The Earth's biosphere is characterized by an enormous diversity of living things that have evolved across many generations in response to natural selection acting on individuals and populations within ecosystems. The abundances and distributions of these diverse taxa reflect the dynamic mosaic of physical, climatological and ecological features that affect an organism's ability to grow and reproduce. Thus patterns of biodiversity are not static but are changing as the elements of the mosaic vary.
Within the ecological sphere, a recurring phenomenon affecting the configuration of taxa is invasion by individuals into environments and communities where they had previously never existed. Biological invasions have been common in natural ecosystems over evolutionary time and have had significant effects on the composition of and interactions within the resulting communities (Mooney and Drake, 1986). However, the advent of human populations and associated development of mechanized transport between population centers has provided an unparalleled opportunity for organisms to cross physical barriers that would have been impossible to traverse naturally.
Humans, therefore, are not only altering the mechanism for biological invasions but through land use and habitat disturbance are also transforming the physical environment within which invaders act and potentially succeed. This is providing new ecological landscapes for invasions that favor organisms with the attributes that make them successful in disturbed and/or variable systems. The net result, according to Mooney and Drake (1989), is the creation of a "new biological order" (p. 503) with favored organisms being found in increasing abundances all around the globe.
An important trait that may enhance survival of invaders under disturbed conditions is the ability to be phenotypically plastic. Recent investigations have suggested that plasticity itself may be a heritable trait and evolve independently from other traits (Scheiner 1993b). Thus understanding how plasticity may be regulated and acted upon by selection is a crucial step toward understanding how invading organisms can be successful and how the taxa within the invaded communities may subsequently evolve.
Patterns of Biological Invasion
Before humans began moving themselves and cargo over long distances, organisms invaded new habitats through natural dispersal modes such as "hitch-hiking" on birds and animals who seasonally migrated out of their typical range (Mooney &;Drake 1986) or clinging to rafts of vegetation carried by rivers and ocean currents (di Castri 1989). In the last two centuries, however, invading organisms have been able to hitch-hike across entire continents and ocean basins due to intentional or unwitting transport by humans. Today the dominant mode of invasion is through human activity, most notably in ship ballast water (Carlton 1985), adhesion to domesticated animals (Baker 1986), or through intentional introductions for various economic and management reasons (e.g. Ewel 1986).
Upon arrival, invading organisms must be able to utilize the existing resources and produce viable offspring in order to be successful. The characteristics of a successful invader will then depend in large part on the nature of the habitat being invaded. Although several observational and empirical studies of successful invasions have been described (e.g. Williams &;Moore 1989, Elton 1958), there is little consensus on what attributes of a habitat make it more or less invasible. There are, however, a number of ecosystem characteristics which are common to invaded habitats. In general, the probability of successful invasion is related to the extent and type of disturbance, the number of invading propagules deposited per year, the length of time over which invading propagules are deposited, and the amount of biomass or cover in the initial habitat (Rejmanek 1989).
In particular, the level of disturbance (measured as both temporal and spatial stochasticity) may have the most influence over who can successfully invade; consequently, regions of low disturbance will likely be more susceptible to invasion from organisms with one set of traits while regions of higher disturbance may be invaded by organisms with a distinctly different suite of characteristics (Bazzaz 1986). According to Bazzaz (1986), the traits of invading taxa that tend to make them successful in more disturbed habitats are:
1. High population growth rate.
The Role of Phenotypic Plasticity in Successful Invasions
The primary benefit of being phenotypically plastic as an invading taxon lies in the ability of plasticity to broaden the population's niche width and therefore its range of potentially available resources (Mac Nally 1995; Sultan &;Bazzaz 1993; Roughgarden 1972). Particularly in a variable environment, being able to exploit a wider range of nutrients or habitats may afford the invading individuals a selective advantage over the native taxa.
For instance the introduced European wild rabbit has spread over a wide range of climatic regions in Australia, due to the fact that these rabbits are able to change their body shape for thermoregulation and the suggestion that plasticity itself seems to be selected (Williams and Moore 1989). Behavioral plasticity in the timing of migration in the American shad, Alosa sapidissima, allows this anadromous fish species to better respond to warming river temperatures in the Pacific northwest and thus produce more viable offspring than the native sockeye salmon, Oncorhynchus nerka (Quinn and Adams 1996). In addition, the self-pollinating annual grass, Bromus tectorum, was first introduced to the northeast U.S. in 1861 and has since proliferated over a diverse array of grasslands in the intermontane region of the western U.S. Bromus shows significant phenotypic variation in flowering time which allows the grass to persist on adjacent sites that vary widely in soil pH, soil depth, and water-holding capacity (Rice and Mack 1991).
