Models development to estimate rate passage in goats

• Grant number: 11/24073-1
• Support Opportunities: Scholarships abroad - Research Internship - Doctorate
• Start date: March 17, 2012
• End date: July 16, 2012
• Field of knowledge: Agronomical Sciences - Animal Husbandry - Animal Nutrition and Feeding
• Principal Investigator: Izabelle Auxiliadora Molina de Almeida Teixeira
• Grantee: Simone Pedro da Silva
• Supervisor: Antonello Cannas
• Host Institution: Faculdade de Ciências Agrárias e Veterinárias (FCAV). Universidade Estadual Paulista (UNESP). Campus de Jaboticabal. Jaboticabal , SP, Brazil
• Institution abroad: Università degli Studi di Sassari (UNISS), Italy
• Associated to the scholarship: 10/00956-9 - Solid and fluid passage rate in goats subject to different nutritional levels, BP.DR

Abstract

Passage rate studies help to understand the processes involved in goats nutrition and thus may enable the most appropriate nutritional management to obtain better production rates. The mathematical models developments are able to properly estimate the rate passage in goats is very important, but they are limited. Therefore, the objective of this project is to develop mathematical models to estimate passage rate in goats and evaluate these models developed based on independent data. To achieve this goal, will be used variables that can be routinely measured and calculated and presenting no collinearity. The database to be used to develop equations for predicting the passage rate will be from the candidate's doctoral thesis (FAPESP - 2010/00956-9), which includes goats in the final stage of growth. The equations are developed using PROC MIXED (SAS, 2002), which will be tested in the matrices of variance and covariance, seeking those that provide the best fit model, based on the criteria of convergence and interaction variances. After selection of articles containing information needed to test the models, these models are evaluations using different mathematical tools. The accuracy of the model will be evaluated by the coefficient of determination (R2) of the linear regression between the predicted and observed values, the concordance correlation coefficient (CCC) and the prediction of the mean square error (msep) and its decomposition into mean bias, bias systematic and random error. (AU)