Evolution and the consequences of species introductions and deletions

Ecology, July, 1996 by Peter A. Abrams

INTRODUCTION

Ecologists and evolutionary biologists since Darwin have been aware that evolution is of paramount importance in determining the nature of the interactions we observe in ecological communities. In spite of this, there has been surprisingly little progress in understanding the specific features that evolution is likely to produce in ecological communities. Does evolution strengthen or weaken interspecific interactions? Does it stabilize or destabilize the population dynamics of the component species? Does it favor specialization or generalization? convergence or divergence? escalation or stasis? Unfortunately, we do not have more than the beginnings of answers to these questions (see, e.g., Thompson 1994, Vermeij 1994). This is clearly an unsatisfactory state of affairs. Ecologists are increasingly being called upon to predict the consequences of species additions and/or deletions, since these are occurring at a rapid rate, and have often had unanticipated consequences. At least some of the unanticipated consequences may have been the result of evolutionary change. A necessary first step in understanding the consequences of species addition or deletion is the development of theory that identifies the realm of possible consequences, both for trait values and population densities. This should be followed by empirical work to identify which of the possibilities have occurred or are likely to occur in natural communities.

This article reviews two specific questions related to this general theme. The first question is whether we can predict the change in traits related to resource utilization in a given species following the addition or removal of a competitor. This question has probably received more attention than any other question about trait evolution driven by interspecific interactions. The second question is whether we can predict the change in population density of one species in a food web as the result of the addition or removal of another species; here, attention is directed at simple three-species food webs. Attempts to examine the second question for larger food webs are briefly reviewed. In discussing both of these problem areas, the general themes will be that previous theoretical work is likely to have uncovered only a small number of the possible outcomes, and that previous empirical work has concentrated on only a small fraction of the theoretical possibilities that have already been identified. The only type of evolutionary force that is considered here is natural selection. The effects of gene flow, mutation, and genetic drift on ecological interactions have received almost no attention.

THE EFFECT OF AN INTRODUCED COMPETITOR ON TRAITS RELATED TO RESOURCE UTILIZATION

Brown and Wilson (1956) first suggested that the coevolution of competitors could cause the evolutionary divergence of characteristics determining resource use. In theory, divergence is driven by the differential reduction in abundance of those resources most used by the competitor. This may favor a shift in resource use in the focal species that reduces its ability to acquire these resources, if, by such a shift, the focal species can increase its ability to acquire other resources. Strong evidence for the occurrence of divergent character displacement in the field is provided by Grant (1975, 1986), Case (1979), Schluter et al. (1985), Schluter (1986, 1988), Losos (1990), and Schluter and MacPhail (1992), among others. These and other studies are summarized in Taper and Case (1992b). An extensive list of studies suggesting such displacement in fish is provided by Robinson and Wilson (1994), and recent experimental evidence of selection for divergence in sticklebacks in experimental ponds is provided by Schluter (1994). Collectively, this work makes it difficult to argue now, as Arthur (1982) did, that character displacement lacks empirical support.

The emphasis on strengthening the evidence for divergent character displacement has unfortunately shifted attention away from the question of whether divergence is the only possible response of resource utilization traits to the addition of a competing species. Suggestions have repeatedly been made that character displacement may have other forms, such as parallel shifts or convergence (Grant 1972, Pacala 1988, Abrams 1986a, 1987a, b, 1990a, b, Abrams and Matsuda 1994). These have not been accompanied by a correspondingly large number of attempts to find evidence that such nondivergent forms of displacement actually occur. It has become increasingly evident that theoreticians have only uncovered a fraction of the many mechanisms that can result in nondivergent shifts in characters. In the hope of stimulating more empirical study of such possibilities, I review some of these mechanisms, and some of the reasons that they are likely to occur. For each mechanism for displacement following species addition, there is a corresponding mechanism for character release upon the removal of that species.