Published December 5, 2014
Effect of dl-malic acid supplementation on feed intake, methane emission, and rumen fermentation in beef cattle P. A. Foley,1 D. A. Kenny, J. J. Callan, T. M. Boland, and F. P. O’Mara University College Dublin (UCD) School of Agriculture, Food Science and Veterinary Medicine, UCD, Lyons Research Farm, Newcastle, Co. Dublin, Ireland
ABSTRACT: The objective of this study was to determine the effect of dietary concentration of dl-malic acid (MA) on DMI, CH4 emission, and rumen fermentation in beef cattle. Two Latin square experiments were conducted. In Exp. 1, six beef heifers (19 ± 1 mo old) were assigned in a duplicated Latin square to 1 of 3 dietary concentrations of MA on a DMI basis (0%, MA-0; 3.75%, MA-3.75; or 7.5%, MA-7.5) over 3 periods. In Exp. 2, four rumen-fistulated steers (48 ± 1 mo old) were assigned to 1 of 4 dietary concentrations of MA (0%, MA-0; 2.5%, MA-2.5; 5.0%, MA-5.0; or 7.5%, MA-7.5) on a DMI basis, over 4 periods. Both experimental diets consisted of grass silage and pelleted concentrate (containing MA). Silage was fed ad libitum once daily (a.m.), whereas concentrate was fed twice daily (a.m. and p.m.) with the aim of achieving a total DMI of 40:60 silage:concentrate. In both Exp. 1 and 2, experimental periods consisted of 28 d, incorporating a 13-d acclimatization, a 5-d measurement period, and a 10-d washout period. In Exp. 1, enteric CH4, feed apparent digestibility, and feed intake were measured over the 5-d measurement period. In Exp. 2, rumen fluid was collected on d 16 to 18, immediately before (a.m.)
feeding and 2, 4, 6, and 8 h thereafter. Rumen pH was determined and samples were taken for protozoa count, VFA, and ammonia analysis. Enteric CH4 emissions were estimated by using the sulfur hexafluoride tracer technique and feed apparent digestibility was estimated by using chromic oxide as an external marker for fecal output. In Exp. 1, increasing dietary MA led to a linear decrease in total DMI (P < 0.001) and total daily CH4 emissions (P < 0.001). Compared with the control diet, the greatest concentration of MA decreased total daily CH4 emissions by 16%, which corresponded to a 9% reduction per unit of DMI. Similarly, in Exp. 2, inclusion of MA reduced DMI in a linear (P = 0.002) and quadratic (P < 0.001) fashion. Increasing dietary MA led to a linear decrease in molar proportion of acetic (P = 0.004) and butyric acids (P < 0.001) and an increase in propionic acid (P < 0.001). Ruminal pH tended to increase (P = 0.10) with increasing dietary MA. Dietary inclusion of MA led to a linear (P = 0.01) decrease in protozoa numbers. Increasing supplementation with MA decreased CH4 emissions, but DMI was also decreased, which could have potentially negative effects on animal performance.
Key words: beef cattle, dl-malic acid, methane, rumen ©2009 American Society of Animal Science. All rights reserved.
INTRODUCTION Globally, ruminant livestock produce approximately 80 million t of CH4 annually, accounting for approximately 28% of anthropogenic emissions (Beauchemin et al., 2008). Agriculturally derived greenhouse gas (GHG) emissions total 26.4% of the national GHG emissions in Ireland, of which enteric fermentation is responsible for 49%, with enteric fermentation alone accounting for almost 13% of the total GHG emissions of Ireland (McGettigan et al., 2008). Production of CH4
1 Corresponding author:
[email protected] Received March 11, 2008. Accepted November 14, 2008.
