Influence of carbohydrates on feed intake, rumen fermentation and milk performance in high-yielding dairy cows

Food for human consumption originates directly from plants, after processing, or indirectly by conversion of plant materials into food of animal origin through livestock. An important example of food of animal origin are dairy products such as milk, cheese, butter, yoghurt, etc.During the last decades milk production from Dutch dairy herds has increased considerably. This increase in production, yield and content, was the result of a combination of improvements in genetic potential, due to breeding and progress made in nutrition. Within the area of nutrition better quality roughage and increased usage of concentrates were at the root of this progress.Initially the concentrates were based mainly on grains and oil cakes. These feedstuffs showed, slight variation in energy and protein value. Large quantities of by-products, varying in origin, are produced as a consequence of food production. Without alternative use, these by-products would be wasted and as such provide an additional source of environmental pollution. Fortunately, many by-products of food production have considerable value as feedstuffs and are used in ruminant nutrition.For the profitable utilization of by-products in dairy cow feeding a precise evaluation of their feeding value is essential, because of the high degree of variation involved. Product classification according to the processing methods is a help, but considerable emphasis should be given to chemical composition and predictive methods of evaluation using this or alternative sources of information. Some of these by-products are produced using high moisture techniques. These products sometimes become available after artificially drying or are delivered with a high moisture content and have to be ensiled.One of the chemical components, which varies the most in by-products is the carbohydrate fraction. Sometimes, some of the carbohydrates (starch, sugars) are removed during processing, increasing the concentration of cell wall constituents, ash, fat and protein content in the remaining product. By selecting and combining various by- product ingredients, large variations in the carbohy drate composition of concentrates can be achieved, even when feeding isocaloric diets. Although these concentrates are similar in their energy value, the fermentation pattern in the rumen may differ widely. If the products contain large quantities of starch such concentrates may even cause rumen acidosis, because of their rapid fermentation and accumulation of lactic acid in the rumen. By replacing starch with highly digestible cell wall constituents the pattern changes in favour of more acetic acid instead of propionic and/or lactic acid. These changes in fermentation pattern had a positive influence on total DM intake, which resulted in an increase in milk yield (Chapter 2).In a subsequent study, comprising a feeding trial (Chapter 3) and fermentation study (Chapter 4) an investigation was made of the influence of the dry matter content of some concentrates (dried versus pressed ensiled beet pulp) and replacements of some concentrates with extra roughage (maize silage). Total DM intake was highest on the dried beet pulp diet. Milk production did not differ between treatments, as with fat yield and content. Milk protein yield and content tended to be lowest for the group fed maize silage. The rumen fermentation study showed similar concentrations of major VFA's on all diets. The concentration of branched-chain fatty acids and ammonia were highest for the cows fed maize silage, indicating reduced microbial protein synthesis, which confirmed the tendency towards lower milk protein yield measured in the feeding trial. The degradation characteristics of the OM showed the lowest rate with maize silage. The undigestible fraction (U) was highest with maize silage. The results of both experiments demonstrated the importance of the balance between energy and protein availability for rumen fermentation.Higher levels of intake caused an increase in the concentration of volatiles and reduced rumen fluid pH. Feed intake level influenced the pattern of consumption, which changed from an intake pattern of two large meals at low intake level, towards several smaller meals at the high level of intake. Although concentrations were highest at the high intake level, diurnal variation was highest on the low level of intake, due to the meal size.Grass silage can be fed as wilted or wet ensiled material. In the latter case some carbohydrates will be replaced by fermentation end products during the ensiling process, reducing the amount of easily fermentable carbohydrates and increasing the soluble fraction and rate of N degradation (Chapter 6). In a feeding trial (Chapter 5), in which a comparison was made between the DM content and the amount of fermentation end products, reduced DM intakes were found on diets low in DM content as well as those high in fermentation end products. The effects of DM content as such were minor, compared to those of the combination of low DM content and increased amounts of fermentation end products. Reduction in DM intake, due to a low DM content, was restricted to the first 6 weeks of lactation, whereas the combination effect remained throughout the experiment. Total DM intake was lowest for high moisture diets, which was reflected in lower energy and protein intakes. Milk production was lowest on high moisture diets, reflecting the lower energy intake. Milk fat content and yield were not affected, partly because of the increased mobilization of body fat during the period of negative energy balance. Milk protein content and yield were lowest on high moisture diets. The fermentation and kinetic studies (Chapter 6) showed