Abstract
The present study conducted non-parametric test on 22 distance characters of 42 samples of three-month-old Labeo rohita from a fish pond in Chitwan district, Nepal. ImageJ software was used to create landmark on the fish samples and established 22 distance characters along the dorso-ventral and antero-posterior body axes. Univariate and multivariate data analysis was used with PAST software. All the 22 distance characters were positively correlated with correlation coefficient greater than zero. Shapiro-Wilk test calculated p-value less than 0.05 for two distance characters. There was high variation in the means and medians of the data based on Levene’s test and Kruskal Wallis test. Tukey's pairwise test and Dunn’s post hoc test analyzed medians of the data for the significant relationship among the distance characters. Multivariate analysis with Principal Components Analysis (PC1 and PC2) of the 22 distance characters calculated the total Eigenvalue of 19.18944 and 87.2247% of variance under correlation matrix and Bootstrap N: 100. The fish samples varied in external morphology mainly in the body depth shape in dorso-ventral axis which is supposed to be the effect of culture conditions, food and mode of feeding.