Effects of grain variability and processing on starch utilization by lactating dairy cattle

Variation in site of digestion and lactation performance among lactating dairy cows fed different grain sources was evaluated and reviewed. The enzymatic assay for starch contributes to variation among and possibly even within experiments. Genetic differences within grain source affect digestibility, but effects of grain processing are considerable. When quantifying response criteria among numerous experiments with varying experimental procedures, the variation among experiments is very high and should be considered in regressions used to evaluate treatment responses. Some of the variation among experiments could be explained by variable DMI (positive relationship with milk production and microbial N flow to the duodenum but negative relationship with milk fat percentage and ruminal digestibility of starch) and other chemical measurements of dietary composition (especially forage and NDF percentages of the diet). Other sources of variation that could not be quantified among experiments were accounted for in regression models as experiment effects. After these adjustments, least squares means were calculated as the average of all effects remaining in the models. Our results documented that processing procedures to enhance the ruminal degradability of starch generally decreased NDF digestibility by a lesser degree, partially negating responses in true ruminal OMdigestibility. However, lactating cows fed diets with low ruminal degradability of starch had higher, nearly complete compensatory digestion of starch in the hindgut, thereby resulting in relatively minor improvements in total tractOMdigestibility or lactation performance for the more highly available starch sources, assuming similar DMI. These results highlight the importance of maximization of DMI compared with grainprocessing method and the need for more production experiments comparing grain sources at similar intakes of ruminally digestible starch (i.e., lower percentages in the diet for more highly degradable starch sources) for better evaluation of overall efficacy. In particular, the long-term responses and potential for residual responses over time (i.e., treatment × time interactions) are needed. Until better mechanistic prediction of lactation response is available, the least squares means of ruminal and total tract digestibilities of nutrients should be useful for empirical formulation and evaluation of dairy rations.