Estimation of Age Composition using Maximum Likelihood from Length Frequency Distribution of Nile Tilapia (Oreochromis niloticus) in Lake Tana, Ethiopia

Abstract

Estimation of age composition of fish using Length frequency distribution is common if there is no aging of individual fish using otoliths or tag and recapture methods. It can be done by using maximum likelihood. The objectives of this study, therefore, to estimate age compositions; and the Von Bertalanffy growth coefficient and asymptotic length of the Nile tilapia stock. Available length-frequency data from Bahirdar Fisheries and Other Aquatic Life Research Center’s annual progressive reports for Nile tilapia fishery was compiled into 63 groups with  0.5-cm size classes (11-11.5 cm to 41.5-42cm). Length-frequency observations consisted of the actual number of catch at length  measured of the specious by year was used for this study. Estimation of age composition in to age cohorts and estimation of the proportions each age cohort was done using maximum likelihood. The number of age classes of the stock was selected using an index of AIC.  According the result of minimum AIC, a stock having three cohorts was selected with the proportion in the population of 0.14, 0.285, and 0.875 for the 1^st, 2^nd and 3^rd cohorts, respectively. It was found also that this estimation is statistically significant at a significant level of less than 1%. The Von Bertalanffy growth coefficient and asymptotic length were also estimated. The growth parameter measured in the change in average length per cohort was 0.435cm and highly significant; and the asymptotic fish length was 44.11 cm.

Keywords: length frequency distribution, Maximum Likelihood, , age composition, Nile tilapia,  asymptotic length, growth parameter.