Effects of partial mixed rations and supplement amounts on milk production and composition, ruminal fermentation, bacterial communities, and ruminal acidosis.

Late-lactation Holstein cows (n=144) that were offered 15kg dry matter (DM)/cow per day of perennial ryegrass to graze were randomized into 24 groups of 6. Each group contained a fistulated cow and groups were allocated to 1 of 3 feeding strategies: (1) control (10 groups): cows were fed crushed wheat grain twice daily in the milking parlor and ryegrass silage at pasture; (2) partial mixed ration (PMR; 10 groups): PMR that was isoenergetic to the control diet and fed twice daily on a feed pad; (3) PMR+canola (4 groups): a proportion of wheat in the PMR was replaced with canola meal to produce more estimated metabolizable protein than other groups. Supplements were fed to the control and PMR cows at 8, 10, 12, 14, or 16kg of DM/d, and to the PMR+canola cows at 14 or 16kg of DM/d. The PMR-fed cows had a lower incidence of ruminal acidosis compared with controls, and ruminal acidosis increased linearly and quadratically with supplement fed. Yield of milk fat was highest in the PMR+canola cows fed 14 or 16kg of total supplement DM/d, followed by the PMR-fed cows, and was lowest in controls fed at these amounts; a similar trend was observed for milk fat percentage. Milk protein yield was higher in the PMR+canola cows fed 14 or 16kg of total supplement DM/d. Milk yield and milk protein percentage were not affected by feeding strategy. Milk, energy-corrected milk, and milk protein yields increased linearly with supplement fed, whereas milk fat percentage decreased. Ruminal butyrate and d-lactate concentrations, acetate-to-propionate ratio, (acetate + butyrate)/propionate, and pH increased in PMR-fed cows compared with controls for all supplement amounts, whereas propionate and valerate concentrations decreased. Ruminal acetate, butyrate, and ammonia concentrations, acetate-to-propionate ratio, (acetate + butyrate)/propionate, and pH linearly decreased with amounts of supplement fed. Ruminal propionate concentration linearly increased and valerate concentration linearly and quadratically increased with supplement feeding amount. The Bacteroidetes and Firmicutes were the dominant bacterial phyla identified. The Prevotellaceae, Ruminococcaceae, and Lachnospiraceae were the dominant bacterial families, regardless of feeding group, and were influenced by feeding strategy, supplement feeding amount, or both. The Veillonellaceae family decreased in relative abundance in PMR-fed cows compared with controls, and the Streptococcaeae and Lactobacillaceae families were present in only minor relative abundances, regardless of feeding group. Despite large among- and within-group variation in bacterial community composition, distinct bacterial communities occurred among feeding strategies, supplement amounts, and sample times and were associated with ruminal fermentation measures. Control cows fed 16kg of DM of total supplement per day had the most distinct ruminal bacterial community composition. Bacterial community composition was most significantly associated with supplement feeding amount and ammonia, butyrate, valerate, and propionate concentrations. Feeding supplements in a PMR reduced the incidence of ruminal acidosis and altered ruminal bacterial communities, regardless of supplement feeding amount, but did not result in increased milk measures compared with isoenergetic control diets component-fed to late-lactation cows.