Estimating ruminal crude protein degradation with in situ and chemical fractionation procedures

Predicting nitrogen excretion of cattle kept under tropical and subtropical conditions using semimechanistic models

The present study aims at evaluating whether current semimechanistic models developed for temperate cattle systems can be adopted for cattle under (sub-) tropical husbandry systems to adequately (accurately and precisely) predict total nitrogen (TN), urine nitrogen (UN), faecal nitrogen (FN) excretion and its partition into different FN fractions. Selected models were built based on the feeding recommendations for ruminants of the British (Model A), German (Model G) and French (INRA; Model I) system. Model evaluation was conducted using eight nitrogen balance studies performed in El Salvador, Kenya and Peru (n = 392 individual observations including lactating cows, heifers and steers). Concordance correlation coefficient, root mean square errors (RMSE), and mean biases were estimated to evaluate the models' adequacy in predicting nitrogen excretion. Input variables causing greatest variation in nitrogen excretion prediction were identified by a sensitivity analysis and adjusted. Model G was able to adequately (i.e., RMSE of <25% of observed mean, systematic error of <5% of the mean square error) predict TN excretion through a compensation between overestimation of UN excretion and underestimation of FN excretion. None of the models were able to adequately predict UN, FN, and different FN fractions. Model I adequately predicted FN (RMSE = 18%) when duodenal microbial crude protein flow was increased, and the intercept used to predict FN excretion was reduced from 4.30 to 3.82 g of nitrogen per kilogram of dry matter intake. These adjustments, however, were not sufficient to predict adequately UN excretion (RMSE = 38%), individual FN fractions (RMSE > 56%), and TN (RMSE = 22%) excretion, by Model I.

Effect of Cereals and Legumes Processing on In Situ Rumen Protein Degradability: A Review

The determination of the ruminal degradability rate of feeds, mainly starch and crude protein, is one of the most common methods to evaluate the nutritional value of ruminant feed. The protein requirements for ruminants are met from microbial protein and undegraded dietary protein digested in the small intestine. In order to reach maximum productivity, high-quality proteins are needed, and the requirement for undegraded dietary protein increases with the performance of the animal. This protein can be supplied by reducing the ruminal degradation to increase the amount of protein digested post-rumen, but the form in which a feed is administered influences degradability, and grain processing, especially, is a common practice to improve feed efficiency. Despite these aspects, studies on the effects of feed processing methods on protein degradability are limited, even though more and more ruminants are fed with processed feeds. For these reasons, this review investigated the protein degradability of different processed cereals and legumes in ruminants based on the analysis of available literature in order to take stock of the state of the art on this topic. Results showed that: First, the majority of the papers are focused on the energy aspects mainly due to carbohydrate-rich feeds; second, the majority of the studies in the literature are quite old, probably because the changes occurred in the animal testing legislation that made in vivo studies more and more difficult in the last 20 years; third, as a consequence, the few data available in recent years concern in vitro experiments; fourth, we found a high variability of the experimental conditions thus affecting protein degradability and making it quite difficult to compare the different results.

The effects of rumen nitrogen balance on nutrient intake, nitrogen partitioning, and microbial protein synthesis in lactating dairy cows offered different dietary protein sources

The aim was to study the effects of rumen N balance (RNB), dietary protein source, and their interaction on feed intake, N partitioning, and rumen microbial crude protein (MCP) synthesis in lactating dairy cows. Twenty-four lactating Holstein cows were included in a replicated 4 × 4 Latin square experimental design comprising four 20-d periods, each with 12 d of adaptation to the experimental diets and 8 d of sampling. The dietary treatments followed a 2 × 2 factorial arrangement (i.e., 4 treatments) with 2 main protein sources [faba bean grain (FB) and SoyPass (SP; Beweka Kraftfutterwerk GmbH, Heilbronn, Germany)] offered at 2 dietary RNB levels each [0 g/kg of dry matter, DM (RNB0) and -3.2 g/kg of DM (RNB-)]. The RNB was calculated as the difference between dietary crude protein (CP) intake and the rumen outflow of undegraded feed CP and MCP and divided by 6.25. Composition of concentrate mixtures was adjusted to create diets with desired RNB levels. Each of these protein sources supplied ≥35% of total dietary CP. Both diets for each protein source were isoenergetic but differed in CP concentrations. The DM intake (kg/d) was lower for RNB- than for RNB0 in diets containing FB, whereas no differences were seen between the RNB levels for SP diets. The RNB- decreased N intake and urinary N excretion but increased milk N use efficiency in both FB and SP diets, with greater differences between the RNB levels for FB diets than for SP diets. Similarly, duodenal MCP synthesis (g/kg of digestible organic matter intake) estimated from purine derivatives in the urine was lower for RNB- than for RNB0 in FB diets but similar between the RNB levels in diets containing SP. Low RNB of approximately -65 g/d (approximately -3.2 g/kg of DM) in diets reduced feed intake, N balance, and performance in high-yielding dairy cows with possibly more pronounced effects in diets containing rapidly degradable protein sources.

