New species form when groups of individuals in a population become reproductively isolated and can no longer mate with each other to produce living, healthy offspring. For decades, evolutionary biologists have sought to understand the links between an individual’s choice of mate and reproductive isolation between populations and species1. Writing in Proceedings of the National Academy of Sciences, Grant and Grant2 provide evidence suggesting that two species of Darwin’s finch learn features of their parents early in life and use this knowledge to inform their choice of mate in adulthood, a process known as sexual imprinting. Their study raises fascinating questions about the roles of learning and genetics in mate choice, and how matings between similar individuals (assortative mating) drive the evolution of new species.
Darwin’s finches live in the Galapagos archipelago. They are an iconic group of approximately 15 bird species that have contributed hugely to our understanding of natural selection and speciation3–5. Previous work has shown that the cultural inheritance of song can promote reproductive isolation between different species of Darwin’s finch6. However, it was not known whether sexual imprinting based on morphological features such as body size and beak characteristics could similarly promote reproductive isolation, or play a part in the rare cases of mating between species that produce hybrid individuals.
If sexual imprinting is key in directing mate choice, then individuals should choose mates that resemble their parents, and also themselves. In addition, if sexual imprinting contributes to matings between species, then the parents of the hybrid individuals that result from such matings should more closely resemble the other species than their own. To test these hypotheses, Grant and Grant analysed 22 years of data on body size, beak size and beak shape in two finch species — Geospiza fortis and Geospiza scandens — living on the same island.
Grant and Grant found significant positive associations between certain features of the birds’ chosen mates and those of their parents. For G. fortis, the body size of the chosen mate was strongly correlated with the body size of the chooser’s father, but weakly correlated with that of the chooser’s mother. The researchers did not explicitly test whether this imprinting was stronger in male or female offspring. For G. scandens, the beak length of male mates chosen by females was significantly associated with the beak shape and length of the female’s father, although there were no other significant associations. Grant and Grant suggest that these imprinting patterns can promote assortative mating by body size and beak length in both species.
For matings between species, the results were less straightforward. In some cases, hybridizing individuals or their fathers from a given species were similar in size and shape to individuals from the other species. Although such results are intriguing, hybridization events are rare and sample sizes are small, so further research is needed to confirm whether they can be generalized to other species.
Grant and Grant frame their study as a test of sexual imprinting, but acknowledge that the correlations they observed could also arise if mating preferences were inherited genetically. There has been increasing support for the idea that sexual imprinting can reinforce reproductive isolation in birds and other species5. However, evidence for the existence of genetically inherited mating preferences in birds is limited. It is not yet clear whether learnt behaviour has a greater effect on mate choice than does genetic inheritance, or whether these inherited effects have been under-studied. Disentangling the roles of inherited and learnt mate preferences, and their consequences for speciation, is a key challenge for the future7.
The most powerful tests for identifying sexual imprinting use an experimental ‘cross-fostering’ approach, in which offspring are swapped early in life and reared by unrelated individuals of the same or a different species8. There is also increasing interest in directly quantifying the genetic basis of mate choice using DNA sequencing9. We anticipate that future studies will combine experimental and genetic approaches to understand when and why learnt and inherited mating preferences evolve.
Grant and Grant’s findings hint that sexual imprinting might have different effects in males and females and across different species. In a cross-fostering experiment in wild mice published last year, the strength of sexual imprinting differed substantially between two species8. Furthermore, in the mouse species in which imprinting was weaker, only males showed signs of imprinting. Why imprinting might be weaker and inherited preferences stronger in females of some species is not clear. This could occur if matings between species require more investment in time or effort from females than males, or if mate-choice patterns are influenced by differences in the extent of parental care or the social environment. The effects of inherited and learnt mate preferences are likely to be complex, and studies of a broad range of biological systems might be required to uncover their relative roles in nature.
The work by Grant and Grant links individual variation in mating preferences in Darwin’s finches to the evolution of reproductive isolation, which is central to speciation. Sexual imprinting could have a role both in the ‘classic’ model of speciation, in which one species separates into two, and in the rarer process of speciation through hybridization, in which two different species mix to create a new one (Fig. 1). The possibility that Darwin’s finches show sexual imprinting should encourage further experimental tests in other species to determine the role of imprinting in natural populations. Understanding how mating preferences evolve will shed light on the processes shaping past, present and future biodiversity.
Nature 564, 196-197 (2018)