Advertisement

Feeding up to 91% concentrate to Holstein and Jersey dairy cows: Effects on enteric methane emission, rumen fermentation and bacterial community, digestibility, production, and feeding behavior

Open AccessPublished:October 04, 2022DOI:https://doi.org/10.3168/jds.2021-21676

      ABSTRACT

      Due to climate change, periods of drought might be longer and occur more frequently, which challenges roughage production and requires changed feeding of dairy cattle by increasing the grain content of the diet. This study investigated the effect of diets with concentrate proportions up to 91% of dry matter on dry matter intake (DMI), milk production, enteric methane emission, rumen fermentation, rumen bacterial community structure, nutrient digestibility, and feeding behavior of Holstein and Jersey dairy cows. Twelve Danish Holstein and 12 Danish Jersey cows were fed ad libitum with one of 3 total mixed rations differing in concentrate proportion in a continuous design with staggered approach over 19 to 29 d. Dietary concentrate proportions were 49% (C49), 70% (C70), and 91% (C91) on dry matter basis, and were based on increasing proportions of chopped barley straw, dried beet pulp, barley, NaOH-treated wheat, dried distillers grain, and rapeseed cake at the expense of grass/clover silage, corn silage and soybean meal. Cows were adapted to the diets over a 12- to 19-d period, before rumination activity was measured over 3 d. Subsequently, spot samples of feces were collected for digestibility determination over 2 d, and gas exchange was measured on the last 3 d of the experimental period. Shortly after chamber stay, rumen liquid was collected using an oro-ruminal device. Dry matter intake was higher for Holstein than Jersey. Methane emissions (all expressions) were affected by the interaction between breed and diet. Methane per kilogram of DMI was lowered by 18 and 48% for Holstein fed C70 and C91, respectively, compared with C49, whereas this was 17 and 22% respectively for Jersey. Rumen propionate molar proportion increased more, rumen bacterial community was less diverse, and rumination time and rumination chews relative to DMI reduced less for Holstein than for Jersey cows with increasing concentrate level. In conclusion, Holstein dairy cows responded stronger to increased dietary concentrate level regarding methane mitigation, changes in rumen VFA profile, and effect on the rumen bacterial community structure than Jersey cows, whereas Jersey cows responded stronger with regard to rumination time and rumination chews (per kilogram of DMI and per kilogram of neutral detergent fiber intake) than Holstein cows. Thus, diets high in concentrates are a less effective methane mitigation strategy for Jersey than for Holstein.

      Key words

      INTRODUCTION

      The agricultural sector is faced by challenges related to global warming and climate change, which affect human and animal food security. Changing climatic conditions, such as unexpected seasonal droughts during the growing season, become more frequent and negatively affect the quality and quantity of roughage production (
      • Rojas-Downing M.M.
      • Nejadhashemi A.P.
      • Harrigan T.
      • Woznicki S.A.
      Climate change and livestock: Impacts, adaptation, and mitigation.
      ). Challenged roughage production requires a changed feeding of dairy cattle in the short term and potentially also in a long-term perspective. For example, temporary increases in the concentrate content of rations could substitute grass silage and corn silage when availability is low. Obviously, also the production of crops for concentrate can be negatively affected by drought, but usually crop yield will not be affected simultaneously in the world. Because transportation of concentrate from other parts of the globe is easier than transportation of roughage, importing concentrate is a potential way to overcome shortage of feed. Depending on the level and composition of concentrate, especially with regard to the level of readily fermentable carbohydrates, in the ration, DMI, and milk yield increase with increasing concentrate proportion (
      • Huhtanen P.
      • Hetta M.
      Comparison of feed intake and milk production responses in continuous and change-over design dairy cow experiments.
      ;
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ), whereas, digestibility of NDF might be reduced (
      • Nousiainen J.
      • Rinne M.
      • Huhtanen P.
      A meta-analysis of feed digestion in dairy cows. 1. The effects of forage and concentrate factors on total diet digestibility.
      ;
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ). In addition, enteric methane emission declines, especially when concentrate inclusion is above 35 to 40% (
      • Sauvant D.
      • Giger-Reverdin S.
      Modélisation des interactions digestives et de la production de méthane chez les ruminants.
      ). Thus, rations high in concentrate proportion will reduce enteric methane mitigation.
      The effects of feeding diets moderately high in concentrate proportion, especially grains (starch), on methane production of dairy cattle is well established, whereas the effect of feeding only concentrate combined with a small amount of fiber from straw is not well studied. Studies investigating high concentrate diets in relation to methane emission for dairy cows have included concentrate up to 72% of DM (
      • Ferris C.P.
      • Gordon F.J.
      • Patterson D.C.
      • Porter M.G.
      • Yan T.
      The effect of genetic merit and concentrate proportion in the diet on nutrient utilization by lactating dairy cows.
      ;
      • Agle M.
      • Hristov A.N.
      • Zaman S.
      • Schneider C.
      • Ndegwa P.M.
      • Vaddella V.K.
      Effect of dietary concentrate on rumen fermentation, digestibility, and nitrogen losses in dairy cows.
      ;
      • Aguerre M.J.
      • Wattiaux M.A.
      • Powell J.M.
      • Broderick G.A.
      • Arndt C.
      Effect of forage-to-concentrate ratio in dairy cow diets on emission of methane, carbon dioxide, and ammonia, lactation performance, and manure excretion.
      ;
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ) and have mainly focused on the Holstein breed. Jersey is another widely used dairy cattle breed in some countries and differs in gastrointestinal tract size and physiology from larger cattle breeds (
      • Aikman P.C.
      • Reynolds C.K.
      • Beever D.E.
      Diet digestibility, rate of passage, and eating and rumination behavior of Jersey and Holstein cows.
      ;
      • Beecher M.
      • Buckley F.
      • Waters S.M.
      • Boland T.M.
      • Enriquez-Hidalgo D.
      • Deighton M.H.
      • O'Donovan M.
      • Lewis E.
      Gastrointestinal tract size, total-tract digestibility, and rumen microflora in different dairy cow genotypes.
      ). This difference suggests that the same diet might differ in effectiveness between breeds. Previously, we demonstrated that enteric methane emission was reduced to a larger extent for Holstein than Jersey cows when increasing forage-to-concentrate ratio from 68:32 to 39:61 (
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ). Here, we study the effect of 3 diets containing up to 91% concentrate on DM basis. This intensive study is a follow-up to a production study by
      • Børsting C.F.
      • Hellwing A.L.F.
      • Weisbjerg M.R.
      • Østergaard S.
      • Raun B.L.
      • Røjen B.A.
      • Kristensen N.B.
      Feeding dairy cows with no or reduced amounts of forage.
      , who investigated feed intake and production performance of Holstein and Jersey cows fed 5 rations differing in concentrate proportion and type of concentrate. Three of the 5 rations were subsequently included in the current study to investigate DMI, milk production, enteric methane emission, rumen fermentation, rumen bacterial community structure, nutrient digestibility, and feeding behavior in the same 2 breeds. The 3 diets were based on increasing levels of starch, mainly from rolled barley and NaOH-treated whole kernel wheat, and decreasing levels of NDF.
      • Børsting C.F.
      • Hellwing A.L.F.
      • Weisbjerg M.R.
      • Østergaard S.
      • Raun B.L.
      • Røjen B.A.
      • Kristensen N.B.
      Feeding dairy cows with no or reduced amounts of forage.
      showed the largest decrease in milk fat percentage and in rumen acetate:propionate (A:P) ratio for this type of concentrate, which indicates a large effect on rumen fermentation. The aim of this study was to examine the effect of diets with varying concentrate proportions up to 91% of DM on DMI, milk production, enteric methane emission, rumen fermentation, rumen bacterial community structure, nutrient digestibility, and feeding behavior of Holstein and Jersey dairy cows. We hypothesized that feeding increased dietary proportions of concentrate to dairy cows will decrease enteric methane emission, A:P ratio in rumen liquid, total-tract digestibility of nutrients, rumination time, and rumen bacterial diversity, with more pronounced effects for Holstein than Jersey.

      MATERIALS AND METHODS

      Experimental Design

      The experiment was conducted at the Danish Cattle Research Centre (AU Foulum, Tjele, Denmark) and in accordance with the guidelines of the European Union directive 2010/63/EU and current Danish legislation on animal experimentation (law no. 474, May 14, 2014; license no. 2018-15-0201-1495). Twenty-four dairy cows (12 Danish Holstein and 12 Danish Jersey) were fed one of 3 diets (4 Holstein and 4 Jersey per diet) differing in concentrate proportion in a continuous design with staggered approach. To allocate the diets, cows were divided into 2 blocks of 3 cows within breed and parity (12 primiparous and 12 multiparous) according to DIM (8 blocks in total). Within block, cows were allocated randomly to a specific diet that was fed throughout the experiment. Next, animals were rearranged into 6 new blocks of 4 animals to have 1 cow from each block to be allocated to 1 of 4 respiration chambers available. Each new block consisted of 2 Holstein cows and 2 Jersey cows, and 1 primiparous and 1 multiparous cow (second or third parity) for each breed. The first 4 blocks of cows started the experimental feeding on the same day and the next 2 blocks started the experiment 13 d later. There was capacity to measure feeding behavior in 8 cows at a time. Two of the first 4 blocks began these measurements after 12 d of adaptation, and the remaining 2 of these blocks after 19 d of adaptation, whereas the 2 blocks, that started the experimental feeding later, were measured after 12 d of adaptation. Feeding behavior of all cows were measured for 3 d. Spot samples of feces were collected for digestibility determination over 2 d beginning 14 d after start of the experimental feeding for all cows. Gas exchange was measured using 4 respiration chambers on the last 3 d of the experimental period for each block of 4 cows, resulting in onset of these measurements 16, 19, 23, or 26 d after start of experimental feeding. At the end of the chamber measurements, rumen liquid was collected once. One primiparous Jersey cow on a diet containing 49% concentrate had to be removed from the experiment due to difficulties with milking and was replaced by another cow, which went directly to the 49% concentrate diet 11 d into the experiment, equivalent to 5 d before feces sampling and 10 d before it was moved to the respiration chamber. Feeding behavior and digestibility were not measured for this cow.

      Animals, Diets, and Feeding

      At the start of the experiment, Holsteins were on average (±SD) 140 ± 28 DIM with a milk yield of 37.8 ± 5.1 kg/d, whereas Jerseys were 107 ± 33 DIM and produced 25.3 ± 2.6 kg/d of milk. Holsteins weighed on average 656 ± 52 kg and Jerseys 460 ± 33 kg. The animals were housed in individual tiestalls with ad libitum access to feed and water. Cows were milked at 0515 and 1630 h and fed ad libitum a TMR at 0700 and 1600 h. Feed refusals were removed and weighed daily before feeding at 1600 h.
      The Nordic Feed Evaluation System (NorFor;
      • Volden H.
      NorFor – The Nordic Feed Evaluation System. EAAP Publication No. 130.
      ) was used to formulate the rations assuming an expected milk yield of 10,500 kg of ECM/yr. The 3 diets were adapted from
      • Børsting C.F.
      • Hellwing A.L.F.
      • Weisbjerg M.R.
      • Østergaard S.
      • Raun B.L.
      • Røjen B.A.
      • Kristensen N.B.
      Feeding dairy cows with no or reduced amounts of forage.
      and differed in proportions of concentrate, but with similar DM and net energy (NEL20) contents (Table 1). The roughage-to-concentrate ratios (% of dietary DM) were: 51:49 (diet C49), 30:70 (diet C70), and 9:91 (diet C91; Table 1). The roughage part of the rations was based on grass/clover silage, corn silage, and chopped barley straw for diet C49 and C70, and only chopped barley straw in diet C91. With increasing concentrate proportion in the diet, increasing proportions of chopped barley straw, dried beet pulp, barley, NaOH-treated wheat, dried distillers grain, and rapeseed cake were included at the expense of grass/clover silage, corn silage, and soybean meal. All dry ingredients, except NaOH-treated wheat and a standard concentrate mixture, were mixed into a premix by DLG (Aarhus, Denmark) to which also titanium dioxide was added (TiO2; 1.25 g/kg of dietary DM). Titanium dioxide served as an external marker to determine nutrient digestibility. The DM concentration of the diets was adjusted by the addition of water to obtain similar DM contents between diets (approximately 400 g/kg of fresh matter). The increase from 49 to 91% concentrate in DM led to an increase in starch from 173 to 223 g/kg of DM, and a decrease in NDF from 306 to 248 g/kg of DM. At the same time, there was a slight increase in fat content from 36 to 42 g/kg of DM, and a slight increase in CP content from 159 to 171 g/kg of DM.
      Table 1Dietary and chemical composition of the diets (g/kg of DM, unless stated otherwise)
      ItemDiet
      Dietary concentrate proportions on a DM basis were 49% (C49), 70% (C70), and 91% (C91).
      C49C70C91
      Dietary concentrate proportion (%)497091
      Dietary composition
       Primary growth grass/clover silage11356.50.0
       First regrowth grass/clover silage14271.10.0
       Corn silage2431210.0
       Barley straw12.650.287.9
       Concentrate mixture
      Ingredient composition of the concentrate mixture per kilogram of DM: 170 g of dried beet pulp, 168 g of rapeseed meal, 146 g of barley, 146 of wheat, 90 g of dehulled soybean meal, 70 g of dried citrus pulp, 70 g of dehulled sunflower seed meal, 50 g of grass pellets, 50 g of wheat bran, 22 g of sugar beet molasses, 8 g of palm fatty acids distillates, 7 g of salt, 2 g of vitamins, and 1 g of magnesium sulfate.
      109109109
       Dried beet pulp120160201
       Barley112121130
       Wheat, NaOH treated0.077.4155
       Dried distillers grain0.068.7138
       Rapeseed cake78.6106134
       Soybean meal53.827.10.0
       Molasses (sugarcane)4.1412.520.9
       Palm fatty acids distillate2.112.873.64
       Vitamin and mineral premix3.433.192.93
       Salt3.234.485.73
       Limestone0.833.486.15
       Sodium bicarbonate1.612.653.68
       Magnesium oxide0.00.981.97
       Titanium dioxide1.241.251.25
      Chemical composition
       DM
      To obtain a DM content of approximately 400 g/kg fresh matter, water was added in the following amounts: 0.25 L/kg DM, 0.75 L/kg DM, and 1.26 L/kg DM for C49, C70, and C91, respectively.
      (g/kg of fresh matter)
      404408400
       Ash58.160.863.1
       CP159164171
       Crude fat363942
       Starch173194223
       NDF306278248
       INDF
      Indigestible NDF.
      76.478.078.2
       DNDF
      Digestible NDF.
      230200170
       Titanium dioxide1.321.251.15
       Gross energy
      Calculated according to NorFor (Volden and Nielsen, 2011).
      (MJ/kg of DM)
      19.019.019.1
       NEL20 (MJ/kg of DM)
      Net energy for lactation calculated for 20 kg of DMI/d (Volden and Nielsen, 2011).
      6.576.596.60
       AAT20 (g/MJ NEL20)
      Amino acids absorbed in the small intestine (MP) available for milk production calculated for 20 kg of DMI/d (Volden and Nielsen, 2011).
      16.115.314.3
       PBV20 (g/kg of DM)
      Protein balance in the rumen calculated for 20 kg of DMI/d (Volden and Larsen, 2011).
      132029
      1 Dietary concentrate proportions on a DM basis were 49% (C49), 70% (C70), and 91% (C91).
      2 Ingredient composition of the concentrate mixture per kilogram of DM: 170 g of dried beet pulp, 168 g of rapeseed meal, 146 g of barley, 146 of wheat, 90 g of dehulled soybean meal, 70 g of dried citrus pulp, 70 g of dehulled sunflower seed meal, 50 g of grass pellets, 50 g of wheat bran, 22 g of sugar beet molasses, 8 g of palm fatty acids distillates, 7 g of salt, 2 g of vitamins, and 1 g of magnesium sulfate.
      3 To obtain a DM content of approximately 400 g/kg fresh matter, water was added in the following amounts: 0.25 L/kg DM, 0.75 L/kg DM, and 1.26 L/kg DM for C49, C70, and C91, respectively.
      4 Indigestible NDF.
      5 Digestible NDF.
      6 Calculated according to NorFor (
      • Volden H.
      • Nielsen N.I.
      Energy and metabolizable protein supply.
      ).
      7 Net energy for lactation calculated for 20 kg of DMI/d (
      • Volden H.
      • Nielsen N.I.
      Energy and metabolizable protein supply.
      ).
      8 Amino acids absorbed in the small intestine (MP) available for milk production calculated for 20 kg of DMI/d (
      • Volden H.
      • Nielsen N.I.
      Energy and metabolizable protein supply.
      ).
      9 Protein balance in the rumen calculated for 20 kg of DMI/d (
      • Volden H.
      • Larsen M.
      Digestion and metabolism in the gastrointestinal tract.
      ).
      Cows receiving diet C49, were fed this diet from d 1 in the experiment. Adaptation to diet C70 and C91 was achieved gradually by mixing with decreasing proportions of diet C49 and increasing proportions of C91. For diet C70, C49 constituted 83% of DM at d 1 and 2, 67% of DM at d 3 and 4, and 50% of DM at d 5 and onward. For adaptation to diet C91, C49 constituted 67% of DM at d 1 and 2, 33% of DM at d 3 and 4, and 0% from d 5 and onward.

