Existing equations to estimate chemical composition of cattle were developed over 40 years ago using different cattle genetics. The objective of this analysis was to evaluate existing equations using cattle genetics from then to today. A literature search was performed to identify experiments measuring carcass chemical composition by proximate analysis resulting in 106 treatment means from 10 experiments since 1976. Four sets of equations were used to compute carcass chemical composition assuming carcass water was known: 1) Garrett and Hinman (1969; 10.2527/jas1969.2811), 2) Gil et al. (1970: 10.2527/jas1970.313459x), 3) Preston et al. (1974; 10.2527/jas1974.38147x), and 4) Ferrell et al. (1976; 10.2527/jas1976.4251158x). Mean (SD) carcass water, fat and protein were 57.19 (7.20), 20.49 (8.87), and 17.80 (2.90) % of HCW, respectively. For all equations, simultaneous testing of intercept equal to zero and slope equal to 1 indicated that the intercept and slope for fat, protein, ash and energy were different than zero and one, respectively. The concordance correlation coefficient (CCC) was high (> 0.85) for fat and energy, but low (< 0.40) for protein and ash for all equations. Mean bias for fat, protein, ash, and energy ranged from -8.84 to -0.09%, 2.02 to 8.68%, -13.35 to 12.35%, and -3.50 to 0.79%, respectively, with Eq. 1 having the MB closest to zero for ash, Eq. 3 having the MB closest to zero for fat and protein, and Eq. 4 having the MB closest to zero for energy. Publication year as a continuous variable was a significant (P < 0.05) predictor of the difference between observed and predicted values for all components using all equations. Fat and energy had greater overprediction whereas, protein and ash had greater underprediction in more recent publications. In conclusion, existing equations provide similar precision, but differ in accuracy to predict carcass chemical composition. The trend for greater bias in more recent publication years indicates the need to reevaluate relationships among carcass chemical components using current cattle genetics.

中文翻译：

---

215 评估有关生长/育肥牛胴体化学成分的方程

用于估计牛化学成分的现有方程是 40 多年前使用不同的牛遗传学开发的。该分析的目的是评估从那时到今天使用牛遗传学的现有方程。进行了文献检索，以确定通过近似分析测量胴体化学成分的实验，从 1976 年以来的 10 次实验中得到了 106 种处理方法。假设胴体水已知，使用四组方程来计算胴体化学成分：1) Garrett 和 Hinman (1969) ; 10.2527/jas1969.2811), 2) Gil 等人。（1970：10.2527/jas1970.313459x），3）普雷斯顿等人。(1974; 10.2527/jas1974.38147x)，和 4) Ferrell 等人。（1976；10.2527/jas1976.4251158x）。平均 (SD) 胴体水、脂肪和蛋白质分别为 HCW 的 57.19 (7.20)、20.49 (8.87) 和 17.80 (2.90)%。对于所有方程，截距等于 0 和斜率等于 1 的同时测试表明，脂肪、蛋白质、灰分和能量的截距和斜率分别不为零和一。对于所有方程，脂肪和能量的一致性相关系数 (CCC) 高 (> 0.85)，但蛋白质和灰分的一致性相关系数 (CCC) 低 (< 0.40)。脂肪、蛋白质、灰分和能量的平均偏差范围分别为 -8.84 到 -0.09%、2.02 到 8.68%、-13.35 到 12.35% 和 -3.50 到 0.79%，方程式。1 对于灰分，MB 最接近于零，方程式。3 脂肪和蛋白质的 MB 最接近于零，方程式。4 具有最接近于零能量的 MB。作为连续变量的出版年份是使用所有方程的所有组件的观察值和预测值之间差异的显着（P < 0.05）预测因子。在最近的出版物中，脂肪和能量的预测值更高，而蛋白质和灰分的预测值更低。总之，现有方程提供了相似的精度，但预测胴体化学成分的精度不同。最近出版年份出现更大偏差的趋势表明，需要使用当前的牛遗传学重新评估胴体化学成分之间的关​​系。