Ruminal fermentation characteristics and related feeding values of compound feeds and their constituting single feeds studied by using in vitro techniques

Single concentrate feeds are mixed together forming compound feeds for cattle. However, knowledge regarding the potential interactions (associative effects) between the feeding values of single feeds in compound feeds is lacking. The main objective of the present study was to evaluate ruminal fermentation characteristics and feeding values of eight industrially produced compound feeds in mash form from their constituent single feeds for dairy cows through in vitro assays. Additivity was given for gas production (GP), digestibility of organic matter (dOM) and utilisable CP at the duodenum (uCP). Additivity of CP fractions (determined using the Cornell Net Carbohydrate and Protein System (CNCPS)) was dependent on the fraction and compound feed type; however, the effective degradation calculated from CP fractions (EDCNCPS) showed additivity. Additivity was not given for intestinal digestibility of rumen-undegraded protein (IDRUP) for five out of eight compound feeds. Precise calculation of metabolisable energy (ME) of compound feeds from ME of single feeds was possible when using the same ME equations for all single and compound feeds. Compound feeds are often provided in pellet form; therefore, our second objective was to evaluate the effects of pelleting on ruminal fermentation characteristics and feeding values of compound feeds. Pelleting affected GP at 24 h (GP24; up to 2.4 ml/200 mg DM), dOM (up to 2.3 percentage point (pp)) and ME (up to 0.3 MJ/kg DM), but these differences were overall small. More considerable effects of pelleting were observed for uCP, which was increased in all compound feeds except the two with the highest CP concentrations. The IDRUP was lower in most compound feeds following pelleting (up to 15 pp). Pelleting also affected CP fractions in a non-systematic way. Overall, the effects of pelleting were not considerable, which could be because pelleting conditions were mild. Our third objective was to compare in situ ruminal CP degradation (EDIN_SITU) of compound feeds with ED using two prediction methods based on CP fractions. EDIN_SITU reference data were obtained from a companion study using the same feeds. Prediction accuracy of EDIN_SITU and EDCNCPS was variable and depended on the compound feed and prediction method. However, future studies are needed as to date not enough data are published to draw overall conclusions for the prediction of EDIN_SITU from CP fractions.

Ruminal protein digestibility of Australian produced oilseed meals

Initially, samples of Australian canola, soybean, cottonseed and flaxseed meal produced by solvent-extraction, expeller and cold-press technologies collected from late 2014 to early 2015 were analysed for general chemical composition, protein and ruminal digestibility characteristics. The oilseed meals had levels of ash, neutral-detergent insoluble crude protein, total intestinal digested protein, B1, B2, B3 and C protein content similar to those in previous reports, but lower Fraction A (non-protein N) levels than in previous reports. Acid-detergent insoluble fibre, metabolisable energy, total digestible nutrients, ash (P &lt; 0.05), neutral-detergent fibre, in vitro dry matter digestibility, crude protein (CP), true protein, acid-detergent insoluble CP, soluble protein, in vitro rumen-undegradable protein (RUP), total tract digested protein, Fraction A, B2 and C (P &lt; 0.01) differed among oilseed types. Dry matter, the ratio of RUP to total tract digested protein (P &lt; 0.05), CP, lipid, soluble protein, RUP, Fraction A, B1 and B2 (P &lt; 0.01) differed among oil-extraction techniques. Utilising an in vitro simulated rumen proteolysis procedure by Krishnamoorthy et al. (1983), mean ranges of in vitro RUP were greater and varied more so in canola (18.8–70.1%) than soybean (40.3–54.0%), cottonseed (31.5–33.9%) and flaxseed (18.5–21.8%) meals. Estimated RUP was lower (P &lt; 0.01) in cold-press than expeller and solvent-extraction oilseed meals.

