Contribution of dietary nitrogen and purine bases to the duodenal digesta: comparison of duodenal and polyester-bag measurements

Protein requirements for pregnant dairy cows

We aimed to estimate the protein requirements of pregnant Holstein × Gyr cows. A total of 61 Holstein × Gyr (HG) cows were used, with an average initial body weight (BW) of 480 ± 10.1 kg and age 5 ± 0.5 yr. Cows were divided into 3 groups: pregnant (n = 43), nonpregnant (n = 12), and baseline (n = 6). Baseline animals were slaughtered before starting the experiment to estimate the initial body composition of the remaining animals. Pregnant and nonpregnant cows received 2 diets: maintenance and ad libitum. Pregnant cows were slaughtered at 139, 199, 241, and 268 gestation days (GD). First, we used data only from nonpregnant cows to determine the requirements for maintenance and growth in adult cows. The requirements of metabolizable protein for maintenance (MPm; grams of empty metabolic BW [EBW0.75] per day) were estimated using a linear regression between the metabolizable protein intake (MPI, g/d) and average daily gain (g/d), and the MPm was defined as the intercept divided by the average metabolic BW. Net protein requirements for gain (NPg; g/d) were estimated by the first derivative of the allometric equation between final crude protein in the body (kg) and the final empty BW (EBW; kg). The efficiency of use of metabolizable protein for gain (k) was calculated from the regression between the retained protein (g EBW0.75/d) and the MPI (g EBW0.75/d), and k was the slope of this regression. The MPI was estimated by combining microbial protein synthesis (purine derivatives) with the digestible rumen undegradable protein [(total protein intake - rumen-degradable protein) × intestinal digestibility]. Second, an exponential model was used to fit the protein accumulation in the gestational components in the function of GD. The first derivative of that model was considered the net requirement for gestation (NPgest). The efficiency of protein utilization for gestation (kgest) was calculated by the iterative method using the equation Δ = MPI - (MPm + NPg/kg + NPgest/kgest), where kg is efficiency of protein utilization for gain. The iteration was performed aiming at a zero deviation between observed MPI and metabolizable protein (MP) estimated by the requirements determined herein. We obtained a value of 3.88 g EBW0.75/d for MPm. The estimation of NPg can be calculated according to the following equation: NPg = 0.716 × (EBWopen-0.308) × EBGc, where EBWopen is the EBW (kg) for nonpregnant animals and EBGc is the empty body gain (kg/d) corrected for the gestational component. The k was determined as 0.347. The NPgest requirements were determined as NPgest (g/d) = 0.0008722 × exp(0.01784 × GD) × (calf weight/35). The kgest was 0.625. It is important to highlight that different methods of MP estimates should not be mixed and that the proposed method requires the estimation of microbial protein (estimated via urinary estimates), which might limit practical application. In conclusion, new studies should be conducted to validate our results and the methodology adopted to determine protein requirements for pregnancy in dairy cows. Due to the pattern of protein accumulation in the gestational components, we suggest an exponential model to describe protein requirements for pregnancy for dairy cows.

Invited review: Nitrogen in ruminant nutrition: A review of measurement techniques

