Effects of phenolic acid structures on meadow hay digestibility

Animal Feed Science and Technology 136 (2007) 297–311

Effects of phenolic acid structures on meadow hay digestibility M.A.M. Rodrigues a,∗ , C.M. Guedes a , J.W. Cone b , A.H. van Gelder b , L.M.M. Ferreira a , C.A. Sequeira a a

CECAV-Universidade de Tr´as-os-Montes e Alto Douro, Department of Animal Science, Apartado 1013, 5001-801 Vila Real, Portugal b Animal Sciences Group, WUR, P.O. Box 65, NL-8200 AB Lelystad, The Netherlands

Received 18 April 2006; received in revised form 5 September 2006; accepted 7 September 2006

Abstract The objectives were to evaluate effects of phenolic acid content and composition on the digestibility of six meadow hays from Northern Portugal. Digestibility was assessed by gas production, in vitro and in situ degradation methods. Four cows fed diets at energy maintenance were used for in situ incubations and to provide rumen fluid for in vitro incubations. There were no relationships between phenolic acid concentrations and other cell wall components. The dry matter (DM) potential degradation (a + b) was positively related to the etherified fractions of ferulic acid (FAeth, P=0.012) and p-coumaric acid (PCAeth, P<0.001) and to the total amount of ferulic acid (FAtotal; P=0.033). The insoluble, but potentially degradable, constant (b) of DM had a positive relationship with PCAeth (P=0.008). The in vitro neutral detergent fibre digestibility (IVNDFD), and the estimated asymptotic Abbreviations: a, immediately soluble fraction; A, estimated asymptotic gas production; a + b, potential degradation; ADFom, ADF expressed exclusive of residual ash; b, insoluble but potentially degradable fraction; B, time of incubation at which half of the asymptotic gas production has been formed; c, rate constant for the degradation of fraction b; C, sharpness of the switching characteristic for the profile; CP, crude protein; DM, dry matter; ED, effective degradability; FAest, esterified fraction of ferulic acid; FAeth, etherified fraction of ferulic acid; FAtotal, total amount of ferulic acid; IVDMD, in vitro dry matter digestibility; IVNDFD, in vitro NDF digestibility; NDFom, NDF not assayed with stable amylase expressed exclusive of residual ash; PCAest, esterified fraction of p-coumaric acid; PCAeth, etherified fraction of p-coumaric acid; PCAtotal, total amount of p-coumaric acid; R, fractional rate of substrate fermentation; Rmax G, maximum rate of gas production; TRmax G, time at which maximum rate of gas production is reached ∗ Corresponding author. Tel.: +351 259 350 422; fax: +351 259 350 482. E-mail address: [email protected] (M.A.M. Rodrigues). 0377-8401/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.anifeedsci.2006.09.009

298

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

gas production of the second phase (A2 ), were positively related to the esterified fraction of ferulic acid (FAest, P<0.05), FAeth (P<0.01) and FAtotal (P<0.05). The total amount of p-coumaric acid (PCAtotal) had a positive relationship with IVNDFD (P=0.047). In vitro DM digestibility (IVDMD) was positively correlated with FAeth and FAtotal (P<0.05). In contrast, lignin (pm) concentrations negatively correlated with DM effective degradability (ED; P=0.005) and the maximum rate of gas production of the second phase (Rmax G2 ; P=0.009). Degradation kinetic constants i.e., (c) of DM and NDF tended to be negatively correlated to the phenolic acid concentrations, mainly with the PCAeth fraction (P=0.056 and P=0.006, respectively). The same trend occurred for the fractional rate of fermentation of the second phase (R2 ; P=0.09). Principal component analysis confirmed that lignin (pm) concentration (principal component 3) is one of the major limiting factors to in vitro DM digestibility of these hays. Thus it seems that for these meadow hays, both lignin (pm) content and cross-linkages between cell wall polymers influenced rate and extent of DM degradation. © 2006 Elsevier B.V. All rights reserved. Keywords: Digestibility; Meadow hays; Lignin; Phenolic acids

1. Introduction It is widely accepted that diets containing large amounts of forage negatively affect animal production due to low digestibility of structural carbohydrates. This limitation is mainly attributed to lignification (Jung, 1989) and to covalently bound hydroxycinnamic acids (Jung and Fahey, 1983). However, recent studies have supported negative effects of lignin (Jung et al., 1997; Lavrencic et al., 1997; Fonseca et al., 1998), although published data on the effect of ferulic (FA) and p-coumaric (PCA) acids on forage digestibility are limited, and research has largely been directed towards individual plant species, distinct plant parts and even specific tissues. There is evidence that these compounds effect on digestibility are dependent on the content and bonding mode in the cell wall structure; but results are not consistent. If the majority of PCA is esterified to lignin (Jung and Deetz, 1993), and if PCA ethers are only linked to lignin (Lam et al., 1992a), it is probable that these components do not directly affect digestion. Nevertheless, the concentrations of PCA, and the ratio of PCA to FA, have been reported to have a negative effect on cell wall digestibility (Gabrielson et al., 1990; Grabber et al., 1992). Etherified FA, being a measure of cross-linking between lignin and arabinoxylans, have a negative effect on cell wall digestibility (Casler and Jung, 1999; Lam et al., 2003). However, for esterified FA the results are not consistent (Jung and Casler, 1990, 1991). More recently, Casler and Jung (2006) reported negative effects of esterified FA on in vitro 24 h neutral detergent fibre (NDF) digestibility of smooth bromegrass and reed canarygrass, but the relationship changed to positive values when digestibility was measured at 96 h. Given the relevance of these compounds and the shortage of results regarding its influence on hay digestibility, the objectives of this study were to evaluate the influence of phenolic compounds on digestibility of meadow hays, with special relevance to the study of degradation kinetics.

