Search

In natural biological communities, species interact with many other species. Multiple species interactions can lead to indirect ecological effects that have important fitness consequences and can cause nonadditive patterns of natural selection. Given that indirect ecological effects are common in nature, nonadditive selection may also be quite common. As a result, quantifying nonadditive selection resulting from indirect ecological effects may be critical for understanding adaptation in natural communities composed of many interacting species. We describe how to quantify the relative strength of nonadditive selection resulting from indirect ecological effects compared to the strength of pairwise selection. We develop a clear method for testing for nonadditive selection caused by indirect ecological effects and consider how it might affect adaptation in multispecies communities. We use two case studies to illustrate how our method can be applied to empirical data sets. Our results suggest that nonadditive selection caused by indirect ecological effects may be common in nature. Our hope is that trait-based approaches, combined with multifactorial experiments, will result in more estimates of nonadditive selection that reveal the relative importance of indirect ecological effects for evolution in a community context.

The study of evolutionary constraint is an active and important area, and genetic correlations and quantitative genetic techniques more generally have been the dominant approach to constraint. Here, I argue that genetic correlations are not very useful for studying constraint, review recent alternative approaches, and briefly discuss the state of our knowledge of the evolutionary importance of constraints caused by genetic variance and covariance.

Genetic correlations are the most commonly studied of all potential constraints on adaptive evolution. We present a comprehensive test of constraints caused by genetic correlation, comparing empirical results to predictions from theory. The additive genetic correlation between the filament and the corolla tube in wild radish flowers is very high in magnitude, is estimated with good precision (0.85 ± 0.06), and is caused by pleiotropy. Thus, evolutionary changes in the relative lengths of these two traits should be constrained. Still, artificial selection produced rapid evolution of these traits in opposite directions, so that in one replicate relative to controls, the difference between them increased by six standard deviations in only nine generations. This would result in a 54% increase in relative fitness on the basis of a previous estimate of natural selection in this population, and it would produce the phenotypes found in the most extreme species in the family Brassicaceae in less than 100 generations. These responses were within theoretical expectations and were much slower than if the genetic correlation was zero; thus, there was evidence for constraint. These results, coupled with comparable results from other species, show that evolution can be rapid despite the constraints caused by genetic correlations.

Although conflicting selection from different resources is thought to play a critical role in the evolution of specialized species, the prevalence of conflicting selection in generalists is poorly understood. Plants may experience conflicting selection on floral traits by different pollinators and between genders. Using artificial selection to increase phenotypic variation, we tested for conflicting and nonadditive selection on wild radish (Raphanus raphanistrum) flowers. To do this, we measured selection by each of the major pollinator taxa through both male and female fitness, and tested for a single-generation response to selection by a subset of these pollinators. We found some evidence for conflicting selection on anther exsertion—-sweat bees exerted stabilizing selection and larger bees selected for increased exsertion. Stamen dimorphism was only under selection by honey bees, causing a response to selection in the next generation, and flower size was under similar selection by multiple pollinators. Selection differed significantly between genders for two traits, but there was no evidence for stronger selection through male fitness or for conflicting selection between genders. Our results suggest wild radish flowers can adapt to multiple pollinators, as we found little evidence for conflicting selection and no evidence for nonadditive selection among pollinators.

