Effects of treating high forage and high concentrate diets with exogenous fi brolytic enzymes on their in vitr o ruminal fermentation

Batch cultures of mixed rumen microorganisms were used to study the effects of a 24 h pretreatment with a xylanase from Trichoderma viride (XYL) and a cellulase from Trichoderma longibrachiatum (CEL) on the in vitro fermentation of two diets composed by grass hay and concentrate in the proportions of 70:30 (HF) and 30:70 (HC). In 8 h incubations with HF, CEL increased (P<0.05) the production of acetate, propionate and total volatile fatty acids (VFA) and gas, but no effects (P>0.05) were observed for XYL. In contrast, with HC both enzymes increased (P<0.05) the production of propionate and total VFA and the true DM degradability of substrate, and decreased the CH4/VFA ratio (P<0.05). After 24 h incubation, some of the observed effects disappeared, but CEL still increased (P<0.05) the production of acetate, propionate and total VFA for HF. For HC, both enzymes increased (P<0.05) the production of total VFA and XYL also increased (P<0.05) acetate production. The results indicate that effects of enzymes on in vitro fermentation were infl uenced by the nature of the diet, and that this infl uence varied with the incubation time, being more pronounced at short incubation times.


INTRODUCTION
The use of fi brolytic enzymes as feed additives in ruminant diets has received considerable attention in the last years, despite observed responses have been highly variable.The effects of enzymes are infl uenced by factors such as type and dose of enzyme, type of diets fed to the animals, and enzyme application method (Beauchemin et al., 2003).To our knowledge, however, no study has been conducted to investigate how the forage:concentrate ratio in the diet can affect the response to enzyme treatment.The objective of this study was therefore to evaluate the effects of two fi brolytic enzymes on the in vitro ruminal fermentation of two diets differing in their forage:concentrate ratio.The study examined the effects of the enzymes at two incubation times, since it has been shown that the response to enzymes usually becomes less marked as incubation time increases (Beauchemin et al., 2003).

MATERIAL AND METHODS
The diets were composed of grass hay and a commercial concentrate based on barley, maize, soyabean meal and a vitamin-mineral mixture in the proportions (DM basis) of 70:30 (high-forage diet; HF) and 30:70 (high-concentrate diet; HC).Crude protein (CP), neutral-detergent fi bre (NDF) and acid-detergent fi bre (ADF) contents of diets were 143, 496 and 253 g/kg DM for HF and 176, 376 and 161 g/kg DM for HC.Two enzyme preparations commercialized by Fluka Chemie GmbH (Germany) were tested: xylanase from Trichoderma viride (XYL) and cellulase from Trichoderma longibrachiatum (CEL).Enzyme preparations were assayed for endoglucanase, exoglucanase, xylanase and amylase activities following the procedures described by Colombatto and Beauchemin (2003).All activities were measured at pH 6.5 and 39ºC in order to resemble optimal ruminal conditions.At pH 6.5 and 39ºC, 1 mg of XYL liberated per min 15.0 µmol of xylose from oat spelt xylan and 0.237 µmol of glucose from carboxymethylcellulose, but no amylase and exoglucanase activities were detected.One mg of CEL liberated 1.72 µmol of xylose from oat spelt xylan, and 2.40, 0.385 and 0.040 µmol of glucose from carboxy-methylcellulose, soluble starch and Avicel PH-101, respectively.One enzymatic unit was defi ned as the amount of enzyme required to release 1 µmol of xylose or glucose per min from the corresponding substrate at 39ºC and pH 6.5, and both enzyme preparations were added at a rate of 40 enzymatic units/g substrate DM.
Samples (500 mg) of ground diet (1-mm screen) were accurately weighed into 120-ml serum bottles.Solutions of each enzyme preparation were prepared in 0.1 M sodium phosphate buffer (pH 6.5), 0.5 ml of the corresponding solution were added directly to each bottle 24 h before starting the incubation, and bottles were kept at room temperature (21-23ºC) until incubation.The inoculum was obtained from four rumen-cannulated Merino sheep fed 500 g of grass hay and 500 g of concentrate per day administered in two equal portions at 09.00 and 18.00 h.Ruminal contents of each sheep were obtained immediately before the morning feeding, mixed and strained through four layers of cheesecloth.Particle-free fl uid was mixed with the buffer solution of Goering and Van Soest (1970; no trypticase added) in a proportion 1:4 (vol/vol) at 39ºC under continuous fl ushing with CO 2 .Fifty ml of buffered ruminal fl uid were added into each bottle under CO 2 fl ushing, bottles were sealed with rubber stoppers and aluminium caps, and incubated at 39ºC for 8 and 24 h.Four incubation runs were performed on different days, and in each of them one bottle per treatment was included.
After 8 and 24 h of incubation, total gas production was measured in all bottles using a pressure transducer and a calibrated syringe, and a gas sample (about 15 ml) from each bottle was stored in a vacuum tube before analysis for CH 4 concentration.Bottles were then uncapped, the pH was immediately measured and the fermentation was stopped by swirling the bottles in ice.Two ml of the bottle content were taken for VFA and NH 3 -N analyses.Finally, the content of the bottles was transferred to previously weighed fi lter crucibles, the solid residue was washed with 50 ml of hot distilled water (50ºC) and the crucibles were dried at 50ºC for 48 h.Residues were analysed for NDF to estimate true DM degradability (TDMD) and NDF degradability (NDFD).Procedures for analysis of CP, NDF, ADF, VFA, CH 4 and NH 3 -N have been described by Carro et al. (1999).
Data from each incubation time were analysed independently.When data from both enzymatic treatments and diets were analysed together, signifi cant enzyme x diet interactions (P<0.05) were observed for some variables.Therefore, data were analysed separately for each diet as an ANOVA with three enzyme treatments (control (CON), XYL and CEL) and four incubation runs (blocking factor) as main effects.When a signifi cant effect of the enzyme treatment (P<0.05) were detected, differences between means were assessed by the LSD test.

