Mimetismo

PERSPECTIVES
microevolutionary construction of a tightly linked supergene now seems untenable. Batesian polymorphism requires new colour genes to be tightly linked from the beginning. Without this, tighter linkage will not evolve because recombination will destroy the adaptive polymorphism before it is stabilized6,15. Most genetic architectures are probably not suitable, resulting in mostlymonomorphic mimicry from which novel, poorly adapted mutants cannot escape. Batesian mimicry encounters what Turner6 calls an evolutionary ‘sieve’, beyond which few species will pass to become polymorphic.

Diversity in mimicry: paradox or paradigm?
Mathieu Joron James L.B. Mallet
Visual mimicry is a textbook case of natural selection because it is both intuitively understandable and hasrepeatedly evolved in a range of organisms: it is the ultimate example of parallel evolution. In many mimetic groups, particularly butterflies, a huge variety of colour patterns has arisen, even in closely related species. There has been much recent controversy over explanations of this variety. Mimicry is today a broad field of evolutionary study; here we discuss the evolution of its diversity inpredator–prey systems.
Mathieu Joron is at Génétique et Environnement, CC065 ISEM, Université de Montpellier 2, Place Bataillon, F-34095 Montpellier Cedex 5, France (joron@isem.univ-montp2.fr); James Mallet is at the Galton Laboratory, University College London, 4 Stephenson Way, London, UK NW1 2HE (http://abacus.gene.ucl.ac.uk./jim/jim.html).

The coevolutionary chase
One of the main attributes ofBatesian mimicry, and a reasonable part of its definition, is that the mimic is deleterious to its model2,13,16,17. The model could therefore escape its mimic by evolving a new warning pattern. However, this escape is probably transient, because mutant models could soon attract new mimics. A coevolutionary chase might therefore arise between models and mimics, similar to the evolutionary ‘arms race’described for host–parasite interactions2,18. Huheey16 suggested a coevolutionary chase could cause morph and race differentiation in models (as in Heliconius erato and H. melpomene), eventually leading to the coexistence of several mimetic patterns in a community. Similarly, although new hypotheses are now being tested19, Smith et al.20 explained polymorphism in the distasteful Danaus chrysippusas an escape from an overload of palatable Hypolimnas mimics. More recently, equations for the coevolution of a pair of mimetic species have been developed21, showing that cyclical coevolution of model and mimic is possible if interspecific interactions (i.e. benefit of mimicry to the mimic and cost of Batesian mimicry to the model) are stronger than intraspecific interactions (i.e. aposematism andpalatability).

lthough mimicry is often presented merely as an undergraduate showcase for darwinian selection, it is a microcosm for evolutionary theory in general, with ramifications into the evolution of polymorphism, transitions between adaptive peaks, origins of biodiversity, and the evolution and maintenance of community patterns. The classical view of mimicry and frequency-dependentselection, found in most textbooks, is as follows: predator aversion to the appearance of warningly coloured (aposematic) species is exploited by edible prey – Batesian mimics – which evolve to resemble their models1,2. Batesian mimics should lose protection when common because they hide among superficially similar models whose unpalatability maintains the validity of their warning colours – if theedible mimics become too common the predators would soon learn to ignore the colouration. Rare mimetic morphs should be fittest because of this diversifying frequency-dependent selection, leading to the possibility of stable polymorphisms if several distinct models are imitated by a single species. Conversely, aposematic models are best protected when common (purifying frequency-dependent...