Assessing genetic diversity between breeds for conservation

Abstract

In domestic species, genetic diversity is partitioned between and within breeds, roughly in equal proportions. While some of the variation is deleterious, the main dichotomy of interest to us is neutral versus adaptive variation. The latter has been utilized by livestock keepers in selecting for specific economic characteristics or by natural selection to adapt the animal to specific environments. There is some correlation between both types of variation created by random drift or genetic linkage (hitch-hiking) but the correlation is probably low because of variable mutation rates and selection regimes. Reed and Frankham (Evolution, 55, 1095–1103) suggest that the correlation between molecular and quantitative variation is weak (Z about 0.20), indicating that molecular variation explains only 4% of variation in quantitative traits. When taking decisions on conservation, this dichotomy plays a paramount role. In analysing population structure and history, strictly neutral variation allows us to identify ancestral populations holding a possible source of alleles that are of economic value and which still may have been lost by chance during domestication. Moreover, strictly neutral variation provides tools for detecting the populations that are genetically more distinct. The classical approach of using genetic distances may, however, present several problems in ranking breeds for conservation. First, it completely ignores genetic variation within populations. Secondly, admixed populations, common in livestock, are not suitable for constructing phylogenetic trees. And thirdly, genetic distances vary greatly according to the marker used and the recent demographic history of a breed (e.g. effect of population bottleneck). There are other more robust tools (see the review by Toro & Caballero, Philos Trans R Soc Lond B, Biol Sci, 360, 1367–1378.). Adaptive variation, based on functional rather than neutral differences between populations, can provide new criteria to backup conservation decisions. There are two ways of approaching the problem: by examining known genes, and by identifying regions altered through natural or artificial selection. It is presumed that estimates of diversity that are based on known genes, such as those affecting growth and reproduction, better reflect the additive genetic variance in quantitative traits. However, note that contrary to what may be expected a priori, estimates of quantitative variation obtained with the loci controlling a quantitative trait loci are not necessarily better than the ones based on the neutral variation. This is a consequence of the multilocus nature of quantitative traits versus single locus estimates from neutral markers or individual genes. Covariances of allelic effects among selected loci do not contribute to single locus estimates. Thus, differentiation for economically important loci might not be more informative than differentiation for neutral markers. The second way to picture the adaptive variation is to identify regions that have been subject to artificial or natural selection. Demographic processes affect the entire genome, whilst selection affects specific important loci and their close neighbourhood leaving its signature in the region (selective sweep and differentiation between populations). Therefore, when a locus shows high levels of genetic differentiation, this may be interpreted as positive selection. Following an old proposal by R.C. Lewontin and J. Krakauer in 1973 (recently addressed by Luikart et al., Nature Rev Genet, 4, 981–994) comparing the observed distribution of FST (population differentiation) for different loci with the values expected for neutral loci, outlier markers are identified which deviate significantly from neutrality. These putative adaptive markers can be removed from quantifying the neutral differentiation and used as indicators of adaptive differentiation. It seems likely that the characterization of genetic diversity in future research will increasingly use adaptive variation, through the analysis of specific genes or outlier markers, and quantitative traits in combination with neutral variation.