The interplay between plasticity and evolution in response to human-induced environmental change

Climate change

Humans have made lasting impacts on the environment for thousands of years, from megafaunal extinctions during the late Quaternary period³³ to rapid industrialization and global temperature rise within the last century³⁴. The magnitude of recent human-induced changes, particularly to climate, has already led to the global redistribution of plants and animals in space and time^35,36. For those organisms that cannot get out of the way of climate change, they must rely instead on plastic and evolved responses. A major focus has been on whether current amounts of phenotypic plasticity and evolutionary potential are sufficient to keep pace with projected climate velocities^37,38. Of course, there are many effects of contemporary global climate change—from atmospheric deposition of elemental compounds to shifts in precipitation patterns—but temperature is perhaps the best studied and one that impacts organisms from biochemical rate processes up through species interactions, community dynamics, and ecosystem function³⁹.

By the end of the century, global temperature is expected to rise by over 4°C under the highest greenhouse gas emissions scenarios, which we are currently tracking³⁴. Studies of organismal capacities to cope with this change, either through plasticity or evolution, yield mixed evidence for whether organismal changes will be large enough and proceed fast enough to keep pace with climate change^18,38,40,41. In general, there has been a strong research emphasis on existing plasticity in phenology, morphology, physiology, and behavior, perhaps because of the assumption that plasticity will operate over a timetable commensurate with rapid global climate change and perhaps also stemming from the added challenges of assessing evolutionary responses⁴². This emphasis is not without merit, as plasticity has been broadly shown to be a critical component of the response to climate change⁴³. However, existing plasticity is not always sufficient to cope with climate change. For example, Anderson and colleagues found a substantial role for phenological plasticity in plant flowering time under recent climate change but concluded that plasticity alone is insufficient and evolutionary change is needed for these populations to keep pace⁴⁴. While such variance partitioning studies of responses to climate change among plastic and evolved components is growing, there remain comparatively few studies assessing how this plasticity shapes evolutionary responses³.

Those studies that have explored the role of plasticity in facilitating adaptive evolutionary responses to climate change have found little evidence in support of this mechanism, though this may be as much a product of the organisms and systems for which these data are available as it is a general biological pattern. For example, research on lizard responses to climate change from Buckley and colleagues indicates that plasticity in thermoregulatory behavior constrains lizard evolutionary responses to warming by shielding variation in thermal performance from selection⁴⁵. As a notable contrast, Logan and colleagues found evidence of selection on thermal performance traits when lizards were transplanted to a novel thermal environment⁴⁶, but plasticity’s role in facilitating or constraining evolutionary change was not reported. It is worthwhile to consider why there are so few studies on genetic accommodation under climate change. One important limitation may be how the ancestral and novel populations are assessed under a temporal change in climate. Resurrection studies or studies which compare the leading edge of a rapid range expansion to the center range seem to be strong candidates for such an analysis⁴⁷. Next, we consider two environmental changes—land-use and pollution—where the ancestral and novel populations are readily accessible.

Land-use change

Although there are many types of land-use changes occurring, including deforestation and habitat fragmentation, to which organisms respond through plasticity and evolutionary change⁴⁸, urbanization is an increasingly important source of land-use change. Rates of urbanization are accelerating globally, with current levels of urbanization at three percent of the Earth’s landmass, excluding Greenland and Antarctica⁴⁹. This may not seem like an exceptionally large number, but over half of the world’s population lives within these urbanized areas, and pockets of urbanization dot almost every corner of the globe⁵⁰. Although studies have begun to quantify community composition and phenotypic changes in populations across urban and nearby rural habitats, few have explored the mechanisms that contribute to these changes. Because the urbanization process offers both novel urban habitats in close spatiotemporal proximity to ancestral rural habitats and the potential for replication across independent ancestral-novel comparisons^51,52, urbanization gradients are excellent, but underused, sources of ancestral-novel comparisons needed for assessing plasticity’s role in shaping evolutionary change. Of particular interest is the rapid change in environmental temperature over space and time under the urbanization process, especially as urban warming can serve as a space-for-time proxy to understand responses to global climate change. Urban heat island effects are relatively consistent across many regions, with temperature increases in excess of several °C in the air column and over 10°C at the surface being possible (though there are exceptions, especially urbanization in desert habitats⁵³). Though, of course, many other changes accompany the urbanization process, including changes in habitat structure, nutrient availability, and pollution⁵⁴.

In general, studies that explore plastic and evolutionary responses to urbanization have lagged behind climate change studies. This is surprising, since one of the best-known and earliest examples of rapid evolutionary change, industrial melanism in peppered moths, was itself driven by the fast progression of urbanization (and environmental pollution) associated with the industrial revolution^55,56. More recently, Donihue and Lambert⁵⁴ provided an excellent overview of how to assess and disentangle plastic and evolved responses to urban land-use change in addition to an exhaustive review of the studies that have done so in urban environments. They defined three criteria for demonstrating adaptive evolutionary change in response to urbanization, including quantifying phenotypic shifts in traits, measuring the fitness effects and genetic basis of those traits, and identifying drivers of trait changes. Only two studies met these criteria at the time of the review, including one study on killifish responses to elevated polychlorinated biphenyl pollution in urban environments⁵⁷ and another on the reduction of dispersing versus non-dispersing plant seeds in urban environments⁵⁸. Notably, the killifish study highlights the fact that the environmental stressors we consider in our review—climate change, land-use change, and environmental toxins—are not mutually exclusive, as, for example, killifish responses to polychlorinated biphenyl compounds may be considered in context of both land-use change and environmental toxicity. Relatedly, structural changes in urban environments, specifically replacing grassland and forest with impervious surfaces like roads, sidewalks, and buildings, contributes to the urban heat island effect; as the study on plant seed production demonstrated, these structural changes can also serve as an agent of selection themselves. Dispersing seed pods are selected against in urban environments because there are large gaps in suitable habitat owing to intervening sidewalks and roads⁵⁸.

