Diptera: Culicidae)

Foraging Behavior and Reproductive Success of the Malaria Mosquito Anopheles gambiae s.s. (Diptera: Culicidae)

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Christopher M. Stone. M.S.

Graduate Program in Entomology

The Ohio State University

2011

Dissertation Committee:

Woodbridge A. Foster, Advisor

David Denlinger

Ian M. Hamilton

Christopher M. Stone

2011

Abstract

The malaria mosquito Anopheles gambiae Giles s.s. has a diet consisting of nectar meals taken from plants and blood from warm-blooded animals, particularly humans.

Foraging theory predicts that diets include only those items that maximize energetic intake (a proxy for fitness), yet previous studies indicate that for mosquitoes specializing on human blood, such as the yellow fever mosquito, Aedes aegypti, and An. gambiae, sugar has a negative effect on their fitness. The objective of this dissertation was to find how the presence of sugar- and blood-hosts in the environment affect the foraging decisions made by An. gambiae mosquitoes, their reproductive success, and their potential to transmit malaria.

The experiments reported on in this dissertation were conducted to study the mosquito in a more (―semi‖-)natural environment than more commonly used laboratory cage environments, to more aptly reflect the energetic expenditures that come with meal-, mate-, and oviposition-site seeking behaviour in nature. Chapter 1 gives a description of the mesocosm, and develops the rationale for using such set-ups.

The importance of male mosquitoes to population processes has long been overlooked, but their mating capability is strongly dependent upon access to sugar sources. Chapter 2 investigates the insemination rates of females per the presence or absence of environmental sugar; a matrix population model suggested that An. gambiae

iii

populations will not be viable in environments devoid of sugar sources, due to the hampering of male mating ability.

Rather than maximizing energy intake and thereby directly increasing fitness, female sugar-feeding may reflect a behavioural constraint, e.g., finding and locating a swarm of males and selecting a mate may require energy best provided by a sugar meal, or sugar may be required to enable (blood-)host seeking. Thus, in Chapter 3 the sugar- feeding behaviour of females is placed in a behavioural context. Results indicated that females may take either a blood or a sugar meal as their first meal before mating and subsequently seek a blood meal.

In Chapter 4 the opportunistic nature of the first meal of this species is investigated further, by altering nectar-bearing plant abundance and blood host accessibility, and assessing the meal choices of large- and small-bodied 1-d-old mosquitoes. With unrestricted access to a human throughout the night, blood was the preferred meal type. When access to a human was limited by the use of a bed net, a proportion of mosquitoes switched to sugar, and smaller (i.e., containing less energetic reserves) females were more likely to do so.

When cohorts of An. gambiae were presented with various ―sugar-poor‖ plants or

―sugar-rich‖ plants, they experienced different, age-dependent, levels of mortality, and the plant community affected their human-biting rates. This resulted in mosquitoes in environments where sugar was less readily available having a higher vectorial capacity – a measure of their ability to propagate malaria (Chapter 5). Measures of fitness slightly

favoured mosquitoes in sugar-rich rooms, supporting the notion of opportunistic sugar- feeding by this highly efficient vector of Plasmodium.

Dedication

To Keith, Paulette and Karel Stone, and Holly Tuten.

Acknowledgements

Many people have contributed immensely to this dissertation, intellectually, through their support, or just through hard work. I feel fortunate to have been able to progress through this Ph.D. program under the guidance of my primary advisor, Woodbridge Foster, to whom many of the ideas and sentences in this document are to be credited. I am grateful for his support and encouragement these past 5 years. I thank the other members of my dissertation committee, David Denlinger and Ian Hamilton, for their suggestions and help, and in the case of Ian in particular for his contributions to Chapter 4. I was fortunate to have worked with Robin Taylor at the start of this program. Chapters 2-4 would not have been possible without her hard work, particularly on developing and setting up the mesocosms. I am also grateful to have collaborated with Bernard Roitberg on Chapter 3, which would have been much less appealing without his model. The last two years I have been lucky to be able to work with Bryan Jackson, without whose help the larger mesocosms used in Chapters 5 and 6 would not have been designed and set up, and performing the final experiment would have been much more grueling than it was. Robert

Aldridge, Jonathan Terbot and Ashley Jackson eased several experiments by helping provide adult mosquitoes, and Jonathan Corbett provided useful assistance with the experiment described in Chapter 6. It was a pleasure to have worked alongside Babak

vii

Ebrahimi and Philip Otienoburu, and I am grateful for their help and support in various ways throughout. I am thankful for the support and encouragement received from my parents, Keith and Paulette Stone, and brother, Karel Stone. Finally, I thank Holly Tuten for support, inspiration, and her comments on, and careful editing of, several chapters.

The author gratefully acknowledges the financial support of NIH grants R21-AI062857 and R01-AI077722.

viii

Vita

January 29, 1980…………………………………….Born-IJsselstein, The Netherlands

1998………………………………………………... Oosterlicht College

2004………………………………………………... Doctoraal, Wageningen University

2006-2011…………………………………………. Graduate Teaching / Research Associate, The Ohio State University

Publications

. 1. Christopher M. Stone, Robin M. Taylor, and Woodbridge A. Foster. 2009. An effective indoor mesocosm for studying populations of Anopheles gambiae in temperate climates. Journal of the American Mosquito Control Association 25(4): 514-516.

2. C. M. Stone, R. M. Taylor, B. D. Roitberg, and W. A. Foster. 2009. Sugar deprivation reduces insemination of Anopheles gambiae (Diptera: Culicidae), despite daily recruitment of adults, and predicts decline in model populations. Journal of Medical Entomology 46(6): 1327-1337.

3. C.M. Stone, I.M. Hamilton and W.A. Foster. 2011. A survival and reproduction trade-off is resolved in accordance with resource availability by virgin female mosquitoes. Animal Behaviour 81(4): 765-774.

Field of Study

Major field: Entomology

Table of Contents

Abstract…………………………………………………………………….. ……... ii

Dedication………………………………………………………………….. ……... v

Acknowledgements………………………………………………………………… vi

Vita…………………………………………………………………………………. viii

List of Tables………………………………………………………………………. xi

List of Figures……………………………………………………………………… xii

Chapters

1. Introduction: Plant feeding and vectorial capacity………………….……... 1 Abstract……………………………………………..……………… 1 Introduction………………………………………………………... 3 1.1 Taxa involved and evidence…………………………………… 5 1.2 General features of plant feeding behaviour….……………….. 11 1.3 Obligatory or facultative nature of sugar feeding.……………... 18 1.4 Vectorial capacity…………………………..………………….. 29 1.5 Fitness…………………………………………………..……… 64 1.6 Male insemination capacity and competitiveness………………70 1.7 Flight activity and range……………………………………...... 75 1.8 Learning………………………………………………………... 76 1.9 Plant-Based Techniques for Vector Control……………...... 76 Conclusion………………………………………………………..... 92 Illustrations………………………………………………………… 94

2. An effective indoor mesocosm for studying populations of Anopheles gambiae (Diptera: Culicidae) in temperate climates…..……………………..……....96 Abstract…………………………………………………………….. 96 Body of text…………………………………………………………97 Illustrations………………………………………………………….102

3. Sugar deprivation reduces insemination of Anopheles gambiae (Diptera: Culicidae), despite daily recruitment of adults, and predicts decline in model populations…………………………………………………………………. 104 Abstract…………………………………………………………….. 104 Introduction………………………………………………………… 105 Materials and Methods……………………………………………... 107 Results……………………………………………………………… 118 Discussion………………………………………………………….. 121 Illustrations………………………………………………………….127

4. A survival and reproduction trade-off is resolved in accordance with resource availability by virgin female mosquitoes …………………………………. 133 Abstract…………………………………………………………….. 133 Introduction………………………………………………………… 134 Materials and Methods……………………………………………... 138 Results……………………………………………………………… 145 Discussion………………………………………………………….. 151 Illustrations………………………………………………………….159

5. The first meal choice (blood vs. sugar) of the malaria mosquito Anopheles gambiae s.s. is affected by bed net use and female size, but not plant abundance ………………………………………………………………………164 Abstract…………………………………………………………….. 164 Introduction………………………………………………………… 165 Methods……………………………………………...... 170 Results……………………………………………………………… 176 Discussion………………………………………………………….. 179 Illustrations………………………………………………………….187

6. Plant community composition affects the vectorial capacity and fitness of the malaria mosquito Anopheles gambiae s.s. ………………………………… 194 Abstract…………………………………………………………….. 194 Introduction………………………………………………………… 195 Materials and Methods……………………………………………... 198 Results……………………………………………………………… 204 Discussion………………………………………………………….. 209 Illustrations………………………………………………………….215

Conclusions………………………………………………………………………… 225

Bibliography……………………………………………………………………….. 227

List of Tables

Table Page

1.1 Summary of published effects of a blood-and-sugar diet, compared to a blood- only diet, on fitness components of mosquitoes …………………………….95

3.1 Parameters used in equations for stage transitions………………………… 127

5.1 Estimated coefficients, standard errors, χ2 values, p-values, odds ratios and 95% confidence intervals for the logistic regression model for feeding on sugar vs. blood ……………………………………...... 187

5.2 Estimated coefficients, standard errors, χ2 values, p-values, odds ratios and 95% confidence intervals for the logistic regression model for feeding on one or two resources (sugar and blood) ……………………………………………….. 187

6.1 Mean age at death for females in sugar rich and poor areas, mortality functions and estimated parameter values……………………………………………..218

6.2 Mean age at death for males in sugar rich and poor areas, mortality functions and estimated parameter values………………………………………………….219

6.3 Values for measures of reproductive success of An. gambiae in sugar-poor and sugar-rich mesocosms ………………………………………………………221

6.4 Vectorial capacity of cohorts in sugar-poor and sugar-rich environments, calculated for age-dependent, and constant, biting rates and mortality …….221

xii

List of Figures

Figure Page

1.1 Behavioral choice diagram for An. gambiae females ……….………………94

2.1. Measurements of relative humidity (%) and temperature (oC)…………….. 102

2.2 A schematic drawing of the mesocosm ….…………………………………103

2.3 Resting tubes within a mesocosm …………………………………………. 103

3.1 Stage structures for male and female An. gambiae…………………………128

3.2 Mean percentage of four replicates of females inseminated after 10 d with or without sugar ………………………………………………….……………128

3.3 Survival of males and females after 10 d with or without sugar …………. 129

3.4 Mean size of females surviving after 10 d in large or small enclosures …... 129

3.5 Mean time (± SE) of median bite recorded within 20 min host-exposure period each night ………………………………………………………………….. 130

3.6 Average number of bites per female (± SE) per night ………………….…..130

3.7 Population decline following the removal of carbohydrate sources in the environment ……………………………………………………………….. 131

3.8 Population size after 183 d at different levels of reduced male insemination rate and survival, when males are the only sex affected by a lack of carbohydrates in the environment..……………………………………………………………131

3.9 Population size after 183 d at different levels of reduced male insemination rate and survival, when both males and females are affected by a lack of carbohydrates in the environment..…………………………………………………………132

xiii

4.1 Mean proportions of females in internal-state categories per day after emergence in the mesocosms with sugar available ……………………………….. 158

4.2 Path diagrams of internal states and transition-matrix values for Model 1…159

4.3 Path diagrams of internal states and transition-matrix values for Models 2, 3, 4 and 5.………………………………………………………………………..160

4.4 Mean proportions of females in internal-state categories per day after emergence in the mesocosms without sugar …………………...... 161

4.5 Mean proportions of females in internal-state categories per day after emergence in the cages with sugar…………………………………………………….. 161

4.6 Mean proportions of females in internal-state categories per day after emergence in the cages without sugar………………………………………………….. 162

4.7 Mean glycogen levels + SE of females in different stages of blood meal digestion ……………………………………………………………………………... 162

4.8 Mean lipid content + SE of females in different stages of blood meal digestion ………...... 163

5.1 Diagram & photograph of the mesocosm………………………………….. 188

5.2 Proportion of small and large females over all replicates and treatments that fed on something (i.e., blood or sugar) or nothing …………………………..…189

5.3 The percentages of small (a) and large (b) females that fed on blood, sugar, both or neither when a blood host was accessible for 8 hours, for 2 half hours, or was not present.………………………………………………………………….190

5.4 The percentages of females that fed on blood, sugar, both or neither when 1 or 6 Senna didymobotrya were present …………………...... 191

5.5 Proportion of large and small males that were positive for fructose after 1 night of exposure to either 1 or 6 Senna didymobotrya ……………………………..192

5.6 Mean log amounts (µg) ± sem of fructose per large or small male with 1 or 6 Senna didymobotrya ………………………………………………………. 192

5.7 Differences in amounts of sugar ingested by females according to size, presence of a blood host, and the number of plants present in the mesocosm……..... 193

xiv

6.1 An image of the mesocosm…… ……...... 215

6.2 Kaplan-Meier survivorship curves for females in nectar-poor and –rich mesocosms …………………………………………………………………216

6.3 Kaplan-Meier survivorship curves for males in nectar-poor and –rich mesocosms …………………………..………………………………………………….217

6.4 Mean human biting rates per female per day.……………………………...220

6.5 Effects of proportionally reducing initial cohort size, biting rates and adult survival through conventional means on the vectorial capacity in sugar-poor and sugar-rich mesocosms………………………..……...... 222

6.6 Vectorial capacity in sugar-rich environments as a percentage of that in sugar- poor environments, when mortality functions are as in replicate 4...... 223

6.7 Vectorial capacity of cohorts after 41 days in sugar-rich environments as a percentage of that in sugar-poor environment, when an additional age- independent mortality factor is incrementally increased from 0 to 0.1…………………..………………………..……...... 224

Introduction: Plant Feeding and Vectorial Capacity*

Abstract

Sugar feeding is a common behaviour of male and female mosquitoes. In some species it is essential to one or both sexes; in others it is facultative. Even among females of anthropophilic species that are predisposed to a diet of frequent blood meals sugar is often taken, perhaps opportunistically, depending on internal state and environmental conditions. This opportunism is expressed as an increased likelihood of feeding on nectar when opportunities to oviposit or blood feed are scarce. Newly emerged An. gambiae females sometimes may show a preference for sugar before mating even when blood hosts are available, likely depending on both the strength of plant and animal kairomones and on the quality of the attractive properties of each.

Incorporation of sugar in the diet by mosquitoes affects certain components of their vectorial capacity. Environmental conditions, such as bed net coverage and abundance of certain plants that mosquitoes obtain nectar from, will affect the extent to which mosquitoes feed on sugar. If the effect of sugar feeding on vectorial capacity is significant, these conditions will affect transmission rates of vector-borne diseases and should be considered for inclusion in epidemiological models.

* Stone & Foster, submitted 1

Vectorial capacity is pulled in opposing directions by sugar feeding, through its effect on the two most important components, survival and biting rate. Survival of females feeding on sugar and blood is greater than that of females restricted to a blood-only diet, according to the vast majority of studies, whereas biting rates usually are depressed when sugar is available, but field evidence is scarce.

Vector density results from survival and fecundity. Most studies on vectors suggest that while fecundity per gonotrophic cycle is enhanced by sugar feeding, measures of fitness in anthropophilic species are slightly depressed. For such species, population density at the beginning of the rainy season should rise faster when females only feed on blood.

Vector competence appears to be negatively affected by sugar feeding. In certain cases plant nectar contains factors that inhibit development of the parasite in the vector. More common may be positive effects on the vector‘s immune response, but this appears to depend heavily on the host-parasite system, condition of the vector, and possibly genotype-by-environment interactions. Other aspects suggest vector competence is increased when sugar is ingested, but evidence for a role of sugar feeding is not entirely convincing for these. These aspects include enhanced repair of damaged epithelial tissues; decreased biting and related avoidance of host defensive behaviour; a lack of accelerated digestion of blood meals; and use of energetic reserves by the parasite.

Estimating the effect of these factors on vector competence at different levels of sugar intake remains difficult at this point, but an overall impression is that vectorial capacity is

somewhat decreased in environments where sugar is readily accessed. One recent field study contradicts this. The notion that reduced biting rates and delay of oviposition are laboratory artifacts is worth considering; semi-field studies on these subjects should be illuminating.

Sugar-feeding behaviour can be exploited for control purposes, most promising being the application of sugar solutions in combination with an attractant and oral insecticide for direct control, and the use of attractive phytochemicals for surveillance. Main questions facing both approaches are their suitability in verdant areas where attractants will compete with a diverse flora. For females of anthropophilic species in settings with abundant blood hosts, the question may be whether populations can be effectively suppressed by targeting male mosquitoes.

Introduction

Overviews of insect-vector sugar feeding (Downes 1958, Yuval 1992, Foster 1995) identified important gaps in our understanding of the process and its implications. One particularly important gap was the influence of available plant sugar on vector populations: whether it is a limiting resource, so that its restriction can affect the sustained density of adults in a particular community (i.e., the carrying capacity of the environment). Another gap, equally important, was the effect of plant sugar on pathogen

transmission, including how sugar‘s availability can affect vectorial capacity by altering vector competence or by changing biting frequency and survival. Because density of adults also contributes indirectly to vectorial capacity, these two gaps in our knowledge are part of the same question for disease ecologists: Is plant-sugar feeding by vectors a critical component of pathogen transmission?

Investigations of sugar-feeding vector biology, both in the laboratory and in the field, made since those comprehensive reviews, are beginning to provide details that can fill these gaps. This review will mostly focus on newly published work and on aspects of the plant-vector subject not previously discussed. For the earlier literature supporting conclusions and generalizations about mosquito sugar feeding, summarized in the present review but not fully referenced, the reader is referred to Foster (1995). For malaria transmission in particular, the possibility that sugar feeding by Anopheles mosquitoes may be an essential component of the epidemiological process is gaining wide recognition (Ferguson et al. 2010). Its importance for leishmaniasis transmission by phlebotomine sand flies in desert regions also is strongly supported. Experiments in both disease systems are providing direct evidence for plant-sugar‘s pivotal role in vector biology and for ways by which the connection between plant and vector can be manipulated to weaken or eliminate pathogen transmission, either by itself or as a valuable component of integrated-control approaches (Beier et al. 2008; Shaukat et al.

2010).

1 Taxa Involved and Evidence

1A. Taxa Covered.

Most blood-feeding Diptera also ingest plant sugar. For this reason, it is accurate to say that they have two types of hosts: vertebrate animals and vascular plants. These dipterans include the blood-feeding species of the families Culicidae, Ceratopogonidae, Simuliidae,

Psychodidae, Tabanidae, Rhagionidae, and some blood-feeding Muscidae. Notable exceptions are the tsetse flies (Glossinidae) and the ectoparasitic pupiparous dipterans

(Hippoboscidae, Streblidae, and Nycteribiidae). For a few poorly known haematophagous flies, e.g., Corethrellidae and Carnidae, the role of sugar feeding has not yet been established. No other blood-feeding insects are known to take plant sugar. This includes the Siphonaptera, such Hemiptera as the triatomine Reduviidae, Cimicidae, and

Polyctenidae, and Phthiraptera such as the anoplurans, amblycerans and rhynchophthirines. This also appears to be true of all haematophagous Acari. So far, the great majority of sugar feeding investigations have targeted mosquitoes (Culicidae) and the phlebotomine sand flies (Psychodidae). However, recent important plant-related studies also have been conducted on horse flies and deer flies (Tabanidae), black flies

(Simuliidae), biting midges (Ceratopogonidae), and stable flies (Muscidae).

1B. Plant Food Sources and Composition.

The sugar-feeding haematophagous Diptera obtain their sugar from a variety of plant sources, the most common being from floral and extrafloral nectar and from honeydew; the latter is plant-derived but homopteran-produced. Other sources frequently used are damaged or decaying fruit and seeping sap from plant wounds. Typically, these flies direct the ingested vertebrate blood straight to the midgut, where most digestion and all absorption occurs. They shunt all but the smallest sugar meals to the foregut diverticula, blind sacs where sugar is stored prior to being doled, a little at a time, into the midgut for digestion and absorption. Most of a sugar meal is stored in the large ventral diverticulum, also known as the ―crop.‖ Although the sugars sucrose, fructose, and glucose are the primary constituents of nectar, minor sugars also occur, and various oligosaccharides are common in honeydews. Glycosidases in mosquito and sand fly saliva (Marinotti et al.

1990, James and Rossignol 1991, Jacobson and Schlein 2001) and midgut (Souza-Neto et al. 2007, Jacobson and Schlein 2001) cleave sucrose into its constituent hexoses: fructose and glucose. In addition, plant-sugar meals usually contain amino acids and are considered to be part of the flower-pollinator and extrafloral gland-mutualist syndromes

(Shuel 1992). The amino acids by themselves are insufficient to stimulate or support mosquito egg development, but they do appear to promote survival (Eischen and Foster

1983, Vrzal et al. 2010) and also may serve as a flight substrate (Scaraffia and Wells

2003). Many other nectar constituents have been found, some presumably nutritional, others distasteful or toxic.

Tissue Piercing

Healthy plant tissue is sometimes pierced, and sugars and other nutrients are then extracted from phloem sap or tissue fluids. Tissue feeding has been examined most extensively and intensively in desert sand flies, where it appears to be essential to their survival. In mosquitoes, this phenomenon has been reported many times in the literature, yet it has not been explored intensively, and its significance for them globally remains unclear. The most recent evidence for mosquito and sand fly tissue feeding comes from studies in Israel, where indirect evidence comes from the detection of calcofluor-stained cellulose particles in the midgut (Schlein and Müller 1995, Schlein and Jacobson 1999,

Müller and Schlein 2005, Müller et al. 2010b, Junnila et al. 2010), and from identification of chloroplast DNA (Junnila et al. 2010). Amylase activity has been detected in many haematophagous flies (reviewed by Gooding 1975), including sand flies and mosquitoes, and amylase gene expression has been explored in mosquitoes

(Grossman et al. 1997). In phlebotomine sand flies it is used to digest starch granules obtained during plant tissue feeding (Ribeiro et al. 2000, Jacobson and Schlein 2001,

Jacobson et al. 2001). The fluid ingested by sand flies during plant piercing is transferred directly to the midgut (i.e., in the ―blood-feeding mode‖), rather than being shunted into the crop (Schlein and Warburg 1986). The presence of amylase in mosquitoes and other haematophagous Diptera suggests that starch, derived from tissue feeding, is part of the diet. In both sand flies and mosquitoes it is reported to be particularly common during

seasons and localities when and where sugar sources are rare and plants are under heat and water stress (e.g., Schlein and Müller 1995, Schlein and Jacobson 1999, Müller et al.

2010c), and it may provide water as well as sugar, but it also affects Leishmania infections in sand flies (Schlein and Jacobson 2001) (see below).

1C. Methods for evaluating plant feeding

The evidence for sugar feeding in the field, and much of what we know about it, comes either from direct observations of insects on plants or from chemical tests of gut contents.

Observations of vector behaviour on plants are sometimes difficult to interpret, because landing, aggregating, and even probing do not necessarily result in ingestion of sugars.

Furthermore, the failure to observe plant-feeding behaviour, and therefore deduce its absence, is notoriously misleading. Sugar feeding often occurs rapidly and over broad periods of the insect‘s activity period, spread over broad sweeps of a landscape that supports a vector‘s host plants. This is unlike blood-feeding behaviour, which tends to be concentrated on relatively scarce hosts and consequently is more obvious to the human observer.

Chemical tests are less susceptible to sampling biases than direct observations. The one that revolutionized sugar feeding studies of vectors is Van Handel‘s cold-anthrone test

(Van Handel 1965, 1972) for fructose, a plant sugar not synthesized de novo within the insect. Many variations on it have been published. Other methods for detecting fructose

have been developed that are reported to give greatly increased sensitivity (Somani 1987,

Nunes et al. 2008). Simple chromatographic methods also have proved satisfactory for detecting undigested sugar meals, provided that they distinguish between plant-derived and metabolically generated sugars (Laarman 1968, Watanabe 1973, Magnarelli and

Anderson 1977, Nayar 1978, Magnarelli 1978, 1979, 1980). Particularly useful have been thin-layer and gas chromatography that identifies sugars distinctive to honeydew, as opposed to floral and extrafloral nectars (Burgin and Hunter 1997a,b,c, Janzen and

Hunter 1998, Burkett et al. 1998, 1999, Hunter and Ossowski 1999), and that can help determine the likely plant-host species (Hamilton and El Naiem 2000, Manda 2007a).

Finally, evidence for the penetration of undamaged plant tissue, such as leaves and stems, can be deduced from the presence of dyes and cellulose (e.g., Schlein and Müller 1995,

Schlein and Jacobson 1999, Müller et al. 2010b), and the plant-host species can be identified from chloroplast DNA (Junnila et al. 2010). The principal problem we confront when evaluating the results of any chemical test of gut contents is that the sugars and other materials either gradually or rapidly disappear, depending on the amount consumed and on the digestion and egestion rates of each species at a particular temperature and according to the physiological status of the individual insect. Many negatives will be recorded, even among species that plant-feed at frequent intervals. So the tests provide only an approximation of the proportion of vectors that have fed on plants during a particular period of time.

2 General features of plant feeding behaviour

2A. Autogeny and Diapause

A good baseline of knowledge about sugar-feeding behaviour in mosquitoes exists, derived from a variety of studies, both lab and field. In the case of females of autogenous species (i.e., those not requiring a blood meal for initial egg development), a sugar meal often is necessary for the development of the first batch of eggs (O‘Meara 1985,1987). A few species are exclusively autogenous and feed only on sugar. Even among anautogenous mosquitoes there are species that rarely or never seek blood until they take at least one sugar meal (Renshaw et al. 1994, 1995, Hancock and Foster 1997, 2000,

Briegel et al. 2001). Sugar-feeding frequency may diminish in females of some anautogenous species, once insemination is achieved and blood feeding commences, whereas in males it remains constant throughout life. On the other hand, where winters are severe, females that have entered a state of adult diapause prior to overwintering either do not take blood at all (Culex) or take non-ovigenic blood meals from animals or humans close to their hibernacula (Anopheles). However, Culex females entering diapause up-regulate genes for fatty acid synthase (Robich and Denlinger 2005), and indirect evidence indicates that they feed on sugar at very frequent intervals prior to entering hibernacula (Jaenson and Ameneshewa 1991, Bowen 1992). During this period they accumulate large reserves of fat before foraging becomes impossible. Where winters are milder, some sugar feeding may occur among diapausing Culex populations

throughout the winter (e.g., Reisen et al. 1986), perhaps explaining the sporadic expression of the fatty acid synthase gene throughout simulated hibernation.

2B. Food Utilization.

Ingestion of sugar is directly correlated with flight range, and sugar can be consumed directly as a flight-energy substrate. Alternatively, it may be converted to glycogen for storage in the insect‘s fat body and flight muscles. Although previously thought to be used primarily for survival, stored lipid derived from either blood or sugar also can serve as a flight substrate in Anopheles gambiae Giles, which has an exceptional ability to mobilize lipid during long flights (Kaufmann and Briegel 2004). There is evidence that even amino acids derived directly from the blood meal may be used in flight in Aedes aegypti Linneaus (Scaraffia and Wells 2003). In addition to its stimulatory effect on blood feeding, sugar depresses the frequency of blood feeding, at least in the laboratory.

Possible reasons for this are discussed under Vectorial Capacity.

2C. Timing and Frequency.

The first adult food of both sexes of anautogenous species is likely to be plant sugar, and both sexes continue to take sugar throughout their reproductive lives. Sugar feeding by

non-diapausing anautogenous or post-autogenous females appears to be least likely to occur when they are digesting a blood meal, and most often occurs either while they are gravid or between oviposition and the next blood feeding—especially if the time between between egg maturation and egg laying, or between egg laying and blood feeding, is extended (Gary and Foster 2006). But there are many exceptions to this, depending both on species and on environmental circumstances. We know that sugar feeding has a characteristic time, or times, in the diel activity cycle. Cycles of sugar feeding often share a general activity rhythm with other behaviours, so that the phases of different behaviours are the same or nearly so (e.g., blood feeding and sugar feeding may have simultaneous or largely overlapping time frames). In some species, the two activities are disjunct. A field study of An. gambiae by Müller et al. (2010b) demonstrated that the times of sugar seeking and blood seeking, though to some extent overlapping, occurred in distinctly different parts of the diel cycle: attraction to sugar baits occurred mainly early in the night (but with a second peak shortly before dawn), whereas attraction to blood-host odour occurred mainly in the second half of the night, which is in accord with several other studies based on landing or biting rates in this species.

The average frequency of sugar feeding is difficult to infer from field data, because of strong temperature fluctuations and narrow periods when feeding occurs. Rough estimates are based on the time for all insects to completely digest naturally acquired nectar meals and on the fraction of resting insects that contain a meal still in some stage of digestion. By extrapolation, the time spent without sugar is calculated, and the total

time with and without sugar provides the sugar-feeding interval. Typical values indicate that males in the field during warm weather may take sugar every 1-2 days, whereas mature females of anthropophilic species may take it as infrequently as every 6-9 days.

These species are less dependent on sugar for successful reproduction, flight, and extended life in the laboratory (Harrington et al. 2001, Kaufmann and Briegel 2004,

Fernandes and Briegel 2005), and they less often contain undigested sugar meals in the field (as explained below). Females of typical zoophilic species, by contrast, often contain undigested sugar and probably feed on sugar at least as frequently as they feed on blood, i.e., every 3-4 days. Without sugar, they can die rapidly, despite frequent access to animal or even human blood (Nayar and Sauerman 1971, Wittie 2003, Fernandes and

Briegel 2005).

2D. Limited and Limiting Availability in the Field.

In laboratory cages, it is easy to demonstrate that sugar availability alters survival and reproduction. Evidence from the field is much harder to come by, primarily because of difficulties in measuring the accessibility of sugar in nature. Field samples support the general notion that more sugar feeding occurs (indicated by fructose-positivity rates) when more sugar sources, or just more plants, are available (Van Handel et al. 1994,

Martinez-Ibarra et al. 1997, Müller et al. 2010d). Based on these studies, it is possible to hypothesize that plant sugar is a limited, and potentially fitness-limiting, resource. The

underlying assumption is that up to some unknown point, it is advantageous for vectors to take more sugar if they can find it. In several studies (e.g., Hocking 1953, 1968,

Gadawski and Smith 1992), vector population density was low if fewer preferred nectar sources were available, or it declined seasonally, when the sugar sources declined. Some of the best evidence comes from the few studies of sand flies and mosquitoes in isolated areas in which a suspected plant host either can be marked with a sugar-baited dye or can be sprayed with insecticidal bait (Müller and Schlein 2006, Schlein and Müller 2008).

The dye, if it subsequently appears in a large proportion of the local vector population, demonstrates that the vectors have been preferentially visiting the plant species in large numbers, confirming independent studies of odour-based plant preference. And the significant decline in that population, when the same plant is treated with an insecticide, also demonstrates that the plant was attractive. Another approach has been to study the reproductive performance of cohorts in mesocosms with or without sugar availability

(Stone et al. 2009), or to measure the density and survival of populations of vectors in isolated areas that appear to differ only in the availability of certain preferred plant hosts

(Gu et al. 2011). Drastic differences in survivorship and biting frequency can be attributed to the presence or absence of a single species of host plant. (See more under

Vectorial Capacity, below.)

2E. Plant-Host Preference

Mosquitoes and other vectors collected while standing, crawling, or probing on a variety of plants in bloom appear to show preferential attraction to certain plant species, because they occur disproportionately on those plants. This apparent degree of host specificity is reinforced by the many anecdotal observations that link an insect to only one or a few plant species. Another source of evidence for plant-host selectivity is the ability of vectors to obtain fluids, secretions, or nutrients from a small subgroup of the species in the plant community. For example, Abdel-Malek and Baldwin (1961) were the first to suggest selective removal of sugar-bearing plants as a means of control (see Selective

Removal of Plants below). They found that Ae. aegypti and three indigenous Canadian mosquitoes (Ae. punctor Kirby, Ae. diantaeus Howard Dyar and Knab, Ae. implacibilis

Walker) fed on only three (Viola conspersa, Trillium grandiflorum, Spiraea latifolia) of

24 native plant species offered to them. A study in a natural setting in Egypt, an oasis with gardens of crops, fruits and vegetables around breeding sites, revealed that An. sergentii Evans males fed on a very select number of plants. Out of 40 plant species representing 24 families, nearly all the flora present, only three were used by males for sugar meals (Salicornia fruticosa Guckel, Alhagi maurorum Awmack and Locke, Juncus arabicus Adamson). These plants were common in the oasis, and were predictors of the presence of larvae in pools. Field collections of males were successful on these plants, and very few An. sergentii were collected from other plants (Abdel-Malek 1964). Similar laboratory experiments demonstrating differences in mosquito and sand fly ingestion and survival on various plant species have been conducted by Patterson et al. (1969), Schlein

and Warburg (1986), Alexander and Usma (1994), Gary and Foster (2004), Impoinvil et al. (2004) and Manda et al. (2007b).

Even behavioural studies that take into account the relative availability of all possible host plants suffer from misinterpeting insect aggregation—possibly the result of behavioural arrest—as attraction. Recent sand fly and mosquito experiments have eliminated some of this bias and confirm that vectors have plant preferences. These experiments have used plant-baited or plant-associated traps (Schlein and Yuval 1987,

Müller and Schlein 2004), selective insecticide dye-marking or insecticide treatment of plants in the field (Schlein and Jacobson 1994, Schlein and Müller 1995, 2008, Müller et al. 2010b, Müller et al. 2011), radioactively tagged plants (Abdel-Malek and Baldwin

1961, Abdel-Malek 1964, Patterson et al 1969) or landing and probing rate comparisons in small and large cages (e.g., Manda et al. 2007a), and wind-tunnel olfactometers

(Gouagna et al. 2010). That strong differences in attraction of An. sergentii to plants exists was convincingly demonstrated by baiting miniature CDC-light traps without lights with branches in a date plantation in a cultivated oasis near the Dead Sea in Israel.

Branches of Acacia raddiana Savi, Tamarix nilotica Bunge and Ochradenus baccatus

Delile with flowers were found to be 70-130 times more attractive than branches without flowers, and these three perennial plants were also considerably more attractive than 6 other annual plants tested. While A. raddiana, the least attractive of the three perennials, caught an average of 22.5 mosquitoes, Picris longorostris Sch. Bip., the most attractive of the annuals only caught only 5.2 mosquitoes per trap (Müller and Schlein 2006).

Müller et al. (2010a) also found evidence for differences in attraction to fruits and flowering plants in Mali for An. gambiae using wire-mesh constructions covered in glue surrounding 26 plants and 26 fruits and seedpods, then counting the number of males and females that were attracted overnight. Of the fruits and seedpods, 5 caught significantly more mosquitoes than a water-soaked sponge. Guava (Psidium guajava Linnaeus) was the most attractive fruit, trapping an average of 14 females and 9 males per night.

Flowering plants were considerably more attractive and the most attractive plant, A. macrostachya, trapped an average of 105 females and 41 males per night. Nine out of the

26 plants tested were attractive, with only minor differences between males and females.

A completely unbiased form of evidence for selective plant-feeding (i.e., the host- utilization rate) in nature must come from the guts of random samples of vectors collected in the field. These kinds of data, when combined with knowledge of the proportions of different plant species available to the vector, provide a measure of plant preference (i.e., the forage ratio or feeding index). Rigorous studies of this kind have yet to be conducted. However, chromatographic profiles of sugars in Phlebotomus orientalis

Parrot sand flies in the Sudan were able to demonstrate the relative importance of fruit and honeydew (Hamilton and El Naiem 2000) and in several species of mosquitoes

(Burkett et al. 1999), black flies (Burgin and Hunter 1997a,b,c), and horse flies (Janzen and Hunter 1998, Hunter and Ossowski 1999). Identification of species-specific chloroplast nucleotide sequences have shown that An. sergentii mosquitoes in Israel fed primarily on flower nectars of two common species during the flowering season, but

during the dry season they mainly tissue-fed on three succulent species that formed less than 1% of the vegetation (Junnila et al. 2010). As would be expected due to the uncommon occurrence of usable amounts of chloroplast DNA in nectar, this DNA-based technique appeared to be most efficient when the mosquitoes, lacking nectar sources were obliged during the dry season to pierce plant tissue. Although frequently fed upon in cages and nature in several regions of the world, judging from its signature sugars, honeydew appears not be attractive to mosquitoes (Schlein and Müller 2008, Müller et al.

2011) and sand flies (Müller and Schlein 2004, Müller et al. 2011).

3 Obligatory vs. Facultative Nature of Sugar Feeding

3A. Anthropophilic and Generalist Species

For females of some mosquito species, and males of all species, sugar is a necessity.

Without it, males typically die within a few days, having engaged in minimal mating activity and or having had limited reproductive success. Females of these species slowly die without sugar, even when offered frequent access to blood, presumably because they are incapable of handling the costs of intensive protein metabolism or gaining sufficient somatic reserves from blood meals. Among mosquito species there appears to be a continuum of female reliance upon sugar feeding. At one extreme end of the spectrum,

Toxorhynchites females feed exclusively on plant sugars, relying for egg production upon the protein reserves built up in the larval stage. A more typical mosquito female will feed

on sugar soon after emergence and at regular intervals from then on, both during and between gonotrophic cycles. At the other extreme, where human blood is readily available, certain species infrequently or rarely contain undigested plant sugar (Gillies

1968, Edman et al. 1992, Costero et al. 1998, Beier 1996, Spencer et al. 2005). Females of these species may rely on blood entirely, often feeding multiple times per gonotrophic cycle (e.g., Foster and Eischen 1987, Edman et al. 1992, Beier 1996, Braks et al. 2006) or having overlapping gonotrophic cycles (Briegel and Hörler 1993). Nonetheless, such females readily feed on sugar in the laboratory and a modest proportion can be found to contain sugar in the field. On the other hand, among zoophilic species (e.g., Reisen et al.

1986, Haramis and Foster 1990, Müller and Schlein 2005) fructose-positive rates are usually quite high.

Species for which facultative sugar feeding (i.e., they can use it if there is a shortfall of blood sources, but usually do not need to, and sometimes perhaps should not) has been proposed are An. gambiae and Ae. aegypti. These species occupy a specialized niche and share a number of characteristics that predisposes them to a diet limited to blood. In particular, both are strongly anthropophilic in most populations, and endophilic.

It is most likely this close association with humans that allows for a sugar-limited diet.

Due to their tendency to remain indoors after feeding on blood, energy required for flight is limited to that needed for mate location and oviposition site selection, and even this may require negligible amounts if oviposition sites are plentiful and proximate to domiciles. The mating behaviour of Ae. aegypti is notable in that mating occurs around

the blood host, instead of in normative swarms. Yuval (2006) proposed that this behaviour may be an adaptation to dispersed and non-synchronous emergence of adults

(though it is tempting to invoke inter-sexual selection pressure). A side effect is that this may have further decreased their need to feed on sugar. An. gambiae do mate in swarms, but little is known about the amount of flight energy females expend during this process.

It is conceivable that to meet this energetic demand, females are prone to feed on sugar soon after emergence. The intriguing finding that mating sometimes occurs indoors in

West Africa (Dao et al. 2008) implies that in some areas the energetic cost to mating and host seeking may overlap.

The lower incidence of sugar feeding, and higher rate of multiple blood feeding per gonotrophic cycle is likely one of the reasons both Ae. aegypti and An. gambiae are such efficient vectors of human disease, and may have evolved in response to an oddity of human blood composition. One of the amino acids essential for vitellogenesis, isoleucine, is notably limited in human blood compared to that of other vertebrates (Dimond et al.

1956; Lea et al. 1958; Briegel 1985). Consequently, a smaller proportion (up to 30% less) of each blood meal can be used for gametic functions, allowing for a greater investment of blood-meal carbon to somatic functions. Kaufmann and Briegel (2004) provide an elegant example of this differing physiological reliance on sugar by comparing the flight distances of An. gambiae and An. atroparvus van Thiel when fed on sugar or blood. An. gambiae females mobilized their lipid reserves for flight, and were capable of flying similar distances after being fed on sugar or on two blood meals. The more generalist

blood feeder An. atroparvus, in contrast, did not use lipids to fuel its flight, and achieved greater flight distances when sugar-fed.

Mosquitoes feeding on blood hosts with higher isoleucine contents may thus be expected to have a higher reliance on sugar for the maintenance of energetic reserves. Whether prior sugar feeding by generalist mosquitoes, in turn, affects blood host choice is not known. Under certain circumstances, non-human animals may form larger proportions of blood meals taken by anthropophilic species. Unclear at this point is whether feeding on such blood elevates the tendency to take sugar meals by these vectors. This notion is supported by a comparison among experiments that have given markedly different survival and fecundity results, depending on whether the host was human, bird, or rodent.

One example from a single study demonstrated that Ae. aegypti females, fed on human blood, had superior lifetime fecundity when sugar was absent; but when fed on mouse blood, fecundity was higher when sugar was present. Individual mouse blood meals, supplemented by sugar, generated by far the largest egg output (Harrington et al. 2001).

Lifetime survival showed a similar relationship except that there was no difference in survival on human blood, with or without sugar access. Studies on Ae. albopictus Skuse, which more often takes animal blood meals in nature, demonstrated that on human blood, females without sugar had moderately reduced survivorship (Braks et al. 2006). But on bird blood, females without sugar had drastically shortened lives (Xue et al. 2010).

3B. Field Evidence

The field evidence for a low incidence of plant sugar feeding by anthropophilic species is not unequivocal. Interpretation of fructose-positivity rates of field-collected females is tricky due to the detection of only very recent sugar meals by the anthrone test, the rapid digestion of sugar by mosquitoes, and the likelihood of collecting host-seeking or blood- fed females in indoor resting catches, potentially under-representing females in states that may be more inclined to sugar feed. Compounding these issues is the high variation reported in fructose rates between geographical areas and seasons, the main question there being whether this variation is best ascribed to differences in plant community composition and abundance, or to the presence and availability of preferred blood-hosts.

Several field and laboratory studies provide insight into the facultative nature of sugar- feeding of Ae. aegypti and An. gambiae, and its variation across different habitats. For instance, wild Ae. aegypti were collected in a rural village in Thailand and tested for the presence of sugar during both the wet and the dry season (Edman et al. 1992).

Seasonality did not affect sugar positivity, and of the females only 3% were sugar positive compared to 35% of males. The evidence of sugar feeding in a third of the males proves that sugar sources were at least available (but conceivably not of a quality that enticed females). In a mark-release-recapture study, sugar-deprived released females remained negative for fructose for 12 d following release. Of sugar-fed released females, the sugar-positivity rate declined over 4 days to 0. This slow rate of decline suggests either a modest re-feeding on sugar, or a long retention time of sugar due to the sedentary nature of this species. Spencer et al. (2005) compared fructose-feeding patterns of Ae.

aegypti females during the rainy and dry season (which correspond with high and low dengue transmission) in the Mae Sot region in Thailand. In contrast to the study by

Edman et al. (1992), seasonality did affect sugar intake of females: in the dry season 16% of females had fructose levels > 4 µg, compared to 5% in the rainy season. The authors reported that flowering plants were abundant in the region throughout both seasons, and during the dry season the predominant ones were bougainvillea, hibiscus and euphorbia.

The plant hosts actually used in the field were unknown. Similar evidence of limited sugar intake was obtained by collecting males and females inside houses in San Juan,

Puerto Rico (Van Handel et al. 1994). Only 2% of the collected females contained fructose, and all females that were blood-fed or gravid were fructose negative. This could not be explained by the absence of sugar sources, as 51% of males were fructose positive.

Plant abundance or suitability as nectar host was not measured or tested, but the authors do report that houses, patios and backyards in San Juan typically contain a large number and variety of plants that could serve as nectar sources. The results from these field collections are in stark contrast with sugar positivity rates of females collected at a rural site (a tire dump) near Vero Beach, Florida. There, collected females contained no eggs or blood, and 74% contained fructose, while 63% of males did so (Van Handel et al.

1994). The authors atribute this striking difference to the difference in blood-host abundance between an urban centre with a dense human population and the rural site where humans are rare, though an alternative explanation might be that newly emerged

Ae. aegypti are more disposed to sugar than older females are. A different field study suggested that the extent to which Ae. aegypti feed on plant nectar depends on

environmental circumstances such as the abundance of plants (Martinezz-Ibarra et al.

