Predição do desempenho de bovinos de corte confinados com base na ingestão de matéria seca e energia mensuradas no inicio do periodo de alimentação
Carregando...
Data
2019-01-18
Autores
Silvestre, Antonio Marcos
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O objetivo deste estudo foi avaliar a classificação das baias com base no consumo logo nas semanas iniciais do confinamento de machos castrados e novilhas, sobre o desempenho, e desenvolver modelos de predição da ingestão de matéria seca (IMS), procurando melhorar sua acurácia por meio da inclusão da fase de confinamento em que os animais se encontram. Isso possibilitaria a identificação de “baias problemas”, ou de desempenho superior, facilitando a tomada de decisão na gestão. Para isso foram utilizados dados de 3.650 baias de confinamentos comerciais dos USA (2.256 de machos castrados e 1.394 de novilhas), os quais foram confinados entre os anos de 2009 a 2014. Os dados de IMS e de energia líquida de ganho (ELg) médios do período de terminação foram ajustados de acordo com o peso vivo inicial e dias em alimentação (DEA) pelo PROC NLIN do SAS (2009) separadamente para machos castrados e novilhas. Tanto a IMS como a ingestão de ELg médias ajustadas do período de terminação foram correlacionados com a ingestão das semanas de 5 a 12 pelo PROC CORR, e de acordo com a semana de maior correlação as baias foram classificadas em quartis: Superior ¼, Alta ¼, Média ¼ e Baixa ¼ do maior para o menor consumo e avaliado o desempenho pelo PROC MIXED. Para o desenvolvimento dos modelos foi utilizado o PROC CORR para identificar a correlação de cada variável (PVI, IMS de 8 a 28 dias, sexo, IMS da semana anterior) com a IMS e ordenar a inclusão delas no modelo, e por meio do PROC MIXED foi testado a permanecia de cada uma como parâmetro no modelo (P<0,10), e a variacia média foi calculada pelo PROC MEANS. A validação do modelo foi realizada pelo PROC REG. A IMS e a ingestão de ELg na sétima semana de alimentação foi a que apresentou maior correlação com as ingestões médias ajustadas do período de terminação, tanto para machos castrados (0.72) como para novilhas (0.66). Como resultado, a classificação das baias pela IMS ou ELg nesta semana foi capaz de inferir o desempenho das baias ao final do confinamento por meio das variáveis testadas (GPD, PVF, eficiência alimentar), se diferenciando entre os tratamento empregados (P<0,05), sendo que as baias nos quartis de menor ingestão tiveram seu desempenho prejudicado, o que deve estar associado a maior incidência de doença e mortalidade nestas baias (P<0,05). As equações desenvolvidas para cada semana do período de terminação apresentaram boa capacidade preditiva com valores médios de r2= 0,78, RMSE = 0,490 e AIC = -723,27. Os dados oriundos da classificação das baias na semana 7 indicam que a identificação de baias de desempenho superior e “baias problemas” pode ser feito neste período, seja com base na IMS ou ingestão de ELg, assim como, a alternativa proposta. Os modelos desenvolvidos possuem boa capacidade preditiva acompanhando a curva de consumo dos animais.
