Acidogenic value of feeds. I. The relationship between the acidogenic value of feeds and in vitro ruminal ph changes

Acidogenic value of feeds. I. The relationship between the acidogenic value of feeds and in vitro ruminal ph changes B. Rustomo 1, J. P. Cant 1, M. Z. Fan 1, T. F. Duffield 2, N. E. Odongo 1, and B. W. McBride 1,3 1 Department of Animal and Poultry Science, 2 Department of Population Medicine, University of Guelph, Guelph, Ontario, Canada N1G 2W1. Received 21 October 2004, accepted 28 December 2005. Rustomo, B., Cant, J. P., Fan, M. Z., Duffield, T. F., Odongo, N. E. and McBride, B. W. 2006. Acidogenic value of feeds. I. The relationship between the acidogenic value of feeds and in vitro ruminal ph changes. Can. J. Anim. Sci. 86: 109 117. The objective of this study was to use an in vitro technique (i) to assess the acidogenic value (AV) of feed ingredients, (ii) to evaluate the relationship between the AV of feed and ruminal ph changes, and (iii) to determine the relationship between the AV of feeds and chemical constituents of feeds. Assessments of AV were based on 24 and 48 h in vitro incubation in rumen liquor. A series of feeds, ranging from energy, fibre and protein sources were evaluated. Ruminal fluid ph changes in the incubation medium were measured at the end of 24 and 48 h incubation in buffered rumen liquor (60% buffer, 40% rumen liquor). The chemical constituents of the feed ingredients were determined using standard procedures. There were no differences (P > 0.05) between 24 and 48 h incubations on apparent AV and rumen fluid ph changes. The best predictors of AV for all classes of feed were non-fibre carbohydrate (NFC) fraction and acid detergent fibre (ADF; R 2 = 0.81; P < 0.001). The best predictor of AV for energy sources were NFC and ADF (R 2 = 0.70; P < 0.027); neutral detergent fibre (NDF) for fibre sources (R 2 = 0.84; P < 0.027) and crude protein (CP) for protein sources (R 2 = 0.73; P < 0.014). The rumen fluid ph changes had stronger relationships with apparent AV of all feeds after 24 h (R 2 = 0.74, P < 0.0001) than starch (R 2 = 0.35, P = 0.04) or NFC (R 2 = 0.56; P < 0.0001). The results indicate that 24 h AV measurements and rumen fluid ph changes are acceptable measures for qualitatively describing or ranking feed ingredients. Key words: Acidogenic value, dairy cow, feed chemical composition, rumen ph Rustomo, B., Cant, J. P., Fan, M. Z., Duffield, T. F., Odongo, N. E. et McBride, B. W. 2006. Valeur acidogène des aliments du bétail. I. Relations entre la valeur acidogène des aliments du bétail et la variation in vitro du ph du rumen. Can. J. Anim. Sci. 86: 109 117. Par le biais d une technique in vitro, l étude devait : (i) évaluer la valeur acidogène (VA) des aliments du bétail, (ii) établir les liens entre la VA des aliments du bétail et les changements de ph dans le rumen et (iii) préciser la relation entre la VA des aliments du bétail et leur composition chimique. La VA a été évaluée par incubation in vitro dans du fluide du rumen pendant 24 h ou 48 h. Les auteurs ont examiné plusieurs types d aliments représentant diverses sources d énergie, de fibres et de protéines. Ils ont mesuré l écart du ph du fluide du rumen dans le milieu d incubation au terme de la période d incubation de 24 h ou de 48 h dans du fluide du rumen tamponné (60 % de solution tampon, 40 % de fluide du rumen). La composition chimique des aliments du bétail a été établie de la manière usuelle. La VA apparente et la variation de ph du fluide du rumen sont les mêmes (P > 0,05) après 24 ou 48 h d incubation. La fraction d hydrates de carbone non cellulosiques (FNC) et les fibres au détergent acide (FDA) prédisent le mieux la VA, peu importe le type d aliment (R 2 = 0,81, P < 0,001). La FNC et les FDA prédisent le mieux la VA pour les sources d énergie (R 2 = 0,70, P < 0,027); pour les sources de fibres, il s agit des fibres au détergent neutre (R 2 = 0,84, P < 0,027) tandis que pour les sources de protéines, la protéine brute donne les meilleurs résultats (R 2 = 0,73, P < 0,014). La variation de ph du fluide du rumen présente des liens plus étroits avec la VA apparente des aliments du bétail après 24 h (R 2 = 0,74, P < 0,0001) que l amidon (R 2 = 0,35, P = 0,04) ou la FNC (R 2 = 0,56, P < 0,0001). Les résultats indiquent qu on pourrait recourir à la quantification de la VA après 24 h d incubation et à la variation du ph dans le fluide du rumen pour décrire ou classer les aliments du bétail qualitativement. Mots clés: Valeur acidogène, vaches laitières, composition chimique des aliments, ph du rumen 3 To whom correspondence should be addressed (e-mail: bmcbride@uoguelph.ca). 