A comparison of purine derivatives excretion with conventional methods as indices of microbial yield in dairy cows.

ABSTRACT

Three multiparous, ruminally and duodenally cannulated Holstein-Friesian milking cows (558 +/- 14 kg BW) with a mean milk yield of 19.9 +/- 1.4 kg/d in their 4th mo of lactation were fed a mixed diet of forage and concentrate at 100, 85, and 75% of ad libitum intake in a 3 x 3 Latin square design. Duodenal digesta flow was estimated using the dual-phase technique in which Cr-EDTA and Yb-acetate were used as liquid and solid markers, respectively. Microbial N (MN) was estimated using the duodenal flow of purine bases (PB); bacterial isolates from the rumen liquid and solid phases were used as references. Additionally, duodenal flow of PB and MN were estimated indirectly using the excretion of purine derivatives (PD) in urine and milk. Duodenal flow of PB and derived MN tended to decrease with feed restriction (from 258 to 154 mmol/d and 123.5 to 74.4 g/d, respectively). Estimates of PB and MN based on urinary PD showed the same trend, and decreases in PB (from 314 to 266 mmol/d, using LAB) were statistically significant. Using LAB, efficiencies of microbial protein synthesis in the ad libitum treatment were 12.9 and 17.0 g of MN/g of organic matter apparently digested in the rumen when estimated using duodenal PB and urinary excretion of PD, respectively. Urinary excretion of PD closely reflected changes in duodenal flow of PB as a result of feed restriction.