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

299

2. Materials and methods 2.1. Hays The meadow hays examined (MH1, MH2, MH3, MH4, MH5, MH6) were gathered during May (1996) in the Northeast region of Portugal from sixrms located in Montalegre, at a latitude varying from 41◦ 43 N to 41◦ 46 N, and longitude ranging from 7◦ 46 W to 7◦ 55 W, at an altitude ranging from 685 to 830 m. Average annual temperature was 11 ◦ C, with an average minimum temperature of 0.4 ◦ C in January, and a maximum medium temperature of 23 ◦ C in July. Average annual precipitation was 1340 mm. Dominant species within the meadow hays were red fescue (Festuca rubra L.), creeping velvetgrass (Holcus mollis L.), common velvetgrass (Holcus lanatus L.), bristle bent (Agrostis setacea Curtis), perennial ryegrass (Lolium perenne L.), and smooth brome (Bromus inermis Leyss). Legumes were present at a low proportion (w/w) representing approximately 120 g/kg of species. These values are consistent with data published by Mascarenhas-Ferreira et al. (1989) on meadow hays from the same region. At harvest, plants were judged to be at heading stage. Samples were collected after baling, choosing a hay lot from a single cutting and a single field for each farm. Bales were chosen at random and a minimum of 30 subsamples, thoroughly mixed to produce about 1 kg of a composite sample, were collected to represent the hay lot. Samples were dried in an air forced oven at 60 ◦ C for 24 h, ground to pass 4 and 1 mm screens (Retsch, Cutting mill, model SM1, Haan, Germany) and stored in airtight flasks at room temperature of 22 ◦ C. 2.2. Animals and diet Four cows, weighing 635 ± 81 kg fitted with ruminal canula (Bar Diamond Inc., Parma, Idaho, USA), were fed a diet of meadow hay shredded to 20 cm particles through a bale gripper (JN Jensen and Sommer APS, model DK 6534 Agerskov, Denmark) and 150 g/kg soybean meal offered in equal proportions at 08:00 h and 16:00 h. Cows had free access to water and to mineral–vitamin blocks. 2.3. In vitro incubations with rumen fluid From each cow, rumen fluid samples were collected 2 h after the morning feeding at 08:00 h and pooled into one pre-warmed insulated bottle filled with CO2 . Before use in the lab, rumen fluid was strained and filtered through cheesecloth. Rumen fluid used to measure gas production was mixed (1:2, v/v) with an anaerobic buffer/mineral solution (Cone et al., 1996). All laboratory handling was under continuous flushing with CO2 . Samples (400 mg) were accurately weighed into 250 ml serum bottles (Schott, Mainz, Germany) and incubated in 60 ml buffered rumen fluid. Each sample was incubated in duplicate in three series colpleted on different days. Gas production was recorded every 20 min for 72 h using a fully automated system (Cone et al., 1996). Except for the composition of the anaerobic buffer/mineral solution, as described above, the procedure proposed by Goering and Van Soest (1970) was followed to determine in vitro DM digestibility (IVDMD). Each sample was incubated in duplicate in three series,

300

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

on different days. After 48 h of incubation, flasks were removed and the NDF contents were determined. The IVDMD was calculated according as: IVDMD = Dw Cw + 0.98 Cs where Dw is the calculated NDF digestibility, Cw the NDF content, and Cs is the cell content. Gas curves were fitted by iteration to a bi-phasic model as described by Groot et al. (1996) as: Y=

Ai (1 + (Bi /t)Ci

where Ai = estimated asymptotic gas production; Bi = time of incubation at which half of the asymptotic gas production has been formed; Ci = sharpness of the switching characteristic for the profile; i = the number of the subcurve. Maximum rate of gas production (Rmax Gi ), and the time at which maximum rate of gas production was reached (TRmax Gi ), were calculated according to Yang et al. (2005) as:   Ci − 1 1/Ci Ai (Bi Ci )Ci (TRmax Gi (−Ci −1) ) Rmax Gi = , TR G = B max i i Ci + 1 (1 + (Bi Ci )(TRmax Gi (−Ci ) ))2 Fractional rate of substrate fermentation (Ri ) was calculated using the equation proposed by Groot et al. (1996) as: Ri =

Ci TRmax Gi (Ci −1) Bi Ci + TRmax Gi Ci

2.4. In situ degradation Measurements of in situ DM and NDF degradation were completed in three rumen fistulated cows using a nylon bag technique (Ørskov et al., 1980) as described by Guedes and Dias-da-Silva (1994). Each incubation was repeated once per cow. In total, there were six replicates for each sample (three cows×two times×one bag). Kinetics of in situ degradation were described by the exponential equation of Ørskov and McDonald (1979) as: Y = a + b(1 − e−ct ) where Y is the degradation after time t, a the immediately soluble fraction, b the insoluble but potentially degradable fraction, a + b the potential degradation and c is the rate constant for the degradation of fraction b (h−1 ). Effective degradability (ED) was calculated according to Ørskov and McDonald (1979) as:   bc ED = a + c+k with a, b and c as described above and k is the ruminal solids outflow rate (h−1 ), which was set arbitrarily at 0.04.

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

301

2.5. Chemical analysis Samples were analyzed for ash (method 942.05, AOAC, 1990) and total N as Kjeldahl N (method 954.01, AOAC, 1990). Neutral detergent fibre (NDFom) and acid detergent fibre (ADFom) fractions were determined by the detergent procedures of Van Soest et al. (1991) and Robertson and Van Soest (1981). The values are expressed on an ash-free basis, and sodium sulfite and heat stable amylase were not added. Lignin (pm) was determined according to Goering and Van Soest (1970). The total amount, and the esterified fraction, of ferulic (FAtotal and FAest, respectively) and p-coumaric acids (PCAtotal and PCAest, respectively) were determined in NDF residues. Both treatments were according to methodology described by Argillier et al. (1996). Chromatography separation was by high-performance liquid chromatography (HPLC) with a diode array detector (Beckman diode array detector 168, Beckman, Fullerton, USA) using a reverse-phase column (Spherisorb ODS2, 150 mm×4.6 mm, 5 ␮m). The column was operated at 30 ◦ C at a flow rate of 1.0 ml/min. Methanol (solvent A) and deionized water (solvent B) were the solvents used under gradient conditions. The first 20 min consisted in an isocratic concentration of 870 ml/l B; from 20 to 25 min a linear gradient concentration from 870 to 750 ml/l B; from 25 to 35 min an isocratic concentration of 750 ml/l B. Hydroxycinnamic acids were detected at 280 nm and the concentrations were calculated using external standards. The etherified fractions of ferulic (FAeth) and p-coumaric (PCAeth) acids were calculated as the difference between the total amount and the esterified content of these phenolic acids. 2.6. Statistical analysis Data were analyzed with the GLM procedure of SAS (1999) as a completely randomized experiment using one-way ANOVA with hay in the model statement. Least square means procedure of SAS (1999) was used to compare means. The relationships between phenolic acid concentrations and cell wall chemical composition and phenolic acid concentrations and in vitro digestion and in situ degradation constants were examined by correlation analysis using the CORR procedure of SAS (1999). Principal component analysis was completed for chemical composition data using the PRINCOMP procedure of SAS (1999). Principal component analysis provides a multivariate projection of data, analyzing the correlation structure in the data set, allowing the expression of the chemical composition variables by several components. In this way, it is possible to obtain consecutive components that are independent of each other. Stepwise regression analyses were performed using the REG procedure of SAS (1999) to develop equations, which described relationships between the principal components and the degradation constants. 3. Results 3.1. Chemical composition and degradation The chemical composition of the meadow hays is in Table 1. As expected, cell walls (NDFom) were the greatest proportion of the samples, ranging between 660 and 702 g/kg