RESULTS AND DISCUSSION
The effects of the treatment with fi brolytic enzymes on in vitro fermentation of diets after 8 and 24 h of incubation with rumen mixed microorganisms are shown in Tables 1 and 2, respectively.For HF, there were no effects (P>0.05) of enzymes on fi nal pH, CH 4 production, NH 3 -N concentration and TDMD at any incubation time.After 8 h of incubation and with HF, CEL treatment increased (P<0.05) the production of gas, acetate, propionate and total VFA by 11, 10, 23 and 13%, respectively, but no effects of XYL (P>0.05) were detected.In contrast, for HC both enzymes decreased (P<0.05)fi nal pH and CH 4 :VFA ratio, and increased (P<0.05)TDMD and the production of propionate and total VFA.These results indicate that CEL was more effective than XYL with HF, but only subtle differences were observed between the effects of both enzymes on in vitro fermentation of HC.The treatment with XYL implied mainly xylanase activity, as endoglucanase and amylase activities were negligible.Compared to XYL, CEL presented lower xylanase activity, but greater endoglucanase, exoglucanase and amylase activities.Eun and Beauchemin (2007) suggested that limiting enzymatic activity is likely substrate dependent.The lack of effect of XYL on fermentation of HF diet might be partly explained if endoglucanase activity would have been the limiting activity for this diet.However, for HC diet XYL and CEL produced similar effects on some ruminal variables at 8 h.These results indicate again that the enzymatic activities required to enhance ruminal degradation are infl uenced by the nature of the incubated substrates.As pointed out by Wallace et al. (2001), a precisely identifi cation of the enzymatic activity causing a positive response in ruminal fermentation might make possible to develop more effective fi brolytic enzyme products.Trichoderma longibrachiatum.Both enzymes were applied at a level of 40 enzymatic units/g of substrate DM 2 standard error of the difference a,b,c for each diet, mean values within a row with unlike superscript letters differ (P<0.05)Several studies (Wang et al., 2001;Giraldo et al., 2007) have reported that treating different feeds with fi brolytic enzymes produced a shift in the molar proportions of VFA, but shifts in pattern of VFA seem to be infl uenced by the type of diet and enzyme preparations.In the present study, CEL treatment decreased signifi cantly the acetate:propionate ratio at 8 h of incubation for HF, indicating a change in fermentation pattern.Wang et al. (2001) suggested that changes in fermentation pattern may refl ect a shift in the species profi le of colonizing bacteria in response to pre-treatment of feed with exogenous enzymes.Trichoderma longibrachiatum.Both enzymes were applied at a level of 40 enzymatic units/g of substrate DM 2 standard error of the diff erence a,b,c for each diet, mean values within a row with unlike superscript letters diff er (P<0.05)Some of the observed effects of enzymes disappeared after 24 h of incubation, but CEL still increased (P<0.05) the production of acetate, propionate and total VFA by 10, 11 and 9.5%, respectively, for HF; in contrast, no effects (P>0.05) of XYL were detected for these diet.Regarding HC, both CEL and XYL increased total VFA by 2.8 and 3.1%, respectively.For both diets, the effects of enzymes were more marked at 8 than at 24 h of fermentation, thus indicating that enzymes produced their effects at early stages of fermentation.Previous studies (Beauchemin et al., 2003;Giraldo et al., 2007) have also shown that effects of fi brolytic enzyme on in vitro fermentation were generally larger during the initial stages of degradation.
In general, effects of CEL on in vitro fermentation of diets were more marked than those of XYL.Although CEL was effective with both substrates, lower effects were observed for HC compared to HF.The results of this study indicate that effects of enzymes on in vitro fermentation were infl uenced by the nature

Table 1 .
Infl uence of two enzymatic treatments on in vitro fermentation of two diets (500 mg DM) in batch cultures of mixed rumen microorganisms for 8 h (n=4) CON -control, XYL -xylanase from Trichoderma viride, CEL -cellulase from

Table 2 .
Infl uence of two enzymatic treatments on in vitro fermentation of two diets (500 mg DM) in batch cultures of mixed rumen microorganisms for 24 h (n=4)