectors, i.e., An. dirus ( Baimai et al. 1987, Green et al. 1992a, Walton et al. 1999), An. minimus (Sucharit et al. 1988, Komalamisara 1989, Green et al. 1990, Baimai et al. 1996a, Sharpe et al. 2000), and An. maculatus (Green et al. 1985, Takai et al. 1987, Chabpunnarat 1988, Baimai et al. 1993, Rongnoparut et al. 1999), and the secondary vectors, i.e., An. pseudowillmori (Green et al. 1992b) and An. sundaicus (Baimai et al. 1996b, Sukowati and Baimai 1996, Sukowati et al.1999). Investigation of the role of karyotype variation in generating pre-mating barriers by using ovarian nurse cell polytene

214

Journal of Vector Ecology

December 2005

chromosomes and isoenzymes, post-mating barriers by hybridization experiments, and DNA sequence variations of some specific genomic markers, have never been performed in a systematic way for An. aconitus, which has three karyotypic forms. Thus, the ovarian nurse cell polytene chromosome and isoenzyme investigations, hybridization, and comparative DNA sequencing of internal transcribed spacer 2 (ITS2) and mitochondrial cytochrome c oxidase subunit I and II (COI and COII) of sympatric and allopatric populations of An. aconitus Form B and C from Thailand are presented. MATERIALS AND METHODS Metaphase karyotype identification Wild-caught, fully engorged females of An. aconitus were collected from the endemic areas of malaria, Ban Pang Mai Daeng, Maetang District, Chiang Mai Province, northern Thailand [the area where Baimai et al. (1996a) incriminated the three karyotypic forms of An. aconitus]; Ban Huai Pong Kan, Muang District, Mae Hong Son Province, northwest Thailand, and Ban Tha Salao, Nong Ya Plong District, Phet Buri Province, southwest Thailand (Figure 1), by using both human-baited and buffalo-baited traps. They were allowed to individually oviposit eggs in isolated ovipots to establish isolines. The larvae were colonized further using the techniques described by Choochote et al. (1983). Metaphase chromosomes were prepared from the ovaries and testes of newly-emerged adult F1- and/or F2-progenies of each isoline

using the method of Choochote et al. (2001). Identification of karyotypic forms followed the cytotaxonomic key of Baimai et al. (1996a). All identified isolines were used for the experiments. Ovarian nurse cell polytene chromosome investigation The technique for chromosome preparation was as described by Green (1972). Females of the F1- and/or F2progenies from each isoline of An. aconitus were allowed to feed on golden hamsters. Engorged females at the semi-gravid stage with Christopher's stage-III ovaries were processed for chromosome preparation. The prepared chromosomes were scrutinized under a compound microscope using a green filter (Olympus: VANOX-AH2-PC) to search for the inverted heterozygote and homozygote of paracentric inversions. Isoenzyme investigation The method of electrophoresis followed that of Komalamisara (1989). Nineteen isoenzymes of 4th stage larvae and adult females from each isoline of An. aconitus were investigated. The enzymes were aldehyde oxidase (Aldox), acid phosphatase (Acp), alkaline phosphatase (Alp), esterase (Est), fumerase (Fum), hexokinase (Hk), isocitrate dehydrogenase (Idh), leucine aminopeptidase (Lap), lactate dehydrogenase (Ldh), malic dehydrogenase (Mdh), malic enzyme (Me), glucose phosphomutase (Gpm), glucose phosphate isomerase (Gpi), xanthine dehydrogenase (Xdh), a-glycerophosphate dehydrogenase (a-Gpdh), peptidase (Pep), hydroxybutyrate dehydrogenase (Hbdh), octanol dehydrogenase (Odh), and 6-phosphogluconate dehydrogenase (6-Pgd). Isoenzymes were numbered with respect to increasing anodal migration. Allozymes were named numerically according to their mobility relative to the most common allele (= 100). Hybridization study Crossing experiments were conducted among the two representative karyotypic forms of four laboratory-raised isolines of three strains (Chiang Mai, Mae Hong Son, and Phet Buri) of An. aconitus. Hybridization followed the method reported by Choochote et al. (2002a). The salivary gland polytene chromosomes of 4th instar larvae from the crosses were prepared using the techniques described by Kanda (1979). Low viability of the crosses (hatchability, survival, pupation, emergence, adult sex-distortion, abnormal morphology, and reproductive system), and asynapsis of polytene chromosomes were the criteria used to establish postmating isolation. DNA extraction, PCR amplification, cloning, and sequencing Genomic DNA was extracted from individual adult mosquitoes using a DNeasy Tissue kit (Qiagen, Co.) according to manufacturer's instructions. The rDNA ITS2, and mitochondrial COI and COII were amplified using the primers described in Park et al. (2003) and Folmer et al. (1994): 18S+1600 (5'-GCG TTG ATT ACG TCC CTG CCC TTT G-3') and 28S-60 (5'-GTT GGT TTC TTT TCC TCC-3')

Figure 1. Map of Thailand showing Chiang Mai (CM), Mae Hong Son (MS), and Phet Buri (PB) Provinces, where mosquito collections were performed. Chiang Mai Province is situated on latitude 18 o 47' N and longitude 98 o 59' E in northern Thailand and is approximately 97 and 647 kilometers away from MaeHong Son Province, northwest Thailand and Phet Buri Province, southwest Thailand, respectively.

Vol. 30, no. 2

Journal of Vector Ecology

215

Figure 2. Metaphase karyotypes of An. aconitus Form B and C (Giemsa staining).Testis chromosomes; Form B: (A) Chiang Mai strain, showing X1, Y2-chromosomes; (B) Phet Buri strain, showing X2, Y2-chromosomes; Form C: (C) Chiang Mai strain, showing X1, Y3-chromosomes; (D) Mae Hong Son strain, showing X2, Y3-chromosomes. Ovary chromosomes: (E) showing homozygous X1, X1-chromosomes, (F) showing homozygous X2, X2-chromosomes, (G) showing heterozygous X1, X2-chromosomes. Note, all types of X-chromosomes were found in all forms and strains of An. aconitus.

Figure 3. Standard map of ovarian nurse cell polytene chromosomes of An. aconitus Form B (Chiang Mai strain).

216

Journal of Vector Ecology

December 2005

for ITS2; LCO1490 (5'-GGT CAA CAA ATC ATA AAG ATA TTG G-3') and HCO2198 (5'-TAA ACT TCA GGG TGA CCA AAA AAT CA-3') for COI; AnoCO2+1 (5'-GAT TAG TGC AAT GAA TTT AAG C-3') and AnoCO2END (5'-GAG ATC ATT ACT TGC TTT CAG TC-3') for COII. PCR conditions were as follows: one cycle of 2 min at 94 oC; 35 cycles of 15 s at 95 oC, 30 s at 56 oC and 2 min at 72 oC and final cycle of 5 min at 72 oC. The PCR fragments were gel purified using a GeneClean kit (Q·BIOgene, Co.). The PCRamplified DNA fragments were cloned using the pGEM®-T Easy Vector Systems (Promega, Co.). Positive clones were selected by clonal PCR using identical primers used in original PCR amplifications. Plasmid DNAs from selected clones were extracted using a QIAprep miniprep kit (Qiagen, Co.). The purified samples were subjected to sequencing in an ABI PRISM® 3700 DNA Analyzer (Applied Biosystems, Co.) using a Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Co.). The newly reported sequences were deposited in the GenBank nucleotide sequence database library under accession numbers DQ000241-DQ000276 (Table 1). Sequence data analysis Both strands were sequenced and aligned using the CLUSTALX multiple alignment program (Thompson et al. 1997). Geographical and cytological types of specimens, symbols used in figures, and their sequence accession numbers within Genbank are denoted in Table 1. Estimates of Kimura two-parameter distances (Kimura 1980) were determined with the MEGA version 3.0. To explore the conflict between data sets (Farris et al. 1994), the partition homogeneity test was applied to the combined data matrix (100 randomizations) with PAUP ver. 4.0 b10 (Swofford 1999). The DnaSP version 3.99 software (Rozas et al. 2003) was used for the analysis of polymorphism and genetic differentiation. The level of polymorphism was estimated as the number of polymorphism (S), the average number of pairwise nucleotide differences (k), nucleotide diversity (p) (Nei 1987), and the average number of nucleotide substitutions per site between groups or populations (d xy). The statistical significance of genetic differentiation between groups, as estimated by Ks* was established by the permutation test (Hudson et al. 1992). RESULTS Metaphase karyotype identification The results of investigations of F1- and/or F2-progenies of three, four, and 89 isolines of An. aconitus strains from Mae Hong Son, Phet Buri, and Chiang Mai Province revealed the two forms of metaphase karyotypes, i.e., Form B (X1, X2, Y2), and C (X1, X2, Y3) (Figure 2). Form B was obtained from four and 48 isolines of Phet Buri and Chiang Mai strains, respectively, and Form C was recovered from three and 41 isolines of Mae Hong Son and Chiang Mai strains, respectively. It was interesting to note that An. aconitus Form A (X1, X2, Y1) collected from Maetang District, Chiang Mai Ovarian nurse cell polytene chromosome investigation Only An. aconitus Form B and Form C were obtained from this study, therefore, the standard homozygote of ovarian nurse cell polytene chromosomes were prepared from F1- and/ or F2-progenies of An. aconitus Form B. Standard map and/ or zone division of the five arms of chromosomes (20 complete sets of chromosomes, one set from each isoline) were based on the former reports of An. aconitus salivary gland polytene chromosomes (Sharma et al. 1980) and followed the convention of White et al. (1975) in Anopheles. Thus, the ovarian nurse cell polytene chromosomes of An. aconitus Form B were divided into a total of 46 zones, numbered from

Figure 4. Salivary gland polytene chromosome of F1-hybrid 4th stage larvae of An. aconitus:(A) Form B female (Chiang Mai strain) x Form B male (Phet Buri strain), (B) Form B female (Chiang Mai strain) x Form C male (Chaing Mai strain), (C) Form B female (Chiang Mai strain) x Form C male (Mae Hong Son strain). All the crosses yielded complete synapsis in all arms, except floating, heterozygous inversion (INV) on 2L (C), which was found in only one preparation from the F1-hybrid 4th stage larvae of Form B female (Chiang Mai strain) x Form C male (Mae Hong Son strain).

Province, and Sadao District, Songkla Province in October 1983, and subsequently reported by Baimai et al. (1996a), was not found in our study.

Vol. 30, no. 2

Table 1. Geographical origin of mosquitoes and their GenBank accession numbers.

Symbol of isoline ITS2 COI COII Region Genbank accession number Geographical origin Reference

Mosquito species

Length of ITS2 (bp)

An. aconitus Form B

Journal of Vector Ecology

An. aconitus Form C

An. aconitus*

An. minimus A

378 379 378 376 381 379 378 379 378 378 377 381 378 373

1BPB 2BPB 3BPB 1BCM 2BCM 3BCM 1CCM 2CCM 3CCM 1CMS 2CMS 3CMS Aconi Aconi Aconi -

ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2, COI, COII ITS2 COI COII ITS2

DQ000241 DQ000242 DQ000243 DQ000244 DQ000245 DQ000246 DQ000247 DQ000248 DQ000249 DQ000250 DQ000251 DQ000252 AJ626946 AF194504

DQ000253 DQ000254 DQ000255 DQ000256 DQ000257 DQ000258 DQ000259 DQ000260 DQ000261 DQ000262 DQ000263 DQ000264 AY423055 -

DQ000265 DQ000266 DQ000267 DQ000268 DQ000269 DQ000270 DQ000271 DQ000272 DQ000273 DQ000274 DQ000275 DQ000276 AJ512744 -

PB PB PB CM CM CM CM CM CM MS MS MS SL XM KB

This study This study This study This study This study This study This study This study This study This study This study This study Garros et al. 2004 Chen et al. 2003 Sharpe et al. 2000

PB = Phet Buri, Thailand; CM = Chiang Mai, Thailand; MS = Mae Hong Son, Thailand; SL = Sri Lanka; XM = Xiangming, China; KB = Kanchanaburi, Thailand. *Karyotypic forms were not determined.

217

218

Journal of Vector Ecology

December 2005

Table 2. Allelic frequencies (allele numbers) observed at 10 isoenzymes 16 loci in 4th stage larvae of An. aconitus Form B (Chiang Mai and Phet Buri strains) and C (Chiang Mai and Mae Hong Son strains).

An. aconitus Form No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Loci Est-5 Idh-2 Lap-2 Lap-3 Lap-4 Lap-5 Lap-6 Ldh-1 Ldh-2 Me Gpi Gpd Acp Alp Pep-1 Pep-2 Allele 100 104 100 105 108 100 102 100 100 100 100 102 100 98 100 100 100 105 100 102 95 100 102 98 100 100 102 100 CM 0.78 (56) 0.22 (16) 0.80 (16) 0.15 (3) 0.05 (1) 0.92 (22) 0.08 (2) 1.00 (44) 1.00 (44) 1.00 (44) 0.73 (32) 0.27 (12) 1.00 (16) 0.11 (6) 0.89 (48) 1.00 (24) 0.82 (18) 0.18 (4) 0.96 (25) 0.04 (1) 0.25 (2) 0.75 (6) 0 (0) 0.40 (4) 0.60 (6) 0.86 (12) 0.14 (2) 1.00 (10) B PB 1.00 (6) 0 (0) 1.00 (6) 0 (0) 0 (0) 1.00 (6) 0 (0) 1.00 (6) 1.00 (6) 1.00 (6) 1.00 (6) 0 (0) 1.00 (6) 1.00 (6) 0 (0) 1.00 (6) 1.00 (6) 0 (0) 1.00 (6) 0 (0) 0.75 (6) 0.25 (2) 0 (0) 0.50 (2) 0.50 (2) 0.33 (2) 0.67 (4) 1.00 (4) CM 0.89 (52) 0.11 (6) 0.83 (15) 0.11 (2) 0.06 (1) 0.83 (10) 0.17 (2) 1.00 (26) 1.00 (26) 1.00 (26) 0.83 (20) 0.17 (4) 1.00 (20) 0.04 (2) 0.96 (46) 1.00 (12) 0.78 (14) 0.22 (4) 1.00 (14) 0 (0) 0.29 (4) 0.43 (6) 0.28 (4) 0.14 (2) 0.86 (12) 0.40 (4) 0.60 (6) 1.00 (4) C MS 0.33 (2) 0.67 (4) 1.00 (6) 0 (0) 0 (0) 1.00 (6) 0 (0) 1.00 (6) 1.00 (6) 1.00 (6) 1.00 (6) 0 (0) 1.00 (6) 1.00 (6) 0 (0) 1.00 (6) 1.00 (6) 0 (0) 1.00 (6) 0 (0) 0.67 (4) 0.33 (2) 0 (0) 0 (0) 1.00 (4) 0.33 (2) 0.67 (4) 1.00 (6)

CM: Chiang Mai; MS: Mae Hong Son; PB: Phet Buri. 1 to 46. In the map (Figure 3), there are six zones (1-6) in chromosome X, 13 zones (7-19) in chromosome 2R (2), nine zones (20-28) in chromosome 2L (3), nine zones (29-37) in chromosome 3R (4), and nine zones (38-46) in chromosome 3L (5). When comparing bands on the same arm of three (Form C: Mae Hong Son strain), four (Form B: Phet Buri strain), and 20 (Form C: Chiang Mai strain) complete sets of chromosomes to standard chromosomes, no major chromosomal rearrangements that related to the karyotype variations were demonstrated. Isoenzyme investigation Sixteen out of 19 isoenzymes could be detected and analyzed. Details of allelic frequencies are shown in Table 2 and 3. The results of investigations on allelic frequencies of 4th stage larvae and adult females of three (Form C: Mae Hong Son strain), four (Form B: Phet Buri strain), 41 (Form C: Chiang Mai strain), and 48 (Form B: Chiang Mai strain) isolines indicated that An. aconitus Form B and C of all strains had similar allelic frequencies observed at ten isoenzymes 16 loci in 4th stage larvae, and 11 isoenzymes 13 loci in adult females. Hybridization study Details of hatchability, pupation and emergences of parental, reciprocal, and back-crosses among four isolines of three strains of An. aconitus Form B and C are displayed in Table 4. Observations on the hatchability, pupation, emergence, and adult sex-ratio of parental, reciprocal, and back-crosses among four isolines that represented two karyotypic forms, revealed that all crosses yielded viable progenies, and no evidence of genetic incompatibility was observed between An. aconitus Form B and C. The hatchability, pupation, emergence rates, and adult female/male ratio of parental, reciprocal, and back-crosses were 66.8586.75%, 81.16-91.87%, 94.39-97.30% and 0.77-1.40; 70.8987.10%, 82.96-92.86%, 86.72-100% and 0.67-1.35; 70.9091.03%, 77.86-97.89%, 93.17-100% and 0.59-1.52, respectively. The salivary gland polytene chromosomes of the 4th stage larvae from all crosses showed complete synapsis along the whole length of all autosomes and the Xchromosome (Figure 4). Sequences analysis of ITS2, COI, and COII regions The rDNA ITS2, and mitochondrial COI and COII

Vol. 30, no. 2

Journal of Vector Ecology

219

Table 3. Allelic frequencies (allele numbers) observed at 11 isoenzymes 13 loci in adult females of An. aconitus Form B (Chiang Mai and Phet Buri strains) and C (Chiang Mai and Mae Hong Son strains).

An. aconitus Form No. 1 2 3 4 5 6 7 8 9 10 11 12 13 Loci Aldox-1 Aldox-2 Est-5 Fum Hk Ldh-2 Mdh-1 Mdh-2 Me Gpm Gpi Xdh Idh Allele 98 100 102 98 100 100 104 95 100 100 100 100 105 100 100 95 100 105 100 105 99 100 95 100 CM 0.07 (4) 0.78 (42) 0.15 (8) 0 (0) 1.00 (54) 0.81 (63) 0.19 (15) 0.60 (6) 0.40 (4) 1.00 (28) 1.00 (44) 1.00 (4) 0 (0) 1.00 (4) 1.00 (4) 0 (0) 0.85 (34) 0.15 (6) 0.80 (40) 0.20 (10) 0.10 (2) 0.90 (18) 0 (0) 1.00 (12) B PB 0 (0) 0.67 (4) 0.33 (2) 0.33 (2) 0.67 (4) 1.00 (6) 0 (0) 0 (0) 1.00 (6) 1.00 (6) 1.00 (8) 1.00 (4) 0 (0) 1.00 (4) 1.00 (4) 0 (0) 1.00 (6) 0 (0) 1.00 (6) 0 (0) 0 (0) 1.00 (6) 0.37 (3) 0.63 (5) CM 0.11 (6) 0.78 (42) 0.11 (6) 0.10 (4) 0.90 (40) 0.83 (43) 0.17 (9) 0.33 (6) 0.67 (12) 1.00 (12) 1.00 (42) 0.75 (3) 0.25 (1) 1.00 (4) 1.00 (4) 0.04 (1) 0.88 (23) 0.08 (2) 0.86 (48) 0.14 (8) 0.60 (12) 0.40 (8) 0 (0) 1.00 (10) C MS 0 (0) 1.00 (6) 0 (0) 0.67 (4) 0.33 (2) 0.33 (2) 0.67 (4) 0.33 (2) 0.67 (4) 1.00 (6) 1.00 (6) 1.00 (4) 0 (0) 1.00 (4) 1.00 (4) 0 (0) 1.00 (6) 0 (0) 1.00 (6) 0 (0) 0 (0) 1.00 (6) 0 (0) 1.00 (6)

CM: Chiang Mai; MS: Mae Hong Son; PB: Phet Buri.

regions were amplified by PCR from genomic DNA of individual mosquitoes of cytologically identified 12 isolines of An. aconitus: half (six isolines) were Form B, and half were Form C. Total sequence lengths (ITS2+COI+COII) of these newly identified An. aconitus isolines varied from 1550bp to 1556bp. Intraspecific distances between them ranged from 0.1% to 1.0%. Their geographical origins and Genbank accession numbers are given in Table 1. Five length polymorphisms were detected from the complete ITS2 sequences of An. aconitus, and they varied from 376bp to 381bp in length. The haplotype diversity (Hd) of ITS2 sequences was 1.000. Twenty-four polymorphic nucleotide sites (corresponding to 26 mutations) were identified (excluding all sites with alignment gaps), of which 22 were singletons (92.7%) and only two were parsimony informative sites (8.3%) (Figure 5). Intraspecific distances in the An. aconitus ranged from 0.0% to 2.9%, whereas the interspecific distances between two species, An. aconitus and An. minimus, were very high with values of 25.0% to 27.0%. We also compared partial mitochondrial COI and COII sequences, and no length variation was detected: COI and COII contained 506 and 671 nucleotides, respectively. Thirteen and four variable sites were identified from COI and COII sequences, respectively, but all nucleotide polymorphisms were silent. If the outgroup sequence of COI (AY423055; Garros et al. 2004) was not considered, only three variation sites remained in the COI alignment (Figure

5). To explore the conflict between data sets, pairwise incongruence was estimated using a partition homogeneity test (Farris et al. 1994), which indicated that our data sets from these three distinct gene regions were heterogeneous (P = 0.24). Therefore the data sets should not be combined for analysis. To measure the nucleotide divergences between isoline groups of An. aconitus Form B and Form C, we used only ITS2 sequences because most of informative variations were in ITS2 (24 out of 30 variable sites: 80%). Summary of polymorphism within- and between-isoline groups of Form B and Form C is given in Table 5. The average nucleotide diversity (p) among Form B isolines (0.01667) was slightly greater than among those of Form C (0.00993). The genetic differentiation between two groups was very low as measured by dxy (0.01322). In fact, there was no fixed variation between two groups, and most of variable sites were singletons. The lack of genetic differentiation between two groups was also supported by the result of the permutation test (Hudson et al. 1992) (P = 0.62 for K*s estimator). DISCUSSION In this study, only two karyotypic forms of An. aconitus were obtained, i.e., Form B from Phet Buri Province, southwest Thailand, and Chiang Mai Province, northern Thailand, and Form C from Mae Hong Son Province,

220

ITS2

1 3 2 * T G A A A G C A T C G G C A C . . . . . . . G . . . . . . . . . . . . . . . . G . . . . . . . . . . . . . . . . T . . A . . . . . . . . . . . . T . C . . . . . . . T . . G . . . . . . . . T . . . . . . . . . . . . . . . . . . . . T . . . . . . . . . . . . . . . . . . . . A . G . T . . G . T . . . . . . . . . . . A . . . T . . . T . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . T . . . . . . . . . . . . . . . . . . A . . . . . . . . . . G . . . . . . . . . . . . . . . . . . . . . . T G . . . . . . . . T . . . . G . T . . . . . . . . . . . . . . . . . . . . . . T . . . . G . . . . G . T . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . T . C . . . . . . . . G . T . . . . . . . . . . . . . . . . . . . . . T T A G C G A T T G T T T C T * * * * T C C C . . . C . . . . . T . . . . . . . . . . . . 3 2 3 1 0 9 3 5 9 4 5 0 5 7 0 4 5 9 1 3 8 6 7 9 3 2 4 2 5 8 6 6 3 5 8 9 0 2 3 3 3 7 0 2 4 4 5 6 7 3 5 9 3 8 2 2 5 6 7 9 9 9 5 7 5 4 * 5 2 * G C A T A T A T . . . . . . A T . . . . . . . . . . 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 3 1 1 3 4 4 4 4 4 4 4 1 2 3 5

COI

COII

1

3

4

4

6

9

0

1

2

4

3

0

5

Consensus

A . . . . . . . .

A G T

T

T

G T

1BCM

.

.

.

.

.

.

2BCM

.

.

.

C

.

.

3BCM

.

A C

.

.

.

1BPB

.

.

.

.

.

.

Form B

2BPB

.

.

.

.

.

.

3BPB

G .

.

.

C

.

Journal of Vector Ecology

1CCM

.

.

.

.

.

.

2CCM

.

.

.

.

.

.

3CCM

.

.

.

.

.

.

1CMS . .

.

.

.

.

.

A .

Form C

2CMS

.

.

.

.

.

.

3CMS

.

.

.

.

.

.

Aconi

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

A T A T .

C A C C A T A

C

C

A T

Out group

December 2005

Figure 5. Variable sites in the sequence alignment of the ITS2, COI and COII sequences. Numbering is relative to the alignment. Only those positions differing from the consensus are shown. Indel (insertion or deletion) sites are not shown. An asterisk shows those sites yielding parsimony informative sites. A dot indicates a base pair identical to that of the consensus sequence. Sequence names are defined in Table1.

Table 4. Cross-mating among isolines of An. aconitus Form B and C.

Vol. 30, no. 2

Cross Embryonation rate** 90 88 72 89 89 90 80 89 75 86 94 88 76 88 79 89 92 87 91 88 74 81 142 (91.03) 151 (81.18) 134 (70.90) 133 (82.10) 122 (71.76) 147 (81.22) 181 (86.19) 131 (81.88) 144 (87.27) 220 (85.94) 116 (71.17) 135 (79.88) 139 (97.89) 128 (84.77) 121 (90.30) 124 (93.23) 115 (94.26) 137 (93.20) 157 (86.74) 102 (77.86) 128 (88.89) 205 (93.18) 91 (78.45) 127 (94.07) 136 (97.84) 126 (98.44) 121 (100) 119 (95.97) 113 (98.26) 137 (100) 157 (100) 102 (100) 128 (100) 191 (93.17) 89 (97.80) 122 (96.06) 139 (82.25) 189 (87.10) 121 (76.10) 135 (83.33) 112 (70.89) 136 (82.93) 128 (92.09) 166 (87.83) 110 (90.91) 112 (82.96) 104 (92.86) 125 (91.91) 111 (86.72) 151 (90.96) 107 (97.27) 112 (100) 104 (100) 120 (96.00) 46 (41.44) 77 (50.99) 51 (47.66) 45 (40.18) 43 (41.35) 69 (57.50) 81 (59.56) 76 (60.32) 52 (42.98) 44 (36.97) 68 (60.18) 71 (51.82) 88 (56.05) 53 (51.96) 62 (48.44) 84 (43.98) 40 (44.94) 59 (48.36) 131 (86.75) 166 (84.26) 123 (66.85) 138 (81.18) 107 (81.68) 148 (89.16) 113 (91.87) 112 (81.16) 101 (94.39) 144 (97.30) 109 (96.46) 108 (96.43) 59 (58.42) 66 (45.83) 59 (54.13) 47 (43.52) 42 (41.58) 78 (54.17) 50 (45.87) 61 (56.48) 65 (58.56) 74 (49.01) 56 (52.34) 67 (59.82) 61 (58.65) 51 (42.50) 55 (40.44) 50 (39.68) 69 (57.02) 75 (63.03) 45 (39.82) 66 (48.18) 69 (43.95) 49 (48.04) 66 (51.56) 107 (56.02) 49 (55.06) 63 (51.64) No. hatched (%) No. pupation (%) No. emergence (%)

Total egg (No.)*

No. females and males from total emergence (%) Female Male

Sex ratio Female/male 1.40 0.85 1.18 0.77 0.71 1.04 0.91 0.67 0.70 1.35 1.47 1.52 0.75 0.59 1.51 1.08 1.28 1.08 0.94 0.79 0.82 0.94

151 (62, 89) 197 (92, 105) 184 (86, 98) 170 (79, 91)

169 (55, 114) 217 (65, 152) 159 (77, 82) 162 (67, 95) 158 (61, 97) 164 (81, 83)

Journal of Vector Ecology

Female x Male Parental crosses BC x BC BP x BP CC x CC CM x CM Reciprocal crosses BC x BP BP x BC BC x CC CC x BC BC x CM CM x BC Back crosses (BC x BP)F1 x BP BC x (BC x BP)F1 (BP x BC)F1 x BC BP x (BP x BC)F1 (BC x CC)F1 x CC BC x (BC x CC)F1 (CC x BC)F1 x BC CC x (CC x BC)F1 (BC x CM)F1 x CM BC x (BC x CM)F1 (CM x BC)F1 x BC CM x (CM x C)F1

156 (57, 99) 186 (88, 98) 189 (94, 95) 162 (67, 95) 170 (79, 91) 181 (83, 98) 210 (65, 145) 160 (58, 102) 165 (79, 86) 256 (84, 172) 163 (67, 96) 169 (76, 93)

BC: An. aconitus Form B (Chiang Mai strain); BP: An. aconitus Form B (Phet Buri strain); CC: An. aconitus Form C (Chiang Mai strain); CM: An. aconitus Form C (Mae Hong Son strain). *Two selective egg-batches of inseminated females from each cross. **Dissection from 100 eggs.

221

222

Journal of Vector Ecology

December 2005

Table 5. Nucleotide polymorphism within and between isoline sequences of Anopheles aconitus Form B and C.

Form B Form C Form B vs Form C 6 6 12 0 0 0 17 (15) 10 (8) 24 (22) 6.263 3.733 3.972 0.01667 0.00993 dxy 0.01322 S: number of segregating site; k: average number of pairwise nucleotide differences; : nucleotide diversity; dxy: average number of nucleotide substitutions per site between groups. Sample size Fixed variation S (singletons) k

northwest Thailand and Chiang Mai Province, northern Thailand. Several intra-taxa of the Asian anopheline species, which were primarily detected with morphological, biological, and cytological differences and/or variations, led to the doubtful status of sibling species and/or subspecies. Subsequently, they were confirmed by polytene chromosome and isoenzyme investigations, and hybridization experiments, e.g., An. culicifacies complex (Green and Miles 1980, Subbarao et al. 1988, Adak et al. 1994), An. dirus complex (Kanda et al. 1981, Baimai et al. 1987, Sawadipanich et al. 1990, Green et al. 1992a), An. maculatus complex (Takai et al. 1987, Chabpunnarat 1988, Green et al. 1992b), An. minimus complex (Komalamisara 1989, Green et al. 1990, Choochote et al. 2002b) and An. sundaicus complex (Sukowati and Baimai 1996, Sukowati et al. 1999). Thus, we performed ovarian nurse cell polytene chromosome and isoenzyme investigations and hybridization experiments to determine the degree of genetic proximity between An. aconitus Form B and C strains from Phet Buri, Mae Hong Son, and Chiang Mai Province. Additionally, their comparative DNA sequences of ITS2, COI and COII were included in this study. The examination of polytene chromosomes of wildcaught adult females and/or progenies of isolines provides unequivocal evidence for the existence of different specific mate recognition systems (SMRS) (Paterson 1980). The total absence or significantly deficient number of heterozygotes for an inversion in a population indicates entirely the presence of reproductive isolation within a taxon (Green et al. 1992b, Subbarao 1998). The results of no major chromosomal rearrangements from the comparative investigations of ovarian nurse cell polytene chromosomes of standard mapping, An. aconitus Form B (Chiang Mai strain), to sympatric An. aconitus Form C (Chiang Mai strain), and allopatric An. aconitus Form B (Phet Buri strain) and Form C (Mae Hong Son strain), indicate no pre-mating isolation between An. aconitus Form B and C. Enzyme electrophoresis is being used extensively in studies of species complexes. Electrophoretic variations at enzyme loci are not only useful for the identification of isomorphic species, but they can also be used for the correct identification of morphologically cryptic Anopheles species. Variations at a locus thus enable the detection of the reproductive isolation within populations, resulting from

positive assortative matings within a population (Green et al. 1990, Subbarao 1998). The similar allelic frequencies observed at ten isoenzymes 16 loci in 4th stage larvae, and 11 isoenzymes 13 loci in adult females of sympatric An. aconitus Form B and C (Chiang Mai strain), and allopatric An. aconitus Form B (Phet Buri strain) and C (Mae Hong Son strain), suggest negative assortative mating between An. aconitus Form B and C. Hybridization experiments and/or the testing of reproductive isolation at the post-mating barriers are still efficient and reliable diagnostic tools for the differentiation of intra-taxon of anopheline species to a sibling species. Hybrid inviability, sterility, or breakdown are the criteria for genetic incompatibility, including lack of insemination, embryonation, hatchability, larva survival, pupation, emergence, adult sex distortion, abnormal morphology, and reproductive system (Kanda et al. 1981). Nonetheless, a point to be remembered is that colonies established from speciesspecific diagnostic characteristics of progeny from isolines have to be used. A laboratory colony established from a mixed, natural population may be a mixture of two or three species (Subbarao 1998). The results of genetic compatibility, providing viable progenies and complete synaptic salivary gland polytene chromosomes from the crossing studies among four isolines of An. aconitus Form B (Chiang Mai and Phet Buri strains) and C (Chiang Mai and Mae Hong Son strains), revealed no post-mating barriers between An. aconitus Form B and C both in sympatric and allopatric populations. Molecular investigation of some specific genomic markers, e.g., ribosomal DNA (ITS1, ITS2, D3) and mitochondrial DNA (COI, COII), has been used extensively as a tool to characterize and/or diagnose the sibling species and/or cryptic members in the intra-taxa of anopheline mosquitoes (Mitchell et al. 1992, Sharpe et al. 2000, Min et al. 2002, Park et al. 2003). From the molecular evidences of genetic distances and nucleotide divergences between isoline sequences of An. aconitus Form B and C, a conspecific relationship of these two An. aconitus forms was wellsupported. Based on the above evidence, we confidently conclude that An. aconitus Form B and C are conspecific cytological races in the Thai population. Similar results have been found previously in An. maculatus Form B and E (Chabpunnarat 1988), An. vagus Form A and B (Choochote et al. 2002a),

Vol. 30, no. 2

Journal of Vector Ecology

223

An. sinensis Form A and B (Choochote et al. 1998, Min et al. 2002), and An. pullus Form A and B (Park et al. 2003). Acknowledgments The authors sincerely thank the Thailand Research Fund (TRF: BRG/14/2545) and the Thailand Research Fund through the Royal Golden Jubilee Ph.D Program (Grant No. PHD/0044/2546) for financially supporting this research project, Supot Wudhikarn, Dean of the Faculty of Medicine, Chiang Mai University, for his interest in this research, and the Faculty of Medicine Endowment Fund for Research Publication for its financial support in defraying publication costs. REFERENCES CITED Adak, T., S.K. Subbarao, V.P. Sharma, and S.R. Rao. 1994. Lactate dehydrogenase allozyme differentiation of species in the Anopheles culicifacies complex. Med. Vet. Entomol. 8: 137-140. Baimai, V., R.G. Andre, B.A. Harrison, U. Kijchalao, and L. Panthusiri. 1987. Crossing and chromosomal evidence for two additional sibling species within the taxon Anopheles dirus Peyton and Harrison (Diptera: Culicidae) in Thailand. Proc. Entomol. Soc. Wash. 89: 157-166. Baimai, V., U. Kijchalao, R. Rattanarithikul, and C.A. Green. 1993. Metaphase karyotypes of Anopheles of Thailand and Southeast Asia: II. Maculatus group, Neocellia series, Subgenus Cellia. Mosq. Syst. 25: 116-123. Baimai, V., U. Kijchalao, and R. Rattanarithikul. 1996a. Metaphase karyotypes of Anopheles of Thailand and Southeast Asia: V. Myzomyia series, Subgenus Cellia (Diptera: Culicidae). J. Am. Mosq. Contr. Assoc. 12: 97105. Baimai, V., U. Kijchalao, and R. Rattanarithikul. 1996b. Metaphase karyotypes of Anopheles of Thailand and Southeast Asia. VI. The Pyretophorus and the Neomyzomyia series, subgenus Cellia (Diptera: Culicidae). J. Am. Mosq. Contr. Assoc. 12: 669-675. Chabpunnarat, S. 1988. Cytogenetic study of the Anopheles maculatus complex. M.Sc. Thesis, Mahidol University, Bangkok. Choochote, W., A. Jitpakdi, Y. Rongsriyam, N. Komalamisara, B. Pitasawat, and K. Palakul. 1998. Isoenzyme study and hybridization of two forms of Anopheles sinensis (Diptera: Culicidae) in Northern Thailand. Southeast Asian J. Trop. Med. Publ. Hlth. 29: 841-847. Choochote, W., A. Jitpakdi, K. Sukontason, U. Chaithong, S. Wongkamchai, B. Pitasawat, N. Jariyapan, T. Suntaravitun, E. Rattanachanpichai, K. Sukontason, S. Leemingsawat, and Y. Rongsriyam. 2002a. Intraspecific hybridization of two karyotypic forms of Anopheles vagus (Diptera: Culicidae) and the related egg surface topography. Southeast Asian J. Trop. Med. Publ. Hlth. 33 (Suppl 3): 29-35. Choochote, W., B. Pitasawat, A. Jitpakdi, E.

Rattanachanpichai, D. Riyong, S. Leemingsawat, and S. Wongkamchai. 2001. The application of ethanolextracted Gloriosa superba for metaphase chromosome preparation in mosquitoes. Southeast Asian J. Trop. Med. Publ. Hlth. 32: 76-82. Choochote, W., S. Sucharit, and W. Abeywickreme. 1983. A note on adaptation of Anopheles annularis Van Der Wulp, Kanchanaburi, Thailand, to free mating in a 30x30x30 cm cage. Southeast Asian J. Trop. Med. Publ. Hlth. 14: 559-560. Choochote, W., Y. Rongsriyam, S. Leemingsawat, A. Jitpakdi, N. Komalamisara, K. Surathin, P. Somboon, B. Chen, S. Wongkamchai, N. Jariyapan, P. Tippawangkosol, B. Pitasawat, and D. Riyong. 2002b. Intraspecific hybridization of Anopheles minimus (Diptera: Culicidae) species A and C in Thailand. Southeast Asian J. Trop. Med. Publ. Hlth. 33 (Suppl 3): 23-28. Farris, J.S., M. Kallersjo, S.G. Kluge, and C. Bult. 1994. Testing significance of incongruence. Cladistics. 10: 315320. Folmer, O., M. Black, W. Hoeh, R. Lutz, and R. Vrijenhoek. 1994. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Molec. Mar. Biol. Biotechnol. 3: 294­299. Garros, C., L.L. Koekemoer, L. Kamau, T.S. Awolola, W. Van Bortel, M. Coetzee, M. Coosemans, and S. Manguin. 2004. Restriction fragment length polymorphism method for the identification of major African and Asian malaria vectors within the Anopheles funestus and An. minimus groups. Am. J. Trop. Med. Hyg. 70: 260-265. Green, C.A. 1972. Cytological maps for the practical identification of females of the three freshwater species of the Anopheles gambiae complex. Ann. Trop. Med. Parasitol. 66: 143-147. Green, C.A., R.F. Gass, L.E. Munstermann, and V. Baimai. 1990. Population-genetic evidence for two species in Anopheles minimus in Thailand. Med. Vet. Entomol. 4: 25-34. Green, C.A., L.E. Munstermann, S.G. Tan, S. Panyim, and V. Baimai. 1992a. Population genetic evidence for species A, B, C and D of the Anopheles dirus complex in Thailand and enzyme electromorphs for their identification. Med. Vet. Entomol. 6: 29-36. Green, C.A., R. Rattanarithikul, and A. Charoensub. 1992b. Population genetic confirmation of species status of the malaria vectors Anopheles willmori and An. pseudowillmori in Thailand and chromosome phylogeny of the Maculatus group of mosquitoes. Med. Vet. Entomol. 6: 335-341. Green, C.A. and S.J. Miles. 1980. Chromosomal evidence for sibling species of the malaria vector Anopheles (Cellia) culicifacies Giles. J. Trop. Med. Hyg. 83: 7578. Green, C.A., V. Baimai, B.A. Harrison, and R.G. Andre. 1985. Cytogenetic evidence for a complex of species within the taxon Anopheles maculatus (Diptera: Culicidae). Biol. J. Linn. Soc. 4: 321-328. Harrison, B.A. 1980. Medical entomology studies: XIII. The

224

Journal of Vector Ecology

December 2005

Myzomyia series of Anopheles (Cellia) in Thailand, with emphasis on intra-interspecific variations (Diptera: Culicidae). Contrib. Am. Entomol. Inst. 17: 1-195. Hudson, R.R., D.D. Boos, and N.L. Kaplan. 1992. A statistical test for detecting geographic subdivision. Molec. Biol. Evol. 9: 138-151. Kanda, T. 1979. Improved techniques for the preparation of polytene chromosome for some anopheline mosquitoes. Mosq. News. 39: 568-574. Kanda, T., K. Takai, C.L. Chiang, W.H. Cheong, and S. Sucharit. 1981. Hybridization and some biological facts of seven strains of the Anopheles leucophyrus group (Reid, 1968). Jpn. J. Sanit. Zool. 32: 321-329. Kimura, M. 1980. A simple method for estimating evolutionary rates of base substitution through comparative studies of nucleotide sequences. J. Molec. Evol. 16: 111-120. Komalamisara, N. 1989. Genetic variability in isoenzymes of Anopheles minimus group from various localities in Thailand. Jpn. J. Sanit. Zool. 41: 69-80. Min, G.S., W. Choochote, A. Jitpakdi, S.J. Kim, W. Kim, J. Jung, and A. Junkum. 2002. Intraspecific hybridization of Anopheles sinensis (Diptera: Culicidae) strains from Thailand and Korea. Molec. Cells 14: 198-204. Mitchell, S.E., S.K. Narang, A.F. Cockburn, J.A. Seawright, and M. Goldenthal. 1992. Mitochondrial and ribosomal DNA variation among members of the Anopheles quadrimaculatus (Diptera: Culicidae) species complex. Genome 35: 939-950. Nei, M. 1987. Molecular evolutionary genetics. Columbia University Press, New York. Park, S.J., W. Choochote, A. Jitpakdi, A. Junkum, S.J. Kim, N. Jariyapan, J.W. Park, and G.S. Min. 2003. Evidence for a conspecific relationship between two morphologically and cytologically different Forms of Korean Anopheles pullus mosquito. Molecul. Cells. 16: 354-360. Paterson, H.E. 1980. A comment on "Mate Recognition Systems" Evolution. 34: 330-331. Rozas, J., J.C. Sanchez-DelBarrio, X. Messeguer, and R. Rozas. 2003. DnaSP, DNA polymorphism analyses by the coalescent and other methods: an integrated program for molecular population genetics and molecular evolution analysis. Bioinformatics 19: 2496-2497. Rongnoparut, P., N. Sirichotpakorn, R. Rattanarithikul, S. Yaicharoen, and K.J. Linthicum. 1999. Estimates of gene flow among Anopheles maculatus populations in Thailand using microsatellite analysis. Am. J. Trop. Med. Hyg. 60: 508-515. Sawadipanich, Y., V. Baimai, and B.A. Harrison. 1990. Anopheles dirus species E: chromosomal and crossing

evidence for another member of the dirus complex. J. Am. Mosq. Contr. Assoc. 6: 477-481. Scanlon, J.E., E.L. Peyton, and D.J. Gould. 1968. An annotated checklist of the Anopheles of Thailand. Thai. Natl. Sci. Pap. Fauna. Ser. 2: 1-35. Sharma, G.P., K.G. Soni, and S. Chaudhry. 1980. Salivary gland chromosome map of Anopheles aconitus Doenitz (Diptera: Culicidae). Cytobios. 27: 35-42. Sharpe, R.G., R.E. Harbach, and R.K. Butlin. 2000. Molecular variation and phylogeny of members of the Minimus group of Anopheles subgenus Cellia (Diptera: Culicidae). Syst. Entomol. 25: 263-272. Subbarao, S.K. 1998. Anopheline species complex in Southeast Asia. WHO Tech. Pub. Ser. 18: 1-82. Subbarao, S.K., K. Vasantha, and V.P. Sharma. 1988. Studies on the crosses between the sibling species of the Anopheles culicifacies complex. J. Hered. 79: 300-303. Sucharit, S., N. Komalamisara, S. Leemingsawat, C. Apiwathanasorn, and S. Thongrungkiat. 1988. Population genetic studies on the Anopheles minimus species complex in Thailand. Southeast Asian J. Trop. Med. Publ. Hlth. 19: 717-723. Sukowati, S. and V. Baimai. 1996. A standard cytogenetic map for Anopheles sundaicus (Diptera: Culicidae) and evidence for chromosomal differentiation in populations from Thailand and Indonesia. Genome. 39: 165-173. Sukowati, S., V. Baimai, S. Harun, Y. Dasuki, H. Andris, and M. Efriwati. 1999. Isozyme evidence for three sibling species in the Anopheles sundaicus complex from Indonesia. Met. Vet. Entomol. 13: 408-414. Swofford, D.L. 1999. PAUP: Phylogenetic analysis using parsimony (and other methods), ver. 4.0 b10. Sinauer Associates, Sunderland, MA. Takai, K., T. Kanda, K.I. Ogawa, and S. Sucharit. 1987. Morphological differentiation in Anopheles maculatus of Thailand accompanied with genetical divergence assessed by hybridization. J. Am. Mosq. Contr. Assoc. 3: 148-153. Thompson, J.D., T.J. Gibson, D.E. Plewniak, F. Jeanmougin, and D.G. Higgins. 1997. The CLUSTAL X-windows interface: Flexible strategies for multiple sequence alignment aided by quality analysis. Nucl. Acids Res. 25: 4876-4882. Walton, C., J.M. Handley, C. Kuvangkadilok, F.H. Collins, R.E. Harbach, V. Baimai, and R.K. Butlin. 1999. Identification of five species of the Anopheles dirus complex from Thailand, using allele-specific polymerase chain reaction. Med. Vet. Entomol. 13: 24-32. White, G.B., M. Coluzzi, and A.R. Zahar. 1975. Review of cytogenetic studies on anopheline vectors of malaria. WHO/VBC/75.538. 35 pp.

Vol. 30, no. 2

Journal of Vector Ecology

225

Geographical distribution of Anopheles minimus species A and C in western Thailand

Ampornpan Kengluecha2, Pornpimol Rongnoparut1,4, Soamrutai Boonsuepsakul1, Ratana Sithiprasasna2, Prinyada Rodpradit1, and Visut Baimai3,4

Department of Biochemistry, Faculty of Science, Mahidol University, Bangkok, Thailand Department of Entomology, Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand 3 Department of Biology, Faculty of Science, Mahidol University, Bangkok, Thailand 4 Center for Vectors and Vector-Borne Diseases, Faculty of Science, Mahidol University, Bangkok, Thailand

2 1

Received 4 January 2005; Accepted 3 April 2005 ABSTRACT: Elucidating vector distribution based on an accurate species identification is important to understanding the nature of the species complex in order to achieve vector control. Morphologically, An. minimus s.l. is difficult to distinguish from both its species complex and its closely related species. A polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) technique and a single multiplex-allele specific PCR developed for species identification were applied in this study in comparison with morphological identification. Both methods were used, combining with geographical information systems to determine the distribution of An. minimus species A and C. The investigation on the breeding habitats was performed in the malarious area of western Thailand. Anopheles larvae were collected from 36 bodies of water among five districts (Sangkhaburi, Thong Pha Phum, Si Sawat, Muang, and Sai Yok) of Kanchanaburi Province, Thailand. In this study, An. minimus A larvae were present in all study districts but the association differed when focusing on study sites within each district. Although there were many reports of An. minimus A in Ban Phu Rat and Ban Phu Toei villages in Sai Yok District, we did not find the breeding sites of species A in those two areas. An. minimus A and C were found in Ban Phu Ong Ka village in Sai Yok District. The breeding habitats of An. minimus C were present covering 30-40 km of distance in northern part of Sai Yok and this species was also found in the central and southern parts of Si Sawat District. Journal of Vector Ecology 30 (2): 225-230. 2005. Keyword Index: Anopheles minimus, species distribution, molecular identification, malaria, Thailand.

INTRODUCTION Thailand, a developing country in Southeast Asia confronting endemic malaria, needs reliable surveillance programs to understand and manage its malaria problems. Understanding the spatial and temporal changes in anopheline mosquito abundance, quantification of transmission potential of vector populations, and description of distributions of host (human) populations are necessary prerequisites for predicting high-risk malaria areas. Anopheles minimus sensu lato (subgenus Cellia, Myzomyia Series and Funestus Group, Minimus subgroup) (Harbach 2004) is a major malaria vector throughout its distribution, extending from Uttar Pradesh in India to Southeast Asia, China, Taiwan, and Japan. In Thailand, its distribution is reduced in the peninsula and was not observed in central plains but remains abundant in forested hilly areas (Harrison 1980). This species plays an important role in malaria transmission during the dry season and during much of rainy season when populations of Anopheles dirus are low at the Thai-Myanmar border (Sithiprasasna et al. 2003). It is known as a species complex comprising two species on mainland Southeast Asia, informally designated An. minimus A and C (Harbach 2004). Species A has been found

predominantly in Thailand and has been recorded in Vietnam, Laos, Cambodia, and China (Chen et al. 2002, Green et al. 1990, Kengne et al. 2001, Van Bortel et al. 1999). Species C was only recorded in three provinces in western and northern Thailand (Kanchanaburi, Tak, and Chiang Mai) (Green et al. 1990, Sharpe et al. 1999), northern Vietnam (Kengne et al. 2001, Van Bortel et al. 1999) and south-central provinces of China (Chen et al. 2002). The differences between the two species in their biting and resting behavior have been studied. Species A is more anthropophilic and endophilic than species C in northern Vietnam (Van Bortel et al. 1999), while in Thailand both species tend to feed from cows rather than humans (Rwegoshora et al. 2002). Misidentification using morphological characters to separate An. minimus s.l. from closely related species could occur (Harrison 1980). Moreover, An. minimus A and C are isomorphic and cannot be distinguished by their morphology (Green et al. 1990). Green et al. (1990) found that a humeral pale (HP) wing spot was present more often in An. minimus C than in An. minimus A in western Thailand. However, the presence of HP spots cannot be used with any degree of confidence to differentiate species C from species A (Chen et al. 2002). Allozyme electrophoresis was used to identify species of the An. minimus complex and related species in

226

Journal of Vector Ecology

December 2005

An. minimus group, including An. aconitus (Green et al. 1990, Van Bortel et al. 1999). Recently, PCR-based methods have been developed to identify this species complex. These include random amplified polymorphic DNA-PCR, allele-specific amplification (ASA), single-strand conformation polymorphism (SSCP), PCR-restriction fragment length polymorphism (RFLP), and multiplex PCR of RAPD and internal transcribed spacer 2 (ITS2) (Garros et al. 2004, Kengne et al. 2001, Sharpe et al. 1999, Sucharit and Komalamisra 1997, Van Bortel et al. 2000). The aim of this study was to identify species of An. minimus complex by using morphological characters and molecular techniques. MATERIALS AND METHODS Mosquito collection Anopheles larvae were collected from different sites in five districts including: Muang, Sangkhaburi, Si Sawat, Sai Yok, and Thong Pha Phum of Kanchanaburi Province (Figure 1) from February to June 2004. The study area is in a malaria endemic area of western Thailand near the Myanmar border. Larvae were collected from breeding habitats and reared to adults for species identification by morphological and molecular techniques. Morphological identification Adult mosquitoes were identified using the morphological key (Harrison 1980). Specimens identified as Anopheles minimus s.l. were scored for the presence or absence of the humeral pale (HP) spot on the wings. An. minimus specimens with ambiguous morphology were identified as An. minimus s.l. All specimens were stored at -20 ºC for until verification by molecular techniques. Molecular identification Genomic DNA of individual mosquito was extracted using Ready AmpTM Genomic DNA Purification System kit (Promega Corporation, Madison, WI, U.S.A.), following the manufacturer's instructions. Isolated genomic ssDNA was subjected to PCR amplification using both multiplex allelespecific PCR assay (Garros et al. 2004) and PCR-RFLP (Van Bortel et al. 2000) to confirm species. The Polymerase Chain Reaction-Restriction Fragment Length Polymorphism (PCRRFLP) assay based on the ITS2 marker and developed by Van Bortel et al. (2000) was used for the identification of An. minimus A, An. minimus C, and closely related species (i.e., An. aconitus). PCR primers, PCR conditions, and restriction digestion with MspI and Sau96I (New England Biolabs, Beverly, MA) were carried out following the previously described method (Van Bortel et al. 2000). Multiplex allelespecific PCR assay (Garros et al. 2004) was performed to confirm the identification of the members of the An. minimus species complex. PCR primers, ITS2A, MIA, and MIC were used to identify An. minimus A and C. The primer sequences were according to those previously described (Garros et al. 2004). PCR amplification using AmpliTaq Gold DNA polymerase (Applied Biosystems, Branchburg, NJ, U.S.A.)

was performed as described (Garros et al. 2004). RESULTS Approximately 350 of some 600 Anopheles specimens collected between February and June 2004 from five districts of Kanchanaburi were identified as being from the Myzomyia series of the subgenus Cellia. Of these, 238 specimens morphologically identified as An. minimus were confirmed by both PCR-RFLP and multiplex allele-specific PCR (Garros et al. 2004) to distinguish them from their closely related species group. As shown on Table 1, a total of 114 field specimens from Muang, Sangkhaburi, Si Sawat, Sai Yok and Thong Pha Phum were identified as An. minimus A, while 123 specimens from Sai Yok, and Si Sawat were An. minimus C. One specimen from Sangkhaburi, identified as An. aconitus, was morphologically misidentified as An. minimus A (Table 1). When using molecular identification, 84% of An. minimus A and 89% of An. minimus C of morphologically identified mosquitoes were in agreement for species identification. A portion (18% of total specimens) of ambiguously identified An. minimus s.l. by their morphology was identified by molecular technique (among these, ~70% were An. minimus A and ~30% were species C). We did not find any natural hybrids between An. minimus A and C in our study, although Van Bortel et al. (1999) first reported the hybrid (< 1% of samples) in the village of Khoi in northern Vietnam. A total of 36 larval breeding habitats categorized into five types among the five study districts (Table 1) were mapped with data on the district locations and are shown in Figure 1. An. minimus A was found in stream margins, stream pools, swamps, and ditches, while the breeding sites of An. minimus C were distributed in stream margins, stream pools and ground pools. All of the 78 An. minimus larvae collected from 16 water bodies in Sangkhaburi, Thong Pha Phum, and Muang were identified as An. minimus A. Although both 36 An. minimus A and 123 An. minimus C were found in Sai Yok and Si Sawat, they were found in sympatry only in the Ban Phu Ong Ka and Sai Yok districts. Three larval collection sites of An. minimus A and two larval collection sites of An. minimus C were found in Ban Phu Ong Ka (Table 1). In this study, An. minimus A was associated with all study districts. However, the association differed by focusing on study sites in the districts. Figure 1 shows localities of breeding habitats of An. minimus A, An. minimus C, and both A and C around the five districts, with flags representing the location of each village. No breeding sites of species A were found in Ban Phu Rat or Ban Phu Toei in Sai Yok. The An. minimus species complex population appeared to be separated in the area. The breeding habitats of An. minimus C were present in Ban Tha Thung Na and Ban Phu Rat, covering the distance of 30-40 km in the northern part of Sai Yok. It is clear that An. minimus C is predominant in the Sai Yok district. An. minimus C was also found in the central parts of Si Sawat District (Figure 1). Figure 2 depicts the number of malaria patients by subdistrict from October 2001 to May 2004. The bar graphs show infection with Plasmodium falciparum, P. vivax, mixed

Vol. 30, no. 2

Table 1. Localities, habitat types, and numbers of An. minimus A, An. minimus C, and An. aconitus identified by molecular techniques (multiplex PCR and PCR-RFLP). Larval habitat types (ditch, stream margin, stream pool, and swamp) are represented by Dit, Stm, Stp, and Sw, respectively.

An.minimus A 10 6 4 19 4 6 4 6 6 24 8

An.minimus C

An. aconitus

Journal of Vector Ecology

9 5 12 6 4 7 2 6 9 1 7 2 4 10 20 6 1 2 1 4 3 1 3

District Muang Muang Muang Muang Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sai Yok Sangkhaburi Sangkhaburi Sangkhaburi Sangkhaburi Sangkhaburi Si Sawat Si Sawat Thong Pha Phum Thong Pha Phum Thong Pha Phum Thong Pha Phum Thong Pha Phum Thong Pha Phum Thong Pha Phum 1

Village Name Ban Chan Ui Ban Thap Si La Ban Thap Si La Ban Tha Wi Ban Tha Sao Ban Phu Rat Ban Phu Rat Ban Phu Rat Ban Phu Rat Ban Phu Rat Ban Phu Ong Ka Ban Phu Ong Ka Ban Phu Ong Ka Ban Phu Ong Ka Ban Phu Toei Ban Phu Toei Ban Phu Toei Ban Phu Toei Ban Tha Thung Na Ban Dong Sak Ban Chong Ua Ban Chong Ua Ban Chong Ua Ban Chong Ua Ban Chong Ua Ban Pong Wai Ban Dan Mae Chalaep Ban Huai Khayeng Ban U Long Ban Tha Khanun Ban Nong Bang Ban Nong Bang Ban Nong Bang Ban Nong Bang

Collection No. K-62 K-37 K-57 K-65 K-41 K-09 K-10 K-08 K-05 K-31 K-36 K-34 K-55 K-56 K-02 K-04 K-03 K-01 K-27 K-26 K-22 K-24 K-23 K-20 K-21 K-44 K-46 K-18 K-19 K-16 K-12 K-11 K-13 K-14

Habitat type Stm Sw Stm Stp Sw Stm Stm Stm Stm Stm Stm Sw Stm Stm Stm Stm Stm Stm Stp Stm Stm Stm Sw Stm Stm Stm Stm Stm Stm Stm Dit Stm Stm Stm

227

228

Journal of Vector Ecology

December 2005

Kanchanaburi

Figure 1. Thematic map showing localities of breeding habitats of An. minimus A (black dots), An. minimus C (white dots), both A and C (asterisks) around the five districts. A flag represents the location of each village. The insert map shows the map of Thailand indicating Kanchanaburi Province, the area covered by the present research.

infection of both falciparum and vivax, and gametocyte. The number of malaria infections showed no correlation with the distribution of An. minimus A and C. DISCUSSION An. minimus species A and species C were correctly identified morphologically at 84% and 89% and misidentified or ambiguously identified at 16% and 11%, respectively. The percentage of morphological identification error when compared with the PCR-based methods in this study agrees with those reported in previous studies that An. minimus A and C cannot be distinguished accurately by morphology due to character variations (Chen et al. 2002, Green et al. 1990). Green et al. (1990) investigated population genetics of An. minimus A and C in Thailand and suggested that the

identification using morphological characters can lead to 37% error. Chen et al. (2002) found that the morphological identification for both species A and C females from southern China could be unreliable. Morphological identification of closely related species in the Minimus group, including An. aconitus, An. varuna, and An. minimus can be easily mixed up due to overlapping morphological characters (Harrison 1980). However, in the present study only one An. aconitus was morphologically misidentified as An. minimus A (Table 1). Our previous studies also revealed that 1% (2 of 200) of the specimens collected from Mae Sot, Tak Province and identified morphologically as An. minimus was proven to be An. varuna by the PCR-RFLP method (unpublished data). This is in contrast with a previous study by Van Bortel et al. (2001) that found as much as 96% (72 of 75) of morphologically identified An. minimus A were misidentified.

Vol. 30, no. 2

Journal of Vector Ecology

229

Figure 2. Thematic map showing the number of malaria patients from October 2001- May 2004 for all species of malaria parasite by subdistrict. In the pie chart, infections with Plasmodium falciparum, Plasmodium vivax, mixed infection of both P. falciparum and P. vivax, and gametocytes are shown. They were identified by the PCR-RFLP method as An. varuna, An. aconitus, and An. pampanai. An. minimus s.l. has been reported to oviposit primarily in clear, running streams in forested areas (Harrison 1980). In this study, stream margin was the main breeding habitat for both An. minimus A and C. We were unable to identify which key environmental factors were associated with species A or C since there were a variety of habitat types. Alternatively, the results imply that these species could be associated with locations of habitat rather than types of habitat. An. minimus A is known to occur throughout Thailand (Green et al. 1990). This is consistent with our finding of An. minimus A in the five districts of Kanchanaburi Province. The occurrence of An. minimus C is restricted to the Sai Yok district. Charlwood and Edoh (1996) explained the distribution of An. gambiae Giles complex in that An. gambiae Giles and An. arabiensis Patton differ in location of breeding site rather than host availability. The association of An. minimus C distribution with more humid areas in Ban Phu Toei has been reported by Sucharit et al. (1988). Beside Ban Phu Toei and Ban Phu Rat, we also found An. minimus C in Ban Tha Thung Na, Ban Dong Sak, and Ban Phu Ong Ka of Sai Yok district (Table 1; Figure 1). It is possible that Ban Phu Ong Ka is the area that separates breeding sites of An. minimus C in the northern part from An. minimus A in the southern part of Sai Yok (Figure 1). An. minimus C extends its distribution eastward into Si Sawat and this is the first report of species C found in Si Sawat district (Figure 1). Further investigation in the factors that may influence the spatial distribution pattern of An. minimus A and C, including satellite imagery of landuse/land-cover classification and proximity analysis of breeding habitats of An. minimus A and C, is underway. Acknowledgments This work was supported by the TRF/BIOTEC Special Program for Biodiversity Research and Training and CNRS (France) grant BRT R-245006.

230 REFERENCES CITED

Journal of Vector Ecology

December 2005

Charlwood, J.D. and D. Edoh. 1996. Polymerase Chain Reaction used to describe larval habitat use by Anopheles gambiae complex (Diptera: Culicidae) in the environs of Ifakara, Tanzania. J. Med. Entomol. 33: 202-204. Chen, B., R.E. Harbach, and R.K. Butlin. 2002. Molecular and morphological studies on the Anopheles minimus group of mosquitoes in Southern China: taxonomic review, distribution and malaria vector status. Med. Vet. Entomol. 16: 253-265. Garros, C., L.L. Koekemoer, M. Coetzee, M. Coosemans, and S. Manguin. 2004. A single multiplex assay to identify major malaria vectors within the African Anopheles funestus and the oriental An. minimus groups. Am. J. Trop. Med. Hyg. 70: 583-590. Green, C.A., R.F. Gass, L.E. Munstermann, and V. Baimai. 1990. Population genetic evidence for two species in Anopheles minimus in Thailand. Med. Vet. Entomol. 4: 25-34. Harbach, R.E. 2004. The classification of genus Anopheles (Diptera: Culicidae): a working hypothesis of phylogenetic relationships. Bull. Entomol. Res. 94: 537553. Harrison, B.A. 1980. The Myzomyia Series of Anopheles (Cellia) in Thailand, with emphasis on intra-interspecific variations (Diptera : Culicidae). Medical entomology studies ­ XIII. Contrib. Am. Entomol. Inst. 17: 1-195. Kengne, P., H.D. Trung, V. Baimai, M. Coosemans, and S. Manguin. 2001. A multiplex PCR-based method derived from random amplified polymorphic DNA (RAPD) markers for the identification of species of the Anopheles minimus group in Southeast Asia. Insect Mol. Biol. 10: 427-435. Rwegoshora, R.T., R.G. Sharpe, K.J. Baisley, and P. Kittayapong. 2002. Biting behavior and seasonal variation in the abundance of Anopheles minimus species

A and C in Thailand. SE. Asian J. Trop. Med. Publ. Hlth. 33: 694-701. Sharpe, R.G., M.M. Hims, R.E. Harbach, and R.K. Butlin. 1999. PCR-based methods for identification of species of the Anopheles minimus group: allele-specific amplification and single-strand conformation polymorphism. Med. Vet. Entomol. 13: 265-273. Sithiprasasna, R., K.J. Linthicum, G.J. Liu, J.W. Jones, and P. Singhasivanon. 2003. Some entomological observations on temporal and spatial distribution on malaria vectors in three villages in northwestern Thailand using a Geographic Information System (GIS). Southeast Asian J. Trop. Med. Publ. Hlth. 34: 505-516. Sucharit, S., N. Komalamisra, S. Leemingsawat, C. Apiwathnasorn, and S. Thongrungkiat. 1988. Population genetic studies on the Anopheles minimus complex in Thailand. Southeast Asian J. Trop. Med. Publ. Hlth. 19: 717-723. Sucharit, S. and N. Komalamisra. 1997. Differentiation of Anopheles minimus species complex by RAPD-PCR technique. J. Med. Assoc. Thail. 80: 598-602. Van Bortel, W., H.D. Trung, N.D. Manh, P. Roelants, P. Verle, and M. Coosemans. 1999. Identification of two species within the Anopheles minimus complex in northern Vietnam and their behavioral divergences. Trop. Med. Int. Hlth. 4: 257-265. Van Bortel, W., H.D. Trung, P. Roelants, R.E. Harbach, T. Backeljau, and M. Coosemans. 2000. Molecular Identification of Anopheles minimus s.l. beyond distinguishing the members of the species complex. Insect Mol. Biol. 9: 335-340. Van Bortel, W., R.E. Harbach, H.D. Trung, P. Roelants, T. Backeljau, and M. Coosemans. 2001. Confirmation of Anopheles varuna in Vietnam, previously misidentified and mistargeted as the malaria vector Anopheles minimus. Am. J. Trop. Med. Hyg. 65: 729-732.

Vol. 30, no. 2

Journal of Vector Ecology

231

Repellent effect of plant essential oils against Aedes albopictus

Pin Yang and Yajun Ma

Department of Etiologic Biology, College of Basic Medical Sciences Second Military Medical University, Shanghai 200433, China Received 5 January 2005; Accepted 9 May 2005 ABSTRACT: Six essential oils: asteraceae oil, rutaceae oil, mentha piperta oil, carvacryl oil, citronella oil, and eucalyptus oil were tested for evaluation of their repellent effects against Aedes albopictus mosquitoes under laboratory conditions. Only citronella oil and eucalyptus oil were tested with human beings. There was considerable protection for mice. Carvacryl oil (7%) provided 100% protection for mice after 7 h. Eucalyptus oil (15%) gave protection to humans for least 3 h; the protection time was prolonged to 5 h after adding 5% vanillin. The mixture could be developed into a practical product after the field evaluation. Journal of Vector Ecology 30 (2): 231-234. 2005. Keyword Index: Repellents, essential oils, Aedes albopictus. INTRODUCTION Several mosquito species can spread pathogens of diseases such as malaria, filariasis, Japanese encephalitis, dengue fever, and yellow fever (Su and Ye 1996). Currently, major approaches to these mosquito-borne diseases are mosquito control and personal protection from mosquito bites. Repellents are a practical and economical means for preventing the transmission of these diseases to humans. Repellents, such as the "gold standard" N,N-diethyl-3methylbenzamide (DEET), has shown significant repellency against mosquitoes and other biting arthropods (Yap 1986, Coleman et al. 1993, Walker et al. 1996). However, toxicity reactions for humans after applications of DEET can be severe (Zadikoff 1979, Robbins and Cherniack 1986, Edwards and Johnson 1987, Qiu et al. 1998). To avoid these adverse effects, many laboratories have tried to replace DEET with repellents that are derived from plant extracts. Various plant extracts, such as neem (from Azadirachta indica A. Juss), basil oil (Ocimum basilicum L., O. basilicum L. fa. citrtum Bach, O. gratissimum L., O. americanum L., O. tenuiflorum L.), citronella grass (Cymbopogon nardus Rendle), galingale (Alpinia galanga L.), clove (Syzygium aromaticum L.), and thyme (Thymus vulgaris L.), have been recorded as mosquito repellents (Sukumar et al. 1991, Sharma et al. 1993, Chokechaijaroenporn et al. 1994, Suwonkerd and Tantrarongroj 1994, Boonyabancha et al. 1997, Barnard 1999). Several natural repellents have demonstrated good efficacy against some mosquito species, but they were evaluated only with mice under laboratory conditions. It would be better to conduct the research on humans, although this is difficult. In this study the repellency of several essential oils against Ae. albopictus was tested on mice and humans, searching for potential essential oils with promising protective effects on humans. MATERIALS AND METHODS Essential oils Six essential oils were chosen for this study: asteraceae oil was extracted from Ajania tenuifolia collected from the Haibei Alpine Meadow ecosystem, Qinhai Province, China; Rutaceae oil was extracted from leaves of Citrus sinensis collected from Deqing County, Guangdong Province, China. These two were both extracted by steam distillation. Mentha piperta oil, carvacryl oil, citronella oil, eucalyptus oil and vanillin were the products of Shanghai Charoma Perfumery and Bio-Chem Co., Ltd. China, and the oils' source plants were Mentha piperita, Mentha spicata, Cymbopogon citrates, and Eucalyptus globulus, respectively. All testing solutions were dissolved in ethanol. Test mosquitoes The mosquitoes were a laboratory colony of Ae. albopictus provided by Jiangsu Institute of Parasitic Diseases Control, China. The mosquitoes were maintained at 26°C ± 1°C, 65% ± 5% (RH), and 12 h:12 h (light:dark) photoperiod. Nulliparous females of 4-5 days-old were used, and all testing was carried out in the rearing room. The same cage of mosquitoes was used to test the same oil solutions. The repeated tests were carried out with different mosquitoes. Toxicity on mosquitoes was not observed during the testing procedure. Repellency test on Kunming mice Kunming mice were fixed supinely on the board with their abdomens cleaned and depilated. An exposed hairless area of 2 × 2 cm was marked and the mouse was put into a mosquito cage (40×30×30 cm, containing 300 female mosquitoes) for 2 min. If more than 20 mosquitoes bit the mouse during a test, the mosquitoes and mouse were then used in the repellency tests. Five essential oils were tested at a concentration of 7%; two of them were also tested at 1% and 15%.

