INVESTIGATIONS OF THE AFRICAN MALARIA MOSQUITO (ANOPHELES GAMBIAE S.L., DIPTERA: CULICIDAE): OVIPOSITIONAL BEHAVIOR AND TOXICITY OF AVERMECTINS By Megan L. Fritz A DISSERTATION Submitted to Michigan State University In partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Entomology and Ecology, Evolutionary Biology and Behavior 2011 ABSTRACT: INVESTIGATIONS OF THE AFRICAN MALARIA MOSQUITO (ANOPHELES GAMBIAE S.L., DIPTERA: CULICIDAE): OVIPOSITIONAL BEHAVIOR AND TOXICITY OF BLOOD-BORNE INSECTICIDES By Megan L. Fritz Anopheles gambiae sensu stricto is a major malaria vector in Sub-Saharan Africa. High malaria morbidity and mortality rates are experienced by humans, mainly children under the age of 5, living in endemic regions. Studiesof the basic biology of this mosquito, including oviposition, provide a background for assessing which attributes might be exploited for suppressing An. gambiae s.s. populations. Ovipositional periodicity of groups assessed using a modified battery-powered wall clock revealed bimodal egg deposition. Oviposition by caged An. gambiae s.s.groups is most probable in early to mid scotophase, or early photophase. Confined An. gambiae s.s. individuals oviposit in a single ca. 2–4 h continuous bout per 24 h. However, some oviposition can occur at any hour during 24 h, especially if females were previously deprived of ovipositional substrate. Many females sit on a dark, wet horizontal substrate while ovipositing. However, vertical resting sites are adequate perches from which oviposition can occur if they provide high humidity and the paramount dark and wet ovipositional cues for An. gambiae s.s. oviposition. Under such conditions, laboratory tests revealed that An. gambiae s.s. were as likely to rain eggs down from a vertical perch, as to oviposit while sitting horizontally on a substrate of moist mud or open water. These studies confirm remarkable ovipositional flexibility by An. gambiae s.s., and I conclude that oviposition is not a suitable target for An. gambiae s.s. management. Host choice appears to be an attribute that is relatively restricted for An. gambiae sensu lato. Despite increasing selection pressure imposed by ITN use in Western Kenya, most blood meals from An. gambiae s.s. were imbibed from human hosts, whereas sibling species, An. arabiensis, blood fed upon cattle. Prior to 1998 An. gambiae s.s. comprised between > 70% of the total An. gambiae s.l. population in Kisian village. In the present study, >50% of the An. gambiae s.l. collected were identified as An. arabiensis, revealing a shift in the dominant vector species. Ivermectin, a cattle dewormer, is highly lethal to both An. gambiae s.s. and An. arabiensis. When cattle are treated with ivermectin, ninety percent of An. gambiae s.s. that fed upon them within 2 wk of treatment failed to survive >10 d post blood meal. No eggs were deposited by An. gambiae s.s. fed on ivermectin-treated cattle within 10 d of treatment. According to the label, commercial formulations of ivermectin cannot be used on pregnant or lactating animals. Laboratory tests of eprinomectin, a commercially available dewormer for pregnant and lactating cattle, revealed equivalent toxicity to An. arabiensis at low concentrations (LC50 of 8.5 ppb). These results suggest that cattle treated with ivermectin or eprinomectin in the prescribed range of low dosages as parasiticides have blood toxic to zoophilic malaria vectors. Regionally-coordinated, seasonal treatment of cattle could suppress An. arabiensis populations, thereby reducing malaria transmission. Dedicated with love to: Brad, who has supported me Oliver and Corrina, who have inspired me My parents, who have encouraged me iv ACKNOWLEDGEMENTS Special thanks to my major professors, Drs. James Miller and Edward Walker, whose guidance and support has given me a love and appreciation for scientific research. Thank you to my graduate committee, Drs. Rufus Isaacs, Linda Mansfield, and Ke Dong for your advice and insight. Your contributions have helped me develop as an aspiring scientist. I would also like to extend special thanks to Dr. Gabriel Ording for his commitment to the instruction of graduate and undergraduate students alike. I would also like to extend a special thanks to Piera Siegert, who has shown me the proverbial mosquito rearing ropes, and given me countless advice, both spoken and unspoken, on the management of scientific studies. Thank you, also to the numberous Miller and Walker lab undergraduate students that I have had the pleasure of working with over the years. Support for this project was funded the National Science Foundation Graduate Research Fellowship Program, Hutson Endowment in the Department of Entomology at Michigan State University, and the Predissertation Travel Award from the International Studies Program at Michigan State University. v TABLE OF CONTENTS LIST OF TABLES ix LIST OF FIGURES x GLOSSARY OF ABBREVIATIONS xii GENERAL INTRODUCTION: BIOLOGY, ECOLOGY, AND BEHAVIOR OF THE SIBLING SPECIES, ANOPHELES GAMBIAE S.S. AND ANOPHELES ARABIENSIS AS THEY PERTAIN TO HUMAN MALARIA TRANSMISSION 1 Impact of malaria in Africa 1 Malaria transmission 1 Biology of the mosquito vectors 5 PART I: STUDIES ON THE OVIPOSITIONAL BIOLOGY OF ANOPHELES GAMBIAE S.S. 8 CHAPTER 1: OVIPOSITIONAL BIOLOGY OF ANOPHELES GAMBIAE S.S. Chemical cues Visual/Physical cues 9 9 10 CHAPTER 2: OVIPOSITIONAL PERIODICITY OF CAGED ANOPHELES GAMBIAE INDIVIDUALS Introduction Methods Mosquitoes Automated Ovipositional Clock Experimental Series 1 – Automated measurement of caged mosquito groups Experiment 2 – Automated measurement of caged individuals Experiment 3 – Assessment of capacity for mid-afternoon oviposition Results Experimental Series 1 – Bimodal oviposition by caged groups Experiment 2 – Caged individuals oviposit in one continuous bout Experiment 3 – An. gambiae are capable of mid-afternoon oviposition Discussion CHAPTER 3: EGG PLACEMENT BY ANOPHELES GAMBIAE (DIPTERA: CULICIDAE) AS INFLUENCED BY RESTING SURFACE SLOPE AND ENCLOSURE Introduction Methods Mosquitoes and bioassay conditions Experiment 1 – Exposed resting surface slope preference Experiment 2 – Resting surface counts Experiment 3 – Enclosed resting surface slope preference Experiment 4 – Presence/absence of surrogate vegetation vi 14 15 16 16 17 19 20 21 22 22 24 28 28 33 33 35 35 36 38 39 39 Results Discussion 41 45 PART II: STUDIES ON THE FEASIBILITY OF AVERMECTIN-TREATED CATTLE AS A VECTOR CONTROL MEASURE IN SUB-SAHARAN AFRICA 48 CHAPTER 4: AVERMECTINS AND MOSQUITO CONTROL Chemical and pharmacological properties of the avermectins Alternatives to ivermectin Mode of action of the avermectins Safety of ivermectin and moxidectin in vertebrates Current vector control methods 49 49 51 52 53 54 CHAPTER 5: TOXICITY OF BLOODMEALS FROM IVERMECTIN-TREATED CATTLE TO ANOPHELES GAMBIAE S.L. Introduction Methods Mosquitoes Delivery of ivermectin-treated blood via artificial membrane feeder Mosquito feeding bioassay on ivermectin- and moxidectin-treated cattle Statistical Analysis Results An. gambiae s.l. LC50 and LC95 estimates from lab feedings of ivermectin-treated blood Survivorship and fecundity of An. gambiae fed on treated bulls Discussion 56 56 58 58 59 60 61 63 63 63 68 CHAPTER 6: LETHAL AND SUBLETHAL EFFECTS OF AVERMECTIN/MILBEMYCIN PARASITICIDES ON THE AFRICAN MALARIA VECTOR, ANOPHELES ARABIENSIS (DIPTERA: CULICIDAE) Introduction Methods Mosquitoes Preparation of treated blood Doramectin extraction and HPLC Data collection and statistical analysis Results Discussion 70 70 72 72 73 75 76 77 82 CHAPTER 7: HOST UTILIZATION BY ANOPHELES GAMBIAE S.S. AND ANOPHELES ARABIENSIS IN AN AREA OF HIGH INSECTICIDE-TREATED BED NET COVERAGE IN WESTERN KENYA AS DETERMINED BY REVERSE DOT BLOT, DNA-DNA HYBRIDIZATION Introduction Methods Study site Mosquito collection and species identification Blood meal identification via direct sequencing Host DNA amplification with Cyto primer pair 85 86 88 88 88 89 89 vii Reverse dot blot analysis 90 93 94 Results Discussion FINAL CONCLUSIONS AND FUTURE RESEARCH DIRECTIONS Part I: Ovipositional flexibility of An. gambiae Part II: An. arabiensis control using eprinomectin 100 100 101 APPENDIX 102 REFERENCES 105 viii LIST OF TABLES Table 2.1. Mid afternoon egg output by Anopheles gambiae as influenced by previous access to an ovipositional resource. 29 Table 3.1. Analysis of variance for Experiment 3. An asterisk denotes a significant p-value. 42 Table 5.1. Viability of eggs laid by Anopheles gambiae s.s. and An. arabiensis fed treated vs. untreated blood meals. No difference in viability was observed (p = 0.78). 64 Table 5.2. Log-rank test for differences in survivorship in Anopheles gambiae s.s. taking different blood meal sizes within the ivermectin treatment group. P-values are unadjusted for multiple comparisons. 67 Table 6.1. Blood concentrations used to determine the effects of each AI on An. arabiensis survivorship and fecundity. 74 Table 6.2. Abbott‟s corrected LC50 and LC95 values for An. arabiensis fed upon eprinomectin-, ivermectin-, doramectin- and moxidectin-treated blood, given in parts per billion of drug in the mosquito blood meal. The  and  values represent the intercept and slope of the logistic regression analyses, respectively. 78 Table 6.3. Mean number of eggs per surviving female with bootstrapped 95% confidence intervals (N=1000). A significant p-value (denoted by *) signifies that increasing the drug dosage significantly reduces egg production in surviving Anopheles arabiensis. 81 Table 6.4. Mean proportion of eggs oviposited out of the total number of eggs produced (± bootstrapped 95% CIs; N=1000) by Anopheles arabiensis females surviving dosages lower than each drug‟s LC50 compared with respective controls. 82 Table 7.1. Oligonucleotide probes imprinted upon membrane strips for reverse dot blot procedure . 91 Table 7.2. Resting sites from which An.gambiae s.s. and An. arabiensis were collected. Indoor resting sites included the sides and underside of furniture, and the interior walls of houses. Outdoor resting sites were the interior of clay pots used for rain catchment outside of houses. Individuals classified as unknown were morphologically identified as An. gambiae s.l., but could not be identified to species level via qPCR. 95 Table 7.3. An. gambiae s.s. and An. arabiensis hosts were identified by at least 1 method of blood meal analysis. Individuals whose blood meals were mixed or avian were classified as “other”. ix 96 LIST OF FIGURES Figure 1.1. Parasite maturation time within the mosquito vector vs. external temperature. Adapted from Patz and Olson (2006). 3 Figure 1.2. Eave-opening in traditional housing of the Luo tribe in Kisian, Kenya. 4 Figure 2.1. Clock apparatus used in automated measurement of Anopheles gambiae ovipositional periodicity. 18 Figure 2.2. Ovipositional periodicity of laboratory strain Anopheles gambiae groups. Mean percent of eggs oviposited per hour by a caged laboratory strain group (500 females per replicate; total eggs = 18,303). 23 Figure 2.3. Ovipositional periodicity of house-collected Anopheles gambiae groups. Mean percent of eggs oviposited per hr by caged house-collected groups (100 females per replicate; total eggs = 11,007). Measurements of outdoor light intensity, represented in the graph as a dark blue line, were taken 5/11/2004 in Kisumu, Kenya. 25 Figure 2.4. Ovipositional patterns of individual Anopheles gambiae over 24 hrs. Each horizontal cluster of rectangles represents a single individual. Shading classifies the number of eggs deposited per individual per hr (n = 56) during the oviposition interval. Gray shading represents the preoviposition interval. 26 Figure 2.5. Accumulated ovipositional patterns of individual Anopheles gambiae over 24 hrs. Percent of eggs oviposited per hr by caged individual An. gambiae (total eggs = 4,815). 27 Figure 3.1. A. Mean percentage of Anopheles gambiae eggs (± 1 SEM) collected from ovipositional resources with exposed resting surfaces at varying inclines with respect to the cage floor (total eggs = 17,134; mean no. eggs/female = 37). B. Mean number of resting mosquitoes (± 1 SEM) counted on ovipositional resources with exposed resting surfaces at varying inclines with respect to the cage floor. Bars within a graph sharing the same letter are not significantly different (p>0.05). 37 Figure 3.2. Mean percentage of eggs (± 1 SEM) deposited on ovipositional resources enclosed by no walls, 2.7cm or 10cm high walls (total eggs = 16,813; mean no. eggs/female = 47). Dark and light bars indicate high and low RH respectively. Mean separation was applied within each RH, and bars within an RH treatment sharing the same letter are not significantly different (p>0.05). 43 Figure 3.3. Map of egg deposition within the tall vertical cylinder adapted from photographed results of the first replicate of Experiment 3. Black dots represent individual eggs, while outer and inner black circles represent the Petri dish and sponge walls respectively. x 44 Figure 4.1. Chemical structure of ivermectin (adapted from Edwards 2003). 50 Figure 5.1. Median survival time (d) of Anopheles gambiae s.s. with bootstrapped confidence intervals (N = 1000). An. gambiae s.s. sample sizes were n= 643, n= 615, and n = 666 for ivermectin, moxidectin and untreated groups respectively. 62 Figure 5.2. Kaplan-Meier survivorship curves for Anopheles gambiae s.s. that bloodfed upon ivermectin-treated cattle. Each curve corresponds to a uinique post-treatment interval. The arrow indicates the required development time for Plasmodium falciparum in the body of a mosquito at an external temperature of 27⁰ C (Patz and Olson 2006). 65 Figure 5.3. Mean (±95% CIs) number of eggs deposited per female per treatment group over 8 different post-treatment intervals (d). 66 Figure 6.1. Uncorrected mean proportion of Anopheles arabiensis deceased (left) and production of eggs by survivors (right) within 9 d of feeding upon eprinomectin- and ivermectin-treated blood. Error bars represent bootstrapped 95% confidence intervals (N=1000). 79 Figure 6.2. Uncorrected mean proportion of Anopheles arabiensis deceased (left) and production of eggs by survivors (right) within 9 d of feeding upon doramectin- and moxidectin-treated blood. Error bars represent bootstrapped 95% confidence intervals (N=1000). Doramectin dose-response curve was adjusted using HPLC analysis of final extraction product. 80 Figure 7.1. Schematic of a membrane strip used for reverse dot blot analysis, including placement of the probes. On the right, a membrane strip used in a test reveals that the unknown host was bovine. 92 Figure 7.2. Relative visualization of probes in a mixed blood meal based upon the proportion of DNA from each host. Gel bands demonstrate a roughly equal quantity of DNA per samples, yet relative intensity of the RDBA probe varies based upon the proportion of DNA from each host. In this figure, background brightness for some strips was adjusted to create equivalent background brightness across all strips. 97 xi GLOSSARY OF ABBREVIATIONS Ae. – Aedes An. – Anopheles AVM – avermectin B1a Cx. – Culex D. – Drosophila d – day GABA - γ-aminobutyric acid h - hour IRS – Indoor residual spraying ITN – Insecticide-treated bed net P. – Plasmodium sensu lato – wide sense sensu stricto – narrow sense wk - week xii GENERAL INTRODUCTION: BIOLOGY, ECOLOGY, AND BEHAVIOR OF THE SIBLING SPECIES, ANOPHELES GAMBIAE S.S. AND ANOPHELES ARABIENSIS AS THEY PERTAIN TO HUMAN MALARIA TRANSMISSION Impact of Malaria in Africa Human malaria is the scourge of the tropics, and is devastating to human health, economic growth, and regional development. Each year, malaria causes 2.7 million deaths world-wide, though 90% is experienced by people, particularly children, living in SubSaharan Africa (Breman et al. 2001, Molyneux 2004). Countries at risk for severe malaria are often impoverished, and experience very slow, or negative economic growth. Between 1965 and 1990, growth of per capita income in countries at low or no risk for malaria was 5 times higher than in malarious countries (Gallup and Sachs 2001). In areas where malaria abounds, development of human society is impeded by medical costs, premature mortality, low worker productivity, absenteeism, and reduced savings and investment (Sachs and Malaney 2002). Malaria transmission The malaria parasites Plasmodium falciparum, P. vivax, P. ovale and P. malariae are transmitted between humans by Anopheles mosquito vectors. P. falciparum causes the majority of life-threatening and deadly malaria cases throughout the tropics. Newly eclosed mosquitoes are only able to transmit malaria after feeding on human blood containing mature male and female gametocytes. Male and female gametocytes unite in the gut of the mosquito and the parasite begins a complex cycle of multiplication and development. Mature parasites (sporozoites) that are infective to humans eventually migrate to the mosquito‟s salivary 1 glands, but the time required for parasite maturation depends on the Plasmodium species and the external temperature. The two most prevalent parasites, P. vivax and P. falciparum, require approximately 8-13 days to complete this cycle within the mosquito if external temperatures are 27ºC (Pampana 1969). Lower external temperatures lead to slower parasite maturation (Figure 1.1). Infected Anopheles vectors only transmit malaria if they outlive the required time for parasite maturation (Zucker 1996). The primary vectors of malaria in Sub-Saharan Africa are the mosquitoes of the Anopheles gambiae Giles sensu lato complex (An. gambiae s.l.), and An. funestus Giles. Anopheles gambiae sensu lato (Latin for “wide sense”) refers to a complex of morphologically indistinguishable species. This complex includes An. gambiae sensu stricto (Latin for “narrow sense”, and hereafter referred to as An. gambiae), An. arabiensis Patton, An. quadriannulatus, An. bwambae, and the salt water species An. merus and An. melas. The sibling species An. gambiae and An. arabiensis, reputed to be two of the most efficient vectors of malaria in the world (Coetzee et al. 2000), are the focus of my research. Malaria transmission is dependent upon the prevalence of competent vector species (Omumbo et al. 1998) and the availability of suitable aquatic vector breeding sites (Holstein 1954). As larval habitat availability increases at the onset of the rainy season, populations of these species grow rapidly (Joshi et al. 1975). Mean indoor resting density of An. gambiae and An. arabiensis may reach as high as 223 females per house, and average human bite rate can reach over 170 bites per human per night in some regions of Sub-Saharan Africa (Molineaux and Gramiccia 1980). Human biting rates can be attributed not only to high vector abundance, but also accessibility of human hosts. For example, mosquitoes easily pass in and out of homes with open eaves (Figure 1.2), and have nightly access to sleeping humans. 2 40 35 Development time (d) 30 25 20 P. falciparum 15 P. vivax 10 5 0 14 16 18 20 22 24 26 28 30 32 Temperature (ºC) Figure 1.1. Parasite maturation time within the mosquito vector vs. external temperature. Adapted from Patz and Olson (2006). 3 Figure 1.2. Eave-opening in traditional housing of the Luo tribe in Kisian, Kenya. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this dissertation. 4 Although the proportion of An. gambiae complex mosquitoes containing mature sporozoites varies only between 1.9 and 11.8 % in Western Africa (Molineaux and Gramiccia 1980), the high human biting rate significantly elevates the risk of malaria infection during the rainy season. Vectorial capacity is positively correlated with both biting frequency (Garrett-Jones 1964b) and especially vector life-span (Gary and Foster 2001). Some An. gambiae and An. arabiensis females live for 50 days or more (Gary and Foster 2001). Longer-lived females take more blood meals during their lifetime, which increases the probability of acquiring and spreading malaria parasites. Biology of the mosquito vectors Populations of An. gambiae and An. arabiensis are widely distributed throughout SubSaharan Africa, and often occur in sympatry. They are morphologically indistinguishable, but can be identified to species level using polymerase chain reaction (Paskewitz and Collins 1990). An. gambiae and An. arabiensis were considered a single species until 1962, when their distinct behaviors and physiologies led to the recognition of the An. gambiae complex (White 1974). Of these two sibling species, An. gambiae predominates in more humid environments (White 1974). In Western Kenya, An. gambiae larvae are found in ground pools, irrigation ditches, puddles, tire tracks and hoof prints (Minakawa et al. 1999; Mutuku et al. 2006) shortly after the onset of the rainy season. Larval development is completed within 5-12 days depending on the temperature (Bayoh and Lindsay 2003), as well as density of larvae per habitat (Gimnig et al. 2002). Adult An. gambiae are well-known for high endophily (indoor resting behavior) and anthropophagy (White and Rosen 1973, White 1974). An. gambiae will preferentially feed on humans, and the human blood index (the proportion of blood meals taken from a human host; Garrett-Jones 1964a) of this species remains at greater than 80 %, 5 even in the presence of alternative hosts (White 1974). Peak indoor biting time for An. gambiae females occurs between 0100 and 0600 h, though bites can occur anytime between 2000 and 0600 h (Githeko et al. 1996). In Western Kenya, freshly blood-fed An. gambiae are more than twice as likely to rest indoors than outdoors (Odiere et al. 2007) while processing their blood meal, although this mosquito is frequently found resting inside houses regardless of physiological state (White 1974). While An. arabiensis often shares larval habitats and ovipositional sites with its sibling species, An. gambiae (White and Rosen 1973, Minakawa et al. 1999), they can also utilize slow-flowing riverine habitats for egg deposition during the dry season, when transient ground pools disappear (Dukeen and Omer 1986). The ability of An. arabiensis to use these year-round habitats for reproduction, along with their higher body water content (Gray and Bradley 2005), allows this mosquito to inhabit drier geographic regions, and predominate during periods of low rainfall. Therefore, the relative proportions of An. gambiae and An. arabiensis vary due to: 1) microclimatic differences in a region, 2) monthly seasonal rainfall, and 3) yearly changes in average relative humidity (Highton et al. 1979). An. arabiensis is also anthropophagic, but will opportunistically feed on other hosts. In Kenya, it was reported that only 39% of An. arabiensis fed on humans, but 59% took blood meals from cattle (Highton et al. 1979). In Ethiopia, however, 51% of outdoor resting, and 66% of indoor resting An. arabiensis contained human blood (Tirados et al. 2006). At this same location, ca. 25% An. arabiensis contained mixed cattle-human blood meals. The resting behavior of An. arabiensis combines endophily and exophily (outdoor resting behavior). It has been reported that populations of An. arabiensis in Northern Sudan are almost entirely endophilic (Dukeen and Omer 1986), but populations from Southern Ethiopia are five times more likely to rest outdoors than indoors (Tirados et al. 2006). Behavioral differences between An. arabiensis populations are likely linked to genetic variability 6 (Coluzzi et al. 1979), and indoor vector control measures, such as indoor residual spraying (IRS), exert heavy selection pressure for exophily in An. arabiensis populations (White 1974). Reduction in malaria transmission by these important vector species relies upon a thorough understanding of the vector‟s life history. Life history research identifies behavioral and physiological traits that can be exploited for vector control. Once potential vector control methods are identified, their efficacy and environmental impacts must be thoroughly investigated before widespread implementation in the field. The following dissertation includes research on the basic ovipositional biology of the mosquito, An. gambiae, as well as analysis of the efficacy of a novel vector control measure for the sibling species pair, with the ultimate goal of reducing malaria transmission and saving human lives. 