Expeller Barrel Dry Heat and Moist Heat Pressure Duration Induce Changes in Canola Meal Protein for Ruminant Utilisation

Simple Summary

Canola meal, a by-product of oil production from canola seed, is a source of protein commonly incorporated into dairy and feedlot rations. Processing conditions and pressure treatments can alter the quality of protein in canola meal. In this study, the impact of expeller dry heat and moist heat pressure duration time on general nutritional properties, in vitro protein degradability, Maillard reaction product formation, and molecular and microscopic structural characteristics of canola meal were investigated. Increased dry heat temperature rapidly increased digestible protein and non-protein nitrogen content, and constricted amide II secondary structure. Increased moist heat pressure treatment duration promoted browning, and the conversion of protein to more intermediately and slowly degradable forms. Dry heat and moist heat pressure affected meal protein solubility and protein and lipid-related functional groups. Moist heat pressure fragmented canola meal into enzyme-resistant aggregates with crevices containing oil bodies. Induced changes may impact the supply of protein and amino acids and subsequently the yield and composition (protein and lipid) of milk produced by dairy cows. These findings benefit producers of canola meal by further describing the effects of processing and treatment conditions on protein characteristics, particularly those which affect the production potential of ruminants fed canola meal as a source of protein.

Abstract

To improve the protein nutritional quality of canola (Brassica napus L.) meal, further investigation of the effects of processing conditions and post-production treatments is desirable. The impact of barrel dry heat temperature (20 °C (cold press) and 100 °C (expeller)) and moist heat pressure (MHP) duration time on general nutritional properties, Maillard reaction product (MRP) formation, in vitro protein degradability, and molecular and microscopic structural characteristics of canola meals were investigated. Increased MHP duration reduced (p < 0.05) dry matter, soluble protein, rapidly degradable protein, yellowness (early MRP), whiteness (late MRPs), absorbance at 294 nm (intermediate MRPs), and amide I; and increased (p < 0.05) non-protein N, neutral detergent fibre, neutral detergent insoluble crude protein (CP), intermediately and slowly degradable protein, in vitro effective CP degradability, redness, degree of colour change, and browning. Increased dry heat temperature reduced (p < 0.01) CP and rapidly degradable protein, constricted amide II, reduced (p < 0.05) protein solubility in 0.5% KOH and increased (p < 0.05) acid-detergent fibre and intermediate MRPs. Browning index and redness exhibited potential as rapid indicators of effective CP degradability and soluble protein, respectively. Dry heat and MHP altered (p < 0.05) lipid-related functional groups. Dry heat affected napin solubility, and MHP altered cruciferin and napin solubility. Application of MHP induced the formation of proteolysis-resistant protein aggregates with crevices containing oil bodies. Induced changes may impact the supply of proteins and amino acids and subsequently the yield and composition (protein and lipid) of milk produced by dairy cows.

Protection of protein from ruminal degradation by alkali-induced oxidation of chlorogenic acid in sunflower meal

Lactating ruminants require an adequate supply of absorbable amino acids for the synthesis of milk protein from two sources, that is crude protein (CP) synthesized microbially in the rumen and ruminally undegraded CP (RUP) from feed which can both be digested in the small intestine. Several chemical and physical methods have been identified as being effective in increasing the proportion of RUP of total CP of a feedstuff, yet there is a continuing need for developing and establishing methods which protect feed protein from ruminal degradation with acceptable expenditure of labour and other costs. The objective of this study was to identify and quantify effects of and interactions between chlorogenic acid and protein in solvent-extracted sunflower meal (SFM) as induced by alkali treatment. Response surface methodology was employed to investigate the influence of pH, reaction time and drying temperature on the resulting SFM and, subsequently, its protein value for ruminants estimated from laboratory values. For this purpose, alkali-treated SFM was subjected to a fractionation of feed CP according to the Cornell net carbohydrate and protein system as a basis for estimating RUP at different assumed ruminal passage rates (K_p ). To estimate the intestinal digestibility of the treated SFM and its RUP, a three-step enzymatic in vitro procedure was applied. Alkaline treatment of SFM increased RUP values with factors ranging from approximately 3 (K_p =.08/hr) to 12 (K_p =.02/hr). Furthermore, the intestinal digestibility of the alkali-treated SFM was enhanced by approximately 10% compared to untreated SFM. Increasing pH and reaction time led to both increasing RUP values and intestinal digestibility. In conclusion, a targeted alkaline treatment of naturally occurring compounds in feedstuffs might be a promising approach to provide high-RUP feeds for ruminants which, at the same time, have improved intestinal digestibility values.