†Background There are a number of difficulties associated with the study of adaptation. One is a lack of variation in the trait, which is common in adaptations because past selection has removed unfit variants. This lack of variation makes it difficult to determine the relationship between trait variation and fitness. Another difficulty is proving causation in this trait‐fitness relationship, because a correlated trait might be the actual adaptation. These difficulties can be ameliorated at least partially by combining studies of natural variation with studies of experimentally manipulated traits and traits whose variance has been augmented by artificial selection. †Scope We review here a number of our studies on the adaptive value of two aspects of anther position in wild radish (Raphanus raphanistrum, Brassicaceae): anther exsertion, i.e. the degree to which anthers protrude from the mouth of the corolla tube, and anther height dimorphism, i.e. the difference in lengths of the filaments between the two short and four long stamens. We have used both functional analyses, in which the response variable is pollen removal, and measurements of selection, in which the response variable is lifetime male fitness estimated by molecular genetic paternity analyses. In these studies we use both the natural variation in populations as well as manipulated variation, the latter through both stamen removal and artificial selection, to recreate the ancestral trait conditions. †Conclusions Our work provides convincing evidence that intermediate anther exsertion values are adaptive, and that this is probably an adaptation to a subset of the pollinator fauna, small bees. The picture for anther height dimorphism is much less clear, as the weight of current evidence suggests that current values of this trait might actually be maladaptive; however, if this is true it is difficult to understand how the dimorphism is maintained across the family Brassicaceae.

Kingsolver et al.'s review of phenotypic selection gradients from natural populations provided a glimpse of the form and strength of selection in nature and how selection on different organisms and traits varies. Because this review's underlying database could be a key tool for answering fundamental questions concerning natural selection, it has spawned discussion of potential biases inherent in the review process. Here, we explicitly test for two commonly discussed sources of bias: sampling error and publication bias. We model the relationship between variance among selection gradients and sample size that sampling error produces by subsampling large empirical data sets containing measurements of traits and fitness. We find that this relationship was not mimicked by the review data set and therefore conclude that sampling error does not bias estimations of the average strength of selection. Using graphical tests, we find evidence for bias against publishing weak estimates of selection only among very small studies (N38). However, this evidence is counteracted by excess weak estimates in larger studies. Thus, estimates of average strength of selection from the review are less biased than is often assumed. Devising and conducting straightforward tests for different biases allows concern to be focused on the most troublesome factors.

The evolutionary mechanisms underlying the maintenance of invariant traits are poorly understood, partly because the lack of variance makes these mechanisms difficult to study. Although the number of cotyledons that plant species produce is highly canalized, populations of plants frequently contain individuals with abnormal cotyledon numbers. In a garden study with 1857 wild radish plants from 75 paternal half-sibling families, 89 (almost 5%) had cotyledon numbers less or greater than two. We found evidence for direct selection on cotyledon number, but no evidence for additive genetic variation for cotyledon number. In spite of the very large sample size, our power to detect variation and selection was hampered by the small number of individuals (10) producing more than two cotyledons. Thus, our results provide support for both a lack of genetic variation and selection as reasons for the current lack of variation in wild radish cotyledon number.

Summary • How species evolve depends on the communities in which they are embedded. Here, we briefly review the ideas underlying concepts of diffuse coevolution, evolution, and selection. • We discuss criteria to identify when evolution will be diffuse. We advocate a more explicitly trait-oriented approach to diffuse (co)evolution, and discuss how considering effects of interacting species on fitness alone tells us little about evolution. We endorse the view that diffuse evolution occurs whenever the response to selection by one interacting species on a given trait is altered by the presence of a second interacting species. • Building on the work of others, we clarify and expand the criteria for diffuse evolution and present a simple experimental design that will allow the detection of diffuse selection. • We argue that a greater focus on selection on specific traits and the evolutionary response to that selection will improve our conceptual understanding of how communities affect the evolution of species embedded within them.

Plant defence against any type of stress may involve resistance (traits that reduce damage) or tolerance (traits that reduce the negative fitness impacts of damage). These two strategies have been proposed as redundant evolutionary alternatives. A late-season frost enabled us to estimate natural selection and genetic constraints on the evolution of frost resistance and tolerance in a wild plant species. We employed a genetic selection analysis (which is unbiased by environmental correlations between traits and fitness) on 75 paternal half-sibling families of annual wild radish [Raphanus raphanistrum (Brassicaceae)]. In an experimental population in southern Ontario, we found strong selection favouring plant resistance to frost, but selection against tolerance to frost. The selection against tolerance may have been caused by a cost of tolerance, as we provide evidence for a negative genetic correlation between tolerance and fitness in the absence of frost damage. Although we found no evidence for the theoretically predicted trade-off between frost tolerance and resistance among our families, we did detect negative correlational selection acting on the two traits, indicating that natural selection favoured high resistance combined with low tolerance and low resistance coupled with high tolerance, but not high or low levels of both traits together. There were few genetic correlations between the measured traits overall, but frost tolerance was negatively correlated with initial seed mass, and frost resistance was positively correlated with resistance to insect herbivory. Periodic episodes of strong selection such as that caused by the late-season frost may be disproportionately important in evolution, and are likely becoming more common because of human alterations of the environment.