302

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

Table 1 Chemical composition (mean ± S.D.) of meadow hays (MH) MH1

MH2

MH3

MH4

MH5

MH6

98 680 669 171

± ± ± ±

0.9 5.9 4.6 5.1

68 702 611 131

± ± ± ±

0.9 6.1 9.1 5.5

53 671 581 121

± ± ± ±

1.0 1.7 5.7 2.1

67 667 550 116

± ± ± ±

0.9 5.7 7.4 4.2

83 679 554 117

± ± ± ±

2.5 5.9 5.0 3.5

82 660 643 151

± ± ± ±

0.6 2.4 9.8 4.1

Phenolic acids (mg/g NDF)a PCAest 3.3 PCAeth 1.8 FAest 1.9 FAeth 2.6 PCAtotal 5.1 FAtotal 4.5

± ± ± ± ± ±

0.03 0.08 0.02 0.04 0.06 0.03

5.9 1.9 2.8 3.4 7.8 6.2

± ± ± ± ± ±

0.08 0.2 0.03 0.09 0.1 0.08

4.9 2.3 4.4 4.1 7.1 8.5

± ± ± ± ± ±

0.02 0.03 0.03 0.02 0.03 0.01

4.8 2.4 4.3 4.2 7.2 8.5

± ± ± ± ± ±

0.01 0.06 0.04 0.04 0.06 0.06

4.5 1.3 2.7 2.8 5.9 5.5

± ± ± ± ± ±

0.03 0.04 0.04 0.02 0.03 0.02

6.5 2.0 4.9 3.7 8.5 8.6

± ± ± ± ± ±

0.05 0.07 0.04 0.06 0.08 0.03

CP (g/kg DM) NDFom (g/kg DM) ADFom (g/kg NDF) Lignin (pm) (g/kg NDF)

All values are means (n = 4) of duplicate analysis. a The total amount and the esterified fraction of ferulic (FAtotal and FAest, respectively) and p-coumaric acids (PCAtotal and PCAest, respectively) were determined in NDF residues of samples. The etherified fractions of ferulic (FAeth) and p-coumaric acids (PCAeth) were calculated as the difference between the total amount and the esterified concentration of these phenolic acids.

DM. MH1 and MH2 had higher lignin (pm) concentrations than the others. Concentrations of the esterified and etherified fractions of PCA and FA varied widely among samples. PCA and FA total concentrations were lowest in MH1 and highest in MH6. In situ degradation characteristics of DM and NDF (Table 2) showed that the rate constant (c) for degradation of insoluble, but potentially, degradable fraction (b) had the largest variation (P<0.05) ranging from 0.021 to 0.033 h−1 and from 0.018 to 0.030 h−1 , respectively. Excepting MH5, the DM potential degradation (a + b) coefficients were higher than Table 2 Constants of ruminal dry matter (DM) and neutral detergent fibre (NDF) degradation and effective degradability (ED) of meadow hays (MH) Constanta

MH1

MH2

MH3

MH4

MH5

MH6

S.E.M.

DM a b a+b c (h−1 ) ED

0.14 b 0.56 c,d 0.70 c 0.025 b 0.36 b

0.17 c 0.55 c 0.72 c 0.028 b,c 0.39 c

0.19 d 0.58 d,e 0.77 d,e 0.021 b 0.39 c

0.17 c 0.61 e 0.78 e 0.025 b,c 0.40 c

0.18 c 0.47 b 0.65 b 0.033 c 0.39 c

0.16 c 0.60 e 0.75 d 0.023 b 0.38 b,c

0.002 0.005 0.005 0.0015 0.006

NDF a b a+b c (h−1 ) ED

0.06 d 0.48 c 0.54 d 0.023 b,c 0.24 d

0.04 c 0.55 d 0.59 e 0.023 b,c 0.24 d

0.002 b 0.41 b 0.42 b 0.018 b 0.13 b

0.0 b 0.55 d 0.55 d 0.019 b 0.18 b,c

0.0 b 0.49 c 0.49 c 0.030 c 0.21 c

0.06 d 0.52 d 0.58 e 0.019 b 0.23 d

0.002 0.006 0.006 0.0015 0.002

Means in the same row with different letters differ (P<0.05). a a: immediately soluble fraction, b: insoluble but potentially degradable fraction; a + b: potential degradation; c: rate constant for the degradation of fraction b; ED: effective degradability was calculated according to Ørskov and McDonald (1979) at ruminal outflow rate set at 0.04 h−1 .