232

Journal of Vector Ecology

December 2005

The testing solution (5 l/cm2) was painted on the exposed part of the mouse abdomen. After 1 h, the mouse was put in the mosquito cage for 2 min, removed, and then put again into cages every hour for 7 h. The number of blood-fed mosquitoes was observed. The blank control consisted of 100% ethanol. Each treatment was repeated three times. The percentage of protection for the mice was calculated by the formula below (Frances et al. 2001): Percentage of protection (100%) = [(Control - Treated) / Control] ×100 Controls consisted of mosquitoes that fed on mice treated with a control solution, compared to the number bitten treated by the testing solution. The data analysis used was a repeated measures analysis of variance (ANOVA) and completed by Statistical Analysis Systems (Version 8.2). Repellency test on humans Two essential oils were tested at concentrations of 15% and 30%. The mixture of 5% vanillin and testing solution was also tested. An area of 4 × 4 cm on volunteers' hands was marked and painted with a testing solution (2 l/cm2), while the other hand was treated with the mixture. The hand was covered by plastic film except for the marked area. After 1 h, each volunteer put his or her hand into a mosquito cage (40×30×30 cm, containing 300 nulliparous females) for 2 min, then removed it and again placed it into the cage for another 2 min every hour for 7 h. The test continued until one mosquito bit a volunteer. The control was 5% vanillin and 4% DEET ("Wen Bu Ding" solution, containing 4% DEET). Proboscis amputated mosquitoes that cannot feed were used on human subjects (Shirai et al. 2000). If there was no single mosquito biting during the time after application of the test solution, it was recorded as protecting (+), otherwise, as not protecting (-) . RESULTS Protection on Kunming mice The protection results on Kunming mice treated by 7% essential oils against Ae. albopictus is shown in Table 1. There was no repellency with the control (100% ethanol). The protection time of carvacryl oil was the longest and the percentage of protection remained 100% after an exposure of 7 h. The protection percentage of five essential oils was more than 90% after 7 h treatments. There were significant differences among the five essential oils by repeated measures analysis of variance (F=9.78, 4 df, P = 0.0009), and there was also significant variance among the different time sections (F = 6.31, 7 df, P=0.0009). The repellency of the five essential oils against Ae. albopictus was high during the first 3 h, and there was no significant difference among those exposed for 4 h to 8 h. There was no significant correlation between the time section and essential oils (P = 0.1742). There was no significant difference in repellency between asteraceae oil solutions (F = 0.22, 2 df, P = 0.8122) and

rutaceae oil solutions (F=2.31, 2 df, P=0.1806). Their protection rate all exceeded 87%. A repeated measures analysis of variance revealed that there were significant differences among the three concentrations (1%, 7%, and 15%) of asteraceae oil (F=3.79, 7 df, P=0.0032) and rutaceae oil (F=11.71, 7 df, P<0.0001) with respect to time. The repellency of asteraceae oil was the highest (F=27.27, P=0.0020; F=2.31, P=0006) at an exposure of 3 h and wore off after 8 h (F=10.07, P=0.0193). Rutaceae oil was most potent (F=53.89, P=0.0003; F=25.53, P=0.0023; F=48.73, P=0.0004; F=23.90, P=0.0027) at an exposure of 4 h and wore off markedly after 7 h (F = 43.41, P = 0.0006). There was no correlation between concentration and exposure time (F=0.49, 14 df, P=0.9205) in asteraceae oil, but this correlation was significant (F=3.51, 14 df, P=0.003814) in rutaceae oil. Protection on humans The results of protection on human beings against Ae. albopictus by the essential oils is shown in Table 2. The range of protecting periods were 1 h, 3 h, and 1 h with 30% citronella oil, 15% and 30% eucalyptus oil solution, respectively. There was no repellency activity with 15% citronella oil solution. However, the repellency activity increased significantly when 5% vanillin was added to the testing solution. As a result, the range of protecting periods was extended to 3 h and 2 h with 15% and 30% citronella oil solution, respectively, and to 5 h and 4 h with 15% and 30% eucalyptus oil solution, respectively. The control solutions were 4% DEET and 5% vanillin solution. The 4% DEET solution provided protection for at least 6 h, and 5% vanillin solution for only 1 h. DISCUSSION Five oil solutions had considerable protection for mice in this study. The repellency was related to the type of essential oil and exposure time, but there was no significant correlation between them, which showed that the pattern of response in mosquitoes to repellent materials over time was not different among the essential oils. The results using the three concentrations of asteraceae oil and rutaceae oil showed that the repellency was not related to concentrations. It is suggested that carvacry oil would be suitable as a candidate for mosquito repellents. However, there was a difference between human and mouse skin after being treated with the same oil solution. There was not any protection to humans with the above oil solutions, except citronella oil, even if the concentration was above 30%. Citronella oil was chosen to test on volunteers after screening on mice. Eucalyptus oil was tested previously (Trigg 1996, Moore et al. 2002, Hadis et al. 2003, Schreck and Leonhardt 1991). Eucalyptus-based repellent containing 30% p-menthane-diol had a protection rate of 96.89% for 4 h against Anopheles darlingi, which was similar to those currently in application (Moore et al. 2002). In the present study, the protection rate of 15% eucalyptus oil with 5% vanillin was similar to that of 4% DEET, even though their effects were different in many respects. Citronella oil cannot reliably protect humans, as reported by Wasuwat et al. (1990),

Vol. 30, no. 2

Table 1. Repellency effects of five essential oils on Kunming mice against Aedes albopictus.

Time of exposure mentha piperta oil (7%) 100.00 100.00 100.00 100.00 99.00 99.33 99.00 97.67 15% 95.00 95.56 97.09 96.31 95.32 96.23 92.85 93.28 control * 20 20 22 21 18 20 20 15 citronella oil (7%) 100.00 100.00 100.00 99.67 99.67 99.67 99.67 99.67 rutaceae oil 1% 7% 99.67 99.66 100.00 99.67 98.66 99.67 98.99 100.00 96.29 97.63 98.00 98.26 96.40 96.58 95.81 98.31

control *

1h 2h 3h 4h 5h 6h 7h 8h

23 21 23 27 27 25 27 20

carvacryl oil (7%) 100.00 100.00 100.00 100.00 100.00 100.00 100.00 98.33

Percentage of protection (%) asteraceae oil control* 1% 7% 97.27 99.67 20 98.31 99.67 26 98.99 99.67 20 94.80 94.55 21 93.74 94.86 20 95.83 96.26 26 93.82 95.80 25 90.35 91.81 18

15% 100.00 99.67 100.00 98.99 99.33 97.72 87.29 95.54

*Number of blood-fed mosquitoes.

Journal of Vector Ecology

Table 2. Repellency effects of two essential oils and control treatment against Aedes albopictus on humans.

Time of 30% + + + 30% + 5% vanillin 15% + + + -

citronella oil

eucalyptus oil 15% + 5% vanillin + + + + + 30% + 30% + 5% vanillin + + + + -

exposure + + + -

15%

15%+ 5% vanillin

vanillin (5%) + -

DEET (4%) + + + + + + 233

1h 2h 3h 4h 5h 6h 7h

-

+: protecting, -: non protecting.

234

Journal of Vector Ecology

December 2005

and Suwonkerd and Tantraongroj (1994). It is important to note that citronella oil in the above studies was obtained from different sources. In fact, repellency of a cream containing 14% citronella oil could keep Ae. aegypti away for 2 h (Wasuwat et al. 1990), while 10% citronella oil could keep An. minimus away for less than 2 h under laboratory conditions (Suwonkerd and Tantraongroj 1994). Unfortunately, this laboratory study was conducted during the day whereas An. minimus feeds at night. We also observed that the essential oil solution had repellency against Cx. pipiens quinquefasciatus and An. stephensi but had no repellency against Ae. albopictus (unpublished data). The protection time of the two oils can be significantly increased by the incorporation of 5% vanillin. This agrees with data that vanillin can prolong protection against Ae. aegypti in most cases (Khan et al. 1975). Based on this study, it is suggested that the mixture of 15% eucalyptus oil with 5% vanillin can be developed into a practical product after field evaluation. Essential oils contain complex constituents. Their components are decided by geographical distribution, harvesting time, and growing conditions like soil, water, and nutritional conditions, and their repellent mechanism is still not very clear. The major components must be analyzed if the repellent mechanism were to be studied. Acknowledgments We thank Prof. Yuguang Du (Dalian Institute of Chemical Physics, Chinese Academy of Science) and Prof. Ping Ding (Guangzhou University of Traditional Chinese Medicine) for providing the essential oils and Laiqing Hua (Department of Health Statistics, Second Military Medical University) for valuable suggestions on statistical design and analysis. This study was supported by the Chinese National Natural Science Foundation (No. 30271161). REFERENCES CITED Barnard, D. R. 1999. Repellency of essential oils to mosquitoes (Diptera: Culicidae). J. Med. Entomol. 36: 625-629. Boonyabancha, S., K. Suphapathom, and A. Srisurapat. 1997. Repellent effect of volatile oils on Ae. aegypti. Bull. Dept. Med. Sci. 39: 61-66. Chokechaijaroenporn, O., N. Bunyapraphatsara, and S. Kongchuensin. 1994. Mosquito repellent activities of ocimum volatile oils. Phytomedicine. 1: 135-139. Coleman, R. E., L. L. Robert, L. W. Roberts, J. A. Glass, D. C. Seeley, A. Laughinghouse, P. V. Perkins, and R. A. Wirtz. 1993. Laboratory evaluation of repellents against four anopheline mosquitoes (Diptera: Culicidae) and two phlebotomine sand flies (Diptera: Psychodidae). J. Med. Entomol. 30: 499-502. Edwards, D.L. and C.E. Johnson. 1987. Insect-repellentinduced toxic encephalopathy in a child. Clin. Pharm. 6: 496-498. Frances, S.P., R.D. Cooper, S. Popat, and N.W. Beebe. 2001.

Field evaluation of repellents containing deet and AI337220 against Anopheles koliensis in Papua New Guinea. J. Am. Mosq. Contr. Assoc. 17: 42-44. Hadis, M., M. Lulu, Y. Mekonnen and T. Asfaw. 2003. Field trials on the repellent activity of four plant products against mainly Mansonia population in western Ethiopia. Phytother. Res. 17: 202-205. Khan, A.A., H.I. Maibach, and D.L. Skidmore. 1975. Addition of vanillin to mosquito repellents in Guinea Bissau, West Africa. Acta Trop. 72: 39-52. Moore, S.J., A. Lenglet, and N. Hill. 2002. Field evaluation of three plant-based insect repellents against malaria vectors in Vaca Diez Province, the Bolivian Amazon. J. Am. Mosq. Contr. Assoc. 18: 107-110. Qiu, H., H.W. Jun, and J.W. McCall. 1998. Pharmacokinetics, Formulation, and safety of insect repellent N,N-diethyl3-methylbenzamide (deet): A review. J. Am. Mosq. Contr. Assoc. 14: 12-27. Robbins, P.J. and M.G. Cherniack. 1986. Review of biodistribution and toxicity of the insect repellent N,Ndiethyl-m-toluamide (deet). J. Toxicol. Environ. Hlth. 18: 503-525. Schreck, C.E. and B.A. Leonhardt. 1991. Efficacy assessment of Quwenling, a mosquito repellent from China. J. Am. Mosq. Contr. Assoc. 7: 433-436. Sharma, V.P., M.A. Ansari, and R.K. Razdan. 1993. Mosquito repellent action of neem (Azadirachta indica) oil. J. Am. Mosq. Contr. Assoc. 9: 359-360. Shirai, Y., K. Kamimura, T. Seki, and M. Morohashi. 2000. Proboscis amputation facilitates the study of mosquito (Diptera: Culicidae) attractants, repellents and host preference. J. Med. Entomol. 37: 637-639. Su, S.Z., and B.H. Ye. 1996. Modern Medical Entomology. High Education Press, Beijing, China, pp. 213-233. Sukumar, K., M.J. Perich, and L.R. Boobar. 1991. Botanical derivatives in mosquito control: A Review. J. Am. Mosq. Contr. Assoc. 7: 210-237. Suwonkerd, W. and K. Tantrarongroj. 1994. Efficency of essential oil against mosquito biting. Commun. Dis. J. 20: 4-11. Trigg, J.K. 1996. Evaluation of a eucalyptus-based repellent against Anopheles spp. in Tanzania. J. Am. Mosq. Contr. Assoc. 12: 243-246. Walker, T.W., L.L. Robert, R.A. Copeland, A.K. Githeko, R.A. Wirtz, J.I. Githure, and T.A. Klein. 1996. Field evaluation of arthropod repellents, deet and a piperidine compound, AI3-37220, against Anopheles funestus and Anopheles arabiensis in West Kenya. J. Am. Mosq. Contr. Assoc.12: 172-176. Wasuwat, S., T. Sunthonthanasart, S. Jarikasem, N. Putsri, A. Phanrakwong, S. Janthorn, and I. Klongkarn-ngan. 1990. Mosquito repellent efficacy of citronella cream. J. Sci. Tech. 5: 62-68. Yap, H. H. 1986. Effectiveness of soap formulations containing deet and permethrin as personal protection against outdoor mosquitoes in Malaysia. J. Am. Mosq. Contr. Assoc. 2: 63-67. Zadikoff, C. M. 1979. Toxic encephalopathy associated with use of insect repellent. J. Pediatr. 95: 140-142.

Vol. 30, no. 2

Journal of Vector Ecology

235

Evaluation of habitat management strategies for the reduction of malaria vectors in northern Belize

John P. Grieco1, Roy C. Vogtsberger2, Nicole L. Achee1, Errol Vanzie3, Richard G. Andre1, Donald R. Roberts1, and Eliska Rejmankova4

1

Department of Preventive Medicine and Biometrics, Uniformed Services University of the Health Sciences, Bethesda, MD 20814, U.S.A. 2 Department of Biology, Hardin-Simmons University, Box 16165, Abilene, TX 79698-6165, U.S.A. 3 Ministry of Health, Belize City, Belize 4 Department of Environmental Science and Policy, University of California, Davis, CA 95616, U.S.A. Received 20 January 2005; Accepted 18 April 2005

ABSTRACT: Mowing and burning of emergent vegetation were evaluated as potential management strategies for the control of the malaria vector, Anopheles vestitipennis, in northern Belize, Central America. The primary aim was reduction of tall dense macrophytes (dominated by Typha domingensis) as preferred larval habitat for An. vestitipennis. Nine experimental plots were established in a Typha marsh in Orange Walk District, Belize. Three plots were burned, three were treated by subaquatic mowing, and three were unaltered controls. After treatment, Typha height was most dramatically affected by the mow treatment. Plant heights at 21 and 95 days post-treatment reflected an 89% and 48% decrease, respectively, compared to pretreatment conditions. The Typha height in the burn plots was not as severely affected. Heights at 21 days post-treatment were 39% lower than those of pre-treatment vegetation, with a return to near pre-test heights by 95 days post-treatment. Both treatments resulted in a significant reduction in the number of An. vestitipennis larvae collected as compared to control plots. Conversely, the treatments resulted in increased larval densities of several other vector and pest mosquito species. Larval population densities of An. albimanus, Ochlerotatus taeniorhynchus, and Culex coronator were significantly higher in burn plots. In mow plots, there were significant increases in An. albimanus and Oc. taeniorhynchus larval populations. Non-target invertebrate species affected by the treatments were adult Tropisternus collaris, larval Corythrella, and adult Paraplea puella. Journal of Vector Ecology 30 (2): 235-243. 2005. Keyword Index: Habitat, management, vectors, Anopheles, malaria, Belize.

INTRODUCTION Evidence for the role of Anopheles vestitipennis as a vector of malaria in Belize has been mounting as seen from thorough documentation of human-vector contact (Grieco et al. 2000, Bangs5), high natural field infection rates (Achee et al. 2000), and comparative susceptibility studies (Grieco et al. 2005). For this reason, a renewed effort has been placed on developing methods for the control of this vector species. One available control option is habitat management by eliminating undesirable aquatic and emergent wetland vegetation. Emergent macrophytes have been shown to enhance survival of Anopheles larvae by providing a favorable microhabitat and protection from predators (Orr and Resh 1989). Other beneficial effects, such as enhanced food resources, shelter from physical disturbance, and favorable oviposition conditions, are all created by associated aquatic vegetation (Collins et al. 1985, Walker et al. 1988, Collins and Resh 1989). A strong association has been documented between tall dense macrophytes (TDM), dominated by Typha

5

domingensis, and An. vestitipennis larval habitats (Rejmankova et al. 1998). An assumption could be made from this association that by seasonal elimination of Typha, one could successfully reduce populations of An. vestitipennis. Typha domingensis is a rhizomatous macrophyte reproducing mainly regeneratively. It has a tendency to form dense monocultures and is rather difficult to control. Several methods such as sub-aquatic mowing (Buele 1979) and burning followed by immediate flooding (Nelson and Deitz 1966, McCoy and Rodriguez 1994) have been examined as potential control strategies for emergent macrophytes. Due to their reduced dependence on chemical methods and their established use for the enhancement of waterfowl habitats (Heitmeyer et al. 1989, Rollins 1981, Batzer and Resh 1992), these practices have gained enormous attention in the U.S. as potential control methods for a variety of mosquito species (de Szalay et al. 1995). This habitat management approach has not yet been investigated for the control of malaria vectors in Belize. In addition to affecting the vector populations, habitat

Bangs, M.J. 1999. The susceptibility and behavioral response of Anopheles albimanus Weidemann and Anopheles vestitipennis Dyar and Knab (Diptera: Culicidae) to insecticides in northern Belize. Ph.D. Thesis, Uniformed Services University of the Health Sciences, Bethesda, MD. 489 pp.

236

Journal of Vector Ecology

December 2005

management strategies also have an impact on non-target invertebrate communities. Invertebrate communities contain natural predators of mosquito larvae and provide an important component of the diet of waterfowl. Environmental management strategies may disrupt naturally occurring predator-prey relationships, eliminate plant refuges, or alter overall community structure. Disruption of any or all of these elements may inadvertently promote vector populations and/ or create niches for other disease-transmitting mosquito species. Therefore, as with any vector control strategy, an attempt to document changes to the overall community structure must be made for an accurate assessment of the strategy's long-term practicality and effectiveness. Little is known, however, of the freshwater invertebrate fauna found in Belize and even less is known of the biology and behavior of these organisms. The present study was designed to evaluate the efficiency of two management strategies, mowing and burning. The primary aim was the reduction of TDM with the ultimate goal of eliminating An. vestitipennis habitat and reducing vector abundance. MATERIALS AND METHODS Site description In the summer of 2002, a Typha domingensis dominated marsh was selected for a small-scale evaluation of management strategies. The marsh was located just outside the town of San Pablo in the Orange Walk District of Belize, Central America (N 18'13 36.9, W 88'33 36.5). The marsh was characterized by a relatively homogeneous coverage of Typha domingensis and the presence of high larval densities of An. vestitipennis. Six experimental plots (each 6 m x 6 m) were measured and partitioned off using nylon rope. Three replicates each of mow, burn, and control treatments were randomly assigned to the plots. A 1 m buffer was cut immediately around each plot to delineate the experimental plots and provide a working area from which to maneuver during sampling. A 1 m buffer of vegetation was maintained between plots to serve as an undisturbed refuge. Treatment All treatments were conducted after heavy rains had flooded the plots. Initial attempts at controlling Typha by mowing indicated that the best control resulted from cutting the vegetation 4 to 5 cm below the water surface (i.e. subaquatic mowing). Vegetation in each randomly assigned mow plot was hand mowed using a machete. Cuts were made approximately 5 cm below the water surface. All cut material was subsequently removed from the plot and deposited at the marsh edge. Plots designated for burn were prepared in advance by adding additional fuel in the form of dried palm leaves. These palm leaves allowed for a more intense burn without using chemical fuel or drip pots. Additional efforts were made to bend over the top 0.5 m of Typha in order to contain the flames within the plot and prevent fire from spreading to neighboring plots or to the entire marsh. Vegetation that did not burn was

allowed to remain in the plot for the duration of the experiment. The remaining three plots were left undisturbed to serve as controls. Vegetation Plant height and density were measured pre-treatment and then on days 21, 66, 95, and 358 days post-treatment. Between the final two sample periods, an intense drought was documented throughout the entire country of Belize by the National Meteorological Service. This resulted in the drying of the entire study area and uncharacteristic stress on the vegetation. Water level at approximately 80 m into the marsh was recorded at 13 cm below ground level on Day 358. Thus, Typha measurements taken 358 days post-treatment do not reflect natural growth patterns under normal climatic conditions. Plants were measured in four randomly placed 1/ 4 m2 quadrates. All plants falling within the quadrate were counted, the number of leaves counted, and the average leaf height per individual measured. Water depth was recorded in the plots to coincide with intervals of invertebrate sampling. Invertebrate sampling Aquatic invertebrate sampling was conducted prior to treating in order to establish a baseline for comparison of treatment effects. Post-treatment sampling was conducted weekly for one month followed by monthly sampling for the next five months. Attempts were made to conduct sampling at 12 months after treatment but due to an unusually severe drought all plots were dry, making aquatic sampling impossible. Three collection methods were used to characterize invertebrate population densities and calculate population estimates. The collection methods that most accurately evaluated the invertebrate community structure were: D-frame aquatic sweep netting (net 1 mm mesh size); standard larval dip cup; and submerged bottle traps. All three sample methods were used in each plot to evaluate both the mosquito populations and the non-target invertebrate populations. A specific sequence was followed at each sampling site to minimize the disturbance for subsequent collection methods. Dipping of all plots was conducted first followed by sweeping with the D-frame aquatic net. After all plots were sampled, bottle traps were set in each plot and retrieved the following day. The use of a standard dip cup was the primary method of collecting mosquito larvae and was also successful in sampling smaller, surface swimming invertebrates that could be missed with the aquatic sweep net. Each plot was sampled using 30 dips (15 dips on each diagonal). All invertebrates were removed from the dipper using a disposable pipette and were immediately place in labeled vials containing 80% alcohol. Collections with D-frame aquatic nets were performed by making six complete passes of each plot sweeping the net in back and forth 1 m swaths. At the end of each pass, all material in the net was removed and placed on wire mesh sheets (0.64 cm mesh size) suspended over water-filled enamel pans. The plant material from the net was spread evenly over the mesh to facilitate drying and encourage all invertebrate

Vol. 30, no. 2

Journal of Vector Ecology

237

Table 1. Overall effects of mow or burn treatments as compared to undisturbed control areas on the mean mosquito densities (± S.D.) by sampling period.

Experimental treatment Species

An. vestitipennis An. albimanus An. crucians Ochlerotatus taeniorhynchus Culex coronator

Mow 0* 25.5 (±3.5) * 1.4 (±1.0) 26.5 (±8.6) * 22.4 (±9.2)

Burn 3.2 (±1.0) * 23.8 (±3.7) * 3.2 (±1.4) 24.3 (±6.3) * 28.8 (± 11.2) *

Control 7.9 (±2.5) 7.5 (±2.9) 2.3 (±1.2) 13.6 (± 3.2) 19.4 (± 4.4)

* Indicates a significant difference (P<0.05) for that species between treatment and control. life to seek refuge in the water-filled pan. Pans remained in the sun for approximately 30 min to maximize drying. After this time the screens were removed and a visual examination was made to remove any additional invertebrates remaining in the dried material. All invertebrates were then removed from the pan using either forceps or a disposable pipette. All invertebrates from individual plots were placed into labeled vials containing 80% ETOH to insure all insects had been killed and that samples were properly preserved until processing. Each plot was visually divided into four equal 3 m by 3 m quadrants with one bottle trap being placed in the center of each quadrant. Bottle traps were placed just below the water surface in a horizontal position. At the time of placement, air bubbles were removed from the bottle in order to deprive collected invertebrates of an oxygen supply. Each bottle trap was tied to a stake in the corresponding quadrant to ensure the bottles remained in a horizontal position and to prevent them from moving during the night. Bottles were retrieved 24 h after placement and the contents poured through small sieve bags cut from pantyhose. The contents from all four bottles from a given plot were placed into a single bag along with a plastic label indicating the plot number. Bags were immediately placed into bottles of alcohol to kill any insects that had recently entered the trap and had remained alive as well as to preserve material before laboratory processing. At the conclusion of each sampling session all material was returned to the field laboratory where it was processed for further identification. Mosquito larvae were removed and identified using taxonomic keys for the region (Faran 1980, Clark-Gil and Darsi 1983). Identified mosquito larvae were separated by species, type of collection method, and plot number and were placed in shell vials containing 80% ETOH for long-term storage. All other material from all other collection methods were taxonomically sorted by family and placed in vials containing 80% ETOH and labeled with collection method, plot number, and date of collection. All invertebrate material except for mosquito larvae was forwarded to collaborators at Hardin-Simmons University for identification. RESULTS Typha domingensis height (Figure 1) was most dramatically affected by the mow treatment and all heights post-treatment were significantly different from those prior to treatment. Plant heights before treatment averaged 281 cm. On Day 21 post-treatment plants were 30.2 cm tall (89% decrease). By Day 95 post-treatment, the Typha had only regenerated to 48% of its original height. The Typha height in the burn plots was not as severely affected, although plant heights up to 6 months post-treatment were significantly different from plant heights before treating the plots. This is due, in part, to a low-intensity burn caused by treatment of the plots after they had already been flooded. The overall result was minimal reduction of emergent vegetation. Plant height was, however, reduced from 273 cm pre-treatment to 165.2 cm 21 days post-treatment. This is a 39% reduction between pre- and post-treatment levels. The regeneration was almost complete by Day 95 with only a slight decrease in overall Typha height. In contrast, the mow plots were still reduced by more than half their original height compared to pre-treatment levels by day 95. The resulting difference in the burn plots, however, remained significant from pretreatment and control plots. Measurements deviated by 13% from pre-treatment levels. The height of the Typha in the control plots was not significantly different from average heights pre-treatment and 21, 66, and 95 days post-treatment measuring 274, 272, 267, and 264 cm, respectively. Plant heights for all treatment plots at 358 days post-treatment were not significantly different. Average heights in the control, burn, and mow plots were 127, 92, and 125 cm, respectively. Seasonal larval collections conducted throughout northern Belize have indicated that larval populations of the dominant anopheline species increase during the late wet season (November-December) and decline in the late dry season (April-May) (Figure 2). These data are the result of ongoing studies conducted in the districts of Orange Walk and Corozal in northern Belize. Larval population increases are in direct association with water depth in the plots. A total of 5,495 mosquito larvae was collected, comprising three genera and five species in the following

238

Journal of Vector Ecology

December 2005

San Pablo

320 240 160 80 0 1 320 240 160 80 0 1 320 240 160 80 0 1 21 C3 66 95 1 21 M3 66 95 1 21 B3 66 95 21 M2 66 95 1 21 B2 66 95 1 21 C2 66 95 21 B1 66 95 1 21 C1 66 95 1 21 M1 66 95

HEIGHT

HEIGHT

DAY

Figure 1. Plant height (y-axis, cm) by treatment plot and by day (x-axis). Sample day 1 is under pretreatment conditions and was conducted on 10 July 2002. B=Burn, M=Mow, C=Control.

1400

An. vestitipennis

1200 1000

1192

An. albimanus An. crucians 736

1162 911

876 670

Number of larvae

800 600

459 367 267 66 106 2

Oct.-Nov Feb-M arch M ay-June

400 200 0

Aug.-Sept.

Season

Figure 2. Seasonal abundance of Anopheles larvae collected throughout the Orange Walk and Corozal Districts of Belize from August 2000 to December 2001.

Vol. 30, no. 2

Journal of Vector Ecology

239

densities: 316 (5.8%) An. vestitipennis, 1,541 (28%) An. albimanus, 28 (0.5%) An. crucians, 1,752 (31.9%) Ochlerotatus taeniorhynchus, and 1,858 (33.8%) Culex coronator. All species could be found in all the plots in varying population densities except for An. vestitipennis which was not found in the mow plots post-treatment. The only mosquito species that was significantly reduced by the treatment was An. vestitipennis (Table 1, Figures 3 and 4). A significant reduction in An. vestitipennis larvae was documented in both the mow (p<0.05) and burn plots (p<0.05) as compared to the control plots. This was most evident in the mow plots where no An. vestitipennis larvae were collected up to 6 months post-treatment (Figure 4). Monthly sampling of the plots showed An. vestitipennis larval populations increased in the control plots in the wet season (October-December). An increase in larval density also occurred in the burn plots during the wet season; however, the increase was much less pronounced in the burn plots than in control plots. Populations of An. albimanus larvae increased significantly in both the mow (P< 0.01) and burn plots (p< 0.01) as compared to control plots (Table 1, Figures 3 and 5). An immediate increase in larval density occurred 1 week posttreatment, and levels in both mow and burn plots reached a peak 6 months post-treatment. Again, a noticeable increase was observed during the wet season to coincide with natural seasonal increases associated with this species. Two pest mosquito species also demonstrated marked increases in population density in one or both of the treatment plots: Oc. taeniorhynchus and Cx. coronator. Both species showed significantly higher population densities in the burn and mow plots as compared to control plots and both showed seasonal increases during the wet season. The most dramatic impact to the non-target invertebrate community identified from this study was seen in the mowed plots. Unfortunately, the invertebrate fauna of Belize has been poorly described in the literature and taxonomic keys for the region are scarce. For this reason, many of the non-target invertebrate species found in the plots have yet to be processed and identified. Several species, however, have been identified and their densities analyzed. The three species that demonstrated pronounced increases in overall density in the burn plots as compared to pre-treatment and control conditions were adult Tropisternus collaris (Coleoptera), larval Corythrella (Diptera), and adult Paraplea puella (Hemiptera) (Figure 5). Both T. collaris and P. puella also showed significant increases in density in the mow plots. DISCUSSION Decreasing budgets and tighter environmental restrictions on international control programs are steering the public health community toward finding new methods for reducing mosquito populations. In Belize, the mounting evidence for the increased potential of An. vestitipennis as a competent malaria vector creates new challenges for control programs. Looking at efforts made in other regions to control mosquito populations under increasing environmental restrictions, one finds successful use of practices such as mowing and burning

of emergent wetland vegetation. Many larval mosquitoes use this vegetation as protection from predators, shelter from wind action, and as a source of nutrients created from developing biofilms (Orr and Resh 1989). Vegetation removal has been shown to reduce the dominant mosquito taxa that inhabit treated areas. The strong association between larval An. vestitipennis and Typha domingensis suggests that vegetation removal is a viable management strategy. Timing of such treatment, however, is critical to ensuring Typha stems are immediately flooded, which cuts off contact with oxygen and prevents regeneration. Unless all of the vegetation is removed, there will continue to be available habitat for further mosquito production. This is evident from the significant differences seen between the burn and mow treatment results in the present study. The burn treatments were conducted after the plots had already been flooded. This resulted in a low-intensity fire that did not sufficiently eliminate the above ground Typha. The heat from the fire did kill much of the Typha and resulted in extensive bending of dead leaves and vegetation into the plot. The resulting effects of the treatment created suitable conditions for algal and bacterial growth and at the same time provided ample protection for developing larvae from predators. While decreases were documented, the burn treatment did not completely eliminate An. vestitipennis populations. It must be stressed, however, that provided proper timing and a second burn approximately 1 week after the initial burn, the desired effect of complete above-ground elimination of Typha might be achieved. Such a burn would eliminate many of the undesirable effects seen in this study and undoubtedly elicit stronger control pressures on all mosquito populations. This study shows that an efficient method of controlling Typha is to mow the vegetation below the water surface in order to deplete the rhizomes of oxygen and induce mortality. This also resulted in the greatest reduction of An. vestitipennis larvae, compared to pre-treatment or control conditions. The open areas created in the absence of Typha did, however, increase the production of another vector species, namely, An. albimanus, as well as two pestiferous mosquitoes species, Oc. taeniorhynchus and Cx. coronator. The production of these other species was significant. In the mow plots, Oc. taeniorhynchus populations demonstrated a 100% increase from the control plots, while in the burn plots they showed a 95% increase as compared to the control plots. In the mow and burn treatments, Cx. coronator populations showed a 30% and 48% increase, respectively. The increase was even greater with regards to An. albimanus populations, demonstrating an increase of over 200% in both the mow and burn plots as compared to the control plots. One other major drawback of this control method is that it is very labor intensive and time consuming. Public acceptance and work involved in subaquatic mowing and subsequent removal of all cut material preclude this as a viable control strategy. Alternative methods of automated vegetation removal, such as mechanized discing, should be examined to achieve the same desired results. The population densities of a number of non-target invertebrate species were also dramatically affected by both

240

35

Journal of Vector Ecology

December 2005

Average number of larvae

30 25 20 15 10 5 0

CONTROL

An. vestitipennis

An. albimanus

BURN

MOW

Treatment

Figure 3. Average number of Anopheles vestitipennis and An. albimanus larvae collected post-treatment from all sample periods in the control, burn, and mow treatment plots.

An. v estitipennis

18 16 14 12 10 8 6 4 2 0 P ret reat W eek 1 W eek 2 W eek 3 W eek 4 Month 2 Month 3 Month 4 Month 5 Mont h 6

Number of larvae / 30 dips

vesti burn

vesti mow

vesti c ontrol

Time post treatment

Figure 4. Average number of An. vestitipennis larvae pre- and post-treatment for control, burn, and mow plots.

An. albimanus

45 40 35 30 25 20 15 10 5 0 Pretreat Week 1 Week 2 Week 3 Week 4 Month 2 Month 3 Month 4 Month 5 Month 6

Number of larvae / 30 dips

albi burn

albi mow

albi control

Time post treatment

Figure 5. Average number of An. albimanus larvae pre- and post-treatment for control, burn, and mow plots.

A

18 16 14 12 10 8 6 4 2 0 B u rn M ow C o n tro l

Average number /p l plot A v e ra g e # per

T rea tm en t

B

Average number per plot

20 15 10 5 0 Burn Mow Control Average #/plo

Treatment

C

Average #/plo Average number per plot 12 10 8 6 4 2 0 Burn Mow Control

Treatment

Figure 6. Average pre- (shaded) and post-treatment (unshaded) population densities of three non-target species that have potential as predator species against mosquito larvae by treatment. A: Average number of Paraplea puella; B: Average number of Tropisternus collaris; and C: Average number of Corethrella spp.

242

Journal of Vector Ecology

December 2005

treatments. These affects, both negative and positive, were most pronounced in the sub-aquatic mow treatment where all above water vegetation was removed. To date, three invertebrate species have been identified as having been positively affected by the treatments: adult P. puella, T. collaris, and larval Corythrella spp. The aquatic invertebrate fauna found in Belize has been poorly documented in the literature and even less is known of the biology and behavior of many of these organisms. Studies are presently being conducted to determine the predation levels of these and other invertebrate species on An. vestitipennis larvae. In addition to altering the invertebrate community structure, the mow treatment effectively removed protective barriers to predator-prey interactions. The effects of the burn treatment, while less pronounced, demonstrate the dangers created if treatments are not timed properly. By essentially killing the upper portions of Typha while not completely removing it from the system, a situation is created in which protective debris remain to shelter the larvae from predators and wind, while at the same time providing increased nutrients for the mosquito larval diet. Studies are presently underway to evaluate other potential treatment methods. For example, studies are being conducted to look at the effects of double burning as compared to mowing. This treatment will be performed during the dry season when the plots are dry and allow for a more intense burn; thereby destroying all above-ground vegetation. Flooding of the plots after such a burn would separate the rhizomes from a source of oxygen and prevent rapid plant regeneration. Timing of such a treatment could reduce Typha density during the three months when An. vestitipennis reaches its peak population density. It is clear from this study that both the burn and mow treatments had a dramatic effect on the plant and invertebrate communities. While both treatments resulted in a reduction in An. vestitipennis larvae, they also created favorable habitats for other mosquito species. Most notable was an increase in the malaria vector, An. albimanus. Dramatic increases in An. albimanus larvae may be the result of a newly-established niche where the predator-prey communities had not yet been established. Once fish and invertebrate communities become established in these plots, predation levels may keep mosquito populations in check. This assessment can only be made through long-term monitoring of treatment effects. Information gained on the biology and behavior of the invertebrate community altered by these treatments will also aid our understanding of the long-term effects of plant management practices. Acknowledgments We are grateful for administrative and logistical support provided by Dr. Jorge Polanco (Ministry of Health, Belize). We would also like to thank Ireneo Briceno, Russell King, and Petr Macek for their help in conducting the field experiments. This research was supported by the NIH-NSF Ecology of Infectious Diseases program, Grant #R01 AI49726, "Environmental Determinants of Malaria in Belize."

Disclaimer: The opinions and assertions contained in this article are not to be considered as official or as reflecting the views of the Department of Defense or the Uniformed Services University of the Health Sciences. REFERENCES CITED Achee, N.L., C.T. Korves, M.J. Bangs, E. Rejmankova, M. Lege, D. Curtin, H. Lenares, Yvette Alonzo, R.G. Andre, and D.R. Roberts. 2000. Plasmodium vivax polymorphs and Plasmodium falciparium circumsporozoite proteins in Anopheles (Diptera: Culicidae) from Belize, C.A. J. Vector Ecol. 25: 203-211. Batzer, D.P. and V.H. Resh. 1992. Wetland management strategies that enhance waterfowl habitats can also control mosquitoes. J. Am. Mosq. Contr. Assoc. 8: 117-125. Buele, J.D. 1979. Control and management of cattails in southeastern Wisconsin wetlands. Wisconsin Department of Natural Resources Tech. Bull. 112, 41 pp. Clark-Gil, S. and R.F. Darsie. 1983. The mosquitoes of Guatemala: Their identification, distribution and bionomics with keys to adult females and larvae. Mosq. Syst. 15: 151-284. Collins, J.N., K.D. Gallagher, and V.H. Resh. 1985. Thermal characteristics of aquatic habitats at Coyote Hills Marsh: implications for simulation and control of Anopheles mosquitoes. Proc. Calif. Mosq. Vector Contr. Assoc. 53: 83-86. Collins, J.N. and V.H. Resh. 1989. Guidelines for ecological control of mosquitoes in non-tidal wetlands of the San Francisco Bay Area. California Mosquito and Vector Contr. Assoc., Sacramento, CA 93 pp. de Szalay, F.A., D.P. Batzer, E.B. Schlossberg, and V.H. Resh. 1995. A comparison of small and large scale experiments examining the effects of wetland management practices on mosquito densities. Proc. Calif. Mosq. Vector Contr. Assoc. 63: 86-90. Faran, M.E. 1980. A revision of the Albimanus section of the subgenus Nyssorhynchus of Anopheles. Contrib. Am. Entomol. Inst. 214 pp. Grieco, J.P., N.L. Achee, R.G. Andre, and D.R. Roberts. 2000. A comparison study of house entering and exiting behavior of Anopheles vestitipennis (Diptera:Culicidae) using experimental huts sprayed with DDT or deltamethrin in the southern district of Toledo, Belize, C.A. J. Vector Ecol. 25: 62-73. Grieco, J.P., N.L. Achee, D.R. Roberts, and R.G. Andre. 2005. Comparative susceptibility of three species of Anopheles from Belize, Central America to Plasmodium falciparum (NF-54). J. Am. Mosq. Contr. Assoc. (In press). Heitmeyer, M.E., D.P. Connelly, and R.L. Pederson. 1989. The Central, Imperial and Coachella Valleys of California. In: Habitat management for migratory and wintering waterfowl in Northern America. Texas Tech University Press, Lubbock, TX. 560 pp. McCoy, M.B. and J.M. Rodriguez. 1994. Cattail (Typha dominguensis) eradication methods in the restoration of a tropical, seasonal, freshwater marsh. In: Global

Vol. 30, no. 2

Journal of Vector Ecology

243

Wetlands: Old World and New. Elsevier Sciences. pp. 469-482. Nelson, N.F. and R.H. Deitz. 1966. Cattail control methods in Utah. Utah Department of Fish and Game Publ. 66-2, 31 pp. Orr, B.K. and V.H. Resh. 1989. Experimental test of the influence of aquatic macrophyte cover on the survival of Anopheles larvae. J. Am. Mosq. Contr. Assoc. 5: 581585. Rejmankova, E., K. Pope, D. Roberts, M. Lege, R. Andre, J. Greico, and Y. Alonzo. 1998. Characterization and

detection of Anopheles vestitipennis and Anopheles punctimacula (Diptera: Culicidae) larval habitats in Belize with field survey and SPOT satellite imagery. J. Vector Ecol. 23: 74-88. Rollins, G.L. 1981. A guide to waterfowl habitat management in Suisun Marsh. Calif. Dept. Fish Game Publ. 106 pp. Walker, E. D., R.W. Merritt, and R.S. Wotton. 1988. Analysis of the distribution and abundance of Anopheles quadrimaculatus (Diptera: Culicidae) larvae in a marsh. Environ. Entomol. 17: 992-999.

244

Journal of Vector Ecology

December 2005

Methods for monitoring outdoor populations of house flies, Musca domestica L. (Diptera: Muscidae)

Christopher J. Geden

USDA, ARS, Center for Medical, Agricultural and Veterinary Entomology, P.O. Box 14565, Gainesville, FL 32607, U.S.A. Received 24 January 2005; Accepted 4 April 2005 ABSTRACT: Relative collections of house flies were compared on two Florida dairy farms using several monitoring methods: sticky cylinders, baited jug traps (Farnam Terminator and Victor Fly Magnet), and bait strips (Wellmark QuikStrike). Bait strips were placed over collecting pans and under 61 cm square plywood roofs to protect the toxicant from sunlight ("sheltered QuikStrike traps"). Sticky cylinders collected the fewest flies (515-679 flies/trap/day) and sheltered QuikStrike traps the most (5,659-8,814 flies/trap/day). The sheltered QuikStrike traps are promising tools for disease surveillance programs. The two baited jugs collected a similar and intermediate number of flies, with collections highest during the first 2 days after placement (2,920-5,462 flies/trap/day). Jug trap collections were low after 4 days of use in the field, apparently due to deterioration in the attractiveness of the bait over time. Jug traps collected mostly females, whereas sticky cylinders and sheltered QuikStrike traps collected mostly males. Exposure of jug trap bait (Farnam) to fly cadavers for 3 days did not increase attractiveness of the bait. Combinations of the Farnam and Victor attractants were more attractive than either attractant alone and 25-43% more attractive than expected based on the sum of collections in the single-attractant jug traps. A 25% solution of farm-grade blackstrap molasses was as effective as either of the two proprietary baits tested, offering a low-cost alternative for fly population monitoring. Journal of Vector Ecology 30 (2): 244-250. 2005. Keyword Index: House fly, Musca domestica, trapping, sampling, monitoring.

INTRODUCTION House flies have long been known to transmit a variety of pathogens to humans and animals (Howard 1911). Recent concerns about food-borne human illnesses have led to renewed interest in the role of flies in spreading diseasecausing organisms, especially Escherichia coli and Shigella spp. (Cohen et al. 1991, Buma et al. 1999, Kobayashi et al. 1999, Agui 2001, Graczyk et al. 2001, Nayduch and Stuzenberger 2001). Pathogen-carrying flies are commonly found in animal agriculture production facilities and landfills, from which they may disperse to more densely populated areas where they pose a greater risk to humans (Goulson et al. 1999, Mian et al. 2002, Sulaiman et al. 2000, Olsen and Hammack 2000, Meek 2001, Moriya et al. 1999). When fly production and dispersal from animal agricultural facilities exceeds levels acceptable to the public, public health workers and entomologists are often consulted to develop fly management programs. A critical element of such programs is the use of monitoring methods to document their effectiveness. Sampling methods for fly populations in enclosed systems such as animal barns and poultry houses are well-established. These methods can include the use of spot cards, baited traps, or sticky tapes and cards, and the numerical relationships among collections with different methods have been determined in several cases (Beck and Turner 1985, Burg and Axtell 1984, Hogsette et al. 1993,

Lysyk and Axtell 1985, 1986, Stafford et al. 1988, Geden et al. 1999). House fly populations also occur outdoors in areas near breeding sites and in open fields where manure has been spread. Monitoring flies in such open systems is more challenging than in closed spaces. Spot cards, which are the most commonly used method for indoor situations, are impractical in open areas. The Scudder grid has often been used as a method for measuring relative abundance of flies outdoors. Although this method has the virtue of providing instantaneous estimates of relative fly abundance (Murvosh and Thaggard 1966, Scudder 1996), grid counts are subject to wide variation depending on positional effects, subjective differences among people making the counts, time of day, and weather conditions. The objective of this study was to evaluate different methods for sampling outdoor populations of house flies using commercially available products that are sold as fly control devices and to assess modifications of these products for sampling purposes. Specifically, the goals were to: 1) compare several commercially available baited traps for outdoor use; 2) compare baited trap collections with a passive sampling method (sticky traps); 3) determine whether QuikStrike bait strips can be used as a sampling tool; 4) observe the performance of trap attractants over several days; and 5) evaluate new configurations of trapping products and attractants.

Vol. 30, no. 2 MATERIALS AND METHODS

Journal of Vector Ecology

245

Sampling methods, product sources, and study sites Alsynite cylinders as described by Broce (1988) with adhesive-coated white sleeves were obtained from Olson Products, Medina, OH and held on stakes ca. 1 m above the ground. Two types of baited jug traps with their respective proprietary attractants were evaluated; the Terminator trap produced by Farnam Co. (Phoenix, AZ) and the Victor Fly Magnet by Woodstream Co. (Lititz, PA). Jug traps were suspended from metal plant hangers ca. 1 m above the ground. Attractant fluids for the jug traps were prepared following product label instructions. QuikStrike bait strips (Wellmark International, Schaumburg, IL) were also used. The strips include an ampule of proprietary attractant and are composed of a plastic strip coated with sugar and the insecticide nithiazine. The ampule is crushed to release the attractant when the strips are deployed. Because nithiazine is inactivated rapidly by sunlight, QuikStrike strips are intended for indoor use. In our trials, the strips were suspended under a 61 x 61 cm sheet of plywood held 25 cm above the ground by wooden stakes. An aluminum metal pan was placed beneath the strip to collect flies, which typically died within several seconds of feeding on the strip. Pans were emptied and flies counted daily. These stations are hereafter referred to as "sheltered QuikStrike" traps. Testing was conducted in the summer of 2002 on two dairy farms in Gilchrist and Alachua Co., FL. Traps on one farm, with ca. 400 milking animals, were located in an open area with high fly activity adjacent to the milking barn. On the second farm, with a milking herd of ca. 250 animals, traps were placed in a shaded location adjacent to a row of outdoor calf hutches. Traps were arranged in a linear series 10 m apart and their positions were rotated each farm visit. Fly collections were made daily during all tests. Trial 1: Sticky traps versus jug traps and sheltered quikstrike traps Terminator jug traps, Fly Magnet jug traps, sticky cylinders, and sheltered QuikStrike traps were evaluated for 5 weeks, 3 traps/farm/day, 4 days per week. New sticky sleeves were placed on the traps daily and the flies counted. Fresh attractants were prepared for the jug traps on the first day of each 4-day test; flies were sieved from the attractant and counted, and the fluids reused in the traps for days 2-4. Fresh QuikStrike strips were placed each week. The strips were replaced during tests when an estimated 50% of the sugar coating had been depleted by fly activity. Trap collection data (number of flies collected and percent female flies) were subjected to two-way analysis of variance using trap type, day (1-4) and trap*day interaction as the grouping variables; Tukey's means separation procedure was used to compare trap collections within days (SAS Institute 1992). Separate regressions for each day after trap placement were also conducted to compare fly collections on sticky cylinders (a passive sampling method) with the three attractant-based traps.

Trial 2: Effect of "fly-conditioning" on attractiveness of baits Terminator traps with freshly prepared Farnam attractant were placed in the field and allowed to collect flies for 3 days. Flies (ca. 15,000 per trap) were separated from the fluid, and the used fluid was compared with fresh attractant over an additional 4 days. Victor Fly Magnet traps with fresh attractant were also included for comparison. Flies were removed daily and the fluids reused as in Trial 1. Tests were conducted for 3 weeks on both farms, 3 traps/treatment/farm. Data were analyzed by two-way ANOVA and Tukey's procedure as in Trial 1. Trial 3: Effect of combining attractants in jug traps Jug traps were baited with either Farnam attractant alone, Victor attractant alone, or in combination. Combinations were done by first preparing Farnam attractant following manufacturer's instructions (30 ml of concentrate in 1,500 ml water). Victor attractant, a pelleted dry product, was added to the resulting fluid at the rate recommended for the Fly Magnet. Tests were conducted for 3 weeks on both farms, 3 traps/treatment/farm. Data were analyzed by two-way ANOVA and Tukey's procedure as in Trial 1. Trial 4: Large pail traps Pail traps were prepared by first cutting holes in the lids of standard 19-liter (5-gallon) white plastic pails and inserting the top (fly entry ports) of a Terminator jug trap. Pail traps were baited with either a single QuikStrike strip (with attractant released by breaking the ampule) or 3 liters of Farnam attractant. A third hybrid trap was tested as well. In this trap, 1.5 liters of Farnam bait was placed in a screen-topped 2-liter plastic container and placed in the bottom of the pail. In addition, a QuikStrike bait strip was placed in the pail. The purpose of this design was to determine whether flies would be drawn into the trap by the liquid attractant within the screened container, killed by feeding on the QuikStrike strip, and collected in a dry condition. Standard jug traps were included in the tests for purposes of comparison. Pail trap tests were conducted on 10 individual days, 3 traps/type, with fresh attractant prepared for each test. Data were analyzed by one-way ANOVA and Tukey's procedure. Trial 5: Molasses as an attractant Fly activity on dairy farms is high in areas where molasses is mixed with grain. To evaluate the relative attractiveness of molasses, Terminator jug traps were baited with either 25% dilutions of farm-grade blackstrap molasses (obtained from a local feed store), Farnam attractant mixed with water at the usual dilution, or Farnam attractant mixed with 25% molasses solution. A small amount (5 g) of detergent (Alcanox) was added to the molasses solutions as a surfactant. Traps were placed on the farms (3 traps/treatment/farm) for 4-day tests as in Trials 1-3. Data were analyzed by two-way ANOVA and Tukey's procedure.

246

Journal of Vector Ecology

December 2005

Table 1. Fly collections and percent female flies collected on sticky cylinder traps, Terminator jug traps, Victor Fly Magnet jug traps, and sheltered QuikStrike traps. New sticky sleeves were placed on cylinder traps daily. Flies were removed daily from jug traps for 4 days after initial trap placement; attractant fluids (minus previous fly collections) were re-used on days 2-4. Day Cylinder Terminator Fly Magnet Sheltered QuikStrike

Mean no. house flies/trap1 1 2 3 4 661 (83)b 679 (76)c 678 (106)c 515 (74)b 5,462 (1512)a 4,356 (955)b 3,080 (721)b 904 (166)b 2,920 (1272)a 2,934 (785)bc 2,520 (672)bc 1,611 (384)b 6,015 (1029)a 8,814 (1428)a 7,366 (938)a 5,659 (988)a

1 2 3 4

21.9 (2.3)c 19.8 (1.5)c 19.8 (1.8)c 15.9 (1.6)d

% Female flies 72.1 (3.4)a 59.1 (4.6)ab 73.8 (3.1)a 78.0 (3.7)a 62.2 (4.1)ab 68.2 (5.1)a 66.5 (3.2)b 76.8 (3.3)a

47.5 (4.4)b 35.2 (4.2)b 46.2 (4.7)b 41.0 (3.3)c

ANOVA F2: Treatment Day Trt* Day

1 2

No. flies collected 50.6** 6.24** 2.21*

% females 269.74** 1.25 ns 4.59 **

Means within rows followed by the same letter are not significantly different (Tukey's means separation method). **P<0.01, *P<0.05, ns P>0.05.

Table 2. Regression analysis of fly catches with Terminator, Fly Magnet, and sheltered QuikStrike traps against sticky cylinder traps with white sleeves. Day Terminator Fly Magnet R-square 0.742 0.266 0.049 0.125 Slope (SE)1 19.4 (2.9) ** 6.9 (2.9) * 1.4 (1.6) ns 2.3 (1.5) ns Sheltered QuikStrike

1 2 3 4

0.876 0.287 0.084 0.192

0.846 0.603 0.212 0.242

1 2 3 4

1

22.8 (3.0) ** 9.3 (3.3) * 2.6 (1.6) ns 1.5 (1.5) ns

10.2 (1.1) ** 12.2 (2.4) ** 2.7 (1.3) ns 5.3 (2.4) *

**P<0.01, *P<0.05, ns P>0.05. N=30 traps of each type for each day.

Vol. 30, no. 2 RESULTS AND DISCUSSION

Journal of Vector Ecology

247

Trial 1: Sticky traps versus jug traps and sheltered quikstrike traps The three attractant-based devices (jug and sheltered QuikStrike traps) collected statistically similar numbers of house flies on the first day after trap placement (ca. 3,0006,000 flies/trap), and all three collected significantly more flies than the sticky cylinders (661 flies) (Table 1). On days 2-4, the QuikStrike traps collected two to six times as many flies as the jug traps. Fly collections on sticky cylinders and in QuikStrike traps changed little over the 4-day study interval, whereas jug trap collections declined substantially during this time. Collections in Terminator jug traps declined from 5,462 to 904 flies/traps on days 1 and 4, respectively, and collections in the Fly Magnet decreased by about half. There were no statistically significant differences between the two types of baited jug traps on any day. Jug traps collected higher proportions of female flies (66-78%) than QuikStrike stations (35-48%) or sticky traps (16-22%). Jug traps and sheltered QuikStrike trap counts closely tracked sticky cylinder trap counts on day 1 after trap placement (R-square=0.74-0.88), but sampling precision dropped off sharply on subsequent days (Table 2). The decay in sampling precision was more rapid and pronounced with jug traps than with sheltered QuikStrike traps. These results indicate that the attractants in the jug traps lost effectiveness over the 4-day observation period. The sheltered QuikStrike trap appears to be an effective fly monitoring tool that can be used as an effective alternative to baited jugs and sticky traps. The traps have the additional virtues of being easy to service and leaving flies in a dry condition, thus not requiring handling of liquid baits that can have an objectionable odor. Moreover, because flies are collected in a pan beneath the strip, the QuikStrike traps are not as prone to fly saturation effects as are the other methods, which reach saturation levels at about 2,000 flies for sticky cylinders and 20,000 flies for jug traps. Trial 2: Effect of "fly-conditioning" on attractiveness of baits This trial was conducted to determine whether liquid baits that have already collected flies are more attractive than fresh baits in subsequent collections. Our results indicated no difference in the numbers of flies collected between fresh and fly-conditioned Farnam bait on any of the days in the 4-day observation period (Table 3). Fly-conditioned bait collected significantly higher proportions of male flies than fresh attractants on the first three days after trap placement, suggesting degradation of the compounds in the bait that, when fresh, preferentially attracts female flies. As was observed in the first trial, bait attractiveness decreased markedly between days 1 and 4. In this test, the Fly Magnet traps collected substantially fewer flies than the Terminator traps baited with fresh attractant on days 1-3. The reason for the differences in the relative performance of the two traps types in Trial 1 and 2 is uncertain, but may have been partially due to temperature. Temperatures during Trial

Table 3. Comparison of Terminator jug traps baited with fresh Farnam attractant or fly-conditioned attractant recovered from traps after 3 days of fly collection in the field. Flies were removed from fluid daily for 4 days after initial trap placement. Victor Fly Magnet jug traps were included for comparison. _____________________________________________________ Day Farnam attractant Victor attractant (fresh) Fresh Fly-conditioned _____________________________________________________ Mean (SE) no. flies collected1 1 7,953 (1443)a 8,149 (832)a 766 (499)b 2 6,762 (1901)a 3,337 (604)ab 593 (360)b 3 4,749 (1833)a 2,429 (699)ab 85 (24)b 4 994 (489)a 706 (521)a 72 (32)a % Female flies 33.0 (3.1)b 60.2 (4.2)b 57.0 (15.1)b 66.8 (2.5)a

1 2 3 4

90.9 (0.4)a 75.9 (2.9)a 78.5 (0.6)a 64.7 (2.5)a

86.5 (2.8)a 82.8 (0.7)a 77.6 (3.0)a 56.5 (2.2)a

ANOVA F2: No. flies collected % females Treatment 24.56** 24.63** Day 13.91** 2.30ns Trt*Day3.27* 9.41** _______________________________________________________ 1 Means within rows followed by the same letter are not significantly different (Tukey's means separation method). 2 ** P<0.01, * P<0.05, ns P>0.05.

2 were exceptionally high, with daily highs of 37oC. Trial 1, which was a longer study, was conducted earlier in the summer, when daily temperature maxima rarely exceeded 32oC. It is possible that these extreme temperatures had different effects on the volatility or stability of the two liquid baits. Trial 3: Effect of combining attractants in jug traps Traps baited with mixtures of Farnam and Victor attractants collected significantly more flies than traps baited with either attractant alone on days 2-4 (Table 4). There was also evidence for a synergistic effect of the two attractants. Attractant combinations collected substantially more flies than would be expected based on the sum of the collections with the single attractant on each day. This was most apparent on day 4, when mixtures collected 43% more flies (3,662 flies/ trap) than the sum of flies collected in the traps baited with single attractants (1,037 and 1,059 flies/trap for the Farnam and Victor baits, respectively). Sex ratios of flies collected in the traps were unaffected by type of attractant. Attractant mixtures could be effective for fly monitoring in situations where fly populations are relatively low such as in neighborhoods near animal production facilities.

248

Journal of Vector Ecology

December 2005

Table 4. Comparison of jug traps baited with Farnam attractant, Victor Fly Magnet attractant, or a combination of the two attractants. Flies were removed from fluid daily for 4 days after initial trap placement. Day Farnam attractant Victor attractant Farnam + Victor attractants

left these traps soon after entry without feeding on the strip. Trial 5: Molasses as an attractant A 25% dilution of animal-feed-grade blackstrap molasses was as effective as Farnam attractant on days 2-4 of the study (Table 6), and the addition of Farnam attractant to the molasses solution did not result in increased collections. Proportionally more females responded to the Farnam attractant (73-83% females) than to the molasses-baited traps (41-68%). Molasses is a complex substance that comes in a myriad of grades and has long been noted for its attractiveness to flies (Howard 1911). Not all grades of molasses may be equally attractive, however. During this study we also baited several jug traps with a premium grade molasses sold for human consumption. These traps collected virtually no flies (unpublished data). The blackstrap molasses that was used in the test presented in Table 6 was purchased from a livestock feed store where it is stored in bulk barrels. Proprionic acid is added to this product to retard spoilage. It is uncertain whether the strong response that this molasses product elicited from flies was due to partial fermentation, the propionic acid, or to some of the hundreds of natural compounds present in blackstrap (Binkley and Wolfrom 1953). In summary, the two most common commercially available jug traps are effective for monitoring outdoor house fly populations, especially females, but their attractants degrade at unpredictable rates that may be related to temperature and previous fly collection history. Combining the attractants of the two products synergizes their effectiveness and such combinations could be useful when greater trap sensitivity is desired. Blackstrap molasses offers a low-cost alternative to commercial baits and is widely

1 2 3 4

Mean (SE) no. flies collected1 469 (78)b 1,279 (647)ab 2,295 (591)a 2,631 (638)b 2,114 (564)b 6,847 (934)a 2,457 (669)b 2,030 (369)b 6,812 (1024)a 1,037 (395)b 1,059 (253)b 3,662 (863)a % Female flies 60.6 (5.9)a 80.9 (5.8)a 66.0 (10.4)a 83.9 (3.6)a

1 2 3 4

69.9 (4.5)a 79.3 (5.1)a 59.2 (7.6)a 73.7 (4.2)a

74.6 (2.0)a 62.6 (6.0)a 61.4 (8.7)a 79.0 (2.6)a

ANOVA F2: No. flies collected % females Treatment 39.8** 0.62ns Day 13.77** 7.69** Trt X Day 2.43* 2.94* _______________________________________________________ 1 Means within rows followed by the same letter are not significantly different (Tukey's means separation method). 2 ** P<0.01, * P<0.05, ns P>0.05. Trial 4: Large pail traps Large pails baited with Farnam attractant collected the greatest number of flies (>14,000 flies/trap) (Table 5). These traps collected significantly more flies than any of the other traps with the exception of jug traps baited with the same attractant (6,293 flies/trap). In contrast, QuikStrike baited pails collected negligible numbers of flies, most of which were males. These results were surprising considering the strong response of flies to the QuikStrike strips in Trial 1. It is possible that the effectiveness of the strips includes a visual element or depends on flies visiting the strips in a relatively random fashion. It may also be that the attractant in the strips is effective in short-range exposed settings such as the sheltered QuikStrike trap but was not sufficiently strong to induce flies to enter the pails through the relatively small openings in the trap lids. Fly collections in the hybrid trap that used Farnam bait as an attractant and QuikStrike strips as the killing agent collected substantially fewer flies (949 flies/trap/day) than either pail or jug traps baited with the same attractant. Flies were observed entering these traps at a rate similar to the Farnam-attractant-baited pail trap, however, many flies were also observed leaving the hybrid traps through the entry port. Because the bait in the hybrid trap was covered with a screened lid, the flies could not fall into the liquid. Flies that entered the trap could either leave, die of dehydration/starvation, or succumb to niathiazine poisoning after feeding on the bait strip. The results suggest that a large proportion of the flies

Table 5. Evaluation of 5-gallon traps baited with Farnam (Terminator) attractant, QuikStrike bait strip, and a hybrid trap using Farnam bait as the attractant (in screen-topped container) and QuikStrike (QS) strips as the killing agent. Standard Terminator and Victor jug traps included for comparison. Tests run for single days only (bait not "recycled"). Trap type Mean (SE) no. house flies collected1 27.5 (4.6)c 14,214.2 (6,327.4)a 948.6 (210.6)bc 6,269.3 (1,142.7)ab 5,219.0 (1,183.1)b % (SE) female

QuikStrike-baited 5-gallon-pail trap Farnam-baited 5-gallon-pail trap QS/Farnam hybrid 5-gallon-pail Terminator jug trap Victor jug trap

33.7 (3.6)c 68.5 (7.6)ab 89.3 (2.0)a 65.5 (3.1)b 69.3 (3.5)ab

ANOVA F2 8.03** 20.30** 1 Means within columns followed by the same letter are not significantly different (Tukey's means separation method). 2 ** P<0.01, * P<0.05, ns P>0.05.

Vol. 30, no. 2

Journal of Vector Ecology

249

Table 6. Comparison of molasses (25% diluted blackstrap), Farnam attractant, and molasses plus Farnam attractant. Flies were removed from fluid daily for 4 days after initial trap placement. Day Molasses Farrnam attractant Molasses + Farnam attractant

1 2 3 4

Mean no. house flies/trap1 6,251 (904)b 9,835 (1,103)a 8,588 (850)ab 4,407 (490)a 6,946 (1,407)a 7,021 (1,204)a 16,417 (4540)a 19,378 (4,726)a 16,053 (3,382)a 8,895 (1610)a 13,778 (2,530)a 10,347 (1,745)a % Female flies 83.0 (1.4)a 72.6 (6.7)a 81.4 (1.8)a 76.4 (4.5)a

1 2 3 4

62.0 (5.2)b 41.0 (3.7)b 68.2 (3.8)b 44.2 (6.0)b ANOVA F2: Treatment Day Trt X Day

68.4 (5.2)ab 70.7 (4.7)a 69.1 (2.6)b 56.7 (5.7)b

No. flies collected 2.02 ns 11.61 ** 0.16 ns

% females 29.04 ** 6.88 ** 2.39 *

1

Means within rows followed by the same letter are not significantly different (Tukey's means separation method). 2 ** P<0.01, * P<0.05, ns P>0.05.

available. The sheltered QuikStrike trap is a highly promising tool for fly surveillance as it is very effective, easy to service, and preserves flies in a dry and odorless condition for ease of counting. This method may have particular utility in pathogen surveillance programs. Acknowledgments The author thanks H. McKeithen and H. Brown for assisting with field work and fly counts. REFERENCES CITED Agui, N. 2001. Flies carrying enterohemorrhagic Escherichia coli (EHEC) O157 in Japan: A nationwide survey. Med. Entomol. Zool. 52: 97-103. Beck, A.F. and E.C. Turner, Jr. 1985. A comparison of five house-fly (Diptera: Muscidae) population monitoring techniques. J. Med. Entomol. 22: 346-348. Binkley, W.W. and M.L. Wolfrom. 1953. Composition of cane juice and cane final molasses. Adv. Carbohyd. Chem. 3: 1-24. Broce, A.B. 1988. An improved alsynite trap for stable flies, Stomoxys calcitrans (Diptera: Muscidae). J. Med. Entomol. 25: 406-409. Buma, R., H. Sanada, T. Maeda, M. Kamei, and H. Kourai. 1999. Isolation and characterization of pathogenic

bacteria, including Escherichia coli O157: H7, from flies collected at a dairy farm field. Med. Entomol. Zool. 50: 313-321. Burg, J.G. and R.C. Axtell. 1984. Monitoring house fly, Musca domestica (Diptera: Muscidae), populations in cagedlayer poultry houses using a baited jug-trap. Environ. Entomol. 13: 1083-1090. Cohen, D., M. Green, C. Block, R. Slepon, R. Ambar, S.S. Wasserman, and M.M. Levine. 1991. Reduction of transmission of shigellosis by control of houseflies (Musca domestica ). Lancet 337: 993-997. Geden, C.J., J.A. Hogsette, and R.D. Jacobs. 1999. Effect of airflow on house fly (Diptera: Muscidae) distribution in poultry houses. J. Econ. Entomol. 92: 416-420. Goulson, D., W.O.H. Hughes, and J.W. Chapman. 1999. Fly populations associated with landfill and composting sites used for household refuse disposal. Bull. Entomol. Res. 89: 493-498. Graczyk, T. K., R. Knight, R.H. Gilman, and M.R. Cranfield. 2001. The role of non-biting flies in the epidemiology of human infectious diseases. Microb. Infect. 3: 231-235. Hogsette, J.A., R.D. Jacobs, and R.W. Miller. 1993. The sticky card: Device for studying the distribution of adult house fly (Diptera: Muscidae) populations in closed poultry houses. J. Econ. Entomol. 86: 450-454. Howard, L.O. 1911. The house fly - disease carrier. Frederick A. Stokes, NY. Kobayashi, M., T. Sasaki, N. Saito, K. Tamura, K. Suzuki, H. Watanabe, and N. Agui. 1999. Houseflies: not simple mechanical vectors of enterohemorrhagic Escherichia coli O157:H7. Am. J. Trop. Med. Hyg. 61: 625-629. Lysyk, T.J. and R.C. Axtell. 1985. Comparison of baited jugtrap and spot cards for sampling house fly, Musca domestica (Diptera: Muscidae), populations in poultry houses. Environ. Entomol. 14: 815-819. Lysyk, T.J. and R.C. Axtell. 1986. Field evaluation of three methods for monitoring populations of house flies (Musca domestica ) (Diptera: Muscidae) and other filth flies in three types of poultry housing systems. J. Econ. Entomol. 79: 144-151. Meek, F. 2001. Study finds new fly pathogens. Pest Notes 3: 1-3. Mian, L.S., H. Maag, and J.V. Tacal. 2002. Isolation of Salmonella from muscoid flies at commercial animal establishments in San Bernardino County, California. J. Vector Ecol. 27: 82-85. Moriya, K., T. Fujibayashi, T. Yoshihara, A. Matsuda, N. Sumi, N. Umezaki, H. Kurahashi, N. Agui, A. Wada, and H. Watanabe. 1999. Verotoxin-producing Escherichia coli O157:H7 carried by the housefly in Japan. Med. Vet. Entomol. 13: 214-216. Murvosh, T.M. and C.W. Thaggard. 1966. Ecological studies of the house fly. Ann. Entomol. Soc. Am. 59: 533-547. Nayduch, D. and F. Stutzenberger. 2001. The housefly (Musca domestica) as a vector for emerging bacterial enteropathogens. Rec. Res. Devel. Microbiol. 5: 205-209. Olsen, A.R. and T.S. Hammack. 2000. Isolation of Salmonella spp. from the housefly, Musca domestica L., and the dump

250

Journal of Vector Ecology

December 2005

fly, Hydrotaea aenescens (Wiedemann) (Diptera: Muscidae), at caged-layer houses. J. Food Prot. 63: 958960. SAS Institute. 1992. SAS users guide: statistics. SAS Institute, Cary, NC. Scudder, H.I. 1996. Use of the fly grill for assessment of house fly populations: An example of sampling techniques that create rough fuzzy sets. J. Vector Ecol. 21:167-172.