1997). More females in the outskirts of Huixtla, Chiapas, Mexico, were sugar-fed (21%) than in the midtown area (8%). There was no difference in number of inhabitants per house between these areas, but there was a significant difference in the number of flowering plants between the areas. Besides the difference in abundance of plants, sugar feeding may have been affected by the availability of specific plants, because bougainvillea and hibiscus occurred in 71% and 53% of the houses with sugar-positive mosquitoes, though it is not clear whether these plants were also more common in houses with sugar-positive mosquitoes than in houses without (sugar-positive) mosquitoes.

Field evidence for the use of sugar by An. gambiae s.s., and the effects of environmental conditions on this behaviour, are scarcer than they are for Ae. aegypti, but laboratory and semi-field studies do provide some insight. Most field evidence suggests that sugar feeding is rare, although a study by Laarman (1968) does suggest that feeding on sugar is a normal component of this species‘ behaviour. Muirhead-Thompson (1951) considered sugar to be an unnatural food source for this species. Gillies (1968) found little evidence of sugar feeding in females collected indoors, as these specimens lacked fluid in their crops. McCrae suggested the habitats of An. gambiae were characterized by a paucity of sugar sources (1989), but did observe 6 female and 1 male An. gambiae s.l. feeding on the extra-floral nectaries of Acacia macrostachya Rolfe, out of a total of 315 specimens belonging to 31 species of mosquitoes (personal communication to W.A.F.). Indoor- resting catches and indoor-biting catches in Kisian and Saradidi, Kenya, revealed

―surprisingly low‖ proportions of female An. gambiae s.l. and An. funestus Giles with detectable fructose (Beier 1996). In Kisian the average proportion of anophelines that were fructose positive was 9.7% and 10.4% for host-seeking and indoor-resting, respectively. In Saradidi 17.5% and 4.3% of the females in those behavioral states were fructose positive.

3C. Laboratory Studies on the Blood/Sugar Choice

Field observations, like those mentioned above, are not easy to interpret in terms of how likely a female is to feed on sugar throughout her life, in particular if an increased likelihood of doing so is negatively correlated with host-seeking behaviour. A number of laboratory studies provide insight on this subject and suggest that while An. gambiae may indeed not be an obligate sugar feeder, sugar is a viable option for their first meal (Foster and Takken 2004), and is increasingly used later in life where blood hosts or oviposition sites are less accessible or more distant from one another (Gary and Foster 2006). Based on the finding that An. gambiae would feed on a human 24 hrs after emergence, and that non-oogenic females were able to convert as much protein and lipid into maternal reserves as oogenic females transferred to egg yolk, Fernandes and Briegel (2005) suggested that this species may be opportunistic in terms of its feeding behaviour, i.e., both sugar and blood meals allow for rapid increases in reserve levels, and whichever is encountered first may be taken. This idea was supported by studies in mesocosms that

showed that the pre-mating meal taken by this species favours sugar, but a proportion does feed on blood instead, and a greater proportion does so in the absence of sugar sources (Stone et al. 2011). Sugar-related and human-related volatiles thus clearly compete and the determination of the initial meal choice may depend on both the strengths of the competing stimuli and their qualities.

In Cx. nigripalpus Theobald, a mosquito that strongly prefers to feed on sugar before feeding on blood, this situational decision-making was demonstrated by Hancock and

Foster (1997). In a wind tunnel choice test between sugar sources (dishes of honey) and blood (budgerigars), the response to either food increased by increasing the number of dishes of honey or the number of birds. A comparable study on An. gambiae in a mesocosm showed that blood-host presence (a human sleeping in the mesocosm throughout the night, or being available only for 1 hr per night) and female size, but not abundance of Senna didymobotrya (Fresen) Irwin and Barneby, affected the sugar/blood choice of 1-d old females, which strongly favoured blood in this case (Stone et al. unpubl.). This promotes the notion that females of this species are, rather than obligatory sugar feeders, opportunistically inclined to use this resource at various points in their lives, depending on environmental conditions and resource accessibility and quality.

3D. Sugar Feeding by Mosquitoes, According to Optimal-Foraging Theory

Whether or not an animal should include in its diet an item with a particular energetic gain, probability of encounter, and handling time cost, is a question of classic foraging theory (Stephens and Krebs 1986). An assumption inherent to this theory is that animals will maximize their rate of energy input, which is then essentially a proxy for fitness. As most animals have limited time-budgets and operate under constraints, and thus face trade-offs, using energy intake as a proxy for fitness is not always informative or free from error. An extension of classical foraging theory, dynamic state variable modelling

(Mangel and Clark 1988, Clark and Mangel 2000), employs a fitness function that allows the maximization of a more relevant parameter. Typically this will be lifetime reproductive output. Theories based on these ideas could provide relevant insight into the facultative nature of sugar feeding in certain mosquito species. In ecological terms,

―facultative‖ implies that the tendency to perform an act is dependent on environmental factors (e.g. resource availability). Additionally, mosquitoes in different behavioural states (age, mating status, body size, energetic reserve level, diapause status, etc.) may differ in their tendencies toward taking sugar. Combining these organismal and environmental factors allows for a subtler division than the dichotomous outcome of obligatory or facultative sugar feeding. To date, general theories applied to this have been developed (Roitberg and Friend 1993, Roitberg et al. 1994), and one study has investigated when to sugar feed for the specific case of An. gambiae (Ma and Roitberg

2008). In that study, larval development sites and blood hosts were assumed to be spatially segregated, and the availability of blood hosts and sugar sources around domiciles and of sugar sources near oviposition sites, were manipulated. The main

predictions from their model are that after emergence, a female is likely to take a sugar meal to increase its energetic state before moving indoors to seek a blood host. Around houses, sugar is only sought when energy levels become very low, and frequency of sugar feeding in such localities becomes negligible with increasing blood host availability. Following oviposition, sugar is again commonly taken—unless her energy reserves are high, in which case she returns indoors immediately. Longevity and fecundity both increased with increasing availability of sugar hosts outdoors and blood hosts indoors, whereas presence of peridomestic sugar hosts had negligible effects on these parameters.

Several aspects that may be relevant to these predicted feeding choices have not yet been theoretically explored. The first is mating behaviour. In one study, it was shown that females should opportunistically decide to blood feed before or after mating, but sugar feeding was not considered (Onyabe et al. 1998). It would be relevant to contemporary control methods and our understanding of mosquito biology to determine whether females in nature have an advantage by feeding on blood or sugar before they have located a male swarm; at the very least it would provide hypotheses about the energetic and time costs involved. Another consideration is how expected longevity, host-seeking capabilities and the need for an additional blood meal to initiate oogenesis, influence the sugar-feeding decisions of large or small mosquitoes. Finally, another realm of questions that would benefit from model-driven hypothesis generation relate to how female mosquitoes infected with parasites may be expected to modify their feeding decisions

(see Vector competence, below). Comparison of optimal behaviour as predicted by theory to behavioural observations of infected and uninfected mosquitoes may shed light on whether female mosquitoes modify their behaviour in response to infection, or instead are manipulated by the parasite.

4 Vectorial capacity

4A. Components of Vectorial Capacity.

Vectorial capacity (C) is a simplified measure of a vector‘s power for pathogen propagation. It is expressed as the total number of new vertebrate cases of an infection that can arise directly from one original infection in a particular environmental setting, due to the insect or other carrier in question. It is a subset of the equations originally developed by Ross (1910) and Macdonald (1957) to provide a quantitative epidemiological description of pathogen transmission and spread among humans by anopheline mosquitoes, in terms of the reproduction rate of cases of malaria. Vectorial capacity was introduced by Garrett-Jones (1964) and Garrett-Jones and Shidrawi (1969) as a means of singling out the vector components useful to entomologists in evaluating the potential ability of a particular insect population to spread a disease. Its basic components are the same as those in the classic Ross-Macdonald models. While vectorial capacity may be easier to measure in the field, the full epidemiological models provide more insight into how mosquito population dynamics and behavior may impact disease

transmission. In its simplest form, vectorial capacity is a function of female density relative to humans (m), biting frequency (a), survival rate (p), and duration of the extrinsic cycle (n), according to this simple expression:

Note that to make the vector density more tractable, the ―man-biting habit‖ (a) was conveniently combined with density as the ―man-biting rate‖ (ma), which can be measured as number of bites per person per day or night. Vectorial capacity is most sensitive to changes in survival rate (p) and biting frequency (a) of females. Because of their magnification by powers of n or 2, even small changes result in large effects. In mosquitoes and probably most other dipteran vectors, plant-sugar feeding appears to affect both of these coefficients, whose adjustment up or down changes vectorial capacity exponentially. Sugar also promotes egg production and insemination performance, which indirectly enhances a third component of vectorial capacity, density (m), through accelerating vector reproduction during periods of biotic release of the vector population, i.e., when density is not constrained by density-dependent factors.

Vectorial capacity is an oversimplified measure of a vector‘s power (Dye 1992), because in its original form it assumes perfect vector competence (see below). Yet in practice, even the simplest version of vectorial capacity is difficult to employ, because its parameters are not easy to measure. Still, this formula has tremendous heuristic value. It 30

allows an entomologist to focus on control efforts that are likely to achieve maximum effect. For example, it becomes clear that the density of vectors, which can be manipulated by larval suppression, whether by insecticides or source reduction, is not nearly as important as reduced adult survival, whether by residual insecticide applied to resting sites, insecticide-treated bed nets, or reduced availability of sugar sources. It also shows, in a quantitative way, which coefficients are most important to measure and worth the effort to investigate by detailed study.

Thus, to increase its heuristic utility, researchers have suggested including an approximation of vector competence (b) as a factor in the numerator of vectorial capacity.

In addition, to provide a more realistic value for the probability of survival, researchers have introduced survival rates as various sorts of non-linear functions (e.g., Styer et al.

2007b, Dawes et al. 2009, Bellan 2010), so that the probability of death may be high or low early in adult life, decline or gradually increase subsequently, then either increase greatly or decelerate at advanced vector age, based on field experiments (Harrington et al. 2008), meta-analyses of field data (Clements and Paterson 1981) and laboratory experiments (Styer et al. 2007a,b, Dawes 2009). Vector survivorship in nature remains difficult to describe with confidence. Investigators have made some progress in determining the very important effect that the pathogen has on vector survival curves

(e.g., Dawes et al. 2009), a critical component of any realistic formulation of vectorial capacity. All of these issues are discussed further, below.

Plant-Affected Components.

The following features of vectorial capacity have been shown to be directly affected by plant feeding. The vector density (m), is indirectly affected through sugar-feeding‘s influence both on reproduction (e.g., fecundity and male reproductive capacity) and on survival. The biting rate (a) is affected by sugar directly, through its influence on supplemental feeding within a gonotrophic cycle, its delaying of the primary (ovigenic) blood feeding while the crop contains a large sugar meal, and its delaying of oviposition.

Survival (p) depends both on the frequency and quality of sugar feeding. Very little is known about whether energetic reserves affect host choice and the duration of the extrinsic cycle (n). Vector competence (b), is reported to be affected by the inclusion of sugar in the diet by female mosquitoes, but the manner of the effect seems to depend on the vectors, pathogens and plants involved. Flight activity and flight range also are vulnerable to plant-sugar availability, and their unfettered performance is an implicit assumption of vectorial capacity, so they also are mentioned below. A related assumption is that host use is random, though evidence suggests otherwise. Finally, a factor in age- dependent models of vectorial capacity is the age at which blood feeding commences

(Styer et al. 2007a), which relates directly to the blood/sugar feeding choices of young females, discussed under ‗Obligatory or facultative nature of sugar feeding‘.

4B. Vector Competence

For the purposes of this chapter we consider vector competence to be the product of the vector‘s characteristics, which will determine the success rate of a particular parasite to infect that vector, to develop, and then to be transmitted to the extrinsic host (Hardy et al.

1983). There is a gauntlet of challenges parasites must run (be they Plasmodium, filarial worms, or arboviruses) in order to infect and develop in the vector. Both the efficiency with which they do so and the strength of the challenge they face may be altered by environmental conditions, such as plant-sugar availability, and hence energetic reserves of vectors.

The main challenges parasites face in their vectors are exposure to the proteinases and trypsins present in the midgut environment after ingestion with blood. For parasites spending a long period here (such as Plasmodium spp.), coagulation of the blood bolus and concomitant formation of the peritrophic matrix interfere with infection. The main bottleneck, however, is associated with invasion of the midgut epithelial cells and passage through the basal lamina to the abluminal side of the mesenteron. It is at this point that certain Plasmodium spp. are subject to melanization (e.g., Chun et al. 1995).

Further challenges from the insect immune system are faced by parasites during migration through the hemacoel to either the lumen of the salivary glands or the head region (Beerntsen et al. 2000). Here we provide an overview of the ways in which plant feeding may affect these processes, in particular focusing on direct toxic effects of plant material on the parasite while it is in the midgut environment, on the effect of energy

status on the immune response of the vector, and on the vector‘s energetic reserves, which serve as a nutrient source for the metabolic demands of the parasite as well as the defenses of the vector. It may be difficult to tease apart whether energy is expended on repair, immune function, or a different drain caused by the parasite.

Toxic effects on Leishmania and Plasmodium

Phlebotomus-Leishmania Plant-Meal Interactions.

The most immediate way in which feeding on particular plants could affect vector competence is by an inhibition of the infection by compounds present in the plant. To date, the most compelling evidence for the occurrence of a toxic effect on a parasite resulting from plant feeding by a vector is that of mortality and agglutination of

Leishmania major, the cause of zoonotic cutaneous leishmaniasis, after plant feeding by

Phlebotomus papatasi Scopoli (Schlein 1986, Schein and Jacobsen 1994, Jacobsen and

Schlein 1999). This widespread blood feeder inhabits semi-arid regions of the

Mediterranean and Middle East and appears to depend on plants, utilizing both their floral nectar and their tissue juices. Some of the host plants provide the sand flies with lectins and toxins. That these cause either agglutination or lysis of the parasites within the sand flies‘ midguts (Jacobson and Schlein 1999) became evident when sand flies artificially infected with promastigotes were kept for 7 d with access to branches of

Capparis spinosa Maire, Ricinus communis L., Solanum luteum L., Malva nicaeensis

Allioni, to branches of Spartium junceum L. with honeydew secretions of Icerya purchasi

Maskell or Aphis craccivora Koch, or to 20% sucrose. Only 7% of females that had fed on M. nicaeensis, and 17% that had fed honeydew had decreased infections (less than

1000 parasites), whereas 35-65% of sand flies fed the other plant species had reduced infection loads. After feeding on C. spinosa, R. communis, or S. luteum, many parasites were agglutinated in clumps and had disintegrated organelles or other aberrations, and after feeding on R. communis, 88% of infected females showed post-infection damage (>

50% dead or damaged parasites) (Schlein and Jacobsen 1994).

Extracts of R. communis, C. spinosa, Prosopis farcta Macbr. and T. nilotica agglutinated

Leishmania parasites in-vitro. The strongest inhibition was found for the R. communis extract: a 211 dilution still affected 25% of promastigotes. The inhibition of this toxic effect in in-vitro assays was prevented by the presence of various sugars, indicating plant lectins as the causative agent (Jacobsen and Schlein 1999). Lectins prevalent in plants had previously been shown to agglutinate promastigotes of Leishmania spp. (Dwyer

1974, Dawidowicz et al. 1975, Jacobsen et al. 1982).

Additionally, drought-induced sugar shortages in plants can affect vector population size and parasite survival. With short rasping stylets, the sand flies can cut into healthy plant tissues and ingest sugars, starch granules, and cellulose particles. They generate amylase in their saliva and elsewhere to digest the starch to simple sugars, and the parasites

generate both amylase and glucosidase, the latter being capable both of cleaving sucrose and of partially digesting cellulose (Jacobson and Schlein 2001, Jacobson et al. 2001).

The resulting sugars benefit both sand fly and parasite. During the summer period, the plants are stressed by high temperatures and lack of rain, and the plants produce much less sugar. This causes shortened sand fly lifespans, thereby greatly reducing reproduction and also curtailing the probability that a female will become infected after feeding on an infected rodent, survive long enough to allow the parasite to complete its extrinsic cycle in the sand fly, and then be transmitted to uninfected rodents. This effect appears to be offset by the natural selection for deprivation-resistant sand flies, which live longer under these conditions. The increased drought tolerance also has the side-effect of weakening the sand flies‘ ability to eliminate their parasite infections (Schlein and

Jacobson 2001).

Mosquito Pathogen Plant-Meal Effects.

Vector-produced lectins, proteins that bind with a parasite‘s structural carbohydrates important for invasion, are implicated in the outcome of several other pathogen-vector associations. For example, in Ae. aegypti, addition of N-acetyl-D-glucosamine to a blood meal containing the filarial nematode Brugia pahangi facilitated migration of the microfilariae into the haemocoel, apparently because this sugar blocked the action of gut lectins (Ham et al. 1991). Unknown at this point is whether sugars with similar effects

may be present in plant nectar. Sugar meals have been reported to either enhance or retard the development of malaria and filariasis parasites within mosquitoes (Samish and

Akov 1972, Weathersby and Noblet 1973, Vaughan et al. 1994, Pumpuni et al. 1996,

Kelly and Edman 1997, Basseri et al. 2008), but these effects and their mechanisms have not been investigated in depth. It has also been reported that plant substances ingested by

An. gambiae during plant feeding on several attractive species greatly curtail the production of P. falciparum oocysts (Manda et al. 2005, 2007c, and pers. comm. to

W.A.F.). After exposure to Ricinus communis, Parthenium hysterophorus, Lantana camara, and Senna didymobotrya, before and/or after the infectious blood meal, all had lower infection prevalence and the last three also had lower oocyst densities than glucose controls. Exposure to P. hysterophorus, which had the greatest effect, caused complete inhibition of oocyst development if the gametocyte densities in the blood were below

200/µl. P. hysterophorus is a common noxious weed known to contain parthenin, an antimalarial sesquiterpenoid (Herz et al. 1962, Hooper et al. 1990). This suggests that, as in the case of sand flies, plant feeding by mosquitoes will, to some extent, affect vector competence directly.

Effects of parasites on sugar feeding (metabolic demands).

Parasitic infection with filarial nematodes or Plasmodium spp. has been shown to affect mosquito fecundity negatively (Hurd et al. 1995), suggesting that mosquitoes harbouring

such infections bear a considerable cost. For instance, a reduced egg output as high as

33% has been reported for Ae. trivitattus Coquillett infected with Dirofilaria immitis

(Christensen 1981). The exact cause underlying this trade-off may be difficult to pinpoint, as the costs can be mediated either by the metabolic demands of the parasite or by the immune response of the vector to the parasite (Tripet et al. 2008). In either case, plant-sugar feeding may play an important role in compensating for energetic losses due to infection. An example of a parasite-mediated cost is an increased susceptibility to infection with bacteria following penetration of the midgut epithelium and development of the Plasmodium oocyst between the basement membrane and basal lamina of the midgut. One such bacterium is Serratia marcescens, which mosquitoes may obtain from sugar feeding on contaminated wicks in insectaries, which, in conjunction with

Plasmodium infection, increases mortality strongly (Maier et al. 1987). This necessitates efficient midgut repair of invaded epithelial cells. In An. stephensi this begins a few hours after infection with P. falciparum through the activation of nitric oxide synthase, resulting in apoptosis or necrosis of the invaded cells, and their subsequent extrusion and replacement. Besides the direct energetic cost, this process uses arginine, a component required for egg production in insects obtained through the diet, thus suggesting a conflict over other limiting compounds between fecundity and the response to infection (Tripet et al. 2008).

Evidence that growing oocysts use or rely on the energetic reserves of the mosquito to meet their metabolic demands is mostly indirect, and rather scant. An. stephensi Liston

infected with P. cynomolgi have reduced flight capability compared to uninfected females, as measured by their distance flown (2 km less than uninfected controls), speed and duration of flight. Furthermore, pre-flight weight between uninfected and infected females was different, indicating that this could be due to differential carbohydrate reserve use. This is supported by the intriguing finding that isolated midguts of An. stephensi infected with P. cynomolgi used up to 8 times the amount of glucose as uninfected midguts over a 2 hour period (Schiefer et al. 1977).

A number of studies have investigated how prior nutritional regimes of mosquitoes affect establishment of infection and development of parasites. Kelly and Edman (1996) provided Ae. aegypti with either a sucrose solution or water before an infectious (P. gallinaceum) blood meal, and sugar or additional blood meals afterward. Oocyst counts were highest in the group with water before, and sugar after, the infective blood meal, but there were no significant differences in sporozoite load between nutritional regimes.

Infectivity rate was lowest for females that had no access to sugar, but access to additional blood meals after infection, suggesting that perhaps a lack of nutrients from sugar negatively affects parasite establishment, or rather that subsequent blood meals and increased enzymatic activity in the midgut interferes with oocyst development. Vaughan et al. (1994), indeed, showed that the accelerated blood meal digestion resulting from prior blood feeding in An. gambiae had a detrimental effect on the production of P. falciparum oocysts. Mosquitoes with access to only sugar developed the most oocysts, those with two prior blood meals the least. And this was only evident when gametocyte

fertility (i.e., ookinete abundance) was low. However, when An. gambiae were fed blood from human volunteers naturally infected with P. falciparum, a different result was found

(Okech et al. 2004). Mosquitoes that had two prior blood meals (4 days apart) had a higher infection rate than did females provided one blood meal or only a 10% glucose solution, but oocyst loads did not differ between treatments. Whether females were kept with access to sugar or just water for the first 2 d after emergence did not affect oocyst infection rates. When Culex pipiens pipiens L. were provided access to sucrose solutions of different concentrations (2, 10, 20, 40%), survivorship and energetic reserves increased with increasing sucrose concentration. Susceptibility to West Nile virus was not affected by this, but a higher proportion of females transmitted the virus when they had access to low-concentration sucrose meals (Vaidyanathan et al. 2008).

Effects of energy state on immune (melanization) response

A robust body of work exists on the insect melanization of parasites, a specialized component of the immune response that sometimes occurs during filarial nematode and

Plasmodium infections. In a Plasmodium-refractory strain of An. gambiae (Collins et al.

1986) late ookinetes/early oocysts are readily encapsulated and melanized, and negatively charged C-25 Sephadex beads elicit a very similar response. Due to the straightforward nature of this method for quantifying the level of response (the proportion of beads melanized, the degree of melanization), and its proposed value as a general model for the

strength of the immune response (Schwarz and Koella 2002), it can offer precise insights into the interactions between mosquito diet, energetic reserves, and immunity.

Additionally, as with midgut repair in response to infection, the involvement of limiting resources besides energy reserves hints at an immune-reproductive system trade-off. For instance, it has been suggested that l-tyrosine, an amino-acid precursor of melanin involved with chorion tanning, may be diverted to melanization of a parasite. A case in point is Armigeres subalbatus Coquillett, which when challenged with Brugia malayi, experiences a delay of oviposition (Ferdig et al. 1993). Both the before mentioned arginine and tyrosine can be found in nectar (e.g., in Lantana camara [Vrzal et al. 2010]), and thus it is plausible that nectar feeding by infected vectors positively mediates such trade-offs.

A number of studies have looked into factors such as age, prior adult diet, and larval nutrition on the efficacy of the adult melanization response, and they are briefly summarized here. In laboratory strains of An. gambiae that were either susceptible or refractory to certain Plasmodium spp., melanization of negatively charged C-25 beads was highest the day after emergence, and then dropped rapidly. In the two days following a blood meal, melanization was elevated in comparison to non-blood-fed mosquitoes of the same age in the refractory strain, whereas it had a negligible role in the response of susceptible mosquitoes (Chun et al. 1995). In the refractory strain, melanization decreased with increasing temperature, and with restricted larval diet and consequent smaller body size (but not with body size within larval diet treatments, suggesting the

immune response is more closely tied to nutrient reserves rather than body size per se)

(Suwanchaichinda and Paskewitz 1998). In Ae. aegypti, melanization was positively correlated with age at pupation and with body size (Koella and Boëte, 2002), but this contrasts with the finding that Ae. aegypti refractory to P. gallicineum had shorter development times and smaller body sizes than susceptible mosquitoes (Yan et al. 1997).

Possibly, a different aspect of the immune response was involved in the latter study, which would then indicate that nutritional status does not uniformly affect the immunocompetence of vectors. Body size did not affect melanization in An. stephensi

(Koella and Sørensen 2002), but adult diet did. If females had taken a blood meal one day prior to being injected with a bead, the likelihood of complete melanization went up with increased concentration of the sugar solution available ad lib. With 2% sucrose no females completely melanized beads, but with 6% sugar 38% of mosquitoes did.

However, if females had not obtained blood, 9% on average completely melanized their bead, and this did not depend on the sucrose concentration available. This suggests that to mount an effective immune response, females must locate and feed from a high quality sugar source to augment the energy and nutrients obtained from a blood meal.

An age-dependent aspect of the melanization-promoting effects of sugar and blood meals also has been found, both in laboratory and field-collected An. gambiae (Schwartz and

Koella 2002). The strength of the melanization response decreased with age, over the first four days of life. For very young females, glucose concentration increased melanization, but by 4 days of age this was no longer the case, and instead a blood meal increased the

immune response. Both the proportion of females melanizing beads and the intensity of their response was higher in field-collected specimens. Those with longer wings (i.e., larger bodies) had a higher likelihood of carrying oocysts, and post-collection glucose concentration (2% or 8%) did not seem to affect whether oocysts developed. Only 2 of

425 infected females had melanized oocysts, suggesting that the melanization response may not be relevant for certain natural vector-malaria interactions, as Plasmodium spp. may evade this response. The authors suggest that therefore sugar feeding by An. gambiae does not affect the immune response in a significant way, as young females will rely mostly on their teneral reserves, and older females on blood meals. However, this may be a premature conclusion if studies in laboratory cages misrepresent energetic expenditures mosquitoes face in the field, or if females that do or do not include sugar in their diet experience differential energetic increases over multiple gonotrophic cycles.

Such questions are best investigated under semi-field conditions.

Clearly, the immune response (considering both apoptosis and replacement of infected epithelial cells and melanization) is energetically costly, and there is limited evidence that this, rather than the metabolic demands of developing parasites, is the main metabolic burden borne by infected vectors. In the black fly Simulium ornatum Meigen infection with the filarial nematode Onchocerca lienalis reduced vitellin contents of the ovaries from 47 to 23 µg per ovary 36 hrs after infection. Because this occurred even at very low levels of infection, it would seem that a costly immune response, rather than energetic drain by the parasite, is competing with fecundity (Hurd et al. 1995). Additionally,

Rivero and Ferguson (2003) tested whether Plasmodium-infected An. stephensi females depleted their energetic reserves more rapidly than uninfected females. Levels of whole body glycogen and lipid did not differ, but glucose was greatly increased in females with developing oocysts. However, the number of oocysts present did not influence the amount of glucose. This suggested to the authors that An. stephensi increases its sugar intake when infected with Plasmodium, irrespective of the infection load. If this interpretation is correct, the question remains whether the parasite induces this behaviour directly (e.g., to increase host survivorship through the extrinsic incubation period), or indirectly (e.g., the female requires more sugar to boost her immune response).

Observational corroboration of a change in sugar-feeding behaviour related to time since infection would be useful.

A cost of infection unrelated to oocyst burden was also found for An. stephensi infected with two genotypes of P. chabaudi known as CR and ER (Ferguson and Read 2002).

Under conditions of sugar deprivation (10% glucose solution on alternate days, instead of ad libitum) the CR clone of P. chabaudi was more virulent, despite its producing a significantly lower oocyst burden than the ER clone or a mixed CR/ER infection. With ad libitum glucose, a mixed infection had the highest survivorship cost, and there was no difference between the single infections. Virulence of different parasite genotypes therefore was concluded to depend on environmental circumstances (i.e., presence of sugar). Lambrechts et al. (2006) investigated how An. stephensi genotype (8 isofemale lines, separated for 4 generations) and environmental conditions (availability of 2, 4, or

6% glucose solution) affected infection with P. yoelii yoelii. Infection rates did not differ

among mosquito lines or glucose treatments, but they did find a genotype level difference for infection load, and the number of oocysts was greater with 4% glucose than with 2% or 6%. Mortality of infected mosquitoes was greater after 8 days (23%) than for females fed on an uninfected mouse (7%), showing a mortality cost of infection, and it was higher on the low-glucose concentration. Additionally, there was an interaction between infection status of the blood meal and glucose concentration, suggesting that infected mosquitoes suffer more under low glucose concentrations than uninfected females do.

As indicated by the contradictory findings above, the manner in which sugar intake influences immunocompetence, infection rates, infection loads, and virulence is complex.

The main point is that the parasite-vector relationship, and therefore possibly disease transmission in a given area, is highly specific to mosquito genotypes, parasite genotypes, environmental conditions, and sometimes to environment-by-genotype interactions. It is therefore important to place mosquito immune responses in an environmental context, and one aspect of environment (differing both geographically and seasonally) is the presence and abundance of attractive nectariferous plants. It‘s an exciting time to research nectar use by vectors, and many more studies on natural host-parasite systems under semi-field conditions will be required before plant-vector-parasite relations can be elucidated and generalized.

4C. Survival

Minor decreases in daily survivorship of adult mosquitoes result in stark declines in vectorial capacity. It is therefore of interest to know how this variable depends on an environmental factor that has so far not been taken into account in epidemiological models, namely the abundance and composition of the plant community. Only a few studies have investigated how different nectar sources and their presence in the field affect mosquito longevity; most are performed in laboratory or semi-field settings, whose results require careful extrapolation to field settings. Another aspect laboratory investigations have focused on is the difference in survivorship of mosquitoes on diets of sugar and blood or on blood only.

Clements and Paterson (1981) reanalyzed survival data of mosquitoes and came to the conclusion that the commonly made assumption of constant mortality throughout life rarely held. Therefore, Mcdonald‘s (1957) assumption that senescence of mosquitoes in nature is either absent or entirely overshadowed by the high daily mortality may thus not hold. Instead, a more realistic value for the probability of survival may be derived from a non-linear function, so that the probability of death increases at advanced vector age.

This is of particular relevance for vectorial capacity, because the age at which a female takes an infected blood meal determines the probability of her living through the extrinsic incubation period, and models have been proposed that take this into account (Rasgon et al. 2003, Styer et al. 2007a, Dawes et al. 2009). Furthermore, it is not just that senescence is important, but different mortality functions can have very different

outcomes for vectorial capacity For example, compare a Gompertz model, where mortality increases exponentially after a certain stage, to a logistic model, where mortality of older individuals decreases, in which case the end result is a small, but highly infective, proportion of old mosquitoes (Carey 2001, Dawes et al. 2009). While it is well known that sugar feeding will increase longevity, more important may be whether mortality of females that have sugar in their diet is best described by a different function than females feeding on blood only. For Ae. aegypti, Styer et al. (2007a) found that this was not the case, because mortality functions were best described by a logistic fit for females kept on only sugar, blood only, or both blood and sugar. Okech et al. (2004) do suggest that for female An. gambiae kept on blood only, senescence sets in faster than for females with access to sugar. A follow-up question is whether different sugar sources affect mortality curves in different ways.

To date, only a handful of studies have delved into the survivorship of mosquitoes when exposed and allowed to feed on different natural plants, and these have mostly focused on the malaria vector An. gambiae. This is an oversight, given the wealth of knowledge concerning many non-vector taxa on how strongly plants differ in nectar rewards and secondary compounds, and how pollinators or nectar thieves respond to plant cues and make use of these resources (Goulson 1999). As a result, basic behavioural investigations on the nectar-foraging behaviour of medically and veterinarily important taxa (e.g., whether foraging vectors ignore certain plants in nature, whether responses to plant volatiles are fixed or affected by individual experience with particular plant species, and

whether certain vegetation types provide inadequate amounts of nectar to serve as a relevant resources) are just beginning.

First steps in this direction were studies performed on An. gambiae where longevities of mosquitoes were compared, when they were kept in cages with cuttings of plants that occurred in their natural habitat. In such a study comparing survival on several sugar sources, Gary and Foster (2004) found greatest survival on a 50% sucrose solution and cassava (Manihot esculenta Crantz), followed by castor bean (R. communis) with extra- floral nectaries exposed, honeydew, lantana (L. camara) with flower heads, and water.

Tests of sugar positivity and obtained amounts show that overnight females were mostly able to obtain sugar (80-90%) in comparable amounts from honeydew, castor bean, and cassava, whereas on lantana and on castor beanvwith extra-floral nectaries covered with nail polish, there were no sugar-positive mosquitoes. A roughly similar proportion of mosquitoes obtained similar amounts of sugar from honeydew, castor bean and cassava, yet mean survival times differed significantly, suggesting that other components or the sugar composition of nectar affected mosquitoes. Similar differences in survivorship on different plants were also found in a study where eight plant species were tested. Of these, only castor bean increased survival to an extent comparable to a 6% sucrose solution. Lantana and sweet potato (Ipomoea batatus Lam.) resulted in high percentages of sugar positive mosquitoes, but only extended survival by an average of only one day over a water only diet, likely due to the lower amounts of sugar mosquitoes obtained from these plants. The other plants either did not differ from, or were associated with

slightly worse (Amaranthus hybridus L.) survival, than the water treatment (Impoinvil et al. 2004).

The sugar- and amino-acid composition of both floral or extra-floral nectar varies from plant to plant, and these characteristics may influence survival of mosquitoes. Yet, how they do so has been studied only in laboratory settings. For example, Vrzal et al. (2010) assessed the effect of amino acids on survival of Cx. quinquefasciatus Say males and females by adding certain amino acids (l-alanine, l-asparagine, l-glutamic acid, l- glutamine, glycine, l-proline, l-serine, l-arginine, l-threonine, l-tyrosine, l-valine) to water or to a sugar mixture based on lantana nectar. While adding amino acids to water did not increase survival, adding these amino acids to the sugar solution increased survival by

5% for females, but not males. Andersson (1992) tested longevity of female Ae. communis de Geer of different sizes and with access to different concentrations of sucrose or fructose. No difference in longevity was found between sucrose and fructose solutions, but the mean time of survival was shorter on 10% solutions than on the higher concentrations.

These results strongly suggest that mosquitoes should be discriminatory in their plant feeding. That this is indeed the case was demonstrated by observing how frequently mosquitoes rested, probed and fed on a panel of thirteen different plants, and what percentage of mosquitoes had detectable sugar in the crop the morning after exposure in a cage to these plants. Based on these parameters, plants could fairly consistently be placed

in preferred or non-preferred categories (Manda et al. 2007a). When the survival and fecundity of females feeding on a number of consistently preferred plants was compared to less preferred plants and a water control (Manda et al. 2007b), of the plants that were consistently favoured by females, only P. hysterophorus did not significantly extend survival beyond that of the water treatment. Survival was greatest on Tecoma stans Juss. ex Kunth, S. didymobotrya and castor bean, in that order (11-13 d mean survival). On the less-preferred plants Hamelia patens Jacq. and lantana mean survival was approximately

7 days. After one blood meal fecundity of females fed on P. hysterophorus and lantana was less than that of those fed on the other plants. After three consecutive blood meals there were no differences in fecundity, although the number of females ovipositing was lower on lantana and P. hysterophorus than on S. didymobotya, castor bean and T. stans.

While not assessed, this last result could be explained by fewer females kept with these poor nectar sources becoming inseminated, due to poor mating performance of males on these plants.

Thus, at least in terms of fecundity, anthropophilic species such as An. gambiae may be able to compensate for a lack of sugar feeding opportunities by increasing their blood intake frequency. Survival of female mosquitoes on diets consisting of just blood meals, typically offered once a day, compared to females offered blood meals in addition to having ad libitum access to a sugar solution, has been thoroughly explored for a limited number of species, though most studies have occurred in laboratory cages. Sugar promoted survival of female mosquitoes and sand flies in nearly every study, regardless

of species (recent references: Day et al. 1994, Bowen and Romo 1995, Straif and Beier

1996, Curtis et al. 1996, Kelly and Edman 1997, Costero et al. 1998, Schlein and

Jacobson 1999, Canyon 1999, Gary and Foster 2001, 2004, Impoinvil et al. 2004, Okech et al. 2003, Kaufmann and Briegel 2004, Fernandes and Briegel 2005, Braks et al. 2006,

Styer et al. 2007a, Manda et al. 2007, Xue et al. 2008, 2010, Stone et al. 2009). A few additional studies that looked at this under less conventional scenarios are summarized in the following paragraph.

Survival of female An gambiae in cages that were placed inside huts in a semi-field environment with ad libitum access to sugar solutions, access to a blood host every other day, or both was significantly higher whenever sugar was present (sugar and blood: 33 d; sugar only: 29 d mean survival) than when blood was the only nutritional source (14 d)

(Okech et al. 2003). The authors suggested that this large difference could be because limited access to blood impeded the ability of females to make up for a lack of sugar by increasing their blood feeding rate. A similar result was reported for An. aquasalis Curry, which did not show a difference in survival with access to a sugar-only or a sugar + blood diet, whereas females on a blood only diet faced severe mortality by their 4th day (Souza-

Neto et al. 2007). However, blood availability was restricted in this study as well, as the first blood meal was not offered until the third day of life, and every four days after that.

Differences between a sugar-only or a sugar + blood diet may have been obfuscated by the decision to cut off the experiment after 10 days. Oddly, while most studies indicate that on a mixed sugar-and-blood diet survival is greater than on a diet consisting of sugar

by itself (Styer et al. 2007a, Xue et al. 2008, 2010), Joy et al. (2010) found that lifespan was increased by restricting access to blood, so that female Ae. aegypti offered single or no blood meals lived longer than females offered blood once a week. The positive effect of dietary restriction was not limited to the adult stage. Females reared under a 0.25x larval diet had slightly lower survival on sugar alone, but increased survivorship after one blood meal; whereas those fed twice the standard larval diet had the lowest survivorship.

4D. Biting frequency

General

As noted above, nearly all studies indicate that sugar availability reduces mosquito blood- feeding frequency. And the most recent investigations confirm that sugar deprivation tends to promote blood feeding or makes females more responsive to host stimuli or less deterred by repellents (Bowen and Romo 1995a,b, Bowen et al. 1995, Renshaw 1995,

Fernandez and Klowden 1995, Takken et al. 1998, Canyon et al. 1999, Xue and Barnard

1999, Gary and Foster 2001, Braks et al. 2006). This effect probably is the manifestation of one or more of these phenomena: a) crop sugar raises the threshold of responsiveness to blood-host stimuli, b) a large store of energy reserves depresses the likelihood of supplemental blood feeding within the gonotrophic cycle, and c) either crop sugar or a large energy reserve causes reduced activity and (and possibly as a result) a delay in

oviposition, thus retarding the initiation of each new gonotrophic cycle. The overall result, taken by itself, suggests that the absence of sugar sources increases vectorial capacity. However, it must be noted that all these studies were conducted in the laboratory, where abnormal amounts and frequencies of sugar ingestion may delay initiation of gonotrophic cycles and inhibit supplementary blood feeding. Field evidence for sugar‘s depressing effect is equivocal. Ae. provocans Walker females with large amounts of sugar generally were at rest, those with moderate meals were biting, and those with small meals were nectar feeding (Smith and Gadawski 1994). Resting An. freeborni

Aitken more often contained plant sugar than those seeking blood hosts (Holliday-

Hanson et al. 1997), similar to earlier field work on salt marsh mosquitoes by Magnarelli

(1978, 1979, 1980). Female Ae. aegypti released after receiving both sugar and blood meals completed their gonotrophic cycles 2 days later than those receiving only blood

(Morrison et al. 1999). An exceptional result was reported by Gu et al. (2011), who found that An. sergentii in an oasis with a prominent sugar source had shorter gonotrophic cycles over much longer life spans than a population at a similar oasis without that sugar source, although possible differences in blood host presence between the sites were apparently not accounted for. A similar conclusion was drawn from field data by Gadawski and Smith (1992). However, in the latter case it is likely that preliminary sugar meals were necessary simply to hasten sexual maturation and initiate the blood-feeding mode.

Energetic reserves may push biting rates of female mosquitoes in opposing directions. If responsiveness of nutritionally deprived mosquitoes to hosts is diminished, females with low levels of reserves may seek a sugar meal before initiating host seeking. In An. gambiae, however, females that have a wing length below a certain size (~ 3 mm) as a result of developing in a nutritionally poor larval habitat, require an additional blood meal to bring their ovaries to the pre-vitellogenic resting stage, and it has been suggested for both An. gambiae and Ae. aegypti that energetic deficits are compensated by supplementary blood meals. Such additional, gonotrophically discordant, blood meals would greatly increase vectorial capacity. If females with low energetic reserves are also challenged in their immunocompetence, and, after blood feeding do not suffer further survival reduction due to small size, these females would be highly efficient vectors. If, however, their host seeking response is inhibited, leading them to take sugar more frequently, delaying the time between gonotrophic cycles, then it may be the larger females that pose a greater danger to human health. To elucidate these matters, we attempt to review and synthesize how sugar feeding and energetic reserves affect these different aspects related to biting frequencies (i.e., host responsiveness, supplementary and pre-vitellogenic meals, delay of oviposition, and biting persistence).

Host responsiveness

A number of studies have investigated the initiation of responsiveness to blood hosts by poorly nourished mosquitoes. Female Ae. aegypti when reared on a low diet showed a weaker response (59-61%) to a host odour than did those given a standard amount (83-

87%) when tested in an olfactometer at 4-6 days of age (Klowden et al. 1988). Whether females had access to a 1% or a 10% sucrose solution did not affect this behaviour. A different aspect was shown for Ae. bahamensis Dyar and Knab, which developed sensitivity to blood hosts faster if they received more nutrition as larvae (Bowen et al.

1995). Initiation of host-seeking in Ae. cantans Meigen and Ae. punctor was related to accumulated lipid reserves, either through larval feeding or nectar feeding. Whereas Ae. cantans would not take a blood meal before 192 hours of age, Ae. punctor is capable of synthesizing lipids more rapidly and was willing to blood feed after 48 hours (Renshaw et al. 1995).

The wing lengths of newly emerged An. gambiae were smaller than those of the host- seeking population in the field (Lyimo and Takken 1993), suggesting that a large portion of smaller females dies between emergence and host location, or never expresses host- seeking behaviour. Takken et al. (1998) found that responsiveness of large and small females to a human hand in an olfactometer increased in a similar fashion over the course of 6 days, but large-bodied females were always more responsive, despite the availability of 10% sucrose to both body sizes at all times, except for the 12 hrs preceding the tests.

Similar body-size differences in olfactometer performance during attraction to a sugar source one day after emergence have been observed (Foster and Takken 2004). These

5555

results indicate that small-bodied females are debilitated by more than just a small energy reserve.

Supplementary blood meals and ovarian development

The taking of multiple blood meals per gonotrophic cycle has most notably been observed and linked to the vector status of, Ae. aegypti and An. gambiae (Scott et al.

1993, Beier 1996, Norris et al. 2010). Multiple blood meals may occur in very young gonoinactive females as a pre-gravid meal, and this has been linked to small females that developed in nutritionally poor larval habitats (Feinsod and Spielman 1980, Lyimo and

Takken 1993) or later in life throughout regular gonotrophic cycles (Briegel and Hörler

1993) as supplementary meals. The influence of teneral reserves and reserves accumulated through sugar feeding on either type of multiple feeding may be different.

For instance, female Ae. aegypti kept with ad libitum access to sugar show host-seeking inhibition during oogenesis (Klowden and Lea 1979), but this is not the case when they are kept in the absence of sucrose (Klowden 1986), as a large proportion of gravid females then continues to seek a host. This depressing effect of sugar is also reflected in biting rates. For example, Ae. aegypti kept with only water had a higher biting rate, explained mainly by an increased frequency of supplementary feeding, than did females kept with honey. The opposite effect occurred in An. quadrimaculatus Say when kept on water. These females had an almost equal rate of supplementary feeding, but a higher

biting rate compared to those with honey, likely due to the absence of a variable delay in oviposition caused by honey feeding (Foster and Eischen 1987). Supplementary feeding in anopheline mosquitoes is proposed to be more common and used as a tactic to compensate for a smaller midgut capacity and low teneral reserves, with additional meals increasing fecundity and maternal deposits (Briegel and Hörler 1993). In An. gambiae, presence of sugar does not inhibit the host seeking response after blood feeding in the same way as it does for Ae. aegypti. While the response to a host was inhibited for 12 hrs following a blood meal, after 24 and 48 hrs females showed no sign of inhibition

(Klowden and Briegel 1994). Straif and Beier (1996) found that it was only the older females (> 20 days) that showed an increased biting rate when kept in the absence of sugar. That was not due to an increase in biting rate with age, but rather to a subgroup of the sugar-deprived females, which fed more often and therefore survived longer.

