Identification of dispersal barriers for a colonising seagrass using seascape genetic analysis
Science of the Total Environment
School of Science / Centre for Marine Ecosystems Research / Centre for Ecosystem Management
Department of Biodiversity, Conservation and Attractions Western Australian Marine Science Institution Dredging Science Node Edith Cowan University
© 2020 Seagrasses are important habitats providing many ecological services. Most species have broad distributions with maximum dispersal distances of 100's of kms, however there is limited understanding of dispersal distances of colonising species like Halodule uninervis. It commonly grows in disturbed environments and could disperse to other meadows via clonal fragments. Effective conservation management requires greater understanding of genetic structure, dispersal barriers, and connectivity timescales to predict recovery following disturbance. Despite fragment viability of up to 28 days in a congenera, this theory remains untested in situ. Using 80 neutral single nucleotide polymorphisms, we investigated genetic diversity, gene flow patterns and structure among 15 populations of H. uninervis along 2000 km of Western Australian coastline. These data were combined with a multi-generational oceanographic dispersal model and a barrier dispersal analysis to identify dispersal barriers and determine which fragment dispersal duration (FDD) and timescale over which stepping-stone connectivity occurred, best matched the observed genetic structure. The 2-7 day FDD best matched the genetic structure with 4–12 clusters, with barriers to dispersal that persisted for up to 100 years. Modelling suggested greater fragmentation of metapopulations towards the southern edge of the species distribution, but genetic diversity did not decline. Several long-term boundaries were identified even with fragment viability of up to 28 days. This suggests H. uninervis dispersal is spatially limited by factors like oceanographic features and habitat continuity which may limit dispersal of this species. This study reiterates that potential dispersal does not equal realised dispersal, and management scales of 10's of kilometers are required to maintain existing meadows. Recruitment from distances further than this scale are unlikely to aid recovery after extreme disturbance events, particularly towards the range edge of H. uninervis distribution.