Artificial neural networks for discovering characteristics of fishing surveillance areas

Abstract

The demographic pressure entails over-exploitation of the coastal regions and the consumption of marine resources in a non-sustainable manner, jeopardizing the species renewal. Several species are currently facing great threat of disappearing from Portuguese coastal waters, namely the Sardina pilchardus, due to illegal, unregulated or not reported fishing. The Portuguese Navy performs regular surveillance and monitoring of fishing activities for law enforcement. Those actions gather useful information about the fishing activity, specifically about the types of fishing gear used. Since the geo-spatial data on a regular map, by itself, was not enough to present a clear picture regarding the predominant type of fishing gears used for captured sardine in the Portuguese coastal areas, we applied an artificial neural network to georeferenced information in order to derive a new layer with the areas where the fishing gears used for Sardina pilchardus fishing are most likely to be found.