      Sampling and Measurements

      Feed intake was monitored daily throughout the experiment. Dry matter content of TMR and feed residues was determined for all days allocated to determination of digestibility, feeding behavior, and gas exchange. Samples of TMR were collected on 4 d during the experiment: on d 13 and 14 after experimental onset for the first 4 blocks of cows (16 animals) and on d 14 and 15 for the other 2 blocks (8 animals). These samples were pooled and stored at –20°C for further chemical analysis. To determine apparent total-tract digestibility of nutrients, 2 fecal samples (0900 and 1500 h) of 250 mL were collected and pooled for each cow at the time of sampling. Cows were equipped with a RumiWatch halter with noseband sensor (ITIN+HOCH GmbH, Liestal, Switzerland;
      • Zehner N.
      • Niederhauser J.J.
      • Nydegger F.
      • Grothmann A.
      • Keller M.
      • Hoch M.
      • Haeussermann A.
      • Schick M.
      Validation of a new health monitoring system (RumiWatch) for combined automatic measurement of rumination, feed intake, water intake and locomotion in dairy cows.
      ,
      • Zehner N.
      • Umstätter C.
      • Niederhauser J.J.
      • Schick M.
      System specification and validation of a noseband pressure sensor for measurement of ruminating and eating behavior in stable-fed cows.
      ) to determine rumination and eating time, and number of eating and rumination chews. The noseband sensor was attached to a halter and continuously recorded pressure of jaw movements related to eating and rumination at a frequency of 10 Hz. Rumination behavior is considered as chewing of a bolus and characterized by a steady frequency in jaw movements, and eating behavior is considered as the intake and chewing of feed at unsteady frequency (
      • Zehner N.
      • Umstätter C.
      • Niederhauser J.J.
      • Schick M.
      System specification and validation of a noseband pressure sensor for measurement of ruminating and eating behavior in stable-fed cows.
      ). The measurements lasted for 3 d (from the morning of d 12 until the morning of d 15 for 16 cows, and d 19–22 for the remaining 8 cows). Before the measurements, animals were habituated to the halter for 4 to 5 h. The RumiWatch Converter software version V0.7.3.2 (ITIN+HOCH GmbH) was used to convert the raw data into hourly data and summed over the day to obtain daily estimates.
      Each cow within a block was allocated to one of 4 open circuit respiration chambers, where the distribution of cows on chambers was balanced for breed, parity, and diet. Gas exchange (methane, carbon dioxide, oxygen, and hydrogen) was measured for the last 3 d (except for 4 cows where measurements were omitted for d 1 due to technical issues) of the experiment based on indirect calorimetry using the system described by
      • Hellwing A.L.F.
      • Lund P.
      • Weisbjerg M.R.
      • Brask M.
      • Hvelplund T.
      Technical note: Test of a low-cost and animal-friendly system for measuring methane emissions from dairy cows.
      . Measurements within a chamber lasted 30 s and occurred at a 12.5 min sampling frequency between measurements within a chamber. The feed bins of the chambers were automatically regulated to open 30 min after closing the chambers to enable stabilization of gas concentrations before feeding commenced. Cows were confined to the chambers throughout the measurement period and the chambers remained closed except during the twice daily occasions for feeding, milking, and cleaning (approximately 25 min per occasion). Gas measurements recorded during these events were deleted and replaced by average values of the remaining hours of the day to obtain 24 h in total. The airflow rates were set at approximately 2,000 L/min for Holstein cows and approximately 1,500 L/min for Jersey cows. The respiration chamber system was routinely checked for recovery of gases by infusing a known amount of reference gas into each chamber and measuring recovered gas concentrations. The average recovery rates were breed specific, due to the different airflow rates in the chambers. The acquired recovery rates for Holstein were 99.3 ± 0.65% for methane (based on 14 tests in total; 3 or 4 tests per chamber) and 99.3 ± 0.76% for carbon dioxide (based on 32 tests in total; 8 tests per chamber). For Jersey, applied recovery rates were 98.7 ± 0.57% for methane (based on 12 tests in total; 3 tests per chamber) and 98.8 ± 0.43% for carbon dioxide (based on 22 tests in total; 5 or 6 tests per chamber). These recovery rates were used to correct the gas measurements with. All calculations involving gases were based on standard temperature and pressure (0°C or 273.15 K; 101.325 kPa). Densities of 0.716 g of methane/L, 0.090 g of hydrogen/L, 1.965 g of carbon dioxide/L, and 1.429 g of oxygen/L were used to calculate respective gas emission in grams. An oro-ruminal sampling device (FLORA rumen scoop, Geishauser, Wittibreut, Germany;
      • Geishauser T.
      • Linhart N.
      • Neidl A.
      • Reimann A.
      Factors associated with ruminal pH at herd level.
      ) was used to collect liquid samples from the rumen (<40 mL;
      • Larsen M.
      • Hansen N.P.
      • Weisbjerg M.R.
      • Lund P.
      Technical note: Evaluation of the ororuminal FLORA sampling device for rumen fluid sampling in intact cattle.
      ) shortly after cows exited the respiration chambers and before afternoon feeding. Rumen liquid was filtered through 1 layer of cheesecloth and transferred to 4 Eppendorf tubes (1 mL), except for samples intended for microbial analysis, which were not filtered. Samples were stored frozen at −80°C until further analysis. Only molar proportions of VFA are reported, whereas VFA concentrations and pH are omitted due to the risk of saliva contamination of the samples (
      • Larsen M.
      • Hansen N.P.
      • Weisbjerg M.R.
      • Lund P.
      Technical note: Evaluation of the ororuminal FLORA sampling device for rumen fluid sampling in intact cattle.
      ).
      Milk yield was recorded daily on the last 7 d of the experiment, including the 3 d cows stayed in respiration chambers. Milk was sampled during 4 subsequent milkings during chamber stay and analyzed for protein, lactose, and fat content. Milk composition data were used to calculate ECM for the last 7 d of the experiment. Body weight was recorded at the start of the experiment when cows were moved from the loose-housing research farm to the intensive research facilities, and before and after chamber stay.

      Analytical Methods

      Samples of TMR and feces were freeze-dried before grinding using a 1 mm screen, except for subsamples to determine starch content which were ground at a 0.5-mm screen. Dry matter content was determined by drying at 60°C for 48 h (
      • AOAC International
      AOAC Official Methods of Analysis.
      ). Samples were analyzed for ash content by combustion at 525°C for 6 h and nitrogen by the Dumas method (
      • Hansen B.
      Determination of nitrogen as elementary-N, an alternative to Kjeldahl.
      ) using a Vario MAX CN apparatus (Elementar Analysensysteme GmbH, Hanau, Germany). Crude protein was calculated by multiplying the nitrogen content with the factor 6.25. Crude fat was determined by hydrolysis with hydrochloric acid using a Hydrotherm HT6 apparatus (C. Gerhardt GmbH & Co. KG) followed by Soxhlet extraction using petroleum ether with a Soxtherm SOX 416 apparatus (C. Gerhardt GmbH & Co. KG;
      • Stoldt O.W.
      Vorschlag zur vereinheitlichung der fettbestimmung in lebensmitteln.
      ) at an external laboratory (Eurofins Steins Laboratories, Vejen, Denmark). Samples were also analyzed for NDF and indigestible NDF (INDF) using heat-stable amylase and sodium sulfite (
      • Mertens D.R.
      Gravimetric determination of amylase-treated neutral detergent fiber in feeds with refluxing in beakers or crucibles: Collaborative study.
      ) following the Ankom procedure (
      • ANKOM
      Analytical methods. ANKOM Technology. Analytical Methods, ADF NDF and Crude Fiber, Automated Fiber Analyzer. ANKOM Technology.
      ) and data are presented as ash-free NDF. Before analyzing INDF, TMR samples were first incubated in F57 Ankom bags for 288 h (12 d) in the rumen of 3 dry cows fed a standard ration at maintenance (for the ration description see
      • Brask M.
      • Lund P.
      • Hellwing A.L.F.
      • Poulsen M.
      • Weisbjerg M.R.
      Enteric methane production, digestibility and rumen fermentation in dairy cows fed different forages with and without rapeseed fat supplementation.
      ). Starch was analyzed enzymatically using heat-stable α-amylase and amyloglucosidase and measured as liberated glucose (YSI model 2900 analyzer, YSI Inc.;
      • Kristensen N.B.
      • Storm A.
      • Raun B.M.L.
      • Røjen B.A.
      • Harmon D.L.
      Metabolism of silage alcohols in lactating dairy cows.
      ). Titanium dioxide was analyzed spectrophotometrically (Lamba 900, PerkinElmer Inc.) as described by
      • Myers W.D.
      • Ludden P.A.
      • Nayigihugu V.
      • Hess B.W.
      Technical note: A procedure for the preparation and quantitative analysis of samples for titanium dioxide.
      with an adjustment of the method by adding 15 mL of 30% hydrogen peroxide instead of 10 mL and 5 additional drops before measurement of absorbance. Rumen liquid (4 mL) for VFA analysis was stabilized with 1 mL of 25% metaphosphoric acid (MPA) solution to reach 5% MPA in the stabilized sample and analyzed by gas chromatography according to
      • Kristensen N.B.
      • Danfær A.
      • Tetens V.
      • Agergaard N.
      Portal recovery of intraruminally infused short-chain fatty acids in sheep.
      with some modifications. The VFA concentrations were determined in stabilized ruminal liquid after methanol-chloroform extraction using 2-ethylbutyrate as internal standard. The gas chromatograph (Trace 1310, Thermo Scientific) was operated with split/splitless injector at 225°C and a flame ionization detector at 250°C. A 30 m × 0.53 mm × 1 µm HP-FFAP column (Agilent Technologies Inc.) was used with helium as carrier gas at 0.3405 atm. The oven was programmed to increase from 100 to 200°C at 10°C/min. Contents of protein, lactose monohydrate, and fat in milk were analyzed using an infrared analyzer (Milkoscan Msc4000, Foss Analytical) at Eurofins Steins Laboratories (Vejen, Denmark).

