Facultative adjustment of the sex ratio in an insect (Planococcus citri, Pseudococcidae) with paternal genome loss.
We worked with the citrus mealybug, Planococcus citri (Risso), which has a life history typical of most members of the family Pseudococcidae. The adult female is a small, obscurely segmented, wingless insect covered in wax that feeds by sucking juices from the phloem of many species of plant. Females, which often live in large colonies, produce numerous eggs that hatch into first instar "crawlers," which disperse to find new feeding sites. The development of the two sexes is similar until the third instar when males enter a quiescent stage in a cocoon followed by a fourth instar when wing buds are visible; the adult male is winged. Juvenile females only have three instars before becoming adults. Planococcus citri has a PGL sex-determination system of the lecanoid type (Schraeder 1921; Nut 1980, 1990): the paternal genome in somatic tissue is inactivated at the blastula stage of development but can be detected as a heterochromatized body in the nucleus of all cells; in germ-line tissue, the paternal genome is eliminated during spermatogenesis.
Several studies have examined the sex ratio of P. citri and probably more is known about this species than any other mealybug. James (1937, 1938) and Nelson-Rees (1960) discovered that females kept virgin for the first part of their adult life produced a highly male-biased sex ratio when eventually mated. A similar phenomenon has been reported in armoured scale insects (Diaspididae) (reviewed by Nur 1990). Charnov (1982) has suggested that the insect may use the delay in mating as an indication of a low frequency of males in the population and respond by producing more of the rarer sex. Nelson-Rees (1960) also found evidence that there may be cycles in female production over the reproductive life of the mother. Sex ratios were also influenced by temperature, humidity, and ionizing radiation (James 1937, 1938; Nelson-Rees 1960). There thus seems to be evidence for biased and possibly facultative sex ratios, although these results are difficult to interpret and are potentially biased by differential mortality. This risk is particularly great in mealy-bugs where juvenile mortality is very high and sexual differences are manifested early in development.
The aims of our study were twofold: first, to establish whether the primary sex ratio of the mealybug differed from equality by sexing the eggs using a cytological technique; second, to determine whether female mealybugs were able to exert facultative control over offspring sex ratio in response to environmental cues. We chose to study the response of females to crowding. We hypothesized that females would respond to increased crowding by producing a more male-biased sex ratio. Mealybugs are colonial and females tend to remain on the plant where they are born. Males, on the other hand, are winged and disperse. As colony size and hence crowding increases, there is likely to be greater competition among sibling females than among sibling males, leading to selection on mothers increasingly to bias their sex ratio toward males (Clark 1978; Taylor 1981; Charnov 1982). We tested the influence of crowding in a two-way experimental design with two levels of crowding during development and two levels of crowding during oviposition as the experimental treatments.
A culture of P. citri was obtained courtesy of Wye College, University of London. Mealybugs were cultured on sprouting potatoes (Maris Piper cultivar) kept in a constant temperature room at 25 [degrees] C and 75% relative humidity under a 16 h light: 8 h dark regime. Ovipositing females were removed from the main culture and allowed to continue laying eggs on fresh potato sprouts. Egg masses laid by these females were transferred to fresh potatoes either singly or in groups of five to create two different levels of what we shall refer to as juvenile crowding. The mealybugs were allowed to continue developing for 20 d at which time the sexes could be distinguished and females were in their late second instar. Females from both juvenile crowding treatments were then transferred to fresh potatoes, either singly or in groups of 20, to create two levels of what we shall refer to as adult crowding. Individual mealybugs thus experience one of four combinations of juvenile and adult crowding.
Males were introduced on day 30, after the females had become reproductively mature. Eggs were subsequently collected from four individual females from each of the four treatments every 2 d. Mealybug eggs are laid beneath the female's body and are easily removed by lifting the mother with a toothbrush bristle. Experimental females on crowded potatoes were identified by a small drop of enamel paint on their dorsal surface. Eggs were counted and placed for 4 d in a fixative made from four parts glacial acetic acid, three parts 90% alcohol, and one part water. The sex of the eggs was determined by staining with lacmoid stain made up in glacial acetic acid, lactic acid, and ethanol. When the eggs are squashed and examined under a low power light microscope, the condensed paternal genome is easily seen as a dark staining body in the nucleus of males only. Approximately 7500 eggs were sexed by this method during the experiment.
