Advertisement

Abomasal infusion of corn starch and β-hydroxybutyrate in early-lactation Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis

Published:September 02, 2021DOI:https://doi.org/10.3168/jds.2021-20323

      ABSTRACT

      The objectives of this study were to induce hindgut and metabolic acidosis via abomasal infusion of corn starch and β-hydroxybutyrate (BHB), respectively, and to determine the effects of these physiological states in early-lactation dairy cows. In a 6 × 6 Latin square design, 6 rumen-fistulated Holstein-Friesian dairy cows (66 ± 18 d in milk) were subjected to 5 d of continuous abomasal infusion treatments followed by 2 d of rest. The abomasal infusion treatments followed a 3 × 2 factorial design, with 3 levels of corn starch and 2 levels of BHB. The infusions were water as control, 1.5 kg of corn starch/d, 3.0 kg of corn starch/d, 8.0 mol BHB/d, 1.5 kg of corn starch/d + 8.0 mol BHB/d, or 3.0 kg of corn starch/d + 8.0 mol BHB/d. A total mixed ration consisting of 35.0% grass silage, 37.4% corn silage, and 27.6% concentrate (on a dry matter basis) was fed at 90% of ad libitum intake of individual cows. The experiment was conducted in climate respiration chambers to facilitate determination of energy and N balance. Fecal pH decreased with each level of corn starch infused into the abomasum and was 6.49, 6.00, and 5.15 with 0.0, 1.5, and 3.0 kg of corn starch/d, respectively, suggesting that hindgut acidosis was induced with corn starch infusion. No systemic inflammatory response was observed and the permeability of the intestine or hindgut epithelium was not affected by the more acidic conditions. This induced hindgut acidosis was associated with decreased digestibility of nutrients, except for crude fat and NDF, which were not affected. Induced hindgut acidosis did not affect milk production and composition and energy balance, but increased milk N efficiency. Abomasal infusion of BHB resulted in a compensated metabolic acidosis, which was characterized by a clear disturbance of acid-base status (i.e., decreased blood total CO2, HCO3, and base excess, and a tendency for decreased urinary pH), whereas blood pH remained within a physiologically normal range. Abomasal infusion of BHB resulted in increased concentrations of BHB in milk and plasma, but both remained well below the critical threshold values for subclinical ketosis. Induced compensated metabolic acidosis, as a result of abomasally infused BHB, increased energy retained as body fat, did not affect milk production and composition or inflammatory response, but increased intestinal permeability.

      Key words

      INTRODUCTION

      In early lactation, the rapid increase in energy requirements for milk production by dairy cattle cannot be met by an increase in feed intake alone (
      • van Knegsel A. T.M.
      • van den Brand H.
      • Dijkstra J.
      • van Straalen W.M.
      • Heetkamp M. J.W.
      • Tamminga S.
      • Kemp B.
      Dietary energy source in dairy cows in early lactation: Energy partitioning and milk composition.
      ). To compensate for this energy deficit, dairy cows mobilize body fat and, as a result, the concentration of nonesterified fatty acids (NEFA) is elevated in blood. When the capacity of the liver to oxidize these NEFA is exceeded, remaining NEFA are stored as triglycerides in the liver or are converted into ketone bodies that enter circulation (
      • Bell A.W.
      Regulation of organic nutrient metabolism during transition from late pregnancy to early lactation.
      ). Although ketone bodies can be used as an alternative form of energy by some tissues (e.g., brain;
      • Laffel L.
      Ketone bodies: A review of physiology pathophysiology and application of monitoring diabetes.
      ), the utilization of ketone bodies is limited. Excess concentration of ketone bodies in blood increases the risk for ketosis (
      • Duffield T.
      Subclinical ketosis in lactating dairy cattle.
      ). Subclinical ketosis in early lactation affects 40 to 60% of cows in herds undergoing repeated testing (
      • Emery R.S.
      • Burg N.
      • Brown L.D.
      • Blank G.N.
      Detection, occurrence, and prophylactic treatment of borderline ketosis with propylene glycol feeding.
      ;
      • Simensen E.
      • Halse K.
      • Gillund P.
      • Lutnœs B.
      Ketosis treatment and milk yield in dairy cows related to milk acetoacetate levels.
      ), which is much higher than the typical 2 to 15% incidence of clinical ketosis (
      • Duffield T.
      Subclinical ketosis in lactating dairy cattle.
      ). The ketone bodies BHB and acetoacetate are organic acids (pKa of 3.6 and 4.7, respectively) that dissociate almost fully at physiological pH (
      • Laffel L.
      Ketone bodies: A review of physiology pathophysiology and application of monitoring diabetes.
      ). If the associated rise in hydrogen ion concentration in blood exceeds blood buffering capacity during periods of high blood ketone concentrations, metabolic acidosis may result (
      • Hood V.
      • Tannen R.
      Maintenance of acid-base homeostasis during ketoacidosis and lactic acidosis.
      ).
      It has only been relatively recently recognized that strategies to prevent diseases such as ketosis in early-lactation cows (i.e., increasing the energy density of rations with fermentable carbohydrates) are closely linked with the gastrointestinal function and health of dairy cows (e.g.,
      • Enemark J. M.D.
      The monitoring, prevention and treatment of sub-acute ruminal acidosis (SARA): A review.
      ;
      • Mulligan F.J.
      • Doherty M.L.
      Production diseases of the transition cow.
      ). Hindgut fermentation has been associated with high intake of starch and other fermentable substrates that are not fully degraded in the rumen (
      • Gressley T.F.
      • Hall M.B.
      • Armentano L.E.
      Ruminant nutrition symposium: Productivity, digestion, and health responses to hindgut acidosis in ruminants.
      ). As dairy cows have relatively limited amylolytic activity in the small intestines (
      • Matthé A.
      • Lebzien P.
      • Hric I.
      • Flachowsky G.
      • Sommer A.
      Effect of starch application into the proximal duodenum of ruminants on starch digestibility in the small and total intestine.
      ), incompletely degraded and digested substrates flow into the hindgut where fermentation of these substrates occurs. This increases the risk of hindgut acidosis, which can be defined as an accumulation of organic acids and a subsequent decrease in digesta pH (
      • Gressley T.F.
      • Hall M.B.
      • Armentano L.E.
      Ruminant nutrition symposium: Productivity, digestion, and health responses to hindgut acidosis in ruminants.
      ). Acidosis in the hindgut can potentially cause dysbiosis in the microbial community and epithelial damage, and consequently can affect animal production and health (
      • Plaizier J.C.
      • Danesh Mesgaran M.
      • Derakhshani H.
      • Golder H.
      • Khafipour E.
      • Kleen J.L.
      • Lean I.
      • Loor J.
      • Penner G.
      • Zebeli Q.
      Review: Enhancing gastrointestinal health in dairy cows.
      ).
      Recently,
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      showed that both hindgut acidosis and metabolic acidosis were associated with changes in feed intake, apparent total-tract digestibility (ATTD) of nutrients, milk production and composition, and N and energy partitioning. Importantly,
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      demonstrated that metabolic acidosis became more severe when dairy cows also experienced hindgut acidosis, suggesting that a relation may exist between acidosis in the hindgut and the occurrence of metabolic acidosis. In the study of
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      , metabolic acidosis was induced by infusing ammonium chloride and hindgut acidosis was induced by infusing ground corn. It is unclear whether the relation between hindgut acidosis and metabolic acidosis also exists when metabolic acidosis is induced via accumulation of ketone bodies (i.e., a different mechanism that can affect the acid-base status of a dairy cow). It also remains unclear from the study of
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      whether there is a dose effect of the amount of starch infused and the severity of hindgut acidosis. Therefore, the aims of the current study were to (1) induce hindgut and metabolic acidosis via abomasal infusion of corn starch and BHB, respectively, and (2) determine the effects of these physiological states on feed intake, ATTD of nutrients, energy and N partitioning, milk production and composition, acid-base status, respiratory and metabolic status, and ruminal and hindgut fermentation characteristics. We hypothesized that abomasally infused corn starch would be fermented in the hindgut, produce VFA, and subsequently decrease digesta pH, resulting in hindgut acidosis and compromised intestinal permeability. Abomasal infusion of BHB was expected to disturb the acid-base status of the dairy cows, resulting in metabolic acidosis (i.e., a decreased blood pH). When corn starch and BHB were infused simultaneously into the abomasum, we expected to induce hindgut and metabolic acidosis. We expected the metabolic acidosis to be more severe with both corn starch and BHB infusion compared with only BHB infusion because of a combined effect of both types of acidosis on the acid-base status of the cow.

      MATERIALS AND METHODS

      Experimental Design and Housing

      The experiment was conducted from August until October 2019 at the animal research facilities of Wageningen University & Research (Wageningen, the Netherlands), under the Dutch Law on Animal Experiments in accordance with European Union Directive 2010/63, and was approved by the Central Committee of Animal Experiments (The Hague, the Netherlands; 2018.D-0013.002).
      Six rumen-fistulated, multiparous Holstein-Friesian dairy cows (second parity n = 5, third parity n = 1) producing 28.8 ± 4.96 kg/d at 66 ± 18 DIM (average ± SD) at the beginning of the experiment were randomly assigned to a 6 × 6 Latin square design with 6 treatments. Each experimental period (n = 6) consisted of 5 d of continuous abomasal infusion followed by 2 d of rest. Cows were adapted to the experimental conditions for 19 d before the first experimental period. For the first 14 d of adaptation, cows were housed individually in tiestalls to become adapted to the basal diet as well as the restriction in movement. For the last 5 d of adaptation as well as for the 6 consecutive experimental periods, cows were housed individually in identical climate respiration chambers (CRC) to measure, among other variables, the effect of abomasally infused BHB and corn starch on the energy and N partitioning of the dairy cows.

      Diet and Feeding

      Cows were fed a TMR consisting of 35.0% grass silage, 37.4% corn silage, and 27.6% concentrate on a DM basis (Table 1). To determine the ATTD of nutrients, TiO2 was included in the concentrate (5 g/kg DM) as an external marker. The concentrate was in meal form and produced by Research Diet Services (RDS BV) in a single batch. The diet was formulated to meet 101 and 96% of NEL (
      • Van Es A. J.H.
      Feed evaluation for ruminants. I. The systems in use from May 1977-onwards in the Netherlands.
      ) and intestinal digestible protein requirements (
      • Van Duinkerken G.
      • Blok M.C.
      • Bannink A.
      • Cone J.W.
      • Dijkstra J.
      • van Vuuren A.M.
      • Tamminga S.
      Update of the Dutch protein evaluation system for ruminants: The DVE/OEB2010 system.
      ), respectively, for cows consuming 19.0 kg of DM/d and producing 33 kg/d of milk containing 4.5% fat and 3.5% protein (i.e., mean DMI and milk production measured during the first weeks of lactation of the cows that would potentially enter this study).
      Table 1Chemical composition (g/kg of DM, unless stated otherwise) of the TMR ingredients (grass silage, corn silage, and concentrates; analyzed) and of the complete TMR
      The TMR was composed of 31.5% grass silage, 43.9% corn silage, 11.3% concentrate, and 13.3% water on a product basis, which equals 350 g/kg DM grass silage, 374 g/kg DM corn silage, and 276 g/kg DM concentrate.
      (calculated)
      ItemGrass silage
      Ensiling characteristics: acetic acid = 7 g/kg DM, lactic acid = 18 g/kg DM, ammonia-N = 7% total N, and pH = 5.8.
      Corn silage
      Ensiling characteristics: acetic acid = 11 g/kg DM, lactic acid = 63 g/kg DM, ammonia-N = 6% total N, and pH = 3.8.
      Concentrate
      Concentrate contained (g/kg DM): sugar beet pulp = 325, barley = 300, rumen-protected soybean meal (Mervobest, NuScience) = 200, palm kernel flakes = 60, molasses = 50, sunflower oil = 20, CaCO3 = 15, MgO = 8.0, trace mineral and vitamin mix = 8.0, NaCl = 7.0, NaHCO3 = 2.0, and TiO2 = 5.0.
      TMR
      DM (g/kg of product)404310887364
      OM899953903920
      CP19597187156
      Crude fat45313938
      Gross energy (MJ/kg of DM)19.619.118.019.0
      NDF405383235350
      ADF214230120194
      ADL7121210
      Starch
      Not determined.
      270170148
      Sugar120
      Not determined.
      11273
      NEL
      van Es, 1978.
      (MJ/kg of DM)
      6.96.87.57.0
      DVE
      Intestinal digestible protein (van Duinkerken et al., 2011).
      735716292
      OEB
      Rumen-degraded protein balance (van Duinkerken et al., 2011).
      54−36−44−7
      1 The TMR was composed of 31.5% grass silage, 43.9% corn silage, 11.3% concentrate, and 13.3% water on a product basis, which equals 350 g/kg DM grass silage, 374 g/kg DM corn silage, and 276 g/kg DM concentrate.
      2 Ensiling characteristics: acetic acid = 7 g/kg DM, lactic acid = 18 g/kg DM, ammonia-N = 7% total N, and pH = 5.8.
      3 Ensiling characteristics: acetic acid = 11 g/kg DM, lactic acid = 63 g/kg DM, ammonia-N = 6% total N, and pH = 3.8.
      4 Concentrate contained (g/kg DM): sugar beet pulp = 325, barley = 300, rumen-protected soybean meal (Mervobest, NuScience) = 200, palm kernel flakes = 60, molasses = 50, sunflower oil = 20, CaCO3 = 15, MgO = 8.0, trace mineral and vitamin mix = 8.0, NaCl = 7.0, NaHCO3 = 2.0, and TiO2 = 5.0.
      5 Not determined.
      6
      • Van Es A. J.H.
      Feed evaluation for ruminants. I. The systems in use from May 1977-onwards in the Netherlands.
      .
      7 Intestinal digestible protein (
      • Van Duinkerken G.
      • Blok M.C.
      • Bannink A.
      • Cone J.W.
      • Dijkstra J.
      • van Vuuren A.M.
      • Tamminga S.
      Update of the Dutch protein evaluation system for ruminants: The DVE/OEB2010 system.
      ).
      8 Rumen-degraded protein balance (
      • Van Duinkerken G.
      • Blok M.C.
      • Bannink A.
      • Cone J.W.
      • Dijkstra J.
      • van Vuuren A.M.
      • Tamminga S.
      Update of the Dutch protein evaluation system for ruminants: The DVE/OEB2010 system.
      ).
      Cows were fed ad libitum for the first 10 d of the 19-d adaptation period. Intake during the final 5 d of this 10-d ad libitum intake period was used to calculate a 10% daily intake restriction for individual cows. From d 11 onward, cows were fed this fixed amount for the remainder of the adaptation and experimental periods. The grass silage, corn silage, and water were mixed 3 times weekly using a self-propelled mixer wagon (Strautmann Verti-Mix 500) equipped with a cutter loader system and an electronic weighing scale. The forage mixture was fed directly after preparation or stored in a cooling unit at 6°C to be used for future feedings for a maximum of 2 d. Forage and concentrate portions were manually mixed into a TMR for individual cows and the fresh TMR was fed twice daily at 0500 and 1530 h. Before providing fresh feed, feed refusals were collected and weighed to determine daily feed intake. For a 60-h period over d 3 to 6 of each experimental period (1700 h on d 3 until 0500 h on d 6), cows were fed using an automated feeding system that dispensed equal portions of feed every 2 h to promote metabolic steady-state conditions in preparation for the blood sampling protocol described later. Cows had free access to clean drinking water throughout the entire experiment.

