Predicting voluntary forage intake of supplemented beef cattle

A major priority of beef cattle production is to meet animal nutrient requirements in order to achieve a desired level of productivity. Accurately predicting voluntary forage intake (VFI) is necessary to accurately predict the total nutrient intake of grazing or forage-fed beef cattle that are also supplemented with other sources of nutrients. Therefore, the objectives of this experiment were to utilize data from published literature to 1) identify factors that explain variation in VFI, and 2) develop and validate one or more mathematical models that predict VFI or total nutrient intake of grazing or forage-fed and supplemented beef cattle. A comprehensive literature review was conducted to retrieve experimental means (n=609) and descriptive information from 131 feeding trials that measured VFI and supplement intake. Simple regressions identified 43 continuous and 7 categorical variables that were related (P < 0.05) to forage DMI. Following randomization, 70% of published observations were used to develop predictive models, while the remaining 30% were used for validation. Categorical explanatory variables used to predict forage or total dry matter (DM) intake (DMI) included forage classification, forage harvest method, forage stem length, cattle production stage, and supplement feeding frequency, while continuous explanatory variables included shrunk body weight (BW), supplement neutral detergent fiber (NDF) intake (NDFI), supplement hemicellulose (HEM) intake (HEMI), supplement crude protein (CP) intake (CPI), forage CP content, forage NDF content, and forage HEM content, where supplement intake information was expressed in kg x hd^-1 x d^-1 and forage nutrient content was expressed as a % of forage DM. Development equations explained 70% (RMSE = 1.31; P < 0.0001) and 77% (RMSE = 1.31; P < 0.0001) of the variation in forage and total DMI, respectively. When applied to the validation dataset, these equations explained approximately 68% (RMSE = 1.32; P < 0.0001) and 72% (RMSE = 1.31; P < 0.0001) of the variation in forage and total DMI, respectively. These models explained a substantial portion of the variation in forage and total DMI, and therefore can be used in production systems to aid in predicting total DMI.