In studying the effect of treatment on sex ratio, we included in the analysis adult mealybug length, longevity, and total fecundity, as well as the weight of the potato on which they oviposited. All potatoes used were between 100-150 g in weight with a 10-15 cm shoot. Sex ratios were analysed using generalized linear modeling techniques implemented on the GLIM statistical package (McCullagh and Nelder 1989). GLIM allows standard regression and analysis of variance techniques to be used on data with non-normal errors and without transformation. In the case of sex ratios, the natural error distribution is likely to be binomial. However, quite frequently the error distribution is found to have a higher variance than expected for the binomial distribution (termed overdispersion). In such cases, we corrected for overdispersion using Williams's algorithm (Williams 1982), checking the appropriateness of the model by examining the distribution of residuals.
The mealybugs lived on average for 19.8 d (SE = 1.2) during which time they produced 689 eggs (SE = 61). Body size and longevity together explained approximately 59% of the variation in fecundity. There was a significant effect of treatment on body size ([F.sub.(2,12)] = 4.76, P = 0.03) and below body size is included as a covariate.
TABLE 1. Summary of analysis of sex ratio using GLIM (see methods). The first section of the table gives the deviance around the mean and the residual deviance after fitting the full model. The second section shows the reduction in deviance on dropping each main effect from the full model, together with the associated drop in degrees of freedom, an estimate of the magnitude of the loss of explanatory power (the drop in deviance expressed as a percentage of the deviance around the mean), and the approximate significance of the removal. Deviance df Magnitude Approx. P Deviance around mean 90.5 15 Full model 9.0 9 Adult life history 25.5 3 28.2% [less than] 0.0001 Potato weight 2.3 1 2.4% N.S. (= 0.13) Juvenile crowding 10.8 1 11.9% [less than] 0.01 Adult crowding 48.1 1 53.1% [less than] 0.0001
In the analysis, we explored the effects of the two treatments, potato weight, mealybug size, longevity, and the total number of eggs laid. The last three variables were highly correlated and in what follows they were treated as a single variable "adult life history" (i.e., they were entered or dropped from the statistical model as a single unit): the analysis thus included four explanatory factors or variables. No interaction terms were significant and the full model consisted of the four main effects. The ratio of the residual deviance and the residual degrees of freedom after fitting the full model was 1.8 indicating some overdispersion that was corrected using Williams's algorithm ([Phi] = 0.0017; see methods). The importance of the main effects was explored by dropping and replacing each in turn (see Table 1).
The overall mean sex ratio, 0.316 (SE = 0.008), was significantly female biased. The effects of juvenile and adult crowding are shown in Figure 1. Females that were crowded as adults produced the most female-biased sex ratio, significantly more female biased than insects that were solitary as adults. However, within the two categories of adults, there was a significant tendency for mealybugs that had experienced solitary juvenile conditions to produce a relatively more female-biased sex ratio than insects that had been crowded as juveniles. Thus, against a background of an overall female-biased sex ratio, adult crowding decreased and juvenile crowding increased the observed sex ratio. The conditions experienced by the adults were more important than the conditions experienced by the offspring in influencing sex ratio (Table 1). The adult life history of the mealybug also influenced sex ratio with larger, more fecund, and long-lived individuals producing a more male-biased sex ratio. Host plant quality (potato weight) did not influence sex ratio.
