LIFE HISTORY TRAITS AND NATURAL SELECTION FOR SMALL BODY SIZE IN A POPULATION OF DARWIN'S FINCHES.

Abstract

Application of the theory and methods of quantitative genetics to life history studies has led to a focus on the measurement of phenotypic and genetic covariances between life history traits (Stearns, 1980, 1982a, 1982b; Dingle and Hegmann, 1982 p. 231-244; Etges, 1982; Giesel et al., 1982; Hegmann and Dingle, 1982; Lande, 1982; Rose, 1983). These covariances are the essential parameters in theoretical models describing the evolution of such traits. Estimation of genetic covariance is based on the resemblance between relatives. Genetic covariances have usually been estimated in the laboratory where the environment is controlled and the main cause of mortality is often senescence. Covariances are less easy to estimate in nature because environmental correlations are likel.y to fluctuate greatly in magnitude and sign with changing conditions (Lande, 1982) and because causes of mortality also vary. Thus direct estimates of genetic covariances between life history traits will vary (Giesel et al., 1982; Stearns, 1982a, 1982b). In this paper we adopt an alternative, indirect, approach to estimating covariances between life history traits in natural populations. It depends on the measurement of selectioni on a morphological character at two life history stages, and the assumption that much of the variance in fitness at each stage is due to variance in the character or characters upon which selection acts. Two life history traits, such as probability of survival over two given developmental stages, may be said to covary if they are each determined by the same morphological character or characters. The size and direction of the covariance can be qualitatively estimated from the magnitude and directions of selection. Further, if the morphological character is found to be highly heritable it can be inferred that the two life history traits covary genetically. Boag and Grant (1981) observed intense natural selection favoring large beak and body size in a population of Darwin's Medium Ground finches, Geospiza fortis, associated with adult mortality during a drought. The main cause of mortality was starvation, and differential survival with respect to morphology was at least partly a reflection of differential handling ability of the foods remaining in the environment (Boag and Grant, 1981). Since size measures are highly heritable, with some highly repeatable measures such as beak depth having heritabilities of over .8 (Boag and Grant, 1 978; Boag, 1983), the probability of adult survival with respect to size characters is also expected to be a highly heritable character, provided that adult mortality based on size results from the same general causes in different generations. In this paper we consider the possibility of selection on body size at other stages of the life cycle. Specifically we ask, was the directional selection an isolated event with long term evolutionary consequences, or is it balanced by selection favoring small body size at other stages of the life cycle? To answer this question we examine the possibility of selection on body size associated with juvenile mortality. If selection at this life history stage is found we can follow the reasoning developed above. In particular we can argue that probability ofjuvenile survival is heritable, and genetically negatively covaries with pirobability of adult survival. In doing so, we assume juvenile