Performance Evaluation of Manure Nitrogen Output Models Suitable for Lactating Dairy Cows in China
Is Version Of
Manure nitrogen (N) output from dairy cattle is a major environmental concern in China. Various empirical models are available to predict manure N output from dairy cattle, but accuracy and precision of these models has not been assessed for Chinese conditions. The objective of this study was to evaluate the performance of extant models that predict different forms of manure N output for lactating dairy cows in China with the aim of identifying the best-fit and most suitable prediction models. A total of 35 empirical models were evaluated for their ability to predict N excretion of dairy cows in China fed a wide range of diets. The data set consisted of 99 treatment means from 32 publications with information on animal and dietary characteristics and N output flows. Performance of models was evaluated using root mean square prediction error (RMSPE) and concordance correlation coefficient (CCC) analysis. The N intake (NI) based model of Kebreab et al. (2010) was selected as best for predicting fecal N excretion (RMSPE = 15.8% and CCC = 0.75). The Reed et al. (2015) model, which also used NI as an input variable, was most suitable for predicting urinary N (RMSPE = 26.0% and CCC = 0.63) and total N (RMSPE = 15.8% and CCC = 0.81). Models predicting urinary urea N (UUN) and urinary N / total N performed poorly. Overall, the deviation of regression line from the equality line (y = x line) for even the best-fit urinary, fecal, and total N excretion models demonstrated the need to develop improved models for use under Chinese conditions. Using N output data from dairy cows in China to develop manure N output models may help improve environmental stewardship of the dairy industry in China.
This is a manuscript of the article Dong, Ruilan, Hongmin Dong, Karen A. Beauchemin, and Hongwei Xin. "Performance Evaluation of Manure Nitrogen Output Models Suitable for Lactating Dairy Cows in China." Transactions of the ASABE (2018). DOI: 10.13031/trans.12710. Posted with permission.