Updating analysis of nitrogen balance experiments in dairy cows.

Nitrogen balance (NB) experiments allow calculation of N retention in the body by subtracting N excreted in feces (NF), urine (NU) and milk (NM) from N intake (NI). In a previous study, we found that NB data from experiments with lactating dairy cows were generally high and, in the current meta-analysis, we update our earlier study with experiments from the last 2 decades and investigate probable causes of error. A total of 83 publications, with 86 experiments and 307 dietary treatments, were selected from top-ranked scientific journals that reported all NB components. The NB and NB components were analyzed by linear regression with a model that used NI as an independent variable and experiment as a random effect. The NF, NU and NM each represented 27 to 34% of NI, and the remaining N accumulated in the body was equal to 38.5 g/d (overall SD = 43.2 g/d). Retained N (as g/d or % of NI) increased linearly with NI, and this led to unlikely high N retentions, especially at high NI. Both NF and NU (g/d) increased with increasing NI, and we assume that some N in feces and urine were unaccounted. Only ~22% of experiments measured N in wet feces samples and, when analysis used dry samples, no mention of corrections due to potential volatile N losses during drying were reported. No experimentalists preserved feces immediately to prevent volatilization during collection. Moreover, ~27% of experiments estimated urine volumes by concentration of creatinine in spot samples, and in these experiments, NU was ~12% lower than those where total urine was collected (168 vs. 191 g/d). Only 40 experiments reported the volume and concentration of acids used for urine preservation, 33 furnished incomplete information, and the remainder did not describe the urine preservation method. In conclusion, the results of NB experiments using lactating dairy cows overestimate N retention, and the losses of N from feces and urine are the most probable reason.