MATERIALS AND METHODS

Preparation of TMRs
The Burma padauk(Plerocarpus indicus), rain tree (Samanea saman (Jacg. Merr.) and Siamese rough bush (Streblus asper) leaves and corn cop (control) were used in this study. Theywere collected from the Mahasarakham province area in the North-East of Thailand. Fresh samples (1 kg) were hand harvested from three site specimens. Duplicate fresh samples (0.5 kg/replicate) were dried in a hot, dry air force oven at 65°C for 72 h and weighed. All feed samples (Table 1) were ground to pass through a 1 mm screen for chemical analysis. The feedstuff samples were analyzed for Dry Matter (DM), Crude Protein (CP) and ash, neutral Detergent Fiber (NDF), Acid Detergent Fiber (ADF) and Acid Detergent Lignin (ADL) (Van Soest et al., 1991).

Four TMRs were formulated, to have similar total digestible nutrient (TDN), CP, NDF, ADF, but differ in fiber source (Table 2). The experiment was done at division of animal science, Faculty of veterinary medicine and animal science, Mahasara Kham University, Thailand, from August 25, 2007 to June 5, 2008. The experiment was designed in CRD with five replicates per treatment. The fiber sources of the total mixed ration were corn cop (control), Burma padaukleaf, rain tree leaf and Siamese rough bush leaf. Four TMRs for the gas production test were ground to pass through a 1 mm screen in a hammer mill.

In vitro Gas Production Test
Strict anaerobic techniques were used in all steps during the rumen fluid transfer and incubation period. Rumen fluid inoculums was removed before the morning feeding under vacuum pressure via the rumen fistula into a 2 L glass flask and transferred into two pre-warmed 1 L thermos flasks which were then transported to the laboratory. The medium preparation was as described by Makkar et al. (1995). Mixed rumen fluid inoculums were obtained from two fistulated Brahman-Thai native crossbred steers (weighing about 250±15 kg). The animals were offered rice straw on ad libitum and fed 0.5% body weight of concentrate (concentrate mixture: 49.80% cassava chip, 17.5% rice bran, 14.60% palm meal, 7.0% soybean meal, 1.40% urea, 0.4% salt, 1.0 % mineral mix and 8.30% sugarcane molasses). The animals were fed twice daily; water and a mineral lick were available ad libitum for 14 days.

The feed sample of approximately 500 mg on a fresh weight basis was transferred into a 50 mL serum bottle (Sommart et al., 2000). The bottles were pre-warmed in a hot air oven at 39°C for about 1 h prior to injection of 40 mL of rumen fluid medium (using a 60 mL syringe) to each bottle. The bottles were stoppered with rubber stoppers, crimp sealed and incubated in a hot air oven set at 39°C.

The rate of gas production was measured by reading and recording the amount of gas volume after incubation using a 20 mL glass syringe connected to the incubation bottle with a 23 gauge, 1.5 inch needle. Readings of gas production were recorded from 1 to 72 h (hourly from 1-12 h, every 3 h from 13-24 h, every 6 h from 25-48 h and every 12 h from 49-72 h) after incubation periods. Amounts of cumulative gas volume at 2, 4, 6, 12, 24, 48 and 72 after incubations were fitted using the equation (Orskov and McDonald, 1979):

In vitro digestibility of dry matter and organic matter was measured at 72 h after incubation. The residues of the TMRs were removed by filtering through a glass filtering crucible, residue was washed with 250 mL boiled distilled water and the amount of DM and OM in the residue was estimated. Calculation of in vitro DM and OM digestibility as a percent of total DM and OM followed the equation:

In vitro Fermentation Measurement
The bottles were sampling at 0, 3, 6, 9 and 12 h after incubation. Rumen fluid medium pH was measured immediately after sampling using a portable pH meter. The rumen fluid medium was acidified with 5 mL 6 N HCl and centrifuged at 3000 rpm for 15 min and the clear supernatant was stored in plastic tubes at -20°C until analyzed for ammonia nitrogen (Bremner and Keeney, 1965) and total volatile fatty acid concentration (Briggs et al., 1957).

Statistical Analysis
All data obtained from the trials were subjected to the analysis of variance procedure of statistical analysis system according to a completely randomized design. Means were separated by Duncan New`s Multiple Range Test. The level of significance was determined at p<0.05.

RESULTS AND DISCUSSION

Chemical Composition of Feedstuffs and TMRs
The feed ingredients varied widely in terms of composition. The leaves have high NDF content, more than 40.5% (Table 1). All TMRs had a similar chemical composition. The ration CP, ash and NDF content were approximately 11.8, 7.8 and 41.7 %, respectively (Table 2).

Gas Production Characteristics of TMRs
A comparison of the gas production characteristics of different treatments indicated significant differences between treatment (p<0.05). The a intercept value for all TRMs ranged from -4.2 to -8.7 mL. The values for a, intercept, were negative in the incubations of all TMRs in this study (Table 3). These data suggested that a lag phase due to a delay in microbial colonization of the substrate may occur in the early state of incubation. Blummel and Becker (1997) have also reported negative values with various substrates when using mathematical models to fit gas production kinetics.

