Antagonistic effects of energy status on meal size and egg-batch size of Aedes aegypti (Diptera: Culicidae)

Transcription

1 84 Journal of Vector Ecology June, 2004 Antagonistic effects of energy status on meal size and egg-batch size of Aedes aegypti (Diptera: Culicidae) Wieslaw M. Mostowy and Woodbridge A. Foster * Department of Entomology, The Ohio State University, 318 W. 12th Avenue, Columbus, OH *Corresponding author Received 3 June 2003; Accepted 13 August 2003 ABSTRACT: The effect of sugar feeding on egg-batch size in Aedes aegypti was examined in a way that would distinguish between the roles of a recent sugar meal (full vs. empty crop) and of repeated sugar feeding and digestion (high vs. low energy reserves). Egg numbers of females representing the four combinations of these characteristics (full crop/high reserves, full crop/low reserves, empty crop/high reserves, empty crop/low reserves) were counted during their first gonotrophic cycle. In addition, the sizes of their replete sugar meals and human blood meals were measured to understand the interactions between them. Results demonstrated that blood-fed mosquitoes with full crops and low energy reserves produced the fewest eggs (mean = 56.2); those with empty crops and high energy reserves produced the most eggs (mean = 84.6); and those with the opposite combinations (full/high and empty/ low) had an intermediate fecundity (mean = 75.2 and 76.9, respectively). This ranking of fecundities did not correspond to blood-meal size ranks, owing to direct and indirect effects of energy reserves on meal sizes and egg number. Full-crop females with low reserves ingested the smallest blood meals (mean = 0.62 mg) and had the lowest fecundities. Full-crop females with high reserves ingested more blood (mean = 0.82 mg) and produced more eggs. But empty-crop females with low reserves ingested the largest blood meals (mean = 1.19 mg), yet produced significantly fewer eggs than their high-reserve counterparts, which took smaller blood meals (mean = 0.99 mg). These results demonstrate extremes in the reproductive penalty of crop sugar and the reproductive reward of digested sugar. Energy reserves and an empty crop are similarly valuable in promoting fecundity at the time blood is taken. Journal of Vector Ecology 29 (1): Keyword Index: Aedes aegypti, energy reserves, sugar meal size, blood meal size, fecundity. INTRODUCTION Vertebrate blood and plant sugar contribute both to egg production and to somatic energy reserves used for flight and survival of anautogenous adult female mosquitoes (Foster 1995). Blood meals of sufficient size set in motion the endocrine events that lead to egg maturation and supply the amino acids used in that process. Over a range of blood meal sizes the amount of blood ingested is positively correlated with the number of eggs produced (Woke et al. 1956, Colless and Chellapah 1960, Jalil 1974), yet the proportion of amino acids converted into yolk proteins decreases with increasing meal size (Briegel 1985, 1990). The remaining amino acids are catabolized and their components either excreted or converted to somatic energy reserves, principally lipid and glycogen (Van Handel 1965). These reserves are available for flight, survival, and the nonprotein constituents of eggs. Sugar meals also contribute to this lipid-glycogen energy pool (Van Handel 1965) and thus contribute to egg production indirectly. Imbibed sugar is stored temporarily in the esophageal diverticula, mainly the large ventral diverticulum (crop) that extends into the abdomen. The sugar is transferred gradually to the midgut where absorption occurs. Sugar that exceeds immediate metabolic needs, or is required to maintain the trehalose pool in the hemolymph, is converted to reserves (Van Handel 1965) and stored principally in the fat body but also in thoracic muscle (Clements 1955). As these reserves approach an upper limit, sugar intake decreases (Nayar and Sauerman 1974), and the crop is emptied more slowly (Miles 1977). It appears that the non-protein constituents of eggs are derived primarily from the pre-existing somatic energy pool (Ziegler and Ibrahim 2001, Briegel et al. 2002), though a sugar meal taken even right before or after a blood meal can influence the egg-making decision (Klowden and Chambers 1989). As a result, females with depleted energy reserves often either produce no eggs at all

