Mate choice for neutral and MHC genetic characteristics in Alpine marmots: different targets in different contexts?

Abstract

Sexual selection through female mate choice for genetic characteristics has been suggested to be an important evolutionary force maintaining genetic variation in animal populations. However, the genetic targets of female mate choice are not clearly identified and whether female mate choice is based on neutral genetic characteristics or on particular functional loci remains an open question. Here, we investigated the genetic targets of female mate choice in Alpine marmots (Marmota marmota), a socially monogamous mammal where extra-pair paternity (EPP) occurs. We used 16 microsatellites to describe neutral genetic characteristics and two MHC loci belonging to MHC class I and II as functional genetic characteristics. Our results reveal that (1) neutral and MHC genetic characteristics convey different information in this species, (2) social pairs show a higher MHC class II dissimilarity than expected under random mate choice, and (3) the occurrence of EPP increases when social pairs present a high neutral genetic similarity or dissimilarity but also when they present low MHC class II dissimilarity. Thus, female mate choice is based on both neutral and MHC genetic characteristics, and the genetic characteristics targeted seem to be context dependent (i.e., the genes involved in social mate choice and genetic mate choice differ). We emphasize the need for empirical studies of mate choice in the wild using both neutral and MHC genetic characteristics because whether neutral and functional genetic characteristics convey similar information is not universal.

Keywords: Extra‐pair paternity; Marmota marmota; inbreeding avoidance; major histocompatibility complex; mate choice; sexual selection.

Figure 1

Residual of the number of EPY in a litter as a function of Queller and Goodnight's relatedness between social pairs (A) and as a function of MHC class II protein dissimilarity between social pairs (B). Residuals are corrected for confounding factors (the number of sexually mature male subordinates present in a given family in (A) and for the number of sexually mature male subordinates present in a given family, the social pair Queller and Goodnight's relatedness and its associated quadratic term in (B). Open circles represent the observed residual number of EPY in a litter. Thick lines represent predictions of the model and the gray surface represents standard errors of the fitted model. The black dots represent observed data averaged over classes and their size is proportional to the number of litters within each class (class width 0.2 in A).