Artificial selection has been practiced by humans since the dawn of agriculture, but only recently have evolutionary ecologists turned to this tool to understand nature. To perform artificial selection, the phenotypic trait of interest is measured on a population, and the individuals with the most extreme phenotypic values are bred to produce the next generation. The change in the mean of the selected trait across each generation is the response to selection, and other traits can also evolve due to genetic correlations with the selected trait. Artificial selection can directly answer the question of how quickly a trait will evolve under a given strength of selection. This kind of result can help ecologists determine whether range or niche boundaries are determined by a lack of variation for a key phenotypic trait or trade-offs due to genetic correlations with other fitness-related traits. In a related approach, controlled natural selection, the organisms are not selected according to their values for a given trait, but rather are allowed to evolve for one to several generations under experimentally imposed environmental treatments such as temperature, light, nutrients, presence or absence of predators or competitors, etc. The results of this kind of study can tell us how quickly a population can adapt to a given environmental change, either natural or anthropogenic. Finally, artificial selection can create more variation for measurements of natural selection or can be coupled with QTL mapping; both these combinations provide new insights into adaptation. I discuss advantages and disadvantages of these approaches relative to other kinds of studies and highlight case studies showing how these tools can answer a wide range of basic and applied questions in ecology, ranging from niche and range boundaries and character displacement to climate change and invasive species.

Plant traits that show little variation across higher taxa are often used as diagnostic traits, but the reason for the stasis of such traits remains unclear. Wild radish, Raphanus raphanistrum, exhibits tetradynamous stamens (four long and two short, producing a dimorphism in anther height within each flower), as do the vast majority of the more than 3,000 species in the Brassicaceae. Here we examine the hypothesis that selection maintains the stasis of dimorphic anther height by investigating the effects of this trait on pollen removal, seed siring success, and seed set in R. raphanistrum using both experimental and observational methods. Observational selection gradient analysis based on lifetime seed siring success provided evidence for an optimum dimorphism that was greater than zero in one of three years. In both experimentally manipulated and unmanipulated flowers, more pollen was removed in single visits from flowers with less dimorphism. There was no significant effect of anther dimorphism on female fitness (seed set). Therefore, there is some evidence to suggest that selection is maintaining anther dimorphism in wild radish, and that higher male fitness might result from restriction of single-visit pollen removal. We discuss these results in light of pollen presentation theory.

Genetic correlations among traits are important in evolution, as they can constrain evolutionary change or reflect past selection for combinations of traits1,2. Constraints and integration depend on whether the correlations are caused by pleiotropy or linkage disequilibrium3, but these genetic mechanisms underlying correlations remain largely unknown in natural populations4. Quantitative trait locus (QTL) mapping studies do not adequately address the mechanisms of within-population genetic correlations because they rely on crosses between distinct species, inbred lines or selected lines (see ref. 5), and they cannot distinguish moderate linkage disequilibrium from pleiotropy because they commonly rely on only one or two episodes of recombination6. Here I report that after nine generations of enforced random mating (nine episodes of recombination), correlations between six floral traits in wild radish plants are unchanged, showing that pleiotropy generates the correlations. There is no evidence for linkage disequilibrium despite previous correlational selection acting on one functionally integrated pair of traits7. This study provides direct evidence of the genetic mechanisms underlying correlations between quantitative traits in a natural population and suggests that there may be constraints on the independent evolution of pairs of highly correlated traits.

