Exploring the optimum layout of artificial reefs based on species distribution models and computer simulations

Artificial reefs have been widely used to gather fishes in certain sea areas. The previous studies on the layout of artificial reef mainly depend on flow field simulation methods. This study, based on spatial and temporal distribution of dominant species in artificial reefs area, put forward a method to optimize artificial reefs layout by utilizing species
distribution models and computer simulation methods. The experimental area is located in the south-central of Bohai Strait in China, a region of rocky and muddy substrate. By now, twenty-four reinforced concrete reefs of 3m×7m×3m have been put in place. The process of the method is to: (1) Establish machine learning models on the spatial distribution and driving factors of important species, and select the optimal model based on the biodata, water quality and sediment environment of the artificial reef area; (2) Divide the experimental area into regular cells with a resolution
of 50m×50m; (3) Change the substrate types’ score of cells randomly to detect the effects of layout of reefs on the species distribution with all other driving variables fixed, because the layout of reefs would change the substrate type, and simulate this process for 100,000 times; (4) Find out the optimal numbers and layout of artificial reefs based on the predicted occurrence probability of the dominant economical species and harmful species, in every simulation of reefs layout. Our result shows that the optimal number of artificial reefs in the experimental area is 150. We found two optimal layouts among the 100,000 times of simulation. Because it is unrealistic to move the artificial reefs after the
placement, this method made the arrangement of artificial reefs more reasonable ecologically. Consequently, the method proposed by us to optimize the number and layout of artificial reefs is efficient, viable and replicable.