Use of different approach in finding catch effort relationship in hook and line fishery in Kombuthurai village of Thoothukudi district of Tamil Nadu

Abstract

Kombuthurai has been described as a fishing village in Thoothukudi district, Tamil Nadu which has gained immense importance by practicing sustainable fishing practices. In Kombuthurai village during 2009-2013, catch per unit effort (CPUE) has varied from 0.20 tonnes to the maximum of 0.25 tonnes during this period. It is interesting to know that despite the range of effort varying from 1598hrs to 1720hrs, the catch remained in the range of 347.4 tonnes to a maximum of 430.8tonnes. The influence of various variables (latitude, longitude, distance, depth, seasons, years and fishing gear type) on CPUE was studied by General Linear Model (GLM) and Generalized Linear Model (GLMs). Normal distribution with identity link function and a gamma distribution with log link function were used in GLMs. The variables that were found to be significant in both of the models and distributions are year and season. GLM model was good fit with R2 ranging from 67.60% to a high of 91.7%.The low and high abundance of CPUE were categorized taking the below and above average CPUE. A binary logistic regression model was used between the CPUE abundance and with above mentioned independent variables. The binary logistic regression model explained 75.4% (Nagelkerke R2) of the variance in high CPUE abundance and correctly classified 88% of cases. The independent variable such as latitude, longitude, depth, and distance added significantly to the model, but the year, season and hooks did not add significantly to the model.