To study the genetic architecture of anther exsertion, a trait under stabilizing selection in wild radish, artificial selection on anther exsertion was applied for 11 generations. Two replicate lines each of increased and decreased exsertion plus two randomly-mated controls were included. Full pedigree information is available from generation five. To estimate correlated responses to selection, 571 plants from all lines and matrilines were grown in the greenhouse and a number of floral, growth, and phenology traits were measured. To create an outbred F2 mapping population, all possible crosses among the two high and two low exsertion lines were made, using a multiple-family design to capture the genetic variance still present after 11 generations of selection. Six floral traits were measured on 40 parents, 240 F1, and 4,868 F2 offspring. Opportunities for reuse of these data include traits not previously analyzed, other analyses, especially using the pedigree and fitness data, and seeds from all generations and photos of flowers in the later generations are available.

We present an analysis of Raphanus raphanistrum and simulations illustrating the utility of directly estimating male phenotypic selection gradients using genetic markers. The method offers a much more refined characterization of selection than attempting to assign paternity to individual progeny. Our analysis of R. raphanistrum reveals selection on remarkably fine features of floral morphology, including anther exsertion, that were opaque to previous approaches. The new results also undermine a previous conclusion that selection on wild radish floral morphology acts primarily through female fitness. Simulation results show that selection gradients on the order of beta = 0.1-0.2 can be readily detected with allozyme markers in moderate-sized (200 paternal individuals) populations. Highly polymorphic (e.g., microsatellite) markers will likely detect fine scale selection (beta0.1) in larger populations (or = 400 individuals). Increased progeny sample size, by sampling either additional maternal families or more progeny per maternal parent, partly compensates for low exclusion probability. Increasing the number of possible fathers without changing progeny sample size decreases the ability to detect selection, especially at lower exclusion probabilities. Sampling only some male genotypes reduces the power to detect selection and biases (underestimates) the magnitude of the selection gradient estimate.

