- 1.This study describes spatial patterns in the biodiversity (species, assemblages) of rocky reef fishes at a spatial scale relevant to management, and compared the outcomes for this biodiversity from alternative procedures for selecting marine protected areas (MPAs) and from the selection of MPAs for fisheries-related objectives.
- 2.The study area included 104 species in two assemblage types; 36 species and 14 species occurred only in one or two locations respectively.
- 3.MPAs selected by hotspot richness, greedy richness complementarity, and summed irreplaceability included similar percentages of species and significantly more species than randomly selected MPAs. A combined species-assemblage selection ensured representation of assemblage diversity. Representation of all species and assemblage types required 92% of locations.
- 4.MPAs chosen using density of all fishes or density of exploitable fishes as selection criteria included fewer species (than MPAs selected using species identity) and the percentage of species accumulated did not differ from a random selection.
- 5.Use of an established MPA as the seed for an expanded network was inefficient, leading to additional locations being required and an accumulation of species that did not differ from a random selection.
- 6.The smallest MPA network that fulfilled multiple management objectives (representation of assemblage diversity and majority of species, population viability, support for fisheries, connectivity) required 30% of the surveyed locations.
- 7.This study concluded that: MPAs selected without the benefit of data on intra-habitat variation in species assemblages will be unrepresentative; the upper range of currently promoted targets for MPA establishment (i.e. 30%) should be regarded as a minimum for biodiversity conservation; MPAs selected for fisheries-related reasons may not provide expected benefits for the remainder of the fish assemblage.
Copyright © 2006 John Wiley & Sons, Ltd.