MH1

MH2

MH3

MH4

MH5

MH6

S.E.M.

digestibilitya

In vitro IVDMD IVNDFD

In vitro gas productionb A1 (ml/g OM) B1 (h) C1 Rmax G1 (ml h−1 ) TRmax G1 (h) R1 (h−1 ) A2 (ml/g OM) B2 (h) C2 Rmax G2 (ml h−1 ) TRmax G2 (h) R2 (h−1 )

0.61 c 0.45 c

0.67 e 0.55 e

0.69 f 0.55 e

0.72 g 0.59 f

0.66 d 0.49 d

0.68 e 0.56 e

0.002 0.003

21 c 0.68 c 1.34 c 20.0 c 0.16 c 1.08 e 183 c 25.7 e 1.94 e 4.6 c 14.2 f 0.033 d,e

29 c,d 0.82 c,d,e 1.40 c,d 21.8 c,d 0.22 d 0.89 c 230 d,e 24.5 d 1.86 c,d 5.9 d 12.8 c,d,e 0.034 c,d

48 f 0.93 e 1.49 d,e 31.7 e 0.31 e 0.79 c 262 e 26.3 d 1.68 c 6.1 d 11.4 c,d 0.029 c

38 e 0.77 cd 1.45 c,d,e 30.8 e 0.23 d 0.95 c,d 253 d,e 24.8 d 1.94 d 6.5 d 13.7 d,e 0.034 c,d

36 d,e 0.79 c,d,e 1.42 c,d,e 28.0 d,e 0.23 d 0.91 c,d 223 d 20.6 c 1.81 c,d 6.8 d 10.4 c 0.039 d

30 d 0.88 d,e 1.54 e 20.4 c 0.32 e 0.84 c 239 d,e 25.5 d 1.88 c,d 6.0 d 13.5 d,e 0.032 c

1.8 0.033 0.026 1.48 0.014 0.042 7.4 0.83 0.058 0.20 0.61 0.0014

Means in the same row with different letters differ (P<0.05). a IVDMD: in vitro DM digestibility; IVNDFD: in vitro NDF digestibility. b A : estimated asymptotic gas production of the first phase; B : time of incubation at which half of the asymptotic gas production of the first phase has been formed; 1 1 C1 : sharpness of the switching characteristic for the profile of the first phase; Rmax G1 : maximum rate of gas production of the first phase; TRmax G1 : time at which maximum rate of gas production of the first phase is reached, R1 : fractional rate of fermentation of first phase; A2 : estimated asymptotic gas production of the second phase; B2 : time of incubation at which half of the asymptotic gas production of the second phase has been formed; C2 : sharpness of the switching characteristic for the profile of the second phase; Rmax G2 : maximum rate of gas production of the second phase; TRmax G2 : time at which maximum rate of gas production of the second phase is reached, R2 : fractional rate of fermentation of second phase.

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

Table 3 In vitro digestibility coefficients and in vitro gas production constants of meadow hays (MH)

303

304

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

0.70. The in vitro digestibility of DM and NDF are in Table 3. MH4 had 15.3% higher IVDMD and 23.7% higher IVNDFD than the hay with the lowest values, MH1 (P<0.05). The same trend occurred for the estimated asymptotic gas production of the second phase (A2 ; Table 3) with MH1 having the lowest values (P<0.05). On the contrary, the time at which maximum rate of gas production of the second phase is reached (TRmax G2 ) was higher for MH1 (Table 3). Results are in accordance with data for feeds with a high contribution of the NDF fraction to overall degradation and fermentation. Similar results were obtained by Rodrigues et al. (2002) who studied kinetics of digestion of 12 forages, including 7 cereal straws and 2 meadow hays, 2 oat grain hays and 1 Italian ryegrass hay. 3.2. Relationships of cell wall composition with degradation The correlation between phenolic acid concentrations and in situ degradation constants (Table 4) shows that the DM a + b constant was positively related to FAeth (P=0.012) and PCAeth (P<0.001) and with FAtotal (P=0.033). The b constant of DM also had a positive relationship with PCAeth (P=0.008). In contrast, there was a negative relationship (P=0.005) of effective degradability (ED) with lignin (pm) concentration. Even though not statistically significant, rate of DM disappearance (c) tended (P=0.056) to have a negative correlation with the PCAeth fraction (r = −0.80). But the rate of NDF disappearance

Table 4 Pairwise correlation coefficients between phenolic acid and lignin concentrations and in situ DM and NDF degradation constants Constantb

Phenolic acid and lignin (pm) concentrationsa FAest

FAeth

PCAest

PCAeth

FAtotal

PCAtotal

Lignin (pm)

DM a b a+b c ED

0.34 0.64 0.77 −0.60 0.35

0.51 0.71 0.91* −0.62 0.59

0.23 0.29 0.38 −0.16 0.38

0.15 0.92** 0.99*** −0.80 0.28

0.41 0.69 0.85* −0.62 0.45

0.25 0.53 0.63 −0.38 0.42

−0.89* 0.20 −0.14 −0.28 −0.94**

NDF a b a+b c ED

−0.14 −0.01 −0.07 −0.71 −0.15

−0.39 −0.02 −0.16 −0.79 −0.20

0.23 0.42 0.43 −0.32 0.28

−0.11 0.07 −0.01 −0.94** 0.12

−0.24 −0.01 −0.10 −0.76 −0.17

0.17 0.39 0.38 −0.57 0.28

0.90* −0.01 0.39 −0.10 0.77

* P<0.05; ** P<0.01; *** P<0.001. a The total amount and the esterified fraction of ferulic (FAtotal and FAest, respectively) and p-coumaric acids (PCAtotal and PCAest, respectively) were determined in NDF residues of samples. The etherified fractions of ferulic (FAeth) and p-coumaric acids (PCAeth) were calculated as the difference between the total amount and the esterified content of these phenolic acids. b a: immediately soluble fraction; b: insoluble but potentially degradable fraction; a + b: potential degradation; c: rate constant for the degradation of fraction b; ED: effective degradability was calculated according to Ørskov and McDonald (1979) at ruminal outflow rate set at 0.04 h−1 .