Table 1:	Chemical analysis of feedstuffs used for feed formulation in the experiments (% of DM)
DM = Dry Matter, OM = Organic Matter, CP = Crude Protein, NDF = Neutral Detergent Fiber, ADF = Acid Detergent Fiber, ADL = Acid Detergent Lignin

Table 2:	Feed formulation and chemical composition of dietary treatments
*Calculated value, 1C-TMR = Corn cob as fiber source, B-TMR = Burma padauk leaf as fiber source, R-TMR = Rain tree leaf as fiber source, S-TMR = Siamese rough bush leaf as fiber source, 2DM = Dry matter, OM = Organic matter, CP = Crude protein, NDF = Neutral detergent fiber, ADF = Acid detergent fiber, ADL = Acid detergent lignin

Table 3:	Gas production characteristics and in vitro digestibility of dietary treatments using in vitro gas production technique
a, b, c,d Means within a row different superscripts differ (p<0.05), 1C-TMR = Corn cob as fiber source, B-TMR = Burma Padauk leaf as fiber source, R-TMR = Rain Tree leaf as fiber source, S-TMR = Siamese Rough Bush leaf as fiber source, 2a = the intercept (mL), which ideally reflects the fermentation of the soluble fraction, b = The fermentation of the insoluble fraction (asymptote) (mL), c = Rate of gas production (%/h), |a|+b = Potential extent of gas production (mL)

This is due to either a deviation from the exponential cause of fermentation or delays in the onset of fermentation due to the microbial colonization. It is well known that the value for absolute |a|, described ideally, reflects the fermentation of the soluble fraction. In this study the |a| was highest for C-TMR and significance difference (p<0.05) with B-TMR and R-TMR. It is indicated that the soluble fraction in C-TMR was also highest. The soluble fraction makes it easily attachable by ruminal microorganisms and leads to much gas production (Table 3). The soluble fraction of corn cop was higher than leaves. It`s possibly structure and solubility characteristics of carbohydrate and protein in corn cop which easily attach with microorganisms in the rumen. While fodder trees had high fibrous content it is difficult to attach by microorganisms. Generally, leaves have a large proportion of lignified cell walls (Table 1) with low fermentation rates and digestibility. Therefore, Neutral Detergent Fiber (NDF), Acid Detergent Fiber (ADF) and Acid Detergent Lignin (ADL) content in fodder tree are the crucial parameters affecting fodder utilization.

The gas volume at asymptote (b) described the fermentation of the insoluble fraction. The gas volume at asymptote was significantly higher in C-TMR than that B-TMR, R-TMR and S-TMR (p<0.05). The gas volumes at asymptote have the advantage of predicting feed intake. Blummel and Ørskov (1993) found that the gas volume at asymptote could account for 88% of variance in intake. Sommart et al. (2000) suggested that gas volume is a good parameter from which to predict digestibility, fermentation end-product and microbial protein synthesis of the substrate by rumen microbes in the in vitro system. Additionally, in vitro dry matter and organic matter digestibility were shown to have high correlation with gas volume (Sommart et al., 2000). In this study, C-TMR showed the highest gas volume (Table 3). Low gas volumes at asymptote were observed in B-TMR and R-TMR. It`s possibly reflected by phenolic compounds, which are toxic to rumen microbes (D`Mello, 1992). Getachew et al. (2000) reported that the concentration of phenolic compounds (particularly tannin) in tree leaves is generally high. Lowry et al. (1996) have also reported that many of the multi propose tree are problematic as feed supplements because they often contain anti-nutritional compounds, such as tannin, saponin and non-protein amino acid, which are either toxic to rumen microbe or to the animal, or their metabolic products are toxic.

Rate of gas production (c) expressed in %/h as ranked from the fastest to the slowest were; B-TMR, S-TMR, C-TMR and R-TMR. Fast rates of gas production were observed in B-TMR and S-TMR. This result might have been influenced by the carbohydrate fraction that was readily available to the microbial population.

The potential extent of gas production (|a|+b) of C-TMR was the highest and significantly different (p<0.05) with B-TMR, R-TMR and S-TMR. This implies that C-TMR was highly fermentable in the rumen. Remarkably, the potential of gas production for B-TMR and R-TMR was slightly lower compared with C-TMR and S-TMR, possibly due to the influence of carbohydrate fraction in the TMR. Fibrous constituents had negatively influenced in vitro gas production (Melaku et al., 2003). The fiber content of the B-TMR and R-TMR were high, when compared with C-TMR and S-TMR. The high fibrous content is difficult to attach by microorganisms. Therefore, less gas production was shown in the B-TMR and R-TMR.