7 PART I: STUDIES ON THE OVIPOSITIONAL BIOLOGY OF ANOPHELES GAMBIAE S.S. 8 CHAPTER 1: OVIPOSITIONAL BIOLOGY OF ANOPHELES GAMBIAE S.S. Mosquitoes use a diverse, sometimes species-specific combination of physical, chemical, and temporal cues to find a suitable ovipositional substrate. While ovipositional strategies vary, the outcome is identical across mosquito taxa; females provide their offspring with access to aquatic habitats. Mosquito larvae must ultimately find themselves in these aquatic habitats, containing required food resources and suitable water quality, in order to maximize their fitness. Egg placement into these larval habitats represents one way by which the mother improves the likelihood that some of her offspring survive. Thus, ovipositing females may be sensitive to chemical and physical cues correlated with the quality of a larval habitat. Larvae of An. gambiae are commonly found in small, sunlit pools that are typically the result of human activity (i.e. tire tracks, cattle hoof prints) (Gimnig et al. 2001). These habitats are transient, and typically devoid of vegetation. Attributes of these larval habitats that serve as ovipositional stimulants to An. gambiae have been, and continue to be investigated. Chemical cues Most species of mosquito spend their immature stages in aquatic habitats rich in decaying organic matter. The volatile chemicals 4-methylcyclohexanol, indole, 3methylindol, phenol, m-cresol, o-cresol, and p-cresol are all associated with such habitats (Collins and Blackwell 1998). Metabolism of organic matter by microorganisms in the water is thought to produce these volatiles (Stahl and Parkin 1996). Oviposition by gravid females of some mosquito species appears to be stimulated by water containing one or more of these chemicals (Bentley et al. 1981, Millar et al. 1992, Davis and Bowen 1994, Collins and 9 Blackwell 1998). In one case, the key bacteria and associated volatiles that stimulate oviposition have been identified (Ponnusamy et al. 2008). Gravid An. gambiae oviposit more of their eggs in water from a typical larval habitat, than in tap or distilled water (McCrae 1984, Takken and Knols 1999, Bray 2003, Sumba et al. 2004). Chemicals associated with decaying organic matter that are stimulatory to other mosquito species, particularly 3-methylindole and indole, also elicit an electrophysiological response from An. gambiae antennae (Blackwell and Johnson 2000). Water containing pcresol or phenol is favored compared to distilled water (Bray 2003). Interestingly, water containing either 3-methylindole or indole inhibits oviposition by An. gambiae in the laboratory. Furthermore, bacterial strains that are isolated from native larval habitats are also inhibitory or neutral to ovipositing females (Huang et al. 2006b). It has been postulated that algal volatiles may be more important than bacterial volatiles for stimulating An. gambiae oviposition (Otienoburu et al. 2007). Otienoburu et al. (2007) demonstrated that water from Lake Victoria, near Kisumu, Kenya, is the most potent stimulus for An. gambiae oviposition tested to date. Many species of algae are present in Lake Victoria water (Ochumba and Kibaara 1989), and volatiles from algae are known to influence the behavior of nematodes (Höckelmann 2004). Further research into the stimulatory attributes of algal volatiles may reveal their importance to An. gambiae oviposition. Visual/Physical cues The presence of moisture is paramount to mosquito oviposition. Many insects have thermo- and hygroreceptors that enable them to respond to temperature and humidity gradients (Tichy and Loftus 1996, Tichy and Kallina 2010). These receptors have been found on the tarsi and antennae of Ae. aegypti (Bar-Zeev 1960). In mosquitoes, egg 10 deposition is stimulated by the presence of water, although egg placement behavior at a water body varies with species. For example, some Anopheles, Coquilletidia, and Culex species oviposit directly into the water, while Aedes and Psorophora species often oviposit near aquatic habitats, but above the water line (reviewed in Bentley and Day 1989). Early studies of An. gambiae revealed that gravid females would oviposit directly into the water while either resting on the water surface (Hocking and MacInnes 1948), or hovering over the water (McCrae 1984). More recent investigations of their ovipositional behavior demonstrated that many An. gambiae females deposit their eggs while sitting near a water source. The preponderance of eggs are deposited on moist mud rather than in open water (Miller et al. 2007). In fact, Huang et al. (2005) demonstrated that some An. gambiae will oviposit on sand that feels dry to the human hand, even when standing water is available. Soil moisture is positively correlated with An. gambiae egg deposition, yet the presence of standing water does not appear to be necessary for oviposition (Huang et al. 2005). Color also serves as an important stimulatory cue for gravid female mosquitoes foraging for ovipositional resources (Williams 1962, Beehler et al. 1993, Li et al. 2009). For example, ovipositing Ae. aegypti are most sensitive to yellow-green wavelengths (Snow 1971), while Ae. triseriatus are maximally stimulated to oviposit by blue wavelengths (Williams 1962). Under natural lighting conditions, more Cx. quinquefasciatus deposit their egg rafts in water darkened with India ink (Beehler et al. 1993) than in water alone. McCrae (1984) concluded that egg deposition by An. gambiae was also influenced by ovipositional substrate color. Black substrates received more eggs than either white or grey substrates. In another series of experiments, however, he concluded that turbid water, appearing lighter in color, was maximally stimulatory to ovipositing females (McCrae 1984). Huang et al. (2007) resolved this contradiction by demonstrating that a black ovipositional substrate is highly stimulatory to gravid An. gambiae, but must be placed over a light 11 background to maximize oviposition. Black ovipositional substrates over a white floor received 5-fold more eggs than did black ovipositional substrates over a black floor (Huang et al. 2007). Contrast is likely critical to An. gambiae during ovipositional resource location which occurs under night-time low-lighting conditions. Interestingly, Huang et al. (2007) report that all treatments in their contrast experiment, including a white ovipositional resource over a white background, received some An. gambiae eggs. This indicates a remarkable flexibility by An. gambiae in their acceptance of ovipositional resources. The daily light cycle influences the timing of oviposition among mosquito taxa. Daylight stimulates Ae. aegypti to oviposit, and most oviposition occurs just before dusk (Corbet and Chaddee 1990, Chaddee 2010). Culex (Suleman and Shirin 1981) and Anopheles (Chaddee et al. 1993, Chaddee et al. 1998, Sumba et al. 2004) species oviposit during scotophase, although the peak ovipositional times vary by species. For example, three species within the genus Anopheles have distinct peak ovipositional windows; An. homunculus deposits the preponderance of their eggs in the middle of the night (Chaddee et al. 1998), whereas egg deposition by An. freeborni peaks ca. 4 h prior to the onset of photophase. Oviposition by An. albimanus peaks immediately after the onset of scotophase (Chaddee et al. 1993). Mosquito populations rarely exhibit ovipositional windows that are restricted to a couple of hours. Instead, oviposition by individuals in a population is spread over multiple hours of the day or night. Conflicting data have been reported for the the time of peak oviposition by An. gambiae (Causey et al. 1942, Haddow and Ssenkubuge 1962, McCrae 1983, Sumba et al. 2004). Some studies have demonstrated two ovipositional peaks (Haddow and Ssenkubuge 1962, Sumba et al. 2004) for An. gambiae populations, while others have demonstrated only one (Causey et al. 1942, McCrae 1983). Furthermore, no study has investigated the ovipositional patterns of individual An. gambiae to determine whether the observed 12 bimodality can be attributed to multiple individuals ovipositing at different times, or a single individual splitting oviposition between two different times. The presence or absence of vegetation influences mosquito oviposition. For some species of mosquito, such as Wyeomyia mitchelli, attributes of the vegetation influence oviposition (Frank 1985, reviewed in Clements 2000). In other cases, mosquitoes may oviposit more readily among vegetation due to the low lighting conditions or shade provided (Muirhead Thomson 1940). Some Anophelines, such as An. melas and An. minimus, deposit their eggs in aquatic habitats among vegetation (Giglioli 1965, Clements 2000, Overgaard 2007). Oviposition among vegetation has also been documented for An. gambiae (reviewed in Clements 2000, Minakawa et al. 2004, Huang et al. 2006a). However, there are no studies that investigate how An. gambiae navigate the three-dimensional structures present in their environment during oviposition, and whether any aspects of these structures are stimulatory. The following studies aimed to: 1.) Examine the influence of photoperiod on An.gambiae egg deposition, and compare the daily ovipositional patterns of An. gambiae populations with those of individuals. 2.) Investigate the ways in which ovipositing An. gambiae interact with the threedimensional structures, including surrogate vegetation, in their environment. 13 CHAPTER 2: OVIPOSITIONAL PERIODICITY OF CAGED ANOPHELES GAMBIAE INDIVIDUALS Published in: Journal of Circadian Rhythms, 2008; 6:2. Abstract Anopheles gambiae is a major malaria vector in Sub-Saharan Africa. Studies of the basic biology of this mosquito, including oviposition, provide a background for assessing which attributes might be exploited for suppressing An. gambiae populations. Here, we report on when during the diel cycle An. gambiae individuals deposit eggs as compared to the ovipositional patterns of groups. Battery-powered wall clocks were modified so as to present a unique section of dark and wet ovipositional substrate at hourly intervals over two consecutive 12 h periods. Ovipositional periodicity of mosquito groups (Kisumu laboratory strain or feral females) and individuals was determined by counting the number of eggs present on each section of the ovipositional substrate. Capacity for mid-afternoon oviposition by groups of Kisumu laboratory strain An. gambiae was determined by presenting hypergravid females with an ovipositional substrate exclusively between 1200 and 1600 h. On equatorial time, caged laboratory strain An. gambiae groups deposited 65% of their total eggs between 1800 and 0 h, and the remaining 35% were spread between 0 and 1000 h. Caged house-collected An. gambiae groups deposited 74% of their total eggs between 1800 and 200 h, ceased oviposition for 3 h, and then spread the remaining 26% of their eggs near or after dawn. Ninety-six percent of individual An. gambiae females spread their eggs over a continuous 2-4 h period without interruption. In tests of capacity for mid-afternoon oviposition, females given evening access to an ovipositional resource deposited 2% of their total eggs between 1200 and 1700 h. An. gambiae females given only access to an 14 ovipositional resource between 1200 and 1700 h deposited 3 times more eggs during that time period than did females previously given evening access. Confined individual An. gambiae oviposit in a single ca. 2-4 h continuous bout per 24 h. Oviposition is most probable in early scotophase, mid scotophase, or early photophase. However, some oviposition can occur at any hour during 24 h, especially if females were previously deprived of ovipositional substrate. Introduction Diel ovipositional patterns have been studied in various Diptera, including Drosophila melanogaster (Allemand 1976a, 1976b, 1977, Fleugel 1978), D. pseudoobscura (Fleugel 1984), Delia antiqua (Havukkala and Miller 1987), Chrysomya bezziana (Spradbery 1979), Aedes aegypti (Haddow and Gillett 1957), Anopheles albimanus (Chaddee et al. 1993), An. freeborni (Chaddee et al. 1993), An. albitarsis (Chaddee 1995) and An. gambiae (Causey et al. 1942, Haddow and Ssenkubuge 1962, McCrae 1983). All of these species are reported to deposit the preponderance of their forthcoming eggs within a 2-4 h period, but the time of maximal egg deposition varies interspecifically. However, in no case was oviposition reported to occur strictly within that 2-4 hour window. For An. albimanus (Chaddee et al. 1993) and D. melanogaster (Fleugel 1978), ovipositional rhythm was reported to be bimodal with unequal modes; light intensity during photophase was reported to influence the modality of D. melanogaster oviposition (Allemand 1977). Fluegel (1978) found that light levels furnished by a 40 W white fluorescent bulb resulted in bimodal egg deposition by D. melanogaster individuals. Chadee et al. (1993) reported that individual An. albimanus laid their entire complement of eggs at once rather than splitting them between two different periods. 15 Outcomes of research on the ovipositional periodicity of groups of An. gambiae held in a common cage (Causey et al. 1942, Haddow and Ssenkubuge 1962, McCrae 1983, Sumba et al. 2004) have been divergent. Causey et al. (1942) suggested that An. gambiae was capable of oviposition at any time during the night. However, they observed that five of a total of nine batches of eggs were laid between 2000 and 2300 h. Under equatorial conditions, McCrae (1983) reported wide variation in nocturnal peaks for An. gambiae oviposition. He postulated that the time of peak oviposition during a night was related to the time at which the blood meal was taken. However, Sumba et al. (2004) were unable to confirm this effect. Haddow and Ssenkubuge (1962) and Sumba et al. (2004), reported that oviposition of An. gambiae commenced at scotophase (1800 h) and peaked between 18002100 h. A second but smaller ovipositional peak was documented by both research teams, but at inconsistent times. In all cases, some oviposition occurred throughout scotophase. In one case (Sumba et al. 2004) a feral population deposited about 3% of the total eggs after the onset of photophase. In other cases, it was unclear whether any attention was paid to the possibility of oviposition throughout photophase. Left unknown in all of these studies is whether individual females spread their oviposition across many hours, or whether some individuals deposit all of their eggs early in the night while others deposit all of their eggs near morning. Methods Mosquitoes Two sources of mosquitoes were used: 1) The Kisumu laboratory strain of An. gambiae s.s. originated from the Kenya Medical Research Institute (KEMRI) of Kisumu, Kenya. It was reared at Michigan State University according to Huang et al. (2005). 2) Blood-fed feral females were aspirated from the walls of houses near the KEMRI compound. 16 As previously reported, ca. 90% of the females were An. gambiae s.s. and the remainder were An. arabiensis as determined by PCR (Huang et al. 2005). Cages of mosquitoes were held in an environmental chamber maintained at 28 ± 1º C and 80 ± 10% RH under a LD 12:12 h photoperiod. Indirect light of about 0.17 lx was provided during scotophase by a shaded 4 W tungsten bulb; it was intended to mimic moonlight. Mosquitoes were offered defibrinated horse blood via a membrane feeder 2-3 days before ovipositional tests. These females were held in laboratory cages under high humidity for 24 h before ovipositional tests. Automated Ovipositional Clock The method of egg collection may influence the oviposition rhythm of some Diptera (David and Fouillet 1973). Use of a mechanized egg collector may be less disruptive to egg deposition than manually changing an ovipositional resource at hourly or two-hour intervals. Mechanized egg collectors have previously been utilized in the study of ovipositional rhythms for both D. melanogaster (Fleugel 1984, David and Fouillet 1973) and Agrotis segetum (Byers 1987). In the current study, we developed a new automated mechanical apparatus to sample oviposition over time. This apparatus was used to compare ovipositional periodicity by An. gambiae individuals vs. groups. Battery-powered wall clocks (DSA Incorporated, Farmington Hills, MI, U.S.A.) measuring 31 cm in diameter were modified to progressively present a unique section of dark and wet substrate onto which mosquitoes could oviposit over 12 h (Figure 2.1). Each clock was positioned horizontally and its original face and hands were removed. The clock body was filled with moist sand of particle size 250 – 425μm. The sand was topped with Envision® high-capacity brown paper towelling (Georgia Pacific, Camas, WA, U.S.A). This 17 One hour open section exposing the moist ovipositional resource Figure 2.1. Clock apparatus used in automated measurement of Anopheles gambiae ovipositional periodicity. 18 paper towelling appears light when dry, but dark when wet. Thus, it provided the two key stimuli (dark and wet) necessary and sufficient to strongly stimulate An. gambiae to oviposit (Huang et al. 2005). A new and removable clock face was fashioned from a circular piece of thin plastic, from which had been cut a section equivalent to one h on the clock (Figure 2.1). The opening and perimeter of the face-plate were lined with Parafilm ® (Pechiney Plastic Packaging, Menasha, WI, U.S.A.) flaps to prevent mosquitoes from depositing eggs on any unexposed section of the substrate. The clock face was mounted on the hour-hand driver, so that the opening in the face plate made one revolution every 12 h. The clock face was covered with white paper so as to maximize contrast between background vs. the actual ovipositional site (Huang et al. 2005) Prior to insertion into the clock apparatus, the paper towel substrate was divided by pencil marks into 12 equal wedge-shaped sections. After being exposed to gravid female mosquitoes for 12 h, a clock was removed from the cage of mosquitoes and another was immediately inserted so as to extend the study over a full 24 h. The face of an exposed clock was carefully removed and the paper toweling bearing eggs was carefully peeled off the sand for egg counting under a dissecting microscope. Clock sections open at the beginning and end of a given measurement were exposed to females for a total of 1 h. However, it took 1 h for each intervening section to fully open and another 1 h for each to fully close. Thus, some of the eggs on each intervening section could have been laid over a span of 2 h. Experimental Series 1 - Automated measurement of caged mosquito groups Clocks were presented in white BugDorm-2 insect rearing cages (Mega View Science Education Services Co., Taiwan) measuring 60 x 60 x 60 cm and containing approximately 500 laboratory-reared females of the Kisumu strain varying in reproductive stages. The light 19 cycle in the environmental chamber was set at 12:12 LD, to approximate the natural light cycle found in Kisumu, Kenya. A small tungsten bulb continued to burn in the laboratory at night so as to provide the equivalent light from the night sky. Light levels during scotophase -3 -2 were slightly less than full moonlight (10 W m ) (Gibson 1995). Ovipositional clocks were also presented to groups of house-collected gravid females as described above. In this experiment, the BugDorm-2 cages housing approximately 100 females were placed just inside a screened porch of a house in Kisumu, Kenya. Egg recording sessions for both house-collected and laboratory-reared groups were replicated 8 and 6 times respectively, using a different set of females for each test. Each recording session began at 1700 h, one h prior to the onset of scotophase, and continued for 24 h. At 0500 h, a clock apparatus containing fresh paper towelling was exchanged for the loaded clock. The numbers of eggs laid within each hourly period were counted under a dissecting microscope and incorporated into frequency histograms. The proportion of eggs deposited by laboratory-strain An. gambiae at 1800, 2100, 0, 0400, and 0500 h were arcsine-square-root transformed prior to analysis of variance. Tukey‟s test of significant differences (HSD) was used to separate means. Experiment 2 - Automated measurement of caged individuals The bottom of the enclosure for these tests was the clock apparatus over which sat a 12 cm high cylindrical wire frame. Nylon netting (18 intersections/cm) was placed over the frame and secured by a drawstring. Six 2 cm diameter wet cotton balls (Kendall, Mansfield, MA) were placed on the roof of the cage as a source of moisture. Blood meals were offered to females between 1200 and 1700 h three to four days prior to use. Three or four days after a blood meal, an individual female was gently transferred to the clock cage by aspirator before scotophase. After the female had been exposed to the ovipositional resource for 4-5 h, 20 the female was removed from the cage and fresh paper toweling was substituted for the previously exposed paper toweling within the clock. Then, the female was carefully reinserted. The exchange of the ovipositional resource was repeated 11 h later. The numbers of eggs laid within each hourly period were counted and incorporated into a histogram. Correlation analysis (SAS software version 9.1) was used to test for a correlation between the length of the preoviposition interval (defined as the time interval between a female‟s first exposure to the clock and the time oviposition was initiated) and the length of the oviposition interval (time interval during which oviposition occurred). It was also used to test for a correlation between the length of the preoviposition interval and the total number of eggs deposited per female. A Levene‟s test, as modified by Brown and Forsythe (1974), was used to compare the variances of the preoviposition and oviposition intervals. The ovipositional periodicity was measured for a total of 56 individual females, all of the Kisumu laboratory strain. The terms gravid and hypergravid as used by Sumba et al. (2004) refer to the condition of the female mosquito when they are presented with an ovipositional resource three and four days, respectively, after obtaining a blood meal. Differences in oviposition by gravid vs. hypergravid females were examined by comparing the mean numbers of eggs oviposited per female per h of each respective group using a paired t-test (SAS software version 9.1). After each trial, females were dissected under a dissecting microscope to check for residual eggs. Experiment 3 - Assessment of capacity for mid-afternoon oviposition Engorged females were randomly selected from newly blood-fed cages of mosquitoes and placed in groups of 20 into 8 cages made from 15 cm high and 19 cm in diameter white cardboard cartons. The top of the cage was covered with white netting (8 intersections/cm) 21 and a sleeve of the same netting was fitted to a 10.5 cm hole cut in the side for mosquito and ovipositional resource insertion. Females were provided with a constant source of 10% honey solution and six wet cotton balls (Kendall, Mansfield, MA) were placed on the top of the cage to provide extra moisture. Two days after blood-feeding, an ovipositional resource was provided to half of the cages approximately 2 h before the lights were turned off to record egg deposition during scotophase. The ovipositional resource was a 100 x 35 mm clear plastic Petri dish containing 20 mL of distilled water, placed over a circular piece of black paper. At 1200 h the following day, the loaded ovipositional resources were replaced with new Petri dishes containing fresh filtered water. Four ovipositional resources were also introduced into the 4 cages from which an ovipositional resource had been withheld. These resources, identical to those previously mentioned, were used to record oviposition by gravid females during photophase. After the initial introduction, ovipositional resources were changed hourly from 1200 -1600 h in the latter half of the cages and all exposed ovipositional resources were examined for the presence of eggs. Using a small brush, eggs present were brushed into lines on a piece of white paper and counted. Results Experimental Series 1 – Bimodal oviposition by caged groups An. gambiae of the Kisumu laboratory strain revealed two ovipositional pulses (Figure 2.2). The first occurred from 1800 to 0 h, peaked at 2100 to 2200 h, and accounted for 65% of the total eggs deposited. A second but smaller pulse occurred between 0 and 1000 h, and peaked at 400 h. It is notable that some oviposition by females in groups occurred throughout scotophase. Moreover, a few eggs were deposited in the early hours of photophase. The proportion of eggs deposited by the laboratory strain during the 22 25 a Percent Eggs 20 15 a 10 b b 5 b 0 14 16 18 20 22 0 2 4 6 8 10 12 Clock Interval Figure 2.2. Ovipositional periodicity of laboratory strain Anopheles gambiae groups. Mean percent of eggs oviposited per hour by a caged laboratory strain group (500 females per replicate; total eggs = 18, 303). Bars bearing a common letter are not significantly different (Tukey‟s HSD, α = 0.05). An. gambiae exposure to a 12:12 LD photoperiod during testing is represented by the alternating light and dark horizontal bar above the figure. 