Plant samples for extracting condensed tannins

Blackcurrant (Ribes nigrum) and redcurrant (Ribes rubrum) leaves were collected at Hildred pick your own (PYO)-farm, Goring-on-Thames, UK; goat willow (Salix caprea) leaves and twigs were collected from Goring-on-Thames, UK; weeping willow (Salix babylonica) catkins were collected on Evesham Rd., Emmer Green, Reading, UK; white clover (Trifolium repens) flowers were collected from the National Institute of Agricultural Botany (NIAB, Cambridge, UK); whole sainfoin (Onobrychis viciifolia, var. Esparsette) plants were collected from Barham, Kent, UK; pine bark was provided by University of Turku, Finland. After collection, the plant materials were freeze-dried and then ground to pass a 1-mm sieve using an impeller mill (Retsch GmbH, SM1, Haan, Germany).

Extraction of condensed tannins

The CT extracts were prepared according to a slightly modified method of Gea et al. (Reference Gea, Stringano, Brown and Mueller-Harvey2011). Finely ground plant powder (50 g) was extracted with acetone/water (500 ml; 7 : 3; v/v) under constant stirring for 60 min. The mixture was transferred into a Buchner funnel, fitted with a Whatman filter paper and filtered under vacuum. The filtrate was extracted with dichloromethane (CH₂Cl₂) to remove lipids and chlorophyll and concentrated in a rotary evaporator at <37 °C. The remaining aqueous solution was centrifuged for 6 min at 4500 rpm (Thermo Electron Corporation, Jouan CR3i Multifunction Centrifuge, Basingstoke, UK) in order to remove residual chlorophyll and any other insoluble material. Extracts were frozen, freeze-dried and stored at −20 °C.

Analysis of condensed tannin extracts

Extracts were analysed for CT content and structural properties by thiolytic degradation, followed by high performance liquid chromatography (HPLC) analysis (Novobilský et al. Reference Novobilský, Mueller-Harvey and Thamsborg2011; Williams et al. Reference Williams, Fryganas, Ramsay, Mueller-Harvey and Thamsborg2014). This provided information on the proportion of flavan-3-ols (catechin, epicatechin, gallocatechin and epigallocatechin) in the CT terminal and extension units (Fig. 1). In addition, it allowed calculation of the mean degree of polymerization (mDP), prodelphinidins : procyanidins (PD : PC) and cis : trans flavan-3-ol ratio in the CT polymers based on the following formulae (1, 2, 3) (Gea et al. Reference Gea, Stringano, Brown and Mueller-Harvey2011) (Table 1):

(1) $${\rm mDP} = \displaystyle{\matrix{\hbox{sum of extension and terminal} \hfill \cr \quad \hbox{flavan-3-ol units} ({\rm mol}) \hfill} \over {\hbox{sum of terminal flavan-3-ol units} ({\rm mol})}}$$

(2) $${\rm PD} = \displaystyle{{{\rm Proportion\; of\; GC} + {\rm EGC\; units}} \over {{\rm Proportion\; of\; C} + {\rm EC} + {\rm GC} + {\rm EGC\; units}}}$$

(3) $$cis = \displaystyle{{{\rm Proportion\; of\; EC} + {\rm EGC\; units}} \over {{\rm Proportion\; of\; C\;} + {\rm GC} + {\rm EC} + {\rm EGC\; units}}}$$

where C = catechin, EC = epicatechin, GC = gallocatechin and EGC = epigallocatechin flavan-3-ols, proportion of PD + PC = 100; proportion of cis + trans = 100.

Fig. 1.

Thiolytic degradation of condensed tannin polymers. Extension subunits are released as flavan-3-ol benzyl mercaptan (BM) adducts, terminal subunits are released as the free flavan-3-ols (Gea et al. Reference Gea, Stringano, Brown and Mueller-Harvey2011).

Table 1.

Condensed tannin (CT) content, mean degree of polymerization (mDP), proportion of prodelphinidins and cis in extracts obtained from eight plant sources

Values are means ± standard deviation.