Stafford, K.C., III, C.H. Collison, and J.G. Burg. 1988. House fly (Diptera: Muscidae) monitoring method comparisons and seasonal trends in environmentally controlled highrise, caged-layer poultry houses. J. Econ. Entomol. 81: 1426-1430. Sulaiman, S., M.Z. Othman, and A.H. Aziz. 2000. Isolations of enteric pathogens from synanthropic flies trapped in downtown Kuala Lumpur. J. Vector Ecol. 25: 90-93.

Vol. 30, no. 2

Journal of Vector Ecology

251

Bionomics and distribution of species of Hystrichopsylla in Arizona and New Mexico, with a description of Hystrichopsylla dippiei obliqua, n. ssp. (Siphonaptera: Hystrichopsyllidae)

Michael W. Hastriter1 and Glenn E. Haas2

1

Monte L. Bean Life Science Museum, Brigham Young University, 290 MLBM, P.O. Box 20200, Provo, UT 84602-0200, U.S.A. 2 557 California Avenue, PMB O7, Boulder City, NV 89005, U.S.A. Received 3 February 2005; Accepted 12 April 2005

ABSTRACT: More than 450 specimens of Hystrichopsylla were collected from nests and hosts of species of Microtus, Neotoma, Tamiasciurus, and Peromyscus in Arizona and New Mexico from 1981-2004. A new subspecies, Hystrichopsylla dippiei obliqua, is described and a map illustrating the distribution of the three taxa (Hystrichopsylla dippiei truncata Holland, H. d. obliqua, and H. occidentalis sylvaticus Campos and Stark) occurring in Arizona and New Mexico is provided. Hystrichopsylla. o. sylvaticus is reported in New Mexico for the first time and H. d. truncata is a new record in Rio Aribba County, NM. Relationships of Mexican species are also discussed. These large fleas are seldom collected from the fur of their mammalian hosts (usually singly) but are prevalent in moist nests. The maximum number of the new subspecies collected from a single nest was 54. Dry nests, particularly those that are not subterranean or otherwise protected from desiccation, do not support development of Hystrichopsylla fleas. Minimum elevation at which H. dippiei ssp. is found in Arizona and New Mexico appears to be about 2,100 m. Journal of Vector Ecology 30 (2): 251-262. 2005. Keyword Index: Hystrichopsylla dippiei, Hystrichopsylla occidentalis, Siphonaptera, flea.

INTRODUCTION Lewis and Lewis (1994) discuss the two nominate species, H. dippiei dippiei Rothschild, 1902 and H. occidentalis occidentalis Holland, 1949, and their subspecific taxa in the western United States. Specimens of H. dippiei ssp. were recorded from New Mexico by Traub and Hoff (1951) and Ford et al. (2004) and from Arizona by Beer et al. (1959), Holland (1957), and Hubbard (1947). A male reported from Arizona by the latter author was further identified by Hopkins and Rothschild (1962) as H. dippiei truncata Holland, 1957. Holland (1957) and Haas et al. (1973) reported H. d. truncata from New Mexico, and Campos and Stark (1979) reported H. o. sylvaticus Campos and Stark, 1979 from Arizona. Fagerlund et al. (2001) reported H. o. linsdalei Holland, 1957 from New Mexico, which is extralimital and likely represented H. o. sylvaticus. The present study is one of several (Kucera and Haas 1992, Haas and Wilson 1998, Lewis and Haas 2001, Haas and Kucera 2004, and Haas et al. 2004) related to an expansion within the Southwest of an earlier (1969-1971) survey of fleas concentrated in and close to the Jemez Mountains, NM (Haas et al. 1972, 1973, Haas 1972, 1973, Méndez and Haas 1972, 1973). Lewis and Eckerlin (2004) described a new species from Guatemala and provided a key for the known Mexican and Central American species of Hystrichopsylla. During the 1980s and 1990s, the emphasis shifted from trapping rodent hosts for their fleas to extracting and rearing fleas from the nests of chipmunks [Tamias (Neotamias) spp.], red squirrels [Tamiasciurus hudsonicus Erxleben)], mice (Peromyscus spp.), woodrats (Neotoma

spp.), and voles (Microtus spp.). The data from these extractions and rearings are reported in this paper. Some findings are presented for hystrichopsyllid material collected from host animals, and relevant discussions of Mexican species are included. MATERIALS AND METHODS The junior author collected most of the fleas in this study from Arizona and New Mexico during the months of May through November 1981-2004. Most recent collections (2002 and 2004) were provided by James R. Kucera. Specimens from collections of other institutions were also examined and included herein. Table 1 illustrates the number of nests examined and the number of each sex recovered from the specified host's nest. Although more nests were collected, the rodent species was uncertain. Such nests and the numbers of fleas recovered are not included in Table 1, but specimens were examined. Figure 1 provides distribution records for H. dippiei ssp. in Arizona and New Mexico and H. o. sylvaticus is included to illustrate the spatial relationships of all three taxa. Most of our flea specimens came from mixed coniferous montane forests and meadows therein. The dominant associated members of these plant communities occurring in our study areas are described in the biotic community classification system of Brown et al. (1998). Four of these areas described by Brown et al. characterize the habitats from which collections were obtained. They are described below and may be associated with each collection made by either

252

Journal of Vector Ecology

December 2005

the junior author or James R. Kucera. These (habitats 1-4 below) are annotated within parentheses immediately after the collector in the "Materials Examined" sections, i.e. "...T. hudsonicus nest, 30 September 1998, G.E. Haas (1), 4 males, 14 female, (RN-143);...". Habitat 1: Rocky Mountain and Great Basin subalpine conifer forest (Brown et al., plate 13), with dominant stands of Englemann spruce (Picea engelmanni Parry) and the alpine fir [Abies lasiocarpa (Hooker)]. Higher north facing slopes and narrow east-west facing canyons were particularly important habitats for red squirrels. Non-coniferous trees frequently used by rodents for nesting included the quaking aspen (Populus tremuloides Michaux). Habitat 2: Rocky Mountain alpine and subalpine grassland (plate 53), with expanses of Thurber fescue (Festuca thurberi Vasey) and other forbs. Such habitats provided ample niches for Microtus montanus (Peale). Habitat 3: Rocky Mountain montane grassland and its conifer associations (plate 59), with meadows of herbaceous forbes were favored habitats for the long-tailed vole [Microtus longicaudus (Merriam)]. Associated communities of Douglas fir [Pseudotsuga menziesii (Mirbel)] and white fir [Abies concolor (Gordon & Glendinning)] generally surrounded these meadows and were commonly associated (unlisted by Brown et al.) with the Blue spruce (Picea pungens Engelmann), a favorite of the red squirrel in New Mexico (Findley et al. 1975). The Limber pine (Pinus flexilis James), although present, was less common. Ponderosa pine (Pinus ponderosa Douglas) dominated in the driest sites having prolonged exposure to the sun (see below). Habitat 4: Rocky Mountain montane conifer forest (plate 22), which often has nearly pure stands of ponderosa pine/ yellow pine (P. ponderosa). These sites are drier and at lower elevations than the three previous communities and provide nesting habitats for Abert's squirrel (Sciurus aberti Woodhouse) and the Mexican vole [Microtus mexicanus (Saussure)]. A description of this biotic community is included to illustrate a habitat with favorable hosts on which few specimens of Hystrichopsylla spp. were collected. Some common non-conifer associates of ponderosa pines listed by Brown et al. (1998) included the Gambel oak (Quercus gambelii Nuttall) found in our northern study areas, Arizona white oak (Quercus arizona Sarg.) in the southern study areas, and the widespread quaking aspen. Little (1950) tabulated the annual rainfall for 12 vegetation types. Two of these include "Douglas fir-mixed conifer-aspen forests" (maximum of 635-732 mm) and "spruce-fir forests" (maximum of 762-889 mm). In general, our habitats 1-3 above receive an annual rainfall in excess of 600 mm (mostly as winter snows). Habitats at lower elevations that do not support these plant communities (habitat 4) receive less annual moisture and are not suitable for species of Hystrichopsylla.

Habitats of the red squirrel and other selected rodents were searched for likely nest sites. These sites were scrutinized for rodent signs, e.g., middens, droppings, and runways. Calls and sightings confirmed the presence of certain species. Helpful guides to distributions, locality records, food habits, behavior, and nest ecology included the monographs of Bailey (1932), Hoffmeister (1956, 1986), Findley (1987), Findley et al. (1975), and Brown (1984). Each nest was removed as intact as possible from its site, inserted into an empty plastic bag, labeled, sealed, and taken to a place where it could be examined immediately or stored temporarily in a sheltered place out of direct sunlight. Each was meticulously examined for adult fleas and larvae over a light-colored pan. Adults and some of the larvae were preserved in ethanol, while most of the larvae were returned to the same nesting material (coarse material discarded) for rearing additional adults. These rearing bags were left open and inserted into larger bags containing moistened paper towels. The outer bags were sealed and stored in a dark place at room temperature (18-24 °C). The inner bags were re-examined for newly-emerged adult fleas at 1-4 week intervals. Teneral fleas were captured and moist paper towels were renewed periodically until no more fleas appeared (usually within 2 months). Trapped animals were obtained using Sherman ® aluminum collapsible traps baited with oatmeal and snap traps baited with peanut butter. Trapping was a minor part of this study (only 12 hosts opposed to 85 nests) and no trapping was done in New Mexico. It was necessary to trap in Arizona because T. hudsonicus does not occur in the Chiricahua Mountains, Cochise County, AZ (Hoffmeister 1986) and nests of other rodents could not be found at high elevations. The endangered status of the endemic Mt. Graham red squirrel [T. h. grahamensis (J.A. Allen)] within the Pinaleno Mountains, Graham County, AZ precluded trapping or nest collections (see Istock and Hoffman 1995). However, 35 fleainfested nests of the long-tailed vole (M. longicaudus) were collected in Arizona, as the vole and its nests were abundant. Each animal was brushed over a light-colored pan and fleas were preserved in 70% ethanol. Collections included numerous fleas from lower elevations in the Sonoran Desert to higher elevations of the alpine valleys and slopes. Although many species of fleas were obtained, this report addresses only species of Hystrichopsylla. They were found only at elevations above 2,164 m. Voles were predominantly found in wet areas along streams and in meadows, woodrats in rocky ecotones at higher elevations, Abert's squirrels in ponderosa forests (lower elevations), and red squirrels in mixed coniferous forests, particularly densely wooded north facing slopes and shaded ravines. In addition to specimens collected during this study, numerous specimens were examined from different institutions. A listing of institutions or individual flea collections designating repository of specimens with associated acronyms follow. Those parenthetical acronyms that are deposited in the respective institution or collection are listed after those specimens in the sections entitled "Materials Examined".

Vol. 30, no. 2

Journal of Vector Ecology

253

BYU--Brigham Young University Flea Collection, Monte L. Bean Life Science Museum, Provo, UT; CMNH--Section of Invertebrate Zoology, Carnegie Museum of Natural History, Pittsburgh, PA; CNC--Canadian National Collections of Insects, Arachnids and Nematodes, Ottawa, Ontario, Canada; GEH--Personal collection of Glenn E. Haas, Boulder City, NV; JRK--Personal collection of James R. Kucera, Salt Lake City, UT; MWH--Personal collection of Michael W. Hastriter, Monte L. Bean Life Science Museum, Provo, UT; USNM--Division of Systematic Biology-Entomology, National Museum of Natural History, Smithsonian Institution, Washington, D.C.; and UNM--Museum of Southwestern Biology, University of New Mexico, Albuquerque, NM. Adults were mounted using conventional flea mounting techniques. Because of their large size, it was often necessary to shim the cover glass with glass shards to hold it parallel to the slide. Dissections of males were ultimately necessary for more detailed studies. Since all male specimens were mounted years before these final studies, slide mounted specimens were selected and placed in petri dishes in xylene until the Canada balsam media was dissolved sufficiently to free the specimens (usually overnight). Dissections were conducted in xylene by removing the right terga and sterna IX (t. IX, st. IX) and the aedeagus with minuten nadeln mounted at the tip of applicator sticks. Because of their extremely small size, these anatomical parts were picked up under one dissecting scope with a minute spatula and transferred into a small drop of Canada balsam on a microscope slide under an adjacent dissecting microscope. The remaining whole flea was mounted adjacent to the dissected parts under a second coverslip to facilitate associated morphology. An Olympus BX61compound microscope, Olympus SZX-12 stereo microscope, Olympus CV12 digital camera, Olympus MicrosuiteTM B3SV program, and Adobe Photoshop 7.0 were used to prepare images. Mammal taxa and their synonymies follow those of Wilson and Reeder (1993) and are specified in the text where synonymies occur. RESULTS AND DISCUSSION Hystrichopsylla dippiei obliqua new subspecies, Hastriter and Haas Figure 1, Figure 2A-C, Figures 3A-G, Table 1 Type Material.---United States. New Mexico: Catron County. Mogollon Mountains: Snow Canyon, 4.3 km NW Snow Lake Dam, 3.3 km W up canyon from tank, 2,347 m, (33°26'30"N 108°31'39"W), T. hudsonicus nest, 30 September 1998, G.E. Haas (3), male holotype, female allotype, 4 males, 14 females (RN-143); same data except 4 km NW Snow Lake Dam, 2,341 m, (33°26'32"N 108°31'40"W), 11 June 1998, 1 female (RN-139); S rim Snow Canyon, 4 km NW Snow Lake Dam, 2,408 m, (33°26'27"N 108°31'40"W), Neotoma mexicana Baird, 28 September 1996, G.E. Haas (3), 1 male (WN-46); 21.3 km E Mogollon, Gilita Creek, 2,393 m, (33°24'46"N 108°34'10"W), N. mexicana + Peromyscus maniculatus (Wagner) nest, 8

September 1995, G.E. Haas (3), 1 male (WN-44 + P-14); 18.5 km E Mogollon, Willow Creek, 2,444m (33°23'48"N 108°35'35"W), N. mexicana nest + P. maniculatus nest, 23 September 1991, G.E. Haas (3), 1 female (WN-35 + P-10); 19.3 km ENE Mogollon, Quaking Aspen Creek, Bear Wallow Lookout Road, FR 119, 2,524 m, (33°27'24"N 108°36'18"W), T. hudsonicus nest, 1 October 1998, G.E. Haas (3), 1 female (RN-144); 16.9 km NE Mogollon, SE slope of Corner Mountain, Bill Lewis Cienega, 2,719 m, (33°27'34"N 108°56'37"W), T. hudsonicus nest, 29 September 1996, G.E. Haas (3), 1 female (RN-134); Tularosa Mountains: 10.1 km SE Apache Creek, Five Springs Canyon between Springs and head of canyon, 2,621 m, (33°15'N 108°29'14"W), N. mexicana nest, 3 October 1999, G.E. Haas (3), 2 females (WN50/51); 9.1 km SE Apache Creek, Five Springs Canyon, 0.3 km W Five Springs tank, 2,195 m, (33°15'N 108°30'31"W), T. hudsonicus nest, 27 September 1996, G.E. Haas (3), 1 male, 3 females (RN-132). Grant County. Mimbres Mountains: 16.6 km NE San Lorenzo post office, Iron Creek Canyon, Iron Creek Campground, 2,195 m, (32°54'30"N 107°48'23"W), M. mexicanus , nest, 9 September 1992, G.E. Haas, 1 male, 1 female (V-24). Socorro County. San Mateo Mountains: 1.7 km, WNW Mt. Withington, Bear Trap Canyon, 2,658 m, (33°53'06"N 107°30'11"W), T. hudsonicus nest, 22 August 1993, G.E. Haas (1), 4 males, 1 female (RN-98); same data, 2 males (14 males, 12 females in alcohol)(RN97). Torrance County. Manzano Mountains: 8.1 km SW Manzano, Ox Canyon, Forest Trail 190, 2,499 m, (34°36'27"N 106°25'06"W), T. hudsonicus nest, 9 August 1991, G.E. Haas (3), 13 males, 11 females (RN-70) (GEH). The holotype, allotype and six paratypes (3 males, 3 females) are deposited in the National Museum of Natural History, Washington, D.C., five paratypes (2 males, 3 females) in the Carnegie Museum of Natural History, Pittsburgh, PA, four paratypes (2 males, 2 females) in the junior author's collection, 6 paratypes (3 males, 3 females) in the senior author's collection and the remaining paratypes in the Brigham Young University collection, Monte L. Bean Life Science Museum, Provo, UT. Other material examined: Mexico. Nuevo León. Sierra Madre Oriental Range: Cerro Potosi, 14 km N Galeana, 2,896 m, (~24°57'N 100°05'W), Microtus sp., 18 May 1963, A.D. Stock and J.H. Shaw, 1 male; 18 km N Galeana, 2,896 m, (~24°59'N 100°05'W), Peromyscus melanotus J.A. Allen and Chapman, 1897, 16-17 February 1963, A.D. Stock and J.H. Shaw, 2 females; El Potosi, 2,896 m, (~24°51'N 100°19'W), Microtus sp., 21 May 1963, A.D. Stock and J.H. Shaw, 1 male, 3 females (CMNH). United States. Arizona: Graham County. Pinaleno Mountains: 4 km W Mt. Graham, 0.4 km S USFS Columbine Work Center, 2,877-2,894 m, (32°42'N 109°54'49"W), M. longicaudus nests (5), 8-11 September 1987, G.E. Haas (3), 11 males, 8 females; same data except 2,918 m, M. longicaudus + N. mexicana nests, 21 May 1990, 1 male, 2 females; M. longicaudus, 17 May 1990, 1 male; 2,915 m, N. mexicana nest, 18 May 1990, 2 males, 5 females; 1.1 km NW Heliograph Peak, Shannon Park, near Heliograph Peak Road, 2,774 m, (32°39'19"N 109°51'30"W), M. longicaudus

254

Journal of Vector Ecology

December 2005

nest, 18 November 1989, G.E. Haas (3), 1 male; same data except N. mexicana nest, 19 November 1989, 16 males; same data except 20 October 1991, 4 males, 3 females; same data except 2,714 m, 19 October 1991, 3 males, 17 females; (GEH); 0.3 km S Heliograph Peak, N. mexicana, 9 August 1951, R.G. van Gelder, 1 female (CMNH); 9 km W Mt. Graham, Riggs Flat near E shore of Riggs Lake, 2,652 m (32°42'30"N 109°57'45"W), M. longicaudus, 15 May 1990, G.E. Haas (3), 1 male 2 females; 3.6 km WSW Mt Graham, 1.2 km SSE Columbine Work Center, 2,896 m (32°41'58"N 109°54'46"W), M. longicaudus nest, 17 May 1990, G.E. Haas (3), 8 males, 3 females; 6.4 km WNW Mt. Graham, Chesley Flat, 2,816 m, (32°42'45"N 109°56'08"W), M. longicaudus nests (2), 12 September 1987, G.E. Haas (3), 13 males 6 females, (GEH). New Mexico: Catron County. Mogollon Mountains: Snow Canyon, 4.3 km NW Snow Lake Dam, 2,344 m, (33°26'31"N 108°31'38"W), T. hudsonicus nest, 11 June 1998, G.E. Haas (3), 1 female (GEH); and 32 km NE Mogollon, 2,409 m, (~33°35'N 108°28'W), Microtus sp., 7 November 1950, 1 female (CNC). Lincoln County. Sacramento Mountains: 11.3 km N Ruidoso, Sierra Blanca east foothills, (~33°25'N 105°41'W), Microtus pennsylvanicus (Ord), 4 December 1951, 1 male; 9.7 km N Ruidoso, Sierra Blanca east foothills, (~33°25'N 105°41'W), Neotoma albigula Hartley, 4 December 1951, 1 male (USNM). Socorro County. Magdalena Mountains: near S Baldy Peak at Langmuir Labs, host unlisted, 12 September 1990, 1 female (UNM). Diagnosis: Material from Arizona and New Mexico is represented by taxa belonging to both the H. dippiei and the H. occidentalis complexes. These may be distinguished by the number of spines in the genal ctenidium. Hystrichopsylla dippiei ssp. usually has six spines per side in the genal ctenidium (infrequently seven) (Figures 2A-B), while H. occidentalis ssp. usually have eight or more on each side. The H. dippiei subspecies include H. d. dippiei Rothschild, 1902 (north central U.S. and Alberta, Canada), H. d. neotomae Holland, 1957 (coastal central California), H. d. spinata Holland, 1949 (Vancouver, British Columbia, Canada through Alaska's southern archipelago), H. d. truncata (east of the Cascades and the Sierras from British Columbia, Canada to Arizona and New Mexico), and H. d. obliqua (Arizona, New Mexico, and Mexico). Males are separable from nominate subspecies by the presence of tubercles on the dorsal angle of st. IX. The oblique angle of the apex of the distal arm of the ninth sternum (DA9) is distinctive in the new subspecies opposed to a more truncate apex in H. d. spinata and H. d. truncata, and less oblique in H. d. neotomae. The latter species also has prominent tubercles versus the flattened scale-like tubercles found in H. d. obliqua. The new subspecies usually possess one or two small spiniform setae between the proximal and penultimate larger spiniforms along the caudal margin of the st. IX (Figures 3D-E). This character occurs in other taxa as well; however, 76/78 males of H. d. obliqua possess the small spiniform between the proximal and penultimate larger spiniforms versus only 7/42 males of H. d. truncata. The anterior margin of st. IX is essentially straight in the new subspecies, whereas it is strongly reflected forward in other

H. dippiei ssp. (Figures 3B, 3D-L). This forward inflection creates a concavity along the anterior margins. Females in the absence of males can not be distinguished for certain from H. d. truncata in Arizona and New Mexico; however, the margin of the t. VIII tends to have a marked concavity or flattened region opposite the dorsal most marginal seta, whereas it is rounded in H. d. truncata (Figure 3C). Females of the new subspecies also have a greater number of marginal spinelets on t. II (Figure 2B) than H. d. truncata (range: 612, mean: 8.3 (n=38) vs. range: 4-8, mean: 6 (n=38)]. In addition to H. d. obliqua, Mexico taxa possessing six genal ctenidia include H. orophila Barerra, 1952 (females unknown) and H. kris Traub and Johnson, 1952 (males unknown). Separable from the female of H. kris by its much smaller size and the abdominal spiracle VIII extends to the margin of tergum IX and from the male of H. orophila by the broadly rounded apical margin of the basimere, rounded apex on securifer of Ford's sclerite, and broader distal arm of sternum IX. Description: Reference to the numbers of setae, spinelets, spiniforms, etc. refer to one side only. General anatomy of females may be assumed to be the same as males unless otherwise specified. Head (Figures 2A-B). Preantennal area with an ocular row of three setae, central single long seta at margin of antennal fossa, frontal row of six slender setae, with numerous scattered small setae all posterior to frontal row. Submarginal area of frons (anterior to frontal row) finely punctate with two placoids near margin. Well-developed tentorial arch anterior to eye; area communis large. Small pigmented eye with ventral sinus. Area between frontal tubercle and oral angle with thick sclerotization. Falx of female complete; much longer than male. Post-antennal area with three rows of setae; ventral two setae in posterior row much longer than others. Male with occipital groove, both sexes with two placoids (anterior placoid of male located in antennal groove). Group of setae dorsal to antennal fossa. Antennal scape with basal patch of small setae (seven to ten), apex with three to four long setae; pedicel with fringe of short setae extending ~1/4 length of clavus (slightly longer in female). Male clavus extends onto prosternum. Genal ctenidium usually with six pigmented spines (rarely seven). Ventral margin of genal lobe extending ventrad over basal coxa in female; not expanded in male. Maxilla acutely pointed, labial palpus of five segments not reaching apex of coxa. Thorax (Figures 2A-B). Pronotal ctenidium of 38-40 spines; two rows of setae. Prosternum and proepimeron divided. Mesonotum with main row of setae, five anterior rows. Mesopleural rod bifurcate dorsally; mesopleuron with numerous scattered small setae anteriorly, larger setae caudally. Metanotum with main row, three anterior rows. Lateral pleural area divided by sclerotized ridge, two setae in each area. Metepisternum with single large seta, metasternum without setae. Metepimeron with vertical line of six setae in main row; ten scattered setae anteriorly in male, 15 in female. Spiracular fossa long and pointed. Legs. In general, legs very setiferous. Proximal ventral margin of each coxa with two large setae. Two femoral pit setae each leg. Setae guarding femoral-tibial joint: foreleg with large lateral, small medial

Vol. 30, no. 2

Journal of Vector Ecology

255

setae; mid- and hind leg, each with small lateral, large medial setae. Dorsal margin of tibia with eight notches in fore leg; mid- and hind legs each with nine notches. Five lateral plantar bristles on each distotarsomere. Unmodified Abdominal Segments (Figures 2A-B). Marginal spinelets on t. II-IV of female (average: seven, four and two, respectively in Pinaleno specimens, n=30; eight, four, and three in other areas, n=38) and male t. II-IV (average: seven, four, and three, respectively in Pinaleno specimens, n=40; eight, four, and three in other areas, n=24). Three specimens in the Pinalenos had a single spinelet on t. V. Tergites each with three rows of setae, main row in females strongly curved cephalad on t. IV-VI. Spiracular fossae round; rows of setae extend far below level of spiracles, especially in females. Three antesensilial bristles, middle longest, mesal shortest. Sternum II with single seta per side with several smaller setae cephalad; female with expansion or lobe on ventro-anterior angle of st. II. Male st. III with four setae in main row, st. IV-VI each with three, st. VII each with three-four, patch of smaller setae anterior to main rows. Female same except five-six setae in main row st. III-VI; with many smaller setae in front of main rows. Modified Abdominal Segments, male. Tergum VIII reduced; patch of five to seven setae dorsad to terminal spiracle. Basimere and telomere as in Figure 3B. Patch of setae along caudal base of manubrium well below acetabulum. Acetabular bristles missing. Sternum VIII with fringe of many long, posteriorly directed, slender setae on posterior half; anterior portion with many very short setae directed towards base of sternite. Shape of apex subject to orientation. Sternum IX adorned with six-seven pairs of spiniform setae along caudal margin. Smaller supernumerary setae are interspersed particularly between the ventral most pair and the penultimate pair. Small patch of fine setae below spiniforms; upper third of anterior margin st. IX with seven-eight caudally directed fine setae. Diagonal lateral line from base to apex; generally very straight in this species (Figures 3B, D-G). Numerous scale-like tubercles on the dorsal swelling where distal and proximal arms merge. Apodemal rod of st. IX absent (Figure 3B). Aedeagus. Aedeagus as in Figure 2C. Ford's sclerite (F.sc) well developed. Median dorsal lobe surrounding F.sc as extended hyaline margin. Median dorsal lobe flanked by accessory lateral lobes that extend to middle of aedeagal apodeme. Lateral lobes enveloping sclerotized inner tube (s.i.t.); dorsal armature (d.ar.) attaching s.i.t. to F.sc. by a hyaline membrane. Aedeagal pouch (ae.p.) thin, membranous on distal portion; well-sclerotized caudally. Penis rods extend just beyond apex of aedeagal apodeme. Elliptical patch of minute setae beneath aedeagal pouch on ventro-lateral surface near proximal portion of virga ventralis (Figure 2C). Modified Abdominal Segments, female (Figure 3C). Sensilium with heavily sclerotized posterior margin. Sternum VII emarginate with ventral lobe; numerous small setae anterior to main row of eight long setae. Sternum VIII extends dorsad to terminal spiracle; bearing many setae, five-six marginal. Margin of st. VIII variable; most specimens with flattened, or concave margin opposite uppermost long marginal seta. Group of sixeight variable sized setae on mesal surface of st. VIII. Side of anal stylet parallel, three times as long as wide; single long

apical bristle with minute seta on dorsum at base of bristle. Ventral anal lobe setiferous with two long undulating apical setae. Two spermathecae, bulga slightly broader at base than near hilla (Figure 3C). Bursa copulatrix membranous and indiscernible. Remarks.--Specimens from the Pinaleno Mountains represent the most extreme southwestern distribution of this species and are isolated from the more contiguous mountain ranges to the north and east. Hoffmeister (1956) also considered the vole population in the Pinalenos as a subspecies (M. longicaudus leucophaeus) distinct from other Arizona and New Mexico species of M. longicaudus. Evolution of this flea subspecies may follow that of its host; however, definitive studies (DNA or karyotype) have not been conducted to validate Hoffmeister's subspecific designation in support of this proposal. Although we place the Pinaleno population with H. d. obliqua, note that the oblique apex of DA9 is more variable than material from the Mogollon, Tularosa, Mimbres, San Mateo, and Manzano Mountains of New Mexico, and the mountains of Nuevo León, Mexico where this feature is constant. In general, the two arms of DA9 of all members of this genus are fused from base to apex, forming a trough between the two distal arms. There are usually six or seven pairs of dark spiniform setae along the ventral margin, the ventral two pairs separated typically by more space than is found in that of H. d. truncata. This space has one or two short spiniform setae in 76 of the 78 males examined. The two specimens that lack this spiniform are considered intergrades and they occur in the Pinaleno (Figures 3F-G) and San Mateo Mountains. Conversely, seven of 42 males of H. d. truncata also possessed this spiniform setal pattern (interestingly in the zone of intergradation of specimens just north and south of New Mexico Interstate Highway 40 east of the Rio Grande in Bernalillo County). One of these also occurred in the northern portion of the White Mountains west of the Blue Range Primitive Area, five in the Jemez Mountains, and one in the Sandia Mountains (Figure 3J, paratype). The zone of intergradation of H. d. obliqua and H. d. truncata appears to occur in a region just north and south of U.S. Interstate 40, and to the southwest. Habitats in New Mexico regions east of 105°W latitude are too dry to support populations of either H. dippiei ssp. or H. occidentalis ssp. Several records were provisionally reported in New Mexico as H. d. truncata by Holland (1957). "Mogollon Mountains" was indicated on one such slide of a female that was examined [referenced in Holland (1957) as Catron Co., 7900', 2 males, 1 female] and is within the distribution of the new subspecies. His records listed from San Miguel County (1 male, 2 females examined) and those "25 miles E Albuquerque" (Bernalillo County, 1 male, 2 females examined) belong to H. d. truncata. The latter three specimens are labeled as "paratypes" but are not listed as paratypes in Holland's (1957) description of H. d. truncata. The male recorded from Pinos Altos, Grant County, NM, was not examined, but surely belongs to the new subspecies, since it occurs in the southern part of the Mogollon Volcanic Plateau (Ungnade 1972). Only three (1 male, 2 females) specimens of H. d. obliqua

256

Journal of Vector Ecology

December 2005

were collected from trapped hosts, i.e., three of seven M. longicaudus (Pinaleno populations considered the subspecies leucophaeus by early workers) in the Pinaleno Mountains during spring and fall. The yield of Hystrichopsylla on P. maniculatus was negative. The infestation of voles but not mice may be attributed to higher populations of fleas in vole nests than those in mouse nests. Voles constructed their nests on and in moist soil, but mice preferred drier sites whose microhabitats were unsuitable for development of this flea. Ford et al. (2004) cite records of Hystrichopsylla dippiei from a dozen hosts in four New Mexico counties (Bernalillo, Sandoval, Santa Fe, and Socorro). Based on their distribution, those from the latter county are likely H. d. obliqua, those from Sandoval and Santa Fe H. d. truncata, and those from Bernalillo could belong to either subspecies. Tipton and Méndez (1968) discuss 36 (9 males, 27 females) specimens of Hystrichopsylla from Cerro Potosi, Nuevo León, Mexico, indicating that they were to be described by Traub and Barrera. This description was never published. Seven of these specimens (2 males, 5 females) from the Traub collection (CMNH) were examined and are conspecific with this new subspecies. A single female (not examined) with the same data as one of the CMNH specimens is listed by AyalaBarajas et al. (1988) as "Hystrichopsylla dippiei Rothschild, 1902". We were unable to locate the other 28 specimens. The presence of H. d. obliqua so far south in Mexico is puzzling, since there are no linking mountain ranges for 1,200 km that have sufficient elevation necessary for the ecology of this flea. Perhaps this species will be discovered in the Sierra Madre Occidental Range to the west, which provides a geographical link between the Sierra Madre Oriental Range and the southern

montane regions of New Mexico. Additional collection of nests in alpine habitats of the Sierra Madre Occidental is needed to understand the association of Mexican and United States populations of this new subspecies. Note of clarification: Tipton and Méndez (1968) inadvertently indicated that Cerro Potosi was in the "Sierra Madre Occidental Range" in the first sentence of their introduction instead of the Sierra Madre Oriental Range. This is pointed out to preclude confusion for future workers. Holland (1957) suggested that H. kris was probably conspecific with H. orophila based on the premise that the spiracle of tergum VIII might be an anomaly. The authors examined a single female collected from Microtus sp. between Mexico City and Puebla, 3,140 m, leg. M.D. Tuttle (CMNH) and propose to dispel Holland's observation. The spiracle of this enormous flea (7 mm) did not reach the margin of tergum VIII and its position and shape was indistinguishable from that of the female holotype illustrated by Traub and Johnson (1952). Although the female of H. orophila has not been described, we examined a pair of Hystrichopsylla (both with six genal ctenidia) from the same host (Microtus sp.), 26 km SSW of Toluca, State of Mexico, leg. A.D. Stock and R. Traub (CMNH). The male is clearly H. orophila. The associated female is likely the same taxon, but a description of the single female would be premature. Future collections of the nests of Microtine and Sigmodontine rodents in subalpine habitats are needed to discover the male of H. kris and additional females of H. orophila to enable their description. Etymology: As with the sister taxon, H. d. truncata, the subspecific name is given for the characteristic shape of the apex of the st. IX, which is distinctly oblique as opposed to truncate.

Figure 1. Map illustrating distribution of Hystrichopsylla ssp. in Arizona and New Mexico [AP=Apache National Forest (NF), CB=Cibola NF, CON=Coconino NF, COR=Coronado NF, GI=Gila NF, LI=Lincoln NF, SF=Santa Fe NF, and TON=Tonto NF].

Vol. 30, no. 2

Journal of Vector Ecology

257

Hystrichopsylla dippiei truncata Holland, 1957 Figure 1, Figures 3H-L, Table 1 Type Material.---Hystrichopsylla dippiei truncata Holland, 1957. The Canadian Entomologist, 84: 318-319, Figure 23. Type Locality.--Canada, British Columbia, Okanagan Landing, ex Peromyscus maniculatus artemisiae [= P. maniculatus], 15 September 1950, J.D. Gregson (holotype No. 6547, Canadian National Collection of Insects, Ottawa). Hystrichopsylla dippiei truncata (see Lewis and Lewis 1994 for listing of synonymies) Material Examined: Arizona: Apache County. White Mountains: 5.5 km SSW Greer, NW Winn Campground, 2,737 m, (33°58'13"N 109°29'17"W), M. montanus, 28 September 1989, N. Wilson, 1 male; 2.2 km SE Big Lake, Big Lake Knoll near lookout on summit, 2,860 m, (33°51'43"N 109°23'27"W), T. hudsonicus nest, 13 September 1986, G.E. Haas (3), 1 male; 5.5 km SSW Greer, NW Winn Campground, 2,835 m, (33°58'09"N 109°28'59"W), T. hudsonicus nest, 18 August 1987, G.E. Haas (3), 1 male; 5.5 km SSW Greer,

Figure 2. A-C. Hystrichopsylla d. obliqua n. sp. A. Head, thorax and abdomen (holotype). B. Head, thorax and abdomen (allotype). C. Apex of aedeagus (paratype, Mogollon Mts., NM) (ae.p., aedeagal pouch; d.ar., dorsal armature; F.sc., Ford's sclerite; l.l., lateral lobe; m.d.l., median dorsal lobe; and s.i.t., sclerotized inner tube).

0.2 km NW Winn Campground, W side of Creek from Sam Hale Reservoir, 2,737 m, (33°58'12"N 109°29'18"W), N. mexicana nest, 25 September 1989, G.E. Haas (3), 1 female; 6 km SSW Greer, 0.4 km S Winn Campground, 2,853 m (33°57'43"N 109°24'06"W), T. hudsonicus nest, 27 September 1991, G.E. Haas (3), 2 males, 2 females; Escudilla Mt., 8.5 km ENE Alpine, Toolbox Draw near trail head #308, 2,941 m, (33°55'14"N 109°07'W), T. hudsonicus nest, 19 August 1993, G.E. Haas (3), 1 male; Lee Valley near NE shore Lee Valley Reservoir, 2,872 m, (33°56'35"N 109°29'49"W), M. longicaudus nest, 18 August 1987, G.E. Haas (3), 2 males, 4 females; 5.7 km SSW Greer, S entrance to Winn Campground, 2,842 m, (33°57'43"N 109°24'06"W), M. montanus nest, 26 September 1989, G.E. Haas (2), 1 male; 5.8 km SSW Greer, S Winn Campground, E Forest Road 554, 2,853 m, (33°57'50"N 109°29'05"W), M. montanus nest, 28 September 1989, G.E. Haas (2), 2 males; Alpine Divide (pass), 5.6 km NNW Alpine, 0.5 km NW Alpine Divide Campground, W highways 180/191, 2,595 m, (33°53'54"N 109°09'28"W), M. longicaudus nest, 25 September 1993, G.E. Haas (3), 1 male (GEH) Cochise County. Chiricahua Mountains: 0.8 km E Cima Ranger Station, 2,590 m, (~31°50'N 109°18'W), P. maniculatus or Peromyscus truei (Shufeldt), 30 July 1960, R.W. Thorington, 1 male, 1 female; Rustlers Park, (31°54'N 109°16'W), Peromyscus sp., 6 September 1960, H. Howden, 1 female (CMNH); 5.5 km WSW Paradise, rock slide S side of Barfoot Peak, 2,560 m, (31°55'08"N 109°16'31"W), Peromyscus boylii (Baird) (2 specimens), 21-22 October 1994, G.E. Haas (3), 1 male, 2 females; 6 km WSW Paradise, rock slide S side of Barfoot Peak, 2,525 m, (31°55'06"N 109°16'42"W), P. maniculatus, 25 November 1989, G.E. Haas (3), 1 female (GEH); Portal (Onion Gap to Rustlers Park), 2,323-2,576 m, (31°54'05"N 109°15'30"W to 31°56'N 109°16'20"W), P. boylii, August 1952, J. Beer, 1 male (UMINN). Coconino County. Kaibab Plateau: Marble Point, 2,725 m, (36°24'05"N 112°02'58"W), M. longicaudus, 3 September 1998, J.R. Kucera, 1 male, 1 female; Dog Lake, 2,688 m, (36°24'42"N 112°04'33"W), P. maniculatus, 5 June 1992, J.R. Kucera, 1 female; VT Lake, 2,686 m, (36°25'13"N 112°07'30"W), T. hudsonicus nest, 26 May 2002, J.R. Kucera, 1 male; along road between Watts and Quaking Aspen Springs, 2,380 m, (~36°22'42"N 112°16'14"W), M. longicaudus nest, 24 September 2004, J.R. Kucera, 1 female; Quaking Aspen Springs, 2,370 m, (~36°22'42"N 112°16'14"W), M. longicaudus nest (3 adult voles present), 24 September 2004, J.R. Kucera, 1 male; same except pool of two nests, 25 September 2004, J.R. Kucera, 1 male, 1 female; same except "rodent nest under log", 1 male (JRK); 24 km SSE Jacob Lake, VT Lake, E side highway 67, 2,646 m, (36°25'54"N 112°07'33"W), M. longicaudus nests (2), 4 September 1991, G.E. Haas (3), 2 male, 2 females; De Motte Park (in forest), 2,740 m, (36°24'30"N 112°08'30"W), T. hudsonicus nest, 19 August 1986, G.E. Haas (3), 1 female (GEH). Greenlee County. Blue Range: 0.6s km SSW Hannagan Meadow, Hannagan Meadow Campground, 2,788 m, (33°38'13"N 109°19'42"W), M. longicaudus nest, 8 May 1981, G.E. Haas (3), 1 male; same data except 2,784 m, 7.2 km SSW of Hannagan Meadow, NW KP Cienega, NW tributary of KP

258

Journal of Vector Ecology

December 2005

Figure 3. A-G. Hystrichopsylla d. obliqua n. sp. A. Distal arm of ninth sternum and eighth sternum. B. Distal arm of ninth sternum, basimere and telomere (paratype, San Mateo Mts., NM). C. Terminal segments of female (allotype). DG. Distal arm of ninth sternum (paratypes). D. Manzano Mts., NM. E. Mogollon Mts., NM. F-G. Pinaleno Mts., AZ. H-L. Hystrichopsylla d. truncata, distal arm of ninth sternum. H. Chiricahua Mts., AZ. I. White Mts., AZ. J. Sandia Mts., NM. K. Henry Mts., UT. L. Kaibab Plateau, AZ. Creek, 2,731 m, (33°34'45"N 109°21'22"W), M. longicaudus nest, 19 August 1987, G.E. Haas (3), 2 males, 1 female; 2 km SSW Hannagan Meadow, NW KP Cienega Campground, 2,755 m, (33°34'40"N 109°21'30"W), M. longicaudus nest, 25 May 1990, G.E. Haas (3), 2 males; 7.6 km SSW Hannagan Meadow near site # 3 at KP Cienega Campground, 2,739 m, (33°34'36"N 109°21'24"W), M. longicaudus nest, 11 September 1997, G.E. Haas (3), 2 males, 2 females (GEH). New Mexico: Bernalillo County. Sandia Mountains: 40 km E Albuquerque, P. maniculatus, 22 May 1949, C.C. Hoff and H.H. Lewis, 1 male (paratype), same data except Eutamias quadrivittatus = Tamias quadrivittatus (Say), 1 female (paratype), same data except Peromyscus nasutus (J.A. Allen), 23 April 1949, H.H. Lewis, 1 female (paratype) (CNC); 2.5 km ENE Sandia Crest, Ninemile Picnic Area, 2,810 m, (35°12'49"N 106°25'19"W), P. maniculatus nest, 10 August 1991, G.E. Haas (3), 1 female; same data except 12 September 1994, 2 males, 1 female (GEH). Los Alamos County. Jemez

Mountains: south cliff of Boulder Mt., 2,316 m, (35°54'27"N 106°19'45"W), M. longicaudus, 23 January 1970, G.E. Haas (3), 1 female; same data except P. truei (n=3), 23-24 January 1970, 2 males, 1 female; Limber Pine Line, W Boulder Mt., 2,347 m, (35°54'30"N 106°19'45"W), P. truei (n=2), 23-24 January 1970, G.E. Haas (3), 2 males, 1 female; same data except 16 May 1970, 1 female; overlook at highway 4 and Water Canyon, 2,426 m, (35°50'N 106°21'20"W), P. truei (n=2), 30 November 1969, G.E. Haas (3), 1 male, 3 females; Los Alamos Canyon, north side of Omega Road, 2,164 m, (35°16'30"N 106°21'42"W), P. truei, 26 April 1970, G.E. Haas (3), 1 male; and Pajarito Mt., north side ski slope, 3,016 m, (35°53'23"N 106°23'38"W), T. hudsonicus nest, 28 July 1970, G.E. Haas (1), 1 female (GEH). Rio Arriba County. San Pedro Mountains: 48.3 km E Cuba, (~36°03'N 106°26'W), M. pennsylvanicus, 20 September 1951, 1 male (CNC). Sandoval County. Jemez Mountains: N of junction highway 4 and St. Peters Dome Road, Forest Road 142, 2,737 m, (35°51'02"N 106°25'23"W), M. montanus, 16 June 1970, G.E. Haas (3), 1 female; same data except 21 July 1970, 1 male; W side Frijoles Canyon along St. Peters Dome Road, Forest Road 142 in rock fill, 2,699 m, (35°50'13"N 106°25'43"W), P. maniculatus, 13 May 1970, G.E. Haas (3), 1 male; SE junction highway 4 and St. Peters Dome Road, Forest Road 142, 2,725 m, (35°50'48"N 106°25'27"W), P. maniculatus, 7 May 1970, G.E. Haas (3), 1 female; Rockslide, Sierra de los Valles at SE corner of Valle Grande, 2,682 m, (35°51'49"N 106°25'43"W), Tamias minimus Bachman, 18 September 1970, G.E. Haas (3), 1 male, 1 female (GEH); M. montanus, 3, August 1971, G.E. Haas (3), 1 male, 2 females; same data except E. quadrivittatus = T. quadrivitatus, 4 August 1971, 1 male, 1 female; and same data except P. maniculatus, 5 August 1971, 1 male, 1 female (CNC). San Miguel County. Sangre de Cristo Mountains: 27 km N Pecos, 2,440 m, (~35°49'N 105°41'W), Spermophilus lateralis (Say), 28 July 1950, H. Stark, 1 male, 1 female; Holy Ghost Canyon, (~35°46'N 105°42'W), S. lateralis, 17 June 1953, H.B. Morlan, 1 female (CNC); 4.5 km SW Cowles, Holy Ghost Canyon, 2,542 m, (35°46'49"N 105°42'20"W), M. longicaudus nest, 26 August 1993, G.E. Haas (1), 1 female; 4.5 km SW Cowles, Holy Ghost Canyon, 2,524 m, (35°46'46"N 105°42'18"W), M. longicaudus nest, 26 August 1993, G.E. Haas (1), 1 female; 10.5 km NNW Pecos, Dalton Canyon, 2,262 m, (35°39'54"N 105°42'05"W), T. hudsonicus nest, 27 August 1993, G.E. Haas (3), 2 males (GEH); Grass Mt., Winsor Ranch, 2,743 m, (~35°49'N 105°38'W), S. lateralis, 9 August 1961, 1 male (USNM). Santa Fe County. Sangre de Cristo Mountains: Ski Basin, (~35°48'N 105°48'W), M. longicaudus, 5 November 1952, H.B. Morlan, 1 male; same data except 6 August 1953, 1 male; Pacheco Canyon, (~35°48'N 105°54'W), E. quadrivittatus = T. quadrivittatus, 30 August 1951, H.B. Morlan, 2 females; same data except Neotoma cinerea (Ord), 1 female; Hyde Park, (~35°45'N 105°49'W), Neotoma sp., 3 April 1951, H.B. Morlan, 1 female (CNC). Utah. Garfield County. Henry Mountains: 2,287 m (~38°07'N 110°45'W), "mouse nest", 20 July 1968, J.E.H. Martin, 1 male, 1 female; and 39 km S Hanksville, 2,287 m, (~38°07'N 110°45'W), Peromyscus sp.,

Vol. 30, no. 2

Table 1. Number of male and female Hystrichopsylla dippiei subspecies found in a specified number of nests of host mammals in Arizona and New Mexico. TOTAL

mexicanus Hystrichopsylla dippiei obliqua 23N/51 males, 109 females Hystrichopsylla dippiei truncata 15N/16 males, 25 females TOTAL 38N/67 males, 134 females 10N/52 males, 69 females 1N/1 male 9N/52 males, 68 females

Microtus montanus longicaudus

Neotoma mexicana

Tamiasciurus hudsonicus

1N/1 male, 1 female*

9N/38 males, 57 females

42N/142 males, 235 females

Journal of Vector Ecology

4N/5 males, 11 females

7N/6 males,6 females

27N/27 males, 43 females

1N/1 male, 1 female

4N/5 males, 11 females

16N/44 males, 63 females 69N/169 males, 278 fermales

*N=number of rodent nests examined/number of male fleas, number of female fleas.

259

260

Journal of Vector Ecology

December 2005

20 July 1968, J.E.H. Martin, 10 males, 8 females (CNC). San Juan County. Abajo Mountains: W Dalton Springs Campground, N North Creek Road, 2,316 m, (37°54'N 109°26'W), T. hudsonicus nest, 22 September 1996, G.E. Haas (3), 1 male, 1 female; La Sal Mountains: 20.5 km N La Sal, 0.12 km S Oowah Lake, 2,682 m, (38°29'56"N 109°16'23"W), T. hudsonicus nest, 15 June 1991, G.E. Haas and J.R. Kucera (3), 1 male, 1 female. (GEH). Remarks: Hystrichopsylla d. truncata is broadly distributed in the mountainous regions west of the great plains and east of the Cascade and Sierra Nevada ranges from British Columbia, Canada to the northern half of Arizona and New Mexico. This subspecies is replaced by H. d. obliqua to the south and H. d. neotomae to the extreme west (California). It is indiscriminate in its host selection, developing in the moist nests of chipmunks, mice, voles, red squirrels, and occasionally wood rats. Like others belonging to the H. dippiei complex, it is primarily a nest flea, collected only infrequently in the fur of host animals. Additional specimens of H. d. truncata were examined from Garfield, Grand, and San Juan Counties, UT for comparison of Arizona specimens. Specimens from each of these isolated Utah mountain ranges were conspecific with H. d. truncata from the northern regions of Arizona and New Mexico; however, it should be noted that the DA9 of all specimens from the Kaibab Plateau were markedly much broader than other typical specimens of H. d. truncata (or H. d. obliqua) (Figure 3L). Populations from the Kaibab Plateau deserve additional collecting and assessment. The occurrence of H. d. truncata in Rio Arriba County, NM is a new record for that county. Only three males exist from the Chiricahua Mountains. Although they are separated from the northern population of H. d. truncata by the isolated Pinaleno population of H. d. obliqua, they are rightfully assigned to H. d. truncata. Two such specimens reported as H. dippiei by Beer et al. (1959) were collected from a "Douglas fir-yellow pine association" between Onion Gap and Rustlers Park. Lewis and Lewis (1994) included this male in their synonymies as H. dippiei dippiei. All three of these males possess flattened scales on the dorsal angle of the DA9. The DA9 is also very truncate and arched anteriorly, forming a concavity on the dorsal margin. Two of the three possess a small spiniform between the proximal and penultimate pair of spiniforms on the DA9, a characteristic of H. d. obliqua. This might be expected in any population of H. d. truncata that is geographically associated with H. d. obliqua. The sex ratio (males:females) of H. d. truncata recovered from nests was 0.7:1.0 (n=85) compared to H. d. obliqua, which was 0.66:1.0 (n=266). The sex ratio for the combined total number of Hystrichopsylla d. ssp. was 0.67:1.0 (n=351). Marshall (1981) proposed that at the time of emergence of adult fleas, the sex ratio is usually about equal (based on sample sizes of 100 or more). He concluded that male fleas are shorter-lived and eventually females predominate on hosts and in their nests. Immediate extraction of fleas (dead and alive) from nests and subsequent tabulation of all reared fleas would effectively eliminate this differential. Therefore, our findings would suggest that H. dippiei species produce more

females than males as a biological entity inherent in the species studied. The disparity between the number of fleas extracted from nests illustrated in Table 1 and those used in calculating sex ratios can be explained by two criteria: 1) fleas from all nests were not used to determine sex ratios. Nests that were considered dry or medium dry were not included, although a few fleas may have been present in such nests. An assumption was made that only a moist nest thriving with numerous fleas would reflect a true sex ratio, hence, only those nests were selected; and 2) only nests were included in Table 1 for which the identity of each flea positive nest could be accurately associated with a specific host species (some were uncertain). Hystrichopsylla occidentalis sylvaticus Campos and Stark, 1979 Figure 1 Type Material: Hystrichopsylla occidentalis sylvaticus, Campos and Stark, 1979, J. Med. Entomol. 15: 442, Figures 1-15 (USNM 104685, primary type missing). Type Locality: Holotype male, Foothills Research Campus, 2.5 km W Ft. Collins, CO, 1,580 m, ex Peromyscus difficilis (J.A. Allen), 24 October 1969, E.G. Campos. Hystrichopsylla occidentalis sylvaticus (see Lewis and Lewis 1994 for listing of synonymies) Lewis and Lewis (1994) discussed the distribution of the H. occidentalis complex. The H. o. linsdalei reported by Fagerlund et al. (2001) from Colfax County, NM is certainly a misidentification and probably represents H. o. sylvaticus. Unfortunately, the specimen was apparently used for plague analysis (proved positive for Yersinia pestis) and was destroyed in the procedure. This dubious record is not illustrated in Figure 1. Hystrichopsylla o. sylvaticus is mainly allopatric with H. d. ssp. in Arizona and New Mexico (Figure 1). This might be attributed to the drier habitats of the Rocky Mountain Montane conifer forest where only a few specimens have been collected by Haas and Kucera (unpublished data) and four specimens that were reported by Campos and Stark (1979) near Williams (Coconino County), and Payson (Gila County), AZ. We examined two females from Colorado (label data = "Antonito, Conejos County, CO A.P.#8049A, [ex:] Says Gr. Sq., 8-6-32)", which is likely H. o. sylvaticus based on seven genal ctenidia on each side of the head of both specimens. These were located just north of the New Mexico border near Fagerland's record in "Colfax County" (exact location unknown). Nest Requirements of Hystrichopsylla dippiei ssp. The complex of Hystrichopsylla species inhabiting the southwestern United States occupy extremely different ecological niches. Hystrichopsylla d. neotomae and H. o. linsdalei are restricted to California, each occurring at lower elevations, whereas H. d. truncata, H. d. obliqua, and H. o. sylvaticus are found at higher elevations, usually in subalpine environments of conifer forests. Except for personal communication from H.J. Egoscue cited by Campos and Stark (1979) regarding H. o. sylvaticus, previous knowledge of the

Vol. 30, no. 2

Journal of Vector Ecology Acknowledgments

261

habitat or microenvironmental requirements for these subspecies has been lacking. Although humidity measurements by instrumentation were not taken in our study, the junior author recorded his observations of the visible condition of each nest as it was collected. Across the range of collections in Arizona and New Mexico, 147 nests of voles, woodrats, red squirrels, and Peromyscus spp. were categorized as either "dry" or "moist". Thirty-seven nests were noted as moist and 110 were designated as dry. Thirty (81%) of the 37 moist nests contained H. d. ssp., whereas the yield for dry nests was only five of 110 nests (4.5%). Little difference was noted in nests from different locations of either H. d. truncata (dry = 4.3%, moist = 80%), or H. d. obliqua (dry = 5%, moist = 83%). In cases where dry nests were positive for fleas, there was rarely more than one flea. Positive moist nests frequently yielded many specimens. A number of nests that were not included in these totals were designated as dry but moist on the bottom and/or the surrounding area was very wet. Invariably these nests were positive for live fleas. One or two dead fleas were occasionally found in dry nests, perhaps indicating that the flea had died there after abandoning its host, as opposed to developing in those nests. The location of nests was also carefully noted, e.g., tree cavity, under a log, in a burrow, in branches of tree, etc. In general, nests that were not subterranean, or shielded from dehydration, did not support or harbor hystrichopsyllid fleas. Different host species had no observable influence on the yield of these large fleas in the nests of their respective hosts. The moisture in nests appeared to be the decisive factor in propagation of the species. Based only on empirical results, our findings parallel those of Stark (2002) for H. o. linsdalei. Stark (2002) found that nests of the California vole [M. californicus (Peale)] required more moisture during the dry season to sustain H. o. linsdalei than they required during the wet season. He also found that this flea was more numerous in underground nests than in surface nests. Successful collection of H. dippiei ssp. requires collecting moist nests and preferably in moist situations such as meadows and subterranean or microclimates protected from desiccation. The climatic conditions of lower elevations (below 2,100 m) are apparently too dry for the species to develop successfully. Collections below this elevation is unlikely in future collections. The largest yield of Hystrichopsylla d. ssp. from any single nest of N. mexicana was 54 (29 males, 25 females), T. hudsonicus 35 (13 males, 22 females), M. longicaudus 25 (8 males, 17 females), M. montanus 12 (7 males, 5 females), and P. maniculatus 3 (2 males, 1 female). It is interesting that a diurnal, arboreal squirrel belonging to a family (Sciuridae) differing from all other hosts (Muridae) may support the development of Hystrichopsylla within its nests. Live adults or larvae of Hystrichopsylla were not found in nests of this arboreal host unless the nest was found to be moist (see above), in subterranean or protected from desiccating situations. Habitat preference is the dominant factor in perpetuation of these species and host specificity appears far less important.

For the loan of specimens, the authors are deeply grateful to Nancy Adams, National Museum of Natural History, Smithsonian Institution, Washington, D.C., Valerie BehanPelletier and Barbara Beamer, Biosystematics Research Institute, Ottawa, Ontario, Canada, Sandra L. Brantley, Museum of Southwestern Biology, University of New Mexico, Albuquerque, NM, Philip J. Clausen, Department of Entomology, University of Minnesota, St. Paul, MN, John E. Rawlins, Carnegie Museum of Natural History, Pittsburgh, PA, and Gary Steck, Division of Plant Industry, Florida Department of Agriculture, Gainsville, FL. In addition to his loan of specimens, we are grateful to James R. Kucera, Salt Lake City, UT for kindly permitting us to report his records from the Kaibab Plateau. For identification of various mammals and mammal parts, we are indebted to J.S. Findley and J.A. Cook, University of New Mexico, Albuquerque, NM, and A.H. Harris, University of Texas, El Paso, TX. We are also grateful to the Arizona Game and Fish Department, the New Mexico Department of Game and Fish, and the Utah Division of Wildlife Resources for granting permits to perform the requisite fieldwork. To Michael F. Whiting and the staff of the Monte L. Bean Life Science Museum, Provo, UT, we give thanks for their continued provision of working space, equipment, supplies, and all the general support required for accomplishing this work. Partial funding for this publication was provided by the Monte L. Bean Life Science Museum. REFERENCES CITED Ayala-Barajaas, R., J.C. Morales-Muciño, N. Wilson, J.E. Llorente-Bousquets, and H.E. Ponce-Ulloa. 1988. Colección Alfredo Barrera, catálogo de pulgas (Insects: Siphonaptera), Serie Catálogos del Museo de Zoología "Alfonso L. Herrera" Catalog No. 1. Universidad Nacional Autónoma de México, Ciudad Universitaria, México, D.F., 102 pp. Bailey, V. 1932. Mammals of New Mexico. North American Fauna. 53: 1-412. Barrera, A. 1952. Notas sobre Siphonápteros. IV. Descripción de Hystrichopsylla orophila nov. sp. (Siph., Hystrichips.). Ciencia 12: 39-42. Beer, J.R., E.F. Cook, and R.G. Schwab. 1959. The ectoparasites of some mammals from the Chiricahua Mountains, Arizona. J. Parasit. 45: 605-613. Brown, D.E. 1984. Arizona's Tree Squirrels. Arizona Game and Fish Department, Phoenix, AZ. 114 pp. Brown, D.E., F. Reichenbacher, and S.E. Franson. 1998. A classification of North American biotic communities. University of Utah Press, Salt Lake City, UT. 141 pp. Campos, E.G. and H.E. Stark. 1979. A revaluation of the Hystrichopsylla occidentalis group, with description of a new species (Siphonaptera: Hystrichopsyllidae). J. Med. Entomol. 15: 431-444. Fagerlund, R.A., P.L. Ford, and P.J. Polechla, Jr. 2001. New records for fleas (Siphonaptera) from New Mexico with notes on plague-carrying species. Southwest. Nat. 46: 94-96.

262

Journal of Vector Ecology

December 2005

Findley, J.S. 1987. The natural history of New Mexican mammals. University of New Mexico Press, Albuquerque, NM. 164 pp. Findley, J S., A.H. Harris, D.E. Wilson, and C.Jones. 1975. Mammals of New Mexico. University of New Mexico Press, Albuquerque, NM. 360 pp. Ford, P.L., R.A. Fagerlund, D.W. Duszynski, and P.J. Polechla. 2004. Fleas and lice of mammals in New Mexico. General Technical Report (RMRS-GTR-123), Fort Collins, CO, U.S. Department of Agriculture, Rocky Mountain Research Station, 57 pp. Haas, G.E. 1972. Partial castration in Monopsyllus vison (Baker)(Siphonaptera). Entomol. News 83: 275-278. Haas, G.E. 1973. Morphological notes on some Siphonaptera (Leptopsyllidae and Ceratophyllidae) of New Mexico. Am. Midl. Nat. 90: 246-252. Haas, G.E. and J.R. Kucera. 2004. Fleas (Siphonaptera) in nests of voles (Microtus spp.) in montane habitats of three regions of Utah. Western North Am. Nat. 64: 346-352. Haas, G.E., R.P. Martin, M. Swickard, and B.E. Miller. 1973. Siphonaptera-mammal relationships in northcentral New Mexico. J. Med. Entomol. 10: 281-289. Haas, G.E., R.P. Martin, M. Swickard, and N. Wilson. 1972. Bird fleas (Siphonaptera) of New Mexico. Canad. Entomol. 104: 881-883. Haas, G.E. and N. Wilson. 1998. Polygenis martinezbaezi (Siphonaptera: Rhopalopsyllidae) reared from a rodent nest found in the Peloncillo Mountains of southwestern New Mexico. J. Med. Entomol. 35: 431-432. Haas, G.E., N. Wilson, and C.T. McAllister. 2004. Fleas (Siphonaptera: Ceratophyllidae, Ctenophthalmidae) from rodents in five southwestern states. Western North Am. Nat. 64: 515-517. Hoffmeister, D.F. 1956. Mammals of the Graham (Pinaleno) Mountains, Arizona. Am. Midl. Nat. 55: 257-288. Hoffmeister, D.F. 1986. Mammals of Arizona. Arizona Game and Fish Department, University of Arizona Press, Tucson, AZ. 602 pp. Holland, G.P. 1957. Notes on the Genus Hystrichopsylla Rothschild in the New World, with descriptions of one new species and two new subspecies (Siphonaptera: Hystrichopsyllidae). Canad. Entomol. 89: 309-324. Hopkins, G.H.E. and M. Rothschild. 1962. An illustrated catalogue of the Rothschild collection of fleas (Siphonaptera) in the British Museum (Natural History). Vol. III. Hystrichopsyllidae. Trustees of the British Museum, London, 560 pp. Hubbard, C.A. 1947. Fleas of Western North America, their relation to the public health. Iowa State College Press, Ames, IA, 533 pp.

Istock, C.A and R.S. Hoffmann (eds.). 1995. Storm over a Mountain Island: conservation biology and the Mt. Graham affair. University of Arizona Press, Tucson, AZ. 291 pp. Kucera, J.R. and G.E. Haas. 1992. Siphonaptera (fleas) collected from small mammals in montane southern Utah. Great Basin Nat. 52: 382-384. Lewis, R.E. and R.P. Eckerlin. 2004. A new species of Hystrichopsylla Taschenberg, 1880 (Siphonaptera: Hystrichopsyllidae) from Guatemala. Proc. Entomol. Soc. Wash. 106: 757-760. Lewis, R.E. and G.E. Haas. 2001. A review of the North American Catallagia Rothschild, 1915, with the description of a new species (Siphonaptera: Ctenophthalmidae, Neopsyllinae, Phalacropsyllini). J. Vect. Ecol. 26: 51-69. Lewis, R.E. and J.H. Lewis. 1994. Siphonaptera of North America north of Mexico: Hystrichopsyllidae s. str. J. Med. Entomol. 31: 795-812. Little, E.L., Jr. 1950. Southwestern trees: A guide to the native species of New Mexico and Arizona. U.S. Department of Agriculture Handbook, No. 9, 109 pp. Marshall, A.G. 1981. The sex ratio in ectoparasitic insects. Ecological Entomol. 6: 155-174. Méndez, E. and G.E. Haas. 1972. A new flea of the genus Megarthroglossus Jordan and Rothschild from New Mexico (Siphonaptera: Hystrichopsyllidae, Anomiopsyllinae). J. Med. Entomol. 9: 285-288. Méndez, E. and G.E. Haas. 1973. Megarthroglossus wilsoni, new species, with notes on the genus in New Mexico (Siphonaptera: Hystrichopsyllidae). Ann. Entomol. Soc. Am. 66: 1129-1139. Stark, H.E. 2002. Population dynamics of adult fleas (Siphonaptera) on hosts and in nests of the California vole. J. Med. Entomol. 39: 818-824. Tipton, V.J. and E. Méndez. 1968. New species of fleas (Siphonaptera) from Cerro Potosi, Mexico, with notes on ecology and host parasite relationships. Pac. Insects 10: 177-214. Traub, R. and C.C. Hoff. 1951. Records and descriptions of fleas from New Mexico (Siphonaptera). Am. Mus. Novit. No. 1530: 1-23. Traub, R. and P.T. Johnson. 1952. Four new species of fleas from Mexico (Siphonaptera). Am. Mus. Novit. Number 1598: 1-23. Ungnade, H.E. 1972. Guide to New Mexico Mountains. University of New Mexico Press, Albuquerque, NM. 235 pp. Wilson, D.E. and D.M. Reeder. 1993. Mammal species of the world, a taxonomic and geographic reference. 2nd ed., Smithsonian Institution Press, Washington, D.C. and London, 1206 pp.

Vol. 30, no. 2

Journal of Vector Ecology

263

Abundance and diversity of human-biting flies (Diptera: Ceratopogonidae, Culicidae, Tabanidae, Simuliidae) around a nickel-copper smelter at Monchegorsk, northwestern Russia

M. V. Kozlov1 , N. K. Brodskaya2, A. Haarto3, K. Kuusela4, M. Schäfer5, and V. Zverev1

Section of Ecology, Department of Biology, University of Turku, FIN-20014 Turku, Finland 2 Zoological Institute, Universitetskaja 1, 199034 St. Petersburg, Russia 3 Zoological Museum, Section of Biodiversity and Environmental Science, University of Turku, FIN-20014 Turku, Finland 4 Oulanka Research Station, University of Oulu, Liikasenvaarantie 134, FIN-93999 Kuusamo, Finland 5 Population Biology / Department of Ecology and Evolution, Uppsala University, Norbyvägen 18D, SE-75236 Uppsala, Sweden Received 9 March 2005; Accepted 3 May 2005 ABSTRACT: In the summers of 2001 and 2002, we quantitatively sampled human-biting flies in twelve sites located 1.6 to 63 km from a large copper-nickel smelter at Monchegorsk on the Kola Peninsula, Russia. We collected 429 specimens of three species of Ceratopogonidae, 92 specimens of seven species of Culicidae, 76 specimens of seven species of Tabanidae, and 4,788 specimens of 19 species of Simuliidae. Culicoides chiropterus was for the first time reported from the Kola Peninsula. Catches of Culicidae and Simuliidae decreased near the smelter, presumably due to the combined action of toxicity of pollutants, pollution-induced forest damage, and decline in vertebrate density. An abundance of Ceratopogonidae and Tabanidae, the size of the most common black fly species, Simulium pusillum, and the diversity of all families did not change along the pollution gradient. Journal of Vector Ecology 30 (2): 263-271. 2005. Keyword Index: Environmental contamination, faunistics, blood-sucking flies, Kola Peninsula, landscape degradation, size variation.