2 June, 2004 Journal of Vector Ecology 85 (Klowden 1986) or produce fewer eggs (Nayar and Sauerman 1975, Harrington et al. 2001, Briegel et al. 2002) when they take a blood meal. Thus, the size of the energy reserve establishes the relationship between blood meal size and egg-clutch size. The positive effect of sugar feeding on egg production contrasts with what appears to be a negative interaction between sugar feeding and blood meal size or egg-clutch size. Nayar and Sauerman (1975) found that mosquitoes of several species took larger blood meals if they had been held without sugar for 2-5 d after emergence when compared with mosquitoes maintained on 10% sucrose for 11 d after emergence. In this case, the groups taking the larger blood meals produced fewer eggs. Lachmajer and Antonowicz (1983) also noticed an inverse relation between sugar feeding and blood meal size. Although the sugar-fed females of Nayar and Sauerman were starved for 24 h prior to blood feeding, it is not known whether they took less blood because they contained substantial energy reserves in the fat body or because they had crops distended with sucrose. In a release-recapture study, Morrison et al. (1999) attributed the lower fecundity of honey-fed females to the smaller blood meals taken before release, as a result of their enlarged crops. It is important to know whether crop sugar or the somatic reserve is responsible for smaller blood meals and how they act separately and together to influence egg production. These conditions bear directly on the significance of the interval between sugar feeding and blood feeding in nature. That interval determines whether imbibed sugar will still be in the crop or will have been converted to somatic energy reserves by the time blood is taken. Thus, the sugar-blood interval might determine the fecundity of the individual. MATERIALS AND METHODS Mosquito preparation Experimental mosquitoes were derived from a colony of domestic Aedes aegypti (L.) (Rockefeller strain) maintained in this laboratory for >15 yr. Females of this common synanthropic form are noted for being strongly anthropophilic, for converting a large proportion of their primary (oogenic) human blood meals into energy reserves, and for taking one or more supplementary blood meals during a gonotrophic cycle, meals that contribute to reserves only. As a result, females can survive for extended periods on blood alone, and field samples indicate that plant sugar may be ingested infrequently in the domestic environment (Macfie 1915, Edman et al. 1992, Van Handel et al. 1994, Costero et al. 1998). Nonetheless, field studies also show that sugar feeding remains a part of the species repertoire of females even in domestic environments and that it is quite common where humans are less available or where plant sugar is more so (Nayar 1981, Van Handel et al. 1994, Martinez- Ibarra et al. 1997). Therefore, sugar-blood interactions are relevant to this species. Mosquitoes were reared at low larval density and high diet according to methods described by Klowden and Lea (1978) and Haramis and Foster (1983) to produce nearly synchronized emergence of uniformly large adults. Adults were held at 27±1 o C and 70-80% RH. All those used in experiments emerged within a 48- h period and were allowed continuous access to 10% sucrose solution and to water from wicks, starting within 24 h of emergence. Water wicks and vials were changed every 3-4 d, sucrose wicks and vials every day. Females were held together with equal numbers of males, 0-3 d older, in 86-l acrylic cages until 5 d after emergence, to allow mating. Females were then divided into two groups and transferred to clean cages. After the 5-d mating and sugar-feeding period, one group continued to receive 10% sucrose to build energy reserves still further. Beyond that time, the other group was allowed only water, to deplete energy reserves until >50% were dead. These groups will be referred to as high-reserve and low-reserve mosquitoes, respectively (Figure 1). Two to three days before the low-reserve group surpassed 50% mortality, sucrose was removed from the high-reserve females, to allow the latter to empty their crops. The appropriate time to start depriving highreserve females had been determined previously in pilot experiments. At this time, which was 6-7 d after the start of deprivation of low-reserve females, depending on the replicate, 21-42% had died, and the high-reserve mosquitoes had had access to sucrose for a total of d. After the additional 2-3 d with only water, >50% of the low-reserve females had died and >70% of the high-reserve females had emptied their crops (no liquid visible, but gas bubbles usually present), as determined by daily samples of 10 females. High-reserve females