SEARCH

SEARCH BY CITATION

Keywords:

  • Culex quinquefasciatus;
  • body size;
  • narrow-sense heritability;
  • genotype-by-environment interaction

ABSTRACT:

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES CITED

Body size is an important trait involved in overall fitness through its effects on mating success, fecundity, resource acquisition and mortality, and desiccation resistance. In this study, we raised inbred Culex quinquefasciatus mosquito cohorts at different developmental temperatures of 20°, 23°, and 27° C. As an indicator of the amount of genetic variation in body size, we estimated the narrow-sense heritability of body sizes defined as wing aspect ratios. Our results show that narrow-sense heritability of the body size increased as the developmental temperature increased. We also detected the presence of strong genotype-by-environment (G × E) interaction from low cross-environmental correlations. The body size of each temperature regime followed the general rule that higher temperatures produce smaller individuals. We suggest that the increase in genetic variation with increasing temperature might be due to an unleashing of the cryptic genetic variation of the putative genes affecting body size. We conclude that this increase in genetic variation tracking the environmental (developmental temperature) change could have considerable implications for the distribution and range expansion of Cx. quinquefasciatus, especially in warmer environments.


INTRODUCTION

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES CITED

Members of the Culex (Culex) pipiens L. complex are among the best-studied mosquitoes because of their worldwide distribution, close association with humans, and importance as nuisances or vectors of pathogens of human and animal diseases. There are extensive behavioral, physiological, and morphological variations and intricate interfertility among local populations of the forms in this complex. Culex quinquefasciatus Say, a vector of avian malaria (Goff and van Riper III 1980), West Nile virus (Reisen et al. 2004), bancroftian filariasis, Chikungunya, and St. Louis encephalitis (Marquardt et al. 2005), is widely distributed in the tropical and subtropical areas of the world. It occurs in all climatic zones, ranging from forest to semi-desert.

Variation in growth rates in insects may often be adaptive (Arendt 1997). Body size varies continuously because of the effects of natural selection on the size-dependency of resource acquisition and mortality rates (Chown and Gaston 2010). In addition, temperature is a critical factor for insects, directly affecting lifespan, mortality, and development rates, which can govern phenotypic alterations including changes in morphology (Sibly and Atkinson 1994, Debat et al. 2003, Maharaj 2003).

The fecundities of different species show very different types of dependence on environmental variables. Small individuals may survive and reproduce better when food is limited because they need less food to sustain themselves (Dingle 1992, Blanckenhorn et al. 1994). On the other hand, larger individuals may survive better when there is no food at all, e.g., during hibernation if body size is correlated with nutrient reserves (Ohgushi 1996). As a generally accepted guideline, increased temperature results in higher growth rates, shorter development times, and smaller adult size in insects and other ectotherms (Sibly and Atkinson 1994). The direct effect of temperature on metabolic rates sets limits for growth rates and, since size is typically less flexible, for development times (Nylin and Gotthard 1998). Reaction norm is an important and highly operative concept about the change of genotypic expression through different environments (Schlichting and Pigliucci 1998). It also defines a panel of environmental expression profiles that can be used to assess genotype-by-environment (G × E) interaction (Mackay and Anholt 2007).

Narrow-sense heritability (h2) is a useful and measurable concept to assess the amount of genetic contribution to phenotypic variation in a trait that is influenced by many genes with additive effects (Falconer and Mackay 1996, Lynch and Walsh 1998). It also provides an idea of the capacity to respond to selection when the environment changes. When the amount of genetic variation that is expressed as narrow-sense heritability is relatively low, it is assumed that considerable selection operates on the trait in question (Falconer and Mackay 1996, Lynch and Walsh 1998).

In this study we present our estimations of narrow-sense heritability (h2) of body sizes that were determined as wing aspect ratios in mosquitoes raised at different developmental temperatures and, as the heritabilities were obtained in particular environments, we show them in the perspective of a reaction norm with a focus on the degree of G × E. We also determine the degree of G × E quantitatively from the magnitude of the cross-environmental correlations obtained for the different developmental temperatures in our experiments.