1

INTRODUCTION Blood-sucking flies are among the most common and widespread invertebrates in boreal forests. They play an important role in both aquatic (streams/rivers) and terrestrial (riparian/upland) northern ecosystems, and several groups may serve as indicators of water quality. Although the effects of heavy metals and acidification on mosquitoes and black flies are reasonably well documented (Chmielewski and Hall 1992a,b, Rayms-Keller et al. 1998, Courtney and Clements 2002), most studies report effects of water contamination on fly larvae. Almost nothing is known about the changes in abundance and diversity of blood-sucking flies in the impact zones of large industrial polluters, where toxicity problems and indirect effects of habitat deterioration co-occur. Pronounced landscape transformation, from forest damage to development of industrial barrens near the polluter (Rigina and Kozlov 1999), are likely to influence availability of both oviposition and resting sites and blood meal. We are aware of only one publication directly related to the topic of our study: Nekrasova (1995) reported no correlation between the abundance of mosquitoes and the distance to the nickel smelter in Karabash, southern Ural. The nickel-copper smelter at Monchegorsk, northwestern Russia, has been the focus of studies of pollution-induced changes in northern boreal ecosystems during the past decades (Kozlov and Barcan 2000). Biotic effects observed near this smelter are described in more than 400 publications (Kozlov

and Zvereva 2003), providing much of the background information for studying the impact of pollution on different species and communities. The present study documents species richness and the abundance of human-biting flies along a strong pollution gradient at Monchegorsk, thus contributing to an integrated analysis of the impact of severe pollution on different groups of insects (Kozlov 1996, 1997, Kozlov et al. 1996a,b, Ruohomäki et al. 1996, Zvereva et al. 1997, Kozlov and Whitworth 2002). We were also interested in checking whether the size of the most common black fly species, Simulium pusillum, changed along the pollution gradient. MATERIALS AND METHODS Study area and study sites The study was conducted in a lowland forest zone of the Kola Peninsula, north of the Polar Circle. Twelve study sites (Figure 1) were located within 63 km of the Severonikel nickel-copper smelter at Monchegorsk (67° 55' N, 32° 48' E). This smelter is one of the largest European sources of aerial pollution, predominantly of sulfur dioxide and heavy metals (Pearce 1994), and the only big polluter in the central part of the Kola Peninsula. In 1990 it emitted 2.33 x 108 kg of sulfur dioxide, 2.71 x 106 kg of nickel, and 1.81 x 106 kg of copper. Since that time, emissions have steadily declined, reaching 0.44 x 108 kg of sulfur dioxide, 1.21 x 106 kg of nickel, and 0.83 x 106 kg of copper in 2001 (Barcan 2002).

264

Journal of Vector Ecology

December 2005

Figure 1. Position of study sites (labeled with distance and direction from the polluter) relative to Severonikel smelter at Monchegorsk. Settlements are shown by shaded rectangles. Inserted: position of study area (open square) in northern Europe.

Although the smelter was officially opened in 1937, the regular work began only in 1946-1947, and signs of forest damage around the smelter were already visible in the early 1950s (Kozlov and Barcan 2000). Before the smelter started operation, the dominant type of lowland vegetation in the areas south of Monchegorsk was spruce forest (Bobrova and Kachurin 1936). Vast quantities of sulfur and heavy metals emitted during the last 60 years have caused widespread destruction of soils and vegetation around the smelter (Rigina and Kozlov 1999). The total area influenced by aerial pollution recently exceeded 10,000 km², with industrial barrens covering tens of square km (Kryuchkov 1993, Tikkanen and Niemelä 1995). For vegetation characteristics, concentrations of pollutants, and levels of environmental disturbances in the study area, see Barcan (1993), Koroleva (1993), Kryuchkov (1993), Rigina and Kozlov (1999), Zvereva and Kozlov (2001), Kozlov & Zvereva (2004) and Kozlov (2005). The study sites were labeled with the distance (km) and direction from the smelter, e. g. 8S for the site at 8 km distance south of the polluter. The sampled habitats were classified as follows: industrial barren (1.6NW, 6.6S, 8S), birch transitional community (5N, 5.2S, 13.8S), damaged forest (14.6N, 11.1N, 27S), and nearly undamaged (healthy) forest (33.6S, 40.9S, 63SE). Note that the same sites in several earlier publications were labelled as 1N, 7S, 9S, 5N, 5S, 14S, 15N, 11N, 29S, 35S, 47S and 65SE, respectively; present estimates of distances are more accurate as based on the GPS data.

Pollution loads Emissions of any smelter consist of dozens of substances, many of which are toxic. The main pollutants in the study area are sulfur dioxide and heavy metals, predominantly nickel and copper. Concentrations of all pollutants in all media (ambient air, soils, plants, animals) decrease proportionally with increase of distance from the smelter (Barcan 1993, Ruohomäki et al. 1996, Valkama and Kozlov 2001, Kozlov 2005). The Monchegorsk pollution gradient is rather long, because the concentrations of metals are increased already at 50-60 km distances; the peak concentrations of nickel and copper in birch foliage exceed regional background by a factor of 20-25, reaching 243 ± 43 and 141 ± 35 g g-1, respectively (Kozlov 2005). Since it is impossible to attribute the effects observed in the course of field studies to any of the individual pollutants, we, in line with earlier studies (Ruohomäki et al. 1996, Valkama and Kozlov 2001, Kozlov and Whitworth 2002), use distance from the smelter as a rough measure of pollution load in our study sites. Sampling protocol Flies were collected on three dates in both 2001 (on July 9, July 24, and August 5) and 2002 (on June 30, August 1, and August 15). The order in which the sites were visited, as well as the time at which sampling was conducted (between 11:00 and 19:00), varied between sampling dates to minimize effects of sampling time on catches.

Vol. 30, no. 2

Journal of Vector Ecology

265

Table 1. Numbers of Simuliidae by species and study site (total of six sampling sessions), and summary statistics for study sites. Propensity of species to bloodsucking on humans: ***vicious, **common, *bloodsucker on humans only locally and accidentally (after Rubzov 1956, Usova 1961).

Species 14.6N Prosimulium hirtipes (Fries)*** Metacnephia lyra (Lundstr.)** Eusimulium vernum (Macq.)** E. beltukovae Rubz.** E. lundstromi (End.)* E. angustipes Edw. E. aureum (Fries)* E. (Nevermannia) sp. Simulium pusillum (Fries)*** S. equinum (L.)** S. murmanum End.* S. ornatum (Meig.)*** S. frigidum Rubz.** S. monticolum (Frieder.)** S. truncatum (Lundstr.)* S. noelleri Frieder.* S. morsitans Edw.* S. rostratum (Lundstr.)** S. reptans (L.)* Total number Species number 1 3 2 62 11 4 2 76 8 1.332 7 3 0.796 1 1 20 5 0.778 56 5 0.356 7 4 1.154 52 5 0.837 62 5 1.219 518 11 0.692 1 4 2 14 45 1 16 1 1 1 1 2 5 4 24 1 2 22 14 1 3 5 1 2 1 1 1 52 4 40 24 446 1 16 3 11 4 10 8 2 3 2 2 902 12 0.273 12 3 1 2439 9 0.134 16 8 587 11 0.960 20 4 1 2 18 1 5 6 863 2389 430 6 92 1 1 3 11 12 14 5 1 1 1 1 1 1 3 4 2 8 8 2 11.1N Study sites (distance and direction from the Monchegorsk smelter) 5N 1.6NW 5.2S 6.6S 8S 13.8S 27S 33.6S 40.9S 63SE 1 12

Shannon-Wiener diversity index 1.741

266

Journal of Vector Ecology

December 2005

During each sampling session we recorded the following parameters: cloudiness (0, 10, 25, 50, 75, 100%), presence of direct sunshine (yes/no), air temperature (to the nearest 1oC), wind speed (twice, at 2 m height, by using KestrelR Pocket Wind Meter, Nielsen-Kellerman, Boothwyn, PA, U.S.A.; speed range 0.3 to 40 m s-1, accuracy 0.1 m s-1 or 3 % of reading for velocities exceeding 3.5 m s-1), and wetness of ground layer vegetation (wet/dry). Sampling was conducted when wind did not exceed 2 Beauforts, temperature was 14oC or higher, and not earlier than 2 h after the rain. For sampling we used an `Open Air' tent for two persons (base 210 x 210 cm, height 130 cm) that has a door opening of approximately 0.3 m2. This tent was assembled on a sampling site and established with the door opening leeward. The collector (the same person for all sampling sessions) entered the tent and left the door open for 5, 10, 15 or 20 min (15, 17, 28 and 12 sampling sessions, respectively), depending on the number of flies arriving. We aimed at collecting at least 20 individuals whenever possible for more accurate diversity estimates; abundance was re-calculated for a 5-min sampling session prior to the analysis. The collector sat motionless during the sampling session and used a mosquito net but did not use repellents on the day of sampling. After the session, the door was closed and the collector sampled all flies with a portable battery-powered vacuum-cleaner, equipped with a special container. Collected insects were narcotized by chloroform, sorted to families, and preserved on cotton layers for identification. Although the described method may have introduced biases in terms of the probabilities of capture of different species, the catches allow comparisons of relative abundances of human-biting flies among study sites.

Size of Simulium pusillum From each sample we measured either all specimens of S. pusillum, if their number was less than 25, or the 25 first specimens (total of 443 specimens). Black flies were soaked in KOH and preserved in alcohol; the length of hind tibia was measured with the ocular scale of the dissecting microscope (accuracy 0.025 mm). Data analysis We used a Shannon-Wiener diversity index to measure diversity. Low numbers of flies in many of the samples prevented us from use of rarefaction analysis, because this method would not provide an estimate of species richness in samples larger than the smallest individual sample (Krebs 1999). Among-site variation in the numbers of individuals in catches (adjusted for sampling time and log-transformed) and among-habitat variation in diversity were analyzed by General Linear Model, type III sum of squares, which accounted for all variables recorded at the time of sampling (SAS Institute 1990); site-specific means were correlated with logtransformed distance (a proxy of pollution load). Measurements of S. pusillum were averaged for site-specific values, and the Pearson correlation coefficient was calculated to explore relationships between black fly size and distance to the polluter. RESULTS We collected 429 specimens of three species of Ceratopogonidae (Table 3), 92 specimens of seven species of Culicidae (Table 2), 76 specimens of seven species of Tabanidae (Table 4), and 4,788 specimens of 19 species of Simuliidae (Table 1).

Abundance, exx/sample

r = 0.44, P = 0.15 50

9 7

r = 0.77, P = 0.0036

30

5 3

10

a

10 30 50 70

1 10 30 50 70

b

Abundance, exx/sample

r = - 0.05, P = 0.86 9 7 5 3 1 10 30 50 70 300 700 500

r = 0.92, P < 0.0001

c

Distance from the smelter, km

100 10 30 50 70

d

Distance from the smelter, km

Figure 2. Abundance of blood-sucking flies in relation to the distance from the Severonikel smelter at Monchegorsk: a, Ceratopogonidae; b, Culicidae; c, Tabanidae; d, Simuliidae. Means are each based on six catches; bars indicate standard errors.

Vol. 30, no. 2

Journal of Vector Ecology

267

In none of the families was abundance influenced by cloudiness, direct sunshine, or wetness of ground layer vegetation (data not shown), and therefore these variables were excluded from the final analysis. The abundance of Simuliidae varied with temperature, sampling time, and wind speed, while number of Culicidae varied with collecting date; catches of Ceratopogonidae or Tabanidae were not affected by sampling conditions (Table 5). The variation in the mentioned factors was not related with distance to the polluter (data not shown). Among-site variation in catches was found in Simuliidae and Culicidae (Table 5): abundance of these flies increased significantly with distance from the smelter,

while abundance of Ceratopogonidae and Tabanidae did not (Figure 2). We detected no changes in Shannon-Wiener index with distance to polluter in either family (Simuliidae: r = -0.15, n = 12, P = 0.64; Culicidae: r = 0.33, n = 11, P = 0.32; Ceratopogonidae: r = 0.32, n = 8, P = 0.44; Tabanidae: r = 0.17, n = 9, P = 0.66); differences in diversity between habitat types were also non-significant (F3, 24 = 1.29, P = 0.30). Size of S. pusillum, measured as the length of hind tibia, did not change with the distance to the polluter (r = -0.21, n = 12 sites, P = 0.50).

Table 2. Numbers of Culicidae by species and study site (total of six sampling sessions), and summary statistics for study sites.

Species 14.6N Aedes cinereus Meigen Ochlerotatus communis (DeGeer) Oc. excrucians (Walker) Oc. hexodontus Dyar Oc. intrudens Dyar Oc. pullatus (Coquillett) Oc. punctor (Kirby) Unidentified Total number Species number Shannon-Wiener diversity index 7 4 17 3 0.898 3 3 8 3 0.950 3 1 8 3 1.004 0 0 . 1 1 3 2 10 4 1.255 1 1 0 9 3 0.684 4 2 0.562 1 7 1 5 1 1 1 3 1 1 1 1 7 3 22 5 1.187 3 2 8 4 1.242 4 4 10 2 0.636 3 1 3 9 1 1 2 1 3 2 17 5 1.287 11.1N Study sites (distance and direction from the Monchegorsk smelter) 5N 1.6NW 5.2S 6.6S 8S 13.8S 27S 33.6S 1 1 2 8 1 40.9S 63SE

Table 3. Numbers of Ceratopogonidae by species and study site (total of six sampling sessions), and summary statistics for study sites.

Species 14.6N Culicoides fascipennis (Staeg.) Cu. pulicaris (L.) Cu. chiopterus (Mg.) Total number Species number 29 2 2 2 0.693 0 0 . 0 0 . 62 2 0.318 0 0 . 25 2 0.279 19 2 0.336 39 2 0.507 9 20 11.1N 1 1 Study sites (distance and direction from the Monchegorsk smelter) 5N 1.6NW 5.2S 56 6 6.6S 8S 23 2 13.8S 17 2 27S 31 8 33.6S 103 1 2 106 3 0.147 147 2 0.125 0 0 . 40.9S 143 4 63SE

Shannon-Wiener diversity index 0.619

268

Journal of Vector Ecology

December 2005

Table 4. Numbers of Tabanidae by species and study site (total of six sampling sessions), and summary statistics for study sites.

Species 14.6N Haematopa pluvialis (L.) Chrysops nigripes Zett. Hybomitra borealis (F.) H. lurida (Fall.) H. apradi (Szilády) H. lundbecki Lyneborg H. bimaculata (Macq.) Total number Species number 1 11 1 4 5 6 1 29 7 30 4 0.531 2 2 0.693 0 0 . 1 1 0 0 0 . 3 2 0.636 1 1 0 1 1 0 2 1 0 5 5 1.609 0 0 . 2 1 1 11.1N 1 26 1 1 Study sites (distance and direction from the Monchegorsk smelter) 5N 1.6NW 5.2S 1 2 1 1 1 1 6.6S 8S 13.8S 1 2 27S 33.6S 40.9S 1 1 63SE

Shannon-Wiener diversity index 1.618

DISCUSSION Faunistic and life histories of human-biting flies Only four species of Ceratopogonidae (Culicoides pulicaris, Cu. fascipennis, Cu. impunctatus, and Cu. subfascipennis) have so far been reported from the Kola Peninsula (Gutsevich 1973, Glukhova 1989); C. chiopterus is recorded in this region for the first time. Additional species are likely to be found on the Kola Peninsula, since faunas of neighboring territories are more diverse: 13 species are listed for Fennoscandia (Szadziewski et al. 1997), and 16 species for Karelia (Glukhova 1962). Gutsevich (1934) reported six species of Culicidae from the Khibiny mountains which are quite close to our sampling sites, and Fridolin (1936) provided detailed data on the ecology of mosquitoes in the Kola Peninsula where about 1418 species of Culicidae could be expected (Utrio 1979). Our collections included the most severe human-biting species that belong to the so-called snow melt mosquitoes with typically one generation per year. Almost the same set of species was attracted by man in northern Finland (Itämies and Lumiaho 1982). Karvonen (1969) reported 14 species of Tabanidae from the Finnish biogeographical region `Lapponia kemensis' situated approximately at the same latitude as Monchegorsk. Tabanidae of the Kola Peninsula are relatively well investigated, with 24 species recorded so far (Lutta and Bykova 1962); we collected only the most common species. The last comprehensive list of Simuliidae of the Kola Peninsula was published by Usova (1961). Yankovsky (2005) recently reported nine species collected from our study area and described one new species, Argentisimulium tshuni. The

unpublished compilation by J. Raastad, Z. Usova and K. Kuusela includes 60 recorded species, among which some 20 species can more or less disturb humans. We therefore conclude that we collected about half of the species expected from the study area (central part of the Kola Peninsula), and presumably over two-thirds of humanbiting species. Thus, our samples seem representative in terms of taxonomic diversity. Migration data are available only for mosquitoes: the collected species are known to stay nearby their breeding sites and do not disperse more than 2 km (Schäfer et al. 1997, and references therein). However, due to small sample size and significant variation in abundance between study sites, our catches are likely to include mostly local breeders. Changes in diversity and abundance in pollution gradient Although abundances of all groups of human-biting flies were lowest in the most polluted sites (Figure 2), only species of Culicidae and Simuliidae demonstrated gradual decline when approaching the polluter. This pattern allows us to suggest that gradual increase in deposition of pollutants (for pollution levels at our study sites consult Kozlov and Zvereva 2004, Kozlov 2005) may have contributed to this pattern, either due to toxic effects of heavy metals or indirectly, due to habitat deterioration. The maximum reported concentrations of heavy metals in the water of small lakes in the industrially contaminated zone in the Kola Peninsula were 145 g l-1 for nickel and 117 g l-1 for copper (Moiseenko et al. 1995). Furthermore, the concentrations of toxic forms of trace (nickel, copper) and rare (iron, aluminum, manganese) elements in contaminated streams around Monchegorsk greatly increase during a flood period and, in combination

Vol. 30, no. 2

Journal of Vector Ecology

269

Table 5. Sources of variation in abundance of blood-sucking flies (SAS GLM procedure, Type III sum of squares; error d.f. = 55).

Source of variation Distance Year Hour Date Temperature Wind speed 11 1 1 1 1 1 d. f. Culicidae F 5.60 2.84 3.59 7.70 1.73 3.07 P <.0001 0.10 0.07 0.0081 0.20 0.09 Simuliidae F 3.66 0.39 20.2 2.80 9.63 11.1 P 0.0010 0.53 <.0001 0.10 0.0034 0.0018 Ceratopogonidae F 1.08 0.06 0.01 0.79 2.99 0.00 P 0.40 0.80 0.94 0.38 0.09 0.97 Tabanidae F 1.73 1.17 1.76 0.15 2.07 1.03 P 0.10 0.29 0.19 0.70 0.16 0.32

with pH depressions, create extremely stressful conditions for the freshwater fauna (Moiseenko et al. 2001). Another factor that may have influenced these groups of flies is availability of blood meal: biomass of potential hosts (vertebrates) declines with proximity to the smelter to 1050% of the values recorded in unpolluted forests (Gilyazov 1993, Kataev et al. 1994, Kozlov et al. 2005). Last but not least, all collected mosquito species prefer to stay in the forest (Gutsevich et al. 1970), therefore forest decline near the smelter may have contributed to the observed decrease in their abundance. Our results for mosquitoes disagree with the report by Nekrasova (1995) who did not detect density changes attributable to pollution, but this discrepancy most likely results from the fact that industrial barrens have not been surveyed in the cited study. The size of insects frequently declines with increased pollution (Heliövaara and Väisänen 1993). Absence of this effect in S. pusillum, in combination with extremely low abundance of black flies near the polluter, hints that at least some of specimens collected in the most polluted sites may represent occasional migrants rather than locally breeding populations. This explanation seems plausible due to existing data on long-distance passive migrations of black flies (up to 150 km: Rubzov 1956) in combination with high abundance of these species in unpolluted forests, relatively small extent of the heavily contaminated area (about 20 km), and a threefold increase in wind velocity near the smelter due to pollutioninduced forest decline (Kozlov 2002). An alternative explanation is that populations of black flies are depressed by top-down factors (which are not expected to decrease insect size); the data on body concentrations of principal pollutants are necessary to discriminate between these hypotheses. The abundance of Ceratopogonidae and Tabanidae were both rather low at most of the sites, in contrast to high densities observed in a couple of localities. This pattern hints that habitat preferences rather than pollution impact play the leading role in shaping the abundance of these groups. In particular, high catches of Tabanidae were associated with sparse pine forests surrounded by peat bogs (sites 14.6NE and 11N), while Ceratopogonidae were most abundant in swampy spruce

forests (sites 33.6S and 40.9S). It has long been accepted that undisturbed sites have the highest species richness (Margalef 1968, Odum 1971). An alternative hypothesis suggests that species richness is maximized at intermediate levels of disturbance (Grime 1973, Connell 1978) because superior competitors and disturbancetolerant species may coexist only at these conditions. The observed diversity pattern of blood-sucking flies fits neither of these predictions, being independent of the distance to polluter and the level of habitat deterioration. Absence of correlation between pollution load and diversity had been already reported for different groups of insects from impact zones of several polluters (Kozlov 1996, 1997, Kozlov and Zvereva 1997, Kozlov and Whitworth 2002), allowing the validity of the widespread opinion that polluted habitats display a reduction in diversity to be questioned (Magurran 1988). More data from the impact zones of the point polluters are badly needed to explore sources of variation in changes of insect communities associated with pollution-induced habitat deterioration. Acknowledgments We thank E. Melnikov, A. Vassiliev, and A. Zvereva for assistance in fieldwork and sorting the catches. E. Zvereva and two anonymous referees provided valuable comments of an earlier draft of the paper. The study was made possible due to logistic support provided by the Lapland Biosphere reserve, for which we are thankful to S. Shestakov and V. Barcan. REFERENCES CITED Barcan, V.S. 1993. Measurement of atmospheric concentrations of sulphur dioxide by passive lead dioxide absorbers. In: M.V. Kozlov, E. Haukioja, and V.T. Yarmishko (eds.) Aerial pollution in Kola Peninsula. Proc. Intern. Workshop, April 14-16, 1992, St. Petersburg, pp. 90-98. Kola Science Centre, Apatity, Russia.

270

Journal of Vector Ecology

December 2005

Barcan, V.S. 2002. Nature and origin of multicomponent aerial emissions of the copper-nickel smelter complex. Environ. Internat. 28: 451-456. Bobrova, L.I. and M.N. Kachurin. 1936. Vegetation of Monchetundra. In: Y.D. Zinzerling (ed.) Materials on vegetation of Northern and Western Parts of the Kola Peninsula, pp. 95-121. Academy of Sciences, Moscow and Leningrad. (In Russian). Chmielewski, C.M. and R.J. Hall. 1992a. Effects of acid deposition on the emergence of blackflies (Diptera: Simuliidae) over 50 years from Algonquin park streams. Ontario Ministry of the Environment, Ontario. 31 pp. Chmielewski, C.M. and R.J. Hall. 1992b. Responses of immature blackflies (Diptera: Simuliidae) to experimental pulses of acidity. Can. J. Fish. Aquat. Sci. 49: 833-840. Connell, J.H. 1978. Diversity in tropical rain forests and coral reefs. Science 199: 1302-1310. Courtney, L.A. and W.H. Clements. 2002. Assessing the influence of water and substratum quality on bentic macroinvertebrate communities in a metal-polluted stream: an experimental approach. Freshwater Biol. 47: 1766-1778. Fridolin, V.J. 1936. Community of animals and plants of Khibiny mountain area: Biocenothical investigations in 1930-1935. Academy of Sciences, Moscow and Leningrad. 296 pp. (In Russian). Gilyazov, A.S. 1993. Air pollution impact on the bird communities of the Lapland biosphere reserve. In: M.V. Kozlov, E. Haukioja, and V.T. Yarmishko (eds.) Aerial pollution in Kola Peninsula. Proc. Intern. Workshop, April 14-16, 1992, St. Petersburg, pp. 383-390. Kola Science Centre, Apatity, Russia. Glukhova, V.M. 1962. Blood-sucking midges (Diptera, Heleidae) of Karelia. Proc. Zool. Inst. Acad. Sci. USSR 31: 197-249 (In Russian). Glukhova, V.M. 1989. Blood-sucking midges of the genera Culicoides and Forcipomyia (Ceratopogonidae). Fauna USSR. Diptera. Vol. 3, part 5a. Nauka, Leningrad. 408 pp. (In Russian). Grime, J.P. 1973. Competitive exclusion in herbaceous vegetation. Nature 242: 344-347. Gutsevich, A.V. 1934. On mosquitoes from Khibiny mountains. Magasin de parasitologie de l'Institut zoologique de l'Académie des Sciences de l'URSS 6: 517 (In Russian). Gutsevich, A.V. 1973. The blood-sucking midges (Ceratopogonidae). Fauna USSR. Diptera. Vol. 3, part 5. Nauka, Leningrad. 270 pp. (In Russian). Gutsevich, A.V., A.S. Monchadskii, and A.A. Shtakelberg. 1970. Diptera. Mosquitoes, Family Culicidae. Fauna USSR. Diptera. Vol. 3, part 4. Nauka, Leningrad. 384 pp. (Translation by Keter Press, Jerusalem, 1974). Heliövaara, K. and R. Väisänen. 1993. Insects and pollution. CRC-Press, Florida. 393 pp. Itämies, J. and I. Lumiaho. 1982. Mosquitoes (Diptera, Culicidae) attracted by man in SW and NE Finland. Aquilo Ser. Zool. 21: 1-5. Karvonen, J. 1969. On Finnish Tabanids (Diptera). Ann.

Entomol. Fenn. 35: 176-183. Kataev, G.D., J. Suomela, and P. Palokangas. 1994. Densities of microtine rodents along a pollution gradient from a copper-nickel smelter. Oecologia 97: 491-498. Koroleva, N. 1993. Pollution-induced changes in forest vegetation structure as revealed by ordination test. In: M.V. Kozlov, E. Haukioja, and V.T. Yarmishko (eds.) Aerial pollution in Kola Peninsula. Proc. Intern. Workshop, April 14-16, 1992, St. Petersburg, pp. 339345. Kola Science Centre, Apatity, Russia. Kozlov, M.V. 1996. Subalpine and alpine assemblages of Lepidoptera in the surroundings of a powerful smelter on the Kola Peninsula, NW Russia. Nota lepidopterol. 18: 17-37. Kozlov, M.V. 1997. Pollution impact on insect biodiversity in boreal forests: evaluation of effects and perspectives of recovery. In: R.M.M. Crawford (ed.) Disturbance and recovery in Arctic Lands: an ecological perspective. Proceedings of the NATO advanced research workshop on disturbance and recovery of Arctic terrestrial ecosystems, Rovaniemi, Finland, 24-30 September 1995, pp. 213-250. (NATO ASI series. Partnership subseries 2. Environment: vol. 25). Kluwer Academic Publishers, Dordrecht. Kozlov, M.V. 2002. Changes in wind regime around a nickelcopper smelter at Monchegorsk, northwestern Russia. Intern. J. Biometeorol. 46: 76-80. Kozlov, M.V. 2005. Sources of variation in concentrations of nickel and copper in mountain birch foliage near a nickelcopper smelter at Monchegorsk, north-western Russia: results of long-term monitoring. Environ. Pollut. 135: 91-99. Kozlov, M.V. and V.S. Barcan. 2000. Environmental contamination in the central part of the Kola Peninsula: history, documentation, and perception. AMBIO 29: 512517. Kozlov, M.V. and T. Whitworth. 2002. Population densities and diversity of Calliphoridae (Diptera) around a nickelcopper smelter at Monchegorsk, Northwestern Russia. Entomol. Fenn. 13: 98-104. Kozlov, M.V. and E.L Zvereva. 1997. Effects of pollution and urbanization on diversity of frit flies (Diptera: Chloropidae) in industrially polluted areas. Acta Oecol. 18: 13-20. Kozlov, M.V. and E.L. Zvereva. 2003. Impact of industrial polluters on terrestrial ecosystems: a research synthesis. In: J.O. Honkanen and P.S. Koponen (eds.) Sixth Finnish Conference of Environmental Sciences: Proceedings, pp. 72-75. Finnish Society for Environmental Sciences and University of Joensuu, Joensuu, Finland. Kozlov, M.V. and E.L. Zvereva. 2004. Reproduction of mountain birch along a strong pollution gradient near Monchegorsk, Northwestern Russia. Environ. Pollut. 132: 443-451. Kozlov, M.V., J. Jalava, A.L. Lvovsky, and K. Mikkola. 1996a. Population densities and diversity of Noctuidae (Lepidoptera) along an air pollution gradient on the Kola Peninsula, Russia. Entomol. Fenn. 7: 9-15.

Vol. 30, no. 2

Journal of Vector Ecology

271

Kozlov, M.V., A.L. Lvovsky, and K. Mikkola. 1996b. Abundance of day-flying Lepidoptera along an air pollution gradient in the northern boreal forest zone. Entomol. Fenn. 7: 137-144. Kozlov, M.V., E.L. Zvereva, A.S. Gilyazov, and G.D. Kataev. 2005. Contaminated zone around a nickel-copper smelter: a death trap for birds and mammals? In: F. Columbus (ed.) Focus on ecology research. Nova Science Publishers, Hauppauge, N.Y. (in press). Krebs, C. J. 1999. Ecological methodology. 2nd ed. Addison Wesley Longman, Menlo Park, CA. 620 pp. Kryuchkov, V.V. 1993. Extreme anthropogenic loads and the northern ecosystem condition. Ecol. Appl. 3: 622-630. Lutta, A.S. and C.I. Bykova. 1982. Tabanidae of the European North of the U.S.S.R. Nauka, Leningrad. 184 pp. (In Russian). Magurran, A.E. 1988. Ecological diversity and its measurement. Chapman and Hall, London. 179 pp. Margalef, R. 1968. Perspectives in ecological theory. University of Chicago Press, Chicago, IL. viii + 111 pp. Moiseenko, T.I., L.P. Kudryavtseva, I.V. Rodyushkin, V.A. Dauvalter, A.A. Lukin, and N.A. Kashulin. 1995. Airborne contaminantion by heavy metals and aluminium in the freshwater ecosystems of the Kola subarctic region (Russia). Sci. Total Environ. 160/161: 715-727. Moiseenko, T., L. Kudrjavzeva, and I. Rodyshkin. 2001. The episodic acidification of small streams in the spring flood period of industrial polar region, Russia. Chemosphere 42: 45-50. Nekrasova, L.S. 1995. Development of larvae of bloodsucking mosquitoes of the genus Aedes in industrial landscapes of Southern Ural. Ekologia [Ecology, Ekaterinburg] 0(5): 400-403 (In Russian). Odum, E.P. 1971. The strategy of ecosystem development. Science 164: 262-270. Pearce, F. 1994. Europe Top 100 Polluters. New Scientist 143: 9. Rayms-Keller, A., K.E. Olson, M. McGaw, C. Oray, J.O. Carlson, and B.J. Beaty. 1998. Effect of heavy metals on Aedes aegypti (Diptera: Culicidae) larvae. Ecotoxicol. Environ. Safety 39: 41-47. Rigina, O. and M.V. Kozlov. 1999. Pollution impact on subArctic northern taiga forests in the Kola peninsula, Russia. In: J.L. Innes, and J. Oleksyn (eds.) Forest dynamics in heavily polluted regions, pp. 37-65. (IUFRO Research Series, 1). CAB International, Wallingford, U.K. Ruohomäki, K., P. Kaitaniemi, M.V. Kozlov, T. Tammaru, and E. Haukioja. 1996. Density and performance of

Epirrita autumnata (Lep., Geometridae) along three air pollution gradients in northern Europe. J. Appl. Ecol. 33: 773-785. Rubzov, I.A. 1956. Black flies (Fam. Simuliidae). Fauna USSR. Diptera. Vol. 6, part 6. 2nd Edn. Academy of Sciences, Moscow and Leningrad. 859 pp. (In Russian). SAS Institute. 1990. SAS/Stat. User's guide, version 6.0. SAS Institute, Cary, N.C. 1686 pp. Schäfer, M. and J.O. Lundström, 2001. Comparison of mosquito (Diptera: Culicidae) fauna characteristics of forested wetlands in Sweden. Ann. Entomol. Soc. Am. 94: 576-582. Schäfer, M., V. Storch, A. Kaiser, M. Beck, and N. Becker. 1997. Dispersal behavior of adult snow melt mosquitoes in the Upper Rhine Valley, Germany. J. Vector Ecol. 22: 1-5. Szadziewski, R., J. Krzywinski, and W. Gilka. 1997. Diptera Ceratopogonidae, Biting Midges. In: A. Nilsson (ed.) Aquatic insects of North Europe: a taxonomic handbook. Vol. 2 : Odonata ­ Diptera, pp. 243-263. Apollo Books, Stenstrup, Denmark. Tikkanen, E. and I. Niemelä (eds.) 1995. Kola Peninsula pollutants and forest ecosystems in Lapland. Final report of the Lapland Forest Damage Project. Finnish Forest Research Institute, Rovaniemi, Finland. 82 pp. Usova, Z.V. 1961. Fauna of black flies of Karelia and Murmansk region (Diptera, Simuliidae). Academy of Sciences, Moscow and Leningrad. 286 pp. (Translation by IPST, Jerusalem 1964). Utrio, P. 1979. Geographic distribution of mosquitoes (Diptera, Culicidae) in eastern Fennoscandia. Notulae Entomol. 59: 105-123. Valkama, J. and M.V. Kozlov. 2001. Impact of climatic factors on developmental stability of mountain birches growing in a contaminated area. J. Appl. Ecol. 38: 665-673 Yankovsky, A.V. 2005. Simuliidae (Diptera) of Lapland biosphere reserve, with description of new species Argentisimulium tshuni Yankovsky, sp. n. Entomol. Obozr. (in press) (In Russian). Zvereva, E.L. and M.V. Kozlov. 2001. Effects of pollution induced habitat disturbance on willow response to simulated herbivory. J. Ecol. 89: 21-30. Zvereva, E.L., M.V. Kozlov, and E. Haukioja. 1997. Population dynamics of a herbivore in an industrially modified landscape: case study with Melasoma lapponica (Coleoptera: Chrysomelidae). Acta Phytopatol. Entomol. Hung. 32: 251-258.

272

Journal of Vector Ecology

December 2005

An annotated checklist of the Anopheles mosquitoes (Diptera: Culicidae) in Iran

Mohammad Mehdi Sedaghat and Ralph E. Harbach1

Department of Medical Entomology, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran 1 Department of Entomology, The Natural History Museum, Cromwell Road, London SW7 5BD, U.K. Received 11 March 2005; Accepted 1 May 2005 ABSTRACT: An annotated checklist of the Anopheles mosquitoes in Iran is provided. The list is based on data contained in Iranian manuals and reports and information published elsewhere. Supplementary information is provided for 24 species. Journal of Vector Ecology 30 (2): 272-276. 2005. Keyword Index: Anopheles, mosquitoes, malaria vectors, checklist, Iran.

INTRODUCTION A number of articles have been published on the Anopheles (Diptera: Culicidae: Anophelinae) of Iran. Most pertain to geographical areas of the country where malaria has been a major problem. The earliest studies, conducted by Iranian and foreign investigators, were related to the fight against malaria and contributed much information on Anopheles species. The first such study involved surveys in northern parts of the country (Latishev 1921) that identified An. maculipennis s.l. as the vector of malaria. Marsh (1933) described Anopheles apoci as a new species from southwestern Iran. Amidzadeh (1941) conducted malaria surveys in northwestern and southeastern areas of the country where he recorded An. maculipennis and An. superpictus as vectors. Lindberg (1941) subsequently recognized nine Anopheles species in Iran, including An. bifurcatus, An. culicifacies, An. maculipennis, An. multicolor, An. pseudopictus, An. rhodesiensis, An. stephensi, An. superpictus, and An. turkhudi. However, An. bifurcatus actually referred to Culex pipiens and the presence of An. rhodesiensis has never been confirmed, but it seems that the name rhodesiensis used in much of the early European literature referred to An. dthali (Pringle 1954). Gutsevich (1943), during an expedition in northern parts of the country, found An. maculipennis, An. sacharovi, An. hyrcanus, An. algeriensis, An. plumbeus, An. superpictus, and An. pulcherrimus. Zolotarev (1945) made a very thorough investigation of eggs of the Maculipennis Complex in the Caspian Sea region and northwestern parts of Iran. Etherington and Sellick (1946) provided some notes on the bionomics of An. maculipennis and An. sacharovi from Iran and Iraq and examined the distribution of the two species in central and northern areas of Iran. Macan (1950) recorded 15 species of Anopheles from Iran and provided a key for the identification of these species in both Iran and Iraq. Dow (1953) described 11 anophelines from Iran, one of which, An. subalpinus, is now considered to be a synonym of An. melanoon (Linton et al. 2002b). Following these earlier reports, Mofidi (1953), Ghaffary (1954), Ansary (1956), and Faghih (1969) recognized 19

species of Anopheles in Iran. Shahgudian (1960) provided a checklist and comprehensive keys to the adult females and 4th instar larvae of the Iranian Anopheles. These keys included 20 species, including three varieties of An. hyrcanus. During an expedition to Iran in 1970, Minar (1974) found An. maculipennis s.l. in central and northern areas and An. pseudopictus in northern areas of the country. Saebi (1987) recorded 17 species in Iran based on the identification of more than 40,000 Anopheles larvae collected around the country during surveys conducted in 1984-1986. Saebi did not find An. algeriensis or An. subpictus, which were reported previously. Based on collections made in 1987-1988, Zaim and Javaherian (1991) reported the presence of An. culicifacies species A in Sistan va Baluchistan Province in the southeastern corner of the country. In the following year, Glick (1992) listed 22 species of Anopheles in Iran based on literature records. His list included An. subalpinus and An. martinius. Anopheles subalpinus, as indicated above, is now recognized as a synonym of An. melanoon. Furthermore, the occurrence of An. martinius in Iran has not been confirmed and is, therefore, removed from the list of Iranian Anopheles even though it could be present in northeastern areas of the country. Most recently, Sedaghat et al. (2003b) described An. persiensis as a new species of the Maculipennis Group in Iran, and Oshaghi et al. (2004) recorded the presence of species B of the An. culicifacies complex. Herein, we recognize a total of 24 Anopheles species in the country. These species are listed below, with information derived from published records and our own data. Subgenus Anopheles Meigen 1. An. (Ano.) algeriensis Theobald. The records of An. algeriensis in Iran are based on studies carried out before 1955. This species has not been found in Iran since then. 2. An. (Ano.) claviger Meigen. Although Dow (1953) only recorded this species from Azerbaijan, it has a wide distribution in northern and central areas of Iran. 3. An. (Ano.) hyrcanus Pallas. The first record of An. hyrcanus was from the southern shores of the Caspian Sea

Vol. 30, no. 2

Journal of Vector Ecology

273

(Ramsdale 2001). The name hyrcanus is derived from Hyrcania, a northern part of the ancient Persian Empire, which corresponds to Mazandaran Province of present-day Iran (Kitzmiller 1982). Some reports of An. pseudopictus and reports of An. nigerrimus in Iran (Dow 1953, Mofidi 1953, Ansary 1956, Shahgudian 1960, Faghih 1969) apply to this species. 4. An. (Ano.) maculipennis Meigen. Anopheles maculipennis is the nominotypical member of the Holarctic Maculipennis Group. Anopheles maculipennis s.l. is the historical vector of malaria in Eurasia. In the early 1920s it was found to represent at least two species based on egg morphology (Falleroni 1926, van Thiel 1927). Current understanding of the composition of the Maculipennis Group in the Palaearctic Region stems from White (1978), de Zulueta et al. (1983), Linton et al., (2002b), Sedaghat et al. (2003b), and Nicolescu et al. (2004). White (1978) suggested that An. lewisi Ludlow may be synonymous with An. messeae or An. beklemishevi, but this nominal form is still regarded as a valid species because its identity has not been resolved. Consequently, 11 species of the Maculipennis Group are currently formally recognized in the Palaearctic Region, namely An. atroparvus, An. beklemishevi, An. daciae, An. labranchiae, An. lewisi, An. maculipennis, An. martinius, An. melanoon, An. messeae, An. persiensis, and An. sacharovi. With the exception of An. sacharovi, these species are impossible to distinguish morphologically in the adult and larval stages and, despite chromosome and isoenzyme differences, egg morphology remains the golden standard by which the species are routinely identified. As a result of recent molecular genetic studies, DNA assays and sequence data are available for identifying eight members of the group, including An. atroparvus, An. daciae, An. labranchiae, An. maculipennis, An. melanoon, An. messeae, An. persiensis, and An. sacharovi (Marinucci et al. 1999, Proft et al. 1999, Linton et al. 2002a, 2002b, 2003, Sedaghat et al. 2003a, 2003b, Nicolescu et al. 2004). At present, only three members of the group, An. maculipennis, An. persiensis, and An. sacharovi, have been identified in Iran based on molecular methods (Sedaghat et al. 2003b). Anopheles melanoon and An. messeae are listed as Iranian species based on egg studies (Dow 1953, Zolotarev 1945, Faghih 1969, Moemeni 1991). Members of the Maculipennis Group are distributed mostly in northern and central areas of the country. 5. An. (Ano.) marteri sogdianus Keshishian. Pringle (1954) recorded this species from Luristan Province in central Iran during studies conducted in 1943. Shahgudian (1956) examined many specimens from Yazd in the central plateau of Iran, Algeria, Tunisia, Spain, Jordan, and Palestine, and explained the differences between An. marteri marteri and An. marteri sogdianus. Saebi (1987) recorded the latter taxon from the central plateau and some western and southwestern areas of the country. 6. An. (Ano.) melanoon Hackett. Records of this species in the northern provinces of Gilan and Mazandaran in the Caspian Sea littoral are based on egg morphology (Dow 1953, Zolotarev 1945, Faghih 1969, Moemeni 1991). There is no evidence to support the role of this member of the

Maculipennis Group in malaria transmission. 7. An. (Ano.) messeae Falleroni. The reported occurrence of this member of the Maculipennis Group in Gilan and Mazandaran Provinces of the Caspian Sea littoral is also based on the morphological identification of eggs (Zolotarev 1945, Faghih 1969, Moemeni 1991). Recent studies, however, did not find any evidence for the presence of this species in the north of Iran (Sedaghat et al. 2003b). 8. An. (Ano.) persiensis Linton, Sedaghat & Harbach. This species is a member of the Maculipennis Group (see An. maculipennis above). Anopheles persiensis has been collected only in the northern Caspian Sea littoral provinces of Gilan and Mazandaran. It seems likely that the species could be responsible for malaria transmission in this region that was previously attributed to An. maculipennis (Sedaghat et al. 2003b). 9. An. (Ano.) plumbeus Stephens. This species occurs mainly in forests of northern areas of the country (Pringle 1954, Saebi 1987). 10. An. (Ano.) pseudopictus Grassi. This species is only known from Gilan and Mazandaran Provinces in the north of Iran (Dow 1953, Saebi 1987, Glick 1992, Azari-Hamidian et al. 2004, Sedaghat et al. unpublished data). 11. An. (Ano.) sacharovi Favre. This is another member of the Maculipennis Group (see An. maculipennis above). Anopheles sacharovi is a major vector of malaria in the central plateau of Iran and is widely distributed in central areas of the country (Faghih 1969, Manouchehri et al. 1992, Sedaghat et al. 2003a). Subgenus Cellia Theobald 12. An. (Cel.) apoci Marsh. Anopheles apoci was originally described from southern Iran (Marsh 1933). The specific name is derived from the acronym for the AngloPersian Oil Company. Pringle (1954) doubted the validity of this species and assumed it was probably a melanic form of An. dthali, but Shahgudian (1960) treated it as a valid species in his study and keys to the Anopheles of Iran. Anopheles apoci occurs in southern areas of the country. 13. An. (Cel.) culicifacies species A. Anopheles culicifacies is a complex of five sibling species, informally designated species A, B, C, D, and E (Kar et al. 1999). Zaim and Javaherian (1991) reported the occurrence of species A in Iran based on cytological study, and this was later confirmed by Oshaghi et al. (2004) (see next species). Anopheles culicifacies s.l. is considered to be a major malaria vector in the southeastern corner of Iran (Manouchehri et al. 1992). 14. An. (Cel.) culicifacies species B. Oshaghi et al. (2004) reported the presence of this species in Iran, and confirmed the presence of species A, based on PCR-RFLP analysis of the mitochondrial cytochrome c oxidase I gene (COI). Anopheles culicifacies s.l. is recorded from the south and southeastern corner of Iran. 15. An. (Cel.) dthali Patton. Anopheles dthali is considered a secondary vector of malaria in southern areas of Iran (Manoochehri et al. 1972, Manouchehri et al. 1992). 16. An. (Cel.) fluviatilis James. Sarala et al. (1994) originally recognized three species within An. fluviatilis

274

Journal of Vector Ecology REFERENCES CITED

December 2005

(provisionally designated species S, T, and U) based on the banding patterns of polytene chromosomes. Manonmani et al. (2001) developed a PCR assay from rDNA ITS2 sequence differences that identified two of these species, which in the absence of chromosomal data were referred to as species X and Y. Manonmani et al. (2003) showed that species X and Y correspond to species S and T, respectively, based on the correlation of chromosomally and PCR-identified mosquitoes. Comparison of DNA sequences in GenBank revealed that the D3 region of 28S rDNA of species S is identical with that of species C of the An. minimus complex (Garros et al. 2005). Consequently, the An. fluviatilis complex includes only two species, species T and U. Naddaf et al. (2003) reported the presence of species T in Iran based on rDNA ITS2 sequence. Anopheles fluviatilis s.l. is generally considered to be a major vector in southern areas of Iran, and a secondary vector in the southeastern corner of the country (Faghih 1969, Manouchehri et al. 1992), but further study is needed to clarify the role of this taxon in malaria transmission. 17. An. (Cel.) moghulensis Christophers. This species is not considered to be a vector of malaria in Iran. Its distribution is limited in the southeastern corner of the country. 18. An. (Cel.) multicolor Cambouliu. Anopheles multicolor mainly occurs in the eastern half of Iran (northeast, southeast, and east-central areas of the country). 19. An. (Cel.) pulcherrimus Theobald. Anopheles pulcherrimus is considered a potential vector of malaria in Sistasn va Baluchistan Province in the southeastern corner of Iran based on natural infections found in this species in Baluchistan (Zaim et al. 1993). 20. An. (Cel.) sergentii Theobald. This species occurs in some central and southern areas of the country. It does not play a role malaria transmission in Iran. 21. An. (Cel.) stephensi Liston. This species has three egg phenotypes. All three, namely mysorensis (Mesghali and Eshghi 1960, Rutledge et al. 1970, Sahabi et al. 1974, Yaghoobi 2004), typical and intermediate (Yaghoobi 2004), have been found in Iran. Anopheles stephensi is generally considered to be a major malaria vector in southern areas of the country. 22. An. (Cel.) subpictus Grassi. Anopheles subpictus is a complex of four sibling species (informally designated species A, B, C, and D) in India (Suguna et al. 1994). The first reports of this species in Iran stem from studies carried out before 1955. There were no reports of An. subpictus in Iran during the last 50 years, but Oshaghi et al. (personal communication) recently found this species in the southeastern corner of the country. It is not known which of the four species of the complex occur(s) in Iran. 23. An. (Cel.) superpictus Grassi. This species has a wide distribution in Iran. Anopheles superpictus is considered to be a major malaria vector in the central plateau and southern areas of the country (Faghih 1969, Manouchehri et al. 1992). 24. An. (Cel.) turkhudi Liston. This species is not considered to be a vector of malaria. It occurs in the southern and some central areas of Iran.

Amidzadeh, G. 1941. Research sur la paludism dans l'Iran. Acta Med. Scand. 107: 579-583. Ansary, N. 1956. Study on distribution of Anopheline of Iran. Tehran. Reports of the Institute of Parasitology and Malariology No. 359. (in Persian). Azari-Hamidian, S., M.A. Joeafshani, A.R. Rassaei, M. Moslem, and E. Mousavi-Evanaki. 2004. Mosquito fauna of the genus Anopheles (Diptera: Culicidae) in Gilan Province. Modarres J. Med. Sci. 6: 11-22. (in Persian). de Zulueta, J., C.D. Ramsdale, R. Cianchi, L. Bullini, and M. Coluzzi. 1983. Observations on the taxonomic status of Anopheles sicaulti. Parassitologia 23: 73-92. Dow, R.P. 1953. Notes on Iranian mosquitoes. Am. J. Trop. Med. 2: 683-695. Etherington, D. and G. Sellick. 1946. Notes on the bionomics of Anopheles sacharovi in Persia and Iraq. Bull. Entomol. Res. 37: 191-195. Faghih, M.A. 1969. Malariology and malaria eradication. Tehran. Tehran University Press. (in Persian). Falleroni, D. 1926. Fauna anofelica italiana e suo `habitat' (paludi, risaie, canali). Metodi di lotta contro la malaria. Riv. Malariol. 5: 553-559. Garros, C., R.E. Harbach, and S. Manguin. 2005. Morphological assessment and molecular phylogenies of the Funestus and Minimus Groups of Anopheles (Cellia). J. Med. Entomol. 42: 42: 522-536. Ghaffary, E.N. 1954. Tentative distributional data on the anophelines of Iran. Distribution. American Mosquito Control Association Meeting, Atlantic City, NJ. Glick, J.I. 1992. Illustrated key to the female Anopheles of southwestern Asia and Egypt (Diptera: Culicidae). Mosq. Syst. 24: 125-153. Gutsevich, A.V. 1943. On the mosquitoes of North Iran. Compt. Rend. Acad. Sci. USSR 40: 123-125. Kar, I., S.K. Subbarao, A. Eapen, J. Ravindran, T.S. Satyanarayana, K. Raghavendra, N. Nanda, and V.P. Sharma. 1999. Evidence for a new malaria vector species, species E, within the Anopheles culicifacies complex (Diptera: Culicidae). J. Med. Entomol. 36: 595-600. Kitzmiller, J.B. 1982. Anopheline names: their derivation and histories. The Thomas Say Foundation, Vol. III. Entomological Society of America, College Park, MD. Latishev, L.N. 1921. Cited by E.N. Povlovsky in: epidemic parasitology mission to Iran and parasitological surveys. Acad. Sci. U.S.S.R. 1948: 235-238. Lindberg, K. 1941. Le paludisme dans l'Iran. Acta Med. Scand. 107: 547-578. Linton, Y.-M., A. Saminidou-Voyadjoglou, and R.E. Harbach. 2002a. Ribosomal ITS2 sequence data for Anopheles maculipennis and An. messeae in northern Greece, with a critical assessment of previously published sequences. Insect Mol. Biol. 11: 379-383. Linton, Y.-M., L. Smith, and R.E. Harbach. 2002b. Observations on the taxonomic status of Anopheles subalpinus Hackett & Lewis and An. melanoon Hackett. Eur. Mosq. Bull. 13: 1-7.

Vol. 30, no. 2

Journal of Vector Ecology

275

Linton, Y.-M., L. Smith, G. Koliopoulos, A. SamanidouVoyadjoglou, A.K. Zounos, and R.E. Harbach. 2003. Morphological and molecular characterization of Anopheles (Anopheles) maculipennis Meigen, type species of the genus and nominotypical member of the Maculipennis Complex. Syst. Entomol. 28: 39­55. Macan, T.T. 1950. The anopheline mosquitoes of Iraq and north Persia, pp. 111-219, Anopheles and malaria in the Near East. Mem. Lond. Sch. Hyg. Trop. Med. No.7. Manonmani, A., N. Nanda, P. Jambulingam, S. Sahu, T. Vijayakumar, J. Ramya Vani, and S.K. Subbarao. 2003. Comparison of polymerase chain reaction assay and cytotaxonomy for identification of sibling species of Anopheles fluviatilis (Diptera: Culicidae). Bull. Entomol. Res. 93: 169-171. Manonmani, A., H. Townson, T. Adeniran, P. Jambulingam, S. Sahu, and R. Vijayakumar. 2001. rDNA-ITS2 polymerase chain reaction assay for the sibling species of Anopheles fluviatilis. Acta Trop. 78: 3-9. Manouchehri, A.V., M. Zaim, and A.M. Emadi. 1992. A review of malaria in Iran, 1975-1990. J. Am. Mosq. Contr. Assoc. 8: 381-385. Manoochehri, A.V., M. Ghiasseddin, and E.R. Shahgudian. 1972. Anopheles dthali Patton 1905, a new secondary vector in southern Iran. Ann. Trop. Med. Parasitol. 66: 537-538. Marinucci, M., R. Romi, P. Mancini, M. Di Luca, and C. Severini. 1999. Phylogenetic relationships of seven Palaearctic members of the maculipennis complex inferred from ITS2 sequence data. Insect Mol. Biol. 8: 469-480. Marsh, F. 1933. A new species of Anopheles (Myzomyia group) from Southwest Persia. Stylops 2: 193-197. Mesghali, A. and N. Eshghi. 1960. Notes on verity of Anopheles stephensi. Reports of Institute of Parasitology and Malariology No. 780. (in Persian). Minar, J. 1974. Results of the Czechoslovak-Iranian Entomological Expedition to Iran 1970 (with enclosed results of collections made in Anatolia). No. 6. Diptera: Culicidae. Acta Entomologica Musei Nationalis. Mofidi, S. 1953. Distribution of the anophelines of Iran. Reports of the Institute of Parasitology and Malariology, Tehran. (in Persian). Moemeni, S. 1991. Distribution of Anopheles maculipennis complex in Gilan and Mazandaran Provinces and determination of susceptibility to insecticides in Mazandaran. MSPH Thesis, Tehran University of Medical Sciences, Tehran. (in Persian). Naddaf Dezfouli, S.R., M.A. Oshaghi, H. Vatandoost, and M. Assmar. 2003. rDNA- ITS based species-diagnostic polymerase chain reaction assay for identification of sibling species of Anopheles fluviatilis in Iran. Southeast Asian J. Trop. Med. Publ. Hlth. 34 (Suppl. 2): 56-60. Nicolescu, G., Y.-M. Linton, A. Vladimirescu, T.M. Howard, and R.E. Harbach. 2004. Mosquitoes of the Anopheles maculipennis group (Diptera: Culicidae) in Romania, with the discovery and formal recognition of a new species based on molecular and morphological evidence.

Bull. Entomol. Res. 95: 525-535. Oshaghi M.A., B. Taghilo, M.T. Moradi, and H. Vatandoost. 2004. Detection of the Anopheles culicifacies complex, species A and B in Baluchistan using mtDNA PCR-RFLP assay; the first report of species B from Iran. Hakim 7: 35-42. (in Persian). Pringle, G. 1954. The identification of the adult anopheline mosquitoes of Iraq and neighbouring territories. Bull. Endem. Dis. 1: 53-76. Proft, J., W.A. Maier, and H. Kampen. 1999. Identification of six sibling species of the Anopheles maculipennis complex (Diptera: Culicidae) by a polymerase chain reaction assay. Parasitol. Res. 85: 837-843. Ramsdale, C.D. 2001. Internal taxonomy of the Hyrcanus Group of Anopheles (Diptera: Culicidae) and its bearing on the incrimination of vectors of malaria in the west of the Palaearctic Region. Eur. Mosq. Bull. 10: 1-7. Rutledge L.C., R.A. Ward, and W.E. Bickley. 1970. Experimental hybridization of geographic strains of Anopheles stephensi (Diptera: Culicidae). Ann. Entomol. Soc. Am. 63: 1024-1030. Saebi, M.E. 1987. Morphological study on anopheline larvae and their distribution in Iran. PhD Thesis. Tehran University of Medical Sciences, Tehran. (in Persian). Sahabi, Z., J.D. Amirkhanian, and E. Shahgoudian. 1974. The polytene chromosome pattern of Anopheles stephensi mysorensis of Kazeroon, Iran. Iran. J. Publ. Hlth. 2: 194198. Sarala, K.S., N. Nutan, K. Vasantha, V.K. Dua, M.S. Malhotra, R.S. Yadav, and V.P. Sharma. 1994. Cytogenetic evidence for three sibling species in Anopheles fluviatilis (Diptera: Culicidae). Ann. Entomol. Soc. Am. 87: 116-121. Sedaghat, M.M., Y.-M. Linton, G. Nicolescu, L. Smith, G. Koliopoulos, A.K. Zounos, M.A. Oshaghi, H. Vatandoost, and R.E. Harbach. 2003a. Morphological and molecular characterization of Anopheles (Anopheles) sacharovi Favre, a primary vector of malaria in the Middle East. Syst. Entomol. 28: 241-256. Sedaghat, M.M., Y.-M. Linton, M.A. Oshaghi, H. Vatandoost, and R.E. Harbach. 2003b. The Anopheles maculipennis complex (Diptera: Culicidae) in Iran: molecular characterisation and recognition of a new species. Bull. Entomol. Res. 93: 527-535. Shahgudian, E.R. 1956. Notes on Anopheles marteri Senevet and Prunelle, 1927. Proc. R. Entomol. Soc. London. (A) 31: 71-75. Shahgudian, E.R. 1960. A key to the anophelines of Iran. Acta Med. Iranica 3: 38-48. Suguna, S.G., K. Gopala Rathinam, A.R. Rajavel, and V. Dhanda. 1994. Morphological and chromosomal descriptions of new species in the Anopheles subpictus complex. Med. Vet. Entomol. l8: 88-94. van Thiel, P.H. 1927. Sur l'origine des variations de taille de l'Anopheles maculipennis dans les Payes-Bas. Bull. Soc. Pathol. Exot. 20: 366-390. White, G.B. 1978. Systematic reappraisal of the Anopheles maculipennis complex. Mosq. Syst. 10: 13-44. Yaghoobi, F. 2004. Investigation on morphological and

276

Journal of Vector Ecology

December 2005

molecular characterization of Anopheles stephensi Liston in Sistan va Baluchistan and Hormozgan provinces. MSPH Thesis. Tehran University of Medical Sciences, Tehran. (in Persian). Zaim, M. and Z. Javaherian. 1991. Occurrence of Anopheles culicifacies A in Iran. J. Am. Mosq. Contr. Assoc. 7: 324326.

Zaim, M., S.K. Subbarao, A.V. Manouchehri, and A.H. Cochrane. 1993. Role of Anopheles culicifacies s.l. and Anopheles pulcherrimus in malaria transmission in Ghassreghand (Baluchistan) Iran. J. Am. Mosq. Contr. Assoc. 9: 23-26. Zolotarev, C. 1945. Anopheles maculipennis of northern Iran. Med. Parasitol. 14: 1-13.

Vol. 30, no. 2

Journal of Vector Ecology

195

Mosquito larvicidal activity of aqueous extracts of long pepper (Piper retrofractum Vahl) from Thailand

Uruyakorn Chansang1, Nayer S. Zahiri2, Jaree Bansiddhi3, Thidarat Boonruad3, Pratom Thongsrirak3, Jiranuch Mingmuang3, Nipa Benjapong1, and Mir S. Mulla2

Division of Medical Entomology, National Institute of Health, Department of Medical Sciences, Ministry of Public Health, 88/7 Tiwanon Road, Nonthaburi 11000, Thailand 2 Department of Entomology, University of California, Riverside, CA 92521, U.S.A. 3 Medicinal Plant Research Institute, Department of Medical Sciences, Ministry of Public Health, 88/7 Tiwanon Road, Nonthaburi 11000, Thailand Received 30 November 2004; Accepted 24 February 2005 ABSTRACT: Aqueous extracts of nine medicinal plants were bioassayed against larvae of Culex quinquefasciatus Say and Aedes aegypt (L.). Among these plants, the long pepper, Piper retrofractum Vahl (Piperaceae), showed the highest level of activity against mosquito larvae. To gain more information on larvicidal activity of P. retrofractum, fresh fruits of this plant were extracted in water and the extracts made into powder and bioassayed against 3rd and 4th instar larvae of Cx. quinquefasciatus and Ae. aegypti in the laboratory. Extracts of unripe (001/3) and ripe (002/3 and 001/4) fruits showed different levels of activity against Cx. quinquefasciatus larvae. Extracts 001/3 and 002/3 were equi-toxic to a Bacillus sphaericus resistant and susceptible strains, both from Thailand. The ripe fruit extract 002/3 was somewhat more active against Ae. aegypti than Cx. quinquefasciatus. Another ripe fruit extract (001/4) was much more toxic to both mosquito species. Diluted solutions of the solid extract (002/3) in distilled water lost their larvicidal activity upon aging. Loss of activity at 25°C was greater than that stored at 4°C, and greater in water than in acetone solution. Journal of Vector Ecology 30 (2): 195-200. 2005. Keyword Index: Aedes aegypti, Culex quinquefasciatus, larvicides, medicinal plants, Piper retrofractum. INTRODUCTION Humans have used plant parts, products, and metabolites in pest control since early historical times. Plants are the chemical factories of nature, producing many chemicals, some of which have medicinal and pesticidal properties. By using plant parts in early historical times and plant extracts and concentrated components in more recent times, man has been able to control certain pests with these remedies quite successfully. The current use and future potential of plants for pest control on farms and homes are detailed in an FAO document (FAO 1999). Most of the earlier information on the pesticidal uses of plants is anecdotal in nature rather than contained in published records. Wood (2003) lists some important phytochemical products such as pyrethrum, derris, quassia, nicotine, hellebore, anabasine, azadirachtin, dlimonene, camphor, and terpenes that have been used as insecticides. These are major groups of insecticides of plant origin that were used in developed countries before the advent of synthetic organic insecticides (Casida and Quistad 1998). More than 2,000 plant species have been known to produce chemical factors and metabolites of value in pest control programs (Ahmed et al. 1984), and among these plants, products of some 344 species have been reported to have a variety of activity against mosquitoes (Sukumar et al. 1991). A recent review by Shaalan et al. (2005) treated this aspect in detail. One of the most studied botanical species with good pesticidal attributes is the neem tree (Azadirachta indica A. Juss.), whose extracts have shown considerable activity and multiple modes of action against agricultural pests, forestry insects, and insects of public health (Schmutterer 2002, Mulla and Su 1999). Sukumar et al. (1991) published an extensive review of phytochemicals from plants with activity against mosquitoes. Members of the plant families Asteraceae, Cladophoraceae, Labiateae, Miliaceae, Oocystaceae, and Rutaceae possess various types of activity against many species of mosquitoes. Recently, Park et al. (2002) carried out a detailed laboratory study on extracts of fruits of Piper nigrum L. (Piperaceae) against larvae of Culex pipiens L., Aedes aegypti (L.), and Ae. togoi Theobald. Four active constituents were identified by spectroscopic analyses as isobutylamide alkaloids pellitorine, guineensine, pipercide, and retrofractamide A. They indicated that the compound most toxic to Cx. pipiens larvae was pipercide and that retrofractamide A had higher larvicidal activity against Ae. aegypti and Ae. togoi than pipercide. Mulla and Su (1999) reviewed neem products with activity and bio-efficacy against arthropods of medical and veterinary importance. They noted that a variety of neem components exhibited effects such as antifeedancy, growth regulation, ovicidal effects, reproduction suppression, and changes in biological fitness in many species of arthropods of medical and veterinary importance. In the genus Piper there are about 1,000 species with worldwide distribution and of these, 112 species have been studied for their phytochemistry and potential use of metabolites for pest control. So far, at

1

196

Journal of Vector Ecology

December 2005

least 611 active ingredients have been isolated and identified from different parts of Piper species (Dyer et al. 2004). To date, P. peepuloides Roxb. (Srivastava 1970), P. longum L. (Lee 2000, Yang et al. 2002), P. nigrum (Park et al. 2002), and another 15 species (Bernard et al. 1995) have been studied in the laboratory for mosquito larvicidal activity. However, the species P. retrofractum, which is common in Thailand and other countries in Southeast Asia, has not been investigated for bioactivity against mosquito larvae. The Institute of Medicinal Plant Research (IMPR) in the Ministry of Public Health, Thailand, has an ongoing research program for studying and promoting plant material for medicinal uses and developing promising products for traditional medicine and commercialization. The industrial development and medicinal uses of plants and plant products in many countries, especially in the Orient and Southeast Asia, represent a major activity in traditional health care programs. In this regard, significant advances have been made in the development and uses of phytomedicinal products. Piper retrofractum Vahl (Piperaceae), grown throughout Thailand, is listed as a source of herbal drugs in the Thai Herbal Pharmacopoeia (THP 2000). It is commonly called "Di pli" (The Forest Herbarium 2001). The plant is a climber, stem cylindrical, glabrous, with prominent nodes. Fruits are berries, broadly rounded, connate, and agnate to stalk of bract with 1 seed. The fruits turn from green to red when they are ripe (Figures 1A, 1B), and have a spicy taste (Backer and Van Den Brink 1963). The fruit of P. retrofractum in Thai traditional medicine textbooks is reported to be useful for treatment of bronchial asthma, bronchitis, muscle pain, and other maladies (Farnsworth and Bunyapraphatsara 1992). In Indonesia, this fruit is mixed in carminative and sudorific remedies (Perry and Metzger 1980). The principle active constituents of the fruit are the alkaloid piperine and volatile oils (THP 2000). Isobutyl amides, which are abundant in this genus, have been evaluated for their medicinal and commercial potential, and a number of compounds found in Piper fruits were reported to have pharmacological and biological activity (Banerji et al. 2002). Recently, the amide piperidine from P. longum fruit has shown activity against Ae. aegypti mosquito larvae (Yang et al. 2002), and isobutylamides in P. nigrum fruits showed larvicidal activity against three mosquito species (Park et al. 2002). To take advantage of the potential larvicidal activity as well as the increasing trend to use herbs as alternative remedies in several developing countries including Thailand, we conducted preliminary laboratory studies on nine medicinal plants against mosquito larvae in the laboratory. Among these plants, the extracts of P. retrofractum were found to be the most active plant product. Limited studies were also carried out on the stability of aqueous solutions of the extract in water and acetone at different temperatures. MATERIALS AND METHODS Plant and plant material All medicinal plants, including the fruits of P.

retrofractum, were collected from the nurseries or gardens of the Medicinal Plant Research Institute in Bangkok, Chantaburi and Chiang Mai Provinces, Thailand. Identification of P. retrofractum was made using the description and keys of Backer and Van Den Brink (1963) and Wu and Raven (1999) and compared with authentic specimens (Maxwell 74-748) at the Bangkok Herbarium (BK), Department of Agriculture, Ministry of Agriculture and Cooperatives, Thailand. A voucher specimen (Bansiddhi 47-1) was deposited at the Botanical Laboratory, Medicinal Plant Research Institute, Department of Medical Sciences, Ministry of Public Health, Thailand. Procedures for extraction and lyophylization For the preliminary testing of the nine species of plants, all plants were washed thoroughly and 50 g wet weight quantities of each plant were crushed with a mortar and pestle and mixed with 1 L of distilled water and allowed to stand for 24 h. The mixture was then filtered and the clear filtrate used for bioassays without obtaining dry solids. In preliminary tests, P. retrofractum showed the highest activity, and detailed studies on the extracts of P. retrofractum were then carried out against mosquito larvae. We found that the dried fruits were more effective than the fresh fruits on a weight volume basis, therefore for subsequent tests, plant material was dried. In addition to the effectiveness of dried fruits, it is easy to keep them in storage. For drying, fresh fruits in 1.8 kg quantities were cleaned and dried in an oven at 50ºC for 5 days. Piper retrofractum has volatile oils that are essential for activity of this plant, but these oils will be destroyed if the plant is dried at greater than 50ºC for 5 days. The dried fruit sample of 790 g (43.89 % yield of fresh fruits) was macerated and powdered in a blender and then saturated and mixed with 1.6 L of distilled water and kept for 24 h. The plant-water mixture was filtered to obtain a clear-brown filtrate. The filtrate was then lyophilized, yielding 37 g of powder (4.68 % yield by dried weight). The solids from the extract were kept in light-resistant and closed bottles and placed in the refrigerator (4º C) for a short time or the freezer (-20º C) for a longer period of time before being used. Laboratory bioassays The nine medicinal plant extracts were bioassayed against 3rd and 4th instar (essentially similar in their susceptibility) larvae of Cx. quinquefasciatus and Ae. aegypti. Among these, P. retrofractum showed the highest level of activity. We therefore studied the extracts of P. retrofractum in more detail against larvae of these two species. In order to gauge the activity of the fruit extract, three extracts of this plant, prepared at different times (to ascertain variability of different batches), were tested. We tested for differences in unripe and ripe fruit extracts to determine if the bioactive principles are present in sufficient quantities in both ripe and unripe fruits. Solids of extracts of unripe and ripe fruits (001/3 and 002/3, respectively) were received on June 25, 2004, and the third extract of ripe fruit 001/4 (fresh extract) was received on October 10, 2004 from Thailand. The solid materials of extracts were bioassayed against late 3rd and early 4th instar (similar in susceptibility) larvae of

Vol. 30, no. 2

Journal of Vector Ecology

197

Bsph-susceptible Cx. quinquefasciatus colony (Bsph S-Th) and also against a highly Bsph-resistant colony of this species (Bsph R-Th), both strains from Thailand. The Bsph R strain was included to see if there is any indication of crossresistance. The susceptible and resistant colonies were established in our laboratory from Sirichai District and Wat Pikul, Nonthaburi Province, Thailand in 2001, respectively. Additionally, the activity of P. retrofractum fruit extracts was determined against Ae. aegypti larvae colonized from dry eggs on filter paper brought to our laboratory from Thailand in 2003. For culturing these colonies, egg rafts (8­10) from each sub-colony of Cx. quinquefasciatus and pieces of filter paper with conditioned and dried Ae. aegypti eggs were placed in a separate enamel pan (50 x 25 x 7 cm) containing 2 L of tap water and 2 g of rabbit pellets (Brookhurst Mill, Riverside, CA) added as larval food. Larvae were reared under a photoperiod of 14:10 (L:D) at 27±2º C. The larvae were reared to the 3rd and 4th instar and allowed to pupate in the pans. The pupae were collected from the pans and transferred into cups in tap water and placed in screen cages (23 x 30 x 23 cm) where the adults emerged. The adults were provided with 10% sucrose solution, and on day 5 after emergence, females were allowed to feed overnight on restrained 2­5 day-old chicks (Animal Use Protocol Number A-S 0205015-2, University of California, Riverside, CA). Third and 4th instar (with similar susceptibility) larvae from these colonies were used in laboratory bioassays. Bioassays were conducted, according to Zahiri et al. (2002). Briefly, 25 late 3rd and early 4th instar (similar in susceptibility) larvae were transferred to 100 ml of distilled water in 120 ml disposable waxed paper cups. One drop of larval food (2 g of ground up rabbit pellets in 20 ml distilled water) was added per cup. The cups were treated with 4­5 concentrations of the 1% lyophilized solids of the plant extract dissolved in water, acetone, or ethanol depending on the test

material and susceptibility of a given strain of mosquitoes, yielding between 10 to 95% mortality in 48 h. Each concentration was replicated 4 times in each test. In preliminary testing we noted that the activity of the extracts as aqueous solution declined on storage and we therefore deemed it was necessary to use organic solvents acetone and ethanol. The bioassay cups were held at 25° C and larval mortality readings taken 48 h post-exposure. Moribund larvae were counted as dead. If mortality in the control exceeded 5%, the test was discarded. Each treatment was replicated four times and the tests run on two or three different days. Stock suspensions (1%) of the fruit extract powder of P. retrofractum were prepared in distilled water, acetone, and ethanol by vigorously shaking 0.2 g of the plant extract in 20 ml of the solvent in screwcap glass vials. Piper retrofractum extract did not dissolve in alcohol, so the alcohol solutions were not tested. To investigate the stability of P. retrofractum extract solutions (1%), fresh solutions were prepared in distilled water and acetone. To study the stability of extract, the aqueous solution was stored in a refrigerator (4° C), room temperature (25° C), and the extract in acetone was stored at room temperature only, for various periods of time. Bioactivities of these solutions were determined by larval bioassays. The dose-response data were subjected to probit regression analysis using POLO-PC (LeOra Software 1987). The slope, lethal concentrations in mg/l (LC50 and LC90), and their 95% confidence intervals were calculated using this software. RESULTS AND DISCUSSION In preliminary comparative studies, P. retrofractum (long pepper) with eight other medicinal plants were tested against Cx. quinquefasciatus and Ae. aegypti larvae. Among these

Table 1. Activity of medicinal plant extracts dissolved in distilled water against 3rd and 4th instar Cx. quinquefasciatus and Ae. aegypti. Plants Piper retrofractum Vahl Curcuma longa L. Syzygium aromaticum (L.) Merr. & Perry Allium sativum L. Spilanthes paniculata Wall. ex DC. Gynura pseudochina DC. Zingiber montanum (Koen.) Link ex Dietr. Rhinacanthus nasutus (L.) Kurz Cleome viscosa L. Family Piperaceae Zingiberaceae Myrtaceae Alliaceae Compositae Compositae Zingiberaceae Acanthaceae Capparidaceae Mosquito species Cx. quinquefasciatus Ae. aegypti Cx. quinquefasciatus Ae. aegypti Cx. quinquefasciatus Ae. aegypti Cx. quinquefasciatus Ae. aegypti Cx. quinquefasciatus Ae. aegypti Cx. quinquefasciatus Ae. aegypti Cx. quinquefasciatus Ae. aegypti Cx. quinquefasciatus Ae. aegypti Cx. quinquefasciatus Ae. aegypti LC50 (mg/l) 135 79 9,530 3,880 2,455 2,656 2,102 4,369 5,101 7,065 11,208 9,208 6,166 9,851 5,124 9,681 8,994 8,550 LC90 (mg/l) 1,079 229 27,429 10,612 3,598 4,135 3,079 8,379 11,149 16,595 33,023 30,023 10,864 17,684 16,473 22,620 14,962 11,858

198

Journal of Vector Ecology

December 2005

Table 2. Activity of three extracts of Piper retrofractum 1% dissolved in distilled water against 3rd and 4th instar Cx. quinquefasciatus and Ae. aegypti, using 48 h mortality. Mosquito species LC50 mg/l*(95% CL) S-Th R-Th LC90 mg/l *(95% CL) Slope ± SE

Cx. quinquefasciatus

Extract 001/3 ­ unripe fruit 49.4a (45.2-58.1) 102.4a (88.1-119.3) a 40.2 (35.2-48.9) 94.9a (66.5-252.6) Extract 002/3 ­ ripe fruit 34.8ab (20.0-64.4) 115.3a (75.5-259.3) ab 33.2 (22.3-48.9) 102.8a (64.3-345.3) 18.3b (10.9-32.8) 56.1ab (31.7-322.2)

2.93 ± 0.41 3.67 ± 0.45

Cx. quinquefasciatus

S-Th R-Th

2.02±0.22 2.01±0.24 2.64±0.22

Ae. aegypti

Cx. quinquefasciatus Ae. aegypti

S-Th

Extract 001/4 ­ ripe fruit 4.33c (2.47-6.05) 24.68 bc (17.39-50.71) 6.30c (5.35-7.38) 27.94 bc (21.20-41.43)

1.69±0.23 1.98±0.20

* Significant at 0.05 level, calculated separately for each test.