J. Anim. Sci. 2009. 87:1048–1057 doi:10.2527/jas.2008-1026
has long been regarded as an inefficiency in the ruminant digestive process, representing a loss of ingested GE of between 2 and 15% (Van Nevel and Demeyer, 1996). The potential of dicarboxylic organic acids (OA) such as fumaric acid (FA) and malic acid (MA) as inhibitors of methanogenesis is well documented in vitro (Carro and Ranilla, 2003; Newbold et al., 2005) and in vivo (Lila et al., 2004; Wallace et al., 2006). These OA are propionate precursors in the succinate-propionate pathway, and by acting as alternative H2 sinks in the rumen, they have the potential to decrease ruminal methanogenesis (Newbold et al., 2005). In vivo responses in CH4 production after OA supplementation have been variable. For example, Wallace et al. (2006) reported CH4 reductions of up to 75% in
1048
dl-Malic acid supplementation to beef cattle
lambs offered FA; similarly, Lila et al. (2004) reported 18% less CH4 emission from steers offered β-cyclodextrin diallyl maleate. Furthermore, other in vivo studies have failed to establish any effect of OA supplementation in either beef heifers (Beauchemin and McGinn, 2006) or dairy cows (McCourt et al., 2008). Thus, in contrast to the well-documented CH4 production response to OA in vitro, responses to dietary supplementation in vivo remain inconclusive. The objective of the current study was to assess the feed intake, CH4 emission, and ruminal fermentation responses to dl-MA supplementation in beef cattle. The study consisted of 2 separate but linked experiments.
MATERIALS AND METHODS All procedures described in this experiment were conducted under experimental license from the Irish Department of Health in accordance with the Cruelty to Animals Act 1876 and the European Communities (Amendments of the Cruelty to Animals Act 1976) Regulations, 1994.
Animals, Experimental Design, and Treatments Exp. 1. This experiment was designed as a duplicated (n = 2) Latin square design with 3 treatments of varied MA supplementation and 3 periods. Six Charolais cross heifers, 452 ± 25 kg of BW and 19 ± 1 mo old, were blocked according to BW and assigned at random according to BW from within blocks to 1 of 3 treatments in each period, with 2 heifers per treatment per period. The basal diet consisted of forage (grass silage) and concentrate on a 40:60 proportion (DM basis). The concentrate portion of the diet was supplemented with MA to give final concentrations of MA in the diet of 0% (MA-0), 3.75% (MA-3.75), and 7.5% (MA-7.5), on a DM basis. Exp. 2. This experiment was also arranged in a Latin square design with 4 treatments of varied MA supplementation and 4 periods, and was carried out in parallel with Exp. 1. Four rumen-fistulated Friesian steers, 895 ± 45 kg of BW and 48 ± 1 mo old, were assigned at random to 1 of the 4 treatments in each period. The basal diet was the same as for Exp. 1 and was prepared to provide final concentrations of MA in the diet of 0% (MA-0), 2.5% (MA-2.5), 5% (MA-5.0), and 7.5% (MA-7.5), on a DM basis. Within both experiments, each experimental period lasted 28 d (d 1 to 28), which included 13 d for acclimatization, 5 d for measurement (d 14 to 18), and 10 d for a washout period. In Exp. 1, enteric CH4, feed apparent digestibility, and feed intake were measured over 5 d. In Exp. 2, feed intake was measured over d 14 to 18, whereas rumen fluid (RF) sampling took place over d 16 to 18. Animals were housed in individual stalls from d 1 to 18 and group housed in slatted floor pens over d 19 to 28 of the experimental periods.
