Dispersal behaviour and riverine network connectivity shape the genetic diversity of freshwater amphipod metapopulations. Issue 24 (10th October 2021)

• Record Type: Journal Article
• Authors: Alther, Roman; Fronhofer, Emanuel A.; Altermatt, Florian
• Is Part Of: Molecular ecology. Volume 30:Issue 24(2021)
• Page Start: 6551
• Page End: 6565
• Publication Date: 2021-10-10
• Subjects: connectivity -- Gammarus fossarum -- landscape genetics -- population genetics -- stochastic simulation; Molecular ecology -- Periodicals; Molecular population biology -- Periodicals; 576
• DOI: 10.1111/mec.16201
• Languages: English
• ISSNs: 0962-1083

Abstract

Abstract: Theory predicts that the distribution of genetic diversity in a landscape is strongly dependent on the connectivity of the metapopulation and the dispersal of individuals between patches. However, the influence of explicit spatial configurations such as dendritic landscapes on the genetic diversity of metapopulations is still understudied, and theoretical corroborations of empirical patterns are largely lacking. Here, we used microsatellite data and stochastic simulations of two metapopulations of freshwater amphipods in a 28, 000 km 2 riverine network to study the influence of spatial connectivity and dispersal strategies on the spatial distribution of their genetic diversity. We found a significant imprint of the effects of riverine network connectivity on the local and global genetic diversity of both amphipod species. Data from 95 sites showed that allelic richness significantly increased towards more central nodes of the network. This was also seen for observed heterozygosity, yet not for expected heterozygosity. Genetic differentiation increased with instream distance. In simulation models, depending on the mutational model assumed, upstream movement probability and dispersal rate, respectively, emerged as key factors explaining the empirically observed distribution of local genetic diversity and genetic differentiation. Surprisingly, the role of site‐specific carrying capacities, for example by assuming a direct dependency of population size on local river size, was less clear cut: while our best fitting model scenario included this feature, over all simulations, scaling of carrying capacities did not increase data‐model fit. This highlights the importance of dispersal behaviour along spatial networks in shaping population genetic diversity.