Body size of the female does not appear to affect this biting inhibition prior to oviposition, as gravid females reared under either standard or deficient larval conditions, then kept with sugar, do not respond to hosts (Klowden et al. 1988). Thus, either body size has no impact on the tendency toward supplementary blood meals later in life, or the energy deficit required to lower the host-seeking inhibition can be made up for by sugar feeding. This does not rule out the possibility that small females in the absence of sugar may exhibit a higher degree of supplementary feeding. In the field, Scott et al. (2000) found a negative relationship between wing length and blood feeding rate of Ae. aegypti in Thailand, but not in Puerto Rico. However, blood feeding also increased with

temperature in Thailand, but not in Puerto Rico, complicating the inference of causality.

Geographic variation in gonotrophic discordance is also apparent among anophelines.

High proportions of freshly blood-fed female An. gambiae s.l. and An. funestus (29-55%) resting in houses in western Kenya contained stage IV or V oocytes (Beier 1996). But,

18.9% of An. arabiensis Patton collected in Zambia were found to have taken multiple meals within one gonotrophic cycle (Norris et al. 2010). Multiple feeding appeared to be more common in a highland site in western Kenya in An. gambiae s.s. and An. funestus

(14, 11%) than in a lowland site (0, 2%) (Scott et al. 2006). The stimulus for such differences is unknown, but may also be driven by nutrition, temperature, or both.

Smaller, nutritionally deprived An. gambiae females often require one or two additional blood meals before their primary ovarian follicles reach the (gonoactive) resting stage

(Christophers‘ stage II). It is not entirely clear whether sugar feeding can make up for this nutrient deficit, and how size and sugar feeding influence biting tendencies at this time.

Takken et al. (1998) showed that despite their lower responsiveness to hosts, small females are in greater need of blood from a reproductive perspective. Small females did not reach the resting stage of ovarian development without a blood meal, whereas 52% of large females did. And with one blood meal small females reached stage II, yet large females reached stage V. In these experiments, mosquitoes had access to sugar. In a different study, access to sugar inhibited responsiveness to blood hosts as measured by the time from emergence until 95% of females had taken a blood meal, and sugar

apparently was able to substitute for a pre-gravid meal. With sugar, females of all sizes were able to mature eggs with just one blood meal (Fernandes and Briegel 2004).

In summary, while small females may have an increased need for blood meals soon after emergence to make up for energetic deficits, their decreased host-seeking capabilities interfere with this, resulting in a higher probability of feeding on sugar than larger females. Sugar feeding inhibits supplementary feeding in some species, but does not do so in Anopheles. Finally, small size may increase the tendency to take supplemental blood meals during later gonotrophic cycles, but more research is required.

Delay of oviposition

De Meillon et al. (1967) discovered that giving Cx. pipiens access to cane sugar in cages resulted in erratic and delayed oviposition, compared to females receiving only a blood meal. Females with access to water only laid most of their eggs within 5 days of receiving a blood meal, but with sugar their egg laying was more spread out and continued to the 10th day after blood feeding. Yet, withdrawing sugar resulted in an increase in oviposition rate. When access to sugar was available only before the blood meal, this effect was far less noticeable. Natural sugar sources such as raisins, honey and nectar induced this delay to a lesser extent. Presence of sugar after feeding on blood did not affect the number of eggs produced by Ae. vexans Meigen females, nor the time between the blood meal and the first eggs, but inseminated females with sugar spread

their oviposition over a slightly longer period, more often failed to oviposit (17% compared to 4% of sugar-starved females), and retained a greater number of eggs

(Shroyer and Sanders 1977). The average biting rate of An. quadrimaculatus was independent of sugar during the first four gonotrophic cycles. Yet, sugar had a slight stimulatory effect on supplementary blood feeding, especially during the first cycle, indicating that some sugar-fed females delayed oviposition in each cycle (Foster and

Eischen 1987). Increased amounts of egg retention by females given access to sugar after blood feeding also was reported for An. nuneztovari Gabaldón (Lounibos and Conn

1991). And Hudson (1970) suggested that feeding on carbohydrates may raise the threshold of chemoreceptors, after recording a similar inhibition of oviposition by sugar feeding in Ae. atropalpus Coquillett. Klowden and Dutro (1990) found that feeding on sugar reduced the responsiveness to oviposition site stimuli, and this inhibition was greater when Ae. aegypti females had access to a higher concentration of sugar, and was possibly initiated through crop distension. It is plausible that the delayed oviposition when sugar is abundant may be an adaptive response to a favourable energy state, i.e., females with a greater flight range and higher prospects for survival may be more selective in their choice of oviposition sites and distribution of eggs among them, whereas females that are close to starvation may prefer to accept oviposition sites of lesser quality. Tsunoda et al. (2010) investigated whether body size and access to sugar affects skip oviposition by Ae. aegypti. The time between blood feeding and peak oviposition was delayed, and the period over which eggs were laid was longer in sugar- fed females. Additionally, both larger and sugar-fed females oviposited higher above the

water surface and visited (oviposited in) a greater number of cups over the course of eight days. An advantage of this behaviour might be that hatching of larvae is more varied in both space and time, and as such is a form of bet-hedging.

Persistence

A special aspect of blood feeding that also is influenced by plant sugar is a vector‘s persistence at a host. Persistence probably has some bearing on the interval between successful blood meals or infectious bites. Differences in persistence in obtaining a primary blood meal have been detected in mosquitoes that differ in their energetic state as a result of access to sugar. It is not always clear whether this is an effect of sugar in the crop, energy reserves in the fat body and flight muscles, or both. Attack duration declined with repeated attacks by Ae. triseriatus Say and Ae. aegypti, but it declined more rapidly in females not having had sugar (Walker and Edman 1985, Nasci 1991). In an An. gambiae study that distinguished between persistence (time elapsed before resting on wall for more than 3 min) and attack number (number of landings per second), persistence was independent of energy state (amount of time since sugar was made inaccessible), but the number of attacks and attack rate were dependent on energy state

(Roitberg et al. 2010). Females not deprived at all had the highest number of attacks. The number of attacks was lowest in those deprived 1 day, and it was somewhat higher in those deprived 2 days, suggesting that feeding tactics vary in a non-linear way with

changing energy status. Whether these effects influence biting frequency over multiple gonotrophic cycles remains to be examined

4E. Reproduction and population density

Effects of sugar feeding and energetic reserves on fecundity

The importance of sugar feeding to a vector‘s reproduction has been evaluated primarily by laboratory studies, though field experiments are gaining prominence. We follow the definition of fecundity as the number of gametes produced by an individual over its lifetime, and fertility as the number of viable offspring produced (Clements 1992). The latter would be affected by factors such as egg retention, viability, and fertilization, whereas fecundity, which we consider here, is simply the number of mature eggs produced.

Blood meals are used both for vitellogenesis and extra-ovarian reserves (Briegel 1990).

The proportion of the blood meal that is utilized for oogenesis differs with female body size, suggesting flexibility in the allocation of proteins and lipids. When females are close to starvation or emerge with low teneral reserves, they prioritize synthesizing extra- ovarian reserves. With higher energy levels, a greater investment in vitellogenesis is possible. Increasing lipid reserves by carbohydrate feeding should thus increase egg-

batch size of females. Laboratory experiments have demonstrated that female mosquitoes with access to sugar usually produce more eggs per gonotrophic cycle (Eischen and

Foster 1983, Harrington et al. 2001, Briegel et al. 2002, Mostowy & Foster 2004, Manda et al. 2007, Gary and Foster unpubl.). Among first-cycle autogenous species there are exceptions to sugar‘s positive effect on egg- batch size (e.g., Telang and Wells 2004). An. nuneztovari that had continuous access or were deprived of sugar after blood feeding did not differ in their fecundity, measured as the number of eggs oviposited plus the number of mature eggs retained (Lounibos and Conn 1991), suggesting that feeding on sugar after a blood meal is taken may have much less of an effect on egg production during that gonotrophic cycle. But some mosquitoes may rely heavily on sugar feeding throughout the gonotrophic cycle. For instance, Ae. communis females lacking access to sugar after a blood meal failed to develop follicles to Christopher‘s stage V (Andersson 1992).

Directly after taking a sugar meal, a female's distended crop will compete for abdominal space with the midgut's future blood meal, resulting in a smaller blood meal. Because blood meal size correlates with fecundity, this suggests that the effect of sugar feeding on fecundity will largely be dependent on the exact timing of both sugar and blood feeding.

This was shown for Ae. aegypti, where females with high levels of energy reserves, built up through sugar feeding, but with an empty crop, had the highest fecundity (a mean of

84.6 eggs), and females with low reserves but a full crop, resulting from a sugar meal taken 6 hours before the blood meal and access to water prior to that, the lowest fecundity

(56.2 eggs/female). Females with high reserves and a full crop or low reserves but an

empty crop had similar, intermediate fecundities. Only in low-reserve females with empty crops was the conversion of blood into eggs considerably less efficient. Thus, in sugar-deprived females fecundity appears limited by energy levels, whereas in sugar-fed females fecundity appears to be limited mostly by blood-meal size (Mostowy and Foster

2004).

Fitness

A case has been made, based on the relatively infrequent nature of sugar feeding, and high proportions of supplementary blood meals taken by both Ae. aegypti and An. gambiae, that sugar is an inconsequential component of the diet of these vectors in settings where blood hosts are common, and that females may actually optimize their fitness by feeding exclusively on blood. This strict interpretation does not necessarily follow from field data showing low levels of fructose positivity, as sugar may be used to a greater degree farther away from indoor sampling sites, or during times of stringency within the life cycle. Even if sugar feeding is indeed uncommon in the field, this should not lead us to automatically conclude that it is unimportant. The reproductive success of females that seldom feed on sugar may be higher than that of females that never do so

(Ma and Roitberg 2008). If this were not the case, it would imply these mosquitoes either do not behave optimally (i.e., they make mistakes), or the selective pressures toward a

blood-only diet are operating, but exclusive blood feeding has not yet reached fixation in certain populations.

It may be useful to briefly define the different measures of fitness that encompass the span of life history parameters that are individually affected by sugar feeding. In general, the most commonly used measure of fitness is Fisher‘s Malthusian parameter, r, the intrinsic rate of increase. For age-structured populations r is obtained by solving the characteristic equation (Roff 1992):

Where l(x) is the proportion of the cohort alive at age x, and m(x) the production of female offspring at age x. Charlesworth (1994) described this as an adequate measure of fitness in the case of weak selection and random mating with respect to age in density- independent and constant environments. If r is close to zero (the population is stationary), the use of R0 may be justified. R0, the net reproductive rate, is the expected number of female offspring produced by a female over her lifetime:

It has been argued that in order for R0 to be a useful indicator of fitness, the growth rate of the population should not just average to zero, but actually be zero (i.e., it would not be a suitable measure for fluctuating populations), and therefore it may be better seen as another component of fitness (Caswell 2001). In environments that are not constant (e.g.,

alternating ‗good‘ or ‗bad‘ years, or seasons) the geometric mean of the finite rate of increase has been proposed as the most accurate indicator of fitness (Roff 1992).

Whether females of anthropophilic mosquitoes have a greater or lesser fitness with access to blood exclusively, or to blood and ad libitum access to a sugar solution has been the subject of several studies. An overview of these, and of how including sugar in the diet affects different fitness parameters, is given in Table 1. Most studies, but not all, report values for daily survival (lx) and fecundity (mx), and fitness measures r and R0. The only published life table data for An. gambiae s.s. to date reported that both r (0.296 versus

0.287) and R0 (180.3 versus 176.2) were slightly higher in cages where sugar was unavailable (Gary and Foster 2001). This resulted from a decrease in daily egg production, and increase in survival. Straif and Beier (1996) did not measure fecundity, but did report a similar effect on survival. Braks et al. (2006) compared Ae. albopictus to

Ae. aegypti, and found both species showed a similar response to the diets. All other studies we are aware of have focused on Ae. aegypti. While there is a general trend for survival to be increased when sugar is available, and daily fecundity decreased, not all studies bear this out. In one, survivorship decreased when females could feed on sugar

(Scott et al. 1997), and others reported no significant differences between the diets

(Naksathit and Scott 1998, Canyon et al. 1999). Styer et al. (2007b) found no difference in survival when females were kept individually, but an increase when kept in cages of

200. Harrington et al. (2001) found that it depended on the blood source, i.e., when fed on human blood the addition of sugar made no difference, when fed on mouse blood

survivorship increased with sugar availability. It is not clear, however, what could be causing the discrepancies between studies that appear similar (caged individually, daily access to a human blood source, etc.).

The effect of sugar feeding on fecundity is mostly negative in these studies, with some exceptions (Day et al. 1994, Scott et al. 1997). When blood was offered daily, Styer et al.

(2007b) found no difference in the daily egg production of females offered blood or both sugar and blood. When blood was offered every 2 days, females with access to sugar had a significantly lower daily number of offspring. This may have been due to increased reliance on sugar feeding—the rarity of blood may then have resulted in missed opportunities to blood feed or smaller blood meal volumes.

The net reproductive rate, R0, is lower for females with access to sugar in the majority of these studies. Braks et al. (2006) and Day et al. (1994) reported no difference in this parameter. The only study where R0 increased when females could sugar-feed was in the one where they were housed in cages together, instead of individually (Styer et al.

2007b). This may be related to increased levels of disturbance and concomitant higher levels of flight activity and energy use in crowded cages. Whether those conditions, or those where mosquitoes are kept individually are a better reflection of the stress factors faced in a more natural setting is debatable. The intrinsic rate of increase most often decreases when sugar is available. Harrington et al. (2001) reported r values of 0.734

(human blood alone) and 0.730 (human blood + sugar). Costero et al. (1998) reported

different results for cool and hot seasons in Puerto Rico. During the hot season females feeding only on blood had the advantage (r = 0.24 versus 0.13), while there was no difference in r during the cool season (r = 0.25 for both diets). In crowded cages r was higher when females had daily access to sugar and blood, but, counterintuitively, lower when females had access to blood every 2 days (Styer et al. 2007b).

Overall, the impression is that in the laboratory the prolonged life of a sugar- and human- fed female is often insufficient to offset sugar‘s negative effect on lifetime fecundity in these anthropophilic species. On the face of it, the depressing effect of sugar on long-term fecundity means that natural selection should favour the absence of sugar feeding in these females. This selection should be particularly strong in species lacking a quiescent egg stage (i.e., are unable to accumulate offspring in an ―egg bank‖ and engage in installment hatching after receiving hatching stimuli), during periods of population growth.

Quiescent eggs occur in Ae. aegypti and other aedines, but only to a very limited extent in

An. gambiae and other anophelines. Therefore, anophelines should take full advantage of opportunities for unrestricted reproduction during periods of population growth, by feeding only on blood. Yet, they do feed on sugar, both in the lab and in the field.

Perhaps the main critical question about these studies is whether the conditions under which they are performed are representative of those faced by mosquitoes in nature. A mark-release-recapture experiment in Puerto Rico, where female Ae. aegypti were kept either with blood only or with blood + sugar for 5 days, then released, did validate these

findings in that regard (Morrison et al. 1999). However, additional validation would be useful, as in this study values for r and R0 were not obtained, fecundity was assessed for only 1 gonotrophic cycle, and survival assessed for just 5 days. It is possible that the difference in environment between nature and laboratory cages does not matter, but the higher energetic expenditures and greater risk of mortality associated with real situations—these become evident even in mesocosms, compared to cages (Stone et al.

2009, 2011)—could enhance the reproductive success of females that do take sugar. In small spaces, flight energy may be consumed at a lower rate, reducing demand for sugar, yet incidentally sugar is encountered frequently in cages and ingested after stimulatory contact. Excessive sugar feeding may exacerbate its negative effects on the volume of blood meals, the frequency of blood feeding, and prompt oviposition. Further, it is worth pointing out that some of these experiments did not start until females were 2-3 d old to allow for mating to occur, and it is not always evident from the methods whether sugar was provided during this period.

More intangible aspects relating to sugar feeding and fitness than are measured and reported in these studies may in reality be important. Delay in oviposition may have the side-effect that females are able to spend more time locating a higher quality oviposition site, which will translate to a greater probability of survival of more competitive offspring

(Tsunoda et al. 2010). In mesocosms, females that fed on sugar 1 day before access to males had a lower insemination rate than did females feeding on blood (Stone et al.

2011). If this higher rate of insemination reflects a decrease in choosiness among half-

gravid females, then this may result in a lower genetic quality of their offspring. Another way in which sugar feeding may subtly alter offspring quality is through a maternal effect. An. gambiae females kept only on blood invested slightly fewer calories in different protein-to-lipid proportions in their eggs, than those kept on sugar (Fernandes and Briegel, 2005). Although it is not known how this affects larval development, it is certainly plausible that under harsh field conditions these subtler details of offspring quality do translate into fitness benefits that may go unnoticed in a restrained and controlled laboratory experiment.

In zoophilic species, whether aedine, culicine, or anopheline, the survival penalty of relying on blood as the sole source of energy appears to be much greater. Though not well documented, the ability of blood (either human or animal) to sustain life in the absence of sugar appears to be much poorer in these animal-feeding mosquitoes (Nayar and Sauerman 1975, Wittie 2003, Fernandes and Briegel 2005), probably because they are less able to cope with the costs of protein catabolism.

Male insemination capacity and competitiveness

Males often are completely overlooked in studies of nutrition‘s effects on survival and reproduction. But, interest in the biology of male mosquitoes has increased over recent years, largely related to concerns about the competitiveness of sterile or genetically

manipulated males compared to wild type males. Experience has shown that competitiveness can be a pitfall for genetic control programs (Reisen 2004).

Due to the commonly-held assumption that in polygynous species—in which the operational sex ratio of male to female will be high (Emlen & Oring 1977)—all females will become inseminated, repercussions of male mating behaviour on population dynamics have not been well studied. Assuming that males can inseminate multiple females, regardless of food intake, probably adds to this deficiency. A recent review of entomological field studies reports that mating failure of females in nature is common, and typically associated with early or late emergence times, high or low population densities, and in some species female size (Rhainds 2010). Often this will be age-related

(i.e., ―temporary wallflowers‖ rather than a total mating failure), with less preferred females having a longer period between maturity and insemination, which will affect their fitness. A preference for certain females has been reported in mosquitoes (Hancock et al. 1990, Okanda et al. 2002, South and Arnqvist 2011).

Mating failure of mosquitoes in nature is difficult to assess because of the lack of an age- grading method that does not rely on female reproduction. However, because males do not ingest blood, their sexual responsiveness, flight activity, and survival--and consequently their insemination potential--is completely dependent on reserves carried over from larval feeding and from post-emergence sugar feeding. Hence, sugar availability may lead to pronounced shifts in the operational sex ratio and, if male

population size declines sufficiently, an Allee effect (i.e., at low numbers, there is a positive relationship between population growth rate and population density) in certain environments or seasons. For this to be the case it has to be true that the insemination rate of females drops with the sex ratio. This will depend largely on the upper bounds of male mating capacity, and on the efficiency with which females locate swarms of different sizes—a subject of which little is known. Howell and Knols (2009) recently reviewed the mating biology of male mosquitoes and suggest a typical anopheline male may mate 0-3 times in its lifetime (in monandrous species with a 1:1 sex ratio, which is typical of mosquitoes, the average must be 1), but the maximum probably is higher.

To have even a chance at reproductive success, males must survive through a period of maturation when their terminalia rotate, and antennal fibrillae can become erect.

Following this they have to engage in the energetically costly and risky behaviour of swarming. In most species this takes place at dusk or both dusk and dawn, though many aedines swarm during the day, typically lasting only for 10-30 minutes, but sometimes for hours. The swarm itself is stationary, with males engaging in a constant to-and-fro, up-and-down movement (Downes 1969), until a female is encountered and clasped.

Having sufficient energy to perform this behaviour for multiple nights would clearly favour a male‘s prospects at mating. Furthermore, a male‘s mating ability improves during its first week of life (Verhoek and Takken 1994).

The effect of body size, an indicator of the reserves a male accumulated as a larva, on the mating performance is somewhat contradictory. This may be because on the one hand it is likely to increase longevity and the ability to join a swarm, and perhaps the length of time a male swarms per night (Yuval et al. 1993, 1994), but on the other hand, with greater body size, agility in flight and ability to grasp a female before another male does may be decreased, compared to smaller males (Ng‘habi et al. 2008).

The ability of male mosquitoes to locate and feed on sugar throughout their life may thus be the prime determinant of mating success. Yuval et al. (1994) found that An. freeborni feed on sugar only after swarming in the evening, as only resting males collected in the morning contained significant proportions of fructose, whereas males collected in the late afternoon or during swarming did not. The same appears to hold true for An. gambiae

(Stone, pers. observ.), and for Cx. tarsalis Coquillett (Reisen et al. 1986). The amount of sugars and glycogen, but not lipids, decreased significantly from the start to the end of swarming. The energetic cost of swarming was calculated to be 0.39-0.51 cal/h, resulting in a consumption of over 50% of available reserves if a male were to swarm for 40 min

(Yuval et al. 1994). In the case of Ae. aegypti, enough males may survive in the absence of sugar to inseminate all females in the same age cohort (Braks et al. 2006), at least among mosquitoes that have developed under ideal conditions and held in small cages.

Yet natural selection should always favour males that take sugar meals and thereby greatly increase their mating potential and competitive abilities.

For An. gambiae males, successfully mating with a female in the absence of sugar is an almost insurmountable task. Increased teneral reserves in combination with a low environmental temperature increase the odds somewhat, which is reflected by an increase in survival. Large males lived for an average of 3.7 d in cages (2.99 d for small males) at

23 oC. At 27oC mean survival was 2.36 and 1.94 d for large and small males, respectively.

Among sugar-fed males, the percentage that erected fibrillae and swarmed increased to almost 100% over the course of 2-3 days, whereas without sugar the percentage doing so was already much diminished by the second day of life. The proportion of females that were inseminated was influenced by both cage size and temperature; in small cages sugar deprivation did not affect insemination rates for the first 2 days of cohabitation, but a smaller proportion was inseminated after 3 days at 23 oC when sugar was absent (at 27oC all males had died by this time). In larger cages the reduction of insemination rate of females due to sugar deprivation was much more pronounced (Gary et al. 2009). In a follow-up study, the effect of sugar availability on insemination rates of females in more natural, energetically demanding, mesocosms was studied with overlapping cohorts of males and females (i.e., groups of 0-d-old males and females were released for 10 consecutive days). Maturing females would therefore have multiple opportunities to mate with maturing males. After 10 days, the cumulative insemination rate of females when sugar was present was 49.7%, compared to 10.9% in the absence of sugar (Stone et al.

2009). It remains to be seen how this relates to natural environments that may consist of a range of poor-to-good host plants. As mentioned above, directly related to insemination rates is the amount of time a female is likely to remain unmated. If this window between

female maturity and time of actual mating is expanded, females will suffer a fitness cost, and this will be absolute if they fail to become inseminated. If this occurs for a significant proportion of females, this may have population-level repercussions. Indeed, simulations of a population projection matrix show just that. When sugar sources are removed from the environment, population sizes are reduced to zero over a wide range of life history parameters (e.g., fecundity), suggesting that An. gambiae populations are not viable in the absence of sugar sources (Stone et al. 2009). It would be valuable to gain deeper insight into the link between male survival and mating ability and their foraging behaviour.

Particularly relevant is how well males are sustained on poor-to-medium-quality sugar hosts, and to what extent males can make up for this by increasing their foraging efforts.

4F. Flight Activity and Range.

Although not a coefficient of vectorial capacity, flight range is important to transmission.

This is partly because vectorial capacity makes the assumption that biting will be random within the vertebrate host population. As flight is restricted, chances increase that vectors will bite the same hosts repeatedly, introducing the complications of redundant infections or superinfections. Flight range also is critical to the successful movements of vectors between oviposition sites and vertebrate hosts when the two are widely separated (e.g.,

Clarke et al. 2002). These transmigrations must be supported by energy derived either from portions of blood meals not used in vitellogenesis or from plant-sugar meals, as

flight-mill studies have shown (e.g., Kaufmann and Briegel 2004). They almost certainly affect two critical vectorial capacity components: blood-feeding frequency and survival.

4G. Learning.

The availability of sugar, and therefore its effect on biting frequency and survival, probably changes with vector age as a result of experience. If this conjecture is valid, then sugar‘s effects on vectorial capacity also change with age. Early experiments with

Cx. quinquefasciatus and Cx. pipiens demonstrated that, by associating plant odours with the presence of sugar, young adults became more responsive to those odours (Tomberlin et al. 2006, Jhumur et al. 2006). This learning ability is expected to be advantageous in environments where sugar production by different plant species changes seasonally and also would allow adjustment to different plant communities. Non-random selection of blood hosts (McCall et al. 2001) and oviposition sites (McCall and Eaton 2001), as a result of experience, also may cause distortions unaccounted for in assessments of vectorial capacity (McCall and Kelly 2002).

5. Plant-Based Techniques for Vector Control and Interruption of Pathogen

Transmission

Reducing the incidence of malaria, the deadliest among mosquito-borne diseases, relies on reducing the entomological inoculation rate (EIR), which is the number of infectious bites received per person over a given time, usually expressed per year. This relies on the biting rate, a, and density, m, of anophelines, as well the sporozoite rate, i.e., the proportion of bites received from mosquitoes with sporozoites in their salivary glands.

The main vector control methods used to prevent malaria are currently indoor residual spraying (IRS) and insecticide-treated bed nets (ITN). A recent review of the impact of control efforts on the EIR reveals that while great reductions in EIR are often achieved with these methods, the only recent control efforts that have produced an EIR < 1, the level required to achieve a sustained reduction in parasite prevalence, made use of integrated vector management (IVM), by combining ITNs with source reduction

(Shaukat et al. 2010). The WHO has recommended the use of IVM for vector-borne diseases (WHO 2004). Its effectiveness comes from combining two or more vector control methods that are most efficacious in a particular setting, and that complement each other in a synergistic manner. Novel control methods that can be applied in such a context are urgently needed, and those targeting components of the mosquito life cycle that are left untouched by current control methods may be especially promising in that regard (Ferguson et al. 2010). We may be able to exploit the sugar-feeding behaviour of mosquitoes for such a purpose, making it all the more poignant that basic knowledge of the feeding decisions and behaviour of even one of the most important malaria vectors,

An. gambiae s.s., remains scant to date. Here we review promising studies that make use of the sugar-feeding behaviour of mosquitoes in control efforts and for surveillance of

pathogens and vectors, and speculate on their use over a range of settings, and on other possible applications, such as the use of sugar-feeding stations to introduce genetically modified bacteria into mosquito populations.

5A. Marking.

Mosquitoes and sand flies may be marked by various methods in order to determine aspects of their behaviour and survival. One efficient method is to allow the mosquitoes to feed on sugar that has been mixed with a dye or radioisotope. An advantage is that the vectors mark themselves, either at emergence (Reeves et al. 1946, Midega et al. 2007) or at a suspected host plant (Abdel-Malek and Baldwin 1961, Abdel-Malek 1964, Müller and Schlein 2006, Schlein and Müller 2008, Müller et al. 2010b), thus avoiding the disruptive effects of handling. Marking has been effective in studies of dispersal, flight range, survival, and plant-host utilization. An unfortunate side-effect of field studies in which a marked sugar solution is provided right at the place of adult emergence is that the vectors have an unnatural, easily accessed, and very early source of energy. Therefore, such studies may obtain misleading data on the timing of mating, gonotrophic cycle events, early mortality, and average distance flown from the emergence site.

5B. Trapping and surveillance of vectors, and detection of pathogens

Plant-based attractants

Vectors find their sugar sources by volatile organic compounds (VOCs) released from the host plants and by some visual cues associated with the plants. The VOCs appear to be the dominant stimuli guiding mosquitoes from intermediate or long distances to the flowers of their plant hosts or to decaying fruit, whereas visual stimuli are principally the showy white or pale petals that can be detected only within a meter or two. The plants gain by releasing VOCs that serve as attractive signals to specialist or generalist pollinators, making the vectors nectar thieves in most cases. It is not yet clear whether vectors are sometimes attracted to specialist-pollinated plants that have nectar inaccessible to vectors. The chemical cues coming from extrafloral nectaries of host plants and from host plants that must be pierced are uninvestigated (see below).

Plant VOCs have not yet been used in surveillance, apart from incidental information gained while evaluating the attractiveness of potential host plants and crude fruit-based attractants (Reisen et al. 1986, Müller and Schlein 2004, 2006, Schlein and Müller 2008,

Müller et al. 2008, 2010a). A synthetic odour, comprising the main constituents of a flower‘s VOC headspace and showing vector attraction in the field, has not yet been created.

The advantages of using phytochemical lures in traps for mosquitoes, and major hurdles that must be cleared, have been reviewed and described in detail in Foster and Hancock

(1994) and Foster (2008). The main advantages are 1) attraction of both males and females of all ages of nearly all species of mosquitoes, 2) early detection, because sugar feeding is often the first activity after emergence, and males tend to emerge first, 3) localization of emergence sites, because males tend to remain more localized, 4) attraction of females in all gonotrophic states, not just those in the blood-seeking mode, and 5) attraction of females in reproductive diapause, when they will not seek blood.

In other words, plant-volatile baited traps would target a much wider segment of the mosquito population than is typically sampled with CO2-baited traps and ovitraps, and do so throughout the year, even in temperate zones.

Because plant-sugar feeding usually occurs in the same general activity period as blood- host seeking, there is likely to be competition between the two resources, and it is generally thought that blood host volatiles will be dominant over floral volatiles.

However, during early life this may not be the case, and the relative strengths do matter.

An appealing option is to combine phytochemicals with vertebrate kairomones, which would combine the strong attraction of blood-hosts to the wider attraction (males, non- host seeking females) of sugar-hosts. But whether these volatiles would be additive or synergistic, or inhibit each other through interference remains to be tested. The combination of phytochemicals with oviposition-site volatiles in gravid traps also might be effective (females may prefer to oviposit near sites where they and their offspring can quickly regain energy) and is worth considering.

The major hurdle is then simply the identification and synthesis of a plant-volatile blend that is attractive to mosquitoes at a high release rate so that it can out-compete naturally occurring plant odours. A complication is that relatively little is known about the attraction of mosquitoes to specific plant volatiles. An example of this is that the chemical cues coming from extrafloral nectaries of host plants and from host plants that must be pierced to obtain sugar, are almost completely unstudied. The latter clearly are attractive at a distance to the sand flies that feed on them (Schlein and Yuval 1987,

Schlein and Müller 1995, Junnila et al. 2010, Müller et al. 2011). And Schlein and Müller

(2008) report that branches of similar plants with or without honeydew did not differ in the numbers of mosquitoes attracted, suggesting honeydew itself is not attractive.

Honeydew, though readily fed upon, appears not to be attractive to sand flies either

(Müller and Schlein 2004, Müller et al. 2011). However, sand fly attraction to an aphid alarm pheromone has been demonstrated (Tesh et al. 1992).

Therefore, to obtain some sugar meals mosquitoes may rely on tarsal contact when resting on plants, or they may be attracted to general plant volatiles and locate the nectaries by random walk. However, activation and orientation of mosquitoes towards floral VOC‘s are indisputable, and greater attraction to specific flowers has been shown.

For instance, there was a greater probing response in a small bioassay chamber of Ae. aegypti to extracts of milkweed (Asclepias syriaca L.) than to extracts of Canada goldenrod (Solidago canadensis L.), the former being the more fragrant flower (Vargo and Foster, 1982). And there is long-range attraction of Ae. aegypti to isolated floral

odours of ox-eye daisy (Leucanthemum vulgare Lam.). When ray and disc florets were both removed landing rates of mosquitoes on this flower were greatly reduced, whereas removal of either by itself did not result in a reduction (Jepson and Healy 1988). But, in their experiments an extract did not elicit a probing response. A role for floral odours in the mediation of long-range nectar-source location also was demonstrated by the attraction of An. arabiensis to an extract of Achillea millefolium L., a temperate plant, in a wind tunnel. The major component of the odour was reported to be a cyclic or bicyclic monoterpene (Healy and Jepson 1988). Mauer and Rowley (1999) found that Cx. pipiens responds to milkweed flowers in an olfactometer, but synthetic blends they created were not attractive. Of 36 odour receptor neurons on type A2 sensilla trichoidea of female Cx. pipiens, 19 were relatively specific to bicyclic monoterpenes containing a ketone group

(thujone and verbenone). The other 17 sensilla were more broadly tuned, and also were sensitive to other compounds, such as green-plant odours. A response in a wind tunnel was not elicited by exposure to each of these terpenes alone, or in combination with CO2, suggesting instead that an odour blend is relied on for plant location (Bowen 1992).

Carey et al. (2010) found that individual odourant receptors of An. gambiae that had strong responses to esters and aldehydes, volatiles common in the headspace of fruits, were all broadly tuned. Their role in discrimination between such volatiles is not entirely clear, but narrowly tuned receptors often appear to be associated with salient, i.e., ecologically highly relevant, odours. A comprehensive study on the breadth of odourant sensitivity to volatiles present in floral headspace has not yet been performed.

If mosquitoes have strong preferences for certain volatile blends, baits may be able to out-compete natural sources, but this baiting could be a difficult task where nectar sources are very abundant. This may determine the usefulness of plant-based attractants in the field. In recent years, more information on the specific attraction of certain plants and sugar sources has come out of Israel and Mali. A general impression of these studies is that there are certain super-attractants, and even natural concoctions based on these sources are effective attractants in sugar baits. The use of plant attractants in combination with insecticides is discussed below.

Salivation to Detect Pathogens

Mosquitoes and sand flies salivate while feeding on sugar, both to break down oligosaccharides through α-glucosidase—and perhaps also to break down starch with amylase—and to dilute very concentrated sugar solutions to facilitate ingestion. While salivating, they release viruses and malaria sporozoites (Russell et al. 1963, Beier et al.

1991, Billingsly et al. 1991, Van den Hurk et al. 2007). Van den Hurk et al. (2007) tested whether this made it possible to determine infectivity rates of vectors (Cx. annulirostris

Theobald and Cx. gelidus Theobald) for Japanese encephalitis, Kunjin virus and Murray

Valley encephalitis virus, without removing and testing salivary glands or testing extracts of whole insects. After mosquitoes were fed on a blood/virus mixture and provided with

0.3 M sucrose soaked cotton pledgets after an extrinsic-cycle period, viral RNA was

detected in the pledgets using RT-PCR, showing that these three arboviruses are indeed secreted during sugar feeding. To see if this could be used as a convenient monitoring system, CO2-baited updraft box traps were deployed in the field, in which mosquitoes would feed on honey-soaked cards that preserve nucleic acids (Hall-Mendelin et al.

2010). These cards were collected once per week and presence of viruses detected using

RT-PCR. In the lab this was shown to be useful for detection of Chikungunya, Ross

River virus and West Nile virus. In the field, Ross River and Barmah Forest viruses were detected. The main advantage of this method over previous methods is its ease, because all you need to test for viruses are the cards, which then can be associated with the species of mosquitoes within the trap, even after the mosquitoes have died and their viral contents corrupted. This should also speed up the turn-around time, allowing for a better early warning system. Possibly this also could be applied as a Plasmodium sentinel, which would be useful when infection rates of mosquitoes are very low.

5C. Reduction of Population Density and Age by Deploying Toxic Sucrose Solutions

Treatment of resting sites, including vegetation

When combined with a 20% sugar solution, Lea (1965) found that surface application of the organophosphate malathion, used as a residual insecticide, killed Ae. aegypti mosquitoes at one-tenth the dosage required otherwise. This effect was apparently

because the irritancy of the mixture was offset by its gustatory stimulation, so that mosquitoes remained in contact with it longer and perhaps also fed on it. The duration of its effectiveness also was extended considerably.

Over recent years, a control method that cleverly exploits sugar feeding has been developed and tested in several different environments with several species of mosquitoes. Attractive toxic sugar baits (ATSB), employ fruit scents to attract both male and female mosquitoes, a sucrose solution to stimulate feeding, and an oral insecticide— either boric acid or spinosad, both having very low vertebrate toxicity. Theoretically, the development of resistance can be avoided by rotating among many oral insecticides, although evolution of behavioural resistance (e.g. avoiding sugar sources) is still a concern.

That this technique is effective in arid areas with relatively few flowering plants was demonstrated by spraying A. raddiana (the only local flowering plants at the time) with dyed toxic sugar solution in a small oasis in Israel. In the control site, where the same solution was applied minus the toxin, between 80-90% of An. sergentii and 72-86% of

Ae. caspius Pallas were marked with the dye, indicating that they rely predominantly on this plant species for their sugar. Both populations were eliminated in the treatment oasis

(Müller and Schlein 2006). In another study, Cx. pipiens was shown to be strongly attracted to flowering T. jordanis branches. This was tested by placing bundles of branches of 26 plant species occurring in Israel around CDC-light traps at the periphery of a sewage pond (Schlein and Müller 2008). Attraction to T. jordanis branches was 6

and 10 times greater than to the second and third most attractive plants. Even in more plant species-diverse oases, most females fed from sugar sprayed on this plant due to the attractant properties of this tree. Treatment of just this plant species with toxic sugar solution reduced Cx. pipiens numbers tenfold, although after 18 days the population did rebound. Importantly, the age structure changed in favour of females having undergone less than 3 gonotrophic cycles. Similarly impressive results were obtained when the attractant was the juice of rotting nectarines and red wine, which was sprayed on vegetation surrounding a sewage pond. After doing so, mean trap catches of Cx. pipiens females went from 78.5 to 9, and of males from 48.5 to 3, while catches in a control site remained stable. The number of females that had gone through more than one oviposition cycle was reduced by a factor of 6 (Müller et al. 2010c).

Further examples of the potential of this technique were demonstrated by applying boric acid in sugar solutions on the leaves and stems of vegetation in outdoor screen cages as well as smaller cages, which resulted in significant mortality (> 96% in small cages, and evident as a reduced rate of recapture in larger cages) of Ae. albopictus and Cx. nigripalpus, as well as a reduction in human landing rates. Ae. taeniorhynchus

Wiedemann was apparently unaffected by the boric acid solution (Xue et al. 2006). Even sub-lethal exposures reduced host seeking, fecundity, and survival of Ae. albopictus (Ali et al. 2006). If it is possible to suppress or reduce the age-structure of populations using low concentrations of oral insecticides, this may be an additional method to avoid the evolution of resistance.

Perhaps the most relevant question pertains to the applicability of this method to An. gambiae s.s., not only because of its status as one of the prime vectors of falciparum malaria in sub-Saharan Africa, but because the reliance of females on sugar in the field is a subject of some debate. A field trial performed in Mali suggests we should be hopeful, because mosquito abundance dropped by 90% following the spraying of ATSB on patches of vegetation of unknown attractiveness surrounding larval sites. The percentage of females reaching at least 3 gonotrophic cycles dropped from 37 to 6% in the treatment area. In the control plot 56.4% of females and 62.2% of males captured were positive for sugar (Müller et al. 2010b).

Overall these results suggests toxic sugar baits may be highly effective in semi-arid areas of Africa, especially where breeding sites are spatially segregated from domestic areas.

The method is cheap and easy to implement, the attractant is easy to produce, and the approach can be used synergistically with ITNs and IRS. Whether this will work with bait stations placed in houses with vertebrate hosts, or lush areas with greater competition of natural nectar sources, or urban areas where humans and breeding sites are closer together remains to be seen. And, its impact on EIR or malaria prevalence has not been studied. A principal concern in application of ATSB is their impact on non-target sugar- feeding insect orders, such as Hymenoptera, Lepidoptera, Coleoptera, and Blattaria. This is unlikely to be an issue when applied within houses, where elimination of ants and cockroaches is welcome, and risk to humans is negligible. But outdoors, lethality to such beneficial insects as parasitoid wasps and pollinating bees, other flies, moths, and beetles

requires special consideration. A partial way around this obstacle is to design attracticide stations that allow access to mosquitoes and other vectors but exclude large-bodied pollinators. An alternative is to provide a toxin, such as Bti, whose action is specific to nematocerous dipterans.

Attractive toxic bait stations

As an alternative to spraying toxic sugar solution on vegetation, the use of sugar feeding stations also has been developed and tested. These consist of soda bottles filled with the solution (overripe nectarine juice, wine and sugar, and an oral insecticide), with a hole cut into them through which a wick keeps a sock wrapped around the bottle moist. A hood is placed atop this construction to shield it from the elements (Müller and Schlein

2008). In one study, these were placed at the openings of cisterns, the resting and larval development sites of An. claviger Meigen. After introduction of the baits the number of females in that area, measured using CDC-light traps, decreased ten-fold. The landing / human biting rate decreased by more than ten-fold. Unexpectedly, the population of males decreased less rapidly, suggesting they rested elsewhere, and came into contact with the baits less frequently (Müller and Schlein 2008). These bait stations also were used in a study in oases, this time suspended from A. raddiana trees. Baits that were laced with an insecticide steadily reduced the An. sergentii population to less than a tenth, and Ae. caspius to a third of the starting population (Müller et al. 2008). It is not immediately clear why the results, while impressive, were not as dramatic as the 2006

study where full elimination was achieved, in particular for Ae. caspius, as both species did feed on the solution readily in the control site (after three days 71% of An. sergentii were labelled with dye, 74% of Ae. caspius). It may indicate that the bait stations are slightly less effective than the spraying of resting vegetation.

5C. Selective Plant Removal or Replacement

A potential alternative use of plant feeding to control vectors, one that does not involve insecticides, is the selective removal and replacement of a vector‘s principal sources of sugar (Abdel-Malek and Baldwin 1961, Abdel-Malek 1964). In mesocosms, removal of sugar sources causes a dramatic reduction in An. gambiae reproduction, to the point of creating an environment that cannot sustain the mosquito population (Stone et al. 2009).

To be practical and to have a minimal impact on the environment, this approach requires that the host plants be both few in diversity and low in density. Obviously, if a vector can use any of 5 or 10 different plant species in a transmission zone, and at least some of them are abundant, their removal and replacement would be onerous (Schaefer and Miura

1972), except close to human habitations. To identify opportunities where this approach is feasible, we first must overcome the obstacle of determining mosquito plant-host breadth due to both innate and learned behaviour. Following removal of sources with highly attractive cues, will the ability of mosquitoes to efficiently locate nectar be

significantly diminished, or will the mosquitoes without much effort shift to the next best plant?

Additionally, if the plant species used by mosquitoes are limited, and hence, limiting on the mosquito population, the presence of those species in an environment would be of epidemiological concern. One natural example where host-plant availability appears to have a profound effect on vector population density and other features of vectorial capacity was demonstrated by Gu et al. (2011). A comparison of two oases, only one of which had just two Tamarix trees in bloom, revealed that the oasis without the trees had a much smaller An. sergentii population, with shorter female lives, a younger age structure, and a greater interval between blood meals. The vectorial capacity of the sugar-poor population was estimated to be 1/250 that of the sugar-rich population, even though the sugar-poor mosquitoes nevertheless were obtaining some plant sugar.

One topic that will be important to plant-based control is the contribution of honeydew to sugar meals of mosquitoes. If the contribution of honeydew is high, manipulating the environment in order to diminish sugar meals will require identifying the plant hosts of the homopterans that make it. The few studies that have attempted to quantify this suggest large differences between species. Honeydew feeding levels in Aedes spp. in

Ontario, Canada, were approximately 9% (Russell and Hunter 2002), comparable to values in northern Florida for Coquilletidia perturbans Walker (10%) and Psorophora ferox von Humboldt (7%), but not for Culiseta melanura Coquillett (31%) or An.

quadrimaculatus (57%) (Burkett et al. 1999). This is an understudied area, and more data are needed in order to make generalizations about the likelihood of feeding on honeydew for a particular mosquito species in a particular environment.

A different tactic is suggested by the occurrence of plants that are naturally toxic to vectors but that are nevertheless attractive to them. The ornamental Bougainvillea has this effect on Ph. papatasi, and circumstantial evidence indicates that in its vicinity sand fly numbers are much lower than in other locations (Schlein et al. 2001). These kinds of plants, if planted around human habitations, might provide sustained natural suppression of vector densities and lower survival, thereby compromising their vectorial capacity.

Such plants have not yet been discovered that work against mosquitoes or other vectors.

5D. Inoculation with Microorganisms.

