The choice of a fitting model for in sacco degradation curves of some temperate and tropical grasses
| Autor: | Benoît Moreau, A.G. Deswysen, Alain Peeters, André Nivyobizi, D. Dehareng, Yvan Larondelle |
|---|---|
| Rok vydání: | 2007 |
| Předmět: | |
| Zdroj: | Grass and Forage Science. 62:198-207 |
| ISSN: | 1365-2494 0142-5242 |
| DOI: | 10.1111/j.1365-2494.2007.00578.x |
| Popis: | The objective was to compare alternative models for fitting to data on degradation of dry matter (DM), organic matter (OM) and crude protein (CP) in the rumen of two harvests of four temperate grasses, made into hay in Belgium, and one harvest of three tropical grasses commonly used in Burundi. Fifteen non-linear models were used and these included generalized Mitscherlich, simple Mitscherlich (or exponential), generalized Michaelis-Menten, Gompertz, logistic, simple Michaelis-Menten, segmented Van Milgen and von Bertalanfly models. Degradation profiles of DM, OM and CP obtained in sacco from the rumen of three mature sheep were fitted to all the models. The convergence success rate along with acceptable parameters was used as a first tool to eliminate models. Comparisons between the remaining twelve models were then made using the run test of sign, the root mean square prediction error, the root mean square error, the lack of fit test and the variance ratio test criteria. On the basis of an overall assessment using these tests, the Gumpertz lagged model was the best suited to fit the degradation data of both temperate and tropical grasses. This study confirmed the superiority of sigmoid type functions over diminishing return-type models. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|