      Abomasal Infusions

      The abomasal infusion treatments followed a 3 × 2 factorial design, with 3 levels of corn starch and 2 levels of BHB, and contained corn starch, BHB, or a combination of both. The abomasal infusion treatments were (1) 0.0 kg of corn starch/d + 0.0 mol BHB/d, (2) 0.0 kg of corn starch/d + 8.0 mol BHB/d, (3) 1.5 kg of corn starch/d + 0.0 mol BHB/d, (4) 1.5 kg of corn starch/d + 8.0 mol BHB/d, (5) 3.0 kg of corn starch/d + 0.0 mol BHB/d, and (6) 3.0 kg of corn starch/d + 8.0 mol BHB/d (Table 2). We based the dose of BHB on the assumption that, on average, 28 kg of body fat would be mobilized over a period of 8 wk (i.e., 500 g of body fat/d;
      • Tamminga S.
      • Luteijn P.A.
      • Meijer R. G.M.
      Changes in composition and energy content of liveweight loss in dairy cows with time after parturition.
      ;
      • van Knegsel A. T.M.
      • van den Brand H.
      • Dijkstra J.
      • van Straalen W.M.
      • Heetkamp M. J.W.
      • Tamminga S.
      • Kemp B.
      Dietary energy source in dairy cows in early lactation: Energy partitioning and milk composition.
      ), resulting in the production of 6.3 mol BHB/d. To increase the likelihood of inducing metabolic acidosis, the BHB dose in the present study is about 25% more than the estimated level of endogenous BHB production. With the corn starch dose, we aimed to achieve an amount of starch entering the small intestine (i.e., abomasally infused corn starch plus starch from basal TMR) that was within the range of total postruminal starch digestion reported for lactation dairy cows fed large amounts of corn grain (maximum of 5.0 kg/d;
      • Reynolds C.K.
      • Sutton J.D.
      • Beever D.E.
      Effects of feeding starch to dairy cattle on nutrient availability and production.
      ), but which would (according to
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      ) likely result in hindgut acidosis. The treatments were administered in 20-L batches that were replenished daily and infused at a rate ranging from 868 to 1,024 g/h (Table 2) to facilitate 5 d of continuous infusion (0900 h on d 1 until 0900 h on d 6 of each experimental period). To compensate the cation Na+ of Na-BHB (potentially affecting the acid-base status of the cows when not compensated), 8 mol of NH4Cl was added to the Na-BHB containing treatments, where Cl could function as the compensating anion. To compensate for the salts in the Na-BHB containing treatments (Na of Na-BHB and Cl of NH4Cl; potentially affecting the water balance of the cows when not compensated), 8 mol NaCl was added to the treatments without Na-BHB. The infusion treatments were subsequently designed to be isonitrogenous, using urea to compensate the N content of NH4Cl and corn starch. Unfortunately, final infused treatments were not completely iso-N, where the amount of N infused varied between 113.0 and 113.9 g/d (Table 2). To each treatment, 48 g of Co-EDTA/d was added as a marker for intestinal permeability [method adopted from
      • Amado L.
      • Berends H.
      • Leal L.N.
      • Wilms J.
      • van Laar H.
      • Gerrits W. J.J.
      • Martin-Tereso J.
      Effect of energy source in calf milk replacer on performance, digestibility, and gut permeability in rearing calves.
      and
      • Wilms J.
      • Berends H.
      • Martín-Tereso J.
      Hypertonic milk replacers increase gastrointestinal permeability in healthy dairy calves.
      ]. Eighty grams of carboxymethylcellulose/d (Acros Organics) was added to increase the viscosity of the treatments (i.e., to minimize settlement of particles) to facilitate homogeneous infusions.
      Table 2Composition of treatments that were infused into the abomasum of early-lactation dairy cows (target and realized)
      Treatment
      48 g of Co-EDTA and 80 g of carboxymethylcellulose (Acros Organics) were added to each treatment.
      (per d)
      Target (g/d, unless stated otherwise)Realized (g/d, unless stated otherwise)
      BHBCorn starch
      Chemical composition (in g/kg DM, unless stated otherwise): DM = 875 g/kg, OM = 999, CP = 2.5, crude fat = 6.1, NDF = 0, ADF = 0, ADL = 0, starch = 934, sugar = 0, and gross energy = 17.5 MJ/kg DM.
      UreaNH4ClNaClWater (L/d)Infusion rate (g/h)Success rate (%)BHBCorn starch
      0.0 kg of corn starch00244.10.0467.520868101.400
      0.0 kg of corn starch + 8.0 mol BHB1,01704.1427.90.02089998.81,0050
      1.5 kg of corn starch01,500242.10.0467.52093195.701,436
      1.5 kg of corn starch + 8.0 mol BHB1,0171,5002.1427.90.02096196.19771,442
      3.0 kg of corn starch03,000240.00.0467.52099396.402,892
      3.0 kg of corn starch + 8.0 mol BHB1,0173,0000.0427.90.0201,024100.01,0173,000
      1 48 g of Co-EDTA and 80 g of carboxymethylcellulose (Acros Organics) were added to each treatment.
      2 Chemical composition (in g/kg DM, unless stated otherwise): DM = 875 g/kg, OM = 999, CP = 2.5, crude fat = 6.1, NDF = 0, ADF = 0, ADL = 0, starch = 934, sugar = 0, and gross energy = 17.5 MJ/kg DM.
      The abomasal infusion technique was identical to the technique described by
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      . Briefly, infusion lines were placed in the abomasum via the ruminal cannula 2 d before the first experimental period and were checked twice daily for patency (i.e., by checking the peristaltic pumps, the realized amount infused in comparison with the target amount at that moment, and the infusion line, inside and outside the cows for knits or knots) and position (i.e., by opening the rumen fistula and following the infusion line to check placement in the abomasum). The infusion device was constructed from 200 cm of braided polyvinyl chloride hose attached to the rumen cannula plug at the proximal end and a flexible disk (equipped with holes to allow digesta passage) at the distal end to secure its placement through the sulcus omasi. The flexible disk was 12 cm in diameter and made of plastisol (Bar Diamond Inc.). The infusion lines (Tygon S3 E-3603, 3.2 mm i.d., 6.4 mm o.d. for the treatments without corn starch; Tygon S3 E-3603, 4.8 mm i.d., 8.0 mm o.d. for treatments with corn starch; VWR) were connected between the treatment solutions and the multichannel peristaltic pumps with luer-to-tubing connectors. The infusion lines were subsequently inserted into the braided polyvinyl chloride hose feeding into the rumen cannula. Treatments containing corn starch were maintained in suspension by continuous stirring.

      Climate Respiration Chambers

      A detailed description of the CRC design and gas measurements has been reported by
      • Heetkamp M. J.W.
      • Alferink S. J.J.
      • Zandstra T.
      • Hendriks P.
      • van den Brand H.
      • Gerrits W. J.J.
      Design of climate respiration chambers, adjustable to the metabolic mass of subjects.
      and
      • van Gastelen S.
      • Antunes-Fernandes E.C.
      • Hettinga K.A.
      • Klop G.
      • Alferink S. J.J.
      • Hendriks W.H.
      • Dijkstra J.
      Enteric methane production, rumen volatile fatty acid concentrations, and milk fatty acid composition in lactating Holstein-Friesian cows fed grass silage- or corn silage-based diets.
      . Briefly, the relative humidity and temperature in each CRC compartment was maintained at 80% and 10°C, respectively. The CRC compartments were separated with thin walls and windows to allow audio-visual contact between cows to minimize the effects of social isolation on cow behavior, feed intake, and milk production. Cows were exposed to 16.5 h of light per d (0500 to 2130 h).
      Gas concentrations and ventilation rates were corrected for pressure, temperature, and relative humidity to obtain standard temperature pressure dew point volumes of inlet and exhaust air. The inlet and exhaust air of each CRC were sampled as described by
      • van Gastelen S.
      • Antunes-Fernandes E.C.
      • Hettinga K.A.
      • Klop G.
      • Alferink S. J.J.
      • Hendriks W.H.
      • Dijkstra J.
      Enteric methane production, rumen volatile fatty acid concentrations, and milk fatty acid composition in lactating Holstein-Friesian cows fed grass silage- or corn silage-based diets.
      , with a second gas analyzer used for the additional 2 CRC compartments as described by
      • Nichols K.
      • Dijkstra J.
      • van Laar H.
      • Pacheco S.
      • van Valenberg H.J.
      • Bannink A.
      Energy and nitrogen partitioning in dairy cows at low or high metabolizable protein levels is affected differently by postrumen glucogenic and lipogenic substrates.
      . Production of CO2 and CH4 and consumption of O2 were calculated from the difference between inlet and exhaust gas volumes. Once daily, calibration gases were sampled for gas analysis instead of the inlet air, and the analyzed and actual values of these calibration gases were used to correct the measured gas concentrations from the inlet air and exhaust air of the 6 CRC compartments. At the start and the end of the experiment, each CRC was checked by releasing known amounts of CO2 and comparing these values with the data from the gas analysis system to determine CO2 recovery. The average recovery of CO2 was 100.1% (ranging between 99.5 and 100.7% for individual compartments). Gas measurements during time points when staff entered the CRC compartments (maximum 30 min for milking, feeding, and checking the abomasal infusion lines) were discarded from the data analysis. Production of CO2 and CH4 and consumption of O2 were assumed to be linear between the last data point before opening and the first data point after closing the CRC.

      Sample Collection and Measurements

      Cows were milked twice daily at 0500 and 1530 h during the adaptation and experimental periods. Milk yield was recorded at each milking. Milk samples were collected at each milking (n = 6) during the last 72 h of the infusion period into tubes containing sodium azide (5 µL) for preservation, and were stored no longer than 4 d at 4°C until analysis. An additional milk sample, representative to the milk production (i.e., 5 g sample per kg of milk produced), was collected at each milking from each cow during the last 72 h of the infusion period, pooled per cow, and stored at −20°C until energy and N analyses.
      Fecal samples were collected immediately before each milking (n = 6) during the last 72 h of the infusion period by rectal grab sampling. Feces was scored for consistency using a fecal consistency scale (1–5 scale) according to
      • Ireland-Perry R.L.
      • Stallings C.C.
      Fecal consistency as related to dietary composition in lactating Holstein cows.
      and
      • Zaaijer D.
      • Kremer W. D.J.
      • Noordhuizen J. P. T.M.
      Consistentie van verse koemest.
      , where 1 = runny, liquid consistency, spreads readily; 2 = thin, soft, and loose consistency, may pile slightly and spreads moderately; 3 = soft and loose, but slightly firm (not hard) consistency, piles but spreads slightly; 4 = thick and firm consistency, piles and hardly spreads; and 5 = dry: hard, dry, and stiff consistency, original form not distorted. Fecal samples were pooled into a composite sample by cow and infusion period, which was stored at −20°C until analysis for composition (i.e., for determination of ATTD of nutrients as well as energy and N balance).
      Before the last 2 milking events of the infusion period, feces, urine, and ruminal fluid samples were collected. The fecal sample (∼250 g) was collected by rectal grab sampling and the urine sample was collected by stimulating the dairy cows to urinate by rubbing the perineum using a vertical movement. The ruminal fluid sample (∼200 mL) was obtained as described by
      • van Zijderveld S.M.
      • Fonken B.
      • Dijkstra J.
      • Gerrits W. J.J.
      • Perdok H.B.
      • Fokkink W.
      • Newbold J.R.
      Effects of a combination of feed additives on methane production, diet digestibility, and animal performance in lactating dairy cows.
      , and was composed of equal volumes collected from the front and middle of the ventral sac and from the cranial sac of the rumen. Immediately after collection, each sample was mixed and their pH measured using an electronic pH meter (HI9024C, Hanna Instruments). Subsequently, the urine sample was discarded and a subsample of feces (∼100 g) was collected. A subsample of the ruminal fluid (600 μL) was collected and acidified with an equal volume of 0.85 mg/L orthophosphoric acid containing 19.68 mM isocaproic acid as an internal standard. The subsamples of feces and ruminal fluid (2 subsamples each per cow per infusion period, representing the last 2 milking events) were stored at −20°C until VFA analysis.
      Samples of grass silage, corn silage, and concentrate were collected 3 times weekly during feed preparation. These samples were pooled per experimental period and stored at −20°C until analysis. During the last 72 h of the infusion period in the CRC, feed refusals were collected twice daily (0500 and 1530 h), weighed, and stored at 4°C. At the end of each experimental period, these daily orts were pooled per cow, mixed, subsampled, and stored at −20°C until analysis.
      Measurements of CH4 and CO2 production and O2 consumption were based on data recorded from d 4 (0800 h) through d 6 (0800 h) of each experimental period, whereas energy and N balance and the ATTD of nutrients were based on manure and fecal samples collected from d 3 (1000 h) through d 6 (0900 h). Cows were weighed at the start and at the end of each balance period. Each CRC compartment was cleaned at 0900 h on d 3 (taking approximately 60 min) to remove manure collected from the end of the previous period to facilitate a new total collection period. The manure (feces and urine) produced during the balance period (i.e., 1000 h on d 3 through 0900 h on d 6) was totally collected, weighed, mixed, subsampled, and stored at −20°C until analysis. To quantify the contribution of N from volatilized ammonia appearing from excreted and mixed feces and urine, samples of condensed water from the chamber heat exchanger and from a 25% sulfuric acid solution (wt/wt), through which aerial ammonia in outflowing air was trapped, were collected. This is according to
      • Mesgaran S.D.
      • Kühla B.
      • Baumont R.
      • Cantalapiedra-Hijar G.
      • Noziére P.
      • Lund P.
      • Humphries D.
      • Dijkstra J.
      Nutrient digestibility and balance studies.
      . The total amount of condensed water was recorded and a representative sample was collected at the end of each balance period for each CRC compartment. For the 25% sulfuric acid solution, we recorded per each balance period and for each CRC compartment the amount of air flowing through and the weight of the sample collected. Both types of samples were stored at 4°C until analysis.
      On d 5 of each experimental period, blood samples were collected from the coccygeal vessels (assumed mixture of arterial and venous blood with negligible metabolism across the tissues of the tail;
      • Emery R.S.
      • Brown L.D.
      • Bell J.W.
      Correlation of milk fat with dietary and metabolic factors in cows fed restricted-roughage rations supplemented with magnesium oxide or sodium bicarbonate.
      ) into 10-mL sodium heparin and potassium EDTA Vacutainers (Becton Dickinson) at 0800, 1000, 1200, and 1400 h. At each sampling moment, a subsample was collected immediately from the heparin Vacutainer and analyzed by a handheld VetScan i-STAT 1 analyzer (ABAXIS Inc.) using disposable G3+ cartridges [analyzing whole blood for pH, partial pressure CO2 (mmHg), partial pressure O2 (mmHg), base excess (mmol/L) HCO3 (mmol/L), total CO2 (mmol/L), and O2 saturation (%); Abbott] and CHEM 8+ cartridges [analyzing whole blood for Na (mmol/L), K (mmol/L), Cl (mmol/L), ionized Ca (mmol/L), glucose (mg/dL), urea nitrogen (mg/dL), creatinine (mg/dL), hematocrit (% PCV), hemoglobin (g/dL), and anion gap (mmol/L); Abbott]. Immediately after the subsample was collected, the heparin and potassium EDTA Vacutainers were placed in ice and subsequently centrifuged at 5,000 × g for 15 min at room temperature. Plasma was aliquoted by time point and stored at −20°C until further analysis (where each time point was analyzed separately).

