Mathematical models describing disappearance of Lucerne hay in the rumen using the nylon bag technique


Palangi V., Macit M., Bayat A. R.

SOUTH AFRICAN JOURNAL OF ANIMAL SCIENCE, cilt.50, sa.5, ss.719-725, 2020 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 50 Sayı: 5
  • Basım Tarihi: 2020
  • Doi Numarası: 10.4314/sajas.v50i5.9
  • Dergi Adı: SOUTH AFRICAN JOURNAL OF ANIMAL SCIENCE
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, BIOSIS, CAB Abstracts, Food Science & Technology Abstracts, Veterinary Science Database
  • Sayfa Sayıları: ss.719-725
  • Atatürk Üniversitesi Adresli: Evet

Özet

It is essential to study the dynamics of rumen degradation of feeds before their potential use in formulating diets for ruminants. Various mathematical models have been developed to describe this degradation. The non-lagged exponential model (Model I), the lagged exponential model (Model II), the Gompertz model (Model III), and the generalized Mitscherlich model (Model IV) were examined using two alternative software (SAS and MATLAB) to determine their efficacy in accounting for variation in ruminal disappearance of dry matter (DM) and crude protein (CP) of lucerne hay from three cuttings. All models described DM degradability well (R-2 >0.98). Only Models I and II converged when fitted to CP degradability data (R-2 >0.98). It was concluded that any of these models could be used to describe the degradation of DM, whereas only Models I and II could be used to describe the degradation of CP from three cuttings of Lucerne hay. All the models that were fitted to the DM degradation data performed reasonably well, with only minor differences in goodness of fit. However, these models differed in values of the parameter estimates. Additionally, SAS failed to converge in the analyses of CP with Models III and IV, and MATLAB converged to nonsensical values with Model III. Model I might be recommended because it fitted the data well and required estimates of the fewest parameters.