Nitrogen is a component of essential nutrients critical for the productivity of ruminants. If excreted in excess, N is also an important environmental pollutant contributing to acid deposition, eutrophication, human respiratory problems, and climate change. The complex microbial metabolic activity in the rumen and the effect on subsequent processes in the intestines and body tissues make the study of N metabolism in ruminants challenging compared with nonruminants. Therefore, using accurate and precise measurement techniques is imperative for obtaining reliable experimental results on N utilization by ruminants and evaluating the environmental impacts of N emission mitigation techniques. Changeover design experiments are as suitable as continuous ones for studying protein metabolism in ruminant animals, except when changes in body weight or carryover effects due to treatment are expected. Adaptation following a dietary change should be allowed for at least 2 (preferably 3) wk, and extended adaptation periods may be required if body pools can temporarily supply the nutrients studied. Dietary protein degradability in the rumen and intestines are feed characteristics determining the primary AA available to the host animal. They can be estimated using in situ, in vitro, or in vivo techniques with each having inherent advantages and disadvantages. Accurate, precise, and inexpensive laboratory assays for feed protein availability are still needed. Techniques used for direct determination of rumen microbial protein synthesis are laborious and expensive, and data variability can be unacceptably large; indirect approaches have not shown the level of accuracy required for widespread adoption. Techniques for studying postruminal digestion and absorption of nitrogenous compounds, urea recycling, and mammary AA metabolism are also laborious, expensive (especially the methods that use isotopes), and results can be variable, especially the methods based on measurements of digesta or blood flow. Volatile loss of N from feces and particularly urine can be substantial during collection, processing, and analysis of excreta, compromising the accuracy of measurements of total-tract N digestion and body N balance. In studying ruminant N metabolism, nutritionists should consider the longer term fate of manure N as well. Various techniques used to determine the effects of animal nutrition on total N, ammonia- or nitrous oxide-emitting potentials, as well as plant fertilizer value, of manure are available. Overall, methods to study ruminant N metabolism have been developed over 150 yr of animal nutrition research, but many of them are laborious and impractical for application on a large number of animals. The increasing environmental concerns associated with livestock production systems necessitate more accurate and reliable methods to determine manure N emissions in the context of feed composition and ruminant N metabolism.

Microbial contribution to the duodenal flow of purine bases in cattle

It is widely accepted that purine bases (PB) could overestimate microbial production when used as microbial markers as result of rumen by-pass of dietary PB. Free nucleic acids are rapidly degraded when incubated in rumen content either in vitro or in vivo (McAllan and Smith, 1973) and nylon bag disappearance of food PB seems to confirm their low contribution to duodenal flow (Pérez et al. 1996). However isotopic labelling of microbial PB (Pérez et al. 1997) have resulted in much higher estimates of non microbial PB arriving to duodenum, which were attributed to differential uptake of dietary PB by bacterial and protozoal populations leading to lack of representativeness of the reference bacterial sample. The present experiment aimed to know if estimation of microbial contribution to duodenal PB could be biased by differences between rumen bacterial and protozoal populations in 15N enrichment of PB.

Microbial contribution to duodenal purine flow in fattening cattle given concentrate diets, estimated by purine N labelling (15N) of different microbial fractions

The origin of duodenal purine bases (PB) was studied in a digestion experiment with four heifers, cannulated in the rumen and duodenum, which received a basal concentrate (152 g crude protein (CP) per kg dry matter (DM)) together with barley straw (85: 15 fresh weight basis) or the same concentrate supplemented with soya-bean meal, carbohydrate-treated soya-bean meal, maize gluten meal or fish meal to increase its protein content to 192 g/kg DM. Tr eatments were assigned to the four animals in five experimental periods according to an incomplete Latin-square design. Each 30-day period included 20 days of change-over adaptation and 10 days of experimental measurements. The flow of digesta entering the duodenum was estimated using Yb and acid-detergent insoluble ash as indigestible markers according to a double-marker system and microbial nitrogen (N) and PB were labelled with15N infused into the rumen. The proportion of duodenal PB of microbial origin estimated from15N enrichment of PB-N averaged 0·66 (s.e. 0·029) and did not differ between treatments nor when protozoa or bacteria associated with liquid (LAB) and solid (SAB) fractions were used as a reference sample. On average microbial contribution to duodenal non-ammonia N was higher when estimated from the PB/N ratio than from15N (0·67 v. 0·55 (s.e. 0·015)) although differences were small and not significant when LAB was the reference sample (0·58 v. 0·52 (s.e. 0·018)) reflecting the higher PB/N ratio of this fraction compared with SAB and protozoa (2·04 v. 1·65 and 1·60 (s.e. 0·04) mmol/g). Considering only the duodenal PB of microbial origin resulted in estimates of microbial N synthesis from the PB/N ratio of SAB similar to those derived from15N enrichment of both bacterial fractions (12·9 v. 13·5 and 13·3 (s.e. 0·83) g/kg of organic matter apparently digested in the rumen OMADR)) but underestimated the values derived from LAB (9·9 g/kg OMADR). Regardless of the estimation method, neither the duodenal flow of microbial N nor the efficiency of microbial synthesis differed between treatments. These results suggest that a significant proportion of duodenal PB have a non-microbial origin which may lead to overestimation of microbial yield when PB are used as a marker. Differences in PB/N ratio between microbial fractions is another important factor to be considered.

