Age modifies the effect of body size on fecundity in Culex quinquefasciatus Say (Diptera: Culicidae)

Fecundity of mosquitoes can vary with many factors and can have a strong effect on population growth. This study reports the effects of body size, blood meal size, and age on the reproductive output of nulliparous Culex quinquefasciatus, a vector of arboviruses and other pathogens. Mated adult female mosquitoes from a colony were reared under standard conditions and fed on chickens at different ages post-eclosion. Blood meal size and wing length were recorded, as well as the number of eggs in the first-cycle egg raft. Each of these factors had a significant influence on fecundity considered in a simple regression context. Multiple regression analysis revealed a significant interaction effect between age and body size on fecundity. Up to 13 days of age, fecundity was positively correlated with body size, but in mosquitoes older than 13 days, this relationship was not significantly different from zero. These results are discussed in terms of the known physiology of this and other species. Culex quinquefasciatus Say is a cosmopolitan mosquito species found in tropical, subtropical, and warm temperate regions. It is a noted vector of vertebrate pathogens, including filarial worms (Ahid et al. 2000, Lima et al. 2003), protozoan parasites (van Riper III et al. 1986), and various arboviruses (Meyer et al. 1983, Sardelis et al. 2001). Due to its status as a vector, much attention has been paid to the bionomics of this species. Many factors, at the levels of landscapes, microhabitats, and the physiology of individual mosquitoes, are known to affect the growth and maintenance of populations of Cx. quinquefasciatus. Less is known about interactions between nutritional and age factors and fecundity. This study focuses on some of the important predictors of reproductive output at the level of adult physiology, as the ultimate determinant of population success is the success of individuals (Briegel 2003). Like all anautogenous mosquitoes, Cx. quinquefasciatus depends on a blood meal for the necessary nutrients (protein) to produce eggs. There have been many studies detailing the relationships between blood feeding and reproductive output (Akoh et al. 1992, Hogg et al. 1996, Roitberg and Gordon 2005). Studies on Cx. quinquefasciatus have demonstrated a positive relationship between blood meal size and fecundity (Akoh et al. 1992, Lima et al. 2003). This indicates that the size of the blood meal, in part, determines the number of eggs that can be laid. Other factors are known to influence mosquito fecundity, including body size and teneral reserves of the adult females (Briegel 2003). A portion of the protein reserves accumulated during the larval period can be used to provision eggs, such that better larval conditions usually result in female mosquitoes capable of greater reproductive output. Body size is usually positively correlated with larval food resources and thus serves as an indicator of both larval habitat quality and teneral reserves (Briegel 2003, Telang and Wells 2004). Many studies have shown a positive relationship between female mosquito body size and egg production (Akoh et al. 1992, Armbruster and Hutchinson 2002, Briegel 1990a, Briegel 1990b, Lima et al. 2003, Lyimo and Takken 1993). This implies that the potential reproduction of a female mosquito is determined in part by the larval habitat. Other members of this species complex, such as Culex pipiens molestus Forskal (Vinogradova 2000), are autogenous and derive the entire protein input for egg production in the first clutch from larval nutrition. Thus, it is likely that a portion of the protein required for egg production in the anautogneous members of this complex, such as Cx. quinquefasciatus, is also derived from nutrition in the larval stage. Age lowers reproductive output in many animal taxa. In animals with high rates of daily mortality such as Culex mosquitoes (Dow 1971, Elizondo-Quiroga et al. 2006), selection is likely to have favored high early-life reproduction, at the expense of late-life fitness and reproductive capacity. A number of studies have examined declines in reproduction with age in mosquitoes, and several of these (Akoh et al. 1992, Gomez et al. 1977, Suleman 1979, Walter and Hacker 1974) have detailed such declines in various strains of Cx. quinquefasciatus. Most of the studies have failed to examine the role of other physiological parameters such as body and blood meal size, or have done so in a manner that did not allow for multivariate models to partition variance. Because of this, it is unclear what the relative contributions of each of these factors to fecundity in an aging population truly are. The implication of this decline in reproductive capacity with age for the population growth rate is unclear, but it may at times be severe (Charlesworth 2000), especially if populations are low and aged, such as during conditions of drought (Shaman et al. 2003). Ultimately, continued transmission of diseases across seasons may depend on a continuous high population of competent vector mosquitoes, and as such, the