The genetics of competition: does resource limitation constrain evolution?

Competition occurs among animals when resources (e.g., food, mates, space) are limited. Although their impacts on growth and reproduction have long interested population biologists, competitive interactions may also have profound, but largely unrecognised, implications for our understanding of how evolution works. This is because while we normally think of competition as arising from purely environmental effects (e.g. lack of food), its outcome (i.e., which animal 'wins') may depend critically on genes. This causes problems for biologists by breaking down the traditional separation of genetic and environmental effects which is the starting point for most studies of evolution. Competition can be thought of as an environmental effect, but we need to acknowledge that it is one with the potential to evolve in its own right. This possibility is not usually accounted for by the theory underpinning evolutionary studies, and may actually have enormous consequences for many fields of pure and applied research - from ecology and animal behaviour, to livestock breeding and the promotion of animal welfare. For example, imagine a gene that influences an animal's growth through effects on foraging - with some versions of the gene (termed 'alleles') making individuals better than average foragers, and therefore likely to grow faster. We might expect that selection of individuals carrying these alleles would result in evolution of faster growth. However, if food is limited, then a gene that increases one individual's growth by allowing it to dominate access to a resource, would actually cause other individuals to get less to eat, and therefore grow more slowly. Seen from this point of view, selection of faster growing individuals would lead to a more competitive environment which would act to counter (and possibly even reverse) our initial expectation that growth rate will increase. This type of process could impose limits on the extent to which traits respond to selection, either natural selection occurring in a wild animal population, or artificial selection being imposed by an animal breeder. Additionally, since competition often involves aggressive behaviours (e.g., animals fighting over a limited resource), any selection that inadvertently leads to a more competitive environment may also be detrimental to animal welfare. Although it is therefore clear that competition could have enormous implications for the evolution of resource-dependent traits, we currently have very little information about how widespread or large these effects actually are. While methodological limitations have restricted progress to date, a number of recently developed statistical approaches mean that these questions can now be addressed. I therefore propose to conduct a comprehensive investigation of the genetics of competition, assessing the role that it plays in shaping, and constraining the evolution of growth and other resource-limited traits. To do this I will take two approaches. Firstly I will perform a series of experiments on a laboratory population of fish (namely the green swordtail, Xiphophorus helleri). This will allow me to directly manipulate key variables such as food availability, the number of competing individuals, and relationships among competitors thereby testing several hypotheses that stem from theoretical predictions. Secondly, I will apply novel statistical methods to test for genetic variation in competitive ability in two commercial livestock populations (Atlantic salmon, chickens). Use of these two high quality data sets will complement the experimental work, and also allow a direct assessment of the applied potential for this direction of research.

If individuals vary in competitive ability, then a limited resource (e.g., food) is likely to be shared unequally among competitors. This will result in increased phenotypic variance for resource-dependent traits (e.g., growth, fecundity). Furthermore, if differences in competitive ability arise from genetic effects (e.g. at loci influencing resource acquisition), then this additional phenotypic variance will include a genetic, and therefore heritable, component. Thus, competition may have important, but rarely considered, consequences for the evolution of resource-dependent traits under selection. Theoretical work has shown that genetic variance arising from competition will generally act as an evolutionary constraint. This is because covariance between direct genetic effects (defined as the impact of an individual's genotype on itself) and indirect genetic effects (defined as the impact of an individual's genotype on the phenotypes of others) is expected to be negative. For example, a gene that increases resource acquisition will increase growth in a focal individual, but will also reduce available resource for competitors, thus having a negatively pleiotropic impact on their growth. Such negative covariance between direct and indirect genetic effects will reduce, and potentially even reverse, an expected response to individual-level selection - whether natural or artificial. This type of effect could have major implications for our understanding of phenotypic evolution, and provides a plausible mechanism for the frequent mismatch between predicted and observed selection responses. However, a lack of empirical scrutiny to date means that we currently know little about the prevalence or magnitude of indirect genetic effects arising from competition. The proposed work would address this empirical gap by developing quantitative genetic models of competition and applying them to data from experimental and commercial animal populations.