高精細枝肉横断面画像を利用した筋肉内脂肪中モノ不飽和脂肪酸割合の推定に関する研究

MUFA ratio might be one of the indicators of palatability for beef. The aim of this study was to develop the prediction method of MUFA ratio in marbling by image analysis method for Japanese Brown and Holstein. Data were collected from 25 ]apanese Brown cattle (19 steers， 6 heifers; 1 investigation day) and 76 Holstein steers (5 investigation days). High-resolution digital images of their carcass cross-sections were taken. Samples for chemical analysis were collected from rib eye surface and actual MUFA ratio was measured by gas chromatography. Marbling area percentage in the rib eye，characteristics of marbling shape (19 traits)，rib eye shape (2 traits)，color of rib eye (84 traits)，and the proportion of subcutaneous and intermuscular fat to the cross section area (2 traits) were calculated by image analysis. The multiple regression analysis was used to predict MUFA ratio from above 108 traits as candidate variables. The average of actual MUFA ratio was 55.32% (50.65 to 62.l0)for Japanese Brown and 51.74% (44.06 to 58.24) for Holstein respectively. R² of the multiple regression equation for Japanese Brown and Holstein were 0.68 and 0.38 without considering the investigation day. When the five investigation days were included in the multiple regression analysis for Holstein cattle， the R² become 0.39 to 0.91. The selected traits were proportion of subcutaneous fat，proportion of intermuscular fat，overall coarseness of marbling，degree of convexo-concave，standard deviations (SD) of green component of lean and SD of green component of marbling. The selected variables were the same between the investigation days. These results indicated that the prediction of MUFA ratio might be possible for Holstein and Japanese Brown cattle by image analysis method.