109 Sub-acute rumen acidosis (SARA) represents lowered ruminal ph as a result of carbohydrate fermentation (Nocek 1997). Effective fibre intake and carbohydrate digestion rate interact to determine ruminal ph (Armentano and Pereira 1997). Rumen acidosis is therefore related to the amount of acid produced as feed is fermented and the ability of the feed to encourage chewing and production of salivary buffers. Although animals suffering from SARA do not typically exhibit clinical signs of illness and often go undetected (Owens et al. 1998), SARA can reduce feed intake, negatively affect rumen fermentation, growth, performance, and contribute to laminitis (Nocek 1997). Sub-acute ruminal acidosis can also damage ruminal and intestinal epithelial tissue, leading to bacterial infection and subsequent liver abscesses (Underwood 1992). Wadhwa et al. (2001) have developed a simple laboratorybased technique for evaluating acid production from feedstuffs Abbreviations: AV, acidogenic value; ADF, acid detergent fibre; BC, buffering capacity; CP, crude protein; DMI, dry matter intake; NDF, neutral detergent fibre; NFC, nonfibre carbohydrate; NPN, non-protein nitrogen; SARA, subacute rumen acidosis

110 CANADIAN JOURNAL OF ANIMAL SCIENCE based on the dissolution of Ca from CaCO 3. This method has been used to rank feeds in terms of the acid-load accumulated within the rumen during fermentation. The concept of effective neutral detergent fibre (endf), which integrates the effects of diet on chewing, saliva production, ruminal acid production and neutralization, uses milk fat composition as a response variable to predict acidosis (Armentano and Pereira 1997). However, Pereira et al. (1999) showed that milk fat depression was not a reliable indicator of pendf or SARA. The concept of physically effective NDF (pendf) to describe the impact of physical effectiveness of NDF in stimulating cud chewing has also been described (Mertens 1997). Although the concept of pendf was developed to avoid the problem of excess production of acid in the rumen (Mertens 1997), this approach does not consider acid production from the feed. Other in vitro studies have determined acid production from feeds by measuring the final ph after incubation (de Smet et al. 1995; Malestein et al. 1982). However, ph measurements alone do not include aspects of buffering (Stewart 1983). In the acidogenic value (AV) approach, the dissolved Ca represents how much total acid is produced and neutralized during fermentation (Wadhwa et al. 2001). However, Wadhwa et al. (2001) did not measure the relationship between AV and ruminal fluid ph changes in the incubation medium for different classes of feeds. Additionally, Wadhwa et al. (2001) did not measure the relationship between AV and the chemical components from different classes of feeds. The objectives of this study, therefore, were: (i) to measure the AV of feeds, (ii) to evaluate the relationship between AV of feeds and ruminal ph changes, and (iii) to examine the relationship between AV or ruminal fluid ph changes and feed chemical composition. MATERIALS AND METHODS The AV of the feed was determined using an in vitro technique developed by Tilley and Terry (1963) and modified by Wadhwa et al. (2001). Feeds were freeze-dried and ground through a 1-mm screen in a laboratory mill (Thomas Wiley, Philadelphia, PA) before being returned to the freeze drier to remove any moisture that had been picked up during grinding. One-gram (DM basis) samples were weighed directly from the freeze drier into 100-mL incubation tubes held at 39 C in a water bath. The samples were incubated in duplicate with 30 ml of buffered rumen liquor comprising 60% buffer and 40% rumen liquor. The buffer (5.880 g L 1 NaHCO 3 ; 5.580 g L 1 Na 2 HPO 4 ; 0.282 g L 1 NaCl; 0.342 g L 1 KCl; 0.028 g L 1 CaCl 2.2H 2 O and 0.036 g L 1 MgCl 2 ) was made up at 20% the strength of the Tilley and Terry (1963) buffer (Wadhwa et al. 2001). Rumen fluid was collected from two rumen-fistulated cows fed alfalfa hay ad libitum 3 h after morning feeding. All experimental procedures using the fistulated cows were done with the approval of the University of Guelph Animal Care Committee in accordance with the guidelines of the Canadian Council on Animal Care. Cysteine hydrochloride monohydrate (0.025% wt/vol) was added into the 100-mL incubation tubes just before incubation and the tubes were closed with gas release valves and shaken continuously. After 24 and 48 h of incubation, 2- ml samples were withdrawn from each tube and transferred in to 8-mL centrifuge tubes containing 50 mg of CaCO 3 powder (Catalogue No. C6763; Sigma Chemical Co., St. Louis, MO). The mixture was shaken manually for 5 s and then centrifuged at 4000 g for 10 min and the Ca content in the supernatant determined using a test kit (Sigma Diagnostics Inc., Calcium Procedure No. 587; Sigma- Aldrich Co., St. Louis, MO) in a laboratory spectrophotometer (HACH, DR/4000, Loveland, CO) set at 575 nm. The absorbance was read to zero using water as reference. All samples were run in duplicates in two separate runs. Apparent AV was calculated as the product of Ca concentration from the analysis and fluid volume (30 ml) divided by the sample weight (1 g; Wadhwa et al. 2001). To eliminate the contribution of Ca from the feed, basal AV was calculated after correcting for Ca dissolved before the addition of CaCO 3. True AV was calculated as apparent AV (after CaCO 3 ) less