Michael G. Ritchie1 and Roger K. Butlin2,3 1Centre for Biological Diversity, University of St. Andrews, College Gate, St. Andrews KY16 9AJ, United Kingdom 2Ecology and Evolutionary Biology, School of Biosciences, The University of Sheffield, Sheffield S10 2TN, United Kingdom 3Department of Marine Sciences, University of Gothenburg, 405 30 Göteborg, Sweden Correspondence: mgrst-andrews.ac.uk; r.k.butlinsheffield.ac.uk

The concept of “genetic coupling” in mate recognition systems arose in the 1960s as a potential mechanism to maintain coordination between signals and receivers during evolutionary divergence. At its most basic it proposed that the same genes might influence trait and preference, and therefore mutations could result in coordinated changes in both traits. Since then, the concept has expanded in scope and is often used to include linkage or genetic correlation between recognition system components. Here we review evidence for genetic coupling, concentrating on proposed examples of a common genetic basis for signals and preferences. Mapping studies have identified several examples of tight genetic linkage between genomic regions influencing signals and preferences, or assortative mating. Whether this extends as far as demonstrating pleiotropy remains a more open question. Some studies, notably of Drosophila, have identified genes in the sex determination pathway and in pheromonal communication where single loci can influence both signals and preferences. This may be based on isoform divergence, in which sex- and tissue-specific effects are facilitated by alternative spicing, or on regulatory divergence. Hence it is not clear that such examples provide compelling evidence of pleiotropy in the sense that “magic mutations” could maintain trait coordination. Rather, coevolution may be facilitated by regulatory divergence but require different mutations or coevolution across isoforms. Reconsidering the logic of genetic coupling, it may be that pleiotropy could actually be less effective than linkage if distinct but associated variants allow molecular coevolution to occur more readily than potentially “unbalanced” mutations in single genes. Genetic manipulation or studies of mutation order effects during divergence are challenging but perhaps the only way to disentangle the role of pleiotropy versus close linkage in coordinated trait divergence.