      Analytical Procedures

      Samples of grass silage, corn silage, concentrate, feed refusals, manure, and feces were thawed at room temperature, freeze-dried until a constant weight was reached, and ground to pass a 1-mm screen using a cross beater mill for the grass silage, corn silage, and feed refusals (Peppink 100AN) and an ultra-centrifugal mill for all other samples (Retsch ZM200, Retsch GmbH). Wet chemical analysis for DM, ash, N, NH3-N, starch, reducing sugars (i.e., all carbohydrates with reducing properties and soluble in 40% ethanol), crude fat, NDF, ADF, ADL, and Ti was performed as described by
      • Nichols K.
      • Bannink A.
      • Pacheco S.
      • van Valenberg H.J.
      • Dijkstra J.
      • van Laar H.
      Feed and nitrogen efficiency are affected differently but milk lactose production is stimulated equally when isoenergetic protein and fat is supplemented in lactating dairy cow diets.
      . Bomb calorimetry (ISO 9831;
      • International Organization for Standardization
      ISO 9831:1998. Animal feedstuffs, animal products, and feces or urine–Determination of gross caloric values–Bomb calorimeter method.
      ; adiabatic bomb calorimeter, IKA-C700, Janke and Kunkel) was used to determine gross energy (GE) content. Crude protein was calculated as N × 6.25.
      Grass silage, corn silage, and concentrate were analyzed for DM, ash, N, crude fat, starch (corn silage and concentrate only), sugars (grass silage and concentrate only), NDF, ADF, ADL, GE, and Ti (concentrate only). Feed refusals were analyzed for DM. Manure samples were analyzed for DM, N, and GE. Fecal samples were analyzed for DM, ash, N, crude fat, starch, NDF, GE, and Ti. In addition, samples of condensed water and the sulfuric acid solution were analyzed for N. Milk samples from individual milking events were analyzed for fat, protein, lactose, urea, and BHB content by mid-infrared spectroscopy (ISO 9622;
      • International Organization for Standardization
      ISO 9622:2013. Milk and liquid milk products. Guidelines for the application of mid-infrared spectrometry.
      ; Vereniging Veehouderijbelangen). Pooled milk samples were analyzed for GE and N in fresh material as described above. Samples of ruminal fluid and feces were analyzed for VFA as described by
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      .
      Blood plasma was analyzed by University Veterinary Diagnostic Laboratory (Utrecht, the Netherlands) for urea, insulin, BHB, NEFA, albumin, serum amyloid A (SAA), and haptoglobin. Plasma NEFA, BHB, insulin, and urea were analyzed as described by
      • van Knegsel A. T.M.
      • van den Brand H.
      • Dijkstra J.
      • van Straalen W.M.
      • Jorritsma R.
      • Tamminga S.
      • Kemp B.
      Effect of glucogenic vs. lipogenic diets on energy balance, blood metabolites, and reproduction in primiparous and multiparous dairy cows in early lactation.
      . Plasma SAA, haptoglobin, and albumin were analyzed as described by
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      . The intraassay coefficients of variation for SAA, haptoglobin, and albumin, were 5.0, 5.3, and 0.8%, respectively. Algemeen Medisch Laboratorium (Antwerp, Belgium) analyzed blood plasma for cobalt content using inductively coupled plasma MS and a NexION 350D ICP-MS (Perkin Elmer).

      Calculations and Statistical Analysis

      Reported values for nutrient content of the TMR were calculated from ration composition and analyzed values obtained for grass silage, corn silage, and concentrate. For the grass silage and corn silage, intestinal digestible protein, rumen degradable protein balance, and NEL contents were calculated based on the chemical composition as obtained by near-infrared spectroscopy analysis (Eurofins Agro, Wageningen, the Netherlands). For the concentrate, intestinal digestible protein, rumen degradable protein balance, and NEL contents were calculated based on table values for composition of the ingredients (
      • CVB
      Chemische Samenstellingen en Nutritionele Waarden van Voedermiddelen (in Dutch).
      ). The ATTD of nutrients was calculated taking into account nutrient inflow from the TMR and the treatment infusions.
      All variables related to feed intake, milk yield, and milk composition were averaged over the 3-d balance period. Whole-blood constituents measured with the i-STAT as well as the plasma constituents, ruminal and fecal VFA, and ruminal, fecal, and urine pH were averaged over the analyzed sampling times. The variables related to energy and N balance were expressed per kilogram of metabolic BW (BW0.75) per day. Cow was considered the experimental unit for all variables. Data were analyzed using the MIXED procedure in SAS (version 9.4, SAS Institute Inc.). The model contained main effects and interaction effects of infusion treatment factors (BHB and corn starch) as fixed effects, and cow and period as random effects. Differences were considered significant at P ≤ 0.050 and tendencies at 0.050 < P ≤ 0.100. Multiple comparisons between treatment means were made using the Tukey-Kramer method when an effect of corn starch or a corn starch × BHB interaction was detected at P ≤ 0.050. Treatment arrangement within the Latin square was balanced for first-order carryover effects in subsequent periods (
      • Williams E.J.
      Experimental designs balanced for the estimation of residual effects of treatments.
      ), where each treatment immediately preceded and followed every other treatment exactly once. We observed carryover effects between periods, assessed by testing for an effect of the previous treatment in the ANOVA, for blood Na (mmol/L; P = 0.023), anion gap (mmol/L; P = 0.013), and SAA (mg/L; P = 0.003).

      RESULTS

      Nutrient Intake and Digestibility

      Nutrient intake and ATTD of nutrients were not affected by corn starch × BHB interactions, with the exception of DMI of the infusate (tendency only; P = 0.088) and the ATTD of starch (P = 0.037; Table 3). The ATTD of starch increased in the presence of BHB, but only when combined with 3.0 kg of corn starch/d.
      Table 3Intake and apparent total-tract digestibility of nutrients in early-lactation dairy cows receiving abomasal infusions of corn starch, BHB, or a combination of both
      ItemCorn starch (kg/d)BHB (mol/d)SEMP-value
      0.01.53.00.08.0StarchBHBStarch × BHB
      Nutrient intake (TMR + abomasal infusions, kg/d unless stated otherwise)
       DM TMR17.2
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      16.9
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      16.0
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      16.916.50.700.0240.1460.849
       DM infusate1.19
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      2.40
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.76
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      2.072.830.024<0.001<0.0010.088
       DM total18.4
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      19.2
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      19.8
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      19.019.30.700.0060.4030.922
       OM16.4
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      17.2
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      17.6
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      16.917.20.630.0100.3000.906
       Digestible OM12.712.812.712.612.80.470.7370.2200.849
       CP3.41
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.32
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.22
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.353.280.1150.0170.1720.919
       Crude fat0.6540.6490.6260.6520.6340.02890.1630.1540.885
       Gross energy (MJ/d)338.3
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      353.1
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      361.4
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      347.5354.413.260.0160.2600.910
       NDF6.07
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      5.95
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      5.67
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      5.985.810.2800.0250.1550.881
       Digestible NDF4.43
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      4.22
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      4.05
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      4.334.140.1880.0050.0310.663
       ADF3.37
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.30
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.14
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.323.220.1540.0240.1550.873
       ADL0.174
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      0.171
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      0.163
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      0.1720.1670.00860.0250.1480.789
       Starch2.51
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.64
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      4.75
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.663.610.108<0.0010.3780.919
       Sugar1.26
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1.23
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1.17
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1.241.200.0550.0190.1210.641
      Apparent digestibility (% of intake)
       DM76.2
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      74.3
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      72.1
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      74.174.30.73<0.0010.7920.390
       OM77.2
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      74.7
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      72.0
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      74.574.70.77<0.0010.6920.439
       CP72.5
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      67.5
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      67.8
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      69.768.80.58<0.0010.0440.157
       Crude fat59.857.058.059.656.91.790.1770.0310.534
       Gross energy74.4
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      72.1
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      70.3
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      72.272.30.80<0.0010.8400.415
       NDF73.070.971.872.471.41.040.1080.2180.448
       Starch
      Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 98.1x, 0.0 kg/d corn starch + 8.0 mol/d BHB = 97.8x, 1.5 kg/d corn starch + 0.0 mol/d BHB = 97.9x, 1.5 kg/d corn starch + 8.0 mol/d BHB = 97.2x, 3.0 kg/d corn starch + 0.0 mol/d BHB = 89.0y, 3.0 kg/d corn starch + 8.0 mol/d BHB = 92.2z; a different superscript (x–z) indicates a significant difference (P < 0.05).
      97.997.690.695.095.70.50<0.0010.2690.037
      a–c Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1 Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 98.1x, 0.0 kg/d corn starch + 8.0 mol/d BHB = 97.8x, 1.5 kg/d corn starch + 0.0 mol/d BHB = 97.9x, 1.5 kg/d corn starch + 8.0 mol/d BHB = 97.2x, 3.0 kg/d corn starch + 0.0 mol/d BHB = 89.0y, 3.0 kg/d corn starch + 8.0 mol/d BHB = 92.2z; a different superscript (x–z) indicates a significant difference (P < 0.05).
      Regardless of BHB infusion, TMR DMI was lower (P = 0.024) with infusion of 3.0 kg of starch/d compared with 0.0 kg of starch/d, with both not differing from 1.5 kg of starch/d. The same was observed for the intake of CP, NDF, digestible NDF, ADF, ADL, and sugar, where the intake of all these nutrients was lower (P < 0.025) with infusion of 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, and with both not differing from infusion of 1.5 kg of corn starch/d. Abomasal infusion of corn starch increased the DMI of the infusate (P < 0.001) due to the design of the infusion treatments. Subsequently, total DMI (DMI TMR + DMI infusate) and the intake of OM and GE were higher (P < 0.016) with infusion of 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with both not differing from 1.5 kg of corn starch/d. The intake of starch increased (P < 0.001) with each increasing level of abomasal infusion of corn starch. The ATTD of DM, OM and GE decreased (P < 0.001) with each increasing level of infused corn starch. The ATTD of CP was lower (P < 0.001) for 1.5 kg and 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with no difference between 1.5 and 3.0 kg of corn starch/d.
      Regardless of the presence of corn starch in the infusion, DMI of the infusate increased (P < 0.001) upon abomasal infusion of BHB due to the design of the infusion treatments. Infusion of BHB decreased (P < 0.044) the intake of digestible NDF, ATTD of CP, and ATTD of crude fat compared with no BHB infusion.

      Milk Production and Composition

      Milk yield (P = 0.075) and milk protein concentration (P = 0.090) tended to be affected and feed efficiency (kg of fat- and protein-corrected milk/kg of total DMI) was affected (P = 0.010) by a corn starch × BHB interaction (Table 4). Abomasal infusion of 3.0 kg of corn starch/d decreased feed efficiency, but only when BHB was not present. Regardless of BHB infusion, abomasal infusion of corn starch tended to decrease (P = 0.084) milk fat concentration and decreased (P < 0.001) milk urea concentration. Milk urea concentration was lower with infusion of 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with both not differing from 1.5 kg of corn starch/d. Regardless of the presence of corn starch in the infusion, milk urea concentration decreased (P = 0.007) and milk BHB concentration increased (P < 0.001) with abomasal infusion of BHB compared with no BHB infusion.
      Table 4Lactation characteristics of early-lactation dairy cows receiving abomasal infusions of corn starch, BHB, or a combination of both
      ItemCorn starch (kg/d)BHB (mol/d)SEMP-value
      0.01.53.00.08.0StarchBHBStarch × BHB
      Milk yield (kg/d)23.123.923.523.323.71.750.3730.3970.075
      FPCM
      Fat- and protein-corrected milk = (0.337 + 0.116 × fat % + 0.06 × protein %) × milk yield (kg/d); CVB, 2018.
      (kg/d)
      24.124.523.823.924.31.310.3120.3460.107
      Feed efficiency
      Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 1.32x, 0.0 kg/d corn starch + 8.0 mol/d BHB = 1.29x, 1.5 kg/d corn starch + 0.0 mol/d BHB = 1.30x, 1.5 kg/d corn starch + 8.0 mol/d BHB = 1.25x, 3.0 kg/d corn starch + 0.0 mol/d BHB = 1.16y, 3.0 kg/d corn starch + 8.0 mol/d BHB = 1.24x; a different superscript (x,y) indicates a significant difference (P < 0.05).
      (kg of FPCM/kg of DMI total)
      1.311.271.201.261.260.047<0.0010.8960.010
      Fat concentration (g/100 g)4.424.274.164.284.290.1240.0840.8690.578
      Protein concentration (g/100 g)3.373.373.353.383.350.1520.6680.1740.090
      Lactose concentration (g/100 g)4.674.704.694.704.670.0550.6910.3740.109
      Urea concentration (mg/dL)22.9
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      19.2
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      18.7
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      21.219.30.76<0.0010.0070.553
      Fat yield (g/d)1,0051,0069649821,00148.40.1490.3440.178
      Protein yield (g/d)77379777778078542.90.4440.7630.527
      Lactose yield (g/d)1,0781,1191,1011,0941,10479.90.3980.6670.104
      BHB (mmol/L)0.0820.0730.0730.0610.0910.00810.198<0.0010.811
      a,b Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1 Fat- and protein-corrected milk = (0.337 + 0.116 × fat % + 0.06 × protein %) × milk yield (kg/d);
      • CVB
      Chemische Samenstellingen en Nutritionele Waarden van Voedermiddelen (in Dutch).
      .
      2 Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 1.32x, 0.0 kg/d corn starch + 8.0 mol/d BHB = 1.29x, 1.5 kg/d corn starch + 0.0 mol/d BHB = 1.30x, 1.5 kg/d corn starch + 8.0 mol/d BHB = 1.25x, 3.0 kg/d corn starch + 0.0 mol/d BHB = 1.16y, 3.0 kg/d corn starch + 8.0 mol/d BHB = 1.24x; a different superscript (x,y) indicates a significant difference (P < 0.05).

      Blood and Plasma Constituents

      Blood partial pressure of O2 and anion gap tended (P ≤ 0.093; Table 5) to be affected by a corn starch × BHB interaction. The results of the anion gap should be interpreted with caution because, as mentioned earlier, this variable was also significantly affected by a carryover effect between periods. Regardless of the presence of BHB, total CO2 tended to decrease (P = 0.072) and ionized calcium tended to increase (P = 0.074) upon infusion of corn starch. Blood pH and HCO3 were lower (P ≤ 0.048) with infusion of 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with both not differing from 1.5 kg of corn starch/d. Blood base excess was lower (P = 0.006) for 3.0 kg of corn starch/d compared with 0.0 and 1.5 kg of corn starch/d, with no difference between the latter 2 levels of corn starch. Blood urea N was lower (P < 0.001) for 1.5 and 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with no difference between the former 2 levels of corn starch. Blood chloride concentration was higher (P = 0.006) with infusion of 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with both not differing from 1.5 kg of corn starch/d.
      Table 5Blood and plasma constituents of early-lactation dairy cows receiving abomasal infusions of corn starch, BHB, or a combination of both
      ItemCorn starch (kg/d)BHB (mol/d)SEMP-value
      0.01.53.00.08.0StarchBHBStarch × BHB
      Blood constituent
       pH7.43
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      7.42
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      7.40
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      7.427.410.0110.0280.2830.155
       pCO2
      pCO2 = partial pressure CO2.
      (mm Hg)
      42.142.243.343.042.11.060.3870.2510.536
       pO2
      pO2 = partial pressure O2.
      (mm Hg)
      758568767613.00.1250.9410.075
       Base excess
      Base excess is defined as the mEq of acid or base that must be added to 1 L of blood to restore the pH to 7.40 at 37°C and at a pCO2 of 40 mmHg (Corey, 2003).
      (mmol/L)
      3.4
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.0
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1.7
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.32.00.580.0060.0020.302
       HCO3 (mmol/L)27.9
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      27.7
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      26.8
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      28.126.90.560.0480.0020.516
       Total CO2 (mmol/L)29292829280.60.0720.0020.498
       sO2
      sO2 = O2 saturation.
      (%)
      85878285843.20.2870.6270.405
       Sodium (mmol/L)1391391391391390.30.2460.9990.288
       Potassium (mmol/L)4.74.74.94.74.80.090.1340.4260.562
       Chloride (mmol/L)101
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      101
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      102
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1011020.60.0060.1730.516
       Ionized calcium (mmol/L)1.261.251.281.261.260.0120.0740.4820.435
       Glucose (mg/dL)66676768651.90.8670.0060.556
       Hematocrit
      PCV = ratio between the volume of red blood cells and the total volume of blood.
      (% PCV)
      26252525250.90.2980.4940.619
       Hemoglobin (g/dL)8.548.498.668.658.470.3320.6440.2560.671
       Urea N (mg/dL)13.8
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      11.0
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      11.5
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      12.411.81.10<0.0010.1690.774
       Creatinine (mg/dL)0.570.550.550.560.550.0380.3310.4840.419
       Anion gap (mmol/L)18.117.618.017.618.30.410.2160.0150.093
      Plasma constituent
       Urea (mmol/L)4.34.04.24.24.10.290.2690.4880.961
       Insulin (mIU/L)12.612.813.013.711.92.800.9400.0310.154
       BHB (mmol/L)0.570.590.580.500.660.0670.9320.0130.387
       NEFA
      NEFA = nonesterified fatty acids.
      (mmol/L)
      0.170.170.170.180.160.0150.9510.2080.708
       Albumin (g/L)33323333330.80.3570.5910.452
       Serum amyloid A (mg/L)93124791148325.40.3400.2210.169
       Haptoglobin (g/L)0.690.490.700.630.620.1460.4750.9430.917
       Cobalt (ug/L)22524626522626529.80.2150.0450.124
      a,b Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1 pCO2 = partial pressure CO2.
      2 pO2 = partial pressure O2.
      3 Base excess is defined as the mEq of acid or base that must be added to 1 L of blood to restore the pH to 7.40 at 37°C and at a pCO2 of 40 mmHg (
      • Corey H.E.
      Stewart and beyond: New models of acid-base balance.
      ).
      4 sO2 = O2 saturation.
      5 PCV = ratio between the volume of red blood cells and the total volume of blood.
      6 NEFA = nonesterified fatty acids.
      Regardless of the presence of corn starch in the infusion, BHB infusion decreased (P ≤ 0.031) blood base excess; blood concentration of HCO3, total CO2, and glucose; and plasma concentration of insulin, compared with no BHB infusion. Infusion of BHB increased (P ≤ 0.045) blood anion gap, and plasma BHB and cobalt concentration compared with no BHB infusion. Also here, the results of the anion gap should be interpreted with caution because this variable was also significantly affected by a carryover effect between periods.