Sugar baits have been used to induce Cx. pipiens mosquitoes to pick up the pathogenic bacterium Bacillus sphaericus and transfer it to larval development sites (Schlein and

Pener 1990). Additionally, attention has been drawn to symbiotic gut bacteria that interfere with the ability of ingested mature Plasmodium gametocytes to transform into oocysts in the gut wall of Anopheles (Pumpuni et al. 1996). Attractive and palatable sugar baits have been tested as a means of spreading such bacteria into the mosquito populations (Lindh et al. 2006). Acetic-acid bacteria (Acetobacteraceae) are acquired in

nectars, fruit sugars, and phloem sap (e.g., Yamada et al. 2000, Suzuki et al. 2010, Crotti et al. 2010), invade most relevant organs of the mosquitoes‘ bodies, and are transmitted both horizontally and vertically within mosquito populations. Such bacteria look particularly promising to be used to deliver anti-parasite molecules to vectors (Riehle and

Jacobs-Lorena 2005, Riehle et al. 2007, Favia et al. 2007, Damiani et al. 2008). One candidate is the osmotolerant bacterium Asaia, strains of which can infect the major vector species of mosquitoes (Chouaia et al. 2010) and can be transformed easily with foreign DNA (Favia et al. 2007, 2008) to produce strains that inhibit malaria development, thereby eliminating its vector competence. Viral paratransgenesis also is being examined with this perspective in mind (Ren et al. 2008).

Conclusion

Our aim in writing this review on the sugar feeding behaviour of mosquitoes in the context of vectorial capacity was to raise awareness of the potential effects vegetation abundance and composition in a given environment may have on pathogen transmission dynamics. Further, this underappreciated aspect of the biology of mosquitoes shows tremendous promise for novel surveillance and control methods. We have drawn parallels from the sand fly literature for certain aspects that have not been sufficiently studied for mosquitoes but are likely to have relevance. In compiling this review, numerous research gaps were identified and often seemingly contradictory results encountered that currently frustrate a holistic view of the epidemiological consequences of sugar feeding in

mosquitoes. Here we briefly reiterate the major questions in need of further research: 1.

What is the nectar-host breadth of vector species? 2. What properties make certain plants attractive? 3. How is vector performance in the absence of highly attractive, nectar-rich plants affected? and 4. How does stimulus strength and perceived quality of plant volatiles interact to mediate host (blood or nectar) choice? Studies on the effect of sugar on vectorial capacity and on reproductive success would benefit greatly from the use of more natural systems (i.e., wild-type mosquitoes, natural vector-parasite interactions, semi-field set-ups, natural blood hosts) to obtain a more realistic estimation of energetic expenditures of mosquitoes. Further, most of these questions remain to be investigated for generalist and zoophilic mosquitoes, which would advantageously broaden the current myopic view we have of mosquito-nectar dynamics due to the excessive reliance on experimentation with An. gambiae s.s. and Ae. aegypti.

Fig. 1.1: Life choice diagram for An. gambiae females, highlighting the behavioural components, the likely sequence in which they move through reproductive cycles, and factors informing the decisions between behaviours. 1.) After emergence, or oviposition, females face a choice between seeking blood or nectar 2.) Depending on their mating status, females may seek out a male swarm. 3.) Females will then seek a gonoinactive (= ―pre-gravid‖) or gonoactive blood meal, unless blood hosts are limited, in which case the likelihood of sugar-feeding increases 4.) After a gonoactive blood meal, gravid females will oviposit, unless suitable sites are unavailable, in which case the likelihood of sugar- feeding increases.

94 Table 1.1: Summary of published effects of a blood-and-sugar diet, compared to a blood-only diet, on fitness components of mosquitoes *(‗-‗ = not measured; ‗↓‘ (decrease), ‗↑‘ (increase), ‗↔‘ (no difference) refer to the effect on parameter by including sugar in the diet compared to blood only

Study Species Blood host Blood host Biting rate Survival Fecundity R0 r comments - availability d (Lx) (Mx) 1 Straif & An gambiae Mouse 15 min ↓ (only for oldest ↑ - - - Beier 1996 age group)

Gary & An gambiae s.s. Human 10 min ↓ (difference bigger ↑ ↓ ↓ ↓ Vectorial Foster 2001 in older females) capacity ↓

Scott et al Ae aegypti Human 10 min - ↓ ↔ ↓ ↓ 1997

Braks et al Ae albopictus Human 10 min ↓ ↑ ↓ ↔ - Same response 2006 & Ae aegypti for both species

Naksathit & Ae aegypti Human 10 min - ↔ ↓ ↓ ↓ Same pattern for Scott 1998 large & small females Harrington Ae aegypti Human & 15 min No statistics human ↔ ↓? ↓? ↓? et al 2001 mouse provided mouse ↑

Styer et al Ae aegypti Human (1) 15 min - 94 ↑ (1) ↓ (1) ↑ (1) ↓ 2007 every other (2) ↔ (2) ↑ (2) ↑

day (2) 10 min 10 min (?) - ↔ ↓ ↓ -?

Costero et al Ae aegypti Human 10 min - (?) ↑ ↓ ↓ (1) ↔ (1) cool season 1998 (2) ↓ (2) hot season

Day et al Ae aegypti Chicken 2 hrs - ↑ (mean ↔ ↔ - 1994 LT50 & LT90)

Canyon et al Ae aegypti Human 10 min, 4x ↓ ↔ ↓ NM - High levels of 1999 egg retention, exp. stopped at 12 d

Chapter 2: An Effective Indoor Mesocosm for Studying Populations of Anopheles

gambiae (Diptera: Culicidae) in Temperate Climates*

Abstract

To study the adult behavior of populations of the Afrotropical malaria vector

Anopheles gambiae (Diptera: Culicidae) in a temperate climate, we have devised a walk- in mesocosm, built within a greenhouse. The structure provides conditions more natural than laboratory cages, including sufficient room for swarming and for flight between resting sites, sugar-bearing plants, a human host, and an oviposition site. These activities impose energy demands closer to those encountered in the field. The structure also has predators, fluctuating temperatures, natural daylight, and an evening crepuscular period.

Most important, its resting sites comprise a bank of tubes that can be inspected or removed individually to obtain, at regular time intervals, random representative samples of an experimental population while all individuals are inactive. Samples from aging cohorts of mosquitoes, released at emergence, can yield information on behavioral sequences, mate competition, reproductive success, and survival under different nutritional regimes.

* C.M. Stone, R.M. Taylor & W.A. Foster. 2009. Journal of the American Mosquito Control Association 25(4): 514-516. 96

Mosquito populations that are confined within large enclosures offer opportunities for experimentation available neither in laboratory cages nor in the field. They eliminate the confounding effects of emigration and immigration. They allow release of mosquito cohorts of known age and physiological status and the reliable retrieval of them, unlike the low recapture rates inherent in mark-release recapture studies in field populations.

Yet, unlike lab cages, they allow space for swarming and male competition and for a more realistic representation of the energy spent on flight used in swarming, mating, and locating water, sugar, blood, and oviposition sites. Recent interest in large but confined mosquito populations has been driven by proposed methods to control malaria and dengue by introducing genetically modified or sterile male mosquitoes into natural populations. Before they are implemented, such measures necessitate transition experiments in large, secure enclosures. Ideally, such enclosures would be very large

―semi-field systems,‖ which are situated in the mosquito‘s natural environment and are exposed to natural conditions (Ferguson et al. 2008). Here, we describe an intermediate- size enclosure that allows the study of the Afrotropical malaria vector Anopheles gambiae

Giles in temperate-climate Ohio.

An. gambiae in Africa is exposed to sustained warmth, periods of intense sun, and a ~12:12 light:dark photoperiod. Ohio, on the other hand, has two strong seasons, including a winter with low light, an 8:16 light:dark photoperiod, and temperatures frequently below -10oC. To provide the mosquitoes with conditions as close to natural as possible, we constructed two large enclosures within The Ohio State University

Biological Sciences Greenhouse. In each of two identical contiguous greenhouse rooms

(cement floors, lower wall of concrete block and upper wall and ceiling of glass) we erected netting on wooden frameworks to create identical enclosures of 2.4 2.1 1.8 m

(= 9.1m3). The netting was backed by a layer of clear polyethylene sheet to maintain humidity and help stabilize temperature. The walls and ceiling were lined with a continuous length of unbleached muslin (244 cm. wide), stapled to the wooden framework. During photophase, the solid white surroundings simulated the bright daytime outdoor environment typical for An. gambiae and had the effect of driving mosquitoes into resting sites at daybreak, as occurs in the field. The white fabric also made any mosquito flying or resting on or near it clearly visible. Overlapping muslin flaps, held together with straight pins, was sufficient to prevent escape at the enclosure entryway. Two screened compartments and entryways between the experimental enclosure and the rest of the greenhouse added further security. Diatomaceous earth distributed at the base of the muslin walls and doorways impeded, but did not entirely exclude, predators such as spiders and centipedes, and scavengers such as cockroaches.

Within the enclosures, humidity was provided by a low flow of water through soaker hoses onto white cotton rugs. A central greenhouse climate-control system opened and closed ceiling vents and operated baseboard heaters, evaporative coolers, and blowers, maintaining temperature between 20 and 26 oC. Figure 1 shows the extent to which these limits were exceeded during summer. Examples of climate conditions in successful semi-field systems for An. gambiae can be found in Knols et al. 2002, and

Ferguson et al. 2008. Temperature and humidity were recorded with Hobo ™ data loggers (Onset Computer Corp. Bourne, MA). The fabric walls mitigated air turbulence,

creating a calm, quiet atmosphere in which An. gambiae exhibited normal behavior, as determined through observations by one of us (C.M.S), made while sitting quietly for 30 min at dusk or early night to provide blood during experiments (Stone et al. in press).

In addition to natural ambient light, the enclosures were illuminated with two 40- watt fluorescent tubes housed in a single 1.2 m shop-light, on which we also draped a string of five, 60-watt incandescent bulbs. The combination of light spectra allowed for plant growth and created a timer-controlled 12-h photophase between 0600 and 1800 hours. Thirty minutes before lights on, a separate timer-controlled single fluorescent light outside the entryway, extended the morning photophase when needed. At natural evening crepuscular light, mosquitoes responded with normal behaviors of host seeking, swarming and mating. In experiments to test the effects of sugar availability on survival and insemination (Stone et al. in press), we have presented sugar in one of two ways: 1) scented sucrose or honey solutions in eight glass vials plugged with cotton wicks and suspended upside down within the terminal twist of four pieces of 14-gauge solid-core coated wire (183–244 cm) draped over the top of and along the length of the shop light; or 2) three or four nectar-producing potted plants on the floor, below the lights. Open water, contained within a shallow plastic tray (26 32 cm) situated on the floor, provided an oviposition site, though eggs may also have been laid on the wet rugs or the damp concrete floor.

In some types of population surveys and experiments, it is essential to have a reliable way to recapture unskewed samples or a large proportion of the mosquitoes released. The best approach is to make resting collections during daylight hours, when

An. gambiae males and females are inactive and characteristically remain in houses and outdoor cavities (Gillies and De Meillon 1968). For our enclosures, we devised a bank of resting tubes, which simulated the secluded corners, hollows, and cavities favored by An. gambiae and allowed for easy and unobtrusive counting or collecting from selected numbers of resting units. This bank consisted of thirty cylindrical 10-cm-diameter cardboard mailing tubes, cut to 20-cm lengths and painted black inside and out and inserted into 30 circular holes in a dark-stained wooden board (80 139 cm). The open end of each tube was flush with the board (blocked by a rubber band from slipping through). The inner end of each tube was capped by its original plastic cap, but with the center cut out and replaced with nylon netting for air circulation. The board and tubes created the fourth wall of a chimney-like structure (80 62 139 cm) made of concrete cinderblocks. Cotton batting served as ―mortar‖ between the blocks and in crevices to eliminate hiding places for marauding greenhouse predators. The chimney was topped by a flat wooden roof, covered in aluminum foil to reflect heat and light. At the bottom of the shaft was a bucket filled with water. The interior of the chimney, into which the netting-capped mail tubes protruded, was therefore cool, dark, and damp.

The resting sites were sufficiently attractive to the mosquitoes that during sampling and monitoring we could assume that during the day, on average, 81% of the population could be found there (Stone et al. in press), as opposed to the concrete blocks, the muslin walls and ceiling, or the damp floor. This reliability allowed us to count mosquitoes daily without disturbing them, and made collecting an experimental population easy, even by mouth aspirator. Alternatively, the open end of the tubes could 100

be plugged (using cheap commercial plastic flower pots) and the entire tube removed from the board. The tubes could then be placed in a freezer at 40 C to arrest metabolic processes, or be put into a cold room (4–7 C) for 15 min to inactivate mosquitoes for sorting, say, to remove females but return males to the enclosure.

Thus, the type of enclosure described here offers a viable temperate-climate alternative to tropical semi-field screenhouses for studying An. gambiae population parameters, when a temperature-controlled greenhouse is available. Experiments can include the effect of environmental resources on mating success, mate competition, biting frequency, survival, and perhaps female reproductive success. Though the environment is far from its natural African counterpart in the field, space is sufficient for the energy- demanding flights associated with swarming, blood feeding, sugar feeding, and oviposition. Particularly important is the availability of attractive resting sites that can be sampled efficiently to obtain daily estimates of population behavior and size of a large cohort of mosquitoes released at emergence, by randomly sampling only a subset of the cohort (Southwood 1976, Pedigo 2002), rather than counting or collecting every individual.

101

Temperature (*C) RH (%)

31 90

30 80

29 70

28 60

27 50

Temperature (C) Temperature Relative Humidity (%)Relative 26 40

25 30 1st 2nd 3rd 4th 5th 6th 7th Date (August '08)

Fig. 2.1: Quarter-hourly measurements of relative humidity (%) and temperature (oC) in the most central resting tube in a mesocosm from August 1st until 7th, 2008.

102

Fig. 2.2: a schematic drawing of the mesocosm.

Fig. 2.3: Resting tubes within a mesocosm.

103

Chapter 3: Sugar Deprivation Reduces Insemination of Anopheles gambiae (Diptera: Culicidae), Despite Daily Recruitment of Adults, and Predicts Decline in Model Populations*

Abstract

Our research tests the hypothesis that the inability to sugar-feed reduces the insemination rate in mosquito populations. To test this, we measured the effects of sugar availability on cumulative insemination performance of male Anopheles gambiae Giles s.s. during 10-d periods of continual emergence of equal numbers of both sexes, and we evaluated the implications at the population level with a matrix population model. On each day of each of four replicates, 20 newly emerged mosquitoes of each sex were recruited into the populations within two mesocosms, large walk-in enclosures with simulated natural conditions. Each mesocosm contained a cage to replicate the experiment on a small scale. Scented sucrose was absent or present (control). A human host was available nightly as a blood meal source in both mesocosms. Sugar availability and enclosure size significantly influenced female insemination. In the mesocosms, with sugar 49.7% of the females were inseminated, as compared to 10.9% of the females without sugar. In the small cages, the insemination rates were 76.0% and 23.5%, respectively. In the mesocosms, cumulative survival of females after 10 d was 51.6%

* C. M. Stone, R. M. Taylor, B. D. Roitberg, and W. A. Foster. 2009. Journal of Medical Entomology 46(6): 1327-1337. 104

with sugar and 25.6% without sugar. In the cages, female survival was 95% and

73%, respectively. Sensitivity analysis of the population projection matrix shows that both reduced male survival and reduced mating capability due to a lack of sugar contributed to lower insemination rates in females, and in the absence of sugar the insemination rate was lowered to an extent that led to population decline.

Introduction

Anopheles gambiae Giles s.s. is one of the main vectors of malaria in sub-Saharan

Africa, yet the behavioral ecology of the males remains insufficiently explored (Ferguson et al. 2005). Current interest in developing sterile male or genetically modified male release programs makes this need for basic knowledge urgent. Here, we examine the mating capability of males in relation to the availability of sugar sources in the environment. Plant sugar is the only exogenous source of energy available to adult An. gambiae males (Foster 1995). In the absence of sugar, survival of males is dependent solely on accumulation of larval reserves, and in this species those reserves are severely limited (Foster and Takken 2004, Walker 2008). For all females of a cohort to become inseminated under these circumstances, either each male must successfully mate at least once within the first few days of his life, before reaching his peak mating capability

(Verhoek and Takken 1994), or some proportion of males must inseminate multiple females during this brief period. This demand on males may increase if polyandry in nature is as common as in some laboratory settings (Klowden 2006, Helinski et al. 2008).

105

A recently growing body of work on the use of plant sugars by An. gambiae indicates sugar feeding in this species is more common than traditionally thought

(Muirhead-Thompson 1951, Gillies and De Meillon 1968, McCrae 1989). A wind tunnel test of newly emerged females showed a preference for honey odor over human foot odor

(Foster and Takken 2004). Furthermore, studies on plant feeding demonstrated that some plant species promote the survival of both males and females of this species (Gary and

Foster 2004, Impoinvil et al. 2004, Manda et al. 2007a). Laboratory studies have revealed that sugar feeding is a common and recurring event in the lives of both males and females, and in females, sugar feeding is strongly influenced by the availability of blood and/or oviposition sites (Gary and Foster 2006). A recent cage study (Manda et al. 2007b) showed An. gambiae to be selective in the plants it feeds upon, which may indicate a preference for particular plant species as sugar sources in the field.

Whether An. gambiae feeds on sugar as a necessity or uses this resource only opportunistically and under limited circumstances (such as in laboratory cages), is a question that remains unanswered. Certainly without sugar, females remain capable of completing gonotrophic cycles (Briegel and Hörler 1993) and appear to have a higher reproductive success despite a diminished expected life-span (Straif and Beier 1996).

Truly adverse effects of sugar deprivation on An. gambiae mating were unknown until

Gary et al. (2009) focused on the behavior of male mosquitoes and found that the presence of sugar, in combination with advanced male age, larger body size, lower temperature, and smaller cage size, significantly enhanced male mating performance.

Under the most favorable conditions (small cage, large males, 23○C) males with or

106

without sugar were equally capable of mating on the first two nights. On the third night males without sugar were less capable of mating, and after the third night had virtually starved to death. Although this clearly indicated that males require sugar to reach their peak mating performance, it did not allow for a direct extrapolation to mating success in nature at the population level, where daily cohorts of both sexes overlap, environmental factors fluctuate, and cumulative sex ratios depend on both male and female survival rates. For example, one could imagine that continuous recruitment of newly emerged males might provide sufficient mating stock for inseminating most or all females. In our study, therefore, we simulated a more natural situation by releasing cohorts of newly emerged males and females daily into mesocosms in the presence or absence of sugar, to determine how sugar-source availability may affect reproductive potential in the field.

Concurrently we ran the same experiment in cages located within the mesocosms, to compare the effect of sugar under the less natural conditions of close confinement on mating behavior in this species. To explore the implications of reduced male performance in the absence of sugar, we developed a population-dynamics theory using a 2-sex population projection matrix (Caswell 1989, Fujiwara and Caswell 2002).

Materials and Methods

Mosquitoes

The mosquitoes used for these experiments came from a colony established in

2001 by staff at the International Centre of Insect Physiology and Ecology from the local population of An. gambiae s.s. in Mbita Point, Suba District, Nyanza, Kenya, and

107

identified by PCR. This Mbita strain has been maintained in acrylic cages at 26.6±1○C and 80±5 % RH since 2006. Water and 10% sucrose solution were available to colony adults ad libitum. Experimental mosquitoes were reared by transferring 100 newly hatched first instar larvae into 22.8 x 33 cm pans filled with 450 ml of aged tap water and feeding them 0.2 mg of finely ground TetraminTM fish flakes per larva during the first 3 d of larval development, 0.4 mg for the next 3 d, and 0.8 mg for subsequent days until pupation. The day before adult emergence, each pupa was placed in a glass test tube three-quarters filled with aged tap water and corked. Adults were released the following day into two types of enclosures: walk-in mesocosms and laboratory cages located within these mesocosms.

Mesocosms and Cages

Two contiguous rooms 2.4 x 1.8 x 2.1 m (= 9.1 m3) in The Ohio State University

Biological Sciences Greenhouse allowed for a more natural situation than a typical cage can provide, particularly for the energetic demands associated with swarming, mating, and locating water, sugar, blood, and dark resting sites within a large space. Prototype mesocosms previously had been employed successfully by Gary et al. (2009) and a fully developed version is described in more detail by Stone et al. (submitted). Briefly, humidity was maintained by a low flow of water through a soaker hose on the floor, while temperature was maintained within a pre-set range (20-26 oC) by a thermostatically controlled greenhouse climate-control system, automatically heated by steam wall

108

radiators and cooled by overhead vents and floor-level evaporative coolers. Temperature and humidity were recorded with Hobo data loggers.

Each mesocosm contained a cement block ―hut‖ (80 x 62 x 139 cm), which provided daytime resting sites for the mosquitoes. The front of the hut was a dark wooden board with 30 circular holes into which were inserted 10-cm diameter cardboard mailing tubes cut to 20-cm lengths and painted black on the inside. Experiments were performed between September and November, when natural photoperiods range from 13:11 h L:D to 10:14 h L:D. A combination of a string of incandescent light bulbs and fluorescent shop lights maintained a minimum 12-h photophase, timed so that after the artificial light switched off, the natural evening crepuscular light could facilitate swarming and coupling. Both were observed during twilight, although they were not quantified. Males engaged in the dance-like, zig-zagging flight typical of swarming male mosquitoes.

Eight glass vials containing either water or sugar solution were suspended upside-down from wires along the length of the 121-cm lighting fixture. The opening of each vial was stopped with a cotton-wool dental wick. A plastic tray (26 x 32 cm), containing a thin layer of water, was placed on the floor to serve as an oviposition site to avoid potential consequences of delayed oviposition.

To compare the performance of mosquitoes in the large mesocosm environment to their performance in cages of a size often used in colony maintenance, a clear acrylic plastic cage (30 x 30 x 45cm) was situated on the floor of each mesocosm, in which the same experiment was performed simultaneously. Black plastic cups, oriented sideways

109

and glued against the cage wall, served as resting sites. A small plastic cup (10 cm diameter, 4 cm deep) 1/3 filled with aged water served as an oviposition site.

Experimental procedures

Each day, 20 male and 20 female mosquitoes were released into each of the mesocosms the afternoon following their emergence to simulate the natural daily emergence of new cohorts. At the same time, five males and five females were released into each of the plastic cages. Field observations by Marchand (1984) showed swarms to typically consist of 50-100 mosquitoes, although with no lower limit. The numbers we released, taking survival into account, allowed for roughly comparable numbers. Typical resting densities in and around households also fall within this range (Minakawa et al.

2002).

In the water-only (experimental) mesocosm and cage, all glass vials were filled with water: eight in the mesocosm, four in the cage. In the sugar-access (control) mesocosm, four were filled with water and four with a 10% sucrose solution that contained 0.05 ml of verbenone per liter to provide an attractive scent (Gary and Foster

2006). In the cage, two were filled with water and two with scented sucrose. The wicks that delivered the fluid were replaced every 2 d to ensure an ad libitum supply of water or fresh sucrose solution.

Each evening, between 2000 and 2300 hours, the females in one of the mesocosms and the cage within it were allowed to blood-feed on the lower legs and feet of a human volunteer (C.M.S.) for 20 min, after which the females in the other mesocosm

110

and its cage were allowed to blood-feed (in accordance with Human Subjects Research protocol number 2004H0193 and Biohazard Research protocol number 2005N0020). The order in which the treatments were given access to blood was alternated each day. Using a headlamp with a red filter to avoid disturbing the mosquitoes, the volunteer observed and recorded the total number of bites that occurred during the exposure period and the time (minutes) between the beginning of the exposure period and the initiation of each bite, i.e., settling on the skin. Settling was invariably followed by blood engorgement.

The time of the median bite after the start of exposure was used to assess female responsiveness to a potential blood meal on each of the 10 nights in each treatment. Mean blood-feeding frequency per female in each treatment during each of the 10 nights was calculated from estimates of female survivorship (as described below).

After 10 d, all surviving mosquitoes were collected by aspirator, frozen, and later counted. Every female was dissected in saline, the spermatheca removed and inspected visually for the presence of sperm at 100 and 450x magnification. Because body size is known to affect behavior and survival of mosquitoes, 20 males and females from each treatment were selected randomly to determine wing length, measured from the axillary incision to the tip, excluding the fringe (Packer and Corbet 1989).

Daily survival rates of all released females and males were based on the number of mosquitoes that were still alive at the end of the experiment (Fig. 2). Daily survival could be derived from this value, because the number alive after 1 to 10 d in an enclosure should equal the sum of the survivors of all 10 released cohorts. The survivors from day one, for example, would be based on n = 20*daily survival (to the power of 10), and so

111

on, for each of the cohorts (assuming no significant variation in cohort survival). In a similar manner the daily male insemination rate could be deduced. For example, for the cohort released on the first day the number of surviving, inseminated females at the end of the experiment would be as follows: 10 nights*sex ratio*male insemination rate *20 females*daily female survival10. The sum over all 10 cohorts should then be equal to the total number of inseminated females at the end of the experiment. The average sex ratio over all 10 nights was used to deduce daily male insemination rates (see further elaboration below).

The experiment was performed four times, each time switching treatments between mesocosms to eliminate any effects of mesocosm differences unrelated to dietary treatment.

Temperature and humidity

Temperature and relative humidity differed slightly between different areas within the experimental enclosure. Mean (± SD) values for temperature and humidity were the following: within the plastic cage set on the floor 23.0 (± 2.67) oC and 63.3 % (± 16.0), within a resting tube in the mesocosm 24.5 (± 2.05) oC and 60.6 % (± 19.8), and on top of the hut in the mesocosm, 25.4 (± 3.21) oC and 56.6 % (± 20.5). The narrower range of temperature and humidity within the resting sites, and the slightly lower temperature and higher humidity on the floor, compared to the top of the hut, illustrate that the mosquitoes would be capable of evading the more inhospitable conditions. This was confirmed observationally; typically all mosquitoes were resting either close to the floor outside of

112

the resting hut, on the soaker hose, or (the majority) inside the resting tubes, with a preference for the lower tubes. Based on all 74 tallies throughout the experimental periods, an average of 81 ± 15% of mosquitoes were found resting in the tubes during the daytime, as opposed to elsewhere in the mesocosm.

Population projection matrix

The goal of developing a population model was to evaluate the consequences of reduced male insemination performance on population growth over many generations.

We chose to employ a population projection matrix, because mosquito populations are structured according to mating status (virgin and mated females, whereas males were not competent until rotation of the terminalia was complete) and, in the case of females, blood-feeding status (i.e., those that have obtained a blood meal, and those that have not).

When reproduction and survivorship are status-dependent, models that explicitly consider population structure are appropriate (Caswell 1989). Our model consisted of a stage- distribution vector and a transition matrix. The stage distribution vector, At, was composed of the elements N1-N6, which represent the numbers of males or females in any given stage at any point in time (Fig. 1). Equations for the transition values, a, between any two elements, that is, aij , are provided below. An overview of the parameters used in these equations is given in Table 1.

The transition matrix, Tt, corresponding to the stage structure in Fig. 1, had 3 regions and can be written as follows:

113

a11 0 0 F4 0 0

a21 a22 0 0 0 0 a 0 a 0 0 0 31 33 0 a42 a43 a44 0 0

0 0 0 F4 a55 0

0 0 0 0 a65 a66

The upper-left sector contains transition values for females (aij) as well as recruitment of females via reproduction from mated, blood-fed females (F4 α), where α is the primary sex ratio that we set at 0.5. The lower-left sector provides recruitment of males via reproduction from mated, blood-fed females. The lower-right sector harbors transition

values for males. Further, there are a number of non-transition elements (e.g., a11), where individuals remain within their current stage, as explained below.

Female transitions are determined by the following three parameters: (i) a mating term, ξ (0,1) which is a function of sex ratio; in the model described here, ξ = where is the sex ratio (i.e. ratio of competent males to virgin females) and s and sf are male insemination rates in sugar-present and sugar-free environments, respectively; (ii) a blood-feeding probability β; and (iii) survivorship terms e-μfs and e-μfsf in sugar-present and sugar-free environments, respectively. Finally, as we assume no natal control of sex

(i.e., α = 0.5), recruitment of females into the population is simply N4 R α, where R is the daily fecundity of a mated, blood-fed female. For males, recruitment is identical to females, and immature males (i.e., adult males that are not yet competent) mature at rate

114

e-η, and survivorship terms are e-μms and e-μmsf in sugar and sugar-free environments, respectively. This yields the following transitions (daily probabilities):

-(μf) a11 = (1- ξ) (1- β) e (a virgin, unfed female will survive but remain in the same state)

-(μf) a21 = (1- ξ) β e (a virgin, unfed female will survive, blood feed, but fail to mate)

-(μf) a31= ξ (1- β) e (a virgin, unfed female will survive, mate, but fail to blood feed)

-(μf) a22 = (1- ξ) e (a virgin, blood-fed female will survive but fail to mate)

-(μf) a33 = (1- β) e (a mated, unfed female will survive but fail to blood feed)

-(μf) a42 = ξ e (a virgin, blood-fed female will survive and mate)

-(μf) a43 = β e (a mated, unfed female will survive and blood feed)

-(μf) a44 = e (a mated, blood-fed female will survive)

-(η+ μm) a55 = (e ) (an immature male will survive but not mature)

-(η) -(μm) a65 = (1-e ) e (an immature male will survive and mature)

-(μm) a66 = e (a mature male will survive)

There are several key assumptions:

1) Males are capable of fewer inseminations per capita, per night, as sugar availability declines.

2) Females and males die at faster rates as sugar availability declines, i.e., μfsf > μfs and

μmsf > μms, respectively.

3) Probability of insemination is a linear positive function of the sex ratio.

115

4) Daily fecundity, R, reflects recruitment of adults into the population and therefore combines fecundity in terms of eggs laid per female and larval survival which is assumed to be density-independent.

We used the model to evaluate the importance of sugar availability (e.g., nectar) on the growth of a theoretical anopheline population. We started with an initial population with a quasi-stable age distribution and under environmental conditions allowing exponential growth. Because there are functions in the transition matrix, we were not able to solve for the stable distribution directly. Instead, we ran the model for 50 generations, obtained a nearly constant R, and used that distribution. The model was iterated over time (1 day at a time) with given values for the different parameters, and we observed population size and sex ratios. This was conducted independently for environments with and without sugar sources.

To confirm that the results of the simulations were not only valid for our experimentally derived values, we ran the simulations separately using values for survivorship and male insemination rate estimated from the literature (Gary and Foster

2001, Midega et al. 2007, Rodriguez, Aldridge, and Foster unpublished), and at different levels of fecundity. For the literature-based simulations, R was determined numerically to generate population growth that would yield an arbitrary 50-fold population increase in one year when sugar was readily available. Although this R (0.099) indicates a fairly low fecundity per blood meal taken, this figure for production of adult mosquitoes incorporates considerable egg and larval mortality. Further note that we do not incorporate the time investment and probability of finding a suitable oviposition site and

116

that females once blood-fed and mated do not revert to a previous stage, so that time spent on obtaining subsequent blood meals is also corrected for in our value of R. For the mesocosm-experiment-based simulations we set R at 1.03, giving a tenfold population- level increase over 100 d, which is comparable to the reported increase in mosquito density in houses over the course of the long rainy season in western Kenya (Minakawa et al. 2001, 2002). The rate of maturation, η, was set at 0.7 to yield 50% maturation within the first day of emergence. Finally, note that field populations normally do not continue to grow unimpeded for such a large number of generations.

In model runs, we arbitrarily implemented a sugar shortage starting at day 183.

Here, we considered 2 scenarios: (i) Only males are impacted in terms of insemination rate (60% of normal, or 27% based on our experiment) and survivorship (10% of normal, or 22% based on our experiment). (ii) Males and females are both impacted by sugar elimination. Here, we assumed that blood-feeding females without sugar survive at 80% longevity relative to females with access to both sugar and blood, whereas females that do not blood-feed with probability (1- β) survive at the same reduced rate as males.

Statistics

Effects of sugar availability on male mating performance, quantified as the proportion of females inseminated at the end of 10 d, were analyzed with Chi-square tests. Survivorship data were analyzed with Mann-Whitney tests. Wing lengths were normally distributed and compared with t tests. Biting data were analyzed with Friedman tests. All facets of the analysis employed SPSS v.16 (2007) software.

117

Results

Experiments

Presence of sugar had a significant and profound effect on the insemination of females (Fig. 2), both in the mesocosms (χ2 = 88.7, n = 598, P<0.05) and plastic cages (χ2

= 89.3, n = 335, P<0.05). In the mesocosms, over four replicates the mean percentage of females that had mated after 10 d was 49.7% when sugar was present and 10.9% when it was absent. This effect also was observed in the cages, yet here the insemination rates were higher, with sugar 76.1% and without sugar 23.5%. Calculations based on survivorship and sex-ratio estimates gave average daily male insemination rates of 0.25 and 0.09 in mesocosms with sugar present and absent, and 0.28 and 0.20 in cages with sugar present and absent, respectively.

Survival of mosquitoes, measured as the number of survivors present after 10 d, depended strongly on whether a sugar source was present or absent in the environment

(Fig. 3). Mann-Whitney tests indicated that this difference was significant for both males and females in both mesocosms and cages (U = 0, n = 8, P = 0.029 for all four comparisons). Average daily survival of females was 0.88 and 0.72 in mesocosms with sugar present and absent, and 0.99 and 0.94 in cages with sugar present and absent, respectively. Corresponding values for males were 0.83 and 0.50 in mesocosms with sugar present and absent, and 0.99 and 0.50 in cages with sugar present and absent, respectively.

118

The body size of females collected after 10 d from the mesocosm with only water was slightly, but not significantly, greater than those collected from the mesocosm where sugar was present, (t-test, df = 158, P = 0.11, 46.5% power) (Fig. 4). Surviving females in the plastic cages, and males in the mesocosms, showed no treatment differences in body size after 10 nights. Survival of males in cages without sugar was too low to allow a comparison.

Across the replicates, median time to initiate biting, averaged among the four replicates, was 8.2 min for females without sugar, and 9.9 min for females with sugar, a difference that is not significant (Friedman test, χ2 = 3.6, df = 1, P>0.05). It appeared that only on the second night, females with sugar present took substantially longer to land on their host (median bite, 16 min) than females without sugar (median bite, 7 min). Fig. 6 depicts the average number of bites per female per night for mesocosms with and without sugar. Estimates of the number of females present in the mesocosms on a given night were derived from survivorship values as described above. On eight of the 10 nights, females without sugar had a higher biting rate than females with sugar. Over all 10 nights females took an average of 0.30 and 0.20 blood meals per female per day, in mesocosms without and with sugar, respectively. Nonetheless, this difference was not significant

(Friedman test, χ2 = 3.6, df = 1, P = 0.058).

Models

We explored the results of eliminating access to sugar on population size after

183 days of growth using a 2-sex population projection matrix using values derived from

119

the literature (Fig. 7a) and values based on mesocosm experiments (Fig. 7b). Both graphs show male-only (dashed lines) and male-plus-female effects (solid lines).

For both scenarios in Fig. 7a there is an exponential increase in the mosquito population until day 183, after which the populations rapidly decline, most dramatically within the scenario with effects on both males and females. In Fig. 7a the male-only effect scenario shows a rapid decrease in mosquito numbers, mostly due to shifts in survivorship of males (μms = 0.034 vs μmsf = 0. 34), i.e., median survivorship drops from

20 to 2 days. From there, both male and female numbers declined due to reduced insemination effects and altered sex ratios, although more weeks were required to drive the population to near-zero than in the male-plus-female effect scenario. The importance of insemination rate in this decline was emphasized by the result that before reducing sugar, virgin females comprised 4% of the population versus 19% afterwards. The same pattern of exponential growth, then a plummeting of numbers, can be seen in Fig. 7b.

Here, the difference between scenarios was less pronounced, and the population decline sharper, due to the lower insemination rates and more severe mortality estimates obtained from our mesocosm experiments.

To explore the impact of reduced insemination rates and survivorship on population growth, we ran a sensitivity analysis with stepwise reductions of 10% for each of these parameters in each of the 2 scenarios. Figs. 8 and 9 show final population size after 183 days with sugar elimination, normalized against the maximum population size in the presence of sugar. As expected, each of the 2 parameters contributed to population reduction, but it was the survivorship parameter that had effects throughout the surface.

120

This was because reduced insemination rates due to behavioral factors could be mitigated by large numbers of males when they were able to survive for extended periods (we assumed no senescence or any loss of fertility across nights). On the other hand, when survivorship was low, there was a strong effect of insemination loss as well. This interaction was less obvious in the plots with 2-sex effects, because female death shifts the sex ratio back towards a 1:1 ratio, thereby reducing the impact of limited per-capita insemination.

Discussion

Insemination of female An. gambiae was shown to depend heavily upon the presence of sugar in the environment. Female An. gambiae, like female Aedes aegypti

(L.) (Harrington et al. 2001), are generally thought to survive primarily by blood feeding and only rarely take sugar meals in nature (Beier 1996). Therefore, the sex that affects insemination rates due to the absence of sugar in the environment can reasonably be assumed to be the males, which do (as adults) rely completely on sugar as a food source

(Yuval 1992, Foster 1995).

Without crop sugar and glycogen, both of which play an important role as fuel for flight in mosquitoes (Nayar and Van Handel 1971, Yuval et al. 1994), males are completely dependent on the reserves carried over from the larval stage to provide them with enough energy to sustain the flight activity necessary for swarming and mating behavior; here they were severely disadvantaged by an absence of sugar.

121

Besides affecting male An. gambiae by limiting flight energy, a lack of sugar also affected their survival. Holliday-Hanson et al. (1997) postulated that lipids converted from sugars may be used for resting metabolism, because starved males of An. freeborni were found to have drastically reduced lipid content. The lack of a chance to build up and replenish these reserves induced such mortality that these males were unlikely to reach an age of more than 2-3 days (Gary and Foster 2004). An. gambiae males show a peak in insemination capability around 7 days of age (Verhoek and Takken 1994), suggesting that sugar-deprived males not only have less energy to sustain patrolling and capturing flight within a swarm, but also may never reach the age at which their mating capabilities are at their full potential. The combination of these factors likely explains the reduced insemination rate of females in the environment where sugar was unavailable, with the decreased survival of males most strongly reducing insemination rates (Fig. 8).

Although the presence of sucrose increased insemination rates in both the plastic cages and the mesocosms, reduced enclosure volume also increased insemination rates, such that rates overall were higher in the cages than in the mesocosms. This may be because chance encounters between males and females are simply more frequent in a confined space; or females may be disturbed more often in these smaller confines, taking flight more frequently, consequently increasing their chance of being heard and seized by a male; or it may reflect the adaptation of the mosquito strain to mating in colony cages.

It is noteworthy that daily male insemination rate, in cages (following the formula ξ =

without sugar was 71% of (0.2/0.28) when sugar was present, whereas in a sugar- deprived mesocosm environment, is only 37% (0.09/0.245) of that when sugar is 122

available, clearly indicating that the effect of sugar availability on male insemination rate is itself influenced by enclosure size.

In a similar mesocosm study, within a single cohort of sugar-deprived mosquitoes, Gary et al. (2009) found lower insemination than we did, while in cages the level of insemination was comparable to that found here. Although Gary et al. (2009) found mesocosm males to be practically incapable of mating with females if sugar was not available, we found that on average 10.9% of females became inseminated even in the absence of sugar. A direct comparison was not possible, however, due to the differences in experimental design: 10 consecutive cohorts of mosquitoes released in this experiment versus a single cohort in theirs. In addition, Gary et al. used the Suakoko strain, established in 1987 by M. Coluzzi from mosquitoes originating in Suakoko, Bong,

Liberia. This strain has been used in laboratories for a significantly longer time than the

2001 Mbita strain used here. Although little is known about how fast An. gambiae adapts to mating in small cages and whether it simultaneously loses the ability to mate under more natural circumstances, it may help explain why the Suakoko strain performed worse under mesocosm conditions than the Mbita strain.

Our observation that sugar-deprived males with a high mortality could still inseminate 10% of the females poses the question whether some males sexually matured quickly and mated with females early in life, or whether a minority of males managed to survive longer and achieved most of the mating. While Verhoek and Takken (1994) found that on the second day after emergence male An. gambiae were capable of inseminating females to some extent, how that holds up in the absence of sugar and in a

123

much larger environment remains unclear. A relevant question that was not addressed in the present study is to what extent, when sugar is in short supply but not absent, the few males that gain access to sugar can make up for the insemination deficit. Survival of females in mesocosms with sugar was lower than survival in cages, but it did reflect mortality estimates of An. gambiae in the field. In our experiments, 52% of females were present at the end of the 10-d experiment, which equals 12% mortality per day. This mortality is close to field estimates obtained from mark-release-recapture experiments, which range from 16% (Gillies 1961) and 12-20% (Costantini et al. 1996) to 22% per day

(Takken et al. 1998). Daily mortality at the Kenyan coast has been estimated to be 5%

(Midega et al. 2007).

Although the differences in energy expenditure between mesocosms and cages may have contributed to the difference in survival, females might be expected to survive well in both types of enclosures in both the presence and absence of sugar, because they could compensate for the absence of sugar by producing fewer eggs or taking more blood. Fig. 6 indicates that the latter indeed occurred (though not significantly), but that this higher rate of blood feeding did not allow females to compensate for the lack of sugar in terms of survival. This observation is consistent with the results of Straif and

Beier (1996) and Gary and Foster (2001). Blood feeding also may have been much easier, and therefore more frequent in both treatments, in the cages. Another difference may have been the absence of predators in cages. Occasionally other arthropods, such as ants, small spiders (Salticidae), and centipedes, were found in the mesocosms and likely

124

disturbed and preyed upon mosquitoes, perhaps contributing to a field-like level of mortality.

Insemination levels for An. gambiae found in the sugar treatment of this experiment were comparable to rates in the field [88-98% for An. gambiae in Tanzania,

(Gillies and Chir 1956); 78% for An. arabiensis in Ethiopia, (Ameneshewa and Service

1996)], which suggests that sugar feeding must be a part of the behavioral repertoire of male An. gambiae in nature. If only 10% of females can become inseminated in the absence of sugar sources, does that mean this species is confined to areas where plants with nectar, or other sugar sources, are present? Our matrix population model certainly suggests that this is the case. Several of the underlying assumptions do require more information on the biology of this species. To obtain a more realistic R, for instance, more information on egg fertility and survivorship of immatures would be useful. How swarm size in nature affects mating [locating a small swarm of males might be considerably easier for females in a mesocosm than under true field conditions, where an

Allee effect may occur (Courchamp et al. 1999)] is relevant to the assumption that mating probability is a linear positive function of sex ratio. Even so, nectar availability in nature must certainly have a considerable impact on the dynamics and size of An. gambiae populations, which suggests a possibility of employing environmental management as a means of population control. The degree of population suppression by eliminating sugar availability, as exhibited in this study, is significant and certainly in line with that of other methods of population control. For example, Fillinger and Lindsay (2006) reported a 90% population reduction from using various Bacillus larvicidal formulations. Similarly,

125

Perez-Pacheco (2005) reported a 30-88% reduction of Anopheles larvae from using mermithid nematodes. Finally, Ansari and Razdan (2004) reported ca. 90% suppression of Anopheles culicifacies Giles from deltamethrin indoor spraying. Our work is similar to that of Gu et al. (2006) insofar as we also considered non-traditional control by manipulation of a life-history component on mosquito population growth rates: in our case, insemination rates, and in Gu et al., oviposition site arrival rates. In both cases, simple changes in rates of one component can have profound effects at the population level.

126

Table 3.1. Parameters used in equations for stage transitions.

Parameter Description Value (alit.) Value (bexp.) Source

ξ Mating term Varies

Sex ratio (proportion m:f) Varies

Male insemination rate 0.5; 0.3 0.245; 0.09 Rodriguez et al.

(♀♀/day) unpublished

β Blood feeding probability / 0.2

female / day

μ Death rate in environments Gary and Foster

with sugar or sugar-free (s/sf) of: 2001; Midega et

Males 0.034; 0.34 0.192; 0.693 al. 2007

Females, blood fed 0.028; 0.034 0.128; 0.16

Females, unfed 0.23 0.46

η Adult male maturation rate 0.7 Gary et al. 2009

R Daily production of offspring 0.099 1.03 Fitted to give

that survive to emerge as adults population growth

of a mated, blood fed female rates of 50x/year;

or 10x in 100 d

alit. = values estimated from literature or other sources bexp. = experimental results from this study

127

Fig. 3.1: Stage structures for male and female An. gambiae, schematically depicting the stage distribution vector (N1-6) and the transition matrix elements (aij).

