Effect of body size and condition factor on whole body composition of immature population hatchery reared mahseer, Tor putitora from Pakistan

Sixty two samples of hatchery reared mahseers (Tor putitora) were sampled for analysis of proximate composition. It was observed that the highly inverse correlation between water contents (%) and percent fat contents (wet weight, r = 0.3105) and percent water content and percent protein content (wet weight, r = 0.7624) exist. There is little bit negative correlation between percent water contents and percent ash contents (wet weight, r = 1244), percent fat contents (dry weight, r = 1.006) and percent protein contents (dry weight, r = 0.1676) while there is positive correlation between percent water contents and percent ash contents (dry weight, r = 0.3239). Condition factor has highly significant negative correlation with percent ash contents (wet weight, r = 0.5637) and percent protein contents (wet weight, r = 3053) but it has no effect on water contents and fat contents of body. Wet body weight has highly significant negative correlation with percent fat contents (wet and dry weight, r = 0.3386; r = 3261), while wet body weight has slightly positive correlation with other body constituents. Total length has highly positive correlation with percent ash content (wet and dry weight, r = 0.287; r = 0.343) and highly inverse correlation with percent fat contents (wet and dry weight, r = 0.312; r = 0.291), while other body constituents remains constant with varying total length. Log edible body weight has highly positive correlation with log water (r = 0.999), log ash (r = 0.961), log fat (r = 0.763) and log protein content (r = 0.979). Log total length also has highly positive correlation with log water (r = 0.947), log ash (r = 0.958), log fat (r = 0.771) and log protein content (r = 0.906). This shows that fish grow isometrically. It is therefore concluded that body composition of Tor putitora can be estimated directly from water content, body weight or length of the fish using predictive regression models developed in this work with a reasonable amount of accuracy.