basal AV (before CaCO 3 ). Blank (with no feed sample) samples and standards (wheat and straw) were included in each run for calibration but these were not used to adjust the AV. The ph was measured before (0 h) and after incubations (24 and 48 h). Feed ingredients chosen for assessment included a range of energy, protein and fibre sources. The chemical composition of the feed ingredients was determined using standard Association of Official Analytical Chemists [(AOAC) 1990] procedures. The DM content was determined by oven drying at 60 o C for 48 h. Crude protein (CP), soluble protein and non-protein nitrogen (NPN) were analyzed using macro-kjeldahl (method 984.13). Starch using IKA analyzer technick (C5000) and Ca and P by inductively coupled plasma spectroscopy (method 945.46). The samples were also analysed for fat, non-fibre carbohydrate (NFC), acid detergent fibre (ADF; method 973.18c) and lignin (AOAC 1990) and neutral detergent fibre (NDF) (Goering and Van Soest 1970). Statistical Analysis The relationship between AV and rumen fluid ph and AV and feed chemical composition was determined using simple and multiple linear regression analysis within the SAS Institute, Inc. (2004). Comparison between 24 and 48 h incubations were conducted using paired t-test procedures in SAS. Effects were considered significant at a probability P < 0.05. RESULTS AND DISCUSSION The chemical composition of the feed ingredients is presented in Table 1. Acidogenic values for feed ingredients after 24 and 48 h of incubation are shown in Table 2. The basal AV of some feed ingredients (e.g. sugar beet pulp and alfalfa hay) were relatively high (6.15 ± 0.29 and 5.44 ± 0.21, respectively; Table 2), which suggests contribution of inherent Ca within these feed ingredients to the measured apparent AV. Energy sources had the highest AV; fibre sources had intermediate AV and protein sources had the lowest AV. Acidogenic values were not different (P > 0.05) between the 24 and 48 h incubation for all feed classes (Table 3). This suggests that there was little further fermentation after 24 h. It has been shown that the rate of degradation of feed ingredients is higher during the first 9 to12 h than during subsequent incu-

RUSTOMO ET AL. ACIDOGENIC VALUE OF FEEDS. I 111 Table 1. Chemical composition of the feed ingredients (% DM). Standard error of the mean (SEM), based on duplicate, analysis are given in parentheses DM CP Sol-P (Sol.P/CP) z NPN (NPN/Sol.P) y Fat Ash NDF ADF Lignin Starch NFC Ca P Sugar beet pulp 88.3 11.6 2.5 21.3 2.3 91.2 0.9 5.7 45.2 24.4 4.0 0.6 41.2 0.7 0.1 (0.02) (0.06) (0.04) (0.49) (0.06) (3.90) (0.11) (0.02) (0.14) (0.83) (0.03) (0.05) (1.32) (0.01) (0.00) Barley 88.2 12.6 1.7 13.2 0.7 42.2 2.1 2.3 30.5 6.2 0.5 53.1 59.7 0.0 0.3 (0.05) (0.13) (0.00) (0.13) (0.03) (1.62) (0.29) (0.06) (0.24) (0.09) (1.30) (0.48) (1.04)) (0.00) (0.04) Oats 89.1 12.4 2.7 22.0 0.8 30.0 4.3 2.5 29.3 13.8 1.9 46.3 47.2 0.1 0.3 (0.00) (0.20) (0.09) (0.36) (0.03) (0.04) (0.11) (0.01) (0.46) (0.30) (0.09) (1.19) (2.56) (0.01) (0.01) Wheat 88.1 13.0 1.2 9.1 0.5 44.8 1.8 1.7 11.9 2.6 1.9 62.0 70.2 0.0 0.3 (0.02) (0.25) (0.09) (0.86) (0.00) (3.40) (0.41) (0.08) (0.40) (0.21) (0.85) (0.94) (0.50) (0.00) (0.00) Wheat middling 87.9 21.0 4.7 22.4 2.0 43.0 5.1 4.0 28.3 7.4 2.0 36.6 44.4 0.2 0.8 (0.01) (0.08) (0.15) (0.63) (0.00) (1.38) (0.18) (0.03) (0.12) (0.03) (0.05) (0.46) (0.20) (0.01) (0.00) High moisture corn 58.0 9.0 3.1 34.1 2.4 77.0 3.5 0.8 5.5 1.3 0.0 74.5 82.7 0.2 0.2 (0.004) (0.23) (0.04) (1.30) (0.04) (2.37) (0.37) (0.03) (0.05) (0.01) (0.00) (0.10) (0.36) (0.02) (0.00) Corn distillers 85.5 35.4 5.2 14.7 4.2 81.3 12.9 4.6 43.1 15.1 5.7 3.0 12.2 0.0 0.7 (0.04) (0.12) (0.18) (0.46) (0.06) (1.75) (0.45) (0.02) (0.16) (0.18) (0.26) (0.02) (1.24) (0.00) (0.14) Wheat shorts 86.9 19.8 4.3 21.9 2.4 55.3 4.5 6.0 39.8 11.5 3.7 24.3 28.5 0.1 1.2 (0.01) (0.07) (0.04) (0.14) (0.09) (1.48) (0.45) (0.03) (0.33) (0.23) (0.12) (0.52) (1.88) (0.00) (0.01) Wheat bran 87.3 17.7 3.9 22.0 2.4 60.8 6.5 7.0 48.8 12.9 3.9 16.3 18.6 0.1 1.5 (0.01) (0.03) (0.04) (0.28) (0.11) (3.47) (1.96) (0.18) (0.12) (0.05) (0.09) (0.20) (2.89) (0.00) (0.01) Corn silage 38.6 9.6 3.9 40.9 3.7 95.0 3.8 3.0 37.1 18.6 1.4 33.4 37.1 0.3 0.2 (0.02) (0.17) (0.11) (1.90) (0.03) (3.56) (0.44) (0.07) (0.13) (0.07) (0.02) (0.33) (0.03) (0.02) (0.10) Alfalfa pellets 93.4 16.0 2.9 18.1 2.6 88.7 2.8 7.5 54.9 32.2 4.9 2.1 15.7 1.2 0.2 (0.03) (0.07) (0.08) (0.43) (0.03) (3.40) (0.16) (0.12) (0.22) (0.10) (0.07) (0.24) (0.39) (0.02) (0.00) Alfalfa hay 86.7 16.7 2.9 17.5 2.5 86.7 2.7 8.0 45.6 28.8 4.8 4.5 26.6 1.0 0.3 (0.05) (0.22) (0.24) (1.66) (0.07) (4.56) (0.08) (0.00) (0.02) (0.25) (0.35) (0.08) (0.06) (0.01) (0.00) Haylage 31.9 18.2 9.8 53.7 8.7 88.9 3.4 7.9 58.8 37.6 5.2 0.3 10.1 0.7 0.3 (0.008) (0.21) (0.00) (0.62) (0.08) (0.84) (0.04) (0.02) (0.21) (0.20) (0.10) (0.21) (5.08) (0.09) (0.00) Wheat straw 92.2 4.0 0.7 16.9 0.2 35.4 2.1 5.0 83.3 67.7 18.5 1.9 10.2 0.1 0.1 (0.04) (0.29) (0.06) (2.65) (0.03) (1.22) (0.55) (0.07) (0.20) (0.01) (0.14) (0.05) (2.25) (0.00) (0.00) Roasted