Genomic analysis of morphometric traits in bighorn sheep using the Ovine Infinium® HD SNP BeadChip

Abstract

Elucidating the genetic basis of fitness-related traits is a major goal of molecular ecology. Traits subject to sexual selection are particularly interesting, as non-random mate choice should deplete genetic variation and thereby their evolutionary benefits. We examined the genetic basis of three sexually selected morphometric traits in bighorn sheep (Ovis canadensis): horn length, horn base circumference, and body mass. These traits are of specific concern in bighorn sheep as artificial selection through trophy hunting opposes sexual selection. Specifically, horn size determines trophy status and, in most North American jurisdictions, if an individual can be legally harvested. Using between 7,994-9,552 phenotypic measures from the long-term individual-based study at Ram Mountain (Alberta, Canada), we first showed that all three traits are heritable (h² = 0.15-0.23). We then conducted a genome-wide association study (GWAS) utilizing a set of 3,777 SNPs typed in 76 individuals using the Ovine Infinium^® HD SNP BeadChip. We found suggestive association for body mass at a single locus (OAR9_91647990). The absence of strong associations with SNPs suggests that the traits are likely polygenic. These results represent a step forward for characterizing the genetic architecture of fitness related traits in sexually dimorphic ungulates.

Keywords: Animal model; GWAS; Ovis canadensis; SNP.

Figure 1. Manhattan plots for morphological characteristics.

Horn length (A), horn base circumference (C), and body mass (E). The blue line represents the genome-wide significance threshold; the red line represents the threshold for suggestive association. Positions are relative to the domestic sheep genome assembly (version 3.1; Jiang et al., 2014). The green arrow indicates the suggestive locus for body mass. Next to each Manhattan plot is the corresponding QQ-plot (B, D, and F), with the genomic inflation factor (λ) and standard error indicated in the bottom right of each plot. The black line shows a 1:1 correspondence while the red line is a regression through the observed data.

Figure 2. Scatterplot of LD estimates versus inter-markers distance.

A non-linear least squares regression line is shown, with the round point indicating the half-length estimate.

Figure 3

Heat maps of expected percent power of a GWAS as a function of sample size and effect size for linkage disequilibrium (LD) estimates of 0.75 (A), 0.50 (B), and 0.25 (C). Light colors indicate higher power to detect associations (D). Dotted red lines correspond to the number of samples used in this study (N = 76).