These examples demonstrate how plasticity may improve the fitness of invading populations over native taxa in variable and/or disturbed environments. However, an important question is the degree to which the observed phenotypic plasticity may evolve upon introduction to and persistence in a new habitat and community. Williams and Moore (1989) concluded that the plasticity of body shape for thermal regulation of introduced rabbits in Australia dominated the genetic variation in body shape and that plasticity was apparently maintained in progeny; therefore an adaptive advantage of the phenotypic response seemed likely. Although Williams and Moore did not explicitly test for this, they judged from their empirical data that the plasticity in body shape was heritable and that plasticity could be acted upon by natural selection. Several investigators have pursued this question directly and their results have elicited considerable debate in recent years.
Selection for and Heritability of Plasticity
The existence of phenotypic plasticity as well as its effect on fitness in variable environments has been recognized for many years by numerous investigators (Waddington 1968), however the adaptive nature of plasticity as a trait separate from the particular morphological or behavioral characteristics it describes is a recent notion and one subject to much debate (e.g. Via 1993; Schlichting and Pigliucci 1993; Scheiner 1993a). At issue is whether or not phenotypic plasticity can be acted upon by natural selection as a trait in and of itself and thereby evolve independently from other traits, or if plasticity is simply a by-product of selection on individual traits within and among variable environments. Either result can have significant implications for the rate and nature of evolution in invaded habitats.
Via (1994) contends that for phenotypically plastic organisms stabilizing selection within each distinct environmental state (or at points along an environmental gradient) experienced by the population drives characters toward an environment-specific optimum. These optima can be different in each environment which then can result in reaction norms (the set of phenotypes produced by a genotype over environments) with slopes ‚ 0, i.e. plasticity. In this way, the observation of plasticity is simply a result of the shifts of population means for character states in varying environments, and that a population responds to selection on several different traits simultaneously.
The key to the evolution of plastic populations in Via and others' models (Via and Lande 1985; Gomulkiewicz and Kirkpatrick 1992) is the degree of genetic correlation between the various character states. If there is little or no genetic correlation between character states then there can be independent evolution of the mean phenotype in each environment toward an optimum. However, if there is too much correlation between character states, then an evolutionary change in the mean phenotype in one environment may lead to a correlated change in the expressed phenotype in another environment. This may prevent the population from attaining anything close to the optimum in some environments and may therefore be maladaptive. What may appear to be evolution in the reaction norm is just the expression of varying population means in specific characters or character sets which have evolved more or less independently in different environments.
The more commonly accepted model for the evolution of plasticity takes the opposing view that plasticity is itself a trait that can be measured in the field and the laboratory using statistical regression techniques. Moreover, there can be genetic variation for plasticity measurable by ANOVA on phenotypic variances, where the total phenotypic variance is partitioned into genotypic effects and plastic effects dependent on the environment, and the interaction (G x E) term describes the genetic variability for plasticity (Scheiner 1993b; Scheiner and Lyman 1991; Schlichting 1986). The existence of genetic variability for plasticity itself provides the raw material upon which selection may act to shape the degree of plasticity that is passed on to future progeny, i.e. plasticity can be heritable. No single measure of the heritability of plasticity exists, but instead it can be partitioned into the heritability of environment-independent genetic variation and environment-dependent genetic variation (Scheiner 1993b). Both of these parameters can be measured from the phenotypic variances of plastic traits over an environmental gradient and the variations in the polynomial coefficients of the reaction norm functions across the gradient. Scheiner and Lyman (1989) describe the relationship as:
h2pl = G * -2P
As described above, the ability to be phenotypically plastic as an invading population upon arrival can provide a selective advantage over native organisms. If plasticity is not only a selectable trait but also heritable, then the regulation of plasticity over time and changes in environment as well as the costs of maintaining plasticity become additional concerns.
Regulation and Costs of Plasticity
For plasticity to be adaptive in populations living within variable environments there must be some level of control over the degree and timing of plasticity expressed in and across generations, plus some trade-offs for the costs of maintaining the genetic machinery for plastic response. Schlichting and Pigliucci (1993) assert that phenotypic plasticity is indeed genetically regulated via specific plasticity genes: regulatory loci that exert environmentally dependent control over structural gene expression. These regulatory loci are different from those responsible for the expression of a single trait in any particular environment.