23 first peak (2100 h) was significantly greater than the proportion of eggs deposited at 0 h when oviposition greatly diminished (p < 0.0001; Tukey‟s HSD test). The proportion of eggs deposited during the second peak was not significantly different from the proportion deposited during the first peak at 2100 h (p = 0.3). The valley between these two peaks was marginally significant (p = 0.054). Two discrete pulses of oviposition were recorded for house-collected An. gambiae groups (Figure 2.3). The first began at dusk, peaked at 1900 h, and ceased after 0100 h. Seventy-four percent of the total eggs were laid between 1800 and 0200 h. The second pulse commenced near dawn, peaked around 0800 h, and ceased before 1300 h. Unlike the laboratory strain, wild-caught females deposited a substantial portion (more than 25%) of their eggs after sunrise. Experiment 2 – Caged individuals oviposit in one continuous bout Ovipositional periodicity of gravid and hypergravid females held individually was similar, although more gravid females contributed to the second ovipositional pulse than did hypergravids. A paired t-test of the total mean number of eggs oviposited per An. gambiae female per h (1700-1900 and 2100-2200 h) revealed no significant difference between gravid and hypergravid states (p = 0.61). However, correlation analysis revealed a significant positive correlation between the length of the preoviposition interval and the total number of eggs deposited (p = 0.03). The time at which individual females initiated oviposition was highly variable (Figure 2.4). The mean length of the preoviposition interval was 3.5 h with a variance of 9.5, and the mean length of the oviposition interval was 2.5 h with a variance of 1.0. According to a modified Levene‟s test, the length of the preoviposition interval was more variable than the length of the oviposition interval (p < 0.0001). However, there was no correlation between 24 45 600 40 500 Percent Eggs 30 400 25 300 20 15 200 Light Intensity (lux) 35 10 100 5 0 0 14 16 18 20 22 0 2 4 6 8 10 12 Clock Interval Figure 2.3. Ovipositional periodicity of house-collected Anopheles gambiae groups. Mean percent of eggs oviposited per hr by caged housecollected groups (100 females per replicate; total eggs = 11,007). Measurements of outdoor light intensity, represented in the graph as a dark blue line, were taken 5/11/2004 in Kisumu, Kenya. 25 Figure 2.4. Ovipositional patterns of individual Anopheles gambiae over 24 hrs. Each horizontal cluster of rectangles represents a single individual. Shading classifies the number of eggs deposited per individual per hr (n = 56) during the oviposition interval. Gray shading represents the preoviposition interval. 26 20 18 16 Percent Eggs 14 12 10 8 6 4 2 0 14 16 18 20 22 0 2 4 6 8 10 12 Clock Interval Figure 2.5. Accumulated ovipositional patterns of individual Anopheles gambiae over 24 hrs. Percent of eggs oviposited per hr by caged individual An. gambiae (total eggs = 4,815). 27 the length of the preoviposition interval and the length of the oviposition interval (p = 0.51). Compiled individual oviposition was similar to patterns of groups (Figure 2.5); egg deposition occurred throughout scotophase and even during certain hours of photophase. Seven percent of individuals commenced egg deposition before lights off and one individual initiated oviposition after lights on. Interestingly, Figure 2.4 documents that most females oviposited without detectable interruption and those females spread their eggs continuously over a few consecutive clock intervals. Only one individual out of 56 exhibited two ovipositional pulses; she commenced at 1900 h, ovipositing 12 eggs, and then paused until 2300 h before depositing another 105 eggs. Experiment 3 – An. gambiae are capable of mid afternoon oviposition Seven out of the 12 cages from which an ovipositional resource had been withheld until mid afternoon produced eggs (Table 2.1). All of the 12 cages provided with an evening ovipositional resource produced eggs. Cages provided only with a mid afternoon ovipositional resource produced 641 total eggs, which is equivalent to 6% of the eggs produced by cages provided an evening ovipositional resource. Eggs from cages provided only a mid afternoon ovipositional resource were spread over the entire 4 h period. Sixtyseven percent were deposited in the first 2 h, and approximately 24% were deposited between 1400 and 1500 h. The remaining 9% were deposited in the last h. Two hundred and three eggs were oviposited between 1200 and 1600 h in cages previously exposed to an ovipositional resource the evening prior. Discussion An. gambiae deposits eggs in two ovipositional pulses per 24 h. Both laboratorystrain individuals and house-collected groups of An. gambiae produced a large pulse of eggs 28 Table 2.1. Mid afternoon egg output by Anopheles gambiae as influenced by previous access to an ovipositional resource. Treatment Total eggs per cage per period With ovipositional resource the previous evening Cage No. Individuals / cage 1 14 2 18 3 17 4 16 5 20 6 20 7 20 8 20 9 19 10 20 11 20 12 20 Total 224 1700 - 1200 h 252 363 77 512 790 818 567 872 1888 1855 1251 1793 11038 29 0 0 0 1 10 38 154 0 0 0 0 0 203 - Without ovipositional resource the previous evening Cage No. Individuals / cage 1 14 2 18 3 15 4 18 5 20 6 20 7 20 8 20 9 18 10 20 11 18 12 20 Total 221 1200 - 1600 h 3 9 0 161 0 128 0 51 6 184 99 0 641 commencing at scotophase and peaking 1-2 h later. These results agree with those of Haddow and Ssenkubuge (1962) and Sumba et al. (2004). In our work with both the laboratory strain and house-collected strain, we also observed a second smaller ovipositional pulse a few h after the first pulse. The second pulse by laboratory strain groups occurred between 0 and 1100 h, while this second pulse occurred between 0500 and 1300 h for the house-collected strain. For individuals, the onset of the second pulse occurred earlier than its occurrence in the group tests (Figure 2.5). Most eggs were deposited between 2300 and 0 h. Between the first and second ovipositional pulses in all groups, egg deposition sharply declined. While both laboratory and house-collected strains decreased ovipositional activity at 0 h, laboratory strain egg deposition resumed at 0100 h, whereas oviposition by housecollected females remained sparse until 0600 h. We speculate that the significant midnight decline in egg deposition may be the result of an endogenous rhythm. The length of the quiescent period between pulses may be a direct result of exposure to environmental conditions, such as early morning low temperatures probably experienced by house-collected, but not laboratory strain females during these tests. Jones and Gubbins (1978) reported that peak flight by An. gambiae occurs immediately after lights off and that a second smaller peak in activity occurs between 6 and 10 h later. This suggests that flight activity is regulated by a circadian rhythm that could secondarily influence ovipositional patterns (Jones and Gubbins 1978, Jones et al. 1972). Increasing flight activity during the onset of scotophase would increase the probability that a female encounters a suitable ovipositional resource. These peak flight times described by Jones et al. (1972) and Jones and Gubbins (1978) may contribute to the dusk and early morning peaks in oviposition that we have recorded. Our research established that individual females rarely split their eggs over two distinct time periods but rather lay eggs steadily after oviposition begins. We conclude that 30 the two pulses in oviposition by groups are not the result of individual females spreading their eggs over two distinct time periods. Instead, some individual females delay the onset of oviposition to create the second peak. There was much greater variability in the preoviposition interval (i.e. the time interval prior to when a female initiated oviposition), than there was in the amount of time devoted to oviposition. In the case of the single female who split her eggs between two ovipositional periods, an interruption caused by the exchange of the paper toweling could explain this single aberration. While the rates of egg deposition by gravid and hypergravid females were not found to be different, a statistically significant positive correlation existed between the length of the preoviposition interval and the total number of eggs deposited per female. Individuals with longer preoviposition intervals tended to deposit slightly more eggs. However, this correlation likely has little biological significance due to the considerable scatter in the data. This is demonstrated by the width of the 95% confidence intervals (CI) surrounding the mean total numbers of eggs per preoviposition interval. The width of the 95% CI surrounding mean egg deposition for individuals with a 14 h preovipositional interval was 64 (actual 95% CI: [90, 154]) eggs, vs. 20 (actual 95% CI:[80, 100]) eggs when the preoviposition interval was 4 h. An. gambiae has the capacity for afternoon oviposition in full light. Females denied an ovipositional resource for 18 h oviposited between 1200 and 1600 h, when the ovipositional resource was introduced. In some cases, eggs were found on the ovipositional resources between 1200 and 1600 h even when the mosquitoes had a resource beginning at 1700 h on the previous night. Visual contrast of the ovipositional substrate is an important stimulus for oviposition and egg placement. Huang et al. (2007) reported that a black ovipositional dish on a white or grey floor received many more eggs than any other white-black or grey-black combination of 31 ovipositional substrate and background. Clay soil in Kisumu, Kenya, appears black when wet and grey when dry, and discrimination between grey and black coloration improves at -3 -2 light levels of 2.1 x 10 W m , which is equivalent to late dusk or early dawn (Huang et al. 2007). When ovipositional resources are sparse, it may benefit An. gambiae to forage for ovipositional sites before full darkness and at or after dawn, when visual contrast would be more detectable. Our overall results establish that oviposition by An. gambiae is not restricted only to one specific time of day, and oviposition is not fully inhibited by high light levels. Gravid females can initiate oviposition as soon as an ovipositional resource becomes available. Thus, ovitraps, a tool to monitor An. gambiae population growth and help predict malaria epidemics, should remain available throughout the full 24 h diel to be maximally effective. Further study of abiotic factors such as daily temperature and relative humidity fluctuations and their contribution to patterns in flight activity in the field may be of interest. High temperature and low RH may limit mid-afternoon oviposition in the field. During the daytime in the tropics, air and soil temperatures typically exceed the optimum temperature for oviposition - > 25 ºC (Huang et al. unpublished data). In conclusion, An. gambiae populations are ovipositionally flexible. Rather than confining oviposition to a specific brief period during 24 h, as is true for many insects, An. gambiae can oviposit at any time after their eggs have fully developed and they have access to an ovipositional resource. But, they most commonly begin oviposition and deposit the majority of eggs shortly after dusk. Once oviposition commences, individual females deposit their eggs over a continuous 2 to 3 h period without interruption. 32 CHAPTER 3: EGG PLACEMENT BY ANOPHELES GAMBIAE (DIPTERA: CULICIDAE) AS INFLUENCED BY RESTING SURFACE SLOPE AND ENCLOSURE Abstract In these experiments, laboratory-reared Anopheles gambiae s.s. were as likely to rain eggs down from a vertical perch, as to oviposit while sitting horizontally on a substrate of moist mud or open water. Tall cylinders with moist, dark walls provided an enclosed vertical resting surface from which An. gambiae deposited eggs. Similar numbers of eggs were deposited in these cylinders as on dark and moist horizontally-positioned ovipositional substrates. Likewise, An. gambiae oviposited equally from a vertical perch among emergent surrogate reeds vs. sitting horizontally on mud. Open ovipositional resources with exposed resting surfaces presenting 45⁰, 90⁰, and 135⁰ angles relative to the cage floor received fewer eggs than a horizontally positioned, moist and dark ovipositional dish, even though more An. gambiae females settled on these angled resting surfaces post-oviposition. We conclude that vertical resting sites are adequate perches from which oviposition can occur if they provide high humidity and the paramount dark and wet ovipositional cues for An. gambiae oviposition. Introduction Many insects accommodate the physical structure and dimensionality of their surrounding environment during ovipositional site selection (Harris and Miller 1982, 1984, Prokopy and Owen 1983, Roessingh and Städler 1990). Mosquito breeding habitats commonly occur at the interface of aquatic and terrestrial environments, and may have 3dimensional structure. Visual/physical cues arising from the structure of the surrounding 33 environment may play an important role in ovipositional site selection for some mosquito species (Overgaard 2007). By definition, egg deposition by Anopheles gambiae s.s. Giles determines the distribution of larvae of this important malaria vector species in nature, which in turn influences resultant production of the adults (Mutuku et al. 2006b). An. gambiae was once thought to deposit most of their eggs in small sunlit pools of rainwater devoid of vegetation, in close proximity to human dwellings (Holstein 1954). However, it is clear that the ovipositional plasticity of An. gambiae extends egg deposition beyond these habitats (Muirhead Thomson 1951, Holstein 1954, Gillies and De Meillon 1968, Huang et al. 2006; Miller et al. 2007). Indeed, An. gambiae larvae have been discovered in such habitats as inundated agricultural fields (Minakawa et al. 1999, Munga et al. 2005), swamp margins (Minakawa et al. 2004), drainage ditches (Mutuku et al. 2006), and tree holes (Omlin et al. 2007). In fact, a cohort of offspring from a single An. gambiae female may even be distributed among multiple mosquito breeding habitats (Chen et al. 2006), suggesting that ovipositing females use multiple sites during an ovipositional cycle. In the laboratory, females readily oviposit within grassy vegetation (Huang et al. 2006a), on sand that is dry to the touch as perceived by a human (Huang et al. 2005), and on agarose media (Huang et al. 2006b). The most important determinants of An. gambiae oviposition described to date are darkness and wetness (McCrae 1984, Huang et al. 2005, Huang et al. 2007). Dark substrate juxtaposed to a lighter background provides visual contrast and promotes An. gambiae oviposition (McCrae 1984, Huang et al. 2007), particularly during peak ovipositional times of dusk and dawn (Sumba et al. 2004, Fritz et al. 2008) when ambient light levels are low. Indeed, dark mud appears to receive the preponderance of An. gambiae eggs (Miller et al. 2007). Larvae eclosing there can crawl into puddles or be carried there during rains. 34 Ovipositional posture has rarely been investigated in An. gambiae, but appears to be flexible. McCrae (1984) reported that coastal Kenyan An. gambiae deprived of a muddy border around a pool of water oviposited while in flight. However, a full third of his test population oviposited from a settled posture onto ovipositional resources. More recent studies (Bray 2003, Huang et al. 2006, Miller unpublished) suggest that oviposition during flight is less common than previously reported, even in wild-caught populations. Upon detection of the ovipositional site, laboratory-reared anophelines, including An. gambiae, land and oviposit on the water surface (Bates 1940, McCrae 1984) or more commonly on mud (Miller et al. 2007). Other anophelines perch vertically on emergent vegetation and drop eggs onto the water below (Giglioli 1965, Clements 2000). Based upon previous reports of their ovipositional flexibility, we postulated that An. gambiae may also exhibit this behavior. Here we extended the analysis of An. gambiae ovipositional site preference to 3 dimensions, and attempt to characterize An. gambiae egg deposition patterns as influenced by the 3dimensional objects to which they are exposed. If such behavior were to be found, such data would suggest that some larval habitats of this important malaria vector have been overlooked. Methods Mosquitoes and bioassay conditions Kisumu strain An. gambiae, reared according to Huang et al. (2005), was used for high and low humidity ovipositional tests conducted in a previously described mosquito rearing facility at Michigan State University (MSU) (Huang et al. 2005). Some low humidity ovipositional tests for Experiments 1 and 3 were conducted at the Entomology laboratory of the Kenyan Medical Research Institute (KEMRI), in Kisian, Kenya, using the KEMRI Kisumu strain An. gambiae colony. 35 Forty 5 d old females were blood-fed via artificial membrane feeder (Huang et al. 2005) 3 d prior to use in ovipositional tests. Once gravid, females were inserted into 60 × 60 × 60 cm white BugDorm-2 dome insect rearing cages (MegaView Science Education Services Co., Taiwan) via mouth aspirator. All tests were performed under a 12L:12D h photoperiod. At KEMRI, low light was provided during scotophase by outdoor security lights shining through the window, while a small tungsten night-light (4 lux) provided low light in the colony room at MSU. For all tests, ambient temperature was ca. 25ºC. The RH for low humidity tests was 35% (±10). A Kenmore Whole House Humidifier (Model 758.154120; Sears, Roebuck and Co. Hoffman Estates, IL), as well as damp white kitchen towels draped over the cages walls increased the RH for high humidity tests to 80% (±10). Mosquitoes were provided access to a 10% honey solution ad libitum throughout tests. Experiment 1- Exposed resting surface slope preference In this choice test, ovipositional resources were composed of a horizontal and a vertical component (Figure 3.1). Three of four ovipositional resources contained vertical components sloped at varying angles relative to the floor. Each vertical component was made using a 13 × 9 × 2 cm plastic dish (The Glad Products Co., Oakland, CA, U.S.A.). Three layers of well-water-moistened Spontex® sponge cloth (Supa Brite, Nairobi, Kenya) were cut to fit inside each rectangular dish, then stapled in place. These were covered with black paper printed from an HP LaserJet 4100 PCL 5 (Hewlett-Packard Co., Palo Alto, CA, U.S.A.), and then moistened to stimulate oviposition (Huang et al. 2005). Rectangular dishes were hot-glued to the horizontal component of the ovipositional resource, the bottoms of dry 100mm diam × 35mm clear plastic Petri dishes, at 45º, 90º, and 135º angles to the floor. A fourth ovipositional resource consisted of a 2cm deep rectangular dish, not secured to a Petri dish base, placed horizontally on the floor. Additionally, the Petri dish base served as a 36 A a Percent eggs 100 80 60 40 20 b b b 0 45 90 135 180 Mean # resting An.gambiae per cage B 4 a 3 ab 2 b b 1 0 1 2 3 4 Figure 3.1. A. Mean percentage of Anopheles gambiae eggs (± 1 SEM) collected from ovipositional resources with exposed resting surfaces at varying inclines with respect to the cage floor (total eggs = 17,134; mean no. eggs/female = 37). B. Mean number of resting mosquitoes (± 1 SEM) counted on ovipositional resources with exposed resting surfaces at varying inclines with respect to the cage floor. Bars within a graph sharing the same letter are not significantly different (p>0.05). 37 moisture reservoir; water was added to the Petri dish bottoms at the start of each trial to replace evaporating moisture from paper and sponges of the vertical component. To discourage egg deposition in the water-filled Petri dish bases, we covered the entire base with one layer of aluminum foil, and then a layer of dry, black paper. The cage floor was covered with dry white paper, and divided into 4 quadrants of equal size; the center of each received a different ovipositional resource. Experiment 1 was replicated with 6 different groups of 40 gravid An. gambiae at both high and low humidity, and ovipositional resources were rotated to a new position in the cage for the testing of each new group. Ovipositional resources were presented to females 1 h prior to the onset of scotophase, and were removed ca. 20 h later. For each cage, oviposition was monitored over two consecutive nights; loaded ovipositional resources were replaced with fresh ones between consecutive nights. Both the horizontal and vertical components of each ovipositional resource were examined under a dissecting microscope daily for the presence eggs. Egg totals from each cage over the two nights were summed for respective treatments. Experiment 2 – Resting surface counts The post-ovipositional sitting preferences of mosquitoes from six cages used in Experiment 1 were quantified. After two nights of testing, but before ovipositional resources were collected, the total numbers of mosquitoes resting on ovipositional resources in these cages were counted. Mosquitoes were dislodged from their resting places by waving a human hand near each ovipositional resource, the walls, and the cage floor. After 15 min, resting counts resumed, and the process was repeated. Three resting counts were taken for each cage of mosquitoes, and mean values calculated. 38 Experiment 3 - Enclosed resting surface slope preference Two vertically positioned cylinders of differing heights were designed to provide an enclosed resting surface around the perimeter of an ovipositional resource. Dry, black paper in a Petri dish bottom, then covered with Parafilm deterred females from resting on the floor of the ovipositional resource while depositing eggs. The walls of the cylinders were made from moistened Spontex ® sponge cloth cut into 20 × 2.7 cm and 20 × 10 cm pieces, then rolled into tubes. These sponge tubes of equal diameter, but differing in height, fit inside the rim of 100mm diam × 35mm high clear plastic Petri dish bases. The outside surfaces of the cylinders were lined with aluminum foil, then white paper, while the insides were lined with black paper, to create a dark, moist enclosed vertical surface. A Petri dish bottom was overlayed with three moist circular pieces of sponge cloth and covered with a circular piece of black paper, so as to provide a horizontal surface on which ovipositing females could rest. The three cylinder treatments were spaced equidistantly from each other on a white paper-covered floor. In tests conducted under both high and low humidity, ovipositional resources were rotated to a new position within the cage for each of 8 replicate groups of An. gambiae. Ovipositional resources were presented to gravid females 1 h prior to the onset of scotophase, and were removed from cages ca. 20 h later. Tests were conducted over two consecutive nights, and each night, fresh ovipositional resources replaced loaded ones. The sum of the eggs found on each ovipositional resource over two consecutive nights was recorded. Experiment 4 – Presence/absence of surrogate vegetation Two rectangular opaque plastic dishes measuring 12 × 10 × 3 cm were used as ovipositional resources. In the center of one dish, 1.4 × 17 mm wire nails were hot-glued vertically, but with the head down within a 10 × 2.5 cm strip in 3 staggered rows. The 39 bottom of the dish was covered with a layer of Michigan soil, followed by a layer of brown paper towel. Glass tubes (12 cm high and 0.3 cm diam) rounded and sealed at one end, and open at the other were painted green and coated with wax (Harris 1987) to make removable surrogate reeds. These surrogate reeds slipped over the nails during oviposition tests, but could be removed and inspected for eggs during data collection. In the bottom of the other rectangular dish, the soil was layered so as to create a shallowly sloped surface leading to a 10 × 2.5cm strip of flat soil substrate. A layer of brown paper towel covered the soil for ease of egg collection. Both rectangular dishes were placed equidistantly from the walls of a 37 × 37 × 29 cm clear Plexiglas cage, and then filled to the brim with filtered well water immediately prior to initiation of oviposition tests. Five gravid An. gambiae were inserted into a cage via mouth aspirator ca. 15 min prior to the onset of scotophase. Tests ran for 20 h, at which time dishes were carefully removed from the cages so as not to spill any water. Water from each ovipositional resource was collected with a disposable plastic Pasteur pipette and filtered through white paper. Eggs from filter paper, as well as any on the brown paper towel were counted and recorded. Surrogate reeds were also removed from the ovipositional resource, and each reed was washed with water from a squirt bottle into a plastic cup. This water was also poured through the filter funnel to ensure that all eggs sticking to the surrogate reeds were collected. Tests were replicated with eight new groups of mosquitoes. Visual observations of mosquito ovipositional behavior were made in four of the replicates. An observer monitored activity in the arena for two h after insertion of the mosquitoes. An Energizer® 6 LED Headlight (Eveready Battery Company, Inc. St. Louis, MO, USA) was set to red night vision and worn by the observer to improve visibility. Data were collected on 1) where oviposition began, 2) ovipositional posture assumed, and 3) 40 whether ovipositing mosquitoes switched ovipositional resources for all individuals beginning oviposition 2 h post-insertion. Data collection and analysis All analyses were performed using R version 2.7.2 (R Development Core Team, Vienna, Austria). For experiments 1 and 3, a likelihood ratio test was used to determine whether ambient humidity influenced egg deposition. Data from Experiments 1 and 3 (Table 3.1) were arc sine square root transformed, and analyzed using ANOVA. Means were separated by Tukey‟s honestly significant difference (HSD) test. Data from Experiment 2 were fitted to the Poisson distribution using Akaike‟s Information Criterion (AIC), and a likelihood ratio test determined the effects of surface angle on An. gambiae resting behavior. Bootstrapped 95% confidence intervals (N=1000) were compared across treatments for separation of means. A paired t-test was used to analyze egg deposition from Experiment 4. Results In Experiment 1, An. gambiae deposited most of their eggs in association with dark, wet horizontal surface of the rectangular dish, and few eggs when resting on any exposed vertical component of an ovipositional resource (df = 3, F = 217.46, p <0.001; Figure 3.1a). Humidity did not influence egg deposition (df = 1, F = 0.35, p = 0.55). In Experiment 2, more females rested on the vertical components of ovipositional resources presenting angles of 90⁰ and 135⁰ relative to the cage floor (df = 1, χ2 = 8.04, p = 0.005; Figure 3.1b). Room RH slightly influenced the site of greatest egg deposition in Experiment 3, resulting in a significant interaction between RH and ovipositional resource (p = 0.01; Table 3.1). At low ambient RH, the greatest proportion of eggs was deposited in the tall cylindrical ovipositional resource (p = 0.0001; Figure 3.2). However, equal numbers of eggs were deposited in the 41 Table 3.1. Analysis of variance for Experiment 3. An asterisk denotes a significant p-value. Source Treatment Residuals Humidity MS F value Pr (>F) 3.05 1.52 28.94 1.98 e-08 * 1 Humidity SS 2 Treatment df 0.02 0.02 0.32 0.57 2 0.54 0.27 5.15 0.01 * 39 2.05 0.05 42 Percent eggs SEM 100 A 80 60 40 a a B b 20 C 0 Horizontal Short Tall Figure 3.2. Mean percentage of An. gambiae eggs (± 1 SEM) deposited on ovipositional resources enclosed by no walls, 2.7cm or 10cm high walls (total eggs = 16,813; mean no. eggs/female = 47). Dark and light bars indicate high and low RH respectively. Mean separation was applied within each RH, represented by upper vs. lower case letters. Bars within an RH treatment sharing the same letter are not significantly different (p>0.05). 