Contrasting performance of marine spatial planning for achieving multiple objectives at national and regional scales

Although spatially explicit decision support tools are widely used in marine conservation planning, only a few studies have used multi-objective approaches to explicitly achieve biodiversity conservation and sustainable use objectives. Here, we use spatial analyses of a comprehensive environment-based classification (developed using Gradient Forest modelling) that summarises spatial patterns in demersal fish composition and turnover in the oceans around New Zealand, and a 20-year record of trawl locations (representing spatial use by fishers), to identify options for biodiversity protection that vary in their relative delivery of conservation outcomes and impacts on fishers. We also compare analyses carried out at a national scale with those generated by aggregating independent regional analyses. The top-ranked 20% of sites identified using fish composition data alone would on average protect 33% of the geographic range of each fish species, but would result in a 58% loss of areas currently fished. By contrast, using the trawl locations as a cost layer reduced average fish protection by 23% in relative terms, but loss of fished areas was almost eliminated. Using different weightings to vary the influence of the trawl footprint, a range of intermediate scenarios were identified that delivered greater fish protection than the fully cost-constrained scenario. However, these intermediate scenarios resulted in greater reductions in the area retained for fishers than the biodiversity optimised scenario. Aggregated regional scenarios delivered lower conservation returns and higher impacts on fisheries than comparable single national scenarios, reflecting the constraining of opportunities for biodiversity protection whilst avoiding overlaps with trawling. While stakeholder MPA design processes often occur at regional scales, our analyses demonstrate the value of considering both regional and national analyses to identify optimal reserve configurations.