that still retained sugar solution in the crop had negligible amounts. This procedure produced two groups of approximately 14-d-old females differing substantially in the amount of reserves in the abdominal fat-body (Figure 1). Mean total energy content of mosquitoes in the two groups, based on lipid and carbohydrate analysis of individuals prepared as described above, was as follows: low reserve, 0.53 cal (n = 62); high reserve, 3.34 cal (n = 47) [unpublished data of W.A.Foster and R.G. Hancock, following methods of Van Handel (1985a,b)]. The baseline value for total lipids and carbohydrates unavailable for survival, based on analyses of females that died during sugar deprivation, was 0.37 cal (n =

3 86 Journal of Vector Ecology June, ). In nature, Ae. aegypti females normally would be much younger than 14 d when taking their first blood meals. However, extreme energy-reserve differences compared in these experiments are, in nature, produced in young females primarily as a result of differences in available larval food. These correlate with body-size differences (Briegel 1990), which would have introduced a confounding variable in the experiments. Energyreserve differences used here, in older females, would be the result of energy accumulation during several gonotrophic cycles. Feeding and egg production High-reserve and low-reserve groups were each divided into two sub-groups, one of which was given a replete 50% sucrose meal from wicks containing the solution, by providing all females tarsal contact with the wick. These sub-groups will be referred to as emptycrop and full-crop mosquitoes. The selected sucrose concentration falls within the typical natural range of sugars in nectars and honeydews (Foster 1995) and results in the ingestion of large meals and a gradual crop emptying, thus maximizing the meals effects on blood meal size. Six ± 0.5 h after the sugar meal, females of all four treatment groups were presented with a human hand either at close range or in physical contact and were allowed to engorge to repletion on blood. Repletion was defined as unprovoked withdrawal of the proboscis from the sugar wick or human skin. Females refusing either sugar or blood meals were discarded. Some blood-fed females from each treatment were pooled in 7-l acrylic cages of individuals, killed 3 d after the blood meal, and dissected to count the number of mature ovarian follicles (chorionated eggs). Others, to be used simultaneously in another study, were isolated individually in such cages and allowed to oviposit before dissection. The entire experiment was replicated three times. Total fecundity values, i.e., total numbers of mature eggs developed by each female from one blood meal, were derived from both pooled and isolated females, including both oviposited and retained eggs in the latter case. Differences among treatment groups in the proportions of females producing at least some eggs were analyzed by a G-test for independence. Effects of three factors (crop status, energy reserve status, and experimental replicate) on fecundity were tested by threeway analysis of variance. Differences among individual and combined treatments were tested by Student- Newman-Keuls multiple comparisons. Statistical procedures followed those of Sokal and Rohlf (1969). Meal weights Dry weights of sucrose and blood meals were calculated as the difference between mean dry weights of samples of fed and unfed mosquitoes, weighed individually, based on single replicates. Mosquitoes were prepared exactly as described for fecundity experiments, and replete females were killed with chloroform min after withdrawing their mouthparts, both after the 50% sucrose meal and after the blood meal taken 6 h later. Their unfed counterparts were killed at the end of the same feeding session, which lasted up to 0.5 h for each treatment group. Killed mosquitoes were dried to a constant weight (3 d) at 60 o C, then weighed individually to the nearest 0.01 mg. The unfed controls for the sucrose-fed group were not offered a sucrose meal; the unfed controls for the blood-fed group were presented with the human hand but were prevented from blood feeding by being collected with an aspirator as fast as they landed on the hand. Because unfed individuals refusing to feed were treated as still another group, this procedure resulted in nine categories, each, of high- and low-reserve mosquitoes, or 18 weight categories in all. The difference in body weights of unfed mosquitoes in the high-reserve and low-reserve categories served as a general confirmation of the differences