      Energy and Nitrogen Balance

      None of the variables associated with energy and N balance were affected by a corn starch × BHB interaction (Table 6). Regardless of the presence of BHB, metabolic BW tended to decrease in response to corn starch infusion. The respiratory quotient was higher (P < 0.001) with infusion of 1.5 and 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with no difference between the former 2 levels of corn starch. Intake of GE (GEI) was higher (P = 0.013) with infusion of 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with both not differing from 1.5 kg of corn starch/d. Energy excreted in manure increased (P < 0.001) with every increasing level of infused corn starch. The ratio between metabolizable energy intake (MEI) and GEI was lower (P < 0.001) for 3.0 kg of corn starch/d compared with 0.0 and 1.5 kg of corn starch/d, with no difference between the latter 2 levels of corn starch.
      Table 6The energy and nitrogen balance of early-lactation dairy cows receiving abomasal infusions of corn starch, BHB, or a combination of both
      ItemCorn starch (kg/d)BHB (mol/d)SEMP-value
      0.01.53.00.08.0StarchBHBStarch × BHB
      Respiratory quotient1.120
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1.147
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1.147
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1.1451.1310.0100<0.0010.0010.786
      Metabolic BW
      The mean BW per cow per balance period was used to calculate metabolic BW (BW0.75).
      (kg0.75)
      1241241231241242.30.0960.3060.847
      Energy balance (kJ/kg of BW0.75 per day, unless stated otherwise)
       GEI
      GEI = gross energy intake (TMR plus infusate).
      2,729
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      2,842
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      2,930
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      2,8092,859110.30.0130.3340.923
       CH4 production1771731651751687.60.1240.1460.807
       Energy in manure854
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      944
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1,044
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      95094546.6<0.0010.8580.858
       DEI
      DEI (digestible energy intake) = GEI × apparent total-tract digestibility of GE (% of intake).
      2,0312,0522,0582,0292,06584.40.7840.2920.846
       MEI
      MEI (metabolizable energy intake) = GEI − methane production − energy in manure (feces + urine).
      1,6991,7261,7221,6851,74665.00.6250.0220.885
       MEI to GEI ratio (%)62.3
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      60.8
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      58.8
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      60.161.10.64<0.0010.1060.962
       Heat production
      Heat production (kJ/d) = 16.175 × VO2 (L/d) + 5.021 × VCO2 (L/d), where VO2 = volumes of O2 consumed, and VCO2 = volumes of CO2 produced (Gerrits et al., 2015).
      95796195895296628.30.9870.4520.999
       Energy in milk60261159960160729.10.6440.5450.195
       ER total
      Energy retention total = MEI − heat production − energy in milk.
      13915416513317236.50.6940.1200.484
       ER protein
      Energy retention protein = protein gain (N × 6.25) × 23.6 kJ/g (energetic value of protein).
      73606170599.00.5230.3390.383
       ER fat
      Energy retention fat = energy retention total − energy retention protein.
      66941046311331.90.4140.0480.721
      Nitrogen balance (mg/kg of BW0.75 per day)
       N intake
      N intake from TMR plus infusate.
      4,400
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      4,279
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      4,173
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      4,3354,234149.00.0350.1380.929
       N manure2,8492,8112,7162,8072,77785.20.1650.6040.496
      N feces
      N feces = N intake × [100 − apparent total-tract digestibility of N (% of intake)]/100.
      1,211
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1,388
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1,344
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1,3101,31945.7<0.0010.7670.499
      N urine
      N urine = N manure − N feces.
      1,638
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1,423
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1,373
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1,4971,45868.20.0070.5560.331
       N milk97399898598498753.00.6420.8740.629
       N condensed water + N acid82
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      65
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      60
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      726610.3<0.0010.1420.399
       N balance49640541247240361.40.5230.3390.383
      N efficiency
      N efficiency = N milk/N intake (%).
      22.1
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      23.3
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      23.6
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      22.723.30.900.0040.1050.185
      a–c Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1 The mean BW per cow per balance period was used to calculate metabolic BW (BW0.75).
      2 GEI = gross energy intake (TMR plus infusate).
      3 DEI (digestible energy intake) = GEI × apparent total-tract digestibility of GE (% of intake).
      4 MEI (metabolizable energy intake) = GEI − methane production − energy in manure (feces + urine).
      5 Heat production (kJ/d) = 16.175 × VO2 (L/d) + 5.021 × VCO2 (L/d), where VO2 = volumes of O2 consumed, and VCO2 = volumes of CO2 produced (
      • Gerrits W. J.J.
      • van den Borne J. J. G.C.
      • Labussière E.
      Deriving heat production from gaseous exchange: Validity of the approach.
      ).
      6 Energy retention total = MEI − heat production − energy in milk.
      7 Energy retention protein = protein gain (N × 6.25) × 23.6 kJ/g (energetic value of protein).
      8 Energy retention fat = energy retention total − energy retention protein.
      9 N intake from TMR plus infusate.
      10 N feces = N intake × [100 − apparent total-tract digestibility of N (% of intake)]/100.
      11 N urine = N manure − N feces.
      12 N efficiency = N milk/N intake (%).
      Nitrogen intake was lower with infusion of 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with both not differing from 1.5 kg of corn starch/d. Nitrogen excreted via urine and N trapped in condensed water and acid were higher (P ≤ 0.007) for 0.0 kg of corn starch/d compared with 1.5 and 3.0 kg of corn starch/d, with no difference between the latter 2 levels of corn starch. Infusion of corn starch increased (P ≤ 0.004) N excreted via feces and N efficiency, where both were lower with 0.0 kg of corn starch/d compared with 1.5 and 3.0 kg of corn starch/d, with no difference between the latter 2 levels of corn starch.
      Regardless of the presence of corn starch, infusion of BHB decreased (P = 0.001) the respiratory quotient, but increased (P ≤ 0.048) MEI and energy retained as fat compared with no BHB infusion. Abomasal infusion of BHB did not affect any of the N balance variables.

      Ruminal Fluid, Feces, and Urine

      Ruminal butyrate proportions, fecal total VFA concentration, and fecal valerate proportions were affected (P ≤ 0.031) by a corn starch × BHB interaction (Table 7). Despite the significant corn starch × BHB interaction for ruminal butyrate, no significant differences were found between treatment means using the Tukey-Kramer method. Infusion of BHB decreased the fecal total VFA concentration, but only when combined with 3.0 kg of corn starch/d. Fecal valerate proportions decreased with every increasing level of infused corn starch, but only in the absence of BHB. With BHB infusion, the fecal valerate proportion was higher for 0.0 kg of corn starch/d compared with 1.5 and 3.0 kg of corn starch/d, with no difference between the latter 2 levels of corn starch.
      Table 7pH and VFA of ruminal fluid, feces, and urine (pH only) of early-lactation dairy cows receiving abomasal infusions of corn starch, BHB, or a combination of both
      ItemCorn starch (kg/d)BHB (mol/d)SEMP-value
      0.01.53.00.08.0StarchBHBStarch × BHB
      Rumen
       pH6.306.206.313.296.250.0780.2250.4480.923
       Total VFA (mM)1061151141101134.50.0690.3660.188
       VFA (% of total VFA)
      Acetate65.565.065.765.565.40.840.3000.8720.368
      Propionate18.919.019.019.118.80.830.9430.4980.147
      Butyrate
      Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 11.6, 0.0 kg/d corn starch + 8.0 mol/d BHB = 10.9, 1.5 kg/d corn starch + 0.0 mol/d BHB = 10.8, 1.5 kg/d corn starch + 8.0 mol/d BHB = 11.7, 3.0 kg/d corn starch + 0.0 mol/d BHB = 10.4, 3.0 kg/d corn starch + 8.0 mol/d BHB = 11.4; no significant differences (P > 0.05).
      11.311.310.910.911.30.410.4310.1940.031
      Isobutyrate0.830.830.850.830.840.0280.6770.9170.228
      Valerate1.581.651.581.601.600.0620.3110.9890.505
      Isovalerate1.912.181.962.041.990.2380.3480.7540.431
       Acetate to propionate ratio3.53.53.53.53.50.170.9190.4270.265
      Feces
       Consistency2.5
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      2.8
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      3.0
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      2.92.60.190.0500.1310.186
       pH6.49
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      6.00
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      5.15
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      5.955.810.094<0.0010.1320.123
       Total VFA
      Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 54xy, 0.0 kg/d corn starch + 8.0 mol/d BHB = 61x, 1.5 kg/d corn starch + 0.0 mol/d BHB = 60xy, 1.5 kg/d corn starch + 8.0 mol/d BHB = 62x, 3.0 kg/d corn starch + 0.0 mol/d BHB = 61x, 3.0 kg/d corn starch + 8.0 mol/d BHB = 48y; a different superscript (x,y) indicates a significant difference (P < 0.05).
      (mM)
      58615558573.20.1900.5660.015
       VFA (% of total VFA)
      Acetate75.373.572.074.972.31.060.1470.0690.894
      Propionate13.4
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      11.9
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      4.9
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      10.79.51.25<0.0010.3310.534
      Butyrate8.4
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      13.4
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      22.5
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      12.816.71.22<0.0010.0090.305
      Isobutyrate1.12
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      0.48
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      0.24
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      0.710.520.101<0.0010.1310.838
      Valerate
      Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 1.05x, 0.0 kg/d corn starch + 8.0 mol/d BHB = 0.97x, 1.5 kg/d corn starch + 0.0 mol/d BHB = 0.48y, 1.5 kg/d corn starch + 8.0 mol/d BHB = 0.22yz, 3.0 kg/d corn starch + 0.0 mol/d BHB = 0.01z, 3.0 kg/d corn starch + 8.0 mol/d BHB = 0.42yz; a different superscript (x–z) indicates a significant difference (P < 0.05).
      1.010.350.210.510.540.077<0.0010.8010.022
      Isovalerate0.67
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      0.34
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      0.05
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      0.380.330.053<0.0010.4480.813
       Acetate to propionate ratio5.7
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      8.1
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      14.2
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      9.29.51.21<0.0010.8170.330
      Urine
       pH7.64
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      7.65
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      7.38
      Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      7.617.500.066<0.0010.0570.431
      a–c Least squares means within a row with a different superscript indicate a significant difference (P < 0.05) between 0.0, 1.5, and 3.0 kg of corn starch/d.
      1 Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 11.6, 0.0 kg/d corn starch + 8.0 mol/d BHB = 10.9, 1.5 kg/d corn starch + 0.0 mol/d BHB = 10.8, 1.5 kg/d corn starch + 8.0 mol/d BHB = 11.7, 3.0 kg/d corn starch + 0.0 mol/d BHB = 10.4, 3.0 kg/d corn starch + 8.0 mol/d BHB = 11.4; no significant differences (P > 0.05).
      2 Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 54xy, 0.0 kg/d corn starch + 8.0 mol/d BHB = 61x, 1.5 kg/d corn starch + 0.0 mol/d BHB = 60xy, 1.5 kg/d corn starch + 8.0 mol/d BHB = 62x, 3.0 kg/d corn starch + 0.0 mol/d BHB = 61x, 3.0 kg/d corn starch + 8.0 mol/d BHB = 48y; a different superscript (x,y) indicates a significant difference (P < 0.05).
      3 Starch × BHB interaction: 0.0 kg/d corn starch + 0.0 mol/d BHB = 1.05x, 0.0 kg/d corn starch + 8.0 mol/d BHB = 0.97x, 1.5 kg/d corn starch + 0.0 mol/d BHB = 0.48y, 1.5 kg/d corn starch + 8.0 mol/d BHB = 0.22yz, 3.0 kg/d corn starch + 0.0 mol/d BHB = 0.01z, 3.0 kg/d corn starch + 8.0 mol/d BHB = 0.42yz; a different superscript (x–z) indicates a significant difference (P < 0.05).
      Regardless of the presence of BHB in the infusion, starch infusion did not affect any of the ruminal fermentation variables. Fecal consistency score was higher (P = 0.050) for 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with both levels of corn starch not differing from 1.5 kg of corn starch/d. Fecal pH and fecal isovalerate proportion decreased (P < 0.001) with every increasing level of corn starch infused. Fecal propionate proportion and urine pH were lower (P < 0.001) with 3.0 kg of corn starch/d compared with 0.0 and 1.5 kg of corn starch/d, with no difference between the latter 2 levels of corn starch. Infusion of corn starch also decreased (P < 0.001) fecal isobutyrate proportions, which were lower with 1.5 and 3.0 kg of corn starch/d compared with 0.0 kg of corn starch/d, with no difference between the former 2 levels of corn starch. Fecal butyrate proportions increased (P < 0.001) with every increasing level of infused corn starch. The acetate to propionate ratio in feces increased (P < 0.001) with infusion of 3.0 kg of corn starch/d compared with 0.0 and 1.5 kg of corn starch/d, with no difference between the latter 2 levels of corn starch.
      Regardless of the presence of corn starch, fecal acetate proportion (P = 0.069) and urine pH (P = 0.057) tended to decrease with abomasal infusion of BHB. Fecal butyrate proportions were higher (P = 0.009) when BHB was present in the infusion compared with no BHB.

      DISCUSSION

      The aim of this study was to induce hindgut and metabolic acidosis via abomasal infusion of corn starch and BHB, respectively, and to determine the effects of these physiological states in early-lactation cows. Ruminal fermentation characteristics were largely unaffected by either starch or BHB infusion, suggesting that there was no appreciable back-flow of infusates into the rumen. Therefore, our aim to induce acidosis only posterior to the rumen was achieved.