The effects of synchronizing the rate of dietary energy and nitrogen supply to the rumen on the production and metabolism of sheep: food characterization and growth and metabolism of ewe lambs given food ad libitum

The effects of diets formulated to have two rates of organic matter (OM) release and to be either synchronous or asynchronous with respect to their hourly release of nitrogen (N) and OM in the rumen on the growth, intake and metabolism of ewe lambs was studied. In experiment 1, the in situ degradation characteristics of N and OM for 16 food ingredients was determined. The foods varied considerably in their chemical composition and degradability coefficients of N and OM. Based on this information, four diets were formulated to differ in their rate of N and OM release in the rumen but to have a similar content of metabolizable energy (10·4 MJ/kg dry matter (DM)), crude protein (140 g/kg DM), daily ratio of N: OM released (34 g N per kg OM) and digestible undegradable protein (32 g/kg DM) but to be synchronous or asynchronous with respect to hourly N: OM ratio. The four diets were slow energy, synchronous (SS), slow energy asynchronous (SA), fast energy, synchronous (FS) and fast energy asynchronous (FA).

In experiment 2 the four diets were offered ad libitum to 24 ewe lambs of an initial live weight of 25 kg in a 2 × 2 factorial design. Lambs were slaughtered at 40 kg live weight. Animals offered diets FS and FA had a higher growth rate than those offered diets SA and SS (266 v. 225 g/day respectively; P < 0·05). There was a significant interaction between rate ofOM and N release on DM intake (DMI) with lambs offered diet FS consuming less than those offered FA (1·47 v. 1·67 kg DM per day; P < 0·05). Rate of energy release and synchrony had an effect on food conversion efficiency (FCE; kg gain per kg DMI) with lambs offered diets FS and FA having a greater FCE than those offered SS and SA (0·170 v. 0·146 respectively; P < 0·001) and those offered diets SS and FS had a greater FCE than those offered diets SA and FA (0·164 v. 0·152 respectively; P < 0·05). Diet had little effect on carcass composition. The proportion of propionate in rumen fluid was greater in lambs offered diets FS and FA than in those offered diets SS and SA (222 v. 168 mmol/mol respectively; P < 0·01). Plasma urea concentrations exhibited a cyclical trend throughout the day with highest concentrations occurring 2 h after fresh food was offered. Lambs offered diet FS had lower plasma urea concentrations at 6, 10 and 14 h after feeding. Plasma concentrations of ß-hydroxybutyrate were lower throughout the day in lambs offered diet FS. The current findings are consistent with the view that a rapid release of OM and synchronizing hourly N: OM release in the rumen can improve the efficiency of growth through improvements in protein and/or energy metabolism.

