Application of Bayesian belief networks to marine ecosystem services assessment to support MPA management

Marine protected areas (MPAs) are a key management tool to halt the global decline in marine biodiversity and fish stocks. Evidence is mounting on the societal benefits they can support through enhanced ecosystem service... [ view full abstract ]

Marine protected areas (MPAs) are a key management tool to halt the global decline in marine biodiversity and fish stocks. Evidence is mounting on the societal benefits they can support through enhanced ecosystem service provision. In the UK almost all MPAs allow some activities to occur - protection is seldom complete. Here we explore how the level of management of MPAs can influence the delivery of marine ecosystem services using a geospatial Bayesian belief network (BBN) approach. Spatial pattern of seabed habitat potential to supply ecosystem services (ES) of four types was modelled: nursery habitat for commercial fish species; carbon burial; waste processing and primary production. Across the case study area, Lyme Bay, SW England, ES provision was highly heterogeneous and varied by ES type. Results also revealed that the current network of MPAs support higher flows of ecosystem services compared with surrounding waters, and MPAs with mobile demersal fishing restrictions had the greatest ES provision. Practical applications for this approach were tested: 1) mapping high risk areas to prioritise management interventions and ensure the most important areas were targeted; 2) applying scenarios for future management; 3) identifying trade-offs between ES provision with different management interventions /developments; and 4) focusing engagement and communication with managers and stakeholders. The BBN modelling approach provided a framework to build up understanding of how complex marine ecological systems respond to changes resulting from management and has considerable application as a practical tool to enable the consideration of marine ES in marine management.