Adaptation to fragmentation: evolutionary dynamics driven by human influences

Abstract

Fragmentation-the process by which habitats are transformed into smaller patches isolated from each other-has been identified as a major threat for biodiversity. Fragmentation has well-established demographic and population genetic consequences, eroding genetic diversity and hindering gene flow among patches. However, fragmentation should also select on life history, both predictably through increased isolation, demographic stochasticity and edge effects, and more idiosyncratically via altered biotic interactions. While species have adapted to natural fragmentation, adaptation to anthropogenic fragmentation has received little attention. In this review, we address how and whether organisms might adapt to anthropogenic fragmentation. Drawing on selected case studies and evolutionary ecology models, we show that anthropogenic fragmentation can generate selection on traits at both the patch and landscape scale, and affect the adaptive potential of populations. We suggest that dispersal traits are likely to experience especially strong selection, as dispersal both enables migration among patches and increases the risk of landing in the inhospitable matrix surrounding them. We highlight that suites of associated traits are likely to evolve together. Importantly, we show that adaptation will not necessarily rescue populations from the negative effects of fragmentation, and may even exacerbate them, endangering the entire metapopulation.This article is part of the themed issue 'Human influences on evolution, and the ecological and societal consequences'.

Keywords: anthropogenic fragmentation; dispersal; evolutionary rescue; island biogeography; metapopulation; short-term adaptation.

Figure 1.

Connectivity and patch size interact to determine how local (patch) dynamics affect landscape-level dynamics. X-axis: low connectivity decouples local demography from landscape-level dynamics (i.e. patches are largely independent from each other in both demographic and evolutionary trajectory; non-equilibirum dynamics); medium connectivity results in metapopulation dynamics, where local demography is largely independent among patches but recolonization balances local extinction events; high connectivity leads to high occupancy (recolonization and/or demographic rescue outweigh local extinctions,) and little to no genetic differentiation. Y-axis: asymmetry in patch size means some (large) patches are highly persistent and dominate landscape dynamics. Adapted from Harrison & Taylor [14]. (Online version in colour.)

Figure 2.

Pathways by which fragmentation could lead to adaptation. Ecological effects on the physical environment, population demography and biotic communities can create additional ecological effects (thin grey arrows; e.g. smaller patches lead to smaller populations which may lead to species losses) and selective effects (thick black arrows) at the local/patch scale (thin boxes) and landscape/metapopulation scale (thick boxes). Altered selection can then produce a variety of adaptive responses (bottom boxes). (Online version in colour.)