reduced pH and increased concentrations of total VFA, acetic acid, ammonia and BCFA on high moisture diets. The rates of clearance of the OM fraction were significant or showed a tendency towards lower values for high moisture diets, which agreed with the lower total DM intake found in the feeding trial. These results were negatively influenced by the type of concentrates fed in both experiments, a very low amount of easily fermentable carbohydrates were fed (ensiled by-products low in starch and sugars). Animals fed these diets were unable to balance the availability of nitrogen and energy for microbial protein synthesis. This resulted in a lower yield of milk protein and a reduction in milk protein content.An attempt was made to compare the effects on feed intake and milk performance of the type of carbohydrate (starch versus cell wall constituents) and rate of degradation (rapidly versus slowly). This was performed in a feeding trial (Chapter 7) accompanied by fermentation and kinetic studies (Chapter 8). Starch reduced milk fat content, which in the case of rapidly fermentable starch can be explained by a decrease in the ratio of non- glucogenic to glucogenic volatile fatty acids (NGR) found in the rumen. The shift from fermentation in the rumen towards digestion in the small intestine, increase in by-pass starch, partly eliminated the lower NGR in the rumen, but the total amount of glycogenic precursors was found to be higher in animals fed by-pass starch. Milk fat content was highest from animals fed high levels of cell wall constituents. Milk protein content was highest on both starch diets, while the diets rich in cell wall constituents displayed the lowest values. Rumen fermentation characteristics showed an increase in the concentration of ammonia and BCFA and a tendency towards lower amounts of microbial protein in the rumen, which confirmed the results of the feeding trial. Differences occurred between both cell wall diets, because one of these diets did not have an optimal balance between cell wall and nitrogen degradation rates.In the general discussion (Chapter 9) an attempt is made to predict feed intake, milk yield (lactose), milk fat and milk protein production. Different models were used to predict the total DM intake as measured in the feeding trials. Relationships were poor, when predicting DM intake using the energy requirements and the energy density per kg DM. Relationships were improved when body weight, stage of lactation and percentage concentrate in the diet were included as independant variables. Results were much improved, when DM intake was predicted by means of equations derived from rumen kinetic parameters. However, all predictions were poor for diets with a dry matter content below 35 percent. This was due to the lower intake measured during the first weeks of lactation, and to the fact that these types of diets were not included in the data set, from which the equations were derived. An attempt was made to estimate the production of nutrients using a rumen fermentation model. Differences in predicted nutrient supply occurred between the diets fed in the different experiments described in the various chapters. The major differences were found between starch-rich diets and diets rich in cell wall constituents in the amount of glycogenic precursors (propionate and glucose) and from the total level of feed intake, which increased the amount of propionic acid produced in the rumen. Relationships between rumen fermentation parameters (NGR) and milk fat content were adequate within experiments. However, relationships between experiments were poor. Large between treatment differences were found in the amount of ketogenic precursors available for milk fat production from mobilization of body reserves. Estimation of the amount of milk fat produced from rumen ketogenic nutrients (acetic and butyric acid) by de novo synthesis and the production of ketogenic nutrients in the rumen predicted by the model showed a good relationship. Differences between total milk fat production and de novo synthesis could be explained by the availability of longchain fatty acids (LCFA) available to the animal directly from the feed or from mobilization of body reserves. Milk protein was related to the amount of DVE available for milk production. The DVE intake between treatments was related to the microbial protein synthesis in the rumen and could be explained by the balance between nitrogen and energy sources. Milk lactose was related to the availability of glycogenic precursors and showed a very low variation between treatments indicating that the high yielding dairy cow in early lactation is probably hormonally geared towards the production of milk lactose, which might even be preferred above the production of milk protein, due to gluconeogenesis. Although the results of the feeding trials could be explained afterwards from the total energy intake, milk production and milk composition and the deposition and/or repositioning of body reserves, prediction of milk yield and composition was impossible. However, when including the chemical composition of the diets, its rate of degradation, the kinetics of the dietary ingredients and the extent of deposition and reposition, it became possible to predict milk production and composition.In conclusion it can be stated that milk composition could be explained reasonably well from the estimated supply of individual nutrients. However, it should be realised that more information is required concerning the partitioning of nutrients to the various tissues in the animal, and the relationships involved with fermentation and digestion before evaluation based on net energy (VEM) and protein absorbed from the small intestine (DVE) can be replaced by more accurate predictive methods for practical use.