We also analyzed the daily pattern of offspring sex ratio, treating the number of days after mating as a continuous variable. Because of variation among females, we fitted insect as a main effect and then explored the additional explanatory power of days since mating and the interaction between day since mating and the two treatments. Overall, the mealybug sex ratios changed from highly female biased soon after mating to less female biased or even male biased in old individuals. However, the pattern was different among treatments [ILLUSTRATION FOR FIGURE 2 OMITTED]. The three-way interaction, day x juvenile treatment x adult treatment, was not significant but the two-way interactions were (day x adult treatment: [[Chi].sup.2] [similar to] 11.1, 1 df, P [less than] 0.001; day x juvenile treatment: [[Chi].sup.2] [similar to] 6.31, 1 df, P = 0.012; residual deviance 140.1, 135 df, Williams's [Phi] = 0.108). However, the two interaction terms together accounted for only 7.7% of the deviance about the mean. Both individuals that were crowded as juveniles and crowded as adults tended to produce more males late in life than equivalent uncrowded individuals.
The average number of eggs laid by each female in our experiment (689) was much higher than the average number of progeny recorded per female (292) by Nelson-Rees (1960). The discrepancy is almost certainly due to very high early juvenile mortality. The magnitude of juvenile mortality is important as even a relative small bias toward male or female deaths could influence the observed secondary sex ratio. Our average sex ratio was 0.31, slightly more female biased than the value of 0.37 reported by Nelson-Rees (1960) and much more female biased than the 0.5 noted by James (1937). It is clear that the primary sex ratio produced by P. citri has a substantial female bias.
We found that the composite explanatory variable "adult life history" and the two experimental treatments all influenced mealybug sex ratio. Large, long-lived, fecund individuals tend to produce more sons than smaller, short-lived individuals of low fecundity. The mechanistic explanation for this result is probably associated with the increasing proportion of sons produced late in life. Large individuals tend to live longer and so reach an age when more male eggs are laid leading to a less female-biased sex ratio. As was mentioned in the introduction, it has been found in a number of mealybugs that a delay in mating leads to a more male-biased sex ratio and this has been interpreted as an adaptive response to a paucity of males in the environment (Charnov 1982). The observed relationship between age (irrespective of insemination) and the sex of offspring may be the mechanism generating this response, or the influence of a delay in mating may be simply an epiphenomenon if the age/sex ratio relationship is caused by a nonadaptive physiological constraint.
Before performing the experiments, we hypothesized that the biology of mealybugs would result in greater potential resource competition among sisters than among sons. We based this prediction on the greater dispersal ability of males (there is a juvenile crawler stage and the adults have wings) and the fact that males do not feed after the second instar. Greater competition among siblings of one sex for limiting resources is predicted to give rise to selection for a bias in sex ratio toward the other sex, a process termed local resource competition. We expected the sex ratio to become more male biased as the degree of crowding increased.
The results clearly show that crowding influences sex ratio, but not in the simple way we predicted. The overall sex ratio is female rather than male biased and while the proportion of males produced does increase with juvenile crowding, in line with our prediction, the reverse occurs with an increase in adult crowding. For local resource competition to explain the effects of crowding, the sign of the correlation between crowding and sibling competition would have to change between the juvenile and adult stages, something that seems unlikely.
The treatments had a significant effect on sex ratio after other life-history traits had been controlled for. It appears that there is a proximate mechanism that allows the environment to influence the sex ratio produced by individual mealybugs, although we do not know what that mechanism is. Recently, Sabelis and Nagelkerke (1987, 1988, 1993), Dinh et al. (1988), and Nagelkerke and Sabelis (1991) have established that phytoseiid mites with PGL are able to change their sex ratio in response to the environment. In mealybugs and other homopterans with PGL, facultative control has been suspected based on studies of the secondary sex ratio (Nelson-Rees 1960; Charnov 1982; Nur 1990), but the large amount of juvenile mortality has meant that differential mortality could not be excluded. Our results conclusively demonstrate facultative control of the sex ratio in mealybugs with PGL. Bull's (1983) suggestion that the inability of species with PGL to change their sex ratio in response to local conditions might explain the rarity of PGL in comparison with haplodiploidy must at least be modified. Species with PGL may still be relatively disadvantaged if the mechanisms of sex ratio manipulation are coarse compared with haplodiploidy, where the sex of an egg is under the behavioral control of the mother. However, Nagelkerke and Sabelis (1991) were able to demonstrate precise sex-ratio control (i.e., with less than binomial variance) in phytoseiid mites.