in their somatic energy reserves at the time of feeding of their fed counterparts, though baseline irreducible dry weights were not determined. Weights of those refusing to feed were treated separately. Differences in meal weight among treatment groups were analyzed by comparing the populations of weights in the fed and unfed groups in one-contrast and six-contrast simultaneous comparisons procedures with t-tests (Brown and Hollander 1977) and analysis of variance (Scheffé 1959), using log-transformed data (sucrose meals) and squareroot-transformed data (blood meals). To check for inherent bias in body weight of mosquitoes choosing to feed on sucrose, a t-test was used to compare weights of mosquitoes refusing a meal with those of controls not offered one. RESULTS Fecundity Mosquitoes with high energy reserves produced significantly more eggs in their first gonotrophic cycle than those with low energy reserves (F = 38.28; df = 1), and mosquitoes with empty crops produced significantly more eggs than those with full crops (F = 47.88; df = 1) (both P<0.001). A multiple comparison of the four treatment groups that detected significant differences at the 0.05 level of a type-i error confirmed that the emptycrop/high-reserve combination provided the highest

4 June, 2004 Journal of Vector Ecology 87 Table 1. Subsequent egg production of females differing in crop condition and energy reserves at time of human blood meal. Mean number of eggs per female and sample size (n)* Replicate Crop Energy condition** reserves I II III Totals ± SEM Full Low 44.0 (22) 70.9 (30) 50.5 (30) a 56.2 ± 2.6 (82) High 74.8 (29) 75.7 (29) 75.2 (30) b 75.2 ± 2.1 (88) Empty Low 67.6 (19) 78.7 (39) 79.6 (40) b 76.9 ± 2.5 (98) High 82.6 (39) 86.2 (40) 84.9 (37) c 84.6 ± 1.6 (116) * Samples include females not developing any eggs (<1% of females with high reserves, 4% of females with low reserves). ** Full, 50% sucrose meal 6 h before blood meal. Empty, sucrose not available for previous 2-3 d (high energy reserves) or previous d (low energy reserves). Means not preceded by same letter are significantly different (P<0.05, SNK multiple-comparisons test). Table 2. Meal size of females differing in crop condition and energy reserves, based on dry weights of individuals taking or not taking the meal. Mean dry body weight (mg) ± SEM and sample size (n) Previous Mean meal weight crop Energy (body weight Meal type condition reserves Meal No meal* difference) ± SEM** 50% sucrose Empty Low 1.23 ± 0.04 (30) 0.59 ± 0.01 (28) 0.64 ± 0.04 High 1.21 ± 0.03 (42) 0.87 ± 0.02 (40) 0.34 ± 0.03 Blood Full Low 1.86 ± 0.05 (29) 1.24 ± 0.04 (27) 0.62 ± 0.07 High 1.95 ± 0.04 (42) 1.13 ± 0.03 (47) 0.82 ± 0.04 Blood Empty Low 1.80 ± 0.06 (22) 0.61 ± 0.01 (40) 1.19 ± 0.05 High 1.82 ± 0.02 (90) 0.83 ± 0.02 (58) 0.99 ± 0.03 * Sucrose meal not available; blood meal attempted but prevented. **Sucrose meal weights significantly different (P<0.001). All blood meal weights significantly different in onecontrast comparison (P<0.005) and also in six-contrast comparison (P<0.005), except in full-low vs. full-high contrast (P>0.1).

5 88 Journal of Vector Ecology June, 2004 fecundity (mean = 85), and the full-crop/low-reserve combination provided the lowest fecundity (mean =56) (Table 1). The opposite pair of combinations (full/high and empty/low) provided egg batches that were intermediate size (mean = 75, 77), significantly different from the other two but not from each other. A significant interaction occurred between crop status and energyreserve status (F = 6.63; df = 1; P<0.02), with crop status having a greater effect on fecundity when energy reserves were low, and reserves having a greater effect when the crop was full. Replicates differed significantly (F = 6.41; df = 2; P<0.002), but the relative mean fecundities of the four treatment groups were about the same within each replicate (Table 1). Medians (and lower and upper quartiles) of the egg-clutch sizes were as follows: full/ low, 61 (38-75); full/high, 79 ( ); empty/low, 79.5 (64-91); empty/high, 82.5 (75-94). The two intermediate groups, despite similar means and medians, differed strongly in skewness and kurtosis. The proportion of females failing to produce any eggs was small (2%). Those with low reserves included significantly more infecund females (4%) than those with high reserves (<1%) (G = 5.95; df = 1; P<0.02). The difference between infecund females with full and empty crops was not significant (G = 1.09; df = 1; P>0.2). Meal size Sucrose meals taken by low-reserve females were almost twice as large as those