reproductive capacity of small aged populations is of interest. Consideration of other factors such as blood meal size and body size may improve the robustness and predictive qualities of models of age-dependant fecundity. This study was designed to determine the contributions of age, body size, and blood meal size to the fecundity of Cx. quinquefasciatus, considered together in a multiple regression context. Larvae were reared in enameled metal pans measuring 24 × 36 × 5 cm containing approximately 700 ml of tap water. Each pan was set with three egg rafts of colonized Cx. quinquefasciatus, from the USDA ARS Gainesville, FL colony, established in 1995 from wild-caught mosquitoes from Gainesville, FL (Allan et al. 2006). Food was provided daily to each pan as 20 ml of a slurry containing 20 mg/ml 1:1 Brewer's yeast/liver powder in water. This somewhat food-limiting rearing regimen was chosen to generate some variation in size while still achieving relatively simultaneous adult emergence. Pupae were placed in 500 ml cups in a large cage measuring 57 × 57 × 57 cm and sugar was provided to the emerging adults as 10% sucrose solution on cotton wicks. Adults were allowed to emerge for 12 h following the emergence of the first female, whereupon the cups were removed. The final density of mosquitoes in the cage was estimated to be about 700 females and 700 males. Mean temperature was 26.8 ±0.8° C for the duration of the experiment. Relative humidity was greater than 90% and the light cycle was 16:8 L:D. Beginning at five days post-eclosion, approximately 36 host-seeking females were captured using a small cylindrical trap with a funnel, placed with the funnel end inside the cage, and introduced into the screened end to attract host-seeking female mosquitoes. These mosquitoes were removed from the cage and blood fed to repletion on a restrained chicken. Animal use followed a University of Florida Institutional Animal Care and Use Committee approved protocol (D509). This was repeated at four-day intervals until six groups were obtained. The age range thus produced was five to 25 days post-eclosion. We assumed all females were mated, as observations suggest mating occurs during the first five days, and the male:female ratio was approximately 50:50. Mating was further confirmed by assessing viability of egg rafts laid. We used the spectrophotometric method (Briegel 1980) of quantifying hematin content in feces as a measure of blood meal size. Immediately after blood feeding, individual mosquitoes were placed in separate 40 ml vials (2.5 cm diameter, 9.5 cm deep) covered with screen. Ten percent sucrose was provided on small cotton balls. Following a four-day period for egg maturation, females were transferred to different, individual vials (see below) for oviposition. The time required for digestion and egg development for this species has been found to be two to three days at a temperature of 31.5-34° C (Elizondo-Quiroga et al. 2006), and hence four days was found to be a good balance between maximal survival in the vials and maximal ovarian development. Fecal material in the egg maturation vials was rinsed from the vial with 2.0 ml 1% LiCO3. The resulting solutions were decanted into spectrophotometric cuvettes and the absorbance at 387 nm was read using a spectrophotometer. Absorbance readings were converted to micrograms of hematin using a standard curve previously prepared. After egg maturation, gravid females were transferred into a second set of 40 ml vials containing 4.0 ml of 10% (by volume) hay infusion in tap water for oviposition. The following morning, egg rafts deposited were removed, placed on water under a microscope, and photographed at high magnification with a digital camera. The number of eggs in the raft was counted on the photograph. Eggs were then allowed to hatch to confirm female mating status. After hatching, larvae were filtered out of the water and counted to obtain a count of percentage hatch. Following oviposition, females were removed, killed, and their wings excised and mounted on slides for measurement. Abdomens were dissected to count retained eggs. The wings on the microscope slides were photographed with a standard size reference (a 7.752 mm length of steel measured to the thousandth of a millimeter with a caliper). The photographs were opened in SigmaScan Pro 5 (Systat Software, Inc., Point Richmond, CA), calibrated for size, and measured from the alular notch to the distal end of R2, excluding fringe hairs (Packer and Corbet 1989). The distal end of R2 was used for measurements, as it is an unambiguous standard landmark. Regression analyses were conducted using S-Plus© 7.0 for Windows© (Insightful 2005). Analysis of covariance was conducted using SAS 9.0 (SAS Institute 2001). A one-way analysis of variance testing for differences in wing length with age was performed to determine if the size of the females that oviposited significantly varied with age, as might be expected if there were an effect of size on mortality rate. Size was measured only for females that survived the egg development period and oviposited. Simple linear regressions of each predictor on fecundity were