Fig 3.2: Mean percentage of four replicates of females inseminated after 10 d with or without sugar, both with blood nightly, in large or small enclosures (mesocosm vs. cage, respectively). Newly emerged males and females were introduced daily.

128

Fig. 3.3: Survival of males and females after 10 d with or without sugar, both with blood nightly, in large or small enclosures (mesocosm vs. cage, respectively), based on four replicates and daily releases of newly emerged males and females.

Fig. 3.4: Mean size of females (wing length, n = 80) surviving after 10 d, with or without sugar in large or small enclosures (mesocosm vs. cage, respectively).

129

Fig. 3.5: Mean time (± SE) of median bite recorded within 20 min host-exposure period each night, for females in mesocosms with 10% sucrose vs. only water available.

Fig. 3.6: Average number of bites per female (± SE) per night, for females in mesocosms with 10% sucrose vs. only water available.

130

Fig. 3.7: Population decline following the removal of carbohydrate sources in the environment assuming only males are affected (dashed line) or both sexes are affected

(solid line) by a lack of sugar, for simulations based on values from the literature (a) or from the mesocosm experiments (b).

Fig. 3.8: Population size after 183 d at different levels of reduced male insemination rate and survival, expressed as a proportion of population size after 183 d when sugar is readily available, under the assumption that males are the only sex affected by a lack of carbohydrates in the environment.

131

Fig. 3.9: Population size after 183 d at different levels of reduced male insemination rate and survival, expressed as a proportion of population size after 183 d when sugar is readily available, under the assumption that both males and females are affected by a lack of carbohydrates in the environment.

132

Chapter 4: A survival and reproduction trade-off is resolved in accordance with

resource availability by virgin female mosquitoes*

Abstract

The first 2-4 days after an Anopheles gambiae female mosquito emerges are critical to her survival and reproductive success. Yet, the order of behavioural events

(mating, sugar feeding, blood feeding) during this time has received little attention. We discovered that among female cohorts sampled from emergence, sugar feeding has a higher probability than blood feeding of occurring first, and mating rarely occurs before a meal is taken. The night after emergence, 48% of females fed on sugar in mesocosms, and 25% fed on human blood; in the absence of sugar, 49% of females fed on human blood. After 5 days, 39% of the sugar-supplied females had blood fed, mated, and now were fructose negative, whereas by that time only 8% of the sugar-denied females had both blood fed and mated. The model that best explains the transitions suggests that females made use of two distinct behavioral pathways, the most common one being to sugar feed, then mate, and then seek blood. Other females sought blood first, then mated, and forewent a sugar meal. Lipid levels were higher in females with access to sugar than

* C.M. Stone, I.M. Hamilton and W.A. Foster. 2011. Animal Behaviour 81(4): 765-774. 133

in those without, particularly for females in later gonotrophic stages, while glycogen levels in the sugar-supplied group were higher throughout. In single-night experiments with females having had access to sucrose since emergence, those given a blood meal one day before spending a night with males had higher insemination rates than those not receiving the blood meal. These results indicate that the trade-off between survival and immediate reproduction is resolved by young adult females in accordance with availability of resources and gonotrophic state.

Introduction

The acquisition and maintenance of energetic reserves by animals serve as insurance against starvation. Reserve size is expected to vary with the costs and benefits of its maintenance (Witter and Cuthill 1993). For example, juvenile salmon (Salmo salar

Linnaeus) are thought to regulate energy intake according to a low target level of lipid reserves at the end of the winter according to a trade-off between starvation and predation-risk (Bull et al. 1996; Finstad et al. 2010). Predictability of the food supply also exerts an influence on the optimal reserve level, allowing, for instance, dominant great tits (Parus major L.) to maintain lower lipid reserves, and be less exposed to predation, than subordinates (Gentle and Gosler 2001).

For females of most anautogenous mosquitoes (i.e., requiring a blood meal to develop their first batch of eggs) the most efficient way to create and maintain reserves is

134

by feeding on plant sugar (Van Handel 1965). The main cost of recent sugar feeding is a reduction in subsequent blood-meal size, and therefore egg-batch size, as both compete for space in the mosquito abdomen (Mostowy and Foster 2004). Thus, females are faced with a choice between sugar feeding to rapidly increase their energetic reserves, or blood feeding, which contributes to fecundity through the development of eggs and to a smaller extent also supplements energetic reserves (Nayar and Sauerman 1975). Optimal diet choice of mosquitoes will thus have to balance somatic and reproductive demands

(Roitberg and Friend 1992). Females of certain species may fare just as well, or better, by excluding sugar from their diets. Aedes aegypti L., the yellow fever mosquito, appears to fit this last description. Life table experiments showed that age-specific survivorship

(Scott et al. 1997), reproductive potential and the basic reproduction rate (Scott et al.

1997; Harrington 2001) were all higher in females offered only blood and water than in females offered both blood and continuous access to a 10% sugar solution. This phenomenon was also shown by Braks et al. (2006) for A. albopictus Skuse.

Similarly, females of the highly anthropophilic species Anopheles gambiae Giles s.s. were shown to have a lower survival rate in the absence of sugar, but an increased blood-feeding rate (Gary and Foster 2001; Straif and Beier 1996). As a result, lifetime fecundity is roughly equivalent with and without sugar, but females lacking sugar achieve this fecundity in a shorter timeframe. For growing populations (Houston and McNamara

1999), sugar feeding therefore appears to have a fitness cost, begging the question: why partake in it?

135

Several studies indicate that some A. gambiae females do indeed use sugar. Beier

(1996) found that 10-17% of female A. gambiae s.l. collected during indoor biting catches in western Kenya tested positive for fructose, indicating recent plant-sugar ingestion. Foster and Takken (2004) found in olfactometer experiments that 1-d-old females strongly preferred honey-related volatiles over human-related volatiles, while the reverse was true for 5-d-old sugar-fed females. Several cage studies show an increased survival of this species when offered sugar solution (Gary and Foster, 2001) or access to nectar-bearing plants (Gary and Foster 2004; Impoinvil et al. 2004; Manda et al. 2007).

And females with access to putative host plants had higher fecundity after one blood meal

(compared to blood only access) yet not after three blood meals (Manda et al. 2007).

Thus, feeding on sugar before the first blood meal may be advantageous for females of this species. In laboratory cages, females fed on sugar between gonotrophic cycles, and they fed at higher frequencies if either an oviposition site or blood availability was delayed (Gary and Foster 2006). A recent theoretical study (Ma and Roitberg 2008) supports the expectation that sugar feeding in this species will be most prevalent after emergence and after oviposition. Field-based confirmation of these insights into the occurrence of sugar feeding remains absent at this point.

The following hypotheses, summarized in Table 1, may explain female A. gambiae‘s use of plant sugar early in life: a) Females take sugar predominantly at the first available opportunity and prefer it over a blood meal. Reasons for this might be to increase expected survival directly by supplementing meager energy reserves. The sugar priority may be related to blood-host-

136

seeking behaviour if the expression of host-seeking (Clements 1999) is enabled by sugar feeding. Alternatively, the sugar preference may stem from its facilitation not just of survival, but also of mating behavior (reviewed by Clements 1999; Yuval 2006). It fuels the swarming flight of males (Gary et al. 2009; Stone et al. 2009b) and, possibly, the mating activity of females, such as locating swarms and selecting mates (Cator et al.

2009; Gibson and Russell 2006; Pennetier et al. 2010; Warren et al. 2009). If females favour sugar initially, we expect that when both blood and sugar sources are present in the environment, behavioural sequences will start with sugar feeding. b) Females should not feed on sugar unless opportunities to blood feed are uncommon.

Thus, only those that would otherwise remain unfed and starve should accept a sugar meal; the proportion of females taking a blood meal initially is independent of the availability of sugar in the environment. c) Females are opportunistic feeders. Limiting the time spent searching for either food is the most important factor, because blood and sugar meals early in life have similar fitness values: they both supplement energetic reserves, allow for development of ovarian follicles to the resting stage (Fernandes and Briegel 2004), and provide fuel for flight.

Thus, food preference will merely reflect abundance of host types, ease in locating them, or access to them. If females are opportunistic, we expect that when both sugar and blood are available that behavioural sequences starting with either meal will be observed.

Without access to sugar, we expect the proportion of females feeding on blood initially to increase, compared to a situation where both resources are available.

137

The principal objective of this study was to establish the behavioural routine of young adult female A. gambiae under simulated natural conditions. This allowed us to determine whether they are obligate sugar-feeders, opportunistic sugar-feeders, or preferential blood feeders. By also investigating two plausible reasons that females include sugar in their diet, to enhance energy reserves or to facilitate mating, we gained further insight into how this species resolves the trade-off between enhanced survival prospects and earlier reproduction.

Materials & Methods

The mosquitoes used for these experiments came from a colony established in

2001 by staff at the International Centre of Insect Physiology and Ecology from the local population of A. gambiae s.s. in Mbita Point, Suba District, Nyanza, Kenya, and identified by PCR. This Mbita strain has been maintained in acrylic cages at 26.6±1 ○C and 80±5 % RH since 2006. Water and 10% sucrose solution were available to colony adults ad libitum. Blood feeding of adults on humans, for colony maintenance and for experiments, was covered under The Ohio State University‘s Biosafety protocol No.

2005R0020 and Biomedical IRB protocol No. 200440193, FWA No. 00006378.

Experimental mosquitoes were reared by transferring 100 newly hatched 1st instar larvae into 22.8 x 33 cm pans filled with 450 ml of aged tap water and feeding them 0.2 mg of finely ground TetraminTM fish flakes per larva during the first 3 d of larval development,

0.4 mg on each of the next 3 d, and 0.8 mg on subsequent days until pupation.

138

Experimental mosquitoes inhabited two mesocosms at The Ohio State University, moist-floored 9.1 m3 greenhouse enclosures designed to simulate conditions of endemic habitats in equatorial Africa. A detailed description is given in Stone et al. (2009a), with a few adjustments described here. A container of water was placed inside a chimney-like construct that housed resting sites, to elevate humidity within this structure. Mosquitoes rested in 30 black cardboard tubes projecting into the chimney. To compare the behaviour and die-off of mosquitoes in the large mesocosm environment to that in laboratory cages, a clear acrylic cage (36 x 52 x 46 cm) was situated on 3 bricks in a water-filled tray on the floor of each mesocosm, in which the same experiment was performed simultaneously. A black cup (10 cm in diameter, 8 cm in depth) oriented sideways and glued to the wall served as a resting site.

Lights in the mesocosms came on at 07.00 and remained on until 18.00, leaving enough time for the natural, diminishing light of the evening crepuscular period at this latitude to facilitate mating.

Treatments consisted of the presence or absence of sugar in the environment. In the sugar-supplied mesocosm, four potted plants with extra-floral nectaries (two Senna didymobotrya and two Ricinus communis) were placed in aluminium-wrapped trays on the floor; water in the trays served as a moat to keep ants from competing with the mosquitoes for nectar. The Ricinus plants produced a large amount of visible extra-floral nectar. To ensure that sugar availability would not be limited in the sugar treatment, in addition to the plants, wicks of sucrose and honey solution were suspended from a bank of lights in the mesocosm. Wicks protruded from six vials of 10% sucrose solution with

139

0.001 mg/L verbenone added for scent (Gary and Foster 2006), and six wicks protruded from vials of 10% multifloral honey solution. In cages, one sucrose wick and one honey wick were available to mosquitoes. These solutions and wicks were replenished every 2 days. In the sugar-denied mesocosm and cage, the wicks held water, and an artificial

(plastic) Ficus tree was placed in the mesocosm, to provide a plant-like structure.

Behavioural Sequences and Reserve Acquisition With Blood but With or Without Sugar

At the start of each replicate experiment, 400 pupae were placed in each mesocosm, and 200 in each cage. They emerged as adults overnight. Females were allowed to blood feed for 30 min on a human host (C.S.) each evening between 21:00 and

22:00, starting the day after emergence. The order in which mesocosms were given access to a host was switched each evening.

Every morning, between 07.00 and 08.00, 20 females per mesocosm, and 10 per cage, were removed from resting sites by mouth aspirator, killed and fixed at -40 ○C for later dissection and chemical analysis. After 5 days of sampling, the number of females in the water-only mesocosm had become depleted (only with considerable effort could a full sample be obtained), and the experiment was stopped at this point. It is likely that some predation occurred, because a small plump spider was found in one of the mesocosms on one occasion, and a centipede was found outside its perimeter on another.

In addition to the 400 pupae, on the first night, 200 5-d-old virgin males that had been kept in cages with ad libitum access to sugar, i.e. sexually mature and well fed, were released into the mesocosms. To ensure that adequate numbers of competent males would

140

be present throughout the experimental period in both treatments (preliminary cage studies had revealed good survival of such males for only 2 days after sugar was withheld), on the third night another 150 5-d-old sugar-fed males were released in both mesocosms. The same procedures were applied to the cages, but with half the number of males.

Because females taking blood one or two nights after emergence would have been gravid before the end of the experiment, an oviposition site (a plastic tray with a layer of water in the mesocosms; a petri dish with water in the cages) was made available to avoid behavioral aberrations associated with delayed oviposition. In the mesocosms females could also use the moats holding the plants, or even the moist carpet on the floor, for this purpose.

The following information was gathered for each individual female removed from the experiment: a) Sella‘s stage of blood digestion and gonotrophic progress, based on the relative sizes of the digesting blood meal and swelling ovaries; b) Christopher‘s stage of a few primary ovarian follicles, to further assess if and when a blood meal was taken; c) wing length, as an indicator of body size; d) presence or absence of tracheolar skeins on an ovary of any non-blood-fed female (Detinova method), to distinguish nulliparous from parous females (only of females collected on days 4 and 5); e) the presence or absence of sperm in the spermatheca; and f) energetic reserves: each body, minus the ovaries of non- blood-fed and non-gravid (i.e., ―empty‖) females collected on days 4 and 5, was then stored individually in an Eppendorf tube and frozen until its energetic reserves were measured, using the methods of Van Handel (1985) and Van Handel & Day (1988).

141

Females were crushed in 0.2 ml of 2% sodium sulfate solution, and centrifuged for 15 min after 1.5 ml of chloroform:methanol (1:2) was added to each tube. Half of the supernatant was used for lipid analysis, half for fructose, and the precipitate for glygocen analysis. Lipids were assayed by evaporating the supernatant set aside for this, adding 0.2 ml sulfuric acid and heating it at 90°C for 10 min. After adding 4.8 ml of vanillin- phosphoric acid reagent, absorbance was read at 525 nm on a spectrophotometer. The amount of fructose was determined by evaporating the second half of the supernatant to about 0.1 ml, adding 4.9 ml of anthrone reagent and incubating at 26°C for one hour, after which the absorbance was read at 625 nm. Glycogen was determined by adding 5 ml of anthrone reagent to the precipitate, heating at 90°C for 15 min, and measuring absorbance at 625 nm. Standards were glucose, sucrose and soybean oil solutions.

Amounts (µg) of carbohydrates and lipids were transformed into calories by multiplying by 0.004 and 0.009, respectively. Fructose quantities, which indicate undigested crop sugar from recent sugar feeding, were measured for all females; lipid and glycogen amounts per female were measured for the final two replicates. The glycogen and lipid values obtained included both the maternal reserves and, in the case of blood-fed and gravid females, also those materials contained within the developing egg follicles and fully developed eggs.

Females that had taken a blood meal, regardless of the state of digestion of the meal and consequent development of vitellogenic eggs, were scored as blood fed, as were females that appeared unfed but were parous, having already laid the eggs. Females were classified according to their internal state: eight combinations (N 1-8), consisting of their

142

blood-feeding (gonotrophic and post-gonotrophic) status, insemination status, and fructose positivity. Altogether, four replicates of the experiment were performed.

For analysis of behavioural sequences, population vectors for the 5 days of the experiment were based on the proportions of females in each internal state. The corresponding behavioural transitions between internal states over 5 days were estimated by utilizing Wood‘s quadratic programming method (Caswell 2001). A number of different models were considered, and the best among these was selected by projecting the population using the estimated transition matrix, obtaining the residual sum of squares by comparing the real with the simulated data, and calculating the Akaike‘s

Information Criterion (AICc) for small samples for each model. AIC is a relative measure of the information lost when using a model, which can then be used to select among alternative models. The model or models with the lowest AIC values are those that best explain the data with a minimum of free parameters. It allows for a ranking of competing models based on fit, while accounting for the number of estimated parameters (Burnham and Anderson 2002).

Bootstrap-selection frequencies were calculated to provide information on the uncertainty of the best model. Bootstrap samples were obtained by combining data from all four replicates and sampling with replacement from this data set for each day separately. A single data point consisted of the scored internal state of a female. One bootstrap sample then consisted of a matrix of proportions of eight states for each of 5 days. Subsequently, for each of 2000 bootstrap samples a transition matrix and an AICc value were calculated for each model, and the best model selected. The selection

143

frequencies were obtained by tallying the number of times each model was selected as the best model.

Effect of Blood Meal or Sugar Feeding on Insemination

Further experiments were done to obtain independent evidence that blood-fed only, sugar-and-blood-fed, and sugar-fed only females are equally capable of mating while digesting a meal. We separated males and females 1 day after emergence and provided them with continuous access to 10% sucrose in clear acrylic cages. Of 100 sugar-fed females, 50 were allowed to take a replete blood meal on day 3. In the late afternoon of day 4, 100 males and all 100 females were released into one mesocosm with the four Ricinus and Senna plants in it. The following morning all mosquitoes were collected by aspirator and frozen. Presence or absence of blood in the midgut indicated the meal history of each female. The spermatheca was removed and inspected by compound microscope for the presence or absence of sperm. This experiment was replicated five times. In a variation of that experiment, 50 females had access only to sucrose from day 1, whereas the other 50 had no sucrose but took their blood meals on day 2. Then all 100 females were released into the mesocosm along with 100 same-age sugar-fed males on day 3 and were collected the following morning. This experiment also was replicated five times.

Statistics & data analysis

144

Differences in energy reserves and proportions of females in different behavioural states were tested for normality and analyzed with t tests or Mann-Whitney U tests. An effect of fructose content on insemination status in mesocosms was analyzed with binary logistic regression. Insemination-rate differences between females that had taken blood or not, and the proportions of females surviving for 5 days were analyzed with Pearson‘s chi-square tests. All facets of the analysis employed SPSS v.16 (2007) software.

Calculations of AICc values and bootstrap selection frequencies were performed in

MATLAB (2009a, The Mathworks).

Results

Mesocosms with Sugar

Sugar feeding in the sugar-supplied mesocosm was observed during early scotophase. It occurred mainly on the plants rather than on the sugar and honey wicks.

Mosquitoes probed near the tips of Senna leaflets, as well as on the extra-floral nectaries located at the base of the petioles. The number of males and females probing on the plants increased after swarming had stopped. Approximately 1 hr after sunset large numbers of mosquitoes were on the plants, primarily Senna, upon the leaflets of which tiny droplets of sweet liquid were visible.

Judging from the internal states on day 1 (Fig. 1), a small proportion of females already had taken sugar during the night of emergence, but the vast majority (92.5%) had not engaged in any of the three behaviours under study. After the following night, 24%

145

remained in the initial state, but the majority had at this point acquired a meal, either blood (25%) or sugar (48%); the differences in proportions of females feeding on either blood or sugar was not significant (t 6 = -2.05, P = 0.086). On day 3 the largest proportions of females were those that were both fructose positive and had mated (24%), and those that were fructose positive but had not mated (21%). Females that had taken a blood meal and mated successfully comprised 19% of the population at this time, while

14% had taken a blood meal without having mated. By day 4 the largest proportion consisted of females who had mated, blood fed, and were fructose negative (39%), which was not significantly different from the 23% of females who had mated and were fructose positive, but had not fed on blood (t 6 = 0.96, P = 0.37).

Several models were constructed to elucidate the transitions between these states.

The initial model (Model 1) allowed females to perform one action (blood feed, sugar feed or mate) per night, or to remain within their current states. The transitions for this model are shown in Fig. 2. Notable is the low estimate for a21, suggesting that mating before taking a meal was rare. Low values for a63 and a64 confirm that feeding on sugar was typically not followed directly by feeding on blood, and vice versa, while the value of a88 appears artificially low (which might be interpreted as low survival rate of N8 = blood-fed, mated, fructose positive).

Model 2 (Fig. 3) was used to refine Model 1, by removing the pathways to states

N6 (blood-fed, virgin, fructose positive) and N8 (blood-fed, mated, fructose positive) from other states, and from the unfed-virgin to unfed-mated state (a21). As the anthrone test detects only recent sugar meals, Model 2 allowed for the possibility that females

146

could digest a sugar meal completely, and it opened the pathways from the sugar-fed- mated state to the mated state (a25), and also to the blood-fed-mated state (a75).

The following three models differed slightly and were constructed to answer three questions: 1) whether sugar feeding necessarily precedes blood feeding, which would suggest that sugar has a role in supplying energy for host-seeking behavior (Model 3); 2) whether sugar and blood feeding are both viable options that precede mating (Model 4), which is the same as Model 2 but lacks the pathway from the sugar-fed-mated state to the unfed-mated state; and 3) whether females are entirely opportunistic, i.e., unfed virgins seek either sugar or blood, and those that take sugar subsequently mate or take a blood meal (Model 5).

Among these five models, Model 4 had the most support, based on the AICc values (Table 2). Model 5, where sugar-feeding precedes either mating or a blood meal, had considerably less support. The first three models had virtually no support. That mating occurred only following a meal suggests that the expression of mating behavior depends on energy reserves. The strong support for Model 4 suggests that feeding on sugar specifically enables the behavioral sequence of mating followed by host-seeking, rather than providing energy for opportunistically mating or finding a blood meal. It also suggests that both blood and sugar are viable options for an initial meal, though feeding on sugar first is more likely under the experimental circumstances.

The bootstrap-selection frequencies of the five tested models reinforce this conclusion and provide information on the robustness of Model 4, which was selected as the best model in 80.5% of bootstrap samples, whereas Model 5, the model that allows

147

for more opportunistic behaviour, was selected as the best model in 18.2% of the samples. Model 3 was the best model in only 1.3% of the cases, while Models 1 and 2 were never selected.

This analysis suggests that over the first 5 days of A. gambiae females‘ lives the most common sequence of behaviors is to rest on the night of emergence, feed on sugar on the following night, then find a mate, and take a blood meal. However, a fairly large proportion appears to make use of a different pathway, by taking blood first, then mating, and forsaking sugar altogether. Among females that, during the first 3 days, had taken sugar but had not yet mated, the mean wing length (2.92mm) did not differ significantly from that of females who fed on blood instead (2.87mm) (ANOVA, F1, 96 = 1.46, P =

0.229). This indicates that female size was not a factor that determined initial meal choice.

Mesocosm without Sugar

The behavioral pattern of females in mesocosms lacking sugar was more straightforward (Fig. 4). On the night of emergence, most females (99%) again engaged in neither mating nor blood feeding. On the following night, however, 49% had taken a blood meal, which was a significantly greater proportion than females who had taken a blood meal on that night when sugar was available (U = 16, N = 8, P < 0.02). On days 3 and 4, 75% were blood-fed virgins. On day 5, 8% of females were blood fed and had successfully mated; this did not differ significantly from the proportion of females in that

148

state when sugar was available (U = 13.5, N = 8, P = 0.09). Females that mated before obtaining a meal were very rare throughout, as in the sugar-supplied mesocosms.

Cages with and without Sugar

In the cage study, behavioural transition patterns were largely the same as in mesocosms, with a few marked differences (Fig. 5). Where sugar was available, a significantly greater proportion of females had already taken a sugar meal (25%) on the night of emergence (U = 16, N = 8, P < 0.02). On the last 2 days, more than 50% of females tested negative for fructose, but they had taken a blood meal and had mated.

Thus, the pattern was similar to that in the mesocosm, but the end point, when females have mated and obtained a blood meal, was reached 1 day faster than in the larger space of the mesocosm. The main difference between females in cages and mesocosms with access only to water and blood was an only marginally higher percentage of females that had both blood fed and mated (U = 2, N = 8, P = 0.074). On day 4 this was 18% of the females (cf. 8% for females in mesocosms) (Fig. 6).

Energy Reserves

The mean glycogen and lipid reserves per day differed sharply between females in mesocosms with and without sugar (Table 3). Throughout the experimental period, lipid levels of females with access to sugar were higher than those without, the difference being most distinct on the last day. Glycogen levels also were substantially higher in females in the presence of sugar sources from day 1 onwards. The same trend occurred in

149

females kept in cages. Here, however, on the last 2 days of the experiment, glycogen levels were not different between treatments.

From a comparison of the energetic reserves of females according to gonotrophic state, in sugar-supplied and sugar-denied treatments, it is evident that glycogen and lipid levels were usually higher in females with environmental sugar (Figs. 7 & 8). The exception was the lipid levels of freshly blood-fed females, which were insignificantly different.

In mesocosms, fructose-positive females that had mated contained lower levels of fructose than those that had not (Table 3). The most likely explanation is that mated females took their sugar meals earlier, and had used or converted more of it, which is congruent with the interpretation that females usually took a sugar meal before mating.

This pattern was not evident for females in cages.

Survival

Daily survival values were derived from the average number of surviving females after 5 days, assuming a constant rate of mortality and accounting for removal through sampling. Females in mesocosms containing or lacking sugar had daily survival values of

2 0.94 and 0.86, respectively. The difference in the number of survivors was significant (χ 1

= 95.1, P<0.05). Females in cages containing or lacking sugar had daily survival values of 0.96 and 0.94, respectively, and the number of surviving females differed significantly

2 ((χ 1 = 7.67, P<0.05).

150

Effect of Blood Meal on Insemination

Of the sugar-fed females allowed to mate with 5-d-old, sugar-fed males in a

2 mesocosm on night 4, a significantly greater proportion (χ 1 = 31.8, P<0.05) that took a blood meal on day 3 were inseminated (71%), compared to females that had had only sugar (41%). Likewise, the insemination rate of sugar-deprived females that had taken blood on day 2, then allowed to mate on night 3, was significantly greater (56%) than

2 those with sugar (21%) (χ 1 = 35.5, P<0.05). All five replicates of both experiments showed the same trend.

Discussion

The analysis and evaluation of data derived from examining the status of blood feeding, sugar feeding, and insemination of females, sampled on sequential days after emergence in mesocosms, revealed the following: When sugar was present in the mesocosm, most females fed first on sugar, then mated, then fed on blood. But a substantial proportion instead took blood first, then mated, bypassing sugar feeding. The significance of this duality of sequences was supported by the model based on transition frequencies. In the absence of sugar, a much larger proportion fed first on blood, then mated. Is it possible then, to determine the timing and necessity of sugar feeding in this species? We hypothesized three distinct early life behaviours: females strongly prefer sugar, strongly prefer blood, or are entirely opportunistic with regards to their initial meal

151

(Table 1). The observed data favour the opportunistic hypothesis. Were females to favour sugar initially (hypothesis 1) we would expect all observed behavioural sequences to start with sugar feeding. The presence of two sequences, one that starts with sugar feeding and one that starts with blood feeding thus does not lend support for hypothesis 1. It also does not appear that sugar feeding is a requirement for successful mating in females, as sugar- deprived, blood-fed females became inseminated at a higher rate.

The results do not support the hypothesis that females favour blood initially, taking sugar only if they fail to obtain a blood meal (hypothesis 2). If they were to do so, one would not expect such a large difference between treatments in the proportion of females feeding on blood. That approximately one quarter of females fed on blood when sugar was available on the night after emergence, and half did so when sugar was not available, lends strong support for hypothesis 3: females are opportunistic with regards to their initial meal choice. Females were willing to feed on either blood or sugar initially when both were available, and the absence of environmental sugar resulted in an increased likelihood of feeding on blood.

The appearance of a sugar priority—i.e., when both resources were present, either one might be used, but sugar more often—could be primarily a reflection of environmental access, based on random encounter with the stimuli associated with each food, because blood was available for a relatively brief period each night but abundant sugar sources were available throughout the night. In nature, such a situation is common for newly eclosed females emerging from habitats distant from human habitations, or in habitations where bed nets insulate humans from mosquitoes a few hours after dark.

152

Blood-host availability has been shown to affect the response of mosquitoes (Hancock and Foster 1997) and biting midges (Garcia-Saenz et al. 2010) to blood hosts, and it probably affects blood-feeding patterns of mosquitoes (Chaves et al. 2010). The extent to which the host response of A. gambiae depends on the presence and strength of sugar cues remains to be investigated, but it is likely to affect food choice (i.e., the exact values of parameters a21 and a31 may differ).

Wing lengths of 1-d-old females that had taken either a blood meal or a sugar meal first did not differ, indicating that initial meal choice of females is not size- dependent. However, mosquitoes used in these experiments were all reared under the same densities and food levels, and most likely were comparable to the large females reported by Takken et al. (1998) and Fernandes and Briegel (2004) to use their first blood meal for vitellogenesis. Our failure to observe any gonotrophic discordance (Briegel and

Hörler 1993), a gonoinactive state (―pre-gravid‖ state, sensu Gillies 1954), or other deviations from the typical blood-digestion and egg-development pattern in our females

(unpublished data) supports this conclusion. Females with shorter wings, indicating crowding or food stress during immature development, may choose their initial meal differently as a result of differing, size-dependent benefits of sugar- and blood-meals

(Roitberg et al. 2010).

Finally, we note that studies on fitness consequences of female mosquito diet

(feeding on blood, or on blood and sugar) have all been performed in laboratory cages

(e.g., Gary and Foster 2001; Scott et al. 1997; Fernandes and Briegel 2004). Lifetime fecundity in these small-scale environments may differ from that in more energetically

153

demanding large-scale surroundings. Life tables derived in more realistic environments may well indicate that females do in fact benefit by including sugar in their diet. Our comparison of the daily survival values derived from cages and mesocosms suggest that the difference in survival on diets with and without sugar is underestimated in cage studies.

Glycogen levels of females in mesocosms with and without access to sugar differed more pronouncedly than did lipid levels, at least until the final day. According to Nayar and

Van Handel (1971) only sugars and glycogen are used for flight in mosquitoes, whereas triglycerides contribute to resting metabolism. If glycogen is mainly used to sustain flight, this would support the hypothesis that females take a sugar meal early on to aid them in engaging in mating behavior, and/or host seeking. This fits with the most probable behavioral sequence constructed from the data. However, results of Kaufmann and Briegel (2004) muddle this picture somewhat, showing that in A. gambiae carbohydrates contribute significantly to survival, and lipid is mobilized for long-range flight, but hardly at all for survival. Among the late-stage blood-fed and gravid females in our mesocosms, those with sugar had higher reserves of both glycogen and lipid (Figs. 7

&8). Individuals having more energy reserves before taking a blood meal may be able to convert a larger part of the meal into eggs. This is supported by Manda et al. (2007), who found that after one blood meal, females with access to sugar had higher fecundity.

Tests of the effect of blood-feeding on mating showed that blood-fed females, whether sugar-fed or not, had higher insemination rates than non-blood-fed females. This unexpected result suggests that blood is a sufficient energetic resource for A. gambiae to

154

locate and join a male swarm. The promoting effect of blood may be behavioral rather than energetic, if the blood meal causes a female to prioritize insemination so that eggs will be fertilized as soon as they mature, as few as 2-3 days later. Alternatively, she may be easier to grasp by swarming males or may be more attractive to them. The plausibility of explanations based on maneuverability depends in part on females having control over mating, but the extent to which anopheline females do so (beyond the decision to enter a swarm) is unresolved. The higher insemination rate of blood-fed females fails to support the hypothesis that females should initially feed on sugar in order to enable mating behavior itself. In nature, the meal used to facilitate mating may depend more on the proximity of mating arenas relative to adult emergence sites, vegetation and blood hosts, than on the flight-enabling effects of a blood or sugar meal.

Gillies (1954) suggested that females typically take a non-vitellogenic (―pre- gravid‖) blood meal before mating, but among mosquitoes caught in houses he considered only blood-fed females, discarding unfed (though possibly sugar-fed) females.

In a theoretical study Onyabe et al. (1997) concluded that females should be entirely opportunistic with regards to taking a blood meal before or after mating. A field study on the islands of São Tomé and Príncipe by Charlwood et al. (2003) supports this: A. gambiae females mated either before or after blood feeding in roughly equal proportions.

Our results confirm the optional nature of the blood-mate sequence, but in addition indicate that a meal—be it blood or sugar—is not optional and is a normal precondition to mating in this species. The choice between these two resources is, in effect, similar to the decision faced by parasitoids that use hosts for either oviposition or feeding, reflective

155

of a trade-off between current and future reproduction (Heimpel and Rosenheim 1995).

Collier et al. (1994) suggest that the decision to host-feed depends on a threshold egg load, and that this threshold in turn depends on the probability of encountering hosts.

How then do mosquitoes resolve the trade-off between increasing expected survival and more immediate reproduction? With regards to foraging for blood hosts,

Costantini et al. (1998) suggested that the outcome of the decision-making process relies on the stimulus strength of the chosen host, the stimulus strengths of competing hosts, the internal state of the female, and a genetic disposition to respond to the various stimuli

(i.e., an intrinsic ranking of host types). That genetic variation underlies this behaviour was elegantly shown by Gillies (1964), who was able to select for host preference in A. gambiae and significantly change females‘ preference after a few generations. It is feasible that the degree to which A. gambiae respond to nectar-related or human volatiles has a similar genetic basis.

A female‘s internal state (e.g. gonotrophic state, size, energetic reserves) probably also influences her response to various stimuli. Houston & McNamara (1985) suggest the energetic state of animals may allow the incorporation of sub-optimal items into a diet, if the energetic gain of the preferred food type is highly variable. The prediction is that individuals with high levels of reserves should be risk averse, and display a higher degree of opportunism. For mosquitoes, the degree of defensive behaviour of potential blood- hosts, with its associated costs, introduces variability in the likelihood of successfully obtaining a meal. This is concomitant with the—at first sight surprising—theoretical results showing that female mosquitoes that are close to starvation should seek blood

156

hosts but ignore sugar sources (Roitberg and Friend 1992). This suggests that the use of two behavioral pathways, as found in this study, may be a result of variance in energetic reserves at emergence.

157

0.9

0.8 Virgin, unfed, non sugar 0.7 fed 0.6 Mated, unfed, non sugar fed 0.5 Virgin, blood fed, non 0.4 sugar fed Virgin, unfed, sugar fed 0.3

0.2

0.1

0 0 1 2 3 4

0.9

0.8

0.7 mated, blood fed, non sugar fed 0.6 mated, unfed, sugar fed

0.5 virgin, blood- & sugar 0.4 fed mated, blood- & sugar 0.3 fed

0.2

0.1

0 0 1 2 3 4

Fig. 4.1: mean proportions +SE (four replicates) of females in internal-state categories per day after emergence in the mesocosms with sugar available (0 or 1 actions above; 2 or 3 below).

158

Fig. 4.2: path diagrams of internal states and transition-matrix values for Model 1. Arrows without a value indicate zero.

159

Fig. 4.3: path diagrams of internal states and transition-matrix values for Models 2, 3, 4 and 5.

160

0.9

0.8

0.7

0.6 virgin, unfed mated, unfed 0.5 virgin, blood fed 0.4 mated, blood fed 0.3

0.2

0.1

0 0 1 2 3 4

Fig. 4.4: mean proportions + SE (four replicates) of females in internal-state categories per day after emergence in the mesocosms without sugar.

1 0.9

0.8 virgin, unfed, non 0.7 sugar fed 0.6 mated, no blood- & sugar feeding 0.5 virgin, blood fed, non 0.4 sugar 0.3 virgin, unfed, sugar fed 0.2

0.1 0 0 1 2 3 4

Fig. 4.5: mean proportions + SE (of four replicates) of females in internal-state categories per day after emergence in the cages with sugar.

161

1 0.9 0.8 0.7 virgin, unfed 0.6 mated, unfed 0.5 virgin, blood fed 0.4 mated, blood fed 0.3 0.2 0.1 0 0 1 2 3 4

Fig. 4.6: mean proportions + SE (of four replicates) of females in internal-state categories per day after emergence in the cages without sugar.

0.25

0.2

0.15 w ater plants 0.1

0.05 Mean glycogen cal / female glycogen/ Mean cal 0 Unfed Freshly fed Late stage through Gravid

Fig. 4.7: mean glycogen levels + SE of females in different stages of blood meal digestion, for mesocosms with sugar / nectar or mesocosms with only water and blood. Differences were significant for all three groups (t tests, N = 46 – 166, P < 0.05).

162

0.9 0.8 0.7 0.6 0.5 w ater 0.4 plants 0.3

0.2 Mean lipid cal / female / lipid cal Mean 0.1 0 Unfed Freshly fed Late stage through Gravid

Fig. 4.8: mean lipid content + SE of females in different stages of blood meal digestion, for mesocosms with sugar / nectar or mesocosms with only water and blood. Differences were significant for all but freshly fed females (t tests, N = 41 – 160, P < 0.05).

163

Chapter 5: The first meal choice (blood vs. sugar) of the malaria mosquito

Anopheles gambiae s.s. is affected by bed net use and female size, but not plant

abundance*

Abstract

We investigated how the sugar or blood meal choice of Anopheles gambiae females in a mesocosm one day after emergence is influenced by aspects of blood-host presence and accessibility (i.e., untreated bed net usage), nectariferous plant abundance

(1 or 6 potted Senna didymobotrya), and female size as indicator of energetic state.

Rather than obligatory sugar feeders after emergence, females were opportunistic with regard to their initial meal. When a blood host was accessible for 8 hours per night, 92 % fed on blood, and only 3.7 % fed on sugar. Even with the use of an untreated bed net, 78

% managed to obtain a blood meal during the 30 min in the evening or at dawn when a host was present but not covered by the net, and 14 % of females were now fructose positive. Mosquitoes that fed on both resources were more often small females that had taken a sugar meal earlier in the night. This indicates that a change in biting times as a result of (untreated) bed net usage can be explained largely by the inherent plasticity of

* C.M. Stone, B.T. Jackson and W.A. Foster, in preparation. 164

mosquito behaviour. The diversion of energetically deprived mosquitoes to sugar sources suggests a possible synergy between bed nets and sugar-based control methods.

Introduction

The sub-Saharan malaria vector Anopheles gambiae Giles s.s. uses two nutrient sources, plant sugar and blood (Gary and Foster 2006; Manda et al. 2007a). A female will take a relatively small number of large meals throughout her life, imparting every single feeding decision with importance (Foster 1995). But how the abundance of sugar-bearing plants and the accessibility of blood hosts in an environment affect the decisions to feed on either sugar or blood is currently not well understood, even though it pertains to malaria epidemiology through its effects on vectorial capacity (Gary and Foster 2001; Gu et al.

2011). It also pertains to vector control through its implications for behaviour of mosquitoes around bed nets and for novel sugar-based control methods that exploit the sugar-feeding habit of mosquitoes. A promising example of such a method uses attractive toxic sugar baits (ATSB) that employ fruit scents to attract both male and female mosquitoes, a sucrose solution to stimulate feeding, and an oral insecticide (Müller and

Schlein 2006). This method has been shown to be highly effective in arid areas for a variety of mosquito species (Müller et al. 2010b; Müller and Schlein 2008; Schlein and

Müller 2008), and, of importance for malaria control prospects, in an area in Mali its use resulted in effective control of a population of An. gambiae (Müller et al. 2010a).

165

In recent years access to insecticide treated nets (ITNs) has increased greatly in sub-

Saharan Africa, sufficient to cover an estimated 76% of persons at risk by 2010 (WHO

2010), which has resulted in a decrease of malaria incidence in many countries. Perhaps the greatest threat to this progress is the selection for resistance to pyrethroid insecticides

(N‘Guessan et al. 2007), and additional methods that could aid in managing resistance will become increasingly important (Curtis et al. 1998). If mosquitoes can be diverted from nets to sugar sources, the use of oral insecticide-laced sugar baits placed alongside nets to manage resistance may be worth considering. Very little is currently known of the likelihood of feeding on sugar in the presence of a human covered by a net, save one report that mortality due to starvation in artificial huts could be minimized by placing glucose pads in the hut (Curtis et al. 1996). Behavioural resistance to bed nets—meaning that vectors may respond to a high degree of bed net coverage by changing their feeding behaviour accordingly, by biting at different times when people are not protected by nets or by developing a higher degree of exophagy, as a result of either adaptation or inherent plasticity in behaviour—is also a concern (Russell et al. 2011a; Takken 2002). Potentially this type of resistance could be facilitated by feeding on sugar, if females make use of this resource to avoid starvation after being repelled from a hut. Indeed, one of the great knowledge gaps of mosquito behaviour is what decisions females make after exiting a hut with a net, and how likely they are to be diverted to a non-human, a different human, nectar, or instead to a resting site. Yet it has been realized that to reduce infection levels beyond that which is attainable with a high degree of treated net coverage, it will be

166

necessary to employ additional control methods that target exophagic and early biting mosquitoes (Russell et al. 2010a). Thus there is a need for novel control methods that are synergistic with ITNs (Ferguson et al. 2010; Shaukat et al. 2010). A prerequisite to assessing the potential of sugar-based control methods to complement bed net programmes is knowledge of whether mosquitoes are likely to obtain sugar after a failed attempt to feed on a human under a net.

How access to blood hosts and nectariferous plants affects the feeding decisions of An. gambiae also relates to the question in which type of environments sugar-based control methods are likely to be fruitful. The feasibility of using ATSB as a stand-alone control effort or as a part of integrated vector management (Beier et al. 2008) in a given area will depend to a large degree on the frequency of sugar feeding by female anophelines in a variety of sub-Saharan environments, and, especially in more verdant areas, the likelihood of feeding on these stations rather than on natural sugar sources. Further, in environments where larval development sites are interspersed among, or located in the close vicinity of, human habitations, sugar baits may face competition with blood hosts instead of natural sugar sources. The relevance of the latter concern for a species such as

An. gambiae s.s. will be directly related to whether newly emerged females are obligatory sugar feeders (which would imply a high vulnerability of this species to the method), or whether even at this age sugar constitutes a facultative part of the diet, and blood may be taken instead.

167

An. gambiae emerges from the aquatic pupal stage with energy reserves near a critical minimum, so that both sexes die within a few days without some form of adult nutrition.

It lives in close association with humans, often resting inside houses, and most populations use humans primarily or almost exclusively as their source of blood, a trait that has unusual consequences. Human blood is low in one of the essential amino acids, isoleucine, thus limiting the amount of all amino acids derived from haemoglobin that can be converted into egg yolk protein. The excess amino acids are catabolised and make a large contribution to the energy reserve. Some authors suggest that sugar is rarely or never taken by An. gambiae females (Beier 1996; McCrae 1989; Muirhead-Thompson

1951), which makes sense in light of the energy derived from human blood and the mosquitoes‘ easy access to it.

State-dependent behavioural models developed to elucidate the decision to feed on sugar or on blood for female mosquitoes (Ma and Roitberg 2008; Roitberg and Friend 1992;

Roitberg et al. 1994) provide a framework by which resource availability (access to blood and sugar) and the risks associated with each behaviour can be factored in. These make the assumption that plant sugar is a normal part of any mosquito‘s natural history

(including An. gambiae, supported by recent field research; Manda, Foster et al. unpubl.) and that, by extension, a sugar-or-blood decision is determined by internal state, a built-in assessment of costs associated with food availability, and the stimulus situation. One can infer from the model of Ma & Roitberg that An. gambiae females would choose to feed

168

first on sugar when they are away from the blood feeding habitat, but around domiciles the likelihood of sugar feeding declines sharply with increasing blood host availability.

To date, experimental studies on the behaviour of 1-d-old An. gambiae females support the notion that this feeding decision is opportunistic, but favours sugar. In an olfactometer, 1-d-old females showed a higher response towards honey-baited ports than to ports baited with human-related volatiles (a soiled sock), and showed a preference for honey when both were presented simultaneously, suggesting that sugar is a viable food choice for young females, and may be preferred when nectar-related stimuli are strong

(Foster and Takken 2004). In a mesocosm with sugar sources, both sugar-bearing plants and sucrose- or honey-solutions, where females had nightly access to a human, the majority of 1-d-olds had taken a sugar meal, but a proportion had taken a blood meal instead (Stone et al. 2011). And when sugar was absent, a greater proportion of females had taken a blood meal on their first night. This suggests that their initial meal choice is flexible and depends on availability of both types of resources, but very little is known of the factors, and their interactions, that determine the outcome of the decision-making process.

