Effects of learning on evolution: robustness, innovation and speciation

All animals are highly plastic and rely on the modulation of gene action, physiology and behaviour to continuously modify their phenotypes. Compared to other types of plasticity, learning, defined as the internal representation of novel information, allows animals to better exploit environmental features unique to certain times and places. This distinctive property of learning gives it an enormous potential to promote evolution through increased robustness, innovation and speciation rate. First, learning can enhance robustness because it allows individuals to adopt new resources and avoid novel threats. Empirical examples include the modification of egg-laying timing and nesting site selection in birds and of egg-laying substrate choice in insects. Second, learning can lead to innovation because it often has an exploratory stage that can lead to the discovery and refinement through trial and error of new, fitness-enhancing features. The best examples are cases of social learning that lead to the exploitation of novel food sources followed by genetic changes that optimize use of the new diet. Finally, learning can increase the levels of assortative mating that lead to population divergence either when young imprint on their parents or when individuals restrict their mate choice criteria based on interactions with prospective mates. While the notion that learning can have strong effects on evolution is backed by theory and some data, we currently lack broad experimental evidence to support that claim.