Comparing spatial conservation prioritization methods with site versus spatial dependency‐based connectivity

Larval dispersal is an important component of marine reserve networks. Two conceptually different approaches to incorporate dispersal connectivity into spatial planning of these networks exist, with an open question as to when either would be most appropriate. Candidate reserve sites can be selected individually based on local properties of connectivity, or with a spatial dependency-based approach of selecting clusters of strongly connected habitat patches. The first acts upon individual sites, whilst the second acts upon linked pairs of sites. We used a combination of larval dispersal simulations representing different seascapes and case studies of biophysical larval dispersal models in the Coral Triangle region and the province of Southeast Sulawesi, Indonesia, to compare the performance of these two methods in the spatial planning software Marxan. We explored the reserve design performance implications of different dispersal distances and patterns, based on the equilibrium settlement of larvae in protected and unprotected areas. We further assessed different assumptions about metapopulation contributions of unprotected areas, including a worst-case scorched earth of 100% depletion and more moderate scenarios. Our results suggest that the spatial dependency method is suitable when dispersal is limited, a high proportion of the area of interest is substantially degraded, or the target amount of habitat protected is low. Conversely, if subpopulations are well-connected, the scorched earth assumption is relaxed, or more habitat is protected, protecting individual sites scoring highly in metrics of connectivity is a better strategy. Spatial dependency methods generally produced more spatially clustered solutions with more benefits inside than outside reserves compared to site-based methods. Therefore, spatial dependency methods potentially provide better results for ecological persistence objectives over enhancing fisheries objectives, and vice versa. Different spatial prioritisation methods of using connectivity are appropriate for different contexts, depending on dispersal characteristics, unprotected area contributions, habitat protection targets, and specific management objectives. Article Impact statement: Performance of different methods for using larval dispersal information in spatial conservation prioritization are highly context dependent . This article is protected by copyright. All rights reserved.