The approach to the sugar-blood decision that we describe here, taken with a sense of immediacy for understanding a mosquito species of immense importance to human health, is a compromise between a field experiment and a laboratory one. This empirical study examined the influence of environmental conditions, i.e., the use of a bed net by a

169

blood host and the abundance of nectar-bearing plants, as well as female size—an indicator of her energetic reserves—by recreating several semi-natural conditions and observing a mosquito‘s decisions under those conditions. Our main objectives were to see whether the initial meal of female An. gambiae favours sugar even when a blood host is available throughout the night, and, if not, whether the use of an untreated net is likely to divert mosquitoes to sugar sources.

Methods

Mosquitoes (An. gambiae s.s., Mbita strain) were reared according to standard methods, as previously described (Stone et al. 2009b), with the following modifications: larvae were either kept at a high or low density to increase the size range of experimental mosquitoes. High density pans received 300 1st instar larvae, and a food regime of 0.13 mg of finely ground Tetramin fish flakes per larva during the first 3 d of larval development, 0.26 mg for the next 3d, and 0.53 mg for subsequent days until pupation, while low density pans received 50 1st instar larvae and 0.4 mg of food per larva during the first 3 d of larval development, 0.8 mg for the next 3d, and 1.6 mg for subsequent days until pupation, on average 9 days after eggs were laid.

Pupae, approximately 200 of each density, were placed in small cages. Pupae were not sexed. The day after emergence these were placed in a mesocosm, and released in the

170170

afternoon, approximately 4 hrs before sunset. That night the mosquitoes would have access to differing levels of plant and blood hosts (depending on treatment, see below) upon which they could feed. Mosquitoes were recovered by backpack and mouth aspirators the following morning between 8 and 9 o‘clock, killed immediately in a -40° C freezer to stop metabolic processes, and kept there until processed.

The mesocosm, a customized insect cage manufactured by Megaview LLC, had a bank of resting sites for mosquitoes (Stone et al. 2009a), plants, and the inner vinyl-and-mesh part of a 2-person camping tent, which could be closed off (excluding mosquitoes) or opened

(allowing mosquitoes to feed on the sleeping human inside) (Fig 1a,b). The sides of the mesocosm, ceiling, and sleeves were made with white polyester netting (108 x 32 mesh/per sq. in), the floor material was white vinyl. To protect the vinyl floor, a 15 x 20‘ white tarp (Tarpaflex US, Naples, FL) was placed inside of the cage. The dimensions of the cage were 5.66 x 4.87 x 3.00 m (L x W x H) for a total of 82.69 m3. The structure was supported by a PVC pipe framework and connectors. Pairs of 18‖ ties that were located along the seams of the structure as well as across the roof, allowed the cage to be tied to the PVC framework and the ceiling support beams. Nine cylindrical sleeves (0.45 m diameter and 0.5 m length) where located in the ceiling to allow growing lights to be positioned inside of the cage. These growing lights were suspended by chains from the metal structure of the greenhouse. The cylindrical sleeves were then wrapped around the chains and held in place with plastic zip ties. Also located in the ceiling was a 1 m long

D-shaped zippered door made from the same material as the ceiling. This was to allow

171

access to mechanical parts above the cage. Located on the floor were two openings. Each was 0.18 sq. meters with netting to allow water to drain through and had a vinyl flap covering. The position of the drains corresponded to drains in the greenhouse floor.

Access to the cage was through a large D-shaped zippered door (1.8 m high and 1.2 m wide) made from clear vinyl, positioned in the middle of one of the short walls. Outside of the door was a small antechamber to prevent mosquito escape. The antechamber measured 1 x 2 x 2 m and had two zippered doors, a vinyl floor, and one wall and the ceiling of netting. Located to the left of the door were two sleeves. The top sleeve was used for the exhaust pipe of the humidifier and the bottom sleeve was for humidistat controls.

Temperature was controlled through the greenhouse heating and cooling system. Wall- mounted steam radiators heated the room while louvered roof vents, an exhaust fan, and an evaporative cooling system maintained cooler temperatures all of which were controlled through a thermostat positioned in the middle of the room approximately 1 m from the floor. Humidity was maintained with an Ocean Mist® MH3 industrial ultrasonic humidifier (Mico Inc., El Monte, CA) which was controlled with the included humidistat suspended from the room‘s thermostat.

Treatments consisted of two levels of plant availability [either one, or six, potted Senna didymobotrya (Fabaceae)—a common plant in Kenya that this species readily feeds on

(Manda et al. 2007b)—were present in the mesocosms] and three levels of blood-host presence: 1) No blood host was present. 2) A human volunteer (C.S.) sat in the mesocosm

172

with lower legs exposed and allowed mosquitoes to feed for half an hour in the evening

(22:30 – 23:00) and in the morning (ranging from 06:30-07:00 to 07:30-08:00, so as to approximately coincide with sunrise at 39˚57‘ 40‖ northern latitude between mid-July and October), and slept in the closed tent during the inter-feeding period to simulate the use of an untreated net. A human landing catch was performed during the morning biting period, and females were aspirated as they bit, so that size of females biting in the morning and evening (i.e., any blood-fed female collected from the resting sites) could be compared. 3) A human volunteer (C.S.) slept in the open tent from 23:00–07:00, allowing mosquitoes to feed throughout the night. Four replicates were performed of each of the six plant-host abundance/blood-host presence permutations.

The following information was recorded for each subsample (n = 3167, 80-230 per night) of all mosquitoes collected from resting sites after sunrise and biting catches: sex, size, as indicated by wing length and measured from allular notch to the edge of the wing, excluding the fringe, using an ocular micrometer (mosquitoes above the median wing length of their sex are classified as large, below the median as small); whether or not a female had taken a blood meal, determined by visual inspection; sugar positivity and the amount of fructose ingested, using the cold-anthrone method (Haramis and Foster 1983).

Females were scored as unfed, sugar fed, blood fed, or blood and sugar fed.

Analysis of sugar & blood feeding:

173

There are 4 possible outcomes of the feeding decisions made by females: they can 1) not feed on anything, 2) feed on sugar, 3) feed on blood, or 4) feed on both sugar and blood.

These outcomes may reflect different processes. For instance, remaining unfed rather than feeding on either resource may have more to do with a mosquito failing to feed successfully than with choice, while taking a sugar or a blood meal implies a choice. To tease these components apart, we analyzed these outcomes at the following levels, using logistic regression in JMP v9 (SAS Institute Inc., Cary, NC) to see which factors determined the following:

1) Whether female mosquitoes fed on something (sugar and/or blood) or nothing. One

would expect that with both increased plant and blood host availability, feeding on

something becomes more likely, and that larger females, due to their higher energy

reserves, may be more successful at locating a source and feeding on it.

2) Whether a female mosquito fed on sugar or on blood. This is an analysis on a subset

of the whole data set; only females that fed on either sugar or on blood are

considered, and the treatment where no blood host is available was not examined, as

mosquitoes did not have a relevant choice there. One would expect that sugar feeding

increases with sugar availability, and blood feeding with blood host availability. Due

to their greater reserves, large females may prioritize blood feeding, while small

females may prioritize sugar feeding to replenish their more stringent reserves.

3) Whether a female feeds on one resource (sugar or blood) or on two resources (sugar 174

and blood) in the same night. Again the analysis is on a subset: the treatment without

a blood host is not relevant, and females that did not feed at all are excluded. Feeding

on both sugar and blood was expected to be most common in the scenario when blood

hosts were not accessible throughout the night, so that females might be diverted to a

sugar meal instead, with some of those then taking a blood meal at dawn.

On each experimental night the mean Relative Humidity and mean Temperature were recorded, and as these might influence activity of mosquitoes, or nectar production of plants, they were included as potential variables. Size (large or small) was included as a variable to see how energetic reserves might affect the blood/sugar choice. The possible independent variables therefore were plant availability (2 levels), blood host presence (3 levels, or 2 for analyses 2 & 3), size (large and small), mean RH, and mean temperature.

The effect of the independent variables on the dependent variable was initially assessed for each variable separately, and only variables with p < 0.25 entered into the logistic regression model (Hosmer and Lemeshow 2000). The decision to include interactions was made based on a graphical examination, using interaction profiles. Final models were those with the lowest AIC values, and the independent variables included in them was decided by a mixed (backward and forward) selection procedure.

Further, we asked the following questions:

175

4) Does the probability of male sugar feeding depend on the presence of the plants, male

size, temperature and humidity? As males rely only on sugar, this will provide more

accurate insight into whether there is an appreciable effect of plant abundance at the

tested levels on the likelihood of finding sugar.

5) Does the amount of fructose ingested for sugar positive males and females depend on

the variables mentioned? If so, this would indicate the possibility of competition for a

limited amount of nectar by mosquitoes.

6) Are there differences, in terms of size and the tendency to feed on > 1 resource,

between females biting between 22:30-23:00 and 6:30-7:00 when blood host access is

restricted by an untreated net? Both aspects of responsiveness to hosts (Takken et al.

1998) and persistence (Nasci 1991) may affect this outcome.

Results

The median wing length of female mosquitoes was 3.0 mm (mean 2.96 ± 0.24 mm), that of males 2.88 mm (mean 2.88 ± 0.19 mm).

I. What determines whether females feed on something (either blood or sugar) or nothing?

Both the size of females and the accessibility of human blood had strong effects on whether females succeeded to obtain a meal of either type, or remained unfed (Fig. 2).

176

Small females were more likely to have taken a meal than were large females (χ2 = 32.8, p < 0.0001). There was a significant effect of blood-host availability on the propensity to feed on something, both when the absence of a blood host is compared to host access only before and after he has slept under a net, and compared to when he was accessible the whole night (χ2 =501.9, p < 0.0001). There was also a significant difference in feeding between the restricted and unrestricted levels of blood host accessibility (χ2

=26.1, p < 0.0001). An interaction between mosquito size and blood-host presence (none vs. 2.5 hrs or 8 hrs) was included in the regression model, but was only marginally statistically significant (χ2 =3.7, p = 0.053). Whether one or six S. didymobotrya were present had no significant effect on whether females took a meal of blood or sugar (χ2 =

0.16, p = 0.68).

II. What determines whether a female chooses to feed on sugar or blood?

A comparison of Figures 3 and 4 leads to the conclusion that the feeding choices of female An. gambiae were determined to a great degree by the presence and accessibility of the blood host, and seemingly not by the abundance of potential nectar sources in the mesocosm. In a comparison of the sugar or blood choices by females when a blood host was present (restricted or unrestricted), the final regression model confirmed this observation, because it included blood-host presence as well as mosquito size (Table 1), but not plant abundance, temperature or humidity, or any interactions.

III. What determines whether a female feeds on just one or both resources?

177

Variables included in the model were blood-host presence, size and mean relative humidity (Table 2). Small females were more likely to take both a sugar and a blood meal in one night than were larger females, and feeding on both foods happened more often when access to blood was restricted (Fig 3).

IV. Does the probability of male mosquitoes obtaining sugar depend on male size, plant abundance, temperature and humidity?

Both sugar host presence (χ2 = 35.1, p < 0.0001) and size (χ2 = 64.9, p < 0.0001) had a significant effect on the proportion of males testing fructose-positive (Fig 5). With just one plant in the mesocosm, males were more likely to remain unfed (OR = 2.32, CI =

1.76 – 3.07), and small males were more likely to be sugar positive than large males (OR

= 3.1, CI = 2.36 – 4.09).

V. Amount of fructose

The final model on the size of the sugar meal taken by males included the number of plants present in the mesocosm (F = 12.8, p < 0.001) and male size (F = 3.01, p = 0.08), though the effect of the latter was not significant (Fig. 6). The amount of sugar taken by females was greater when more plants were available in the mesocosm (t = 3.81, p <

0.001). Female size and the presence of a blood host in the mesocosm did not significantly affect the size of the sugar meal (Fig. 7). The size of the sugar meal was also not different between females that had fed only on sugar and those that fed on both sugar and blood (t = -0.3, p = 0.74).

178

VI. Differences between females taking blood in the evening and at dawn

Females biting between 22:30-23:00 had a mean wing length of 2.87 mm and were significantly smaller than females biting at dawn, with a mean wing length of 2.96 mm (F

= 23.2, p < 0.0001). Over 8 nights, 44.2% of females that blood fed bit between 22:30-

23:00. Of the females that bit in the evening, 6% was positive for fructose as well, whereas of those biting at dawn 12.5% was fructose positive, indicating that females that bit earlier in the night were less likely to also obtain a sugar meal (χ2 = 35.1, p < 0.0001).

And of the females that bit at dawn, those that had obtained a sugar meal earlier in the night were significantly smaller (wing length of 2.72 mm) than those that were fructose negative (2.99 mm) (Z = -8.38, p < 0.0001).

Discussion

The main results of this study were that in a mesocosm with a sleeping human present, regardless of bed net usage, the majority of 1-d-old females obtained a blood meal. When a blood host was not present, or access was restricted through the use of an untreated net, sugar meals became more frequent, and smaller females were more likely to take a sugar meal under these circumstances, and were more likely to take both a sugar and a blood meal on the same night. The number of plants present in the room affected sugar positivity of males, but did not appear to affect female meal choice in this experiment.

179

For the use of sugar baits as a malaria vector control tool, the strong tendency to feed on blood, even at 1 d post-emergence, suggests that in areas where larval development sites are close to human habitations (Minakawa et al. 2002), the method may be useful mainly as a complement to bed nets.

The response of females to plants in this experiment was considerably lower than in two previous experiments (Foster and Takken 2004; Stone et al. 2011). This cannot be entirely explained by the differences in blood-host presence between experiments, because even without a human in this experiment only approximately one third of females fed on sugar, while approximately half did so in (smaller) mesocosms even when a blood host was present for 30 min per night (Stone et al. 2011). One explanation is that the plants used in this experiment provided a lower amount of sugar than the sources used previously. Only a third of the males were fructose positive when only one plant was present, and the proportion of fructose-positive males, as well as the amount of sugar obtained by both males and females, increased when six plants were present. For females, this level of sugar availability or plant stimuli may have been below a response threshold.

Possibly the difference in response to nectar was due to the quality of these nectar sources. Honey was present in both previous experiments, and its volatiles may be more attractive to mosquitoes than those of the plant. Both these explanations are not entirely satisfying, because this plant ranked among the most ―preferred‖ in experiments by

Manda et al. (Manda et al. 2007b), is attractive to An. gambiae in olfactometers

180

(Nikbakht Zadeh, pers. comm.) and was present as a sugar source in a previous mesocosm experiment, in which—though this was not quantified—it appeared to be favoured by male and female mosquitoes (C.S., pers. obs.). Possibly, nectar production and attractiveness to mosquitoes of this plant is condition-dependent, e.g., age, amount of new growth. Thus, an aspect of the blood-sugar choice that remains to be investigated is that of plant quality (presumably reflected in the volatile organic compounds released by the plant), which is pertinent to sugar baits competing in the field with blood hosts.

One may question whether certain decisions made by mosquitoes in this experimental set-up are relevant to the choices faced by mosquitoes in the field. For instance, if larval development sites are located a considerable distance from humans, the preferences of females may be irrelevant. Whether females already enter dwellings through eaves or other openings at this age is not certain, thus, it may be that while females are apparently willing to feed on humans as early as their 1st night after emergence, in nature they may not actually come into contact with humans that early. Our mesocosm also did not have an artificial hut, and whether females that are unable to bite a human due to a net would in reality leave a hut to locate a sugar source is also not known. Further, when a net was used in this experiment, mosquitoes still were allowed to blood feed for 30 min in the evening, and 30 min at dawn, reflecting periods where mosquitoes might bite humans before they sleep, or after waking up. But the absence of defensive behaviour may well have inflated the feeding success of females under this scenario.

181

The effect of body size on likelihood of taking a meal may have been due to, or exaggerated by differential survival of mosquitoes from high- and low- density larval environments that failed to obtain a meal of either kind on their 1st night as an adult, but the mean wing length of females recovered when a blood host was absent or present for 8 hr was 2.98 mm in both cases, and was slightly lower (2.92 mm) when blood was restricted, a pattern that does not match the above scenario. While it does make sense that females with low energetic reserves should take sugar more readily, being closer to starvation (as has been found previously for Cx. nigripalpus (Hancock and Foster 1997)), while females with larger reserves may prefer to not feed on sugar and thereby keep their options open, for males such an explanation makes little sense. An alternative explanation is that larger males responded to plant cues faster, obtained sugar earlier in the night, and had converted a greater amount of fructose to trehalose by morning than small males.

Assuming that defensive behaviour of humans before going to bed or after waking up does not preclude An. gambiae from biting entirely, it also suggests that obtaining a blood meal when untreated nets are used is likely even if the nets are used properly and are in pristine condition (i.e., no tears or holes), though the biting rate was reduced, corresponding to findings in the field (Clarke et al. 2001; Rehman et al. 2011).

We can conclude from this that a change in biting time of mosquitoes in response to the use of (untreated) nets in an area would not need to evolve, but may result from behavioural plasticity.

182

While changes in biting times have been observed, these have mostly been associated with a shift in species (Mbogo et al. 1996; Russell et al. 2011a). Given the enormous selective pressure that presumably comes with broad ITN coverage, it is surprising that a shift in biting times of An. gambiae s.s. has not been more commonly observed (Pates and Curtis 2005), and this may be related to whether in the field behavioural avoidance is caused by selection, or by plasticity. If a sufficient proportion of mosquitoes is repelled by nets, rather than killed outright, and females are likely to obtain sugar or blood elsewhere (i.e., if the costs of searching for an alternate host are low), one would not necessarily expect a change in host seeking times to evolve. For instance, if a female starts host seeking around 2am, and in the absence of bed nets obtains a blood meal shortly thereafter, but in the presence of nets has to search for a longer period, but does manage to obtain blood at dawn, the selective pressure for a shift in host seeking periodicity may be quite small. Therefore, if behavioural avoidance is mostly a result of plasticity, assessing such changes in biting time (i.e., the time of day at which females obtain blood, rather than the time at which host seeking commences) would not be accurately measured by human landing catches, which is the common method of assessing biting times in the field. Thus, the amount of behavioural avoidance in the field due to plasticity of An. gambiae s.s. may be currently underestimated.

With the scaling-up of long-lasting insecticidal nets (LLIN) distribution across sub-

Saharan Africa since 2008 (WHO 2010), one may question whether the behaviour of mosquitoes around untreated nets, as tested here, is of interest. We think that, with certain

183

caveats, this mesocosm set-up may be a useful proxy for studying mosquito behavioural plasticity relating to bed net use, even if in the field untreated nets will be largely restricted to those areas where LLIN coverage lags, or to households where one or more treated nets are used, but certain family members still use an untreated net (Vanden Eng et al. 2010), or to nets where the efficacy has worn off. Further, the behaviour around untreated nets may also give insight into the behaviour of mosquitoes resistant to pyrethroids around treated nets (N‘Guessan et al. 2007).

The decisions made in this set-up also provides clues to the behaviour of females after being repelled from a hut, which to date remains an area of speculation. In this set-up, mosquitoes were able and likely to obtain blood at sunrise, after being frustrated a large part of the night. In the field females may either continue seeking a blood host until an accessible host is found, or cease host seeking and rest (possibly taking a sugar meal), and presumably then biting early the following evening. Such a shift to earlier biting following the introduction of impregnated bed nets has been reported in a number of studies (Charlwood and Graves 1987; Mbogo et al. 1996) but has not been found in others (Mathenge et al. 2001). Information on what happens to females repelled from huts is especially relevant to control in areas that have already achieved a high degree of bed net coverage, but where this is not resulting in further reductions of parasitemia

(Eisele et al. 2011). This may be a result of a change in behaviour of a species, plasticity in behaviour, or a replacement with a different vector that is already predisposed to biting at different times or outdoors. To achieve further reductions in malaria incidence under

181844

these scenarios, it has been suggested that indoor control options will have to be supplemented with control methods targeting mosquitoes outdoors (Russell et al. 2010b), for example by the use of zooprophylaxis, treated resting stations, or larvicides. This study suggests that in particular small females are likely to seek a sugar meal when access to blood hosts is restricted, suggesting that a sugar-based method, as a supplement to ITNs, may be a useful control tool for such endgame scenarios. The combination of such baits with treated nets (for instance placed indoors, or near a hut) is also of interest as an additional tool against resistant mosquitoes. The feasibility of such a control option will likely require a sugar bait that is substantially more attractive than the plants used in this experiment, such as those used in Mali (Müller et al. 2010a), and in particular be attractive to large mosquitoes as well, which may become more prominent as population numbers decline (Russell et al. 2011b).

Conclusions

The initial meal choice of An. gambiae s.s. strongly favoured human blood over plant sugar in these mesocosm experiments. Sugar feeding by this species is then thought to be opportunistic; with reduced accessibility of humans, and energetic reserves of mosquitoes, feeding on sugar becomes more likely. For sugar-based control methods this implies that in certain environments these will have to compete with human hosts as well as with natural sugar sources, but a potentially synergistic interaction with bed nets was found. The diversion from humans sleeping under bed nets to sugar sources may play a

185

role in the behavioural adaptation to bed nets, allowing mosquitoes to survive until an alternate host can be fed on; our results highlight the importance of behavioural plasticity in this adaptation. Further studies on these questions would benefit from more natural circumstances, with a wider range of Anopheles species (e.g., An. arabiensis) and sugar hosts of differing qualities.

186

Table 5.1: Estimated coefficients, standard errors, χ2 values, p-values, odds ratios and 95% confidence intervals for the logistic regression model for feeding on sugar vs. blood

Variable Coeff. se χ2 p > χ2 OR 95 % CI

Blood host (2 ½ hrs) -1.219 0.25 28.57 <.0001* 0.08 0.02 0.21

Size (small) -0.931 0.17 22.01 <.0001* 0.15 0.07 0.29

Intercept 4.039 0.28 199.01 <.0001*

(Log likelihood = 42.48, p < 0.0001*)

Table 5.2: Estimated coefficients, standard errors, χ2 values, p-values, odds ratios and 95% confidence intervals for the logistic regression model for feeding on one or two resources (sugar and blood)

Variable Coeff. se χ2 p > χ2 OR 95 % CI

Blood host (2 ½ hrs) 0.544 0.13 17.23 <.0001* 2.97 1.82 5.11

Size (small) 0.799 0.13 36.37 <.0001* 4.94 3.01 8.57

Mean RH 0.037 0.01 8.73 0.0031* 4.87 1.01 1.06

Intercept -5.614 0.89 39.55 <.0001*

(Log likelihood = 41.68, p < 0.0001*)

187

D B

A C

F B E B

Fig 5.1a,b: Diagram & photograph of the mesocosm, [with A) bank of resting sites, B) S. didymobotrya plants, C) sleeping pad & mesh netting D) antechamber, E) lights, F) temperature & humidity sensor.]

188

Fig 5.2: Proportion of small and large females over all replicates per treatment that fed on something (i.e., blood or sugar) or nothing.

189

8 hrs

both blood 2 half hrs sugar none

none

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Proportion of females (small)

8 hrs

both blood 2 half hrs sugar none

none

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Proportion of females (large)

Fig 5.3a,b: The proportions of small (a) and large (b) females that fed on blood, sugar, both or neither when a blood host was accessible for 8 hours, for 2 half hours, or was not present.

190

1 both blood sugar none

Number Number of plants 6

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Proportion of females

Fig 5.4: The proportions of females that fed on blood, sugar, both or neither when 1 or 6

Senna didymobotrya were present.

191

Fig 5.5: Proportion of large and small males that were positive for fructose after 1 night of exposure to either 1 or 6 Senna didymobotrya.

Fig 5.6: Mean log amounts (µg) ± sem of fructose per large or small male with 1 or 6

Senna didymobotrya.

192

Fig 5.7: Differences in amounts of sugar ingested by females according to size, presence of a blood host, and the number of plants present in the mesocosm.

193

Chapter 6: Plant Community Composition Affects the Vectorial Capacity and

Fitness of the Malaria Mosquito Anopheles gambiae s.s.6

Abstract

Dynamics of Anopheles gambiae abundance and malaria transmission potential rely strongly on environmental conditions. Both female and male An. gambiae use sugar and are affected by its absence, but how the presence or absence of nectariferous plants affects An. gambiae abundance and vectorial capacity has not been studied. We report on

4 replicates of a cohort study performed in mesocosms with sugar-poor and sugar-rich plants, where we measured mosquito survival, biting rates and fecundity. Survivorship was greater with access to sugar-rich plant species, and mortality patterns were age- dependent; sugar-poor populations experienced Weibull mortality patterns, while 2 and 3 out of 4 female and 4 male populations, respectively, in sugar-rich environments were better fitted by Gompertz(-Makeham) functions. A trend of higher biting rates in sugar- poor mesocosms, particularly for young females, was found, and this resulted in a higher vectorial capacity in such environments. Fitness parameters were not significantly different between the environments, indicating that mosquito abundance is unlikely to be affected by plant species composition. The consequences of these differences in

6 C.M. Stone, B.T. Jackson and W.A. Foster, in preparation.

194

survivorship patterns and biting rates between sugar-rich and sugar-deprived environments on vector control measures are considered.

Introduction

Spatial and temporal heterogeneity in malaria transmission depends on climatic and environmental factors. Perhaps most well-studied are the effects of rainfall and availability of breeding sites, and temperature fluctuations. Additionally, certain models have included vegetation as a predictive factor, either as an indicator of rainfall and thus suitable conditions for Anopheles larval development (Hay et al. 1998), while others have considered the impact of, e.g., tree canopy cover on temperature, affecting adult survival or larval development (Afrane et al. 2008; Zhou et al. 2007). But vegetation also provides protective cover and nectar meals for vectors, whilst providing forage for potential blood hosts (in the case of non-anthropophilic species), and as such can be an effective predictor of vector dispersion, and potentially disease transmission (Reisen 2010). In this study we focus on the question whether the plant community species composition, and in particular the presence of nectariferous plant species, should be considered as a component of the landscape that influences mosquito abundance and malaria transmission. If so, this is a factor that not only changes geographically and seasonally, but potentially with land development and agricultural practices.

The effects of environmental factors, or of vector control methods, on malaria transmission are best qualitatively investigated through their effects on vectorial capacity

195

(Dye 1992; Garrett-Jones 1964), a measure which encompasses the entomological aspects of the basic reproductive number of malaria (R0) (Macdonald 1957). It describes the number of secondary infections caused by a population of mosquitoes per daily exposure to an infected host, and is, at its most basic, a function of mosquito density relative to humans, biting frequency, survival rate, and duration of the extrinsic cycle.

Survival and biting rate are particularly important components as they affect vectorial capacity exponentially.

One aspect where concern about the accuracy of vectorial capacity has been expressed relates to the assumption of a constant mortality factor, as analyses show that hazard functions of mosquito populations in nature and in laboratory settings are typically better described by age-dependent mortality functions (Clements and Paterson 1981; Dawes et al. 2009; Styer et al. 2007). These concerns would be particularly serious if using an exponential mortality function results in qualitatively different outcomes when comparing environments or applications of control measures, and one theoretical investigation so far suggests that this may be the case (Bellan 2010).

Presence of sugar in the environment of the malaria mosquito, Anopheles gambiae Giles s.s., affects most of the components of vectorial capacity (Stone & Foster, submitted). For instance, biting rates are reported to be higher when sugar is absent (Gary and Foster

2001; Straif and Beier 1996), while for survivorship the opposite is true (Gary and Foster

2001). Mosquito density, or cohort size is affected both by female fecundity (Manda et al.

2007a), and by the proportion of females that are inseminated. In the absence of sugar,

196

male reproductive performance may be affected to the extent that too few females can become inseminated to sustain a viable population (Gary et al. 2009; Stone et al. 2009).

Thus sugar affects vectorial capacity of An. gambiae in opposing directions, by simultaneously decreasing biting rates and increasing survival. However, while in laboratory cages, these factors balance out such that females attain a greater vectorial capacity in the absence of sugar (Gary and Foster 2001), in the field these factors may balance out very differently as a result of higher levels of mortality, both due to increased energetic expenditures related to host seeking and location of oviposition sites and mates, and a higher level of background mortality related to predation, defensive behaviour, and environmental factors. Further, in nature sugar will have to be obtained from a variety of sources. That these sources differ in both quality and quantity of nectar is clear from experiments where access to different plants resulted in a wide degree of variation in survival times of mosquitoes (Gary and Foster 2004; Impoinvil et al. 2004; Manda et al.

2007a). In one field study, populations of Anopheles sergentii Theobald in two oases were reported to differ in vectorial capacity by a factor of 250, which the authors attributed to a difference in availability of sugar-bearing plants between the oases (Gu et al. 2011).

What remains unknown is whether access to sugar-bearing plants of certain qualities results in different age-dependent patterns of mortality. In addition to patterns of survival, biting rates may be affected by sugar presence; for instance, when sugar is readily available young female An. gambiae may prefer to feed on sugar (Foster and Takken

2004), whereas when sugar is restricted and blood hosts are readily available even young

197

females may seek a blood meal (Stone et al. 2011). Thus, the age at which mosquitoes obtain their first blood meal may depend on environmental access to sugar.

Consequently, vectorial capacity may differ qualitatively between environments with sugar-rich plants and sugar-poor plants. Here we explore these questions, as well as measure fitness parameters to see if mosquito abundance may be influenced by plant communities, through cohort studies in mesocosms.

Materials and Methods

Mosquitoes (An. gambiae s.s., Mbita strain) were reared according to standard methods, as previously described (Stone et al. 2009). Experiments were performed between March and June 2011 in mesocosms set up for this purpose in The Ohio State University

Biological Sciences Greenhouse. An advantage over field mark-release-recapture studies is a lack of confounding emigration or immigration of mosquitoes, and better control over experimental factors such as blood host presence; a disadvantage is that certain other mortality factors (e.g., predation, extreme weather) are absent, and the energetic costs associated with foraging and mating are probably still underestimated, although less so than in small laboratory cages.

The mesocosms (Fig. 1) were customized insect cages manufactured by Megaview LLC.

A full description is given elsewhere (Stone and others, in prep.). Briefly, the sides of the mesocosm, ceiling, and sleeves were made with white polyester netting (108 x 32

198

mesh/per sq. in), the floor material was white vinyl. The floor was protected with a fiber- reinforced plastic tarp of similar size. The dimensions of the cage were 5.66 x 4.87 x 3.00 m (L x W x H) for a total of 82.69 m3. Nine cylindrical sleeves (0.45 m diameter and 0.5 m length) were located in the ceiling to allow 500 W growing lights to be positioned inside of the cage; the cylindrical sleeves were then wrapped around the chains and held in place with plastic zip ties. These lights were on between 09:00 and 17:00, in order to add to the natural ambient daylight in the mesocosms, but not interfere with crepuscular light conditions. Access to the cage was through a large D-shaped zippered door (1.8 m high and 1.2 m wide) made from clear vinyl. Outside of the door was a small antechamber to prevent mosquito escape. Temperature was controlled through the greenhouse heating and cooling system: wall-mounted steam radiators heated the room while louvered roof vents, an exhaust fan, and an evaporative cooling system maintained cooler temperatures all of which were controlled through a thermostat positioned in the middle of the room approximately 1 m from the floor. Humidity was maintained with an

Ocean Mist® MH3 industrial ultrasonic humidifier (Mico Inc., El Monte, CA) which was controlled with the included humidistat suspended from the room‘s thermostat.

To simulate environments with plant communities that consisted of either nectar-rich plants, or plants that were poor nectar sources for mosquitoes, we selected plant species endemic to western Kenya of which survival and sugar intake by An. gambiae exposed to them has previously been studied (Gary and Foster 2004; Impoinvil et al. 2004; Manda et al. 2007a; Manda et al. 2007b). Plants that showed a high level of sugar intake and survivorship in those studies were used in the sugar-rich environment. These were Senna

199

didymobotrya Fresen. (Fabaceae), Ricinus communis L. (Euphorbiaceae), and Tecoma stans L. (Bignoniaceae); Senna occidentalis (Fabaceae) was also included in this category due to prior observations of mosquitoes, in laboratory and field, feeding on its abundant and visible droplets of nectar (B.T. Jackson, W.A. Foster, B. Njiru, unpubl.). Plants that were used in the sugar-poor environment were Tithonia diversifolia Hemsl. (Asteraceae),

Parthenium hysterophorus L. (Asteraceae), Lantana camara L. (Verbenaceae) and

Datura stramonium L. (Solanaceae). While P. hysterophorus was reported to give a high level of fructose positivity in one study (Manda et al. 2007b), survival was comparable to that of negative (water-only) controls (Manda et al. 2007a), matching our preliminary personal observations. Seven plants were present and watered daily throughout the experiments in each mesocosm. In the 1st two replicates, the sugar-rich environment consisted of 3 S. didymobotrya, 2 R. communis, and 2 T. stans. In the 3rd and 4th replicates a S. occidentalis replaced one R. communis. The sugar-poor environment consisted of 2 plants each of T. diversifolia, P. hysterophorus and L. camara, and 1 D. stramonium during the 1st 2 replicates. During the 2nd replicate, copious amounts of nectar were observed on the new growth of T. diversifolia, indicating that, at least sometimes, it is not a nectar-poor plant. For that reason, for the 3rd and 4th replicates two T. diversifolia were replaced with an extra P. hysterophorus and D. stramonium.

Resting sites were provided in the form of 4 terracotta pots (36 cm diameter), the opening of which was closed off with a thin sheet of wood, with a circular hole (12.75 cm diameter) cut into the middle. To prevent mosquitoes from resting and dying on the soil of potted plants, where their bodies were more likely to go unnoticed, the plastic pots

200

were covered from rim to the stem of the plant by white nylon fabric. Two clear plastic containers per mesocosm were present throughout and held aged tap water to serve as oviposition sites. A few leaves of the plants present were strewn on the surface of the water, to break its surface tension.

At the start of each replicate ca. 1000 pupae were placed, without first being separated by sex, in the mesocosms. Survival of male and female mosquitoes was estimated by removal and counting of dead bodies each morning from resting sites and on the white vinyl floor, rather than by aspirating and counting the survivors every day, in order to minimize disturbance of the mosquitoes. A human blood-host (CS) was available for 30 minutes per mesocosm in the hour following sunrise, before the overhead lights came on, from day 1 onwards (the day after pupae were placed in the mesocosms was designated day 0). Biting rate was assessed by counting engorging females, and relating this to the estimate of surviving females present on that day. Each morning the oviposition sites were inspected for presence of eggs. If present, these were transferred to a round white filter paper. The eggs were then photographed with a Sony a-300 digital camera with 50 mm macro lens mounted on a copy stand, and their number estimated in ImageJ (Mains et al. 2008). Each replicate was run for 21 d, after which survivors were collected by backpack aspirator and counted. To account for differences between the number of surviving mosquitoes collected and the number of those released (minus the dead bodies collected and counted throughout the experiment), we assumed that a constant proportion of dead bodies went unnoticed (e.g., around the base of the plants), and multiplied the bodies counted by this factor. If subtracting the body counts over the duration of the

201

experiment from the number of males or females released resulted in a negative number, we assumed a slight deviation from a 1:1 sex ratio, and adjusted the numbers of males and females in the pupae released accordingly.

A subsample of surviving females of replicates 3 and 4 were dissected and their spermathecae inspected by compound microscope for presence of sperm, to determine insemination status after cohabitation with males for 21 d in sugar-poor and –rich mesocosms.

The questions we specifically wanted to answer with these experiments were whether productive plant hosts enhance or depress vectorial capacity of An. gambiae, and by what means. To assess this, we investigated the effects of two simulated plant-species communities, differing in their nectar availability, on two components of this measure: biting rate and survivorship. In the case of survivorship, we wished to determine at the same time whether mortality is best described by an exponential (i.e., constant) model, or, one of several, age-dependent functions, and whether mortalities in sugar-poor and sugar- rich environments are best described by the same or different functions. If their mortality and biting rates differ between such environments how do their combined effects determine vectorial capacity, and in which direction? The answer has implications for potential methods of vector control that reduce population biting rate and age structure.

A third component of vectorial capacity is the density of mosquitoes, which often is determined by their fecundity and rates of population increase. Male reproductive

202

performance and female insemination rates (presumably affecting female reproductive output) are enhanced by availability of sugar (Gary et al. 2009, Stone et al 2009). Thus, we measured the following female fitness parameters during the 21 d of the experiment: daily fecundity, net replacement rate (Ro), and intrinsic rate of increase (r) in nectar-rich and nectar-poor environments.

Survivorship of mosquitoes in sugar-poor and –rich environments were analyzed by constructing Kaplan-Meier survivorship curves per replicate for males and females, and testing for differences in survivorship using a Cox proportional-hazards analysis in R

(Crawley 2007; R Core Development Team 2010). Mortality functions describing the distribution of ages at death (Pletcher 1999; Pletcher et al. 2000) were fitted to the data and their parameters estimated using the ‗Survomatic‘ package for R (Bokov and Gelfond

2010). Differences in biting rates between environments were analysed with a repeated measures MANOVA in JMP 9 (SAS Institute Inc., Cary, NC).

Reproductive fitness of females in sugar-poor and –rich environments was assessed by creating life tables for each replicate, which allowed for the calculation of the net reproductive rate, R0 (Carey 1993). The rate of increase, r, was calculated by taking the natural log of the dominant eigenvalue (i.e., λ) of corresponding Leslie matrices (Skalski et al. 2007), which were obtained by using Mathematica 7 (Wolfram, Champaign, IL).

To calculate the vectorial capacity, allowing for age-dependent mortality and biting rates, we used a formula similar to that for ―total population vectorial capacity‖ (Rasgon et al.

203

2003; Styer et al. 2007). We calculated, in Mathematica, the expected number of potentially infective bites by cohorts in either environment, as follows:

T n x 1 n x 2 T

C m x i 'i i i i 1 x 1 i 1 i x i n x

Where m is the cohort size; T is the terminal time, i.e., the end point under consideration

(here: day 21); n is the extrinsic incubation period; εx is the probability of biting at time x;

ε’ is 1- ε; and µi is probability of survival on a given day. Thus, we calculated and summed over day x = 1 through x = T-n, the probability of not biting but surviving until day x, the probability of biting on day x and surviving through the extrinsic incubation period, and the expected number of infective bites from a female taking her 1st infected meal on day x. The main assumptions are that the first blood meal taken will invariably infect a female, and that all bites after the incubation period will be infective. This assumption is likely to be violated in nature, and our formula would therefore overestimate the number of infective bites. However, this should not affect a qualitative comparison between environments, unless vector competence of mosquitoes in sugar- poor and sugar-rich environments differs. Schwartz & Koella (2002) suggested that sugar-feeding by An. gambiae does not play a major role in the immune response to infection with Pl. falciparum, but for other mosquito-parasite systems this may differ

(reviewed in Stone & Foster, submitted).

204

Results

Survival of female An. gambiae was significantly greater in mesocosms with sugar-rich plants than in mesocosms with sugar-poor plants in three out of four replicates (Table 1), despite females having similar access to a blood host each day. The same pattern was found for male mosquitoes, though here the difference in survival was greater between treatments than for females (Figs. 2 & 3). A series of semi-hierarchical mortality functions (exponential, Weibull, Gompertz, Gompertz-Makeham, Logistic and Logistic-

Makeham) was fitted against observed mortality patterns (see Pletcher et al. (2000) for a description of these functions and parameters). In none of the replicates, for either sex, was an exponential mortality function the best model (i.e., the model with the lowest

Akaike Information Criterion [AIC] value). Best models and the estimated parameter values are given in Tables 1 and 2, and the corresponding survivorship functions

(Pletcher et al. 2000) plotted against the survivorship values in Figures 2 and 3. Females in poor environments displayed mortality patterns best described by Weibull distributions in three out of four replicates. In the fourth a logistic function gave the best fit, but both the Gompertz and Weibull models had AIC differences ( i) of only 1, suggesting all three models should be considered (Burnham and Anderson 2002). In the sugar-rich environments of replicates 1 and 2 mortality was best described by a Weibull function, while the 3rd and 4th replicate were better described by Gompertz-Makeham and

Gompertz functions, respectively. Male survival patterns were comparable to the extent that in sugar-poor environments Weibull distributions gave the best fits, while versions of

205

the Gompertz function gave the best fit to three out of four replicates in the sugar-rich mesocosms.

The mean biting rate over all days for sugar-poor mesocosms was 0.198 ± 0.08 bites per day per female mosquito; in sugar-rich mesocosms this was 0.142 ± 0.07 bites per female per day (Fig. 4). This difference in biting rate over 4 replicates between these environments was not quite significant (repeated measures MANOVA, F = 4.83, P =

0.07). Figure 4b indicates a more pronounced difference in biting rates between treatments when sugar-poor mesocosms did not include T. diversifolia, in particular during the 1st 2 d of the experiment. To test for differences between the slopes of regression lines fitted to sugar-poor and –rich mesocosms an analysis of covariance was done, taking a significant interaction factor between days and treatment as indication of differing slopes. In the 1st two replicates, where the sugar-poor treatment included T. diversifolia, there was no difference between the slopes (t = 0.39, P = 0.69), both which increased slightly, but non-significantly, with time. In the final two replicates the biting rate of females in the sugar-poor treatment decreased, but not significantly (t = 0.89, P =

0.38), whereas the biting rate in the sugar-rich room tended to increase with time (t =

2.01, P = 0.051); there was a trend for a difference in slopes (t = 1.86, P = 0.066).

Table 3 gives values of fitness parameters for females in both environments. There were no significant differences in mean daily fecundity between treatments. In replicate three in both environments mean daily fecundity was remarkably low, particularly in the sugar- poor environment. Both the mean net reproductive rate and the intrinsic rate of increase

206

were higher in the sugar-rich mesocosms, but not significantly so. In samples of surviving females of replicates 3 56 of 57 (98.2%) from the sugar-rich mesocosm had been inseminated, while only 2 out of 55 (3.6%) from the sugar-poor treatment had been inseminated. In replicate 4 these figures were 53 out of 54 (98.1%), and 51 out of 59

(86.4%), respectively. Thus, the poor reproductive performance of sugar-poor females in replicate 3 may be attributable to the poor mating capacity of males.

For vectorial capacity calculations, we used the mean temperature (25.28°C) during the experiments, which leads to an extrinsic incubation period of Plasmodium falciparum of

11.96 d (Detinova 1962), rounded up to 12 d. The outcomes, representing the potential number of infectious bites stemming from one cohort of 500 females, are provided in

Table 4. The calculations for each replicate used the daily survival probabilities (px) from the life tables of that cohort, and the daily biting rates (i.e., the actual number of bites counted on a given day, divided by the estimated number of female mosquitoes present) for the age-dependent vectorial capacity. We also calculated vectorial capacity using constant values (i.e., the mean px and biting rates per replicate). On average, the vectorial capacity of cohorts in sugar-poor environments was higher by 25% than those in sugar- rich environments, though the difference was not significant (t = 0.6, P =0.56). Using constant measures of survival and biting gave a roughly similar vectorial capacity for sugar-poor mesocosms, but it underestimated the vectorial capacity in the sugar-rich treatments. As a comparative measure the constant values therefore disagreed with the age-dependent values, because the cohorts in sugar-poor mesocosms were estimated to have vectorial capacities 46% greater than in sugar-rich mesocosms.

207

Figure 5 shows how the vectorial capacity is expected to decrease with a reduction of either the cohort size (e.g., through larval control), biting rate (e.g., through the deployment of bed nets), and adult survivorship (e.g., insecticidal spraying), for both the sugar-rich and sugar-poor environments. Despite their different mortality functions, and differences in biting rate, the sensitivity of vectorial capacity to these different measures is approximately the same.

Finally, we explored the effects on vectorial capacity theoretically, and based daily survivorship on the mortality functions that best fit replicates 3 and 4, allowing extrapolation beyond the 21 d the experiments ran. We also simplified biting patterns and based these on the average values, but allowing a peak of biting (0.5 bites / female) to occur in sugar-deprived environments 1 day after emergence. By doing so we could explore the effects of different mortality patterns and parameter values and of different biting rates separately, i.e. changing one while holding the other constant. This exercise confirmed that the effects of a higher biting rate overrode the effects of differences in survival probabilities between the two environments (Fig. 6, 7). Under the scenario in

Fig. 5 the vectorial capacity in a sugar-rich environment is approximately 40% lower than in a sugar-poor one. To achieve a similar reduction of vectorial capacity by conventional means, e.g., the use of bed nets, requires an overall reduction of biting rates of 26%.

