Get access

Facing change: forms and foundations of contemporary adaptation to biotic invasions


  • Box 1 Sources of phenotypic variation in human-altered environments

    Trait values differ as a result of the environment, as a result of evolved genetic differences, or both. Both environmental and genetic variation can lead to phenotypic changes that are adaptive (or not) in human-altered environments. Fig. 1 shows the influence of such variation on the probability of population persistence (‘+’ positive, ‘–’ negative). Developmental flexibility and behavioural plasticity may enhance the performance of individuals in changed circumstances. For example, threatened prey species may learn to avoid introduced predators by associating novel signals with the threat (Griffin & Evans 2003).

    • image(1)

    [ Fates of populations via individual and evolutionary responses to environmental change. Adaptive responses that increase the probability of population persistence may be induced or genetically based; they may also operate simultaneously, and plasticity may evolve as well (e.g. Carroll & Corneli 1999; Pigliucci 2001). Adaptive change is not guaranteed (Price et al. 2003; Ghalambor et al. 2007). ]

    Nonetheless, cue-dependent trait expression may be disrupted in altered environments. Changed circumstances may also present individuals with signals that induce maladaptive actions or phenotypes. Misleading signals, or ‘evolutionary traps’ (Schlaepfer et al. 2005), range from such phenomena as the attraction of nocturnal insect to artificial lights, to native predators attacking toxic toads introduced to Australia (Phillips et al. 2006) and birds nesting preferentially in less productive habitats formed by invasive plants (Misenhelter & Rotenberry 2000). These examples show that altered habitats induce naïve organisms to make mistakes, though the prospect remains that learning may permit individuals to extract themselves from traps as well. Even when adaptive, plasticity at one juncture may later constrain adaptive options (Weinig & Delph 2001).

    Like environmentally induced phenotypic variation, genetic sources of phenotypic variation may also be a double-edged sword from the standpoint of population persistence (Fig. 1). For example, while differential survival and reproduction in altered habitats may result in adaptive evolution in heritable traits, hard selection may undermine fitness increments by dangerously reducing population size (e.g. Stockwell et al. 2003). In rapidly or unpredictably changing environments, local adaptation may have reduced value with respect to population viability. Indeed, responses to natural selection may be against the interests of the species as a whole (Ferriere et al. 2004). Likewise, mutations may be beneficial or deleterious, but may in general be too infrequent to commonly play a role in tipping an imperiled population toward survival or extinction. In contrast, changes in the frequencies of extant alleles, whether through selection, gene flow or inbreeding, may strongly impact population persistence (e.g. Hedrick & Kalinowski 2000; Garant et al. 2007). Such effects may be positive if they increase the frequencies of beneficial alleles. A particularly interesting concept is the founder-flush model, which postulates that the altered allele frequencies of bottlenecked populations may be magnified through inbreeding to create derivative populations with frequencies far different from those of the parental population (Willis & Orr 1993; Goodnight 2000). Such novel genotypes may theoretically express unprecedented phenotypes resulting from novel gene–gene (dominance, epistasis) interactions, though the conditions in which they are expected to be important are under debate (e.g. Naciri-Graven & Goudet 2003; Turelli & Barton 2006). Unanticipated, recently evolved dominance and epistatic variation is highlighted later in this paper. Likewise, stress-augmented phenotypic (and genotypic) variation may place genes in novel developmental environments and contribute both to reduced fitness and to adaptive evolutionary potential (Badyaev 2005).

  • Box 2 Evolutionary analysis with reciprocal comparisons

    Testing hypotheses about population differences across environments benefits from a focus on traits whose values likely influence fitness in each. Reaction norms measured within populations can be contrasted and compared in several ways to explore the rate and structure of ongoing adaptive evolution.

    inline image

    1 Evolution of reaction norms during the transition from an ancestral phenotype in the original environ-ment to the derivative state in the altered environment. Performance in each environment is indicated by the closed (original environment) or open (altered environment) circles. In this hypothetical example, each ecotype performs better in its current environment, with the arrows linking each to their performance when experimentally exposed to the alternative environment.

    inline image

    2 Three types of inferential links between reaction norms of the two populations in panel A. By itself, the current ecological (or ‘populational’ or ‘phenotypic’) contrast is devoid of genetic information. The simplest explanations for the performance similarity between environments are that the phenotype is strongly canalized, or the environments differ trivially. The evolutionary path shows performance is not canalized, and reveals the true course of evolution. The reciprocal cross shows that the loss of performance of the derived population in the original environment (tradeoff) is at least as great as the increment in the new. (Based on Carroll et al. 2001.)

    inline image

    3 Within and among traits that are evolving in soapberry-bug host races, the rates (in haldanes) and structure of evolutionary change differ substantially. The colours scheme for the evolutionary path (blue), current ecological contrast (yellow) and evolved tradeoff (red) matches that of panel B above. Beak length in the ancestral population is relatively unaffected by the host, but is much shorter in the derived population, especially in bugs reared on the native host. Thus the evolved tradeoff slightly exceeds the evolutionary path, and both exceed the current ecological contrast. Body size (measured by thorax width) is smaller in the derived race, but no more so than the decrement from rearing the ancestral race on the introduced host. A reciprocal host effect in the derived race means that the tradeoff appears to have evolved, in contrast, substantially. For development time, current values are very similar, but host-based tradeoffs in both populations reveal that substantial evolution has taken place to restore a development time close to the original. Without reciprocal rearing, that ‘cryptic’ counter-gradient evolution would be undetectable.

Scott P. Carroll, Fax: 530-297-6080; E-mail:


Ongoing adaptation in native populations to anthropogenic change both facilitates and challenges ecologically appropriate and sustainable management. Human disturbance promotes adaptive responses at the genomic, individual and population levels. Traits vary widely in whether adaptation occurs through plasticity or evolution, and these modes interact within and among traits. For example, plasticity in one trait may be adaptive because it permits homeostasis and lessens the intensity of selection in another. Both opportunity and catastrophe generate adaptive responses. Recently evolved adaptations characterize the responses of many native species to biotic invasions. Several well-known examples involve native phytophagous insects colonizing introduced plants. For example, our studies of North American and Australian soapberry bugs on nonindigenous plants demonstrate both diversifying and homogenizing contemporary evolution. Modes of adaptation differ among traits and populations and as a function of the host on which they develop. The genetic architecture of the evolving adaptations involves a substantial degree of nonadditive genetic variation. One important consequence of contemporary adaptation may be an enhanced capacity of native communities to provide adaptive biological control of invasive species. Conservation scientists may manipulate adaptation to achieve conservation goals, but must also decide how deeply they wish to attempt to control the phenotypes and genotypes of other species.