      Bacterial DNA Extraction and Analyses

      To extract DNA from rumen samples, a Nucleospin Soil DNA extraction kit (Machery-Nagel) was used as described by
      • Noel S.J.
      • Olijhoek D.W.
      • McLean F.
      • Løvendahl P.
      • Lund P.
      • Højberg O.
      Rumen and fecal microbial community structure of Holstein and Jersey dairy cows as affected by breed, diet, and residual feed intake.
      . Amplicon libraries covering the V3-V4 region of the 16S rRNA gene were also prepared according to
      • Noel S.J.
      • Olijhoek D.W.
      • McLean F.
      • Løvendahl P.
      • Lund P.
      • Højberg O.
      Rumen and fecal microbial community structure of Holstein and Jersey dairy cows as affected by breed, diet, and residual feed intake.
      using universal primers Bac341F and Bac805R as recommended by
      • Klindworth A.
      • Pruesse E.
      • Schweer T.
      • Peplies J.
      • Quast C.
      • Horn M.
      • Glöckner F.O.
      Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies.
      . Amplicon libraries were sequenced on the Illumina MiSeq (Illumina) using 300 bp paired end reads. Bioinformatics on sequence reads were performed in the QIIME2 pipeline (Qiime2 core 2019.7;
      • Bolyen E.
      • Rideout J.R.
      • Dillon M.R.
      • Bokulich N.A.
      • Abnet C.C.
      • Al-Ghalith G.A.
      • Alexander H.
      • Alm E.J.
      • Arumugam M.
      • Asnicar F.
      • Bai Y.
      • Bisanz J.E.
      • Bittinger K.
      • Brejnrod A.
      • Brislawn C.J.
      • Brown C.T.
      • Callahan B.J.
      • Caraballo-Rodriguez A.M.
      • Chase J.
      • Cope E.
      • Da Silva R.
      • Dorrestein P.C.
      • Douglas G.M.
      • Durall D.M.
      • Duvallet C.
      • Edwardson C.F.
      • Ernst M.
      • Estaki M.
      • Fouquier J.
      • Gauglitz J.M.
      • Gibbons S.M.
      • Gibson D.L.
      • Gonzalez A.
      • Gorlick K.
      • Guo J.
      • Hillmann B.
      • Holmes S.
      • Holste H.
      • Huttenhower C.
      • Huttley G.A.
      • Janssen S.
      • Jarmusch A.K.
      • Jiang L.
      • Kaehler B.D.
      • Kang K.B.
      • Keefe C.R.
      • Keim P.
      • Kelley S.T.
      • Knights D.
      • Koester I.
      • Kosciolek T.
      • Kreps J.
      • Langille M.G.I.
      • Lee J.
      • Ley R.
      • Liu Y.-X.
      • Loftfield E.
      • Lozupone C.
      • Maher M.
      • Marotz C.
      • Martin B.D.
      • McDonald D.
      • McIver L.J.
      • Melnik A.V.
      • Metcalf J.L.
      • Morgan S.C.
      • Morton J.T.
      • Naimey A.T.
      • Navas-Molina J.A.
      • Nothias L.F.
      • Orchanian S.B.
      • Pearson T.
      • Peoples S.L.
      • Petras D.
      • Preuss M.L.
      • Pruesse E.
      • Rasmussen L.B.
      • Rivers A.
      • Robeson II, M.S.
      • Rosenthal P.
      • Segata N.
      • Shaffer M.
      • Shiffer A.
      • Sinha R.
      • Song S.J.
      • Spear J.R.
      • Swafford A.D.
      • Thompson L.R.
      • Torres P.J.
      • Trinh P.
      • Tripathi A.
      • Turnbaugh P.J.
      • Ul-Hasan S.
      • van der Hooft J.J.J.
      • Vargas F.
      • Vázquez-Baeza Y.
      • Vogtmann E.
      • von Hippel M.
      • Walters W.
      • Wan Y.
      • Wang M.
      • Warren J.
      • Weber K.C.
      • Williamson C.H.D.
      • Willis A.D.
      • Xu Z.Z.
      • Zaneveld J.R.
      • Zhang Y.
      • Zhu Q.
      • Knight R.
      • Caporaso J.G.
      Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2.
      ). First, raw sequence data were demultiplexed and quality filtered using the q2-demux plugin with the following options: forward reads truncated after 300 bases and reverse reads truncated after 266 bases, primers were removed, max ee = 2 and trunc_q = 2. This was followed by denoizing and grouping into amplicon sequence variants (ASV) with DADA2 (
      • Callahan B.J.
      • McMurdie P.J.
      • Rosen M.J.
      • Han A.W.
      • Johnson A.J.A.
      • Holmes S.P.
      DADA2: High-resolution sample inference from Illumina amplicon data.
      ). Representative sequences of ASV were aligned with mafft (q2-alignment;
      • Katoh K.
      • Standley D.M.
      MAFFT multiple sequence alignment software version 7: Improvements in performance and usability.
      ) and used to construct a phylogentic tree with fasttree2 (q2-phylogeny;
      • Price M.N.
      • Dehal P.S.
      • Arkin A.P.
      FastTree 2 - approximately maximum-likelihood trees for large alignments.
      ). Samples were rarefied (subsampled without replacement) to 36,806 sequences per sample before α-diversity metrics (ASV richness and Shannon diversity; within sample diversity), β-diversity [weighted UniFrac (
      • Lozupone C.A.
      • Hamady M.
      • Kelley S.T.
      • Knight R.
      Quantitative and qualitative β diversity measures lead to different insights into factors that structure microbial communities.
      ) and unweighted UniFrac (
      • Lozupone C.
      • Knight R.
      UniFrac: A new phylogenetic method for comparing microbial communities.
      ); between sample diversity], and principle coordinate analysis were calculated using core diversiy metrics (q2-diversity). Group significance on α-diversity metrics were perfomed using Kruskal-Wallis with pairwise comparisons and Benjamini-Hochberg correction to control for the false discovery rate (presented as q-values). Group significance on β-diversity metrics were performed with PERMANOVA with 999 permutations and pairwise comparisons (q2-diversity). Taxonomy was assigned using the q2-feature-classifier (
      • Bokulich N.A.
      • Kaehler B.D.
      • Rideout J.R.
      • Dillon M.
      • Bolyen E.
      • Knight R.
      • Huttley G.A.
      • Caporaso J.G.
      Optimizing taxonomic classification of marker-gene amplicon sequences with QIIME 2's q2-feature-classifier plugin.
      ) to classify-sklearn naïve Bayes taxonomy classifier trained on the Greengenes 13_8 99% operational taxonomic unit reference sequences (
      • McDonald D.
      • Price M.N.
      • Goodrich J.
      • Nawrocki E.P.
      • DeSantis T.Z.
      • Probst A.
      • Andersen G.L.
      • Knight R.
      • Hugenholtz P.
      An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea.
      ). Alpha rarefaction was performed at sampling depth 30,000 to determine if the sampling depth was high enough to cover the observed diversity (q2-diversity). Alpha rarefaction graphs indicated sufficient sampling depth to cover the sequence variation (data not shown).
      Raw microbiome sequence reads were deposited in the National Center for Biotechnology Information (NCBI) short-read archive database under BioProject ID: PRJNA786778 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA786778). After quality filtering and assigning to ASV, a total of 1,214,636 sequence reads were obtained for 24 rumen samples. The samples had an average of 50,610 reads each (minimum = 36,806; maximum = 69,963), which fell into 6,386 ASV. Principal coordinate plots and α-diversity plots were drawn in R (version 4.0.5,
      • R Core Team
      R: A language and environment for statistical computing.
      ) from distance matrices derived in QIIME2.

      Calculations and Statistical Analyses

      Dry matter intake was calculated as the total amount of DM offered minus the amount of DM in the refusals. The average daily amount of titanium dioxide supplied in the feed was used as a marker to calculate fecal DM flow. Apparent total-tract digestibility of DM, OM, CP, starch, and NDF were calculated from the respective nutrient intake and fecal flow. Gross energy intake (GEI), NEL20, AA absorbed in the small intestine (AAT20; metabolizable protein), protein balance in the rumen (PBV20) were calculated in regard to a standard of 20 kg of DMI/d according to NorFor (
      • Volden H.
      • Larsen M.
      Digestion and metabolism in the gastrointestinal tract.
      ;
      • Volden H.
      • Nielsen N.I.
      Energy and metabolizable protein supply.
      ). Energy-corrected milk yield (3.14 MJ/kg) was calculated as ECM yield (kg/d) = milk yield (kg/d) × [(38.3 × milk fat (g/kg) + 24.2 × milk protein (g/kg) + 15.71 × milk lactose (g/kg) + 20.7)/3,140], where lactose is lactose monohydrate (
      • Sjaunja L.O.
      • Baevre L.
      • Junkkarinen L.
      • Pedersen J.
      • Setala J.
      A Nordic proposal for an energy corrected milk (ECM) formula.
      ). Feed-conversion efficiency (FCE) was calculated as kg of ECM/kg of DMI.
      All variables were averaged per cow (24 observations in total). For VFA and bacterial community data, one observation of a Holstein receiving diet C49 was deleted (23 observations in total) because of a low feed intake on the last day of gas measurements. Another observation was omitted for digestibility and feeding behavior measurements, resulting in 23 observations in total. This observation belonged to a Jersey receiving diet C49, replacing another cow in the experiment, and was omitted because of an insufficient adaptation to the diet before measurements of feeding behavior and digestibility were taken. For this specific cow, data collected later in the experiment during gas measurements were retained in the analysis, because the adaptation length was considered minimal, but sufficient (i.e., 10 d). Feeding behavior data for another Jersey receiving diet C49 was based on 1 d of observation, due to issues with the RumiWatch halter. Proc MIXED in SAS (version 9.4, SAS Institute Inc.) was used to analyze the data including fixed effects for breed (Holstein and Jersey), diet (C49, C70, and C91), parity (primiparous or multiparous), and the interaction between breed and diet, and a random effect for block (6 levels). Cow was the experimental unit. The model included containment as degrees of freedom method. Another mixed model was made for each breed separately including fixed effects for parity and diet and a random effect for block to obtain linear and quadratic polynomial contrasts across diets within breed (quadratic contrasts are not presented in tables, but significant quadratic contrasts and tendencies are presented in footnotes of the tables). Least squares means are reported in tables. The raw sequence read counts from the ASV abundance table were collapsed at the species taxonomic rank and normalized to the relative abundance counts. Spearman rank correlations between the relative abundance of individual rumen bacterial species and molar proportions of VFA in rumen liquid, hydrogen production, and methane emission were made across breed and diet and presented as a heatmap (Supplemental Figure S1, https://doi.org/10.5281/zenodo.7074417;
      • Olijhoek D.W.
      • Hellwing A.L.F.
      • Noel S.J.
      • Lund P.
      • Larsen M.
      • Weisbjerg M.R.
      • Børsting C.F.
      Feeding up to 91% concentrate to Holstein and Jersey dairy cows: Effects on enteric methane emission, rumen fermentation and bacterial community, digestibility, production, and feeding behavior.
      ). The heatmap was created in R using the gplots package (
      • Warnes G.R.
      • Bolker B.
      • Bonebakker L.
      • Gentleman R.
      • Liaw W.H.A.
      • Lumley T.
      • Maechler M.
      • Magnusson A.
      • Moeller S.
      • Schwartz M.
      • Venables B.
      gplots: Various R programming tools for plotting data. R package version 3.0.1.1.
      ). Statistical significance was declared at P ≤ 0.05 and tendencies at 0.05 < P ≤ 0.10.

