Sexual size dimorphism and egg-size allometry in birds.
If sexual size dimorphism in adults results in costs to juveniles, selection should favor any adaptation that reduces those costs. Although the costs are borne directly by juveniles, their parents share those costs (directly through provisioning the young and indirectly through decreased fitness when their offspring starve), thus selection should favor adaptations in adults that increase the survival of their offspring. One way in which the cost of growing large could be reduced is for nestlings to start growing at a higher proportion of their adult size (i.e., for their mothers to lay larger eggs). Egg size should respond to selection because, although egg size varies allometrically with body size, substantial variation within that general relationship exists (Gill 1990), and egg size is highly heritable (Boag and van Noordwijk 1987). Egg size appears to influence growth (Schifferli 1980; Howe 1976; Williams 1980; Teather 1990) and survival (Davis 1975; Howe 1976; Thomas 1983; although see Bolton  concerning potentially confounding effects of egg size and parental quality); thus by laying larger eggs, females should lose fewer of their larger-sexed offspring to starvation. Evidence from strongly dimorphic species suggests that males and females hatch from eggs of similar size (Fiala 1981; Blank and Nolan 1983; Bancroft 1984a; Weatherhead 1985; Teather 1989), thus females do not lay larger eggs when producing the larger sex. The alternative we test here is that females of sexually dimorphic species in which males are larger could lay larger eggs than expected for their body size. Because females would benefit from any reduction in mortality of their progeny, a female trait (egg size) could easily evolve in response to selection acting directly on a male trait (i.e., male body size when males are the larger sex).
If egg size has been modified to reduce the cost of sexual size dimorphism, two predictions follow. First, in species in which males are larger, the deviation in egg size from that predicted by the female's body size should increase as sexual dimorphism increases. Thus, females of the most dimorphic species should lay the largest eggs relative to their body size. Second, in species in which females are larger than males, no relationship between egg size and sexual size dimorphism is predicted. Rather, females should continue to lay eggs in proportion to their body size. This follows because as females become larger, so do their eggs, thereby automatically compensating for the increased energetic cost of raising daughters.
Alternatively, egg size may not reflect the costs of sexual size dimorphism, but rather egg size may simply vary with body size. Two allometric relationships seem possible. First, egg size could vary with female body size alone. In this case, egg-size deviations should be independent of sexual size dimorphism. Second, egg-size deviations could vary as some function of both male and female body size. For species in which males are larger, this hypothesis makes the same prediction as the cost-reduction hypothesis, namely that as size dimorphism increases, females should lay larger eggs relative to their body size. For species in which females are larger, however, this hypothesis predicts that relative egg size will decrease as size dimorphism increases.
MATERIALS AND METHODS
To determine how egg size varied with sexual size dimorphism in birds, we assembled data on adult and egg weights for species in six taxonomic groups in which the constituent species displayed substantial variation in sexual dimorphism. Three of the taxa (waterfowl, shorebirds, and galliformes) have precocial young and the other three (raptors, owls, and icterines) have altricial young. Sexual size dimorphism is predominantly male-biased (i.e., males are larger) in the waterfowl, galliformes, and icterines, and female-biased in the owls and raptors, whereas biases in both directions occur among the shorebirds. Thus, our data included substantial variation in both the patterns of dimorphism and in nesting ecology.
We obtained weight data for males and females of 446 species from the following sources: waterfowl--Johnsgard (1978), Palmer (1976a,b); raptors--Palmer (1988a,b); shorebirds--Johnsgard (1981); owls--Johnsgard (1988a), Mikkola (1983); galliformes--Johnsgard (1986, 1988b); icterines--various sources. Values for adult and egg weights for species not provided were obtained from Dunning (1984) and Schonwetter (1960-1972), respectively.
By referring to different sources, we attempted to obtain the best estimates for adult and egg sizes for each species. In several cases, ranges rather than mean values were provided for adult weights; in such instances we used the midpoint of this range. Where discrepancies occurred in values between sources, or within the same source (because of seasonal or geographic variability), final values were determined in one of the following ways. If the mean mass for different races or subspecies was provided, we used the one (usually nominate) for which egg mass was reported. Where values for egg size could not be matched with any of the races, we used mean values from all races. Similarly, where substantial geographic variation existed, we again tried to match the locality from which adult masses were obtained to those from which egg masses came. Where body mass exhibited substantial seasonal variation, we used breeding weights when available.