      Establishment of Hindgut Acidosis

      Currently, there is no clear fecal pH threshold defined for hindgut acidosis, but fecal pH was previously reported to be 6.30 (
      • Li S.
      • Khafipour E.
      • Krause D.O.
      • Kroeker A.
      • Rodriguez-Lecompte J.C.
      • Gozho G.N.
      • Plaizier J.C.
      Effects of subacute ruminal acidosis challenges on fermentation and endotoxins in the rumen and hindgut of dairy cows.
      ) to 6.49 (
      • Danscher A.M.
      • Li S.
      • Andersen P.H.
      • Khafipour E.
      • Kristensen N.B.
      • Plaizier J.C.
      Indicators of induced subacute ruminal acidosis (SARA) in Danish Holstein cows.
      ) for control-fed animals. Considering these values as a threshold in combination with the observed fecal pH, hindgut acidosis was achieved upon infusion of 1.5 and 3.0 kg/d starch, and became more severe with a higher corn starch level.
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      reported a fecal pH of 6.86 without abomasal infusion of ground corn and a fecal pH of 6.00 when 3.0 kg of ground corn/d (equals ∼1.5 kg of starch/d, when corrected for DM and starch content) was infused into the abomasum. This is in agreement with the results of the present study, where infusion of 1.5 kg of corn starch/d resulted in an average fecal pH of 6.00 and infusion of 3 kg of corn starch/d decreased fecal pH even further.
      The decrease in fecal pH with 1.5 and 3.0 kg of corn starch/d suggests that part of the infused corn starch escaped intestinal digestion and was fermented in the hindgut (
      • Reynolds C.K.
      • Cammell S.B.
      • Humphries D.J.
      • Beever D.E.
      • Sutton J.D.
      • Newbold J.R.
      Effects of postrumen starch infusion on milk production and energy metabolism in dairy cows.
      ). This is supported by the decrease in ATTD of CP, likely as a result of increased microbial protein synthesis, and by the increase in respiratory quotient, likely as a result of anaerobic fermentation of dietary carbohydrates (
      • Gerrits W. J.J.
      • van den Borne J. J. G.C.
      • Labussière E.
      Deriving heat production from gaseous exchange: Validity of the approach.
      ). Contrary to
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      , we did not observe an increase in total VFA concentration in feces upon abomasal infusion of corn starch. This lack of effect on fecal total VFA concentrations combined with decreased fecal pH suggests that the buffering capacity of the large intestine may have changed with abomasal infusion of corn starch.
      Infusion of corn starch in the abomasum appeared to disturb the acid-base status of the dairy cows, likely related to increased VFA absorption from the hindgut during corn starch infusion. In the rumen, absorption of VFA is suspected to be primarily regulated by the anion exchange pathway, where VFA are absorbed in exchange for the secretion of HCO3 (
      • Aschenbach J.R.
      • Bilk S.
      • Tadesse G.
      • Stumpff F.
      • Gabel G.
      Bicarbonate-dependent and bicarbonate-independent mechanisms contribute to nondiffusive uptake of acetate in the ruminal epithelium of sheep.
      ). An acid load in the rumen depleted HCO3 from the blood and decreased blood pH in beef steers (
      • Brown M.S.
      • Krehbiel C.R.
      • Galyean M.L.
      • Remmenga M.D.
      • Peters J.P.
      • Hibbard B.
      • Robinson J.
      • Moseley W.M.
      Evaluation of models of acute and subacute acidosis on dry matter intake, ruminal fermentation, blood chemistry, and endocrine profiles of beef steers.
      ). This mechanism is likely valid for the hindgut as well, especially when excessive fermentation takes place.
      The increase in the molar proportion of fecal butyrate upon abomasal infusion of corn starch is consistent with what has been observed in monogastric animals during hindgut fermentation of slowly fermentable starch (
      • Lv S.
      • Li D.
      • Xing J.
      • Ma Y.
      • Huang S.
      • Li J.
      Effects of extrusion of corn on growth performance, nutrient digestibility and short-chain fatty acid profiles in the hindgut of weaned piglets.
      ), as well as in dairy cattle fed a high concentrate diet (
      • Mao S.
      • Zhang R.
      • Wang D.
      • Zhu W.
      The diversity of the fecal bacterial community and its relationship with the concentration of volatile fatty acids in the feces during subacute rumen acidosis in dairy cows.
      ). It has been suggested that elevated colonic butyrate levels may support a healthy gut environment because of the use of butyrate by the intestinal epithelium as an energy source (
      • Brouns F.
      • Kettlitz B.
      • Arrigoni E.
      Resistant starch and “the butyrate revolution”.
      ). Increased fecal concentration of butyrate, assuming absorption of butyrate has not changed largely, is an indication of increased butyrate production in the hindgut. Thus, more butyrate may have been available for hindgut epithelia, and could have provided a supportive effect on epithelial function in the hindgut.
      Based on the data from the present study and that of
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      , we determined the relation between fecal pH and the amount of corn starch infused into the abomasum (i.e., on individual cow level, not treatment means). The Pearson correlation between fecal pH and the amount of corn starch abomasally infused was negative (P < 0.001, r = −0.84), where fecal pH = 6.74 (±0.063) − 0.59 (±0.047) × kg of corn starch abomasally infused. Assuming that a fecal pH of 6.00 (i.e., lower threshold reported by
      • Plaizier J.C.
      • Danesh Mesgaran M.
      • Derakhshani H.
      • Golder H.
      • Khafipour E.
      • Kleen J.L.
      • Lean I.
      • Loor J.
      • Penner G.
      • Zebeli Q.
      Review: Enhancing gastrointestinal health in dairy cows.
      ) indicates hindgut acidosis, hindgut acidosis would be induced as soon as 1.25 kg of extra corn starch/d enters the abomasum, irrespective of the source of corn starch. Based on the variation in ruminal starch digestion in ruminants (
      • Mills J. A.N.
      • France J.
      • Dijkstra J.
      A review of starch digestion in the lactating dairy cow and proposals for a mechanistic model: 1. Dietary starch characterisation and ruminal starch digestion.
      ;
      • Allen M.S.
      Effects of diet on short-term regulation of feed intake by lactating dairy cattle.
      ;
      • Moharrery A.
      • Larsen M.
      • Weisbjerg M.R.
      Starch digestion in the rumen, small intestine, and hind gut of dairy cows - A meta-analysis.
      ), we assume that 60% (i.e., common for corn) to 80% of starch intake is fermented in the rumen. Considering this alongside the above-described equation, daily total starch intake would have to reach at least 3.1 to 6.3 kg/d before hindgut acidosis may possibly be induced. The average DMI of the present study and that of
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      was 16.7 kg/d. Hence, the intake of 3.1 and 6.3 kg of starch/d would be achieved at a dietary starch content of 186 to 371 g/kg of DM, respectively. These levels, particularly the lower dietary starch content, is easily achieved in practice under many farming conditions.

      Effects of Abomasal Corn Starch Infusion and Associated Hindgut Acidosis

      The model used to induce hindgut acidosis in the present study differs from some studies where hindgut acidosis was induced.
      • Li S.
      • Khafipour E.
      • Krause D.O.
      • Kroeker A.
      • Rodriguez-Lecompte J.C.
      • Gozho G.N.
      • Plaizier J.C.
      Effects of subacute ruminal acidosis challenges on fermentation and endotoxins in the rumen and hindgut of dairy cows.
      and
      • Plaizier J.C.
      • Li S.
      • Danscher A.M.
      • Derakshani H.
      • Andersen P.H.
      • Khafipour E.
      Changes in microbiota in rumen digesta and feces due to a grain-based subacute ruminal acidosis (SARA) challenge.
      induced hindgut acidosis via a grain-based SARA challenge. In this model, feeding large amounts of grain is combined with poor rumen mat function associated with SARA conditions. This generally leads to larger amounts of fermentable substrates that bypass rumen fermentation. If these substrates are not digested in the small intestine, these will be fermented in the hindgut (
      • Plaizier J.C.
      • Danesh Mesgaran M.
      • Derakhshani H.
      • Golder H.
      • Khafipour E.
      • Kleen J.L.
      • Lean I.
      • Loor J.
      • Penner G.
      • Zebeli Q.
      Review: Enhancing gastrointestinal health in dairy cows.
      ). In contrast, the abomasal infusion model used in the present study induced hindgut acidosis but did not affect ruminal conditions. Hence, differences in the models applied to induce hindgut acidosis may lead to differences in results (i.e., the effects of hindgut acidosis) across studies.

      Nutrient Intake and Digestibility

      We observed a depressed voluntary TMR DMI with abomasal infusion of 3.0 kg of corn starch/d, but not with the abomasal infusion of 1.5 kg of corn starch/d. This is in agreement with
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      , who did not observe a decreased voluntary DMI when 3.0 kg of ground corn/d (equal to ∼1.5 kg of starch/d) was infused in the abomasum. On the contrary though, both
      • Reynolds C.K.
      • Humphries D.J.
      • Cammell S.B.
      • Benson J.
      • Sutton J.D.
      • Beever D.E.
      Effects of abomasal wheat starch infusion on splanchnic metabolism and energy balance of lactating dairy cows.
      and
      • Knowlton K.F.
      • Dawson T.E.
      • Glenn B.P.
      • Huntington G.B.
      • Erdman R.A.
      Glucose metabolism and milk yield of cows infused abomasally or ruminally with starch.
      observed a decreased voluntary DMI when 1.2 kg of wheat starch/d and 1.5 kg of hydrolyzed starch/d was infused, respectively.
      In the present study, ATTD of nutrients was determined by including TiO2 in the concentrate as an external marker and collecting fecal samples twice daily during the infusion period.
      • Morris D.L.
      • Rebelo L.R.
      • Dieter P.A.
      • Lee C.
      Validating intrinsic markers and optimizing spot sampling frequency to estimate fecal outputs.
      stated that sampling feces twice daily (as we have done) might result in different fecal marker and nutrient concentrations compared with sampling feces 12 times daily. However, cows were fed only once daily in the study of
      • Morris D.L.
      • Rebelo L.R.
      • Dieter P.A.
      • Lee C.
      Validating intrinsic markers and optimizing spot sampling frequency to estimate fecal outputs.
      , whereas the cows in the present study were fed according to a steady-state pattern during the period of sampling. Steady-state feeding will reduce diurnal variation in fecal marker and nutrient concentrations (
      • Owens F.N.
      • Hanson C.F.
      External and internal markers for appraising site and extent of digestion in ruminants.
      ). Furthermore,
      • Titgemeyer E.C.
      • Armendariz C.K.
      • Bindel D.J.
      • Greenwood R.H.
      • Löest C.A.
      Evaluation of titanium dioxide as a digestibility marker for cattle.
      showed that TiO2 recoveries that did not differ from 100% when cattle were adapted to marker administration for 17 to 21 d before measurement of nutrient ATTD. In the present study, cows were adapted to marker administration for at least 21 d. Hence, we consider the duration the duration of marker administration as well as the frequency of fecal sampling (in combination with steady-state feeding) sufficient to draw conclusions from the results concerning ATTD of nutrients.
      The decreased ATTD of most nutrients upon abomasal infusion of corn starch is generally in agreement with
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      but contrary to
      • Knowlton K.F.
      • Dawson T.E.
      • Glenn B.P.
      • Huntington G.B.
      • Erdman R.A.
      Glucose metabolism and milk yield of cows infused abomasally or ruminally with starch.
      and
      • Reynolds C.K.
      • Cammell S.B.
      • Humphries D.J.
      • Beever D.E.
      • Sutton J.D.
      • Newbold J.R.
      Effects of postrumen starch infusion on milk production and energy metabolism in dairy cows.
      .
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      proposed that overall fermentation and absorption are lower for postruminally infused nutrients than for nutrients entering the rumen. By avoiding ruminal fermentation, it is likely that starch, and subsequently, DM, OM, and GE (i.e., because starch contributes to these fractions), were digested to a lesser extent. In contrast, nutrients that originate entirely from the TMR fed (i.e., fat and NDF) undergo rumen fermentation, explaining the unaffected ATTD of these nutrients upon abomasal infusion of corn starch. Decreased ATTD of CP was likely the result of increased microbial protein synthesis upon fermentation of the infused corn starch, and was also observed by
      • Gressley T.F.
      • Armentano L.E.
      Effects of low rumen degradable protein or abomasal fructan infusion on diet digestibility and urinary nitrogen excretion in lactating dairy cows.
      and
      • Gressley T.F.
      • Hall M.B.
      • Armentano L.E.
      Ruminant nutrition symposium: Productivity, digestion, and health responses to hindgut acidosis in ruminants.
      in response to increased starch supply. If this microbial protein remains undigested in the hindgut, it is excreted as microbial N and thereby decreases ATTD of N (
      • Owens F.N.
      • Zinn R.A.
      • Kim Y.K.
      Limits to starch digestion in the ruminant small intestine.
      ). As N contributes to the DM, OM, and GE fraction, decreased ATTD of N likely also results in decreased ATTD of DM, OM, and GE.

      Energy Partitioning

      The decreased ATTD of GE, as discussed above, in combination with increased GEI, resulted in increased energy excreted via manure upon abomasal infusion of corn starch. When corn starch is digested or fermented in the hindgut, increased MEI is expected from greater glucose and VFA absorption (
      • Reynolds C.K.
      Production and metabolic effects of site of starch digestion in dairy cattle.
      ). In the present study, however, both MEI and blood glucose concentration were unaffected by abomasal infusion of corn starch, contrary to
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      but in line with
      • Knowlton K.F.
      • Dawson T.E.
      • Glenn B.P.
      • Huntington G.B.
      • Erdman R.A.
      Glucose metabolism and milk yield of cows infused abomasally or ruminally with starch.
      . The lack of effect on blood glucose concentration, despite the large quantities of starch that were infused, can be explained by greater utilization of arterial glucose by tissues of the portal-drained viscera relative to absorbed glucose (
      • Reynolds C.K.
      • Humphries D.J.
      • Cammell S.B.
      • Benson J.
      • Sutton J.D.
      • Beever D.E.
      Effects of abomasal wheat starch infusion on splanchnic metabolism and energy balance of lactating dairy cows.
      ), oxidation of luminal glucose by enterocytes (
      • Knowlton K.F.
      • Dawson T.E.
      • Glenn B.P.
      • Huntington G.B.
      • Erdman R.A.
      Glucose metabolism and milk yield of cows infused abomasally or ruminally with starch.
      ), decreased hepatic gluconeogenesis in response to the increased glucose absorption (
      • Clark J.H.
      • Spires H.R.
      • Derrig R.G.
      • Bennink M.R.
      Milk production, nitrogen utilization and glucose synthesis in lactating cows infused postruminally with sodium caseinate and glucose.
      ), or a combination of these mechanisms.
      Milk yield is largely regulated through the osmotic properties of lactose, where synthesis of lactose requires the transport of glucose from blood into the cytosol of mammary epithelial cells (
      • Cant J.P.
      • Trout D.R.
      • Qiao F.
      • Purdie N.G.
      Milk synthetic response of the bovine mammary gland to an increase in the local concentration of arterial glucose.
      ). The unaffected milk yield, milk lactose concentration, and lactose yield in response to abomasal corn starch infusion, are therefore in line with the increased energy excretion in manure and the unaffected blood glucose concentrations. Milk fat concentration tended to decrease when corn starch was infused into the abomasum, which is contrary to
      • Reynolds C.K.
      • Cammell S.B.
      • Humphries D.J.
      • Beever D.E.
      • Sutton J.D.
      • Newbold J.R.
      Effects of postrumen starch infusion on milk production and energy metabolism in dairy cows.
      ,
      • Rius A.G.
      • Appuhamy J. A. D. R.N.
      • Cyriac J.
      • Kirovski D.
      • Becvar O.
      • Escobar J.
      • McGilliard M.L.
      • Bequette B.J.
      • Akers R.M.
      • Hanigan M.D.
      Regulation of protein synthesis in mammary glands of lactating dairy cows by starch and amino acids.
      , and
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      , but similar to
      • Knowlton K.F.
      • Dawson T.E.
      • Glenn B.P.
      • Huntington G.B.
      • Erdman R.A.
      Glucose metabolism and milk yield of cows infused abomasally or ruminally with starch.
      . When elevated in circulation, glucose promotes insulin secretion (
      • Bauman D.E.
      • Griinari J.M.
      Regulation and nutritional manipulation of milk fat: Low-fat milk syndrome.
      ). Insulin stimulates the uptake of lipogenic precursors into adipose tissue and decreases lipolysis, resulting in a reduction of circulating fatty acids for milk fat synthesis (
      • Bauman D.E.
      • Griinari J.M.
      Regulation and nutritional manipulation of milk fat: Low-fat milk syndrome.
      ). Although energy retained as fat, blood glucose concentration, and concentrations of insulin, NEFA, and BHB in plasma were not affected by abomasal infusion of corn starch in the present study, the tendency for a decreased milk fat concentration suggests a reduced availability of long-chain fatty acids for milk lipid synthesis during infusion of corn starch.