MATERIALS AND METHODS

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES CITED

Maintenance of laboratory colonies

The S. Lab colony used in this study was previously reared in France (ISEM), which originated from California in the 1980s. Egg rafts from the colony were transferred to the Ecological Research Laboratory of Hacettepe University (ESRL) in Ankara, Turkey, in 2005. The rearing and feeding of adults and larvae followed the methods of Kasap and Kasap (1983) with a temperature 27±1° C, 60±5 RH%, and 14:10 h (L:D) photoperiod. Individuals used in the experiments were obtained from the same generation (F30).

Maintenance of cohorts in the climate chambers

The colonization history and techniques to maintain them are as described in Gunay et al. (2010). Individuals were all obtained from the same generation (F30). Three replicates of 750 1st instar larvae were transferred into standard polyethylene 27×16×17 cm cups containing 1 liter of distilled water on the subsequent day of their oviparity. The cups were placed in five climate chambers programmed at five different temperatures (15°, 20°, 23°, 27°, and 30° C) and exposed to a 14:10 (L:D) photoperiod with a constant relative humidity of 60%. Based on their developmental stage, the larvae were fed each day with 0.01–0.1 g sinking TetraminÒ fish food. Pupal development was checked daily and the pupae were counted. After pupation, 100 female and 100 male adults were selected when both sexes were emerging evenly, and then placed in 20×20×20 cm cloth cages to reproduce in each chamber, with plastic cups containing distilled water provided as oviposition sites for each replicate at each temperature. Females were fed with fresh chicken blood every four days for 2 h. Experiments for this study lasted for three generations, (F0, F1, and F2), thus we generated as many fourth generation (F3) adults as possible using the same standard rearing method. All F3 pupae were separated into glass vials before eclosion. From the virgin adults obtained, the following procedure was monitored.

Maintenance of F3 cohorts in climate chambers

One virgin male and five virgin females were transferred to 7×7 cm polyethylene cups with distilled water at the bottom for oviposition and were placed into the rearing chambers. In order to minimize their stress, females were artificially fed with fresh chicken blood using the glass apparatus from Kasap et al. (2003), during their optimum time preferences from midnight to 02:00 every day until at least one of them produced fertilized eggs. Despite all our efforts due to the decrease in the survival rates through generations (Gunay et al. 2010), we did not succeed in producing enough replicates from the extreme temperatures of 15° and 30° C. We recovered 15 replicates for 20°, 23°, and 27° C; 15 virgin males were mated with virgin females and the offspring of just one female were used for wing aspect ratio estimations.

Body size: wing measurements

As an indicator of body size, wing aspect ratios (WAR) were estimated. The left wings of F3 males (fathers, 4th generation) and F4 males (sons, 5th generation) were mounted on slides in Entellane and photographed. Length of the left wing was measured as the distance from the basal of the alula to the apical of the third radius vein (R3), which is shown as a line from 4th to 11th landmarks in Figure 1. In addition, centroid sizes were estimated using 20 landmarks out of those 22 suggested by Aytekin et al. (2009) (Figure 1). The wing aspect ratio was defined as the ratio of the squared wing length to the wing centroid size. Centroid size was preferred to wing area in these estimations to minimize the measurement error (Debat et al. 2003). Only the wing of the males were used in our estimations, as the environmental variance is known to be reduced 20% in males compared to that of females (Reeve and Robertson 1954). The heritability design rested on the male measurements as below. Mean body sizes and coefficients of variation (CV) of fathers and sons developed at respective temperatures were estimated. Coefficient of variation were compared using F ratio statistics, CV2X/CV2Y, for NX-1 and NY-1 (N: sample size) degrees-of-freedom, in which X and Y were the coefficients of two samples (of the fathers and sons developed at different temperature regimes in our case) (Lewontin 1966).

image

Figure 1. Location of the 20 landmarks on the wing of male Cx. quinquefasciatus.