of high-reserve females (F = 53.60; df = 3; P<0.001) (Table 2). This is shown in differences in mean dry body weight of sucrose-fed and non-sucrose-fed females at two points: the time of the sugar meal (low-reserve difference minus high-reserve difference: = 0.30 mg larger) and 6 h later, at the time of blood feeding (low-reserve difference minus high-reserve difference: [ ] - [ ] = [ ] = 0.33 mg larger). Blood meals were significantly larger among emptycrop females than full-crop females (t = 220; df = 351; P<0.001) (Table 2). Within the empty-crop category, females with low reserves took significantly larger blood meals than females with high reserves (six-contrast: F = 3.82; df = 18; P<0.005). In the full-crop category, it was the reverse; low-reserve females tended to take smaller blood meals. This latter difference was significant in a one-contrast comparison (F = 8.49; df = 3; P<0.004) but not in the conservative experimentwise six-contrast comparison (F = 1.42; df = 18; P>0.1 (Table 2). Few mosquitoes (9 of 364, 2%) refused to feed on blood, regardless of crop status or energy-reserve status. This, and the fact that all unfed controls attempted to feed, obviates any bias with regard to energy reserves, and therefore body weight, of blood-fed females. The proportion of low-reserve mosquitoes refusing a 50% sucrose meal was 4 of 34 (12%). These four did not weigh significantly more than the sample not offered sucrose (n = 28)(mean = 0.66 mg and 0.59 mg, respectively; t = 1.69, df = 30, P>0.1), indicating that there was little or no energy-reserve bias in the crop-full treatment group. The number of high-reserve mosquitoes refusing a 50% sucrose meal was 8 of 50 (16%). These eight weighed significantly less (mean = 0.76 mg) than the sample not offered sucrose (n = 40, mean = 0.87 mg) (t = 2.36, df = 46, P<0.03). This counterintuitive difference suggests that the high-reserve group may have been biased toward higher reserves compared to the controls not offered sucrose, resulting in a slight overestimate of sugar-meal size. More likely, the observed bias was a result of random sampling error. DISCUSSION Sugar feeding and fecundity Results demonstrated that both sucrose stored in the crop and sucrose metabolized and stored principally in the fat body had significant effects on Ae. aegypti fecundity during a single gonotrophic cycle. Those effects were opposite and approximately equal, the crop sugar reducing egg production, metabolized energy reserves increasing it. Thus, mosquitoes that contained either a recent full sugar meal or depleted energy reserves produced an intermediate and similar number of eggs after taking a blood meal. Mosquitoes possessing both of these characteristics produced fewest eggs, and those possessing neither of them produced the most eggs. The 50% sucrose meal was available for digestion and conversion to reserves 6 h before blood feeding. Even up to 12 h after a small blood meal, sugar feeding can promote the decision to develop eggs (Klowden and Chambers 1989), though crop emptying is reduced substantially during the digestion of a replete blood meal (Miles 1977). Nevertheless, any sugar digestion that may have occurred did not offset the apparent effect of smaller blood meal size in lowering the fecundity of low-reserve females. The mean rate of conversion of blood into eggs was about the same in low/full females (90.3 eggs/mg blood) as in the high/full females (91.5 eggs/mg) and high/empty females (85.9 eggs/mg), whereas in low/ empty females it was considerably lower (64.7 eggs/mg). These values indicate that enough energy from the 50% sucrose meal was available by the time of egg commitment so that fecundity was blood-meal-limited in the low/full females and was energy-reserve-limited in the low/empty females.

6 June, 2004 Journal of Vector Ecology 89 Emergence 10% SUCROSE Mating 5 days High-Reserve Prep. Low-Reserve Prep. 10% SUCROSE 6-7 days 6-7 days WATER ONLY WATER ONLY 2-3 days 2-3 days WATER ONLY >50% dead 50% SUCROSE Full-Crop WATER ONLY Empty-Crop 50% SUCROSE Full-Crop WATER ONLY Empty-Crop 6 hr 6 hr 6 hr 6 hr BLOOD BLOOD 3 days 3 days BLOOD BLOOD 3 days 3 days EGGS EGGS EGGS EGGS Full/High Group Empty/High Group Full/Low Group Empty/Low Group Figure 1. Preparation of the four experimental groups of female Aedes aegypti, from emergence until egg maturation. The times for killing groups and sub-groups of mosquitoes that were unfed, fed, or refused-to-feed, which were used to determine meal size and body weight bias, correspond to the feeding times, as described in the text.