43 Figure 3.3. Map of egg deposition within the tall vertical cylinder adapted from photographed results of the first replicate of Experiment 3. Black dots represent individual eggs, while outer and inner black circles represent the Petri dish and sponge walls respectively. 44 horizontal and tall cylindrical ovipositional resources when ambient RH was high (p = 0.65). Most eggs deposited in the tall cylindrical treatment were found within 1 cm from the cylinder wall (Figure 3.3). At both high and low RH, the short vertical cylinder received the fewest eggs. In Experiment 4, egg counts did not differ between the ovipositional resource containing emergent surrogate reeds vs. the flat moist surface (df = 7, t = -0.19, p = 0.86; mean no. eggs/female = 60). The mean number of eggs per ovipositional resource was 142 and 155, respectively. Three individuals were observed ovipositing in two of the replicates that were observed. All began oviposition within the surrogate reeds while perched on a stem, and dropped eggs onto the surface of the water below. One individual hopped from stem to stem between ovipositional bouts, but none switched to the soil treatment. No individuals were observed dropping eggs while hovering over either treatment. Discussion An. gambiae is ovipositionally flexible, and uses varied ovipositional sites. Previously, An. gambiae was reported to oviposit in flight while hovering over an ovipositional resource (McCrae 1984). Yet egg deposition from flight has neither been quantified, nor has it been confirmed that hovering females always drop eggs. Subsequent studies suggest this behavior is rare in the laboratory (Bray 2003). There, females oviposit from a sitting position on moist mud (Huang et al. 2006a, Miller et al. 2007), and can hover over an ovipositional resource without depositing eggs (Miller et al. unpublished). The An. gambiae used in the present study were tested for their propensity to oviposit while in flight. In a Plexiglas cage measuring 128 × 63 × 50 cm, females deposited eggs while sitting on a 15 45 cm diam Petri dish containing moist mud twice as often as they deposited eggs while hovering (Huang et al. unpublished). Perhaps the discrepancy between these results and those reported by McCrae (1984) can be explained by the use of different An. gambiae populations; McCrae used An. gambiae s.l., while An. gambiae s.s. was used in the present study. A more likely explanation is that the Petri dishes used by McCrae did not provide a suitable resting surface onto which An. gambiae could settle during an ovipositional bout. Furthermore, egg deposition patterns within the tall vertical cylinder (Figure 3.3) revealed the location from which An. gambiae oviposited. If eggs were deposited in flight, the distribution of eggs within the tall vertical cylinder should have been random; it would have been equally or more likely that eggs occurred in the center of the cylinder floor than along the cylinder wall. Instead, greater than 95% of the eggs deposited within the tall vertical cylinder were found within 1 cm from the wall (Figure 3.3). These results demonstrate that An. gambiae females do oviposit from a settled posture, even while sitting on a vertical resting surface over an ovipositional resource. In Experiment 1, 8-fold more eggs were deposited on the horizontal ovipositional resource than on any other resource. Yet in Experiment 2, this same resource was least likely to be accepted by An. gambiae as a resting site. Furthermore, An. gambiae was equally or more likely to oviposit from a vertical perch in an enclosure, vs. while sitting on a horizontal resting surface in Experiment 3. Our interpretation of the broader distribution of eggs in Experiment 3 relative to Experiment 1 is that the tall cylinder received eggs that would have been deposited in the horizontal treatment if the tall cylinder had not been present. These data, in conjunction with An. gambiae resting site preferences found in Experiment 2 suggest that egg deposition can be influenced by resting site. An. gambiae is more likely to accept a vertical slope vs. a horizontal surface for resting. If such a resting site contains the critical 46 ovipositional cues of darkness and wetness, females may simply deposit their eggs from their vertical resting position. We speculate that higher humidity levels, resulting from a greater wet surface area, and darker conditions within the tall cylinder enhanced An. gambiae willingness to oviposit while resting inside the cylinder. In the field, vegetation can present vertical resting sites near dark and moist ovipositional sites. Laboratory-reared females used in Experiment 4 equally accepted surrogate emergent vegetation vs. flat, bare soil, thus confirming the results of Huang et al. (2006a) who used natural vegetation. In the current study, An. gambiae was more likely to oviposit on vegetation over bare soil than the An. gambiae used by Huang et al. (2006a). Such differences could be attributed to: 1) the larger stem diameter of our surrogate vegetation providing a more stable perch from which An. gambiae females could oviposit, or 2) the sparse surrogate vegetation density providing greater access to resting sites for gravid females. We speculate that the willingness to oviposit while resting on a vertical surface may be correlated with probability of ovipositing in vegetation. Like An. melas and An. funestus, feral An. gambiae may also perch on vegetation and release eggs onto water and damp soil below. Although most larval production occurs in small sunlit pools of open water, even putatively unproductive larval habitats can enhance malaria transmission (Le Menach et al. 2005). Understanding the potential suitability of different aquatic habitats for An. gambiae oviposition would improve vector control efforts. Our results suggest that larval sampling and control efforts for this vector of important human diseases should be broadened to include pools containing, or surrounded by emergent vegetation, or other vertical resting surfaces. 47 PART II: STUDIES ON FEASIBILITY OF AVERMECTIN-TREATED CATTLE AS A VECTOR CONTROL MEASURE IN SUB-SAHARAN AFRICA 48 CHAPTER 4: AVERMECTINS AND MOSQUITO CONTROL Chemical and pharmacological properties of the avermectins The avermectins and milbemycins, otherwise known as the macrocyclic lactones, belong to a group of related chemicals derived from Streptomyces microorganisms (Martin et al. 2002). This group is named for the base structure of its compounds: they are 16membered lactone rings, onto which are appended different functional groups (Figure 4.1). Avermectins and milbemycins are known for their broad-spectrum endo- and ectoparasitic activity. These lipophilic compounds dissolve easily in most organic solvents and are remarkably safe for use in animals and humans (McKellar and Benchaoui 1996). In veterinary and human medicine, ivermectin and similar compounds are used to manage arthropod pests (Wilson 1993), e.g. bovine ectoparasites, such as Chorioptes bovis, Sarcoptes scabiei var. bovis, Haematobia irritans, and Haematopinus eurysternus as well as to control the tick, Boophilus microplus (Benz et al. 1989). Ivermectin is also used to rid cattle of gastrointestinal nematodes (Benz et al. 1989), which are generally thought to have severe world-wide economic consequences for the cattle industry (Gibbs and Herd 1986). In 1982, ivermectin was first introduced for the treatment of Onchocerca volvolus in humans (Aziz et al. 1982), and is now extensively used for Onchocerciasis control in endemic regions. Ivermectin is also used for the treatment of humans affected by filariasis, scabies, strongyloidiasis, and other gastrointestinal nematodes (Dourmishev et al. 2005). For the treatment of cattle, ivermectin is either applied topically, by subcutaneous injection, intravenously, or in a bolus. In humans, ivermectin is administered as an oral tablet. Bioavailability of the macrocyclic lactones, and thus the efficacy and persistence 49 22, 23 – dihydroavermectin B1a (Ivermectin) Figure 4.1. Chemical structures of ivermectin (Adapted from Edwards 2003). 50 of the drug, varies depending on the formulation used for treatment (Edwards et al. 1988, Fink and Porras 1989, Escudero et al. 1999). For example, the biological half-life of ivermectin is 2.8 d in cattle treated intravenously, but 8.3 d in cattle treated via subcutaneous injection (Fink and Porras 1989). Regardless of treatment formulation, the main route of excretion is in the feces (Chiu and Lu 1989). Some mosquitoes die after taking a bloodmeal containing ivermectin (Tesh and Guzman 1990, Gardner et al. 1993, Bockarie et al. 1999). However, the lethal and sublethal effects on An.gambiae s.l. after blood feeding directly on ivermectin-treated cattle or humans have never been investigated. Alternatives to ivermectin Moxidectin, doramectin, and eprinomectin are all used for the treatment of bovine endo- and ectoparasites. Moxidectin is thought to more effectively control nematodes than arthropods (Mckellar and Benchaoui 1996). It differs from ivermectin by three functional groups: 1) the C-13 disaccharide found in ivermectin is absent in moxidectin, 2) a methoxime moiety is found at C-23 in moxidectin, but absent in ivermectin, and 3) an olefinic side chain is found at C-25 in moxidectin, but is absent from ivermectin (Rock et al. 2002). Moxidectin is more lipophilic than ivermectin. In cattle, the half-lives of moxidectin and ivermectin in fatty tissue are 12-14 d (Rock et al. 2002) and 7-8 d (Chiu and Lu 1989) respectively. This property may give moxidectin a longer mean plasma residence time than ivermectin (Lanusse et al. 1997). However, cattle treated with equivalent dosages of these two drugs do not have significantly different peak plasma concentration values (Lanusse et al. 1997). Doramectin was derived via mutational biosynthesis of avermectin A1, whereas ivermectin is a derivative of avermectin B1 (Goudie et al. 1993). The presence of a cyclohexyl group on C-25 can be used to distinguish doramectin from ivermectin (Conder and Baker 2002). It is highly efficacious against nematode parasites (Jones et al. 1993), and 51 in some cases outperforms ivermectin (Goudie et al. 1993). After cattle are administered subcutaneous injections of either doramectin or ivermectin at 200μl/kg, examination of their blood plasma reveals that doramectin peaks higher and persists for longer than does ivermectin (Lanusse et al. 1997, Toutain et al. 1997). Eprinomectin was discovered in 1997, and is only commercially available as a topical formulation. It was purposefully derived to create a product that was highly effective against livestock gastro-intestinal parasites, yet had low milk partitioning (Shoop and Soll 2002). Like ivermectin, it is derived from avermectin B1, but is distinguished by the presence of an amino group at C-4 (Shoop et al. 1996). Cattle and deer are treated topically with Eprinex (containing eprinomectin) at 500 μg/kg, while other avermectins are typically only applied at 200 μg/kg. Interestingly, eprinomectin is three-fold more effective at reducing endoparasites in cattle than ivermectin, even when the drugs are applied at the same rate. This may be, in part, due to the greater bioavailability of eprinomectin; blood plasma availability of eprinomectin is twice that of ivermectin in cattle (Alvinerie et al. 1999). The modes of action (Arena et al. 1995) and routes of excretion (Mehlhorn 2008) are similar among the avermectins. Unlike ivermectin, moxidectin and eprinomectin do not have meat or milk withdrawal times, and all three ivermectin alternatives are approved for use in pregnant animals. Hosts o f An. gambiae s.l. that cannot be treated with ivermectin, such as pregnant or lactating cattle, could safely be treated with moxidectin, doramectin, or eprinomectin. Mode of action of the avermectins It was previously suggested that avermectins and milbemycins are activators of both GABA-gated and glutamate-gated chloride channels. Glutamate is a stimulatory transmitter found at nerve-muscle junctions (only in invertebrates) (Fox and Lloyd 1999), whereas γ- 52 aminobutyric acid (GABA) is an inhibitory postsynaptic transmitter of the central nervous system (Nation 2002). Fritz et al. (1979) were the first to report that the inhibitory postsynaptic potentials in lobster stretcher muscles perfused with avermectin B1a (AVM) were irreversibly blocked. The inhibitory neurotransmitter at the lobster stretcher muscle neuromuscular junction was believed to be GABA (Fritz et al. 1979). Later, Kass et al. (1980) reported that AVM inhibits transmission between interneurons and excitatory motoneurons in the ventral nerve chord of nematodes; they suggested that AVM acted as a GABA agonist in these animals. Feng et al. recently showed that AVM does interact with nematode GABA receptors (2002), though others report that AVM inhibits rather than activates these receptors (Martin and Pennington 1988). The high concentration of AVM needed to activate GABA receptors (Wolstenholme and Rogers 2005) in nematodes, and the inactivity of AVM on mammalian peripheral GABA receptors also indicate that the GABAnergic receptors (Martin et al. 2002) are not the main target of the avermectins. The macrocyclic lactones act mainly on glutamate-gated chloride channels, which are important in signal transmission to locomotory and pharyngeal muscles in nematodes (Martin et al. 2002, Wolstenholme and Rogers 2005). They cause paralysis and death in target organisms, and can also cause sterility in some arthropods by blocking oviposition (Hollingworth 2006). Safety of ivermectin and moxidectin in vertebrates The macrocyclic lactones are widely used in human and veterinary medicine for the treatment of endo- and ecto-parasites. They powerfully affect glutamate-gated chloride channels, yet this group of ligand-gated Cl channels is not found in vertebrates (Martin et al. 2002). Cattle can tolerate dosages of ivermectin and moxidectin well above the recommended therapeutic doses. The acute toxicity of ivermectin applied orally and subcutaneously is 10 mg/kg. Single concentrated doses of 6.0 mg/kg can be administered to 53 cattle without ill effects (Pulliam and Preston, 1989). Injectable moxidectin has been used safely in cattle at dosages to 1.0 mg/kg (Yazwinski et al. 2006). In humans, therapeutic dosages of ivermectin are well tolerated up to 200 μg/kg (Greene et al. 1989). To date, no other avermectin has been approved for human use. Current vector control methods In sub-Saharan Africa, current vector control strategies include: 1) indoor spraying of insecticides targeting adults, 2) personal protection measures such as ITNs and insecticidetreated curtains, 3) larvicidal compounds such as Bacillus thuringiensis var. israeliensis and 4) environmental manipulation for prevention of successful breeding (Touré 2001). Yet ITNs and IRS continue to be the most inexpensive and effective tools used for community-wide vector control and prevention of malaria epidemics (Rozendaal 1997). ITNs successfully limit survivorship of mosquitoes that contact the net, and can reduce malaria transmission by 90 percent in villages where nets are deployed to cover all sleeping spaces (Gimnig et al. 2003). ITN use can reduce childhood mortality by 17% or more in Sub-Saharan Africa (Lengeler 2004). Mass distribution programs in rural Africa have the potential to significantly increase ITN coverage to even the poorest families. Between 2004 and 2006, ITN use by rural Kenyan children increased from 7 to 67% due to expansion of a subsidized clinical distribution system (Noor et al. 2007). Even unprotected families receive some of the benefit of high mosquito mortality from ITN-protected neighbors; increased malaria protection is experienced in villages where ITN coverage is greater than 50% (Hawley et al. 2003). In Western Kenya, An. gambiae is being extirpated in areas where ITN coverage reaches over 90% (Bayoh et al. 2010). However, malaria transmission still occurs in areas where ITN coverage is high (Walker, personal communication). Although An. gambiae populations are declining, the entomologic inoculation rate remains high enough for malaria 54 infection to recur at least once per person per year (Bayoh, personal communication). Furthermore, ITNs apply selection pressure and push An. arabiensis populations toward exophily, which may lead to behavioral resistance to indoor control measures in these vector populations. Tirados et al. (2006) found that only 16% of An. arabiensis females entering a house at night rested there the following morning, suggesting that females feed indoors but leave immediately after feeding. While indoor vector control measures successfully control endophilic vector populations, a vector control measure targeting exophilic populations is needed. In the 1950s and „60s, over-reliance on a single vector control measure, DDT, led to rapid evolution of insecticide resistance (Nabarro and Tayler 1998). Heavy reliance on pyrethroids used in ITN treatment, and lack of affordable alternatives to ITNs, could potentially lead to pyrethroid resistance. Currently, An. gambiae s.l. populations resistant to both pyrethroids, and DDT used in IRS have been documented in several African countries (Chandre et al. 1999). To improve malaria control, a multi-tactic approach that supplements ITNs and IRS is urgently needed. The following investigations aimed to: 1) Demonstrate that vector populations were feeding upon, and likely persisting due to cattle hosts present in areas of high ITN coverage, 2) Examine the cost of blood feeding upon ivermectin- and moxidectin-treated cattle to An. gambiae s.l. 3) Compare the toxicities of all four cattle dewormers on An. arabiensis in a controlled laboratory setting. 55 CHAPTER 5: TOXICITY OF BLOODMEALS FROM IVERMECTIN-TREATED CATTLE TO ANOPHELES GAMBIAE S.L. Published in: Annals of Tropical Medicine and Parasitology, 2009; 103(6): 539–547 Abstract Two cattle dewormers, ivermectin and moxidectin, were tested for lethal and sublethal effects on the malaria vectors Anopheles gambiae s.s. and An. arabiensis. In the laboratory, direct addition of ivermectin to bovine blood reduced mosquito survivorship and fecundity. The LC50 of ivermectin for An. gambiae s.l. laboratory populations was 19.8 ± 2.8 ppb. In the field, commercially available formulations containing ivermectin and moxidectin were injected into cattle at three times the recommended rate. Ninety percent of An. gambiae s.s. that fed on ivermectin-treated cattle within 2 wk of treatment failed to survive >10 d post blood meal. No eggs were deposited by An. gambiae s.s. fed on ivermectin-treated cattle within 10 d of treatment. Survivorship and egg production of mosquitoes feeding on moxidectin-treated cattle were no different from those feeding on untreated cattle. These results suggest that treatment of cattle with ivermectin could be used as part of an integrated control program to reduce zoophilic vector populations that contribute to malaria transmission. Introduction Malaria continues to burden human populations throughout the tropics, annually killing over 1 million African children. Two mosquitoes of the Anopheles gambiae complex, An. gambiae sensu stricto and An. arabiensis, are widely distributed throughout Sub-Saharan 56 Africa, and along with An. funestus, are the principle vectors of malaria. Although morphologically indistinguishable, the species in the An. gambiae s.l. complex differ in blood host utilization, resulting in variable vectorial capacity. An. gambiae is predominantly anthropophagic, and An. arabiensis feeds primarily on cattle, yet both species will opportunistically feed on either host (Highton et al. 1979, Githeko et al. 1996, Tirados et al. 2006). Where densely deployed, pyrethroid-treated bed nets reduce An. gambiae s.l. populations. However, it is unlikely that malaria will be suppressed to tolerable levels unless the cattle-mosquito connection is simultaneously broken. This is attributable to An. arabiensis; despite feeding primarily upon cattle throughout its broad distribution, this mosquito is still an efficient vector of malaria (Trape et al. 1992, Mendis et al. 2000). Concerns over insecticide (Chandre et al. 1999, Müller et al. 2008, Munhenga et al. 2008, Awolola et al. in press) and behavioral resistance management (Molineaux and Grammiccia 1980), also call for the implementation of novel vector control measures that can augment bed nets and indoor residual sprays. Ivermectin and similar macrocyclic lactones are used in veterinary and human medicine against a range of helminths and arthropod pests (Wilson 1993). It is ingested by humans for the treatment of ectoparasites Sarcoptes scabies and Pediculus humanus, as well as endoparasites such as Strongyloides stercoralis, Wuchereria bancrofti and Onchocerca volvulus. Ivermectin is administered to cattle via subcutaneous injection, topically as a pouron, or as a bolus to treat the ectoparasites Chorioptes bovis, Sarcoptes scabiei var. bovis, Haematobia irritans, Haematopinus eurysternus, and Boophilus microplus (Benz et al. 1989), as well as gastrointestinal nematodes. Some mosquitoes die after taking a blood meal containing ivermectin. All An. stephensi Liston fed on the blood of mice treated with 2.8 mg/kg ivermectin died 57 (Pampiglione et al. 1985). Tesh and Guzman (1990) reported that blood-ivermectin levels greater than 10 μg/ml were lethal to Ae. aegypti, Ae. albopictus, and Cx. quinquefasciatus. Even at sublethal doses, surviving mosquitoes suffered reduced egg production and hatchability (Tesh and Guzman 1990). Seventy percent of blood-fed An. punctulatus collected from human dwellings 4 d after ivermectin treatment of human inhabitants, died 24 h post-bloodmeal (Bockarie et al. 1999). More than 90% of An. farauti fed on the arm of an ivermectin-treated human volunteer died when the feeding occurred within 10 d after ivermectin ingestion (Foley et al. 2000). To our knowledge, the lethal and sublethal effects of ivermectin-treated blood on An. gambiae s.l. have never been investigated. Therefore, in this study we quantified the effects of anthelmintic-treated cattle on the survivorship and fecundity of the important malaria vector An. gambiae s.l. Methods Mosquitoes We used Kisumu strain An. gambiae laboratory colonies: one was reared at Michigan State University (MSU) according to the methodology of Huang et al. (2005), and the other at the Centre for Global Health Research, Kenya Medical Research Institute (KEMRI), Kisumu, Kenya. One experiment included An. arabiensis of the Dongola strain. This colony was provided to MSU in January of 2008 by the Centers for Disease Control and Prevention, Malaria Research and Reference Reagent Resource Center (MR4; American Type Culture Collection, Manassas, VA). These laboratory-reared An. arabiensis were handled identically to An. gambiae at Michigan State University. Eclosed mosquitoes were placed in 30cm3 BugDorm-1 (MegaView Science Co. Ltd., Taiwan) rearing cages and were always held at 28±1⁰C and 80±10% RH under a 12:12 LD photoperiod. 