      Nitrogen Partitioning

      The transfer efficiency of feed N into milk N may increase when glucogenic energy supply to the animal is increased. This is because of reduced catabolism of AA for gluconeogenesis, leaving more available for milk protein synthesis. Contrary to the findings of
      • Rius A.G.
      • Appuhamy J. A. D. R.N.
      • Cyriac J.
      • Kirovski D.
      • Becvar O.
      • Escobar J.
      • McGilliard M.L.
      • Bequette B.J.
      • Akers R.M.
      • Hanigan M.D.
      Regulation of protein synthesis in mammary glands of lactating dairy cows by starch and amino acids.
      and
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      , milk protein content or yield did not increase in response to corn starch infusion in the present study. These results are, however, in line with the lack of effect of starch infusion on blood glucose or plasma insulin concentrations and in agreement with
      • Knowlton K.F.
      • Dawson T.E.
      • Glenn B.P.
      • Huntington G.B.
      • Erdman R.A.
      Glucose metabolism and milk yield of cows infused abomasally or ruminally with starch.
      and
      • Reynolds C.K.
      • Cammell S.B.
      • Humphries D.J.
      • Beever D.E.
      • Sutton J.D.
      • Newbold J.R.
      Effects of postrumen starch infusion on milk production and energy metabolism in dairy cows.
      . The increased N efficiency observed with corn starch infusion was due to lower N intake as a result of a lower DMI of the TMR combined with unaffected milk N secretion.
      We observed a clear shift in N excretion from urinary N to fecal N upon abomasal infusion of corn starch, similar to
      • Reynolds C.K.
      • Cammell S.B.
      • Humphries D.J.
      • Beever D.E.
      • Sutton J.D.
      • Newbold J.R.
      Effects of postrumen starch infusion on milk production and energy metabolism in dairy cows.
      and
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      . This is likely related to increased microbial protein synthesis in the hindgut arising from fermentation of the infused corn starch in the large intestine. The N source for this microbial protein may have partly been derived from blood next to N sources in the digesta. Thus, the decreased urinary N is in line with the decreased blood urea N concentration and decreased milk urea concentration upon abomasal infusion of corn starch.

      Immune Response

      Similar to
      • van Gastelen S.
      • Dijkstra J.
      • Nichols K.
      • Bannink A.
      Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
      , the lack of effect of abomasally infused corn starch on the plasma concentration of acute phase proteins suggests that hindgut fermentation of corn starch did not result in a systemic inflammatory response detectable by these plasma markers. An alteration in the condition of intercellular tight junctions (e.g., structural damage in response to acidic conditions), can result in increased permeability of the intestine to large molecules. Damaged tight junctions become leaky, allowing pathogens and toxins to enter the bloodstream (
      • Kameda H.
      • Abei T.
      • Nasrallah S.
      • Iber F.L.
      Functional and histological injury to intestinal mucosa produced by hypertonicity.
      ). Intestinal permeability measured by the appearance of Co (of the indigestible marker Co-EDTA) in blood appeared to be unaffected, suggesting that abomasal infusion of up to 3.0 kg of corn starch/d did not result in a leaky gut.

      Establishment of Metabolic Acidosis

      Blood pH was not affected by abomasal infusion of BHB and was within the normal physiological range for BHB and non-BHB treatments (i.e., between 7.38 and 7.43 pH units;
      • Indrova E.
      • Dolezel R.
      • Novakova-Mala J.
      • Pechova A.
      • Zavadilova M.
      • Cech S.
      Impact of acute metabolic acidosis on the acid-base balance in follicular fluid and blood in dairy cattle.
      ). This suggests that daily infusion of 8 mol BHB did not result in metabolic acidosis. However, in clinically healthy cows, acid-base disturbances are compensated by various regulatory mechanisms (
      • Gärtner T.
      • Zoche-Golob V.
      • Redlberger S.
      • Reinhold P.
      • Donat K.
      Acid-base assessment of post-parturient German Holstein dairy cows from jugular venous blood and urine: A comparison of the strong ion approach and traditional blood gas analysis.
      ). Respiratory compensation is commonly characterized during suspected acid-base disturbances by a reduced partial pressure of CO2 (
      • Gant R.G.
      • Sanchez W.
      • Kincaid R.L.
      Effect of anionic salts on selenium metabolism in nonlactating, pregnant dairy cows.
      ). Respiratory compensation, however, is of minor importance in dairy cows because of their relatively small lung capacity (
      • Enemark J.M.
      • Jørgensen R.J.
      • Enemark P.S.
      Rumen acidosis with special emphasis on diagnostic aspects of subclinical rumen acidosis: A review.
      ), so it follows that pCO2 was maintained within normal range (i.e., between 35 and 45 mmHg;
      • Constable P.D.
      Clinical assessment of acid-base status. Strong ion difference theory.
      ) during abomasal infusion of BHB in the present study. When a respiratory response is not present, total CO2, HCO3, and base excess can be used as indicators for an acid-base disturbance (
      • Constable P.D.
      Clinical assessment of acid-base status. Strong ion difference theory.
      ). Although these 3 variables decreased upon abomasal infusion of BHB, they remained within their normal physiological ranges (e.g., between −0.5 and 4.5 mmol/L for base excess;
      • Indrova E.
      • Dolezel R.
      • Novakova-Mala J.
      • Pechova A.
      • Zavadilova M.
      • Cech S.
      Impact of acute metabolic acidosis on the acid-base balance in follicular fluid and blood in dairy cattle.
      ). These results, combined with a tendency for a decreased urinary pH, suggest that a compensated metabolic acidosis was induced via abomasal BHB infusion. This is characterized by an acid-base disturbance but un unaffected blood pH (
      • Vagnoni D.B.
      • Oetzel G.R.
      Effects of dietary cation-anion difference on the acid-base status of dry cows.
      ;
      • Zimpel R.
      • Poindexter M.B.
      • Vieira-Neto A.
      • Block E.
      • Nelson C.D.
      • Staples C.R.
      • Thatcher W.W.
      • Santos J. E.P.
      Effect of dietary cation-anion difference on acid-base status and dry matter intake in dry pregnant cows.
      ).
      Concentration greater than 1.2 mmol BHB/L of blood (
      • Enjalbert F.
      • Nicot M.C.
      • Bayourthe C.
      • Moncoulon R.
      Ketone bodies in milk and blood of dairy cows: Relationship between concentrations and utilization for detection of subclinical ketosis.
      ) and 0.2 mmol BHB/L of milk (
      • Geishauser T.
      • Leslie K.
      • Tenhag J.
      • Bashiri A.
      Evaluation of eight cow-side ketone tests in milk for detection of subclinical ketosis in dairy cows.
      ) have been proposed as critical threshold values indicating subclinical ketosis. The observed concentrations of BHB in milk and plasma were well below these thresholds, irrespective of the infusion treatment. This suggests that 8 mol BHB/d infused into the abomasum was not sufficient to induce a state resembling subclinical ketosis or ketoacidosis (i.e., accumulation of ketone bodies causing metabolic acidosis;
      • Ewaschuk J.B.
      • Zello G.A.
      • Naylor J.M.
      • Brocks D.R.
      Metabolic acidosis: Separation methods and biological relevance of organic acids and lactic acid enantiomers.
      ). We based our BHB dose on the assumption that, on average, 28 kg of body fat would be mobilized over a period of 8 wk (i.e., 500 g of body fat/d;
      • Tamminga S.
      • Luteijn P.A.
      • Meijer R. G.M.
      Changes in composition and energy content of liveweight loss in dairy cows with time after parturition.
      ;
      • van Knegsel A. T.M.
      • van den Brand H.
      • Dijkstra J.
      • van Straalen W.M.
      • Heetkamp M. J.W.
      • Tamminga S.
      • Kemp B.
      Dietary energy source in dairy cows in early lactation: Energy partitioning and milk composition.
      ). Mobilization of 500 g of body fat/d equals some 1.6 mol C18 fatty acids/d, considering the molar weight of triglycerides (i.e., 946 g if consisting of glycerol and 3 fatty acids with 18C). We assumed that 100% of these C18 fatty acids would be converted into BHB as a C4 fatty acid (i.e., 1 mol C18 fatty acids delivering 4 mol BHB), and that the capacity of the liver to oxidize NEFA is exceeded and all NEFA are converted into ketone bodies. Hence, 6.3 mol BHB/d would be produced by the liver on average for the first 8 wk of lactation. In the current study, we realized an infused supply of 7.7 to 8.0 mol BHB/d, which is about 25% more than this estimated level of endogenous BHB production. This dose was apparently not sufficient to induce a subclinical ketosis or ketoacidosis.
      • Herrick K.J.
      • Hippen A.R.
      • Kalscheur K.F.
      • Schingoethe D.J.
      • Ranathunga S.D.
      • Anderson J.L.
      • Moreland S.C.
      • van Eys J.E.
      Infusion of butyrate affects plasma glucose, butyrate, and β-hydroxybutyrate but not plasma insulin in lactating dairy cows.
      stated that ketosis-indicating threshold concentrations of BHB in milk or plasma are based on elevated levels of endogenously produced ketones arising from liver dysfunction and excessive body fat mobilization. There was no evidence in the current study that these conditions were occurring upon BHB infusion (i.e., increased energy retention in body fat, no effect on plasma NEFA concentrations or TMR DMI). Therefore, it is logical that exogenously supplied BHB produced a metabolic response characterized as a compensated metabolic acidosis rather than a classical ketoacidosis.

      Effects of Abomasal BHB Infusion and Associated Compensated Metabolic Acidosis

      Nutrient Intake and Digestibility

      Ketosis is generally associated with reduced feed intake (e.g.,
      • Benedet A.
      • Manuelian C.L.
      • Zidi A.
      • Penasa M.
      • De Marchi M.
      Invited review: β-hydroxybutyrate concentration in blood and milk and its associations with cow performance.
      ). This is likely related to the stimulation of oxidative metabolism and signals of energy status and satiety in the brain by ketones (
      • Laeger T.
      • Metges C.C.
      • Kuhla B.
      Role of β-hydroxybutyric acid in the central regulation of energy balance.
      ) as well as to changes in acid-base balance (
      • Zimpel R.
      • Poindexter M.B.
      • Vieira-Neto A.
      • Block E.
      • Nelson C.D.
      • Staples C.R.
      • Thatcher W.W.
      • Santos J. E.P.
      Effect of dietary cation-anion difference on acid-base status and dry matter intake in dry pregnant cows.
      ). Abomasal infusion of BHB did not affect voluntary DMI of the TMR in the current study, likely because only a compensated metabolic acidosis was observed and because plasma BHB concentration, although elevated, may not have reached critical levels affecting satiety. In agreement,
      • Zarrin M.
      • De Matteis L.
      • Vernay M. C. M.B.
      • Wellnitz O.
      • van Dorland H.A.
      • Bruckmaier R.M.
      Long-term elevation of β-hydroxybutyrate in dairy cows through infusion: Effects on feed intake, milk production, and metabolism.
      ,
      • Zarrin M.
      • Wellnitz O.
      • van Dorland H.A.
      • Gross J.J.
      • Bruckmaier R.M.
      Hyperketonemia during lipopolysaccharide-induced mastitis affects systemic and local intramammary metabolism in dairy cows.
      ) found no reduction in DMI of dairy cows when BHB was intravenously infused for 48 and 56 h, respectively. The decreased digestible NDF intake following abomasal infusion of BHB appears to be the result of a numerical decrease in TMR DMI and ATTD of NDF. The decrease in ATTD of CP and fat might be related to the increased intestinal permeability (suggested by the increased Co concentration in blood) upon abomasal infusion of BHB. An increased permeability may negatively affect hindgut fermentation and the stability of the intestinal microbiota. Taking this into account, it is unclear why the ATTD of starch increased with abomasal infusion of BHB when combined with 3.0 kg of corn starch/d, but not when combined with the lower infusion level of corn starch.

