Pendekatan bayesian berhirarki untuk penganggaran indeks kepelbagaian Shannon bagi macrobenthos

Abstract

Water pollution is a matter of serious concern of the authority. The level of water pollution in an ecosystem is usually measured using Shannon’s diversity index for a particular organism. In this study, the level of pollution at the Bintulu waters over time period of August 2002 to September 2006 were investigated based on Shannon’s diversity index using macro benthic organism. Sampling were done for each month from August 2002 to October 2002, and continued once every three months except for the period of February 2005 to Mac 2005. Data of diversity index for three sampling stations are computed and analyzed using Bayesian models. Time series data for each station are analyzed separately to obtain the posterior mean of diversity index for the study period. The results from individual analysis are combined using Hierarchical Bayesian model to allow for the variation between stations. When allowing for these variations, the posterior mean of diversity index for station 1 is 1.522, station 2 is 1.420 and station 3 is 1.476 and standard deviations are 0.6953, 0.6964 and 0.6974 respectively. The results show that the overall mean is 1.471 implying that the average water quality level at the project site reaches the warning level because the value is close to 1.5 since the acceptable level of diversity index indicating no pollution is between 1.5 to 3.5