58 Delivery of ivermectin-treated blood via artificial membrane feeder Ivermectin (> 90% B1a; Sigma-Aldrich Corp., St. Louis, MO) was dissolved in dimethyl sulfoxide (DMSO) to make a stock solution of 10 ppm. A 10:1 serial dilution of stock solution in DMSO was used to make 6 treatments (1000, 100, 10, 1, 0.1, and 0.01 ppb). Heparinized bovine blood (Lampire Biological Laboratories, Inc., Pipersville, PA) was added to 1.5 ml of each treatment solution to make 6 treated blood solutions of 15 ml each. The negative control was 1.5 ml of DMSO in 13.5 ml of bovine blood. Treatments of 2, 4, 6 and 8 ppb ivermectin in bovine blood were offered to both An. gambiae and An. arabiensis, so as to refine LC50 estimates. Each treatment was presented to 20 mosquitoes transferred via mouth aspirator to cages made from circular, clear plastic food containers measuring 10 cm in diameter. The center was removed from the lid of each container and replaced with netting material through which mosquitoes could blood feed. The bottom of each container was cut off and replaced with a new, removable cage floor filled with moist sand, and covered in brown paper towel (Georgia Pacific, Atlanta, GA) for oviposition. Treated blood was offered to cages of mosquitoes for 30 min via artificial membrane feeder (Huang et. al 2005). Then, all unfed females were removed. Mosquitoes were monitored daily for mortality and oviposition for 9 d post blood-meal. Dead mosquitoes were counted and removed, and ovipositional substrates were inspected and eggs counted daily. Ten percent honey solution was provided to all mosquitoes ad libitum during the post-treatment period; six 1.5 cm diam cotton balls were soaked in the solution, then placed on top of each cage. Approximately 100 eggs from each loaded ovipositional substrate were separated according to treatment, and submerged in clear plastic 100 x 15 mm petri dishes containing filtered tap H2O. Petri dishes were held on the laboratory bench top at ca. 22⁰C under 12:12 LD photoperiod. First instars were counted and removed daily with a Pasteur pipette. Larval counts ceased when no new first instars were found for 3 consecutive d. 59 Mosquito feeding bioassay on ivermectin- and moxidectin-treated cattle Nine zebu bulls between one and three years old were purchased in Kisumu, Kenya. These bulls were incorporated into a local herd. They grazed daily and were housed as in traditional Luo husbandry: tied from sun-down until 10 am within the open air center of a housing compound. These bulls were weighed on a WB5000 portable scale (Salter Brecknell, Fairmont, MN). They were partitioned into 3 groups of 3 bulls each, so as to block by weight. Groups were randomly assigned to one of three treatments: injectable Ivomec (ivermectin; Merial, Duluth, GA) at 3 mL/50 kg (600 μg/kg AI), injectable Cydectin (moxidectin; Fort Dodge Animal Health, Fort Dodge, IA) at 3 mL/50 kg (600 μg/kg AI), and a 0.9% sterile saline injection at 3 mL/50 kg. Prior to treatment, cages of An. gambiae were fed on each animal to establish a baseline for mortality not due to chemical. Cages consisted of a clear plastic cone with a cotton-plugged port at the narrow end for insertion of 30 mosquitoes, and covered at the wide end with white netting. Prior to blood feeding, a ca. 6 cm diam circular patch was shaved (or reshaved) from the rib cage immediately behind the foreleg of each animal. A loaded cage was snuggly affixed by means of a size 32 brassiere. Cattle were loosely tethered to a tree trunk during the 30 min blood-feeding, and up to 3 bulls at one time were donating blood under the care of one or two human attendants. Starting 1 day post treatment, cohorts of 3-d-old mosquitoes were fed on the treated and untreated cattle every three days for up to 23 d post-treatment. After transport to the KEMRI insectary, cattle-fed mosquitoes within one of the three chemical treatment groups were segregated into two sub-groups according to blood meal size. Females with distended abdomens in which blood was clearly visible in greater than 50% of the abdominal length were classified as having taken a large blood meal. Females lacking distended abdomens, with blood visible in less than 50% of the abdominal length were classified as having taken a small blood meal. Cages were monitored daily for mortality and 60 oviposition until all females in each cage died. Each day, dead females were counted and removed, and eggs deposited were counted under a dissection microscope. Statistical Analysis When ivermectin-treated blood was delivered via artificial membrane feeder, the LC50 and LC95 estimates for An. gambiae and An. arabiensis were identical. Mortality data for the two species were therefore pooled, and a single LC50 and LC95 was calculated using PROBIT analysis. Residual variation in egg production was modeled using a gaussian distribution. Parameter estimates were calculated using maximum likelihood, and treatment effect, in groups where eggs were produced, was determined by an F-test. Egg viability residuals were modeled using a normal distribution, and an F-test tested for differences in viability due to treatment. When mosquitoes fed upon treated cattle, survivorship data were compared across treatment groups using the Cochran-Mantel-Haenszel χ2 analysis (Parmar and Machin 1995). Kaplan-Meier survival analysis (Parmar and Machin 1995) was used to compare mosquito survivorship curves across pre- and post-treatment intervals within the ivermectin treatment group. Bootstrapping was used to calculate the 95% CIs surrounding the median survivorship estimates at each post-treatment interval in Figure 5.1. Where mortality was substantial, survivorship as influenced by blood meal size was compared across posttreatment intervals using a log-rank test. Maximum likelihood was used to estimate egg deposition at each post-treatment interval across treatments. An F-test tested for the effect of treatment on egg deposition at each post-treatment interval. All data were analyzed using R statistical software version 2.7.2 (R Development Core Team 2008). 61 Median Survival (d) Ivermectin Moxidectin Untreated Post-treatment Interval (d) Figure 5.1. Median survival time (d) of Anopheles gambiae s.s. with bootstrapped confidence intervals (N = 1000). An. gambiae s.s. sample sizes were n= 643, n= 615, and n = 666 for ivermectin, moxidectin and untreated groups respectively. 1 62 Results An. gambiae s.l. LC50 and LC95 estimates from lab feedings of ivermectin-treated blood Laboratory tests of ivermectin-treated bovine blood established that the LC50 and LC95 for An. gambiae s.l. at 19.8 ± 2.8 ppb, and 77.7 ± 8.1 ppb respectively. No eggs were produced by either species in the three highest ivermectin-in-blood treatments. The numbers of eggs produced per blood fed female in the three lowest treatment groups and the untreated group were indistinguishable (p = 0.85). Nearly all deposited eggs were viable; hatchability reached > 86% across treatments (Table 5.1). Survivorship and fecundity of An. gambiae fed on treated bulls The blood of ivermectin-treated cattle reduced survivorship and limited the fecundity of laboratory-reared An. gambiae, while moxidectin did not (p < 0.001; Figure 5.1). The median survivorship of An. gambiae feeding on the blood of ivermectin-treated cattle 1 d post-treatment was only 2 d. Thereafter, survivorship of An. gambiae ingesting ivermectin increased slowly over post –treatment intervals (Figure 5.1). Survivorship was significantly lower after feeding on ivermectin-treated cattle than on untreated cattle for up to 20 d post treatment (Figure 5.1). Furthermore, the 10 d survival probability was <10% for An. gambiae fed on ivermectin-treated cattle within two wk after cattle treatment (Figure 5.2). Relative to partial feeding, taking a full meal of ivermectin-containing blood significantly reduced mosquito survivorship for > 2 weeks after cattle treatment (Table 5.2). For 1-10 d post-treatment, neither the fully-fed nor the longer-lived partially-fed mosquitoes produced eggs (Figure 5.3). Oviposition was reduced in mosquitoes feeding on ivermectin-treated cattle (p = 0.07; Figure 5.3), though egg deposition 20 d post-treatment and thereafter of mosquitoes that fed on ivermectin-treated cattle no longer differed from the negative control. 63 Table 5.1. Viability of eggs laid by Anopheles gambiae s.s. and An. arabiensis fed treated vs. untreated blood meals. No difference in viability was observed (p = 0.78). Treatment (ppb) Species 1.00 An. gambiae 496 565 An.arabiensis 528 621 An. gambiae 574 593 An.arabiensis 377 457 An. gambiae 621 722 An.arabiensis 268 300 An. gambiae 458 481 An.arabiensis 297 332 0.10 0.01 DMSO only Eggs Hatched Total Eggs 64 Percent Hatch 86.3 89.2 87.1 93.1 1.0 Survival probability 0.8 0.6 0.4 0.2 0.0 0 10 20 30 Time post-bloodmeal (days) 40 Figure 5.2. Kaplan-Meier survivorship curves for Anopheles gambiae s.s. that blood-fed upon ivermectin-treated cattle. Each curve corresponds to a unique post-treatment interval. The arrow indicates the required development time for Plasmodium falciparum in the body of a mosquito at an external temperature of 27⁰ C (Patz and Olson, 2006). 65 50 Ivermectin Moxidectin 40 30 20 10 0 Mean eggs deposited per female Untreated 1 4 7 10 13 17 20 23 PTI by treatment Figure 5.3. Mean (±95% CIs) number of eggs deposited per female per treatment group over 8 different post-treatment intervals (d). 66 Table 5.2. Log-rank test for differences in survivorship in Anopheles gambiae s.s. taking different blood meal sizes within the ivermectin treatment group. P-values are unadjusted for multiple comparisons. Post-treatment Interval (d) Blood meal size N Observed Expected 1 Full Partial 59 27 59.0 27.0 2 Full Partial 23 31 7 Full Partial 10 χ2 χ2 df p 64.0 22.0 4.3 1 0.0373 23.0 31.0 17.8 36.2 4 1 0.0459 38 39 38.0 39.0 27.8 49.2 10.4 1 0.00127 Full Partial 68 26 68.0 26.0 69.3 24.7 0.2 1 0.698 13 Full Partial 38 29 38.0 29.0 31.1 35.9 4.3 1 0.0377 17 Full Partial 51 17 51.0 17.0 48.3 19.7 0.6 1 0.436 20 Full Partial 57 11 57.0 11.0 59.8 8.2 1.2 1 0.267 23 Full Partial 42 19 42.0 19.0 45.2 15.8 1 1 0.306 67 Discussion Ivermectin persisted in the blood of Zebu cattle, and reduced survivorship and fecundity of laboratory-reared An. gambiae populations for nearly 3 wk post treatment. If applied at the onset of the rainy season, ivermectin-treated cattle could delay the An. gambiae s.l. population explosion, possibly precluding malaria epidemics. During those 3 wk, up to 90% of An. gambiae s.l. feeding on cattle would die prior to transmitting parasites (Figure 5.2). Based upon our LC50 estimates of laboratory strain An. gambiae s.l., we speculate that ivermectin-treated cattle would similarly reduce survivorship in An. arabiensis populations. In a laboratory setting, it takes several days for a lethal blood meal to kill laboratory-reared An. gambiae. During this time, these females were 1) unable to digest their blood meals normally, 2) lethargic, and 3) unable to fly. These conditions would likely hasten mortality in the field, where hazards of desiccation, extreme temperatures, and predation are expected (Foley et al. 2000). Our results were similar to those of Bockarie et al. (1999); after 28 d elapsed between treatment and blood feeding, humans treated with ivermectin at a rate of 400 μg/kg were no longer lethal to An. punctulatus. Foley et al. (2000) document a convincing reduction in survivorship of An. farauti laboratory populations for 14 d after human ingestion of 250 μg/kg ivermectin. Insecticide-treated nets (ITNs) are reducing vector populations, as well as malaria transmission. In Western Kenyan villages, where ITNs are deployed to cover all sleeping spaces, malaria transmission has been reduced by up to 90% (Gimnig et al. 2003). Furthermore, An. gambiae is being extirpated in villages where ITN coverage reaches 90% or more (Walker et al. unpublished). In these same villages, however, both An. arabiensis and malaria transmission persist (Walker et al. unpublished). Mass treatment of humans with 68 ivermectin could reduce malaria transmission and preclude epidemics (Bockarie et al. 1999). It has also been suggested that mass treatment of livestock may reduce malaria transmission by zoophilic vectors, such as An. farauti (Iakubovich et al. 1989, Foley et al. 2000). The use of anthelmintic drugs to supplement ITNs represents a novel, integrated approach to malaria vector control. Previously, DDT and pyrethroids have been used in indoor residual spray programs and ITNs respectively for suppression of vector populations (Beier et al. 2008). However, cross-resistance between these classes of insecticide has been reported (Chandre et al. 1999). Furthermore, there is no widely-used malaria management strategy that simultaneously targets both zoophagic and anthropophagic vector populations. Additionally, no convincing cross-resistance of avermectins with pyrethroids has been reported to date (Cochran 1990, Rugg et al. 2007), giving this combination of vector control measures unique insecticidal resistance management potential. Treatment of cattle with ivermectin fulfills the need for new insecticidal compounds that can safely be used in proximity to humans, while targeting overlooked zoophagic vector populations. 69 CHAPTER 6: LETHAL AND SUBLETHAL EFFECTS OF AVERMECTIN/MILBEMYCIN PARASITICIDES ON THE AFRICAN MALARIA VECTOR, ANOPHELES ARABIENSIS (DIPTERA: CULICIDAE) Abstract Four cattle parasiticides of the avermectin/milbemycin class were examined for their lethal and sublethal effects on the zoophilic, African malaria vector Anopheles arabiensis. Ivermectin, moxidectin, doramectin, and eprinomectin were mixed with bovine blood and provided to laboratory-reared An. arabiensis in a membrane feeder. Ivermectin and eprinomectin were lethal to An. arabiensis at low concentrations (LC50s of 7.9 ppb and 8.5 ppb, respectively). While the lethality of doramectin (LC50 of 23.9 ppb), was less than that of ivermectin and eprinomectin, it markedly reduced egg development. The concentration of moxidectin required to reduce survivorship and egg production in An. arabiensis was >100 fold greater than for ivermectin or eprinomectin. Moxidectin was weak in its actions against An. arabiensis relative to the other three chemicals. These results suggest that cattle treated with ivermectin or eprinomectin in the prescribed range of low dosages as parasiticides have blood toxic to zoophilic malaria vectors. Regionally-coordinated, seasonal treatment of cattle could suppress An. arabiensis populations, thereby reducing malaria transmission. Doramectin (although less toxic) would have population level effects on egg production if used in this manner. Introduction Avermectins are glutamate-gated chloride channel activators that cause reduced motor activity and paralysis in both insects and nematodes (Martin et al. 2002). Ivermectin is a 70 member of the avermectin class, capable of reducing endo- and ectoparasite burden in vertebrates. In humans, ivermectin is ingested or applied topically to treat endo- and ectoparasitic infestations of the following: Sarcoptes scabies, Pediculus humanus, Strongyloides stercoralis, Filaria bancrofti and Onchocerca volvulus (Dourmishev et al. 2005). In cattle, the species spectrum of ivermectin activity ranges from parasitic nematodes, mites, ticks, lice, and fly larvae (Benz et al. 1989) to certain species of mosquito (Iakubovich et al. 1989, Tesh and Guzman 1990, Bockarie et al. 1999, Foley et al. 2000, Fritz et al. 2009, Chaccour et al. 2010). Circulating ivermectin in the blood of treated hosts reduces survivorship of the mosquitoes that fed on them in both laboratory and controlled field settings (Bockarie et al. 1999, Fritz et al. 2009, Chaccour et al. 2010). For example, treatment of cattle at 600 μg/kg AI resulted in sufficiently high ivermectin blood plasma titers to produce >90% mortality in the Anopheles mosquitoes feeding upon them within 2 wk posttreatment (Fritz et al. 2009). The most widespread vector control methods currently in use are insecticide-treated bed nets and indoor residual spraying, both of which reduce mosquito survivorship and successful human blood meal acquisition (Molineaux and Gramiccia 1980, Killeen and Smith 2007, Siegert et al. 2009). These control measures are relatively less effective at reducing blood meal acquisition (Killeen and Smith 2007), and vector abundance (Bayoh et al. 2010) for zoophilic vectors like An. arabiensis, compared with anthropophilic vectors like An. gambiae. In some cases, cattle serve as adequate hosts when humans are unavailable (Lefevre et al. 2009), allowing vector populations to persist. In malaria endemic areas, timed mass treatments of livestock with ivermectin could reduce zoophilic vector populations at the onset of the rainy season, precluding epidemics (Foley et al. 2000, Fritz et al. 2009). In addition, benefits for cattle shareholders would extend beyond malaria control, by reducing parasite burden of the cattle (Ciordia et al. 1984, Roncalli and Benitez Usher 1988, Dimander 71 et al. 2003), and thus increase the health and productivity of cattle herds (Ciordia et al. 1984, Dimander et al. 2003). The utility of ivermectin, however, is limited for malaria control because it currently is not labelled for use in lactating animals, which can comprise ca. 20% of a cattle population in traditional agro-pastoral regions of Africa (Voh and Otchere, 1988). It also requires post-treatment milk and meat withdrawal periods for treated animals, limiting the application of this parasiticide for vector control even further. The parasiticides moxidectin, doramectin, and eprinomectin are closely related to ivermectin and part of the avermectin/milbemycin family. The mean residence time for moxidectin and doramectin is longer than for ivermectin in the blood of cattle (Lanusse et al. 1997), and commercially available pour-on formulations of moxidectin and eprinomectin do not require meat or milk withdrawal intervals following treatment. The potential use of these compounds for malaria vector control has not been thoroughly investigated, however. Moxidectin-treated cattle do not reduce An. gambiae s.s. mosquito survivorship as effectively as did cattle treated with ivermectin at the same rate (Fritz et al. 2009). Still, the toxic dose of moxidectin required for An. gambiae complex mosquitoes has not been determined. The aim of this study was to determine the concentrations of moxidectin, doramectin, and eprinomectin that reduced survivorship and fecundity of the opportunistic malaria mosquito, An. arabiensis. Methods Mosquitoes In 2009, An. arabiensis of the Dongola strain were acquired from the Malaria Research and Reference Reagent Resource Center (MR4), and reared according to Huang et al. (2005). Thirty min prior to blood feeding, groups of twenty 3-5 d old females were aspirated into small, circular plastic cages (10 cm diam.). White mesh covering the top of the 72 cage permitted blood feeding via artificial membrane feeder, while brown paper towel over moist mud served as an ovipositional resource covering the cage floor. Throughout the study, cages of adults were always provided cotton moistened with 10% honey solution and held at 25⁰C and 80±10% RH under a 12L: 12D cycle. Preparation of treated blood Ivermectin [22,23-Dihydroavermectin B1], doramectin [25-cyclohexyl-5-O-demethyl25-de(1-methylpropyl) avermectin A1a; extracted as below], eprinomectin [4”-(epiacetylamino)-4” deoxyavermection B1], and moxidectin [(2aE,4E 5‟R,6R 6‟S,8E,11R,13S,15S,17aR,20R, 20aR,20bS)-6‟-[(E)-1,2-dimethyl-1butenyl]5‟6,6‟,7,10,11,14,15,17a,20,20a,20b-dodecahydro-20,20b-dihydroxy-5‟6,8,19,tetramethylspiro[11,15-methano-2H,13H,17H-furo[4,3,2-pq][2,6]benzodioxacyclooctadecin13,2‟-[2H]pyrano]-4‟,17(3‟H)-dione,4‟-(E)-(O-methyloxime)] were dissolved in dimethyl sulfoxide (DMSO), then added to defibrinated bovine blood (Hemostat Laboratories, Dixon, CA, USA) in a 15 mL plastic conical centrifuge tube and inverted 6 times. Initial tests of each drug consisted of a 10-fold serial dilution of drug in DMSO, so as to achieve these final concentrations of active ingredient (AI) in blood: 1000, 100, 10, and 1 ppb. Control blood contained equivalent concentrations of DMSO. For each drug, cages of mosquitoes were randomly assigned to a single blood concentration, and fed only once throughout the course of the study. The results of the initial test determined the blood concentrations of each AI used in subsequent analyses of survivorship and fecundity (Table 6.1). Mosquitoes were blood-fed via artificial membrane feeder, using parafilm as a membrane for tests of the drugs eprinomectin and moxidectin. Pork sausage casing (Great Lakes Butcher Supply, Howell, MI, USA) was substituted for parafilm during ivermectin and doramectin tests, because it enhanced mosquito acceptance of the artificial feeder. After blood-feeding was complete, 73 Table 6.1. Blood concentrations used to determine the effects of each AI on An. arabiensis survivorship and fecundity. Drug Concentration AI in blood (ppb) Eprinomectin 100, 10, 7, 4, 1, 0.1 Ivermectin 100, 10, 7, 4, 1, 0.1 Doramectin 100, 70, 40, 10, 7, 4 Moxidectin 50000, 10000, 5000, 1000, 500, 100 74 unfed mosquitoes were removed from each cage using a mouth aspirator, and the total number of blood-fed mosquitoes was recorded. Cages of mosquitoes were held for 9 d post blood-feeding. Doramectin extraction and HPLC Pure doramectin was unavailable, so it was extracted from a commercial injectable formulation for cattle (Dectomax, Pfizer Inc., New York, NY, USA). Five mL of ethanol was added to 5 mL Dectomax, and the mixture was allowed to separate after shaking in a separatory funnel (15 min). The top layer was reserved, while 5 mL EtOH was again added to the bottom layer. This process was repeated 3 times, for a total collection of 15 mL EtOH and doramectin. Thin-layer chromatography confirmed the presence of doramectin and absence of other solutes in the extract. Five μl of the doramectin extract was blotted onto a precoated silica-gel 60 F254 aluminum backed sheets. These were developed at room temperature using a mixture of hexane-acetone-decane-methanol (59:30:10:1, v/v). Sheets were viewed in a dark room under a 340-380 nm UV lamp. After doramectin was confirmed present, the extract was dried under a stream of nitrogen for 12 h in a dark hood. The final product was dissolved in DMSO to make a stock solution of expected concentration 10 ppm. The entire extraction process was replicated 3 times to produce 3 different doramectin extracts, each used twice during the study. HPLC was used to quantify total doramectin in two of the extracts. Samples of each extract were diluted with DMSO to a working level for the HPLC/MSD. The LC system was a Waters 2695 HPLC and analytical column Waters XBridge (3.0 x 50mm) packed with 3.5μm C18 stationary phase. Samples (10μl) were eluted using an isocratic mobile phase consisting of 50% acetonitrile, 50% water and 0.1% triethylamine at a flow rate of 0.25ml min-1. Mass spectrometry (MS) was run at basic pH 75 conditions in order to promote the formation of analyte anions. Detection by MS in negative ion mode required an atmospheric pressure chemical ionization interface (APCI), and was performed on a Waters MicroMass ZQ mass spectrometer. Ions 591.5 and 815.6 were monitored for detection of doramectin; ion 591.5 was used for quantitation and 815.6 for confirmation. Data collection and statistical analysis Beginning 1 d post-blood-meal, cages were examined for the presence of mosquito carcasses and egg production every other day. Dead females were counted and removed from the cage floor; eggs were counted under a dissection microscope. After 9 d of observation, mosquitoes from each cage were frozen at -20⁰C, and then dissected for the presence of eggs. Eggs present of Christopher‟s stages IV and V were counted under a dissection microscope. Logistic regression was used to model survivorship and egg deposition data for each drug. For all drugs, survivorship residuals were specified as binomially distributed, and Abbott‟s correction was applied to the mean proportion of individuals deceased at each dose. A log transformation was applied to dosage within each drug treatment, and maximum likelihood estimation (MLE) was used to estimate parameters for mortality models. LC50 and LC95 estimates from these models were calculated using the MASS library dose.p program for R statistical software (v. 2.7.2; Anon. 2010). Residual variation in egg production was modeled using a Poisson distribution for each drug. The effect of drug dosage on egg development was explored using a mixed effects model; drug dosage was specified as fixed, while date of feeding was specified as random. A likelihood ratio test (LRT) determined the effect of drug dosage on egg production. To determine whether drug and drug dosage influenced oviposition relative to 76 total egg development, all egg deposition below each LC50 was expressed as a proportion over total eggs developed (oviposited + dissected). Within each treatment group, the proportion of eggs deposited by females fed drug dosages below the LC50 was compared with the proportion of eggs deposited by females fed blood treated only with DMSO. Data were fit to a generalized linear model with a quasibinomial distribution, and drug contributions to the proportion of eggs oviposited relative to the total eggs developed (oviposited + dissected) were assessed with a t-test. Results A summary of the lethality of the parasiticides on An. arabiensis can be found in Table 6.2. Mortality elicited by treated blood was highly variable among the four compounds. Eprinomectin and ivermectin were highly lethal to An. arabiensis; LC50 values were 8.48 ppb and 7.85ppb, respectively. While the concentration required to kill 95% of blood feeding An. arabiensis tended to be lower for eprinomectin than for ivermectin, the difference was not statistically significant based upon their overlapping 95% confidence intervals. Average recovery of doramectin was 80% as determined by HPLC. Thus, expected doramectin dosages (Table 6.1) were multiplied by a correction factor of 0.8 prior to statistical analysis. Doramectin and moxidectin were comparatively less lethal, requiring, respectively, 3 and 148 times the blood concentration to kill An. arabiensis than either eprinomectin or ivermectin. As expected, increasing the amount of drug in the proffered blood increased mosquito mortality across all four dewormers (Figures 6.1and 6.2). Three of the four drugs also exerted sub-lethal effects. Surviving individuals having fed upon ivermectin-, doramectin-, and moxidectin-treated blood produced fewer eggs than 77 Table 6.2. Abbott‟s corrected LC50 and LC95 values for An. arabiensis fed upon eprinomectin-, ivermectin-, doramectin- and moxidectin-treated blood, given in parts per billion of drug in the mosquito blood meal. The  and  values represent the intercept and slope of the logistic regression analyses, respectively. Drug N df α β LC50 (95% CI) LC95 (95% CI) Eprinomectin 451 5 -3.66 1.71 8.5 (7.2, 10.0) 47.4 (28.2, 79.8) Ivermectin 518 5 -2.17 1.06 7.9 (6.2, 9.9) 128.1 (62.1, 264.4) Doramectin 585 5 -3.18 1.00 23.9 (19.1, 30.0) 453.0 (236.2, 868.7) Moxidectin 621 5 -0.25 1.51 1181.0 (999.4, 1395.7) 8305.3 (5758.7, 11978.1) 78 Mortality Mean no. eggs/ survivor 1 0.8 0.6 0.4 0.2 Ivermectin Proportion Deceased Egg Production 0 0 0.0001 0.01 1 100 10000 1 Mean no. eggs/ survivor Proportion Deceased Eprinomectin 0.8 0.6 0.4 0.2 0 0 0.0001 0.01 1 100 10000 50 40 30 20 10 0 0 0.0001 0.01 1 100 10000 0.01 1 100 10000 50 40 30 20 10 0 0 0.0001 Dose (ppb) Dose (ppb) Figure 6.1. Uncorrected mean proportion of Anopheles arabiensis deceased (left) and production of eggs by survivors (right) within 9 d of feeding upon eprinomectin- and ivermectin-treated blood. Error bars represent bootstrapped 95% confidence intervals (N=1000). 