      Energy Partitioning

      Abomasal infusion of BHB decreased blood glucose and plasma insulin concentrations. This suggests that insulin did not play a role in the inhibiting effect of BHB on glucose concentrations, which is in agreement with studies in sheep (
      • Schlumbohm C.
      • Harmeyer J.
      Hypocalcemia reduces endogenous glucose production in hyperketonemic sheep.
      ) and dairy cows (
      • Zarrin M.
      • De Matteis L.
      • Vernay M. C. M.B.
      • Wellnitz O.
      • van Dorland H.A.
      • Bruckmaier R.M.
      Long-term elevation of β-hydroxybutyrate in dairy cows through infusion: Effects on feed intake, milk production, and metabolism.
      ). Alongside insulin, gluconeogenesis is also hormonally regulated by glucagon (
      • Zarrin M.
      • De Matteis L.
      • Vernay M. C. M.B.
      • Wellnitz O.
      • van Dorland H.A.
      • Bruckmaier R.M.
      Long-term elevation of β-hydroxybutyrate in dairy cows through infusion: Effects on feed intake, milk production, and metabolism.
      ). Glucagon increases plasma glucose concentration through stimulation of gluconeogenesis and glycogenolysis (
      • Aronoff S.L.
      • Berkowitz K.
      • Shreiner B.
      • Want L.
      Glucose metabolism and regulation: Beyond insulin and glucagon.
      ). Ketone bodies and NEFA can suppress glucagon secretion (
      • Gerich J.E.
      • Charles M.A.
      • Grodsky G.M.
      Characterization of the effects of arginine and glucose on glucagon and insulin release from the perfused rat pancreas.
      ), and
      • Zarrin M.
      • De Matteis L.
      • Vernay M. C. M.B.
      • Wellnitz O.
      • van Dorland H.A.
      • Bruckmaier R.M.
      Long-term elevation of β-hydroxybutyrate in dairy cows through infusion: Effects on feed intake, milk production, and metabolism.
      reported decreased glucagon concentrations when BHB was intravenously infused in dairy cattle. Although not measured in the current study, glucagon secretion may have been depressed by the elevated plasma BHB concentration upon abomasal infusion of BHB, which may have resulted in less stimulation of gluconeogenesis or glycogenolysis and subsequently reduced blood glucose concentrations.
      The effects of elevated plasma BHB concentrations on milk yield in literature are controversial. Several studies observed a decrease of daily milk production between 1% and 18% (e.g.,
      • Duffield T.F.
      • Lissemore K.D.
      • McBride B.W.
      • Leslie K.E.
      Impact of hyperketonemia in early lactation dairy cows on health and production.
      ;
      • McArt J. A.A.
      • Nydam D.V.
      • Oetzel G.R.
      Epidemiology of subclinical ketosis in early lactation dairy cattle.
      ), whereas others reported an increase of daily milk yield between 5% and 11% (e.g.,
      • Vanholder T.
      • Papen J.
      • Bemers R.
      • Vertenten G.
      • Berge A. C.B.
      Risk factors for subclinical and clinical ketosis and association with production parameters in dairy cows in the Netherlands.
      ;
      • Ruoff J.
      • Borchardt S.
      • Heuwieser W.
      Short communication: Associations between blood glucose concentration, onset of hyperketonemia, and milk production in early lactation dairy cows.
      ; reviewed by
      • Benedet A.
      • Manuelian C.L.
      • Zidi A.
      • Penasa M.
      • De Marchi M.
      Invited review: β-hydroxybutyrate concentration in blood and milk and its associations with cow performance.
      ) in cows with an elevated plasma BHB concentration (>1.2 mmol/L). Ketone bodies can have a glucose-sparing effect in tissues where ketone bodies can be used as an energy source (
      • Zarrin M.
      • De Matteis L.
      • Vernay M. C. M.B.
      • Wellnitz O.
      • van Dorland H.A.
      • Bruckmaier R.M.
      Long-term elevation of β-hydroxybutyrate in dairy cows through infusion: Effects on feed intake, milk production, and metabolism.
      ), but glucose is required as a unique source for mammary lactose synthesis (
      • Zarrin M.
      • Wellnitz O.
      • van Dorland H.A.
      • Gross J.J.
      • Bruckmaier R.M.
      Hyperketonemia during lipopolysaccharide-induced mastitis affects systemic and local intramammary metabolism in dairy cows.
      ). Despite the decreased blood glucose concentrations observed in the present study, milk yield and lactose yield were not affected by abomasal BHB infusion. This is in agreement with
      • Zarrin M.
      • De Matteis L.
      • Vernay M. C. M.B.
      • Wellnitz O.
      • van Dorland H.A.
      • Bruckmaier R.M.
      Long-term elevation of β-hydroxybutyrate in dairy cows through infusion: Effects on feed intake, milk production, and metabolism.
      ; only milk yield reported) as well as
      • Herrick K.J.
      • Hippen A.R.
      • Kalscheur K.F.
      • Schingoethe D.J.
      • Ranathunga S.D.
      • Anderson J.L.
      • Moreland S.C.
      • van Eys J.E.
      Infusion of butyrate affects plasma glucose, butyrate, and β-hydroxybutyrate but not plasma insulin in lactating dairy cows.
      . As demonstrated by
      • Mebane D.
      • Madison L.L.
      The hypoglycemic effect of ketone bodies.
      and
      • Madison L.L.
      • Mebane D.
      • Unger H.R.
      • Lochner A.
      The hypoglycemic action of ketones. II. Evidence for a stimulatory feedback of ketones on the pancreatic beta cells.
      , if the increased plasma BHB concentration in response to BHB infusion inhibited nonmammary peripheral glucose utilization, sufficient glucose may have been available to support milk production in the current study.
      It was expected that abomasal infusion of BHB would result in an increased milk fat concentration, because BHB can be used as precursor for milk fat synthesis (
      • Bauman D.E.
      • Griinari J.M.
      Nutritional regulation of milk fat synthesis.
      ), and because increased milk fat concentrations are often observed in ketotic cows (e.g.,
      • Duffield T.F.
      • Lissemore K.D.
      • McBride B.W.
      • Leslie K.E.
      Impact of hyperketonemia in early lactation dairy cows on health and production.
      ;
      • Vanholder T.
      • Papen J.
      • Bemers R.
      • Vertenten G.
      • Berge A. C.B.
      Risk factors for subclinical and clinical ketosis and association with production parameters in dairy cows in the Netherlands.
      ;
      • Benedet A.
      • Manuelian C.L.
      • Zidi A.
      • Penasa M.
      • De Marchi M.
      Invited review: β-hydroxybutyrate concentration in blood and milk and its associations with cow performance.
      ). However, an increase in milk fat concentration was not observed in the present study, perhaps because the increase in circulating BHB concentration was below threshold levels characteristic of ketotic dairy cows.
      We observed an increase in energy retained as body fat in the present study upon abomasal infusion of BHB, suggesting that more body fat was accreted with BHB infusion. This may simply be due to the numerically increased GEI and decreased methane production upon BHB infusion, resulting in a significantly increased MEI. Despite increase MEI, milk energy output was unaffected by BHB infusion, leading to more energy being available for body accretion.

      Nitrogen Partitioning

      Abomasal BHB infusion had minimal effects on N partitioning. Interestingly, milk urea concentration decreased in response to BHB infusion. This was also observed by
      • Santschi D.E.
      • Lacroix R.
      • Durocher J.
      • Duplessis M.
      • Moore R.K.
      • Lefebvre D.M.
      Prevalence of elevated milk β-hydroxybutyrate concentrations in Holstein cows measured by Fourier-transform infrared analysis in dairy herd improvement milk samples and association with milk yield and components.
      , where milk urea N concentration was lower for ketotic cows compared with nonketotic cows, and this response was assigned to a lower protein intake for the cows with ketosis. In the present study, N intake decreased numerically and ATTD of N decreased significantly upon abomasal infusion of BHB. Hence, it is likely that less N was absorbed from the gastrointestinal tract, resulting in a smaller surplus (indicated as urea in milk or blood).

      Immune Response

      The increased molar proportion of fecal butyrate upon BHB infusion might be related to the observed increased intestinal permeability, where elevated colonic butyrate levels are thought to have the ability to support a healthy gut environment (
      • Brouns F.
      • Kettlitz B.
      • Arrigoni E.
      Resistant starch and “the butyrate revolution”.
      ). Assuming that starch fermentation in the hindgut results in increased levels of butyrate (e.g.,
      • Mao S.
      • Zhang R.
      • Wang D.
      • Zhu W.
      The diversity of the fecal bacterial community and its relationship with the concentration of volatile fatty acids in the feces during subacute rumen acidosis in dairy cows.
      ), BHB infusion may also have caused a shift from starch digestion in small intestines to starch fermentation in hindgut.
      Increased intestinal permeability (suggested by the increased Co concentration in blood) appears not to be related to the potential hyperosmolality that could have been caused by infusion of BHB as a ketone salt (
      • Kameda H.
      • Abei T.
      • Nasrallah S.
      • Iber F.L.
      Functional and histological injury to intestinal mucosa produced by hypertonicity.
      ). This is because the non-BHB treatments were formulated to contain the same amount of Na compared with the BHB treatments (Table 2). Although, increased permeability does appear to be attributable to BHB, it is unclear why the epithelium of the large intestine became more permeable or susceptible to damage upon BHB infusion.

      CONCLUSIONS

      Fecal pH decreased with each level of corn starch infused into the abomasum (pH of 6.49, 6.00, and 5.15 with 0.0, 1.5, and 3.0 kg of corn starch infused/d, respectively), suggesting hindgut acidosis was induced when corn starch was abomasally infused. Abomasal infusion of BHB resulted in a compensated metabolic acidosis, which was characterized by a disturbance of the acid-base status (i.e., decreased blood total CO2, HCO3, and base excess, and trend for decreased urinary pH), whereas blood pH remained within a physiologically normal range. Overall, in this short-term experiment, induced hindgut acidosis was not associated with negative effects on milk production and composition, energy balance, or inflammatory response, and was positively associated with milk N efficiency. Similarly, the induced compensated metabolic acidosis was not associated with negative effects on milk production and composition or inflammatory response, and was being positively associated with energy retained as body fat. Long-term studies are needed to confirm these short-term findings.

      ACKNOWLEDGMENTS

      Lianne Alderkamp, Nisola Ayanfe, and Christy de Wildt (students of Wageningen University & Research, the Netherlands), and the staff of the experimental facilities (Teus Bleyenberg, Wim van Ommeren, and Ries Verkerk) are gratefully acknowledged for their assistance during the implementation of the experiment, as are the laboratory staff of the Animal Nutrition Group. This research was conducted by Wageningen University & Research (Wageningen Livestock Research, Wageningen, the Netherlands), commissioned and funded by the Ministry of Agriculture, Nature and Food Quality (The Hague, the Netherlands) within the framework of Policy Support Research theme ‘Feed4Foodure' (BO-31.03-005-001; TKI-AF12039) and by the Vereniging Diervoederonderzoek Nederland (Rijswijk, the Netherlands). The authors have not stated any conflicts of interest.