Likewise, Fig. 6 shows that the lower vectorial capacity in sugar-rich areas remains proportional with increasing adult mortality caused by vector control when modelled by increasing an age-independent constant factor in the survivorship functions.

208

Discussion

Plant community composition clearly affects An. gambiae‘s survival and biting rate, two major components of vectorial capacity. Increased accessibility to plant sugar increases survivorship of both male and female mosquitoes, whereas it depresses the biting rate of females. These opposing effects of sugar on vectorial capacity have been documented in cage studies. What is surprising is that, over the 21 d measured, these effects resulted in a potential number of infectious bites that was on average 25% higher in environments with poor sugar-hosts, despite the reduced survivorship. The difference was not significant, which we attribute to few replicates with great variation.

Given equal climatic conditions, the differences in male survival between replicates are a reasonable indication of variation in nectar production or state of the plants, leading to, for instance, a higher survival rate in the 3rd replicate in the sugar-rich mesocosms compared to the other replicates. Although we attributed sugar-poor and sugar-rich labels a priori to the different plant species used in these experiments, the variation in survival between replicates that we observed suggests that this division is too simplistic, as both sugar-poor and sugar-rich plants may provide, at times (likely depending on plant age and condition), sufficient nutrients for An. gambiae males and females. For instance, T. diversifolia was observed to produce copious amounts of nectar during the 2nd replicate, leading to its removal from the sugar-poor environment in subsequent replicates. To come to a better understanding of the nutritive value of certain plant species for An. gambiae it will be necessary to assess their range of nectar production.

209

While it is evident that the plant community can affect the age-distribution of a mosquito population, less clear from these results is whether this will strongly affect the abundance of mosquitoes in an area. If it does, it appears that it would not be due to the female‘s reliance on sugar, but rather to a severe hampering of male reproductive performance when sugar is rare or absent, as suggested by prior experiments (Gary et al.

2009; Stone et al. 2009). It is evident from Table 3 that such a reduction in male mating ability occurred only during the 3rd replicate in the sugar-poor treatment. In that instance, only 3.6% of surviving females were inseminated, and their fitness measures were correspondingly depressed. It is surprising that minute differences in mean age of males at death (e.g., 3.88 d and 4.07 d in replicates 3 and 4, respectively) could result in such dramatic differences in female insemination rates. It raises the question whether male mating performance responds in a binary manner to a threshold of environmental sugar, instead of as a linear function of male mortality.

Given our lack of knowledge of mosquito plant foraging behaviour, it is difficult to extrapolate these results to a natural situation. But based on the similarities in reproductive rates in sugar-rich and sugar-poor environments in three of the replicates, we may assume that mosquito populations can be sustained even if only ―sugar-poor‖ hosts are present, as long as these still provide sufficient nectar to fuel male mating activity. Future studies should investigate how male mating performance depends on the

210

sugar concentration and accessibility of different host plants, e.g., if tissues are being pierced, will phloem sap sustain male mating activity?

In this regard, with fitness parameters being equivalent or slightly higher when nectar is readily available, An. gambiae differs from Aedes aegypti L., another anthropophilic mosquito also reported to achieve higher reproductive success when it does not feed on sugar sources (Costero et al. 1998; Harrington et al. 2001; Scott et al. 1997). Even for this species the results are not unequivocal, and spatial constraints may matter. For An. gambiae this certainly appears to be the case, as in laboratory cages fitness (R0 and r) of

An. gambiae were slightly higher for sugar-deprived females (Gary and Foster 2001), suggesting that the value of sugar to females increases in more realistic settings. Whether this can be extrapolated to even more energetically-demanding field environments remains to be investigated, though it would help explain why this behaviour is retained by this mosquito.

Comparing the final two sugar-poor replicates we notice that despite suffering comparable mortality, vectorial capacity was drastically different, resulting from differences in biting rates. While both had a peak on day 1—a phenomenon we have observed previously in sugar-deficient habitats,18 due to opportunistic blood-feeding after emergence if acceptable sugar sources are absent—the biting rate remained stable after that in the 4th replicate, but showed a decline with age in the 3rd. This can be ascribed to the marked difference in insemination rates between the two replicates. In the 3rd replicate 86.8% of surviving, unmated females collected at the end of the experiment retained Christopher‘s stage V eggs, i.e., were gravid and unable to make more eggs,

211

likely causing the decline in biting rate with age by limiting the gut‘s capacity for blood and the need for further protein.

The outcome of replicate 4, in which sugar had opposite effects on biting rate and survival, does agree with a laboratory-cage study using blood or blood plus a 10% sucrose solution (Gary and Foster 2001). The results are in stark contrast with a mark- release-recapture study performed with An. sergentii, where vectorial capacity was reported to be 250 times higher in a sugar-rich oasis than in a sugar-deficient oasis (Gu et al. 2011). Whether this is due to differences in the biology of the two species, or a reflection of the difference between the field and confined environments remains to be seen.

The utility of vectorial capacity when mortality rates are not constant, as they are in nature, has been questioned. As a qualitative term, this is not necessarily a hindrance, if, e.g., a control- and treatment-area are over- or underestimated by a similar margin.

The main contribution of this study is to show that different environments, if they differ in plant species composition and abundance, may differ in the pattern of age-dependent mortality that mosquitoes experience. A consequence of this is that one can no longer assume that constant mortality values will provide qualitatively sensible answers when comparing the vectorial capacity in two regions. Complicating matters further is that the age at which mosquitoes first bite does matter when using age-dependent mortality

(Clements and Paterson 1981; Styer et al. 2007). Here we find that this may also differ depending on the availability of sugar sources. Further studies on whether and when there is a decline or increase in biting rates as mosquitoes age would justify the use of the more

212

elaborate formula we used in this paper. Some studies do suggest that this is the case; in these the oldest age-cohort of sugar-deprived females (given the opportunity to mate prior to the withdrawal of sugar) had increased biting rates (Gary and Foster 2001; Straif and

Beier 1996). An increase in biting activity has also been observed in Plasmodium- infected mosquitoes (Koella et al. 1998), which typically coincides with increased age

(Smith et al. 2004).

Field studies will therefore have to take these variable mortality patterns and parameter values, and biting rates, into account. One way to do so would be to make use of mark-release-recapture methods with synthetic cohorts (Harrington et al. 2008), with human landing catch or equivalent as the method of recapture.

Environments differing in composition of nectar-producing plants will have vector populations with different biting habits, age distributions, and vectorial capacities.

A question relating to vector control is whether this might make certain control measures more or less effective in a given area. Our data suggest that the best way to control vectors, regardless, is to focus on reducing adult survival, because sensitivity to reductions in biting rate, adult survival or cohort size did not differ between the environments (Fig. 5).

The other question is whether malaria transmission rates in the field can be altered by manipulating the plant-species composition in an area. Despite the variance in nectar production and vectorial capacities in our experiments, we can extract pattern by

213

exploring the differences in survivorship and biting data between treatments. Figures 6 &

7 in particular suggest that forms of environmental enrichment, i.e., the planting or maintenance of nectariferous host plants for mosquitoes, may have a profound impact and work in an additive manner with conventional control measures. This would be akin to the suggestion to enrich agricultural lands with floral food sources for parasitoids, to enhance biological control (Wäckers et al. 2007). The other possible direction to take environmental manipulation of sugar sources is to replace them with poor sugar-hosts and reduce male mating performance below the threshold of population viability, as appeared to occur in the third replicate—though this will depend on a finer understanding of male energetic needs and foraging than we currently possess. Before manipulation of plant- species composition can be seriously considered, responses to environmental sugar of other vectors, particularly in the An. gambiae and Anopheles funestus Giles complexes, should be studied, because it is possible that more exophagic species such as Anopheles arabiensis Patton have a more stringent reliance on sugar than its sibling species. These interpretations, at this point, should thus be made with great care, and mostly indicate the potential impact of environmental sugar on mosquito behaviour. These studies should encourage further exploration of such effects in natural settings.

214

Fig. 6.1: An image of the mesocosm, with resting pots, oviposition sites, lights, temperature and humidity sensors, a chair for the blood host, and nectariferous plants.

215

Fig. 6.2: Kaplan-Meier survivorship curves for females in nectar-poor and –rich mesocosms, per replicate (left, upper: 1; left, lower: 2, right, upper: 3; right, lower: 4), and (fitted lines) associated estimated survivorship functions, see Table 1.

216

Fig. 6.3: Kaplan-Meier survivorship curves for males in nectar-poor and –rich mesocosms, per replicate (left, upper: 1; left, lower: 2, right, upper: 3; right, lower: 4), and (fitted lines) associated estimated survivorship functions; see Table 2.

217

Table 6.1: Mean age (days) at death for females in sugar rich and poor areas for replicates 1 to 4, whether survival is different between environments according to Cox proportional hazards, the mortality function that best describes the distribution of ages at death, and estimated parameter values of that function.

Replicate Treatment Mean age Cox prop. hazard Mortality function Parameters at death z P λ Β or γ c or s

1 Poor 10.8 -9.326 <0.0001 *** Weibull 0.095 1.91 Rich 14.5 Weibull 0.0676 2.79

2 Poor 13.6 11.7 0.00089 *** Weibull 0.074 1.858

218 Rich 11.7 Weibull 0.084 2.02

3 Poor 9.46 -18.14 <0.0001 *** Weibull 0.11 1.2 Rich 33.8 Gompertz-Makeham 5.5e-06 0.50 0.015

4 Poor 10.6 -4.255 <0.0001 *** logistic 0.123 0.15 1.7 Rich 14.0 Gompertz 0.026 0.11

218

Table 6.2: Mean age (days) at death for males in sugar rich and poor areas for replicates 1 to 4, whether survival is different between environments according to Cox proportional hazards, the mortality function that best describes the distribution of ages at death, and estimated parameter values of that function.

Replicate Treatment Mean age Cox prop. hazard Mortality function Parameters at death z P λ Β or γ c or s

1 poor 6.07 -23.83 <0.0001 *** Weibull 0.16 1.87 rich 20.8 Gompertz-Makeham 0.00065 0.27 0.02

2 poor 21.4 6.222 <0.0001 *** Weibull 0.045 1.566

219 rich 15.8 Weibull 0.063 1.96

3 Poor 3.88 -26.02 <0.0001 *** Weibull 0.26 1.7 Rich 36.7 Gompertz 8.3e-03 0.068

4 Poor 4.07 -17.55 <0.0001 *** Weibull 0.24 1.282 Rich 12.14 Gompertz-Makeham 1.1e-05 0.5 0.074

219

0.35 Sugar-rich 0.3

0.25 Sugar-poor

0.2

0.15

Human biting rate 0.1

0.05

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Day

0.45 0.4 Sugar-rich 0.35 0.3 Sugar-poor 0.25 0.2

0.15 Human biting rate 0.1 0.05 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Day

Fig 6.4a,b: a) upper; mean biting rate per female per day of the 1st two replicates, where the sugar-poor mesocosm included Tithonia diversifolia. Lower; mean biting rate per female per day for the final two replicates.

220

Table 6.3: Values for measures of reproductive success of An. gambiae in sugar-poor and sugar-rich mesocosms Treatment Replicate Mean Daily Net Reproductive Rate of increase Fecundity (Mx) Rate (R0) (r) Sugar-poor 1 25.50 118.42 0.57 2 18.86 151.15 0.64 3 3.84 21.93 0.41 4 26.6 150.89 0.55 Mean 18.7 110.6 0.54

Sugar-rich 1 18.12 167.87 0.56 2 22.08 135.44 0.71 3 8.38 131.84 0.5 4 18.4 139.32 0.56 Mean 16.75 143.62 0.58

t ratio 0.32 1.04 0.59 P 0.75 0.36 0.57

Table 6.4: Vectorial capacity of cohorts in sugar-poor and sugar-rich environments, calculated for age-dependent, and constant, biting rates and mortality. replicate age-dependent constant poor rich poor : rich poor rich poor : rich 1 72.6 129.1 0.56 55.3 86.8 0.64 2 138.8 88.1 1.57 135.9 66.9 2.02 3 68.5 77.4 0.88 97.5 86.7 1.12 4 215.9 109.9 1.96 200.6 97.6 2.06

mean 123.9 101.1 1.25 122.3 84.5 1.46

221

Fig. 6.5: Effects of proportionally reducing initial cohort size, biting rates and adult survivorship through conventional means on the vectorial capacity in sugar-poor (light lines) and sugar-rich (dark lines) environments. The difference between treatments is indicated by the shaded areas between lines.

222

Fig 6.6: Vectorial capacity in sugar-rich environments as a percentage of that in sugar- poor environments, when mortality functions are as in replicate 4 (see Table 1), and daily biting rates are 0.15 / ♀ / d in sugar-rich, and 0.5/ ♀ on d 1, and 0.2 / ♀ / d for other days in sugar-poor areas. The dotted lines indicate the extent to which overall biting rates would have to be reduced in a sugar-poor environment to obtain a vectorial capacity found in a sugar-rich area.

223

Fig 6.7: Vectorial capacity of cohorts after 41 days in sugar-rich environments as a percentage of that in sugar-poor environments, when an additional age-independent mortality factor is incrementally increased from 0 to 0.1 in both environments, representing reduction of adult survival due to control measures. Mortality functions are as in replicate 3 (see Table 1), and daily biting rates are 0.15 / ♀ / d in sugar-rich, and 0.5/ ♀ on d 1, and 0.2 / ♀ / d for other days in sugar-poor areas.

224

Conclusion

Sugar-feeding by one of the most important transmitters of Plasmodium falciparum, the causal agent of the deadliest form of malaria in sub-Saharan Africa, An. gambiae s.s., appeared paradoxal (though this was mostly inferred from studies on its ecological analogue, Ae. aegypti): with access to sugar females readily incorporate it in their diet, and by so doing appear to decrease their fitness.

The importance of sugar feeding for this species was the focus of this dissertation. A critical component was to study the behaviour in the more natural, energetically demanding (for the mosquitoes) environment of mesocosms (Chapter 2). The second was to not just focus on female (i.e., biting) mosquitoes, but also study male mosquito behaviour and sugar reliance (Chapter 3). The strong correlation between male fitness and the ability to locate plant-sugar in the environment (the exact nature of this correlation remains to be investigated) raises the question whether the ability and tendency of females to do so may in part be explained by a genetic constraint, e.g. perhaps females possess the tendency and ability to sugar feed due to indirect selection on males. For example, females that possess odorant receptors that are narrowly tuned to nectar sources occurring in their environment may have more successful sons.

225

The chapters focusing on female mosquitoes reveal a likelier explanation: compared to small laboratory cages, mesocosms reflect the challenges faced by mosquitoes in natural environment better, and in this arena, females enjoyed an approximately similar, or slightly higher fitness with a greater amount of sugar in the environment (Chapter 6). A true measure of fitness of wild type An. gambiae in actual natural settings will be required to fully settle the issue, although obtaining such measures will be challenging.

The results from Chapters 4 & 5 suggest that the advantage of sugar-feeding to female mosquitoes comes with being opportunistic; obtaining the highest level of fitness on average over varying situations that differ in their availability and accessibility of nectar and blood sources. A greater insight into how these decisions are made can be gained by incorporating quality of the respective resources into the equation. Doing so will lead to an improved understanding of the malaria transmission potential of this species in different environments, as well as increase the potential of sugar-based control methods.

One such example coming out of this dissertation is the peak of biting activity 1 day after emergence as a result of opportunistic feeding choices, which strongly affects the vectorial capacity of mosquitoes. This raises the tantalizing question whether malaria transmission may be affected by manipulation of the plant community (e.g., enrichment through the planting of attractive nectar-hosts), but before this can be seriously considered similar experiments on wild-type An. gambiae in natural settings will have to be performed, as well as on other co-occurring mosquito species, which may depend on sugar quite differently.

226

Bibliography

Abdel-Malek, A. A. 1964. Study of the feeding habits of Anopheles sergentii Theo. at Siwa Oasis using radiophosphorus. Bulletin of The World Health Organization, 30, 137- 139. Abdel-Malek, A. A. & Baldwin, W. F. 1961. Specificity of plant feeding in mosquitoes as determined by radioactive phosphorus. Nature, 192, 178-179. Afrane, Y. A., Little, T. J., Lawson, B. W., Githeko, A. K. & Yan, G. 2008. Deforestation and vectorial capacity of Anopheles gambiae Giles mosquitoes in malaria transmission, Kenya. Emerging infectious diseases, 14, 1533-1538. Alexander, B. & Usma, M. C. 1994. Potential sources of sugar for the phlebotomine sandfly Lutzomyia youngi (Diptera: Psychodidae) in a Colombian coffee plantation. Annals of Tropical Medicine and Parasitology, 88, 543. Ali, A., Rui-de Xue, R. & Barnard, D. R. 2006. Effects of sublethal exposure to boric acid sugar bait on adult survival, host-seeking, bloodfeeding behavior, and reproduction of Stegomyia albopicta. Journal of the American Mosquito Control Association, 22, 464- 468. Ameneshewa, B. & Service, M. W. 1996. The relationship between female body size and survival rate of the malaria vector Anopheles arabiensis in Ethiopia. Medical and Veterinary Entomology, 10, 170-172. Andersson, I. H. 1992. The effect of sugar meals and body size on fecundity and longevity of female Aedes communis (Diptera: Culicidae). Physiological Entomology, 17, 203-207. Ansari, M. A. & Razdan, R. K. 2004. Follow-up studies after withdrawal of deltamethrin spraying against Anopheles culicifacies and malaria incidence. Journal of the American Mosquito Control Association, 20, 424-428. Basseri, H. R., Doosti, S., Akbarzadeh, K., Nateghpour, M., Whitten, M. & Ladoni, H. 2008. Competency of Anopheles stephensi mysorensis strain for Plasmodium vivax and the role of inhibitory carbohydrates to block its sporogonic cycle. Malaria Journal, 7, 131. Beerntsen, B. T., James, A. A. & Christensen, B. M. 2000. Genetics of mosquito vector competence. Microbiology and Molecular Biology Reviews, 64,

227

115-137. Beier, J. C. 1996. Frequent Blood-Feeding and Restrictive Sugar-Feeding Behavior Enhance the Malaria Vector Potential of Anopheles gambiae s.l. and An. funestus (Diptera: Culicidae) in Western Kenya. Journal of Medical Entomology, 33, 613-618. Beier, J. C., Davis, J. R., Vaughan, J. A., Noden, B. H. & Beier, M. S. 1991. Quantitation of Plasmodium falciparum sporozoites transmitted in vitro by experimentally infected Anopheles gambiae and Anopheles stephensi. The American Journal of Tropical Medicine and Hygiene, 44, 564. Beier, J. C., Keating, J., Githure, J. I., Macdonald, M. B., Impoinvil, D. E. & Novak, R. J. 2008. Integrated vector management for malaria control. Malaria Journal, 7, S4. Bellan, S. E. 2010. The importance of age dependent mortality and the extrinsic incubation period in models of mosquito-borne disease transmission and control. PLoS One, 5, e10165. Billingsley, P. F., Hodivala, K. J., Winger, L. A. & Sinden, R. E. 1991. Detection of mature malaria infections in live mosquitoes. Transactions of the Royal Society of Tropical Medicine and Hygiene, 85, 450-453. Bokov, A. F. & Gelfond, J. 2010. Survomatic: Analysis of longevity data. R Package version 1.4.0.0. Bowen, M. F. 1992. Patterns of sugar feeding in diapausing and non-diapausing Culex pipiens (Diptera: Culicidae) females. Journal of Medical Entomology, 29, 843-849. Bowen, M. F. 1992. Terpene-sensitive receptors in female Culex pipiens mosquitoes: electrophysiology and behaviour. Journal of Insect Physiology, 38, 759-764. Bowen, M. F., Haggart, D. & Romo, J. 1995. Long-distance orientation, nutritional preference, and electrophysiological responsiveness in the mosquito Aedes bahamensis. Journal of Vector Ecology, 20, 203-210. Bowen, M. F. & Romo, J. 1995. Host-seeking and sugar-feeding in the autogenous mosquito Aedes bahamensis (Diptera: Culicidae). Journal of Vector Ecology, 20, 195- 202. Bowen, M. F. & Romo, J. 1995. Sugar deprivation in adult Aedes bahamensis females and its effects on host seeking and longevity. Journal of Vector Ecology, 20, 211-215. Braks, M. A. H., Juliano, S. A. & Lounibos, L. P. 2006. Superior reproductive success on human blood without sugar is not limited to highly anthropophilic mosquito species. Medical and Veterinary Entomology, 20, 53-59. Briegel, H. 1985. Mosquito reproduction: incomplete utilization of the blood meal protein for oögenesis. Journal of Insect Physiology, 31, 15-21. Briegel, H. 1990. Fecundity, metabolism, and body size in Anopheles (Diptera: Culicidae), vectors of malaria. Journal of Medical Entomology, 27, 839-850. Briegel, H., Hefti, M. & DiMarco, E. 2002. Lipid metabolism during sequential gonotrophic cycles in large and small female Aedes aegypti. Journal of Insect Physiology, 48, 547-554. Briegel, H. & Hörler, E. 1993. Multiple bloodmeals as reproductive strategy in Anopheles (Diptera: Culicidae). Journal of Medical Entomology, 30, 975-985. Briegel, H., Waltert, A. & Kuhn, R. 2001. Reproductive physiology of Aedes (Aedimorphus) vexans (Diptera: Culicidae) in relation to flight potential. Journal of Medical Entomology, 38, 557-565.

228

Bull, C. D., Metcalfe, N. B. & Mangel, M. 1996. Seasonal matching of foraging to anticipated energy requirements in anorexic juvenile salmon. Proceedings of the Royal Society of London, Series B: Biological Sciences, 263, 13-18. Burgin, S. G. & Hunter, F. F. 1997. Evidence of honeydew feeding in black flies (Diptera: Simuliidae). Canadian Entomologist, 129, 859-869. Burgin, S. G. & Hunter, F. F. 1997. Nectar versus honeydew as sources of sugar for male and female black flies (Diptera: Simuliidae). Journal of Medical Entomology, 34, 605-608. Burgin, S. G. & Hunter, F. F. 1997. Sugar-meal sources used by female black flies (Diptera: Simuliidae): a four-habitat study. Canadian Journal of Zoology, 75, 1066-1072. Burkett, D. A., Butler, J. E. & Kline, D. L. 1998. Field evaluation of colored light- emitting diodes as attractants for woodland mosquitoes and other Diptera in north central Florida. Journal of the American Mosquito Control Association, 14, 186-195. Burkett, D. A., Kline, D. L. & Carlson, D. A. 1999. Sugar meal composition of five North Central Florida mosquito species (Diptera: Culicidae) as determined by gas chromatography. Journal of Medical Entomology, 36, 462-467. Burnham, K. P. & Anderson, D. R. 2002. Model selection and multimodel inference; a practical information-theoretical approach, second edition: Springer. Canyon, D. V., Hii, J. L. K. & Muller, R. 1999. Effect of diet on biting, oviposition, and survival of Aedes aegypti (Diptera: Culicidae). Journal of Medical Entomology, 36, 301-308. Carey, A. F., Wang, G., Su, C. Y., Zwiebel, L. J. & Carlson, J. R. 2010. Odorant reception in the malaria mosquito Anopheles gambiae. Nature, 464, 66-71. Carey, J. R. 1993. Applied demography for biologists with special emphasis on insects: Oxford University Press, USA. Carey, J. R. 2001. Insect biodemography. Annual Review of Entomology, 46, 79-110. Caswell, H. 1989. Matrix population models: construction, analysis, and interpretation.: Sinauer, Sunderland, MA. Caswell, H. 2001. Matrix population models, second edition. Sunderland, Massachusetts: Sinauer Associates. Cator, L. J., Arthur, B. J., Harrington, L. C. & Hoy, R. R. 2009. Harmonic convergence in the love songs of the dengue vector mosquito. Science, 323, 1077-1079. Charlesworth, B. 1994. Evolution in age-structured populations: Cambridge Univ Pr. Charlwood, J. D. & Graves, P. M. 1987. The effect of permethrin impregnated bednets on a population of Anopheles farauti in coastal Papua New Guinea. Medical and Veterinary Entomology, 1, 319-327. Charlwood, J. D., Pinto, J., Sousa, C. A., Ferreira, C., Petrarca, V. & do Rosario, V. E. 2003. 'A mate or a meal' – Pre-gravid behaviour of female Anopheles gambiae from the islands of São Tomé and Príncipe, West Africa. Malaria Journal, 2, 9. Chaves, L.F., Harrington, L.C., Keogh, C.L., Nguyen, A.M. & Kitron, U.D. 2010. Blood feeding patterns of mosquitoes: random or structured? Frontiers in Zoology, 7, 3. Chouaia, B., Rossi, P., Montagna, M., Ricci, I., Crotti, E., Damiani, C., Epis, S., Faye, I., Sagnon, N. F. & Alma, A. 2010. Molecular evidence for multiple infections as revealed by typing of Asaia bacterial symbionts of four mosquito species. Applied and Environmental Microbiology, 76, 7444.

229

Christensen, B. M. 1981. Effect of Dirofilaria immititus on the fecundity of Aedes trivittatus. Mosquito News, 41, 78-81. Chun, J., Riehle, M. & Paskewitz, S. M. 1995. Effect of mosquito age and reproductive status on melanization of Sephadex beads in Plasmodium-refractory and -susceptible strains of Anopheles gambiae. Journal of Invertebrate Pathology, 66, 11-17. Clark, C. W. & Mangel, M. 2000. Dynamic state variable models in ecology. New York: Oxford university press. Clarke, S. E., Bogh, C., Brown, R. C., Pinder, M., Walraven, G. E. L. & Lindsay, S. W. 2001. Do untreated bednets protect against malaria? Transactions of the Royal Society of Tropical Medicine and Hygiene, 95, 457-462. Clarke, S. E., Bøgh, C., Brown, R. C., Walraven, G. E. L., Thomas, C. J. & Lindsay, S. W. 2002. Risk of malaria attacks in Gambian children is greater away from malaria vector breeding sites. Transactions of the Royal Society of Tropical Medicine and Hygiene, 96, 499-506. Clements, A. N. 1992. The biology of mosquitoes. Volume 1: development, nutrition and reproduction: Chapman & Hall. Clements, A. N. 1999. The biology of mosquitoes. Volume 2: sensory reception and behaviour. Wallingford: CABI Publishing. Clements, A. N. & Paterson, G. D. 1981. The analysis of mortality and survival rates in wild populations of mosquitoes. Journal of Applied Ecology, 18, 373-399. Collier, T. R., Murdoch, W. W. & Nisbet, R. M. 1994. Egg load and the decision to host-feed in the parasitoid, Aphytis melinus. Journal of Animal Ecology, 63, 299-306. Collins, F. H., Sakai, R. K., Vernick, K. D., Paskewitz, S., Seeley, D. C., Miller, L. H., Collins, W. E., Campbell, C. C. & Gwadz, R. W. 1986. Genetic selection of a Plasmodium-refractory strain of the malaria vector Anopheles gambiae. Science, 234, 607. Costantini, C., Li, S., Della Torre, A., Sagnon, N., Coluzzi, M. & Taylor, C. E. 1996. Density, survival and dispersal of Anopheles gambiae complex mosquitoes in a West African Sudan savanna village. Medical and Veterinary Entomology, 10, 203-219. Costero, A., Edman, J. D., Clark, C. G. & Scott, T. W. 1998. Life table study of Aedes aegypti (Diptera: Culicidae) in Puerto Rico fed only human blood versus blood plus sugar. Journal of Medical Entomology, 35, 809-813. Courchamp, F., Clutton-Brock, T. & Grenfell, B. 1999. Inverse density dependence and the Allee effect. Trends in Ecology and Evolution, 14, 405-410. Crawley, M. J. 2007. The R book: John Wiley & Sons Inc. Crotti, E., Rizzi, A., Chouaia, B., Ricci, I., Favia, G., Alma, A., Sacchi, L., Bourtzis, K., Mandrioli, M. & Cherif, A. 2010. Acetic Acid Bacteria, Newly Emerging Symbionts of Insects. Applied and Environmental Microbiology, 76, 6963. Curtis, C. F., Miller, J. E., Hodjati, M. H., Kolaczinski, J. H. & Kasumba, I. 1998. Can anything be done to maintain the effectiveness of pyrethroid–impregnated bednets against malaria vectors? Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 353, 1769. Curtis, C. F., Myamba, J. & Wilkes, T. J. 1996. Comparison of different insecticides and fabrics for anti mosquito bednets and curtains. Medical and Veterinary Entomology, 10, 1-11.

230

Damiani, C., Ricci, I., Crotti, E., Rossi, P., Rizzi, A., Scuppa, P., Esposito, F., Bandi, C., Daffonchio, D. & Favia, G. 2008. Paternal transmission of symbiotic bacteria in malaria vectors. Current Biology, 18, 1087-1088. Dao, A., Adamou, A., Yaro, A. S., Maiga, H. M., Kassogue, Y., Traore, S. F. & Lehmann, T. 2008. Assessment of alternative mating strategies in Anopheles gambiae: does mating occur indoors? Journal of Medical Entomology, 45, 643-652. Dawes, E. J., Churcher, T. S., Zhuang, S., Sinden, R. E. & Basáñez, M. G. 2009. Anopheles mortality is both age- and Plasmodium-density dependent: implications for malaria transmission. Malaria Journal, 8, 228. Dawidowicz K., Hernandez A.G., Infante R.B. and Convit J. (1975) The surface membrane of leishmania. I. The effects of lectins on different stages of leishmania braziliensis. The Journal of Parasitology, 61, 950-953. Day, J. F., Edman, J. D. & Scott, T. W. 1994. Reproductive fitness and survivorship of Aedes aegypti (Diptera: Culicidae) maintained on blood, with field observations from Thailand. Journal of Medical Entomology, 31, 611-617. De Meillon, B., Sebastian, A. & Khan, Z. H. 1967. Cane-sugar feeding in Culex pipiens fatigans. Bulletin of The World Health Organization, 36, 53-65. Detinova, T. S. 1962. Age grouping methods in Diptera of medical importance with special reference to some vectors of malaria. In: W.H.O. Mongr. Ser. 47. Dimond, J. B., Lea, A. O., Hahnert, W. F. & DeLong, D. M. 1956. The amino acids required for egg production in Aedes aegypti. Can. Entomol, 88, 57–62. Downes, J. A. 1958. The feeding habits of biting flies and their significance in classification. Annual Review of Entomology, 3, 249-266. Downes, J. A. 1969. The swarming and mating flight of Diptera. Annual Review of Entomology, 14, 271-298. Dwyer, D. M. 1974. Lectin binding saccharides on a parasitic protozoan. Science, 184, 471. Dye, C. 1992. The analysis of parasite transmission by bloodsucking insects. Annual Review of Entomology, 37, 1-19. Edman, J. D., Strickman, D., Kittayapong, P. & Scott, T. W. 1992. Female Aedes aegypti (Diptera: Culicidae) in Thailand rarely feed on sugar. Journal of Medical Entomology, 29, 1035-1038. Eischen, F. A. & Foster, W. A. 1983. Life span and fecundity of adult female Aedes aegypti (Diptera: Culicidae) fed aqueous extracts of pollen. Annals of the Entomological Society of America, 76, 661-663. Eisele, T. P., Miller, J. M., Moonga, H. B., Hamainza, B., Hutchinson, P. & Keating, J. 2011. Malaria infection and anemia prevalence in Zambia's Luangwa district: an area of near-universal insecticide-treated mosquito net coverage. The American Journal of Tropical Medicine and Hygiene, 84, 152-157. Emlen, S. T. & Oring, L. W. 1977. Ecology, sexual selection, and the evolution of mating systems. Science, 197, 215-223. Favia, G., Ricci, I., Damiani, C., Raddadi, N., Crotti, E., Marzorati, M., Rizzi, A., Urso, R., Brusetti, L. & Borin, S. 2007. Bacteria of the genus Asaia stably associate with Anopheles stephensi, an Asian malarial mosquito vector. Proceedings of the National Academy of Sciences, 104, 9047.

231

Favia, G., Ricci, I., Marzorati, M., Negri, I., Alma, A., Sacchi, L., Bandi, C. & Daffonchio, D. 2008. Bacteria of the genus Asaia: a potential paratransgenic weapon against malaria. In: Transgenesis and the management of vector-borne disease. (Ed. by Aksoy, S.): Landes Bioscience and Springer Science+Business Media. Feinsod, F. M. & Spielman, A. 1980. Nutrient-mediated juvenile hormone secretion in mosquitoes. Journal of Insect Physiology, 26, 113-117. Ferdig, M. T., Beerntsen, B. T., Spray, F. J., Li, J. & Christensen, B. M. 1993. Reproductive costs associated with resistance in a mosquito-filarial worm system. The American Journal of Tropical Medicine and Hygiene, 49, 756. Ferguson, H. M., Bernadette, J., Ng'habi, K. & Knols, B. J. G. 2005. Redressing the sex imbalance in knowledge of vector biology. Trends in Ecology and Evolution, 20, 202-209. Ferguson, H. M., Dornhaus, A., Beeche, A., Borgemeister, C., Gottlieb, M., Mulla, M. S., Gimnig, J. E., Fish, D. & Killeen, G. F. 2010. Ecology: a prerequisite for malaria elimination and eradication. PLoS Medicine, 7, e1000303. Ferguson, H. M., Ng'habi, K., Walder, T., Kadungula, D., Moore, S. J., Lyimo, I., Russell, T. L., Urassa, H., Mshinda, H., Killeen, G. F. & Knols, B. J. G. 2008. Establishment of a large semi-field system for experimental study of African malaria vector ecology and control in Tanzania. Malaria Journal, 7, 158. Ferguson, H. M. & Read, A. F. 2002. Genetic and environmental determinants of malaria virulence in mosquitoes. Proceedings of the Royal Society B, 269, 1217-1224. Fernandes, L. & Briegel, H. 2005. Reproductive physiology of Anopheles gambiae and Anopheles atroparvus. Journal of Vector Ecology, 30, 11-26. Fernandez, N. M. & Klowden, M. J. 1995. Male accessory gland substances modify the host-seeking behavior of gravid Aedes aegypti mosquitoes. Journal of Insect Physiology, 41, 965-970. Fillinger, U. & Lindsay, S. 2006. Suppression of exposure to malaria vectors by an order of magnitude using microbial larvicides in rural Kenya Tropical Medicine and International Health, 11, 1629-1642. Finstad, A. G., Berg, O. K., Forseth, T., Ugedal, O. & Næsje, T. F. 2010. Adaptive winter survival strategies: defended energy levels in juvenile Atlantic salmon along a latitudinal gradient. Proceedings of the Royal Society of London, Series B: Biological Sciences, 277, 1113-1120. Foster, W. A. 1995. Mosquito sugar feeding and reproductive energetics. Annual Review of Entomology, 40, 443-474. Foster, W. A. 2008. Phytochemicals as population sampling lures. Journal of the American Mosquito Control Association, 24, 138-146. Foster, W. A. & Eischen, F. A. 1987. Frequency of blood-feeding in relation to sugar availability in Aedes aegypti and Anopheles quadrimaculatus (Diptera: Culicidae). Annals of the Entomological Society of America, 80, 103-108. Foster, W. A. & Hancock, R. G. 1994. Nectar-related olfactory and visual attractants for mosquitoes. Journal of the American Mosquito Control Association, 10, 288-296. Foster, W. A. & Takken, W. 2004. Nectar-related vs. human-related volatiles: behavioural response and choice by female and male Anopheles gambiae (Diptera: Culicidae) between emergence and first feeding. Bulletin of Entomological Research, 94,

232

145-157. Fujiwara, M. & Caswell, H. 2002. Estimating Population Projection Matrices from Multi-Stage Mark-Recapture. Ecology, 83, 3257-3265. Gadawski, R. M. & Smith, S. M. 1992. Nectar sources and age structure in a population of Aedes provocans. Journal of Medical Entomology, 29, 879-886. Garcia-Saenz, A., McCarter, P. & Baylis M. 2010. The influence of host number on the attraction of biting midges, Culicoides spp., to light traps. Medical and Veterinary Entomology, no. doi: 10.1111/j.1365-2915.2010.00904.x Garrett-Jones, C. 1964. Prognosis for interruption of malaria transmission through assessment of the mosquito's vectorial capacity. Nature, 204, 1173-1175. Garrett-Jones, C. & Shidrawi, G. R. 1969. Malaria vectorial capacity of a population of Anopheles gambiae: an exercise in epidemiological entomology. Bulletin of the World Health Organization, 40, 531. Gary, R. E., Cannon, J. W. & Foster, W. A. 2009. Effect of sugar on male Anopheles gambiae Giles (Diptera: Culicidae) mating performance, as modified by temperature, space, and body size. Parasites and Vectors, 2, 19. Gary, R. E. & Foster, W. A. 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. Gary, R. E. & Foster, W. A. 2004. Anopheles gambiae feeding and survival on honeydew and extra-floral nectar of peridomestic plants. Medical and Veterinary Entomology, 18, 102-107. Gary, R. E. & Foster, W. A. 2006. Diel timing and frequency of sugar feeding in the mosquito Anopheles gambiae, depending on sex, gonotrophic state and resource availability. Medical and Veterinary Entomology, 20, 308-316. Gentle, L. K. & Gosler, A. C. 2001. Fat reserves and perceived predation risk in the great tit, Parus major. Proceedings of the Royal Society of London, Series B: Biological Sciences, 268, 487-491. Gibson, G. & Russell, I. 2006. Flying in tune: sexual recognition in mosquitoes. Current Biology, 16, 1311-1316. Gillies, M. T. 1954. The recognition of age-groups within populations of Anopheles gambiae by the pre-gravid rate and the sporozoite rate. Annals of Tropical Medicine and Parasitology, 48, 58-74. Gillies, M. T. 1961. Studies on the dispersion and survival of Anopheles gambiae Giles in East Africa, by means of marking and release experiments. Bulletin of Entomological Research, 52, 99-127. Gillies, M. T. 1964. Selection for host preference in Anopheles gambiae. Nature, 203, 852-854. Gillies, M. T. 1968. Journal of meetings. Proceedings of the Royal Entomological Society of London, 33, 27. Gillies, M. T. & Chir, B. 1956. A new character for the recognition of nulliparous females of Anopheles gambiae. Bull. WHO, 15, 451-459. Gillies, M. T. & De Meillon, B. 1968. The Anophelinae of Africa South of the Sahara (Ethiopian Zoogeographical Region): South African Institute for Medical Research. Gooding, R. H. 1975. Digestive enzymes and their control in haematophagous

233

arthropods. Acta Tropica, 32, 96. Gouagna, L. C., Poueme, R. S., Dabire, K. R., Ouedraogo, J., Fontenille, D. & Simard, F. 2010. Patterns of sugar feeding and host plant preferences in adult males of Anopheles gambiae (Diptera: Culicidae). Journal of Vector Ecology, 35, 267-276. Goulson, D. 1999. Foraging strategies of insects for gathering nectar and pollen, and implications for plant ecology and evolution. Perspectives in Plant Ecology, Evolution and Systematics, 2, 185-209. Grossman, G. L., Campos, Y., Severson, D. W. & James, A. A. 1997. Evidence for two distinct members of the amylase gene family in the yellow fever mosquito, Aedes aegypti. Insect Biochemistry and Molecular Biology, 27, 769-781. Gu, W., Muller, G., Schlein, Y., Novak, R. J. & Beier, J. C. 2011. Natural plant sugar sources of Anopheles mosquitoes strongly impact malaria transmission potential. PLOS One, 6, e15996. Gu, W., Regens, J. L., Beier, J. C. & Novak, R. J. 2006. Source reduction of mosquito larval habitats has unexpected consequences on malaria transmission. Proceedings of the National Academy of Sciences, 103, 17560 –17563. Hall-Mendelin, S., Ritchie, S. A., Johansen, C. A., Zborowski, P., Cortis, G., Dandridge, S., Hall, R. A. & van den Hurk, A. F. 2010. Exploiting mosquito sugar feeding to detect mosquito-borne pathogens. Proceedings of the National Academy of Sciences, 107, 11255-11259. Ham, P. J., Phiri, J. S. & Nolan, G. P. 1991. Effect of N acetyl D glucosamine on the migration of Brugia pahangi microfilariae into the haemocoel of Aedes aegypti. Medical and Veterinary Entomology, 5, 485-493. Hamilton, J. G. C. & El Naiem, D. 2000. Sugars in the gut of the sandfly Phlebotomus orientalis from Dinder National Park, Eastern Sudan. Medical and Veterinary Entomology, 14, 64-70. Hancock, R. G. & Foster, W. A. 1997. Larval and adult nutrition effects on blood/nectar choice of Culex nigripalpus mosquitoes. Medical and Veterinary Entomology, 11, 112- 122. Hancock, R. G. & Foster, W. A. 2000. Exogenous juvenile hormone and methoprene, but not male accessory gland substances or ovariectomy, affect the blood/nectar choice of female Culex nigripalpus mosquitoes. Medical and Veterinary Entomology, 14, 376-382. Hancock, R. G., Yee, W. L. & Foster, W. A. 1990. Tests of Sabethes cyaneus leg paddle function in mating and flight. Journal of the American Mosquito Control Association, 6, 733-735. Haramis, L. D. & Foster, W. A. 1983. Visual quantification of sugar in mosquitoes using anthrone reagent. Mosquito News, 43, 362-364. Haramis, L. D. & Foster, W. A. 1990. Gonotrophic cycle duration, population age structure, and onset of sugar feeding and insemination of Aedes triseriatus (Diptera: Culicidae). Journal of Medical Entomology, 27, 421-428. Hardy, J. L., Houk, E. J., Kramer, L. D. & Reeves, W. C. 1983. Intrinsic factors affecting vector competence of mosquitoes for arboviruses. Annual Review of Entomology, 28, 229-262. Harrington, L. C., Edman, J. D. & Scott, T. W. 2001. Why do female Aedes aegypti (Diptera: Culicidae) feed preferentially and frequently on human blood? Journal of

234

Medical Entomology, 38, 411-422. Harrington, L. C., Jones, J. J., Kitthawee, S., Sithiprasasna, R., Edman, J. D. & Scott, T. W. 2008. Age-dependent survival of the dengue vector Aedes aegypti (Diptera: Culicidae) demonstrated by simultaneous release-recapture of different age cohorts. Journal of Medical Entomology, 45, 307-313. Hay, S. I., Snow, R. W. & Rogers, D. J. 1998. Predicting malaria seasons in Kenya using multitemporal meteorological satellite sensor data. Transactions of the Royal Society of Tropical Medicine and Hygiene, 92, 12-20. Healy, T. P. & Jepson, P. C. 1988. The location of floral nectar sources by mosquitoes: the long-range responses of Anopheles arabiensis Patton (Diptera: Culicidae) to Achillea millefolium flowers and isolated floral odour. Bulletin of Entomological Research, 78, 651-657. Heimpel, G. E. & Rosenheim, J. A. 1995. Dynamic host feeding by the parasitoid Aphytis melinus: the balance between current and future reproduction. Journal of Animal Ecology, 64, 153-167. Helinski, M. E. H., Hood, R. C. & Knols, B. J. G. 2008. A stable isotope dual-labelling approach to detect multiple insemination in un-irradiated and irradiated Anopheles arabiensis mosquitoes. Parasites & Vectors, 1, 9. Herz W., Watanabe H., Miyazaki M. & Kishida Y. (1962) The structures of parthenin and ambrosin. Journal of the American Chemical Society, 84, 2601- 2610. Hocking, B. 1953. The intrinsic range and speed of flight of insects. Hocking, B. 1968. Insect-flower associations in the high Arctic with special reference to nectar. Oikos, 19, 359-387. Holliday-Hanson, M. L., Yuval, B. & Washino, R. 1997. Energetics and sugar-feeding of field collected anopheline females. Journal of Vector Ecology, 22, 83-89. Hooper M., Kirby G.C., Kulkarni M.M., Kulkarni S.N., Nagasampagi B.A., O'Neill M.J., Phillipson J.D., Rojatkar S.R. & Warhurst D.C. (1990) Antimalarial activity of parthenin and its derivatives. European Journal of Medicinal Chemistry 25: 717-723. Hosmer, D. W. & Lemeshow, S. 2000. Applied logistic regression: Wiley-Interscience. Houston, A. I. & McNamara, J. M. 1985. The choice of two prey types that minimises the probability of starvation. Behavioral Ecology and Sociobiology, 17, 135-141. Houston, A. I. & McNamara, J. M. 1999. Models of adaptive behaviour. An approach based on state. Cambridge: Cambridge University Press. Howell, P. & Knols, B. G. J. 2009. Male mating biology. Malaria Journal, 8, S8. Hudson, A. 1970. Factors affecting egg maturation and oviposition by autogenous Aedes atropalpus. Canadian Entomologist, 102, 939-949. Hunter, F. F. & Ossowski, A. M. 1999. Honeydew sugars in wild-caught female horse flies (Diptera: Tabanidae). Journal of Medical Entomology, 36, 896-899. Hurd, H., Hogg, J. C. & Renshaw, M. 1995. Interactions between bloodfeeding, fecundity and infection in mosquitoes. Parasitology Today, 11, 411-416. Impoinvil, D. E., Kongere, J. O., Foster, W. A., Njiru, B. N., Killeen, G. F., Githure, J. I., Beier, J. C., Hassanali, A. & Knols, B. J. G. 2004. Feeding and survival of the malaria vector Anopheles gambiae on plants growing in Kenya. Medical and Veterinary

