THE CAUSES OF NATURAL SELECTION

We discuss the necessary and sufficient conditions for identifying the cause of natural selection on a phenotypic trait. We reexamine the observational methods recently proposed for measuring selection in natural populations and illustrate why the multivariate analysis of selection is insufficient for identifying the causal agents of selection. We discuss how the observational approach of multivariate selection analysis can be complemented by experimental manipulations of the phenotypic distribution and the environment to identify not only how selection is operating on the phenotypic distribution but also why it operates in the observed manner. A significant point of departure of our work from recent discussions is in regard to the role of the environment in the study of natural selection. Instead of viewing the environment as a source of unwanted variation that obscures the relationship between phenotype and fitness, we view fitness as arising from the interaction of the phenotype with the environment. The biotic and abiotic environment is the context that gives rise to the relationship between phenotype and fitness (selection). The analysis of the causes of selection is in essence a problem in ecology. The experimental study of the association between selection gradients and environmental characteristics is necessary to identify the agents of natural selection. We recommend research methods for identifying the agency of selection that depend upon a reciprocity between the observational approach of multivariate selection analysis and the manipulative approach of field experiments in evolutionary ecology.     

THE FUNDAMENTAL THEOREM OF NATURAL SELECTION

The fundamental theorem of natural selection was a valiant attempt by R. A. Fisher to fit a quart into a pint pot
The present author reviewing Population Genetics by Ewens, 1969 a (Edwards, 1969).
Many authors have maintained that the theorem holds only under very special conditions, while only a few (e.g. Edwards, 1967) have thought that Fisher may have been correct – if only we could understand what he meant! It will be shown here that this latter view is correct
Price, 1972.
My own reason for not including a planned chapter on the fundamental theorem is… [ that ] in spite of many efforts I am still not satisfied that I am able to provide an account that does the theorem justice
The present author in the Epilogue to Foundations of Mathematical Genetics (Edwards, 1977).     

THE PHYSIOLOGICAL BASIS OF NATURAL SELECTION AT THE LAP LOCUS

An extensive research program was undertaken to evaluate the contribution of genetic variation at the Lap locus to variation in physiological traits under natural conditions. Rates of carbon and nitrogen metabolism were monitored in a population of the mussel Mytilus edulis near the center of the Lap allele frequency cline on the north shore of Long Island. The goal of this research was to establish whether the previously described genotype-dependent differences in physiological phenotype are meaningful in ecologically relevant circumstances. It was predicted from laboratory studies that, in nature, genotype-dependent differences will exist for rates of nitrogen excretion and that other aspects of the animal's physiology, particularly rates of carbon metabolism, will be unaffected by Lap genotype. Rates of amino acid and ammonia excretion were significantly dependent upon Lap genotype; individuals with the Lap⁹⁴ allele exhibited greater rates of nitrogen loss. These differences among genotypes were most evident in the fall, between September and December. The genotype-dependent component of rates of nitrogen loss were also largest relative to the total rate of excretion during the fall period. As predicted, other aspects of the nitrogen metabolism (acquisition) and rates of carbon metabolism were independent of Lap genotype. There was a striking congruity among a variety of observations that all indicate that phenotypic differences in nitrogen metabolism are the basis of natural selection at the Lap locus in Long Island Sound. Rates of growth were minimal during the fall months (Hilbish, 1985) and mussels are known to lose weight in a genotype-specific manner during this period (Koehn et al., 1980). Rates of elemental gain and loss were summed to produce carbon and nitrogen budgets; these data show the fall to be a period of extended deficit in carbon and nitrogen balance. Genotype-dependent losses of ammonia and amino acids were greatest during the fall months. Finally, selection against the Lap⁹⁴ allele occurs predominantly in the fall (Hilbish, 1985). The data indicate that the depletion of nitrogen resources provides the basis for selection against Lap⁹⁴ genotypes during the fall months.     

