Genetic variation of loci potentially under selection confounds species–genetic diversity correlations in a fragmented habitat

Abstract

Positive species-genetic diversity correlations (SGDCs) are often thought to result from the parallel influence of neutral processes on genetic and species diversity. Yet, confounding effects of non-neutral mechanisms have not been explored. Here, we investigate the impact of non-neutral genetic diversity on SGDCs in high Andean wetlands. We compare correlations between plant species diversity and genetic diversity (GD) calculated with and without loci potentially under selection (outlier loci). The study system includes 2188 specimens from five species (three common aquatic macroinvertebrate and two dominant plant species) that were genotyped for 396 amplified fragment length polymorphism loci. We also appraise the importance of neutral processes on SGDCs by investigating the influence of habitat fragmentation features. Significant positive SGDCs were detected for all five species (mean SGDC = 0.52 ± 0.05). While only a few outlier loci were detected in each species, they resulted in significant decreases in GD and in SGDCs. This supports the hypothesis that neutral processes drive species-genetic diversity relationships in high Andean wetlands. Unexpectedly, the effects on genetic diversity GD of the habitat fragmentation characteristics in this study increased with the presence of outlier loci in two species. Overall, our results reveal pitfalls in using habitat features to infer processes driving SGDCs and show that a few loci potentially under selection are enough to cause a significant downward bias in SGDC. Investigating confounding effects of outlier loci thus represents a useful approach to evidence the contribution of neutral processes on species-genetic diversity relationships.

Partitioning genetic and species diversity refines our understanding of species–genetic diversity relationships

“…Location of the 21 high Andean wetlands sampled in Chile's Norte Chico by Bertin et al. () and Troncoso et al. ().…”

“…Samples from site 13 (Figure ) used by Bertin et al. () and Troncoso et al. () were not included in this study because DNA concentrations did not meet GDF requirements.…”

“…SGDC studies have traditionally focused on neutral genetic diversity, although adaptive diversity has been occasionally considered (Bertin et al., ; Kahilainen, Puurtinen, & Kotiaho, ; Vellend et al., ; Watanabe & Monaghan, ). While neutral processes affect the whole genome uniformly, selection acts on specific regions, which bear the footprint of selection (Holderegger, Kamm, & Gugerli, ).…”

“…In SGDC studies, this approach can provide a clarified portrayal of similarity in the role of neutral processes on species and genetic diversity, and can thus help determine whether neutral processes are participating in the production of species–genetic diversity correlations. Two recent studies have applied this approach (Bertin et al., ; Watanabe & Monaghan, ). Bertin et al.…”

“…Bertin et al. () found that the SGDC did not hold when the effects of wetland connectivity, which explained about 50% of the variation in both diversity components, were factored out ( r = 0.25, p > 0.05). (b) High Andean wetlands in this region are highly fragmented and experience highly variable environmental conditions due to large‐scale climatic variations and local abiotic fluctuations resulting from the sharp orography of the Andes.…”