      RESULTS

      Interaction Between Breed and Dietary Concentrate Proportion

      Dry matter intake, nutrient intake, and nutrient digestibility during the 2 d of feces collection showed no interaction between breed and diet; however, the apparent total-tract digestibility of NDF tended to decrease linearly for Holstein (P = 0.06) and significantly decreased in a linear way for Jersey (P = 0.04) with increasing concentrate proportion (Table 2). In addition, apparent total-tract digestibility of DM and OM tended to decrease linearly for Jersey (P = 0.09 for both), whereas being unaffected for Holstein. Acetate molar proportion decreased linearly with increased level of concentrate for Holstein (P = 0.03) and Jersey (P = 0.05), and there was no interaction with breed (Table 3). The interaction was significant for propionate molar proportion (P = 0.04). There was a linear (P < 0.001) and a quadratic (P = 0.04) increase in propionate molar proportion for Holstein (P < 0.001), whereas there were no linear or quadratic effects for Jersey. The A:P ratio showed a linear decline across diets only for Holstein (P = 0.001), even if there was no interaction between breed and diet. For butyrate, there was a tendency for a linear decrease for Holstein (P = 0.10) and a tendency for an interaction between breed and diet (P = 0.08). For Holstein a linear decline with increasing concentrate proportion was observed for molar proportions of iso-butyrate (P = 0.02), whereas for Jersey there was both a linear (P < 0.001) and a quadratic decline (P = 0.04) with increased proportion of concentrate, in combination with a tendency toward an interaction (P = 0.07). For iso-valerate there was a tendency (P = 0.07) for interaction between breed and diet. Additionally, a breed and diet interaction was found for caproate molar proportion (P = 0.04), and for Jersey, a linear (P = 0.01) and a tendency for a quadratic (P = 0.09) decline was observed. Daily methane production (interaction P-value and linear effect P-value for Holstein are 0.001) and methane intensity of Holstein (interaction P = 0.03 and linear effect for Holstein P = 0.01) were decreased with increased concentrate proportion, but was unaffected for Jersey. Significant interactions between breed and diet were found for methane yield and methane energy losses in percentage of GEI (P = 0.03 and 0.04, respectively; Table 4). Methane yield of Holstein was lowered with 18 and 48%, respectively, for diet C70 and C91 relative to diet C49, whereas the reductions for Jersey were only 17 and 22% for diet C70 and C91 relative to diet C49, respectively. Linear declines in methane emission (all expressions) and CH4:CO2 were found for Holstein together with quadratic effects for methane production (P = 0.01) and CH4:CO2 (P = 0.001) as well as tendencies for methane yield (P = 0.06) and methane energy loss percentage of GEI (P = 0.06), whereas for methane intensity there was no quadratic effect. For Jersey, linear declines across diets were observed for methane yield (P = 0.03) and methane energy loss percentage of GEI (P = 0.03). Hydrogen emission only increased linearly for Holstein (P = 0.03). For Jersey, the linear increase was not significant even if hydrogen production increased by 57 and 104% (C70 and C91 vs. C49). For BW, milk yield, milk composition, and FCE there was no interaction between breed and diet, but linear declines in milk fat percentage (P = 0.06 for Holstein and P = 0.08 for Jersey) and FCE (P = 0.07 for Jersey) were observed (Table 5). The total eating and rumination time and chews were unaffected by the interaction term, but linear declines were observed. The total time spent eating or ruminating (P = 0.02) and the total number of chews during eating and rumination declined linearly for Holstein (P = 0.01) with increasing concentrate proportion, whereas these variables tended to decline linearly for Jersey (P = 0.09 and P = 0.08, respectively; Table 6). Rumination time and rumination chews (all expressions) decreased linearly across diets for both breeds; however, larger declines in rumination time and rumination chews per kilogram of DMI and NDF intake were observed for Jersey than Holstein (interaction: P = 0.02 for rumination time per kilogram of DMI and NDF intake, and P = 0.01 and P = 0.001 for rumination chews per kilogram of DMI and NDF intake, respectively). Eating time and eating chews (all expressions) were unaffected by the interaction term and no significant linear contrasts were observed. The quadratic contrast across diets was significant for Jersey at rumination time (min/d: P = 0.02; per kilogram of DMI: P = 0.04; per kilogram of NDF intake: P = 0.01) and rumination chews (number per day: P = 0.01; number per kilogram of NDF intake: P = 0.03). The ASV richness (i.e., the number of unique sequence types) of the bacterial community in rumen liquid tended to be lower for Holstein than for Jersey for diet C91 (q = 0.06) and was not different for diet C49 (q = 0.31) and C70 (q = 0.39; Figure 1). The Shannon diversity (an indicator of evenness and richness in the community structure), was lower for Holstein than for Jersey for diet C70 and C91 (q = 0.05 for both), but not for diet C49. The weighted and unweighted UniFrac distances (visualized as principle coordinate analysis plots) show separation according to breed and diet (Figure 2). The difference between the breeds was visually less at the lowest concentrate inclusion level (C49) than at the highest concentrate inclusion level (C91) for both weighted and unweighted UniFrac distances.
      Table 2Nutrient intake and apparent total-tract nutrient digestibility of Holstein and Jersey cows fed concentrate at 49 (C49), 70 (C70), or 91% of dietary DM (C91)
      Based on 23 observations (missing observation is for Jersey receiving diet C49).
      ItemHolsteinJerseySEMP-value
      C49C70C91C49C70C91BreedDietBreed × DietLinear for Holstein
      Linear contrast for diet within breed. Quadratic contrasts for diet within breed were nonsignificant for all variables.
      Linear for Jersey
      Linear contrast for diet within breed. Quadratic contrasts for diet within breed were nonsignificant for all variables.
      Intake (kg/d)
       DM
      Dry matter intake during the 2 d of feces collection.
      22.324.421.917.018.219.41.510.0010.520.390.850.32
       OM21.022.920.516.017.118.21.420.0010.540.380.810.33
       CP3.544.013.752.703.003.310.2560.0010.200.430.580.18
       Starch3.854.734.902.943.544.330.3250.0010.010.550.090.07
       NDF6.816.725.555.205.034.910.3950.0010.090.270.090.30
      Digestibility (%)
       DM70.670.769.272.770.568.70.840.560.020.220.180.09
       OM71.972.070.674.271.770.10.880.520.020.200.180.09
       CP63.762.062.763.260.460.11.440.100.150.670.430.13
       Starch98.898.797.899.498.498.10.690.730.200.740.180.12
       NDF54.751.546.660.953.545.52.470.250.0010.290.060.04
      1 Based on 23 observations (missing observation is for Jersey receiving diet C49).
      2 Linear contrast for diet within breed. Quadratic contrasts for diet within breed were nonsignificant for all variables.
      3 Dry matter intake during the 2 d of feces collection.
      Table 3Molar proportions of VFA in rumen liquid of Holstein and Jersey cows fed concentrate at 49 (C49), 70 (C70), or 91% of dietary DM (C91)
      Based on 23 observations (missing observation is for 1 Holstein receiving diet C49).
      ItemHolsteinJerseySEMP-value
      C49C70C91C49C70C91BreedDietBreed × DietLinear for Holstein
      Linear contrast for diet within breed.
      Linear for Jersey
      Linear contrast for diet within breed.
      Acetate (mol/100 mol)60.357.752.763.659.458.01.250.001<0.0010.290.030.05
      Propionate (mol/100 mol)22.527.134.119.723.723.51.730.0010.0010.040.001
      The quadratic contrast across diets was significant for Holstein at propionate molar proportion (P = 0.04) and for Jersey at iso-butyrate molar proportion (P = 0.04), and showed a tendency for Jersey at caproate molar proportion (P = 0.09).
      0.26
      Butyrate (mol/100 mol)12.911.69.4912.913.314.91.250.020.910.080.100.30
      Valerate (mol/100 mol)1.661.662.101.491.651.750.1710.160.080.530.180.08
      Iso-butyrate (mol/100 mol)0.820.620.350.810.590.540.0570.20<0.0010.070.020.001
      The quadratic contrast across diets was significant for Holstein at propionate molar proportion (P = 0.04) and for Jersey at iso-butyrate molar proportion (P = 0.04), and showed a tendency for Jersey at caproate molar proportion (P = 0.09).
      Iso-valerate (mol/100 mol)1.241.010.761.020.781.180.1720.990.320.070.110.43
      Caproate (mol/100 mol)0.560.320.370.540.460.140.0720.460.0010.040.220.01
      The quadratic contrast across diets was significant for Holstein at propionate molar proportion (P = 0.04) and for Jersey at iso-butyrate molar proportion (P = 0.04), and showed a tendency for Jersey at caproate molar proportion (P = 0.09).
      A:P
      Acetate-to-propionate ratio.
      2.702.151.553.242.522.640.2420.0010.010.240.0010.23
      (A + B):P
      Acetate plus butyrate-to-propionate ratio.
      3.272.571.833.893.093.350.3270.0010.020.190.010.38
      1 Based on 23 observations (missing observation is for 1 Holstein receiving diet C49).
      2 Linear contrast for diet within breed.
      3 The quadratic contrast across diets was significant for Holstein at propionate molar proportion (P = 0.04) and for Jersey at iso-butyrate molar proportion (P = 0.04), and showed a tendency for Jersey at caproate molar proportion (P = 0.09).
      4 Acetate-to-propionate ratio.
      5 Acetate plus butyrate-to-propionate ratio.
      Table 4Dry matter intake and gas exchange of Holstein and Jersey cows fed concentrate at 49 (C49), 70 (C70), or 91% of dietary DM (C91)
      ItemHolsteinJerseySEMP-value
      C49C70C91C49C70C91BreedDietBreed × DietLinear for Holstein
      Linear contrast for diet within breed.
      Linear for Jersey
      Linear contrast for diet within breed.
      DMI
      Dry matter intake during chamber stay.
      (kg/d)
      22.223.922.016.618.719.00.98<0.0010.190.390.830.21
      DMI
      MBW = metabolic BW (BW0.75).
      (kg/kg of MBW per day)
      0.180.190.170.170.200.190.0100.490.200.400.650.18
      CH4 (g/d)40736421334732730616.30.940.0010.0010.001
      The quadratic contrast across diets for Holstein was significant for daily methane production (P = 0.01) and CH4:CO2 ratio (P = 0.001), and showed tendencies for methane per kilogram of DMI (P = 0.06) and methane percentage of gross energy intake (GEI; P = 0.06).
      0.16
      CH4 (g/kg of DMI)18.515.29.7221.017.416.30.80<0.001<0.0010.030.001
      The quadratic contrast across diets for Holstein was significant for daily methane production (P = 0.01) and CH4:CO2 ratio (P = 0.001), and showed tendencies for methane per kilogram of DMI (P = 0.06) and methane percentage of gross energy intake (GEI; P = 0.06).
      0.03
      CH4 (g/kg of ECM)12.010.27.2611.710.510.90.690.050.010.030.010.54
      CH4 (% of GEI)5.364.402.816.105.044.700.233<0.001<0.0010.040.001
      The quadratic contrast across diets for Holstein was significant for daily methane production (P = 0.01) and CH4:CO2 ratio (P = 0.001), and showed tendencies for methane per kilogram of DMI (P = 0.06) and methane percentage of gross energy intake (GEI; P = 0.06).
      0.03
      H2 (g/d)0.941.502.370.951.491.940.3290.600.010.760.030.18
      CO2 (g/d)14,68714,82514,40711,16411,64211,243424<0.0010.610.890.610.92
      O2 (g/d)9,5729,5829,6907,0267,2566,984317<0.0010.930.840.990.46
      CH4:CO2 ratio
      CH4-to-CO2 ratio based on gas production in g/d.
      0.0280.0240.0150.0310.0280.0270.0011<0.001<0.0010.001<0.001
      The quadratic contrast across diets for Holstein was significant for daily methane production (P = 0.01) and CH4:CO2 ratio (P = 0.001), and showed tendencies for methane per kilogram of DMI (P = 0.06) and methane percentage of gross energy intake (GEI; P = 0.06).
      0.14
      1 Linear contrast for diet within breed.
      2 Dry matter intake during chamber stay.
      3 MBW = metabolic BW (BW0.75).
      4 The quadratic contrast across diets for Holstein was significant for daily methane production (P = 0.01) and CH4:CO2 ratio (P = 0.001), and showed tendencies for methane per kilogram of DMI (P = 0.06) and methane percentage of gross energy intake (GEI; P = 0.06).
      5 CH4-to-CO2 ratio based on gas production in g/d.
      Table 5Body weight, milk production and composition, and feed efficiency during chamber stay of Holstein and Jersey cows fed concentrate at 49 (C49), 70 (C70), or 91% of dietary DM (C91)
      ItemHolsteinJerseySEMP-value
      C49C70C91C49C70C91BreedDietBreed × DietLinear for Holstein
      Linear contrast for diet within breed. Quadratic contrasts for diet within breed were nonsignificant for all variables.
      Linear for Jersey
      Linear contrast for diet within breed. Quadratic contrasts for diet within breed were nonsignificant for all variables.
      BW (kg)62162364845042745417.5<0.0010.250.710.440.99
      Milk yield (kg/d)33.736.435.821.924.723.22.18<0.0010.450.970.670.53
      ECM yield (kg/d)34.236.229.629.831.028.82.150.070.160.570.320.67
      Fat (%)3.953.822.496.515.525.410.333<0.0010.010.210.060.08
      Protein (%)3.643.593.494.274.274.420.155<0.0010.980.600.720.53
      Lactose (%)4.924.905.034.694.844.690.0620.0010.520.140.330.98
      FCE
      FCE = feed-conversion efficiency, using DMI during chamber stay.
      (kg of ECM/kg of DMI)
      1.541.511.351.811.661.520.0790.010.030.740.200.07
      1 Linear contrast for diet within breed. Quadratic contrasts for diet within breed were nonsignificant for all variables.
      2 FCE = feed-conversion efficiency, using DMI during chamber stay.
      Table 6Feeding behavior of Holstein and Jersey cows fed concentrate at 49 (C49), 70 (C70), or 91% of dietary DM (C91)
      Based on 23 observations (missing observation is for 1 Jersey receiving diet C49). Data for 1 Jersey receiving diet C49 is based on 1 d of observation.
      ItemHolsteinJerseySEMP-value
      C49C70C91C49C70C91BreedDietBreed × DietLinear for Holstein
      Linear contrast for diet within breed.
      Linear for Jersey
      Linear contrast for diet within breed.
      DMI
      DM and NDF intake during feeding behavior measurements.
      (kg/d)
      22.024.022.417.018.718.11.29<0.0010.310.910.800.62
      NDF intake
      DM and NDF intake during feeding behavior measurements.
      (kg/d)
      6.716.615.685.195.174.590.341<0.0010.030.750.090.10
      Total eating plus rumination time (min/d)96190178393684365654.30.090.0010.580.020.09
      Total eating plus rumination chews (n/d)72,98665,00953,56867,96557,85543,8343,9000.02<0.0010.800.010.08
      Eating time (min/d)44143443146543042735.20.880.750.880.750.58
      Eating time (min/kg of DMI)20.118.319.427.323.024.12.340.010.420.800.730.52
      Eating time (min/kg of NDF intake)65.766.476.589.383.495.48.610.010.330.910.210.72
      Eating chews (n/d)34,80731,69630,00035,53631,48130,4912,9300.880.200.980.110.41
      Eating chews (n/kg of DMI)1,5841,3331,3422,0941,6841,7301950.010.170.900.070.41
      Eating chews (n/kg of NDF intake)5,1904,8355,3026,8516,1076,8347050.010.580.950.980.98
      Rumination time (min/d)51745536645742022827.2<0.001<0.0010.070.010.001
      The quadratic contrast across diets was significant for Jersey at rumination time (min/d: P = 0.02; per kg of DMI: P = 0.04; per kg of NDF intake: P = 0.01) and rumination chews (number per day: P = 0.01; number per kg of NDF intake: P = 0.03).
      Rumination time (min/kg of DMI)23.819.116.127.322.112.71.390.33<0.0010.020.010.001
      The quadratic contrast across diets was significant for Jersey at rumination time (min/d: P = 0.02; per kg of DMI: P = 0.04; per kg of NDF intake: P = 0.01) and rumination chews (number per day: P = 0.01; number per kg of NDF intake: P = 0.03).
      Rumination time (min/kg of NDF intake)77.969.463.789.480.350.24.980.450.0010.020.050.001
      The quadratic contrast across diets was significant for Jersey at rumination time (min/d: P = 0.02; per kg of DMI: P = 0.04; per kg of NDF intake: P = 0.01) and rumination chews (number per day: P = 0.01; number per kg of NDF intake: P = 0.03).
      Rumination chews (n/d)37,81132,56924,68031,51726,83213,1401,688<0.001<0.0010.060.0010.001
      The quadratic contrast across diets was significant for Jersey at rumination time (min/d: P = 0.02; per kg of DMI: P = 0.04; per kg of NDF intake: P = 0.01) and rumination chews (number per day: P = 0.01; number per kg of NDF intake: P = 0.03).
      Rumination chews (n/kg of DMI)1,7371,3671,0871,8761,41273882.20.22<0.0010.010.0010.001
      Rumination chews (n/kg of NDF intake)5,6904,9604,2876,1525,1302,9052940.14<0.0010.0010.010.001
      The quadratic contrast across diets was significant for Jersey at rumination time (min/d: P = 0.02; per kg of DMI: P = 0.04; per kg of NDF intake: P = 0.01) and rumination chews (number per day: P = 0.01; number per kg of NDF intake: P = 0.03).
      1 Based on 23 observations (missing observation is for 1 Jersey receiving diet C49). Data for 1 Jersey receiving diet C49 is based on 1 d of observation.
      2 Linear contrast for diet within breed.
      3 DM and NDF intake during feeding behavior measurements.
      4 The quadratic contrast across diets was significant for Jersey at rumination time (min/d: P = 0.02; per kg of DMI: P = 0.04; per kg of NDF intake: P = 0.01) and rumination chews (number per day: P = 0.01; number per kg of NDF intake: P = 0.03).
      Figure thumbnail gr1
      Figure 1Measurements of rumen bacterial community diversity (α-diversity) as amplicon sequence variants (ASV; number of unique sequence types) richness (A) and Shannon diversity (B) of Holstein and Jersey cows fed concentrate at 49 (C49), 70 (C70), or 91% of dietary DM (C91). Group significance based on Kruskal-Wallis for ASV richness: P = 0.42 for breed, P < 0.001 for diet, q = 0.31 for Holstein versus Jersey for diet C49, q = 0.39 for Holstein versus Jersey for diet C70, and q = 0.06 for Holstein versus Jersey for diet C91; for Shannon diversity: P = 0.12 for breed, P < 0.001 for diet, q = 0.72 for Holstein versus Jersey for diet C49, q = 0.05 for Holstein versus Jersey for diet C70, and q = 0.05 for Holstein versus Jersey for diet C91.
      Figure thumbnail gr2
      Figure 2Principal coordinate (PCo) analysis of weighted (A) and unweighted UniFrac (B) of the rumen bacterial community (β-diversity) of Holstein and Jersey cows fed concentrate at 49 (C49), 70 (C70), or 91% of dietary DM (C91). Group significance based on PERMANOVA with 999 permutations for weighted UniFrac: P = 0.001 for breed and P = 0.001 for diet; for unweighted UniFrac: P = 0.01 for breed and P = 0.001 for diet.