ESTIMATING THE FORM OF NATURAL SELECTION ON A QUANTITATIVE TRAIT

The fitness function f relates fitness of individuals to the quantitative trait under natural selection. The function is useful in predicting fitness differences among individuals and in revealing whether an optimum is present within the range of phenotypes in the population. It may also be thought of as describing the ecological environment in terms of the trait. Quadratic regression will approximate the fitness function from data (e.g., Lande and Arnold, 1983), but the method does not reliably indicate features of f such as the presence of modes (stabilizing selection) or dips (disruptive selection). I employ an alternative procedure requiring no a priori model for the function. The method is useful in two ways: it provides a nonparametric estimate of f, of interest by itself, and it can be used to suggest an appropriate parametric model. I also discuss measures of selection intensity based on the fitness function. Analysis of six data sets yields a variety of forms of f and provides new insights for some familiar cases. Low amounts of variation and a low density of data points near the tails of many phenotype distributions emerge as limitations to gaining information on fitness functions. An experimental approach in which the distribution of a quantitative trait is broadened through manipulation would minimize these problems.     

UNIFICATION OF REGRESSION‐BASED METHODS FOR THE ANALYSIS OF NATURAL SELECTION

Regression analyses are central to characterization of the form and strength of natural selection in nature. Two common analyses that are currently used to characterize selection are (1) least squares–based approximation of the individual relative fitness surface for the purpose of obtaining quantitatively useful selection gradients, and (2) spline‐based estimation of (absolute) fitness functions to obtain flexible inference of the shape of functions by which fitness and phenotype are related. These two sets of methodologies are often implemented in parallel to provide complementary inferences of the form of natural selection. We unify these two analyses, providing a method whereby selection gradients can be obtained for a given observed distribution of phenotype and characterization of a function relating phenotype to fitness. The method allows quantitatively useful selection gradients to be obtained from analyses of selection that adequately model nonnormal distributions of fitness, and provides unification of the two previously separate regression‐based fitness analyses. We demonstrate the method by calculating directional and quadratic selection gradients associated with a smooth regression‐based generalized additive model of the relationship between neonatal survival and the phenotypic traits of gestation length and birth mass in humans.     

THE FUNCTIONAL BASIS OF NATURAL SELECTION FOR VERTEBRAL TRAITS OF LARVAE IN THE STICKLEBACK GASTEROSTEUS ACULEATUS

Previous studies have demonstrated selective predation for vertebral traits of larvae in the stickleback Gasterosteus aculeatus. I tested the hypothesis that this selection results from a direct functional advantage to particular vertebral phenotypes by direct measurement of the burst swimming performance of larvae. Within a narrow window of lengths, burst speed did depend on vertebral phenotype. As in the previous predation experiments, performance was related more directly to the ratio of abdominal to caudal vertebrae (VR) than to the total number of vertebrae (VN), and the optimal VR decreased as larval length increased. Changes with length in the vertebral phenotype frequencies of wild larvae provided evidence of selection for VR and for VN in the wild. Larvae with particular VR increased in frequency in the wild at just those lengths when their relative performance was superior in the laboratory. The observed pattern of length-dependent selection for vertebral number provides an explanation for the widespread trends in vertebral number that occur among populations of related fishes.     

PATTERNS OF VARIATION IN THE STELLARIA LONGIPES COMPLEX: EFFECTS OF POLYPLOIDY AND NATURAL SELECTION

Patterns of morphological variation were studied in herbarium specimens of Stellaria longipes, an herbaceous perennial and subsequently in a growth chamber experiment using three cytotypes (4x, 6x, 8x) of S. longipes and diploids of its proposed progenitor S. longifolia. Despite extensive phenotypic plasticity in many traits, patterns of variation resulting from ecotypic differentiation within S. longipes could be detected in the field. A distinct form of S. longipes, which is restricted primarily to arctic and alpine tundra locations, shows genetic differentiation for the following traits: few flowers per ramet, a low proportion of flowering ramets, and ovate leaves. The three cytotypes of S. longipes could be distinguished by their mean genotypic value for leaf length and number of flowers per ramet. The extent of phenotypic plasticity in these traits makes it unlikely that the cytotypes could be distinguished in the field. The direction and extent of morphological divergence between S. longifolia and S. longipes suggest an alloploid origin for S. longipes. Variational trends (among-habitat types and cytotypes) in trait means are similar to those reported previously for the pattern of plasticity. This supports the argument that similar forces guide evolution of the mean and pattern of plasticity of a trait.     