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

305

Table 5 Pairwise correlation coefficients between phenolic acid and lignin concentrations and in vitro digestion constants Constant

Phenolic acid and lignin (pm) concentrationsa FAest

In vitro digestibilityb IVDMD IVNDFD

0.79 0.82*

In vitro gas productionc A1 (ml/g OM) 0.57 B1 (h) 0.74 0.94** C1 0.37 Rmax G1 (ml h−1 ) 0.89* TRmax G1 (h) −1 R1 (h ) −0.72 A2 (ml/g OM) 0.85* B2 (h) 0.37 −0.28 C2 0.47 Rmax G2 (ml h−1 ) TRmax G2 (h) 0.02 −0.46 R2 (h−1 )

FAeth

PCAest

PCAeth

FAtotal

PCAtotal

0.91* 0.93**

0.52 0.66

0.67 0.79

0.86* 0.89*

0.66 0.82*

0.70 0.61 0.68 0.59 0.67 −0.62 0.92** 0.47 −0.28 0.46 0.07 −0.55

0.11 0.67 0.72 −0.18 0.71 −0.72 0.54 0.10 −0.15 0.40 −0.03 −0.19

0.39 0.31 0.42 0.34 0.39 −0.29 0.65 0.76 0.01 0.04 0.48 −0.73

0.64 0.71 0.88* 0.46 0.84* −0.71 0.90* 0.42 −0.29 0.48 0.04 −0.51

0.22 0.69 0.75 −0.05 0.75 −0.72 0.67 0.32 −0.13 0.36 0.12 −0.38

Lignin (pm) −0.78 −0.56 −0.80 −0.42 −0.30 −0.84* −0.35 0.53 −0.73 0.44 0.46 −0.92** 0.65 −0.25

* P<0.05; ** P<0.01. a

The total amount and the esterified fraction of ferulic (FAtotal and FAest, respectively) and p-coumaric acids (PCAtotal and PCAest, respectively) were determined in NDF residues of samples. The etherified fractions of ferulic (FAeth) and p-coumaric acids (PCAeth) were calculated as the difference between the total amount and the esterified content of these phenolic acids. b IVDMD: in vitro DM digestibility; IVNDFD: in vitro NDF digestibility. c A : estimated asymptotic gas production of the first phase; B : time of incubation at which half of the asymptotic 1 1 gas production of the first phase has been formed; C1 : sharpness of the switching characteristic for the profile of the first phase; Rmax G1 : maximum rate of gas production of the first phase; TRmax G1 : time at which maximum rate of gas production of the first phase is reached, R1 : fractional rate of fermentation of first phase; A2 : estimated asymptotic gas production of the second phase; B2 : time of incubation at which half of the asymptotic gas production of the second phase has been formed; C2 : sharpness of the switching characteristic for the profile of the second phase; Rmax G2 : maximum rate of gas production of the second phase; TRmax G2 : time at which maximum rate of gas production of the second phase is reached, R2 : fractional rate of fermentation of second phase.

(c) had a negative relationship (r = −0.94) with PCAeth (P=0.006) and tended (P=0.061) to have a negative relationship (r = −0.76) with FAtotal. The same tendency for negative relationships between degradation kinetics and phenolic concentrations and positive correlation coefficients between total degradation measurements and phenolic acids was observed for the in vitro digestibility and gas production (Table 5). The IVNDFD and A2 values were positively related to FAest (P<0.05), FAeth (P<0.01) and FAtotal (P<0.05). PCAtotal also had a positive relationship (P=0.047) with IVNDFD. IVDMD was positively correlated with FAeth and FAtotal (P<0.05). In contrast, PCAeth concentration tended (P=0.09) to be negatively correlated to the fractional rate of fermentation of second phase (R2 ). The time at which maximum rate of gas production of the second phase is reached (Rmax G2 ) showed a negative relationship with the lignin (pm) concentration (P=0.009).

306

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

4. Discussion 4.1. Cell wall composition Similar phenolic acid concentrations of meadow hays were observed by Susmel et al. (1997) in 33 Mediterranean hays and straws, who reported maximum values of 6.5 and 6.2 g/kg NDF for ferulic and p-coumaric acids of wheat, barley an triticale straws, and substantially higher values of 7.7 and 9.3 g/kg NDF for ferulic and p-coumaric acid concentrations in switchgrass and beesgrass hays. Our data for FAest and PCAest are lower than the values reported by Bourquin et al. (1994) for orchardgrass hay (4.8 and 7.5 g/kg NDF, respectively) but, combined with FAeth, are within the range of concentrations described by Casler and Jung (2006) for smooth bromegrass, cocksfoot, and reed canarygrass. For these three perennial grasses, the authors reported values ranging from 3.55 to 7.02, 1.56 to 5.35, and 1.69 to 5.35 g/kg NDF for FAest, PCAest and FAeth, respectively. Data on PCAeth of temperate grasses are rare, and our values are similar to the concentrations reported for tropical grasses. Morrison and Mulder (1994) found PCAeth concentrations of 2.1 g/kg cell wall and Mandebvu et al. (1999a) noted values of approximately 1.0 g/kg cell wall for coastal bermudagrass. Our higher values may be due to the advanced stage of maturation of the hays since it is accepted that PCA concentrations will increase with the vegetative cycle. Moreover, there are some difficulties in hydrolyzing PCA esters when lignified cell walls are analyzed (Hartley and Morrison, 1991), which would lead to high levels of PCAether due to incomplete release of esterified PCA by treatment with sodium hydroxide. No relationships occurred between phenolic acid concentrations and other cell wall components. However, the correlation matrix showed somewhat surprising results between lignin (pm) and PCAest (r = −0.20) and PCAeth (r = −0.15) concentrations. Ever week negative relationships between these components were not expected since it is accepted that most PCA is esterified to lignin and that the concentration of etherified PCA increases with the maturation process of grasses (Morrison et al., 1998), the same trend as lignin deposition, and that PCA is incorporated to the lignin polymer through ether linkages during secondary cell wall development (Lam et al., 1992a). Our meadow hays were mainly grasses at late stages of growth, and so the extensive lignification process, resulting from development of secondary cell walls, should have been positively related to PCA concentrations. In spite of similar negative relationships between lignin, determined as either Klason lignin or as lignin (sa), and the esterified and etherified fractions of p-coumaric acid were reported by Casler and Jung (1999) with 32 smooth bromegrass clones, our data may reflect the solubility of lignin due to application of the acid detergent solution prior to the permanganate, a serious drawback already mentioned by Hatfield et al. (1994). Since PCA may represent a considerable portion of the Klason lignin fraction, with values up to 180 g/kg of bromegrass lignin (Fukushima and Hatfield, 2004), this process, that dissolves part of the composite between PCA and the lignin molecule, could lead to underestimation of the lignin concentrations of lignified samples. These authors also reported lignin (pm) values of smooth bromegrass, setaria, and oat and wheat straws that were about half of the values obtained with Klason and acetyl bromide soluble lignin methods. In this way, considering the lignin (pm) and PCA concentrations (Table 1), it is possible that small variations in concentrations of lignin could have lead to negative relationships between those components.

