Catch-based multi-model assessment of Acetes fishery along the Gujarat coast of India

The North-west coast of India is a unique marine ecosystem characterised by sustained productivity and a wide continental shelf. The region is known for several unique resources with restricted latitudinal distribution. Acetes are one of the major fishery resource of the Gujarat coast and are often considered the keystone species for the region’s marine ecosystem. Despite its ecological and fishery importance, a minimal assessment has been done of the area. The present study uses a multi-model approach towards the evaluation of the stock status of Acetes. Four models, namely CMSY, BSM, zBRT, and OCOM, were applied to the time series catch data to arrive at the stock status indicators. Artificial intelligence or machine learning tools like boosted regression trees and neural networks are used in model fitting. Output was compared across all the four models. CMSY and BSM gave the most comparable and conservative estimates and are adopted in the present study. The stock can be classified as sustainable or fully exploited (B/BMSY = 1.19-1.20, F/FMSY = 0.79-0.80; B/ B0 = 0.59-0.65). Continuous monitoring of the resource is recommended for its r-selected life history traits. Keywords: Artificial intelligence, Catch only methods Machine learning, Neural networks