Portal de Pesquisa da BVS Veterinária

Tese em Português | VETTESES | ID: vtt-207299

Resumo

Equations to estimate total digestible nutrients (TDN) of small ruminants diets from chemical composition, in tropical conditions, are still not available. The main objective of this study was to develop equations, by means of a meta-analysis, that predict the digestible fractions of ether extract (EE), non-fiber carbohydrate (NFC), crude protein (CP) and neutral detergent fiber (NDF) from different goat and sheep diets. True digestibility coefficients and metabolic fecal contribution were the main parameters fitted to the models. Non-fibrous components were evaluated by the Lucas test to obtain true digestibility coefficients. Data from 27 trials, 21 with sheep (n = 766) and 6 with goats (n = 156) were used. In total, seven equations were developed, but three were only for digestible CP (CP-Lucas, CP-Bicomp and CP-Exp): considering Lucas nutritional entity test, a bicompartmental model that approaches different fractions of protein (soluble and undegradable), and an exponential model as an attempt to reproduce a dynamic degradation system. Two more models were for digestible NDF (NDF-linear and NDF-nonlinear): the first considered chemical components, intake and diet composition, the latter a nonlinear relationship between NDF and lignin. There were no effect of species on true digestibility coefficient or fecal contribution in the equations that estimated digestible NFC and EE, generating joint models for goat and sheep. All three CP models had adequate adjustments, but because of higher MSEP (mean square error of prediction), some variation was perceived in the CP-Exp. Both NDF models had low accuracy as demonstrated by the concordance correlation coefficient (CCC=0.5055 and 0.4197 for NDF-linear and NDF-nonlinear, respectively), meaning that more data might be necessary to increase model accuracy. The conditions to consider NFC and EE fractions as homogeneous nutritional entity were ensured. All three CP equations can be used to estimate the digestible fraction; it will depend on the available input data. As to NDF models, it is recommended some caution because resulting data will alter the energy estimates until further studies can be added to the data bank.

---

Equations to estimate total digestible nutrients (TDN) of small ruminants diets from chemical composition, in tropical conditions, are still not available. The main objective of this study was to develop equations, by means of a meta-analysis, that predict the digestible fractions of ether extract (EE), non-fiber carbohydrate (NFC), crude protein (CP) and neutral detergent fiber (NDF) from different goat and sheep diets. True digestibility coefficients and metabolic fecal contribution were the main parameters fitted to the models. Non-fibrous components were evaluated by the Lucas test to obtain true digestibility coefficients. Data from 27 trials, 21 with sheep (n = 766) and 6 with goats (n = 156) were used. In total, seven equations were developed, but three were only for digestible CP (CP-Lucas, CP-Bicomp and CP-Exp): considering Lucas nutritional entity test, a bicompartmental model that approaches different fractions of protein (soluble and undegradable), and an exponential model as an attempt to reproduce a dynamic degradation system. Two more models were for digestible NDF (NDF-linear and NDF-nonlinear): the first considered chemical components, intake and diet composition, the latter a nonlinear relationship between NDF and lignin. There were no effect of species on true digestibility coefficient or fecal contribution in the equations that estimated digestible NFC and EE, generating joint models for goat and sheep. All three CP models had adequate adjustments, but because of higher MSEP (mean square error of prediction), some variation was perceived in the CP-Exp. Both NDF models had low accuracy as demonstrated by the concordance correlation coefficient (CCC=0.5055 and 0.4197 for NDF-linear and NDF-nonlinear, respectively), meaning that more data might be necessary to increase model accuracy. The conditions to consider NFC and EE fractions as homogeneous nutritional entity were ensured. All three CP equations can be used to estimate the digestible fraction; it will depend on the available input data. As to NDF models, it is recommended some caution because resulting data will alter the energy estimates until further studies can be added to the data bank.

Biblioteca responsável: BR68.1