Variation in promiscuity and sexual selection drives avian rate of Faster-Z evolution

Abstract

Higher rates of coding sequence evolution have been observed on the Z chromosome relative to the autosomes across a wide range of species. However, despite a considerable body of theory, we lack empirical evidence explaining variation in the strength of the Faster-Z Effect. To assess the magnitude and drivers of Faster-Z Evolution, we assembled six de novo transcriptomes, spanning 90 million years of avian evolution. Our analysis combines expression, sequence and polymorphism data with measures of sperm competition and promiscuity. In doing so, we present the first empirical evidence demonstrating the positive relationship between Faster-Z Effect and measures of promiscuity, and therefore variance in male mating success. Our results from multiple lines of evidence indicate that selection is less effective on the Z chromosome, particularly in promiscuous species, and that Faster-Z Evolution in birds is due primarily to genetic drift. Our results reveal the power of mating system and sexual selection in shaping broad patterns in genome evolution.

Keywords: Faster-Z evolution; effective population size; genetic drift; sexual selection.

Figure 1

Relationship between effective population size (N_E) and variance in male reproductive success. Schematic outlining the predicted relationship between variance in male reproductive success and relative N_EZ and N_EX. When variance in reproductive success is the same in males and females, under monogamy, both N_EZ and N_EX = ¾ N_EA. As variance in male mating success increases, N_EZ < ¾ N_EA and N_EX > ¾ N_EA.

Figure 2

Phylogenetic relationship of the Galloanserae species in this study.

Figure 3

Estimates of mean d_N/d_S for loci on autosomes and the Z chromosome across the Galloanserae. Synonymous and nonsynonymous divergence estimates were calculated using the branch model in paml (Galloanserae analysis). 95% confidence intervals were calculated by bootstrapping with 1000 replicates, and significant differences in d_N/d_S between autosomal and Z‐linked orthogroups (permutation test, 1000 replicates) are indicated (*).

Figure 4

Phylogenetically controlled regression between proxies of sperm competition and Faster‐Z Effect. Data points are raw species values but P‐values and r ² estimates were calculated using phylogenetic generalized least squares regression with maximum likelihood and 1000 runs for each analysis. Autosomes refers to macrochromosomes (autosomes 1–10).

Figure 5

Estimates of mean Faster‐Z across sex‐biased gene expression categories. Sex bias was defined using fold change thresholds and t‐tests. 95% confidence intervals were calculated by bootstrapping with 1000 replicates. Autosomal orthologs were limited to chromosomes 1–10.