79 Mortality Egg Production Mean no. eggs/ survivor Proportion Deceased Doramectin 1 0.8 0.6 0.4 0.2 0 0 0.0001 0.01 1 100 10000 Mean no. eggs/ survivor Proportion Deceased Moxidectin 1 0.8 0.6 0.4 0.2 0 0 0.0001 0.01 1 100 Dose (ppb) 10000 50 40 30 20 10 0 0 0.0001 0.01 1 100 10000 1 100 10000 50 40 30 20 10 0 0.0001 0 0.01 Dose (ppb) Figure 6.2. Uncorrected mean proportion of Anopheles arabiensis deceased (left) and production of eggs by survivors (right) within 9 d of feeding upon doramectin- and moxidectin-treated blood. Error bars represent bootstrapped 95% confidence intervals (N=1000). Doramectin dose-response curve was adjusted using HPLC analysis of final extraction product. 80 Table 6.3. Mean number of eggs per surviving female with bootstrapped 95% confidence intervals (N=1000). A significant p-value (denoted by *) signifies that increasing the drug b dosage significantly reduces egg production in surviving Anopheles arabiensis. Drug N χ2 df p-value Dose Mean no. eggs per survivor (95% CI) Eprinomectin 38 1 0.6082 0 0.1 1 4 7 10 27.2 (15.5, 39.2) 27.0 (20.0, 34.0) 20.4 (12.9, 27.5) 9.3 (2.9, 16.1) 2.2 (0, 4.8) 9.4 (0, 22.2) Ivermectin 32 1 0.0003 0 0.1 1 4 7 10 25.7 (12.6, 40.6) 34.0 (22.0, 45.0) 24.2 (17.4, 33.6) 13.2 (4.2, 24.0) 2.8 (0, 7.8) 0 (0, 0) Doramectin 35 1 0.0042 0 0.1 1 4 7 10 40 70 100 23.93 (19.07, 30.02) 19.0 (11.3, 24.8) 30.5 (26.0, 35.0) 24.0 (9.8, 39.2) 17.8 (8.4, 29.4) 10.2 (3.5, 17.8) 0 (0, 0) 0 (0, 0) 1.0 (0, 2.0) Moxidectin 26 1 0.0031 0 100 500 1000 5000 40.2 (22.4, 51.8) 30.3 (24.5, 39.3) 13.2 (6.3, 20.0) 0.2 (0, 0.5) 0 (0, 0) 81 Table 6.4. Mean proportion of eggs oviposited out of the total number of eggs produced (± bootstrapped 95% CIs; N=1000) by Anopheles arabiensis females surviving dosages lower than each drug‟s LC50 compared with respective controls. Drug Eprinomectin Treatment n Mean proportion of eggs deposited ( 95% CI) 92 Treated below LC50 120 0.52 (0.35, 0.69) 91 0.69 (0.40, 0.93) Treated below LC50 155 0.27 (0.16, 0.39) 82 0.32 (0.08, 0.58) Treated below LC50 111 0.08 (0.01, 0.16) DMSO only 51 0.05 (0, 0.15) p-value 0.16 (0.01, 0.34) DMSO only Moxidectin 0.23 (0.06, 0.44) DMSO only Doramectin 144 DMSO only Ivermectin Treated below LC50 t-value 82 0.37 0.71 0.07 0.95 1.02 0.32 -0.56 0.58 those that fed upon blood treated with only DMSO (Figures 6.1and 6.2). While egg production tended to be lower for eprinomectin survivors than for mosquitoes fed upon DMSO-treated blood, this difference was not statistically significant (Table 6.3). An. arabiensis fed upon drug-treated vs. DMSO-treated blood deposited the same proportion of their eggs (Table 6.4). Therefore, capacity for egg deposition was not influenced by drug treatment. Discussion The activity of eprinomectin-treated blood against An. arabiensis is equal to ivermectin at low dosages (Table 6.2). It is highly active against An. arabiensis, and labeled for use in pregnant and lactating cattle, qualifying it as a potential tool for management of zoophilic malaria vectors. However, plasma availability of eprinomectin is low in zebu Gobra cattle (Bos indicus), compared with Holsteins (B. taurus) treated at the same rate, though the mechanism for this difference remains unknown (Bengone-Ndong et al. 2006). When treated topically at a rate of 200 μg/kg active ingredient (AI), Holstein cattle achieve eprinomectin plasma concentrations of 43-76 ng/mL (Alvinerie et al. 1999), while zebu Gobra only reach plasma concentrations of ca. 8 ng/mL (Bengone-Ndong et al. 2006). The milk/plasma ratio of 0.094 in eprinomectin-treated zebu Gobra (Bengone-Ndong et al. 2006) might permit an increase the amount of AI applied to zebu Gobra cattle to increase plasma deposition; the maximum concentration of eprinomectin permitted in milk is 30 ng/mL (Alvinerie et al. 1999). The effects of increased eprinomectin exposure on both zebu Gobra cattle health and safety, and meat withdrawal times should be thoroughly examined. Interestingly, doramectin-treated blood was less lethal to An. arabiensis than ivermectin or eprinomectin, but caused egg production to severely decline at concentrations higher than 10 ppb. Thus it functioned well as a sterilant (Table 6.3). Conversely, 83 eprinomectin-treated blood was highly lethal to An. arabiensis, but did not reduce egg production as strongly in survivors. Moxidectin required higher dosages to reduce survivorship and egg production for An. arabiensis. These results suggest variation in mosquito target site sensitivity to each of the drugs. Furthermore, drug interaction with the target site, or in the case of doramectin, the target site itself, may vary according to drug. Some insecticides also modulate insect behaviors in addition to their lethal effects (Haynes 1988, Siegert et al. 2009). While it was beyond the scope of this study, it would be of interest to know whether, and for how long, the avermectins reduce host-seeking behaviors in mosquitoes that have imbibed sub-lethal doses. Use of doramectin and eprinomectin is not restricted in pregnant animals, and pregnant and lactating animals respectively, whereas ivermectin has such restrictions. The broader applicability of both eprinomectin and doramectin make them better candidates for malaria mosquito control than ivermectin, against those more zoophilic vectors such as An. arabiensis whose primary host is often cattle. Field tests of doramectin- and eprinomectintreated cattle should be the next step in the investigation of this novel vector control measure. The mass treatment of cattle populations with ivermectin, doramectin, or eprinomectin at the onset of the rainy season, in combination with ITN use, is a promising management technique for zoophilic vectors like An. arabiensis. 84 CHAPTER 7: HOST UTILIZATION BY ANOPHELES GAMBIAE S.S. AND ANOPHELES ARABIENSIS IN AN AREA OF HIGH INSECTICIDE-TREATED BED NET COVERAGE IN WESTERN KENYA AS DETERMINED BY REVERSE DOT BLOT, DNA-DNA HYBRIDIZATION Abstract A DNA-DNA hybridization technique previously used in sand flies, reverse dot blot analysis (RDBA), was adapted to the An. gambiae s.l. system. Blood meal identification for An. gambiae s.l. via RDBA was comparable to conventional PCR and direct sequencing techniques. However, RDBA identified mixed blood meals, whereas PCR and direct sequencing could not. Blood fed An. gambiae s.l. collected in June and July 2008 in Kisian, Kenya were 58% An. arabiensis, and 24% An. gambiae s.s. Eighteen percent could not be identified to species via qPCR. Blood meals imbibed by An. arabiensis were mostly (>90%) bovine in origin, whereas An. gambiae s.s. fed upon humans > 90% of the time. Two An. arabiensis of 160 took blood meals from more than one host species. Recent insecticidetreated bed net (ITN) use in Kisian village has likely caused the shift in the dominant vector species from An. gambiae s.s. to An. arabiensis. We conclude that vector populations will not decline to levels whereby malaria transmission is eliminated until the cattle-mosquito connection is broken. Additional management tools targeting cattle-feeding vectors are badly needed. 85 Introduction Host-selection patterns influence transmission in vector-borne disease systems, and increase efficiency of pathogen transmission among host populations when vectors exhibit high host fidelity (McCall and Kelly 2002). Analyzing vector populations for patterns of host-selection achieves several objectives: 1) identifies hosts important to pathogen amplification (Molaei et al. 2006a, Hamer et al. 2009), 2) determines the relative contribution of vector species to disease transmission (Apperson et al. 2004, Molaei et al. 2006b), and 3) elucidates seasonal host-use shifts (Edman and Taylor 1968, Kilpatrick et al. 2006). Tools used to examine mosquito host-selection include immunological- and DNA-based technologies to analyze the gut contents of mosquitoes post-blood meal (Symondson 2002). Enzyme-linked immunosorbent assays (ELISA) largely replaced the precipitin and related immunological methods to assess host-selection patterns (Beier et al. 1988, Scott et al. 1993, Rubio-Palis et al. 1994, Bogh et al. 1998, Diatta et al. 2001, Mwangangi et al. 2003, Apperson et al. 2004, Lefèvre et al. 2009). However, immunological methods suffer from several drawbacks including lack of specific reagents, and low sensitivity especially for older blood meals (Symondson 2002). More recently, DNA-based techniques have been successfully used for vector blood meal analysis (Lee et al. 2002, Cupp et al. 2004, Kent and Norris 2005, Kilpatrick et al. 2006, Molaei et al. 2006a, Molaei et al. 2006b, Hamer et al. 2008, Hamer et al. 2009, Kent 2009). In many studies (Kilpatrick et al. 2006, Molaei et al. 2006a, Molaei et al. 2006b, Hamer et al. 2008, Hamer et al. 2009), DNA from the gut contents of mosquitoes is extracted, amplified via PCR, directly sequenced and compared with a reference data-base for host identification. While DNA-based assays are efficient and the results more reliable than those of ELISA, tools used for direct sequencing are not widely available in field laboratories. Additional tools for analyzing host-selection in field settings would be useful. 86 The mosquitoes Anopheles gambiae sensu stricto and An. arabiensis are highly important to the perpetuation of the human malaria transmission cycle in Sub-Saharan Africa. While host-selection by An. gambiae s.s.(hereafter An. gambiae) tends toward anthropophily (White and Rosen 1973, White 1974, Coluzzi et al. 1979), and An. arabiensis tends toward zoophily (Highton et al. 1979, Duchemin et al. 2001), regional differences in host-selection by An. gambiae (Diatta et al. 1998, Bøgh et al. 2001) and An. arabiensis (Tirados et al. 2006, Kent et al. 2007), as well as phenotypic plasticity (Lefèvre et al. 2009) have recently been described. Vector control measures executed within human dwellings, such as insecticidetreated bed nets (ITNs) preclude human blood meal acquisition (Molineaux and Gramiccia 1980, Gimnig et al. 2003, Killeen and Smith 2007), thereby causing An. gambiae populations (White 1974, Bayoh et al. 2010) and malaria transmission (Lindblade et al. 2004) to decline. Some studies have examined the shifts in An. gambiae and An. arabiensis abundance postITN distribution (Bayoh et al. 2010) but the influence of ITNs on host-selection by these vectors remains understudied. In the present study, we adapted a novel blood meal analysis tool, reverse dot blot analysis (RDBA), previously used in Phlebotomine sand flies (Abbasi et al. 2009) to the An. gambiae s.l. system, and examined host-selection in an area of high ITN use. We postulated that the acquisition of human blood meals by An. gambiae and An. arabiensis following a mass distribution of ITNs in Kisian in 2006, would decline, and the proportion of blood meals taken from non-human hosts would increase. Moreover, we postulated that the diversity of hosts utilized by An. gambiae and An. arabiensis would increase as human hosts become unavailable. 87 Methods Study site Sampling was conducted in Western Kenya, in the village of Kisian (Nyanza Province), during the months of June and July in 2008. Lindblade et al. (2004) previously described the demography, physiography, and geography of this locale. ITNs were owned by ca. 68% of households within the sampling area (Hightower et al. 2010). Mosquito collection and species identification Blood fed mosquitoes were collected indoors from the walls of human dwellings, and outdoors from clay pots (Odiere et al. 2007) via mouth aspirator. The clay pots had been placed by local residents to collect roof water and were typically situated near the houses. Mosquitoes were collected on 16 dates from 57 housing compounds. After collection, mosquitoes were frozen, and then morphologically identified as An. gambiae s.l. Engorged abdomens were separated from the thorax with a scalpel, then placed individually in vials and dried in a desiccator containing Drierite for transport to Michigan State University. There, DNA was extracted from the mosquito abdomen (DNeasy Tissue Kits; Qiagen, Valencia, CA) using sterile technique. Extracted DNA could be used in the identification of both host and vector depending on the primers used for amplification. To identify whether the vector was An. gambiae or An. arabiensis, extracted DNA was amplified via quantitative PCR according to Walker et al. (2007). Taqman mastermix (Applied Biosystems P/N 4304437) and An. gambiae s.l. universal primers (Forward: 5‟-GTGAAGCTTGGTGCGTGCT-3‟, and Reverse: 5‟-GCACGCCGACAAGCTCA-3‟) amplified a ca. 150 bp region of the An. gambiae s.l. intergenic spacer region (IGR) of sample DNA. Fluorecently-labeled speciesspecific Taqman probes (An. gambiae: 5‟VIC – CGGTATGGAGCGGGACACGTA-3‟, An. arabiensis: 5‟FAM - TAGGATGGAGAAGGACACTTA) were coupled to a minimum 88 groove binding ligand and a 3‟quencher, and were added to the reaction. Results were identified via spectrophotometer, after the 5‟ nuclease activity of the Taq polymerase released the fluorescent label, causing it to fluoresce. Amplification and detection were accomplished with an ABI Model 7900 HT workstation (Applied Biosystems, Foster City, California, USA). Blood meal identification via direct sequencing Host DNA was initially identified using an established protocol for blood meal identification following Hamer et al. (2009). Extracted DNA used in mosquito identification was amplified via PCR, purified and directly sequenced (ABI Prism 3700 DNA Analyzer; Applied Biosystems, Foster City, CA). For host identification, sequences were queried in GenBank (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi) with a BLAST search. Host identity was accepted for a sample when the following criteria were met: 1) the sample produced an amplicon after PCR, and 2) the sample sequence was matched with >95% similarity according to the BLAST search. Host DNA amplification with Cyto primer pair The forward and reverse primers from Abbasi et al. (2009) (hereafter, “Cyto” primers) are universal primers that amplify a 344 bp segment in a conserved region of the mitochondrial cytochrome b gene. A biotin group was added to the 5‟ end of Cyto primer 1 (5‟ – CCA TCA AAC ATC TCA GCA TGA TGA AA-3‟) and Cyto primer 2 (5‟ – CCC CTC ATA ATG ATA TTT GTC TCT-3‟), forward and reverse primers respectively. Biotinlabeled DNA hybridized to probes on the membrane could be detected using a Biotin Chromogenic Detection Kit (Fermentas International Inc., Burlington, Ontario, Canada) These primers were examined for An. gambiae complex host identification suitability. 89 Extracted host DNA was amplified in a 100 μl reaction (4 μl DNA extract with 3.3μl of 25 pM primer per sample) using the Failsafe PCR system (Epicenter Biotechnologies, Madison, WI), under the following reaction conditions: an initial denaturation step of 5 min at 94⁰C, followed by 35 cycles at 94⁰C for 30 sec, 55⁰C for 30 sec, and an elongation step at 72⁰C for 1 min. Electrophoresis (E-gel system; Invitrogen, Carlsbad, CA) revealed amplicons that would be submitted for identification via direct sequencing and RDBA. Amplicons were visualized and scored based upon band intensity (0 = no product and 4 = bold product). An aliquot (45 μL) of samples yielding a visible amplicon was purified (QIAquick PCR Purification Kit; Qiagen), submitted for direct sequencing, and identified as above. The remaining 45 μL of amplified and unpurified host DNA was directly subjected to RDBA. Reverse dot blot analysis Candidate hosts were selected based upon 1) Organism-specific oligonucleotide probes (Table 7.1) were diluted with Tris-EDTA buffer to a concentration of 5 pM/μL, and imprinted onto a nylon membrane in a double line. Probes were covalently linked at the 5‟ amino modified end to Pall Biodyne B nylon membranes (Nagle Nunc International, Rochester, NY) using a 96 pin high-density replicating tool on a Biomek 2000 robot (Beckman Coulter, Inc.). Printed membranes were baked for 30 min at 80⁰C to fix the probes. Membranes were cut into strips containing each probe (Fig. 7.1), and strips placed individually into a 15 mL conical tube. Membrane strips were washed with 3mL of 2xSSC [0.15 M NaCl, 0.015 M sodium citrate, 0.1% sodium dodecyl sulfate (SDS)] solution in preparation for hybridization with DNA samples. Tubes were held at 45⁰C for 30 min, and gently rocked to allow the prehybridization solution to wash over the membranes. Amplified 90 Table 7.1. Oligonucleotide probes imprinted upon membrane strips for reverse dot blot procedure. Species Oligonucleotide sequence 1 Bovine 1 (Bos indicus) ATT ATG GGT CTT ACA CTT T 2 Bovine 2 (Bos taurus) ATT ACG GGT CTT ACA CTT T 3 Brown rat (Rattus rattus) CAG TCA CCC ACA TCT GC 4 Goat (Capra hircus) ATA CAT ATC GGA CGA GGT CTA 5 Sheep (Ovis aries) TCC TAT TTG CGA CAA TAG CTT CCT 6 Domestic mouse (Mus musculus) TGG AGT ACT TCT ACT GTT CGC AGT 7 Domestic cat (Felis domesticus) CAT TGG AAT CAT ACT ATT 8 Human (Homo sapiens) ATG CAC TAC TCA CCA GAC GC 9 Domestic chicken (Gallus gallus) CAT CCG GAA TCT CCA C 10 Domestic dog (Canis familiaris) CAG ATT CTA ACA GGT TTA 11 General Avian 2 GCC TCA TTC TTC TTC AT 12 General Avian 1 TAC ACA GCA GAC AC 91 BOV 1 BOV 2 Rat Goat Sheep Mouse Cat Human Chicken Dog AV 2 AV 1 Tab for handling 1 cm Figure 7.1. Schematic of a membrane strip used for reverse dot blot analysis, including placement of the probes. On the right, a membrane strip used in a test reveals that the unknown host was bovine. 92 host DNA, suspended in a total volume of 45 μL, was heated to 94⁰C for 5 min as a melting step, and then added to the conical tube for hybridization. Hybridization of host DNA to the membrane took place at 43⁰C for 1 hr. Following hybridization, the membrane was washed with 2 mL of 0.7xSSC solution for 20 min at room temperature. A Biotin Chromogenic Detection Kit (Fermentas International Inc., Burlington, Ontario, Canada) was used for visual detection of the results. Tests of membrane strip accuracy for detecting host DNA were first conducted using DNA samples extracted from 5 μL blood samples of known origin. Blood was also combined from two hosts in proportions ranging from 1:7- 7:1, and DNA was extracted and amplified. Mixed DNA treatments were then subjected to RDBA to determine the sensitivity with which this method detects a mixed blood meal. The remaining 45ul of amplified DNA from unknown samples were identified via RDBA, and a Pearson‟s chi-squared test was used to compare success rates of RDBA and direct sequencing. Results In total, 299 blood-fed anopheline mosquitoes were collected in June and July of 2008. Molecular identification of the species established that An. arabiensis comprised 58% of the mosquitoes collected, while 24% were An. gambiae. The remaining 18% could not be identified to species, but were morphologically An. gambiae s.l. Culex spp. were discarded and Anopheles funestus was absent. An. gambiae were four times more abundant from indoor resting sites than outdoor resting sites, whereas An. arabiensis were twice as abundant from outdoor resting sites (Table 7.2). In 46 of the samples, there was insufficient DNA available for both Reverse Dot Blot Analysis (RDBA) and direct sequencing. In these cases, only blood meal analysis via direct 93 sequencing was performed. By at least 1 of the 3 methods described above, hosts were identified for ca. 75% of the mosquitoes. Only 3 host animals were fed upon by An. gambiae and An. arabiensis: humans, cattle, and birds. Nearly eighty percent of the total blood meals identified were bovine. An. arabiensis fed upon cattle 50 times more frequently than upon humans, while most An. gambiae fed upon humans (Table 7.3). Identification of visible amplicons by direct sequencing and RDBA were equivalent (p = 0.327). Direct sequencing identified the utilized hosts in 147 of the field collected samples, whereas RDBA identified 139. Furthermore, analysis of field samples identified by both RDBA and direct sequencing with Cyto primers always yielded identical results. DNA from mixed blood meals was identifiable by RDBA, even when 1 blood type represented as little as 1/4th of the total blood meal (Figure 7.2). Three of the field samples analyzed via RDBA revealed mixed blood meals: 2 An. arabiensis fed upon both human and bovine hosts, and 1 unidentified An. gambiae s.l. individual fed upon human and avian hosts. The avian host was identified by BLAST to a Turdus species. Discussion The proportion of An. gambiae relative to An. arabiensis collected in June and July of 2008 suggests a shift in the predominant An. gambiae s.l. complex vectors present in human dwellings. Prior to 1998, An. gambiae comprised between > 70% of the total An. gambiae complex population in Kisian village (Joshi et al. 1975, Service et al. 1978, Petrarca et al. 1991, Githeko et al. 1994), yet comprised < 25% in the present study. Widespread ITN use is likely contributing to this phenomenon, which has been demonstrated in this part of Western Kenya (Bayoh et al. 2010). 94 Table 7.2. Resting sites from which An.gambiae s.s. and An. arabiensis were collected. Indoor resting sites included the sides and underside of furniture, and the interior walls of houses. Outdoor resting sites were the interior of clay pots used for rain catchment outside of houses. Individuals classified as An. gambiae s.l. could not be identified to species level via qPCR. Species Resting Site Indoor Outdoor An. gambiae s.s. 58 14 An. arabiensis 67 105 An. gambiae s.l. 40 15 95 Table 7.3. An. gambiae s.s. and An. arabiensis hosts were identified by at least 1 method of blood meal analysis. Individuals whose blood meals were mixed or avian were classified as “other”. Species Host Resting Site Indoor 29 4 0 0 Bovine 49 105 1 2 Other 1* 2** Bovine 14 6 Human 4 1 Other * ** *** 1 Human An. gambiae s.l. 2 Other An. arabiensis Bovine Human An. gambiae s.s. Outdoor 0 1*** Turdus spp. Mixed bovine and human blood meals Mixed human and avian blood meal 96 1:3 Bovine:Goat 1:1 Bovine:Goat 3:1 Bovine:Goat Goat only Bovine only Ladder Figure 7.2. Relative visualization of probes in a mixed blood meal based upon the proportion of DNA from each host. Gel bands demonstrate a roughly equal quantity of DNA per samples, yet relative intensity of the RDBA probe varies based upon the proportion of DNA from each host. In this figure, background brightness for some strips was adjusted to create equivalent background brightness across all strips. 