soybeans 93.6 42.3 3.0 7.1 1.0 33.8 23.4 5.0 29.2 6.6 1.0 1.1 16.8 0.2 0.6 (0.04) (0.03) (0.06) (0.15) (0.08) (1.85) (0.07) (0.02) (0.27) (0.26) (0.02) (0.02) (0.20) (0.00) (0.00) Soybean meal 87.2 56.0 5.5 9.9 1.5 26.7 3.5 6.4 8.6 4.6 0.4 1.1 24.9 0.3 0.7 (0.13) (0.05) (0.01) (0.01) (0.11) (1.90) (0.47) (0.01) (0.19) (0.15) (0.12) (0.18) (1.43) (0.01) (0.00) Canola meal 86.3 42.1 3.5 8.3 2.2 64.4 4.9 8.1 26.5 19.1 9.7 1.3 15.8 0.8 1.1 (0.09) (0.38) (0.01) (0.10) (0.11) (3.43) (0.40) (0.00) (0.00) (0.20) (0.13) (0.13) (2.31) (0.01) (0.01) Feather meal 89.5 86.0 6.9 8.0 5.3 77.1 13.4 2.1 29.7 8.7 6.4 0.4 ND x 0.4 0.2 (0.01) (1.11) (0.19) (0.32) (0.09) (3.39) (0.07) (0.03) (0.33) (0.15) (0.49) (0.07) (ND) (0.01) (0.01) Corn gluten 87.4 71.4 4.7 6.6 3.9 82.4 1.0 2.5 9.2 3.4 1.4 13.3 18.8 0.0 0.59 (0.02) (0.10) (0.09) (0.11) (0.05) (0.37) (0.15) (0.04) (0.57) (0.00) (0.05) (1.35) (0.01) (0.00) (0.01) Blood meal 95.6 94.7 3.6 3.8 1.2 34.1 1.3 2.1 ND ND ND 0.2 ND 0.1 0.3 (0.03) (0.07) (0.08) (0.08) (0.04) (1.86) (0.09) (0.00) (ND) (ND) (ND) (0.15) (ND) (0.00) (0.00) Herring meal 94.3 77.5 7.2 9.3 5.7 79.2 15.4 10.5 ND 0.6 0.6 0.3 ND 2.2 1.9 (0.04) (0.30) (0.02) (0.01) (0.10) (1.20) (0.20) (0.04) (ND) (0.02) (0.15) (0.03) (ND) (0.04) (0.01) z % soluble protein/cp. y % NPN/Sol. P. x ND = not determined.

112 CANADIAN JOURNAL OF ANIMAL SCIENCE Table 2. Acidogenic value [AV; dissolved Ca (mg Ca g 1 feed DM)] of feed ingredients after 24 and 48 h incubations. Values are given for Ca dissolved before (basal AV) and after the addition of CaCO 3 (apparent AV) and the difference (true AV). Standard error of the mean is given in parentheses 24 h 48 h Feed ingredients Apparent AV Basal AV True AV Apparent AV Basal AV True AV Energy sources Sugar beet pulp 14.68 (0.43) 6.15 (0.29) 8.53 (0.66) 13.14 (0.34) 5.37 (0.06) 7.77 (0.38) Barley 12.67 (0.36) 3.60 (0.34) 9.06 (0.54) 11.28 (0.11) 3.07 (0.06) 8.21 (0.09) Oats 12.41 (0.22) 2.98 (0.21) 9.43 (0.41) 11.13 (0.27) 2.91 (0.12) 8.22 (0.16) Wheat 11.65 (0.62) 3.28 (0.28) 8.37 (0.87) 11.06 (0.12) 2.74 (0.13) 8.32 (0.21) Wheat middling 10.84 (0.94) 3.52 (0.23) 7.31 (0.17) 10.56 (0.37) 3.92 (0.19) 6.64 (0.54) High moisture corn 9.59 (0.24) 2.14 (0.23) 7.44 (0.31) 11.53 (0.25) 2.43 (0.22) 9.10 (0.45) Corn distillers 8.38 (0.22) 2.67 (0.30) 5.71 (0.29) 9.26 (0.17) 2.58 (0.20) 6.68 (0.24) Wheat shorts 8.04 (0.81) 3.17 (0.34) 4.87 (0.78) 9.67 (0.41) 2.97 (0.11) 6.70 (0.31) Wheat bran 5.70 (0.19) 2.84 (0.27) 2.86 (0.26) 8.35 (0.17) 2.70 (0.18) 5.64 (0.20) Fibre sources Corn silage 12.86 (0.25) 3.60 (0.22) 9.24 (0.29) 12.20 (0.18) 3.73 (0.11) 8.46 (0.11) Alfalfa pellet 12.63 (0.67) 4.92 (0.42) 7.71 (0.26) 10.41 (0.28) 4.01 (0.39) 6.41 (0.25) Alfalfa hay 11.71 (0.66) 5.44 (0.21) 6.09 (0.50) 11.46 (0.15) 4.27 (0.31) 7.18 (0.31) Haylage 9.30 (0.50) 5.28 (0.34) 4.02 (0.28) 7.95 (0.36) 3.86 (0.17) 4.09 (0.21) Wheat straw 6.27 (0.90) 3.04 (0.27) 3.22 (0.64) 6.78 (0.56) 2.69 (0.17) 4.09 (0.40) Protein sources Roasted soybean 7.62 (0.68) 2.34 (0.34) 5.28 (0.39) 5.45 (0.32) 1.57 (0.18) 3.88 (0.17) Soybean meal 6.43 (0.39) 3.54 (0.33) 2.89 (0.39) 5.48 (0.40) 2.69 (0.08) 2.78 (0.37) Canola meal 5.13 (0.43) 3.65 (0.13) 1.48 (0.32) 4.09 (0.16) 2.54 (0.30) 1.55 (0.46) Feather meal 3.03 (0.20) 0.79 (0.08) 2.24 (0.17) 1.40 (0.18) 0.34 (0.03) 1.06 (0.16) Corn gluten 2.40 (0.24) 2.13 (0.30) 0.27 (0.21) 3.10 (0.16) 2.07 (0.28) 1.02 (0.34) Blood meal 1.66 (0.11) 1.33 (0.12) 0.33 (0.16) 1.31 (0.19) 0.57 (0.05) 0.74 (0.15) Herring meal 1.42 (0.22) 0.97 (0.09) 0.46 (0.28) 1.87 (0.24) 0.63 (0.07) 1.24 (0.29) Table 3. Effects of incubation time on basal, apparent and true acidogenic value [AV; dissolved Ca (mg Ca g 1 feed DM)] and rumen fluid ph changes during 24 and 48 h incubations 24 h 48 h SEM z P value All feed classes Apparent AV (after the addition of CaCO 3 ) 8.30 7.97 0.31 0.461 Basal AV (before the addition of CaCO 3 ) 3.21 2.75 0.11 0.003 True AV y 5.09 5.23 0.24 0.680 Rumen fluid ph changes x 1.30 1.39 0.07 0.478 Energy sources Apparent AV (after the addition of CaCO 3 ) 10.44 10.66 0.27 0.570 Basal AV (before the addition of CaCO 3 ) 3.37 3.19 0.13 0.338 True AV y 7.07 7.47 0.24 0.250 Rumen fluid ph changes x 1.75 1.90 0.05 0.107 Fibre sources Apparent AV (after the addition of CaCO 3 ) 10.52 9.76 0.42 0.207 Basal AV (before the addition of CaCO 3 ) 4.46 3.71 0.17 0.003 True AV y 6.06 6.05 0.35 0.986 Rumen fluid ph changes x 1.36 1.53 0.07 0.199 Protein sources Apparent AV (after the addition of CaCO 3 ) 3.96 3.24 0.27 0.086 Basal AV (before the addition of CaCO 3 ) 2.11 1.49 0.15 0.005 True AV y 1.85 1.75 0.22 0.757 Rumen fluid ph changes x 0.68 0.65 0.09 0.884 z Standard error of the mean. y True AV was calculated as apparent AV (AV after addition of CaCO 3 ) basal AV (AV before addition of CaCO 3 ). Rumen fluid ph changes = decrease in rumen fluid ph from the start of incubation to the end of incubation. bation (de Smet et al. 1995). The differences in AV were likely related to the rapid initial fermentation of the feeds. de Smet et al. (1995) also showed that the rate of degradation varied with the type of feed ingredient, being higher for feeds with high cell content and lower for feeds with high cell wall content. Energy source feeds resulted in similar AV with fibre sources (Table 2) but had the greatest rumen fluid ph changes (Table 3 and Fig. 1a).

