Statistically Classified Marine Epifaunal Bioregions on the Southern Benguela Margin, South Africa

Marine bioregional maps often underpin effective management and conservation planning. Bioregional maps determined by statistical models can predict the most probable bioregion at unsampled locations and have many advantages over expert‐derived approaches, such as including measures of uncertainty. Here, we aim to statistically determine and map marine epifaunal bioregions on the Southern Benguela margin of South Africa, including their associated spatial uncertainty, utilising survey collected abundance data (70‐800 m). Regions of Common Profile (RCP) models were implemented using the ‘ecomix’ R package to model and map bioregions of epifaunal assemblages and their environment. Epifaunal abundance data from 46 species collected across 325 sites by research trawl surveys between 2017 and 2020, were used to inform bioregions. Environmental predictors were bottom dissolved oxygen and temperature measured in situ and slope derived from bathymetry. Five bioregions were identified. Two well‐defined depth breaks at ~150 m and ~400–500 m separated the mid shelf from the outer shelf and the outer shelf from the upper slope, respectively. The mid shelf and upper slope bioregions were associated with the highest estimated certainty, responding strongly to low oxygen concentration and low temperature, respectively. Outer shelf bioregions had the highest species richness, were dominated by the hermit crab Sympagurus dimorphus and were predicted with greater spatial uncertainty. This was the first study to statistically classify marine epifaunal bioregions using abundance data for the Southern Benguela margin of South Africa and furthers understanding of offshore epifaunal distributions and their potential drivers. Methods applied here produced practical outputs for spatial planning and conservation management by identifying five distinct epifaunal bioregions and associated environmental gradients. This study provides a foundation for conducting statistical approaches to marine ecosystem classification and mapping in South Africa.