Evolutionary trade-offs under conditions of resource abundance and scarcity: experiments with bacteria

Ecology, June, 1999 by Gregory J. Velicer, Richard E. Lenski

INTRODUCTION

Evolutionary trade-offs in performance from one environment to another have long been thought to be important in limiting the distribution and abundance of organisms. These trade-offs are the product of genetic, physiological, and material constraints that prevent an organism from simultaneously optimizing different traits (Stearns 1992). Trade-offs have been hypothesized, and sometimes demonstrated, between such traits as maximum growth rate and carrying capacity (MacArthur and Wilson 1967, Pianka 1970, Solbrig and Simpson 1974), longevity and fecundity (Medawar 1952, Rose and Charlesworth 1980), reproduction and growth (Warner 1984, Reznick 1985, Bell and Koufopanou 1986), competitive ability and resistance to exploitation (Lubchenco 1978, Lenski and Levin 1985), and others. Indeed, trade-offs are central to almost all hypotheses concerning the evolution of generalists vs. specialists (Levins 1968, Huey and Hertz 1984, Jaenike 1990, Van Tienderen 1991).

Among these trade-off theories, the notion of r vs. K selection has been perhaps the most historically prominent. K selection is presumed to optimize performance when a population is near its carrying capacity and resources are scarce, whereas r selection occurs when a population is sparse, resources are abundant, and its per capita growth rate is near its maximum (MacArthur and Wilson 1967, Pianka 1970, Roughgarden 1971, Mueller and Ayala 1981). Although most ecologists no longer think that this simple dichotomy provides an adequate classification for life history strategies (Stearns 1992), the more general hypothesis of an evolutionary trade-off in performance under conditions of resource scarcity vs. abundance remains widely held.

Like many of their counterparts who study animals and plants, microbial ecologists often assume that trade-offs in performance under conditions of resource abundance vs. scarcity are pervasive in the organisms that they study (Matin and Veldkamp 1978, Konings and Veldkamp 1980, Kuenen and Harder 1982, Veldkamp et al. 1984, Andrews and Harris 1986, Andrews 1991, Greer et al. 1992). The short generation times of microbes, their large population sizes, and ease of manipulation make them excellent model organisms for examining trade-off hypotheses such as this one. Indeed, in a well-known set of experiments using the bacterium Escherichia coli, Luckinbill (1978, 1984) allowed populations to evolve under two distinct regimes that he thought differed substantially with respect to selection on growth parameters related to resource availability. However, other authors have recently argued that, in fact, the two selection regimes used by Luckinbill were not very different in their selective character, rendering moot his finding against any trade-off (Vasi et al. 1994). Thus, additional experiments are needed to address convincingly the question of whether there is an evolutionary trade-off in the competitive ability of bacteria at high vs. low resource levels. Addressing this issue in a rigorous manner not only is important for ecological theory, but also may be important for identifying those bacterial strains that are the most useful for bioremediation of polluted sites (National Research Council 1993). Can one find a generalist that will efficiently degrade a certain pollutant over a wide range of concentrations? Or must an array of specialists, each adapted to a different substrate concentration, be employed?

In principle, one could test the hypothesis of fitness trade-offs under conditions of resource scarcity vs. abundance by several different methodologies, such as examining the distribution of genetically distinct organisms across habitats that differ in resource availability or comparing relevant life history traits across populations or species. But both these approaches have inherent limitations that weaken the resulting inferences. In the former case, the complexity of ecological habitats and communities means that many variables other than just resource availability must also affect the distribution of organisms. Identifying, and then excluding, variables that confound the hypothesis at hand is, at best, a formidable task. The comparative method is similarly fraught with inferential limitations. These include the problem of distinguishing adaptive from phylogenetic explanations for differences among organisms (Harvey and Pagel 1991) as well as meaningless estimates of performance traits due to differential "pre-adaptation" of organisms to the arbitrary laboratory environment in which they are measured (Mueller and Ayala 1981, Service and Rose 1985). An experimental evolutionary approach circumvents these difficulties. First, ecological conditions are precisely controlled to minimize confounding selective factors between replicates and treatments. Second, comparisons are made between organisms that have known and direct ancestor-descendant relationships. Also, replication of evolutionary treatments allows the extent of heterogeneity in response to selective conditions to be quantified and included in the statistical analysis (e.g., Bennett et al. 1992). Finally, performance traits are measured in environments that correspond to an organism's most recent evolutionary history, except for the specific variable of interest that is manipulated. Therefore, we have used an experimental evolutionary approach to test the hypothesis of a trade-off in performance under ecological conditions of resource scarcity vs. abundance. Populations of bacteria were propagated under each type of condition for many generations, and changes in the relative fitness of the derived strains were then measured by competition experiments, both in the environment in which they had evolved and in the alternative environment. If fitness gains in the selective environment usually correspond with losses in the alternative environment, this would support a fundamental role for trade-offs in adaptation to ecologically distinct environments. OVERVIEW OF THE EXPERIMENTAL DESIGN