      Effect of Breed

      Intakes of DM and nutrients were higher for Holstein than Jersey (P = 0.001 for all; Table 2). Apparent total-tract digestibility of DM and nutrients were unaffected by breed, except for a tendency for a lower CP digestibility for Jersey than Holstein (P = 0.10). Further, Jersey had higher molar proportions of acetate (P = 0.001; Table 3). As a result, the A:P ratio as well as acetate plus butyrate-to-propionate [(A+B):P] ratio were higher for Jersey than Holstein (P = 0.001 for both ratios). All other VFA were unaffected by breed. Holstein had a consistently lower methane yield (methane per kilogram of DMI; P < 0.001), methane intensity (methane per kilogram of ECM; P = 0.05; numerically not lower for diet C49), methane loss as a percentage of GEI (P < 0.001), and CH4:CO2 ratio than Jersey (P < 0.001); however, there was also an interaction between breed and diet caused by a larger dietary effect for Holstein (Table 4). Carbon dioxide production and oxygen consumption were higher for Holstein than Jersey (P < 0.001 for both), whereas hydrogen production was unaffected by breed. Holstein had a higher milk yield (P < 0.001), tended to have a higher ECM yield (P = 0.07), and had a higher milk lactose percentage than Jersey (P = 0.001), whereas milk fat and protein percentage were lower for Holstein (P < 0.001 for both; Table 5). Holstein had a larger BW than Jersey (P < 0.001). Feed-conversion efficiency was also lower for Holstein than for Jersey (P = 0.01). Further, the total number of chews during eating and rumination was lower (P = 0.02) and the total eating plus rumination time tended to be lower for Jersey than for Holstein (P = 0.09; Table 6). Jersey spent more time eating per kilogram of DMI and NDF intake (P = 0.01 for both) and had a greater number of eating chews per kilogram of DMI and NDF intake than Holstein (P = 0.01 for both); however, daily eating time and daily number of chews were unaffected by breed. In contrast, daily rumination time (P < 0.001) and number of rumination chews per day were higher for Holstein than Jersey for all diets (P < 0.001) even if there was a tendency for an interaction for both parameters (P = 0.07 and P = 0.06, respectively).
      Representative sequences from the 6,386 ASV were assigned to 23 phyla and 110 genera. The relative abundance of dominant species in each treatment group are shown in Figure 3. The most dominant genera (>2% of abundance) for both breeds were Prevotella (17.5%), Succinivibrionaceae (12.6%), Bacteroidales (8.4%), Clostridiales (6.35%), Ruminococcus (5.65%), Lachnospiraceae (3.84%), Ruminococcaceae (3.34%), Succiniclasticum (2.73%), Treponema (2.70%), and Coprococcus (2.59%). The heatmap showing correlations between relative abundances of rumen bacterial species with methane emission, hydrogen production, and molar proportions of VFA across breeds and diets is presented in Supplemental Figure S1. Two clusters in bacterial species are observed: one block of 26 species is positively correlated with hydrogen production and molar proportions of propionate and valerate, whereas being negatively correlated with methane emission and molar proportions of acetate and butyrate. An opposite pattern is observed for a block of 55 different species. The α-diversity measures were unaffected by breed (Figure 1), whereas, the β-diversity measures showed clustering according to breed (P = 0.001 for both weighted and P = 0.01 unweighted UniFrac; Figure 2).
      Figure thumbnail gr3
      Figure 3Relative abundance (%) of the major (>1.5% abundance) bacterial species in rumen samples of Holstein and Jersey cows fed concentrate at 49 (C49), 70 (C70), or 91% of dietary DM (C91). Species level taxa groups are labeled by phylum followed by the lowest identified taxonomic level. Species with <1.5% relative abundance are grouped together.

      Effect of Increased Concentrate Proportion

      With increased proportion of concentrate in the diets, the intake of starch increased (P = 0.01) and the intake of NDF tended to decrease (P = 0.09), whereas intake of DM, OM, and CP were unaffected by diet (Table 2). Apparent total-tract digestibility of DM, OM, and NDF decreased with increasing dietary concentrate proportion (P = 0.02, P = 0.02, and P = 0.001, respectively). Digestibility of CP and starch were unaffected by diet. Furthermore, molar proportion of acetate (P < 0.001) decreased and proportion of valerate tended to increase (P = 0.08) with increasing concentrate proportion in the diet. The decrease in acetate proportion is reflected in the reduced A:P ratio and (A+B):P ratio (P = 0.01 and P = 0.02, respectively; Table 3). For proportions of other VFA (i.e., propionate, caproate, butyrate, iso-butyrate, and iso-valerate; tendencies for the latter 3) there was an interaction between diet and breed. Daily methane emission, methane yield, methane intensity, methane energy loss, and CH4:CO2 ratio all decreased with increasing proportion of concentrate in the diet (P = 0.001, P < 0.001, P = 0.01, P < 0.001, and P < 0.001, respectively; Table 4), but for all of these parameters there was an interaction between diet and breed. Hydrogen emission increased with increased concentrate (P = 0.01), whereas carbon dioxide production and oxygen consumption were unaffected by diet. Further, milk yield and milk composition were unaffected by diet, except for milk fat percentage, which was markedly decreased with increasing concentrate proportion (P = 0.01; Table 5). Feed-conversion efficiency also declined with increasing dietary concentrate proportion (P = 0.03). Total eating plus rumination time and the total number of chews during eating and rumination decreased with increased proportion of concentrate (P = 0.001 and P < 0.001, respectively; Table 6). Increasing dietary concentrate proportion decreased the ASV richness and Shannon diversity measure (P < 0.001 for both; Figure 1). The weighted and unweighted UniFrac distances showed clustering according to diet (P = 0.001 for both weighted and unweighted UniFrac; Figure 2).

      DISCUSSION

      Dietary Composition

      The diets were formulated using different types and levels of concentrates with diet C70 being intermediate to diet C49 and C91. Across diets, starch content had a large increase, NDF and digestible NDF (DNDF) content a large decrease, CP and fat content a slight increase, AAT20 a slight decrease, whereas DM and NEL20 content remained constant when dietary concentrate level increased. Thus, observed effects of the diets on variables relates to the NDF-to-starch ratio, DNDF, and AAT20 content of diets, rather than energy density or fat content.

      Methane Emission

      Methane emission expressed on DMI and GEI basis were substantially lowered with increasing dietary concentrate proportion as supported by other studies (
      • Agle M.
      • Hristov A.N.
      • Zaman S.
      • Schneider C.
      • Ndegwa P.M.
      • Vaddella V.K.
      Effect of dietary concentrate on rumen fermentation, digestibility, and nitrogen losses in dairy cows.
      ;
      • Aguerre M.J.
      • Wattiaux M.A.
      • Powell J.M.
      • Broderick G.A.
      • Arndt C.
      Effect of forage-to-concentrate ratio in dairy cow diets on emission of methane, carbon dioxide, and ammonia, lactation performance, and manure excretion.
      ;
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ). It is well known, that addition of dietary fat can also reduce methane emission. However, in the present experiment increased fat level could only account for a minor part of the 48 and 22% reduction in methane yield found for Holstein and Jersey, respectively, because the increase of 6 g of crude fat/kg of DM in diet C91 compared with C49 is expected to give a decrease in methane yield of only about 2% according to a review by
      • Niu M.
      • Kebreab E.
      • Hristov A.N.
      • Oh J.
      • Arndt C.
      • Bannink A.
      • Bayat A.R.
      • Brito A.F.
      • Boland T.
      • Casper D.
      • Crompton L.A.
      • Dijkstra J.
      • Eugène M.A.
      • Garnsworthy P.C.
      • Haque M.N.
      • Hellwing A.L.F.
      • Huhtanen P.
      • Kreuzer M.
      • Kuhla B.
      • Lund P.
      • Madsen J.
      • Martin C.
      • McClelland S.C.
      • McGee M.
      • Moate P.J.
      • Muetzel S.
      • Muñoz C.
      • O'Kiely P.
      • Peiren N.
      • Reynolds C.K.
      • Schwarm A.
      • Shingfield K.J.
      • Storlien T.M.
      • Weisbjerg M.R.
      • Yáñez-Ruiz D.R.
      • Yu Z.
      Prediction of enteric methane production, yield, and intensity in dairy cattle using an intercontinental database.
      , who found a decrease of around 3.5% per additional 10 g of crude fat per kilogram of DM. The increase in CP level of about 1% of DM is not expected to influence methane emission.
      For daily methane emission, there was a clear interaction between breed and diet, because Jersey had the lowest emission on the conventional diet (i.e., C49) compared with Holstein, whereas Holstein had the lowest emission for the diet with extremely high concentrate proportion (i.e., C91) compared with Jersey. Lower daily methane production for Jersey than for Holstein for diets up to 70% of DM concentrate has been reported in previous work on lactating dairy cows (
      • Münger A.
      • Kreuzer M.
      Methane emission as determined in contrasting dairy cattle breeds over the reproduction cycle.
      ;
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ;
      • Uddin M.E.
      • Santana O.I.
      • Weigel K.A.
      • Wattiaux M.A.
      Enteric methane, lactation performances, digestibility, and metabolism of nitrogen and energy of Holsteins and Jerseys fed 2 levels of forage fiber from alfalfa silage or corn silage.
      ) and dairy heifers (
      • Flay H.E.
      • Kuhn-Sherlock B.
      • Macdonald K.A.
      • Camara M.
      • Lopez-Villalobos N.
      • Donaghy D.J.
      • Roche J.R.
      Hot topic: Selecting cattle for low residual feed intake did not affect daily methane production but increased methane yield.
      ). Despite the interaction between diet and breed for daily methane emission and emission as a percentage of GEI, differences in DMI between breeds explain the consequently lower methane emission per kilogram of DMI and as a percentage of GEI for Holstein than Jersey for all diets, which is supported by
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      , but in contrast with other studies on dairy cattle (
      • Münger A.
      • Kreuzer M.
      Methane emission as determined in contrasting dairy cattle breeds over the reproduction cycle.
      ;
      • Uddin M.E.
      • Santana O.I.
      • Weigel K.A.
      • Wattiaux M.A.
      Enteric methane, lactation performances, digestibility, and metabolism of nitrogen and energy of Holsteins and Jerseys fed 2 levels of forage fiber from alfalfa silage or corn silage.
      ) and dairy heifers (
      • Flay H.E.
      • Kuhn-Sherlock B.
      • Macdonald K.A.
      • Camara M.
      • Lopez-Villalobos N.
      • Donaghy D.J.
      • Roche J.R.
      Hot topic: Selecting cattle for low residual feed intake did not affect daily methane production but increased methane yield.
      ). Further, Holstein cows reduced methane yield to a much larger extent than Jersey cows for diet C91 relative to C49 (48 and 22% for Holstein and Jersey, respectively). The decline in methane yield when concentrate level is increased is in line with our previous studies (Figure 4; R2 = 0.95 for each breed;
      • Hellwing A.L.F.
      • Weisbjerg M.R.
      Effect of digestibility of grass-clover silage and concentrate to forage ratio on methane emission from dairy cows.
      ;
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ).
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      reported larger reductions in methane yield and methane energy loss for Holstein cows than Jersey cows for a diet with 32% concentrate relative to 61% concentrate. For a less drastic dietary intervention than in the current study, the Holstein and Jersey cows in the study by
      • Uddin M.E.
      • Santana O.I.
      • Weigel K.A.
      • Wattiaux M.A.
      Enteric methane, lactation performances, digestibility, and metabolism of nitrogen and energy of Holsteins and Jerseys fed 2 levels of forage fiber from alfalfa silage or corn silage.
      responded similarly regarding methane emission to diets differing in forage NDF level and source (alfalfa and corn silage). In addition to methane emission, Holstein and Jersey cows in the current study responded differently to the increase in dietary concentrate level with respect to rumen VFA profile, rumination, and bacterial community structure, whereas no significant interaction between breed and concentrate level were observed for DMI, nutrient intakes, apparent total-tract digestibility of nutrients, milk yield, FCE, eating time and eating chews, as will be described below.
      Figure thumbnail gr4
      Figure 4Methane yield as a function of dietary concentrate proportion for Holstein (closed symbols and solid regression line) and Jersey cows (open symbols and dashed regression line) from the current study (triangles),
      • Hellwing A.L.F.
      • Weisbjerg M.R.
      Effect of digestibility of grass-clover silage and concentrate to forage ratio on methane emission from dairy cows.
      ; diamonds for Holstein fed early-cut grass silage and squares for Holstein fed late-cut grass silage), and
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ; circles); R2 = 0.95 for each breed.

