Abstract
In a method agreement study, as part of a larger randomized controlled trial, the total-tract apparent digestibilities (TTAD) of DM, starch, and ash-corrected neutral detergent fiber (aNDFom) were calculated and compared between using the orts-adjusted diet, which accounts for the nutrient content of the TMR and orts, versus using the TMR-adjusted diet, which accounts only for the nutrient content of the TMR. Multiparous Holstein cows (n = 28) between 32 and 71 DIM were enrolled and moved to a freestall pen with Calan gates (American Calan, Northwood, NH) and fed a lactating diet high in corn silage for 10 weeks. Total-tract apparent digestibility was measured 3 times during the study, with samples collected on 2 consecutive days at each time point. Bland-Altman plots were generated and data were analyzed using SAS software (v. 9.4, SAS Institute Inc.) with a paired t-test to evaluate mean bias between the methods and with Pearson correlations to determine the agreement between methods and between the means and differences of the 2 methods. Lin's concordance correlation coefficient (CCC) was calculated to assess the agreement between the 2 methods. Significant correlation was found between the orts- and TMR-adjusted diets for DM, starch, and aNDFom TTAD and excellent agreement, as measured by CCC, between the methods for all nutrient TTAD. Evidence suggested that the mean difference between methods was different from 0 for DM, starch, and aNDFom TTAD; however, the mean bias was minimal. In addition, we did observe a correlation between the bias and the magnitude for DM and starch TTAD, but not for aNDFom. Overall, there was excellent agreement between methods, and the mean bias and correlation between the difference and mean TTAD values observed are minimal and likely not to be analytically significant. In conclusion, analyzing the nutrient composition of the orts may not be necessary when calculating TTAD for lactating cow diets.