Variability in snow-urine assays
Urea nitrogen: creatinine ratios in snow urine have become popular for assessing the extent of winter nutritional deprivation in ungulates. During winter 1992 – 1993, we collected 10–17 sequential snow-urine samples (402 total) from 27 individually identifiable free-ranging adult female elk. Within-animal variance accounted for 91% of the total variance (351.66 mg2/dL2) in creatinine, 86% of the total variance (637.03 mg2/dL2) in urea nitrogen, and 82% of the total variance (0.56) in urea nitrogen: creatinine ratios. This substantial within-animal variability was unexpected and led to experiments that examined whether the variability was due to sample collection and measurement technique or actually reflected biological variability. Factors examined included dilution effects, measurement (assay) repeatability, and short-term (<24 h) within-animal constancy in metabolite excretion. No dilution effects were detected when the initial concentrations of snow-urine samples were diluted <75% with water. Measurement variability accounted for 0.78, 0.37, and 27.7% of the total variance in creatinine, urea nitrogen, and urea nitrogen: creatinine ratios, respectively. Within-animal metabolite excretion was reasonably constant within 24 h, suggesting that creatinine provides a valid index for comparing urinary metabolites. We conclude that variability in urea nitrogen: creatinine ratios due to dilution, measurement variability, and short-term temporal variability in metabolite excretion was small compared with the total within-animal variance. Urea nitrogen: creatinine ratios should provide an accurate estimate of the true ratios of these metabolites in an elk's bladder urine. However, the interpretation of urea nitrogen: creatinine ratios is often complex, since they reflect the immediate dynamics between fat depletion, protein catabolism, and dietary intake. Differences in ratios between collections may be partially due to variations in recent dietary intake or restriction, in addition to true differences in long-term nutritional status. The best method for statistically analyzing snow-urine data remains unresolved.