      REFERENCES

        • Allen M.S.
        Effects of diet on short-term regulation of feed intake by lactating dairy cattle.
        J. Dairy Sci. 2000; 83 (10908065): 1598-1624
        • Amado L.
        • Berends H.
        • Leal L.N.
        • Wilms J.
        • van Laar H.
        • Gerrits W. J.J.
        • Martin-Tereso J.
        Effect of energy source in calf milk replacer on performance, digestibility, and gut permeability in rearing calves.
        J. Dairy Sci. 2019; 102 (30852011): 3994-4001
        • Aronoff S.L.
        • Berkowitz K.
        • Shreiner B.
        • Want L.
        Glucose metabolism and regulation: Beyond insulin and glucagon.
        Diabetes Spectr. 2004; 17: 183-190
        • Aschenbach J.R.
        • Bilk S.
        • Tadesse G.
        • Stumpff F.
        • Gabel G.
        Bicarbonate-dependent and bicarbonate-independent mechanisms contribute to nondiffusive uptake of acetate in the ruminal epithelium of sheep.
        Am. J. Physiol. Gastro. Liver Physiol. 2009; 296: G1098-G1107
        • Bauman D.E.
        • Griinari J.M.
        Regulation and nutritional manipulation of milk fat: Low-fat milk syndrome.
        Livest. Prod. Sci. 2001; 70: 15-29
        • Bauman D.E.
        • Griinari J.M.
        Nutritional regulation of milk fat synthesis.
        Annu. Rev. Nutr. 2003; 23 (12626693): 203-227
        • Bell A.W.
        Regulation of organic nutrient metabolism during transition from late pregnancy to early lactation.
        J. Anim. Sci. 1995; 73 (8582872): 2804-2819
        • Benedet A.
        • Manuelian C.L.
        • Zidi A.
        • Penasa M.
        • De Marchi M.
        Invited review: β-hydroxybutyrate concentration in blood and milk and its associations with cow performance.
        Animal. 2019; 13 (30854998): 1676-1689
        • Brouns F.
        • Kettlitz B.
        • Arrigoni E.
        Resistant starch and “the butyrate revolution”.
        Trends Food Sci. Technol. 2002; 13: 251-261
        • Brown M.S.
        • Krehbiel C.R.
        • Galyean M.L.
        • Remmenga M.D.
        • Peters J.P.
        • Hibbard B.
        • Robinson J.
        • Moseley W.M.
        Evaluation of models of acute and subacute acidosis on dry matter intake, ruminal fermentation, blood chemistry, and endocrine profiles of beef steers.
        J. Anim. Sci. 2000; 78 (11132830): 3155-3168
        • Cant J.P.
        • Trout D.R.
        • Qiao F.
        • Purdie N.G.
        Milk synthetic response of the bovine mammary gland to an increase in the local concentration of arterial glucose.
        J. Dairy Sci. 2002; 85 (11949851): 494-503
        • Clark J.H.
        • Spires H.R.
        • Derrig R.G.
        • Bennink M.R.
        Milk production, nitrogen utilization and glucose synthesis in lactating cows infused postruminally with sodium caseinate and glucose.
        J. Nutr. 1977; 107 (845699): 631-644
        • Constable P.D.
        Clinical assessment of acid-base status. Strong ion difference theory.
        Vet. Clin. North Am. Food Anim. Pract. 1999; 15 (10573806): 447-471
        • Corey H.E.
        Stewart and beyond: New models of acid-base balance.
        Kidney Int. 2003; 64 (12911526): 777-787
        • CVB
        Chemische Samenstellingen en Nutritionele Waarden van Voedermiddelen (in Dutch).
        CVB (Centraal Veevoederbureau), 2018
        • Danscher A.M.
        • Li S.
        • Andersen P.H.
        • Khafipour E.
        • Kristensen N.B.
        • Plaizier J.C.
        Indicators of induced subacute ruminal acidosis (SARA) in Danish Holstein cows.
        Acta Vet. Scand. 2015; 57: 1-14
        • Duffield T.
        Subclinical ketosis in lactating dairy cattle.
        Vet. Clin. North Am. Food Anim. Pract. 2000; 16 (11022338): 231-253
        • Duffield T.F.
        • Lissemore K.D.
        • McBride B.W.
        • Leslie K.E.
        Impact of hyperketonemia in early lactation dairy cows on health and production.
        J. Dairy Sci. 2009; 92 (19164667): 571-580
        • Emery R.S.
        • Brown L.D.
        • Bell J.W.
        Correlation of milk fat with dietary and metabolic factors in cows fed restricted-roughage rations supplemented with magnesium oxide or sodium bicarbonate.
        J. Dairy Sci. 1965; 48: 1647-1651
        • Emery R.S.
        • Burg N.
        • Brown L.D.
        • Blank G.N.
        Detection, occurrence, and prophylactic treatment of borderline ketosis with propylene glycol feeding.
        J. Dairy Sci. 1964; 47: 1074-1079
        • Enemark J.M.
        • Jørgensen R.J.
        • Enemark P.S.
        Rumen acidosis with special emphasis on diagnostic aspects of subclinical rumen acidosis: A review.
        Vet. ir Zootech. 2002; 20: 16-29
        • Enemark J. M.D.
        The monitoring, prevention and treatment of sub-acute ruminal acidosis (SARA): A review.
        Vet. J. 2008; 176 (18343172): 32-43
        • Enjalbert F.
        • Nicot M.C.
        • Bayourthe C.
        • Moncoulon R.
        Ketone bodies in milk and blood of dairy cows: Relationship between concentrations and utilization for detection of subclinical ketosis.
        J. Dairy Sci. 2001; 84 (11286410): 583-589
        • Ewaschuk J.B.
        • Zello G.A.
        • Naylor J.M.
        • Brocks D.R.
        Metabolic acidosis: Separation methods and biological relevance of organic acids and lactic acid enantiomers.
        J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2002; 781 (12450652): 39-56
        • Gant R.G.
        • Sanchez W.
        • Kincaid R.L.
        Effect of anionic salts on selenium metabolism in nonlactating, pregnant dairy cows.
        J. Dairy Sci. 1998; 81 (9684171): 1637-1642
        • Gärtner T.
        • Zoche-Golob V.
        • Redlberger S.
        • Reinhold P.
        • Donat K.
        Acid-base assessment of post-parturient German Holstein dairy cows from jugular venous blood and urine: A comparison of the strong ion approach and traditional blood gas analysis.
        PLoS One. 2019; 14 (30650142)e0210948
        • Geishauser T.
        • Leslie K.
        • Tenhag J.
        • Bashiri A.
        Evaluation of eight cow-side ketone tests in milk for detection of subclinical ketosis in dairy cows.
        J. Dairy Sci. 2000; 83 (10714863): 296-299
        • Gerich J.E.
        • Charles M.A.
        • Grodsky G.M.
        Characterization of the effects of arginine and glucose on glucagon and insulin release from the perfused rat pancreas.
        J. Clin. Invest. 1974; 54 (4430717): 833-841
        • Gerrits W. J.J.
        • van den Borne J. J. G.C.
        • Labussière E.
        Deriving heat production from gaseous exchange: Validity of the approach.
        in: Gerrits W. J.J. Labussière E. Indirect Calorimetry. Techniques, Computations and Applications. Wageningen Academic Publishers, 2015: 19-34
        • Gressley T.F.
        • Armentano L.E.
        Effects of low rumen degradable protein or abomasal fructan infusion on diet digestibility and urinary nitrogen excretion in lactating dairy cows.
        J. Dairy Sci. 2007; 90 (17297109): 1340-1353
        • Gressley T.F.
        • Hall M.B.
        • Armentano L.E.
        Ruminant nutrition symposium: Productivity, digestion, and health responses to hindgut acidosis in ruminants.
        J. Anim. Sci. 2011; 89 (21415422): 1120-1130
        • Heetkamp M. J.W.
        • Alferink S. J.J.
        • Zandstra T.
        • Hendriks P.
        • van den Brand H.
        • Gerrits W. J.J.
        Design of climate respiration chambers, adjustable to the metabolic mass of subjects.
        in: Gerrits W. J.J. Labussière E. Indirect Calorimetry. Techniques, Computations and Applications. Wageningen Academic Publishers, 2015: 35-56
        • Herrick K.J.
        • Hippen A.R.
        • Kalscheur K.F.
        • Schingoethe D.J.
        • Ranathunga S.D.
        • Anderson J.L.
        • Moreland S.C.
        • van Eys J.E.
        Infusion of butyrate affects plasma glucose, butyrate, and β-hydroxybutyrate but not plasma insulin in lactating dairy cows.
        J. Dairy Sci. 2018; 101 (29409601): 3524-3536
        • Hood V.
        • Tannen R.
        Maintenance of acid-base homeostasis during ketoacidosis and lactic acidosis.
        Diabetes Rev. (Alex.). 1994; 2: 177-194
        • Indrova E.
        • Dolezel R.
        • Novakova-Mala J.
        • Pechova A.
        • Zavadilova M.
        • Cech S.
        Impact of acute metabolic acidosis on the acid-base balance in follicular fluid and blood in dairy cattle.
        Theriogenology. 2017; 89 (28043369): 41-46
        • International Organization for Standardization
        ISO 9831:1998. Animal feedstuffs, animal products, and feces or urine–Determination of gross caloric values–Bomb calorimeter method.
        International Organization for Standardization, 1998
        • International Organization for Standardization
        ISO 9622:2013. Milk and liquid milk products. Guidelines for the application of mid-infrared spectrometry.
        International Organization for Standardization, 2013
        • Ireland-Perry R.L.
        • Stallings C.C.
        Fecal consistency as related to dietary composition in lactating Holstein cows.
        J. Dairy Sci. 1993; 76 (8387550): 1074-1082
        • Kameda H.
        • Abei T.
        • Nasrallah S.
        • Iber F.L.
        Functional and histological injury to intestinal mucosa produced by hypertonicity.
        Am. J. Physiol. 1968; 214 (5647186): 1090-1095
        • Knowlton K.F.
        • Dawson T.E.
        • Glenn B.P.
        • Huntington G.B.
        • Erdman R.A.
        Glucose metabolism and milk yield of cows infused abomasally or ruminally with starch.
        J. Dairy Sci. 1998; 81 (9891270): 3248-3258
        • Laeger T.
        • Metges C.C.
        • Kuhla B.
        Role of β-hydroxybutyric acid in the central regulation of energy balance.
        Appetite. 2010; 54 (20416348): 450-455
        • Laffel L.
        Ketone bodies: A review of physiology pathophysiology and application of monitoring diabetes.
        Diabetes Metab. Res. Rev. 1999; 15 (10634967): 412-426
        • Li S.
        • Khafipour E.
        • Krause D.O.
        • Kroeker A.
        • Rodriguez-Lecompte J.C.
        • Gozho G.N.
        • Plaizier J.C.
        Effects of subacute ruminal acidosis challenges on fermentation and endotoxins in the rumen and hindgut of dairy cows.
        J. Dairy Sci. 2012; 95 (22192209): 294-303
        • Lv S.
        • Li D.
        • Xing J.
        • Ma Y.
        • Huang S.
        • Li J.
        Effects of extrusion of corn on growth performance, nutrient digestibility and short-chain fatty acid profiles in the hindgut of weaned piglets.
        Arch. Anim. Nutr. 2006; 60 (16649579): 170-179
        • Madison L.L.
        • Mebane D.
        • Unger H.R.
        • Lochner A.
        The hypoglycemic action of ketones. II. Evidence for a stimulatory feedback of ketones on the pancreatic beta cells.
        J. Clin. Invest. 1964; 43 (14135491): 408-415
        • Mao S.
        • Zhang R.
        • Wang D.
        • Zhu W.
        The diversity of the fecal bacterial community and its relationship with the concentration of volatile fatty acids in the feces during subacute rumen acidosis in dairy cows.
        BMC Vet. Res. 2012; 8 (23217205): 237
        • Matthé A.
        • Lebzien P.
        • Hric I.
        • Flachowsky G.
        • Sommer A.
        Effect of starch application into the proximal duodenum of ruminants on starch digestibility in the small and total intestine.
        Arch. Anim. Nutr. 2001; 55 (12357594): 351-369
        • McArt J. A.A.
        • Nydam D.V.
        • Oetzel G.R.
        Epidemiology of subclinical ketosis in early lactation dairy cattle.
        J. Dairy Sci. 2012; 95 (22916909): 5056-5066
        • Mebane D.
        • Madison L.L.
        The hypoglycemic effect of ketone bodies.
        J. Clin. Invest. 1962; 411383
        • Mesgaran S.D.
        • Kühla B.
        • Baumont R.
        • Cantalapiedra-Hijar G.
        • Noziére P.
        • Lund P.
        • Humphries D.
        • Dijkstra J.
        Nutrient digestibility and balance studies.
        in: Mesgaran S.D. Baumont R. Munksgaard L. Humphries D. Kennedy E. Dijkstra J. Dewhurst R. Ferguson H. Terré M. Kühla B. Methods in Cattle Physiology and Behaviour - Recommendations from the SmartCow Consortium. Publisso, 2020
        • Mills J. A.N.
        • France J.
        • Dijkstra J.
        A review of starch digestion in the lactating dairy cow and proposals for a mechanistic model: 1. Dietary starch characterisation and ruminal starch digestion.
        J. Anim. Feed Sci. 1999; 8: 291-340
        • Moharrery A.
        • Larsen M.
        • Weisbjerg M.R.
        Starch digestion in the rumen, small intestine, and hind gut of dairy cows - A meta-analysis.
        Anim. Feed Sci. Technol. 2014; 192: 1-14
        • Morris D.L.
        • Rebelo L.R.
        • Dieter P.A.
        • Lee C.
        Validating intrinsic markers and optimizing spot sampling frequency to estimate fecal outputs.
        J. Dairy Sci. 2018; 101 (30007807): 7980-7989
        • Mulligan F.J.
        • Doherty M.L.
        Production diseases of the transition cow.
        Vet. J. 2008; 176 (18342556): 3-9
        • Nichols K.
        • Bannink A.
        • Pacheco S.
        • van Valenberg H.J.
        • Dijkstra J.
        • van Laar H.
        Feed and nitrogen efficiency are affected differently but milk lactose production is stimulated equally when isoenergetic protein and fat is supplemented in lactating dairy cow diets.
        J. Dairy Sci. 2018; 101 (30007811): 7857-7870
        • Nichols K.
        • Dijkstra J.
        • van Laar H.
        • Pacheco S.
        • van Valenberg H.J.
        • Bannink A.
        Energy and nitrogen partitioning in dairy cows at low or high metabolizable protein levels is affected differently by postrumen glucogenic and lipogenic substrates.
        J. Dairy Sci. 2019; 102 (30527986): 395-412
        • Owens F.N.
        • Hanson C.F.
        External and internal markers for appraising site and extent of digestion in ruminants.
        J. Dairy Sci. 1992; 75 (1452861): 2605-2617
        • Owens F.N.
        • Zinn R.A.
        • Kim Y.K.
        Limits to starch digestion in the ruminant small intestine.
        J. Anim. Sci. 1986; 63 (3539905): 1634-1648
        • Plaizier J.C.
        • Danesh Mesgaran M.
        • Derakhshani H.
        • Golder H.
        • Khafipour E.
        • Kleen J.L.
        • Lean I.
        • Loor J.
        • Penner G.
        • Zebeli Q.
        Review: Enhancing gastrointestinal health in dairy cows.
        Animal. 2018; 12 (30139397): s399-s418
        • Plaizier J.C.
        • Li S.
        • Danscher A.M.
        • Derakshani H.
        • Andersen P.H.
        • Khafipour E.
        Changes in microbiota in rumen digesta and feces due to a grain-based subacute ruminal acidosis (SARA) challenge.
        Microb. Ecol. 2017; 74 (28175972): 485-495
        • Reynolds C.K.
        Production and metabolic effects of site of starch digestion in dairy cattle.
        Anim. Feed Sci. Technol. 2006; 130: 78-94
        • Reynolds C.K.
        • Cammell S.B.
        • Humphries D.J.
        • Beever D.E.
        • Sutton J.D.
        • Newbold J.R.
        Effects of postrumen starch infusion on milk production and energy metabolism in dairy cows.
        J. Dairy Sci. 2001; 84 (11699457): 2250-2259
        • Reynolds C.K.
        • Humphries D.J.
        • Cammell S.B.
        • Benson J.
        • Sutton J.D.
        • Beever D.E.
        Effects of abomasal wheat starch infusion on splanchnic metabolism and energy balance of lactating dairy cows.
        in: McCracken K.J. Unsworth E.F. Wylie A. R.G. Energy Metabolism of Farm Animals, Proceedings of the 14th Symposium on Energy Metabolism. CAB International, 1998: 39
        • Reynolds C.K.
        • Sutton J.D.
        • Beever D.E.
        Effects of feeding starch to dairy cattle on nutrient availability and production.
        in: Garnsworthy P.C. Wiseman J. Recent Advances in Animal Nutrition, Proc. 30th University of Nottingham Conference for Feed Manufacturers. University of Nottingham Press, 1997: 105-134
        • Rius A.G.
        • Appuhamy J. A. D. R.N.
        • Cyriac J.
        • Kirovski D.
        • Becvar O.
        • Escobar J.
        • McGilliard M.L.
        • Bequette B.J.
        • Akers R.M.
        • Hanigan M.D.
        Regulation of protein synthesis in mammary glands of lactating dairy cows by starch and amino acids.
        J. Dairy Sci. 2010; 93 (20630229): 3114-3127
        • Ruoff J.
        • Borchardt S.
        • Heuwieser W.
        Short communication: Associations between blood glucose concentration, onset of hyperketonemia, and milk production in early lactation dairy cows.
        J. Dairy Sci. 2017; 100 (28501411): 5462-5467
        • Santschi D.E.
        • Lacroix R.
        • Durocher J.
        • Duplessis M.
        • Moore R.K.
        • Lefebvre D.M.
        Prevalence of elevated milk β-hydroxybutyrate concentrations in Holstein cows measured by Fourier-transform infrared analysis in dairy herd improvement milk samples and association with milk yield and components.
        J. Dairy Sci. 2016; 99 (27544853): 9263-9270
        • Schlumbohm C.
        • Harmeyer J.
        Hypocalcemia reduces endogenous glucose production in hyperketonemic sheep.
        J. Dairy Sci. 2003; 86 (12836930): 1953-1962
        • Simensen E.
        • Halse K.
        • Gillund P.
        • Lutnœs B.
        Ketosis treatment and milk yield in dairy cows related to milk acetoacetate levels.
        Acta Vet. Scand. 1990; 31 (2099620): 433-440
        • Tamminga S.
        • Luteijn P.A.
        • Meijer R. G.M.
        Changes in composition and energy content of liveweight loss in dairy cows with time after parturition.
        Livest. Prod. Sci. 1997; 52: 31-38
        • Titgemeyer E.C.
        • Armendariz C.K.
        • Bindel D.J.
        • Greenwood R.H.
        • Löest C.A.
        Evaluation of titanium dioxide as a digestibility marker for cattle.
        J. Anim. Sci. 2001; 79 (11325180): 1059-1063
        • Vagnoni D.B.
        • Oetzel G.R.
        Effects of dietary cation-anion difference on the acid-base status of dry cows.
        J. Dairy Sci. 1998; 81 (9684172): 1643-1652
        • Van Duinkerken G.
        • Blok M.C.
        • Bannink A.
        • Cone J.W.
        • Dijkstra J.
        • van Vuuren A.M.
        • Tamminga S.
        Update of the Dutch protein evaluation system for ruminants: The DVE/OEB2010 system.
        J. Agric. Sci. 2011; 149: 351-367
        • Van Es A. J.H.
        Feed evaluation for ruminants. I. The systems in use from May 1977-onwards in the Netherlands.
        Livest. Prod. Sci. 1978; 5: 331-345
        • van Gastelen S.
        • Antunes-Fernandes E.C.
        • Hettinga K.A.
        • Klop G.
        • Alferink S. J.J.
        • Hendriks W.H.
        • Dijkstra J.
        Enteric methane production, rumen volatile fatty acid concentrations, and milk fatty acid composition in lactating Holstein-Friesian cows fed grass silage- or corn silage-based diets.
        J. Dairy Sci. 2015; 98 (25582590): 1915-1927
        • van Gastelen S.
        • Dijkstra J.
        • Nichols K.
        • Bannink A.
        Abomasal infusion of ground corn and ammonium chloride in early lactating Holstein-Friesian dairy cows to induce hindgut and metabolic acidosis.
        J. Dairy Sci. 2021; 104 (33485681): 4174-4191
        • van Knegsel A. T.M.
        • van den Brand H.
        • Dijkstra J.
        • van Straalen W.M.
        • Heetkamp M. J.W.
        • Tamminga S.
        • Kemp B.
        Dietary energy source in dairy cows in early lactation: Energy partitioning and milk composition.
        J. Dairy Sci. 2007; 90 (17297120): 1467-1476
        • van Knegsel A. T.M.
        • van den Brand H.
        • Dijkstra J.
        • van Straalen W.M.
        • Jorritsma R.
        • Tamminga S.
        • Kemp B.
        Effect of glucogenic vs. lipogenic diets on energy balance, blood metabolites, and reproduction in primiparous and multiparous dairy cows in early lactation.
        J. Dairy Sci. 2007; 90 (17582125): 3397-3409
        • van Zijderveld S.M.
        • Fonken B.
        • Dijkstra J.
        • Gerrits W. J.J.
        • Perdok H.B.
        • Fokkink W.
        • Newbold J.R.
        Effects of a combination of feed additives on methane production, diet digestibility, and animal performance in lactating dairy cows.
        J. Dairy Sci. 2011; 94 (21338809): 1445-1454
        • Vanholder T.
        • Papen J.
        • Bemers R.
        • Vertenten G.
        • Berge A. C.B.
        Risk factors for subclinical and clinical ketosis and association with production parameters in dairy cows in the Netherlands.
        J. Dairy Sci. 2015; 98 (25497823): 880-888
        • Williams E.J.
        Experimental designs balanced for the estimation of residual effects of treatments.
        Aust. J. Sci. Res., A. 1949; 2: 149-168
        • Wilms J.
        • Berends H.
        • Martín-Tereso J.
        Hypertonic milk replacers increase gastrointestinal permeability in healthy dairy calves.
        J. Dairy Sci. 2019; 102 (30471909): 1237-1246
        • Zaaijer D.
        • Kremer W. D.J.
        • Noordhuizen J. P. T.M.
        Consistentie van verse koemest.
        in: Hulsen J. Koe signalen. Praktijkgids voor koegericht management. Roodbont Uitgeverij, 2001: 54
        • Zarrin M.
        • De Matteis L.
        • Vernay M. C. M.B.
        • Wellnitz O.
        • van Dorland H.A.
        • Bruckmaier R.M.
        Long-term elevation of β-hydroxybutyrate in dairy cows through infusion: Effects on feed intake, milk production, and metabolism.
        J. Dairy Sci. 2013; 96 (23498021): 2960-2972
        • Zarrin M.
        • Wellnitz O.
        • van Dorland H.A.
        • Gross J.J.
        • Bruckmaier R.M.
        Hyperketonemia during lipopolysaccharide-induced mastitis affects systemic and local intramammary metabolism in dairy cows.
        J. Dairy Sci. 2014; 97 (24679930): 3531-3541
        • Zimpel R.
        • Poindexter M.B.
        • Vieira-Neto A.
        • Block E.
        • Nelson C.D.
        • Staples C.R.
        • Thatcher W.W.
        • Santos J. E.P.
        Effect of dietary cation-anion difference on acid-base status and dry matter intake in dry pregnant cows.
        J. Dairy Sci. 2018; 101 (29970257): 8461-8475