Prediction of weaning weight in Santa Inês lambs using the body volume formula

• Published in: Tropical animal health and production 2023-02, Vol.55 (1), p.29-29, Article 29
• Main Authors: Gurgel, Antonio Leandro Chaves, dos Santos Difante, Gelson, Neto, João Virgínio Emerenciano, Ítavo, Luís Carlos Vinhas, Ítavo, Camila Celeste Brandão Ferreira, Costa, Carolina Marques, dos Santos, Geraldo Tadeu, Chay-Canul, Alfonso Juventino
• Format: Article
• Language: English
• Subjects: Algorithms; Animal Husbandry - methods; Animals; Biomedical and Life Sciences; Body Composition; Body volume (biology); Body Weight; Breast feeding; Correlation coefficient; Correlation coefficients; data collection; Datasets; Life Sciences; linear models; Meat; Model accuracy; Models, Biological; prediction; Reproducibility of Results; Root-mean-square errors; Sheep; Sheep, Domestic; Short Communications; standard deviation; Veterinary Medicine/Veterinary Science; Weaning; Weight; Zoology
• ISSN: 0049-4747; 1573-7438
• DOI: 10.1007/s11250-022-03445-8

Weaning weight (WW) is one of the most important information within production systems, as it is a reflection of management during the breastfeeding phase and will influence the performance of animals in subsequent phases. This study aimed to develop and evaluate linear, quadratic, and exponential models to predict WW using the body volume (BV) formula in Santa Inês lambs for meat. Eighty-five lambs at 90 days of age with WW 17.52 ± 3.79 kg and BV 13.29 ± 2.86 dm 3 were evaluated. The quality of fit of the models was evaluated using the coefficient of determination ( R 2 ), mean squared error (MSE), and root MSE (RMSE). For the external evaluation of the models, an independent dataset from 43 lambs at 90 days of age was used. The first-degree linear model showed the lowest values of MSE (1.02) and RMSE (1.01). In the external evaluation, all models exhibited estimates of mean WW and standard deviation of this weight similar to the external dataset, as well as high values (above 0.89) for the R 2 of predicted vs. observed data. Concordance correlation coefficient (CCC) analysis also revealed that all models showed accuracy and precision (CCC > 0.90). There was no difference between the models in terms of accuracy ( P  > 0.05). The comparison in terms of precision indicated that the linear model is more precise than the exponential model and that the quadratic model is as precise as the linear model. The first-degree linear model should be used due to its simplicity of interpretation and ease of estimation.