97 Surprisingly, increased ITN coverage does not appear to have shifted host-use by An.gambiae. Neither An. gambiae nor An. arabiensis expanded their host range, despite the reduced availability of human hosts. Over 90% of An. gambiae fed upon humans, yet ITNs had been present in nearly 70% of human dwellings since 2006. These results are consistent with those of Fornadel et al. (2010), which demonstrated that host-selection of an anthropophilic An. arabiensis population remained unchanged in the two years following mass ITN distribution. In Burkina Faso, long-term inaccessibility of human hosts (Robert et al. 1991) has led to the evolution of phenotypic plasticity, such that nearly 90% of An. gambiae feed upon cattle, though preference for human odor-baited traps is maintained in the population (Lefèvre et al. 2009). Perhaps the latency between ITN distribution and the present study, household ITN ownership, nightly ITN use, or a combination of these was insufficient to generate selection pressure for phenotypically plastic host-use in the Kisian An. gambiae population. In the present study, the proportion of blood meals taken from cattle reveals their importance for malaria control. In some areas where ITN use is widespread, the An. gambiae population declines, but malaria transmission persists (Bayoh et al. 2010). We postulate that An. gambiae s.l. populations persist by feeding on cattle when deprived of human hosts, yet feed on humans whenever the opportunity arises. Additional vector control measures may be necessary to break the cattle-mosquito connection if the goal of malaria eradication is to be achieved. Treatment of cattle with the systemic insecticides, such as ivermectin (Foley et al. 2000, Fritz et al. 2009), or topical insecticides, such as pyrethroids (Hewitt and Rowland 1999) has been demonstrated as an effective way to reduce zoophilic vector longevity. Further research is necessary to determine whether deprival of both human and cattle hosts in a field setting, using ITNs and cattle treatment, could reduce An. gambiae s.l. populations and thereby improve malaria control. 98 RDBA was as successful at identifying host DNA, as direct nucleotide sequencing, and the two methods consistently returned identical results for our field collected samples. One major advantage of RDBA is that it can detect mixed blood meals, while direct sequencing cannot. While the present study only addresses An. gambiae interspecies hostuse, questions remain about how mosquitoes interact with, and feed upon multiple human hosts present in a single dwelling. RDBA could be extended to address this question; unique probes could be identified for individuals living in the same household, then imprinted on a membrane strip and used for blood meal identification. While RDBA appears to be an accurate and sensitive blood meal identification tool, it is best suited for systems in which host-use by vectors is restricted to relatively few host species (Abbasi et al. 2009). In the current study, the RDBA probes represented potential hosts that were most abundant around human dwellings. Furthermore, direct sequencing did not reveal any additional host animals that could not be detected via RDBA around Kisian village. However, blood meals from vectors with a broad range of hosts are still most accurately identified via direct nucleotide sequencing. 99 FINAL CONCLUSIONS AND FUTURE RESEARCH DIRECTIONS Part I: Ovipositional flexibility of An. gambiae An. gambiae is highly flexible in its ovipositional behavior. Previous work reveals a lack of stimulatory chemical cues for oviposition. While darkness and wetness are key physical stimuli for gravid females locating an ovipositional resource (Huang et al. 2007), An. gambiae accept a broad range of moistures, textures, and substrates onto which they deposit their eggs. The present studies demonstrate that An.gambiae oviposition peaks at dusk, and to a lesser extent, dawn, yet egg deposition is not restricted to any particular time of day. These ovipositional peaks are attributable to multiple individuals that deposit all their eggs at different hours of the night. Furthermore, oviposition can occur in the middle of the day if gravid females are previously deprived of an ovipositional resource. Most An.gambiae deposit eggs while sitting horizontally on a moist substrate, but gravid females are also capable of flying and perching on a vertical structure near a water source and raining onto the water below. Understanding the ovipositional flexibility of this species is key to understanding its distribution and potential for management. The present, as well as previous studies suggest that An. gambiae is an ovipositional generalist, capable of broadcasting eggs over a broad range of aquatic environments (Minakawa et al. 1999, Minakawa et al. 2004, Huang et al. 2005, 2006a, 2006b, 2007, Miller et al. 2007, Omlin et al. 2007). This behavior would reduce the effectiveness of ovitraps for control of Ae. aegypti (Sithiprasasna et al. 2003) and other container-breeding mosquitoes. Furthermore, small larval populations may be distributed beyond the small, sunlit pools devoid of vegetation previously described as larval habitats for An. gambiae. During the rainy season, daily heavy rainfalls that cause water to flow over the ground surface may carry broadly cast eggs and larvae to low-lying areas where water collects. I speculate that flowing rain water may 100 facilitate generalistic ovipositional behavior in this important vector species. To date, no study has examined how flowing water during the rainy season influences the distribution of An. gambiae eggs and larvae. Further research on this aspect of the ecology of An. gambiae is critical to successful management of this important vector species. Part II: An. arabiensis control using eprinomectin In areas of where ITNs are broadly distributed and used to cover sleeping humans, An. arabiensis populations persist. We conclude that Western Kenyan populations of An. arabiensis still remain in close proximity to human dwellings, and continue to bite cattle despite ca. 70% ITN coverage. Interestingly, in the present studies, we demonstrate that An. gambiae and An. arabiensis are highly sensitive to the parasiticide, ivermectin. Although ivermectin-treated blood significantly reduces survivorship and fecundity of An. gambiae s.l. that feed upon it, strict milk and meat withdrawal times, as well as restrictions against its use in pregnant animals preclude widespread cattle treatment for malaria control under current registration guidelines. The present studies demonstrate that eprinomectin is an acceptable alternative to ivermectin, causing equivalent lethality in An. arabiensis that feed upon it. To determine the feasibility of this promising vector control measure, investigation of eprinomectin-treated cattle and their effects on An. arabiensis should expand to the village level. An. arabiensis population growth could be quantified in villages where cattle were treated with eprinomectin at the onset of the rainy season and compared to An. arabiensis population growth in villages with untreated cattle. If used in combination with ITNs, I suggest that eprinomectin could significantly reduce vector abundance. The results generated from such research would be important not only to the scientific community, and our understanding of vector control, but could also have life-saving consequences for people living in malaria endemic areas. 101 APPENDIX 102 APPENDIX 1: RECORD OF DEPOSITION OF VOUCHER SPECIMENS* The specimens listed on the following sheet(s) have been deposited in the named museum(s) as samples of those species or other taxa, which were used in this research. Voucher recognition labels bearing the Voucher No. have been attached or included in fluid-preserved specimens. Voucher No.: 2011-03 Title of thesis or dissertation (or other research projects): INVESTIGATIONS OF THE AFRICAN MALARIA MOSQUITO (ANOPHELES GAMBIAE S.L., DIPTERA: CULICIDAE): OVIPOSITIONAL BEHAVIOR AND TOXICITY OF AVERMECTINS Museum(s) where deposited and abbreviations for table on following sheets: Entomology Museum, Michigan State University (MSU) Investigator‟s Name(s): Megan L. Fritz Date: August 12, 2011 *Reference: Yoshimoto, C. M. 1978. Voucher Specimens for Entomology in North America. Bull. Entomol. Soc. Amer. 24: 141-42. 103 APPENDIX 1.1 Voucher Specimen Data Investigators Name: Date of deposition: Megan L. Fritz August 12, 2011 Voucher No. 2011-03 The following specimens have been received by the Michigan State University Entomology Museum. 4 Adult females 10 10 Adult males Pupae Larvae Anopheles gambiae sensu stricto Anopheles arabiensis MSU colony MSU colony Species or other taxon Eggs Label data for specimens collected or used and deposited No. specimens: Other 104 REFERENCES 105 REFERENCES Abbasi, I., R. Cunio, and A. Warburg (2009). Identification of blood meals imbibed by Phlebotimine Sand Flies using cytochrome b PCR and Reverse line blotting. Vector-borne Zoonotic Diseases 9(1):79-86. Allemand, R. (1976a). Les rythmes de vitellogenese et d‟ovulation en photoperiode LD 12:12 de Drosophila melanogaster. Journal of Insect Physiology 22: 1031-1035. Allemand, R. (1976b). Influence de modifications des conditions lumineuses sur les rythmes circadiens de vitellogenese et d‟ovulation chez Drosophila melanogaster. Journal of Insect Physiology 22: 1075-1080. Allemand, R. (1977). Influence de l‟intensite d‟eclairement sur l‟expression du rythme journalier d‟oviposition de Drosophila melanogaster en conditions lumineuses LD 12:12. C R Acad Sci Paris D 1977, 284: 1553-1556. Alvinerie, M., J.F. Sutra, P. Galtier, and C. Mage (1999). Pharmacokinetics of eprinomectin in plasma and milk following topical administration to lactating dairy cattle. Research in veterinary Science 67: 229-232. Anon. (2010). R: a Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing. Apperson, C.S., H.K. Hassan, B.A. Harrison, H.M. Savage, S.E. Aspen, A. Farajollahi, W. Crans, T.J. Daniels, R.C. Falco, M. Benedict, M. Anderson, L. McMillen, and T.R. Unnasch (2004). Host feeding patterns of established and potential mosquito vectors of West Nile Virus in the Eastern United States. Vector-borne and Zoonotic Diseases 4(1): 7182. Arena, J. P., K. K. Liu, P. S. Paress, E. G. Frazier, D. F. Cully, H. Mrozik, and J. M. Schaeffer. 1995. The mechanism of action of avermectins in Caenorhabditis elegans: correlation between activation of glutamate-sensitive chloride current, membrane binding, and biological activity. Journal of Parasitology 81:286-294. Awolola T.S., O.A. Oduola, C. Strode, L.L. Koekemoer, B. Brooke, and H. Ranson. (2008). Evidence of multiple pyrethroid resistance mechanisms in the malaria vector Anopheles gambiae sensu stricto from Nigeria. Transactions of the Royal Society of Tropical Medicine and Hygiene, doi:10.1016/j.trstmh.2008.08.021 Aziz, M.A., S. Diallo, I.M. Diop, M. Lariviere, and M. Porta (1982). Efficacy and tolerance of ivermectin in human onchocerciasis. Lancet 2:171-173. Bar-Zeev, M. (1960). The location of hygroreceptors and moisture receptors in Aedes aegypti (L.). Entomologia Experimentalis et Applicata 3(3): 251-256. Bates, M. (1940). Oviposition experiments with anopheline mosquitoes. American Journal of Tropical Medicine 20:569-583. 106 Bayoh M.N., and S.W. Lindsay (2003). Effect of temperature on the development of the aquatic stages of Anopheles gambiae sensu stricto (Diptera: Culicidae). Bulletin of Entomological Research 93:375-381. Bayoh, M.N., D.K. Mathias, M.R. Odiere, F.M. Mutuku, L. Kamau, J.E. Gimnig, J.M. Vulule, W.A. Hawley, M.J. Hamel, and E.D. Walker (2010). Anopheles gambiae: historical population decline associated with regional distribution of insecticide-treated bed nets in western Nyanza Province, Kenya. Malaria r Journal 9:62. Beehler, J.W., J.G. Millar, and M.S. Mulla (1993). Synergism between chemical attractants and visual cues influencing oviposition of the mosquito, Culex quinquefasciatus (Diptera: Culicidae). Journal of Chemical Ecology 19(4): 635-644. Beier, J.C., P.V. Perkins, R.A. Wirtz, J. Koros, D. Diggs, T.P. Gargan, D.K. Koech (1988). Bloodmeal identification by direct Enzyme-linked Immunosorbent Assay (ELISA), tested on Anopheles (Diptera: Culicidae) in Kenya. Journal of Medical Entomology 25(1): 9-16. Beier, J.C., J. Keating, J.I. Githure, M.B. MacDonald, D.E. Impoinvil, and R.J. Novac. (2008). Integrated vector management for malaria control. Malaria Journal 7(Suppl 1):S4. Bengone-Ndong, T., M. A. Ba, Y. Kane, I. Sane, J. F. Sutra, M. Alvinerie. (2006). Eprinomectin in dairy zebu Gobra cattle (Bos indicus): plasma kinetics and excretion in milk. Parasitology Research 98: 501-506. Bentley, M.D., I.N. McDaniel, M. Yatagai, H.P. Lee and R. Maynard (1981). Oviposition attractants and stimulants of Aedes triseriatus (Say) (Diptera: Culicidae). Environmental Entomology 10: 186-189. Bentley, M.D. and J.F. Day (1989). Chemical ecology and behavioral aspects of mosquito oviposition. Annual Reviews in Entomology 34:401-421. Benz, G. W., R. A. Roncalli, and S. J. Gross (1989). Use of Ivermectin in Cattle, Sheep, Goats, and Swine, pp. 215-229 In W. C. Campbell [ed.], Ivermectin and Abamectin. Springer-Verlag, New York. Blackwell, A. and S.N. Johnson (2000). Electrophysiological investigation of larval water and potential oviposition chemo-attractants for Anopheles gambiae s.s. Annals of Tropical Medicine and Parasitology 94(4):389-398. Bockarie, M. J., J. L. K. Hii, N. D. E. Alexander, F. Bockarie, H. Dagoro, J. W. Kazura, and M. P. Alpers (1999). Mass treatment with ivermectin for filariasis control in Papua New Guinea: impact on mosquito survival. Medical and Veterinary Entomology 13:120-123. Bøgh, C., E.M. Pedersen, D.A. Mukoko, and J.H. Ouma (1998). Permethrin-impregnated bednet effects on resting and feeding behavior of lymphatic filariasis vector mosquitoes in Kenya. Medical and Veterinary Entomology 12(1): 52-59. 107 Bøgh, C., S.E. Clarke, M. Pinder, F. Sanyang, and S.W. Lindsay (2001). Effect of passive zooprophylaxis on malaria transmission in the Gambia. Journal of Medical Entomology 38(6): 822-828. Bray, A.M. (2003). Ovipositional behavior of Anopheles gambiae as influenced by variable substrates and chemicals. MSc thesis, Michigan State University, USA. Breman, J. G., A. Egan, and G. T. Keusch (2001). The intolerable burden of malaria: a new look at the numbers. American Journal of Tropical Medicine and Hygiene 64:iv-vii. Brown, M.B. and A.B. Forsythe (1974). Journal of the American Statistical Association, 69: 364–367. Byers, J.A. (1987). Novel fraction collector for studying the oviposition rhythm in the turnip moth. Chronobiology International 4(2): 189-194. Causey, O.R., L.M. Deane, and M.P. Deane (1942). Ecology of Anopheles gambiae in Brazil. American Journal of Tropical Medicine 23: 73-94. Chaccour, C., J. Lines, and C.J. Whitty (2010). Effect of ivermectin on Anopheles gambiae mosquitoes fed on humans: the potential of oral insecticides in malaria control. Journal of Infectious Disease 202(1): 113-6. Chadee, D.D., C. Mendis, and J.C. Beier (1993). Diel oviposition periodicity of Anopheline mosquitoes (Diptera: Culicidae) from the Americas: Anopheles albimanus Wieldemann and Anopheles freeborni Aitken. Annals of Tropical Medicine and Parasitology 87(5): 501-507. Chadee, D.D. (1995). Diel oviposition patterns of Anopheles albitarsus in Trinidad, West Indies. Journal of the American Mosquito Control Association 11(1): 103-106. Chaddee, D.D., J.C. Beier, R.T. Mohammed (1998). Laboratory studies of diel oviposition, fecundity, survival and gonotrophic cycles in Anopheles homunculus. Journal of the American Mosquito Control Association 14(2): 153-158. Chaddee, D.D. (2010). The diel oviposition periodicity of Aedes aegypti (L.) (Diptera: Culicidae) in Trinidad, West Indies: effects of forced egg retention. Bulletin of Entomological Research 100(5): 599-603. Chandre, F., F. Darrier, L. Manga, M. Akogbeto, O. Faye, J. Mouchet, and P. Guillet (1999). Status of pyrethroid resistance in Anopheles gambiae sensu lato. Bulletin of the World Health Organization 77:230-234. Chen, H., U. Fillinger, and G. Yan (2006). Oviposition behavior of female Anopheles gambiae in Western Kenya inferred from microsatellite markers. American Journal of Tropical Medicine and Hygiene 75(2): 246-250. Chiu, S. L. and A. Y. H. Lu (1989). Metabolism and Tissue Residues, pp. 131-143 In W. C. Campbell [ed.], 8. Springer-Verlag, New York. 108 Ciordia, H., H.C. McCampbell, G.V. Calvert, and R.E. Plue (1984). Effect of ivermectin on performance of beef cattle on Georgia pastures. American Journal of Veterinary Research 45(11): 2455-7. Clements, A. N. (2000). The Biology of Mosquitoes. Chapman & Hall, New York. Coetzee, M., M. Craig, and D. le Sueur (2000). Distribution of the African Malaria Mosquitoes Belonging to the Anopheles gambiae Complex. Parasitology Today 16:74-77. Cochran, D.G. (1990). Efficacy of abamectin fed to German cockroaches (Dictyoptera: Blattellidae) resistant to pyrethroids. Journal of Economic Entomology 83(4):1243-5. Collins, L.E., and A. Blackwell (1998). Electroantennogram studies of potential oviposition attractants for Toxorhynchites moctezuma and T. amboinensis mosquitoes. Physiological Entomology 23:214-219. Coluzzi, M., A. Sabatini, V. Petrarca, and A. M. Di Deco (1979). Chromosomal differentiation and adaptation to human environments in the Anopheles gambiae complex. Transactions of the Royal Society of Tropical Medicine and Hygiene 73:483-497. Conder, G.A., and W.J. Baker (2002). Chemistry, Pharmacology and Safety: Doramectin and Selamectin, pp. 30-51 In J. Vercruysse and R. S. Rew [eds.], Macrocyclic Lactones in Antiparasitic Therapy. CABI Publishing. Corbet, P.S., and D.D. Chaddee (1990). Incidence and diel pattern of oviposition outdoors of the mosquito, Aedes aegypti (L.) (Diptera: Culicidae) in Trinidad, W.I. in relation to solar aspect. Annals of Tropical Medicine and Parasitology 84(1): 63-78. Cupp, E.W., D.H. Zhang, X. Yue, M.S. Cupp, C. Guyer, T.R. Sprenger, and T.R. Unnasch (2004). Identification of reptilian and amphibian blood meals from mosquitoes in an eastern equine encephalomyelitis virus focus in central Alabama. American Journal of Tropical Medicine and Hygiene 71:272-276. David, J. and P. Fouillet (1973). Enregistrement continu de la ponte chez Drosophila melanogaster et importance des conditions experimentales pour l‟etude du rythme circadian d‟oviposition. Revue du Comportement Animal 7:197-202. Davis, E.E. and M.F. Bowen (1994). Sensory physiological basis for attraction in mosquitoes. Journal of the American Mosquito Control Association 10: 316-325. Diatta, M., A. Spiegel, L. Lochourn, and D. Fontenille (1998). Similar feeding preferences of Anopheles gambiae and A. arabiensis in Senegal. Transactions of the Royal Society of Tropical Medicine and Hygiene 92(3): 270-272. Dimander, S., J. Höglund, A. Uggla, E. Spörndly, and P.J. Waller (2003). Evaluation of gastro-intestinal nematode parasite control strategies for first season grazing cattle in Sweden. Veterinary Parasitology 111(2-3): 193-209. 109 Dourmishev, A. L., L. A. Dourmishev, and R. A. Schwartz (2005). Ivermectin: Pharmacology and Application in Dermatology. International Journal of Dermatology 44:981-988. Duchemin, J.B., J.M., L.P. Tsy, P. Rabarison, J. Roux, M. Coluzzi, and C. Costantini (2001). Zoophily of Anopheles arabiensis and An. gambiae in Madagascar demonstrated by odour-baited entry traps. Medical and Veterinary Entomology 15(1): 50-57. Dukeen, M. Y. H. and S. M. Omer (1986). Ecology of the malaria vector Anopheles arabiensis Patton (Diptera: Culicidae) by the Nile in northern Sudan. Bulletin of Entomological Research 76:451-467. Edman, J.D. and D. J. Taylor (1968). Culex nigripalpus: seasonal shift in the bird-mammal feeding ratio in a mosquito vector of human encephalitis. Science 161: 67-68. Edwards, G., A. Dingsdale, N. Helsby, M. L. Orme, and A. M. Breckenridge (1988). The Relative Systematic Availability of Ivermectin After Administration as Capsule, Tablet, and Oral Solution. European Journal of Clinical Pharmacology 35:681-684. Edwards, G. (2003). Ivermectin: does P-glycoprotein play a role in neurotoxicity? Filaria Journal 2: doi:10.1186/1475-2883-2-S1-S8. Escudero, E., C. M. Carceles, M. S. Diaz, J. F. Sutra, P. Galtier, and M. Alvinerie (1999). Pharmacokinetics of moxidectin and doramectin in goats. Research in Veterinary Science 67:177-181. Feng, X., J. Hayashi, R. N. Beech, and R. K. Prichard (2002). Study of the nematode putative GABA type-A receptor subunits: evidence for modulation by ivermectin. Journal of Neurochemistry 83:870-878. Fink, D. W. and A. G. Porras (1989). Pharmacokinetics of Ivermectin in Animals and Humans, pp. 113-130 In W. C. Campbell [ed.], Ivermectin and Abamectin. Springer-Verlag, New York. Fleugel, W. (1978). Oviposition rhythm of individual Drosophila melanogaster. Experientia 34: 65-66. Fleugel ,W. (1984). Oviposition rhythm of individual Drosophila pseudoobscura. Experientia 40: 250-252. Foley, D.H., J.H. Bryan, and G.W. Lawrence (2000). The potential of ivermectin to control the malaria vector Anopheles farauti. Transactions of the Royal Society of Tropical Medicine and Hygiene 94:625-628. Fornadel, C.M., L.C. Norris, G.E. Glass, and D.E. Norris (2010). Analysis of Anopheles arabiensis blood feeding behavior in southern Zambia during the two years after introduction of insecticide-treated bed nets. Am erican Journal of Tropical Medicine and Hygiene 83(4): 848-853. 110 Fox, L. E. and P. E. Lloyd (1999). Glutamate is a fast excitatory transmitter at some buccal neuromuscular synapses in Aplysia. Journal of Neurophysiology 82:1477-1488. Frank, J.H. (1985). Use of an artificial bromeliad to show the importance of color value in restricting colonization of bromeliads to Aedes aegypti and Culex quinquefasciatus. Journal of the American Mosquito Control Association 1(1): 28-32. Fritz, L. C., C. C. Wang, and A. Gorio (1979). Avermectin B1a irreversibly blocks postsynaptic potentials at the lobster neuromuscular junction by reducing muscle membrane resistance. Proceedings of the National Academy of Sciences, USA 76:2062-2066. Fritz, M.L., J. Huang, E.D. Walker, M.N. Bayoh, J. Vulule, and J.R. Miller (2008). Ovipositional periodicity of caged Anopheles gambiae individuals. Journal of Circadian Rhythms 6:2. Fritz, M.L., P.Y. Siegert, E.D. Walker, M.N. Bayoh, J.R. Vulule, and J.R. Miller. (2009). Toxicity of bloodmeals from ivermectin-treated cattle to Anopheles gambiae s.l. Annals of Tropical Medicine and Parasitology 103(6):539-47. Gallup, J. L. and J. D. Sachs (2001). The economic burden of malaria. American Journal of Tropical Medicine and Hygiene 64:85-96. Gardner, K., M. V. Meisch, C. L. Meek, and W. S. Biven (1993). Effects of Ivermectin in Canine Blood on Anopheles quadrimaculatus, Aedes albopictus, and Culex salinarius. Journal of the American Mosquito Control Association 9:400-402. Garrett-Jones, C. (1964a). The human blood index of malaria vectors in relation to epidemiological assessment. Bulletin of the World Health Organization 30: 241-261. Garrett-Jones, C. (1964b). Prognosis for the interruption of malaria transmission through assessment of the mosquito's vectorial capacity. Nature 24:1173-1175. Gary, R. E. and W. A. Foster (2001). Effects of Available Sugar on the Reproductive Fitness and Vectorial Capacity of the Malaria Vector Anopheles gambiae (Diptera: Culicidae). Journal of Medical Entomology 38:22-28. Gibbs, H.C., and R.P. Herd (1986). Nematodiasis in cattle. Importance, species involved, immunity, and resistance. Veterinary Clinics of North America Large Animal Practice 2:211-224. Gibson, G. (1995). A behavioural test of the sensitivity of a nocturnal mosquito, Anopheles gambiae, to dim white, red and infra-red light. Physiological Entomology 20:224-228. Giglioli, M.E.C. (1965). Oviposition by Anopheles melas and its effect on egg survival during the dry season in the Gambia, West Africa. Annals of the Entomological Society of America 58(6): 885-891. Gillies, M.T., and B. De Meillon (1968). The Anophelinae of Africa South of the Sahara. Publication of the South African Institute for Medical Research No. 54, 343pp. 111 Gimnig, J.E., M. Ombok, L. Kamau, and W.A. Hawley (2001). Characteristics of larval anopheline (Diptera: Culicidae) habitats in western Kenya. Journal of Medical Entomology 38(2): 282-288. Gimnig, J. E., M. Ombok, S. Otieno, M. G. Kaufman, J. M. Vulule, and E. D. Walker (2002). Density-dependent development of Anopheles gambiae (Diptera: Culicidae) larvae in artificial habitats. Journal of Medical Entomology 39:162-172. Gimnig, J. E., J. M. Vulule, T. Q. Lo, L. Kamau, M. S. Kolczak, P. A. Phillips-Howard, E. M. Mathenge, F. O. Ter Kuile, B. L. Nahlen, A. W. Hightower, and W. A. Hawley (2003). Impact of permethrin-treated bed nets on entomologic indices in an area of intense year-round malaria transmission. American Journal of Tropical Medicine and Hygiene 68:1622. Githeko, A.K., M.W. Service, C.M. Mbogo, F.K. Atieli, and F.O. Juma (1994). Origin of blood meals in indoor and outdoor resting malaria vectors in western Kenya. Acta Tropica 58: 307-316. Githeko, A. K., N. I. Adungo, D. M. Karanja, W. A. Hawley, J. M. Vulule, I. K. Seroney, A. V. O. Ofulla, F. K. Atieli, S. O. Ondijo, I. O. Genga, P. K. Odada, P. A. Situbi, and J. A. Oloo (1996). Some observations on the biting behavior of Anopheles gambiae s.s., Anopheles arabiensis, and Anopheles funestus and their implications for malaria control. Experimental Parasitology 82:306-315. Goudie, A.C., N.A. Evans, K.A. Gration, B.F. Bishop, S.P. Gibson, K.S. Holdom, B. Kaye, S.R. Wicks, D. Lewis, A.J. Weatherley, et al. (1993). Doramectin – a potent novel endectocide. Veterinary Parasitology 49(1): 5-15. Gray, E. M. and T. J. Bradley (2005). Physiology of desiccation resistance in Anopheles gambiae and Anopheles arabiensis. American Journal of Tropical Medicine and Hygiene 73:553-559. Greene, B. M., K. R. Brown, and H. R. Taylor (1989). Use of Ivermectin in Humans, pp. 311-323 In W. C. Campbell [ed.], Ivermectin and Abamectin. Springer-Verlag, New York. Haddow, A.J. and J.D. Gillett (1957). Observations on the oviposition-cycle of Aedes (Stegomyia) aegypti (Linnaeus). Annals of Tropical Medicine and Parasitology 51: 159-169. Haddow, A.J., and Y. Ssenkubuge (1962). Laboratory observations on the ovipositioncycle in the mosquito Anopheles (Cellia) gambiae Giles. Annals of Tropical Medicine and Parasitology 56: 352-355. Hamer, G. L., U. D. Kitron, J. D. Brawn, S. R. Loss, M. O. Ruiz, T. L. Goldberg, and E. D. Walker (2008). Culex pipiens (Diptera: Culicidae): A Bridge Vector of West Nile Virus to Humans. Journal of Medical Entomology 45:125-128. Hamer, G.L., U.D. Kitron, T.L. Goldberg, J.D. Brawn, S.R. Loss, M.O. Ruiz, D.B. Hayes, and E.D. Walker (2009). Host selection by Culex pipiens mosquitoes and West Nile Virus amplification. American Journal of Tropical Medicine and Hygiene 80(2): 268-278. 