Modulation of starch digestion for productive performance in dairy cows

[摘要] ENGLISH ABSTRACT: In this thesis, a series of experiments were conducted to ultimately investigate the effects of modulating site of starch digestion by varying ruminal fermentability of various starch sources on feed intake, production and metabolic response of transition cows. The first experiment was performed (Chapter 3) to evaluate the effects of different particle sizes on chemical composition and in vitro ruminal starch degradability of cereal grains commonly used in dairy cow diets. Four starch sources (maize 1 and 2, sorghum, barley and wheat) were ground through 1 and 2-mm screens and fractionated by sieving to obtain the following sizes: <250 (very fine), 250-500 (fine), 500-1180 (medium) and 1180-2000 μm (coarse). The generated particle size fractions and unsieved samples were separately analysed for chemical composition and fermented in vitro using rumen fluid for 0, 3, 6, 9, 12 and 24 h to determine starch degradability (Sd), and rate of starch degradation (kd), assuming a first order decay. Particle size affected (P<0.0001) the chemical composition of all grains, with the highest starch in the smallest particles and highest NDF in the largest particles. For all grains, Sd and kd increased with decreasing particle size. Results from this in vitro study suggest that starch digestion could be potentially shifted post-rumen by controlling particle size and reducing the amount fermented in the rumen. In Chapter 4 (experiment 2), we compared two mathematical approaches for determining the rate of starch degradation. The objective was to evaluate the accuracy and precision of the 7 h-kd's by comparison with rates obtained using a non-linear first order decay model as a reference. Higher accuracy and precision were obtained by using a non-linear estimation. There is a need for using a non-linear estimation, using multiple time points or the development of alternative estimations, especially when quantifying rates of starch ddegradation for high producing cows. Experiment 3 was performed (Chapter 5) to quantify the potential of a starch binding agent (BioProtect™) to reduce in vitro rumen starch degradation of cereal grains of varying particles size. Maize and sorghum fractions used in experiment 1 were treated by spraying with BioProtect™ 24 h before in vitro fermentation to quantify starch degradability (Sd). Both treated and untreated (no BioProtect™) maize and sorghum samples were fermented in vitro. BioProtect™ was effective in decreasing starch degradability for both grains, with effects more pronounced for smaller particle sizes, by reducing Sd 17%-units compared to 7%-units for the largest particles. Simulations with the NDS software indicated that the use of BioProtect™ can reduce rumen starch digestibility, increase rumen starch escape and post rumen starch digestibility. Simulated total tract digestibility was not decreased by the use of BioProtect™ and indicated slightly reduced microbial protein production. Although BioProtect™ showed positive effects on reducing rumen starch degradation, in our simulations, larger particles were more effective at shifting the site of digestion and it could, therefore, be a more cost-effective option for our aim.Based on the in vitro results, two starch sources were selected for further in vivo investigation, to study a possible shift in the site of starch digestion. In experiment 4 (Chapter 6) the effects of starch sources and particle sizes on digesta flow, starch digestibility, ruminal fermentation parameters and production performance of dairy cows were investigated. Four ruminally-cannulated multiparous Holstein cows were used in a 4 × 4 Latin square design with a 2 x 2 factorial arrangement of treatments: maize or sorghum (M or S) either finely or coarsely ground (using a 1- or 4-mm screen sieve, F or C). Diets were formulated to contain similar starch concentration. Digesta flow was quantified using the reticular sampling technique, applying the triple-marker method. Dry matter (DM) intake, milk yield and composition were not affected by dietary treatments in exception of MUN. Milk urea nitrogen concentration was lower for cows fed maize diets: 14.36, 14.89, 16.99 and 17.09 mg/dL for MF, MC, SF and SC, respectively. Rumen pH and reticulum pH were higher for the SC diet (6.20 and 6.56, respectively) when compared to the other treatments. Rumen and reticulum pH were 5.98 and 6.33 for MF, 5.96 and 6.32 for MC, and 5.92 and 6.36 for SF, respectively. Propionate concentration was greater for both maize diets (33.21 and 32.96 vs. 31.22 and 28.68 mM; P < 0.0001) and ruminal ammonia N was lower for the fine maize diet compared to the SF and SC diets. Dietary treatments did not affect (P > 0.05) organic matter (OM) and NDF intake, nutrient flow of DM, OM and NDF, or ruminal digestibility of OM. Starch from the coarser maize was less ruminally digested (83.76 vs. 88.77% of intake) and had a greater flow to the abomasum when compared to the fine particles (1.04 vs 0.76 kg/d). However, the apparent total-tract digestibility of starch was greater in MF than MC cows (96.29 vs. 87.84%). This study confirms that coarser particles can allow part of starch digestion to be shifted from the rumen to the small intestine, but total tract starch digestibility could be decreased if ruminal digestion is not compensated postruminally. The objective of experiment 5 (Chapter 7) was to evaluate the effects of starch fermentability of diets fed during the early postpartum (PP) period on feeding behaviour, dry matter intake (DMI), lactation performance and body metabolism of fresh dairy cows. Jersey cows (n =117) were used in a randomized complete block design. Treatment diets were formulated to similar starch concentration, with ground maize (3 or 6-mm screen sieve) as the primary starch source. Treatments were fed as TMR from calving to 30 d PP before switching to a common lactation diet. Throughout the experiment DMI, milk yield and body weight were recorded daily, and milk composition, body condition score (BCS) and blood metabolites were measured weekly. Feeding coarsely ground maize (MC) increased dry matter intake (16.08 vs. 17.13 kg/d) and milk yield (20.41 vs. 21.70 kg/d) compared to finely ground maize (MF). Diet did not affect (P > 0.05) eating and rumination time and had no effects on milk composition, but milk lactose was increased in the MC compared to the MF diet (4.70 vs. 4.61%) and milk fat percentage tended to be greater (5.57 vs. 5.27%) in the MF than MC diet. Decreases in BW and BCS were greater in cows fed the MF (39.92 vs 32.24 kg and 0.23 vs. 0.14 units) than in cows fed the MC diet, resulting in increased plasma NEFA concentration (0.71 vs. 0.56 mmol/L) in MF cows. Blood glucose levels were not affected (P > 0.05). The increased DMI in cows fed the MC diets could possibly be attributed to reduced production of propionate in the rumen, resulting from shifting starch digestion postruminally and by the decreased plasma NEFA concentration.Overall, in conclusion, our results confirm that starch digestibility increases with decreasing particle size, suggesting that starch digestion could be potentially shifted post-rumen by controlling the grain particle size fed and thus reducing the amount fermented in the rumen. BioProtect™, a starch binding agent was effective in reducing in vitro rumen starch degradation, with effects more pronounced for smaller particle sizes. Shifting the site of starch digestion postruminally in early postpartum cows increased DMI, milk production and decreased mobilization of body reserves as indicated by the decreased concentration of plasma NEFA. The results of this study apparently support the hepatic oxidation theory of the control of feed intake, particularly during the early postpartum period. More processing alternatives should be investigated to reduce loss of digestibility postruminally for larger particles.

[发布机构] Stellenbosch University