Reinforcement of pre-zygotic isolation and karyotype evolution in Agrodiaetus butterflies

Speciation is among the most important and poorly understood areas in evolutionary biology. One controversial aspect is reinforcement, in which natural selection favours mate discrimination when incipient species are still capable of hybridizing, but with reduced fitness. Recent studies have questioned several of the classic examples of reinforcement in natural populations; once part of darwinian orthodoxy, reinforcement was relegated to the margins. But a new study puts it back on the agenda. An analysis of the evolutionary history of Agrodiaetus, an unusual group of butterflies with variable chromosome numbers, shows that when closely related species occur together, they almost always differ in wing colour. This is a character critical in mate choice and colour differentiation occurs primarily between young and closely related species, arguing in favour of reinforcement as a mechanism generating diversity. The reinforcement model of evolution argues that natural selection enhances pre-zygotic isolation between divergent populations or species by selecting against unfit hybrids1,2 or costly interspecific matings3. Reinforcement is distinguished from other models that consider the formation of reproductive isolation to be a by-product of divergent evolution4,5. Although theory has shown that reinforcement is a possible mechanism that can lead to speciation6,7,8, empirical evidence has been sufficiently scarce to raise doubts about the importance of reinforcement in nature6,9,10. Agrodiaetus butterflies (Lepidoptera: Lycaenidae) exhibit unusual variability in chromosome number. Whereas their genitalia and other morphological characteristics are largely uniform, different species vary considerably in male wing colour, and provide a model system to study the role of reinforcement in speciation. Using comparative phylogenetic methods, we show that the sympatric distribution of 15 relatively young sister taxa of Agrodiaetus strongly correlates with differences in male wing colour, and that this pattern is most likely the result of reinforcement. We find little evidence supporting sympatric speciation: rather, in Agrodiaetus, karyotypic changes accumulate gradually in allopatry, prompting reinforcement when karyotypically divergent races come into contact.