COS 192-3 - Estimating dispersal patterns of an endangered stream fish with implications for habitat restoration

Abstract: Restoration ecologists are routinely plagued with the problem of where to implement conservation efforts to best maintain spatial connectivity and population structure for a focal species. Knowledge about dispersal patterns of a regional metapopulation could offer valuable insight to answer this question and allocate limited resources more effectively. However, direct estimation of dispersal is challenging because species can disperse widely within a landscape. Here, we developed a hierarchical Bayesian model to estimate asymmetric dispersal patterns and the associated spatial connectivity from spatial occurrence data of the endangered fish Topeka shiner (Notropis topeka) to guide restoration efforts of this species. Topkea shiner inhabits off-channel habitats, such as oxbows, in lower order streams of the prairie and midwestern regions of the US and oxbow restoration are implemented as a conservation tool. We estimated land use, climatic, and dispersal model parameters with Topeka shiner occurrence in streams (i.e., dispersal corridor) and examined whether spatial connectivity predicts the occurrence of Topeka shiner at restored oxbows. We also examined the usefulness of network centrality measures as a proxy for the estimated connectivity that can be used outside of the study range.



Dispersal patterns of Topeka shiner is estimated to be biased towards upstream movement, and the connectivity measure from this movement pattern had a positive effect on the occurrence of the shiner in streams. We extrapolated this connectivity across the entire network (396 stream reaches; total distance > 1,675 km) using the estimated dispersal parameters, highlighting stream reaches with high (31 reaches; 81 km), medium (116 reaches; 360 km) and low (249 reaches, 1,220 km) immigration potential. This spatial effect had a positive influence on shiner occurrence at restored oxbows. Topeka shiners were more likely to appear in restored oxbows located within reaches with high immigration potential. Although our metric of connectivity requires extensive spatial data, it was reasonably approximated by the network centrality measure, betweenness, which only requires physical information of the stream network. Network betweenness exhibited a positive relationship with immigration potential, thereby predicting shiner occurrence in oxbows. Producing connectivity maps is data intensive and not all regions may have prior data of occurrence. Therefore, using network centrality measures could offer insight about a species dispersal patterns without having the extensive occurrence data. While there are limitations due to data availability, our results show network centrality may be a useful tool when examining for dispersal corridors of a focal species.