7 90 Journal of Vector Ecology June, 2004 Digested Sugar + Energy Reserve Size Sugar-Meal Size Blood-Meal Size + + Egg Number Figure 2. Summary of direct and indirect effects of sugar feeding on meal size and the number of eggs matured. Promoting effects are indicated by a +, retarding effects by a. Energy reserves and meal size The physiological basis for the observation that high reserves depress the size of both sugar meals and blood meals is not clear. Conceivably, fat body volume might place a physical limit on meal size, because fat body, crop, and midgut all occupy abdominal space. Alternatively, energy-reserve status may set a behavioral threshold for the volume of food imbibed. The ultimate consequences for mosquito survival and short-term fecundity, however, are more obvious. A female with more reserves can afford to take less sugar, if not avoid it altogether, when blood sources are available. It thereby retains its ability to ingest adequate blood meals, promotes its agility while obtaining supplementary blood meals, and enhances the range of its foraging and ovipositing activities. It can afford to take less blood per meal because blood in excess of the amount necessary to develop the maximum number of eggs possible (Woke et al. 1956) may be superfluous in a high-reserve female. Crop status and blood meal size In each energy-reserve category, full-crop females took smaller blood meals than their empty-crop counterparts. The seemingly anomalous result within the full-crop category, that high-reserve females took more blood than low-reserve females (the opposite of the case in crop-empty females), is explained by the smaller sucrose meals taken by high-reserve females 6 h previously. In low-reserve females, whose sugar meals were almost twice that size, crop emptying would be expected to occur at a greater rate (Miles 1977), perhaps lessening the difference in crop volume between them and their high-reserve counterparts by the time of the blood meal 6 h later. Thus, the blood meals of low-reserve females might have been even smaller, had the blood been offered sooner after sucrose was ingested. The smaller blood meals taken by females with full crops may be explained by the same alternative mechanisms proposed for mediating the effects of energy reserves. It is clear in this case that the maximum amount of blood taken was not commensurate with maximum reproductive benefit. The amount of blood ingested is positively correlated with fecundity (Briegel 1985) unless the blood meal volume is above the level needed to produce the maximum number of eggs (Woke et al. 1956). That level was not reached in full-crop mosquitoes, judging from the fact that they produced fewer eggs than their empty-crop counterparts. Thus,

8 June, 2004 Journal of Vector Ecology 91 more blood would have increased fecundity, suggesting that a physical limit on meal size was reached in lowreserve, full-crop females. The sucrose meal weights measured in this study were subject to intrinsic bias, because high-reserve females refusing sucrose were significantly lighter than an unselected sample of females not offered sucrose. However, the proportion refusing to feed was small, so their exclusion from the group taking sucrose probably had a negligible effect. In any case, that effect would be to overestimate sugar meal size in high-reserve females, thus underestimating the influence of energy reserves in reducing sugar meal size. When Ae. aegypti females contain substantial energy reserves, they imbibe less sugar in a single meal. When they contain less crop sugar or contain minimal energy reserves, they imbibe more blood in a single meal. Both substantial energy reserves and large blood meals increase the size of the egg clutch. Thus, the energy reserves derived from sugar meals (and presumably also reserves carried over from larval feeding and from previous blood meals) act on fecundity in three different ways, two of them indirect (Figure 2). One consequence of this multi-level action of energy reserves is that the physiological state causing females to ingest the largest blood meals does not result in the maximum egg-clutch size. Caution should be observed in extrapolating from the present results to the effect