Determination of spatio-temporal distribution of Rastrelliger kanagurta using modelling techniques for optimal fishing

Abstract

The commercial Indian mackerel, Rastrelliger kanagurta (R. kanagurta) is widely distributed on the east coast of the Peninsular of Malaysia. Monsoon variations influence its occurrences and abundance. Understanding the variability of oceanographic physical processes and formation of habitats is vital in exploring fish resources. Temperature differences can have an important impact on ocean circulation and mechanisms that can affect the distribution and availability of fish. This study used fishing locations data of 2008 and 2009 and derived chlorophyll-a (chl-a) and sea surface temperature (SST) from satellite data (MODIS-Aqua). The objectives of this study were to determine potential fishing grounds of R. kanagurta using a presence-absence data model, Generalized Additive Model (GAM) and presence-only data model, maximum entropy (Maxent) and to assess the impact of temperature rise on its seasonal distribution based on the IPCC-AR5-RCP temperature forecast. Results showed that the GAM model had higher prediction accuracy. The constructed model predicted larger distribution areas of R. kanagurta. Maxent model, however, predicted limited distribution range concentrated only areas surrounding the point of presence. Increase of SST projected across all RCPs resulted in a decreased extent of suitable fishing habitats. Potential habitats were observed to shift out of the EEZ. Applicability of the GAM model to understand the spatio-temporal distribution of fish habitat positively contributes to optimal fishing and sustainability in the management of marine resources.