Table 3. Activity of extract (002/3) of ripe Piper retrofractum fruit 1% dissolved in acetone against 3rd and 4th instar larvae and ratio of LC50 and LC90 of distilled water to acetone (dw/ac) against mosquito larvae. Mosquito species Cx. quinquefasciatus S-Th Ae. aegypti *Significant at 0.05 level. LC50 mg/l* (95% CL) 15.33a (7.3-26.9) 12.45a (10.6-14.5) LC50 dw/ac 2.3 LC90 mg/l* (95% CL) 73.06a (37.9-469.9) 50.12a (39.4-69.7) LC90 dw/ac 1.6 Slope ± SE

1.89±0.16

1.5

1.1

2.12±0.20

Table 4. Loss of activity of 6-day-old 1% solution of ripe Piper retrofractum fruit extract (002/3) dissolved in distilled water stored at 4° C and 25° C against 3rd and 4th instar Cx. quinquefasciatus larvae (S-Th). Age of solution (days old) 0 6 at 4° C 6 at 25° C *Significant at 0.05 level. LC50 mg/l* (95% CL) 17.7a (4.4-29.7) 62.1b (45.6 ­ 74.7) 147.7b (22.4 ­ 237.6) LC90 mg/l 89.9a (30.9-157.7) 170.8a (142.4 ­ 230.5) 485.4b (345.5 ­ 855.2) Slope ± SE

2.24±0.22 2.55±0.19 2.95±0.28

Vol. 30, no. 2

Journal of Vector Ecology

199

A

B

u

r r r

u

u

Figure 1. The medicinal plant, Piper retrofractum Vahl, (Piperaceae) that was extracted and tested against larvae of Aedes aegypti and Culex quinquefascistus. A: Botanical detail including the young fruit. B: Unripe (u) and ripe (r) fruits of P. retrofractum.

nine plants, only P. retrofractum showed a high level of activity against larvae of both mosquitoes, the others having low levels of activity (Table 1). Three extracts of long pepper found to be the most active were then tested, the 001/3 from unripe fruit, and the 002/3 and 001/4 from ripe fruits. All extracts dissolved readily in distilled water and they showed variable but high levels of activity against the larvae of Bsph S-Th and R-Th of Cx. quinquefasciatus (Table 2). The Bsph R and susceptible strains

showed similar levels of susceptibility to both unripe and ripe fruit extracts. Aedes aegypti larvae were somewhat more susceptible to the extract than Cx. quinquefasciatus. The fresh extract of ripe fruit (001/4) showed much higher activity than the old extracts, against both Cx. quinquefasciatus and Ae. aegypti. This extract (001/4) showed 11.4 and 8.1 times higher activity than other two extracts (001/3 and 002/3) against Bsph S-Th Cx. quinquefasciatus at the LC50 level respectively, and 4.1 and 4.7 times higher activity at LC90 levels respectively. The slope of the regression lines for the 001/4 extract was not as steep as the other two extracts (Table 2). The data in Table 2 show that different batches of extracts of different age could exhibit variable levels of activity in laboratory bioassays. It is also evident that both ripe and unripe fruits contain similar levels of bioactive compounds. It is of great interest that long pepper bioactives have good activity against Bsph R mosquito larvae. Only two extracts of P. retrofractum (002/3 and 001/4) were tested against Ae. aegypti (Table 2). Both species of mosquitoes were more susceptible to the 001/4 extract. A possible explanation for this difference could be the degree of ripeness of the fruit or the storage stability and shipping mode of fruits. Also the time between harvest and extraction could be a factor in the difference in activity of the extracts, factors which were not examined here. A solution of extract 002/3 (ripe fruit) in acetone was highly toxic to both Cx. quinquefasciatus and Ae. aegypti. Solutions in acetone against Cx. quinquefasciatus at the LC50 and LC90 levels were 2.3 and 1.6 times more active than distilled water, respectively (Table 3), while against Ae. aegypti they were 1.5 and 1.1 times more active. To determine the storage stability of P. retrofractum extract in solution, we made a 1% solution of extract 002/3 (ripe fruit) in distilled water. The activity of a fresh solution of the extract was compared with 6-day-old solutions stored in a refrigerator and at room temperature (25° C). It was noted that the color of the 6-day-old solutions at room temperature turned lighter than the fresh or refrigerated ones. In bioassays, the fresh solution of the extract showed a high level of activity against Cx. quinquefasciatus compared with the solution stored for 6 days in the refrigerator and room temperature (Table 4). From this experiment, it appeared that storage of dilute solutions suffered a marked loss of activity. The higher

Table 5. Effect of storage on activity of 1% of Piper retrofractum extract (002/3) dissolved in distilled water and acetone and aged at room temperature (25° C), against 3rd and 4th instar Cx. quinquefasciatus larvae (S-Th). Age of solution (days old) Fresh (0 days) 2 days 4 days 8 days LC50 mg/l(95% CL) 21.6a (6.8-31.4) 71.1b (52.1-89.7) 237.5c (183.5-315.4) 274.4 c(152.8 ­ 438.2) LC90 mg/l(95% CL) 1% aqueous solution * 67.2a (49.2-143.7) 151.3a (112.4-356.9) 502.8b (366.7-917.6) 937.9b (547.2-1566.6) Slope ± SE

2.61 ± 0.36 3.91 ± 0.42 3.93 ± 0.30 2.40 ± 0.23

1% acetone solution * Fresh (0 days) 18.1a (14.8-21.5) 75.9a (56.2-193.8) 3 days 40.4b (35.9-44.9) 97.8a (81.7-158.8) 6 days 61.1c (56.6-65.1) 101.0a (92.2-125.6) *Significant at 0.05 level, calculated separately for each solution.

2.05 ± 0.27 3.33 ± 0.41 4.18 ± 0.48

200

Journal of Vector Ecology

December 2005

storage temperature (25° C) appeared to cause greater degradation of the active compounds in the solution. Additional studies on the storage stability of P. retrofractum extract 002/3 solutions (1%) in distilled water and acetone were carried out. The aqueous solution was stored at room temperature (25° C) for 0, 2, 4, and 8 days. On storage at room temperature, the activity of the aqueous solution declined progressively to a low level (Table 5). The test on the storage stability of P. retrofractum extract in acetone at 25° C for up to 6 days also indicated a negative relationship with storage duration of solutions and their levels of activities; older solutions had much lower activity as compared to the fresh ones. The LC50 and LC90 values of the fresh solution were much lower than those of solutions stored for 3 and 6 days at room temperature (Table 5). The degradation rate of P. retrofractum extract dissolved in acetone was lower than that in distilled water. This study shows that lyophilized aqueous extracts of P. retrofractum have good potential for the control of mosquito larvae, especially those are resistant to the microbial control agent B. sphaericus. At the local and village levels, however, aqueous extracts that extract lower quantities of bioactive compounds are more cost effective than the extracts made by organic solvents. It has been shown that the pure compounds pipericide, dibutylamides, and other amide bioactive compounds in the species of Piper have much higher activity than the crude extracts. The use and application of pure compounds or refined crude extracts will require intensive research on extractability, efficacy, toxicology, and environmental acceptability. Crude plant extracts are more cost effective in community-based control programs and may be employed in localized situations where mosquitoes have become tolerant to synthetic or microbial larvicides. REFERENCES CITED Ahmed, S., M. Graivge, J.W. Hylin, W.C. Mitchell, and J.A. Listinger. 1984. Some promising plant species for use as pest control agents under traditional farming system. In: Schmutterer, H. and K.R.S. Ascher (eds.) Proc. 2nd International Neem Conference, Rauischolzhausen, Germany, 1983. pp. 565-580 GTZ, Eschborn, Germany. Backer, C.A. and R.C. Bakhuizen Van Den Brink. 1963. Flora of Java, Vol.1, Gronigen, N.V.P. Noordhoff. 167-173 pp. Banerji, A., M. Sarkar, R. Datta, P. Sengupta, and K. Abraham. 2002. Amides from Piper brachystachyum and P. retrofractum. Phytochemistry 59: 897-901. Bernard, C.B., H.G. Krishnamurty, D. Chauret, T. Durst, B.J.R. Philogene, P. Sanchez, C. Hasbun, L. San Roman, L. Poveda, and J.T. Arnason. 1995. Insecticidal defenses of Piperaceae from the Neotropics. J. Chem. Ecol. 21: 801814. Casida, J.E. and G.B. Quistad. 1998. Golden age of insecticide research: Past, Present or Future? Annu. Rev. Entomol. 43: 1­16. Dyer, L.A., J. Richard, and C.D. Dodson. 2004. Isolation,

synthesis, and evolutionary ecology of Piper amides. In: L.A. Dyer and A.D.N. Palmer (eds.) A model genus for studies of phytochemistry, ecology, and evolution. pp. 117-139. Kluwer/Plenum. New York. FAO. 1999. Use and potential of wild plants in farm households. FAO information division. Food and Agriculture organization of the United Nations. http:// www.fao.org/docrep/003/w8801E/w8801eoo.htm. Farnsworth, N.R. and N. Bunyapraphatsara. 1992. Thai Medicinal Plants: recommended for primary health care system. Mehidol University, Bangkok. pp. 197­198. Forest Herbarium, The. 2001. Thai Plant Names (Tem Smitinand revised edition). Royal Forest Department, Bangkok. 417 pp. Lee, S.E. 2000. Mosquito larvicidal activity of pipernonaline, a piperidine alkaloid derived from long pepper, Piper longum. J. Am. Mosq. Contr. Assoc. 16: 245­247. LeOra Software. 1987. POLO­PC: A user's guide to probit or logit analysis. LeOra Software: Berkeley, CA. Mulla, M.S. and T. Su. 1999. Activity and biological effects of neem products against arthropods of medical and veterinary importance. J. Am. Mosq. Contr. Assoc. 15: 133­152. Park, I.K., S.G. Lee, S.C. Shin, J.D. Park, and Y.J. Ahn. 2002. Larvicidal activity of isobutylamides identified in Piper nigrum fruits against three mosquito species. J. Agric. Food Chem. 50: 1866­1870. Perry, L.M. and J. Metzger. 1980. Medicinal plants of East and Southeast Asia. MIT Press, Cambridge. 314 pp. Schmutterer, H. 2002. The neem tree and other meliaceous plants. 2nd ed. 893 pp. Neem Foundation, Mumbai, India. Shaalan, E., D. Canyon, M.W. Farid, H. Abdel-Wahab, and A.H. Mansour. 2005. A review of botanical phytochemicals with mosquitocidal potential. Environmental International. (In press). Srivastava, J.B. 1970. Insecticide & larvicide activity in the extract of Piper peepuloides Royle (Piperaceae). Indian J. Exp. Biol. 8: 224 ­ 225. Sukumar, K., M.J. Perich, and L.R. Boombar. 1991. Botanical derivatives in mosquito control: A review. J. Am. Mosq. Contr. Assoc. 7: 210­237. THP, Department of Medical Sciences. 2000. Di-Pli. In: Thai Herbal Pharmacopoeia. Ministry of Public Health, Northaburi, Thailand. 2: 9-15. Wood, A. 2003. Compendium of pesticide common names: Insecticides. (http://www.alanwood.net/pesticides/ index.htm). Wu, Z. and P.H. Raven. 1999. Piperaceae. Flora of China. Science Press, Beijing. 4: 110­134. Yang, Y.C., S.G. Lee, H.K. Lee, M.K. Kim, and S.H. Lee. 2002. A piperidine amide extracted from Piper longum L. fruit shows activity against Aedes aegypti mosquito larvae. J. Agricult. Food Chem. 50: 3765­3768. Zahiri, N.S., T. Su, and M.S. Mulla. 2002. Strategies for management of resistance in mosquitoes to the microbial control agent Bacillus sphaericus. J. Med. Entomol. 39: 513­520.

Vol. 30, no. 2

Journal of Vector Ecology

295

Effects of sub-lethal concentrations of synthetic insecticides and Callitris glaucophylla extracts on the development of Aedes aegypti

Essam Abdel-Salam Shaalan1, Deon Vahid Canyon2 , Mohamed Wagdy Faried Younes3, Hoda Abdel-Wahab1, and Abdel-Hamid Mansour1

2

Zoology Department, Aswan Faculty of Science, South Valley University, Aswan, Egypt Anton Breinl Centre, School of Public Health, Tropical Medicine and Rehabilitative Sciences, James Cook University, Townsville, Australia 3 Zoology Department, Faculty of Science, Menoufia University, Shebin El-Kom, Egypt Received 9 March 2005; Accepted 28 July 2005

1

ABSTRACT: Synthetic and botanical insecticides can have a profound effect on the developmental period, growth, adult emergence, fecundity, fertility, and egg hatch, resulting in effective control at sub-lethal concentrations. This paper investigated sub-lethal concentrations of fenitrothion, lambda-cyhalothrin, and Callitris glaucophylla Joy Thomps. & L.P. Johnson (Cupressaceae) extract to characterize their effects on the development of Aedes aegypti L. (Diptera: Culicidae) mosquito larvae. The LC25, LC50, and LC75 (four replicates) were used for each synthetic insecticide and the LC25 and LC75 (four replicates) were used for C. glaucophylla. Observations of larval mortality, duration of larval stage, pupal mortality, duration of pupal stage, adult emergence, sex ratio, and malformations were recorded over 14 days. A dose-response effect was observed for all insecticides. Although C. glaucophylla extract doses were higher than synthetic insecticide doses, the LC75 treatment outperformed synthetics by completely prohibiting adult emergence. Consequently, this botanical is recommended for field application either in combination with synthetic or natural insecticides or alone. Journal of Vector Ecology 30 (2): 295-298. 2005. Keyword Index: Botanical, phytochemical, insecticide, mosquito, sub-lethal. INTRODUCTION In toxicity studies, the gentle dose-response slope observed for botanicals over 24 h renders many of them unusable by economic standards despite them causing significant mortality at sub-lethal concentrations. Besides toxic larvicidal activity, botanical extracts have been shown to induce pupicidal activity, effects on larval and pupal duration, and often reduce adult emergence. For instance, an extract of Callistemon lanceolatus induced concentration-dependent mortalities in juvenile Culex quinquefasciatus (Mohsen et al. 1990), while an extract of Ipomoea carnea caused mortality and disrupted the development and growth of Anopheles stephensi (Saxena and Sumithra 1985). Neem oil and neem seed kernel extract markedly reduced the percentage of pupation and adult emergence of An. stephensi (Murugan et al. 1996). Botanical extract-induced malformations, particularly larval-pupal intermediates and half-ecdysed adults, were common (Al-Sharook et al. 1991, Jayaprakasha et al. 1997 and Karmegam et al. 1997). Synthetic insecticides also induce sub-lethal effects, however, these can be unpredictable. For instance, three pyrethroids (d-phenothrin, d-allethrin, and tetramethrin) reduced Aedes aegypti egg production, while only dphenothrin and d-allethrin reduced blood engorgement (Liu et al. 1986). Topically applied dieldrin caused dose-dependant effects on feeding and affected egg-laying capacity in Ae. aegypti, but progeny were unaffected (Duncan 1963). Sublethal concentrations of the botanical, Callitris glaucophylla, induced significant larvicidal activity against Ae. aegypti compared with fenitrothion and lambda-cyhalothrin (Shaalan et al., unpublished data). Thus, in the absence of highly toxic natural botanical compounds, the growth or emergence inhibiting activity of a botanical phytochemical may be essential to its uptake by the insecticide industry. Indeed, the rational application of exceptional phytochemicals may not only lead to new IPM strategies but may inhibit the development of insect resistance to existing synthetic insecticides. This study investigated the effects of sub-lethal concentrations of a liquefied refrigerant gas extract of C. glaucophylla, fenitrothion, and lambda-cyhalothrin on the development of Ae. aegypti mosquitoes and determined a concentration that led to satisfactory control. MATERIALS AND METHODS Test mosquitoes Aedes aegypti were obtained from a colony initiated from mosquitoes collected in 2002 from Townsville, Australia. The colony of mosquitoes were maintained at conditions of 27 ± 2 Cº and 70 % ± 5 R.H. under 14L - 10D cycles. Ae. aegypti larvae were kept in plastic buckets half filled with tap water and fed on goldfish flakes. Water in rearing containers was refreshed every two days. Male and female adult mosquitoes were maintained on a 10 % sugar solution while female adults were also provided the opportunity to feed on rat blood.

296

Journal of Vector Ecology

December 2005

Test insecticides Technical grades of the organophosphorous insecticide fenitrothion (96.8 %) and the pyrethroid insecticide lambdacyhalothrin (90.99 %) were provided by Nufarm Ltd (North Victoria, Australia). Liquefied refrigerant gas extract of C. glaucophylla was supplied by Michael Kennedy, Department of Primary Industries, Queensland, Australia (details of extraction awaiting IP protection). It is possible, but not considered likely, that interaction with the compressed refrigerant gas solvent caused or catalyzed chemical changes in the extractive compounds, just as this could possibly happen with conventional solvents and is known to happen during steam distillation. The refrigerant was eliminated from the extract by spontaneous distillation as the pressure was reduced. Minute quantities could remain in the crude extract, but given the volatility of the refrigerant, would be readily lost during application of the extract to the test material during the screening process. Any residual effect on extract activity has not been evaluated. Bioassays Insect growth regulator testing instructions (WHO 1996) were followed to investigate and determine the effects of sublethal doses of test insecticides on Ae. aegypti larvae. Larvae were subjected to two to three sub-lethal concentrations (LC25, LC50, and LC75) in glass beakers. The sub-lethal concentrations for fenitrothion were 0.0025 mg/l (LC25), 0.0044 mg/l (LC50), and 0.0062 mg/l (LC75); for lambda-cyhalothrin they were 0.00004 (LC25), 0.00015 (LC50), and 0.00026 mg/l (LC75); and for C. glaucophylla they were 2.6 mg/l (LC25) and 14.7 mg/l (LC75). All test chemicals were diluted in ethanol. One ml of stock solution was added to 99 ml of de-ionized water. Controls received 1 ml of ethanol only. Four replicates of 25 newly molted 4th instar larvae for each concentration were conducted. For accurate determination of the sub-lethal

effects, the larval and pupal mortality as well as adult emergence were recorded daily up to emergence of the adults or death of the last larva or pupa. Due to the long duration of the test, larvae were provided with food at 2-day intervals during the test period. From the overall results of the test, percentages of both emerged and dead pupae and percentage of adult emergence, sex ratio, larval duration, pupal duration, average developmental period, and growth index were determined. Growth index was calculated according to Saxena and Sumithra (1985): Growth Index (GI) = percentage adult emergence / average developmental period (days). Data analyses were performed using a one-way ANOVA in SPSS version 12.0.1 and significant differences were determined at P<0.05. RESULTS Effects of sub-lethal concentrations of C. glaucophylla, fenitrothion, and lambda-cyhalothrin on juvenile and adult mosquitoes are shown in Tables 1 and 2. Fenitrothion results indicated little significant difference between controls and the LC25 dose and between the LC50 and LC75 doses. At the two higher doses, virtually all measured variables were significantly different from controls. Larval survival and adult emergence were reduced 10-fold, total mortality ranged from 88-98 % and growth was significantly limited. Lambda-cyhalothrin results indicated more of a doseresponse with LC25s often being significantly different from controls. Only LC50 and LC75 doses, however, significantly reduced larval survival and adult emergence. Total mortality was high at all doses and the GI was comparable to that observed for fenitrothion. C. glaucophylla results also showed clear dose-response

Table 1. Effects of sub-lethal concentrations of fenitrothion, lambda-cyhalothrin, and Callitris glaucophylla extract on juvenile Aedes aegypti development and mortality. Early 4th instar larvae were exposed continuously until adulthood.

Average larval period (days) 5.0±0.1c 4.3±0.8

bc

Insecticide Fenitrothion

LC 0 25 50 75

Larval mortality (%) 0.0±0.0a 13.0±8.1

a

Average pupal period (days) 2.0±0.3b 1.9±0.2

b

Pupal mortality (%) 6.5±3.8* 10.5±4.4 10.0±4.3 4.0±2.8 2.0±2.3a 53.5±18c 53.5±10c 23.0±6.2

b

Total average development (days) 6.98±0.3b 6.16±0.7b 3.31±2.0a 1.55±1.1a 13.20±0.6d 8.36±1.2c 5.27±0.4b 1.82±0.5a 6.9±0.6c 4.5±0.6b 0.38±0.5a

2.8±1.6ab 1.3±0.5a 7.1±0.1d 6.6±0.2c 4.8±0.2b 1.7±0.5

a

74.5±19.2b 92.0±5.2c 0.5±1.0a 7.0±3.5a 33.5±3.0b 69.5±9.3

c

0.6±0.5a 0.2±0.3a 6.1±0.6c 1.8±1.0b 0.4±0.4a 0.1±0.1

a

Lambdacyhalothrin

0 25 50 75

Callitris glaucophylla

0 25 75

4.2±0.4c 3.4±0.6b 0.6±0.5a

2.0±2.8a 9.5±4.4b 95.0±2.6c

2.7±0.2c 1.1±0.7b 0.0±0.0a

9.0±2.0a 25.0±12.5b 3.0±2.6a

Values for the same insecticide followed by a different letter within the same column are statistically different (P<0.05). * Group not significant.

Vol. 30, no. 2

Journal of Vector Ecology

297

Table 2. Effects of sub-lethal concentrations of fenitrothion, lambda-cyhalothrin, and Callitris glaucophylla extract on adult Aedes aegypti development and mortality. Early 4th instar larvae were exposed continuously until adulthood.

Insecticide Fenitrothion LC 0 25 50 75 Lambdacyhalothrin 0 25 50 75 Callitris glaucophylla 0 25 75 Adult mortality (%) 0.0±0.0* 1.5±1.9 0.5±1.0 0.5±1.0 4.0±5.7b 13.5±3.0

ab a a

Adult emergence (%) 95.5±3.8c 72.5±13.3

b a a

Emergent females 54.0±9.4c 39.0±8.2

b a a

Emergent males 39.5±7.7b 33.5±6.4

b a a

Malformations 0.0±0.0* 2.5±1.9 3.3±3.6 1.0±2.0 0.0±0.0a 0.0±0.0 0.5±1.0 5.6±3.0

a a

Growth Index 6.72±0.5b 5.86±0.6b 1.21±1.3a 0.47±0.7a 7.1±0.3 c 3.0±1.5b 1.1±0.9a 0.7±0.9a 11.8±1.2b 7.7±6.0b 0.0±0.0a

11.5±17.8 2.5±3.8

4.0±5.4 2.0±4.0

7.5±12.4 0.5±1.0

94.0±6.9c 26.5±15.9

b a a

55.5±8.9c 20.0±10.5

b a a

43.0±14.1b 6.5±6.8 1.0±2.0 0.5±1.0

a a a

7.25±6.7 0.5±1.0

6.0±49 1.5±1.9

5.0±4.2 1.0±1.2

b

7.5±5.3a 30.5±19.2

b

81.5±5.3c 35.0±26.1

b

45.0±2.6c 22.5±16.0

b

36.5±3.0c 12.5±10.1

b

0.0±0.0* 0.0±0.0 2.0±2.3

0.0±0.0a

0.0±0.0a

0.0±0.0a

0.0±0.0a

Values for the same insecticide followed by a different letter within the same column are statistically different (P<0.05). *Group not significant. Growth Index = Adult emergence (%) / Average developmental period (days).

in most measured categories. The LC75 produced exceptional results, killing all larvae and pupae and preventing any emergence. Overall, there was no significant difference in the sex of emergent mosquitoes and malformations, for the most part, were uncommon and not related to dose. Mortality experienced by emergent adults was not significant except in the C. glaucophylla LC25 dose where a third perished. DISCUSSION The tested synthetic insecticides and botanical extract induced a wide range of sub-lethal effects on larval mortality, larval duration, pupicidal activity, pupal duration, adult emergence, sex ratio, adult mortality, and malformation. The insecticide concentrations estimated to cause 25, 50, and 75 % larval mortality in 24 h did not produce expected mortalities with LC25 doses consistently causing lower mortality and LC 75 doses causing higher mortality in fenitrothion and C. glaucophylla. The Probit estimation of LC25 and LC75 doses are less accurate unless large data sets are used, which is why publications typically report LC50s and not LC90s. Pupal mortality did not exhibit a linear relationship with the applied sub-lethal concentrations but was more associated with lower doses. Pupal mortality was lower than controls at the LC 75 dose in fenitrothion and C. glaucophylla, but lambda-cyhalothrin was more effective against pupae. Adult mortality results did not present anything of note. Dead adults were mostly half-ecdysed adults. Total mortality was consistently positively correlated with insecticide concentrations and the duration of exposure (Marcard et al. 1986). Beside immediate toxic larvicidal effects, all insecticides significantly reduced the average larval period compared to controls and, to a large extent, with each other. Larvae were

observed to pupate faster as their environment increased in toxicity. This is clearly a self-preservation mechanism since the pupal form is less susceptible to the environment. All concentrations markedly disrupted pupal duration except for the fenitrothion LC25 dose. Consequently, the average development period (a factor in the Growth Index formula) was consistently negatively correlated with insecticide concentrations and the duration of exposure. This effect can vary, however, with some researchers showing no effect on the larval and pupal developmental periods (Saxena et al. 1993, Sharma and Saxena 1994) and other researchers showing prolongation of the larval and pupal developmental periods (Karmegam et al. 1997, Saxena and Yadav 1983, Zebitz 1984, Saxena and Sumithra 1985, Mwangi and Rembold 1988, Robert and Olson 1989, Mohsen et al. 1990a, b ; Pushpalatha and Muthukrishnan 1995, Pushpalatha and Muthukrishnan 1999). In another study, Melia volkensii was observed to prolong the lifespan of An. arabiensis larvae but not the pupal period (Mwangi and Mukiama 1988). Conversely, Supavarn et al. (1974) reported on 11 of 36 botanicals that significantly inhibited pupal development while only a few botanicals affected larval development. Successful adult emergence is conversely proportional with the insecticide concentration and larval mortality. Of note, C. glaucophylla completely inhibited adult emergence compared to fenitrothion (2 ­ 8 %) and lambda-cyhalothrin (2 ­ 4 %) at the LC75 dose. This effect is expected since several studies have shown that botanical extracts either reduce or inhibit adult emergence. For instance, Descurainia sophia inhibited Cx. quinquefasciatus emergence (Mohsen et al. 1990b) and Tagetes erectes significantly reduced adult emergence in An. stephensi (Sharma and Saxena 1994). Changes in the sex ratio of emergent adults tended towards favoring females, however, results were not significantly different. This is not always the case, since Robert

298

Journal of Vector Ecology

December 2005

and Olson (1989) found a change in the sex ratio towards more males in Cx. quinquefasciatus after sub-lethal exposure propoxur and resmethrin. The percentage of malformations was very low for all insecticides. The only observed abnormalities were larvalpupal intermediates, half-ecdysed adults, and adults with malformed wings. These morphogenetic abnormalities are commonly caused by botanical extracts and are thought to result from a disturbance to growth regulating hormones (Zebitz 1984; Mwangi and Mukiama 1988; Pereira and Gurudutt 1990; Saxena et al. 1993). The growth indices of larvae treated at LC50 and LC75 doses were markedly shorter than controls and LC25 doses for all insecticides with negligible difference between controls and LC25 doses. Similar results were obtained by Saxena and Sumithra (1985) and Saxena et al. (1993), who found that the GI of mosquitoes treated with Annona squamosa alkaloids was longer in controls. In conclusion, for most measured developmental effects, the response was dose and exposure duration dependant. Significant developmental effects were observed for fenitrothion, lambda-cyhalothrin, and the botanical, C. glaucophylla. The latter induced responses at a LC75 dose that were exceptional and worthy of consideration for field trials pending non-target assessment. Acknowledgments We are grateful to Dr. Michael Kennedy, Department of Primary Industries, Queensland, Australia, for providing us with the C. glaucophylla extract. REFERENCES CITED Al-Sharook, Z, K. Balan, Y. Jiang, and H. Rembold. 1991. Insect growth inhibitors from two tropical meliaceae. Effects of crude seed extracts on mosquito larvae. J. Appl. Entomol. 111: 425-430. Duncan, J. 1963. Post-treatment effects of sublethal doses of dieldrin on the mosquito Aedes aegypti L. Ann. Appl. Biol. 52: 1- 6. Jayaprakasha, G.K, R.P. Singh, J. Pereira, and K.K. Sakariah. 1997. Limonoids from Citrus reticulata and their moult inhibiting activity in mosquito Culex quinquefasciatus larvae. Phytochemistry. 44: 843-846. Karmegam, N., M., Sakthivadivel, V. Anuradha, and T. Daniel. 1997. Indigenous plant extracts as larvicidal agents against Culex quinquefasciatus Say. Biores. Technol. 59: 137-140. Liu, W., R.G. Todd, and E.J. Gerberg. 1986. Effect of three pyrethroids on blood feeding and fecundity of Aedes aegypti. J. Am. Mosq. Contr. Assoc. 2: 310- 313. Marcard, M., C.P.W. Zebitz, and H. Schmutterer. 1986 The effect of crude methanol extracts of Ajuga spp. on postembryonic development of different mosquito species. J. Appl. Entomol. 101: 146-154. Mohsen, Z.H., A.L.M. Jawad, B.M. AL-Chalabi, and A. ALNaib. 1990a. Biological activity of Callistemon

lanceolatus against Culex quinquefasciatus. Fitoterapia. 61: 270-274. Mohsen, Z.H., A.L.M. Jawad, M. Al-Saadi, and B.A. Al-Naib. 1990b. Mosquito larvicidal and ovipositional activity of Descurania sophia extract. Intl. J. Crude Drug Res. 28: 77-80. Murugan, K., R. Babu, D. Jeyabalan, N.S. Kumar, and S. Sivaramakrishnan. 1996. Antipupational effect of neem oil and neem seed kernel extract against mosquito larvae of Anopheles stephensi (Liston). J. Entomol. Res. 20: 137-139. Mwangi, R.W and T.K. Mukiama. 1988. Evaluation of Melia volkensii extract fractions as mosquito larvicides. J. Am. Mosq. Contr. Assoc. 4: 442-447. Mwangi, R.W. and H. Rembold. 1988. Growth-inhibiting and larvicidal effects of Melia volkensii extracts on Aedes aegypti larvae. Entomol. Exp. Appl. 46: 103-108. Pereira, J. and K.N. Gurudutt. 1990. Growth inhibition of Musca domestica L. and Culex quinquefasciatus (Say) by (levo)-3-epicaryoptin isolated from leaves of Clerodendron inerme (Gaertn) (Verbenaceae). J. Chem. Ecol. 16: 2297-2306. Pushpalatha, E. and J. Muthukrishnan. 1995. Larvicidal activity of a new plant extracts against Culex quinquefasciatus and Anopheles stephensi. Ind. J. Malariol. 32: 14-23. Pushpalatha, E. and J. Muthukrishnan. 1999. Efficacy of two tropical plant extracts for the control of mosquitoes. J. Appl. Entomol. 123: 369-373. Robert, L.L. and J.K. Olson. 1989. Effects of sublethal dosages of insecticides on Culex quinquefasciatus. J. Am. Mosq. Contr. Assoc. 5: 239-246. Saxena, R.C., V. Harshan, A. Saxena, and P. Sukumaran. 1993. Larvicidal and chemosterilant activity of Annona squamosa alkaloids against Anopheles stephensi. J. Am. Mosq. Contr. Assoc. 9: 84-87. Saxena, S.C. and L. Sumithra. 1985. Laboratory evaluation of leaf extract of a new plant to suppress the population of malaria vector Anopheles stephensi Liston (Diptera: Culicidae). Curr. Sci. 54: 201-202. Saxena, S.C. and Yadav, R.S. 1983. A new plant extract to suppress the population of yellow fever and dengue vector Aedes aegypti (Diptera: Culicidae). Curr. Sci. 52: 713-715. Sharma, M. and R.C. Saxena. 1994. Phytotoxicological evaluation of Tagetes erectes on aquatic stages of Anopheles stephensi. Ind. J. Malariol. 31: 21-26. Supavarn, P., R.W. Knapp, and R. Sigafus. 1974. Biologically active plant extracts for control of mosquito larvae. Mosq. News. 34: 398-402. World Health Organization. 1996. Report of the WHO informal consultation on the evaluation and testing of insecticides. CTD/WHOPES/IC/96. 69 pp. Zebitz, C.P.W. 1984. Effect of some crude and azadirachtinenriched neem (Azadirachta indica) seed kernel extracts on larvae of Aedes aegypti. Entomol. Exp. Appl. 35: 1116.

Vol. 30, no. 2

Journal of Vector Ecology

299

Chemical detection of the predator Notonecta irrorata by ovipositing Culex mosquitoes

Leon Blaustein1,2, Jonathan Blaustein1, and Jonathan Chase2

1

Community Ecology Laboratory, Institute of Evolution, University of Haifa, Haifa, 31905, Israel 2 Department of Biology, Washington University, Saint Louis, MO 63130, U.S.A. Received 22 March 2005; Accepted 20 May 2005

ABSTRACT: We tested the oviposition response of Culex mosquitoes to the predator Notonecta irrorata in an outdoor artificial pool experiment employing equal numbers of control and predator pools. There was a strong oviposition avoidance by Culex of Notonecta pools; 83% of egg rafts were found in control pools during the period in which Notonecta were present. After removing Notonecta, mosquitoes continued to avoid ovipositing in the former Notonecta pools for two additional days suggesting a predator-released kairomone as the cue used by the mosquitoes to detect the presence of this predator. Journal of Vector Ecology 30 (2): 299-301. 2005. Keyword index: Oviposition, predation risk, Notonecta irrorata, kairomone. INTRODUCTION Natural selection should favor female mosquitoes, whose aquatic larvae are vulnerable to predation, that are able to detect aquatic predators and avoid ovipositing in sites with high risk of predation. A small but growing body of literature indicates that this is the case for some mosquito-predator systems. Experiments have shown evidence for oviposition avoidance of aquatic predators in the following mosquitopredator combinations: Anopheles punctipennis-bluegill sunfish (Petranka and Fakhoury 1991); Culex pipiens complex-mosquitofish (Angelon and Petranka 2002), Culex pipiens quinquefasciatus-Notonecta hoffmani (Chesson 1984); Culex laticinctus- Notonecta maculata (Kiflawi et al. 2003), Culiseta longiareolata-Anax imperator (Stav et al. 1999; Stav et al. 2000), C. longiareolata-N. maculata (Blaustein et al. 2004) and C. longiareolata-Anisops sardea (Eitam et al. 2002). In some cases (Angelon and Petranka 2002; Blaustein et al. 2004), but not all (Stav et al. 2000), one important cue used by the mosquitoes was predatorreleased kairomones. In this study, we experimentally examined the possible oviposition habitat selection in response to risk of predation for another prey-predator combination: Culex species and the backswimmer Notonecta irrorata Uhler. We also experimentally determined whether any oviposition avoidance was due to a predator-released kairomone. N. irrorata is common in woodland pools near St. Louis, MO, where this study took place (L. Blaustein personal observation). It is a general predator that can potentially have large impacts on pond community structure (Wilbur 1997). Mosquito densities have been shown to be lower in experimental plots containing N. irrorata (Neri-Barbosa et al. 1997), but the contributions of this reduction due to direct consumptive predation and due to behavioral avoidance (i.e., oviposition habitat selection) have not been investigated. MATERIALS AND METHODS On 7 September 2004 (day 0), eight circular red plastic tubs (45 cm diameter, 35 cm height) were set up in a residential area of St Louis, MO under a thick, mixed tree canopy of Carya cordiformis, Ulmus americana, and Alanthus altissima. Inter-pool distances ranged from 0.5 to 2 m. On the same day, 15 cm (volume = 24 L) of tap water and 60 g of dry leaf litter (consisting of a mixture of the three canopy species) were added to each pool. We counted and removed mosquito (Culex) egg rafts for two days (days 1-2) prior to adding Notonecta. This "preNotonecta period" allowed us to determine if the pools randomly assigned to treatment and control were, by chance, different in attracting oviposition by the mosquitoes. On day 2, we added to each Notonecta pool two adult N. irrorata collected from nearby natural ponds three days earlier. They were deprived of food during these three days prior to adding them to the experimental pools. This density was high enough to elicit strong oviposition avoidance responses by some mosquitoes to other Notonecta species in Israel (e.g., Eitam and Blaustein 2004). We did not observe any colonization of predators into the experimental pools. The absence of predator colonization was expected given considerable distance (>1 km) from any ponds (Wilcox 2001). Egg rafts collected the morning after Notonecta introduction (day 3) were excluded from analysis, an a priori decision because of the possibility that Notonecta kairomones could not sufficiently build up by the first night. The next four days (days 4-7) constituted the Notonecta period in which egg rafts were counted on the first, second, and fourth day of this period. As this short period demonstrated a highly significant effect (see Results), we terminated this part of the experiment. We then removed Notonecta from the pools and followed oviposition over the next six days (days 8-13). This "PostNotonecta period" allowed us to assess whether any effects of Notonecta on oviposition remained even after the predators

300

Journal of Vector Ecology

December 2005

Table. 1. Analysis of variance (pre-Notonecta period) and repeated measures analyses of variance (Notonecta and postNotonecta periods) comparing pools assigned to predator and control pools. Statistically significant P-values (<0.05) are in bold.

Period Pre-Notonecta Notonecta Post-Notonecta

Treatment d.f. F 1,6 0.40 1,6 17.82 1,6 0.60

P 0.552 0.006 0.468

Time d.f. -1,6 2,12

F -5.21 2.26

P -0.062 0.146

Treatment x Time d.f. F P ---1,6 0.01 0.923 2,12 6.52 0.012

were no longer present. Fewer egg rafts in former Notonecta pools would indicate a predator-released kairomone (Blaustein et al. 2004). Because egg rafts were not always counted every day, we analyzed data as egg rafts per pool per two days. A oneway analysis of variance (ANOVA) was performed on egg raft data during the pre-Notonecta period. During the Notonecta and post-Notonecta periods, we analyzed data as repeated measures ANOVA. Data were natural-log transformed (ln[y+1]) due to violations of homogeneity of variance. Mauchly's criterion test for sphericity indicated that no adjustments to degrees of freedom were necessary in within-subjects factors. RESULTS Of the six egg rafts collected in control pools, three were Culex pipiens and three were Culex restuans. Likewise, there were three of each species for the six egg rafts collected from predator pools. We thus refer to the egg raft counts generically as "Culex" egg rafts. During the pre-Notonecta period, Culex egg raft

abundance was not significantly different in predator and control pools (Table 1, Figure 1). During the Notonecta period, egg raft abundance was considerably lower in the predator pools; 83% of the egg rafts were found in control pools during this period (Table 1, Figure 1). Upon removing the Notonecta, egg raft abundance was still considerably lower in the former predator pools during the first two days but not during four and six days after predator removal (Treatment x time interaction: P=0.012; Figure 1). DISCUSSION We demonstrated that Culex mosquitoes can detect and avoid the predatory backswimmer Notonecta irrorata in an artificial pool experiment. Because we identified larvae to species level from only a small fraction of egg rafts, and both C. pipiens and C. restuans were present, we can only be certain that at least one of the species responded to the predator when ovipositing. Growing evidence suggests that many mosquitoes can detect and avoid some, but not all, predators. Of the two species ovipositing in this experiment, at least some members of the C. pipiens complex have been shown to avoid

8

Notonecta added Notonecta removed Control Notonecta

Number of Culex egg rafts per pool per 2 days

7

6

5

4

3

2

1

0 1-2 4-5 6-7 8-9 10-11 12-13

Days after pools inundated

Figure 1. Mean Culex egg rafts per pool per two-day periods. Pools were created and filled with water on day 0. Notonecta irrorata were added on day 2 and removed on day 7. Error bars are + one standard error.

Vol. 30, no. 2

Journal of Vector Ecology

301

mosquitofish (Angelon and Petranka 2002) and another notonectid species (Chesson 1984) when ovipositing. The effects of predators on oviposition habitat selection in C. restuans have not been investigated, although this mosquito has been shown to selectively oviposit in response to food level and conspecific density (Reiskind and Wilson 2002). We also demonstrated that the cue is a chemical. The cue remained detectable for two and possibly three days after removal of the predator itself. Blaustein et al. (2004) demonstrated that a Notonecta released kairomone, under similar experimental conditions, remained detectable by Culiseta longiareolata for seven-eight days. This mosquito cannot chemically detect other predators such as odonates and urodeles (Stav et al. 2000, Blaustein 1999, Blaustein unpublished data). We suspect that the chemical(s) released by different backswimmer species and detected by different mosquitoes in California (Chesson 1984), Israel (Eitam et al. 2002, Blaustein et al. 2004), and now the midwestern U.S.A. is (are) the same or similar, although this remains to be determined. These demonstrations of oviposition habitat selection by mosquitoes in response to predators have implications for how we experimentally assess the effects of aquatic predators on mosquito populations and the consequences of this behavior on mosquito populations (Spencer et al. 2002). Experiments assessing effects of predators on mosquito larvae that assume random oviposition when selective oviposition occurs likely overestimates the effect of the predator on mosquito populations. The kairomone, once chemically characterized, may be used as a mosquito control method. However, how effective a mosquito management method the application of such a kairomone to mosquito breeding habitats might be remains to be determined with empirical works and modeling. Acknowledgments We thank Linda McCuiston and Eric Larsen for verifying mosquito and backswimmer identifications, respectively. Comments by Avi Eitam, Marc Klowden, and an anonymous reviewer improved the manuscript. The study was supported by Israel Science Foundation Grant 600/02 awarded to LB. REFERENCES CITED Angelon, K.A. and J.W. Petranka. 2002. Chemicals of predatory mosquitofish (Gambusia affinis) influence selection of oviposition site by Culex mosquitoes. J. Chem. Ecol. 28: 797-806. Blaustein, L. 1999. Oviposition habitat selection in response to risk of predation: Consequences for populations and community structure. In: S. P. Wasser (ed.) Evolutionary processes and theory: Modern perspectives. pp. 441456. Kluwer Academic Publishers.

Blaustein, L., M. Kiflawi, A. Eitam, M. Mangel, and J.E. Cohen. 2004. Oviposition habitat selection in response to risk of predation: Mode of detection and consistency across experimental venue. Oecologia 138: 300-305. Chesson, J. 1984. Effect of notonectids (Hemiptera: Notonectidae) on mosquitoes (Diptera: Culicidae): Predation or selective oviposition? Environ. Entomol. 13: 531-538. Eitam, A., L. Blaustein, and M. Mangel. 2002. Effects of Anisops sardea (Hemiptera: Notonectidae) on oviposition habitat selection by mosquitoes and other dipterans and on community structure in artificial ponds. Hydrobiologia 485: 183-189. Eitam, A. and L. Blaustein. 2004. Oviposition habitat selection by mosquitoes in response to predator (Notonecta maculata) density. Physiol. Entomol. 29: 188-191. Kiflawi M, L. Blaustein, and M. Mangel. 2003. Predationdependent oviposition habitat selection by the mosquito Culiseta longiareolata: a test of competing hypotheses. Ecol. Lett. 6: 35-40. Neri-Barbosa, J.F., H. Quiroz-Martinez, M.L. RodriguezTovar, L.O. Tejada, and M.H. Badii. 1997. Use of Bactimos briquets (B.t.i. formulation) combined with the backswimmer Notonecta irrorata (Hemiptera: Notonectidae) for control of mosquito larvae. J. Amer. Mosq. Contr. Assoc. 13: 87-89. Petranka, J.W. and K. Fakhoury. 1991. Evidence of a chemically-mediated avoidance response of ovipositing insects to bluegills and green frog tadpoles. Copeia 1: 234-239. Reiskind, M.H., E.T. Walton, and M.L. Wilson. 2004. Nutrient-dependent reduced growth and survival of larval Culex restuans (Diptera: Culicidae): Laboratory and field experiments in Michigan. J. Med. Entomol. 41: 650-656. Spencer, M., L. Blaustein, and J.E. Cohen. 2002. Oviposition habitat selection by mosquitoes (Culiseta longiareolata) and consequences for population size. Ecology 83: 669679. Stav, G., L. Blaustein, and J. Margalith. 1999. Experimental evidence for predation risk sensitive oviposition by a mosquito, Culiseta longiareolata. Ecol. Entomol 24: 202207. Stav, G., L. Blaustein, and Y. Margalith. 2000. Influence of nymphal Anax imperator (Odonata: Aeshnidae) on oviposition by the mosquito Culiseta longiareolata (Diptera: Culicidae) and temporary pool community structure. J. Vector Ecol. 25: 190-202. Wilbur, H. M. 1997. Experimental ecology of food webs: Complex systems in temporary ponds - The Robert H. MacArthur Award Lecture. Ecology 78: 2279-2302. Wilcox, C. 2001. Habitat size and isolation affect colonization of seasonal wetlands by predatory aquatic insects. Israel J. Zool. 47: 459-475.

302

Journal of Vector Ecology

December 2005

Chemical composition and anti-mosquito potential of rhizome extract and volatile oil derived from Curcuma aromatica against Aedes aegypti (Diptera: Culicidae)

Wej Choochote , Dana Chaiyasit, Duangta Kanjanapothi1, Eumporn Rattanachanpichai, Atchariya Jitpakdi, Benjawan Tuetun, and Benjawan Pitasawat

Department of Parasitology, Faculty of Medicine, Chiang Mai University, Chiang Mai 50200, Thailand 1 Chulabhorn Research Institute, Chiang Mai 50200, Thailand Received 23 March 2005; Accepted 23 May 2005 ABSTRACT: Crude rhizome extracts and volatile oils of Curcuma aromatica were evaluated for chemical composition and anti-mosquito potential, including larvicidal, adulticidal, and repellent activities against the Aedes aegypti mosquito. Chemical identification achieved by GC/MS analysis revealed that xanthorrhizol, 1H-3a, 7­methanoazulene and curcumene at 35.08 and 13.65%, 21.81 and 30.02%, and 13.75 and 25.71%, were the main constituents in hexane extracts and volatile oils, respectively. Volatile oil of Cu. aromatica possessed a significantly higher larvicidal activity against the 4th instar larvae of Ae. aegypti than that of hexane extracts, with LC50 values of 36.30 and 57.15 ppm, respectively. In testing for adulticidal activity, on the other hand, hexane-extracted Cu. aromatica (LC50: 1.60 g/mg female) was found to be slightly more effective against female Ae. aegypti than volatile oil (LC50: 2.86 g/mg female). However, the repellency of these two products against Ae. aegypti adult females differed significantly. The hexane-extracted Cu. aromatica, with a median complete protection time of 1 h (range = 1-1.5 h) when applied at a concentration of 25%, appeared to have significantly higher repellency than that of distillate oil (0.5 h, range = 0-0.5 h). The different results obtained from both products of Cu. aromatica were probably due to variety in quantity and type of active ingredients as well as the biological and physiological characteristics that differed between both developmental stages of mosquitoes, larvae, and adults. Journal of Vector Ecology 30 (2): 302-309. 2005. Keyword Index: Curcuma aromatica, Aedes aegypti, chemical composition, larvicidal, adulticidal, repellents.

INTRODUCTION Dengue and dengue hemorrhagic fever are endemic and becoming increasingly serious public health problems in many tropical and subtropical developing countries. Approximately two-fifths of the world's population is now at risk of dengue transmission according to the World Health Organization (WHO 2003). In Thailand, it is estimated that more than 50,000 people suffer annually from dengue attacks, and the number of reported cases continue to rise (Annual Reports 2003-2004). Exacerbated by urbanization, increasing population movement, and lifestyles that contribute to the proliferation of artificial larval habitats of the mosquito, the worsening epidemiological trends seem likely to increase (Corbel et al. 2004). In the absence of a vaccine, the eradication or control of the main mosquito vector, Aedes aegypti, is regarded essentially as the only option available in preventing and controlling dengue. Insecticide applications, although highly efficacious against target species such as mosquitoes, pose a substantial hazard to a variety of animal life and the environment in the form of biomagnification. The rising drawbacks of synthetic substances, which include high cost, possible health risks, and environmental pollution, has contributed to a revived interest in naturalistic agents and methods for controlling mosquito vectors. Furthermore, development of insect resistance to conventionally synthetic insecticides (WHO 1995, Ansari et

al. 2000), and even biopesticides such as Bacillus sphaericus (Tabashnik 1994), has been increasingly reported, while evolution of the resistance to plant-derived compounds has rarely been (Sharma et al.1992). Although considerable efforts have focused on botanicals for potentially useful products as commercial insecticides (Arnason et al. 1989, Sukumar et al. 1991, Wink 1993), few plant products have reached the market and their efficacy is still lower than that of currently used synthetics. Nevertheless, plants remain a potential alternative that constitutes bioactive phytochemicals, which are expected to have a high chance of supplanting or replacing conventional insecticides. Insecticides of plant origin are expected to be target selective and biodegradable to nontoxic compounds, leading to fewer harmful effects on humans and other mammals and more environmental safety when compared to synthetic compounds (Jeyabalan et al. 2003, Prabakar and Jebanesan 2004). Plant-derived materials are also potentially suitable for use in the continuation of integrated mosquito control programs (Alkofahi et al. 1989) because they minimize the accumulation of harmful residues in the environment. Curcuma aromatica Salisb (family Zingiberaceae), a traditional Chinese herb, is now widely cultivated and used as a traditional herbal drug in India, China, and Southeast Asia. In Thailand, the rhizome and roots of Cu. aromatica are frequently used in cosmetics and spas for skin nourishment. Pharmacological study on Cu. aromatica volatile oil reveals various medical activities such as promotion of blood

Vol. 30, no. 2

Journal of Vector Ecology

303

circulation to remove blood stasis and treatment of cancer (Shi et al. 1981). Cu. aromatica oil consists of several major anti-tumor active ingredients such as elemici n, curcumol, and curione (Yang et al. 1996, Dong et al. 1997) and has been reported to have inhibitory effects on the proliferation of hepatoma in mice (Wu et al. 2000). Few studies on the potential of Cu. aromatica against mosquitoes have been reported. In our search for new bioactive products against mosquito vectors, we previously presented the remarkable repellency of ethanolic extract of Cu. aromatica, both in laboratory and field studies (Pitasawat et al. 2003). Although no adverse effects on skin condition occurred in these studies, a brown stain could be found on the skin after application which easily soiled clothing and was considered a drawback for this application. However, in addition to the repellent action of the ethanolic-extracted Cu. aromatica, its cosmetic acceptability is also worth consideration. Investigations of other products that originate from Cu. aromatica, which are extracted differently and provide an effective and practical use as agents, are of interest. With the aim of recommending Cu. aromatica as a probable source for developing mosquito control compounds, this study deals with plant preparation and extraction, chemical identification and investigation of repellency, and other anti-mosquito effects including larvicidal and adulticidal activities against Ae. aegypti mosquitoes. MATERIALS AND METHODS Extraction of Cu. aromatica Rhizomes of the aromatic turmeric, Cu. aromatica, were obtained from E.A.R. Samunpri, a commercial supplier in Chiang Mai province. The voucher specimen, PARA-CU-002/ 2, was deposited at the Department of Parasitology, Faculty of Medicine, Chiang Mai University. The extraction of plant materials was divided into two categories: extraction by volatile solvent (hexane) yielding crude extract and extraction by steam distillation for volatile oil. Shade-dried rhizomes of Cu. aromatica (1 kg) were powdered with an electrical blender, then successively extracted three times by maceration, with 3 liters of hexane at room temperature for 7 days. After suction filtering through a Büchner funnel, the hexane filtrates were separated, concentrated, and dried by a rotary evaporator at 40° C until the solvent was completely eliminated, and then they were lyophilized to yield hexane extract. Six hundred g of dried and finely ground Cu. aromatica was extracted by steam distillation for volatile oil. The liquid formed, together with volatile oil, was collected in a separating funnel. The mixture was allowed to settle for 1 day, after which the water (lower) layer was slowly drawn off until only the oil layer remained. The volatile oil was dried over anhydrous sodium sulfate and then collected and kept in a light-protected bottle. Both hexane extracts and volatile oils were investigated for chemical composition and anti-mosquito potential. Analysis of chemical composition Cu. aromatica samples of both hexane extract and essential oil were qualitatively analyzed for chemical

composition by gas chromatography coupled with mass spectrometry using a GC/MS-MSD instrument (6890 Agilent Technologies-Hewlett Packard model 5973 EI). Gas chromatography conditions were as follows: injection of 0.1 l of Cu. aromatica sample (hexane extract or volatile oil), with a split ratio of 100:1 for hexane extract and 400:1 for oil; capillary column HP 19091J-433 HP-5 -5% phenyl methyl siloxane (30 m x 0.25 mm x 0.25 m, film thickness). The carrier gas was helium (1.0 ml/min). The oven temperature program was initially 70° C, then rising to 245° C (10° C/ min, 10.5-min hold). The injector and detector temperatures were 50° C and 230° C, respectively. Percentage of the identified compound was computed from a total ion chromatogram (TIC). Mosquitoes Laboratory reared mosquito colonies of Aedes aegypti established from the Department of Parasitology, Faculty of Medicine, Chiang Mai University, Chiang Mai province, northern Thailand were utilized in this study. The colonies had been reared and maintained continuously for over 20 generations at 25-30° C and 80-90% relative humidity under a photoperiod of 14:10 h (light/dark) in an insectarium free of exposure to pathogens, insecticides, or repellents. Larvae were fed on finely ground dog biscuit. Adults were provided continuous access to 10% sucrose and 10% multivitamin syrup. Rats were used as a source of blood meal for egg producing females. Two developmental stages, larvae and adult females, were continuously available for the experiments. Human volunteers Healthy adult volunteers of both sexes (age 16-28 years; weight 43-65 kg) were recruited from the students and staff of the Department of Parasitology, Faculty of Medicine, Chiang Mai University. The volunteers for the repellency test had no history of dermatological disease or allergic reaction to arthropod bites or repellents. All volunteers signed an informed consent form after having received a full explanation of the test objectives, procedures, foreseeable risks to subjects, and remedial arrangements. The Research Ethics Committee of the Faculty of Medicine, Chiang Mai University, formally approved the research proposal. Larvicidal bioassay Larvicidal activity of hexane extract and volatile oil derived from Cu. aromatica rhizome against Ae. aegypti was assessed by using a slightly modified version of the WHO standard method (WHO 1981). Batches of 25 early 4th instar larvae of Ae. aegypti were exposed in distilled water treated with a series concentration of each plant product dissolved in dimethylsulphoxide (DMSO) and kept in a 500-ml enamel bowl. Four replicate sets were tested simultaneously with a final total of 100 larvae for each concentration. The toxicity of each plant sample was determined, with at least six concentrations ranging from 20 to 120 ppm that provided a range of 0-100% mortality. Solutions containing distilled water and DMSO, but without the plant sample, served as controls,

304

Journal of Vector Ecology percentage.

December 2005

while the untreated solution was distilled water only. Observations were performed immediately after treatment and then at intervals of 24 h. No food was provided to the larvae during the exposure period. Mortality and survival were monitored after 24 h of treatment. The moribund and dead larvae in four replicates were combined and expressed as a percentage mortality of each concentration. The larvae were considered dead or moribund if they were unresponsive within a reasonable period of time, even after gentle prodding with a needle. Experiments were performed at 25-30° C, replicated three times, and the result reported as the average of individual plant product. Adulticidal bioassay For adult mosquitoes, activity of the volatile oil or hexaneextracted Cu. aromatica was determined by topical application of the plant material to the adult female mosquitoes, following a minor modification of the WHO standard methods (WHO 1996). Non-blood-fed females (2-5 days old) were briefly anesthetized with carbon dioxide, weighed and placed on a cold plate standing under a dissecting microscope. Plant solutions (0.1 l) dissolved in acetone were dropped with a Hamilton digital syringe (700 series MICROLITER TM, Hamilton Company, U.S.A.) onto the upper part of the pronotum of each adult mosquito. Dosages were expressed in g of plant material per mg of mosquito body weight. In total, 25 individuals were used per dose, with at least six doses providing between 0 and 100% mortality. Controls were treated with 0.1 l of acetone alone. Anesthetized mosquitoes with no treatment were considered as untreated. After treatment, mosquitoes were immediately transferred to a clean plastic cup, maintained in an environmentally controlled room (27±3oC and 80%±10% RH), and provided with 10% sucrose and 10% multivitamin syrup ad libitum. Symptoms of treated adults were observed and recorded immediately and at timeintervals after application. Mortality counts were done after 24 h of a holding period. Mosquitoes lying at the bottom of the plastic cup and unresponsive to mechanical stimulation were counted as dead. For each plant product, the experiments were replicated eight times with mosquitoes from different rearing batches, and the pooled data were calculated in

Repellency bioassay The human-topical application technique (Schreck and McGovern 1989, WHO 1996), which was conducted to determine the repellent efficacy of the rhizome extract and volatile oil of Cu. aromatica against Ae. aegypti, a daytimebiting mosquito, was carried out between 08.00 h and 16.00 h in a 10x10x3 m room, at 25-30° C and relative humidity of 60-80%. Pure oil and an ethanolic formulation of Cu. aromatica hexane extract prepared at a concentration of 25% were used in the test. After cleaning with distilled water, each forearm of the volunteer was wrapped in a plastic sleeve attached with double-sided tape, and a cutout was aligned with a 3x10 cm area on the ventral portion of the forearm. Therefore, only a restricted area of skin was exposed to the mosquitoes, with the hand protected by a rubber glove. An aliquot of 0.1 ml of the test solution was applied evenly on the 30 cm2 exposed area of one forearm of each volunteer. The other forearm, acting as a control, was treated with absolute ethanol by the same procedure as that used for the test repellent. After application, the solution was allowed to dry on the skin for 1 min before the test arm was placed in a mosquito net cage (30x30x30 cm) containing 250 bloodstarved 5 to 7 day-old female mosquitoes for the first three min of every half-hour exposure. The mosquitoes that landed on the tested area and attempted to probe and imbibe blood were recorded. Exposure experiments continued at 30-min intervals until at least two bites occurred in a three-min period, or until a first bite occurred followed by a confirming bite (second bite) in the next observation period. During the experiment, successive introductions of the control arm were performed in the same manner before each insertion of the treated arm, in order to confirm and standardize the readiness of the mosquitoes to bite. The time between application of the plant sample and the first two consecutive bites was considered as the complete protection time, which is the criterion used to determine the repellent efficacy of a plant sample. Identical tests were replicated four times on different days on each subject of six human volunteers (three females, three males), providing a total of 24 tests performed for each

Table 1. Physical characters and percentage yields of Cu. aromatica products.

Cu. aromatica product Hexane extract Appearance Semi-solid Color Light brown Odor Cineolic-like Solubility Soluble in DMSO and acetone Volatile oil Fluid liquid Weak yellow Cineolic-like Soluble in DMSO and acetone 1.48 Yield (%) 6.77

Vol. 30, no. 2

Journal of Vector Ecology

305

product. Repellent products were applied randomly to each participant, and no one tested more than one preparation per day. Data management and statistical analysis The results of mosquitocidal bioassays were reported as percentage mortality, which should be obtained in not less than three mortality counts of between 10% and 90%. In cases where the mortality in the control of larvicidal or adulticidal tests ranged from 5-20%, the observed percentage mortality (%M) was corrected by Abbott's formula (Abbott 1925): %M = % test mortality -% control mortality x 100- % control mortality

100

When the control mortality was over 20%, the tests were discarded and repeated. The corrected percentage of mortality was subjected to a computerized log-probit analysis (Harvard Programming; Hg1, 2), providing the lethal dosage of 50, 95, and 99% (LD 50, LD 95, and LD 99) as well as their 95% confidence intervals (CI). Significance differences were determined by comparing the CI of each plant product. The median complete-protection time was used as a standard repellency measure of the plant samples against Ae. aegypti in the laboratory bioassay. The range of protection time was used to measure differences between plant products. RESULTS AND DISCUSSION Extraction of Cu. aromatica rhizomes by two distinct processes, hexane extraction and steam distillation, provided crude extract and volatile oil, respectively, with different physical characters and percentage yields, as shown in Table 1. The hexane-extracted Cu. aromatica, once lyophilized, was a semi-solid non-polar material and light brown in color. Cu. aromatica oil was less dense (0.87 g/ml) than water and exhibited a weak yellow color. Both products of Cu. aromatica

gave a pleasant odor similar to cineolic and a high solubility in DMSO and acetone. The percentage yield of the hexane extract of Cu. aromatica (6.77%: w/w) was higher than that of its volatile oil (1.48%: v/w) by more than four-fold. However, the process of solvent extraction requires a further separation step that would include the removal of solvent by using an evaporator and a lyophilizer in order to get a purified extract. In contrast, steam distillation is a simple technique that requires low-cost equipment. The selection of the extraction process best suited to each experiment should depend not only on the total yield obtained, but also other variables such as the quality and efficacy of the product and simplification of the preparation method. Chemical identification achieved by GC/MS analysis revealed the presence of three major compounds with varying amounts in both products of Cu. aromatica (Table 2). The main constituents of the hexane extract and volatile oil of Cu. aromatica, in terms of relative percentage of total area in the chromatogram, seemed to be xanthorrhizol, 1H-3a, 7­ methanoazulene and curcumene at 35.08 and 13.65%, 21.81 and 30.02%, and 13.75 and 25.71%, respectively. Compounds occurring in trace amounts are not shown in the table. Although the chemical profiles for the hexane extract and volatile oil of Cu. aromatica obtained from different extraction processes tended to be similar, the most abundant constituent of the hexane-extracted Cu. aromatica was xanthorrhizol, while that of the oil was 1H-3a, 7 ­methanoazulene. Testing for the anti-mosquito activity of these compounds, which clarifies their potential against mosquito vectors, has been of interest. The chemical composition of three essential oils that were derived from C. aromatica, cultivated in Japan (two samples) and India (one sample), and analyzed by GC and GC/MS were extremely different (Kojima et al. 1998). The major constituents in both oils from Japanese samples were curdione, germacrone, 1,8-cineole, (4S, 5S)-germacrone-4, 5-epoxide, -elemene, and linalool, whereas those of Indian oil were -

Table 2. Chemical composition of Cu. aromatica products as determined by combined gas chromatography and mass spectrometry.