Evidence consistent with an interpretation of urban evolution continues to accumulate. For example, Winchell and colleagues⁵⁹ have found phenotypic shifts in lizard morphology in urban areas where lizards use artificial surfaces as perches as opposed to vegetation in unaltered habitats (an interesting parallel with the structural changes in urban habitats altering plant seed dispersal); a common garden experiment suggested a genetic basis for these changes in morphology, though the adaptive nature of these changes has yet to be demonstrated, and maternal effects cannot be fully excluded with their use of first-generation offspring from field-collected parents. Interestingly, evolutionary responses to urban heat islands appear to be one of the least-studied aspects of the urbanization process despite the ubiquity of the urban heat island signal among cities. Angilletta and colleagues⁶⁰ demonstrated an increase in upper thermal tolerance in city ants compared with rural ants, though it is unclear whether this shift is adaptive and has a genetic basis. McLean and colleagues used a common garden experiment to demonstrate a genetic basis for shifts in growth rate of a chitinolytic fungus along an urbanization gradient, though, again, the adaptive nature of these changes is unclear, and only one urban-rural comparison was made⁶¹. Although urban evolution is gaining traction in the literature, the use of urban environments to assess the role of plasticity in shaping evolutionary responses has been used only to a marginal extent and remains open for future development. In particular, replicated urban-rural comparisons with multigenerational common garden studies are needed to explore the potential for urban evolution and its interplay with plasticity.

Environmental toxins

In conjunction with increasing CO₂ emissions and changes in land-use, more than a century of industrialization has led to the accumulation of toxins in the environment, notably heavy metal and pesticide contaminants⁶². Environmental toxins can have severe negative consequences for organismal fitness and population survival through a diverse range of effects and mechanisms of action⁶³. Despite this diversity and in contrast to urbanization and climate change, the evolution of resistance to environmental toxins is well established^64–66: for instance, the classic example of divergent evolution and speciation of plant populations living in and adjacent to toxic mine tailings^67–70.

Phenotypic plasticity also contributes to organismal responses to environmental toxins, as prior exposure to heavy metals⁷¹ and pesticides⁷² can reduce the negative effects of later exposure. Interestingly, exposure to one toxin may induce a plastic response with positive fitness effects (i.e. an adaptation) to other toxins^73,74, but the generality of this inducible “cross-tolerance” is currently unknown. This shared adaptive plasticity does not appear to be restricted to toxins with common mechanisms of action, contrary to expectations⁷³. Indeed, this result suggests an important role of plasticity in the evolution of toxin tolerance, as cross-tolerance would induce initially adaptive rather than maladaptive plasticity in response to novel toxins, possibly facilitating subsequent evolutionary refinement.

Like urbanization, the spatiotemporal heterogeneity of environmental toxins provides an opportunity to evaluate the role of plasticity in shaping evolution. Indeed, comparisons among populations differing in their exposure to pesticides provide some of our best evidence that genetic assimilation plays a role in evolutionary responses to anthropogenic change. Historically, studies focused on the evolution of tolerance in pest species directly targeted by pesticide application. A recent emphasis on non-pest species incidentally affected by pesticides has provided novel insights into the mechanisms underlying pesticide tolerance. By comparing pesticide tolerance among naïve populations to populations exposed to pesticides in contaminated environments (e.g. populations close to agriculture), numerous studies have now found evidence for the evolution of greater resistance to pesticides in affected populations of anurans and aquatic invertebrates compared to naïve populations^73,75–83.

Using an agricultural land-use gradient in a space-for-time substitution, Hua and colleagues recently extended these findings to test key predictions of genetic assimilation (Box 1): the canalization of inducible phenotypic variation by the action of natural selection. By quantifying both evolved and inducible (i.e. plastic) tolerances of wood frog populations varying in their distance from agriculture to an insecticide, Hua and colleagues found that evolved tolerance to a potentially lethal exposure of insecticide decreased with distance from agriculture^76,79. In contrast, the degree of plastic tolerance induced from an initial sublethal exposure increased with distance from agriculture⁷⁶. Together, these results are consistent with the predictions of genetic assimilation; exposure to novel environments induced the expression of phenotypic variation in ancestral populations, and this variation was then canalized in populations with a history of exposure to the novel environment^28,29. We emphasize that while these studies provide indirect evidence of genetic assimilation, more direct tests are needed. To accomplish this goal, resurrection studies and the study of extant populations varying in their time of exposure to novel environments may provide powerful means of testing the role of plasticity and evolution for responses to human-induced environmental change.