235

Entomology, 18, 108-115. Jacobsen, R. L., Slutsky, G. M., Greenblatt, C. L. & Schnur, L. F. 1982. Surface reactions of Leishmania I. lectin mediated agglutination. Annals of Tropical Medicine and Parasitology, 76, 45-52. Jacobson, R., Schlein, Y. & Eisenberger, C. 2001. The biological function of sand fly and Leishmania glycosidases. Medical Microbiology and Immunology, 190, 51-55. Jacobson, R. L. & Schlein, Y. 1999. Lectins and toxins in the plant diet of Phlebotomus papatasi (Diptera: Psychodidae) can kill Leishmania major promastigotes in the sandfly and in culture. Annals of Tropical Medicine and Parasitology, 93, 351-356. Jacobson, R. L. & Schlein, Y. 2001. Phlebotomus papatasi and Leishmania major parasites express α-amylase and α-glucosidase. Acta Tropica, 78, 41-49. Jaenson, T. G. & Ameneshewa, B. 1991. Prehibernation diet and reproductive condition of female Anopheles messeae in Sweden. Medical and Veterinary Entomology, 5, 243- 252. James, A. A. & Rossignol, P. A. 1991. Mosquito salivary glands: parasitological and molecular aspects. Parasitology Today, 7, 267-271. Janzen, T. A. & Hunter, F. F. 1998. Honeydew sugars in wild-caught female deer flies (Diptera: Tabanidae). Journal of Medical Entomology, 35, 685-689. Jepson, P. C. & Healy, T. P. 1988. The location of floral nectar sources by mosquitoes: an advanced bioassay for volatile plant odours and initial studies with Aedes aegypti (L.) (Diptera: Culicidae). Bulletin of Entomological Research, 78, 641-650. Jhumur, U. S., Dotterl, S. & Jurgens, A. 2006. Naïve and conditioned responses of Culex pipiens pipiens biotype molestus (Diptera: Culicidae) to flower odors. Journal of Medical Entomology, 43, 1164-1170. Joy, T. K., Arik, A. J., Corby-Harris, V., Johnson, A. A. & Riehle, M. A. 2010. The impact of larval and adult dietary restriction on lifespan, reproduction and growth in the mosquito Aedes aegypti. Experimental gerontology, 45, 685-690. Junnila, A., Müller, G. C. & Schlein, Y. 2010. Species identification of plant tissues from the gut of An. sergentii by DNA analysis. Acta Tropica, 115, 227-233. Kaufmann, C. & Briegel, H. 2004. Flight performance of the malaria vectors Anopheles gambiae and Anopheles atroparvus. Journal of Vector Ecology, 29, 140-153. Kelly, R. & Edman, J. D. 1997. Infection and transmission of Plasmodium gallinaceum (Eucoccida: Plasmodiidae) in Aedes aegypti (Diptera: Culicidae): effect of preinfection sugar meals and postinfection blood meals. Journal of Vector Ecology, 22, 36-42. Klowden, M. J. 1986. Effects of sugar deprivation on the host-seeking behaviour of gravid Aedes aegypti mosquitoes. Journal of Insect Physiology, 32, 479-483. Klowden, M. J. 2006. Switchover to the mated state by spermathecal activation in female Anopheles gambiae mosquitoes. Journal of Insect Physiology, 52, 679-684. Klowden, M. J., Blackmer, J. L. & Chambers, G. M. 1988. Effects of larval nutrition on the host-seeking behavior of adult Aedes aegypti mosquitoes. Journal of the American Mosquito Control Association, 4, 73-75. Klowden, M. J. & Briegel, H. 1994. Mosquito gonotrophic cycle and multiple feeding potential: contrasts between Anopheles and Aedes (Diptera: Culicidae). Journal of Medical Entomology, 31, 618-622. Klowden, M. J. & Dutro, S. M. 1990. Effects of carbohydrate ingestion on the pre-

236

oviposition behavior of the mosquito Aedes aegypti (L.). Bulletin of the Society for Vector Ecology, 15, 59-62. Klowden, M. J. & Lea, A. O. 1979. Abdominal distention terminates subsequent host- seeking behaviour of Aedes aegypti following a blood meal. Journal of Insect Physiology, 25, 583-585. Knols, B. J. G., Njiru, B. N., Mathenge, E. M., Mukabana, W. R., Beier, J. C. & Killeen, G. F. 2002. MalariaSphere: A greenhouse-enclosed simulation of a natural Anopheles gambiae (Diptera: Culicidae) ecosystem in western Kenya. Malaria Journal, 1, 19. Koella, J. C. & Boëte, C. 2002. A genetic correlation between age at pupation and melanization immune response of the yellow fever mosquito Aedes aegypti. Evolution, 56, 1074-1079. Koella, J. C. & Sørensen, F. L. 2002. Effect of adult nutrition on the melanization immune response of the malaria vector Anopheles stephensi. Medical and Veterinary Entomology, 16, 316-320. Koella JC, Sørensen FL, Anderson RA, 1998. The malaria parasite, Plasmodium falciparum, increases the frequency of multiple feeding of its mosquito vector, Anopheles gambiae. Proceedings of the Royal Society of London. Series B: Biological Sciences 265: 763-768. Laarman, J. J. 1968. The intake of sugars by females of wild Anopheles gambiae and Anopheles funestus. Acta Leidensia, 36, 137–144. Lambrechts, L., Chavatte, J., Snounou, G. & Koella, J. C. 2006. Environmental influence on the genetic basis of mosquito resistance to malaria parasites. Proceedings of the Royal Society B, 273, 1501-1506. Lea, A. O. 1965. Sugar-baited insecticide residues against mosquitoes. Mosquito News, 25, 65-66. Lea, A. O., Dimond, J. B. & DeLong, D. M. 1958. Some nutritional factors in egg production by Aedes aegypti. Proc. 10th Int. Congr. Entomol., 1956, 793-796. Lindh, J. M., Terenius, O., Eriksson-Gonzales, K., Knols, B. G. J. & Faye, I. 2006. Re-introducing bacteria in mosquitoes--A method for determination of mosquito feeding preferences based on coloured sugar solutions. Acta Tropica, 99, 173-183. Lounibos, L. P. & Conn, J. 1991. Fecundity, parity, and adult feeding relationships among Nyssorhynchus malaria vectors from Venezuela. Memórias do Instituto Oswaldo Cruz, 86, 57-66. Lyimo, E. O. & Takken, W. 1993. Effects of adult body size on fecundity and the pre- gravid rate of Anopheles gambiae females in Tanzania. Medical and Veterinary Entomology, 7, 328-332. Ma, B. O. & Roitberg, B. D. 2008. The role of resource availability and state- dependence in the foraging strategy of blood-feeding mosquitoes. Evolutionary Ecology Research, 10, 1111-1130. Macdonald, G. 1957. The Epidemiology and Control of Malaria: Oxford University Press. Magnarelli, L. A. 1978. Nectar-feeding by female mosquitoes and its relation to follicular development and parity. Journal of Medical Entomology, 14, 527-530. Magnarelli, L. A. 1979. Diurnal nectar-feeding of Aedes cantator and A. sollicitans

237

(Diptera: Culicidae). Environmental Entomology, 8, 949-955. Magnarelli, L. A. 1980. Bionomics of Psorophora ferox (Diptera: Culicidae): seasonal occurrence and acquisition of sugars. Journal of Medical Entomology, 17, 328-332. Magnarelli L.A. & Anderson J.F. (1977) Follicular development in salt marsh Tabanidae (Diptera) and incidence of nectar feeding with relation to gonotrophic activity. Annals of the Entomological Society of America 70, 529-533. Maier, W. A., Becker-Feldman, H. & Seitz, H. M. 1987. Pathology of malaria-infected mosquitoes. Parasitology Today, 3, 216-218. Mains, J. W., Mercer, D. R. & Dobson, S. L. 2008. Digital Image Analysis to Estimate Numbers of Aedes Eggs Oviposited in Containers. Journal of the American Mosquito Control Association, 24, 496-501. Manda, H., Gouagna, L. C., Foster, W. A., Jackson, R. R., Beier, J. C., Githure, J. I. & Hassanali, A. 2007. Effect of discriminative plant-sugar feeding on the survival and fecundity of Anopheles gambiae. Malaria Journal, 6, 113. Manda, H., Gouagna, L. C., Kabiru, E. W., Foster, W. A., Beier, J. C., Hassanali, A. & Githure, J. I. 2007. Discriminative feeding behavior of Anopheles gambiae s.s. on different plant species and effects on its survival, fecundity, and vector competence in a malaria endemic area of western Kenya. Proc. 56th Meeting of Am. Soc. Trop. Med. Hyg. (Abstract). Manda, H., Gouagna, L. C., Kabiru, E. W., Hassanali, A., Yan, G., Beier, J. C. & Githure, J. I. 2005. Plasmodium falciparum development in the midgut of Anopheles gambiae s.s. feeding on some predominant plants in Western Kenya. Acta Tropica, 95S, S17-18 (Abstract). Manda, H., Gouagna, L. C., Nyandat, E., Kabiru, W., Jackson, R. R., Foster, W. A., Githure, J. I., Beier, J. C. & Hassanali, A. 2007. Discriminative feeding behaviour of Anopheles gambiae s.s on endemic plants in Western Kenya. Medical and Veterinary Entomology, 21, 103-111. Mangel, M. & Clark, C. W. 1988. Dynamic modeling in behavioral ecology: Princeton Univ Pr. Marchand, R. P. 1984. Field observations on swarming and mating in Anopheles gambiae mosquitoes in Tanzania. Netherlands Journal of Zoology, 34, 367-387. Marinotti, O., James, A. A. & Ribeiro, J. M. 1990. Diet and salivation in female Aedes aegypti mosquitoes. Journal of Insect Physiology, 36, 545-548. Martinez-Ibarra, J. A., Rodriguez, M. H., Arredondo-Jimenez, J. I. & Yuval, B. 1997. Influence of plant abundance on nectar feeding by Aedes aegypti (Diptera: Culicidae) in southern Mexico. Journal of Medical Entomology, 34, 589-593. Mathenge, E. M., Gimnig, J. E., Kolczak, M., Ombok, M., Irungu, L. W. & Hawley, W. A. 2001. Effect of permethrin-impregnated nets on exiting behavior, blood feeding success, and time of feeding of malaria mosquitoes (Diptera: Culicidae) in western Kenya. Journal of Medical Entomology, 38, 531-536. Mauer, D. J. & Rowley, W. A. 1999. Attraction of Culex pipiens pipiens (Diptera: Culicidae) to flower volatiles. Journal of Medical Entomology, 36, 503-507. Mbogo, C. N. M., Baya, N. M., Ofulla, A. V. O., Githure, J. I. & Snow, R. W. 1996. The impact of permethrin impregnated bednets on malaria vectors of the Kenyan coast. Medical and Veterinary Entomology, 10, 251-259.

238

McCall, P. J. & Eaton, G. 2001. Olfactory memory in the mosquito Culex quinquefasciatus. Medical and Veterinary Entomology, 15, 197-203. McCall, P. J. & Kelly, D. W. 2002. Learning and memory in disease vectors. Trends in Parasitology, 18, 429-433. McCall, P. J., Mosha, F. W., Njunwa, K. J. & Sherlock, K. 2001. Evidence for memorized site-fidelity in Anopheles arabiensis. Transactions of the Royal Society of Tropical Medicine and Hygiene, 95, 587-590. McCrae, A. W. R. 1989. Differences in sugar-feeding activity between tropical and temperate mosquitoes: field observations and their implications. Vector Ecology Newsletter, 20, 16. Midega, J. T., Mbogo, C. M., Mwambi, H., Wilson, M. D., Ojwang, G., Mwangangi, J. M., Nzovu, J. G., Githure, J. I., Yan, G. & Beier, J. C. 2007. Estimating dispersal and survival of Anopheles gambiae and Anopheles funestus along the Kenyan coast by using mark–release–recapture methods. Journal of Medical Entomology, 44, 923-929. Minakawa, N., Githure, J. I., Beier, J. C. & Yan, G. 2001. Anopheline mosquito survival strategies during the dry period in western Kenya. Journal of Medical Entomology, 38, 388-392. Minakawa, N., Seda, P. & Yan, G. 2002. Influence of host and larval habitat distribution on the abundance of African malaria vectors in western Kenya. The American Journal of Tropical Medicine and Hygiene, 67, 32-38. Morrison, A. C., Costero, A., Edman, J. D. & Scott, T. W. 1999. Increased fecundity of female Aedes aegypti (L.)(Diptera: Culicidae) fed only human blood prior to release in Puerto Rico. Journal of the American Mosquito Control Association, 15, 98-104. Mostowy, W. M. & Foster, W. A. 2004. Antagonistic effects of energy status on meal size and egg-batch size of Aedes aegypti (Diptera; Culicidae) Journal of Vector Ecology, 29, 84-93. Muirhead-Thompson, R. C. 1951. Mosquito Behavior in Relation to Malaria Transmission in the Tropics. London: E. Arnold. Müller, G. & Schlein, Y. 2004. Nectar and honeydew feeding of Phlebotomus papatasi in a focus of Leishmania major in Neot Hakikar oasis. Journal of Vector Ecology, 29, 154-158. Müller, G. & Schlein, Y. 2005. Plant tissues: the frugal diet of mosquitoes in adverse conditions. Medical and Veterinary Entomology, 19, 413-422. Müller, G. & Schlein, Y. 2006. Sugar questing mosquitoes in arid areas gather on scarce blossoms that can be used for control. International Journal for Parasitology, 36, 1077- 1080. Müller, G. C., Beier, J. C., Traore, S. F., Toure, M. B., Traore, M. M., Bah, S., Doumbia, S. & Schlein, Y. 2010. Field experiments of Anopheles gambiae attraction to local fruits/seedpods and flowering plants in Mali to optimize strategies for malaria vector control in Africa using attractive toxic sugar bait methods. Malaria Journal, 9, 262. Müller, G. C., Beier, J. C., Traore, S. F., Toure, M. B., Traore, M. M., Bah, S., Doumbia, S. & Schlein, Y. 2010. Successful field trial of attractive toxic sugar bait (ATSB) plant-spraying methods against malaria vectors in the Anopheles gambiae complex in Mali, West Africa. Malaria Journal, 9, 210-210.

239

Müller, G. C., Junnila, A. & Schlein, Y. 2010. Effective Control of Adult Culex pipiens by Spraying an Attractive Toxic Sugar Bait Solution in the Vegetation Near Larval Habitats. Journal of Medical Entomology, 47, 63-66. Müller, G. C., Kravchenko, V. D. & Schlein, Y. 2008. Decline of Anopheles sergentii and Aedes caspius populations following presentation of attractive toxic (spinosad) sugar bait stations in an oasis. Journal of the American Mosquito Control Association, 24, 147- 149. Müller, G. C., Revay, E. E. & Schlein, Y. 2011. Relative attraction of the sand fly Phlebotomus papatasi to local flowering plants in the Dead Sea region. Journal of Vector Ecology, 36, S187-S194. Müller, G. C. & Schlein, Y. 2008. Efficacy of toxic sugar baits against adult cistern- dwelling Anopheles claviger. Transactions of the Royal Society of Tropical Medicine and Hygiene, 102, 480-484. Müller, G. C., Xue, R. D. & Schlein, Y. 2011. Differential attraction of Aedes albopictus in the field to flowers, fruits and honeydew. Acta Tropica, 118, 45-49. Müller, G. C., Xue, R. D. & Schlein, Y. 2010. Seed pods of the carob tree Ceratonia siliqua are a favored sugar source for the mosquito Aedes albopictus in coastal Israel. Acta Tropica, 116, 235-239. N’Guessan, R., Corbel, V., Akogbéto, M. & Rowland, M. 2007. Reduced efficacy of insecticide-treated nets and indoor residual spraying for malaria control in pyrethroid resistance area, Benin. Emerging Infectious Diseases, 13, 199-206. Naksathit, A. T. & Scott, T. W. 1998. Effect of female size on fecundity and survivorship of Aedes aegypti fed only human blood versus human blood plus sugar. Journal of the American Mosquito Control Association, 14, 148-152. Nasci, R. S. 1991. Influence of larval and adult nutrition on biting persistence in Aedes aegypti (Diptera: Culicidae). Journal of Medical Entomology, 28, 522-526. Nayar, J. K. 1978. The detection of nectar sugars in field-collected Culex nigripalpus and Its application. Annals of the Entomological Society of America, 71, 55-59. Nayar, J. K. & Sauerman, D. M. 1975. The effects of nutrition on survival and fecundity in Florida mosquitoes part 2: utilization of a blood meal for survival. Journal of Medical Entomology, 12, 99-103. Nayar, J. K. & Van Handel, E. 1971. The fuel for sustained mosquito flight. Journal of Insect Physiology, 17, 471-481. Ng'habi, K., Huho, B. J., Nkwengulila, G., Killeen, G. F., Knols, B. J. G. & Ferguson, H. M. 2008. Sexual selection in mosquito swarms: may the best man lose? Animal Behaviour, 76, 105-112. Norris, L. C., Fornadel, C. M., Hung, W. C., Pineda, F. J. & Norris, D. E. 2010. Frequency of multiple blood meals taken in a single gonotrophic cycle by Anopheles arabiensis mosquitoes in Macha, Zambia. The American Journal of Tropical Medicine and Hygiene, 83, 33-37. Nunes, R. D., de Oliveira, R. L. & Braz, G. R. C. 2008. A novel method for measuring fructose ingestion by mosquitoes. Journal of Vector Ecology, 33, 225-231. O’Meara, G. F. 1987. Nutritional ecology of blood-feeding Diptera. Nutritional Ecology of Insects, Mites and Spiders. Wiley, New York, 741–764. Okanda, F. M., Dao, A., Njiru, B. N., Arija, J., Akelo, H. A., Toure, Y., Odulaja, A.,

240

Beier, J. C., Githure, J. I., Yan, G., Gouagna, L. C., Knols, B. J. G. & Killeen, G. F. 2002. Behavioural determinants of gene flow in malaria vector populations: Anopheles gambiae males select large females as mates. Malaria Journal, 1, 10. Okech, B. A., Gouagna, L. C., Killeen, G. F., Knols, B. G. J., Kabiru, E. W., Beier, J. C., Yan, G. & Githure, J. I. 2003. Influence of sugar availability and indoor microclimate on survival of Anopheles gambiae (Diptera: Culicidae) under semifield conditions in western Kenya. Journal of Medical Entomology, 40, 657-663. Okech, B. A., Gouagna, L. C., Walczak, E., Kabiru, E. W., Beier, J. C., Yan, G. & Githure, J. I. 2004. The development of Plasmodium falciparum in experimentally infected Anopheles gambiae (Diptera: Culicidae) under ambient microhabitat temperature in western Kenya. Acta Tropica, 92, 99-108. O'Meara, G. F. 1985. Ecology of autogeny in mosquitoes. Vero Beach, Fl.: Fla. Med. Entomol. Lab. Onyabe, D. Y., Roitberg, B. D. & Friend, W. G. 1997. Feeding and mating in Anopheles (Diptera: Culicidae): theoretical modeling approach. Journal of Medical Entomology, 34, 644-650. Packer, M. J. & Corbet, P. S. 1989. Size variation and reproductive success of female Aedes punctor (Diptera: Culicidae). Ecological Entomology, 14, 297-309. Pates, H. & Curtis, C. 2005. Mosquito behavior and vector control. Annual Review of Entomology, 50, 53-70. Patterson, R. S., Smittle, B. J. & DeNeve, R. T. 1969. Feeding habits of male southern house mosquitoes on P-labeled and unlabeled plants. Journal of Economic Entomology, 62, 1455-1458. Pedigo L.P. 2002. Entomology and pest management (4th ed). Prentice Hall, Upper Saddle River, New Jersey. Pennetier, C., Warren, B., Roch Dabire, K., Russell, I. & Gibson, G. 2010. "Singing on the wing" as a mechanism for species recognition in the malarial mosquito Anopheles gambiae. Current Biology, 20, 131-136. Perez-Pacheco, R. 2005. Control of the mosquito Anopheles pseudopunctipennis (Diptera : Culicidae) with Romanomermis iyengari (Nematoda : Mermithidae) in Oaxaca, Mexico. Biological control, 32, 137-142. Pletcher, S. D. 1999. Model fitting and hypothesis testing for age-specific mortality data. Journal of Evolutionary Biology, 12, 430-439. Pletcher, S. D., Khazaeli, A. A. & Curtsinger, J. W. 2000. Why do life spans differ? Partitioning mean longevity differences in terms of age-specific mortality parameters. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 55, B381. Pumpuni, C. B., Demaio, J., Kent, M., Davis, J. R. & Beier, J. C. 1996. Bacterial population dynamics in three anopheline species: the impact on Plasmodium sporogonic development. The American Journal of Tropical Medicine and Hygiene, 54, 214-218. Rasgon, J. L., Styer, L. M. & Scott, T. W. 2003. Wolbachia-induced mortality as a mechanism to modulate pathogen transmission by vector arthropods. Journal of Medical Entomology, 40, 125-132. Reeves, W. C., Brookman, B. & Hammon, W. 1948. Studies on the flight range of certain Culex mosquitoes using a fluorescent dye marker, with notes on Culiseta and

241

Anopheles. Mosquito News, 8, 62. Rehman, A. M., Coleman, M., Schwabe, C., Baltazar, G., Matias, A., Gomes, I. R., Yellott, L., Aragon, C., Nchama, G. N. & Mzilahowa, T. 2011. How much does malaria vector control quality matter: the epidemiological impact of holed nets and inadequate indoor residual spraying. PLoS One, 6, e19205. Reisen, W. K. 2004. Lessons from the past: historical studies by the University of Maryland and the University of California, Berkeley. Wageningen UR Frontis Series, 2, 25-32. Reisen, W. K. 2010. Landscape epidemiology of vector-borne diseases. Annual Review of Entomology, 55, 461-483. Reisen, W. K., Meyer, R. P. & Milby, M. M. 1986. Patterns of fructose feeding by Culex tarsalis (Diptera: Culicidae). Journal of Medical Entomology, 23, 366-373. Ren, X., Hoiczyk, E., Rasgon, J. L. & Schneider, D. S. 2008. Viral paratransgenesis in the malaria vector Anopheles gambiae. PLoS Pathogens, 4, e1000135 Renshaw, M., Service, M. W. & Birley, M. H. 1994. Host finding, feeding patterns and evidence for a memorized home range of the mosquito Aedes cantans. Medical and Veterinary Entomology, 8, 187-193. Renshaw, M., Silver, J. B. & Service, M. W. 1995. Differential lipid reserves influence host seeking behaviour in the mosquitoes Aedes cantans and Aedes punctor. Medical and Veterinary Entomology, 9, 381-387. Rhainds, M. 2010. Female mating failures in insects. Entomologia Experimentalis et Applicata, 136, 211-226. Ribeiro, J., Rowton, E. D. & Charlab, R. 2000. Salivary amylase activity of the phlebotomine sand fly, Lutzomyia longipalpis. Insect Biochemistry and Molecular Biology, 30, 271-277. Riehle, M. A. & Jacobs-Lorena, M. 2005. Using bacteria to express and display anti- parasite molecules in mosquitoes: current and future strategies. Insect Biochemistry and Molecular Biology, 35, 699-707. Riehle, M. A., Moreira, C. K., Lampe, D., Lauzon, C. & Jacobs-Lorena, M. 2007. Using bacteria to express and display anti-Plasmodium molecules in the mosquito midgut. International Journal for Parasitology, 37, 595-603. Rivero, A. & Ferguson, H. M. 2003. The energetic budget of Anopheles stephensi infected with Plasmodium chabaudi: is energy depletion a mechanism for virulence? Proceedings of the Royal Society of London. Series B: Biological Sciences, 270, 1365- 1371. Robich, R. M. & Denlinger, D. L. 2005. Diapause in the mosquito Culex pipiens evokes a metabolic switch from blood feeding to sugar gluttony. Proceedings of the National Academy of Sciences of the United States of America, 102, 15912-15917. Roff, D. A. 1992. The evolution of life histories: theory and analysis: Springer. Roitberg, B. D. & Friend, W. G. 1992. A general theory for host seeking decisions in mosquitoes. Bulletin of mathematical biology, 54, 401-412. Roitberg, B. D., Keiser, S. & Hoffmeister, T. 2010. State dependent attacks in a mosquito. Physiological Entomology, 35, 46-51. Roitberg, B. D., Smith, J. J. & Friend, W. G. 1994. Host response profiles: a new theory to help us understand why and how attractants attract. Journal of the American

242

Mosquito Control Association, 10, 333. Ross, R. 1910. The prevention of malaria. London: John Murray. Russell, C. B. & Hunter, F. E. 2002. Analysis of nectar and honeydew feeding in Aedes and Ochlerotatus mosquitoes. Journal of the American Mosquito Control Association, 18, 86-90. Russell, P. F., West, L. D., Manwell, R. D. & Macdonald, G. 1963. Practical malariology: Oxford University Press. Russell, T. L., Govella, N. J., Azizi, S., Drakeley, C. J., Kachur, S. P. & Killeen, G. F. 2011. Increased proportions of outdoor feeding among residual malaria vector populations following increased use of insecticide-treated nets in rural Tanzania. Malaria Journal, 10, 80. Russell, T. L., Lwetoijera, D. W., Knols, B. G. J., Takken, W., Killeen, G. F. & Ferguson, H. M. 2011. Linking individual phenotype to density-dependent population growth: the influence of body size on the population dynamics of malaria vectors. Proceedings of the Royal Society B: Biological Sciences. doi: 10.1098/rspb.2011.0153 Russell, T. L., Lwetoijera, D. W., Maliti, D., Chipwaza, B., Kihonda, J., Charlwood, J. D., Smith, T. A., Lengeler, C., Mwanyangala, M. A., Nathan, R., Knols, B. G. J., Takken, W. & Killeen, G. F. 2010. Impact of promoting longer-lasting insecticide treatment of bed nets upon malaria transmission in a rural Tanzanian setting with pre- existing high coverage of untreated nets. Malaria Journal, 9, 187. Samish, M. & Akov, S. 1972. Influence of feeding on midgut protease activity in Aedes aegypti. Israel Journal of Entomology, 7, 41-48. Scaraffia, P. Y. & Wells, M. A. 2003. Proline can be utilized as an energy substrate during flight of Aedes aegypti females. Journal of Insect Physiology, 49, 591-601. Schaefer, C. H. & Miura, T. 1972. Sources of energy utilized by natural populations of the mosquito, Culex tarsalis, for overwintering. Journal of Insect Physiology, 18, 797- 805. Schiefer, B. A., Ward, R. A. & Eldridge, B. F. 1977. Plasmodium cynomolgi: effects of malaria infection on laboratory flight performance of Anopheles stephensi mosquitoes. Experimental Parasitology, 41, 397-404. Schlein, Y. 1986. Sandfly diet and Leishmania. Parasitology today, 2, 175. Schlein, Y. & Jacobson, R. L. 1994. Mortality of Leishmania major in Phlebotomus papatasi caused by plant feeding of the sand flies. The American Journal of Tropical Medicine and Hygiene, 50, 20. Schlein, Y. & Jacobson, R. L. 1999. Sugar meals and longevity of the sandfly Phlebotomus papatasi in an arid focus of Leishmania major in the Jordan Valley. Medical and Veterinary Entomology, 13, 65-71. Schlein, Y., Jacobson, R. L. & Muller, G. C. 2001. Sand fly feeding on noxious plants: a potential method for the control of leishmaniasis. The American Journal of Tropical Medicine and Hygiene, 65, 300. Schlein, Y. & Muller, G. 1995. Assessment of plant tissue feeding by sand flies (Diptera: Psychodidae) and mosquitoes (Diptera: Culicidae). Journal of Medical Entomology, 32, 882-887. Schlein, Y. & Müller, G. C. 2008. An approach to mosquito control: Using the dominant attraction of flowering Tamarix jordanis trees against Culex pipiens. Journal of Medical

243

Entomology, 45, 384-390. Schlein, Y. & Pener, H. 1990. Bait fed adult Culex pipiens carry the larvicide Bacillus sphaericus to the larval habitat. Medical and Veterinary Entomology, 4, 283-288. Schlein, Y. & Warburg, A. 1986. Phytophagy and the feeding cycle of Phlebotomus papatasi (Diptera: Psychodidae) under experimental conditions. Journal of Medical Entomology, 23, 11-15. Schlein, Y. & Yuval, B. 1987. Leishmaniasis in the Jordan Valley IV. Attraction of Phlebotomus papatasi (Diptera: Psychodidae) to plants in the field. Journal of Medical Entomology, 24, 87-90. Schmidt, J. M. & Friend, W. G. 1991. Ingestion and diet destination in the mosquito Culiseta inornata: effects of carbohydrate configuration. Journal of Insect Physiology, 37, 817-828. Schwartz, A. & Koella, J. C. 2002. Melanization of Plasmodium falciparum and C-25 sephadex beads by field-caught Anopheles gambiae (Diptera: Culicidae) from southern Tanzania. Journal of Medical Entomology, 39, 84-88. Scott, T. W., Amerasinghe, P. H., Morrison, A. C., Lorenz, L. H., Clark, G. G., Strickman, D., Kittayapong, P. & Edman, J. D. 2000. Longitudinal studies of Aedes aegypti (Diptera: Culicidae) in Thailand and Puerto Rico: blood feeding frequency. Journal of Medical Entomology, 37, 89-101. Scott, T. W., Clark, G. G., Lorenz, L. H., Amerasinghe, P. H., Reiter, P. & Edman, J. D. 1993. Detection of multiple blood feeding in Aedes aegypti (Diptera: Culicidae) during a single gonotrophic cycle using a histologic technique. Journal of Medical Entomology, 30, 94-99. Scott, T. W., Githeko, A. K., Fleisher, A., Harrington, L. C. & Yan, G. 2006. DNA profiling of human blood in anophelines from lowland and highland sites in western Kenya. The American Journal of Tropical Medicine and Hygiene, 75, 231. Scott, T. W., Naksathit, A., Day, J. F., Kittayapong, P. & Edman, J. D. 1997. A fitness advantage for Aedes aegypti and the viruses it transmits when females feed only on human blood. American Journal of Tropical Medicine and Hygiene, 57, 235-239. Shaukat, A. M., Breman, J. G. & McKenzie, F. E. 2010. Using the entomological inoculation rate to assess the impact of vector control on malaria parasite transmission and elimination. Malaria Journal, 9, 122-122. Shroyer, D. A. & Sanders, D. P. 1977. The influence of carbohydrate-feeding and insemination on oviposition of an Indiana strain of Aedes vexans (Diptera: Culicidae). Journal of Medical Entomology, 14, 121-127. Shuel, R. W. 1992. The production of nectar and pollen. The Hive and the Honey Bee, 4th ed., Dadant and Sons, Hamilton, Illinois, 401–425. Skalski, J. R., Millspaugh, J. J., Dillingham, P. & Buchanan, R. A. 2007. Calculating the variance of the finite rate of population change from a matrix model in Mathematica. Environmental Modelling & Software, 22, 359-364. Smith D.L., Dushoff J., McKenzie F.E., 2004. The Risk of a mosquito-borne infection in a heterogeneous environment. PLoS Biology 2: e368. Smith, S. M. & Gadawski, R. M. 1994. Nectar feeding by the early spring mosquito Aedes provocans. Medical and Veterinary Entomology, 8, 201-213. Somani, B. L., Khanade, J. & Sinha, R. 1987. A modified anthrone-sulfuric acid

244

method for the determination of fructose in the presence of certain proteins. Analytical Biochemistry, 167, 327-330. South, S. H. & Arnqvist, G. 2011. Male, but not female, preference for an ornament expressed in both sexes of the polygynous mosquito Sabethes cyaneus. Animal Behaviour, 81, 645-651. Southwood T.R.E. (1976). Ecological methods, with particular reference to the study of insect populations. (2nd ed.) Chapman & Hall, London. Souza-Neto, J. A., Machado, F. P., Lima, J. B., Valle, D. & Ribolla, P. E. M. 2007. Sugar digestion in mosquitoes: Identification and characterization of three midgut α- glucosidases of the neo-tropical malaria vector Anopheles aquasalis (Diptera: Culicidae). Comparative Biochemistry and Physiology-Part A: Molecular & Integrative Physiology, 147, 993-1000. Spencer, C. Y., Pendergast Iv, T. H. & Harrington, L. C. 2005. Fructose variation in the dengue vector, Aedes aegypti, during high and low transmission seasons in the Mae Sot region of Thailand. Journal of the American Mosquito Control Association, 21, 177- 181. Stephens, D. W. & Krebs, J. R. 1986. Foraging theory: Princeton Univ Pr. Stone, C. M., Hamilton, I. M. & Foster, W. A. 2011. A survival and reproduction trade-off is resolved in accordance with resource availability by virgin female mosquitoes. Animal Behaviour, 81, 765-774. Stone, C. M., Taylor, R. M. & Foster, W. A. 2009. An effective indoor mesocosm for studying populations of Anopheles gambiae in temperate climates. Journal of the American Mosquito Control Association, 25, 514-516. Stone, C. M., Taylor, R. M., Roitberg, B. D. & Foster, W. A. 2009. Sugar deprivation reduces insemination of Anopheles gambiae (Diptera: Culicidae), despite daily recruitment of adults, and predicts decline in model populations. Journal of Medical Entomology, 46, 1327-1337. Straif, S. C. & Beier, J. C. 1996. Effects of sugar availability on the blood feeding behaviour of Anopheles gambiae (Diptera: Culicidae). Journal of Medical Entomology, 33, 608-612. Styer, L. M., Carey, J. R., Wang, J. L. & Scott, T. W. 2007. Mosquitoes do senesce: departure from the paradigm of constant mortality. The American Journal of Tropical Medicine and Hygiene, 76, 111-117. Styer, L. M., Minnick, S. L., Sun, A. K. & Scott, T. W. 2007. Mortality and reproductive dynamics of Aedes aegypti (Diptera: Culicidae) fed human blood. Vector- Borne And Zoonotic Diseases, 7, 86-98. Suwanchaichinda, C. & Paskewitz, S. M. 1998. Effects of larval nutrition, adult body size, and adult temperature on the ability of Anopheles gambiae (Diptera: Culicidae) to melanize sephadex beads. Journal of Medical Entomology, 35, 157-161. Suzuki, R., Zhang, Y., Iino, T., Kosako, Y., Komagata, K. & Uchimura, T. 2010. Asaia astilbes sp. nov., Asaia platycodi sp. nov., and Asaia prunellae sp. nov., novel acetic acid bacteria isolated from flowers in Japan. The Journal of General and Applied Microbiology, 56, 339-346. Takken, W. 2002. Do insecticide treated bednets have an effect on malaria vectors? Tropical Medicine & International Health, 7, 1022-1030.

245

Takken, W., Charlwood, J., Billingsley, P. & Gort, G. 1998. Dispersal and survival of Anopheles funestus and A. gambiae s.l. (Diptera: Culicidae) during the rainy season in southeast Tanzania Bulletin of Entomological Research, 88, 561-566. Takken, W., Klowden, M. J. & Chambers, G. M. 1998. Effect of body size on host seeking and blood meal utilization in Anopheles gambiae sensu stricto (Diptera: Culicidae): the disadvantage of being small. Journal of Medical Entomology, 35, 639- 645. R Development Core Team. 2010. R: A language and environment for statistical computing. In: R Foundation for Statistical Computing Vienna Austria. Telang, A. & Wells, M. A. 2004. The effect of larval and adult nutrition on successful autogenous egg production by a mosquito. Journal of Insect Physiology, 50, 677-685. Tesh, R. B., Guzman, H. & Wilson, M. L. 1992. Trans-beta-farnesene as a feeding stimulant for the sand fly Lutzomyia longipalpis (Diptera: Psychodidae). Journal of Medical Entomology, 29, 226-231. Tomberlin, J. K., Rains, G. C., Allan, S. A., Sanford, M. R. & Lewis, W. J. 2006. Associative learning of odor with food-or blood-meal by Culex quinquefasciatus Say (Diptera: Culicidae). Naturwissenschaften, 93, 551-556. Tripet, F., Aboagye-Antwi, F. & Hurd, H. 2008. Ecological immunology of mosquito- malaria interactions. Trends in Parasitology, 24, 219-227. Tsunoda, T., Fukuchi, A., Nanbara, S. & Takagi, M. 2010. Effect of body size and sugar meals on oviposition of the yellow fever mosquito, Aedes aegypti (Diptera: Culicidae). Journal of Vector Ecology, 35, 56-60. Vaidyanathan, R., Fleisher, A. E., Minnick, S. L., Simmons, K. A. & Scott, T. W. 2008. Nutritional stress affects mosquito survival and vector competence for West Nile virus. Vector-Borne And Zoonotic Diseases, 8, 727-732. Van Den Hurk, A. F., Johnson, P. H., Hall-Mendelin, S., Northill, J. A., Simmons, R. J., Jansen, C. C., Frances, S. P., Smith, G. A. & Ritchie, S. A. 2007. Expectoration of flaviviruses during sugar feeding by mosquitoes (Diptera: Culicidae). Journal of Medical Entomology, 44, 845-850. Van Handel, E. 1965. The obese mosquito. The Journal of Physiology, 181, 478. Van Handel, E. 1972. The detection of nectar in mosquitoes. Mosquito News, 32, 458- 458. Van Handel, E. 1985. Rapid determination of glycogen and sugars in mosquitoes. Journal of the American Mosquito Control Association, 1, 299-301. Van Handel, E. & Day, J. F. 1988. Assay of lipids, glycogen and sugars in individual mosquitoes: correlations with wing length in field-collected Aedes vexans. Journal of the American Mosquito Control Association, 4, 549-550. Van Handel, E., Edman, J. D., Day, J. F., Scott, T. W., Clark, C. G., Reiter, P. & Lynn, H. C. 1994. Plant-sugar, glycogen, and lipid assay of Aedes aegypti collected in urban Puerto Rico and rural Florida. Journal of the American Mosquito Control Association, 10, 149-153. Vanden Eng, J. L., Thwing, J., Wolkon, A., Kulkarni, M. A., Manya, A., Erskine, M., Hightower, A. & Slutsker, L. 2010. Assessing bed net use and non-use after long- lasting insecticidal net distribution: a simple framework to guide programmatic strategies. Malaria Journal, 9, 133.

246

Vargo, A. M. & Foster, W. A. 1982. Responsiveness of female Aedes aegypti (Diptera: Culicidae) to flower extracts. Journal of Medical Entomology, 19, 710-718. Vaughan, J. A., Noden, B. H. & Beier, J. C. 1994. Prior blood feeding effects on susceptibility of Anopheles gambiae (Diptera: Culicidae) to infection with cultured Plasmodium falciparum (Haemosporida: Plasmodiidae). Journal of Medical Entomology, 31, 445-449. Verhoek, B. A. & Takken, W. 1994. Age effects on the insemination rate of Anopheles gambiae s.l. in the laboratory. Entomologia Experimenta et Applicata, 72, 167-172. Vrzal, E. M., Allan, S. A. & Hahn, D. A. 2010. Amino acids in nectar enhance longevity of female Culex quinquefasciatus mosquitoes. Journal of Insect Physiology, 1659-1664. Wäckers F.L., Romeis J., van Rijn P., 2007. Nectar and pollen feeding by insect herbivores and implications for multitrophic interactions. Annual Review of Entomology, 52, 301-323. Walker, E. D. & Edman, J. D. 1985. The influence of host defensive behavior on mosquito (Diptera: Culicidae) biting persistence. Journal of Medical Entomology, 22, 370-372. Walker, K. 2008. Nutritional ecology of the malaria vector Anopheles gambiae, Simon Fraser University. Warren, B., Gibson, G. & Russell, I. 2009. Sex recognition through midflight mating duets in Culex mosquitoes is mediated by acoustic distortion. Current Biology, 19, 485- 491. Watanabe, M., Kamimura, K. & Kubota, K. 1973. A simplified method for the analysis of various sugars fed the mosquito, Culex pipiens sl Jap. Journal of Sanitary Zoology, 24, 17-21. Weathersby, A. B. & Noblet, R. 1973. Plasmodium gallinaceum: Development in Aedes aegypti maintained on various carbohydrate diets. Experimental Parasitology, 34, 426- 431. WHO. 2004. Global strategic framework for integrated vector management. pp. 1-12. Geneva: World Health Organization. WHO. 2010. World Malaria Report, 2010. Witter, M. S. & Cuthill, I. C. 1993. The ecological costs of avian fat storage. Philosophical Transactions of the Royal Society of London, Series B: Biological Sciences, 340, 73-92. Wittie, J. 2003. The effect of sugar and blood-host type on the survival, fecundity, biting frequency, and persistence of the eastern tree-hole mosquito, Ochlerotatus triseriatus (Diptera: Culicidae), The Ohio State University. Xue, R. D., Ali, A. & Barnard, D. R. 2008. Host species diversity and post-blood feeding carbohydrate availability enhance survival of females and fecundity in Aedes albopictus (Diptera: Culicidae). Experimental Parasitology, 119, 225-228. Xue, R. D. & Barnard, D. R. 1999. Effects of partial blood engorgement and pretest carbohydrate availability on the repellency of deet to Aedes albopictus. Journal of Vector Ecology, 24, 111. Xue, R. D., Barnard, D. R. & Muller, G. C. 2010. Effects of body size and nutritional regimen on survival in adult Aedes albopictus (Diptera: Culicidae). Journal of Medical

247

Entomology, 47, 778-782. Xue, R. D., Kline, D. L., Ali, A. & Barnard, D. R. 2006. Application of boric acid baits to plant foliage for adult mosquito control. Journal of the American Mosquito Control Association, 22, 497-500. Yamada, Y., Katsura, K., Kawasaki, H., Widyastuti, Y., Saono, S., Seki, T., Uchimura, T. & Komagata, K. 2000. Asaia bogorensis gen. nov., sp. nov., an unusual acetic acid bacterium in the alpha-Proteobacteria. International Journal of Systematic and Evolutionary Microbiology, 50, 823. Yan, G., Severson, D. W. & Christensen, B. M. 1997. Costs and benefits of mosquito refractoriness to malaria parasites: implications for genetic variability of mosquitoes and genetic control of malaria. Evolution, 51, 441-450. Yuval, B. 1992. The other habit: sugar feeding by mosquitoes. Bulletin of the Society of Vector Ecology, 17, 150-156. Yuval, B. 2006. Mating systems of blood-feeding flies. Annual Review of Entomology, 51, 413-440. Yuval, B. & Bouskila, A. 1993. Temporal dynamics of mating and predation in mosquito swarms. Oecologia, 95, 65-69. Yuval, B., Holliday-Hanson, M. L. & Washino, R. 1994. Energy budget of swarming male mosquitoes. Ecological Entomology, 19, 74-78. Zhou, G., Munga, S., Minakawa, N., Githeko, A. K. & Yan, G. 2007. Spatial relationship between adult malaria vector abundance and environmental factors in western Kenya highlands. The American Journal of Tropical Medicine and Hygiene, 77, 29-35.

248