Repeated evolution and the impact of evolutionary history on adaptation.

Terry J Ord,Thomas C Summers

doi:10.1186/s12862-015-0424-z

Abstract

BackgroundWhether natural selection can erase the imprint of past evolutionary history from phenotypes has been a topic of much debate. A key source of evidence that present-day selection can override historically contingent effects comes from the repeated evolution of similar adaptations in different taxa. Yet classic examples of repeated evolution are often among closely related taxa, suggesting the likelihood that similar adaptations evolve is contingent on the length of time separating taxa. To resolve this, we performed a meta-analysis of published reports of repeated evolution.ResultsOverall, repeated evolution was far more likely to be documented among closely related than distantly related taxa. However, not all forms of adaptation seemed to exhibit the same pattern. The evolution of similar behavior and physiology seemed frequent in distantly related and closely related taxa, while the repeated evolution of morphology was heavily skewed towards closely related taxa. Functionally redundant characteristics—alternative phenotypes that achieve the same functional outcome—also appeared less contingent.ConclusionsIf the literature provides a reasonable reflection of the incidence of repeated evolution in nature, our findings suggest that natural selection can overcome contingent effects to an extent, but it depends heavily on the aspect of the phenotype targeted by selection.Electronic supplementary materialThe online version of this article (doi:10.1186/s12862-015-0424-z) contains supplementary material, which is available to authorized users.

Highlights

Whether natural selection can erase the imprint of past evolutionary history from phenotypes has been a topic of much debate
If the literature provides a reasonable reflection of the incidence of repeated adaptive evolution in nature, our results imply that closely related taxa tend to be predisposed to adapt in similar ways, and that Gould [10] was generally right: evolution does not tend to repeat itself over large macroevolutionary time scales
If we assume that researchers are studying repeated adaptive evolution in similar ways, our findings suggest that behavior and physiology exhibit a higher degree of adaptability than morphology

Summary

Introduction

Whether natural selection can erase the imprint of past evolutionary history from phenotypes has been a topic of much debate. Spectacular examples include the replicated morphologies that reflect similarities in habitat use among Caribbean Anolis lizards [6, 7] or three-spined stickleback fish [8]. Such examples are striking because they appear to override the idiosyncrasies inherent in the process of evolutionary differentiation [9]. Supporters of Gould’s view might argue that Anolis lizards and stickleback fish—and most other examples of adaptive convergence for that matter—are cases where the same adaptation has evolved independently among closely related taxa. Members of the same genus or species tend to occupy similar environments [17] and tend to share

Objectives

Methods

Results

Discussion

Conclusion