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

307

Another possible explanation relates to the presence of legumes in the meadow hays. This possible contamination may also be important in explaining the results as legumes have smaller amounts of ferulate and p-coumarate esters (Jung and Deetz, 1993) than mature grasses. In addition, NDF is also lower in legumes while lignin is higher (Buxton and Russel, 1988). Moreover, no ferulate-mediated cross-links of lignin to cell wall polysaccharides have been observed in legumes. Therefore, the possible presence of legumes will add lignin without increasing p-coumarates and ferulates, thereby distorting the accumulation relationships of lignin with p-coumarates and ferulate in these meadow hays. 4.2. Relationships of cell wall composition with degradation The positive correlations between measures of total degradation and FAest, and the negative relationships between total degradation and lignin concentrations (Tables 4 and 5), are consistent with data of other authors. The relationship between lignin concentration and forage digestibility has been characterized for over 60 years, and the negative effects of cell wall lignification have been extensively discussed (Jung and Fahey, 1983; Jung, 1989; Jung and Deetz, 1993). The advanced stage of maturity of our meadow hays probably explains the negative correlations obtained. In a study comparing different bromegrass plants, Jung and Allen (1995) observed that plants with higher concentrations of FAest also had higher digestibilities. More recently, Mandebvu et al. (1999b) reported that cultivars of bermuda grass hays, with higher concentrations of ester-linked ferulic acid, also had higher OM, NDF and ADF digestibilities. These data show that FAest does not limit extent of digestion, and the positive relationships may only reflect deposition of this hydroxycinnamic acid during primary cell wall development that accompanies incorporation of other cell wall components. The positive effect of FAeth fraction on cell wall digestibility is less clear. Opposite correlations to the ones expected were also obtained for relationships between PCAest and PCAeth fractions and the degradation constants of DM and NDF. Contrary to the FAest fraction that is susceptible to removal by microbial esterases (Deetz et al., 1993), the linkage between FAeth and the lignin structure cannot be broken under anaerobic conditions. As ferulate molecules cross-link arabinoxylans and lignin through ester and ether linkages (Jung and Deetz, 1993), thereby creating a structure that prevents hydrolysis, the FAeth fraction should have a negative effect on forage digestibility. Nevertheless, available data on forages does not always show such direct evidence. Jung and Vogel (1992) demonstrated that ferulate ether concentrations only had occasional negative relationships with NDF digestibility for a series of observations within and among maturity stages and plant parts of switchgrass and big bluesteam. More recently, Jung et al. (1998) reported negative correlations between etherified ferulic acid and cell wall polysaccharides degradability in young maize internodes. However, for mature internodes that had ceased to elongate, no relationship occurred. One of the possible explanations for these poor correlations, suggested by these authors, is the existence of a masking effect of secondary cell wall development over the ether fraction influence on digestibility. During the initial phases of cell wall development, FAest is deposited and may serve as nucleation sites for lignin polymerization (Ralph et al., 1995). During maturation of the cell wall cross-linkages with lignin through incorporation of FAeth will occur, but concentrations of both components will decrease during

308

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

plant growth while deposition of lignin and other polysaccharides increase. This dilution effect can mask the impact of ferulate cross-linking, measured by FAeth concentration, on digestibility. Our data may not only reflect this hypothesis, but also the interdependence that exists between the hydroxycinnamic acid concentrations in the cell wall (data not shown) given the correlation between FAest and FAeth (r = 0.87; P=0.025). In the same way, the correlation between PCAeth and Faeth (r = 0.89; P=0.018) may support the relationship between p-coumaric acid fractions and degradation constants (Tables 4 and 5). In fact, in spite of it being considered that the effect of PCAest on polysaccharide digestion is indirect due to the linkages with lignin (Jung and Allen, 1995) and that the PCAeth has no effect on digestibility, it may only be linked to lignin and not act as a linkage to polysaccharide (Lam et al., 1992b), and so the relationship we obtained should have been negative. Data in Tables 4 and 5 show a tendency to negative relationships between hydroxycinnamic acids and rates of DM and NDF degradation. Independent of the methodology used, it seems that these acids exert an inhibitory effect on degradation kinetics. Nevertheless, the same reasoning can be made regarding the lack of independence within hydroxycinnamic acids concentrations. In this way, the correlation studies may not reflect a cause and effect relationship. To explain the inter-relationships found, principal component analysis was conducted on cell wall components. To create new variables that combine into several consecutive components that maximize variability. Each consecutive component is defined to maximize the variability that is not captured by the preceding component, with consecutive components being independent of each other, allowing the utilization of these new variables to explain differences on degradation mechanisms of meadow hays. The coefficients for the cell wall components in the first five principal components are in Table 6, and the first three principal components explained 91% of the variation in the data set. Based on the size and sign of the coefficients of principal components the following interpretations of cell wall structure characteristics are proposed. The first principal component (PC1) appears to be a measure of phenolic acid concentrations in the cell wall, given the Table 6 Coefficients for the four primary principal components for cell wall composition Cell wall componenta

NDF Lignin (pm) FAest FAeth PCAest PCAeth FAtotal PCAtotal Proportion of variation (R2 ) a

Principal component 1

2

3

4

−0.301 −0.180 0.427 0.407 0.304 0.336 0.434 0.368 0.65

0.266 0.151 −0.040 −0.246 0.682 −0.344 −0.116 0.498 0.14

−0.422 0.874 0.127 −0.168 −0.027 0.115 0.023 0.010 0.12

0.632 0.296 −0.201 0.239 −0.041 0.624 −0.047 0.150 0.08

The total amount and the esterified fraction of ferulic (FAtotal and FAest, respectively) and p-coumaric acids (PCAtotal and PCAest, respectively) were determined in NDF residues of samples. The etherified fractions of ferulic (FAeth) and p-coumaric acids (PCAeth) were calculated as the difference between the total amount and the esterified content of these phenolic acids.