      Volatile Fatty Acids and Nutrient Digestibility

      Reduction in methane emission is generally attributed to shifts in VFA profile and reduced fiber digestibility in the rumen. Indeed, with increasing concentrate proportion and disregarding breeds, increased propionate and decreased acetate molar proportions were observed in connection with lowered total-tract digestibility of DM, OM, and NDF in the current study and agrees with
      • Agle M.
      • Hristov A.N.
      • Zaman S.
      • Schneider C.
      • Ndegwa P.M.
      • Vaddella V.K.
      Effect of dietary concentrate on rumen fermentation, digestibility, and nitrogen losses in dairy cows.
      . Molar proportion of propionate in the rumen increased with increasing concentrate proportion in Holstein, whereas there was no effect in Jersey, and indicates a larger hydrogen consumption facilitated by production of propionate for Holstein cows. Propionate synthesis requires hydrogen, and therefore increased propionate synthesis per se will lead to a larger decrease in methane emission. Nevertheless, there was a significant linear increase in hydrogen emission for Holstein with increasing level of concentrate, whereas for Jersey the increase was not significant, which indicates that there was a larger buildup of hydrogen in Holsteins on diet C91, because of a reduced capacity of the altered rumen microbiome for diet C91 to convert hydrogen into methane in these cows. Moreover, the lack of a linear increase in hydrogen emission for Jersey despite high numerical percentage increments, indicates a high variability and therefore a low power.
      Jersey cows had a higher rumen A:P ratio than Holstein cows for all diets. Similar results were found by
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      , but are in contrast to
      • Uddin M.E.
      • Santana O.I.
      • Weigel K.A.
      • Wattiaux M.A.
      Enteric methane, lactation performances, digestibility, and metabolism of nitrogen and energy of Holsteins and Jerseys fed 2 levels of forage fiber from alfalfa silage or corn silage.
      , who did not find differences in VFA profile between breeds when rumen samples were obtained by rumenocentisis. The greater acetate molar proportion might suggest more digestion of NDF in the rumen for Jersey than for Holstein cows. Even though the apparent total-tract digestibility of NDF did not differ between breeds, a significant linear decrease was observed for Jerseys, whereas this was a tendency for Holsteins. This finding perhaps suggests that Jerseys might respond stronger to increased concentrate proportion regarding decreased NDF digestibility than Holsteins. The apparent total-tract digestibility of DM and other nutrients also did not differ between breeds, despite very low SEM values for digestibilities (except for NDF digestibility). It should be noted that feces were collected from 4 samplings over 48 h and therefore the results on digestibility should be interpreted with care. Nevertheless, the overall lack of difference in digestibility between breeds agrees with
      • Uddin M.E.
      • Santana O.I.
      • Weigel K.A.
      • Wattiaux M.A.
      Enteric methane, lactation performances, digestibility, and metabolism of nitrogen and energy of Holsteins and Jerseys fed 2 levels of forage fiber from alfalfa silage or corn silage.
      , but in contrast to our previous study, where we found that Jersey cows had a higher apparent total-tract digestibility of DM, OM, NDF, and crude fat than Holstein cows (
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ).
      • Aikman P.C.
      • Reynolds C.K.
      • Beever D.E.
      Diet digestibility, rate of passage, and eating and rumination behavior of Jersey and Holstein cows.
      also reported a higher apparent total-tract digestibility of NDF for Jersey compared with Holstein, but not of DM and OM.
      The more pronounced effects of increased level of concentrate in Holsteins than in Jerseys were also observed for the molar proportions of some other VFA in addition to propionate (i.e., the interaction between breed and diet was significant for caproate, and there were tendencies for butyrate, iso-butyrate, and iso-valerate) as well as for α- and β-diversity measures. This jointly indicates that the rumen of Jersey cows was less affected by increased concentrate levels than the rumen of Holstein cows, perhaps related to a better buffering of organic acids and rumen pH related to more intense chewing per kilogram DMI during feed ingestion by Jersey. Another more speculative explanation could be a more regular feed intake pattern over the day by Jersey cows (
      • Aikman P.C.
      • Reynolds C.K.
      • Beever D.E.
      Diet digestibility, rate of passage, and eating and rumination behavior of Jersey and Holstein cows.
      ). Both suggestions imply a lower occurrence of subacute rumen acidosis for Jersey cows than for Holstein cows and could explain why the rumen bacterial community of Jersey cows was less affected by increased dietary concentrate levels. Unfortunately, rumen pH cannot provide reliable information on rumen acidosis when using oro-ruminal sampling devices, because this rumen liquid collection method increases the risk for significant contamination of the sample with saliva making pH and concentrations of organic acids unreliable (
      • Larsen M.
      • Hansen N.P.
      • Weisbjerg M.R.
      • Lund P.
      Technical note: Evaluation of the ororuminal FLORA sampling device for rumen fluid sampling in intact cattle.
      ); however, molar proportions of VFA can be regarded as valid. An increase in propionate molar proportion for diet C91 relative to C49, reduced A:P ratio, and observed milk fat depression, indicates that some degree of subacute rumen acidosis might have occurred in Holstein cows.

      Rumen Bacterial Community Structure

      The lowered NDF digestibility with increasing dietary concentrate level can be attributed to a lower NDF digestibility for barley straw and concentrate, which both have a higher proportion of INDF in total NDF compared with silages and other factors, such as substrate preference by rumen microbes and inhibition of fibrolytic bacteria at lower rumen pH. The latter suggestion can be supported by the trend in lower abundances of the major fibrolytic bacteria (i.e., Fibrobacter succinogenes and Ruminococcus species) as visualized in Figure 3. The bacterial community was evaluated with α- and β-diversity measures. Alpha-diversity is based on the number of features (i.e., ASV richness, closest definition to species) in a sample and additionally the evenness of features in case of the Shannon diversity index, and provides information how diverse the microbial community in a sample is. Beta-diversity shows differences in the microbial community between the treatments. For the weighted UniFrac distances, the relative abundances of ASV are accounted for and therefore mainly show differences in the abundance of dominant species. The unweighted UniFrac distances ignore abundancy of species and thereby provide information on the presence and absence of all species. Dietary concentrate level clearly affected the rumen bacterial community as evident by α- and β-diversity measures, which is in accordance with
      • Noel S.J.
      • Olijhoek D.W.
      • McLean F.
      • Løvendahl P.
      • Lund P.
      • Højberg O.
      Rumen and fecal microbial community structure of Holstein and Jersey dairy cows as affected by breed, diet, and residual feed intake.
      . In other words, the bacterial community was less diverse with increasing dietary concentrate proportion, which can lead to lower methane emission when fibrolytic bacteria become less abundant. Diet has previously been ascribed as the main factor for differences in microbial community structure (
      • Henderson G.
      • Cox F.
      • Ganesh S.
      • Jonker A.
      • Young W.
      • Janssen P.H.
      Global Rumen Census Collaborators
      Rumen microbial community composition varies with diet and host, but a core microbiome is found across a wide geographical range.
      ) and is likely a contributing factor for the distinct patterns of associations seen in correlation of bacterial species with phenotypes observed in Supplemental Figure S1. This figure shows that certain bacterial species are associated with increased hydrogen gas production and hydrogen consuming pathways in the rumen, such as propionate and valerate production, whereas other species are associated with increased methane emission and hydrogen producing pathways, such as acetate and butyrate production. The α-diversity of the rumen bacterial community was not significantly different between Holsteins and Jerseys fed diet C49, which is in contrast with
      • Paz H.A.
      • Anderson C.L.
      • Muller M.J.
      • Kononoff P.J.
      • Fernando S.C.
      Rumen bacterial community composition in Holstein and Jersey cows is different under same dietary condition and is not affected by sampling method.
      , who reported higher richness based on the number of observed operational taxonomic units for Holstein cows than for Jersey cows fed with 49% concentrate of DM. For C91, both α-diversity measures were lower for Holsteins than Jerseys. Both β-diversity measures in the current study were affected by breed, which agrees with
      • Paz H.A.
      • Anderson C.L.
      • Muller M.J.
      • Kononoff P.J.
      • Fernando S.C.
      Rumen bacterial community composition in Holstein and Jersey cows is different under same dietary condition and is not affected by sampling method.
      and
      • Noel S.J.
      • Olijhoek D.W.
      • McLean F.
      • Løvendahl P.
      • Lund P.
      • Højberg O.
      Rumen and fecal microbial community structure of Holstein and Jersey dairy cows as affected by breed, diet, and residual feed intake.
      . More specifically, the abundance of cellulolytic bacteria was reported to differ between breeds (
      • Paz H.A.
      • Anderson C.L.
      • Muller M.J.
      • Kononoff P.J.
      • Fernando S.C.
      Rumen bacterial community composition in Holstein and Jersey cows is different under same dietary condition and is not affected by sampling method.
      ). Overall, these findings indicate that the rumen bacterial community differs in diversity and structure between Holsteins and Jerseys as affected by diet, which can have implications for rumen fermentation. Though, this data set with extreme differences in dietary concentrate proportion gives valuable insight into the relation between specific bacterial species and rumen metabolism, it is beyond the scope of this article to go into detail of specific bacterial species and their functions in rumen fermentation.

      Dry Matter Intake and Milk Production

      Holstein and Jersey differ in body size, which causes a higher intake of DM and nutrients for Holstein than Jersey. Generally, DMI and ECM yield is increased when dietary concentrate proportion is increased to moderately high levels (
      • Xue B.
      • Yan T.
      • Ferris C.F.
      • Mayne C.S.
      Milk production and energy efficiency of Holstein and Jersey-Holstein crossbred dairy cows offered diets containing grass silage.
      ;
      • Huhtanen P.
      • Hetta M.
      Comparison of feed intake and milk production responses in continuous and change-over design dairy cow experiments.
      ;
      • Olijhoek D.W.
      • Løvendahl P.
      • Lassen J.
      • Hellwing A.L.F.
      • Höglund J.K.
      • Weisbjerg M.R.
      • Noel S.J.
      • McLean F.
      • Højberg O.
      • Lund P.
      Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
      ) followed by a lower rumen fill and increased energy intake. In the current study, DMI and ECM yield were not significantly affected by diet, with diet C91 being substantially higher in concentrate proportion than applied in most other studies on dairy cattle. It is worth noting, that care should be taken when interpreting milk yield data due to the experimental design of the current study. Net energy content was equal between diets and in agreement,
      • Sutton J.D.
      • Dhanoa M.S.
      • Morant S.V.
      • France J.
      • Napper D.J.
      • Schuller E.
      Rates of production of acetate, propionate, and butyrate in the rumen of lactating dairy cows given normal and low-roughage diets.
      reported unaffected milk yield when Friesian dairy cows were fed restricted with a diet containing 90% concentrate (barley and soybean meal) compared with 60% concentrate and with similar digestible energy content between diets. Perhaps DMI did not increase in the current study due to rumen imbalance caused by mild rumen acidosis for Holstein cows or the slightly decreasing AAT20 with increasing concentrate proportion. Lowered ruminal pH together with elevated molar proportions of propionate can negatively affect the number of precursors available for lipogenesis in the mammary gland and cause milk fat depression and changes in milk fatty acid profile (
      • Sandri E.C.
      • Lévesque J.
      • Marco A.
      • Couture Y.
      • Gervais R.
      • Rico D.E.
      Transient reductions in milk fat synthesis and their association with the ruminal and metabolic profile in dairy cows fed high-starch, low-fat diets.
      ). Other studies also showed a reduction in milk fat content with increasing dietary concentrate proportion (
      • Sutton J.D.
      • Dhanoa M.S.
      • Morant S.V.
      • France J.
      • Napper D.J.
      • Schuller E.
      Rates of production of acetate, propionate, and butyrate in the rumen of lactating dairy cows given normal and low-roughage diets.
      ;
      • Aguerre M.J.
      • Wattiaux M.A.
      • Powell J.M.
      • Broderick G.A.
      • Arndt C.
      Effect of forage-to-concentrate ratio in dairy cow diets on emission of methane, carbon dioxide, and ammonia, lactation performance, and manure excretion.
      ;
      • Huhtanen P.
      • Hetta M.
      Comparison of feed intake and milk production responses in continuous and change-over design dairy cow experiments.
      ).