The objective of this study was to evaluate the classification of the pens based on the intake in the early weeks of feedlot of sterss and heifers, on the performance, and to develop models of prediction of dry matter intake (IMS), seeking to improve its accuracy through inclusion of the feedlot phase in which the animals are found. This would allow the identification of "problem pens", or of higher performance, facilitating the decision making in the management. Data from 3,650 commercial feedlot pens from the USA (2,256 steers and 1,394 heifers) were used, which were confined between the years 2009 to 2014. The mean DMI and net energy gain (NEg) data of the termination period were adjusted according to the initial body weight and days on feed (DOF) by PROC NLIN of SAS (2009) separately for steers and heifers. Both the DMI and the intake of NEg adjusted means of the termination period were correlated with the intake of weeks 5 to 12 by PROC CORR, and according to the week of greatest correlation the pens were classified into quartiles: Top ¼, High ¼ , Lower ¼ and Lowest ¼ from highest to lowest consumption and evaluated the performance by PROC MIXED. For the development of the models, PROC CORR was used to identify the correlation of each variable (IBW, DMI from 8 to 28 days, gender, DMI from the previous week) with the DMI and to order their inclusion in the model, and through PROC MIXED was tested for the presence of each as a parameter in the model (P <0.10), and the mean variance was calculated by PROC MEANS. The validation of the model was performed by PROC REG. The DMI and NEg intake in the seventh feeding week showed the highest correlation with the mean intakes of the finishing period, for both steers (0.72) and heifers (0.66). As a result, the classification of the pens by DMI or NEg this week was able to infer the performance of the pens at the end of the feedlot using the tested variables (ADG, FBW, feed:gain), differentiating between the treatments used (P <0, 05), and the stalls in the quartiles of lower intake had their performance impaired, which should be associated with a higher incidence of disease and mortality in these stalls (P <0.05). The equations developed for each week of the termination period presented good predictive capacity with mean values of r2 = 0.78, RMSEP = 0.490 and AIC = -723.27. The data from the classification of the pens at week 7 indicate that the identification of superior performance pens and "problem pens" can be done in this period, either based on DMI or NEg intake, as well as the proposed alternative. The models developed have good predictive capacity, following the consumption curve of the animals.
The objective of this study was to evaluate the classification of the pens based on the intake in the early weeks of feedlot of sterss and heifers, on the performance, and to develop models of prediction of dry matter intake (IMS), seeking to improve its accuracy through inclusion of the feedlot phase in which the animals are found. This would allow the identification of "problem pens", or of higher performance, facilitating the decision making in the management. Data from 3,650 commercial feedlot pens from the USA (2,256 steers and 1,394 heifers) were used, which were confined between the years 2009 to 2014. The mean DMI and net energy gain (NEg) data of the termination period were adjusted according to the initial body weight and days on feed (DOF) by PROC NLIN of SAS (2009) separately for steers and heifers. Both the DMI and the intake of NEg adjusted means of the termination period were correlated with the intake of weeks 5 to 12 by PROC CORR, and according to the week of greatest correlation the pens were classified into quartiles: Top ¼, High ¼ , Lower ¼ and Lowest ¼ from highest to lowest consumption and evaluated the performance by PROC MIXED. For the development of the models, PROC CORR was used to identify the correlation of each variable (IBW, DMI from 8 to 28 days, gender, DMI from the previous week) with the DMI and to order their inclusion in the model, and through PROC MIXED was tested for the presence of each as a parameter in the model (P <0.10), and the mean variance was calculated by PROC MEANS. The validation of the model was performed by PROC REG. The DMI and NEg intake in the seventh feeding week showed the highest correlation with the mean intakes of the finishing period, for both steers (0.72) and heifers (0.66). As a result, the classification of the pens by DMI or NEg this week was able to infer the performance of the pens at the end of the feedlot using the tested variables (ADG, FBW, feed:gain), differentiating between the treatments used (P <0, 05), and the stalls in the quartiles of lower intake had their performance impaired, which should be associated with a higher incidence of disease and mortality in these stalls (P <0.05). The equations developed for each week of the termination period presented good predictive capacity with mean values of r2 = 0.78, RMSEP = 0.490 and AIC = -723.27. The data from the classification of the pens at week 7 indicate that the identification of superior performance pens and "problem pens" can be done in this period, either based on DMI or NEg intake, as well as the proposed alternative. The models developed have good predictive capacity, following the consumption curve of the animals.
Descrição
Palavras-chave
Confinamento, Ingestão de Matéria Seca, Monitoramento, Predição, Semana, Feedlot, Dry matter intake, Monitoring, Prediction, Week