To assess inter-species, habitat effects, gender and collectionperiodvariability in Hg levels, variance (ANOVA) and covariance(ANCOVA) analyses were performed after checking assumptionsof normality and homoscedasticity of the data. Where homogeneitywas violated, THg data was log-transformed to stabilize the variance,and the non-parametric Mann–Whitney U test was used.Where assumptions were met, a parametric t-test (assuming unequalvariance) was applied to test for significant differencesamong samples. This included assessing the variations in totalmercury, between species, sex, site, maturity and size. Where thenumber of individuals per species was low (<10), Kruskal–Wallisnon-parametric tests were used to make statistical comparisonsbetween species. To explore the relationship between Hg leveland shark length, linear regression analysis was conducted for eachspecies. In each test, p < 0.05 was considered significant.
การแปล กรุณารอสักครู่..