      Feeding Behavior

      Measurements of feeding behavior began 12 d after the beginning of experimental feeding (i.e., 8 d after feeding 100% of the experimental diet) for 16 of the 24 cows and 19 d after the beginning for the remaining 8 cows. The adaptation period is considered sufficiently long to rely on the results for rumination and feeding behavior in the present study, because rumination time, feeding time, and feeding rate took up to 4 d to stabilize following a change from a normal lactation diet to an energy-reduced lactation diet (30% of DM dilution with straw) in a recent study by
      • Franchi G.A.
      • Larsen M.L.V.
      • Herskin M.S.
      • Foldager L.
      • Larsen M.
      • Jensen M.B.
      Effects of changes in diet energy density and milking frequency and a single injection of cabergoline at dry-off on feeding behavior and rumination time in dairy cows.
      ; see their graphical abstract). A point to consider when interpreting the results of feeding behavior is that the RumiWatch system is a relatively new development. The validation studies involving the RumiWatch Converter version 0.7.3.2 or earlier showed good performance for rumination time when comparing observations obtained with the RumiWatch system and visually (
      • Kröger I.
      • Humer E.
      • Neubauer V.
      • Kraft N.
      • Ertl P.
      • Zebeli Q.
      Validation of a noseband sensor system for monitoring ruminating activity in cows under different feeding regimens.
      ;
      • Ruuska S.
      • Kajava S.
      • Mughal M.
      • Zehner N.
      • Mononen J.
      Validation of a pressure sensor-based system for measuring eating, rumination and drinking behaviour of dairy cattle.
      ;
      • Rombach M.
      • Münger A.
      • Niederhauser J.
      • Südekum K.-H.
      • Schori F.
      Evaluation and validation of an automatic jaw movement recorder (RumiWatch) for ingestive and rumination behaviors of dairy cows during grazing and supplementation.
      ). However, the number of rumination chews were underestimated (
      • Kröger I.
      • Humer E.
      • Neubauer V.
      • Kraft N.
      • Ertl P.
      • Zebeli Q.
      Validation of a noseband sensor system for monitoring ruminating activity in cows under different feeding regimens.
      ) and a small overestimation of eating time occurred for cows kept in tiestalls (
      • Ruuska S.
      • Kajava S.
      • Mughal M.
      • Zehner N.
      • Mononen J.
      Validation of a pressure sensor-based system for measuring eating, rumination and drinking behaviour of dairy cattle.
      ). These findings should be kept in mind when interpreting the data.
      Holstein and Jersey responded differently to increased dietary concentrate proportion regarding rumination and chewing, whereas eating was unaffected by the interaction between breed and diet. Jersey had a much larger decrease in rumination time and rumination chews per kilogram of DMI and NDF intake for diet C91 than Holstein. Explanations may be found when detailing the effect of diet and breed separately. All 3 expressions of rumination time and rumination chews were reduced with increasing concentrate proportion (diet effect) and is in accordance with
      • Kröger I.
      • Humer E.
      • Neubauer V.
      • Kraft N.
      • Ertl P.
      • Zebeli Q.
      Validation of a noseband sensor system for monitoring ruminating activity in cows under different feeding regimens.
      , who used the RumiWatch system to investigate diets containing 0 and 65% concentrate of dietary DM. Reduced rumination is associated with a reduced physically effective NDF content of high concentrate diets, which also negatively affected rumination and total chewing time (per day, per kilogram of DMI, and per kilogram of NDF intake), but not eating time (
      • Cao Y.
      • Wang D.
      • Wang L.
      • Wei X.
      • Li X.
      • Cai C.
      • Lei X.
      • Yao J.
      Physically effective neutral detergent fiber improves chewing activity, rumen fermentation, plasma metabolites, and milk production in lactating dairy cows fed a high-concentrate diet.
      ). Rumination processes are therefore likely more affected by physical structure of the feed than is ingestion of feed and the decrease in rumination time can negatively affect the rumen environment and nutrient digestibility. Differences between breeds were observed for rumination and eating behavior. Daily rumination time was longer and the daily number of rumination chews were higher for Holstein than Jersey and agrees with
      • Prendiville R.
      • Lewis E.
      • Pierce K.M.
      • Buckley F.
      Comparative grazing behavior of lactating Holstein-Friesian, Jersey, and Jersey × Holstein-Friesian dairy cows and its association with intake capacity and production efficiency.
      . Jersey cows also spent more time eating per unit of feed for all diets during feed ingestion than Holstein cows.
      • Aikman P.C.
      • Reynolds C.K.
      • Beever D.E.
      Diet digestibility, rate of passage, and eating and rumination behavior of Jersey and Holstein cows.
      found that lactating Jersey cows had a 36% longer total chewing time per kilogram of BW than Holstein cows when offered TMR, and that Jersey cows spend more time eating per unit of DMI and NDF intake, but also more time for rumination. In the current study, the number of chews during eating per kilogram of DMI and NDF intake was 29% (average across diets) higher for Jersey cows than for Holstein cows, which was associated with a lower daily DMI and NDF intake for Jersey.
      • Prendiville R.
      • Lewis E.
      • Pierce K.M.
      • Buckley F.
      Comparative grazing behavior of lactating Holstein-Friesian, Jersey, and Jersey × Holstein-Friesian dairy cows and its association with intake capacity and production efficiency.
      reported similar findings for grazing Holstein and Jersey cows, which was attributed to the smaller physical size of Jersey cows. Collectively, these findings indicate a higher degree of mastication and efficient particle size reduction of a given diet during ingestion for Jersey cows than Holstein cows, which would lower the necessity for particle size reduction during rumination for Jersey cows (
      • Beauchemin K.A.
      Invited review: Current perspectives on eating and rumination activity in dairy cows.
      ). Responses in eating and rumination time of Holstein and Jersey to similar diets may profoundly be diet specific (e.g., chemical composition) and could occur when chewing behavior is likely to be affected, such as for diets high in concentrate because of a small particle size. More intense mastication during ingestion and rumination stimulates saliva production, which buffers rumen organic acid production and rumen pH (
      • Beauchemin K.A.
      Invited review: Current perspectives on eating and rumination activity in dairy cows.
      ). Consequently, rumen digestibility of fiber may improve and passage rate may increase. A higher passage rate of particles from the rumen and hence shorter retention times in the rumen and total-tract for Jersey cows than Holstein cows at similar digestibility of OM or increased digestibility of NDF have been reported previously (
      • Ingvartsen K.L.
      • Weisbjerg M.R.
      Jersey cows have a higher feed intake capacity and higher rate of passage than Friesian cows.
      ;
      • Aikman P.C.
      • Reynolds C.K.
      • Beever D.E.
      Diet digestibility, rate of passage, and eating and rumination behavior of Jersey and Holstein cows.
      ) and indicate a higher rate of digestion for Jersey cows. Overall, this efficient digestive process for Jersey cows together with more intense mastication during eating can possibly explain the higher methane emission per unit of DMI compared with Holstein cows facilitated by higher (A + B):P ratio. In general, differential responses in feeding behavior, methane emission, and rumen and digestive processes between Holstein and Jersey cows to similar diets is little explored to date and warrants further investigation.

      CONCLUSIONS

      Diets containing concentrates up to 91% of dietary DM were effective in lowering enteric methane emission and Holstein cows responded stronger than Jersey cows to increased dietary concentrate level regarding methane mitigation. Therefore, feeding diets high in concentrates are a less effective methane mitigation strategy for Jersey cows than for Holstein cows. In case of shortage in the availability of roughage, diets high in concentrates might be suitable for Jersey cows, because their rumen environment is less affected; however, Jersey cows responded stronger with regard to rumination than Holstein cows. Despite that the diet with concentrates at 91% of DM was a very efficient methane mitigation strategy for Holstein, it cannot be recommended as a long-term strategy due to the risk of rumen acidosis, as indicated by a high propionate molar proportion and low A:P ratio.

      ACKNOWLEDGMENTS

      The authors thank the barn staff, technicians (Torkild N. Jakobsen, Ester Bjerregaard, and Amin T. Aljundi), and laboratory technicians at AU Foulum (Tjele, Denmark) for the much appreciated work in the barn and laboratory. Leslie Foldager (AU Foulum, Tjele, Denmark) is gratefully acknowledged for statistical support and Nikolaj Peder Hansen for guidance on the RumiWatch system (AU Foulum, Tjele, Denmark). This project received funding from the Danish Milk Levy Fund (Aarhus, Denmark) and Aarhus University (Aarhus, Denmark). The authors have not stated any conflicts of interest.