We have two candidate explanations for the overall female-biased sex ratio. First, it is possible that crowded mealybugs are more efficient than solitary mealybugs in obtaining resources from the plant. This occurs in several aphid species where groups of individuals may form large nutrient sinks (Way and Banks 1967; Murdie 1969; Way and Cammell 1970). If female siblings tend to remain together and enhance each other's fitness then this will result in selection in favor of a female-biased sex ratio (e.g., Gowaty and Lennartz 1985; Packer 1986; Schwarz 1988). A second factor that may influence mealybug sex ratios is mating structure. If matings in large populations occur among siblings more frequently than random, a female-biased sex ratio is predicted (Hamilton 1967). Like Hamilton (1967), we assumed that mating in mealybugs occurs after dispersal and that this model is unlikely to be appropriate for mealybugs. A model that may be more relevant for mealybugs is the "haystack model" (Maynard Smith 1964), developed by Bulmer and Taylor (1980) and Nagelkerke (1996). Here, populations are founded by a limited number of foundresses and persist for several to many generations, during which time all matings occur between the members of the population. Female-biased sex ratios can be predicted by haystack models, although the relationship between population structure, natural history, and predicted sex ratio is complex (Nagelkerke 1996) and further work is required to see whether this model is appropriate for mealybugs. However, were one of these explanations to be correct, it is still unclear why female sex ratios are influenced, in opposite directions, by juvenile and adult crowding. It is possible that females are responding to changes in crowding rather than crowding levels per se, but further experimentation is needed to explore these possibilities.
To conclude, we have used cytological techniques to show that P. citri produces a strongly female-biased primary sex ratio in our laboratory experiments, confirming previous studies based on secondary sex ratios. The proportion of males produced increases with age so that the overall sex ratio of large, long-lived individuals is relatively more male biased. Insects respond to juvenile and adult crowding by changing their sex ratio, although the effects of the two treatments are in opposite directions. The female-biased sex ratio may be due to resource enhancement or be influenced by population mating structure. Thus, despite a PGL sex determination system, mealybugs, like phytoseiid mites, are capable of changing their sex ratio in response to the environment, although further work is required to understand the nature of the selective forces influencing sex ratio in these insects.
We are grateful to R. Blackman, M. Hunter, C. Nagelkerke. A, Rivero, and M. Sabelis for valuable help and discussions. NPV was supported by a National Environmental Research Council studentship.
ANTOLIN, M. F. 1993. Genetics of biased sex ratios in subdivided populations: Models, assumptions and evidence. Oxf. Surv. Evol. Biol. 9:239-281.
BULL, J. J. 1983, Evolution of sex determining mechanisms. Benjamin Cummings, Menlo Park, CA.
BULMER, M. G., AND P. D. TAYLOR. 1980. Sex ratio under the haystack model. J. Theor. Biol. 86:83-89.
CHARNOV, E.L. 1982. The theory of sex allocation. Princeton Univ. Press, Princeton, NJ.
CLARK, A. B. 1978. Sex ratio and local resource competition in a prosimian primates. Science 201:163-165.
DINH, N. V., A. JANSSEN, AND M. W. SABELIS. 1988. Reproductive success of Amblyseius idaeus and A. anonymus on a diet of two-spotted spider mites. Exp. Appl. Acarol. 4:41-51.
GODFRAY, H. C. J. 1994. Parasitoids: Behavioural and evolutionary ecology. Princeton Univ. Press, Princeton, NJ.
GOWATY, P. A., AND M. R. LENNARTZ. 1985. Sex ratios of nestling and fledgling red-cockaded woodpeckers favor males. Am. Nat. 126:347-353.
HAMILTON, W. D. 1967. Extraordinary sex ratios. Science 156:477-488
JAMES, H. C. 1937. Sex ratios and the status of the male in Pseudococcidae. Bull. Entomol. Res. 28:429-461.
-----. 1938. The effect of the humidity of the environment on sex ratios from over-aged ova of Pseudococcus citri (Risso). Proc. R. Entomol. Soc. Lond. 13:73-79.