The goal of this special issue is to survey the current state of our knowledge of natural selection in plants. Do we need yet another compilation inspired by the sesquicentennial of The Origin (Darwin 1859)? I believe we do, because to my knowledge no other Darwin-year collection has focused on natural selection in plants, and plants were crucially important to Darwin’s development of the theory of natural selection (six of his books were on plants). Plants are still crucially important to the study of natural selection in the field (where all selection occurs) because their sessile nature means that individuals can be followed either in their natural locations or after transplantation to new habitats. Plants are also very well suited to experimental evolution because it is a simple matter to store past generations as seeds. This storage ability also means that plants could be used to make detailed studies of present-day rapid evolution, especially in response to anthropogenic environmental changes, if seeds from natural populations are systematically collected and stored now for future studies (Franks et al. 2008). We conduct our survey of natural selection in plants through a mix of empirical reviews, conceptual reviews, and original research by established leaders in the field as well as younger investigators. We have papers addressing natural selection using both phenotypic and molecular approaches. Latta opens the issue with an overview of the central and closely related concepts of selection, adaptation, and variation. It is useful to begin this way for anyone not already very well-versed in these areas, as the literature is unfortunately rife with misunderstandings. Apparently, natural selection is difficult to understand deeply, despite its inherent simplicity. While the contributors were picked solely to represent excellence across a broad spectrum and were encouraged to write about whatever topics they were most interested in, some themes emerged: Selection on floral traits. The proportion of papers addressing selection on floral traits in the issue (Caruso et al., Ellis and Johnson, Gomez and Perfectti, Horvitz et al., Sletvold and Agren) is surely higher than the overall importance of flowers to plant adaptation. This overemphasis on flowers is not just due to my personal bias, because studies of flowers are overrepresented among all studies of natural selection in plants. This may reflect human affinity with beauty (perhaps the only good explanation why we have two articles addressing selection on spur lengths in orchids), but I think more important is the relative ease of identifying and measuring the key traits and studying their functional interactions with the main selective agents (pollinators). These tasks are more difficult for other plant traits such as physiology and vegetative morphology. However, given their importance to adaptation and survival in sessile organisms, more studies of selection on these latter traits are needed; for an example, see the contribution by Mazer et al. Nonpollinator sources of selection on floral traits. Most studies of floral evolution assume that pollinators are the only selective agent, but this assumption has been called into question by recent studies demonstrating non-pollinatormediated selection on floral traits (e.g., Galen and Cuba 2001; Irwin et al. 2003). Two articles in the issue (Caruso et al., Sletvold and Agren) use a relatively new technique to test this assumption and isolate pollinator-mediated selection. This is done by comparing selection on plants that are saturated with pollen experimentally, which represents selection by agents other than pollinators, to selection on naturally pollinated plants (often called ‘‘open pollinated’’), which represents selection due to all selective agents, including pollinators. The difference between these two is inferred to be pollinator-mediated selection. Both articles found evidence for both pollinatorand non-pollinator-mediated selection on floral traits, although orchid spur length was under selection only by pollinators. The relative importance of pollinatorversus non-pollinator-mediated selection on floral traits is an entirely empirical question, and these articles add substantially to our meager knowledge in this area. In a related vein, Mazer et al. apply the non-pollinator-mediated selection idea to the evolution of mating systems, historically one of the most active areas of research in plant evolution. They provide evidence that selfing evolves as a consequence of selection for shortened life cycles in seasonally dry environments. Spatial and temporal variation in selection. One of the more difficult challenges in understanding natural selection and adaptation in the wild is dealing with the high degree of spatial and temporal variation in selection. Both biotic and abiotic selective agents can vary widely within and among years and over short distances. Sletvold and Agren report significant differences in selection across years and populations. Gomez and Perfectti show that spatial variation in selection on corolla shape matches patterns of preferences and abundances of the different taxa of pollinators and that these differences have resulted in local adaptation in corolla shape. This further suggests that the spatial variation is greater than the temporal variation within populations, as this is necessary for the evolution of local 1 E-mail: connerj@msu.edu.