To test our hypotheses, we compared observed egg weights to those expected based on female weights. The general relationship between egg and body weight is
egg weight = a[(female body weight).sup.b],
where a and b are constants for each taxonomic group (Rahn et al. 1975). We obtained predicted egg weights for each species by substituting female body weights into the general equation and using the values of the constants provided by Rahn et al. (1975) for the appropriate taxonomic group. Values for Anseriformes (a = 0.641, b = 0.673, r = 0.93, N = 149); Galliformes (a = 0.484, b = 0.640, r = 0.96, N = 52); Charadriiformes (a = 0.613, b = 0.726, r = 0.98, N = 60); Strigiformes (a = 0.717, b = 0.603, r = 0.99, N = 15); and Falconiformes (a = 0.741, b = 0.633, r = 0.98, N = 54) were used for the waterfowl, galliformes, shorebirds, owls, and raptors, respectively. For icterines, we used the general values for passerines (a = 0.340, b = 0.677, r = 0.91, N = 295). Having obtained predicted egg weights, we then calculated the deviation of observed values from those predicted as a proportion of the observed values.
Two points need to be made regarding our use of Rahn et al.'s (1975) equations for the relationship between female weight and egg weight. The first concerns our reason for not developing our own equations from the data we compiled. Our objective was to use the equations to determine the expected egg weights for females of dimorphic species were they not dimorphic. Thus, for a group such as the icterines, in which sexual dimorphism is pronounced in most species for which we had data, an equation based on those data alone would have been strongly influenced by any effect that dimorphism had on egg weight (i.e., the very phenomenon we are trying to detect). This would then confound our objective of using residuals from the regression analysis to determine whether dimorphism affects egg weight. By using Rahn et al.'s general equation for passerines, for which pronounced size dimorphism is uncommon, we avoided that problem. Less potential for similar confounding effects existed for the other groups we analyzed, but for consistency we used Rahn et al.'s equations throughout.
The second point concerns the use of Rahn et al.'s least-squares linear regression (LLR) to develop their relationships between female weight and egg weight. Because error exists in both the X and Y variables, LLR consistently underestimates the slope of the relationship (LaBarbera 1989). Nonetheless, LLR remains a more appropriate analysis than reduced major axis regression (RMA) in this situation for two reasons (Harvey and Pagel 1991). First, the [R.sup.2] values for the equations we used were all high (0.83 to 0.96), which minimizes the underestimate of the slope from LLR. Second, and more important, when the objective of the regression analysis is to calculate residuals to control for the effect of weight (as we did here), then LLR is more appropriate, because with RMA, residuals computed parallel to the Y-axis will be correlated with the independent variable.
Multiple species comparisons can be complicated if species share similar attributes because of common ancestry. To reduce this problem, we used the method suggested by Read and Weary (1990). For each taxonomic level within our sample (e.g., species within genera, genera within tribes, etc.), we used Spearman rank correlations to examine the relationship between size dimorphism (expressed as the ratio male mass to female mass) and the deviation from the egg mass expected for a given sized female. If the deviation in egg size varied with the degree of sexual size dimorphism, then we would expect significantly more of our Spearman rank correlations to be the same sign (positive or negative) than expected by chance (P [is less than] 0.05), as determined by a binomial test. For example, 15 species of icterines yielded a total of six comparisons (five between species of the same genera, and one among the means of the different genera within the subfamily Icterinae), five of which indicated a positive correlation between relative egg size and sexual size dimorphism. Taxonomic classifications for all groups were taken from Sibley and Ahlquist (1990). In addition, we performed single sample t-tests on z-scores estimated from the correlation coefficients using 0 as the population mean. Z-scores have the advantage of taking into account the sample size used in calculating the correlation coefficient. Thus, within genera, for example, the z-score corresponding to a correlation coefficient from a genus with many species would be larger than that derived from a coefficient from a genus with few species. A significant positive mean z-score would indicate a positive relationship between size dimorphism and relative egg weight.
All statistical analyses were done using SYSTAT (Wilkinson 1989) and StatView (Abacus Concepts 1987).
Scatter plots of egg-weight deviation relative to sexual size dimorphism illustrate two general trends. First, observed egg weights deviated substantially from values expected based on female body weight. Note that nearly all residuals for the blackbirds were negative, indicating that blackbirds collectively lay relatively light eggs compared with other passerines. Because subsequent comparisons using these data were conducted within the blackbirds, however, it was the relative values of the residuals that were important. Thus, the fact that most of the values were negative had no effect on our analyses. The second general trend was that for several taxonomic groups relative egg weight was associated with sexual size dimorphism. This pattern was confirmed by the hierarchical analysis. In the blackbirds and waterfowl, in which males are heavier in dimorphic species, a significant number of groups showed an increase in relative egg weight as size dimorphism increased (i.e., as species became TABULAR DATA OMITTED more dimorphic, females laid heavier eggs relative to their body weight; table 2). In the galliformes (where males are also heavier), relative egg weight was not significantly associated with size dimorphism, although the trend was positive. In the genus Francolinus, the most speciose genus in our sample of galliformes, relative TABULAR DATA OMITTED egg weight was highly correlated with size dimorphism ([r.sub.s] = 0.658, N = 26, P [is less than] 0.01).