Cu. aromatica product Hexane extract Peak 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 Chemical constituent NI curcumene NI 1H-3a, 7­methanoazulene NI xanthorrhizol NI NI camphor curcumene NI 1H-3a, 7­methanoazulene NI epicurzerenone germacrone xanthorrhizol Retention time (min) 11.67 12.06 12.28 12.42 13.59 15.18 17.80 23.80 7.40 12.08 12.29 12.45 13.09 13.60 14.70 15.18 Area (%)a 1.18 13.75 1.35 21.81 4.46 35.08 4.60 5.06 1.81 25.71 2.58 30.02 2.08 5.47 2.26 13.65

Volatile oil

NI: Not identified. aRelative percentage of total sample.

306

Journal of Vector Ecology

December 2005

Table 3. Larvicidal activity of Cu. aromatica products against 4th instar larvae of Ae. aegypti.

Cu. aromatica product (ppm) Hexane extract 40 50 60 70 80 100 Control Untreated Volatile oil 26.10 30.45 34.80 39.15 43.50 47.85 52.20 Control Untreated % Mortality (Mean±SE) Larvicidal activity (95% C.I., ppm) LC50 LC95 LC99 57.15 108.91 156.34 (54.59-59.72) (97.50-128.07) (132.15-200.60)

16.67±4.16 35.33±1.15 53.67±6.43 73.67±4.51 81.67±2.52 91.00±3.46 0 0 15.67±6.43 23.67±9.29 39.67±10.02 62.67±16.50 72.33±13.28 87.67±4.04 95.00±5.00 0 0

36.30 (35.16-37.45)

59.18 (54.84-65.78)

77.85 (69.42-91.44)

Table 4. Adulticidal activity of Cu. aromatica products against adult female Ae. aegypti.

Cu. aromatica product (g/mg female) Hexane extract 0.5 1.0 2.0 3.0 4.0 5.0 Control Untreated Volatile oil 0.87 1.74 2.61 3.04 3.48 4.35 Control Untreated % Mortality (Mean±SE) Adulticidal activity (95% C.I., g/mg female) LD50 LD95 1.60 10.05 (1.34-1.86) (6.36-24.91)

9.00±1.41 33.00±9.90 56.00±4.24 75.00±14.14 93.50±4.95 98.50±2.12 4.50±4.95 1.00±0.00 4.00±0.00 15.00±7.07 37.00±9.90 54.50±4.95 69.50±4.95 85.00±9.90 1.00±1.41 2.00±2.83

2.86 (2.71-3.02)

6.11 (5.36-7.41)

Vol. 30, no. 2

Journal of Vector Ecology Table 5. Repellency (median complete-protection time) of Cu. aromatica products against adult female Ae. aegypti.

Treatment Hexane-extracted Cu. aromatica (25%) Pure oil of Cu. aromatica Control (ethanol) Median complete-protection time (Range, h) 1.0 (1-1.5) 0.5 (0-0.5) 0

307

curcumene, ar-curcumene, xanthorrhizol, germacrone, champhor, and curzerenone. C. aromatica investigated in this study was grown in Thailand and its constituents seemed to be similar to those obtained from Indian oil. It should be pointed out, however, that chemical components within the same species may be different depending on the genetic characteristics of the plant and the conditions under which it was grown and harvested (Martins et al. 1997, Vieira and Simon 2000, Tawatsin et al. 2001). Mosquito larvicidal activities of two products, hexane extract and volatile oil, of Cu. aromatica against 4th instar larvae of Ae. aegypti are demonstrated in Table 3. No mortality was obtained in either the control or untreated group. The susceptibility of Ae. aegypti to graded concentrations of both Cu. aromatica products was dose dependent. When exposed to a higher dose, more larvae showed toxic symptoms that led to an increase in larval mortality. Essential oil of Cu. aromatica induced 100% larval mortality of Ae. aegypti after 24 h with a dosage of 52.2 ppm, whereas a higher dose (> 100 ppm ) of hexane extract was required to reach complete larval mortality. From these results, it appeared that Cu. aromatica volatile oil (LC50 and LC95 values: 36.30 and 59.18 ppm, respectively) possessed significantly higher larvicidal activity against the 4th instar larvae of Ae. aegypti than that of hexane extract (LC50 and LC95 values: 57.15 and 108.91 ppm, respectively) at the same LC levels. Although the symptoms experienced by larvae exposed to both products were similar, abnormal behavior caused by Cu. aromatica oil was observed in a shorter time (5-10 min), than that caused by hexane extract (20-30 min). Abnormal symptoms such as sluggishness and circular movement observed in treated larva indicated that the toxic effect of Cu. aromatica was probably on the neuromuscular system (Sakthivadivel and Thilagavathy 2003). From apparent results, considering the larvicidal potential and simplification of the preparation method, the distillate oil of Cu. aromatica was more likely appropriate than the hexane extract for use in developing mosquito larvicide agents. In testing for adulticidal activity by topical applications, hexane extract of Cu. aromatica (LD50: 1.60 181 g/mg female) showed a slightly higher, but statistically significant, toxicity against female Ae. aegypti compared to that of volatile oil (LD50: 2.86 g/mg female), as measured by LD50 (Table 4). However, the LD95 value obtained from hexane-extracted Cu. aromatica (LD95: 10.05 g/mg female) reversed and became slightly higher than that of volatile oil (LD50: 6.11 g/mg female). When comparing the percentage yield and

adulticidal potential of volatile oil and hexane extract, derived from Cu. aromatica, the latter appeared to be more suitable for developing natural agents for combat against adult mosquitoes. Since little work has been done on plant toxicity against adult Ae. aegypti or other mosquito species tested by the adult application method, the data on adulticidal activity of plant-derived materials against Ae. aegypti was insufficient for comparison with this study. The adulticidal activity of both products of Cu. aromatica investigated in this study is encouraging, but not as much as that of conventional synthetic insecticides. Corbel et al. (2004) reported the excellent intrinsic toxicity of some conventional insecticides against many mosquito species, as measured by topical application. The LD50 of bifenthrin, permethrin, and temephos, when tested against Ae. aegypti adults, was 0.077, 0.24, and 195 ng/mg female, respectively. When comparing the adulticidal activity of Cu. aromatica to these commonly used insecticides, this plant seemed much less toxic against female Ae. aegypti (more than 10-fold at the LD50 level). However, pyrethroid and organophosphate resistance is now widely developed in mosquitoes of medical importance (Corbel et al. 2004). The investigations on repellent revealed that both products of Cu. aromatica were found to exhibit slight repellent activity that differed in significance when applied to human volunteers against adult female Ae. aegypti (Table 5). The hexane-extracted Cu. aromatica, with a median complete protection time of 1 h (range = 1-1.5 h) when applied at a concentration of 25%, appeared to have significantly higher repellency than that of distillate oil (0.5 h, range = 00.5 h). These findings corresponded to the results obtained from an adulticidal study, where hexane extract of Cu. aromatica was more potent against the adult stage of Ae. aegypti than volatile oil. Pitasawat et al. (2003) revealed the pronounced repellency of ethanol-extracted Cu. aromatica against Ae. togoi with ED50 and ED95 values of 0.061 and 1.55 mg/cm2, respectively. It also provided biting protection for 3.5 h when applied at a concentration of 25%. The repellent activity of both products of Cu. aromatica tested in this study, hexane extract and volatile oil, seemed to be lower than that of ethanol-extracted Cu. aromatica. Although less effective, topical application of both products derived from Cu. aromatica did not induce dermal irritation. No serious harmful effects on the skin of the human volunteers were observed during three months of the study period or in the following three months, after which time observations ceased. In the case of hexane-extracted Cu. aromatica, however, a brown

308

Journal of Vector Ecology

December 2005

stain on the skin was found in a similar manner to that seen in those who applied ethanolic extract. Extraction processes for separating the active ingredients from the pigments or other coloring substances are required for developing Cu. aromatica­derived products as a practical agent. Formulated preparations to improve repellent efficacy, stability, and user acceptability are also needed. The variety in quantity, type of active ingredients in each product of Cu. aromatica and the biological and physiological characteristics, which differed between the developmental stages of mosquito, larva, and adult, were probably responsible for anti-mosquito activity. At present, however, the role and relationship of these multifactors are not known and warrants a more extensive study. Because of its significance in larvicidal, adulticidal, and repellent potential, Cu. aromatica can therefore be recommended as a probable source of some biologically active compounds useful in the development of potential alternatives for vector control, particularly in areas where mosquitoes are resistant to conventional insecticides. Acknowledgments The authors thank the individuals who served as subject volunteers. Our thanks also go to the staff members of the Department of Parasitology, Faculty of Medicine, and the Science and Technology Service Center, Chiang Mai University for their cooperation and chemical analysis of plant products, respectively. Acknowledgment is extended to the Faculty of Medicine Endowment Fund for Research Publication for its financial support in publishing this paper. REFERENCES CITED Abbott, W.S. 1925. A method of computing the effectiveness of an insecticide. J. Econ. Entomol. 18: 265-266. Alkofahi, A., J.K. Rupprecht, J.E. Anderson, J.L. Mclaughlin, K.L. Mikolajczak, and B.A. Scott. 1989. Search for new pesticides from higher plants. In: J.T. Arnason, B.J.R. Philogens, and P. Morand (eds.). Insecticides of Plant Origin. pp. 25-43. In: ACS Symp. Ser. 387. Am. Chem. Soc., Washington, D.C. Annual Reports. 2003-2004. Chiang Mai Provincial Public Health Office. Ministry of Public Health, Thailand. Ansari, M.A., P. Vasudevan, M. Tandon, and R.K. Razdan. 2000. Larvicidal and mosquito repellent action of peppermint (Mentha piperita) oil. Biores. Technol. 71: 267-271. Arnason, J.T., B.J.R. Philogens, and P. Morand. 1989. Insecticides of Plant Origin. ACS Symp. Ser. 387. Am. Chem. Soc., Washington, D.C. 213 pp. Corbel, V., S. Duchon, M. Zaim, and J.M. Hougard. 2004. Dinotefuran: a potential neonicotinoid insecticide against resistant mosquitoes. J. Med. Entomol. 41: 712-717. Dong, J.H., G.B. Cheng, and J.H. Hu. 1997. Isolation and differentiation of A-elemene from volatile oil of Curcuma wenyujin and its anti-cancer activity. Zhongcaoyao. 28: 13-14. Jeyabalan, D., N. Arul, and P. Thangamathi. 2003. Studies on

effects of Pelargonium citrosa leaf extracts on malarial vector, Anopheles stephensi Liston. Biores. Technol. 89: 185-189. Kojima, H., T. Yanai, and A. Toyota. 1998. Essential oil constituents from Japanese and Indian Curcuma aromatica rhizomes. Planta Med. 64: 380-381. Martins, E.R., V.W.D. Casali, L.C.A. Barbosa, and F. Carazza. 1997. Essential oil in the taxonomy of Ocimum selloi Benth. J. Braz. Chem. Soc. 8: 29-32. Pitasawat, B., W. Choochote, B. Tuetun, P. Tippawangkosol, D. Kanjanapothi, A. Jitpakdi, and D. Riyong. 2003. Repellency of aromatic turmeric Curcuma aromatica under laboratory and field conditions. J. Vector Ecol. 28: 234-240. Prabakar, K. and A. Jebanesan. 2004. Larvicidal efficacy of some Cucurbitacious plant leaf extracts against Culex quinquefasciatus (Say). Biores. Technol. 95: 113-114. Sakthivadivel, M. and D. Thilagavathy. 2003. Larvicidal and chemosterilant activity of the acetone fraction of petroleum ether extract from Argemone mexicana L. seed. Biores. Technol. 89: 213-216. Schreck, C.E. and T.P. McGovern. 1989. Repellents and other personal protection strategies against Aedes albopictus. J. Am. Mosq. Contr. Assoc. 5: 247-252. Sharma, R.N., A.S. Gupta, S.A. Patwardhan, D.S. Hebbalker, V. Tare, and S.B. Bhonde. 1992. Bioactivity of lamiaceae plants against insects. Indian J. Exp. Biol. 30: 244-246. Shi, J.H., C.Z. Li, and D.L. Liu. 1981. Experimental research on the pharmacology of Curcuma aromatica volatile oil. Zhong Yao Tong Bao. 6: 36-38. Sukumar, K., M.J. Perich, and L.R. Boobar. 1991. Botanical derivatives in mosquito control: a review. J. Am. Mosq. Contr. Assoc. 7: 210-237. Tabashnik, B.E. 1994. Evolution of resistance to Bacillus thuringiensis. Annu. Rev. Entomol. 39: 47-79. Tawatsin, A., S.D. Wratten, R.R. Scott, U. Thavara, and Y. Techadamrongsin. 2001. Repellency of volatile oils from plants against three mosquito vectors. J. Vector Ecol. 26: 76-82. Vieira, R.F. and J.E. Simon. 2000. Chemical characterization of basil (Ocimum spp.) found in the markets and used in traditional medicine in Brazil. Econ. Bot. 54: 207-216. Wink, M. 1993. Production and application of phytochemicals from an agricultural perspective. In: T.A. van Beek and H. Breteler (eds). Phytochemistry and agriculture. pp. 171-213. Clarendon Press, Oxford, UK. World Health Organization. 1981. Instructions for determining the susceptibility or resistance of mosquito larvae to insecticides. WHO/VBC/81.807. Geneva, Switzerland. World Health Organization. 1995. International travel and health vaccination requirement and health advice. WHO, Geneva, Switzerland. World Health Organization. 1996. Report of the WHO informal consultation on the evaluation and testing of insecticides. WHO/CTD/WHOPES/IC/96.1. Geneva, Switzerland. World Health Organization. 2003. Dengue (online, access in 03/06/2003). Available at http://www.who.int/inf-fs/en/

Vol. 30, no. 2

Journal of Vector Ecology

309

fact117.html. Wu, W.Y., Q. Xu, L.C. Shi, and W.B. Zhang. 2000. Inhibitory effects of Curcuma aromatica oil on proliferation of hepatoma in mice. Wld. J. Gastroenterol. l6: 216-219.

Yang, H., X.P. Wang, L.L. Yu, and S. Zheng. 1996. The antitumor activity of elemene is associated with apoptosis. Zhonghua Zhong Liu Za Zhi.18: 169-172.

Vol. 30, no. 2

Journal of Vector Ecology

277

Evaluation of the present dengue situation and control strategies against Aedes aegypti in Cebu City, Philippines

Milagros M. Mahilum 1,2, Mario Ludwig 1, Minoo B. Madon3, and Norbert Becker1,2

German Mosquito Control Association (KABS/GFS), Ludwigstrasse 99, 67165 Waldsee, Germany 2 ICYBAC GmbH, Georg-Peter-Süß-Str.1, 67346 Speyer, Germany 3 Greater Los Angeles County Vector Control District, 12545 Florence Avenue, Santa Fe Springs, CA 90670-3919, U.S.A.

1

Received 11 November 2004; Accepted 22 July 2005 ABSTRACT: The present dengue situation and methods to control Aedes aegypti larvae in Cebu City, Philippines, were evaluated for the development of an integrated community-based dengue control program. The study included the detection of dengue infection among Filipino patients, surveying mosquito breeding sites to determine larval population density of Aedes aegypti, an evaluation of public knowledge, attitude, and personal protection practices against dengue, and an evaluation of the efficacy of VectoBac® DT/Culinex Tab tablets based on Bacillus thuringiensis israelensis against Ae. aegypti larvae. Of the 173 human sera samples that were assayed for dengue viruses, 94.9% were positive, 2.2% negative and 2.8% equivocal. Thirty households were randomly chosen per Barangay "Villages" (lowest level of formal local administration). Of the 489 breeding sites surveyed, 29.4% were infested with Ae. aegypti larvae, with discarded tires having the highest infestation rate (69.4%). A survey of people's knowledge, attitude, and practices for integrated community-based dengue control showed that 68.7% of the interviewees were aware that dengue is transmitted by mosquitoes, but only 4.3% knew that a virus was the cause of the disease. The efficacy of one and two tablets of VectoBac® DT/Culinex® Tab, based on Bacillus thuringiensis israelensis, was assessed against the larvae of Ae. aegypti exposed to sunshine and shaded water containers in semi-field and field tests. In semi-field tests, 100% mortality was achieved until the 18th and 30th day after the application of one and two tablets, respectively, in sun-exposed containers. In shaded containers, 100% mortality was observed until the 30th and 36th day after the application of one and two tablets, respectively. In field tests, the tablets were effective for approximately 3 weeks. Journal of Vector Ecology 30 (2): 277-283. 2005. Keyword Index: Aedes aegypti, Bacillus thuringiensis israelensis, dengue viruses, Philippines. INTRODUCTION Dengue fever (DF) and dengue haemorrhagic fever (DHF) are major public health problems. Since the first global appearance of DHF in Manila, Philippines in 1953, five years later in Bangkok, Thailand, and 1968 in Surabaya, Indonesia, it is presently considered one of the leading viral diseases causing hospitalization and death among children and adults in 100 tropical and subtropical countries throughout America, South-East Asia, the Western Pacific islands, and Africa. About 500,000 people are hospitalized (95% of those affected are children) and about 24,000 fatalities are reported annually (Halstead 1980, 1982, WHO 1997). Dengue fever outbreaks are rising in South America and Asia. Among the affected countries in Asia, the Philippines is considered as one of the "high risk" zones. The resurgence of dengue can be traced to rapid urbanization, poor sewage systems, and improper disposal of garbage. The persistence and numerical increase of Ae. aegypti (Linnaeus) as a primary vector and Aedes albopictus (Skuse) as a secondary vector is partly attributed to the sanitary and hygienic practices. The urbanization process has left many households with inadequate water supplies and has hastened the spread of the disease. Unfortunately, the majority of the people do not realize the seriousness of the situation until they become infected. Many people dispose of their garbage and waste just outside their door step, onto the streets, and in vacant lots. Eventually, the improperly disposed waste accumulates water during rainfall, becoming potential breeding sites for Ae. aegypti. The rapid increase in the human population in the Philippines has contributed to the rising dengue problem. Between 1991-2003, the highest numbers of dengue cases were recorded in 1998, with 35,648 cases and 514 deaths. The majority of the cases were children between 1-9 years (DOH 2003). This study evaluates the present dengue situation and control strategies against Ae. aegypti for the development of an integrated community-based dengue control program in Cebu City, and includes the detection of dengue viruses among Filipino patients, a survey of mosquito breeding sites and infestation rates of Ae. aegypti larvae, an evaluation of people's knowledge, attitude, and practices towards dengue infection, and an evaluation of the efficacy of Bacillus thuringiensis israelensis (B.t.i.) tablets against Ae. aegypti larvae. MATERIALS AND METHODS Detection of dengue viruses among the patients Blood samples from patients were collected from January to May 1999 at the Cebu City Medical Hospital. One hundred seventy-three sera samples, consisting of 108 female and 65

278

Journal of Vector Ecology

December 2005

male patients suspected of having dengue infection were assayed using a commercially available enzyme-linked immunosorbent assay (ELISA) test (PanBio, Australia, Cat. No. DEG-100). Sera samples were defrosted one night prior to the test according to the standard procedure. Five microwells in triplicate were used, each for positive and negative control as well as the cut-off calibrator. With 10 L of patient's serum, 1000 L of serum diluent (provided in the kit) was added and mixed thoroughly. Alternatively, to 10 L of patient's serum, 90 L of serum diluent was added and from this mixture, 20 L added to 180L of serum diluent and mixed again thoroughly. One hundred L of the mixture was pipetted into each microwell of the positive, negative, and cut-off calibrator. The microwells were covered and incubated for 30 min at 37° C and were washed manually six times. One hundred L of horseradish peroxidase (HRP) conjugated anti-human IgG were added to each microwell, covered and incubated for 10 min at 37° C. The microwells were washed again manually and the bound complexes were visualized through the addition of 100 L tetramethylbenzidine (TMB) in each microwell. After 30 min, the absorbance of each microwell (presence of dengue viruses) was read by using a microtiter plate reader with 450 nm wavelength. PanBio units were computed according to the standard procedure. Positive samples for dengue viruses were defined as having PanBio units of >11, negative with <9, and equivocal (either + or -) with 9-11. Survey of mosquito breeding sites and infestation rates of Aedes aegypti larvae The study was conducted in Cebu City which is located at the center of the archipelago (9°25' N and 11°30' N and between 123°25' E) about 400 miles south of Manila. The topography of Cebu is characterized by narrow coastlines, limestone plateaus, and coastal plains but with predominant rolling hills and rugged mountain ranges traversing the northern and southern lengths of the island. The climate is relatively moderate, having no distinct wet and dry season with a temperature range of between 23-33° C (73-91° F). Coolest temperatures can be felt in January and warmest in May. Cebu has a population of about 3.5 million with the city proper accounting for 718,821. Five Barangays "Villages" in Cebu City were selected for the study because of their high incidence of dengue fever during the past five years: Lahug, Labangon, Guadalupe, San Nicolas, and Pardo (CCHR 2003). The survey was conducted over a one month period. Thirty households were randomly chosen per Barangay. In each household, the specific breeding sites and their numbers were recorded as follows: 20 L and 50 L water containers, tin barrels, discarded tires, flower pots, tin cans, discarded plastics, and other breeding sites such as leaf axils and coconut husks. The larvae were collected randomly from major breeding sites and species determination was conducted in the laboratory according to Schoenig taxonomic key (1971, 1977). Pupae were not included for the evaluation. The number of larvae/breeding sites was estimated using the

following ranges: 0, 1-10, 11-25, 26-50, and above 50 larvae/ breeding site. To quantify the infestation rates, standard indices as House Index (HI), Container Index (CI), and Breteau Index (BI) were determined. Evaluation of people's knowledge, attitude, and practices towards dengue infection The evaluation was conducted from June to November 1998 in Barangay T. Padilla, Cebu City, Philippines. This Barangay has a total surface area of 14 hectares with a population density of 857 people/ha. Interviews at 371 households were conducted randomly. The questionnaires contained information on household composition including socio-economic aspects of the individuals, experiences of dengue fever and related symptoms, and people's knowledge of the symptoms, causes, transmission, and prevention of the disease as well as vector control measures at the household. Evaluation of the efficacy of B.t.i. tablets against Aedes aegypti larvae VectoBac® DT/Culinex Tab tablets based on B.t.i. were used in the tests. The tablets (weight 384 mg) had a potency of 2,700 ITU/mg. A semi-field test was conducted from December 17, 2003 to January 28, 2004 in Cebu City. Thirty plastic water containers (each with a 50 L capacity) were used. Fifteen plastic water containers were exposed to sunshine: five of them were treated with one tablet each, another five with two tablets each and the remaining five served as a control. The water containers were covered with fine nets to prevent egg-laying by adult mosquitoes. Another 15 plastic containers were placed in a shaded area and treated in the same way as described above. Twenty larvae of Ae. aegypti (late 2nd and early 3rd instars) were released in each container before the start of the experiment and every three days during the study. Before the release of the new set of larvae, the remaining living larvae in each container were removed. In the control groups, all the larvae were replaced every three days. Mortality rate of the larvae was evaluated every three days starting from the application of the tablets until a mortality rate of less than 50% was observed. For statistical analyses, data were subjected to Duncan's multiple range test and/or Student's t-test (Köhler et al. 1984). Eleven major breeding sites of Ae. aegypti that were exposed to sunshine and in shaded areas were chosen from various houses. The breeding sites included: drums of 200 L water capacity, pails of 25 L water capacity, cement catch basins of 10 L water capacity, old tires, and water jars (Table 1). Breeding sites that contained Ae. aegypti larvae with approximately 1 to 50 L of water were treated with one tablet, while breeding sites with more than 50 L of water were treated with two tablets. For the control group, three different breeding sites were chosen: drums (each with a 200 L capacity), plastic containers (each with a 200 L capacity) and cans (each with a 25 L capacity). Mortality rates of Ae. aegypti larvae were evaluated every three days starting from the application of the tablets until the breeding sites were re-infested with 4th larval instars of Ae. aegypti.

Vol. 30, no. 2

Journal of Vector Ecology Table 1. Description of breeding sites in the field test.

Breeding site no. 1) Drum 2) Plastic pail 3) Drum 4) Drum 5) Drum 6) cement water catch basin 7) Drum 8) Old tire 9) cement water catch basin 10) Drum 11) water jar 12) Drum 13) Can 14)Plastic container Water Capacity (L) 200 25 200r 200 200 5 200 10 2,000 200 25 200 25 200 Quantity Of Water Present (L) approx. 20 approx. 2 approx. 5 approx. 50 approx. 20 approx. 3 approx. 2 approx. 2 approx. 1,2501,500 approx. 100 approx. 10 approx. 50 approx. 1 approx. 150 Location (Exposed Or Shaded) shaded exposed shaded exposed exposed exposed exposed exposed shaded shaded exposed exposed exposed shaded No. of Ae. aegypti larvae present 40-50 larvae (2nd & 3rd instars) 5 larvae (2nd instars) 100-150 larvae (all instars) 15-20 larvae (2nd, 3rd & 4th instars) 150-200 larvae (all instars) 100-120 larvae (2nd & 3rd instars 40-50 larvae (2nd& 3rd instars) 400-500 larvae (2nd, 3rd, & 4th instars) 3,000-4,000 larvae (all instars) 150-200 larvae (3rd & 4th instars) 200-250 larvae (3rd & 4th instars) 15-20 larvae (2nd & 3rd instars) 10-15 larvae (2nd & 3rd instars) 100-150 larvae (all instars) No. of Tablets Applied 1 1 1 2 1 1 1 1 4 1 1 Control Control Control

279

RESULTS Detection of dengue viruses infection among patients 94.9% (164 out of 173) of the sera samples were positive for dengue viruses, comprising 101 females and 63 males. Only 2.2% (4 out of 173) were negative, comprising 3 females and 1 male and 2.8% (5 out of 173) comprising 4 females and 1 male were equivocal. Survey of mosquito-breeding sites and infestation rates of Aedes aegypti larvae Out of 873 mosquito larvae sampled, 799 were Aedes aegypti (90%), 74 Ae. albopictus (9%), and Ae. scutellaris (1%). Of the 489 breeding sites, 29.4% (142) were found to be infested with mosquito larvae. In terms of percentages of the infested breeding sites, discarded tires had the highest rate with 69.4% (50 out of 72). Discarded plastic containers were 68.4% (13 out of 19), and other breeding sites such as leaf axils, coconut husks, and flower pots were next with >50%. Among the breeding sites with the least number of larvae were 20 L and 50 L capacity containers with 3.9% (7 out of 195) and 5% (1 out of 20), respectively (Table 2). According to the indices, Barangay Lahug ranks first with a house index of 93.3% followed by Guadalupe with 56.7%, San Nicolas with 43.3%, Labangon with 36.7%, and Pardo with 33.3%. For the container and Breteau indices, similar ranking was recorded as Barangay Lahug showed the highest rate of 55.7% and 230, respectively, and Barangay Pardo the lowest with 14.4% and 50, respectively (Table 3).

Evaluation of people's knowledge, attitude, and practices towards dengue The average age of the respondents was 38 years, 79.8% females and 20.2% males. The average number of persons in a household was six and the average number of children was three. Sixty-three percent of the population consisted of children below 15 years of age and the average age of residence was 22. Twenty-one percent of the respondents were housewives and 40% of them attended high school. Fortynine percent of the respondents' homes were located near open canals and 57% were constructed with cement and wood. Furthermore, 64% of the respondents had a direct source of tap water. One hundred sixty-nine of 371 (45.5%) respondents had experienced general febrile conditions during the last six months before the interview. Many of the interviewees (65%) mentioned fever, however, only 9.4% reported the typical symptom of body pains and 17.8% reported headache. While 68.7% of the interviewees were aware that the disease is transmitted by mosquitoes, only 4.3% of the respondents knew that a virus was the cause of the disease. About half (52.3%) of the respondents were aware of the potential danger of mosquito breeding sites and were also aware that adulticiding (fogging) was being done in their community. 48.5% of the respondents commented negatively on fogging and 67.7% agreed that fogging should be done in their Barangays. Vector control measures at the household level reported by the respondents included: cleaning up the surroundings, using mosquito nets, repellents (mosquito coils), and screening their homes. Almost half of the respondents were aware of the Barangay-initiated anti-dengue drive

280

Journal of Vector Ecology

December 2005

Table 2. Specific types of mosquito-breeding sites and larval infestation in five selected Barangays.

Breeding sites 1-10 Water container (up to 20 L) Water container (up to 50 L) Barrels Discarded tires Flower pots Tin cans Discarded plastics Other breeding sites (e.g. leaf axils, etc.) Total Percentage (%) 4 0 12 22 6 8 6 16 Number of larvae 11-25 3 0 4 11 1 1 1 5 26-50 0 1 4 9 0 0 2 8 50 and above 0 0 2 8 1 0 4 3 Total breeding sites w/ larvae 7 1 22 50 8 9 13 32 142 70.96 w/out larvae 188 19 56 22 8 17 6 31 347 29.04 Total sites surveyed % of sites w/ larvae

195 20 78 72 16 26 19 63 489 100

3.59 5.0 28.2 69.44 50 38.46 68.42 50.79

programs and 48% were willing to attend them. Evaluation of the efficacy of B.t.i. tablets against Aedes aegypti larvae VectoBac® DT/Culinex Tab tablets were very effective against the larvae of Ae. aegypti. In the semi-field test, 100% mortality of Ae. aegypti larvae was obtained until the 18th day and 57% mortality on the 27th day by the application of one tablet in containers exposed to sunshine. Comparison of controls with statistical means showed that after the 33rd day, there was no significant (P<0.01) efficacy of the tablets remaining. With the application of two tablets in exposed areas, 100% mortality of the larvae was observed until the 30th day and 55% on day 33. In shaded areas, a mortality rate of 100% was achieved until the 30th day, with 85% on the 33rd day and 58% on the 36th day with the application of one tablet. By comparison with controls (P<0.01), it could be demonstrated that after the 42nd day, the tablets had completely lost their efficacy. With the application of two tablets in shaded areas, 100% mortality was achieved until the 36th day, 82% on the 39th day, and 55% day 42. Using Duncan's multiple range test, it was demonstrated that in the sun exposed area, efficacy of B.t.i. dropped significantly (P<0.01) after the 27th day (one Table 3. House, container, and Breteau index values in each Barangay.

Barangay Guadalupe Lahug San Nicolas Labangon Pardo House index 56.66 93.33 43.33 36.66 33.33 Container index 29.89 55.37 25.27 15 14.44 Breteau index 90 230 83.33 56.66 50.0

tablet) and 33rd day (two tablets) respectively, whereas in the shaded area, 100% efficacy was still observed. In the control group, in shaded and in sunny areas, mortality rates of 1- 5 % were observed during the course of the study (Table 4). In field tests, the tablets were effective for approximately 3 weeks (Table 5). In the control group, the containers were infested from the beginning until the end (3 weeks) of the experiment with increasing numbers of larvae. DISCUSSION Detection of dengue viruses infection among patients The results obtained from the ELISA tests showed a very high rate (94.9%) of patients positive for dengue viruses. These figures correlate with the results obtained by Buerano et al. (2000) who showed that 80.4% of sera samples were positive for dengue viruses. These figures are not surprising because the samples were collected from a hospital where poor patients came for treatment, having no money for proper medication, and perhaps preferring to take some herbal medications. These people usually come to the hospital only when they become severely ill. Survey of mosquito-breeding sites and infestation rates of Aedes aegypti larvae The species composition obtained in this study correlates to the results of Schoenig (1977). The abundance of mosquitobreeding sites and the rate of larval infestations are closely correlated with densely populated areas. Barangays Lahug and Guadalupe are the most densely populated among the five Barangays surveyed and also had the highest number of breeding sites and larval infestation rates, with most of the breeding sites classified as artificial. Thus, the breeding sites of mosquitoes can be diminished by the local community by

Vol. 30, no. 2

Journal of Vector Ecology

281

Table 4. Average mortality rate of Ae. aegypti larvae in the semi-field test: shaded and exposed to sunlight.

Average M ortality Rate Of 5 Containers

Date

3 days 6 days 9 days 12 days 15 days 18 days 21 days 24 days 27 days 30 days 33 days 36 days 39 days 42 days

Shaded Area

1 tablet 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 85% a 58% b 27% c 5% c 2 tablets 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 82% a 55% b Control 2% 1% 2% 0% 2% 2% 0% 0% 2% 1% 1% 3% 1% 3% a a a a a a a a a a a a a a 1 tablet 100% a 100% a 100% a 100% a 100% a 100% a 91% a 77% a 57% b 35% b 4% c

Exposed Area

2 tablets 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 100% a 55% b Control 4% 3% 3% 0% 2% 0% 2% 0% 0% 0% 1% a a a a a a a a a a a

Values followed by the same letter are not significantly different (P<0.01)

Table 5. Mortality rate of Ae. aegypti larvae in the field test.

Duration 1 3 days 100% 2 100% 3 100% 4 100% Treated container number Mortality (%) 5 6 7 100% 100% 100% Control Number of larvae 12 13 14 7 (2nd & 3rd instars; 8 pupae) 20-30 (3rd instars) 18 (3rd & 4th instars; 2 pupae) 100-150 (all instar larvae) 100-150 (all instar larvae) 100-150 (all instar larvae) 100-150 (all instar larvae) 8 (2nd & 3rd instars; 2 pupae) 20-30 (2nd & 3rd instars) 20-30 (2nd & 3rd instars; 5 pupae) 50-100 (all instar larvae; 10-20 pupae) 50-100 (all instar larvae) 50-100 larvae (all instars) 50-100 (all instar larvae) 15-20 (2nd & 3rd instars) 5-10 (3rd instars; 5 pupae) 30-50 (1st & 2nd instars) 50-75 (2nd & 3rd instars) 20-30 (3rd instars; 10 pupae) 10-20 (2nd & 3rd and 4th instars) 20-30 (3rd + 4th instars)

8 100%

9 100%

10 100%

11 100%

6 days 9 days

100% 100%

100% 100%

100% 100%

100% 100%

100% 100%

100% 100%

100% 100%

100% 100%

100% infested

100% 100%

100% 100%

12 days

100%

100%

100%

100%

100%

100%

100%

infested (30-50 2nd & 3rd instars)

infested (1,002000 1st & 2nd instars)

100%

100%

15 days

100%

100%

100%

100%

100%

water dried up

100%

100%

water emptied

18 days

100%

100%

100%

100%

100%

water dried up

100%

21 days

infested infested (100-150 (5-10 1st instars) 2nd instars)

100%

infested (10-20 1st & 2nd instars)

infested (50-75 3rd & 4th instars)

infested (50-100 1st & 2nd instars)

1,3,4,5,7,10, and 12 = drum (each has 200 L water capacity). 2 = plastic pail (each has 25 L water capacity). 6 and 9 = cement water catch basin. 8 = old tire.

282

Journal of Vector Ecology

December 2005

strictly following proper sanitation and hygienic practices. In the Philippines, environmental strategies to reduce the population of mosquito vectors are also practiced. An example is a memorandum entitled "4 o'clock Habit" or "Kill the Mosquito, Knock Out Dengue" that was issued in 1996 by the President of the Philippines. At 4:00 every afternoon, bells at universities, churches, and city halls rang, requesting people to go to their respective areas to clean up and reduce mosquitobreeding sites. However, when recommending source reduction for Ae. aegypti larvae by emptying or destroying discarded containers, the usefulness of these practices to homeowners should be considered (Nathan and Knudsen 1991). The Philippine government has declared June as the month of "Dengue Awareness." Little Dengue Health Brigades have also been established, consisting of students distributing basic information about dengue to schools and the general public. Aside from the program mentioned above, a Dengue Task Force has also been created, composed of trained people that are responsible for educating the public about dengue as well as monitoring the existence of dengue viruses. Press conferences and media campaigns (radio and television) are also held to provide information to the public on the disease and appropriate control strategies. Evaluation of people's knowledge, attitudes, and practices towards dengue The knowledge, attitudes, and practices towards dengue were limited. Many of the interviewees (65%) mentioned fever, however, only 9.4% reported typical symptoms of body pains and 17.8% reported headache. While 68.7% of the interviewees were aware that the disease is transmitted by mosquitoes, only 4.3% of the respondents knew that a virus was the cause of the disease. The respondents' fragmentary knowledge on the disease and transmission appears to be one of the reasons why people do not follow the instructions to reduce mosquito-breeding sites (Gubler, 1998). The results of this study agree with the results obtained by Kroeger et al. (1995) on community-based dengue control in Colombia, South America regarding knowledge and practice and the potential contribution of the biological larvicide B.t.i. In their study, they found that many interviewees (70%) mentioned fever, but only one-third reported the typical symptoms of muscle and rheumatic pain. To address this problem, there is a need to disseminate information, education, and communication activities systematically. Educational activities could be carried by social mobilization organized by the community, interpersonal communication such as women's groups, and mass media broadcasts, radio, newspapers, and videos. Evaluation of the efficacy of B.t.i. tablets against Aedes aegypti larvae B.t.i. tablets showed promising results against Ae. aegypti larvae. In semi-field tests, the duration of the efficacy of B.t.i. tablets was longer in shaded water containers than in containers exposed to sunlight. This result is correlated with that of Melo Santos et al. (2001) on the evaluation of a new

tablet formulation based on B.t.i. for the control of Ae. aegypti. In their study, the authors found that B.t.i. was effective for 35 days in containers exposed to sunlight and for more than 50 days in shaded containers. The same results were obtained from the study done by Kroeger et al. (1995) on B.t.i. tablets in Colombia. The authors found that B.t.i. lasted for more than a month. In water containers exposed to sunlight, the long-term effect of the tablets was reduced because B.t.i. was inactivated by sunlight (Becker et al. 2003). The UV-A/B of sunlight reaching the earth's surface is considered responsible for the photo-degradation and consequent loss of toxicity (Glare and O'Callaghan 1998). Experiments under semi-field conditions also showed that the duration of the efficacy can be increased by higher concentrations of B.t.i. (two tablets instead of one). A lower dose (one tablet/50 L water) was less effective and is not recommended, particularly because of decreasing water temperature (Becker and Margalit 1993) and with increasing numbers of mosquito larvae, the efficacy of B.t.i. decreases (Becker et al. 1992). In field tests, the duration of the efficacy of B.t.i. tablets was shorter (about 21 days) than in semi-field conditions (up to 36 days in shaded containers and up to 30 days in sunexposed containers). This may be because the conditions in natural breeding sites cannot be controlled. Various factors have to be considered, including rainfall and thunderstorms. Heavy rain was experienced during the field study causing the mosquito-breeding containers to overflow, thus having a negative impact on the efficacy of B.t.i.. B.t.i. is a very promising microbial agent and should be incorporated in integrated mosquito control management programs to reduce the incidences of diseases such as DF and DHF in the tropics. In suitable formulations, this microbial agent is a useful supplement or replacement for conventional pesticides. It can also be assumed that with the use of B.t.i. in an integrated program, the onset of resistance can be prevented (Becker et al. 1991). On the other hand, B.t.i. offers an ecologically defensible compromise between the need of humans to protect themselves from nuisance mosquitoes and current environmental policies focusing on protection of sensitive ecosystems by the use of non-selective methods. Acknowledgments Heartfelt thanks go to Dr. Beate Ruch-Heeger, Ms. Agnes Tomayao, and Ms. Abigail Genelsa for their assistance during the field study. The authors especially thank Valent BioSciences, ICYBAC GmbH, German Mosquito Control Association (KABS), Gesellschaft zur Förderung der Stechmückenbekämpfung e.V. (GFS), and Lion's Club (Goldener Hut) for financial support of the study. REFERENCES CITED Becker, N., S. Djakaria, A. Kaiser, O. Zulhasril, and H.W. Ludwig. 1991. Efficacy of a new tablet formulation of an asporogenous strain of Bacillus thuringiensis israelensis against larvae of Aedes aegypti. Bull. Soc.

Vol. 30, no. 2

Journal of Vector Ecology

283

Vector Ecol. 16: 176-182. Becker, N., M. Zgomba, M. Ludwig, D. Petric, and F. Rettich. 1992. Factors influencing the activity of Bacillus thruringiensis var. israelensis, treatments. J. Am. Mosq. Contr. Assoc. 8: 285-289. Becker, N. and J. Margalit. 1993. Use of Bacillus thuringiensis israelensis against mosquitoes and blackflies. Bacillus thuringiensis, An Environmental Biopesticide: Theory and Practice. John Wiley and Sons Ltd. pp. 147-170. Becker, N., D. Petric, M. Zgomba, C. Boase, C. Dahl, J. Lane, and A. Kaiser. 2003. Mosquitoes and their control. Kluwer Academic/Plenum Publishers. New York. 497 pp. Buerano, C., I. Ibraham, R. Contretras, F. Hasebe, R. Matias, F. Natividad, and A. Igarashi. 2000. IgM-capture ELISA of serum samples collected from Filipino dengue patients. Southeast Asian J. Trop. Med. Publ. Hlth. 31: 524-529. Cebu City Health Record (CCHR). 2003. Cebu City, Philippines. 112 pp. Department of Health (DOH). 2003. Cebu City, Philippines. 98 pp. Ellis, R. 2001. Municipal Mosquito Control Guidelines. Health Canada, Bureau of Infectious Diseases. Glare, T.R. and M. O' Callaghan. 1998. Environmental and health impacts of Bacillus thuringiensis israelensis. Report to the Ministry of Health, Wellington, NZ. Gubler, D.J. 1998. Dengue and dengue haemorrhagic fever. Clin. Microbiol. Rev. 11: 480-496.

Halstead, S.B. 1980. Dengue haemorrhagic fever-a public health problem and a field for research. Bull. Wld. Hlth. Org. 58: 1-21. Halstead, S.B. 1982. WHO fights dengue haemorrhagic fever. Wld. Hlth. Org. Chron. 38: 65-67. Köhler, W., G. Schachtel, and P. Voleske. 1984. Biometrie. Springer Verlag, Berlin. 246 pp. Kroeger, A., U. Dehlinger, G. Burkhardt, W. Atehortua, H. Anaya, and N. Becker. 1995. Community based dengue control in Colombia: people's knowledge and practice and the potential contribution of the biological larvicide B.t.i. (Bacillus thuringiensis israelensis). Trop. Med. Parasitol. 4: 241-246. Melo Santos, A., E.G. Sanches, and F. de Jesus. 2001. Evaluation of a new tablet formulation based on Bacillus thuringiensis serovar. israelensis: Control of Aedes aegyypti. Mem. Instit. Oswaldo Cruz. 96: 859-860. Nathan, M.B. and B. Knudsen. 1991. Aedes aegypti infestation characteristics in several Caribbean countries. J. Am. Mosq. Contr. Assoc. 7: 400-404. Schoenig, E. 1971. Mosquitoes in Cebu City and adjacent areas: an ecological survey. Phil. Sci. 8: 21-32. Schoenig, E. 1977. Distribution and abundance of mosquito species in Metropolitan Cebu, their relation to public health and their control. Phil. Sci. 14: 34-51. WHO, 1997. World malaria situation in 1994. Part I. Population at risk. Weekly Epidem. Rec. 72: 269-274.

284

Journal of Vector Ecology

December 2005

Synergistic efficacy of botanical blends with and without synthetic insecticides against Aedes aegypti and Culex annulirostris mosquitoes

Essam Abdel-Salam Shaalan1, Deon Vahid Canyon2 , Mohamed Wagdy Faried Younes3, Hoda Abdel-Wahab1, and Abdel-Hamid Mansour1

2

Zoology Department, Aswan Faculty of Science, South Valley University, Aswan, Egypt School of Public Health and Tropical Medicine, James Cook University, Townsville, Qld 4811, Australia 3 Zoology Department, Faculty of Science, Menoufia University, Shebin El-Kom, Egypt Received 22 February 2005; Accepted 26 July 2005

1

ABSTRACT: Increasing insecticide resistance requires strategies to prolong the use of highly effective vector control compounds. The use of combinations of insecticides with other insecticides and phytochemicals is one such strategy that is suitable for mosquito control. In bioassays with Aedes aegypti and Culex annulirostris mosquitoes, binary mixtures of phytochemicals with or without synthetic insecticides produced promising results when each was applied at a LC25 dose. All mixtures resulted in 100% mortality against Cx. annulirostris larvae within 24 h rather than the expected mortality of 50%. All mixtures acted synergistically against Ae. aegypti larvae within the first 24 h except for one mixture that showed an additive effect. We conclude that mixtures are more effective than insecticides or phytochemicals alone and that they enable a reduced dose to be applied for vector control potentially leading to improved resistance management and reduced costs. Journal of Vector Ecology 30 (2): 284-288. 2005. Keyword Index: Botanical, phytochemical, insecticide, mosquito, synergism.

INTRODUCTION Over the past decade, phytochemicals have received progressively more attention as insecticide alternatives. Roark (1947) described approximately 1,200 plant species that had been listed in the literature as having potential insecticidal value, while Sukumar et al. (1991) listed and discussed 344 plant species that only exhibited mosquitocidal activity. Most studies on the synergistic, antagonistic, and additive toxic effects of binary mixtures involving phytochemicals have been conducted on agricultural pests rather than vectors of diseases. The few studies on the mosquitocidal activity of binary mixtures have investigated the combined effects of phytochemicals with insecticides or microbial control agents. Synergism between synthetic insecticides and phytochemicals appears to be more common than among different phytochemicals, with some phytochemicals producing varied results depending on which synthetic insecticides they are mixed with. For instance, non-lethal concentrations of the volatile oil thymol and an unsaponifiable portion isolated from Thymus capitatus synergized the toxicity of malathion but induced additive or antagonistic effects when mixed with permethrin or pirimiphos-methyl insecticides in assays of Culex pipiens larvae (Mansour et al. 2000). This study thus aimed to identify alternative active botanical substances that could be combined with either existing synthetic mosquito control insecticides or other botanicals to produce synergistic or additive effects.

MATERIALS AND METHODS Mosquito species and maintenance of cultures Aedes aegypti were obtained from a colony initiated from mosquitoes collected in 2002 from Townsville, Australia. Culex annulirostris were bred from a colony maintained by the Queensland Institute of Medical Research in Brisbane, Australia. The colonies of mosquitoes were maintained at conditions of 27 ± 2 Cº and 70% ± 5 R.H. under 14L:10D cycles. Ae. aegypti larvae were kept in plastic buckets half filled with tap water and fed on goldfish flakes while Cx. annulirostris larvae were reared in aerated plastic trays half filled with dechlorinated tap water containing pieces of grass and fed on a mixture of granulated fish food (80%), liver powder (10%), and yeast (10%). Water in rearing containers was refreshed every 2 days. Adult mosquitoes were maintained on a 10% sugar solution while females were fed on rat blood. Plant species and extraction Seeds of Khaya senegalensis were collected from the Botanical Island in Aswan Province, Egypt. Seeds of Daucus carota were provided by Arthur Yates & Co Ltd (21A Richmond Road, Homebush, NSW 2140, Australia). Seeds were washed with running water, dried at room temperature, ground, bottled, and refrigerated until extraction. Essential oils of both K. senegalensis and D. carota seeds were extracted according to the method of Stein and Klingauf (1990). A quantity of ground seed (5 - 15 g) was extracted for 4 h with 150-300 ml of solvent (acetone, ethanol, hexane, and methanol) in a Soxhlet apparatus. The different crude extracts, comprised primarily of essential oils, were separated under

Vol. 30, no. 2

Journal of Vector Ecology

285

vacuum by using a rotary evaporator at temperatures equivalent to the boiling points of the solvents used. The three Callitris extracts were supplied by Michael Kennedy, Queensland Forestry Research Institute, Department of Primary Industries, Queensland Government, Australia. Steam volatile oil was obtained by distillation from cypress sawdust in water at atmospheric pressure. In this method, the most volatile components of the wood float on the condensate water as an oil. The second extract from cypress sawdust was obtained by using liquefied refrigerant gas (i.e., under pressure), followed by removal of the refrigerant by flashing off by reducing the pressure. This extract contained a balance of more volatile and less volatile components. The third extract from cypress sawdust was obtained using methanol under reflux followed by removal of methanol by distillation, further removal under vacuum (rotary evaporator), and freeze drying of the remainder. Much of the more volatile components were lost in this process. These extracts, produced in early 2004, were sealed, refrigerated, and protected from light until use. Synthetic Insecticides and growth regulators Technical grade organophosphorous and pyrethroid insecticides (fenitrothion 96.8% and lambda-cyhalothrin 90.99%) were provided by Nufarm Ltd (103-105 Pipe Road, Laverton, 3026 North Victoria, Australia). Technical grade (s)-methoprene, an insect growth regulator, was obtained from Wellmark International (1100 East Woodfield Road, Suite 500, Schaumburg, IL 60173, U.S.A.). Bioassay Two groups of mixtures were evaluated against both Ae. aegypti and Cx. annulirostris larvae. The first group contained mixtures from botanical extracts and synthetic insecticides, while the second group contained mixtures from botanical extracts only. The mixtures were prepared according to the

method described by Moawed4. All mixtures consisted of a 1:1 (v/v) ratio of the LC25 dose of each compound. Four ml from each test substance were mixed together in a glass vial. Two ml of the mixture were placed in a glass beaker containing 98 ml de-ionized water and 25 newly-emerged 4th instar mosquito larvae. Each test was replicated four times with one control and mortality was recorded daily till the death of all larvae or adult emergence. The equation below (Sun and Johnson 1960) was used to evaluate the joint effect of the different binary mixtures after 24 h.

Co - toxicity factor = observed % mortality - expected % mortality × 100 Expected % mortality

This factor differentiates the results into three categories. A positive factor of 20 indicates potentiation, a negative factor of -20 indicates antagonism, and the intermediate values of >-20 to < 20 indicate an additive effect. Because obtained LC25s are mathematically estimated, they were tested again against mosquito larvae to determine the expected mortality. The expected mortality of the combined pair is the sum of the mortalities of single compound at the given concentration LC25. The observed mortality is the recorded mortality obtained 24 h after using the mixtures. RESULTS All mixtures tested against Ae. aegypti larvae showed synergistic effects except a mixture of K. senegalensis hexane extract and steam distilled C. glaucophylla which showed an additive effect (Table 1). All mixtures tested against Cx.

4

Moawed, H.A.M. 1998. Joint action of some plant extracts against the mosquito larvae of Culex pipiens and their physiological impact. M.Sc. Thesis, Faculty of ScienceDmietta, Mansoura University, Egypt.

Table 1. Effect of mixtures from insecticides and botanical extracts at 1:1 ratios (LC25 + LC25) against newly-molted 4th instar Aedes aegypti larvae.

Callitris glaucophylla extracts Liquefied refrigerant Steam distillation Methanol reflux gas Adult Adult Adult Joint Joint Joint emergence emergence emergence action action action (%) (%) (%) Test Cont. Test Cont. Test Cont. S S S 1 1 0 93 99 93 S S S S 0 1 0 0 93 96 93 96 S S S S S 0 0 0 0 0 96 93 93 99 99 Khaya senegalensis (Hexane extract) Joint action S S S A S S Adult emergence (%) Test Cont. 2 3 0 0 0 0 96 99 99 96 96 99

Insecticides and botanical extracts

Insecticides Callitris glaucophylla

Fenitrothion Lambda-cyhalothrin Methoprene Steam distillation extract Liquefied refrigerant gas extract Methanol reflux extract

S = synergism. A = additive.

286

Journal of Vector Ecology

December 2005

Table 2. Effect of mixtures from insecticides and botanical extracts at 1:1 ratios (LC25 + LC25) against newly-molted 4th instar Culex annulirostris larvae.

Insecticides Botanical extracts Lambda-cyhalothrin Adult emergence L. M. in Adult emergence L. M. in Joint action Joint action (%) Cont. (%) Cont. S S S S S S S S S S S 0 0 0 0 0 0 0 0 0 0 0 1 7 7 3.7* 5.6* 2 2 1 1 1 1 S S S S S S S S S S S 0 0 0 0 0 0 0 0 0 0 0 1 4 7 6.5* 5.6* 6.7* 2.5* 1 1 1 1 Fenitrothion

Callitris glaucophylla Daucus carota Khaya senegalensis

Steam distillation extract Liquefied refrigerant gas extract Methanol reflux extract Acetone extract Ethanol extract Hexane extract Methanol extract Acetone extract Ethanol extract Hexane extract Methanol extract

L. M. in Cont. = larval mortality % in control. S = synergism. * Emerged pupae not included in calculations.

annulirostris larvae showed synergistic effects and caused 100% larval mortality within the first 24 h (Tables 2 and 3). Out of 18 mixtures tested, five did not completely inhibit adult Ae. aegypti emergence. Mixtures of fenitrothion with steam-distilled C. glaucophylla and a K. senegalensis hexane extract resulted in 1 and 2% emergence, respectively. Lambdacyhalothrin mixed with steam-distilled and liquefied refrigerant gas extracts of C. glaucophylla and a K. senegalensis hexane extract led to 1, 1, and 3% emergence, respectively. Both groups of mixtures, botanical extracts with or without insecticides, were observed to completely inhibit adult Cx. annulirostris emergence within the first 24 h. While both mixtures of botanicals and mixtures of botanicals with methoprene effectively inhibited adult Ae. aegypti emergence, mixtures of insecticides with botanicals did not completely inhibit emergence although the emergence rate of 1­3% was negligible compared with 93­99% in untreated controls. All binary mixtures proved to be more effective than all other non-mixed sublethal concentrations (LC25, LC50, and LC75) of both insecticide and phytochemical, and were similar in efficacy to non-mixed LC100 doses. DISCUSSION The synergistic effects observed in bioassays using a combination of botanical extracts and different synthetic insecticides have been observed in several previous studies (Kalyanasundaram and Babu 1982, Kalyanasundaram and Das 1985, Mulla and Su 1999, Thangam and Kathiresan 1990, 1991). Some extracts have also produced synergistic effects

with insect growth regulators. Mulla and Su (1999) showed that neem seed kernel extract has synergistic effects when combined with the juvenile hormone analog methoprene. A few studies have mentioned synergism between different botanical extracts. Mwaiko (1992) reported that a mixture of the peel oils extract of three citrus species (lemon, orange, and bitter orange) was much more effective than for the peel oils extract for the individual species. Moawed (1998) studied the joint action of binary mixtures of some plant extracts with each other and with the synthetic pyrethroid insecticide cypermethrin against Cx. pipiens larvae. In addition to response variations caused by plant species, extraction method, and extraction solvent, the results from this study show that insecticidal efficacy of binary mixtures from botanical extracts with or without synthetic insecticides against mosquitoes varies based on mosquito genera. This difference in efficacy against both mosquito genera could be attributed to a difference in physiological response among mosquito genera and species and not to the effect of the mixtures. Surprisingly, it is the first study to mention the effect of mosquito genera on the efficacy of insecticide mixtures. The mechanism of synergism is not known, but Thangam and Kathiresan (1991) stated that synergism might be due to phytochemicals inhibiting the ability of mosquito larvae to employ detoxifying enzymes against synthetic chemicals. Mixtures may be useful in prolonging the lifetime of various cost-effective synthetic insecticides providing that care is taken not to use botanicals that are affected by the same resistance mechanisms that target the synthetic insecticide. Further cost benefits are possible due to increased potency despite the use of lower concentrations. When this is combined

Vol. 30, no. 2

Table 3. Effect of mixtures from botanical extracts at 1:1 ratios (LC25 + LC25) against newly-molted 4th instar Culex annulirostris larvae.

Daucus carota Methanol reflux MC 5 S S S S S S S S S S 0 4 S 0 4 S 0 5 S 0 5 S 0 0 0 5 S 0 4 S 0 6 3 6 0 6 S 0 1 S 0 6 0 3 S 0 4 S S S S 0 0 0 0 6 2 7 2 S S S S 0 0 0 0 3 3 3 2 S 0 5 S 0 1 S S S S 0 0 0 0 5 6 5 S 0 5 0 6 S 0 4 0 6 S 0 4 S 0 3 0 5 S 0 4 S 0 7 S 0 1 0 S S S 0 0 0 5 6 3 S 0 1 S 0 1 S 0 1 0 7 S 0 S 0 5 S 0 JA AE MC JA AE MC JA AE MC JA AE MC JA Ethanol Hexane Methanol Ethanol AE MC Khaya senegalensis Hexane JA AE MC Methanol JA AE MC

Callitris glaucophylla Liquefied refrigerant JA S 0 AE

Botanical extracts

Steam distillation

JA

AE

MC

Callitris glaucophylla

Steam distillation Liquefied refrigerant gas Methanol reflux Acetone

S

0

5

Ethanol

S

0

3

Journal of Vector Ecology

Hexane

S

0

2

Daucus carota

Methanol

S

0

2

Acetone

S

0

1

Ethanol

S

0

4

Hexane

S

0

4

JA = joint action. AE = adult emergence. MC = percentage larval mortality in control. S = synergism.

Khaya senegalensis

Methanol

S

0

5

287

288

Journal of Vector Ecology

December 2005

with the degree of emergence inhibition observed and the longlasting effects (Chockalingam et al. 1990), the benefits of synergistic mixtures are clear. The identification of botanical compounds that prove to be highly synergistic when combined with currently used synthetic insecticides will lead to more cost-effective mosquito control and have the potential to minimize the development of resistance caused by frequent application of mono insecticides. Acknowledgments We are grateful to Dr Michael Kennedy, Queensland Forestry Research Institute, Department of Primary Industries, Queensland Government, Australia, for providing us with extracts of C. glaucophylla. We are also grateful to Dr Wayne Melrose, School of Public Health and Tropical Medicine, James Cook University, Queensland, Australia, for funding solvents required in this work. We are appreciative of Professor Bruce Bowden, School of Pharmacy and Molecular Sciences, James Cook University, Queensland, Australia, for his help, guidance, and unlimited facilities for conducting phytochemical extraction in his laboratory. REFERENCES CITED Chockalingam, S., S. Thenmozhi, and M.S.N. Sundari. 1990. Larvicidal activity of different products against mosquito larvae. J. Environ. Biol. 11: 101-104. Kalyanasundaram, M. and C.J. Babu. 1982. Biologically active plant extracts as mosquito larvicides. Ind. J. Med. Res. 76: 102-106.

Kalyanasundaram, M. and P. K. Das. 1985. Larvicidal and synergistic activity of plant extracts for mosquito control. Ind. J. Med. Res. 82: 19-23. Mansour, S.A., S.S. Messeha, and S.E. EL-Gengaihi. 2000. Botanical biocides. 4. Mosquitocidal activity of certain Thymus capitatus constituents. J. Nat. Tox. 9: 49-62. Mulla, M.S. and T. Su. 1999. Activity and biological effects of neem products against arthropods of medical and veterinary importance. J. Am. Mosq. Contr. Assoc. 15: 133-152. Mwaiko, G.L. 1992. Citrus peel oil extracts as mosquito larvae insecticides. E. Afr. Med. J. 69: 223-226. Roark, R.C. 1947. Some promising insecticidal plants. Econ. Bot., 1: 437. Cited in: C. F. Curtis, (ed.) Appropriate Technology in Vector Control. CRS Press, Florida, 1990. Stein, U. and F. Klingauf. 1990. Insecticidal effect of plant extracts from tropical and subtropical species. J. Appl. Entomol. 110: 160-166. Sukumar, K., M.J. Perich, and L.R. Boobar. 1991. Botanical derivatives in mosquito control: A review. J. Am. Mosq. Contr. Assoc. 7: 210-237. Sun, Y.P. and E.R. Johnson. 1960. Analysis of joint action of insecticides against house flies. J. Econ. Entomol. 53: 887-892. Thangam, T.S. and K. Kathiresan. 1990. Synergistic effects of insecticides with plant extracts on mosquito larvae. Trop. Biomed. 7: 135-137. Thangam, T.S. and K. Kathiresan. 1991. Mosquito larvicidal activity of marine plant extracts with insecticides. Botanica Marina 34: 537-539.

Vol. 30, no. 2

Journal of Vector Ecology

289

Patterns of insecticide resistance in larval Culex pipiens populations in Israel: dynamics and trends

Laor Orshan, Maria Kelbert, and Hedva Pener

Laboratory of Entomology, Ministry of Health, P.O.Box 34410, Jerusalem 91342 Israel Received 22 February 2005; Accepted 19 July 2005 ABSTRACT: Resistance to insecticides of larval Culex pipiens populations in Israel has been monitored for ten years and the results were used for control planning. The insecticides tested were the organophosphates chlorpyrifos, fenthion, and temephos and the pyrethroids permethrin and cypermethrin. Over the years the relative resistance (R/R) values to chlorpyrifos in most populations tested were between 200-400, with records of up to 700 R/R. Practically no susceptible populations were found and the compound has been withdrawn from use, resulting in a decrease in the relative resistance values. In 1996 when fenthion was reintroduced, no resistant populations were found, but in the following years significant resistance appeared in an increasing number of populations with values lower than recorded for chlorpyrifos but still high (up to 100 R/R). All populations tested were found to be susceptible to temephos and accordingly the compound was reintroduced in 2002. Over the ten years, no significant resistance to cypermethrin was found in the majority of the populations tested, while a trend of increased resistance to permethrin was noted with a few records of >1000 R/R. The results demonstrate the importance of continuous monitoring of insecticide resistance for efficient mosquito control. Alternative insecticides and methods of control are discussed. Journal of Vector Ecology 30 (2): 289-294. 2005. Keyword Index: Culex pipiens, larvae, mosquitoes, relative resistance, insecticides.

INTRODUCTION Mosquitoes of the Culex pipiens complex are the most common species in urban and rural areas of Israel (Kitron and Pener 1986). The species, although considered mainly as a nuisance, is strongly implicated in the transmission of West Nile virus (Nir et al. 1972, Samina et al. 1986). Mosquito control in Israel is directed against the larval stages, and routine monitoring of breeding sites is carried out. Registered residual insecticides belonging to organophosphates (OPs) and/or pyrethroids are employed when source reduction, biological control, and the use of IGRs or oils is not feasible. The policy is to alternately treat the mosquito breeding sites with registered compounds of the two chemical groups. Insecticide resistance is an increasing problem for the control of mosquitoes (Hemingway and Ranson 2000) and very high resistance values to OPs and pyrethroid compounds have been recorded in many countries (Wirth and Georghiou 1996, Bisset et al. 1997, Ben Sheikh et al. 1998, Liu et al. 2004). This is also a serious problem in Israel, where the development of resistance is elicited by continuous use of insecticides against mosquitoes (Pener 1982, Horowitz et al. 1998). The phenomena of cross and multiple resistance to the different insecticides should also be taken into consideration (Scott 1995, Wirth and Georghiou 1996, Rodriguez et al. 2002). The susceptibility status of larval populations of Cx. pipiens in Israel to pesticides in use has been monitored annually. Each year several larval populations from locations chosen according to the success or failure of control efforts were tested. Data regarding the resistance status of the

mosquito populations were forwarded to the relevant field personnel. Over the years, detection of high resistance in most populations resulted in the withdrawal from use of some compounds and the introduction of others. This paper summarizes the resistance status of Cx. pipiens larvae for ten consecutive years, covering locations from all over the country. The compounds tested were the OPs chlorpyrifos, fenthion, and temephos as well as the pyrethroids permethrin and cypermethrin. In the early 1980s, due to failure of control, the use of fenthion and temephos was stopped and chlorpyrifos was introduced. A noticeable increase in the resistance values to chlorpyrifos led to reintroduction of fenthion in 1996. As resistance to chlorpyrifos further increased, the compound was withdrawn at the end of 1999 and temephos was reintroduced in 2001. With the increasing awareness of the negative effect of OPs on the environment, the use of pyrethroids in mosquito control has increased (unpublished data, Laboratory of Entomology, Ministry of Health). The dynamics of resistance to insecticides or its reversal in Cx. pipiens larvae, as well as the possible correlation of resistance to the different compounds in each population, are described. MATERIALS AND METHODS Larvae of Cx. pipiens were collected from breeding sites, generally following complaints regarding the failure of control. Overall, 62 populations were tested, 28 in the last two years of the survey. The larvae (at least 500 specimens) were brought in the original water to the laboratory. In the laboratory the larvae were transferred into clean standing

290

Journal of Vector Ecology

December 2005

water. About 500 larvae were kept in three liters of water and fed daily with 2 g of ground food mixture (dry cat food, rabbit pellets, dry yeast ­ 4:4:2) at 27±2 oC, 65±2% RH and L:D=12:12. Late 3rd or early 4th instar larvae of the first and/or second generation (F1, F2) were tested for susceptibility to insecticides. The bioassays essentially followed the instructions of the World Health Organization (WHO) for testing mosquito larvae (WHO 1981). The insecticides tested were technical grade OPs chlorpyrifos, fenthion, and temephos and technical grade pyrethroids permethrin and cypermethrin. Between 6-10 concentrations of each insecticide were employed ranging from 0.001 ­ 2.5 ppm for the OPs and permethrin, and 0.0005 ­ 0.100 ppm for cypermethrin. Triplicates of 20 larvae were exposed to each concentration of insecticide for 24 h and mortality was recorded. The tests were repeated two to three times. Tests in which control mortality exceeded 10% were discarded. LC50 was calculated by probit analysis (SPSS version 12). The relative resistance R/R factor was calculated by relating the values to corresponding values of a susceptible Cx. pipiens population maintained in the laboratory. The susceptible population of Cx. pipiens was collected in the early 1980s from an isolated site in the north of the country. Since then it has been maintained in the laboratory as a reference population under the standard conditions described above. The linear regression of the results was calculated from the data. In the populations tested, we calculated the correlations between the R/R values of each insecticide against each of the others, employing the Bivariate Pearson's Correlation (SPSS version 12). RESULTS Organophosphates Chlorpyrifos: Figures 1A and 1B show the R/R values to chlorpyrifos of Cx. pipiens larval populations for ten consecutive years. It is clearly seen that from 1993 to 2000 the number of populations showing high R/R values increased each year. In the first two years (1993-1994) the R/R values did not exceed 100 and three out of seven populations tested were susceptible. In the following years (1995-1998), only two out of 14 populations assessed were susceptible, and high R/R values (237-578) were already detected in four populations. From 1999 until 2000, no susceptible populations were detected (12 populations). The lowest R/R value recorded was 144 in one population, 200-400 in nine populations and over 600 in two. (Figure 1A, R2=0.4567). In the three years following chlorpyrifos' withdrawal from use at the end of 1999, a trend of reversal to populations showing lower resistance values was noted. In 2001, out of the 14 populations tested, the R/R values in six were about 100, and only in two exceeded 250. The trend in the decrease of the R/ R values is even more noticeable in 2002 with only one out of the 14 populations tested showing R/R>400 (Figure 1B, R2 = 0.277). Fenthion: Figure 2 represents the R/R values for fenthion from 1996 (the year fenthion was reintroduced) to 2002. It is clearly seen that the number of resistant populations of Cx. pipiens larvae to this compound increased. In 1996-1997 only

six populations were tested. Five were found to be susceptible (R<7); only for one was the R/R value 14. In 1998­1999 resistance was already detected in five out of the ten populations tested with one already showing an R/R value of 50. A drastic rise in the resistance status was noted during the last two years of the survey. Only three out of the 27 populations tested were found to be susceptible (R/R<10), while R/R was >40 in a third of them and in one population R/R=100. In recent years more populations have been screened, thus the evaluation of resistance to fenthion has become more accurate. Temephos: Since 2000 the Cx. pipiens larvae were also tested for temephos, aiming to reintroduce the compound in the control program. The relative resistance values are shown in Table 1. The results show that all but one population (R/R = 26) were found to be susceptible. Consequently temephos was officially registered for mosquito control in 2002. Pyrethroids Permethrin: The susceptibility status of Cx. pipiens populations to permethrin was tested from 2000 to 2002. Until the year 2000, the use of permethrin as a larvicide against Cx. pipiens was quite limited. Control attempts at two breeding sites in 2000 failed. Thus, susceptibility tests were started and yielded high R/R values (69 and 323) indicating the development of resistance to permethrin. In the two subsequent years, 14 populations were tested each year. In 2001, two populations were found to be susceptible (R/R<10), five exhibited 25>R/R>10, and four showed 400>R/R>200. In 2002 no susceptible population was detected and three populations showed very high R/R values of 859-1,692 (Figure 3). Cypermethrin: The overall values of cypermethrin resistance were not high (Figure 4) and there is no obvious trend in the changes of the resistance (R2=0.0081). Over the years about half of the populations tested (26/60) were susceptible (R/R<10). High resistance was detected only in three populations: one in 1996 with R/R 200, one in 1998 with R/R 369, and one in 2002 with a R/R value of 126. The remaining populations showed moderate resistance with R/R values between 10-50, with the majority being below 20. Cypermethrin is still the most effective pyrethroid employed in larval control. Bivariate correlation clearly shows that a significant correlation at P<0.01 exists, as expected, between the R/R values to cypermethrin and permethrin and between fenthion and temephos (Table 2). In addition, significance at P<0.05 was noted between the R/R to cypermethrin and chlorpyrifos. No significant correlation was found between chlorpyrifos and fenthion. DISCUSSION Organized mosquito control programs have been implemented in Israel for many years and monitoring the susceptibility of Cx. pipiens larval populations to the insecticides in use constitutes an integral part of the programs.

