Characterization of fecal pH in commercial finishing cattle relative to feeding time

Abstract Fecal pH has been proposed as a practical indicator of digestive health in cattle. However, variation in fecal pH within or across pens in a commercial feedlot environment has not been defined. The objectives of this study were to 1) quantify within- and between-pen variation in fecal pH, 2) explore whether fecal color could predict pH using digital image analysis. The trial was conducted at a commercial feedyard using six pens (81 ± 8 head/pen) of finishing cattle fed a common total mixed ration three times daily. On three consecutive days, ten freshly voided fecal pats per pen (n = 180) were photographed, collected, and analyzed for pH. Feeding time and sampling time were collected to assess potential time related effects that innately occurred due to the commercial feedlot environment the study was conducted in. Sample collection ranged from 77 minutes prior to and 109 minutes after the first daily feeding, depending on pen. Feed ingredients and TMR samples from each pen were collected and analyzed in addition to fecal samples. Each sample was homogenized 1:1 (wt/vol) with distilled water and measured using a handheld probe. Fecal images were analyzed for color metrics, including RGB, HSV, and L*a*b*, along with calculated luminosity and luminance using the RGB color model. Data were analyzed as repeated measures in R (version 4.4.2) using linear mixed models with fixed effects of time relative to feeding and day, and a random effect of pen. Mean fecal pH across all samples averaged 6.21 ± 0.14 by pen and increased (P = 0.018) with time after feeding, indicating an increase in fecal pH value post-feeding. Sampling day did not affect pH (P 0.10), and Levene’s test confirmed equal variance among pens (P 0.10). Within-pen variability, expressed as daily SD (average by pen 0.28 ± 0.10) and range (average by pen 0.90 ± 0.38), was also unaffected by sampling time or day (P 0.10). When fecal color parameters were included in the model, b*, a*, and S were significant predictors of pH (P = 0.006, 0.009, and 0.03, respectively). Predicted pH values derived from these color metrics were weakly correlated with measured pH (Lin’s CCC = 0.182, 0.170, and 0.178 for b*, a*, and S, respectively), suggesting that color attributes describe characteristics associated with but not directly reflective of fecal pH. Collectively, these findings indicate that fecal pH modestly increased following feeding while within-pen variability remained stable. Within-pen variation along with diurnal changes in pH relative to feeding time suggests that consistent sampling time is critical if attempting to develop fecal pH as an on-farm indicator of digestive health of commercial feedlot cattle.