A data-limited modeling approach for conserving connectivity in marine protected area networks

Marine Protected Areas (MPAs) are areas of marine ecosystems that have some level of protection to support one or more conservation objectives. One characteristic of MPA networks is that MPAs are spatially configured such that they provide the greatest protection possible for multiple species. Yet, it can be difficult to determine optimal MPA network arrangement due to insufficient information on multi-species habitat use and their dispersal abilities as larvae and adults. Here, we propose a modelling approach that involves determining the optimal MPA network configuration for multiple species assemblages, located at different depths and having differing dispersal abilities. As a case study, we applied this methodology in Pacific Canada where we identified optimal MPA configurations to protect 40 species having different pelagic larval duration (proxy for dispersal) at 3 different depth class groupings (proxy for habitat use). Taken together, we found dispersal ability had a larger impact on optimal MPA network configuration for species spending a long time as larvae compared to species spending a short time as larvae. We identify which 10% of this area is most important to conserve to maintain connectivity for a multi-species MPA network and show that half of these sites remain important to conserve in the future as climate change alters connectivity patterns. This model for MPA network design is feasible with limited data which is beneficial for application to other regions and ecosystems.

Automated summary

This study presents a modeling approach to determine the optimal MPA network configuration for species with different dispersal abilities at varying depths, focusing on a case study in Pacific Canada. The research highlights the significant impact of dispersal abilities on the optimal MPA network configuration. By identifying the most important sites for conservation to maintain connectivity for a multi-species MPA network, it demonstrates the feasibility of this model with limited data for application to other regions and ecosystems.