RUSTOMO ET AL. ACIDOGENIC VALUE OF FEEDS. I 113 (a) (b) (c) Fig. 1. (a) Rumen fluid ph changes from 0 h to 24 and 48 h after incubation for the energy sources. (b) Rumen fluid ph changes from 0 h to 24 and 48 h after incubation for the fibre sources. (c) Rumen fluid ph changes from 0 to 24 h and 48 h after incubation for the protein sources.

114 CANADIAN JOURNAL OF ANIMAL SCIENCE Rumen fluid ph changes for different classes of feed ingredients after 24 and 48 h of incubation are shown in Fig. 1. The rumen fluid ph decreased significantly during the first 24 h of incubation, but remained relatively constant after that to 48 h of incubation (Fig. 1 a, b). The rate of change of rumen fluid ph changes appears to follow similar patterns as the AV changes. Energy source feed ingredients showed the greatest ph decrease during the first 24 h of incubation (Fig. 1a). This result is in agreement with those of de Smet et al. (1995) and Hall (2002), that highly fermentable carbohydrates, such as sugars, soluble fibre and some starches, have the capacity to decrease ruminal ph in a relatively short period of time (1 to 5 h). Rumen fluid ph changes were lowest for protein feed sources (Fig. 1c). The differences in rumen fluid ph changes between 24 and 48 h of incubation were also smallest for protein feeds (Table 3). These results suggest that protein feed sources have different effects on ruminal ph changes. It has been shown that high protein feeds generally have higher buffering capacity (McBurney et al. 1983; Jasaitis et al. 1987) due to the ammonia produced during fermentation (Crawford et al. 1983). High protein feeds are therefore capable of neutralizing acid to maintain constant ph in continuous culture (Dewhurst et al. 2001). The relationship between apparent AV and rumen fluid ph changes after 24 h of incubation are presented in Table 4. There was a positive correlation between AV and rumen fluid ph change for all classes of feed (R 2 = 0.74; P < 0.0001; Table 4). High AV was associated with a greater decrease in rumen fluid ph change. Furthermore, the significant relationship between apparent AV and rumen fluid ph changes suggests high AV diets would be expected to increase the risk of rumen acidosis in cows. Several studies have shown decreased ruminal ph when more rapidly fermentable carbohydrates were included in the diet (Keunen et al. 2002; Krause et al. 2002b; Plaizier et al. 2001; Reinhart et al.1993). It has also been suggested that increased fermentation of starch might overload the absorptive capacity of the rumen thus exacerbating the reductions in ruminal ph (Krause et al. 2003). Furthermore, a decrease in ruminal ph was reported to decrease appetite (Briton and Stock 1987), fibre degradability (Hoover 1986; Krajcarski-Hunt et al. 2002), microbial protein synthesis (Strobel and Russell 1986; de Veth and Kolver 2001; Calsamiglia et al. 2002), DMI and subsequent milk production (Aldrich et al. 1993; Krause et al. 2002a). Prediction of potential rumen fluid ph changes based on the AV of feeds would be highly beneficial with regard to diet formulation in order to balance for diets that would optimize ruminal ph and rumen function. However, in vivo studies are needed to clarify the effect of feed AV on actual ruminal ph changes in dairy cows. The relationships between the AV of the feeds and feed chemical composition are presented in Table 5. There was a positive correlation between AV and starch content (R 2 = 0.20; P = 0.042) of feed and AV and NFC content (R 2 = 0.43; P < 0.001) of the feed. Although the starch contents of sugar beet pulp and corn distillers were low, their NFC contents were high (Table 1). Sugar beet pulp contains a significant amount of soluble fibre (17.4 to 30.0% DM), and sugars (12.8 to 24.7% DM), while corn distillers contain soluble fibre in the range of 7.8 to 11.6% DM, and sugars from 3.2 to 14.5% DM (Hall et al. 1999). Sugars and starch ferment to lactic acid, which has a lower pka than the volatile fatty acids. Soluble fibre, such as pectin, ferments rapidly, but its rate of fermentation declines at low ph (Ben-Ghedalia et al. 1989; Strobel and Russel 1986). Although the starch and NFC content of highmoisture corn were the highest among the energy feeds (Table 1), the AV was intermediate (Table 2). Different types of NFC differ in their rate of fermentation and their effects on ruminal ph (de Smet et al. 1995; Hall 2000). Rooeny and Pflugfelder (1986) reported that the starch granules in corn are almost completely contained within a protein matrix, which decreases the availability of starch to hydrolysis. For all classes of feed, the relationship between rumen fluid ph changes and AV were stronger (R 2 = 0.74; P < 0.001) than the