Download figure to PowerPoint

Narrow-sense heritability (h2) and the reaction norm

Narrow-sense heritability was estimated for three temperature regimes (i.e., 20°, 23°, and 27° C) with the offspring-on-parent regression method (Falconer and Mackay 1996). Fifteen single male-female matings were realized for each temperature regime. Among the fourth generation offspring, one adult male per each mating was picked up randomly. Thus, fifteen sons per mating were obtained. Their wing components were measured as defined above and regressed on their fathers. The simple regression equation is Y (sons)=ab+X (fathers), and the regression coefficient (b) can be related to narrow-sense heritability (h2) as b = h2/2, hence h2=2b (Falconer and Mackay 1996). Reaction norms for narrow-sense heritability (h2) were obtained by plotting the respective values of the heritabilities against the different temperature regimes at which cohorts developed.

RESULTS

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES CITED

The sample means and their standard errors, and the coefficient of variations (CV) for wing aspect ratios of the fathers and sons are shown in Table 1. Body size values as measured by wing aspect ratios followed the general rule that the lower the temperature, the greater the size, both in fathers and the sons (Table 1). As for the correlation coefficients (CV), it is evident that the body sizes did not vary much with varying developmental temperatures. We also tested, for both fathers and sons, if there were significant differences in the variability (i.e., CV) of body sizes between pairs of temperature regimes. It seems there is no significant difference between any pair of father or son comparisons, or between father-son comparisons at their developmental temperatures (F ratio values in Table 1, as CV2X/CV2Y). Total phenotypic variability levels thus seem to be very similar across the developmental temperature gradient of our study. However, the most remarkable result for our study context is that heritability shows a distinct pattern of increasing temperature effect. Table 2 shows the offspring- (sons) on-parent (fathers) regression coefficients (b) for body sizes (wing aspect ratios, WAR) and the values of narrow-sense heritabilities (h2) estimated from them. The range of the h2 values lies between 0.152 (20° C) and 0.582 (27° C). The highest value of h2 obtained, i.e., that of 27° C developmental temperature which is almost four times more than that of 20° C, clearly points to a higher temperature – higher heritability effect (Table 2; h2). Figure 2 shows the reaction norms for h2 and CVs of fathers and sons across all developmental temperatures. Figure 2 indicates that although the total (i.e., genetic and environmental components combined) phenotypic variation expressed as coefficient of variation shows little change across the temperatures, the increase of the heritability with increasing temperatures is prominent. Because we have used an inbred strain, genotype-by-environment interaction (G × E) might be revealed from the pattern of the reaction norm in the heritability, taking our strain as genotype (i.e., regarding it an individual sample of the whole genome of the Cx. quinqefasciatus), we performed cross-environment correlation analyses between the phenotypic expressions (i.e., WAR) of body size of the sons raised at different temperatures for the heritability analyses. The idea was that if there were G × E correlations between environmental pairs (i.e., correlations using the WAR values of individuals raised at different temperatures), they would be less than unity (Mackay and Anholt 2007). Table 2 also shows the results of these Pearson product moment correlation estimations (Sokal and Rohlf 1995) under the “Correlation.” All correlation coefficients (r) are less than one, and hence a strong indication of a G × E. While two correlations with 23° C sons give negative values (r: -0.185 and -0.265, for 20–23° C and 23–27° C, respectively), only the correlation between 20° C and 27° C was positive (Table 2).

Table 1.  Wing aspect ratio means (FWAR: Fathers wing aspect ratio, SWAR: Sons wing aspect ratio) and their standard errors, and coefficient of variations as percentages of (CV) fathers and their sons at the respective developmental temperatures. Variability differences as CV test values obtained from F ratio tests are also shown for both fathers and sons, and for father-son pairs (ns: not significant, P>0.05).
 Mean± SECVFathersSonsCVfathers2/CVsons2
FWAR20°C2.261 ± 0.0427.263   
SWAR20°C2.215 ± 0.0356.241CV202/CV232: 1.366 nsCV202/CV232: 0.345 ns20° C: 1.354 ns
FWAR23°C2.220 ± 0.0356.214   
SWAR23°C2.062 ± 0.05610.619CV202/CV272: 1.248 nsCV202/CV272: 0.592 ns23° C: 0.342 ns
FWAR27°C1.998 ± 0.0336.502   
SWAR27°C2.035 ± 0.0428.114CV232/CV272: 0.913 nsCV232/CV272: 1.713 ns27° C: 0.642 ns
Table 2.  Estimations of narrow-sense heritability (h2). b: regression coefficient between fathers and sons, genotype – environment interaction as Correlation (r): cross-environmental correlations between sons of different temperatures; subscripts for r-values signify respective temperatures (WAR: Wing Aspect Ratio). ns: not significant, P>0.05.
 bh2Correlation (r)
20° C WAR0.0760.152r20–23: –0.185 ns
23° C WAR0.2100.420r20–27: 0.005ns
27° C WAR0.2910.582r23–27: –0.265ns
image