1049
Feeds and Feeding Previous studies (Wallace et al., 2006; Castillo et al., 2007) have reported negative effects on feed intake when OA were included in ruminant diets, probably caused by poor palatability. In this study, to minimize palatability issues and feed sorting, the concentrate portion of the diet was pelleted. Two separate concentrates were prepared and pelleted (Table 1). One of the concentrates contained no MA, whereas the other concentrate contained 12.5% MA, on a DM basis. In Exp. 1, animals received either the no-MA concentrate only, the no-MA concentrate and 12.5% MA concentrate in a 50:50 ratio, or the 12.5% MA concentrate only as the concentrate portion of the diet. Thus, because concentrates constituted 60% of the daily DM allowance, the MA inclusion rates of treatments in Exp. 1 were 0% (MA-0), 3.75% (MA-3.75), and 7.5% (MA-7.5) of DMI. Similarly, in Exp. 2, the MA-0, MA-2.5, MA-5.0, and MA-7.5 dietary treatments were achieved by allowances of the no-MA concentrate only, the no-MA concentrate and 12.5% MA concentrate in ratios of 1:2 and 2:1, or the 12.5% MA concentrate only, respectively, as the concentrate portion of the diets. The no-MA and 12.5% MA concentrates were formulated to be isonitrogenous and isoenergetic. Ingredient and chemical compositions of the concentrates as well as of the grass silage are shown in Table 1. Silage was fed once daily at 0830 h (Exp. 1) or 0900 h (Exp. 2) by using a Calan Data Ranger feeder (American Calan, Northwood, NH). Daily feed allowances of individual animals were weighed with an Avery Weigh-Tronix continuous weigh system (Avery Berkel Ireland, Dublin, Ireland), offering between 110 and 120% of the feed intakes recorded the previous day. Concentrates were fed twice daily at 0830 and 1700 h (Exp. 1) or 0900 and 1730 h (Exp. 2) in the same feeding trough as the silage. The amount of concentrates offered was adjusted daily to maintain a constant forage:concentrate ratio.
Measurements and Sampling Feed Intake in Exp. 1 and 2. In both experiments, feed intake was recorded electronically with an Avery Weigh-Tronix continuous weigh system, which determined feed intake as the difference between the weight of daily feed offered and the weight of daily feed refused. Feed refusals were separated into forage and concentrate components by using a wire mesh sieve and were individually weighed. Over d 12 to 18 of each experimental period, samples were taken of fresh silage and concentrate, compiled within animals for the period of measurement, and stored for subsequent analysis. Feed refusals were sampled daily, dried, bulked within animal per period of measurement, and stored for subsequent analysis. Dry matter of forage offered was determined daily by oven-drying at 55°C for 72 h. Forage pH was measured daily with a Mettler Toledo MP 200 pH meter (Mettler Toledo Ltd., Essex, UK).
1050
Foley et al.
Table 1. Ingredient (g/kg, as fed) and chemical composition (g/kg of DM) of the pelleted concentrates (no malic acid and 12.5% malic acid) and grass silage Concentrate Item
No malic acid
12.5% malic acid
Grass silage
440 230 152 93 — 20 50 15
328 214 161 87 125 20 50 15
— — — — — — — —
876 141 81 219 119 13 63 — — —
853 137 76 226 110 13 61 — — —
169 196 — 674 429 55 98 1,528 38.4 5.41
Ingredient Barley Citrus pulp Corn gluten feed Soybean meal Malic acid Soybean oil Molasses Mineral-vitamin mixture1 Chemical composition DM CP Crude fiber NDF ADF ADL Ash Buffering capacity, mEq/kg of DM NH3-N, g/kg of total N pH
1 Mineral-vitamin premix contained vitamin A, 300,000 IU/kg; vitamin D3, 75,000 IU/kg; and vitamin E (α-tocopherol) 2,000 mg/kg. Contained the following (mg/kg): monocalcium phosphate (546), sodium chloride (218), magnesium oxide (102), calcium carbonate (70), molasses (30), iodine (as anhydrous calcium iodate; 600), copper (as cupric sulfate; 3,000); zinc (as zinc oxide; 5,000); manganese (as manganous oxide; 2,000); cobalt (as cobaltous carbonate; 150); and selenium (as sodium selenite; 45). Contained B vitamins.
