農業部-水產試驗所全球資訊網農業部-水產試驗所全球資訊網

海鱺 (Cobia, Rachycentron canadum ) 魚體之一般成分差異大，且不容易由外形辨別，本研究嘗試應用低週電磁波掃瞄 (Electromagnetic scan, EM-scan) 之非破壞性分析，建立海鱺魚體之預估方程式，以評估品質指標。掃瞄所用之海鱺體重2,880 ~ 8,340 g，其一般成分分析結果，水分含量隨魚體體重增加而減少，脂肪含量則隨魚體體重增加而增加，且其與水分含量呈負相關 (R² = 0.965)。至於灰分及蛋白質含量則不因魚體大小不同而有明顯差異 (p > 0.05)。以體導電值、體重、尾叉長及魚體水分、瘦肉值、灰分和蛋白質等一般成分測定值，經統計可得出適用之預估方程式 (R² 值分別為0.981、0.983、0.941和0.984)。脂肪預估方程式之R² 值較低為0.784，為檢驗此預估方程式之實用性，另採海鱺6件樣品，分別測量其體重、體長及TOBEC值，推算出魚體水分、脂肪之預估值，與實測值進行paired t-test檢測，結果仍為可行。故知此等非破壞性分析方法均有助於預測海鱺魚體之一般成分。

The proximate composition of Cobia (Rachycentron canadum) varies greatly and is difficult to discern by their appearance. In this study, a non-destructive electromagnetic (EM) scan method was applied to estimate the proximate composition of cobia. The predictive equations were obtained to assess the index of quality.Cobia, weighing 2880~8340 g, were scanned with an EM scan. The analytical results of the proximate composition indicated that the moisture content decreased while lipid content increased as the body weight of cobia increased. Results also showed a negative relationship between moisture and lipid contents (R² =0.965). Ash and protein contents of cobia showed no significant differences with the body weight level in this study (p > 0.05). The TOBEC (total body electrical conductivity) value, body weight, fork length, and proximate composition were used to develop feasible predictive equations. The R² values of the equation for moisture, lean body mass, ash, and protein were 0.981, 0.983, 0.941, and 0.984, respectively.The R² value of the predictive equation for lipids was lower at 0.784. To evaluate the feasibility of the equation by paired t-test, the body weight, fork length, and TOBEC value of moisture and lipid contents of six new samples were checked by the predictive equation and experimental measurements. Results showed no significant differences between them. It is feasible to use equations for non-destructive analysis to predict the proximate composition of cobia.