Figure 2. The reaction norms of narrow-sense heritability (h2) and coefficient of variation (CV) of fathers and their sons across different developmental temperatures.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES CITED

Body size is one of the most important features of organisms, affecting key fitness traits such as mating success and fecundity (Reeve et al. 2000), and determines the size dependency of resource acquisition and mortality rates on which natural selection can act (Chown and Gaston 2010). Many ecological factors shape body size in insects, growth temperature being one of the most effective through its effects on the length of aquatic life span, mortality, and developmental rates, which are in turn determinative of overall fitness (Sibly and Atkinson 1994, Debat et al. 2003, Maharaj 2003). In this context, genetic variation and its expression underlining body size should be directly related to the response and evolution of the organism in general.

In this study, we raised a highly inbred strain of Cx. quinquefasciatus at different developmental temperatures and estimated body size and its narrow-sense heritability (h2) for each temperature. Body sizes were expressed as the male wing aspect ratios, a very useful index for predicting body size as a function of wing length and an associated geometrical area parameter, the centroid size. Narrow-sense heritabilities were estimated using the offspring-on parent regression method, which is one of the most widespread of the classical heritability measurements available.

Overall, our results point to some remarkable aspects. First, as shown in Table 1, mean body size as measured by wing aspect ratios conform to the general observation (i.e., Bergmann's rule) that when temperature is increased, body size is decreased accordingly. This is true for the fathers across the entire temperature range with the sons less so, with the lowest temperature (20° C) mediating the lowest size and the remaining temperatures (23° C and 27° C) not in decreasing size order. This may be due to the introduction of some experimental variance. Clearly these sons would have shown the same temperature response if that part of the experiment was repeated, as the difference in size between 23° C and 27° C sons is small compared to the difference between the sizes at 23° C (or 27° C) for sons and the size of the sons of the lowest temperature (20° C). Phenotypic variation as expressed by coefficients of variation (CV) is considerably lower across the developmental temperatures in our study. Although the inclusion of more thermal regimes may change the range of the lower CV values, we think that the overall picture would have been mostly the same as we included three developmental temperatures in increasing order in the current study. Another point in this respect is that the Cx. quinquefasciatus colony we used did not live for more than three generations beyond 30° C (Günay et al. 2010), posing a limit on the temperature range expansion that can be used. Both fathers and their sons show the same range of variation and there are no significant differences among the pairs of fathers, sons, and fathers-and-sons at all temperatures (Table 1). The interesting outcome is that the narrow-sense heritability, which is the amount of additive genetic variation that contributes to the phenotypic variation in a trait, follows the temperature increase. For each different developmental temperature there is a correspondingly different heritability estimate and the heritability increases with increasing temperatures. We have estimated the narrow-sense heritability as a function of the regression of the size values of the sons on the fathers’ values. The regression coefficients (b) should show the same increase with increasing temperature (Table 2). This suggests that the amount of causality association (as indicated by regression coefficients, b) changes in a directional way. Hence, there is no fixed heritability for size when the environment changes. This result can be taken as the expression of the well substantiated general observation that a change in environment in which heritability is measured will ensue in different values. As the quantitative trait in question is affected by many genes with small independent allelic contributions, as expected from the classical additive theory of quantitative genetics, it may easily yield different trait values with environmental change (Falconer and Mackay 1996). This is typical in many laboratory results (Mackay 2001). However, because the organisms do not live in isolation and do experience environmental variation continuously in their lifetimes in nature, our findings may also point to natural situations in which temperatures change frequently. This would be the simplest corollary from our results, supporting a general truth reflecting one of the many situations in nature. What is remarkable is the increase in the genetic variation (as expressed by h2). We suggest this increase may be due to a change in the expression of the genetic variation for body size. We have estimated the additive genetic variation in each case (Falconer and Mackay 1996). If the putative genes affecting body size contribute to the trait value equally and independently from each other, some genes hidden because of the trade-offs for physiological threshold reasons could be manifest in their effect in response to temperature raise. That effect is well known and is the result of the unleashing of cryptic genetic variation due to the decanalization of the genetic variance for a trait (Gibson and Dworkin 2004, Gibson 2009). This would be reflected in an incremental raise in the expression of genetic variation, which seems to be the case for our results across the increasing temperature regime.

