Abstract
Abstract Unlike for high forage diets, there has been limited validation of spot urine sampling and the use of urinary creatinine as a marker to estimate urine output in beef cattle fed high grain diets. Therefore, the objective of this study was to validate the use of creatinine as a marker for estimating urine output in cattle fed a finishing (FIN) compared with a backgrounding (BKGN) diet. Six heifers were randomly allocated to 2 treatments in a crossover design with 21 d periods. Dietary treatments were alfalfa silage at 12% (FIN) or 60% (backgrounding; BKGN) of dietary dry matter (DM), with the remainder of the diet consisting of corn grain (67.8 or 38.1% on a DM basis), corn dried distillers grains with solubles (DDGS; 0.75 or 19.0%) and a constant level (1.15%) of supplement. Total collection (TC) of urine was conducted during the last 3 d of each period. Twelve spot urine samples were also collected during TC to represent every 2-h sampling in a 24-h cycle. Data were analyzed using SAS. Creatinine excretion rate varied from 27.2 to 35.0 with an average of 31.3 mg/kg of body weight per day. The creatinine concentration of spot samples collected for each heifer were averaged to simulate urine samples obtained from various spot sampling frequencies [equally spaced 12, 6, 4, and 2 time points (TP) starting at feeding: 12TP, 6TP, 4TP, and 2TP, respectively]. There was diurnal variation in the urinary creatinine concentration on both diets. Although creatinine concentration did not differ (P = 0.82) between 12TP and TC, use of 12TP underestimated (P < 0.01) urine output (11.2 vs. 18.0 kg/d). However, there was no diet effect (P ≥ 0.32) on creatinine concentration and total urine output across methods (12TP vs. TC). When comparing spot sampling TP, there was no diet, method, or diet × method interaction (P ≥ 0.28) on urinary creatinine concentration and urine output. The regression between TC and 12TP had an intercept different from 0, and slope differed from 0 and 1. As for TC and 6TP, the intercept was not different from 0; however, slope was different from 0 and 1. As expected, increasing the frequency of spot sampling improved precision and accuracy of the regression between the estimated and measured urine output according to adjusted R2 (0.27, 0.24, 0.87 and 0.92 for 2TP, 4TP, 6TP, and 12TP, respectively) and root mean square prediction error (8.95, 9.09, 3.75, and 2.94). In summary, the use of the spot sampling technique and urinary creatinine as a marker to estimate urine output in beef cattle fed either FIN or BKGN diets might not be reliable due to the diurnal variation in creatinine excretion and variable creatinine coefficients.