Estimation of utilisable crude protein at the duodenum of dried distillers' grains with solubles using a modified gas test

The objective of this study was to characterise the variation of utilisable crude protein at the duodenum (uCP) of dried distillers' grains with solubles (DDGS) for ruminants using a modified gas test and to predict the uCP in DDGS based on chemical composition. Thirteen samples originating from wheat, maize, barley or blends of different substrates were studied. The in vitro uCP was estimated using the modified Hohenheim gas test (moHGT). Samples were incubated in rumen fluid for 8 h, 24 h and 48 h followed by ammonia distillation. The obtained values were compared to reference values of uCP (based on the contents of crude protein (CP), in situ undegraded CP and metabolisable energy). The reference and in vitro values of uCP were calculated according to passage rates of 2, 5 and 8%/h (i.e., uCP2, uCP5 and uCP8, respectively). The in vitro uCP8 ranged from 214 to 320 g/kg DM and reference values from 158 to 302 g/kg DM. The in vitro uCP2 was on average lower (by 7 g/kg DM) and in vitro uCP8 was higher (by 56 g/kg DM) than their respective reference values. The in vitro uCP5 and uCP8 were correlated with reference values and the correlations were improved with increasing passage rates. When the differences of uCP content between in vitro and reference values were related to CP fractions, they increased with increasing content of CP fraction A and decreasing content of CP fraction B3 for uCP8. The prediction of uCP values from chemical composition was not reliable. It was concluded that uCP can be predicted on the basis of the moHGT method and CP fractions. The accuracy of prediction improved upon the inclusion of CP fractions and neutral-detergent insoluble nitrogen. The present study revealed a significant variation in the uCP content of DDGS, which should be considered when formulating rations for dairy cows.

Prediction of crude protein and neutral detergent fibre concentration in residues of in situ ruminal degradation of pasture samples by near-infrared spectroscopy (NIRS)

An important constraint of in situ degradability studies is the need to analyse a high number of samples and often with insufficient amount of residue, especially after the longer incubations of high-quality forages, that impede the study of more than one nutritional component. Near-infrared spectroscopy (NIRS) has been established as a reliable method for predicting composition of many entities, including forages and other animal feedstuffs. The objective of this work was to evaluate the potential of NIRS for predicting the crude protein (CP) and neutral detergent fibre (NDF) concentration in rumen incubation residues of permanent and sown temperate pastures in a vegetative stage. In situ residues (n = 236) from four swards were scanned for their visible-NIR spectra and analysed for CP and NDF. Selected equations developed by partial least-squares multivariate regression presented high coefficients of determination (CP = 0.99, NDF = 0.95) and low standard errors (CP = 4.17 g/kg, NDF = 7.91 g/kg) in cross-validation. These errors compare favourably to the average concentrations of CP and NDF (146.5 and 711.2 g/kg, respectively) and represent a low fraction of their standard deviation (CP = 38.2 g/kg, NDF = 34.4 g/kg). An external validation was not as successful, with R2 of 0.83 and 0.82 and a standard error of prediction of 14.8 and 15.2 g/kg, for CP and NDF, respectively. It is concluded that NIRS has the potential to predict CP and NDF of in situ incubation residues of leafy pastures typical of humid temperate zones, but more robust calibrations should be developed.

Comparison of fractionation methods for nitrogen and starch in maize and grass silages

In in situ nylon bag technique, many feed evaluation systems use a washing machine method (WMM) to determine the washout (W) fraction and to wash the rumen incubated nylon bags. As this method has some disadvantages, an alternate modified method (MM) was recently introduced. The aim of this study was to determine and compare the W and non-washout (D+U) fractions of nitrogen (N) and/or starch of maize and grass silages, using the WMM and the MM. Ninety-nine maize silage and 99 grass silage samples were selected with a broad range in chemical composition. The results showed a large range in the W, soluble (S) and D+U fractions of N of maize and grass silages and the W, insoluble washout (W-S) and D+U fractions of starch of maize silages, determined by both methods, due to variation in their chemical composition. The values for N fractions of maize and grass silages obtained with both methods were found different (p < 0.001). Large differences (p < 0.001) were found in the D+U fraction of starch of maize silages which might be due to different methodological approaches, such as different rinsing procedures (washing vs. shaking), duration of rinsing (40 min vs. 60 min) and different solvents (water vs. buffer solution). The large differences (p < 0.001) in the W-S and D+U fractions of starch determined with both methods can led to different predicted values for the effective rumen starch degradability. In conclusion, the MM with one recommended shaking procedure, performed under identical and controlled experimental conditions, can give more reliable results compared to the WMM, using different washing programs and procedures.