KING, B. H. 1987. Offspring sex ratios in parasitoid wasps. Q. Rev. Biol. 62:367-396.
MAYNARD SMITH, J. 1964. Group selection and kin selection. Nature 201:1145-1147.
McCULLAGH, P., AND J. A. NELDER. 1989. Generalized linear models, 2d ed. Chapman and Hall, London.
MURDIE, G. 1969. Some causes of size variation in the pea aphid. Trans. R. Entomol. Soc. Lond. 121:423-439.
NAGELKERKE, C. J. 1996. Hierarchical levels of spatial structure and their consequences for the evolution of sex allocation in mites and other arthropods. Am. Nat. 148:16-39.
NAGELKERKE, C. J., AND M. W. SABELIS. 1991. Precise sex ratio control in the psuedo-arrhenotokous phytoseiid mite Tryphlodomus occidentalis. Pp. 193-207 in R. Schuster and E W. Murphy, eds. The Acari: Reproduction life history strategies and development. Chapman and Hall, London.
NELSON-REEs, W. A. 1960. A study of sex predetermination in the mealybug Planococcus citri. J. Exp. Zool. 144:111-137.
NUR, U. 1980. Evolution of unusual chromosome systems in scale insects (Coccoidea: Homoptera). Pp. 97-118 in R. L. Blackman, G. M. Hewitt, and M. Ashburner, eds. Insect cytogenetics. Blackwell, Oxford.
-----. 1990. Chromosomes, sex ratios and sex determination. Pp. 179-190 in D. Rosen, ed. Armoured scale insects, their biology, natural enemies and control. Elsevier Science Publishers, Amsterdam, The Netherlands.
PACKER, C. 1986. The ecology of sociality in felids. Pp. 429-451 in D. I. Rubenstein and R. W. Wrangham, eds. Ecological aspects of social evolution. Princeton Univ. Press, Princeton, NJ.
SABELIS, M. W., AND C. J. NAGELKERKE. 1987. Sex allocation strategies of pseudo-arrhenotokous phytoseiid mites. Neth. J. Zool. 37:117-136.
-----. 1988. Evolution of pseudo-arrhenotoky. Exp. Appl. Acarol. 10:45-51.
-----. 1993. Sex allocation and pseudoarrhenotoky in phytoseiid mites. Pp. 512-541 in D. L. Wrensch and M. A. Ebbert, eds. Evolution and diversity of sex ratio in insects and mites. Chapman and Hall, New York.
SCHRAEDER, F. 1921. The chromosomes of Pseudococcus nipae. Biol. Bull. 40:259-270.
SCHWARZ, M. P. 1988. Local resource enhancement and sex ratios in a primitively social bee. Nature 331:346-348.
TAYLOR, P. D. 1981. Intra-sex and inter-sex sibling interactions as sex ratio determinants. Nature 291:64-66.
WAY, M.J., AND C.J. BANKS. 1967. Intra-specific mechanisms in relation to the natural regulation of numbers of Aphis fabae. Ann. Appl. Biol. 59:189-205.
WAY, M. J., AND M. CAMMELL. 1970. Aggregation behaviour in relation to food utilisation by aphids. Pp. 229-247 in A. Watson, ed. Animal populations in relation to food resources. Blackwell, Oxford.
WERREN, J. H. 1987. Labile sex ratios in wasps and bees. Bioscience 37:81-96.
WILLIAMS, D. A. 1982. Extra-binomial variation in logistic linear models. Appl. Stat. 31:144-148.
WRENSCH, D. L., AND M. A. EBBERT. 1993. Evolution and diversity of sex ratio in insects and mites. Chapman and Hall, New York.
|Printer friendly Cite/link Email Feedback|
|Author:||Varndell, N.P.; Godfray, H.C.J.|
|Date:||Oct 1, 1996|
|Previous Article:||Sexual selection operating in a wild population of Drosophila robusta.|
|Next Article:||The selection differential in quantitative genetics and ESS models.|