performed to compare with published results of these predictors on fecundity for Cx. quinquefasciatus females. A multiple linear regression was performed on untransformed variables regressing fecundity on body size, blood meal size, and age. Non-significant main effects and interactions were discarded in a step-wise fashion. Normality was verified with a Kolmogorov-Smirnov Test of composite normality and homoscedasticity with a plot of residuals vs fitted values. The same analyses were conducted using just the eggs laid as a response, but the results were almost identical so are not considered here. A multiple regression of fecundity on the standardized values (Z-scores) of the predictors was also performed. Z-scores were calculated by subtracting the mean value of each regressor from each observation, then dividing this by the standard deviation of the regressor (regressors were age, hematin voided, and winglength. This produces mean values of zero and standard deviations of one. When used in multiple regression analysis, it allows a more standard interpretation of slope values, i.e., it apportions mean changes in response due to predictor variations of one standard deviation (Marquardt 1980). Doing so allows one to order the predictors in terms of influence on a common scale. Significant interactions with age were explored with an ANCOVA testing for differences in slope of hematin voided or wing length on fecundity associated with the different age groups (Huitema 1980). Fecundity varied from 43 to 231 eggs. Blood meal size and wing length also varied (Table 1), and there was no significant effect of age on the wing length of females tested (One-way ANOVA, F=1.46, df=1,128, p=0.23). This indicates there was no effect of wing length on survivorship that could be detected under these conditions; however, the low sample sizes at older ages may have limited our ability to observe these effects. The number of females in each age group ovipositing was not constant, but varied depending on survival through the egg development and oviposition period, with a maximum of 32 and a minimum of four (Table 1). All egg rafts laid by the females in this experiment produced some viable offspring, with a mean of 81.5 (+/− 18.6), a minimum of 1.2, and a maximum of 100% hatch. The simple linear regressions of age, hematin, and wing length predicting fecundity were significant, with R2 ranging from 0.05 to 0.52 (Table 2). The multiple regressions were also significant, although only the age*winglength interaction remained in the model after stepwise elimination (Table 2, 1-3). Scatterplot of fecundity vs unstandardized age at blood feeding for Cx. quinquefasciatus. Scatterplot of fecundity vs unstandardized blood meal size for Cx. quinquefasciatus. Scatterplot of unstandardized wing length vs fecundity showing the response between ages 5–13 days (solid line, circles), 17–25 days (dashed line, triangles), and the response for all ages combined (dotted line). Blood meal size was significantly correlated with body size (Table 3), but the R2 of this regression was low. In this experiment there was low variation in body size, which likely contributed to the low R2. The fit of the full multiple model was considerably better (R2= 0.73) than that of any of the one-factor linear regressions (Table 2). All individual parameters influenced fecundity with directions predicted by previous work, except for the unstandardized multiple model (Table 2) where the coefficient for age was positive rather than negative (see Discussion). The standardized multiple model indicates the expected direction of influence of these parameters on fecundity (hematin positive, wing length positive, age negative). A significant interaction between age and body size was found in the multiple regression models predicting fecundity. Analysis of this interaction by ANCOVA (slopes for wing length at varying levels of age) showed that in the younger age classes, slopes were positive and significant (Table 4), but after 17 days of age, slopes were not significantly different from zero. For illustration, significant slopes were grouped, and slopes not significantly different from zero were grouped (Figure 3). As expected, both blood meal size and body size had a positive influence on fecundity. The degree of dependence on these two factors considered alone is relatively low, but taken together with age, a more precise model explaining most of the variance in fecundity was achieved. Age was the factor that explained the greatest variance in fecundity, either as a single factor in the simple linear regressions, or as a factor in the standardized multiple regression model. The decline in fecundity with age had the greatest influence on fecundity in the standardized multiple model, indicating that a standard deviation in age has a greater effect than a standard deviation in blood meal size or body size. That there is a noticeable decline in fecundity with age over a span of 25 days post-eclosion accords with one of the expectations of the antagonistic pleiotropy hypothesis of aging, namely that age-associated fitness effects are expected to be large in taxa