Moran (1992) also describes the regulation of a particular form of phenotypic plasticity called polyphenism, when a single genotype produces two or more distinct phenotypes in response to an environmental signal. Polyphenism is controlled by a developmental switch that is cued through particular changes in the environment. Often the cue is the presence of predators which stimulates a change in the development of an organism, usually to produce protective armor or an alternative behavioral strategy (Pfennig 1992; Dodson 1989). A key determinant to maintaining adaptive plasticity is the accuracy with which factors acting as cues predict the states of factors acting as selective agents (Moran 1992). That is, there must be a certain level of accuracy in matching the appropriate phenotype to the new environment. In addition, once a switch is established it can be made more sensitive by the accumulation of modifiers or the addition of more developmental or behavioral events (West-Eberhard 1989).
Another indirect, but important, control over the expression of phenotypic plasticity is the cost to fitness of being plastic. Here the cost of plasticity is not the cost of having a particular phenotype, but instead the cost is the trade-off between the ability to respond to variation in the environment and other traits for fitness, i.e. the fitness expense of maintaining the mechanism of plasticity versus being non-plastic across variable environments (Newman 1992). Plasticity is a liability if it diverts too much energy toward maintaining the physiological and morphological structures necessary for plasticity at the expense of reproduction or other fitness traits. In addition to the fitness trade-offs, pleiotropic and epistatic effects may limit the response to selection exhibited by one or more alternative phenotypes (Moran 1992).
The fact that phenotypic plasticity persists in many taxa, in particular taxa which have successfully invaded new habitats, suggests that the benefits of plasticity often outweigh the costs in variable environments. Plasticity must be selected for and afford the population an advantage over natives. This raises the issue of how plasticity may evolve and whether or not being phenotypically plastic may affect the rate of evolution in the resulting community.
Evolution of Plasticity
In addition to the debate over the notion of plasticity as an independent trait, there has been further disagreement about whether plasticity itself evolves and how plasticity may affect the overall rate of evolution in a population. The traditional view of plasticity was that it decoupled phenotype from genotype and thereby limited the potential for evolutionary change. Thus, plasticity was considered a constraint on evolution by reducing the impact of natural selection (Thompson 1991). Theorists in the late 1960's also suggested that plasticity was inversely related to genic heterozygosity due either to increased developmental stability (and therefore increased plasticity) caused by deleterious homozygous recessive genes or because a population with well-developed plastic responses has no need for genetic variation and vice versa (Schlichting 1986 and references therein). These ideas led many investigators to believe that plasticity would slow the rate of evolution by shielding the genome from selective pressure.
More recent studies suggest that plasticity may not necessarily limit evolutionary change and can in fact enhance it. Stearns initially believed that plasticity dampened the rate of evolution, however when examining the plastic responses of introduced mosquito fish in Hawaii he found that the most rapidly evolving traits were more plastic than slowly evolving traits (Stearns 1983). Based on his results he changed his opinion and asserted that plasticity was not the constraint on evolution that he once thought. West-Eberhard (1989) also discussed the potential for rapid evolution in plasticity; in particular, behavioral plasticity can evolve quickly due to the lability of behavior and the greater abundance of potential cues for regulating the expression of an immediate behaviorally adaptive response.
Similarly, the relationship between phenotypic plasticity and heterozygosity may be equivocal, as Raimondi (1992) found high heterozygosity among a phenotypically plastic barnacle population introduced to the Salton Sea, while other investigators have found no relationship at all in other organisms (Schlichting 1993 and references therein). Based on the potential for rapid evolution in plasticity and plastic populations as described above, as well as the tenuous relationship of plasticity to heterozygosity, it appears that plasticity may play a more important role in modulating the rate of evolution than previously supposed.
Implications of Plasticity for Biological Invasions
The majority of studies examining the nature of phenotypic plasticity suggest that plasticity is a heritable trait which can respond to selection and evolve separately from the mean values of the trait it describes. In addition, investigation of the characteristics of successful invaders into disturbed and/or variable environments also suggests that plasticity may be an advantage for fitness of invading populations over native organisms. Taken together, the introduction of phenotypically plastic taxa to new habitats may have several important implications for the community post-invasion.