relationship between starch and rumen fluid ph changes (R 2 = 0.35; P = 0.004) or rumen fluid ph changes and NFC (R 2 = 0.56; P < 0.001; Table 4). Crude protein content was negatively correlated with AV (Table 5). High protein feeds would therefore be expected to have low fermentability and thus produce less rumen acid load. As noted above, it has been shown that high protein feeds generally have higher buffering capacity (McBurney et al. 1983; Jasaitis et al. 1987) due to the ammonia produced during fermentation (Crawford et al. 1983). It is therefore, important to consider feed ingredient AV when formulating rations rather than just using the starch content or the NFC content. Among the forages tested, corn silage had the highest AV and wheat straw the lowest (Table 2). Corn silage was almost as acidogenic as the high-energy feeds despite its lower starch content and higher NDF content (Table 1). The high AV for corn silage might suggest high levels of free acids (Thomas and Wilkinson 1975) and low CP content (Table 1). Additionally, it has been shown that corn silage required higher alkali to stabilize high rumen fluid ph changes in continuous culture (Crawford 1983; Dewhurst et al. 2001; Wadhwa et al. 2001). The low AV for wheat straw might be explained by its poor NDF degradability. According to Weiss et al. (1989) and de Smet et al. (1995), feedstuffs rich in cell wall content have low degradation rates. These results support the use of an optimum forage mixture when feeding large amounts of starch-rich concentrates to dairy cows (Phipps et al. 1995). Table 6 shows apparent AV predictions from chemical composition of the dietary ingredients based on stepwise multiple linear regression analysis. Greater variation was explained when the factors were included in combination demonstrating the importance of the interactions among chemical constituents in a diet. This suggests that acid production and neutralization in the rumen is a complex phenomenon and involves the interaction between feed components, ruminal environment and microbial populations. More species of microorganisms grow in the rumen when the substrates consist of a more complex composition (Malestein et al. 1982). de Smet et al. (1995) postulated that a synergism between different bacterial species may occur when a substrate is more complex and fermentation prod-

RUSTOMO ET AL. ACIDOGENIC VALUE OF FEEDS. I 115 Table 4. Prediction z of rumen fluid ph changes y (Y) from apparent acidogenic value [AV; dissolved Ca (mg Ca g 1 feed DM)] of feed ingredients (x) after 24 h of incubation Factor (x) b 0 b 1 b 2 b 3 R 2 P value All feeds Y 1 = b 0 + b 1 x 1 (starch ) 1.046 0.014 0.35 0.004 Y 2 = b 0 + b 2 x 2 (NFC x ) 0.795 0.018 0.56 <0.0001 Y 3 = b 0 + b 3 x 3 (AV) 0.313 0.119 0.74 <0.0001 Energy sources Y 1 = b 0 + b 1 x 1 (starch ) 1.71 0.001 0.02 0.72 Y 2 = b 0 + b 2 x 2 (NFC) 1.57 0.004 0.17 0.26 Y 3 = b 0 + b 3 x 3 (AV) 1.19 0.054 0.47 0.04 Fibre sources Y 1 = b 0 + b 1 x 1 (starch ) 1.23 0.015 0.61 0.118 Y 2 = b 0 + b 2 x 2 (NFC) 0.99 0.018 0.62 0.114 Y 3 = b 0 + b 3 x 3 (AV) 0.42 0.089 0.80 0.040 Protein sources Y 1 = b 0 + b 1 x 1 (starch ) 0.59 0.037 0.17 0.351 Y 2 = b 0 + b 2 x 2 (NFC) 0.29 0.036 0.82 0.005 Y 3 = b 0 + b 3 x 3 (AV) 0.17 0.130 0.58 0.048 z Rumen fluid ph changes after 24 h incubations (Y 1 ) = b 0 + b 1 (x 1 ); (Y 2 ) = b 0 + b 2 (x 2 ); where b 0 = intercept, b 1 = slope (increase in ph changes per unit increase in starch); b 2 = slope (increase in ph changes per unit increase in NFC); x 1 = starch content and x 2 = NFC content ; and b 3 = slope (increase in ph changes per unit increase in apparent AV); x 1 = starch content; x 2 = NFC content, and x 3 = apparent AV. y Rumen fluid ph change = rumen fluid ph at the start of incubation rumen fluid ph at the end of 24 h of incubation. x NFC = non-fibre carbohydrate R 2 = coefficient of determination Table 5. Single predictor of apparent acidogenic value z from dietary chemical composition variable Factor (x) b 0 b 1 R 2 P value CP (% DM) 12.221 0.119 0.690 <0.0001 Sol. P (% DM) 11.376 0.769 0.163 0.070 Sol. P/CP (%) 5.466 0.157 0.228 0.029 NPN (% DM) 9.596 0.481 0.058 0.291 NPN/Sol P (%) 6.821 0.024 0.019 0.550 Fat (% DM) 9.607 0.229 0.107 0.148 Ash (% DM) 9.269 0.197 0.017 0.575 NDF (% DM) 6.039 0.071 0.141 0.093 ADF (% DM) 5.976 0.187 0.248 0.021 Lignin (% DM) 8.242 0.022 0.0001 0.953 Starch (% DM) 6.930 0.077 0.200 0.042 NFC (% DM) 5.111 0.115 0.425 0.001 Calcium (% DM) 8.624 0.784 0.011 0.652 Phosphorus (% DM) 10.597 4.006 0.231 0.027 z Apparent acidogenic value (AV) (mg Ca g 1 feed DM) = b 0 + b 1 (x); where b 0 = intercept, b 1 = slope (increase in AV per unit increase in dietary chemical composition variables), and x = dietary chemical composition variables. R 2 = coefficient of determination. ucts of one species are used as substrate by another species. However, some feed ingredients have a poor balance of substrates for rumen microbes. Wadhwa et al. (2001) found that some feed ingredients such as wheat (high starch content), wheat straw (high fibre), and feather meal (high protein content) had deviations from a linear response on AV. Furthermore, the extent of ruminal fermentation for both fibre and non-fibre carbohydrates among feedstuffs is extremely variable and is influenced by interactions between diet composition and rumen microbial activity (Allen 1997). Measures of feed chemical composition alone are therefore inadequate for estimating acid-load in the rumen and may not be useful as AV indicators. CONCLUSIONS This study shows that the acid loads from feeds can be estimated by an in vitro technique. Energy sources and fibre sources had the highest AV whilst protein sources had the lowest AV. There was a positive correlation (P < 0.001) between AV and rumen fluid ph changes for all feeds. A high AV was associated with a greater decrease in rumen fluid ph change. The rumen fluid ph changes had stronger relationship with apparent AV of all feeds after 24 h (R 2 = 0.74, P < 0.0001) than with starch (R 2 = 0.35, P = 0.04) or with NFC (R 2 = 0.56; P < 0.0001). This relationship may allow predictions of ruminal fluid ph change from feed AV. The AV estimates of a range of feed ingredients can be used to formulate

116 CANADIAN JOURNAL OF ANIMAL SCIENCE Table 6. Stepwise multiple linear regression analysis prediction of apparent acidogenic value z from dietary chemical composition variables b 0 b 1 b 2 R 2 P value Model 1 (all feed classes) 0.81 <0.0001 Intercept 1.628 0.076 NFC (% DM) 0.135 <0.0001 ADF (% DM) 0.236 <0.0001 Model 2 (energy sources) 0.701 0.027 Intercept 0.805 0.772 NFC (% DM) 0.130 0.011 ADF (% DM) 0.356 0.022 Model 3 (fibre sources) 0.844 0.027 Intercept 18.738 0.003 NDF (% DM) 0.146 0.027 Model 4 (protein sources) 0.730 0.014 Intercept 10.686 0.002 CP (% DM) 0.100 0.014 z Apparent acidogenic value (AV) (mg Ca g 1 feed DM) = b 0 + b 1 (x 1 ) + b 2 (x 2 ); where b 0 = intercept, b 1, b 2, b 3 = slope (unit increase in AV per unit increase in dietary chemical composition variable), x 1 and x 2 = dietary chemical composition variables (% DM). R 2 = coefficient of determination). diets that would present a low rumen acid load in dairy cows and therefore prevent rumen acidosis. There were no differences in apparent AV and rumen fluid ph change measurements at 24 and 48 h of incubation suggesting that 24 h of incubation may provide an acceptable measure of feed AV and rumen fluid ph change. However, further studies are needed to examine the effect of feed AV on in vivo ruminal ph changes in dairy cows. ACKNOWLEDGEMENTS The authors wish to acknowledge the support of Dairy Farmers of Ontario, Ontario Ministry of Agriculture, Food and Rural Affairs and the Natural Sciences and Engineering Research Council (BWM) for financial support. The technical assistance of Qian Zhang in the laboratory work is also gratefully acknowledged. Aldrich, J. M., Muller, L. D., Varga, G. A. and Griel, L. C. 1993. Nonstructural carbohydrate and protein effects on rumen fermentation, nutrient flow, and performance of dairy cows. J. Dairy Sci. 76: 1091 1098. Allen, M. S. 1997. Relationship between fermentation acid production in the rumen and the requirement for physically effective fiber. J. Dairy Sci. 80: 1447 1462. Armentano, L. and Pereira, M. 1997. Measuring the effectiveness of fiber by animal response trial. J. Dairy Sci. 80: 1416 1425. Association of Official Analytical Chemist. 1990. Official methods of analysis. 15th ed. AOAC Arlington, VA. Ben-Ghedalia, D., Yosef, E., Miron, J. and Est, Y. 1989. The effects of starch- and pectin- rich diets on quantitative aspects of digestion in sheep. Anim. Feed Sci. Technol. 24: 289-295. Britton, R. A. and Stock, R. A. 1987. Acidosis, rate of starch digestion and intake. Pages 25 137 in Okla. Agric. Exp. Stn. Mp 121. Calsamiglia, S., Ferret, Plaixats, A. J. and Devant, M. 2002. Effect of ph and ph fluctuations on microbial fermentation and nutrient flow from a dual-flow continuous culture system. J. Dairy Sci. 85: 574 579. Crawford, R. J., Jr., Shriver, B. J., Varga, G. A. and Hoover, W. H. 1983. Buffer requirements for maintenance of ph during fermentation of individual feeds in continuous cultures. J. Dairy Sci. 66: 1881 1890. de Smet, A. M., de Boover, J. L., de Brabander, D. L., Vanacker, D. L. and Boucque, Ch. V. 1995. Investigation of dry matter degradation and acidotic effect of some feedstuffs by means of in sacco and in vitro incubations. Anim. Feed Sci. Technol. 51: 297 315. de Veth, M. J. and Kolver, E. S. 2001. Diurnal variation in ph reduces digestion and synthesis of microbial protein when pasture is fermented in continuous culture. J. Dairy Sci. 84: 2066 2072. Dewhurst, R. J., Wadhwa, D., Borgida, L. P. and Fisher, W. J. 2001. Rumen acid production from dairy feeds. 1. Effects on feed intake and milk production of dairy cows offered grass or corn silages. J. Dairy Sci. 84: 2721 2729. Goering, H. K. and Van Soest, P. J. 1970. Forage fiber analysis (apparatus, reagents, procedures, and some applications). Agric. Handbook No. 379. ARS-USDA, Washington, DC. Hall, M. B. 2000. Neutral detergent soluble carbohydrate: nutritional relevance and analysis. A laboratory manual. University of Florida Cooperative Extension Bulletin 339. Hall, M. B. 2002. Rumen acidosis: Carbohydrate feeding considerations. 