Abstract:
The nature of dispersal barriers for marine taxa often remains elusive. Here, we studied the relative role of stochastic and deterministic variables influencing population divergence, by comparing multiple replicate populations of sponges from marine lakes. Marine lakes are land-locked bodies of seawater with a variable connection with the surrounding sea via subterranean fissures. Furthermore, each lake has a distinct environmental regime (defined as water temperature, salinity and pH). Marine lakes with their discrete populations provide a unique model to study early stages of evolution in coastal marine taxa. We selected lakes that have comparable ages (~ 8000 years) and sizes (~15000 m2), but which vary in degree of connection to the open sea and environmental regimes. Using population genomic methods (double-digest restriction site associated DNA, ddRAD) we studied populations of Suberites diversicolor from 8 marine lakes and three adjacent sea locations. In this study ddRAD proved to be a useful and cost-effective method for both phylogeographic and population genomic analyses of sponges. We found strong genetic structure and in most cases strong divergence between populations (pairwise FST ranged from 0.54 - 0.63). Admixture analyses furthermore showed little gene flow between marine lakes, even between lakes only 1-10 kilometres apart. We found that at large spatial scales (> 200 km), stochastic dispersal limitation plays a large role, while preliminary analysis showed that environment plays a significant role in the connectivity and divergence of marine lakes populations at smaller scales (< 30 km). Hence, varied environments can lead to rapid divergence of sponge populations. Understanding how gene flow corresponds with environmental gradients will improve predictions on adaptive capacities of marine species under different climate change scenarios.