Determination of in Vitro Digestibility of Tropical Feeds Using Cattle Faeces as Rumen Fluid Alternative

The research was conducted to compare the ability of faeces dissolved in distilled water (P1), saline solution (P2), artificial saliva (P3), and rumen fluid (RF) as sources of inoculant in in vitro organic matter digestibilities (IVOMD) of rice straw, corn stover, napier grass, and pangola grass. The rumen fluid was collected from two rumen fistulated Ongole Crossbred Cows of 306 and 333 kg body weight (BW). The cows were fed 3% of BW consisted of 70% napier grass and 30% concentrate. At the end of 30 days feeding, faecal solution was made out of 350 g fresh faeces dissolved in 1 l of each solvent, homogenized using blender for 30 second, while rumen liquor were collected directly from fistula. After straining with four layers of cheesecloth both faecal solution and rumen liquid were mixed with artificial saliva (1:4 v/v). Fifty ml of each inoculants was pipetted into each incubator tube (100 ml) containing 500 mg sample. The tubes were then incubated at 39 oC for 48 h. Value of IVOMD of napier grass, rice straw, corn stover, and pangola grass did not differ among the faecal solvents, but significantly lower (Pin vitro organic matter to predict digestibility of fibrous feed determination. However, the faecal solvent as inoculant produced lower in vitro digestibility than that of rumen fluid.


INTRODUCTION
Stability of ruminant productivity in Indonesia is determined by forage availability throughout the year (Abdullah & Suharlina, 2010). Information of digestibility values of each feedstuff is essential as asset feed information. The value could be obtained by in vivo, in sacco, or in vitro procedures. Even though the in vivo procedure produce more accurate information, but it is ineffi cient, labor and time consuming, high costs, and labor intensiveness (Adesogan, 2004). Many scholars (Williams, 2000;Mahadevamma et al., 2004;Damiran et al., 2007;Palić & Leeuw, 2009) have developed faster, less cost, and need small quantity of sample to get more numbers of feed used in vitro method. The method needs shorter time to predict digestibility (Utomo, 2010).
One of the most widely used in vitro digestibility method was developed by Tilley & Terry (Harris, 1970). The method depended on the availability of rumen fl uid as inoculant source for fermentative process of digestibility. For some areas, collecting rumen fl uid was diffi cult work because the need for rumen fi stulated animal, which is not only high operational cost, easy to contaminate by diseases especially in tropical warm climate, hurting animal to made fi stula or using stomach tube for rumen liquor collection, but also against animal welfare (Mauricio et al., 2001;Dhanoa et al., 2004).
The results of some experiments in sub tropic region showed that rumen fl uid inoculants (RF) for determining in vitro digestibility could be replaced with faecal solution (FS) from some species of animal (Omed et al., 2000;1988;Akhter et al., 1999). It has been reported that microbial species found in the faeces as inoculant were infl uenced by the kind of feed consumed by animal donor (Bauer et al., 2004). Since, tropical feed are largely diff erent with subtropics ones, the microbial species in ruminant are assumed to be diff erent. Therefore the in vitro procedure using faecal solution as inoculant which has been developed in the subtropics might not be used to predict Indonesia feedstuff value correctly. Artifi cial saliva (buff er) is commonly used as solvent for faeces by previous investigators; the solvent is expensive and less available in some areas.
The aim of this research was to compare the eff ectiveness of faeces and rumen fl uid as inoculant source for in vitro mination and to fi nd more aff ordable faecal solvent as an alternative to artifi cial saliva.