First, being plastic may allow a population of invaders (even one as small as 10-20 individuals - see Mac Nally 1995, chapter 7) to quickly adapt to the new environment and to maintain higher fitness relative to the native taxa, thereby establishing dominance in the community. If the invading population has a broad enough niche width resulting from inherited plasticity and the environment were to stay variable, then the invaders could potentially remain dominant over relatively long time periods. A second and related implication for the invaded community with a dominant and plastic alien population, then, is the potential for preventing new invaders from penetrating or being successful in the habitat by utilizing all or most of the available niche space.
However, if the environmental conditions were to stabilize such that selection favored a particular phenotype or narrower range of phenotypes, then the costs of maintaining plasticity would likely become greater than the received fitness benefits to the introduced population. This could result in either canalization of phenotypes or loss of dominance and perhaps elimination of the plastic population from the community. Canalization could occur by a mechanism similar to that described by Tauber et al (1993) in which some individuals in a newly-established plastic population express characters which make them successful in competing for a native specialist's prey. With heritability and repeatability of those characters this portion of the generalist (plastic) population evolves an adaptation to the specific prey. Finally, gene flow between the original plastic population and the derived specialists is reduced through direct selection on reproductive traits.
The third step in Tauber et al's mechanism, reduced gene flow between the original plastic population and the derived specialists, raises an intriguing implication for a community invaded by a plastic population. West-Eberhard (1989) goes a step further than Tauber et al in her contention that even without a stabilization of the environment phenotypic plasticity may in fact be a diversifying factor in evolution, one that contributes to the origin of novel traits and to altered directions of change. Two aspects of plasticity may facilitate novel phenotypes: the capacity for immediate correlated shifts in related traits, and the occurrence of condition-sensitive expression of phenotypes.
For instance, a single small behavioral change can account for a chain of effects not requiring further genetic change; these changes acting together thus produce a coordinated and adaptive set of phenotypes that can originate rapidly. A condition-sensitive expression of alternative phenotypes means that in a variable environment a novel phenotype can evolve alongside an established specialization without being expressed in the same situations. This can contribute to diversity by increasing the consistency with which particular phenotypes are matched to the contrasting conditions which mold their divergent forms.
West-Eberhard further points to the evolutionary significance of the developmental switch which determines which phenotype may be expressed in a particular environment. This switch permits the persistence and modification of divergent forms, and it means that the divergent forms can become dissociated with little or no genetic change. Ultimately, plasticity can lead to speciation through phenotype fixation, increased divergence due to phenotype fixation, acceleration of reproductive isolation due to divergent specialization, and finally a rapid attainment of compatibility in sympatry among the newly-derived groups (although see Meyer 1987).
The effects of invasion by a phenotypically plastic population of organisms into a variable environment can therefore be quite profound, through potential competition and dominance of the invaders, rapid evolution of the ability to remain plastic and competitive exclusion of new invaders, and potential divergence of the plastic population via canalization or adaptive radiation. Not only may plasticity afford invading taxa an advantage for becoming established into a new and variable environment, but the maintenance of plasticity may significantly change the character of the invaded community for a long period after introduction.
Conclusions
Biological invasion is an increasingly common phenomenon around the globe as human activity provides vectors for dispersal that would otherwise be impossible in the natural system. The activities of humans also tend to increase the level and frequency of disturbance within ecosystems making them more invasible by organisms with particular traits, especially phenotypic plasticity. Although there is still some debate over the mechanics, phenotypic plasticity is generally thought to be a trait independent of the characters it describes, with its own genetic variation. Such variation makes it possible for selection to act directly on plasticity and therefore it may be heritable and can evolve. Plasticity may be controlled directly through the action of regulatory genes which influence the expression of structural genes for plastic traits and/or indirectly when the costs of maintaining the mechanisms for plasticity become greater than the fitness benefits plasticity provides.
Based on these ideas about plasticity, it may be possible to predict whether an introduced plastic population will be successful in a variable or disturbed environment and further how the resulting community may evolve. With the increase in disturbance due to human activity there may indeed be a "new biological order" (Mooney and Drake 1989) developing due to biological invasion, and the introduction of plastic populations may be promoting this phenomenon.
DODSON, S. 1989. Predator-induced reaction norms. Bioscience 39(7): 447-452.
WEST-EBERHARD, M.J. 1989. Phenotypic plasticity and the origins of diversity. Ann. Rev. Ecol. Syst. 20: 249-278.