Experimental design

The effects of CT structural features on CH₄ production and fermentation kinetics during in vitro incubation were evaluated using tannin-free grass silage as a control substrate. The chemical composition of grass silage was: DM = 937·0 g/kg; organic matter (OM) = 911·3 g/kg DM; crude protein (CP) = 151·4 g/kg DM, neutral detergent fibre (NDF) = 498·5 g/kg DM. Condensed tannin were added to the substrate at an effective concentration of 40 g CT/kg of substrate in the presence (+PEG) or absence (−PEG) of polyethylene glycol (PEG 6000) to inactivate the tannins (Makkar et al. Reference Makkar, Blümmel and Becker1995). Condensed tannins (10 mg) and substrate (250 mg) with or without PEG (100 mg; CT : PEG = 1 : 10 w/w) (Pellikaan et al. Reference Pellikaan, Stringano, Leenaars, Bongers, Van Laar-Van Schuppen, Plant and Mueller-Harvey2011b ) were weighed into duplicate 250 ml bottles (Schott bottle, GL45, Mainz, Germany) per tannin extract within each run, with two separate runs. Test substrate was incubated with a mixture of rumen liquid collected from three different rumen fistulated lactating Holstein-Friesian dairy cows per run (i.e. total of six rumen fistulated cows). These cows were fed a grass and maize silage mixture in the morning and afternoon and 9 kg of concentrate according to their requirements. The handling of the animals was approved by the Institutional Animal Care and Use Committee of Wageningen University, Wageningen, the Netherlands and in accordance with Dutch legislation on the use of experimental animals. Rumen fluid was collected before the morning feeding in pre-warmed thermos flasks, which were filled with carbon dioxide (CO₂) and transported directly to the nearby laboratory. All further manipulations were done under CO₂ to ensure anaerobic conditions. The rumen fluid was pooled and filtered through two layers of cheesecloth into a flask flushed with CO₂. Filtered rumen fluid was mixed with a buffer solution with constant stirring and continuous flushing with CO₂ and maintained in a water bath at 39 °C. Buffer solution was made as described by Williams et al. (Reference Williams, Bosch, Boer, Verstegen and Tamminga2005). After adding 30 ml of the buffered rumen liquid mixture, bottles were directly placed in a shaking water bath at 39 °C and connected to an automated time-related gas production (GP) system (Pellikaan et al. Reference Pellikaan, Hendriks, Uwimana, Bongers, Becker and Cone2011a ) and GP was measured over 72 h. Gas production in blanks (i.e. buffered rumen fluid mixture without substrate and CT extracts) was 2·7 ± 0·21 ml for run1 and 9·9 ± 0·35 ml for run 2. Condensed tannin extracts consisted of 46·2–87·4 (g/100 g extracts) of non-CT compounds, which may contribute to fermentation. Therefore, blanks were also included containing CT extracts only, with and without PEG. The average GP in CT extracts with and without PEG in run 1 was 10·4 and 4·8 ml, and in run 2 were 16·7 and 4·8 ml, respectively. After 72 h incubation, the fermentation fluid pH was recorded (Mettler Toledo FE20/EL20 pH meter, Schwerzenbach, Switzerland) and fermentation fluid from each bottle was collected for volatile fatty acid (VFA) and ammonia (NH₃) analysis.

In vitro gas and methane production

Total cumulative gas (GP) and methane (GP_CH4) production were measured using an automated GP system at the laboratory of the Animal Nutrition Group of Wageningen University, the Netherlands (Pellikaan et al. Reference Pellikaan, Hendriks, Uwimana, Bongers, Becker and Cone2011a ). Methane concentration in the headspace of the fermentation bottle was measured by gas chromatography (GC8000Top, CE Instruments, Milan, Italy). Fermentation bottles were modified (Pellikaan et al. Reference Pellikaan, Hendriks, Uwimana, Bongers, Becker and Cone2011a ) to sample CH₄ from the headspace. In brief, bottles were fitted with a glass extension that was sealed with a screw cap and an air-tight septum (Grace, XLB-13 Septa 1/2). Ten μl aliquots of the bottle headspace gas were sampled through the septa at distinct time points of incubation (0, 2, 4, 6, 8, 10, 12, 24, 30, 36, 48, 56 and 72 h) using a gas-tight syringe (Gastight^® # 1701 Hamilton 1701N, 10 µl Syringe, Point style 5, Bonaduz, Switzerland). They were injected directly into the GC, which was equipped with a stainless steel column (6 m long, 0·53 mm internal diameter, 25 µm film thickness and packed with PoraPack Q50-80 mesh, Grace, Breda, the Netherlands) and connected to a flame ionization detector. The temperature of the injector, column and detector were maintained at 150, 60 and 150 °C, respectively. The carrier gas was nitrogen, and the pressure for nitrogen, hydrogen and air was set at 100, 50 and 100 kPa, respectively. The CH₄ concentration was calculated by external calibration using a certified gas mixture with a known CH₄ concentration in synthetic air (Linde Gas Benelux, Schiedam, and the Netherlands). Peak areas were determined by automatic integration system software (Chrom-Card data system Version 2.3·3, September, 2005, Rodano Milan, Italy) for GC.

Cumulative CH₄ production was calculated according to the procedure described by Pellikaan et al. (Reference Pellikaan, Stringano, Leenaars, Bongers, Van Laar-Van Schuppen, Plant and Mueller-Harvey2011b ) by taking the sum of the increase in headspace CH₄ concentration between two successive valve openings and the amount of CH₄ vented from the bottle:

(4) $$CH_4 = \; \mathop \sum \limits_{i = 1}^{i = 1 + n} \left\{ {V_{HS}\left( {C_{i + 1} - C_i} \right) + \; G_{i + 1}C_{i + 1}} \right\}$$

where CH ₄ = cumulative CH₄ production (ml/g of incubated OM); V _HS = the bottle headspace volume (ml); C _i , C _{i+ 1} = CH₄ concentration in the bottle headspace gas at valve opening i and i + 1, respectively; G _i+1 = the amount of gas vented at valve opening i + 1 (ml).