      REFERENCES

        • Agle M.
        • Hristov A.N.
        • Zaman S.
        • Schneider C.
        • Ndegwa P.M.
        • Vaddella V.K.
        Effect of dietary concentrate on rumen fermentation, digestibility, and nitrogen losses in dairy cows.
        J. Dairy Sci. 2010; 93 (20723695): 4211-4222
        • Aguerre M.J.
        • Wattiaux M.A.
        • Powell J.M.
        • Broderick G.A.
        • Arndt C.
        Effect of forage-to-concentrate ratio in dairy cow diets on emission of methane, carbon dioxide, and ammonia, lactation performance, and manure excretion.
        J. Dairy Sci. 2011; 94 (21605777): 3081-3093
        • Aikman P.C.
        • Reynolds C.K.
        • Beever D.E.
        Diet digestibility, rate of passage, and eating and rumination behavior of Jersey and Holstein cows.
        J. Dairy Sci. 2008; 91 (18292266): 1103-1114
        • ANKOM
        Analytical methods. ANKOM Technology. Analytical Methods, ADF NDF and Crude Fiber, Automated Fiber Analyzer. ANKOM Technology.
        • AOAC International
        AOAC Official Methods of Analysis.
        17th ed. AOAC International, 2000
        • Beauchemin K.A.
        Invited review: Current perspectives on eating and rumination activity in dairy cows.
        J. Dairy Sci. 2018; 101 (29627250): 4762-4784
        • Beecher M.
        • Buckley F.
        • Waters S.M.
        • Boland T.M.
        • Enriquez-Hidalgo D.
        • Deighton M.H.
        • O'Donovan M.
        • Lewis E.
        Gastrointestinal tract size, total-tract digestibility, and rumen microflora in different dairy cow genotypes.
        J. Dairy Sci. 2014; 97 (24704226): 3906-3917
        • Bokulich N.A.
        • Kaehler B.D.
        • Rideout J.R.
        • Dillon M.
        • Bolyen E.
        • Knight R.
        • Huttley G.A.
        • Caporaso J.G.
        Optimizing taxonomic classification of marker-gene amplicon sequences with QIIME 2's q2-feature-classifier plugin.
        Microbiome. 2018; 6: 90
        • Bolyen E.
        • Rideout J.R.
        • Dillon M.R.
        • Bokulich N.A.
        • Abnet C.C.
        • Al-Ghalith G.A.
        • Alexander H.
        • Alm E.J.
        • Arumugam M.
        • Asnicar F.
        • Bai Y.
        • Bisanz J.E.
        • Bittinger K.
        • Brejnrod A.
        • Brislawn C.J.
        • Brown C.T.
        • Callahan B.J.
        • Caraballo-Rodriguez A.M.
        • Chase J.
        • Cope E.
        • Da Silva R.
        • Dorrestein P.C.
        • Douglas G.M.
        • Durall D.M.
        • Duvallet C.
        • Edwardson C.F.
        • Ernst M.
        • Estaki M.
        • Fouquier J.
        • Gauglitz J.M.
        • Gibbons S.M.
        • Gibson D.L.
        • Gonzalez A.
        • Gorlick K.
        • Guo J.
        • Hillmann B.
        • Holmes S.
        • Holste H.
        • Huttenhower C.
        • Huttley G.A.
        • Janssen S.
        • Jarmusch A.K.
        • Jiang L.
        • Kaehler B.D.
        • Kang K.B.
        • Keefe C.R.
        • Keim P.
        • Kelley S.T.
        • Knights D.
        • Koester I.
        • Kosciolek T.
        • Kreps J.
        • Langille M.G.I.
        • Lee J.
        • Ley R.
        • Liu Y.-X.
        • Loftfield E.
        • Lozupone C.
        • Maher M.
        • Marotz C.
        • Martin B.D.
        • McDonald D.
        • McIver L.J.
        • Melnik A.V.
        • Metcalf J.L.
        • Morgan S.C.
        • Morton J.T.
        • Naimey A.T.
        • Navas-Molina J.A.
        • Nothias L.F.
        • Orchanian S.B.
        • Pearson T.
        • Peoples S.L.
        • Petras D.
        • Preuss M.L.
        • Pruesse E.
        • Rasmussen L.B.
        • Rivers A.
        • Robeson II, M.S.
        • Rosenthal P.
        • Segata N.
        • Shaffer M.
        • Shiffer A.
        • Sinha R.
        • Song S.J.
        • Spear J.R.
        • Swafford A.D.
        • Thompson L.R.
        • Torres P.J.
        • Trinh P.
        • Tripathi A.
        • Turnbaugh P.J.
        • Ul-Hasan S.
        • van der Hooft J.J.J.
        • Vargas F.
        • Vázquez-Baeza Y.
        • Vogtmann E.
        • von Hippel M.
        • Walters W.
        • Wan Y.
        • Wang M.
        • Warren J.
        • Weber K.C.
        • Williamson C.H.D.
        • Willis A.D.
        • Xu Z.Z.
        • Zaneveld J.R.
        • Zhang Y.
        • Zhu Q.
        • Knight R.
        • Caporaso J.G.
        Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2.
        Nat. Biotechnol. 2019; 37: 852-857
        • Børsting C.F.
        • Hellwing A.L.F.
        • Weisbjerg M.R.
        • Østergaard S.
        • Raun B.L.
        • Røjen B.A.
        • Kristensen N.B.
        Feeding dairy cows with no or reduced amounts of forage.
        in: Proceedings of the 10th Nordic Feed Science Conference, Swedish University of Agricultural Sciences, Uppsala, Sweden. Department of Animal Nutrition and Management, Swedish University of Agricultural Sciences, 2019: 87-91
        • Brask M.
        • Lund P.
        • Hellwing A.L.F.
        • Poulsen M.
        • Weisbjerg M.R.
        Enteric methane production, digestibility and rumen fermentation in dairy cows fed different forages with and without rapeseed fat supplementation.
        Anim. Feed Sci. Technol. 2013; 184: 67-79
        • Callahan B.J.
        • McMurdie P.J.
        • Rosen M.J.
        • Han A.W.
        • Johnson A.J.A.
        • Holmes S.P.
        DADA2: High-resolution sample inference from Illumina amplicon data.
        Nat. Methods. 2016; 13 (27214047): 581-583
        • Cao Y.
        • Wang D.
        • Wang L.
        • Wei X.
        • Li X.
        • Cai C.
        • Lei X.
        • Yao J.
        Physically effective neutral detergent fiber improves chewing activity, rumen fermentation, plasma metabolites, and milk production in lactating dairy cows fed a high-concentrate diet.
        J. Dairy Sci. 2021; 104 (33663818): 5631-5642
        • Ferris C.P.
        • Gordon F.J.
        • Patterson D.C.
        • Porter M.G.
        • Yan T.
        The effect of genetic merit and concentrate proportion in the diet on nutrient utilization by lactating dairy cows.
        J. Agric. Sci. 1999; 132: 483-490
        • Flay H.E.
        • Kuhn-Sherlock B.
        • Macdonald K.A.
        • Camara M.
        • Lopez-Villalobos N.
        • Donaghy D.J.
        • Roche J.R.
        Hot topic: Selecting cattle for low residual feed intake did not affect daily methane production but increased methane yield.
        J. Dairy Sci. 2019; 102 (30639015): 2708-2713
        • Franchi G.A.
        • Larsen M.L.V.
        • Herskin M.S.
        • Foldager L.
        • Larsen M.
        • Jensen M.B.
        Effects of changes in diet energy density and milking frequency and a single injection of cabergoline at dry-off on feeding behavior and rumination time in dairy cows.
        JDS Commun. 2022; 3: 195-200
        • Geishauser T.
        • Linhart N.
        • Neidl A.
        • Reimann A.
        Factors associated with ruminal pH at herd level.
        J. Dairy Sci. 2012; 95 (22818470): 4556-4567
        • Hansen B.
        Determination of nitrogen as elementary-N, an alternative to Kjeldahl.
        Acta Agric. Scand. A Anim. Sci. 1989; 39: 113-118
        • Hellwing A.L.F.
        • Lund P.
        • Weisbjerg M.R.
        • Brask M.
        • Hvelplund T.
        Technical note: Test of a low-cost and animal-friendly system for measuring methane emissions from dairy cows.
        J. Dairy Sci. 2012; 95 (22901487): 6077-6085
        • Hellwing A.L.F.
        • Weisbjerg M.R.
        Effect of digestibility of grass-clover silage and concentrate to forage ratio on methane emission from dairy cows.
        in: Proceedings of the 4th Greenhouse Gases and Animal Agriculture Conference, Banff, Canada. Agriculture and Agri-Food Canada, 2010: 120-121
        • Henderson G.
        • Cox F.
        • Ganesh S.
        • Jonker A.
        • Young W.
        • Janssen P.H.
        • Global Rumen Census Collaborators
        Rumen microbial community composition varies with diet and host, but a core microbiome is found across a wide geographical range.
        Sci. Rep. 2015; 514567
        • Huhtanen P.
        • Hetta M.
        Comparison of feed intake and milk production responses in continuous and change-over design dairy cow experiments.
        Livest. Sci. 2012; 143: 184-194
        • Ingvartsen K.L.
        • Weisbjerg M.R.
        Jersey cows have a higher feed intake capacity and higher rate of passage than Friesian cows.
        Arch. Tierz. 1993; 36: 495-498
        • Katoh K.
        • Standley D.M.
        MAFFT multiple sequence alignment software version 7: Improvements in performance and usability.
        Mol. Biol. Evol. 2013; 30 (23329690): 772-780
        • Klindworth A.
        • Pruesse E.
        • Schweer T.
        • Peplies J.
        • Quast C.
        • Horn M.
        • Glöckner F.O.
        Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies.
        Nucleic Acids Res. 2013; 41 (22933715): e1
        • Kristensen N.B.
        • Danfær A.
        • Tetens V.
        • Agergaard N.
        Portal recovery of intraruminally infused short-chain fatty acids in sheep.
        Acta Agric. Scand. A Anim. Sci. 1996; 46: 26-38
        • Kristensen N.B.
        • Storm A.
        • Raun B.M.L.
        • Røjen B.A.
        • Harmon D.L.
        Metabolism of silage alcohols in lactating dairy cows.
        J. Dairy Sci. 2007; 90 (17297111): 1364-1377
        • Kröger I.
        • Humer E.
        • Neubauer V.
        • Kraft N.
        • Ertl P.
        • Zebeli Q.
        Validation of a noseband sensor system for monitoring ruminating activity in cows under different feeding regimens.
        Livest. Sci. 2016; 193: 118-122
        • Larsen M.
        • Hansen N.P.
        • Weisbjerg M.R.
        • Lund P.
        Technical note: Evaluation of the ororuminal FLORA sampling device for rumen fluid sampling in intact cattle.
        J. Dairy Sci. 2020; 103 (31733868): 447-450
        • Lozupone C.
        • Knight R.
        UniFrac: A new phylogenetic method for comparing microbial communities.
        Appl. Environ. Microbiol. 2005; 71 (16332807): 8228-8235
        • Lozupone C.A.
        • Hamady M.
        • Kelley S.T.
        • Knight R.
        Quantitative and qualitative β diversity measures lead to different insights into factors that structure microbial communities.
        Appl. Environ. Microbiol. 2007; 73 (17220268): 1576-1585
        • McDonald D.
        • Price M.N.
        • Goodrich J.
        • Nawrocki E.P.
        • DeSantis T.Z.
        • Probst A.
        • Andersen G.L.
        • Knight R.
        • Hugenholtz P.
        An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea.
        ISME J. 2012; 6 (22134646): 610-618
        • Mertens D.R.
        Gravimetric determination of amylase-treated neutral detergent fiber in feeds with refluxing in beakers or crucibles: Collaborative study.
        J. AOAC Int. 2002; 85 (12477183): 1217-1240
        • Münger A.
        • Kreuzer M.
        Methane emission as determined in contrasting dairy cattle breeds over the reproduction cycle.
        Int. Congr. Ser. 2006; 1293: 119-122
        • Myers W.D.
        • Ludden P.A.
        • Nayigihugu V.
        • Hess B.W.
        Technical note: A procedure for the preparation and quantitative analysis of samples for titanium dioxide.
        J. Anim. Sci. 2004; 82 (14753360): 179-183
        • Niu M.
        • Kebreab E.
        • Hristov A.N.
        • Oh J.
        • Arndt C.
        • Bannink A.
        • Bayat A.R.
        • Brito A.F.
        • Boland T.
        • Casper D.
        • Crompton L.A.
        • Dijkstra J.
        • Eugène M.A.
        • Garnsworthy P.C.
        • Haque M.N.
        • Hellwing A.L.F.
        • Huhtanen P.
        • Kreuzer M.
        • Kuhla B.
        • Lund P.
        • Madsen J.
        • Martin C.
        • McClelland S.C.
        • McGee M.
        • Moate P.J.
        • Muetzel S.
        • Muñoz C.
        • O'Kiely P.
        • Peiren N.
        • Reynolds C.K.
        • Schwarm A.
        • Shingfield K.J.
        • Storlien T.M.
        • Weisbjerg M.R.
        • Yáñez-Ruiz D.R.
        • Yu Z.
        Prediction of enteric methane production, yield, and intensity in dairy cattle using an intercontinental database.
        Glob. Chang. Biol. 2018; 24 (29450980): 3368-3389
        • Noel S.J.
        • Olijhoek D.W.
        • McLean F.
        • Løvendahl P.
        • Lund P.
        • Højberg O.
        Rumen and fecal microbial community structure of Holstein and Jersey dairy cows as affected by breed, diet, and residual feed intake.
        Animals. 2019; 9: 498
        • Nousiainen J.
        • Rinne M.
        • Huhtanen P.
        A meta-analysis of feed digestion in dairy cows. 1. The effects of forage and concentrate factors on total diet digestibility.
        J. Dairy Sci. 2009; 92 (19762821): 5019-5030
        • Olijhoek D.W.
        • Hellwing A.L.F.
        • Noel S.J.
        • Lund P.
        • Larsen M.
        • Weisbjerg M.R.
        • Børsting C.F.
        Feeding up to 91% concentrate to Holstein and Jersey dairy cows: Effects on enteric methane emission, rumen fermentation and bacterial community, digestibility, production, and feeding behavior.
        J. Dairy Sci. 2022;
        • Olijhoek D.W.
        • Løvendahl P.
        • Lassen J.
        • Hellwing A.L.F.
        • Höglund J.K.
        • Weisbjerg M.R.
        • Noel S.J.
        • McLean F.
        • Højberg O.
        • Lund P.
        Methane production, rumen fermentation, and diet digestibility of Holstein and Jersey dairy cows being divergent in residual feed intake and fed at 2 forage-to-concentrate ratios.
        J. Dairy Sci. 2018; 101 (30197132): 9926-9940
        • Paz H.A.
        • Anderson C.L.
        • Muller M.J.
        • Kononoff P.J.
        • Fernando S.C.
        Rumen bacterial community composition in Holstein and Jersey cows is different under same dietary condition and is not affected by sampling method.
        Front. Microbiol. 2016; 71206
        • Prendiville R.
        • Lewis E.
        • Pierce K.M.
        • Buckley F.
        Comparative grazing behavior of lactating Holstein-Friesian, Jersey, and Jersey × Holstein-Friesian dairy cows and its association with intake capacity and production efficiency.
        J. Dairy Sci. 2010; 93 (20105548): 764-774
        • Price M.N.
        • Dehal P.S.
        • Arkin A.P.
        FastTree 2 - approximately maximum-likelihood trees for large alignments.
        PLoS One. 2010; 5 (20224823)e9490
        • R Core Team
        R: A language and environment for statistical computing.
        R Foundation for Statistical Computing, 2019
        • Rojas-Downing M.M.
        • Nejadhashemi A.P.
        • Harrigan T.
        • Woznicki S.A.
        Climate change and livestock: Impacts, adaptation, and mitigation.
        Clim. Risk Manage. 2017; 16: 145-163
        • Rombach M.
        • Münger A.
        • Niederhauser J.
        • Südekum K.-H.
        • Schori F.
        Evaluation and validation of an automatic jaw movement recorder (RumiWatch) for ingestive and rumination behaviors of dairy cows during grazing and supplementation.
        J. Dairy Sci. 2018; 101 (29290426): 2463-2475
        • Ruuska S.
        • Kajava S.
        • Mughal M.
        • Zehner N.
        • Mononen J.
        Validation of a pressure sensor-based system for measuring eating, rumination and drinking behaviour of dairy cattle.
        Appl. Anim. Behav. Sci. 2016; 174: 19-23
        • Sandri E.C.
        • Lévesque J.
        • Marco A.
        • Couture Y.
        • Gervais R.
        • Rico D.E.
        Transient reductions in milk fat synthesis and their association with the ruminal and metabolic profile in dairy cows fed high-starch, low-fat diets.
        Animal. 2020; 14 (32638696): 2523-2534
        • Sauvant D.
        • Giger-Reverdin S.
        Modélisation des interactions digestives et de la production de méthane chez les ruminants.
        INRA Prod. Anim. 2009; 22 (In French): 375-384
        • Sjaunja L.O.
        • Baevre L.
        • Junkkarinen L.
        • Pedersen J.
        • Setala J.
        A Nordic proposal for an energy corrected milk (ECM) formula.
        in: EAAP Publication 50: Performance Recording of Animals: State of the Art 1990. Performance Centre for Agricultural Publishing and Documentation (PUDOC), 1991: 156-157
        • Stoldt O.W.
        Vorschlag zur vereinheitlichung der fettbestimmung in lebensmitteln.
        Eur. J. Lipid Sci. Technol. 1952; 54 ([In German]): 206-207
        • Sutton J.D.
        • Dhanoa M.S.
        • Morant S.V.
        • France J.
        • Napper D.J.
        • Schuller E.
        Rates of production of acetate, propionate, and butyrate in the rumen of lactating dairy cows given normal and low-roughage diets.
        J. Dairy Sci. 2003; 86 (14672193): 3620-3633
        • Uddin M.E.
        • Santana O.I.
        • Weigel K.A.
        • Wattiaux M.A.
        Enteric methane, lactation performances, digestibility, and metabolism of nitrogen and energy of Holsteins and Jerseys fed 2 levels of forage fiber from alfalfa silage or corn silage.
        J. Dairy Sci. 2020; 103 (32389470): 6087-6099
        • Volden H.
        NorFor – The Nordic Feed Evaluation System. EAAP Publication No. 130.
        Wageningen Academic Publishers, 2011
        • Volden H.
        • Larsen M.
        Digestion and metabolism in the gastrointestinal tract.
        in: Volden H. NorFor – The Nordic Feed Evaluation System. EAAP Publication No. 130. Wageningen Academic Publishers, 2011: 59-80
        • Volden H.
        • Nielsen N.I.
        Energy and metabolizable protein supply.
        in: Volden H. NorFor – The Nordic Feed Evaluation System. EAAP Publication No. 130. Wageningen Academic Publishers, 2011: 81-84
        • Warnes G.R.
        • Bolker B.
        • Bonebakker L.
        • Gentleman R.
        • Liaw W.H.A.
        • Lumley T.
        • Maechler M.
        • Magnusson A.
        • Moeller S.
        • Schwartz M.
        • Venables B.
        gplots: Various R programming tools for plotting data. R package version 3.0.1.1.
        • Xue B.
        • Yan T.
        • Ferris C.F.
        • Mayne C.S.
        Milk production and energy efficiency of Holstein and Jersey-Holstein crossbred dairy cows offered diets containing grass silage.
        J. Dairy Sci. 2011; 94 (21338810): 1455-1464
        • Zehner N.
        • Niederhauser J.J.
        • Nydegger F.
        • Grothmann A.
        • Keller M.
        • Hoch M.
        • Haeussermann A.
        • Schick M.
        Validation of a new health monitoring system (RumiWatch) for combined automatic measurement of rumination, feed intake, water intake and locomotion in dairy cows.
        in: Proc. Int. Conf. Agric. Eng. CIGR-AgEng, Valencia, Spain. 2012: C-0438
        • Zehner N.
        • Umstätter C.
        • Niederhauser J.J.
        • Schick M.
        System specification and validation of a noseband pressure sensor for measurement of ruminating and eating behavior in stable-fed cows.
        Comput. Electron. Agric. 2017; 136: 31-41