experiencing high per-diem mortality (Williams 1957). In the standardized model, each predictor accords with the expected direction of influence on fecundity: i.e., negative for age, positive for blood meal size and body size. The unstandardized multiple model has values for the regression coefficients that seem counterintuitive. The unstandardized form seems to qualitatively differ with respect to the expected sign (+/−) for age, because this model considers the relative contributions of each factor where each is held at zero. In the dataset, no factor ever equals zero, and as such, these regression coefficients are arbitrary descriptions of effects outside the range of observed values. Compounding this, the variance in the model attributed to the main factors can be taken up by their interactions when these are not standardized. Converting the values to their standardized forms helps avoid this confusion (Marquardt 1980). Because the predictors are considered at equivalent intervals (of one standard deviation), one can better evaluate the true relative contributions of each to the overall model. Our results highlight the importance of using the appropriate regression methods, as different conclusions would be drawn if only the unstandardized model was used. On the whole, age had the strongest influence on fecundity, followed by blood meal size and body size. It should be noted that the range of body sizes produced was not great, and this may have contributed to the low contribution of this factor to the models. The estimate for the decline in fecundity associated with age in the unstandardized simple linear regression of age on fecundity is similar to that reported for a wild Vero Beach strain of this species (Walter and Hacker 1974), but greater than those reported for Asian populations of this species (Suleman 1979, Walter and Hacker 1974). It is possible that the effect of age on fecundity may also be modified by temperature and humidity, so direct comparisons between the aforementioned studies are difficult to make, as these other studies held the insects under different conditions. The effect of age-related declines in fecundity are only one impact of age on population growth in an age-structured population. Mosquitoes have high rates of daily mortality (Dow 1971, Reisen et al. 1991), and mortality has been shown to be age-dependent in some species (Clements and Paterson 1981, Harrington et al. 2008). Both age-related mortality and declines in fecundity may contribute to the population dynamics of mosquitoes and should be incorporated into models of mosquito populations when possible. The interaction between adult body size and age accounted for a portion of the variance and increased the R2 of the model moderately. It also led to a more careful consideration of the effect of body size across the different age groups. The effect of body size on fecundity was not constant: in the younger groups, it was significant and positive, whereas in the older groups it had no significant influence. A possible explanation would be that mobilization of teneral protein or lipid reserves for egg production was interrupted as a consequence of age, perhaps due to diversion of resources into somatic repair. If most of these reserves have disappeared, then the only significant factor promoting increased fecundity would be blood meal size. Briegel demonstrated the somewhat surprising fact that larger Aedes aegypti females invested progressively less lipid into their yolk mass with increasing age, to sucrose from which to (Briegel et al. This was not in females. In in both groups, lipid content of the mosquitoes increased with age. with this was an that the negative slope of the fecundity vs age curve was for larger females than it was for indicating that to their body size, larger females to produce eggs in than in If such a also in Cx. quinquefasciatus, then this could for an interaction between age and body size predicting fecundity. possible explanation for our results might be a decline in protein reserves about by the of the females to blood feeding et al. If protein reserves are then positively influence fecundity. If also that there is no or low between body size and blood meal size, then any positive effect of body size on fecundity would when teneral reserves are In this only of the variance in blood meal size can be to differences in body size. Because the between body size and blood meal size is low, there is no to larger in the older age groups as a result of greater capacity to large blood This is the first study to an interaction between adult mosquito body size and age that predicted fecundity has been and it should further into the of interactions with age in predictive models of mosquito fecundity. consideration should be to using a of mosquito as well as perhaps greater of age groups. This may a better of the of any interaction effects. Other should also be performed testing this of interaction between body size and age is in females multiple blood We the of and for their in and on this and for use of the spectrophotometer. This was by

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.