Prediction of fillet yield in hatchery populations of Asian sea bass, Lates calcarifer (Bloch, 1790) using body weight and measurements

Document Type : Original research

Authors

• Suthajaree Yenmak ^{1, 2}
• Weerakit Joerakate ¹
• Supawadee Poompuang ¹

¹ Department of Aquaculture, Faculty of Fisheries, Kasetsart University, Bangkok, 10900, Thailand

² Center for Advanced Study in Agriculture and Food, Institute for Advanced Studies, Kasetsart University, Bangkok, 10900, Thailand

10.1007/s40071-018-0202-9

Abstract

The purpose of this study was to develop and validate prediction equations for fillet traits in Asian sea bass. Ninety-day old fish (average weight = 9.48 ± 0.35 g; length = 8.78 ± 0.6 cm) from four hatchery stocks (Chachoengsao, Chon Buri, Chumphon, and Samut Songkram) were raised in earthen ponds. After 300 days, weight and body measurement data of live fish (n = 400) were collected. Mean individual weight was greatest for the Chacheongsao stock (1166.32 ± 23.42 g) and was similar for the other stocks, ranging from 982.96 ± 25.07 to 997.44 ± 24.71 g. Fillet percentage varied slightly from 47.33 to 49.88%. Positive high correlations were observed for weight and body measurements with fillet weight, whereas there were significant but weak correlations for body measurements with fillet yield. Prediction equations developed from body weight using simple linear regression models yielded R2 values of 0.97–0.98 for fillet weight for each stock. Correlations between values predicted from the body weight models and actual values were 0.98–0.99 for fillet weight. Stepwise regression was performed to develop prediction models for fillet yield from body measurements. The best fillet yield prediction models identified length and an additional 2–3 measurement as potential predictors depending on the population. Prediction biases were close to zero despite low to moderate (r values of 0.20–0.51) degrees of predictive power of the models. However, prediction models for fillet yield should be further developed to increase predictability and be applicable to new data.

Keywords

• Barramundi
• Fillet yield
• Stepwise regression
• Predictive power
• Value-added fish products

---