Five primiparous Holstein cows (55 d in milk) that were fitted with ruminal and duodenal cannulas were used in a 4 x 5 incomplete Latin square to determine the effects of blends of steam-flaked and dry-rolled corn on site and extent of nutrient digestion and milk yield and composition. Diets were fed as total mixed rations and consisted of 45% forage and 55% concentrate; each diet contained 27% corn grain. Dietary treatments were composed of blends of dry-rolled and steam-flaked corn in ratios of 100:0, 67:33, 33:67, and 0:100. Intake of dry matter; digestibilities of dry matter, organic matter, acid detergent fiber, cellulose, neutral detergent fiber, fatty acids, and N; and microbial efficiency were unaffected by diet. Ruminal, postruminal, and total tract digestion of starch increased linearly, and starch passage to the duodenum decreased linearly, as the proportion of dry-rolled corn in the diet decreased. Ruminal propionate and valerate increased linearly, and acetate, butyrate, isovalerate, and the acetate to propionate ratio decreased linearly, as proportions of dry-rolled corn in the diet decreased; however, no changes in total volatile fatty acid concentrations in ruminal fluid were observed. Ruminal fluid pH was similar across diets. A decrease in dry-rolled corn decreased ruminal ammonia N and plasma urea N linearly. Milk yield and composition, as well as milk N fractions, were similar across diets. Although changes in fatty acid composition of milk fat were small, linear decreases in percentages of trans-C16:1 and cis-9- and cis-10-C18:1, as well as a linear increase in the percentage of C18:2 occurred as the proportion of dry-rolled corn in the diet decreased. An increased proportion of dry-rolled corn in the diet decreased digestion of starch in the rumen, and patterns of volatile fatty acid concentrations shifted accordingly. However, no effects on lactational parameters were observed.

The influence of insoluble ash in diet, digesta, and fecal samples was evaluated. Ash-free values of neutral detergent fiber (NDF) and acid detergent fiber (ADF) for feed, digesta, and fecal samples were compared with uncorrected values of NDF and ADF in two digestibility experiments. Six feed samples were selected at random along with four duodenal and two fecal samples; the latter two were a part of two digestion studies. The silica composition of feed samples was low with the exception of oat hay and rice straw. Ash-free values of NDF and ADF for digesta and fecal samples were lower than uncorrected values of NDF and ADF, and silica accounted for the majority of the differences. The use of ash-free values of NDF and ADF in both digestibility experiments with dairy cows resulted in more realistic digestibility estimates and reduced the amount of variability associated with the calculation of the digestibility of NDF and ADF in the rumen and total digestive tract. The source of the insoluble ash was not identified but could have come from individual feedstuffs or the consumption of sand. Sample contamination by silica and other ash components may influence NDF and ADF composition if the fiber residues are not ashed prior to calculation of NDF and ADF.