Relationship between the Methane Production and the CNCPS Carbohydrate Fractions of Rations with Various Concentrate/roughage Ratios Evaluated Using In vitro Incubation Technique

The objective of the trial was to study the relationship between the methane (CH₄) production and the Cornell Net Carbohydrate and Protein System (CNCPS) carbohydrate fractions of feeds for cattle and the suitability of CNCPS carbohydrate fractions as the dietary variables in modeling the CH₄ production in rumen fermentation. Forty-five rations for cattle with the concentrate/roughage ratios of 10:90, 20:80, 30:70, 40:60, and 50:50 were formulated as feed samples. The Menke and Steingass’s gas test was used for the measurement of CH₄ production. The feed samples were incubated for 48 h and the CH₄ production was analyzed using gas chromatography. Statistical analysis indicated that the CH₄ production (mL) was closely correlated with the CNCPS carbohydrate fractions (g), i.e. CA (sugars); CB₁ (starch and pectin); CB₂ (available cell wall) in a multiple linear pattern: CH₄ = (89.16±14.93) CA+ (124.10±13.90) CB₁+(30.58±11.72) CB₂+(3.28±7.19), R² = 0.81, p<0.0001, n = 45. Validation of the model using 10 rations indicated that the CH₄ production of the rations for cattle could accurately be predicted based on the CNCPS carbohydrate fractions. The trial indicated that the CNCPS carbohydrate fractions CA, CB₁ and CB₂ were suitable dietary variables for predicting the CH₄ production in rumen fermentation in vitro.

Variation in amino acid digestibility of rapeseed meal studied in caecectomised laying hens and relationship with chemical constituents

1. The objectives were to study the variation in amino acid (AA) digestibility of rapeseed meal in laying hens and to investigate whether the variation in AA digestibility can be explained by chemical constituents of the rapeseed meal. Nine rapeseed meals from different processing plants operating in Germany were used. The crude protein and crude fibre concentrations of the meals ranged from 367 to 410 and 137 to 175 g/kg dry matter (DM), respectively. The concentrations of total glucosinolates varied between 5.1 and 12.9 mmol/kg DM. Each meal was included in one of the experimental diets at an inclusion level of 200 g/kg at the expense of maize starch. 2. A total of 16 Lohmann Brown laying hens were used for this experiment and were surgically caecectomised. They were housed individually in metabolism cages. Each of the 10 diets was fed to at least 6 hens in 4 periods of two weeks duration. Excreta were collected during the last 5 d of each period. The digestibility coefficient of AA was calculated for each diet. The digestibility of the rapeseed meals was calculated by multiple linear regression analysis. 3. Mean essential AA digestibility of rapeseed meals varied between 0.78 and 0.84. Among the essential amino acids, the ighest level of digestibility was found for arginine (0.88) and methionine (0.87), and the lowest for lysine (0.74) and threonine (0.75). The highest variation between rapeseed meals in digestibility was found for threonine (0.68 to 0.79) and lysine (0.68 to 0.78). The concentration of neutral detergent insoluble nitrogen of the rapeseed meals was negatively correlated with the digestibility of lysine, but was not significantly correlated with digestibility of any other AA. Glucosinolate concentration was significantly correlated with valine digestibility. 4. Multiple regression analysis showed that the concentrations of crude protein and ash together were the major factors considered to explain variation in digestibility of essential AA. Additional consideration of neutral detergent insoluble nitrogen and total glucosinolates improved the goodness of fit only marginally and was unlikely to be of practical significance.

Effects of using thermally treated lupins instead of soybean meal and rapeseed meal in total mixed rations on in vitro microbial yield and performance of dairy cows

The objective was to study whether thermally treated lupins (TTL) can replace solvent-extracted soybean meal (SBM) and rapeseed meal (RSM) in dairy cow rations. Three total mixed rations (TMR) were used. They differed in the inclusion of the main protein feeds (TTL alone, TTL + RSM and RSM + SBM) but were equal in organic matter digestibility and metabolizable energy content. In vitro organic matter fermentation was not significantly different between the TMR. Efficiency of microbial crude protein (MCP) synthesis was significantly higher for TMR RSM + SBM than for TTL. In vitro gas production potential was similar for the three TMR. The maximal rate of gas production was achieved later in TMR TTL than in RSM + SBM. Feed intake of dairy cows was significantly lower when TMR TTL was fed than when TMR TTL + RSM or RSM + SBM were fed. Milk yield was significantly lower with the high inclusion rate of TTL in comparison with the other TMR. The contents of milk protein and milk fat were significantly lower when the two TTL containing TMR were fed in comparison with the RSM + SBM ration. Effects of TTL inclusion on MCP synthesis may affect the amino acid supply to the duodenum of cows to a greater extent than differences in the degradability of feed proteins.

