Fishing activity hotspots in the Mediterranean and Atlantic Seas at 0.5°

Statistically significant hotspots of fishing activities in the Mediterranean and Atlanti Seas were identified by the application of the Getis-Ord Gi statistic (Getis and Ord 2010) though the statistical software R using the globalG.test function (spdep package). The function computes a global test for spatial autocorrelation using a Monte Carlo simulation approach. It tests the null hypothesis of no autocorrelation against the alternative hypothesis of positive spatial autocorrelation. Then the local spatial autocorrelation was tested calculating the Gi statistic, using the local_g_perm function (dfdep package), which indicates the strength of the clustering.

Categorization of hotspots was performed, according to the Gi value and the p-value of a folded permutation test obtained for each grid cell, as follows:

• Gi>0 and p_value <=0.01 as Very hot
• Gi>0 and p_value <=0.05 as Hot
• Gi>0 and p_value <=0.1 as Somewhat hot
• Gi<0 and p_value <=0.1 as Somewhat cold
• Gi<0 and p_value <=0.05 as Cold
• Gi<0 and p_value <=0.01 à Very cold

Grid cells with a p-value > 0.1 were categorized as Insignificant.

The analyses were performed on cumulative fishing activity data at 0.5° resolution of seven different gears separately for the two macroareas.

The dataset presented includes for each area maps of each gear hotspot and spatial layers of the gears hotspots (.shp; .csv)