In vitro Digestibility of Dry Matter and Organic Matter
It can be seen that IVDMD and IVOMD are similar. The IVDMD and IVODM were significantly different (p<0.05) among treatment. The S-TMR gave the highest IVDMD and IVOMD. This result implies that the microbe in the rumen and animal have high nutrient uptake (Table 3). The IVDMD and IVOMD of C-TMR, B-TMR and R-TMR lower than S-TMR. The reason for that is possibly that the fiber fractions of C-TMR, B-TMR and R-TMR have a large proportion of lignified cell walls leading to attachment difficulty by microorganism, with low fermentation rates, low digestibility rate and limited intake (Ibrahim et al., 1995). The higher fiber content (Table 1) of Burma Padauk leaf and rain tree leaf probably resulted in lower in vitro dry matter and organic matter digestibility since high NDF and ADL content in feedstuffs result in lower fiber degradation (Van Soest, 1988).

In vitro Fermentation Pattern
Concentrations of NH₃-N, TVFA and pH in the in vitro fluid were used to monitor the in vitro fermentation pattern (Table 4). The pH was not affected by fiber source in TMRs. When monitoring pH pattern at 0, 3, 6, 9 and 12 h after incubation, the pH values were relatively stable at 7.0-7.3 and all treatment means were within the normal range that has been reported as optimal pH (6.0-7.0) for microbial digestion. The buffer in the rumen fluid medium is the reason for pH remaining stable at all times of fermentation. The buffer is a factor that should be considered when using the gas production technique. The exhaustion of the buffer would lead to a lowering of the pH (Getachew et al., 1998).

Table 4:	Effect of Burma Padauk (Plerocarpus Indicus), Rain Tree (Samanea Saman (Jacg.) Merr.) and Siamese Rough Bush (Streblus Asper) Leaves as Fiber Sources in Total Mixed Ration on pH, ammonia nitrogen (NH3N) and total volatile fatty acid (TVFA) in in vitro
a, b, c, dMeans within a row different superscripts differ (p<0.05), 1C-TMR = Corn cob as fiber source, B-TMR = Burma Padauk leaf as fiber source, R-TMR = Rain tree leaf as fiber source, S-TMR = Siamese rough bush leaf as fiber source

At a lower pH, the celluolytic bacteria becomes less active (Russell and Dombrowski, 1980). In this study the buffer was not exhausted. Therefore, this condition was optimal for microbial activity.

Ammonia nitrogen concentration was significantly different (p<0.05) among treatments at each hour of sampling, excepted at 0 and 12 h of sampling. The difference in NH₃-N concentrations among treatments may have been related directly to urea and degradability of protein in the TMRs. Although, nitrogen recycling in the rumen and in vitro is different, NH₃-N concentration was in the optimal range for rumen ecology, microbial activity (Perdok and Leng, 1990; Wanapat and Pimpa, 1999). At 3 to 9 h after incubation C-TMR had the highest NH₃-N, when compared with other TMRs. When ammonium nitrogen is high it indicates that the soluble fraction of protein is also high. Remarkably, NH₃-N concentration of B-TMR and R-TMR were low at all time of sampling. It may have been that the urea level in both TMRs was lower than C-TMR. In addition, the protein in Burma Padauk leaf and rain tree leaf had low degradability (Chumpawadee et al., 2007). Although, NH₃-N concentration of all TMRs was different with C-TMR (control), it was in the normal range. Therefore, it can be used as fiber sources in the TMRs. Future research should investigate the impact of the ability of leaves feed to replace forage in intact animals.

Total volatile fatty acid concentrations were significantly different (p<0.05) among treatments at all times of sampling, excepting 0 h after incubation. Remarkably, TVFA concentrations in the in vitro medium, from 0 to 12 h after incubation, tend to be increased. The reason for that is possibly VFAs accumulated in the medium. The VFA can not absorb via the in vitro, but most of the VFA can be absorbed into rumen wall. Although, VFA increased in the medium, pH did not change because the buffer in the medium was not exhausted. This is the advantages of the gas production technique. The VFA production of B-TMR are the same with S-TMR and difference from C-TMR and R-TMR, this result might have been influenced by carbohydrate fraction in TMRs. The rate and extent of carbohydrates degradation are influenced by the condition of rumen fermentation and rate and extent of VFAs production (Cheng et al., 1991). Keady and Mayne (2001) also suggest that VFAs concentration is similar when the animal fed diets contained a similar carbohydrate composition. In this study, have difference source of fiber in TMRs, thus VFA concentration was also different.

CONCLUSIONS

In in vitro study, leaves as fiber source in TMRs had an affect on gas production characteristic, in vitro digestibility and in vitro fermentation. The corn cop as a fiber source in TMR gave the highest parameters of gas production characteristic. The Siamese rough bush as fiber source in TMR gave the highest in vitro digestibility. Concentration of NH₃-N, TVFA and pH were different when TMRs contained a different fiber sources, but they are in the normal range. Therefore, leaves can be used as fiber source in TMRs. Future research should investigate the impact of a leaves feed replacement of forage for intact animal.

ACKNOWLEDGMENTS

The authors would like to express their gratitude to all staff and my student for their invaluable help on the farm and laboratory. We are grateful to the Division of Animal Science, Faculty of Veterinary Medicine and Animal Science, Mahasarakham University for supporting experiment facilities. This research project No. MRG5080362 was funded by Thailand Research Fund and Commission on Higher Education under the program New Researcher Grant.