The other point, which is related to the temperature dependent increase in genetic variation in our study, is that the change in the genetic variation could be the result of non-negligible G × E for body size. Indeed, this is the case here. We have made a plot of reaction norm for h2 (Figure 2). This plot provides a clear illustration of the presence of G × E for body size across the temperature range. To assess quantitatively if G × E occurs, we have performed cross-environment correlation analyses between the phenotypic expressions (i.e., wing aspect ratio, WAR) of body size of the sons raised at different temperatures for heritability analysis. The idea was that if there were genotype-by-environment interactions, correlations between environmental pairs would be less than unity (Mackay and Anholt 2007). This is again the case; all correlation coefficients are less than unity, stressing the presence of G × E for body size. Overall, we suggest that the additive genetic variation for body size can strongly respond to temperature variation. Our findings may indicate very important ecological aspects with respect to the distribution of Cx. quinquefasciatus. The global phenomenon of warming is of central importance in this respect. Our findings show that increases in temperature can be reflected by an increase in genetic variance, possibly by unleashing cryptic genetic variation. This increase is of the body size, and body size in turn is determinative in important fitness components such as developmental time and the related overall survival. Along with a trade-off between the food scarcity and larger sizes, small individuals may survive and reproduce better when food is limited because they need less food to sustain themselves (Dingle 1992, Blanckenhorn et al. 1994). This may mean that temperature increases can trigger better adaptations via smaller body sizes even when food is scarce in the changing environment. This could be expected to create greater distribution through warming environments. That the genetic variation can be greater for increasing temperatures may have remarkable distribution effects. Indeed, there is a trade-off between desiccation tolerance and body size. The more tolerant individuals are, the smaller they are in size (Alpert 2006). What is striking is that the higher amounts of genetic variation in desiccation tolerance as measured by narrow-sense heritability can be directly related to a wider distribution. In a recent study, Kellerman et al. (2009) found that, along with cold tolerance, the differences in distribution between the specialist/narrowly distributed species and generalist/widely distributed species of Drosophila are clearly related to the amount of the genetic variation (narrow-sense heritability) in desiccation tolerance. The more tolerant species appear to have higher narrow-sense heritability and wider distribution as generalists. Contrastingly, lower amounts of narrow-sense heritability correspond to the lower desiccation tolerance of the narrowly distributed specialist species (Kellerman et al. 2009). The desiccation is a direct function of the increased temperatures, with organisms having greater genetic variation in desiccation tolerance. The increase of the expression of genetic variation in response to the temperature increases we found in our study may point to the potential of range expansion in Cx. quinquefasciatus when environments become warmer.

In conclusion, we suggest that our study of the effect of temperature change on the amount of genetic variation of body size in Cx. quinquefasciatus can reflect the fitness variation in local adaptive perspectives as well as the more important issue of species distribution. We believe our results contribute to the general explanations of adaptation in the context of global warming. A next step would be to work out the genomic expression profiles in response to temperature shifts in Cx. quinquefasciatus to find the candidate genes affecting body size in relation to their pleiotropies with respect to the correlated traits. But, as a first approximation, we also believe in the still invaluable quality of the classical quantitative genetic assessment of genetic variation as narrow-sense heritability (h2) estimates for the traits correlated with body size, such as developmental time and desiccation resistance. An alternative and a good start, therefore, would be to design experiments of heritability for such correlated traits. The overall scheme would be valid and worthwhile for other species of mosquitoes.

REFERENCES CITED

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES CITED