of energy reserves on reproduction of females of different ages and body sizes, or in subsequent gonotrophic cycles. For example, females that are older before taking their first blood meal ingest less blood, but those taking repeated meals as they age ingest larger blood meals (Klowden and Lea 1980). Furthermore, it appears that less lipid is allocated to egg production as a female ages, resulting in smaller clutches of eggs even when sugar feeding is allowed (Briegel et al. 2002). According to the interactions observed in this study, highest fecundities per gonotrophic cycle in nature would be achieved by mosquitoes that can feed extensively on sugar but digest it before taking a primary blood meal. Therefore, one might expect female Ae. aegypti to organize their activities as follows: a) sugar feeding is particularly likely early in adult life; b) within the diel cycle sugar seeking most often occurs long before or immediately after blood seeking activities, c) recently ingested sugar meals inhibit the response toward bloodhost stimuli. Early post-emergence sugar feeding is supported by one field study (Nayar 1981) but not another (Edman et al. 1992); a separation between sugarseeking and blood-seeking diel activity periods has not been detected in the laboratory (Yee and Foster 1992); but an inhibited host response soon after sugar feeding is well-documented (Jones and Madhukar 1976). However, the innate rules that govern a mosquito s feeding decisions are shaped not just by physiological constraints but also by the immediate and long-term probabilities of obtaining sugar meals and blood meals in a particular environment (Roitberg and Friend 1992). For example, the difference in likelihood of encountering sugar sources, relative to blood sources, in domestic vs. woodland habitats best explains the conflicting evidence for early and frequent sugar feeding by Ae. aegypti in the field. Behavioral rules may also be shaped by the likelihood of obtaining a replete sugar meal of such a high concentration or of accumulating energy reserves as high as those tested in this study (Day and Van Handel 1986). If these extremes seldom occur in nature, their effects on long-term fecundity will be minor in a species that may gain greatest fitness by taking blood whenever the opportunity arises. In many mosquito species, this last consideration will apply only if blood hosts are found infrequently. In the case of domestic Ae. aegypti, it is applicable even when hosts are common, because its primary blood meals from humans supply such a large part of the females energy needs, and its supplementary blood meals are practically the equivalent of sugar meals. This is consistent with the observation that in the presence of blood-host odors, even starving Ae. aegypti females completely ignore sugar-related volatiles (Yee and Foster 1992), whereas most species do not. One important repercussion of this particular nutritional strategy is that blood-feeding frequency will be exceptionally high in domestic environments, which promotes the mosquito s vectorial capacity (Edman et al. 1992). Acknowledgments We thank S. Juang, N. Reichenbach, and G.C. Dobrin for statistical advice and analysis. H.J. Harlan, D.M. Bernstein, R.W. Hall, and J. L. Martin gave vital technical assistance in the laboratory during various phases of this work. P.L. Phelan and R.M. Taylor generously provided critical reviews of the manuscript. REFERENCES CITED Briegel, H Mosquito reproduction: incomplete utilization of the blood meal protein for oogenesis. J. Insect Physiol. 31: Briegel, H Metabolic relationship between female body size, reserves and fecundity of Aedes aegypti. J. Insect Physiol. 36: Briegel, H., M. Hefti, and E. DiMarco Lipid metabolism during sequential gonotrophic cycles in large and small female Aedes aegypti. J. Insect