The matrix diagonalization methods advocated by Blows have a number of statistical advantages, especially for the study of natural selection, but these advantages are usually outweighed by the disadvantage that the results are not very biologically interpretable. This lack of interpretability has been the major impediment to the adoption of these methods over the past 15 years. Blows (2007) makes a number of valuable points that deserve wider attention from evolutionary biologists. The paper advocates a more wholly multivariate approach to understanding adaptive evolution of phenotypic traits as an alternative to the now-standard selection gradient and G-matrix analysis. In the standard analysis, multiple regression techniques are used to measure selection on individual traits, correcting for correlations among all traits included in the analysis. With this correction, selection gradients are an extremely useful way of studying adaptive evolution, as they aim to measure direct selection after removing indirect selection caused by phenotypic correlations. More ambitious studies using the standard analysis include both linear terms to estimate the strength of directional selection, and quadratic (squared) terms to estimate the degree of curvature in the fitness function, which might indicate stabilizing and disruptive selection if the maximum or minimum fitness value occurs at intermediate phenotypic values. Still more ambitious studies include the cross-product terms in the regression to measure correlational selection, that is, whether pairs of traits interact to determine fitness in ways not described by the linear and quadratic terms alone. This accumulation of terms in the regression model is perhaps the biggest problem with standard selection gradient analyses, and thus the biggest advantage of the multivariate methods like canonical analysis (one kind of diagonalization) advocated by Blows. To estimate all the quadratic and correlational selection gradients requires a large number of predictor variables in the analysis with only a modest number of traits. With n traits, the number of predictors in the full model (all linear, quadratic and cross-product terms) is 2n+[n(n)1)]/2, so with only five traits there are 20 variables in the model! With canonical analysis, this number is reduced to 2n, because the number of new variables created is the same as the number of original variables in the analysis, and there is a linear and quadratic term for each. The second term in the formula above represents the correlational selection estimates, which drop out when new axes are created that are orthogonal (i.e. uncorrelated). The main reason that correlational terms are often not included in selection gradient analyses (Kingsolver et al., 2001) is not their difficulty in interpretation (more on this below), but rather that people rarely have the large sample sizes needed to fit so many variables in their regression models. Thus, a method that allows people to explore the multivariate nature of selection with the more modest sample sizes of most studies is a real advantage. Therefore, the multivariate methods that Blows champions will prove to be useful additions to the evolutionary biologist’s toolbox in some cases, providing new insight not available from standard selection gradient analysis. However, it is not yet clear how often this will be the case, and these methods will not replace the nowfamiliar selection gradient and G-matrix analysis. This is not just due to tradition; there are good reasons why the selection gradients and G-matrix framework has become one of the most widely adopted in all of evolutionary biology (allowing meta-analyses like that of Kingsolver et al., 2001): the approach can be undertaken by empiricists with modest quantitative skills, and more importantly it produces results that are eminently interpretable, leading to robust conclusions about the nature of the adaptive process. Selection gradients are often presented in standardized form, that is, in standard deviation units, making them readily comparable across different traits and studies. This makes it possible to make comparisons between, for example, the strength of direct selection on tarsus length in different bird species, or to compare the strength of selection on morphological vs. life-history traits (Kingsolver et al., 2001). Blows performs a good service by reminding us of the potential pitfalls of selection gradient analysis; however, as long as practitioners are aware of these and exercise due caution in analysis and interpretation, they are almost never a problem. The diagonalization techniques advocated by Blows will not replace selection gradient and G-matrix analyses because they have a fundamental problem, which is pointed out at the very end of Blows’ article: the difficulty of biological interpretation. This is the Achilles’ heel of these methods, because the central goal of selection and G matrix analyses has to be biological understanding. Once new axes through multivariate space are created for statistical, rather than biological reasons (for example, to maximize variance explained and to make axes orthogonal), then it is often very difficult to understand what these axes mean in terms of adaptation or organismal function (Endler, 1986, p. 192; Correspondence:Jeffrey K. Conner, Kellogg Biological Station, 3700 East Gull Lake Drive, Hickory Corners, MI 49060, USA. Tel.: 269-671-2269; fax: 269-671-2104; e-mail: connerj@msu.edu

I review my research on the evolution of floral morphology in wild radish, Raphanus raphanistrum. A comparative study of patterns of phenotypic correlations among floral traits in five species indicated that selection may have acted to increase the correlation between filament and corolla tube lengths in wild radish and two other species. This high correlation maintains a similar position of the anthers relative to the opening of the corolla tube, regardless of flower size, which may be important for effective pollination. Quantitative genetic analysis shows that the high phenotypic correlation between these two traits is caused by a high genetic correlation. Preliminary results from a long-term random mating study suggest that this genetic correlation is a result of pleiotropy, not linkage disequilibrium. Measurements of the effects of anther position on pollen removal by pollinators and estimates of natural selection on floral morphology failed to produce evidence for selection to increase the correlati...

The distribution of effect sizes of adaptive substitutions has been central to evolutionary biology since the modern synthesis. Early theory proposed that because large-effect mutations have negative pleiotropic consequences, only small-effect mutations contribute to adaptation. More recent theory suggested instead that large-effect mutations could be favoured when populations are far from their adaptive peak. Here we suggest that the distributions of effect sizes are expected to differ among study systems, reflecting the wide variation in evolutionary forces and ecological conditions experienced in nature. These include selection, mutation, genetic drift, gene flow, and other factors such as the degree of pleiotropy, the distance to the phenotypic optimum, whether the optimum is stable or moving, and whether new mutation or standing genetic variation provides the source of adaptive alleles. Our goal is to review how these factors might affect the distribution of effect sizes and to identify new research directions. Until more theory and empirical work is available, we feel that it is premature to make broad generalizations about the effect size distribution of adaptive substitutions important in nature.

