Focused small-scale fisheries as complex systems using deep learning models

Small-scale fishing (SSF) is a relevant economic activity worldwide, so sustainable development will be essential to assure its contributions food security, poverty alleviation, and healthy ecosystems. However, the wide diversity of fisheries, their complexity, lack information limit ability propose/evaluate management measures plans effects on communities other productive activities. The state Baja California Sur, Mexico, our study case, ranks as third place in national fisheries production, possesses SSF fleets, has variety that share areas, seasons, operating units. In this work, assuming complex system were proposed deep learning models (DLM) forecast catch volumes, evaluate each input variable's importance, find interactions. Environmental variables tested DLM estimate predictive power. Different structures parameters optimal model was used. presented higher power are environmental with R = 0.90. Moreover, when used combination catches from performance 0.95 obtained. Using only catches, an 0.81. This allows use indirectly affect demonstrates useful tool assess system's face disturbances variables.