In owls and raptors, in which females are heavier in dimorphic species, we observed the opposite trend: relative egg weight decreased as sexual dimorphism increased. However, the pattern was significant only for the raptors.
Shorebirds were the only group that included both species with male-biased and species with female-biased size dimorphism. When we restricted our analysis to those species that were monomorphic or in which males were heavier, we did not find a significant association between relative egg weight and sexual size dimorphism. Similarly, the association was not significant when restricted to monomorphic species and those in which females were heavier. However, when all shorebird species were included, a significant number of groups showed an increase in relative egg weight as male weight increased relative to female weight. Similarly, when all the groups in table 2 were combined, relative egg weight increased significantly with increasing relative male weight.
Among the groups we examined, relative egg weight varied substantially. When examined across species, females laid relatively heavier eggs as male weight increased relative to female weight. This pattern was consistent with our initial hypothesis that an increase in relative egg weight would reduce the costs for males of growing large in species with male-biased size dimorphism. However, this pattern was also consistent with the hypothesis that egg weight varies as an allometric function of both male and female size. Furthermore, among species in which females are heavier, females were found to lay proportionately lighter eggs. Because this latter pattern was not predicted by our cost-reduction hypothesis, we conclude that interspecific variation in relative egg weight is best explained as a joint allometric function of both male and female body size.
Although our results did not support our hypothesis predicting adaptive modification of egg size in response to the costs of sexual size dimorphism, that failure is not damaging to the underpinnings of the hypothesis. We had predicted that females should lay heavier eggs in species in which males are larger because heavier eggs would reduce the cost to males of growing large. Thus, females laying heavier eggs would have more sons survive. We found this predicted pattern between relative egg size and size dimorphism, but it was best explained as a consequence of allometry. However, even if our conclusion is correct regarding allometry, males of species with male-biased dimorphism still should survive better by hatching from heavier eggs.
Elsewhere we have explored how sex ratios (Weatherhead and Teather 1991) and nestling growth and development (Teather and Weatherhead 1994) vary with sexual size dimorphism in birds. In both cases, we found that although the patterns were qualitatively consistent with hypotheses that predicted adaptations specifically in response to size dimorphism, the patterns could be explained more parsimoniously as simple energetic or allometric consequences of one sex becoming larger. The results we have reported here for egg size in sexually dimorphic species appear to provide yet a third example, in which a consequence of sexual size dimorphism coincidentally produces an effect predicted to evolve in response to costs associated with size dimorphism.
These results also have implications for the basis of egg-size allometry in birds. We initially proposed two possible patterns of allometry. Egg size could vary either with female body size alone or with both male and female body size. We had considered the former possibility more plausible for several reasons. Most obviously, females produce and lay the egg. If the general allometric relationship of egg size in birds is a consequence of the mechanics of developing, transporting, and laying an egg of a given size, then male size should have no influence on egg size. Among bird species, the general allometric relationship appears to allow considerable flexibility for both the size of eggs that birds can lay and the size of birds that can be produced from eggs of a given size. Interspecifically, for example, a 505 g female ruddy duck (Oxyura jamaicensis) and a 1660 g female king eider (Somateria spectabilis) both lay eggs of about 73 g. Intraspecifically, in great-tailed grackles (Quiscalus mexicanus), 214 g males and 119 g females are both produced from eggs of the same size. This flexibility, coupled with high heritability for egg size, should allow substantial scope for the adaptive modification of egg size. Thus, we viewed egg size as essentially a female trait and did not expect male body size to constrain the size of eggs that females lay.
Our results suggested, however, that male size influences egg size. One might argue that this result is not unexpected because females inherit the genetic basis for the eggs they will lay from both their mother and father. In domestic chickens, for example, estimates of the heritability of egg weight average approximately 0.52 whether they are based on maternal or paternal half-sib correlations (Kinney 1969). However, the fact that males carry genes for egg size does not protect those genes from natural selection, because the effect of those genes will be exposed in daughters. Thus, if selection favored females of a particular size laying eggs of a particular size, that relationship should evolve even though genes coding for egg size are carried by both sexes.