9 92 Journal of Vector Ecology June, 2004 Physiol. 48: Brown, B.W. Jr. and M. Hollander Statistics: a biomedical introduction. Wiley, New York, NY. Clements, A.N The sources of energy for flight in mosquitoes. J. Exp. Biol. 32: Colless, D.H. and W.T. Chellapah Effects of body weight and size of blood-meal upon egg production in Aedes aegypti (Linnaeus) (Diptera: Culicidae). Ann. Trop. Med. Parasitol. 54: Costero, A., G.M. Attardo, T.W. Scott, and J.D. Edman An experimental study on the detection of fructose in Aedes aegypti. J. Am. Mosq. Contr. Assoc. 14: Day, J.F. and E.Van Handel Differences between the nutritional reserves of laboratory-maintained and field-collected adult mosquitoes. J. Am. Mosq. Contr. Assoc. 2: Edman, J.D., D. Strickman, P. Kittayapong, and T.W. Scott Female Aedes aegypti (Diptera: Culicidae) in Thailand rarely feed on sugar. J. Med. Entomol. 29: Foster, W.A Mosquito sugar feeding and reproductive energetics. Annu. Rev. Entomol. 40: Haramis, L.D. and W.A. Foster Survival and population density of Aedes triseriatus (Diptera: Culicidae) in a woodlot in central Ohio, USA. J. Med. Entomol. 20: Harrington, L.C., J.D. Edman, and T.W. Scott Why do female Aedes aegypti feed preferentially and frequently on human blood? J. Med. Entomol. 38: Jalil, M Observations on the fecundity of Aedes triseriatus (Diptera: Culicidae). Entomol. Exp. and Appl. 17: Jones, J.C. and B.V. Madhukar Effects of sucrose on blood avidity in mosquitoes. J. Insect Physiol. 22: Klowden, M.J Effect of sugar deprivation on the host-seeking behaviour of gravid Aedes aegypti mosquitoes. J. Insect Physiol. 32: Klowden, M.J. and G.M. Chambers Ovarian development and adult mortality in Aedes aegypti treated with sucrose, juvenile hormone and methoprene. J. Insect Physiol. 35: Klowden, M.J. and A.O. Lea Blood meal size as a factor affecting continued host-seeking by Aedes aegypti (L.). Am. J. Trop. Med. Hyg. 27: Klowden, M.J. and A.O. Lea Physiologically old mosquitoes are not necessarily old physiologically. Am J. Trop. Med. Hyg. 29: Lachmajer, J. and W. Antonowicz Experimental blood feeding of mosquito females of some Anopheles, Aedes and Culex species: taking of blood meal by starving and glucose-fed insects. Acta Parasitol. Pol. 28: Macfie, J.W.S Observations on the bionomics of Stegomyia fasciata. Bull. Entomol. Res. 6: Martinez-Ibarra, J.A., M.H. Rodriguez, J.I. Arredondo- Jimenez, and B. Yuval Influence of plant abundance on nectar feeding by Aedes aegypti (Diptera: Culcidae) in southern Mexico. J. Med. Entomol. 34: Miles, C.T The effect of blood, nutritional reserves, and age on the rate of crop sugar depletion in the mosquito Aedes aegypti (L.). M.S. Thesis, Ohio State Univ., Columbus. Morrison, A.C., A. Costero, J.D. Edman, G.G. Clark, and T.W. Scott Increased fecundity of Aedes aegypti fed human blood before release in a markrecapture study in Puerto Rico. J. Am. Mosq. Contr. Assoc. 15: Nayar, J.K Aedes aegypti (L.) (Diptera: Culicidae): observations on dispersal, survival, insemination, ovarian development and oviposition characteristics of a Florida population. J. Fla. Anti- Mosq. Assoc. 52: Nayar, J.K. and D.M. Sauerman, Jr Long-term regulation of sucrose intake by the female mosquito, Aedes taeniorhynchus. J. Insect Physiol. 20: Nayar, J.K. and D.M. Sauerman, Jr The effects of nutrition on survival and fecundity in Florida mosquitoes. Part III. Utilization of blood and sugar for fecundity. J. Med. Entomol. 12: Roitberg, B.D. and W.G. Friend A general theory for host seeking decisions in mosquitoes. Bull. Math. Biol. 54: Scheffé, H The analysis of variance. Wiley, New York, NY. Sokal, R.R. and F.J. Rohlf Biometry. Freeman, San Francisco, CA. Van Handel, E The obese mosquito. J. Physiol. 181: Van Handel, E. 1985a. Rapid determination of glycogen and sugars in mosquitoes. J. Am. Mosq. Contr. Assoc. 1: Van Handel, E. 1985b. Rapid determination of total lipids in mosquitoes. J. Am. Mosq. Contr. Assoc. 1: Van Handel, E., J.D. Edman, J.F. Day, T.W. Scott, G.G. Clark, P. Reiter, and H.C. Lynn Plant-sugar, glycogen, and lipid assay of Aedes aegypti collected in urban Puerto Rico and rural Florida. J. Am. Mosq. Contr. Assoc. 10:

10 June, 2004 Journal of Vector Ecology 93 Woke, P.A., M.S. Ally, and C.R. Rosenberger The number of eggs developed related to the quantities of human blood ingested in Aedes aegypti (L.). Ann. Entomol. Soc. Am. 49: Yee, W.L. and W.A. Foster Diel sugar-feeding and host-seeking rhythms in mosquitoes under laboratory conditions. J. Med. Entomol. 29: Ziegler, R. and M.M. Ibrahim Formation of lipid reserves in fat body and eggs of the yellow fever mosquito, Aedes aegypti. J. Insect Physiol. 47: