A new approach of mapping interactions among human stressors for marine spatial planning in the North Sea

Anthropogenic stressors have a profound influence on the biome, leading to changes often degradation of habitat structure and species composition. Traditionally, a data driven process, from measurements and related analyses to developed management plans, is too slow for the general goal of nature protection and conservation. Therefore, it would be useful to introduce a tool that allows the implementation of management plans before knowing the actual impact of anthropogenic stressors. In fact, we believe that the identification of areas where human activities are mostly taking place can already serve for management purposes. However, currently environmental assessments consider stressors cumulatively, as their potential danger has been assumed to be related to a sum of anthropogenic activities. In this study, we developed a tool that identifies the interacting stressors, with their intensity and sign of interaction at each location. This way, we were able to map attributes of interactions while acknowledging their spatial variation across the seascape. The method was designed in R, using Unsupervised Random Forest to select the interacting variables, with the intensity and sign of their interaction being provided by the coefficient in a regression model. These steps were repeated using a moving window approach across our study area. We applied this method in the North Sea and the English Channel, as they are well known human exploited areas. We focused on stressors with a potential impact on the benthic fauna such as bottom trawling, presence of windfarms, protected areas, aquaculture and pollutant distribution. The description of interacting stressors provides a new approach in the framework of environmental assessment and marine spatial planning, as it can contribute to develop early strategies of mitigation before an impact assessment on the entire ecosystem has been achieved