From benthic functional biodiversity to the mapping of ecosystem functions: a case study over the Black Sea northwestern shelf

Keywords: macrozoobenthos, functional biodiversity, ecosystem functioning, benthic-pelagic coupling, northwestern shelf of the Black Sea, modelling.

Benthic biodiversity is of global significance for the provision of ecosystem services and the mediation of global biogeochemical cycles. For instance, the macrozoobenthos plays a key role in marine carbon and nutrient cycling. Yet, current ocean biogeochemical models oversimplify or ignore life at the seafloor and its variability. The absence of detailed spatial distribution of the functions of the benthos, at large-scale (e.g., coastal and shelf scales), partly explains why benthic life characteristics are not taken into account in model formulation of benthic-pelagic exchanges. This lack of knowledge critically prevents our ability to predict the impact of climate change on the functioning of benthic life and its feedback on marine ecosystem and the biogeochemical budget of carbon, nitrogen, oxygen, phosphorus.

Here, we propose to scale up benthic biodiversity data from field sampling to the evaluation of ecosystem functions at large-scale (e.g., carbon sequestration, denitrification), relevant for ecosystem-based management. In our study, we include mechanistic and statistical models to map functional benthic biodiversity in relation to environmental drivers, and ultimately to incorporate its variability into current ocean model.

In more details, we compile macrozoobenthos occurrence from 210 sampling stations, covering constrained benthic habitats, over the northwestern shelf of the Black Sea. We use a functional approach of the biodiversity meaning that species are defined by their traits (e.g., dwelling depth and mobility) with an effect on ecosystem functioning. Then, species traits are upscaled at the community level by crossing species observations and their traits. From punctual values, we map continuous distribution of traits as a proxy of ecological processes (e.g., biomixing and biodeposition), precursors of ecosystem functions. We use a neural network to reconstruct maps of traits by linking them to environmental drivers, provided by a biogeochemical model, at high temporal and spatial resolution, run in an operational mode by Copernicus Marine Service (CMEMS). We use a combination of dozen biogeochemical (e.g., bottom oxygen and flux of organic carbon to the bottom) and physical drivers (e.g., bottom temperature and shear stress) as preliminary predictors of the distribution of traits. Then, we choose the best selection of predictors for our trait distribution models.

Our key findings show that bottom oxygen and stock of organic carbon are strong predictors for the distribution of traits at shelf-scale. Specifically, areas with high suspended materials and nutrients, such as near the Danube Delta, show deeper burrowing depths and greater mobility in benthic communities meaning potentially higher impact on sediment biomixing. In contrast, permanently hypoxic waters are characterized by very low sediment biomixing potential and very low benthic biodiversity.

Thanks to the maps of ecosystem functions, we adapt the parametrization of a current diagenetic model (e.g., depth of mixed layer, bioturbation coefficient) to incorporate the variability of the functional benthic biodiversity. A diagenetic model constrained by seafloor biodiversity, will constitute a significant step for the development of ocean models considering the impact of environmental changes on benthic life and its ability to deliver key marine ecosystem functions.