Vol. 30, no. 2

Journal of Vector Ecology

291

Table 1. Susceptibility of Cx. pipiens larval populations to temephos.

Year 2000 2001 2002 No. of populations 6 14 10 R/R mean 4 5 6 R/R range 7-2 26-1 10-2

Table 2. Bivariate (Pearson's) correlation between R/R values of the insecticides tested*.

permethrin chlorpyrifos fenthion temephos cypermethrin 0.184 (39) NS 0.065 (36) NS 0.256 (29) NS 0.789 (39) S** cypermethrin 0.29 (68) S* -0.038 (55) NS 0.182 (33) NS temephos 0.171 (33) NS 0.495 (30) S** fenthion 0.02 (56) NS

NS: non significant. S*: Correlation significant at 0.05 level (2-tailed). S**: Correlation significant at 0.01 level (2-tailed).

According to the results, recommendations for differential use of insecticides are issued annually before the mosquito season. The policy is to abandon a compound when resistance is detected and to alternatively employ a different insecticide. When a wide-ranging trend for resistance is recorded (i.e. an increasing number of populations from different geographical regions are found to be resistant), a general ban of use is issued. The general picture of our results shows variability in the relative resistance values among the populations tested. As many of the mosquito-breeding sites are not permanent and vary from year to year, it was not possible to follow the developing of resistance in consecutive years in one place. Nevertheless, a general trend concerning mosquito populations in the country could be detected when the effects of specific compounds were considered. The data indicated a very high resistance to chlorpyrifos in practically all populations tested. The use of this compound was discontinued and followed by a very gradual but quite noticeable decrease in the R/R values (Figures 1A, 1B). The opposite trend, i.e. an increase in the number of resistant populations, was noted after the reintroduction of fenthion. The reintroduction of temephos showed that all but one population tested were susceptible (Table 1). We may assume that prolonged use brings a quick reversal to resistance. The two pyrethroids (cypermethrin and permethrin) are still in use, although more and more resistant populations to permethrin are being recorded. The literature on resistance and cross-resistance of mosquitoes to OPs and pyrethroids is abundant, and the reported results vary between countries, history, and policy

of treatment. Ben Cheikh et al. (1998) showed that resistance to temephos in Cx. pipiens populations from Tunisia was uniformly low, while resistance to chlorpyrifos was highly variable, reaching the highest levels (>10,000 fold) globally recorded, as well as very high levels of resistance to permethrin (~ 5,000 fold). In Cyprus, Wirth and Georghiou (1996) detected OP resistance to temephos, chlorpyrifos, and fenthion. Various results and conclusions were also reported for other mosquito species. Liu et al. (2004) detected high resistance values to pyrethroids in Cx. quinquefaciatus in Alabama and Florida, U.S.A. Aedes aegypti populations in the Caribbean exhibited geographical variation in their resistance to temephos, chlorpyrifos, and fenthion (Rawlins and Ou Hing Wan 1995). The variations in susceptibility are so great that Canyon and Hii (1999) concluded that susceptibility investigations should assess mosquito samples collected from many sites within an area, rather then relying on the results obtained from one site. Continuous and/or repeated exposure to insecticides results in the development of resistant populations. The resistance mechanisms have a genetic biochemical basis (Brogdon and McAllister 1998, Hemingway and Ranson 2000). Ben Cheikh et al. (1998) relate the very high resistance values, found in Cx. pipiens populations in Tunisia, to modification of the insecticide targets, acetylcholine esterases, AchE, for chlorpyrifos, and the voltage-dependent sodium channels for permethrin. These are the two major resistance mechanisms for the two most commonly employed insecticide groups (OPs and pyrethroids). Cross-resistance to a compound, which has not been used, might follow previous

292

Journal of Vector Ecology

December 2005

800 700 600 500 R/R 400 300 200 100 0 1991 1993 1995 1997 Year

800 700 600 R/R 500 400 300 200 100 0 1999 2000

Year

Figure 1. R/R values, by year, to chlorpyrifos for the populations tested.

A

1999

2001

2003

B

2001

2002

2003

· = R/R for one population, - the year 1999 chlorpyrifos was withdrawn from use. 1A 1993-2000 (R2 = 0.457). 1B. 2000-2002 (R2 = 0.234).

120 100 80 R/R 60 40 20 0 1995

1996

1997

1998 1999 Year

2000

2001

2002

2003

Figure 2: R/R values by year, to fenthion for the populations tested. · = R/R for one population. (R2 = 0.219).

Vol. 30, no. 2

Journal of Vector Ecology

293

1800 1600 1400 1200 R/R 1000 800 600 400 200 0 1999 2000

Year

Figure 3. R/R values by year, to permethrin. · = R/R for one population. (R2 = 0.073).

2001

2002

2003

400 350 300 R/R 250 200 150 100 50 0 1992 1994 1996 1998 2000 2002 2004

Year

Figure 4. R/R values by year, to cypermethrin. · = R/R for one population. (R2 = 0.0081).

294

Journal of Vector Ecology

December 2005

exposure to another insecticide with a similar resistance mechanism. It should be stressed that there is evidence that in mosquito populations, cross-resistance to pyrethroids might also occur as a result of primary resistance to OPs, because of over-produced esterases (Bisset et al.1997). Based on the two major different mechanisms for resistance (Hemingway and Ranson 2000), mosquito control strategies in Israel include alternate use of OPs and pyrethroids. However, when examining our results, highly resistant populations to OP and permethrin were found and the problem of cross-resistance arose. As expected, there was a significant correlation between the resistance/susceptibility to permethrin and cypermethrin on the one hand, and between temephos and fenthion on the other (Table 2). No correlation no found between resistance to chlorpyrifos and fenthion, nor between temephos and the pyrtehroids. The higher R/R values to fenthion, mostly in the last years of the survey (Figure 3), and at the same time the declining in the R/R values to chlorpyrifos on one hand (Figure 1B) and susceptibility to the newly reintroduced temephos on the other, might explain the results. The positive correlation found between chlorpyrifos and cypermethrin might indicate cross-resistance between the OPs and pyrethroids. This phenomenon has already been reported (Wirth and Georghiou 1996, Rodriguez et al. 2002). From our results, it is clear that the resistance status to insecticides within mosquito populations is a dynamic process depending on the frequency of use, type of insecticide, and possible effect of environmental contamination following agricultural pest control. As long as the employment of conventional insecticides constitutes an unavoidable part of an integrated mosquito management program, the monitoring of resistance is imperative. Acknowledgments We thank the field workers of the Ministry of the Environment for their effort in bringing mosquito larvae to the laboratory. The critical reading and invaluable comments of Prof. Y. Schlein are gratefully acknowledged. We also thank Dr. H. Schnur for editorial help. REFERENCES CITED Ben Cheikh, H., Z. Ben Ali-Haouas, M. Marquine and N. Pasteur. 1998. Resistance to organophosphorus and pyrethroid insecticides in Culex pipiens (Diptera: Culicidae) J. Med. Entomol. 35: 251-260. Bisset, J., M. Rodriguez, A. Soca, N. Pasteur and M. Raymond. 1997. Cross-resistance to pyrethroid and

organophosphorus insecticides in the southern house mosquito (Diptera: Culicidae) from Cuba. J. Med. Entomol. 34: 244-246. Brogdon, W.G. and J.C. McAllister. 1998. Insecticide resistance and vector control. Emerg. Infect. Dis. 4: 605613. Canyon, D.V. and J.K.L. Hii. 1999. Insecticide susceptibility status of Aedes aegypti (Diptera: Culicidae) from Townsville. Aust. J. Entomol. 38: 40-43. Hemingway, J. and H. Ranson. 2000. Insecticide resistance in insect vectors of human disease. Annu. Rev. Entomol. 45: 371-391. Horowitz, A.R., P.G. Weintraub, and I. Ishaaya. 1998. Status of pesticide resistance in arthropod pests in Israel. Phytoparasitica 26: 231-240. Kitron, U. and H. Pener. 1986. Distribution of mosquitoes (Diptera: Culicidae) in northern Israel: a historical perspective. II. Culicine mosquitoes. J. Med. Entomol. 23:182-187. Liu, H., E.W. Cupp, K.M. Micher, A. Guo, and N. Liu. 2004. Insecticide resistance and cross-resistance in Alabama and Florida strains of Culex quinquefaciatus. J. Med. Entomol. 41: 408-413. Nir, Y., A. Avivi, Y. Lasovski, J. Margalit, and R. Goldwasser. 1972. Arbovirus activity in Israel. Israel J. Med. Sci. 8: 1695-1701. Pener, H. 1982. Organophosphorous multi resistance in Culex pipiens molestus in Israel. Rev. Parasitol. 43: 409-414. Rawlins, S.C. and J. Ou Hing Wan. 1995. Resistance in some Caribbean populations of Aedes aegypti to several insecticides. J. Am. Mosq. Contr. Assoc. 11: 59-65. Rodriguez, M.M., J. Bisset, M. Ruiz, and A. Soca. 2002. Cross-resistance to pyrethroids and organophosphorus insecticides induced by selection with temephos in Aedes aegypti (Diptera: Culicidae) from Cuba. J. Med. Entomol. 39: 882-888. Samina, I., J. Margalit, and J. Peleg. 1986. Isolation of viruses from mosquitoes of the Negev, Israel. Trans. R. Soc. Trop. Med. Hyg. 80: 471-472. Scott, J. A. 1995. The molecular genetics of resistance: Resistance as a response to stress. Fla. Entomol. 78: 399414. Wirth, M.C. and G.P. Georghiou. 1996. Organophosphate resistance in Culex pipiens from Cyprus. J. Am. Mosq. Contr. Assoc. 12: 112-118. World Health Organization. 1981. Instructions for determining the susceptibility or resistance of mosquito larvae to insecticides. Mimeograph Document. WHO/ VBC/81/807.

328

Journal of Vector Ecology

December 2005

Laboratory estimation of degree-day developmental requirements of Phlebotomus papatasi (Diptera: Psychodidae)

Ozge Erisoz Kasap and Bulent Alten

Department of Biology, Faculty of Science, Ecology Section, EBAL Laboratories, Hacettepe University, 06800, AnkaraTurkey Received 6 June 2005; Accepted 22 July 2005 ABSTRACT: Cutaneous leishmaniasis is one of the most important vector-borne endemic diseases in Turkey. The main objective of this study was to evaluate the influence of temperature on the developmental rates of one important vector of leishmaniasis, Phlebotomus papatasi (Scopoli, 1786) (Diptera: Psychodidae). Eggs from laboratory-reared colonies of Phlebotomus papatasi were exposed to six constant temperature regimes from 15 to 32° C with a daylength of 14 h and relative humidity of 65-75%. No adult emergence was observed at 15° C. Complete egg to adult development ranged from 27.89 ± 1.88 days at 32° C to 246.43 ± 13.83 days at 18° C. The developmental zero values were estimated to vary from 11.6° C to 20.25° C depending on life stages, and egg to adult development required 440.55 DD above 20.25° C. Journal of Vector Ecology 30 (2): 328-333. 2005. Keyword Index: Sand flies, constant temperature, development, life stages, cutaneous leishmaniasis. INTRODUCTION Both cutaneous and visceral forms of leishmaniasis are caused by a wide range of Leishmania species and are transmitted by sand flies. Leishmaniasis is present in about 88 countries, mainly in tropical and subtropical areas, but it is also widespread in southern Europe. The overall prevalence of leishmaniasis is 12 million cases worldwide, and the annual global incidence of all clinical forms approaches two million new cases (WHO 2004). Turkey is a crossroad between the continents of Europe and Asia and has different ecological and climatic conditions that are important in leishmaniasis epidemics (Ok et al. 2002). According to the records of the Ministry of Health, of the 18,216 cases reported between 1994 and 2000, 11,234 (62%) were reported from the southeastern Anatolia region and 5,297 (33%) from the Mediterranean region. However, it is thought that the number of cases is far more than officially reported (Ok et al. 2002). Previous studies showed that nine of the 18 sand fly species found in Turkey are proven or probable vectors of Old World leishmaniasis (Yagci et al. 1998, Alptekin et al. 1999, Killick-Kendrick 1999). P. papatasi, a well known Old World sand fly species occurring throughout most of the south and southeastern part of Turkey, is one of the most suspectable vectors of cutaneous leishmaniasis. Although the role of P. papatasi in Turkey is not well-established, because of its anthropophilic behavior and domestic resting sites with high population density in urban areas, it is assumed that this species, together with the proven vector, Phlebotomus sergenti Parrot, 1917, may be responsible for cutaneous leishmaniasis in Turkey (Ok et al. 2002, Svobodovà et al. 2003). As ectoterms, ambient temperature significantly affects the developmental rates, survival of preimaginal stages, and the longevity of adult phlebotomine sand flies (Theodor 1934, Ready and Croset 1980, Valevich and Dergacheva 1992, Guzman and Tesh 2000). Specific knowledge of pest phenology is an essential ingredient in effective pest management. Models based on population dynamics and the environmental parameters which drive them almost invariably include the effect of temperature on development time. By using these models, it is possible to predict the number of generations per year, the potential ability of population increase, and distribution along temperature gradients (Pruess 1983, Wagner et al. 1991). Although sand flies have a significant impact on public health in Turkey, only a few studies have been carried out under laboratory conditions (Belen et al. 2004) and consequently, little is known about the external factors such as temperature on their developmental rates and longevity. Information regarding the effects of temperature on the development time of sand fly species will undoubtedly improve our knowledge of their population biology and also play a critical role in planning vector control programs. The aim of this study was to investigate the thermal requirements of P. papatasi under laboratory conditions using a degree day (DD) model, as reviewed by Pruess (1983) and Wagner et al. (1991), and estimate the developmental zero for each life stage. MATERIALS AND METHODS Sand fly colonization Sand flies used in this study were obtained from laboratory cultures maintained in the Ecological Science Research Laboratories (ESRL) of Hacettepe University, Ankara. We used the basic techniques for rearing insects that were described by Endris et al. (1982) and Modi and Tesh (1983), and modified by Ferro et al. (1998). P. papatasi colonies were reared under 27 ± 1° C, 65-75% relative

Vol. 30, no. 2

Journal of Vector Ecology

329

humidity and a 14:10 h (L:D) photoperiod regime. Calculation of thermal constant (K) and developmental zero (z) for P. papatasi Experiments conducted for calculating the thermal constant (K) and developmental zero (z) of P. papatasi were carried out in separate constant temperature growth chambers (Dedeoglu Model Lab-Ref 5, Ankara, Turkey) programmed at six different temperatures (15, 18, 20, 25, 28, 32° C), with a 14:10 h (L:D) photoperiod regime and a constant relative humidity of 60%. To avoid the effects of genetic differences among the populations, individuals used in the experiments were obtained from the second laboratory generations (F2) of field-collected parents (Wagner et al. 1991). To model development time as a function of temperature, groups of individuals (eggs) of identical initial age (cohorts) were reared at different constant temperatures. For this purpose newly emerged adult female sand flies were released into cloth cages (35 x 35 cm) with males and were maintained on a 30% sucrose solution under insectarium conditions for two days to allow mating. On the third day, flies were allowed to feed on a ketamine hydrochloride anesthetized hamster. Fully engorged females and an approximately equal number of males were transferred to small cloth cages (20 x 20 cm) for oviposition for five days. Cages were checked daily and gravid females and also males were confined in 425 ml polymehylpentene rearing jars with a 1.5 cm thick layer of moist Plaster of Paris on the bottom (Modi and Tesh, 1983) to provide eggs. These rearing jars were checked daily and 25 eggs of the same age were transferred by insect pin to 100 ml rearing jars and incubated in separate constant temperatures. Each temperature regime (15°, 18°, 20°, 25°, 28°, and 32° C) contained a total of 100 eggs. Rearing jars were examined at 24 h intervals at 10 x magnification using a binocular dissecting microscope. Once the eggs were hatched and the larvae were visible, a small quantity of standard larval diet was added. Any developmental changes and mortality were recorded. Because sand flies are not very synchronous in their development especially at lower temperatures, data obtained from the experiments were not normally distributed and any of the transformations managed to make data meet the assumptions of analysis of variance. For this reason, developmental times for each immature stage of P. papatasi under the different six temperatures were compared using the Kruskal-Wallis test, a non-parametric analogous to ANOVA. Pair-wise differences between rearing temperatures were compared by using the non-parametric Mann-Whitney U test. The linear regression technique described by Wagner et al. (1991) was used to estimate the developmental zero. Since distribution of data deviated statistically from normality, Model II regression was used. The average rate of development (1/d, where d = time spent in days by each developmental stage) and complete development was plotted versus temperature (T, °C). The developmental zero was determined by extrapolating the regression line to its intercept with the x axis (i. e., DZ = - a / b). The reciprocal of the slope (b) of the regression line was used to determine the degree-

days (DD) over the minimum threshold required for each of the developmental stages (i.e., DD = 1/b). All statistics were calculated in PAST (Hammer et al. 2004). RESULTS Effects of temperature on the development of immature stages of P. papatasi Development times of P. papatasi under different temperature regimes are shown in Table 1. None of the eggs maintained at 15° C managed to reach the pupal stage so no adult emergence was recorded for this temperature. At 18° C, only eight of the 100 eggs (8%) reached the pupal stage and a total of seven adults emerged. At the remaining temperatures, all specimens either emerged as adults or died during the experiments. Results of the Kruskal-Wallis tests showed that total development time of emerged adults revealed significant differences (P<0.05, df = 4) with development time decreasing with increasing temperature. Analyses showed the developmental time of each immature stage and the total duration of development from egg to adult to be significantly different between individuals reared at 20° C from individuals reared at 18° C and 20° C (P<0.001). The fastest development time was observed at 32 ° C for all immature stages as well as total development from egg to adult emergence. Pairwise analyses between investigated temperatures revealed significant differences in rate of egg development between all temperatures except for 15° C and 18° C (P = 0.009). Similarly differences in the rate of larval and pupal development were significant between all temperatures except for 25° C and 28° C for larvae (P = 0.557) and for pupae (P = 0.009). Total development of P. papatasi from egg to adult emergence was also significantly different in terms of examined temperatures (P< 0.001) except at 25° C and 28° C (P = 0.500). Percentage hatching of eggs of P. papatasi differed across temperatures. However, the majority of the eggs (76%) reared at 15° C failed to hatch while the highest hatching rate (72%) was observed at 20° C. For all temperatures, highest mortality rates were observed in the larval stage except for individuals reared under 32° C where the lowest survival rates (35%) were recorded in the egg stage. Pupal mortality was lower at all temperatures and with the highest mortality rates (24%) being observed at 32° C. Survival to adult emergence varied from 7% at 18° C to 38% at 28° C (Figure 1). Degree-day requirements and the developmental zero for P. papatasi Mean developmental rates (1/d) of individual developmental stages as well as total development times were found to be linearly correlated with temperature (Figure 2). Table 2 shows the parameters of linear regression equations for estimating the developmental zero (T0) and cumulative degree-days (DD) [Thermal Constant (K)] required for P. papatasi. The results of the regression analysis based on the development rates (1/d) and experimental temperature values showed that developmental zero values for P. papatasi varied from 11.6° C to 20.25° C depending on life stage (Table 2).

330

Journal of Vector Ecology

December 2005

Table 1. Mean development time (days ± SD) of each life stage of Phlebotomus papatasi reared at six constant temperatures (T,°C).

246.43 ± 13.83 (7) a 110.64 ± 56.96 (28) b 42.43 ± 2.96 (15) c 44.08 ± 11.49 (38) c 27.89 ± 1.88 (19) d

Total duration

(egg to adult)

Developmental zero values were lowest for egg hatching (11.6° C) while highest developmental zero values were recorded for the larval stages (19.81° C). It was estimated that the pupal stage required a minimum of 17.63° C while complete development from egg to adult required 20.25 °C. Calculated degree-day requirements for each life stage was lowest for pupa to adult (79.36 DD). However, development of egg and larvae indicated lower summations of 98.97 DD and 330.66 DD, respectively. Overall egg to adult degreeday requirements summed to 440.55 DD. DISCUSSION Our estimates of development times and survival for P. papatasi compare favorably with other published data for which the development times of immature stages of different sand fly species are highly variable, and ambient temperature is one of the most important factors affecting development time (Chaniotis 1967, Endris et al. 1984, Gosh et al. 1992, Guzman and Tesh 2000). Among the examined temperatures, no adult emergence was observed at the lowest temperature (15° C), while the fastest development occurred at the highest temperature (32° C) and a reduction in development time of immature stages was seen as temperature increased. The results of our studies were not unexpected taking into account the rule of Vant Hoff (Q10) which quantifies temperature dependence across a limited temperature range, since biochemical reactions (metabolic rate) increase with temperature between an upper and lower threshold value (Gillooly et al. 2001). Gosh et al. (1992) observed a similar pattern of temperature dependence of developmental time in colonyreared P. papatasi and P. argentipes; an increase in the temperature of the rearing chamber from 20° to 31° C resulted in faster development of eggs, larvae, and pupae. Similarly, a New World species Lutzomyia anthophora (Addis), which was exposed to different temperature regimes between 20° and 32° C under laboratory conditions, completed its total development from egg to adult faster as temperature increased (Endris et al. 1984). However, in our experiment the developmental time observed from egg to adult of P. papatasi at 25° C (42.93 days) was found to be shorter than expected and in comparison with 28° C, it was found that there was no significant statistical difference between the development Table 2. Linear regression parameters of development rate on temperature, estimated threshold temperatures for development [developmental zeros (z)], and thermal constants (K) [degree ­ days (DD)] for each immature stage of Phlebotomus papatasi.

Stage Egg Larva Pupa Egg to adult Slope Intercept r2 0.59 0. 32 0.61 0.25 z (°C) 11.60 19.81 17.63 20.25 DD (day) 98.97 330.66 79.36 440.55

Mean duration (±SD) of each stage (days)

-

23.68 ± 13.83 (7) a 23.45 ± 56.99 (28) b 9.64 ± 2.69 (15) c 5.40 ± 11.49 (38) c 4.81 ± 1.88 (19) d 18 20 25 28 32 16.01 ± 4.26 (57) a 10.26 ± 2.86 (72) b 8.23 ± 1.48 (60) c 6.76 ± 1.33 (70) d 5.62 ± 1.61 (35) e 206.74 ± 13.49 (8) a 76.92 ± 53.36 (33) b 25.05 ± 3.50 (17) c 31.92 ± 14.68 (47) c 17.46 ± 2.38 (25) d

n: numbers of individuals (n) used in calculations are shown in parentheses. Initial number of eggs was 100 at each temperature. The different letters show statistically significant differences (P<0.05) as deduced by the non ­ parametric Mann-Whitney U test. - denotes complete mortality occurred before molting to 2nd instar.

Pupa

Larva

Temperature (°C)

15

17.33 ± 4.45 (24) a

Egg

-

-

0.01010 - 0.11753 0.00302 - 0.05984 0.01260 - 0.22216 0.00226 - 0.04597

Vol. 30, no. 2

Journal of Vector Ecology

331

100 90 80 70 60 50 40 30 20 10 0

Egg

Larv ae

Pupae

Mortality (%)

15

18

20

25

28

32

Temperature (°C)

Figure 1. Percentage mortality of laboratory-reared Phlebotomus papatasi at a range of constant temperatures, estimated as proportion of individuals that developed into larvae, pupae, and adults.

times of larval stages and complete development (P>0.05). Gosh et al. (1992) also found P. papatasi to complete its total development in an average of 65 days at 25° C, and a minimum of 50 days and a maximum of 80 days have been reported for the same species of sand fly at 25° C (Guzman and Tesh 2000). According to previous studies on sand flies, immature stages may exhibit entirely different rates of development even if produced by two sisters, therefore the average duration of development of immature stages may vary even when exposed to identical conditions of temperature, humidity, and diet (Theodor 1934, Chaniotis 1967). Considering these findings, the observed fast development time at 25° C may be due to individual intrinsic factors, as previously shown. It was found in several studies that temperature, an important factor on development of sand flies, also influences the survival of immature stages. Gosh et al. (1992) observed that the rate of adult emergence at different temperatures between 24°-32° C increased rapidly with increasing temperature, but at 32° the rate of adult emergence was found to be significantly low because of the high mortality at the egg stage. In another study, it was found that high mortality rates occurred, particularly at the larval stages, due to the influence of low temperatures, 15° C and 18° C, while the highest survival of immature stages was observed at 28° C (Guzman and Tesh, 2000). Our results concur with these findings when survival and adult emergence rates are compared. Among the six different examined temperatures, no adult emergence was observed at 15° C since none of the larvae managed to reach the pupal stage. Similarly, adult emergence was low at 18° C (7%) due to high larval mortality. In accordance with other studies, we observed the highest rate of emergence at 28° C while a decrease was recorded at 32° C since most of the eggs (65%) did not hatch. According to the relative mortality values and development rates observed in this study, and based on the results obtained from other studies, we may conclude that the immature stages of P. papatasi have different tolerances to temperature, with the egg stage being more sensitive to higher temperatures while the larval stage is more sensitive to lower temperatures. This

may be the reason that Palearctic phlebotomine species overwinter as diapausing fourth instar larvae, whereas in warmer, wetter habitats diapause is in the egg stage (KillickKendrick, 1999). The results of this study show that the minimum threshold temperatures for egg, larvae, pupae, and total development were 11.6°, 19.81°, 17.63° and 20.25° C, respectively. Although there are no data regarding the exact thermal requirements of sand fly species, Theodor (1934) observed that larvae of P. papatasi maintained at 0 °C managed to live for 4.5 days while individuals maintained at -2 °C died after 2-3 days. The larvae of P. perniciosus exposed to low temperature conditions between 2°-10° C became completely immobile and survived for only a short time (Ready and Croset 1980). Although linear estimates of threshold temperatures based on developmental rate (1/d) tend to overestimate the true threshold (Wagner et al. 1991), our data corroborate those of Toprak1 who conducted a field study on the dynamics of P. papatasi populations in Sanliurfa province, the main cutaneous leishmaniasis focus of Turkey. Based on field samplings of adult P. papatasi, Toprak (2003) determined that the first adult population occurred in early-to-mid May. He also recorded that the highest population density was observed in September, but that in October the population density decreased rapidly, with the adult population disappearing completely in November. When we consider these field samplings alongside our laboratory observations, we can say that the results compare favorably with each other. In Sanliurfa province, diapausing larvae during the winter provides the accumulated degree-days necessary to complete their development, in mid-May when temperatures average 23.12° C. After these larvae complete their development, the first adults appear in May, second and third generations are Toprak, S. 2003. Investigations on sandflies (Diptera: Psychodidae) species occurring in Sanliurfa and the bioecology of cutaneous leishmaniasis vectors. PhD Thesis, Hacettepe Universitesi, Fen Bilimleri Enstitüsü, Ankara.

1

332

Journal of Vector Ecology

December 2005

0,06 0.4

1/y = -0.059844+0.0030243(x) S = 0.018497 R-Sq = 31.9% R-Sq (adj) = 31.7%

1/y = -0.11753+0.010104(x) S = 0.0331952 R-Sq = 59.3% R-Sq (adj) = 59.2%

0,05 0,04 0,03 0,02

0.3

0.2 0,01 0 -0,01

0.1

p y Development rate (1/y)

A

20 30 40

B

20 30 40

0.4

1/y = -0.023071+0.011301(x) S = 0.050199 R-Sq = 23.1% R-Sq (adj) = 22.9%

0,04

1/y = -0.04567+0.0022699(x) S = 0.0102143 R-Sq = 23.6% R-Sq (adj) = 23.4%

0,03

0.3

0,02 0.2 0,01 0.1 0 0

-0,01

C

20 30 40 20 30

D

40

Temperature (°C)

Figure 2. Linear regression analysis of developmental rate ± 95 % CI for Phlebotomus papatasi reared in the laboratory at constant temperatures. A) egg, z: 11.60 oC, K: 98.97 DD, B) larva, z: 19.81 oC, K: 330.66 DD, C) pupa, z: 17.63 oC, K: 79.36 DD, D) egg to adult, z: 20.25 oC, K: 440.55 DD.

Vol. 30, no. 2

Journal of Vector Ecology

333

produced subsequently since the average temperatures in June, July and August are sufficient for complete development (>20.25° C). Near the end of the September, third generation females are ready to oviposit. However, in Sanliurfa, the average temperature in September is still high enough for eggs to hatch (26. 63° C, z = 11.60° C, K = 98.97 degree-days), the average temperature in October cannot provide the degree ­ day accumulation needed for successful development of larvae (19.26 0C, z = 19.81, K = 330.66 degree - days). In response to a decline in temperature or day-length, these larvae probably become dormant and because of the lack of fourth generation adults, the population density start to decrease through October and finally, in November, the adult population of P. papatasi disappears in this province. In attempting to estimate developmental rates in nature, it is important to study the exact conditions of the insects' environment. But because the precise larval breeding sites of most phlebotomine species are unknown, it is difficult to measure or duplicate the exact conditions of their environment (Guzman and Tesh 2000). Although simple linear estimates of developmental zero (threshold) and degree-days required for development may be an oversimplification (Wagner et al. 1991), and sand fly development may be influenced by factors other than temperature, including photoperiod (Theodor 1934, Ready and Croset 1980) and food quality (Guzman and Tesh 2000), these models adequately describe development within the ranges applicable to field conditions (Pruess 1983). The control of vector-borne diseases has become possible on a large scale owing to the remarkable effectiveness of the integrated vector control programs that combine bio ­ ecological and chemical methods (WHO 1972). Our estimates of P. papatasi development can be used to model population dynamics and can help us in determining the total distributional range of the species and the distribution of leihmaniasis cases in correlation with the distribution of this vector species within Turkey. In addition, from a purely epidemiological perspective, our results will be helpful in constructing effective control programs. REFERENCES CITED Alptekin, D., M. Kasap, Ü. Lüleyap, H. Kasap, A. Aksoy, and M.L. Wilson. 1999. Sandflies (Diptera: Psychodidae) associated with epidemic cutaneus leishmaniasis in Sanliurfa, Turkey. J. Med. Entomol. 36: 277-281. Belen, A., B. Alten, and A.M. Aytekin. 2004. Altitudinal variation in morphometric and molecular characteristics of Phlebotomus papatasi populations. Med. Vet. Entomol. 18: 343-350. Chaniotis, B.N. 1967. The biology of California Phlebotomus (Diptera: Psychodidae) under laboratory conditions. J. Med. Entomol. 4: 221-233. Endris, R.G., P.V. Perkins, D.G. Young, and R.N. Johnson. 1982. Techniques for laboratory rearing of sand flies (Diptera: Psychodidae). Mosq. News 42: 400-407. Endris, R.G., D.G. Young, and J.F. Butler. 1984. The laboratory biology of the sand fly Lutzomyia anthophara (Diptera:

Psychodidae). J. Med. Entomol. 21: 656-684. Ferro, C., E. Cardenas, D. Corredor, A. Morales, and L.E. Munstermann. 1998. Life cycle and fecundity analysis of Lutzomyia shannoni (Dyar) (Diptera: Psychodidae). Mem. Inst. Oswaldo Cruz 93: 195-199. Gillooly, J.F., J.H. Brown, G.B. West, V.M. Savage, and E.L. Charnov. 2001. Effects of size and temperature on metabolic rate. Science 293: 2248-2251. Gosh, K.N., D.K. Gosh, A. De, A. Bhattacharya. 1992. Biology of P. argentipes Annandale and Brunetti and P. papatasi (Scopoli) in the laboratory. Ann. Parasitol. Hum. Comp. 67: 55-61. Guzman, H. and R.B. Tesh. 2000. Effects of temperature and diet on the growth and longevity of phlebotomine sand flies (Diptera: Psychodidae). Biomedica 20: 190- 199. Killick-Kendrick, R.K., 1999. The biology and control of plebotomine sand flies. Clin. Dermatol. 17: 279-289. Modi, B. and B.R. Tesh. 1983. A simple technique for mass rearing Lutzomyia longipalpis and Phlebotomus papatasi (Diptera: Psychodidae) in the laboratory. J. Med. Entomol. 2: 558-569. Ok, U. Z., I. C. Balcioglu, A. T. Ozkan, S. Ozensoy, and Y. Ozbel. 2002. Leishmaniasis in Turkey. Acta Tropica. 84: 43-48. Pruess, K.P. 1983. Day-degree methods for pest management. Environ. Entomol. 12: 613-619. Ready, P.D. and H. Croset. 1980. Diapause and laboratory breeding of Phlebotomus pernicious Newstead and Phlebotomus ariasi Tonnoir (Diptera: Psychodidae) from southern France. Bull. Entomol. Res. 70: 511-523. Svobodova, M., J. Sadlova, K. P. Chang, and P. Volf. 2003. Distribution and feeding preference of the sand flies Phlebotomus sergenti and P. papatasi in a cutaneous leishmaniasis focus in Sanliurfa, Turkey. Am. J. Med. Hyg. 68: 6-9. Theodor, O., 1934. Observations on the hibernation of P. papatasi. Bull. of Entomol. Res. 25: 459 ­ 472. Toprak, S. 2003. Investigations on sandflies (Diptera: Psychodidae) species occurring in Sanliurfa and the bioecology of cutaneous leishmaniasis vectors. PhD Thesis, Hacettepe Universitesi, Fen Bilimleri Enstitüsü, Ankara. Valevich, T.A. and T.I. Dergacheva. 1992. Laboratory breeding of P. papatasi reaction of the sandfly to unsatisfactory environmental conditions. Meditsinskaya Parazitologiya I Parazitarnye Bolenzi. 1: 18-21. Wagner, T.L., R.L. Olson, and J.L. Willers. 1991. Modeling arthropod development time. J. Agr. Entomol. 8: 251270. WHO. 1972. WHO Expert Commitee, Vector Ecology, Technical Report Series, No. 501, 41 pp. WHO. 2004. The vector-borne human infections of Europe. Their distribution and burden on public health. World Health Organization, Roll Back Malaria Programme, 144 pp. Yagci, S., S. Dincer, and H. Eren. 1998. Phlebotomus (Diptera : Psychodidae) species in Ankara area. Acta Parasitol. Turcica. 22 : 53-56.

334

Journal of Vector Ecology

December 2005

Scientific Note Vertical distribution of adult mosquitoes in native forest in Auckland, New Zealand

José G. B. Derraik 1, Amy E. Snell1, and David Slaney2

1

Ecology and Health Research Centre, Department of Public Health, Wellington School of Medicine and Health Sciences, University of Otago, P.O. Box 7343, Wellington, New Zealand 2 Institute of Environmental Science & Research Ltd, P.O. Box 50348, Porirua, New Zealand Received 15 July 2004; Accepted 2 December 2004

Mosquitoes have been shown to be active over a wide vertical range within the forest column (e.g. Haddow 1947), and their vertical stratification appears to be related to feeding or oviposition habits. Some mosquito species are active throughout the vertical forest strata from ground level to the top of the canopy, while others may display a marked preference for host-seeking high into the forest canopy or close to the ground (e.g. Snow 1974, Braack et al. 1994). There are 12 native and four established exotic mosquito species in New Zealand (Derraik 2004), and although there are no records of indigenously acquired cases of mosquitoborne disease in this country (Derraik and Calisher 2004), the vector status of most indigenous species is unknown. A comprehensive investigation of Culicidae in New Zealand is necessary, and large numbers of specimens should be collected and processed for virus isolation. However, one difficulty in collecting adults of many indigenous species could be a result of their vertical distribution within native forest habitats, and adults of Culex (Culex) asteliae Belkin, for instance, have never been collected in nature. Although there are isolated records describing the presence of some native species in the tree canopy (Pillai 1968), there is overall very little information on their vertical distribution in New Zealand forests. A study was therefore set up to assess the vertical stratification of adult mosquito activity within an area of native forest in northern New Zealand. This investigation was carried out in a relatively pristine native coniferous-broadleaved forest in the Cascade-Kauri Park (36° 53' 35'' S, 174° 30' 30'' E), located within the Waitakere Ranges Regional Park, West Auckland, New Zealand. The area is protected and undergoes regular pest control against introduced mammals, in particular brushtail possums (Trichosurus vulpecula). Four dry ice-baited light traps were used for sampling, and two native rimu trees (Dacrydium cupressinum; Podocarpaceae) located approximately 200 m apart and at least 20 m away from a walking track were selected for simultaneous placement of traps. On each tree, one trap was set as close to the ground as possible, while the other was set at a height of 10 m. The traps were set over five non-consecutive dry and relatively windless nights in April, 2003 (early austral autumn). Adult

Keyword Index: Mosquito, vertical distribution, New Zealand, forest.

trap catches between ground level and 10 m were compared using non-parametric tests (Kruskal-Wallis), and significance level used was P < 0.05. Note that although the statistical analyses were calculated at a per trap basis, the results are given at a per night basis for the sake of clarity. The total number of mosquitoes collected from both trees was very similar, with 516 and 507 mosquitoes recorded from trees 1 and 2, respectively (a total of 1,023 specimens in the 20 traps set up). The relative abundance was also similar for all individual species from both trees, and there were consequently no location effects on the community as a whole or on any individual species (all P-values were > 0.703). Overall, there were significantly more mosquitoes (P = 0.002) collected overnight at ground level than at a height of 10 m. At the species level however, this particular pattern was only observed for the endemic Ochlerotatus (Ochlerotatus) antipodeus (Edwards), which was over five times more abundant at ground level (623, mean = 124.6, SE = 7.9) than at 10 m (175, mean = 35.0, SE = 6.8) (P < 0.001; Figure 1). Ochlerotatus antipodeus largely dominated the community, making up 78% (798) of all mosquitoes recorded, and excluding this species from the overall yield there were significantly more mosquitoes caught at 10 m (P = 0.001). The only exotic species recorded at the site was Ochlerotatus (Finlaya) notoscriptus (Skuse), of which 22 specimens were recorded. The mean number of Oc. notoscriptus specimens collected overnight at ground level and at 10 m was small (2.8 and 2.4, respectively) and not significantly different (P = 0.342, Figure 1). No specimens of the endemic Cx. asteliae and Culex (Culex) pervigilans Bergroth were recorded at ground level, but both were found at 10 m (Figure 1) with the P-values for height differences being < 0.001 and 0.031, respectively. Only 18 Cx. pervigilans were recorded (mean = 3.4, SE = 1.8), while Cx. asteliae had 86 specimens (mean = 15.6, SE = 4.9) recorded in all but one of the 10 canopy traps set. Culiseta (Climacura) tonnoiri (Edwards) was another endemic species that was significantly more abundant at 10 m than at ground level (P = 0.024), with 41 (mean = 8.2, SE = 1.5) and 18 (mean = 3.6, SE = 1.6) specimens recorded, respectively (Figure 1). The only native species for which there was no significant difference between the two heights was Coquillettidia (Coquillettidia) iracunda (Walker) (P = 0.259),

Vol. 30, no. 2

Journal of Vector Ecology

335

25

140

Mean no. adult specimens

Ground 10 m

20

120 100

15

80 60 40

10

5 20 0

Cq. iracunda Cx. asteliae Cx. pervigilans Cs. tonnoiri Oc. notoscriptus

0

Oc. antipodeus

Figure 1. Mean number of adult mosquitoes collected from two traps at each of 5 nights in native forest at ground level and 10 m. Error bars represent the standard error for the mean.

for which 8 and 15 specimens were collected at ground level and 10 m, respectively (Figure 1). The exotic Oc. notoscriptus was recorded both at ground level and 10 m (Figure 1), and similar results were obtained in a small investigation in the Wellington region (Derraik et al. 2003). This species was also caught in biting catches from a height of 18 m in native forest canopy, at a site where the species was also viciously biting at ground level (Derraik 2005). Therefore, there is no doubt that this species is active over a relatively large vertical stratum within native forests in New Zealand. In its native Australia, Oc. notoscriptus seems to feed almost entirely on mammals, although poultry is also included among its potential hosts (Lee and Bugledish 1999). Whether this species feeds on native birds in New Zeland is unknown, but its presence at Cascade-Kauri Park could be an indication that it does. Among the native species, no Cx. pervigilans were collected in this study from ground traps, in comparison to the 17 specimens recorded within the forest canopy (Figure 1). In contrast, this species was recorded at the Wellington Zoo in both ground level and 10 m traps (Derraik et al. 2003). Culex pervigilans is New Zealand's most abundant and widespread mosquito species (Laird 1990), and it seems to thrive particularly well in urban and rural areas, typically away from forest covers. It is possible that Cx. pervigilans may display heterogeneous patterns of activity as a result of phenotypic or genotypic variation. Populations from modified environments may have adapted to feeding near the ground to exploit the abundant blood meals provided by the numerous animals present at ground level (e.g. humans, cattle, and other domestic animals), in contrast to canopy feeding within areas of native forest. Culiseta tonnoiri was the other native New Zealand species for which published height distribution data were available, and Pillai (1968) frequently captured females in

traps baited with bantam fowl (Gallus domesticus) in native forest canopy as high as 33 m. The results from this study support the idea that this species is active in various forest strata but seems to prefer feeding within the tree canopy (Figure 1). This pattern of activity seems to reflect the wide range of hosts on which the Cs. tonnoiri may feed including humans, many domesticated animals, and both canopyfrequenting and ground-dwelling birds (Pillai 1968, Crosby 1978). One interesting pattern was found in the native mosquito Cx. asteliae, whose adults were collected in nature for the first time in this study. This species was entirely absent from ground traps but was relatively common within the tree canopy (Figure 1). The absence of previous collection records of Cx. asteliae adults is most likely a result of this species' marked preference for canopy activity. Interestingly however, larvae of Cx. asteliae are widespread and abundant in the leaf axils of the native plant Collospermum hastatum (Liliaceae) on the ground and in the forest canopy, where they have been collected as high as 18 m (Derraik 2005). As pointed out by Woodward et al. (1996), the vertical stratification of oviposition preferences may not necessarily be correlated with that of adult mosquito activity. Unlike the other native mosquito species observed, Oc. antipodeus was predominantly found at ground level and Cq. iracunda showed no preference for a particular stratum (Figure 1). Apart from the fact that both species bite man (Derraik and Snell 2004, Holder et al. 1999), no information is available regarding their natural hosts, and for Oc. antipodeus at least, according to Belkin (1968) there are no records of the species biting cattle or other animals. The results indicated that mosquitoes seem to be vertically stratified in the native forest site studied. The nature of New Zealand's pre-human fauna (no native land mammals apart from three bat species) means that mosquito species in native

336

Journal of Vector Ecology

December 2005

forests would have adapted to feeding on birds. The latter seem to be now mostly restricted to the forest canopy as anthropogenic impacts have led to the displacement of the majority of New Zealand's ground-dwelling bird fauna and the extinction of 40% of all avian species (Holdaway et al. 2001). Although some mosquito species are opportunistic and will feed on a variety of hosts, blood-feeding preferences of other species can be very host-specific (Service 1971). As a result, the apparent canopy-feeding habits of many native mosquitoes might be a result of avian host availability occurring mostly in the forest canopy. Unfortunately, as with other aspects regarding the ecology of New Zealand mosquitoes, very limited information is available on their host preferences (Belkin 1968, Holder et al. 1999). It is possible therefore, that in other patches of native forest where introduced mammals are abundant (in particular, brushtail possums), these may also become hosts to native mosquitoes. Acknowledgments Thanks to Dave Galloway and Scott De Silva (Auckland Regional Council) for providing logistical assistance. Thanks also to Craig Williams (James Cook University, Australia) and two anonymous reviewers for valuable comments on a previous version of this manuscript. The University of Otago provided funding support. REFERENCES CITED Belkin, J.N. 1968. Mosquito Studies (Diptera: Culicidae) VII. The Culicidae of New Zealand. Contrib. Am. Entomol. Inst. 3: 1-182. Braack, L.E.O., M. Coetzee, R.H. Hunt, H. Biggs, A. Cornel, and A. Gericke. 1994. Biting pattern and host-seeking behavior of Anopheles arabiensis (Diptera: Culicidae) in Northeastern South Africa. J. Med. Entomol. 31: 333339. Crosby, T.K. 1978. A record of Culiseta tonnoiri (Diptera: Culicidae) biting the penguin Eudyptes pachyrhynchus (Aves: Spheniscidae). N. Z. J. Zool. 5: 811-812. Derraik, J.G.B. 2004. Exotic mosquitoes in New Zealand: a review of species intercepted, their pathways and ports of entry. Aust. N. Z. J. Publ. Hlth. 28: 433-444.

Derraik, J.G.B. 2005. Presence of Culex asteliae larvae and Ochlerotatus notoscriptus adults (Diptera: Culicidae) in native tree canopy in the Auckland region. The Weta 29: (In press). Derraik, J.G.B. and C.H. Calisher. 2004. Is New Zealand prepared to deal with arboviral diseases? Aust. N. Z. J. Publ. Hlth. 28: 27-30. Derraik, J.G.B., D. Slaney, P. Weinstein, P. Lester, and G. Purdie. 2003. Presence of adult Ochlerotatus (Finlaya) notoscriptus (Skuse) and Culex (Culex) pervigilans Bergroth (Diptera: Culicidae) in tree canopy in Wellington, New Zealand. N. Z. Entomol. 26: 105-107. Derraik, J.G.B. and A.E. Snell. 2004. Notes on daytime biting catches of mosquitoes (Diptera: Culicidae) in native forest sites in the Auckland region. The Weta 28: 14-19. Haddow, A.J. 1947. The mosquitoes of Bwamba County, Uganda. V. The vertical distribution and biting-cycle of mosquitoes in rain-forest, with further observations on microclimate. Bull. Entomol. Res. 37: 301-330. Holdaway, R.N., T.H. Worthy, and A.J.D. Tennyson. 2001. A working list of breeding bird species of the New Zealand region at first human contact. N. Z. J. Zool. 28: 119-187. Holder, P., G. Browne, and M. Bullians. 1999. The mosquitoes of New Zealand and their animal disease significance. Surveillance 26: 12-15. Laird, M. 1990. New Zealand's Northern Mosquito Survey, 1988-89. J. Am. Mosq. Contr. Assoc. 6: 287-299. Lee, W. and E.-M.A. Bugledish. 1999. Culicidae. In: E.-M.A. Bugledish (ed.), Diptera: Nematocera. Zoological Catalogue of Australia. pp. 161-239. CSIRO Publishing, Melbourne. Pillai, J.S. 1968. Notes on mosquitoes of New Zealand. II. The male terminalia of Culiseta (Climacura) tonnoiri and its ecology (Diptera: Culicidae). J. Med. Entomol. 5: 355357. Service, M.W. 1971. Feeding behaviour and host preferences of British mosquitoes. Bull. Entomol. Res. 60: 653-661. Snow, K.R. 1974. Insects and disease. Routledge and Kegan Paul Ltd, London. Woodward, D.L., A.E. Colwell, and N.L. Anderson. 1996. Temporal and spatial distribution of Aedes sierrensis oviposition. Proc. Mosq. Vect. Contr. Assoc. Calif. 64: 51-62.

Vol. 30, no. 2

Journal of Vector Ecology

337

Scientific Note Outbreak of dengue in National Capital Territory of Delhi, India during 2003

R.S. Sharma, P.L. Joshi, K.N. Tiwari, Rakesh Katyal, and Kuldip Singh Gill

National Vector Borne Diseases Control Programme, Delhi ­ 110 054, India Received 13 April 2004; Accepted 29 November 2004 Dengue and dengue haemorrhagic fever continue to be major infectious diseases of public health importance in countries of the western Pacific and Southeast Asia regions. These regions are experiencing a geographical spread, both in terms of distribution of the virus and the mosquito vector, with an increase in the frequency of epidemics. Since 1963, outbreaks of dengue/DHF have been recorded in almost all parts of India except the northeastern region. In all the outbreaks the main mosquito involved in transmission was Aedes aegypti. However, during the outbreak in Kerala state during 2004, the vector involved in the transmission was Aedes albopictus. The first outbreak of dengue fever in India with hemorrhagic manifestation was reported in Calcutta City. An increasing trend of dengue outbreaks accompanied by dengue hemorrhagic fever poses a problem of utmost importance to public health in India (WHO 1999). Dengue fever outbreaks have been reported from various parts of the country during the past 30-40 years (Yadav and Narsimham 1992). A severe outbreak of dengue hemorrhagic fever swept National Capital Territory, Delhi in 1996 where 10,252 cases and 423 deaths due to DHF were recorded in various parts of Delhi (Kaul et al. 1998, Sharma et al. 1999), and dengue has remained endemic in Delhi for several years.The first DHF outbreak was reported in 1988 with 33% mortality among children admitted in hospitals (Kabra et al. 1992). The principal vector of dengue fever, Ae. aegypti, is prevalent in all cities and towns of India. The gangetic plain of North India is also infested with Ae. aegypti (Rao, 1967). Krishnamurthy et al. (1965) and Katyal et al. (1996) carried out a comprehensive survey of Ae. aegypti populations in Delhi. Sharma et al. (2001) reported the degree of Aedes prevalence in hospitals and schools in Delhi. Although Ae. aegypti has been known to be widely distributed in several countries of Southeast Asia and its importance as potential dengue fever vector has been long recognized, the information on its prevalence and shifting trend of breeding places is still fragmentary (Kalra et al. 1998). Vector surveillance is an important tool for generating entomological data for suggesting appropriate control strategies and developing an early warning system (Pant and Self 1993). These studies on the incidence of dengue and prevalence of Ae. aegypti were conducted in Delhi during 2003. Delhi, the capital of the Republic of India, is situated on the bank of the river Yamuna at approximately 77.15E and 26.15 N. It occupies a 1,485 km2 area of which 900 km2 is classified as urban. The city is the center of vast economic opportunities, attracting many migrants and registering a phenomenal population growth. The population of Delhi, now estimated to be above 10 million, has grown at the rate of 64.2%, 54.6%, and 57.1% in the decades 1951-1961, 19611971, and 1971-1981 respectively. In Delhi, three agencies, namely the Municipal Corporation of Delhi (MCD), New Delhi Municipal Committee (NDMC), and Defense are responsible for dengue control activities inside their own areas. The MCD, with its 12 zones, covers the largest part. Larval surveys were carried out in all the localities of the city irrespective of the risk of dengue/DHF in that locality. Searches were made for Aedes breeding in all types of breeding habitats during 2003 in the areas covered by the MCD, NDMC, Railways, and Delhi Cantonment. During the past several outbreaks of dengue/DHF, the initial cases have originated mainly from hospital and school premises. Stratification of Delhi was according to areas that differed from one another for Aedes breeding potential. Entomological indices including House Indices (HI), Container Index (CI), and Breateau Index (BI) were used for measuring the larval population:

House Index Container Index Breateau Index No. of houses positive x 100 No. of houses inspected No. of containers positive x 100 No. of containers inspected No. of containers positive x 100 No. of houses inspected

The dengue cases and deaths in Delhi from 1996 to 2003 are shown in Table 1. After the dengue/DHF outbreak in 1996, cases and deaths gradually declined until 2002. During 2003, the cases increased in an outbreak form with 2,882 cases, and 35 deaths and dengue cases were reported from almost all the zones of Delhi (Table 2). The maximum cases were reported from central Shahdra south, Shahdra north, and Central regions. Previously, the rural zones were not contributing as many cases and dengue was mainly an urban disease. The relative abundance of breeding infestation and Ae. aegypti indices are shown in Table 3. The maximum Ae. aegypti breeding was reported in the month of September, a post-monsoon month. The breeding infestation was high from July to September. Key breeding sites of Ae. aegypti larvae differed from area to area. Aedes aegypti larvae were mainly found in coolers

338

Table 1. Dengue cases and deaths in Delhi.

Journal of Vector Ecology

December 2005

Table 3. Breeding infestation of Aedes aegypti in Delhi during 2003.

Year 1996 1997 1998 1999 2000 2001 2002 2003

Cases 10,252 273 332 168 180 322 42 2,882

Deaths 423 1 5 2 2 3 0 34

(45%), earthen pots (15%), tires (21%), cement tanks (5%), junk yards (10%), and flower pots (4%). Entomological indices of Ae. aegypti began to increase from July until October, with peaks in the months of August and September, and thereafter declined. The rise in breeding indices during the post-monsoon season may be attributed to increases in potential breeding containers due to rains. The recent outbreak was reported at a very low vector density, which was below the critical index that Pant and Self (1993) have cited in several references showing the relationship of the `larval house index' to outbreaks of dengue. In Singapore there was only one instance where the house index was 9% during outbreak of dengue/DHF in 1996. In Southeast Asia, a strong association between dengue vector and rainfall has been established (Gould et al. 1970). In Delhi, the main sources of Aedes population are the multi-storeyed buildings with water coolers that are difficult to access for maintenance. The present study showed that water coolers and tires constituted 66% of Aedes larval population. Thus, these breeding habitats should be surveyed at weekly intervals particularly in post-monsoon season followed by effective anti-larval measures.

Month January February March April May June July August September October November December TOTAL

THC* 6,651 5,256 4,996 5,336 5,135 5,699 6,805 5,695 5,830 6,374 5,307 5,153 68,237

THP 2 1 2 2 18 18 235 305 327 142 26 1,024

TCC 9,825 7,844 7,401 7,652 7,617 8,522 10,984 8,935 8,661 8,760 7,519 7,327 10,2047

TCP 2 1 2 2 18 19 339 407 473 160 29 1,452

*THC: Total houses checked; THP: Total houses positive; TCC: Total containers checked; TCP: Total containers positive.

REFERENCES CITED Gould, D.J., J.A. Mount, J.E. Scanlon, H.R. Ford, and M.F. Sullivan. 1970. Ecological control of dengue vectors on and island in the gulf of Thailand. Am. Trop. Med. Publ. Hlth. 18: 295-302. Kabra, S.K., I.C. Verma, N.K. Arora, Y. Jain, and V. Kalra. 1992. DHF in children in Delhi. Bull. Wld. Hlth. Org. 45F: 105-108. Kalra N.L., B.L. Wattal and N.G.S. Raghvan. 1968. Distribution patterns of Aedes aegypti in India ­ some ecological considerations. Bull. Ind. Mal. Com. Dis. 93: 307-344. Katyal, R., K. Singh, and K. Kumar. 1996. Seasonal variations in Aedes aegypti population in Delhi, India, Dengue Bull. 20: 78-81. Kaul S.M., R.S. Sharma, S.N. Sharma, N. Panigrahi, P.K. Phukan, and S. Lal. 1998. Preventing dengue and DHF: The role of entomological surveillance. J. Comm. Dis. 30: 187-192. Krishnamurthy, B.S., N.L. Kalra, G.C. Joshi, and M.M. Singh. 1965. Reconnaissance survey of Aedes mosquitoes in Delhi. Bull. Ind. Soc. Mal. Comm. Dis. 2: 56-67. Pant C.P. and L.S. Self. 1993. Vector ecology and bionomics. In: Monograph on dengue/ DHF WHO-SEARO Publication. 22: 121. Sharma R.S., N. Panigrahi, and S.M. Kaul. 1999. Aedes aegypti prevalence in hospitals and schools, the priority sites for DHF transmission in Delhi. Dengue Bull. 23: 109-112. Rao, T.R. 1967. Distribution, density and seasonal prevalence of Aedes aegypti in India subcontinent and South East Asia. Bull. Wld. Hlth. Org. 36: 547-557. Yadava R.L and M.V.V.L. Narsimham. 1992. Dengue/DHF and its control in India. Dengue News. 17.3.8. WHO prevention and control of dengue and DHF. 1999 WHO- SEARO Regional Publication. 29 pp.

Table 2. Percentage of dengue cases reported from NCT of Delhi by zone during 2003.

Zone Shahdara (North) Shahdara (South) Civil line Rohini City Sadar Pahar Ganj Karol Bagh West Central South Narela Najafgarh NDMC Delhi Cant/Railway

Percentage 13 11 10 8 3 3 6 7 13 9 2 11 3 1

Vol. 30, no. 2

Journal of Vector Ecology

339

Scientific Note Molecular evidence for novel Bartonella species in Trichobius major (Diptera: Streblidae) and Cimex adjunctus (Hemiptera: Cimicidae) from two southeastern bat caves, U.S.A.

Will K. Reeves1, , Amanda D. Loftis1, Jeffery A. Gore2, and Gregory A. Dasch1

Centers for Disease Control and Prevention, Viral and Rickettsial Zoonoses Branch, 1600 Clifton Rd NE, Mailstop G-13, Atlanta, GA 30333, U.S.A. 2 Terrestrial Mammals Research Program, Fish and Wildlife Research Institute, Florida Fish and Wildlife Conservation Commission, 3911 Highway 2321, Panama City, FL 32409, U.S.A.

1

Received 28 March 2005; Accepted 2 August 2005

Bartonella are Gram-negative bacteria that infect the erythrocytes of vertebrates and are transmitted by blood feeding arthropods (e.g. Baker 1946, Chomel et al. 1996, Chang et al. 2001, Comer et al. 2001). There are at least 16 valid species or subspecies of Bartonella, and nine of these are associated with diseases in humans (Ciervo and Ciceroni 2004, Dehio et al. 2004). The arthropod vectors and transmission cycles of most Bartonella spp. are unknown, but hematophagous Diptera are potential vectors and should be investigated if they feed on infected hosts. Blood feeding Muscidae, such as Stomoxys calcitrans (Linnaeus), were implicated in the transmission of Bartonella spp. to cattle in California (Chung et al. 2004). Lutzomyia spp. (Diptera: Psychodidae) are vectors of B. bacilliformis (e.g. Alexander 1995) but have not been shown to transmit other species of Bartonella. Hippoboscid flies are potential vectors of Bartonella spp. to ruminants, including B. chomelii and B. schoenbuchensis (Halos et al. 2004). Bartonella schoenbuchensis was isolated from the hippoboscid Lipoptena cervi and might be the cause of deer ked dermatitis in humans (Dehio et al. 2004). Although bats sometimes roost in buildings occupied by humans, the vector potential of their ectoparasites is poorly studied. Carios kelleyi, a bat tick, was found to harbor Bartonella henselae and a novel spotted fever group Rickettsia sp. (Loftis et al. 2005). The cimicid ectoparasites of bats are vectors of Trypanosma spp. and nycteribiid bat flies transmit Polychromophilus spp. to bats (Gardner and Molyneux 1988a,b). Streblid bat flies are not known to transmit pathogens but feed repeatedly on multiple bats and often fly from host to host in roosts. As a result, streblids might serve as mechanical or biological vectors of pathogens and maintain pathogens within bat colonies. Streblids will bite humans in bat roosts (Lloyd 2002). We screened ectoparasitic arthropods from the southeastern myotis, Myotis austroriparius (Rhoads), for species of Bartonella and Rickettsia. Ectoparasitic arthropods were collected from M. austroriparius, in Florida Caverns

State Park, Jackson County, FL (30.81° N, 85.23° W), on 12 October 2001 and preserved in 90% ethanol. Two species of bat flies, Basilia boardmani Rozeboom (Diptera: Nycteribiidae) and Trichobius major Coquillett (Diptera: Streblidae), were in the collection. Additional collections of four B. boardmani and 14 Cimex adjunctus Barber (Hemiptera: Cimicidae) were made from Santee Caves, Orangeburg County, SC (33.49°N, 80.47°W), from guano below a M. austroriparius colony, on 22 September 2004. One B. boardmani and T. major from Florida Caverns and all of the ectoparasites from Santee Caves were frozen in liquid nitrogen and crushed with a sterile Teflon pestle. Total DNA was extracted from the pulverized remains with an IsoQuick Nucleic Acid Extraction Kit (ORCA Research Inc., Bothell, WA) and rehydrated in DNAse free water. The extracts were tested for the presence of DNA from Bartonella and Rickettsia by polymerase chain reaction (PCR) amplification using QHEV1, QHEV4, 17kDF1, and 17kDR1 primers to amplify a fragment of the 16s rDNA to 23s rDNA intergenic spacer (ITS) region of Bartonella or 17 kD antigenic gene of Rickettsia as described by Houpikian and Raoult (2001) and Carl et al. (1990). We attempted to amplify fragments of the 60 kDa heat shock protein (groEL) and riboflavin synthase (ribC) genes of Bartonella, using the primers described by Zeaiter et al. (2002) and Johnson et al. (2003), from ectoparasites determined to harbor Bartonella. All initial stock PCR and sequencing primers were at 20 m. Each PCR tube contained 12.5 l of Taq PCR Master Mix (Qiagen, Valencia, CA), 7.5 l of nuclease free water, 1.25 l of each primer, and 2.5 l of DNA extract in water. PCR products were separated by 2% agarose gel electrophoresis and visualized under ultraviolet light with ethidium bromide. Positive and negative controls were used in all reactions and consisted of genomic DNA extracts of B. henselae, Rickettsia rickettsii, or distilled water. Products were purified with a QIAquick PCR Purification Kit (Qiagen, Valencia, CA). Sequencing reactions were performed with a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems,

Keyword Index: Bartonella, Cimicidae, Hippoboscidae, Nycteribiidae, Streblidae.

340

Journal of Vector Ecology

December 2005

Foster City, CA) using PCR primers, and excess dye was removed with a DyeEx 2.0 column (Qiagen, Valencia, CA). Sequences were determined using an ABI 3100 capillary sequencer (Applied Biosystems, Foster City, CA) and assembled with Seqmerge (Accelrys, San Diego, CA). Each sequencing reaction was duplicated, primer sequences were removed, and sequences were compared to those in GenBank using the BLAST 2.0 program (NCBI, Bethesda, MD). ITS sequences from the Bartonella spp. were aligned with each other using Clustal W version 1.8 (Supercomputer Laboratory, Kyoto University, Japan). DNA from the ITS region of Bartonella was detected by PCR in a single T. major from Florida Caverns and in one out of 14 C. adjunctus from Santee Caves. The 575-bp and 504bp ITS sequences from these two Bartonella spp. were unique when compared to sequences in GenBank. The unique sequences probably represent separate undescribed taxa, based on the limited 75% similarity shared by the two amplicons. This is the first report of Bartonella from a streblid or cimicid. A 768 bp amplicon of the groEL gene was sequenced from a Bartonella in T. major, but primers failed to amplify products from the bacterium in C. adjunctus. The groEL sequence from Bartonella in T. major was most similar (91%) to that of an undescribed Bartonella from Sciurus carolinensis Gmelin (GenBank # AF449763) in the United Kingdom. PCR amplification, using the riboflavin synthase primers described by Johnson et al. (2003), was not successful in amplifying DNA from either Bartonella spp. DNA from Rickettsia was not detected. The presence of Bartonella in T. major and C. adjunctus does not demonstrate infection of the ectoparasites or vector competence; the DNA from Bartonella could have originated in an undigested blood meal. However, the presence of Bartonella in a blood feeding arthropod suggests that they might transmit this agent and could play a role in the natural transmission cycles of these agents. Bartonella spp. can be transmitted to new hosts in feces from infected arthropods (Durden 2002). The presence of bats and their ectoparasites in buildings might expose people to potentially infectious Bartonella in arthropod feces. An alternative hypothesis is that the Bartonella are endosymbionts similar to those described by Aschner (1946) and Bequaert (1953) in hippoboscid and nycteribiid flies. Further research and experimentation are needed to demonstrate vector capacity of either ectoparasite for Bartonella. The DNA sequences for the ITS regions of Bartonella amplified from T. major and C. adjunctus were assigned GenBank accession numbers AY787021 and AY824947, respectively. The GenBank accession number for the groEL sequence from the Bartonella in T. major is AY843209. Voucher specimens of B. boardmani, C. adjunctus, and T. major were deposited in the Clemson University Arthropod Collection. Acknowledgments We thank P.H. Adler for providing laboratory space used during part of this project, the American Society for

Microbiology for partial funding of this project, the Florida Park Service for access to bat roosts, and W. Wills for assisting with collections in Santee State Park. REFERENCES CITED Alexander, B. 1995. A review of bartonellosis in Ecuador and Colombia. Am. J. Trop. Med. Hyg. 52: 354-359. Aschner, M. 1946. The symbiosis of Eucampsipoda aegyptia Mcq. (Diptera Pupipara: Nycteridiidae). Bull. Soc. Fouad. Entomol. 30: 1-6. Baker, J.A. 1946. A rickettsial infection in Canadian voles. J. Exp. Med. 84: 37-50. Bequaert, J.C. 1953. The Hippoboscidae or louse-flies (Diptera) of mammals and birds. Part I. Structure, physiology and natural history. Entomol. Am. 32: 1-209. Carl, M., C.W. Tibbs, M.E. Dobson, S. Paparello, and G.A. Dasch. 1990. Diagnosis of acute typhus infection using the polymerase chain reaction. J. Infect. Dis. 161: 791793. Chang, C.C., B.B. Chomel, R.W. Kasten, V. Romano, and N. Tietze. 2001. Molecular evidence of Bartonella spp. in questing adult Ixodes pacificus ticks in California. J. Clin. Microbiol. 39: 1221-1226. Chomel, B.B., R.W. Kasten, K. Floyd-Hawkins, B. Chi, K. Yamamoto, J. Roberts-Wilson, A.N. Gurfield, R.C. Abbott, N.C. Pedersen, and J.E. Koehler. 1996. Experimental transmission of Bartonella henselae by the cat flea. J. Clin. Microbiol. 34: 1952-1956. Chung, C.Y., R.W. Kasten, S.M. Paff, B.A. Van Horn, M. Vayssier-Taussat, H.J. Boulouis, and B.B. Chomel. 2004. Bartonella spp. DNA associated with biting flies from California. Emerg. Infect. Dis. 10: 1311-1313. Ciervo, A. and L. Ciceroni. 2004. Rapid detection and differentiation of Bartonella spp. by a single-run real time PCR. Mol. Cell. Probes. 18: 307-312. Comer, J.A., C.D. Paddock, and J.E. Childs. 2001. Urban zoonoses caused by Bartonella, Coxiella, Ehrlichia, and Rickettsia species. Vector Borne Zoon. Dis. 1: 91-118. Dehio, C., U. Sauder, and R. Hiestand. 2004. Isolation of Bartonella schoenbuchensis from Lipoptena cervi, a blood-sucking arthropod causing deer ked dermatitis. J. Clin. Microbiol. 42: 5320-5323. Durden, L.A. 2002. Lice (Phthiraptera). In: G. Mullen and L. Durden, (eds.) Medical and Veterinary Entomology. Academic Press, Amsterdam. pp. 45-65. Gardner, R.A. and D.H. Molyneux. 1988a. Polychromophilus murinus: a malarial parasite of bats: life-history and ultrastructural studies. Parasitology 96: 591-605. Gardner, R.A. and D.H. Molyneux. 1988b. Trypanosoma (Megatrypanum) incertum from Pipistrellus pipistrellus: development and transmission by cimicid bugs. Parasitology 96: 433-447. Halos, L., T. Jamal, R. Maillard, B. Girard, J.Guillot, B. Chomel, M.Vayssier-Taussat, and H.J. Boulouis. 2004. Role of Hippoboscidae flies as potential vectors of Bartonella spp. infecting wild and domestic ruminants. Appl. Envir. Microbiol. 70: 6302-6305.

Vol. 30, no. 2

Journal of Vector Ecology

341

Houpikian, P. and D. Raoult. 2001. 16S/23S rRNA intergenic spacer regions for phylogenetic analysis, identification, and subtyping of Bartonella species. J. Clin. Microbiol. 39: 2768-2778. Johnson, G., M. Ayers, S.C.C. McClure, S.E. Richardson, and R. Tellier. 2003. Detection and identification of Bartonella species pathogenic for humans by PCR amplification targeting the riboflavin synthase gene (ribC). J. Clin. Microbiol. 41: 1069-1072. Lloyd, J.E. 2002. Louse flies, keds, and related flies (Hippoboscoidea). In: G. Mullen L. and Durden. (eds.)

