Determining the influence of dietary roughage concentration and source on ruminal parameters related to sulfur toxicity.

Cattle feedlot diets often include ethanol coproducts that provide excess dietary sulfate, which is reduced to sulfide by ruminal bacteria and can be converted to hydrogen sulfide, which has been correlated to S toxicity. The objective of this study was to determine the impact of feeding varying concentrations of NDF from chopped cornstalks (CS) or chopped bromegrass hay (BH) on ruminal pH, ruminal H2S concentration, and DMI of steers fed a high-S finishing diet. Five ruminally fistulated steers (595 ± 87 kg BW) were used in a 6 × 6 Latin square with 14-d periods and fed diets containing 0.45% S from a mixture of dried distillers grains and condensed corn distillers solubles. The study was a 2 × 3 factorial arrangement of treatments with 2 roughage sources--CS or BH--and 3 concentrations of added roughage NDF (rNDF)--4, 7, or 10%. Steers had individual ad libitum access to feed and adapted to each diet for the first 7 d of each period. Effective NDF linearly increased (P < 0.01) as rNDF increased and did not differ between sources (P = 0.44). There was no effect of concentration or source of rNDF on DMI (P ≥ 0.69). Steer behavior was observed on d 13 of each period for 3 h postfeeding. Source of rNDF did not affect time at bunk, DMI during observation, or rate of DMI (P ≥ 0.42). Time at bunk linearly increased as rNDF increased (P = 0.01), while rate of DMI linearly decreased (P = 0.02). Area under the curve for ruminal pH of 5.4, 5.6, and 5.8, calculated using data from d 8 to 14 via an indwelling ruminal bolus, were linearly decreased (P ≤ 0.03) as rNDF increased. Manual ruminal pH taken 6 h postfeeding on d 14 of each period did not differ by source (P = 0.12) but linearly increased (P < 0.01) as rNDF increased. Ruminal H2S concentrations measured 6 h postfeeding on d 14 of each period did not differ by source (P = 0.47) but linearly decreased (P < 0.01) as rNDF increased (0.62, 0.35, 0.31 g/m(3) for 4, 7, and 10% rNDF, respectively). A segmented linear model was found to best fit the ruminal pH and H2S relationship data, suggesting that at or below a pH of 5.6 ± 0.08 with 95% confidence intervals of 5.4 and 5.8, pH is a strong predictor of H2S (P ≤ 0.05), while above this pH range H2S concentrations are not well correlated with ruminal pH (P > 0.50). In conclusion, adding at least 7% NDF from CS or BH to high-S feedlot cattle diets will increase ruminal pH and decrease H2S concentrations, thus decreasing potential for S toxicity.