Why, then, does male size exert an independent effect on egg size? Our expectation that male body size would not have an allometric effect on egg size was based on the assumption that the basis for egg size allometry was a consequence of the mechanics of egg production and transportation prior to laying. Our results suggest that genes influencing overall body size have an epistatic effect on egg size. This would cause egg size to evolve directly as a function of both male and female body size and would limit the extent to which egg size could be modified independently. The evidence from chickens is consistent with this hypothesis (Kinney 1969). Egg weight and body weight are both highly heritable, and the genetic correlation between egg weight and body weight averages approximately 0.37. The fact that a female trait (egg weight) is allometrically related to both female and male body size suggests that the relationship between egg size and body size in sexually dimorphic birds may provide an interesting opportunity to investigate the basis for allometric relationships in general.
We thank M. L. Forbes, S. M. Yezerinac, F. Rohwer, A. R. Palmer, and an anonymous reviewer for their comments on the manuscript. A. F. Read made several suggestions regarding the hierarchical analysis. Funding for this project came from the Natural Sciences and Engineering Research Council of Canada.
Abacus Concepts. 1987. StatView II. Abacus Concepts, Berkeley.
Bancroft, G. T. 1984a. Patterns of variation in size of boat-tailed grackle Quiscalus major eggs. Ibis 126:496-509.
-----. 1984b. Growth and sexual dimorphism of the boat-tailed grackle. Condor 86:423-432.
Blank, J. L., and V. Nolan. 1983. Offspring sex ratio in red-winged blackbirds is dependent on maternal age. Proceedings of the National Academy of Sciences, USA 80:6141-6145.
Boag, P. T., and A. J. van Noordwijk. 1987. Quantitative genetics. Pp. 45-78 in F. Cooke and P. A. Buckley, eds. Avian genetics. Academic Press, London.
Bolton, M. 1991. Determinants of chick survival in lesser black-backed gulls: relative contributions of egg size and parental quality. Journal of Animal Ecology 60:949-960.
Carter, M.D. 1986. The parasitic behavior of the bronzed cowbird in south Texas. Condor 88:11-25.
Clutton-Brock, T. H., S. Albon, and F. E. Guinness. 1985. Parental investment and sex differences in juvenile birds and mammals. Nature 313:131-133.
Darwin, C. 1871. The descent of man and selection in relation to sex. Murray, London.
Davis, J. W. F. 1975. Age, egg size, and breeding success in the herring gull Larus argentatus. Ibis 117:460-473.
Dunning, J. B., Jr. 1984. Body weights of 686 species of North American birds. Western Bird Banding Association, Monograph 1. Eldon Publishing, Cave Creek, Ariz.
Fiala, K. L. 1981. Reproductive cost and the sex ratio in red-winged blackbirds. Pp. 198-214 in R. D. Alexander and D. W. Tinkle, eds. Natural selection and social behavior: recent research and new theory. Chiron Press, New York.
Fiala, K. L., and J. D. Congdon. 1983. Energetic consequences of sexual size dimorphism in nestling red-winged blackbirds. Ecology 64:642-647.
Gill, F. B. 1990. Ornithology. Freeman, New York.
Harvey, P. H., and M. D. Pagel. 1991. The comparative method in evolutionary biology. Oxford University Press, Oxford.
Holcomb, L. C., and G. Tweist. 1970. Growth rates and sex ratios of red-winged blackbird nestlings. Wilson Bulletin 82:294.
Howe, H. F. 1976. Egg size, hatching asynchrony, sex, and brood reduction in the common grackle. Ecology 57:1195-1207.
-----. 1979. Evolutionary aspects of parental care in the common grackle, Quiscalus quiscula L. Evolution 33:41-51.
Johnsgard, P. A. 1978. Ducks, geese, and swans of the world. University of Nebraska Press, Lincoln.
-----. 1981. The plovers, sandpipers, and snipes of the world. University of Nebraska Press, Lincoln.
-----. 1986. The pheasants of the world. Oxford University Press, Oxford.
-----. 1988a. The quails, partridges, and francolins of the world. Oxford University Press, Oxford.
-----. 1988b. North American owls. Biology and natural history. Smithsonian Institution Press, Washington, D.C.
Kinney, T. B., Jr. 1969. A summary of reported estimates of heritabilities and of genetic and phenotypic correlations for traits in chickens. Agricultural Handbook No. 363. U.S. Department of Agriculture, Washington, D.C.