NATURAL SELECTION ON BILL CHARACTERS IN THE TWO BILL MORPHS OF THE AFRICAN FINCH PYRENESTES OSTRINUS

Evidence for natural selection on seven bill and body characters is examined in the two bill morphs of the African estrildid finch Pyrenestes ostrinus. Two regression methods are used in examining natural selection in association with survivorship: a parametric (Lande and Arnold, 1983) and a non-parametric (Schluter, 1988) method. Selection was estimated in adult males, females and juveniles over a four-year period in a population in south-central Cameroon. Selection was common among groups but patterns differed and depended on the method used in detecting selection. The non-parametric method revealed evidence for disruptive selection occurring on bill width and is explained within the context of known feeding efficiencies and the hardness of important seeds in finch diets. Directional selection was common on bill characters in all groups, but infrequent on other characters. There was no evidence of selection on generalized size or shape characters. Selection on bill characters was common across groups despite low annual variation in rainfall. This contrasts with studies of Galápagos finches in which selection is frequently associated with dramatic changes in food supply caused by high variance in annual rainfall. Patterns of selection on bill traits in P. ostrinus also differ from those in song sparrows and Galápagos finches by exhibiting evidence for natural selection on all bill dimensions.     

MUTATION,SELECTION, AND THE MAINTENANCE OF LIFE-HISTORY VARIATION IN A NATURAL POPULATION

In an effort to provide insight into the role of mutation in the maintenance of genetic variance for life-history traits, we accumulated spontaneous mutations in 10 sets of clonal replicates of Daphnia pulex for approximately 30 generations and compared the variance generated by mutation with the standing level of variation in the wild population. Mutations for quantitative traits appear to arise at a fairly high rate in this species, on the order of at least 0.6 per character per generation, but have relatively small heterozygous effects, changing the phenotype by less than 2.5% of the mean. The mean persistence time of a new mutation affecting life-history/body-size traits is approximately 40 generations in the natural population, which requires an average selection coefficient against new mutations of approximately 3% in the heterozygous state. These data are consistent with the idea that the vast majority of standing genetic variance for life-history characters may be largely a consequence of the recurrent introduction of transient cohorts of mutations that are at least conditionally deleterious and raise issues about the meaning of conventional measures of standing levels of variation for fitness-related traits.     

FIELD MEASUREMENTS OF NATURAL AND SEXUAL SELECTION IN THE FUNGUS BEETLE,BOLITOTHERUS CORNUTUS

Selection on three phenotypic traits was estimated in a natural population of a fungus beetle, Bolitotherus cornutus. Lifetime fitness of a group of males in this population was estimated, and partitioned into five components: lifespan, attendance at the mating area, number of females courted, number of copulations attempted, and number of females inseminated. Three phenotypic characters were measured—elytral length, horn length, and weight; there were strong positive correlations among the three characters. Selection was estimated by regressing each component of fitness on the phenotypic traits. Of the three traits, only horn length was under significant direct selection. This selection was for longer horns and was due mainly to differences in lifespan and access to females. The positive selection on horn length combined with the positive correlations between horn length and the other two characters resulted in positive total selection on all three characters.     

COSTS OF REPRODUCTION IN THE WILD: PATH ANALYSIS OF NATURAL SELECTION AND EXPERIMENTAL TESTS OF CAUSATION

During 1991 through 1994, natural selection on reproductive effort in side-blotched lizards was indexed by measuring total clutch mass produced on the first clutch of the reproductive season and assessing how such effort in current reproduction affects subsequent survival and clutch production. In addition, selection was also experimentally assessed in free-ranging female side-blotched lizards by (1) surgically decreasing total clutch mass (direct ovarian manipulation) and enhancing clutch mass using (2) exogenous gonadotropin, and (3) exogenous corticosterone. Surgical reduction of clutch mass uniformly enhanced survival. However, increasing clutch mass had more complex effects depending on year. Experimentally enhanced clutch mass enhanced survival in 1991, had no effect on survival in 1992, and decreased survival in 1993. Despite the complexity of these experimental results, they are corroborated by our comparative data. It is important to note that local environmental effects can obscure detection of costs arising from natural variation in reproductive effort, and we removed such effects using path analysis. The striking shift in natural selection favoring females laying a large clutch mass (1991) to selection against females laying a large clutch mass (1993) is associated with an end of a severe multiyear drought. Our natural-history observations suggest that the correlated increase in predatory snake activity on our study site, coincident with the end of the drought, is the agent of natural selection. Although the actual agents of selection (e.g., snake predation versus drought-related effects) are not resolved, the patterns of natural selection measured in our comparative and experimental data are also consistent with year-to-year changes in clutch mass and egg size that would be indicative of rapid short-term evolution in these traits.