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

309

rather equivalent positive coefficients for all phenolic acids. Principal component two (PC2) is related to the concentration of PCAest in the cell wall. The third principal component (PC3) is a measure of cell wall lignification. Principal component four (PC4) appears to be a measure of total cell wall deposition, being also related to the PCAeth concentration. When principal components were individually regressed with constants of degradation, PC1 was correlated to a + b (r = 0.79; P=0.037) and A2 (r = 0.87; P=0.016) constants, as well to IVDMD (r = 0.84; P=0.024) and to IVNDFD (r = 0.91; P=0.008). Negative relationships occurred between PC3 and in situ DM ED (r = −0.83; P=0.025). This could indicate that lignin concentration is one of the major limiting factors to in vitro digestion of these hays. In addition, the overall concentration of hydroxycinnamic acids, without any relevant indication of linkages with other cell wall components, does not allow a clear distinction of its effect on cell wall degradation. In order to explain most comprehensively the cell wall degradation mechanisms, a backwards stepwise multiple regression analysis was completed. The best-fitting regression model showed that 94% of the variation in DM potential degradation (a + b) was explained (P=0.037) when PC1, PC2 and PC4 were retained. While PC1 and PC4 had positive relations, PC2 was negatively related to a + b-constant indicating that PCAest has a negative effect in hays DM degradation. Lignin (pm) concentration (PC3) again had a negative dominant effect on degradation and was retained (negatively related) in the model (P=0.002) that explained 98% of the variation in DM effective degradability (ED) together with PC1 (positive relation). Showing the same tendency, PC1 and PC3 were inversely retained in models that explained 94% and 92% of IVDMD (P=0.008) and A2 -constant (P=0.010), respectively. These results agree with the view that, in mature forages, lignin is the primary factor limiting cell wall degradation (Jung and Deetz, 1993). However, the inhibitory influence of hydroxycinnamic acids on degradation was not so clear and it seems that the evaluation of its influence should not be done using only simple correlation procedures. After principal component analysis was used, the effect of PCAest and FAest + FAeth on cell wall degradation showed a negative relationship. These results seem to be consistent with the hypothesis that the effect of hydroxycinnamic acids can easily be confounded by increased deposition of lignin during plant maturation, thereby indicating that the impact of ferulate cross-linking on digestibility cannot be based purely on concentrations, especially when analyzing highly lignified plants. The tendency for a negative effect of the hydroxycinnamic acids on degradation kinetics (Table 4) was not confirmed by multiple regression procedures, which may be related to the overall regulating factor that covalent associations between cell wall components have on degradation.

5. Conclusions Results are consistent with the hypothesis that cross-linking of lignin to polysaccharides may be an important factor limiting the rate and extent of cell wall degradation in the rumen. However, because of the complexity of cell wall development, there are associations between lignin and hydroxycinnamic acids deposition with digestibility measures that cannot be accounted through simple correlations or regressions between variables. Additionally, it

310

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

seems that concentrations of p-coumarates and ferulates do not provide useful information to predict digestibility of meadow hays. References AOAC, 1990. Official Methods of Analysis, 14th ed. Association of Official Analytical Chemists, Washington, DC, USA. Argillier, O., Barri´ere, Y., Lila, M., Jeanneteau, F., G´elinet, K., M´enanteau, V., 1996. Genotypic variation in phenolic components of cell walls in relation to the digestibility of maize stalks. Agronomie 16, 123–130. Bourquin, L.D., Titgemeyer, E.C., Van Milgen, J., Fahey Jr., G.C., 1994. Forage level and particle size effects on orchardgrass digestion by steers. II. Ruminal digestion kinetics of cell wall components. J. Anim. Sci. 72, 759–767. Buxton, D.R., Russel, J.R., 1988. Lignin constituents and cell-wall digestibility of grass and legume stems. Crop Sci. 28, 553–558. Casler, M.D., Jung, H.G., 1999. Selection and evaluation of smooth bromegrass clones with divergent lignin or etherified ferulic acid concentration. Crop Sci. 39, 1866–1873. Casler, M.D., Jung, H.G., 2006. Relationships of fibre, lignin, and phenolics to in vitro fibre digestibility in three perennial grasses. Anim. Feed Sci. Technol. 125, 151–161. Cone, J.W., Van Gelder, A.H., Visscher, G.J.W., Oudshoorn, L., 1996. Influence of rumen fluid and substrate concentration on fermentation kinetics measured with a fully automated time related gas production apparatus. Anim. Feed Sci. Technol. 61, 113–128. Deetz, D.A., Jung, H.J.G., Helm, R.F., 1993. Impact of methyl-5-O-(E)-feruloyl-␣-1-arabinofuranoside on in vitro degradation of cellulose and xylan. J. Sci. Food Agric. 61, 423–427. Fonseca, A.J.M., Dias-da-Silva, A.A., Orskov, E.R., 1998. In sacco degradation characteristics as predictors of digestibility and voluntary intake of roughages by mature ewes. Anim. Feed Sci. Technol. 72, 205–219. Fukushima, R.S., Hatfield, R.D., 2004. Comparison of the acetyl bromide spectrophotometric method with other analytical lignin methods for determining lignin concentration in forage samples. J. Agric. Food Chem. 52, 3713–3720. Gabrielson, B.C., Vogel, K.P., Anderson, B.E., Ward, J.K., 1990. Alkali-labile cell-wall phenolics and forage quality in switchgrass selected for differing digestibility. Crop Sci. 30, 1313–1320. Goering, H.K., Van Soest, P.J., 1970. Forage fiber analyses (apparatus, reagents, procedures, and some applications). In: Agric. Handbook No. 379. ARS-USDA, Washington, DC, USA. Grabber, J.H., Jung, G.A., Abrams, S.M., Howard, D.B., 1992. Digestion kinetics of parenchyma and sclerenchyma cell walls isolated from orchardgrass and switchgrass. Crop Sci. 32, 806–810. Groot, J.C.J., Cone, J.W., Williams, B.A., Debersaques, F.M.A., Latinga, E.A., 1996. Multiphasic analysis of gas production kinetics for in vitro fermentation of ruminant feeds. Anim. Feed Sci. Technol. 64, 77–89. Guedes, C.M., Dias-da-Silva, A.A., 1994. Effects of fish-meal supplementation on the digestion and rumen degradation of ammoniated wheat straw. Ann. Zootech. 43, 333–340. Hartley, R.D., Morrison III., W.H., 1991. Monomeric and dimeric phenolic acids released from cell walls of grasses by sequential treatment with sodium hydroxide. J. Sci. Food Agric. 55, 365–375. Hatfield, R.D., Jung, H.G., Ralph, J., Buxton, D.R., Weimer, P.J., 1994. A comparison of the insoluble residues produced by the klason lignin and acid detergent lignin procedures. J. Sci. Food Agric. 65, 51–58. Jung, H.G., 1989. Forage lignins and their effects on fiber digestibility. Agron. J. 81, 33–38. Jung, H.G., Allen, M.S., 1995. Characteristics of plant cell walls affecting intake and digestibility of forages by ruminants. J. Anim. Sci. 73, 2774–2790. Jung, H.G., Casler, M.D., 1990. Lignin concentration and composition of divergent smooth bromegrass genotypes. Crop Sci. 30, 980–985. Jung, H.G., Casler, M.D., 1991. Relationship between lignin and esterified phenolics to fermentation of smooth bromegrass fibre. Anim. Feed Sci. Technol. 32, 63–68. Jung, H.G., Deetz, D.A., 1993. Cell wall lignification and degradability. In: Jung, H.J., Buxton, D.R., Hatfield, R.D., Ralph, J. (Eds.), Forage Cell Wall Structure and Digestibility. American Society of Agronomy, Madison, WI, USA, pp. 315–346.