Enzymic and in sacco methods to estimate rumen degradation of food protein in cattle

BACKGROUND

Two factorial studies compared enzymic and in sacco methods to estimate degradation of ruminant foods. Enzyme degradation (in vitro = enzyme) was determined from the release of leucine-equivalent amino acid (LA) crude protein (CP) from sunflower meal (SF), maize gluten meal (MG), distillers' dark grain (DG) and field beans (FB) after their separate incubations with Streptomyces griseus enzyme for 0-24 h. In sacco crude protein (CP) degradation of these foods was estimated during washing (0 h) and rumen incubations in fistulated cows for 2-24 h. The LA data were expressed as g LA per either kg of CP (LACP) or acid-hydrolysable LA (HLA) of each food and compared with in sacco data.

RESULTS

These methods showed comparable degradation with time (P < 0.01). The in sacco and HLA were greater than LACP for all foods except MG where in sacco value was either lower or equal to LACP depending upon the incubation time (P > 0.05 or P < 0.05). Conversely, HLA was significantly (P < 0.01) greater than LACP from 2 h onwards. At 0 h, in sacco values were significantly greater than those of enzyme for SF, DG and FB (P < 0.05) but not for MG. The foods differed significantly for degradation constants (a, b, c) in each method (P < 0.05).

CONCLUSIONS

Despite variations between in sacco and enzyme estimates for different foods, the relationships between these estimates suggest that the HLA enzyme method has the potential to estimate food degradation. Copyright © 2007 Society of Chemical Industry.

Zum Einsatz von Luzernesilage bei Kuhen mit hoher Milchleistung

Studies on the use of lucerne silage as a forage source for high-yielding dairy cows. It was the main objective of this study to investigate effects of lucerne silage inclusion in maize silage-based rations on dry matter intake, milk yield and metabolic parameters of high yielding dairy cows. Three total mixed rations (TMR) were used. They contained as the main forage source on a dry matter basis 41.5% maize silage and 12% grass silage (ration maize) or 18% maize silage and 29% of either grass silage (ration grass) or lucerne silage (ration lucerne). Organic matter digestibility, determined with wether sheep, was 77 and 65% for the grass and the lucerne silage, respectively. Each TMR was fed to 4 wether sheep to determine crude nutrient digestibilities. The content of net energy for lactation (NEL), as calculated from digestible crude nutrients, was (MJ/kg dry matter (DM)): 7.1 (maize), 7.2 (grass) and 6.8 (lucerne). Based on the results of a chemical fractionation of crude protein it was calculated that the content of utilisable crude protein (nXP) was not below 167 g/kg DM and that the N balance in the rumen was not negative for all 3 TMR. The TMR were fed for 169 days to at least 28 dairy cows per treatment which were in their first half of lactation. Feed intake, milk performance data and physiological indicators were measured for each individual cow. The DM intake was significantly higher in treatment lucerne (23.2 kg/d) than in treatments grass (21.2 kg/d) and maize (22.6 kg/d). Daily intakes of NEL and nXP were significantly lower in treatment grass than in the other two treatments. Milk yield was significantly higher in treatment maize (41.7 kg/d) than in treatments grass (37.3 kg/d) and lucerne (38.5 kg/d). This was associated with a significantly lower milk fat content in treatment maize (3.5%) in comparison with the other two treatments (3.9% each). Milk protein content was unaffected by treatment (3.4% on average). Similarly, the acetone content of milk was not significantly different between treatments. The beta-hydroxybutyrate content of the deproteinized blood was significantly higher in treatment lucerne at start of lactation, but this difference disappeared in subsequent parts of the experimental period. Measurements of back fat thickness indicated mobilisation and retention of energy by cows during the experimental period to be similar in all treatments. It is concluded that lucerne silage is a suitable forage source for high yielding dairy cows in spite of its low digestibility.