Enteric CH4 Emission (Exp. 1). Methane emissions were measured from d 14 to 18 of each period by using a modification of the sulfur hexafluoride (SF6) tracer gas technique of Johnson et al. (1994) as described previously by Lovett et al. (2003), with canisters designed to half fill during a 24-h period. The technique was modified for stall-housed animals; the collection canisters were located remotely from the animal. An expanded metal filter was inserted into the plastic sheath covering the 15-µm filter located over the nose of the animal to prevent blocking of the breath sampling point by feed particles. A permeation tube containing SF6 was placed into the rumen of each animal approximately 2 wk before the measurements reported here. The permeation tubes were manufactured on site and were filled with >1 g of SF6. The average release rate was 1,546 ± 62 ng/min, which was determined over an 11-wk preexperimental period by weighing each permeation tube at the same time weekly. For each day during the measurements, background concentrations of CH4 and SF6 were measured by placing 5 sampling kits (identical to those used on the animals) at strategic locations in the house, which was naturally ventilated. The CH4 and SF6 concentrations in the animal samples were subsequently adjusted for their background concentrations. Gas concentrations of both CH4 and SF6 within the collection canisters were determined by gas chromatography (Varian 3800; Varian, Mulgrave, Australia). The gas chromatograph was calibrated daily by using 3 Na-
tional Institute of Standards and Technology-certified standards (Scott Marin Gases, Riverside, CA) of both CH4 and SF6, with internal verifications for both gases run every 12 injections. Each collection canister was analyzed in duplicate, with sample injected (50°C) via a 1-mL sample loop. Once injected, the sample was then split (approximate ratio 1:2) for the determination of CH4 and SF6 (Johnson et al., 1994), thus enabling the simultaneous determination of CH4 with a flameionization detector (250°C) and a 3.18 mm × 1.22 m stainless steel Porapak N column, 80 to 100 mesh, and SF6 concentration with an Electron Capture Detector (300°C) with a 3.18 mm × 1.83 m stainless steel column packed with a molecular sieve (5A) of 40 to 60 mesh. The oven temperature was maintained at 50°C throughout the analysis. Feed Digestibility (Exp. 1). Total tract apparent digestibilities of feed DM and nutrients were determined by using the external marker chromic oxide (Williams et al., 1962). In each period, animals were dosed orally before each concentrate feed with 1 g of Cr2O3 (contained within a gelatin bolus) from d 12 to 18. Insofar as possible, total feces was collected from each animal from d 16 to 18 inclusive. Before analysis, daily fecal output was thoroughly mixed, subsampled (10% for composite), and oven-dried at 55°C for 72 h. Dried fecal samples were pooled within animal by period and stored for subsequent chemical analysis. Apparent digestibility was calculated by using the following formula: DM digestibility (DMD) = (x − y)/x,
1051
dl-Malic acid supplementation to beef cattle
where x and y, respectively, are equal to the indicator in feces and the indicator in feed. The apparent digestibilities of the OM, NDF, ADF, and N fractions were calculated subsequently by using the approach for DM digestibility. Ruminal Fermentation and Protozoa Numbers (Exp. 2). To determine rumen pH and concentrations of NH3 and VFA, RF samples were taken immediately before the morning concentrate feeding (0 h) and then at 2, 4, 6, and 8 h after concentrate feeding on d 16, 17, and 18 of each experimental period. At these times, a 150-mL sample of rumen contents was collected via the rumen fistula and strained through a double layer of cheesecloth to obtain RF. The pH of the RF was immediately determined with a Mettler Toledo MP 200 pH meter. A 10-mL sample of RF was added to 0.25 mL of 9 M H2SO4 in a universal container for subsequent NH3 analysis. An additional 10-mL aliquot of the strained RF was preserved in 0.5 mL of 70 mM HgCl2 in a ratio of 1:20 preservative solution:RF and subsequently used for determination of VFA concentrations. Samples were stored at –20°C pending analysis. Exp. 2. On d 16 to 18 of each experimental period, and concurrent with RF sampling for ruminal fermentation measurements, a 1-mL sample of RF was added to 9 mL of methyl green solution and stored in darkness according to the method of Jordan et al. (2006) for subsequent protozoal enumeration.