M.A.M. Rodrigues et al. / Animal Feed Science and Technology 136 (2007) 297–311

311

Jung, H.G., Fahey Jr., G.C., 1983. Nutritional implications of phenolic monomers and lignin: a review. J. Anim. Sci. 57, 206–219. Jung, H.G., Vogel, K.P., 1992. Lignification of switchgrass (Panicum virgatum) and big bluestem (Andropogon gerardii) plant parts during maturation and its effect on fiber degradability. J. Sci. Food Agric. 59, 769–776. Jung, H.G., Mertens, D.R., Payne, A.J., 1997. Correlation of acid detergent lignin and klason lignin with digestibility of forage dry matter and neutral detergent fiber. J. Dairy Sci. 80, 1622–1628. Jung, H.G., Morrison, T.A., Buxton, D.R., 1998. Degradability of cell-wall polysaccharides in maize internodes during stalk development. Crop Sci. 38, 1047–1051. Lam, T.B.T., Iiyama, K., Stone, B.A., 1992a. Cinnamic acid bridges between cell wall polymers in wheat and Phalaris internodes. Phytochemistry 31, 1179–1183. Lam, T.B.T., Iiyama, K., Stone, B.A., 1992b. Changes in phenolic acids from internode walls of wheat and Phalaris during maturation. Phytochemistry 31, 2655–2658. Lam, T.B.T., Iiyama, K., Stone, B.A., 2003. Hot alkali-labile linkages in the wall of the forage grass Phalaris aquatica and Lolium perenne and their relation to in vitro wall digestibility. Phytochemistry 64, 603–607. Lavrencic, A., Stefanon, B., Susmel, P., 1997. An evaluation of the gompertz model in degradability studies of forage chemical components. Anim. Sci. 64, 423–431. Mandebvu, P., West, J.W., Froetschel, M.A., Hatfield, R.D., Gates, R.N., Hill, G.M., 1999a. Effect of enzyme or microbial treatment of bermudagrass forages before ensiling on cell wall composition, end products of silage fermentation and in situ digestion kinetics. Anim. Feed Sci. Technol. 77, 317–329. Mandebvu, P., West, J.W., Hill, G.M., Gates, R.N., Hatfield, R.D., Mullinix, B.G., Parks, A.H., Caudle, A.B., 1999b. Comparison of tifton 85 and coastal bermudagrass for yield, nutrient traits, intake, and digestion by growing beef steers. J. Anim. Sci. 77, 1572–1586. Mascarenhas-Ferreira, A., Guedes, C.V.M., Dias-da-Silva, A.A., 1989. Effects of urea treatment on chemical composition and in vitro digestibility of meadow hays of Northern Portugal. Anim. Feed Sci. Technol. 25, 157–167. Morrison III, W.H., Mulder, M.M., 1994. Pyrolysis mass spectrometry and pyrolysis gas chromatography–mass spectrometry of ester- and ether-linked phenolic acids in coastal bermudagrass cell walls. J. Anim. Sci. 35, 1143–1151. Morrison, T.A., Jung, H.G., Buxton, D.R., Hatfield, R.D., 1998. Cell-wall composition of maize internodes of varying maturity. Crop Sci. 38, 455–460. Ørskov, E.R., McDonald, I., 1979. The estimation of protein degradability in the rumen from incubation measurements weighted according to rate of passage. J. Agric. Sci. 92, 499–503. Ørskov, E.R., DeB Hovell, F.D., Mould, F., 1980. The use of the nylon bag technique for the evaluation of feedstuffs. Trop. Anim. Prod. 5, 195–213. Ralph, J., Grabber, J.H., Hatfield, R.D., 1995. Lignin-ferulate cross-links in grasses: active incorporation of ferulate polysaccharide esters into ryegrass lignins. Carbohydr. Res. 275, 167–178. Robertson, J.B., Van Soest, P.J., 1981. The detergent system of analysis and its application in human foods. In: James, W.P.T., Theander, O. (Eds.), The Analysis of Dietary Fiber in Food. Marcell Dekker Inc., New York, NY, USA, pp. 123–158. Rodrigues, M.A.M., Fonseca, A.M., Sequeira, C.A., Dias-da-Silva, A.A., 2002. Digestion kinetic parameters from an in vitro g´as production method as predictors of voluntary intake of forage by mature ewes. Anim. Feed Sci. Technol. 95, 133–142. SAS, 1999. SAS/STAT User’s Guide. Version 8. SAS Institute, Cary, NC, USA. Susmel, P., Guzzon, P., Stefanon, B., Mills, C.R., 1997. Relationships between plant cell wall composition and in situ degradability of Mediterranean feeds. Agric. Med. 127, 82–91. Van Soest, P.J., Robertson, J.B., Lewis, B.A., 1991. Symposium: carbohydrate methodology, metabolism, and nutritional implications in dairy cattle. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74, 3583–3597. Yang, H.J., Tamminga, S., Williams, B.A., Dijkstra, J., Boer, H., 2005. In vitro gas and volatile fatty acids production profiles of barley and maize and their soluble and washout fractions after feed processing. Anim. Feed Sci. Technol. 120, 125–140.