Medical and Veterinary Entomology. Academic Press, Amsterdam. pp. 349-362. Loftis, A.D., J.S. Gill, M.E. Schriefer, M.L. Levin, M.E. Eremeeva, M.J.R. Gilchrist, and G.A. Dasch. 2005. Detection of Rickettsia, Borrelia, and Bartonella in Carios kelleyi (Acari: Argasidae). J. Med. Entomol. 42: 473-480. Zeaiter, Z., P.E. Fourneir, H. Ogata, and D. Raoult. 2002. Phylogenetic classification of Bartonella species by comparing groEL sequences. Int. J. Syst. Evol. Microbiol. 52: 165-171.

342

Journal of Vector Ecology

December 2005

Scientific Note A method for determining the sex of larval Aedes aegypti mosquitoes

Gail M. Chambers

Division of Entomology, University of Idaho, Moscow, Idaho 83844-2339 U.S.A. Received 14 June 2005; Accepted 23 August 2005

Studies of mosquito development must often determine the sex of preimaginal stages. Although female and male Aedes aegypti (L.) larvae are similar in size at the beginning of the 3rd instar, females subsequently grow more rapidly than males when well-nourished. By the middle of the 4th instar, the sexes can be fairly accurately distinguished on the basis of size. If they are sub-optimally nourished, however, the size differences of female and male larvae and pupae are not as pronounced. Some protocols allow the larva to be reared to the pupal stage where females and males are easily distinguished on the basis of external structures: the cerci and 9th sternite of the female and the gonocoxites of the male (Christophers 1960). Other protocols, however, require that the larva be killed for analysis. I present here details of a method that I refined for accurately determining the sex of intact 4th instar larvae of Aedes aegypti (Chambers and Klowden 1990) and that has been used successfully by others (Timmermann and Briegel 1999). Determination of the sex of intact larvae depends on identification of developing male secondary reproductive structures within the anal (9th) segment (Figure 40 in Christophers 1960). Because female structures cannot be readily seen in intact larvae, females are identified by the absence of male structures. A pipette is used to place the larva on a glass slide. Chilling the slide may help immobilize the

larva. It is important that the larvae are not injured, because the internal structures cannot be seen if turgor is not maintained. After most of the water is blotted away, a fine sable watercolor brush (size 00) is used to gently position the larva ventral side up with the anal segment extended over the respiratory siphon. The anal segment is then examined at 100x under a compound microscope and transmitted light. The male structures can be identified by focusing down through the ventral cuticle and the sclerotized saddle that wraps around all but the mid-ventral portion of the segment. The rudiments of the gonocoxite, ejaculatory duct, and vas deferens are discernable under the cuticle. In the late 3rd and early 4th instars they have the appearance of two fried eggs in the anterior portion of the segment. They may be touching at the midline (Figure 1A) or may be widely separated (Figure 1B), depending on the turgidity of the segment. The gonocoxal rudiments of mid-to-late 4th instar larvae are best identified by focusing on their outlines and on the cleft between them. At mid-4th instar the gonocoxal rudiments are much larger though still rounded in outline (Figure 2A). By the late fourth instar they are more elongate and extend to the posterior end of the segment (Figure 2B). Attempts to determine the sex of larvae of other species were not successful because the saddle and other sclerotized cuticular structures of the anal segment

A

B

Figure 1. Ventral view of anal segment of early 4th instar male Aedes aegypti larvae. Arrows point to the male structures visible beneath the cuticle.

Vol. 30, no. 2

Journal of Vector Ecology

343

A

B

Figure 2. Ventral view of anal segment of mid (A) and late (B) 4th instar male Aedes aegypti larvae.

obscure the gonocoxal rudiments. REFERENCES CITED Acknowledgments I thank Dr. Troy Ott, Department of Animal and Veterinary Sciences, University of Idaho, for the use of his compound microscope with digital imaging equipment. This material is based upon work supported by the National Science Foundation under Grant No. IBN-0210251 to Marc J. Klowden. Chambers, G.M. and M.J. Klowden. 1990. Correlation of nutritional reserves with a critical weight for pupation in larval Aedes aegypti mosquitoes. J. Am. Mosq. Contr. Assoc. 6: 394-399. Christophers, S.R. 1960. Aedes aegypti (L.) the yellow fever mosquito. Its life history, bionomics and structure. Cambridge Univ. Press, Cambridge. Timmermann, S.E. and H. Briegel. 1999. Larval growth and biosynthesis of reserves in mosquitoes. J. Insect Physiol. 45: 461-470.

344

Journal of Vector Ecology

December 2005

Scientific Note The use of an experimental hut for evaluating the entering and exiting behavior of Aedes aegypti (Diptera: Culicidae), a primary vector of dengue in Thailand

Theeraphap Chareonviriyaphap1 , Wannapa Suwonkerd1,2, Piti Mongkalangoon1, Nicole Achee3, John Grieco3, Bob Farlow4, and Donald Roberts3

Department of Entomology, Faculty of Agriculture, Kasetsart University, Bangkean, Bangkok 10900 Thailand 2 Office of Disease Prevention and Control No 10, 18 Boonrungrit Rd., Chiangmai 50200 Thailand 3 Department of Preventive Medicine & Biometrics, Uniformed Services University of the Health Sciences, Bethesda, MD 20714, U.S.A. 4 BASF, 3000 Continental Drive - North, Mount Olive, NJ 07828-1234, U.S.A. Received 6 June 2005; Accepted 22 July 2005 Dengue remains a serious health threat around the world, despite significant gains in its control (Gubler and Kuno 1997). Dengue is transmitted primarily by Aedes aegypti, a day-biting mosquito (Gould et al. 1968, Russell et al. 1969) and the prevention and control of this species has been a long-term problem for endemic areas. The interruption of dengue transmission relies heavily on vector surveillance through the detection of mosquito larvae or pupae and vector control methods. Control strategies focus on the elimination of larval habitats through source reduction, which can be expensive and require continuous community participation often resulting in failure (Kongmee et al. 2004). Efforts focused on adult control strategies using various synthetic compounds have shown success and are commonly being used elsewhere (Somboon et al. 2003). Understanding the behavioral responses of vectors, especially avoidance behavior to residual insecticides, is of particular importance to any vector control program. There have been numerous attempts to accurately measure the behavioral responses of mosquitoes to insecticides using experimental huts, mainly on Anopheles species (Smith 1965, Roberts et al. 1984, Rutledge et al. 1999, Bangs5, Grieco et al. 2000, Pates and Curtis 2005). However, no studies have been published on behavioral responses to investigate the entrance and exit behaviors of Ae. aegypti exposed to chemical insecticides using a portable hut. For this reason, the effect of chemical residue inside homes on reducing man-vector contact needs to be evaluated. The portable huts used in the present study were based on the design of Achee et al. (personal communication) used to evaluate the flight behavior of An. darlingi in Belize, Central America. The dimensions of the huts were 4 m wide x 5 m long x 3.5 m high with three windows (1.125 m x 1.175 m) and one door (0.8 m x 2 m) onto which could be affixed entrance and exit traps. Huts were constructed in the fashion of indigenous Thai homes. Hut frames were made of iron pipe and custom-welded galvanized pipes. Pieces of nontreated wood plank, measuring 1 m x 2.5 m in length served as the side walls. Floors were adjusted and aligned with cement blocks and an A-frame style zinc roof was put in place. The top of the angled roof measured 3.5 m from the ground level. Three windows, one on each of three sides were affixed with entrance traps (Figure 1). In addition, a north-facing door was affixed with an exit trap (Figure 2). The dimensions of the entrance traps were 0.84 m long, 1.065 m wide, and 1.065 m high, with an iron frame. Louvers made of 3/8-in non-treated plywood and fixed vertically at 60 degree angles were placed over the front opening of each entrance trap on each side of the opening, 1.065 m x 1.065 m, with a horizontal row of 10-cm wide slit openings made of 3/ 8-in non-treated plywood fixed vertically to 60 degrees. The louvers were placed in an open state producing a series of horizontal, 10-cm wide openings through which mosquitoes could enter. The traps moved forward and backward during the observation period by sliding them on a support platform (Figure 1). This allowed the collector to capture mosquitoes from the trap without having to be inside the hut. An exit trap, measuring 1.2 m long x 0.845 m wide x 2.10 m high, was fixed to the door opening. Twenty plywood louvers identical to those used in the window traps were installed over the front opening and were again fixed at 60 degree angles to the vertical (Figure 2). These were arranged to facilitate the movement of mosquitoes from the hut into the trap. Both trap types were covered by nylon insect netting. Cotton sleeve material was sewn over several holes in both types of trap to facilitate the removal of mosquitoes. In order to test chemicals in the huts without applying a compound directly to the wall surfaces, a series of panels were developed for holding treated netting which could be positioned around the interior surface of the hut. The aluminum frame that houses the netting contained holes in each corner and were placed onto bolts attached to the hut

1

Bangs, M.J. 1999. The susceptibility and behavioral response of Anopheles albimanus Weidemann and Anopheles vestitipennis Dyar and Knab (Diptera: Culicidae) to insecticides in northern Belize. Ph.D. Thesis. Uniformed Services University of the Health Sciences, Bethesda, Maryland 489 pp.

5

Vol. 30, no. 2

Journal of Vector Ecology

345

Figure 1. Removable entrance traps with a support platform.

walls. A 9-cm gap between the aluminum panel and the wood planks prevented the netting from touching the interior walls; wing nuts were also used to prevent aluminum panels from touching the wood and to assure the netting did not lean on the bolts due to wind or unexpected accidents. Four openings were constructed into the gables at the front and rear of the hut, measuring 0.45 m x 0.45 m. These openings were covered by 1/12-in aluminum wire mesh and served to reduce the temperature inside the huts. On the outside of the hut, a teninch wide and fifteen-inch deep channel filling with water was used as an ant trap. To assemble the hut, side-wall metal frames were prepared. For each one, metal pipes served as poles connected together by metal pipes horizontally, at the bottom and top, into the welded pieces and were then fixed to prevent frame shifting. Similarly, the other three side walls were prepared in the same manner and connected together to make the hut frame. Welded pieces were used for all base support legs, corner and wall joints, and roof angles. Height from the ground level to apex of roof was 5 m. There were four eaves between the top of the wall and roof on the front (2 eaves) and rear (2 eaves) walls. The eaves were sealed by 1/12-in aluminum wire mesh fastened across the eave opening. Once frames had been completed, wood planks were put in place together with window and door traps to make a furnished hut. These portable huts were used to evaluate the endophilic behavior of Ae. aegypti field populations in Kanchanaburi

Figure 2. Exit traps fixed to the door opening.

and Chiangmai provinces, Thailand. Briefly, 100 marked female mosquitoes were released either inside the hut (to measure exit behavior) or outside the hut (to measure entrance behavior). Released populations were marked with different colors following the methods of Muir and Kay (1998) and Tsuda et al. (2001) on Ae. aegypti. Mosquitoes were released at 0500 h and collections were made from the traps at 20 min intervals, from 0600-1800 h. Human hosts were covered by mosquito nets. This prevented humans from being bitten during the study. The movement patterns for natural populations of Ae. aegypti into (entrance) and out (exit) of huts with the presence of human hosts in the huts are presented in Figure 3. Reproducible results were obtained. This was the first time a portable experimental hut was used to document the entrance and exiting behavior of Ae. aegypti. The whole system is easy to assemble and can be disassembled in 3-5 hr, depending on weather conditions and manpower. In brief, both traps indicated a high degree of movement through the windows and doors with peaks entering occurring at 08401040 h and 1240-1320 h and peak exiting occurring at 16401740 h (Figure 3). The portable hut design affixed with entrance and exit traps has demonstrated success in collecting entering and exiting Ae. aegypti in Thailand. The portable hut can serve as a means for testing exiting/entering behavior

346

Journal of Vector Ecology

December 2005

Figure 3. Movement patterns for Aedes aegypti into (entrance) and out (exit) of the huts in the presence of human hosts.

in response to chemical compounds for the control of Ae. aegypti, a notoriously efficient dengue vector. Acknowledgments The authors thank the Disease Control Department, Ministry of Public Health, Nontaburi Province, Thailand for helpful advice during implementation of this work. We also thank the Armed Forces Development Command, Sai Yok District, Kanchanaburi Province for supporting a study area. Special thanks go to the students who helped with the construction of huts. This work was supported by a grant from the National Institutes of Health (grant no. 5U01AI05477702). REFERENCES CITED Gould D.J., T.M. Yuill, M.A. Moussa, P. Simasathien, and L.C. Rutledge. 1968. An insular outbreak of dengue hemorrhagic fever. 3. Identification of vectors and observations on vector ecology. Am. J. Trop. Med. Hyg. 17: 609-618. Grieco J.P., N.L. Achee, R.G. Andre, and D.R. Roberts. 2000. A comparison study of house entering and exiting behavior of Anopheles vestitipennis (Diptera: Culicidae) using experimental huts sprayed with DDT or deltamethrin in southern District of Toledo, Belize, C.A. J. Vector Ecol. 25: 62-73. Gubler M.J. and G. Kuno. 1997. Dengue and dengue hemorrhagic fever. CAB, Oxford. Kongmee M., A. Parbaripai, P. Akratanakul , M.J. Bangs, and T. Chareonviriyaphap. 2004. Behavioral responses of Aedes aegypti (Diptera: Culicidae) exposed to deltamethrin and possible implications for disease control. J. Med. Entomol. 41: 1055-1063.

Muir L.E. and B.H. Kay. 1998. Aedes aegypti survival and dispersal estimated by mark-release-recapture in northern Australia. Am. J. Trop. Med. Hyg. 58: 227-282. Pates H. and C. Curtis. 2005. Mosquito behavior and vector control. Annu. Rev. Entomol. 50: 53-70. Pates, H., and C. Curtis. 2005. Mosquito behavior and vector control. Annu. Rev. Entomol. 50: 53-70. Roberts D.R., W.D. Alecrim, A.M. Tavares, and K.M. Mc Neill. 1984. Influence of physiological condition on the behavioral response of Anopheles darlingi to DDT. Mosq. News. 4: 357-361. Russell P.K., D.J. Gould, T.M. Yuill, A. Nisalak, and P.E. Winter. 1969. Recovery of dengue 4 viruses from mosquito vectors and patients during an epidemic of dengue hemorrhagic fever. Am. J. Trop. Med. Hyg. 18: 580-583. Rutledge L.C., N.M. Echana, and R.K. Gupta. 1999. Responses of male and female mosquitoes to repellents in the World Health Organization insecticide irritability test system. J. Am. Mosq. Contr. Assoc. 15: 60-64. Smith A. 1965. A verandah-trap hut for studying the housefrequenting habits of mosquitoes and for assessing insecticides. 2. The effect of dichlorvos (DDVP) on egress and mortality of Anopheles ganbiae Giles and Mansonia uniformis (Theo.) entering naturally. Bull. Entomol. Res. 56: 275-282. Somboon P., L. Prapanthadara, and W. Suwankerd. 2003. Insecticide susceptibility tests of Anopheles minimus, Aedes aegypti, Aedes albopictus, and Culex quinquefasciatus in northern Thailand. Southeast Asian J. Trop. Med. Publ. Hlth. 34: 87-93. Tsuda Y, M. Takagi, S. Wang, Z. Wang, and L. Tang. 2001. Movement of Aedes aegypti (Diptera: Culicidae) released in a small isolated village on Hainan Island, China. J. Med. Entomol. 38: 93-98.

Vol. 30, no. 2

Journal of Vector Ecology

347

Scientific Note First occurrence of Ochlerotatus japonicus in Missouri

Stephanie Gallitano1 , Leon Blaustein1,2,3, and James Vonesh1

2

Tyson Research Center, Biology Department, Washington University in St. Louis, Eureka MO 63025 U.S.A. Mosquito Research Labs, Entomology Department, Rutgers University, 180 Jones Avenue, New Brunswick NJ 08901 U.S.A. 3 Community Ecology Laboratory, Institute of Evolution, Faculty of Sciences, University of Haifa, Haifa 31905 Israel

1

Received 15 August 2006; Accepted 26 August 2006

The container-breeding mosquito, Ochlerotatus japonicus japonicus (Theobold), is native to Japan and the eastern coast of Asia (Tanaka et al. 1979). In 1998, this species was first reported in the United States, in Ocean County, New Jersey and Suffolk County, New York (Crans 1999, Peyton et al. 1999, Romanowski et al. 1999, Scott et al. 1999). By 2003, it had been reported in at least 19 other states in the eastern US as well as to Quebec, Canada (http://www.rci.rutgers.edu/ ~insects/ocjap.htm), and in 2004 it was reported from western Washington (Roppo et al. 2004). The expansion of this exotic species in North America has been watched with concern as it has been shown to be a potential vector of St. Louis Encephalitis and West Nile virus (Sardelis and Turell 2001, 2003). On June 14, 2005, we collected and identified Oc. j. japonicus larvae from artificial containers at Washington University's Tyson Research Center in western St. Louis County (Geographic: 38°-31' 33.6399"\90°-33' 36.9512"; UTM: 980372.24'\802968.29'). Tyson is characterized by a hilly, karst-dominated landscape about 85% of which is covered in oak-hickory forest. This represents the first record of this species in the state of Missouri. To further evaluate the establishment of Oc. japonicus, on July 22, we haphazardly collected 10 mosquito larvae from 15 artificial pools widely spread throughout Tyson Research Center. These pools were initially established on June 23 with 300L of dechlorinated water, oak and hickory leaf litter, and an inoculation of zoo/phytoplankton as part of an experiment assessing treefrog oviposition site selection. A survey of natural container habitats was not possible during this period Table 1. Population composition of immature mosquitoes collected from 15 artificial pools at Tyson Research Center.

Species Cx. pipiens Cx. restuans Cx. salinarius Oc. japonicus Aedes sp. Number of pools containing species 11 9 8 9 7 Percent of total larvae collected 0.31 0.16 0.17 0.24 0.12

due to a drought. Immature Oc. japonicus were found in the samples in nine of the 15 pools along with several other species, population composition recorded in Table 1. These data suggest that Oc. japonicus is established at Tyson Research Center being the second most common mosquito species overall. This represents the western most record of this exotic species in the central and eastern United States. The nearest previous records place Oc. Japonicus in Michigan (2003, Tom Wilmot pers com) to the east and Washington (Roppo et al.) state to the west of Missouri. Given the likelihood that the occurrence of this species is from a spread from the east, our record suggests the possible presence of this exotic in the nearby states, such as Illinois. More widespread surveys of Oc. japonicus should be taken throughout the Missouri and other midwestern and prairie states in order to determine the extent of Oc. japonicus in this region. Acknowledgments We thank Wayne Crans for confirming our identifications. We would also like to thank Richard Lampman, Linn Haramis, Uriel Kitron, Steve Juliano, Jack Swanson, and Tom Wilmot for their correspondence concerning expansion of Oc. japonicus. REFERENCES CITED Crans, W. J. 1999. Aedes japonicus: accidental introduction to the northeastern United States. Vector Ecol. Newsl. 30: 5. Peyton, E. L., S. R. Campbell, T. M. Candeletti, M. Romanowski, and W. J. Crans. 1999. Aedes (Finlaya)japonicus japonicus (Theobald), a new introduction into the United States. J. Am. Mosq. Contr. Assoc. 15: 238-241. Romanowski, M., T. Candeletti, S. Campbell, D. Ninivaggi, and W. J. Crans. 1999. Aedes japonicus in New Jersey and New York- the first United States records. In: Hamilton G.C., ed. Proceedings of the 86th meeting of

Keyword index: Ochlerotatus japonicus, Missouri, exotic species, invasive species.

348

Journal of Vector Ecology

December 2005

the New Jersey Mosquito Control Association. 1999 March 30- April 1; Atlantic City, NJ. Oxford. NJ: New Jersey Mosquito Control Association. p 38-41. Roppo, M. R., J. L. Lilja, F. A. Maloney, and W. J. Sames. 2004. First occurrence of Ochlerotatus japonicus in the state of Washington. J. Am. Mosq. Contr. Assoc. 20: 8384. Sardelis, M. R. and M. J. Turell. 2001. Ochlerotatus j. japonicus in Frederick County, Maryland: discovery, distribution, and vector competence for West Nile virus. J. Am. Mosq. Contr. Assoc. 17: 137-141. Sardelis, M. R., M. J. Turell, and R. G. Andre. 2003. Experimental transmission of St. Louis encephalitis virus

by Ochlerotatus j. japonicus. J. Am. Mosq. Contr. Assoc. 19: 159-162. Scott, J. J., Ochlerotatus japonicus (Theobald). 2004. Rutgers University/Placer Mosquito. http://www.rci.rutgers.edu/ ~insects/ocjap.htm. Scott, J. J., R. J. McNelly, and W. J. Crans. 1999. Aedes japonicus overwinters in New Jersey. Vector Ecol. Newsl. 30: 6-7. Tanaka, K., K. Mizusawa, and E.S. Saugstad. 1979. A revision of the adult and larval mosquitoes of Japan (including Ryukyu Archipelago and the Ogasawara Islands) and Korea (Diptera: Culicidae). Contrib. Am. Entomol. Inst. (Ann Arbor) 16: 1-987.

168

Journal of Vector Ecology

December 2005

Seasonal and geographical distribution of adult Ixodes scapularis Say (Acari: Ixodidae) in Louisiana

Andrew Mackay and Lane Foil

Department of Entomology, Louisiana State University Agricultural Center, Baton Rouge, LA 70803, U.S.A. Received 5 July 2004; Accepted 28 January 2005 ABSTRACT: The distribution and seasonality of adult black-legged ticks (Ixodes scapularis Say) in Louisiana was measured. The presence of adult ticks was determined by flagging at 106 sites throughout Louisiana. It was concluded that Ixodes scapularis is widely distributed throughout Louisiana. Ticks were also collected twice per month at one site over a 15-month period by flagging and use of CO2 traps to establish the relative seasonal abundance pattern of free-living adult ticks. Hostseeking, black-legged adult ticks were collected from October to May. Peak adult abundance was observed in December. More ticks were collected by the use of CO2 traps compared to flagging in October, November, and February. No blacklegged tick larvae or nymphs were collected in this study using either collection method. Journal of Vector Ecology 30 (2): 168-170. 2005. Keyword Index: Ixodes scapularis, flagging, CO2 trapping, seasonality, black-legged tick. INTRODUCTION The black-legged tick, Ixodes scapularis Say, is an important vector of several human pathogens in North America, including Borrelia burgdorferi, the etiologic agent of Lyme disease. Borrelia burgdorferi has been isolated from black-legged ticks in eastern Texas (Teltow et al. 1991), Oklahoma (Kocan et al. 1992) and Georgia (Oliver et al. 1993), and detected in black-legged ticks collected from Alabama (Luckhart et al. 1991) and North Carolina (Ouellette et al. 1997). Black-legged ticks collected from Georgia and Louisiana have been shown to be competent vectors of B. burgdorferi in the laboratory (Sanders and Oliver 1995, Jacobs et al. 2003). Although the southeastern and southcentral states are considered low risk areas for contracting Lyme disease, several confirmed cases are reported from Louisiana every year. Avoiding exposure to ticks is the only effective means of Lyme disease prevention. Knowledge of the geographical range and seasonality of vector species is critical to determining risk of exposure. In Louisiana, very little is known of the biology of the black-legged tick. The objective of this study was to determine the temporal and geographical distribution of adult black-legged ticks in Louisiana. MATERIALS AND METHODS Seasonality The seasonal activity of adult black-legged ticks at the Port Hudson State Commemorative Area, West Feliciana Parish, LA, was determined from January 6, 2000 to March 7, 2001. Port Hudson SCA is located along the eastern escarpment of the Mississippi River, approximately 20km NNW of Baton Rouge. The terrain is characterized by upland, oak-hickory hardwood forests with deep ravines, steep slopes, and narrow ridgetops. Thirty-six 100 m long transects were set up in parallel rows, with at least 20 m between transects. Three randomly selected transects were each sampled twice per month by flagging and use of CO2 traps. Flagging was performed by dragging a 1 m2, white cotton-denim cloth over the leaf litter and understory vegetation 1.5 m in height along the length of the transect. Ticks were removed from the cloth at 20 m intervals. The CO2 traps were polystyrene ice buckets with four 5 mm diameter holes drilled into the base. The traps were placed on a 0.5 m2 piece of white cotton cloth. Each trap was filled with approximately 2.5 kg of dry ice. Traps were placed at 25 m intervals along the length of each transect (3 traps per transect). Ticks were collected from traps after a 24 h period. All collected ticks were placed in 70% ethanol and retained for identification. Air temperatures were recorded using a digital thermometer placed in a shaded location, approximately 30 cm above the ground surface. Geographic distribution A survey was conducted at 106 different locations in Louisiana. From one to six sites were selected in each of the 49 parishes sampled. Each site was flagged once for a minimum of 15 min, particularly in areas considered to be preferred tick habitat (along the margins of wooded areas, game trails, etc.). The survey was conducted from November 4, 1999 to April 27, 2000 (83 sites), February 14 ­ 21, 2001 (16 sites), January 22 to February 25, 2002 (six sites) and January 31, 2003 (one site). RESULTS Seasonality Adult black-legged ticks were collected from January to May, 2000, and October, 2000 to March, 2001 (Figure 1). No ticks were collected from June through September, 2000. The highest number of adult ticks collected by flagging was observed in December and January. The CO2 traps captured

Vol. 30, no. 2

Journal of Vector Ecology

169

Flagging Mean Daily Avg. Air Temp.

CO2 Trapping Mean Daily Min. Air Temp 35 30

Monthly Mean No. Ticks per Transect

10 8 6 4 2

Air Temperature (C)

25 20 15 10 5 0

0

pr il M ay Ju ne Ju ly st m be r ob er ov em be D r ec em be r Ja nu ar y Fe br ua ry M ar ch Ja nu ar y Fe br ua ry M ar ch ug u A pt e O ct A

-5

Se

Figure 1. Seasonal distribution of Adult Ixodes scapularis at Port Hudson SCA, Louisiana (January 2000 to March 2001).

the higher number of ticks early in October, November, and February. No black-legged tick larvae or nymphs were collected in this study using either sampling method. Geographical distribution Adult black-legged ticks were collected from 60 (56.6%) of the 106 sites sampled (Figure 2). An average of 1.3 ± 2.3 (mean ± std) ticks were collected per 10 min of flagging. DISCUSSION Adult black-legged ticks were found to exhibit hostseeking behavior from October to May in southern Louisiana and peak abundance was observed in December and January. The results of our study are similar to other studies using dragging to estimate the seasonality of a black-legged tick in southern states. (Goddard 1992, Cilek and Olson 2000, Kollars et al. 1999). Adult black-legged ticks have been collected from cattle from October to April in southeastern Oklahoma (Barnard 1981). Koch (1982) reported parasitism of dogs in southeastern Oklahoma and northwestern Arkansas from October to June. Solberg et al. (1992) reported that in New Jersey, CO2 traps collected a greater number of black-legged ticks than dragging. In a later study conducted in the same area, Schulze et al. (1997) collected significantly more adult black-legged ticks by walking surveys and by dragging than by use of CO2 traps. In the current study, CO2 sampling was more effective than flagging in certain months but not in others. Although the black-legged tick has been reported to exhibit questing behavior at temperatures below 4ºC (Schulze et al. 2001a), low temperatures may inhibit taxis to a CO2 trap. Therefore, flagging may be a more effective technique than use of CO2

traps for collecting adult black-legged ticks during periods of low temperatures. The nymph of the black-legged tick is thought to be the most important life stage responsible for transmission of B. burgdoferi to humans in North America. In this study, no larval or nymphal black-legged ticks were collected. In the southern U.S.A., the immature stages of this species are rarely collected on humans (Felz et al. 1996), or by flagging (Cilek and Olson 2000, Piesman 2002). Cilek and Olson (2000) attributed this observation to the possibility that subadult black-legged ticks may limit questing behavior to habitats below the vegetation surface which is sampled by flagging. In this study, we found the adult black-legged tick to be widely distributed throughout northern and central Louisiana but less abundant in the southern coastal region. Many of the sites sampled in southern Lousiana were small, fragmented woodlots surrounded by agricultural land, urban development, or wetland habitat. Schulze et al. (2001b) found significantly greater numbers of adult black-legged ticks in mixed and pine habitats, relative to hardwood habitats. In our study, adult black-legged ticks were most frequently collected in mixed and predominantly coniferous habitats and were rarely collected in seasonally flooded, bottomland areas. Acknowledgments We thank Glen Oremus for assistance during field sampling and Gregg Potts, the site manager at Port Hudson State Commemorative Area. This study was published with approval of the Director of Louisiana Agricultural Experiment Station as Manuscript No. 04-26-0343.

N

170

Journal of Vector Ecology

December 2005

No ticks collected Ixodes scapularis collected

Figure 2.Geographical distribution of Ixodes scapularis in Louisiana.

REFERENCES CITED Barnard, D.R. 1981. Seasonal activity and preferred attachment sites of Ixodes scapularis (Acari: Ixodidae) on cattle in southeastern Oklahoma. J. Kansas Entomol. Soc. 54: 547-552. Cilek, J.E. and M.A. Olson. 2000. Seasonal distribution and abundance of ticks (Acari: Ixodidae) in northwestern Florida. J. Med. Entomol. 37: 439-444. Felz, M. W., L. A. Durden, and J. H. Oliver, Jr. 1996. Ticks parasitizing humans in Georgia and South Carolina. J. Parasitol. 82: 505-508. Goddard, J. 1992. Ecological studies of adult Ixodes scapularis in Central Mississippi: questing activity in relation to time of year, vegetation type and meteorologic conditions. J. Med. Entomol. 29: 501-506. Jacobs, M.B., J.E. Purcell, and M.T. Philipp. 2003. Ixodes scapularis ticks (Acari: Ixodidae) from Louisiana are competent to transmit Borrelia burgdorferi, the agent of Lyme Borreliosis. J. Med. Entomol. 40: 964-967. Kocan, A.A., S.W. Mukolwe, G.L. Murphy, R.W. Barker, and K.M. Kocan. 1992. Isolation of Borrelia burgdorferi (Spirochaetales: Spirochaetaceae) from Ixodes scapularis and Dermacentor variabilis ticks (Acari: Ixodidae) in Oklahoma. J. Med. Entomol. 29: 630-633. Koch, H.G. 1982. Seasonal incidence and attachment sites of ticks (Acari: Ixodidae) on domestic dogs in southeastern Oklahoma and northwestern Arkansas, USA. J. Med. Entomol. 19: 293-298.

Kollars, T.M., Jr., J.H. Oliver, Jr., P.G. Kollars, and L.A. Durdan. 1999. Seasonal activity and host associations of Ixodes scapularis (Acari: Ixodidae) in southeastern Missouri. J. Med. Entomol. 36: 720-726. Luckhart, S., G.R. Mullen, and J.C. Wright. 1991. Etiologic agent of Lyme Disease, Borrelia burgdorferi, detected in ticks (Acari: Ixodidae) collected at a focus in Alabama. J. Med. Entomol. 28: 652-657. Oliver, J.H., Jr., F.W. Chandler, Jr., M.P. Luttrell, A.M. James, D.E. Stallknecht, B.S. McGuire, H.J. Hutcheson, G.A. Cummins, and R.S. Lane. 1993. Isolation and transmission of the Lyme Disease spirochete from the southeastern United States. Proc. Natl. Acad. Sci. U.S.A. 90: 7371-7375. Ouellette, J., C.S. Apperson, P. Howard, T.L. Evans, and J.F. Levine. 1997. Tick-Raccoon associations and the potential for Lyme Disease spirochete transmission in the coastal plain of North Carolina. J. Wildlife Dis. 33: 25-39. Piesman, J. 2002. Ecology of Borrelia burgdorferi sensu lato in North America. In: J.S. Gray, O. Kahl, R.S. Lane, and G. Stanek (eds.) Lyme borreliosis: biology, epidemiology, and control. pp. 223-249. CABI Publishing, U.K. 347 pp. Sanders, F.H., Jr. and J.H. Oliver, Jr. 1995. Evaluation of Ixodes scapularis, Amblyomma americanum and Dermacentor variabilis (Acari: Ixodidae) from Georgia as vectors of a Florida strain of the Lyme Disease spirochete, Borrelia burgdorferi. J. Med. Entomol. 32: 402-406. Schulze, T.L., R.A. Jordan, and R.W. Hung. 1997. Biases associated with several sampling methods used to estimate abundance of Ixodes scapularis and Amblyomma americanum (Acari: Ixodidae). J. Med. Entomol. 34: 615623. Schulze, T.L., R.A. Jordan, and R.W. Hung. 2001a. Effects of selected meterological conditions on diurnal questing of Ixodes scapularis and Amblyomma americanum (Acari: Ixodidae). J. Med. Entomol. 38: 318-324. Schulze, T.L., R.A. Jordan, and R.W. Hung. 2001b. Potential effects of animal activity on the spatial distribution of Ixodes scapularis and Amblyomma americanum (Acari: Ixodidae). Environ. Entomol. 30: 568-577. Solberg, V.B., K. Neidhardt, M.R. Sardelis, C. Hildebrandt, F.J. Hoffmann, and L.R. Boobar. 1992. Quantitative evaluation of sampling methods for Ixodes dammini and Amblyomma americanum (Acari: Ixodidae). J. Med. Entomol. 29: 451-456. Teltow, G.J., P.V. Fournier, and J.A. Rawlings. 1991. Isolation of Borrelia burgdorferi from arthropods collected in Texas. Am. J. Trop. Med. Hyg. 44: 469-474.

Vol. 30, no. 2

Journal of Vector Ecology

171

Tick infestations of the eastern cottontail rabbit (Sylvilagus floridanus) and small rodentia in northwest Alabama and implications for disease transmission

Joseph C. Cooney1, Willy Burgdorfer2, Martin K. Painter3, and Cynthia L. Russell3

2

Northwest Shoals Community College, P. O. Box 101 George Wallace Blvd., Muscle Shoals, AL 35662, U.S.A. U.S. Department of Health, Education, and Welfare, Public Health Service National Institutes of Health, National Institute of Allergy and Infectious Diseases, Epidemiology Branch, Rocky Mountain Laboratories, Hamilton, MT 59840, U.S.A. 3 101 Guntersville Street, Sheffield, AL 35660, U.S.A. Received 27 October 2005; Accepted 13 April 2005

1

ABSTRACT: Studies were conducted over a four-county area of northwest Alabama to determine the association of eastern cottontail rabbits with Dermacentor variabilis, the eastern United States vector of Rocky Mountain spotted fever. A secondary objective was to compare infestations of this tick on rabbits with infestations on commonly encountered rodent species as a means of determining the relative importance of each in the disease transmission cycle. These epidemiologic surveys were conducted in response to reported fatal cases of Rocky Mountain Spotted Fever in two counties of the study area. From 202 eastern cottontail rabbits, 3,956 ticks were collected. Of this total, 79.87% were Haemphysalis leporispalustris, 9.15% Amblyomma americanum, 8.22% Ixodes dentatus, and 2.76% D. variabilis. Only immature stages of D. variabilis were collected from cottontail rabbits. Ticks were collected on rabbits in all months except November, and only one specimen was taken in January. Based on the average number of ticks per host collected in each month, April was the peak month for D. variabilis and I. dentatus. High values for H. leporispalustris also occurred at this time, but even higher values occurred in October and December. The heaviest infestation of A. americanum occurred during the month of August and coincides with the activity period for the larvae of this species. Two hundred sixty-nine of the smaller Rodentia, comprising 13 species, yielded 264 ticks, all D. variabilis, and all but two were immature stages. Five rodent species, Microtus ochragaster, Orozomys palustris, Peromyscus gossypinus, Peromyscus leucopus, and Sigmodon hispidus accounted for 95.83% of the ticks collected, and appeared to be preferred hosts for D. variabilis; all five had higher infestation levels per host than did the eastern cottontail rabbit. Data on host relationships in association with seasonal activity are presented. Journal of Vector Ecology 30 (2): 171-180. 2005. Keyword Index: Tick/host relationships, rodents, lagomorphs.

INTRODUCTION Investigations into the natural history of the spotted fever agent, Rickettsia rickettsii, at the Tennessee Valley Authority's Land Between The Lakes recreation and conservation area (170,000 acre facility) between 1969 and 1973, provided serologic evidence that this agent was widely distributed among a large variety of medium-sized mammals (Burgdorfer et al. 1974). The eastern cottontail rabbit (Sylvilagus floridanus), as shown by complement fixation tests, had the highest number (32 of 66) of seropositives. In the same studies, Dermacentor variabilis was the only species of tick from which R. rickettsii was isolated, although numerous specimens of Amblyomma americanum and D. albipictus were examined, along with considerably smaller numbers of Haemphysalis leporispalustris and Ixodes spp. Other studies, (Hechemy et al. 1990), in Ohio yielded similar results regarding rickettsial infection in ticks. These findings tend to support the hypothesis of earlier investigators (Parker 1923, Jellison 1945, Parker et al. 1951), who suggested a strong relationship between lagomorphs and the spotted fever agent, R. rickettsii. D. variabilis is also an established vector of Francisella

tularensis, the causative agent of tularemia, which is of special interest in national security because of its potent pathogenicity to man and its potential use as a bioterrorism agent (Burgdorfer et al. 1974, Tarnvik and Berglund 2003). Although the incidence of tularemia has declined from peak numbers in 1939, it remains throughout the United States and occurs in endemic foci in Arkansas, Missouri, and Oklahoma, with occasional outbreaks on Martha's Vineyard Island, Massachusetts (Feldman et al. 2001, Hornick 2001). In addition to its association with R. rickettsia and F. tularensis, D. variabilis has been studied for competence as a potential vector of several other pathogenic agents. Investigations to assess its potential for vectoring Borrelia burgdorferi based on laboratory transmission trials or examination of field-collected specimens, demonstrated that D. variabilis was not a competent vector (Piesman and Sinsky 1988, Mather and Mather 1990, Levine et al. 1991, Lindsay et al. 1991, Mukolwe et al. 1992, Sanders and Oliver 1995, Piesman and Happ 1997). Dermacentor variabilis has also been evaluated for its potential to vector Ehrlichia organisms. Studies indicate that it is not a competent vector of either human granulocytic ehrlichiosis or Ehrlichia chaffeensis

172

Journal of Vector Ecology

December 2005

(Anderson et al. 1993, Des Vignes et al. 1999). However, in the first documented trials in the laboratory, D. variabilis effectively transmitted Ehrlichia canis to dogs (Johnson et al. 1998). This transmission represented the first by a species other than Rhipicephalus sanguineus. In addition to the human vector importance of D. variabilis, the other species included in our studies having significant human vector concern is Amblyomma americanum. Although it has been found naturally infected with Borrelia burgdorferi in a Missouri study (Oliver et al. 1998), other investigations of its vector competency based on both field and laboratory assessments suggest that it is of only minor concern to human transmission (Piesman and Sinsky 1988, Mather and Mather 1990, Sanders and Oliver 1995, Piesman and Happ 1997). It has, however, been found naturally infected with Ehrlichia chaffeensis in pools of ticks collected from five states, (Anderson et al. 1993), and more recently, Borrelia lonestari, the causative agent of southern tick associated rash illness, was detected in this tick in specimens taken in southern Alabama (Burkot et al. 2001). Although the remaining two species included in our collections, Haemaphysalis leporispalustris and Ixodes dentatus, are not known to feed on man, they were the most prevalent species to occur on cottontail rabbits. The importance of H. leporispalustris in the natural cycle of Rocky Mountain Spotted Fever has been stated previously, and I. dentatus has been documented in laboratory studies as being capable of transmitting B. burgdorferi to laboratory rabbits (Telford and Spielman 1989). I. dentatus has also been found naturally infected with a new genospecies, B. andersoni (Marconi et al. 1995). To further define the seasonal tick-host relationships of these species, studies were conducted to evaluate more precisely the role of the eastern cottontail rabbit as a host for ticks. Concurrent with these studies were complementary investigations into the seasonal host-parasite relationships of Ixodid ticks infesting 13 species of small rodents occurring in the study area. The smaller Rodentia are reportedly the preferred hosts for D. variabilis, and these studies would further clarify preference among the 13 collected rodent species. These data would also demonstrate the relative importance of rodents and lagomorphs as hosts for Ixodid ticks, especially as hosts for D. variabilis. MATERIALS AND METHODS Eastern cottontail rabbits were collected by gun from a contiguous four county area in northwest Alabama from 1970 through 1973, with most of the specimens taken in 1972 and the first half of 1973. Collections were made in all months of the year although the greatest proportions (76%) were made from April through August, the period of greatest activity and abundance for both cottontail rabbits and ticks. The seasonal abundance of cottontail rabbits was determined through personal observations of their relative numbers during collection periods throughout the year, and the period of greatest activity for the ticks included in this study was determined from personal observations and published research

(Bishopp and Trembley 1945, Cooley 1946, Sonenshine and Stout 1970, Cooney and Hayes 1972, Cooney and Burgdorfer 1974, Zimmerman et al. 1988). Small rodents were collected by live trapping in the same four county area of northwest Alabama from 1970 through 1973 with the greatest proportion of specimens taken during the spring and summer of 1972 and 1973; no trapping occurred during January and February each year. All collected specimens (both rabbits and rodents) were immediately placed in tightly sealed plastic bags to prevent the escape of ticks. A chloroform saturated cotton ball was placed in each bag to kill live rodents and to facilitate the removal of live ticks which were combed from the animal fur or removed with forceps. All collected animal species were handled as humanely as possible in the process. Removed tick species were all identified to species. RESULTS Cottontail rabbit/tick relationships Infestation levels Of 202 eastern cottontail rabbits collected, 158 (78%) were infested with a total of 3,956 ticks comprising the four species Haemphysalis leporispalustris, Amblyomma americanum, Ixodes dentatus, and Dermacentor variabilis. The American dog tick, D. variabilis, ranked last in numbers of ticks collected although 17.82% of all cottontail rabbits were infested. Only 109 ticks, all immature stages, were collected from 36 infested rabbits yielding a mean intensity index, (average number per infested host), of 3.03. The lone star tick, A. americanum, was the second most abundant species in terms of total number of ticks collected from cottontail rabbits; however, only 9.41% were infested, (prevalence index). The percentage of cottontail rabbits infested with A. americanum would probably have been much higher if rabbit collections had been concentrated in the months of August and September, the peak active season for the larvae of A. americanum. However, only 13% of all the rabbits collected were taken during these two months. The rabbit tick, H. 1eporispalustris, was the dominant species in terms of density per host and frequency of occurrence on cottontail rabbits (77.23%). Ixodes dentatus was third in terms of numbers of specimens taken, however, it was the second most frequently occurring species infesting 30.30% of all cottontail rabbits. Of the total number of ticks collected from cottontail rabbits, 79.87% were H. leporispalustris, 9.15% A. americanum, 8.22% I. dentatus, and 2.76% D. variabilis. Seasonal prevalence of ticks Based on values of average numbers of total ticks (all species) per host, cottontail rabbits were most heavily infested from March through October with a mid-summer slump during June and July. Peak numbers occurred in April and October and again in December. The high value obtained in December represented almost exclusively the presence of H. leporispalustris, especially larvae. No specimens of any species were collected during November and values for

Vol. 30, no. 2

Journal of Vector Ecology

173

January and February were very low, represented by only four nymphs of H. leporispalustris in January from four cottontail rabbits, and 35 H. leporispalustris and 10 I. dentatus from six cottontail rabbits in February. In proportion to the number of cottontail rabbits collected per month, June produced the lowest average infestation of ticks, with the exception of January and November which are considered to be months of inactivity for both D. variabilis and A. americanum. Of the total number, including all species of ticks collected, peak numbers occurred during April in which 33.3% of specimens were collected, followed by May with 18.2%. Three species, H. leporispalustris, I. dentatus, and D. variabilis, accounted for the high infestation levels recorded during these months. A. americanum was only minimally represented in the collections until August and September when the larval stages appeared, although 23 nymphs were obtained during April and May. With the exception of November in which no ticks were collected, and January in which only four specimens were obtained, cottontail rabbits appear to be heavily infested with one or more of four species of ticks throughout the year in the southeastern United States. Dermacentor variabilis was collected on rabbits from March through October (Figure 1a). About 70% of all specimens collected were taken in April and were represented by seven larvae and 68 nymphs. Occurrence of specimens remained low for the next two months and was followed by a slight resurgence in July during which time about 15% of the total numbers collected were taken. Of the total 109 specimens collected, 32 were larvae and 77 were nymphs; the adult stage did not occur. Figure 1a illustrates that rabbits were most heavily infested with D. variabilis larvae in March while nymphal infestation was greatest in April. Infestation of both life stages on cottontail rabbits dropped sharply after the peak months and then gradually declined to zero by the first of November. The cottontail rabbit most heavily infested with D. variabilis was collected in April and contained 17 nymphs. Two other cottontail rabbits contained 12 D. variabilis each. One of these collected in April contained 11 nymphs and one larva, the other collected in July contained one nymph and 11 larvae. Infestations were generally low and usually were no more than one per infested host. Amblyomma americanum infestations on cottontail rabbits were highly variable with peak averages of the number of ticks per host occurring in August and a lesser peak in September (Figure 1b). Of the 362 total specimens collected, 73% were taken in August and 19% in September. These highest values coincide with the time of year in which larval specimens of A. americanum enter the population. Of the 306 specimens collected during the two peak months, 92% were larvae. Figure 1b illustrates the seasonal infestation of A. americanum on cottontail rabbits. It should be noted that, as with D. variabilis, adults did not occur on any of the cottontail rabbit hosts. Larvae were the dominant life stage of A. americanum comprising about 86% of the total infestation. Although only 9.41% of all cottontail rabbits collected were infested with A. americanum, one specimen alone taken in August contained 247 larvae and one nymph. This same rabbit

was the most heavily infested rabbit seen during the course of these studies and contained a combined total of 322 ticks. Frequency of cottontail rabbit infestation with A. americanum would be expected to be much higher, however, distribution of A. americanum in Alabama is irregular and usually coincides with high populations of deer, the preferred host for the adult stage. With the exception of a small portion of Colbert and Lauderdale Counties, most rabbit collections were made in areas known to be sparsely inhabited by deer. This situation has changed rapidly since these studies were conducted, and now virtually all portions of the previous study area have harvestable populations of white-tailed deer. Consequently A. americanum and Ixodes scapularis are now occurring in tick collections from these areas, and A. americanum has now become the object of numerous tick complaints throughout the study area (unpublished data). Haemaphysalis leporispalustris, by far the dominant species based on density/host and frequency of occurrence, appeared on cottontail rabbits in all months of the year except November. Based on infestations per rabbit, rates were highest in October followed by December and April. One hundred seventy-seven was the highest number of H. leporispalustris recorded on a single cottontail rabbit. All life stages of H. leporispalustris occurred on cottontail rabbits in the 11 months in which they were infested, except that nymphs were not present in December and larvae and adults were absent in January. The adult stage of H. leporispalustris was considerably more prevalent than either the larva or nymph, occurring 1.6 to 1.4 times more than either respectively; larvae and nymphs were about equally represented. Figure 1c illustrates the seasonal distribution of the various life stages of H. leporispalustris on cottontail rabbits. Based on average number of ticks per host, the adult and nymphal stages had graphic coincidental population peaks during April which dropped abruptly during June. Adults resurged and produced a secondary peak in August, and nymphs had a secondary peak in September. Their numbers then declined to zero by November and essentially remained there through December and January. Conversely, larval infestations were low from February through September, but then abruptly peaked in October at a level much higher than either of the other two life stages. Larval infestation levels then dropped to zero in November as did both the nymphal and adult levels. However, a rapid rise immediately followed so that December larval levels were again very high. Sampling in November may have been inadequate to accurately assess infestation levels. Ixodes dentatus was collected on cottontail rabbits throughout most of the year although no specimens were taken in January, September, and November. All life stages were taken on cottontail rabbits with the greatest number of specimens recorded during April, followed by May and March. Eighty-six percent of the total number of I. dentatus collected were adults while larvae and nymphs occurred in about equal proportions. Based on average number of ticks per host, infestations were highest during March followed closely by April. Infestations were generally light for any one cottontail rabbit with the highest number, 35, taken from a

174

Journal of Vector Ecology

December 2005

a

b

c

d

Figure 1. Seasonal infestation of four species of Ixodid ticks on 202 eastern cottontail rabbits collected from June 1970 to September 1973 in northwest Alabama.

Vol. 30, no. 2

Table 1. Seasonal infestation of Dermacentor variabilis on 269 rodent hosts collected from June 1970 through September 1973 in northwest Alabama.

Rodentia Species Blarina brevicauda Microtus ochragaster Mus musculus Neotoma floridana Oryzomys palustris Peromyscus gossypinus Peromyscus leucopus Peromyscus nuttalli Pitymys pinetorum Rattus norvegicus Sciurus carolinensis Sigmodon hispidus Tamias striatus 0 0 62 84 17 49 17 9 22 6 1

Jan. 0 0 0 0 0 0 0 0 0 0 0 0 0

Feb. 0 0 0 0 0 0 0 0 0 0 0 0 0

Mar. 0 14 7 0 1 4 28 0 0 1 0 7 0

Number Of Rodents Collected/Month Apr. May Jun. Jul. Aug. Sep. Oct. 1 0 1 0 0 0 0 15 0 0 0 0 0 0 11 2 1 0 1 5 4 1 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 2 4 2 2 0 11 10 9 2 0 1 0 2 1 0 0 0 0 0 0 0 0 0 1 1 0 7 0 0 0 0 0 0 0 0 7 0 0 0 0 16 2 24 11 4 12 1 19 2 5 0 0 1 1 Nov. 0 0 0 0 0 0 0 0 0 0 1 0 0 Dec. 1 0 1 0 0 0 0 0 0 0 0 0 0 2

Total 3 29 32 2 1 15 61 3 2 8 8 77 28 269

Journal of Vector Ecology

Total

Average Number of Dermacentor variabilis Collected/Host/Month Jan. 0.00 0.00 0.00 0.00 0.00 1.19 0.00 0.00 0.00 1.10 0.10 0.00 Feb. Mar. Apr. 0.05 0.08 0.02 0.15 May 2.76 0.12 0.00 2.88 Jun. 0.00 0.00 0.00 0.00 Jul. 1.65 0.12 0.00 1.76 Aug. 0.00 0.33 0.00 0.33 Sep. 2.23 2.14 0.00 4.36 Oct. 0.00 0.00 0.00 0.00 Nov. 0.00 0.00 0.00 0.00 Dec. 0.00 0.00 0.00 0.00

Dermacentor variabilis

Larvae Nymphs Adults

Overall Average 0.73 0.25 0.01 0.98

175

Composite Average

176

Journal of Vector Ecology

December 2005

Figure 2. Average number of Dermacentor variabilis collected from 15 mammalian hosts, June 1970 to September 1973, in northwest Alabama.

Vol. 30, no. 2

Journal of Vector Ecology

177

Figure 3. Seasonal infestation of Dermacentor variabilis on a total of 269 rodent hosts collected in northwest Alabama, June 1970 to September 1973. *More than 99% of the total ticks collected were immatures. cottontail rabbit collected in April (29 adults and 6 nymphs). The average number of ticks recorded for infested rabbits was 5.33, however the most frequent infestation level was 1-2 per host. Figure 1d illustrates the seasonal infestation cycle for I. dentatus based on average number of ticks collected per host per month. Adults were frequently observed "in copulo" on the host and most often occurred in even sex ratios of males to females. Adult infestations peaked in March and then gradually receded to low levels through mid-to-late summer, followed by virtually no infestations from fall through midwinter. In contrast, larval infestations peaked in mid-Fall as did nymphal density per host. The presence of larvae, nymphs, and adults on hosts in December, and adults again in February, suggests that this species remains on the host year round. The lack of specimens in January may be attributed to the paucity (4) of cottontail rabbits collected during this month. Considering the seasonal infestation data of all four species of ticks including all life stages during each month of collection, it appears that peak infestations of D. variabilis, I. dentatus, and H. leporispalustris all occurred during the spring months although H. leporispalustris produced additional peaks of even greater magnitude in October and December. A. americanum occurred at relatively low levels until August when infestations per host became high as the larval stage entered the activity cycle. When considering the total number of ticks collected, (3,956), about 35% were taken on rabbits during April. Rodent/tick relationships Infestation levels Thirteen species of the smaller Rodentia were collected concurrently with the cottontail rabbit studies and totaled 269 specimens. Table 1 lists the species and numbers collected of each. Sigmodon hispidus and Peromyscus leucopus were the dominant species represented in our collections, followed by Mus musculus, Microtus ochragaster and Tamias striatus, the latter three at much reduced levels. From all of the rodents collected, 264 D. variabilis ticks were obtained (the only species represented), and of these, more than 99% were immature stages; only one adult specimen was recorded. With the exception of Oryzomys palustris, which was represented in our collections by only one specimen, M. ochragaster was the most frequently infested species (44%), followed by Peromyscus nuttalli, P. leucopus, Peromyscus gossypinus, M. musculus, S. hispidus, and T. striatus. Although the percentage of hosts infested value for P. nuttalli was recorded as 33.33%, this figure tends to inflate the importance of this species as a host for D. variabilis, especially when considering that only three P. nuttalli were collected; only one D. variabilis was taken from these three hosts yielding an average infestation per host value of 0.33. Figure 2 illustrates the relative value of 13 rodent and two lagomorph species as hosts for D. variabilis in northwest Alabama. It is apparent from this chart that at least five rodent species are more preferred hosts for D. variabilis than is Sylvilagus floridanus based on average infestation levels per host.

178

Journal of Vector Ecology

December 2005

Rodents were collected in all months except January and February (Table 1). However collections during November and December included only three rodent specimens, demonstrating that essentially all rodent hosts were taken during the period of greatest activity for most hard ticks. Seasonal prevalence of ticks Ticks were generally collected from March through September although none were collected in June. Infestations on infested hosts (intensity) were relatively low and averaged 4.89 with one to three ticks occurring most often. Two specimens of S. hispidus however harbored 55 and 32 ticks respectively, and one P. leucopus contained 27 ticks. The composite average of ticks collected per rodent hosts (269 rodents) was 0.98. Although the largest numbers of ticks were collected in March and September, per host infestation rates in March were lower than September, May, and July. Based on average infestation rates per rodent, September yielded the highest value followed by May. The seasonal activity period for immature D. variabilis based on average infestation values per host, although fluctuating monthly, displayed an increasing trend from spring to autumn, the period of highest activity (Figure 3). DISCUSSION The present observations on the significance of the eastern cottontail rabbits as a tick host are in general agreement with previously published records (Cooley and Kohls 1944, Bishop and Trembley 1945, Cooley 1946, Clifford et al. 1961, Clymer et al. 1970, Cooney and Hayes 1972, Cooney and Burgdorfer 1974, Zimmerman et al. 1988). The data also support previous findings indicating the preference of the immature stages of D. variabilis for rodents, and considerably less for lagomorphs (Cooney and Hayes 1972, Cooney and Burgdorfer 1974, Sonenshine and Stout 1970, Zimmerman et al.1987). Out of a total of 202 rabbits, only 36 (17.8%) were infested with D. variabilis, yielding 109 ticks and producing an overall average infestation of 0.54 ticks per rabbit. In contrast however, the intensity index (average number per infested host) was 3.03 suggesting the effect of other factors than merely host preference on tick density. Regardless of the cause, these infestation rates for rabbits were still considerably higher than those reported by other authors cited previously in this paper, and reflect primarily infestation with the nymphal stage (70.6% of infesting ticks). The much higher infestation rates for eastern cottontail rabbits reported in this study may be attributed to increased availability of this species as a host for D. variabilis, and inherently higher tick densities as well, associated with higher populations of the preferred hosts cohabiting the study area. Conversely, studies at the TVA's Land Between the Lakes (LBL) area in southwest Kentucky and northwest Tennessee by Cooney and Burgdorfer (1974) and Zimmerman et al. (1988) corroborate this theory. Their mammal collections included proportionately fewer lagomorphs and rodents than did the northwest Alabama collections and were derived from

an area which is not frequented by domestic or feral dogs, the preferred host for the adult stage of D. variabilis. In contrast, several species of rodents had much higher frequencies of infestation and produced proportionately larger numbers of ticks per host as well. Oryzomys palustris, although represented by only one specimen, was infested with five ticks. Forty-four percent of all M.ochragaster were infested and produced an average infestation per host of 1.07, twice the rate for cottontail rabbits. Thirty-two percent of all P. leucopus were infested and produced an average infestation per host of 1.38, almost three times the rate for cottontail rabbits. Peromyscus gossypinus was infested 20% of the time and produced an average infestation rate of 0.67, still higher than the rate for rabbits. Only 12.99% of all S. hispidus were infested, however the average infestation per host was 1.60, about three times the rate of infestation of cottontail rabbits. In summarizing, five of 13 rodent species collected had higher average rates of infestation with D. variabilis than did cottontail rabbits, suggesting a preference for these hosts. Further evidence of preference was indicated by density infestation data per host. The most heavily infested rodent contained 55 ticks, whereas the most heavily infested rabbit contained only 17. Of the D. variabilis ticks infesting rodents, 73.8% were larvae, whereas infestations on rabbits were 70.6% nymphs. Each of the three tick species more commonly associated with eastern cottontail rabbits, namely H. leporispalustris, A. americanum, and I. dentatus, has been found naturally infected with rickettsiae and has been incriminated as an actual or potential vector of the spotted fever agent R. rickettsii (Burgdorfer 1975). However, recent advances in serologic procedures, particularly the application of the indirect microimmuno-fluorescence test to the typing of tick-borne rickettsiae (Philip et al. 1978), has revealed that in many instances the rickettsial agents associated with these ticks are not R. rickettsii but rather closely related agents, such as R. montana, R. rhipicephali, or hitherto unclassified agents. Because of similarities in antigenic makeup, these closely related agents elicit cross reactions with spotted fever antigens, especially in the complement fixation test. Thus, it appears likely that the high percentage of seropositive cottontail rabbits recorded in TVA's Land Between The Lakes (Burgdorfer et al. l974) reflects, in part at least, exposure to spotted fever group rickettsiae that are distinct from R. rickettsii and are transmitted by ticks with a greater predilection for rabbits than the American dog tick, D. variabilis. Ultimally it has been shown (Burgdorfer et al. 1980) that cottontail rabbits, although susceptible to R. rickettsii, are not efficient sources for infecting D. variabilis. Infection of this tick, therefore, has to originate from different sources, possibly mice and voles. In the laboratory, at least, these rodents develop infections with rickettsemias of sufficient concentration to infect simultaneously feeding larval and nymphal ticks (Burgdorfer et al. 1966).

Vol. 30, no. 2 Acknowledgments

Journal of Vector Ecology

179

Sincere thanks are extended to Bobby McDuff and Ben Beard, former employees of the Tennessee Valley Authority, for their tabulation and computation of data and for their assistance in processing specimens. We are also indebted to the dedicated efforts of John Upton and Thomas L. Hill for their untiring work in obtaining specimens under rather rigorous conditions. These studies, and the cost of publication, were supported by Northwest Shoals Community College and the Tennessee Valley Authority. REFERENCES Anderson, B.E., K.G. Sims, J.G. Olson, J.E. Childs, J.F. Piesman, C.M. Happ, G.O. Maupin, and B.J.B. Johnson. 1993. Amblyomma americanum: a potential vector of human ehrlichiosis. Am. J. Trop. Med. Hyg. 49: 239244. Bishopp, F. C. and H. L. Trembley. 1945. Distribution and hosts of certain North American ticks. J. Parasitol. 31: 154. Burgdorfer, W. 1975. A review of Rocky Mountain spotted fever (tick-borne typhus), its agent, and its tick vectors in the United States. J. Med. Entomol. 12: 269-278. Burgdorfer, W., J.C. Cooney, and L.A. Thomas. 1974. Zoonotic potential (Rocky Mountain spotted fever and tularemia) in the Tennessee Valley Region. II. Prevalence of Rickettsia rickettsii and Francisella tularensis in mammals and ticks from Land Between The Lakes. Am. J. Trop. Med. Hyg. 23: 109-117. Burgdorfer, W., J.C. Cooney, A.J. Mavros, W.L. Jellison, and C. Maser. 1980. The role of cottontail rabbits (Sylvilagus spp.) in the ecology of Rickcttsia rickettsii in the United States. Am. J. Trop. Med. Hyg. 29: 686-690. Burgdorfer, W., K.T. Friedhoff, and J.L. Lancaster. 1966. Natural history of tick-borne spotted fever in the U.S.A. Susceptibility of small mammals to virulent Rickettsia rickettsii. Bull. Wld. Hlth. Org. 35: 149-153. Burkot, T.R., G.R. Mullen, R. Anderson, B.S. Schneider, C.M. Happ, and N.S. Zeidner. 2001. Borrelia lonestari DNA in adult Amblyomma americanum ticks, Alabama. Emerg. Infect. Dis. 7: 471-473. Clifford, C.M., G. Anastos, and A. Elbl. 1961. The larval ixodid ticks of the eastern United States. Misc. Publ. Entomol. Soc. Am. 2: 215-244. Clymer, B.C., D.E. Howell, and J.A. Hair. 1970. Animal hosts of economically important ticks (Acarina) in east-central Oklahoma. Ann. Entomol. Soc. Am. 63: 612-614. Cooley, R.A. 1946. The genera Boophilus, Rhipiciphalus, and Haemaphysalis (Ixodidae) of the new world. Natl. Inst. Hlth. Bull. 187 pp. Cooley, R.A. and G.M. Kohls. 1944. The genus Amblyomma (Ixodidae) in the United States. J. Parasitol. 30: 77-111. Cooney, J.C. and W. Burgdorfer. 1974. Zoonotic potential (Rocky Mountain spotted fever and tularemia) in the Tennessee Valley Region. I. Ecologic studies of ticks infesting mammals in Land Between The Lakes. Am. J.

Trop. Med. Hyg. 23: 99-108. Cooney, J.C. and K.L. Hayes. 1972. The ticks of Alabama (Ixodidae: Acarina). Ala. Agric. Exp. Station Bull. No.426. Des Vignes, F., M.L. Levin, and D. Fish. 1999. Comparative vector competence of Deracentor variabilis and Ixodes scapularis (Acari:Ixodidae) for the agent of human granulocytic ehrlichiosis. J. Med. Entomol. 36: 182-185. Feldman, K.A., R.E. Enscare, S.L. Lathrop, B.T. Matyas, M. McGuill, M.E. Schriefer, D. Stiles-Enos, D.T. Dennis, L.R. Peterson, and E.B. Hayes. 2001. An outbreak of primary pneumonic tularemia on Martha's Vineyard. N. Engl. J. Med. 345: 1601-1606. Hechemy, K.E., D. Paretsky, D.H. Walker, and L.P. Mallavia. 1990. Rickettsiology: Current issues and perspectives. 8th Meeting of the American Society for Rickettsiology and Rickettisal Diseases, Diamond, New York, 22-26 September 1989. Ann. N.Y. Acad. Sci. 590 pp. Hornick, R. 2001. Tularemia revisited. N. Engl. J. Med. 345: 1637-1639. Jellison, W.L. 1945. The geographical distribution of Rocky Mountain spotted fever and Nuttall's cottontail in the western United States. Publ. Hlth. Rep. 60: 958-961. Johnson, E.M., S.A. Ewing, R.W. Barker, J.C. Fox, D.W. Crow, and K.M. Kocan. 1998. Experimental transmission of Ehrlichia canis (Rickettsiales:Ehrlichieae) by Dermacentor variabilis (Acari:Ixodidae). Vet. Parasitol. 74: 277-288. Levine, J.F., D.E. Sonenshine, W.L. Nicholson, and R.T. Turner. 1991. Borrelia burgdorferi in ticks (Acari: Ixodidae) from coastal Virginia. J. Med. Entomol. 28: 668-674. Lindsay, L.R., I.K. Barker, G.A. Surgeoner, S.A. McEwen, L.A. Elliott, and J. Kolar. 1991. Apparent incompetence of Deracentor variabilis (Acari:Ixodidae) and fleas (Insecta: Siphonaptera) as vectors of Borrelia burgdorferi in a Ixodes dammini endemic area of Ontario, Canada. J. Med. Entomol. 28: 750-753. Marconi, R.T., D. Liveris, and I. Schwartz. 1995. Identification of novel insertion elements, restriction fragment length polymorphism patterns, and discontinuous 23S rRNA in lyme disease spirochetes: Phylogenetic analyses of rRNA genes and their intergenic spacers in Borrelia japonica sp. nov. and genomic group 21038 (Borrelia andersonii sp. nov.) isolates. J. Clin. Microbiol. 33: 2427-2434. Mather, T.N. and M.E. Mather. 1990. Intrinsic competence of three ixodid ticks (Acari) as vetors of the Lyme disease spirochete. J. Med. Entomol. 27: 646-650. Mukolwe, S.W., A.A. Kocan, R.W. Barker, K.M. Kocan, and G.L. Murphy. 1992. Attempted transmission of Borrelia burgdorferi (Spirochaetales:Spirochaetaceae) (JDI strain) by Ixodes scapularis (Acari:Ixodidae), Dermacentor variabilis, and Amblyomma americanum. J. Med. Entomol. 29: 673-677. Oliver, J.H., T.M. Kollars, F.W. Chandler, A.M. James, E.J. Masters, R.S. Lane, and L.O. Huey. 1998. First isolation and cultivation of Borrelia burgdorferi sensu lato from

180

Journal of Vector Ecology

December 2005

Missouri. J. Clin. Microbiol. 36: 1-5. Parker, R.R. 1923. Transmission of Rocky Mountain spotted fever by the rabbit tick Haemaphysalis leporispalustris. Packard. Am. J. Trop. Med. 3: 39-45. Parker, R.R., E.G. Pickens, D.B. Lackman, E.J. Bell, and F.B. Thraikill. 1951. Isolation and characterization of Rocky Mountain spotted fever rickettsiae from the rabbit tick, Haemaphysalis leporispalustris Packard. Publ. Hlth. Rep. 66: 455-463. Philip, R.N., E.A. Casper, W. Burgdorfer, R.K Gerloff, L.E. Hughes, and E.J. Bell. 1978. Serologic typing of rickettsiae of the spotted fever group by microimmunofluorescence. J. Immunol. 121: 1961-1968. Piesman, J. and C.M. Happ. 1997. Ability of the Lyme disease spirochete Borrelia burgdorferi to infect rodents and thre species of human-biting ticks (blacklegged tick, American dog tick, lone star tick) (Acari:Ixodidae). J. Med. Entomol. 32: 451-456. Piesman, J. and R.J. Sinsky. 1988. Ability of Ixodes scapularis, Dermacentor variabilis, and Amblyomma americanum (Acari:Ixodidae) to acquire, maintain, and transmit Lyme disease spirochetes (Borrelia burgdorferi ). J. Med. Entomol. 23: 336-339. Sanders, F.H. and J.H. Oliver. 1995. Evaluation of Ixodes scapularis, Amblyomma americanum, Dermacentor

variabilis (Acari:Ixodidae) from Georgia as vectors of a Florida strain of the lyme disease spirochete, Borrelia burgdorferi. J. Med. Entomol. 32: 402-406. Sonenshine, D.E. and I.J. Stout. 1970. A contribution to the ecology of ticks infesting wild birds and rabbits in the Virginia-North Carolina Piedmont (Acarina:Ixodidae). J. Med. Entomol. 7: 645-654. Tarnvik, A. and L. Berglund. 2003. Tularemia. Eur. Respir. J. 21: 361-373. Telford, S.R. and A. Spielman. 1989. Competence of a rabbitfeeding Ixodes (Acari:Ixodidae) as a vector of the Lyme disease spirochete. J. Med. Entomol. 26: 118-121. Zimmerman, R.H., G.R. McWherter, and S.R. Bloemer. 1987. Role of small mammals in population dynamics and dissemination of Amblyomma americanum and Dermacentor variabilis (Acari:Ixodidae) at Land Between the Lakes, Tennessee. J. Med. Entomol. 24: 370375. Zimmerman, R.H., G.H. McWherter, and S.R. Bloemer. 1988. Medium sized mammal-hosts of Ambloymma americanum and Dermacentor variabilis (Acari:Ixodidae) at Land Between the Lakes, Tennessee, and effects of integrated tick management on host infestations. J. Med. Entomol. 25: 463-466.

Information

183 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

361928

You might also be interested in

BETA