The strength of selection in nature has long been a controversial subject, partly because there were few quantitative measurements of phenotypic selection available until recently. In a new paper, Kingsolver and colleagues reviewed 63 studies and found that the median standardized directional selection gradient (a measure of the strength of phenotypic selection) was 0.16. Whether this means selection in nature is strong or weak depends both on one's point of view and on the error in selection estimates.

Underlying developmental and genetic relationships cause positive correlations among the sizes of a variety of plant traits. Selection for functional independence among traits, however, can reduce these correlations over evolutionary time. In 1960, R.L. Berg hypothesized that the sizes of flowers in insect-pollinated plants should be selected to remain constant regardless of the size of vegetative structures, so that flowers match the sizes of their pollinators for effective pollination. This hypothesis of functional independence of floral sizes from the size of the rest of the plant predicts that correlations between floral and vegetative traits should be reduced relative to correlations within trait groups. We measured correlations in five species of insect-pollinated plants, including four Brassicaceae and Phlox divaricata. Our results support the hypothesis. The correlations among floral traits and the correlations among vegetative traits were significantly greater than the correlations across these two groups of traits in all five species. Keywords: phenotypic correlations, natural selection, functional independence, pollination, floral evolution.

It has often been suggested that selection on floral traits in hermaphroditic plants should occur primarily through differences in male fitness. However, measurements of selection on floral traits through differences in lifetime male fitness have been lacking. We measured selection on a variety of wild radish floral traits using lifetime male fitness measures derived from genetic paternity analysis. These male fitness estimates were then combined with estimates of lifetime female fitness of the same plants to produce measurements of selection based on lifetime total fitness. Contrary to the prediction above, there was no strong evidence for selection on floral morphology through male fitness differences in any of the three years of the study, but there was strong selection for increased flower size through female fitness differences in one year. The main determinant of both male and female fitness in all years was flower number; this lead to moderately positive correlations between male and female fitness in all three years.

To determine whether genetic constraints on adaptive evolution were operating in a laboratory population of a flour beetle, Tribolium castaneum, we first estimated the direct selection acting on each of several body size traits. Strong selection in males for an increase in pupal weight and a decrease in the ratio of adult to pupal weight occurred. In addition, a non-significant trend for a decrease in adult width was found. No significant selection on females was detected, although there were trends toward an increase in pupal weight and a decrease in adult width. These estimates were then combined with estimates of the genetic variances and covariances of the traits to predict the multivariate response to selection, that is, the evolutionary change in the traits across one generation. These projections showed only a small predicted change in male pupal weight in spite of the strong selection on pupal weight, and a relatively large predicted increase in width in spite of the possible negative direct selection on this trait. Both of these results were due in part to the positive genetic covariance between pupal weight and width, and they therefore suggest the possibility of genetic constraints on adaptive evolution of these traits.

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.

Background A mismatch has emerged between models and data of host-parasite evolution. Theory readily predicts that parasites can promote host diversity through mechanisms such as disruptive selection. Yet, despite these predictions, empirical evidence for parasite-mediated increases in host diversity remains surprisingly scant. Results Here, we document parasite-mediated disruptive selection on a natural Daphnia population during a parasite epidemic. The mean susceptibility of clones collected from the population before and after the epidemic did not differ, but clonal variance and broad-sense heritability of post-epidemic clones were significantly greater, indicating disruptive selection and rapid evolution. A maximum likelihood method that we developed for detecting selection on natural populations also suggests disruptive selection during the epidemic: the distribution of susceptibilities in the population shifted from unimodal prior to the epidemic to bimodal after the epidemic. Interestingly, this same bimodal distribution was retained after a generation of sexual reproduction. Conclusion These results provide rare empirical support for parasite-driven increases in host genetic diversity, and suggest that this increase can occur rapidly.