112 Harris, M.O., and J.R. Miller (1982). Synergism of visual and chemical stimuli in the th oviposition behavior of Delia antiqua. Proceedings of the 5 International Symposium on Insect-Plant Relationships, Wageningen (ed. by J.H. Visser and A.K. Minks), pp. 117-122. Harris, M.O., and J.R. Miller (1984). Foliar form influences ovipositional behavior of the onion fly. Physiological Entomology 9:145-155. Harris, M.O. (1987). Responses of ovipositing flies to authentic and surrogate onions. th Proceedings of the 6 International Symposium on Insect-Plant Relationships, p. 392. Junk, Pau, France. Havukkala, I.J., and J.R. Miller (1987). Daily Periodicity in the Ovipositional Behavior of the Onion Fly, Delia antiqua (Diptera: Anthomyiidae). Environmental Entomology 16: 4144. Hawley, W. A., P. A. Phillips-Howard, F. O. Ter Kuile, D. J. Terlouw, J. M. Vulule, M. Ombok, B. L. Nahlen, J. E. Gimnig, S. K. Kariuki, M. S. Kolczak, and A. W. Hightower (2003). Community-wide effects of permethrin-treated bed nets on child mortality and malaria morbidity in Western Kenya. American Journal of Tropical Medicine and Hygiene 68:121-127. Haynes, K.F. (1988). Sublethal effects of neurotoxic insecticides on insect behavior. Annual Review of Entomology 33:149-168. Hewitt, S., and M. Rowland (1999). Control of zoophilic malaria vectors by applying pyrethroid insecticides to cattle. Tropical Medicine and International Health 4(7): 481-486. Highton, R.B., J.H. Bryan, P.F.L. Boreham, and J.A. Chandler. (1979). Studies on the sibling species Anopheles gambiae Giles and Anopheles arabiensis Patton (Diptera: Culicidae) in the Kisumu area, Kenya. Bulletin of Entomological Research 69:43–53. Hightower, A., R. Kiptui, A. Manya, A. Wolkon, J.L. Vanden Eng, M. Hamel, A. Noor, S.K. Sharif, R. Buluma, J. Vulule, K. Laserson, L. Slutsker, and W. Akhwale (2010). Bed net ownership in Kenya: the impact of 3.4 million free bed nets. Malaria Journal 9:183. Höckelmann, C. (2004). Odor compounds from cyanobacterial biofilms acting as attractants and repellents for free-living nematodes. Limnology and Oceanography 49: 1809-1819. Hocking, K.S. and D.G. MacInnes (1948). Notes on the bionomics of Anopheles gambiae and A.funestus in East Africa. Bulletin of Entomological Research 39:453-465. Hollingworth, R. M. (2006). Physiological Targeting of Insecticides and Acaricides, In J. N. All and M. F. Treacy [eds.], Use and Management of Insecticides, Acaricides, and Transgenic Crops. Entomological Society of America. Holstein, M. H. (1954). Biology of Anopheles gambiae: Research in French West Africa. Geneva, World Health Organization. Monograph Series No. 9, 172 pp. 113 Huang, J., E.D. Walker, P.Y. Giroux, J. Vulule, and J.R. Miller (2005). Ovipositional site selection by Anopheles gambiae: influences of substrate moisture and texture. Medical and Veterinary Entomology 19: 442-450. Huang, J., E.D. Walker, P.E. Otienoburu, F. Amimo, J. Vulule, and J.R. Miller (2006a). Laboratory tests of oviposition by the African malaria mosquito, Anopheles gambiae, on dark soil as influenced by presence or absence of vegetation. Malaria Journal 5:88. Huang, J., J.R. Miller, S. Chen, J.M. Vulule, and E.D. Walker (2006b). Anopheles gambiae (Diptera: Culicidae) oviposition in response to agarose media and cultured bacterial volatiles. Journal of Medical Entomology 43(3): 498-504. Huang, J., E.D. Walker, J.Vulule, and J.R.Miller (2007). The influence of darkness and visual contrast on oviposition by Anopheles gambiae in moist and dry substrates. Physiological Entomology 32: 34-40. Iakubovich, V.I., N.F. Zakharova, A.N. Alekseev, and E.A. Alekseev. (1989). Evaluation of the action of ivermectin on blood-sucking mosquitoes. Medicinskaya Parazitologia Moskva 3:60-64. Jones, M.D.R., C.M. Cubbin, and D. Marsh (1972). Light-on effects and the question of bimodality in the circadian flight activity of the mosquito Anopheles gambiae. Journal of Experimental Biology 57: 347-357. Jones, M.D.R., and S.J. Gubbins (1978). Changes in the circadian flight activity of the mosquito Anopheles gambiae in relation to insemination, feeding and oviposition. Physiological Entomology 3: 213-220. Jones, R.M., N.B. Logan, A.J. Weatherley, A.S. Little, and C.D. Smothers (1993). Activity of doramectin against nematode endoparasites of cattle. Veterinary Parasitology 49(1): 27-37. Joshi, G. P., M. W. Service, and G. D. Pradhan (1975). A survey of species A and B of the Anopheles gambiae Giles complex in the Kisumu area of Kenya prior to insecticidal spraying with OMS-43 (fenitrothion). Annals of Tropical Medicine and Parasitology 69:91-103. Kass, I. S., C. C. Wang, J. P. Walrond, and A. O. W. Stretton (1980). Avermectin B1a, a paralyzing anthelmintic that affects interneurons and inhibitory motoneurons in Ascaris. Proceedings of the National Academy of Sciences, USA 77:6211-6215. Kent, R.J., and D.E. Norris (2005). Identification of mammalian blood meals of mosquitoes by a multiplexed polymerase chain reaction targeting cytochrome b. American Journal of Tropical Medicine and Hygiene 73: 336–342. Kent, R.J., P. Thuma, S. Mharakurwa, and D.E. Norris (2007). Seasonality, bloodfeeding behavior and transmission of Plasmodium falciparum by Anopheles arabiensis following an extended drought in southern Zambia. American Journal of Tropical Medicine and Hygiene 76: 267–274. 114 Kent, R. (2009). Molecular methods for arthropod bloodmeal identification and applications to ecological and vector-borne disease studies. Molecular Ecology Reviews 9:418. Killeen, G.F., and T.A. Smith (2007). Exploring the contributions of bed nets, cattle, insecticides and excitorepellency to malaria control: a deterministic model of mosquito hostseeking behavior and mortality. Transactions of the Royal Society of Tropical Medicine and Hygiene 101(9-4): 867-880. Kilpatrick, A.M., L.D. Kramer, M.J. Jones, P.P. Marra, and P. Daszak (2006). West Nile Virus epidemics in North America are driven by shifts in mosquito feeding behavior. PLoS Biology 4(4): e82. Lanusse, C., A. Lifschitz, G. Virkel, L. Alvarez, S. Sanchez, J. F. Sutra, P. Galtier, and M. Alvinerie (1997). Comparative plasma disposition kinetics of ivermectin, moxidectin and doramectin in cattle. Journal of Veterinary Pharmacology and Therapeutics 20:91-99. Lee, J.H., H. Hassan, G. Hill, E.W. Cupp, T.B. Higazi, C.J. Mitchell, M.S. Godsey, and T.R. Unnasch (2002). Identification of mosquito avian-derived blood meals by polymerase chain reaction heteroduplex analysis. American Journal of Tropical Medicine and Hygiene 66:599-604. Lefèvre, T., L. Gouagna, K.R. Dabiré, E. Elguero, D. Fontenille, F. Renaud, C. Costantini, and F. Thomas (2009). Beyond nature and nurture: Phenotypic plasticity in blood-feeding behavior of Anopheles gambiae s.s. when humans are not readily accessible. American Journal of Tropical Medicine and Hygiene 81(6): 1023-1029. Le Menach, A., F.E. McKenzie, A. Flahault, and D.L. Smith (2005). The unexpected importance of mosquito oviposition behavior for malaria: non-productive larval habitats can be sources for malaria transmission. Malaria Journal 4:23. Lengeler, C. (2004). Insecticide-treated bed nets and curtains for preventing malaria. Cochrane Database of Systematic Reviews. Li, J., T. Deng, H. Li, L. Chen, and J. Mo (2009). Effects of water color and chemical compounds on the oviposition behavior of gravid Culex pipiens pallens females under laboratory conditions. Journal of Agricultural and Urban Entomology 26(1): 23-30. Lindblade, K. A., T. P. Eisele, J. E. Gimnig, J. A. Alaii, F. Odhiambo, F. O. ter Kuile, W. A. Hawley, K. A. Wannemuehler, P. A. Phillips-Howard, D. H. Rosen, B. L. Nahlen, D.J. Terlouw, K. Adazu, J. M. Vulule, and L. Slutsker (2004). Sustainability of reductions in malaria transmission and infant mortality in western Kenya with use of insecticide-treated bednets: 4 to 6 years of follow-up. Journal of the American Medical Association 291: 25712580. Martin, R.J., and A.J. Pennington (1988). Effect of dihydroavermectin-b1a on CI singlechannel currents in Ascaris. Pesticide Science 24:90-91. 115 Martin, R. J., A. P. Robertson, and A. J. Wolstenholme (2002). Mode of Action of the Macrocyclic Lactones, pp. 125-140 In J. Vercruysse and R. S. Rew [eds.], Macrocyclic Lactones in Antiparasitic Therapy. CAB International. McCall, P.J., and D.W. Kelly (2002). Learning and memory in disease vectors. Trends in Parasitology 18(10): 429-433. McCrae, A.W.R. (1983). Oviposition by African malaria vector mosquitoes I. Temporal activity patterns of caged, wild-caught, freshwater Anopheles gambiae Giles sensu lato. Annals of Tropical Medicine and Parasitology 77: 615-625. McCrae, A.W.R. (1984). Oviposition by African malaria vector mosquitoes – II. Effects of site tone, water type and conspecific immature on target selection by freshwater Anopheles gambiae Giles, sensu lato. Annals of Tropical Medicine and Parasitology 78(3):307-318. McKellar, Q. A. and H. A. Benchaoui (1996). Avermectins and milbemycins. Journal of Veterinary Pharmacology and Therapeutics 19:331-351. Mehlhorn, H. (2008). Encyclopedia of Parasitology, Volume 1. pp. 874 Springer-Verlag, New York. Mendis, C. , J. L. Jacobsen , A. Gamage-Mendis , E. Bule, M. Dgedge, R. Thompson, N. Cuamba , J. Barreto, K. Begtrup, R. E. Sinden, and B. Høgh. (2000). Anopheles arabiensis and An. funestus are equally important vectors of malaria in Matola coastal suburb of Maputo, southern Mozambique. Medical and Veterinary Entomology 14(2): 171-180. Millar, J.G., J.D. Chaney, and M.S. Mulla (1992). Identification of oviposition attractants for Culex quinquefasciatus from fermented bermuda grass infusions. Journal of the American Mosquito Control Association 8:11-17. Miller, J.R., J. Huang, J. Vulule, and E.D. Walker (2007). Life on the edge: African malaria mosquito (Anopheles gambiae s. l.) larvae are amphibious. Naturwissenschaften 94(3): 195-199. Minakawa, N., C.M. Mutero, J.I. Githure, J.C. Beier, and G. Yan (1999). Spatial distribution and habitat characterization of Anopheline mosquito larvae in Western Kenya. American Journal of Tropical Medicine and Hygiene 61: 1010-1016. Minakawa, N.,G. Sonye, M. Mogi, and G. Yan. (2004). Habitat characteristics of Anopheles gambiae s.s. larvae in a Kenyan highland. Medical and Veterinary Entomology 18:301-305. Molaei, G., J. Oliver, T.G. Andreadis, P.M. Armstrong, and J.J. Howard (2006a). Molecular identification of blood-meal sources in Culiseta melanura and Culiseta morsitans from and endemic focus of Eastern Equine Encephalitis virus in New York. Am erican Journal of Tropical Medicine and Hygiene 75(6): 1140-1147. Molaei, G., T.G. Andreadis, P.M. Armstrong, J.F. Anderson, and C.R. Vossbrinck (2006b). Host feeding patterns of Culex mosquitoes and West Nile Virus transmission, Northeastern United States. Emerging Infectious Diseases 12(3): 468-474. 116 Molineaux, L. and G. Gramiccia (1980). The Garki Project - Research on the epidemiology and control of malaria in the Sudan savanna of West Africa. Geneva, World Health Organization. Ref Type: Serial (Book,Monograph) Molyneux, D. H. (2004). "Neglected" diseases but unrecognised successes - challenges and opportunities for infectious disease control. Lancet 364:380-383. Muirhead Thomson, R.C. (1940). Studies on the behaviour of Anopheles minimus. Part I. The selection of the breeding place and the influence of light and shade. Journal of the Malaria Institute of India 3: 265-294. Muirhead Thomson, R.C. (1951). Mosquito Behaviour in Relation to Malaria Transmission and Control in the Tropics. London: Edward Arnold & Co. Müller P., E. Warr, B.J. Stevenson, P.M. Pignatelli, J.C. Morgan, A. Steven, A.E. Yawson, S.N. Mitchell, H. Ranson, J. Hemingway, M.J.I. Paine, and J.J. Donnelly. (2008). Field-caught permethrin-resistant Anopheles gambiae overexpress CYP6P3, a P450 that metabolizes pyrethroids. PLoS Genetics 4(11):e1000286. doi:10.1371/journal.pgen.1000286 Munga, S., N. Minakawa, G. Zhou, O.J. Barrack, A.K. Githeko, and G. Yan (2005). Oviposition site preference and egg hatchability of Anopheles gambiae: Effects of land cover types. Journal of Medical Entomology 42(6): 993-997. Munhenga, G., H. T. Masendu, B. D. Brooke, R. H. Hunt, and L. K. Koekemoer. (2008). Pyrethroid resistance in the major malaria vector Anopheles arabiensis from Gwave, a malaria-endemic area in Zimbabwe. Malaria Journal 7:247 Mutuku, F.M., J.A. Alaii, M.N. Bayoh, J.E. Gimnig, J. Vulule, E.D. Walker, E. Kabiru, and W.A. Hawley (2006a). Distribution, description and local knowledge of larval habitats of Anopheles gambiae s.l. in a village in Western Kenya. American Journal of Tropical Medicine and Hygiene 74(1): 44-53. Mutuku, F.M., M.N. Bayoh, J.E. Gimnig, J. Vulule, L. Kamau, E.D. Walker, E. Kabiru, and W.A. Hawley (2006b). Pupal habitat productivity of Anopheles gambiae complex mosquitoes in a rural village in Western Kenya. American Journal of Tropical Medicine and Hygiene 74(1): 54-61. Mwangangi, J.M., C.M. Mbogo, J.G. Nzovu, J.I.Githure, G. Yan, and J.C. Beier (2003). Blood-meal analysis for anopheline mosquitoes sampled along the Kenyan coast. Journal of the American Mosquito Control Association 19(4): 371-375. Nabarro, D. N. and E. M. Tayler (1998). The "Roll Back Malaria" Campaign. Science 280:2067-2068. Nation, J. L. (2002). Insect Physiology and Biochemistry, First ed. CRC Press, Boca Raton. 117 Noor, A. M., A. A. Amin, W. S. Akhwale, and W. F. Snow (2007). Increasing Coverage and Decreasing Inequity in Insecticide-Treated Bed Net Use amon g Rural Kenyan Children. PLoS Medicine 4:1341-1347. Ochumba, P.B.O., and D.I. Kibaara (1989). Observations on blue-green algal blooms in the open waters of Lake Victoria, Kenya. African Journal of Ecology 27: 23-34. Odiere, M., M. N. Bayoh, J. Gimnig, J. Vulule, L. Irungu, and E. Walker (2007). Sampling outdoor, resting Anopheles gambiae and other mosquitoes (Diptera: Culicidae) in Western Kenya with clay pots. Journal of Medical Entomology 44:14-22. Omlin, F.X., C. Carlson, B. Ogbunugafor, and A. Hassanali (2007). Anopheles gambiae exploits the treehole ecosystem in Western Kenya: A new urban malaria risk? American Journal of Tropical Medicine and Hygiene 77(Suppl 6): 264-269. Omumbo, J., J. Ouma, B. Rapouda, M. H. Craig, D. le Sueur, and R. W. Snow (1998). Mapping malaria transmission intensity using geographical information systems (GIS): an example from Kenya. Annals of Tropical Medicine and Parasitology 92:7-21. Otienoburu, P.E., M.N. Bayoh, J. Gimnig, J. Huang, E.D. Walker, M.E. Otieno, J. Vulule, and J.R. Miller (2007). Anopheles gambiae (Diptera: Culicidae) oviposition as influenced by type of water infused into black and red soils of western Kenya. International Journal of Tropical Insect Science 27(1):2-5. Overgaard, H.E. (2007). Effect of plant structure on oviposition behavior of Anopheles minimus s.l. Journal of Vector Ecology 32(2): 193-197. Pampana, E. (1963). A textbook of malaria eradication. Oxford University Press, London. Pampiglione, S., G. Majori, G. Petrangeli, and R. Romi (1985). Avermectins, MK-933 and MK-936, for mosquito control. Transactions of the Royal Society for Tropical Medicine and Hygiene 79: 797-799. Parmar, M.K.B., and D. Machin. (1995). Survival Analysis: A practical approach. Chichester, John Wiley and Sons LTD. Paskewitz, S. M. and F. H. Collins (1990). Use of the polymerase chain reaction to identify mosquito species of the Anopheles gambiae complex. Medical and Veterinary Entomology 4:367-373. Patz, J. A. and S. H. Olson (2006). Malaria risk and temperature: Influences from global climate change and local land use practices. Proceedings of the National Academy of Sciences, USA 103:5635-5636. Petrarca, V. , J.C. Beier, F. Onyango, J. Koros, C. Asiago, D.K. Koech, and C.R. Roberts (1994). Species composition of the Anopheles gambiae complex (Diptera: Culicidae) at two sites in Western Kenya. Journal of Medical Entomology 28(3): 307-313. 118 Ponnusamy, L., N. Xu, S. Nojima, D.M. Wesson, C. Schal, and C.S. Apperson (2008). Identification of bacteria and bacteria-associated chemical cues that mediate oviposition site preferences by Aedes aegypti. Proceedings of the National Academies of Sciences 105(27): 9262-9267. Prokopy, R.J., and E.D. Owens (1983). Visual detection of plants by herbivorous insects. Annual Review of Entomology 28:337-364. Pulliam, J.D. and J.M. Preston (1989). Safety of Ivermectin in Target Animals, pp. 149161. In W. C. Campbell [ed.], Ivermectin and Abamectin. Springer-Verlag, New York. R Development Core Team. (2008). R: A language and environment for statistical computing. R Foundation for Statistical Computing,Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org. Robert V., V. Petrarca, M. Coluzzi, C. Boudin, and P. Carnevale (1991). Etude des taux de parturité et d‟infection du complexe Anopheles gambiae dans la rizière de la vallée du Kou, Burkina Faso. Robert V, Chippaux JF, Diomandé L, eds. Le Paludisme en Afrique de l’Ouest: Études Entomologiques et Épidémiologiquesen Zone Rizicole et en Milieu Urbain , 17–35. Rock, D. W., R. L. Delay, and M. J. Gliddon (2002). Chemistry, Pharmacology and Safety: Moxidectin, pp. 75-96 In J. Vercruysse and R. S. Rew [eds.], Macrocyclic Lactones in Antiparasitic Therapy. CABI Publishing. Roessingh, P., and E. Städler (1990). Foliar form, colour and surface characteristics influence oviposition behaviour in the cabbage root fly Delia radicum. Entomologia Experimentalis et Applicata 57:93-100. Roncalli, R.A., and C. Benitez Usher. (1988). Efficacy of ivermectin against Dermatobia hominis in cattle. Veterinary Parasitology 28(4): 343-6. Rozendaal, J. A. (1997). Vector Control: Methods for use by individuals and communities. Geneva, World Health Organization. Ref Type: Serial (Book,Monograph) Rubio-Palis, Y., C.F. Curtis, C. Gonzales, and R.A. Wirtz (1994). Host choice of anopheline mosquitoes in a malaria endemic area of western Venezuela. Medical and Veterinary Entomology 8(3):275-280. Rugg , D., N. Sales , A. C. Kotze, and G. W. Levot. (2007). Susceptibility to ivermectin of pyrethroid-resistant Lucilia cuprina (Wiedemann) (Diptera: Calliphoridae). Australian Journal of Entomology 34(1): 69 – 70. Sachs, J. D. and P. Malaney (2002). The economic and social burden of malaria. Nature 415:680-685. Scott, T.W., E. Chow, D. Strickman, P. Kittayapong, R.A. Wirtz, L.H. Lorenz, and J.D. Edman (1993). Blood-feeding patterns of Aedes aegypti (Diptera: Culicidae) collected in a rural Thai village. Journal of Medical Entomology 30(5): 922-927. 119 Service, M., G.P. Joshi, and G.D. Pradhan (1978). A survey of Anopheles gambiae (species A) and An. arabiensis (species B) of the An. gambiae Giles complex in the Kisumu area of Kenya following insecticidal spraying with OMS-43 (fenitrothion). Annals of Tropical Medicine and Parasitology 72: 377-386. Shoop, W.L., J.R. Egerton, C.H. Eary, H.W. Haines, B.F. Michael, H. Mrozik, P. Eskola, M.H. Fisher, L. Slayton, D.A. Ostlind, B.J. Skelly, R.K. Fulton, D. Barth, S. Costa, L.M. Gregory, W.C. Campbell, R.L. Seward, and M.J. Turner (1996). Eprinomectin: A novel avermectin for use as a topical endectocide for cattle. International Journal for Parasitology 26(11): 1237-1242. Shoop, W., and M. Soll (2002). Ivermectin, Abamectin and Eprinomectin, pp. 1-29 In J. Vercruysse and R. S. Rew [eds.], Macrocyclic Lactones in Antiparasitic Therapy. CABI Publishing. Siegert, P.Y., E. Walker, and J.R. Miller (2009). Differential behavioral responses of Anopheles gambiae (Diptera: Culicidae) modulate mortality caused by pyrethroid-treated bednets. Journal of Economic Entomology 102(6): 2061-71. Sithiprasasna, R., P. Mahapibul, C. Noigamol, M.J. Perich, B.C. Zeichner, B. Burge, S.L. Norris, J.W. Jones, S.S. Schleich, and R.E. Coleman (2003). Field evaluation of a lethal ovitrap for the control of Aedes aegypti (Diptera: Culicidae) in Thailand. Journal of Medical Entomology 40(4): 455-462. Snow, W.F. (1971). The spectral sensitivity of Aedes aegypti (L.) at oviposition. Bulletin of Entomological Research 60: 683-696. Spradbery, J.P. (1979). Daily oviposition activity and its adaptive significance in the screwworm fly, Chrysomya bezziana (Diptera: Calliphoridae). Journal of the Australian Entomological Society 18: 63-66. Stahl, P.D. and T.B. Parkin (1996). Microbial production of volatile organic compounds in soil microcosms. Soil Science Society of America Journal 60: 821-828. Suleman, M., and M. Shirin (1981). Laboratory studies on oviposition behavior of Culex quinquefasciatus Say (Diptera: Culicidae): choice of oviposition medium on oviposition cycle. Bulletin of Entomological Research 71: 361-369. Sumba, L.A., K. Okoth, A.L.Deng, J. Githure, B.G.J. Knols, J.C. Beier, and A. Hassanali (2004). Daily oviposition patterns of the African malaria mosquito Anopheles gambiae Giles (Diptera: Culicidae) on different types of aqueous substrates. Journal of Circadian Rhythms 2: 6-12. Symondson, W.O.C. (2002). Molecular identification of prey in predator diets. Mol Ecol 11: 627-641. Takken, W., and B.G.J. Knols (1999). Odor-mediated behaviour of Afrotropical malaria mosquitoes. Annual Review of Entomology 44:131-157. 120 Tesh, R. B. and H. Guzman (1990). Mortality and infertility in adult mosquitoes after the ingestion of blood containing ivermectin. American Journal of Tropical Medicine and Hygiene 43:229-233. Tichy, H. and R. Loftus (1996). Hygroreceptors in insects and a spider: humidity transduction models. Naturwissenschaften 83: 255–263. Tichy, H. and W. Kallina (2010). Insect hygroreceptor responses to continuous changes in humidity and air pressure. Journal of Neurophysiology 103(6): 3274-3286. Tirados, I., C. Costantini, G. Gibson, and S. J. Torr (2006). Blood-feeding behaviour of the malarial mosquito Anopheles arabiensis: implications for vector control. Medical and Veterinary Entomology 20:425-437. Touré, Y. (2001). Malaria vector control in Africa: Strategies and challenges. 2001 AAAS Annual Meeting, San Francisco, American Association for the Advancement of Science. Malaria in Africa: Emerging Prevention and Control Strategies. Ref Type: Report Toutain, P.L., D.W. Upson, T.N. Terhune, and M.E. McKenzie (1997). Comparative pharmacokinetics of doramectin and ivermectin in cattle. Veterinary Parasitology 72(1): 3-8. Trape, J., E. Lefebvre-Zante, F. Legros, G. Ndiaye, H. Bouganali, P. Druilhe, and G. Salem. (1992). Vector density gradients and the epidemiology of urban malaria in Dakar, Senegal. American Journal of Tropical Medicine and Hygiene 47(2):181-189. Voh, A.A. and E.O. Otchere. (1988). Reproductive performance of Zebu cattle under traditional agropastoral management in northern Nigeria. Animal Reproduction Science 19(3-4): 191-203. Walker, E. D., A. R. Thibault, A. P. Thelen, B. A. Bullard, J. Huang, M. R. Odiere, M. N. Bayoh, E. E. Wilkins, and J. M. Vulule (2007). Identification of field caught Anopheles gambiae s.s. and Anopheles arabiensis by Taqman single nucleotide polymorphism genotyping. Malaria Journal 6:23. White, G. B. and P. Rosen (1973). Comparative studies on sibling species of Anopheles gambiae Giles complex (Dipt, Culicidae). II Ecology of species A and B in savanna around Kaduna, Nigeria, during transition from wet to dry season. Bulletin of Entomological Research 62:613-625. White, G. B. (1974). Anopheles gambiae complex and disease transmission in Africa. Transactions of the Royal Society of Tropical Medicine and Hygiene 68:278-298. Williams, R.E. (1962). Effect of coloring oviposition media with regard to the mosquito Aedes triseriatus (Say). The Journal of Parasitology 48(6): 919-925. Wilson, M. L. (1993). Avermectins in Arthropod Vector Management - Prospects and Pitfalls. Parasitology Today 9:83-86. 121 Wolstenholme, A. J. and A. T. Rogers (2005). Glutamate-gated chloride channels and the mode of action of the avermectin/milbemycin anthelmintics. Parasitology 131:85-95. Yazwinski, T. A., J. C. Williams, L. L. Smith, C. Tucker, A. F. Loyacano, A. DeRosa, P. Peterson, D. J. Bruer, and R. L. Delay (2006). Dose determination of the persistent activity of moxidectin long-acting injectable formulations against various nematode species in cattle. Veterinary Parasitology 137:273-285. Zucker, J. R. (1996). Changing patterns of autochthonous malaria transmission in the United States: A review of recent outbreaks. Emerging Infectious Diseases 2:37-43. 122