Fermentation characteristics and microbial growth promoted by diets including two-phase olive cake in continuous fermenters

Two-phase olive cake (2POC) is the by-product obtained from the so called 'two-phase' procedure to extract olive oil by mechanical methods. After the mechanical extraction the 2POC is dried and most of the remaining oil extracted by chemical means. The production of the crude by-product may reach more than 4 millions t/year in Spain (MAPA, 2003), most of it in areas (Southeast) with shortage of pastures and conventional feeds for ruminants. Six continuous fermenters, inoculated with ruminal liquor from wethers or goats, were fed lucerne hay (LH); LH plus a concentrate including dried two-phase olive cake (LHCO) and; diet LHCO added with polyethylene glycol (LHCOP). The highest pH values and ammonia N (NH(3)-N) output were found in fermenters fed diet LH (6.19 and 6.35 for pH, and 53.7 and 68.9 mg NH(3)N/day, respectively, in fermentes inoculated with rumen liquor from sheep and goats) without differences due to the inoculum origin. The digestibility of carbohydrates (CHO) was affected (p < 0.001) by inoculum (67.0 and 58.8%, respectively, for goats and wethers) the lowest values being for diet LHCO (53.2 and 57.0% with inoculum from wethers and goats, respectively). The main volatile fatty acid (VFA) was acetic acid with higher (p < 0.01) values in fermenters with inoculum from goats than from wethers (80.2 and 63.0 mmol/day respectively). The efficiency of bacterial protein synthesis (EBS) was not different (p > 0.05) with inoculum from wethers and goats [26.4 and 28.1 and 35.2 and 33.5 g bacterial N/kg digested CHO, respectively, obtained by using diamino pimelic acid (DAPA) and purine bases (PB) as microbial markers]. The lowest (p < 0.05) values were found in fermenters fed diets LHCOP and LH, estimated, respectively, from DAPA and PB (21.9 and 29.0 g bacterial N/kg digested CHO). The substitution of a part of lucerne hay by a concentrate including dried 2POC does not seem to cause important differences in efficiency of VFA production. Results concerning bacterial protein synthesis are not so clear since values estimated from DAPA and PB did not show similar trends. Neither inoculum origin nor PEG had important effects on fermentation characteristics. The DAPA could be an adequate marker in continuous fermenters, with comparative purposes, as protozoa are not present and, estimated protein synthesis values are similar to those obtained in vivo with similar diets using urinary excretion of PD (Yáñez Ruiz et al., 2004b). Further research is needed to state the optimal proportion of 2POC in practical diets for ruminants at both maintenance and production stages.

Validation of use of purine bases as a microbial marker by15N labelling in growing lambs given high-concentrate diets: effects of grain processing, animal age and digesta sampling site

The origin of post-ruminal purine bases (PB) was studied in 24 growing lambs that were given a pelleted concentrate plus barley straw (C) or whole barley grain plus protein supplement (WB). Six lambs from each treatment were slaughtered at 10 and 30 days post weaning after15N labelling of microbial nitrogen (N) and PB. Microbial contribution to digesta non-ammonia N (NAN) and PB was lower (P< 0·01) when estimated from duodenal rather than abomasal samples (0·36 v. 0·52 (s.e.d. 0·021) for NAN and 0·47 v. 0·77 (s.e.d. 0·029) for PB) as a result of endogenous contamination. In comparison with15N, total PB/N led to higher estimates (P< 0·01) of microbial contribution to abomasal NAN in WB treatment (0·62 v. 0·46 s.e.d. 0·049). The difference was removed after correcting for microbial PB, while this effect was not observed with < the C diet, resulting in a marker by diet interaction (P< 0·05). Abomasal PB flow increased (P< 0·1) from 10 to 30 days after weaning mainly due to the higher proportion of microbial PB (0·70 v. 0·81 (s.e.d. 0·047)). Rumen apparent PB degradation did not differ between diets in older lambs, but it was proportionally 0·39 lower for WB treatment (P< 0·05) in younger lambs. When the microbial PB flow was estimated indirectly from labelled microbial N and the PB/N ratio of bacterial extracts the estimates were in agreement with those derived from PB-15N in the WB treatment but resulted in unrealistic values in lambs on diet C. Results suggest that significant proportions of dietary PB can escape rumen degradation which may lead to overestimation of microbial contribution to abomasal NAN when the PB/N ratio is used as marker. The extent of the overestimation is affected by the lamb age and grain processing.

