Evolutionary biology: To mimicry and back again.

M any species that are dangerous or unpleasant to eat have evolved conspicuous signals that warn predators to avoid them. Not surprisingly, many other species that are edible to predators, from birds and butterflies to salamanders and sea slugs, have evolved to resemble these inedible species 1. By doing so, the 'mimics' receive protection from predation, just like their 'models'. This phenomenon is known as Batesian mimicry after the explorer and naturalist Henry Walter Bates, who first described it 2. Batesian mimicry has long fascinated evolutionary biologists , and it is widely used to illustrate the power of natural selection to produce remarkable adaptation 1. Yet we still do not know how common Batesian mimicry is, what its role is in evolutionary diversification, nor whether it can be reversed. Writing in Nature Communications , Davis Rabosky et al. 3 present findings on mimicry of coral snakes that go a long way towards answering these questions. In 1867, the naturalist Alfred Russel Wallace suggested that the striking resemblance between deadly coral snakes and numerous harmless species of red–black-banded (RBB) snakes reflected Batesian mimicry 4 (Fig. 1). However, whether coral-snake mimicry actually occurs has been questioned ever since, primarily because of the (presumed) non-concordance in the geographical distributions and abundances of coral snakes and their mimics. Several studies have attempted to address this issue 5 ; most notably, it has been demonstrated that predators avoid artificial snakes that have RBB patterns 6 but only in geographical regions where coral snakes occur 7 , exactly as predicted by the mimicry hypothesis. Davis Rabosky and colleagues focus on this system, but present a more comprehensive study than these earlier investigations. By integrating colour-pattern, distribution and phylo-genetic data across all 'New World' species of snake, they show that evolutionary shifts to RBB patterns in coral snakes and numerous non-venomous species are highly correlated in space and time. Indeed, they find that every origin of the RBB pattern in non-venomous snakes occurred only after that particular lineage and coral snakes were present together in the New World. Thus, in every case, the warning signal arose first in the model, then in the mimic, which is a key prediction of Batesian-mimicry theory. These data should therefore lay to rest any doubts about whether coral-snake mimicry does occur. The authors' work also shows that coral-snake diversity strongly predicts (and substantially increases) the number of mimic species in a …