LaBarbera, M. 1989. Analyzing body size as a factor in ecology and evolution. Annual Review of Ecology and Systematics 20:97-117.
Mikkola, H. 1983. Owls of Europe. T & A D Poyser, Calton.
Newton, I. 1979. Population ecology of raptors. Buteo Books, Vermillion, S.D.
Palmer, R. S., ed. 1976a. Handbook of North American birds, vol. 2. Yale University Press, New Haven, Conn.
-----. 1976b. Handbook of North American birds, vol. 3. Yale University Press, New Haven.
-----. 1988a. Handbook of North American birds, vol. 4. Yale University Press, New Haven.
-----. 1988b. Handbook of North American birds, vol. 5. Yale University Press, New Haven.
Rahn, H., C. V. Paganelli, and A. Ar. 1975. Relationship of avian egg weight to body weight. Auk 92:750-765.
Read, A. F., and D. M. Weary. 1990. Sexual selection and the evolution of bird song: a test of the Hamilton-Zuk hypothesis. Behavioral Ecology and Sociobiology 26:47-56.
Richter, W. 1983. Balanced sex ratios in dimorphic altricial birds: the contribution of sex-specific growth dynamics. American Naturalist 121: 158-171.
-----. 1984. Nestling survival and growth in the yellow-headed blackbird, Xanthocephalus xanthocephalus. Ecology 65:597-608.
Robinson, S. K. 1986. Competitive and mutualistic interactions among females in a neotropical oriole. Animal Behavior 34:113-122.
Roskaft, E., and T. Slagsvold. 1985. Differential mortality of male and female offspring in experimentally manipulated broods of the Rook. Journal of Animal Ecology 54:261-266.
Schifferli, L. 1980. Growth and mortality of male and female nestling house sparrows (Passer domesticus) in England. Avocetta 4:49-62.
Schonwetter, M. 1960-1972. Handbuch der oologie. Lief. 1-19. W. Meise, ed. Akademie, Berlin.
Selander, R. K. 1958. Age determination and molt in the boat-tailed grackle. Condor 60:353-376.
-----. 1972. Sexual selection and dimorphism in birds. Pp. 180-230 in B. G. Campbell, ed. Sexual selection and the descent of man: 1871-1971. Aldine, Chicago.
Sibley, C. G., and J. E. Ahlquist. 1990. Phylogeny and classification of birds. A study in molecular evolution. Yale University Press, New Haven, Conn.
Slagsvold, T., E. Roskaft, and S. Engen. 1986. Sex ratio, differential cost of rearing young, and differential mortality between the sexes during the period of parental care: Fisher's theory applied to birds. Ornis Scandinavica 17:117-125.
Teather, K. L. 1987. Intersexual differences in food consumption by hand-reared great-tailed grackle (Quiscalus mexicanus) nestlings. Auk 104:635-639.
-----. 1989. Sex and egg size in great-tailed grackles. Condor 91:203-205.
-----. 1990. The influence of sibling gender on the growth and survival of great-tailed grackle nestlings. Canadian Journal of Zoology 68:1925-1930.
Teather, K. L., and P. J. Weatherhead. 1988. Sex-specific energy requirements of great-tailed grackle (Quiscalus mexicanus) nestlings. Journal of Animal Ecology 57:659-668.
-----. 1989. Sex-specific mortality in nestling great-tailed grackles. Ecology 70:1485-1493.
-----. 1994. Allometry, adaptation, and the growth and development of sexually dimorphic birds. Oikos. In press.
Thomas, C. S. 1983. The relationship between breeding experience, egg volume and reproductive success of the kittiwake Rissa tridactyla. Ibis 125:567-574.
Weatherhead, P. J. 1985. Sex ratios of red-winged blackbirds by egg size and laying sequence. Auk 102:298-304.
Weatherhead, P. J., and K. L. Teather. 1991. Are skewed fledgling sex ratios in sexually dimorphic birds adaptive? American Naturalist 138:1159-1172.
Wilkinson, L. 1989. SYSTAT: the system for statistics. Systat Inc. Evanston, Ill.
Williams, A. J. 1980. Variation in weight of eggs and its effect on the breeding biology of the great skua. Emu 80:198-202.
|Printer friendly Cite/link Email Feedback|
|Author:||Weatherhead, Patrick J.; Teather, Kevin L.|
|Date:||Jun 1, 1994|
|Previous Article:||Patterns of fluctuating asymmetry and selection against asymmetry.|
|Next Article:||Polyphyly of the blackbird genus Agelaius and the importance of assumptions of monophyly in comparative studies.|