Predição de consumo alimentar de novilhas leiteiras em condições tropicais

• Ano de defesa: 2015
• Autor(a) principal: Ferreira, Verônica Bandeira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Mato Grosso; Brasil; Faculdade de Agronomia, Medicina Veterinária e Zootecnia (FAMEVZ); UFMT CUC - Cuiabá; Programa de Pós-Graduação em Ciência Animal
• Palavras-chave em Português: Consumo de ração; Meta-análise; Modelagem; Bovino; CNPQ::CIENCIAS AGRARIAS::ZOOTECNIA; Feed intake; Meta-analysis; Modeling; Cattle
• Link de acesso: http://ri.ufmt.br/handle/1/221

A meta-analysis was conducted to develop models for predicting dry matter intake (DMI) in dairy heifers under the tropical conditions and to assess its adequacy compared with four US current DMI prediction models [Quigley; National Research Council (NRC); and two Hoffman models]. The dataset was created using 103 treatments means from 29 studies, and it was randomly divided into 2 subdatasets for statistical analysis. The first sub-dataset was used to develop DMI prediction equations (19 studies; 67 treatment means) and the second sub-dataset was used to assess the adequacy of the predictive models (10 studies; 36 treatment means). The models were developed using nonlinear and linear mixed analysis. Breed (Bos taurus vs. Bos taurus × Bos inducus), body weigth (BW, 236.7 ± 63.7 kg) and average daily gain (ADG; 0.86 ± 0.29 kg/d) were considered as independent variable. There were not effects (P>0,05) of breeds or interaction between breeds and independent variables, BW0.75and ADG. Thus, it was proposed one type model for both breeds: nonlinear model [DMI = 0.1134 × BW0.75 − 3.3534 × e(-2.5885 × ADG)], and linear model [DMI = 6.7455 – 0.1625 × BW0.75 + 0.002 × (BW0.75)2 + 3.7634 × ADG − 1.6025 × ADG2]. Nonlinear model explained 72% of the variation in DMI and predicted with higher accuracy and precision rather than linear model (root mean square error of prediction = RMSEP; 9.24 vs. 11.32 % observed DMI). Quigley model explained only 56% of the variation in DMI but underpredicted it by 0.03 kg/d; it was the third most accurate and precise equation (RMSEP 11.96% observed DMI). NRC model explained 69% of the variation in DMI, but underpredicted it by 0.47 kg/d, with RMSEP of 12.28% of the observed DMI and presence of systematic constant bies. Hoffman exponential model I (BW as input) adequately predicted DMI and with similar accuracy to nonlinear model proposed. This equation explained 68% of the variation in DMI, overpredicted it by 0.14 kg/d; it was the second most accurate and precise equation (RMSEP of the 10.02% observed DMI). However, Hoffman exponential model II (BW and diet NDF as inputs) does not adequately predict DMI, because it explained only 46% of the variation in DMI, underpredicted 0.67 kg/d with higth RMSEP (18.57% of the observed DMI). Only the nonlinear model proposed in the present study, Hoffman exponential model I (BW as input) and Quigley model predict adequately the DMI of dairy heifers under tropical conditions.