12 th International symposium on lameness in ruminant. Orlando, Florida. 2002 Jan. 9 13. Dept. of Anim. Sci. University of Florida, Gainesville, FL. pp. 51 61. Hall, M. B., Hoover, W. H., Jennings, J. P. and Miller Webster, T. K. 1999. A method for partitioning neutral detergent-soluble carbohydrates. J. Sci. Food. Agric. 79: 2079 2085. Hoover, W. H. 1986. Chemical factors involved in ruminal fiber digestion. J. Dairy Sci. 69: 2755 2766. Jasaitis, D. K., Wholt, J. E, and Evans, J. L. 1987. Influence of feed ion content on buffering capacity of ruminant feedstuffs in vitro. J. Dairy Sci. 70: 1391 1403. Keunen, J. E., Plaizier, J. C., Kyriazakis, I., Duffield, T. F., Widowski, T. M., Lindinger, M.L. and McBride, B. W. 2002. Effect of subacute ruminal acidosis model on the diet selection of dairy cows. J. Dairy Sci. 85: 3304 3313. Krajcarski- Hunt, H. K., Plaizier, J. C., Walton, J. P., Spratt, R. and McBride, B. W. 2002. Effect of subacute ruminal acidosis

RUSTOMO ET AL. ACIDOGENIC VALUE OF FEEDS. I 117 on in situ fiber digestion in lactating dairy cows. J. Dairy Sci. 85: 570 573. Krause, K. M., Combs, D. K. and Beauchemin, K. A. 2002a. Effects of particle size and grain fermentability in mid lactation cows. I. Milk Production and diet digestibility. J. Dairy Sci. 85: 1936 1946. Krause, K. M., Combs, D. K. and Beauchemin, K. A. 2002b. Effects of particle size and grain fermentability in mid lactation cows. II. Ruminal ph and chewing activity. J. Dairy Sci. 85: 1947 1957. Krause, K. M., Combs, D. K. and Beauchemin, K. A. 2003. Effect of increasing levels of refined cornstarch in the diet of lactating dairy cows on performance and ruminal ph. J. Dairy Sci. 86: 1341 1353. Malestein, A., Van t Klooster, A. T., Counette, G. H. M. and Prins, R. A. 1982. Concentrate feeding and ruminal fermentation. 2. Influence of concentrate ingredients on ph and on L-Lactate concentration in incubations in vitro with rumen fluid. Neth. J. Agric. Sci. 30: 259 273. McBurney, M. I., Van Soest, P. J. and Chase, L. E. 1983. Cation exchange capacity and buffering capacity of neutral detergent fiber. J. Sci. Food. Agric. 34: 910 916. Mertens, D. R. 1997. Creating a system for meeting the fiber requirements of dairy cows. J. Dairy Sci. 80: 1463 1481. Nocek, J. E. 1997. Bovine acidosis: Implications on laminitis. J. Dairy Sci. 80: 1005 1028. Owens, F. N., Secrist, D. S., Hill, W. J. and Gill, D. R. 1998. Acidosis in cattle: a review. J. Anim. Sci. 76: 275 286. Pereira, M. N., Garret, E. F., Oetzel, G. R. and Armentano, L. E. 1999. Partial replacement of forage with non-forage fiber sources in lactating cow diets. Performance and health. J. Dairy Sci. 82: 2716 2730. Phipps, R. H., Sutton, J. D. and Jones, B. A. 1995. Forage mixtures for dairy cows: the effect on dry matter intake and milk production of incorporating either fermented or urea-treated whole-crop wheat, brewers grain, fodder beet or maize silage into diets based on grass silage. Anim. Sci. 61: 491 496. Plaizier, J. C., Martin, A., Duffield, T., Bagg, R., Dick, P. and McBride, B. W. 2001. Effect of subacute ruminal acidosis on in situ digestion of mixed hay in lactating dairy cows. Can. J. Anim. Sci. 81: 421 423. Reinhart C. D., Brandt, R. T., Jr., Behnke, K. C., Freeman, A. S. and Eck, T. P. 1993. Effect of steam flaked sorghum grain density on performance, mill production rate, and subacute acidosis in feedlot steers. Kansas Cattlemen s Day. Kansas State University, Manhattan, KS. pp. 98 100. Rooney, L. W. and Pfugfelder, R. L. 1986. Factors affecting starch digestibility with special emphasis on sorghum and corn. J. Anim. Sci. 63: 1607 1623. SAS Institute, Inc. 2004. SAS user s guide. Statistics. Version 9.1. ed. 2004. SAS Institute, Inc., Cary, NC. Stewart, P. A. 1983. Modern quantitative acid-base chemistry. Can. J. Physiol. Pharmacol. 61: 1444 1461. Strobel, H. J. and Russell, J. B. 1986. Effect of ph and energy spilling on bacterial protein synthesis by carbohydrate-limited cultures of mixed rumen bacteria. J. Dairy Sci. 69: 2941 2947. Thomas, C. and Wilkinson, J. M. 1975. The utilization of maize silage for intensive beef production. J. Agric. Sci. (Camb.) 85: 255 261. Tilley, J. M. A. and Terry, R. A. 1963. A two stage technique for the in vitro digestion of forage crops. J. Br. Grassl. Soc. 18: 104 111. Underwood, W. J. 1992. Rumen lactic acidosis. Part. 1. Epidemiology and pathology. Compend. Contin. Edu. Pract. Vet. 14: 1127 1133. Wadhwa, D., Beck, N. F. G., Borgida, L. P., Dhanoa, M. S. and Dewhurst, R. J. 2001. Development of a simple in vitro assay for estimating net rumen acid load from diet ingredients. J. Dairy Sci. 84: 1109 1117. Weiss, W. P., Fisher, G. R. and Erickson, G. M., 1989. Effect of source of neutral detergent fiber and starch on nutrient utilization by dairy cows. J. Dairy Sci. 72: 2308 2315.