Modeling and Forecasting of Engraulicypris sardella (Usipa) Yields from Mangochi Artisanal Fisheries of Lake Malawi using Holt Exponential Smoothing Method

Abstract

This study was aimed at modelling and forecasting Engraulicypris sardella (Usipa) species yield from artisanal fishery on Lake Malawi in Mangochi District. The study was based on secondary count time series data on fish catches during the years of 1976 through 2012 provided by Fisheries Research Unit of the Department of Fisheries in Malawi. The study considered Holt exponential smoothing method to select an appropriate stochastic model for forecasting the Usipa species yields. Appropriate model was chosen based on Holt‟s exponential parameters α (alpha) and β (beta) Box – Ljung statistics and distribution of residual errors among others were estimated. The Holt‟s exponential smoothing method showed that the artisanal fishery landings of Usipa would increase to 22,849.4 tonnes in year 2022 which is 2,135.76 tonnes more than the landings in the year 2010 of 20,713.64 tonnes assuming that the biological and ecological factors remain the same. The forecast mean for Usipa is predicted to be 22,849.4 tonnes through the year 2022. The study observed that Usipa fisheries is being exploited sustainably. The confidence interval of 95% does not include a zero meaning that the likelihood that Usipa fishery might be under extinction currently or anytime soon might be very low. The increase of Usipa, could be as a result of a decline in population of competing species in the ecosystem. The Usipa Management Strategy for Southeast and Southwest Arms of Lake Malawi and Lake Malombe developed by the USAID/FISH Project will ensure continued effective and sustainable management of exploitation of the fishery.

Keywords: Engraulicypris sardella, landings, Holt exponential smoothing method, artisanal fishery, Lake Malawi.