Comparison of four markers for quantifying microbial protein flow from the rumen of lactating dairy cows

Eight ruminally cannulated lactating cows from a study on the effects of dietary rumen degraded protein (RDP) on production and N metabolism were used to compare 15N, total purines, amino acid (AA) profiles, and urinary excretion of purine derivatives (PD) as microbial markers for quantifying the flow of microbial protein at the omasal canal. Dietary RDP was gradually decreased by replacing solvent soybean meal and urea with lignosulfonate-treated soybean meal. The purine metabolites xanthine and hypoxanthine were present in digesta and microbial samples and were assumed to be of microbial origin. The sum of the purines and their metabolites (adenine, guanine, xanthine, and hypoxanthine) were defined as total purines (TP) and used as a microbial marker. Decreasing dietary RDP from 13.2 to 10.6% of dry matter (DM) reduced microbial nonammonia N (NAN) flows estimated using TP (from 415 to 369 g/d), 15N (from 470 to 384 g/d), AA profiles (from 392 to 311 g/d), and PD (from 436 to 271 g/d). Averaged across diets, microbial NAN flows were highest when estimated using TP and 15N (398 and 429 g/d), lowest when using PD (305 g/d), and intermediate when using AA profiles (360 g/d) as microbial markers. Correlation coefficients between 15N and TP for fluid-associated bacteria, particle-associated bacteria, and total microbial NAN flows were 0.38, 0.85, and 0.69, respectively. When TP was used as the microbial marker, ruminal escape of dietary NAN was not affected by replacing solvent soybean meal with lignosulfonate-treated soybean meal in the diets. The direction and extent of response of dietary and microbial NAN flow to dietary treatments were similar when estimated using 15N, AA profiles, and PD, and were in agreement with previously published data and National Research Council predictions. Microbial and dietary NAN flows from the rumen estimated using 15N appeared to be more accurate and precise than the other markers. Caution is required when interpreting results obtained using TP as the microbial marker.

Effect of dietary carbohydrate composition and availability on utilization of ruminal ammonia nitrogen for milk protein synthesis in dairy cows

A trial with four ruminally and duodenally cannulated, late-lactation dairy cows was conducted to investigate the effect of dietary carbohydrate (CHO) composition and availability on ruminal ammonia N utilization and transfer into milk protein. Two diets were fed at 8-h intervals in a crossover design. The diets differed in CHO composition: the ruminally fermentable non-structural carbohydrates (RFSS) diet (barley and molasses) contained a larger proportion of ruminally available CHO in the nonstructural carbohydrate fractions and the ruminally fermentable fiber (RFNDF) diet (corn, beet pulp, and brewer's grains) contained a larger proportion of CHO in ruminally available fiber. Nitrogen-15 was used to label ruminal ammonia N and consequently microbial and milk N. Fermentation acids, enzyme activities, and microbial protein production in the rumen were not affected by diet. Ruminal ammonia concentration was lowered by RFNDF. Ruminal and total tract digestibility of nutrients did not differ between diets except that apparent ruminal degradability of crude protein was lower for RFNDF compared with RFSS. Partitioning of N losses between urine and feces was also not affected by diet. Milk yield and fat and protein content were not affected by treatment. Average concentration of milk urea N was lower for RFNDF than for RFSS. Proportion of milk protein N originating from ruminal microbial N (based on the areas under the 15N-enrichment curves) was higher for RFNDF than for RFSS. Cumulative recovery of 15N in milk protein was 13% higher for RFNDF than for RFSS indicating enhanced transfer of 15N-ammonia into milk protein with the former diet. The results suggested that, compared to diets containing higher levels of ruminally fermentable starch, diets providing higher concentration of ruminally fermentable fiber may enhance transfer of ruminal ammonia and microbial N into milk protein.