Improving Large‐Scale Population Estimates and Assessments of the Ecological Importance of Three Epifaunal Bivalve Species by Combining Distribution and Abundance Models

Epifaunal bivalves include reef-building organisms that provide several important ecological functions in coastal marine environments. Evaluating the distributional patterns and population sizes is key in assessing the total and relative contribution of species toward these functions and can aid in improving spatial planning and management. In this paper, the use of species distribution and abundance models as a means to improve population size estimates compared to more simplistic methods (i.e., extrapolation of mean densities) was investigated. Gradient boosting models were fitted to predict occurrences and densities of three, both native and invasive, ecologically important bivalve species (blue mussel, <i>Mytilus edulis</i>; Pacific oyster, <i>Magallana gigas</i>; and European oysters, <i>Ostrea edulis</i>) occurring on the Swedish west coast. Bootstrapping methods were used to estimate model performance and error margins around the final predictions of total population size. Additionally, the spatial predictions were used to extract information on the key habitat types of each species. Results from the distribution models show considerable overlap in the use of shallow habitat by the native blue mussels and the invasive Pacific oyster, while the European oyster resided mostly in deeper habitats. There were, however, differences in larger geographical distribution patterns among all species. While all distribution models performed adequately (AUC ≈ 0.75-0.9 and TSS ≈ 0.45-0.65), the largest difference in estimated population size was observed when involving abundance models, reducing them by ~70% and ~50% for blue mussels and Pacific oysters respectively, while the estimate for European oysters remained similar. Importantly, estimates of total abundance when translated to biomass estimates indicated that the invasive Pacific oyster likely contributes the most to ecosystem functions associated with epibenthic bivalves in this area (~50% of total estimated dry weight biomass).