Curve fitting and calculations

Total cumulative gas (GP) and CH₄ production curves were fitted with biphasic and monophasic Michaelis-Menten equations, respectively (Groot et al. Reference Groot, Cone, Williams, Debersaques and Lantinga1996) using the non-linear least squares regression procedure in SAS (2002):

(5) $$OMCV = \mathop \sum \limits_{i = 1\; or\; 2}^n \; \displaystyle{{A_i} \over {1 + {(C_i/t)}^{B_i}}}$$

where OMCV = GP or CH₄ production (ml/g of incubated OM), A = the asymptotic gas production (ml/g of incubated OM), B = the switching characteristics of the curve, C = time at which half of the asymptotic gas production is reached (half-time, T_½, h) and t = the time (h). The value of i indicates the number of phases (i = 1 or 2). The maximum rate of gas production (R _max, ml/h) was calculated using the estimated A, B and C-values as described by Bauer et al. (Reference Bauer, Williams, Voigt, Mosenthin and Verstegen2001).

(6) $$R_{{\rm max}} = \displaystyle{{A \times C^B \times B \times \left[ {TR_{{\rm max}}^{( - B - 1)}} \right]} \over {{\left[ {1 + C^B \times TR_{{\rm max}}^{( - B)}} \right]}^2}}$$

where TR _max is the time at which R _max occurs;

(7) $$TR_{{\rm max}} = C \times \left\{ {{\left[ {\displaystyle{{B - 1} \over {B + 1}}} \right]}^{\left\{ {1/B} \right\}}} \right\}$$

The maximum rate of substrate degradation (R _M, %/h) was calculated from the A, B and C-values as estimated from the CH₄ production curves (Groot et al. Reference Groot, Cone, Williams, Debersaques and Lantinga1996).

(8) $$R_M = \left( {B \times tR_M^{(B - 1)}} \right)/\left( {C^B + tR_M^B} \right)$$

where tR _M is the time at which R _M occurs;

(9) $$tR_M = C \times (B - 1)^{1/B}$$

Chemical analysis

Grass silage was air dried, ground through a 1-mm sieve using a cross beater mill (Peppink 100 AN, Deventer, The Netherlands) and analysed for DM (ISO 6496, 1999), ash (ISO 5984, 2002) and nitrogen (N) (ISO 5983, 2005). Crude protein content was calculated as: CP = 6·25 × N. Neutral detergent fibre was analysed according to Van Soest et al. (Reference Van Soest, Robertson and Lewis1991) after a pre-treatment with a heat stable amylase and corrected for residual ash.

Fermentation fluid, sampled for VFA analysis (750 µl), was acidified with 750 µl of ortho-phosphoric acid solution. The ortho-phosphoric acid solution was composed of 25 ml of 85% (v/v) ortho-phosphoric acid dissolved in 200 ml Millipore water and 300 ml of a 4 g/l 2-methylvaleric acid solution. Volatile fatty acid concentration was analysed by GC following procedures of Pellikaan et al. (Reference Pellikaan, Stringano, Leenaars, Bongers, Van Laar-Van Schuppen, Plant and Mueller-Harvey2011b ) with the carrier gas modified by using hydrogen instead of helium to enhance baseline separation. Isocaproic acid was included as the internal standard. The total VFA (tVFA) concentration in the fermentation fluid was expressed as mm/g of incubated OM. Fermentation fluid samples for NH₃ analysis (750 µl) were mixed with 750 µl of 10% trichloroacetic acid solution. Ammonia was determined using a colorimetric method (Pellikaan et al. Reference Pellikaan, Stringano, Leenaars, Bongers, Van Laar-Van Schuppen, Plant and Mueller-Harvey2011b ) after deproteinizing the supernatant with 100 g/l trichloroacetic acid and the resulting chromophore was measured at 623 nm using an ultra-violet spectrophotometer (Evolution 201-Themo Scientific).

Statistical analysis

Effects of the CT structural properties on the substrate (grass silage) in combination with (+PEG) or without (−PEG) on fermentation kinetics and fermentation end-products were tested by analysis of variance using the MIXED procedure of SAS (2002) as:

(10) $$Y_{ijk} = \mu + T_i + P_j + R_k + (T \times P)_{ij} + \varepsilon _{ijk}$$

where Y _ijk  = the dependent variable, μ = the overall mean, T _i  = the tannin extract type (i = 1 to 9), P _j  = the effect of PEG (j = 1 to 2), R _k  = run (k = 1 to 2), (T × P) _ij = the effect of tannin extract type and PEG interaction and ε _ijk  = the residual error term. The statistical unit was the average of replicate in vitro bottles within run. Differences among treatment means were analysed using Tukey-Kramer's multiple comparison procedure in the LSMEANS statement in SAS with effects considered significant at a probability value of P < 0·05, and trends at a probability value of P < 0·10.

Relationships between CT content, mDP, proportion of PD and proportion of cis and fermentation parameter estimates were analysed using the multiple stepwise regression procedure in SAS (2002) where CT content (g CT/100 g extract), mDP, proportion of PD and proportion of cis were included as independent variables. The criteria to include variables in the model were a combination of a low value for the Mallow's Cp-criterion, a high coefficient of determination (R ²), and setting the residual degrees of freedom (d.f.) in the regression model at >65% of the total d.f.