Laboratory Analyses Exp. 1 and 2. Dried feed and fecal samples were ground in a hammer mill fitted with a 1-mm screen before being analyzed in duplicate. Dry matter concentration of concentrates was determined in an oven at 104°C for a minimum of 16 h. Crude protein concentrations of concentrates and feces were determined as N × 6.25 with a Leco FP 528 instrument (Leco Instruments UK Ltd., Cheshire, UK) according to the method of Dumas (AOAC, 1990). Crude protein concentration of fresh grass silage was determined in duplicate as Kjeldahl N × 6.25 with a Büchi 435 digestion unit and a Büchi 323 distillation unit (Büchi, Postfach, Flawil/Schweiz, Switzerland) according to AOAC (1990). Crude fiber of the concentrates was measured by using the Weende method (AOAC, 1990), whereas NDF, ADF, and ADL concentrations were determined by using a Fibertec system (Tecator, Höganäs, Sweden) according to the method of Van Soest et al. (1991). Ash concentration was determined after ignition at 550°C for 4 h in a muffle furnace and was used to calculate OM. The buffering capacity of the fresh grass silage was analyzed by using the method of Playne and McDonald (1966), and NH3-N concentrations were determined by using a modification of the 5.5.2 distillation method (European Union, 1971). The Cr in feces was extracted according to the method of Williams et al. (1962), and the concentration was measured by using an inductively coupled plasma
emission spectrophotometer (Varian Liberty 200; Varian, Australia Ltd., Mulgrave, Victoria). In Exp. 2, the relative molar concentrations of VFA (acetic, propionic, butyric, isobutyric, isovaleric, and valeric) in the RF were determined by GLC (Varian 3800) with a wall-coated open tubular fused-silica, 25 m × 0.53 mm capillary column with a CP-Wax 58 FFP coating (Varian BC, Middelburg, the Netherlands), according to the method of Porter and Murray (2001). Rumen protozoa numbers were measured by using a 0.1-mm-depth Bürker counting chamber (Rudolf Brand, Wertheim, Germany), with duplicate counts carried out for each sample. The concentration of NH3-N in the RF was determined by the microdiffusion technique as described by Conway (1957).
Statistical Analysis Data were checked for adherence to a normal distribution, and the natural logarithm of protozoa numbers was calculated. In Exp. 1, DMI, CH4, and diet digestibility measurements were analyzed by mixed models ANOVA (PROC MIXED, SAS Inst. Inc., Cary, NC) with terms included for animal, period, treatment, and replicate with the model Yijkl = animali + periodj + treatmentk + replicatel + εijkl. There was no evidence (P = 0.84) of a treatment × replicate interaction; thus, this term was not included in the final statistical model. Animal was treated as a random effect, whereas all other terms were treated as fixed effects. In Exp. 2, the data were again analyzed using mixed models ANOVA with terms included for animal, treatment, and period as follows: Yijk = animali + periodj + treatmentk + εijk. Where repeated within-day sampling occurred (i.e., VFA, pH, NH3, protozoa counts), repeated measures ANOVA was carried out (PROC MIXED, SAS Inst. Inc.). The type of variance-covariance structure used was chosen depending on the magnitude of the Akaike information criterion for models run under compound symmetry and unstructured, autoregressive, or Toeplitz variance-covariance structures. The model with the least Akaike information criterion was chosen. Orthogonal contrasts for linear, quadratic, and cubic (Exp. 2 only) effects of treatments were used to evaluate treatment effects.
RESULTS Feed Intake and Digestibility Exp. 1 and 2. The incremental inclusion of MA in Exp. 1 resulted in a linear reduction in silage (P =
1052
Foley et al.
Table 2. Effect of malic acid (MA) supplementation on feed intake and CH4 emission (Exp. 1) Dietary MA,1 %
P-value
Variable
MA-0
MA-3.75
MA-7.5
SEM
Linear
Quadratic
DMI, kg/d Silage Concentrate Total CH4, g/d CH4, g/kg of total DMI
3.70 5.69 9.39 246.1 26.2
3.59 5.47 9.06 230.6 25.5
3.50 5.23 8.72 207.3 23.8
0.059 0.055 0.081 4.98 0.59
0.01