MATERIALS AND METHODS
The research was done in the Laboratory of Feed Technology, Department of Animal Nutrition and Feed Science, Faculty of Animal Science, Gadjah Mada University, Yogyakarta. Two fi stulated rumen Ongole crossbred cows, about 5 years old, 333 and 306 kg body weight were kept in concrete fl oor individual stalls. In order to have an optimal and stable rumen ecological condition, the cows were fed 3% of body weight, consisted of 70% napier grass and 30% concentrate for 30 days including adaptation period. The concentrate was fed at 07.00 a.m. and 16.00 pm, while napier grass was given after the concentrate was consumed. The cows were off ered water ad libitum. The chemical compositions of feedstuff used in the experiment were shown in Table 1, and the chemical compositions of the diet were shown in Table 2.
The rumen fl uid and faeces were taken from both cows after three weeks of the ration feeding. Rumen fl uid was taken using aspirator via rumen fi stula and been warmed by keeping hot water prior to collection to keep the temperature around 39 o C. While faeces was taken directly from the caecum (grab sampling), and kept at 39 o C in a thermostatic container.
Rumen fl uid from both donor cows were mixed, strained with 4 layers of cheesecloth and then mixed with artifi cial saliva (McDougall's Artifi cial Saliva), in an Erlenmeyer fl ask at 1 : 4 (v/v) ratio (Harris, 1970). Before pouring into the incubation tubes which were already fi lled with feed sample, the inoculant were pre warmed in incubator at 39 o C and fl ushed with CO 2 . The inoculants of fresh faeces from both donor cows were mixed homogeneously, and dissolved in three diff erent solvent (distilled water, saline, and artifi cial saliva) at concentration 35% (w/v) (Akhter et al., 1999;Sudirman, 2007). The mixture was homogenized using a blender for 30 second. The mixtures were strained with 4 layers of cheesecloth and then processed like rumen fl uid.
Four kind of fi brous feed namely rice straw (Oryza sativa), corn stover (Zea mays), napier grass (Pennisetum purpureum), and pangola grass (Digitaria decumbens) as control sample were used to examine the eff ectivity of faecal in comparison to rumen fl uid for determining in vitro digestibility. All feeds were ground using a Wiley mill with 2 mm screen in order to have same particle size. Feeds samples were stored in scaled container in the laboratory at room temperature.
Fifty ml of each inoculant were poured into 100 ml incubation tubes which contained 500 mg of feed sample. Each tube was incubated for 48 h (one step method) anaerobically at 39 o C waterbath, every 8 h the tubes were shaken by hand. After 48 h incubation, the mixture tubes were strained using predeterminated weight porous crucible (30 ml) with glass wool on it. The crucibles were dried at 105 o C until its weight constant to at 550 o C oven temperature for 5 h. In vitro digestibility was calculated using formula of Tilley and Terry method (Harris, 1970).
The experiment was conducted in completely randomized design. Four kinds of inoculants types were used as treatments with 5 replications. The means were tested by Duncan multiple range test. While the ability of faecal solvent to replace rumen fl uid was examined using simple linear regression analyses (Gomez & Gomez, 2007). The linear regression equation was Y= a + bX, the kind of faecal solution were X and rumen fl uid inoculant was Y.

RESULTS AND DISCUSSION
The IVOMD values of all feeds digested with faeces dissolved in diff erent solvent (P1-P3) were not   signifi cantly diff erent (Table 3). Therefore, since distilled water was more aff ordable, it could be used to replace the more expensive solvent such as saline and artifi cial saliva. The IVOMD obtained using faecal inoculant were signifi cantly (P<0.05) lower compared to that of rumen fl uid inoculant, due to the lower total number of microbes in faecal solution than in the rumen fl uid. Table 3 also showed that fi brous feeds (rice straw) which contained the highest neutral detergent fi ber (Table 1) resulted in the lowest value of in vitro digestibility, followed by napier grass, corn stover, and pangola grass.
Based on simple linear regression analysis, the regression coeffi cient (R 2 ), there was close correlation between inoculant of faeces and that of rumen fl uid. The equation of feed in vitro between inoculant of RF (Y) with faeces (X) using solvent P1 (distilled water) for rice straw was: Y (%)= -10.47 + 2.02 X (R 2 = 0,99), corn stover was: Y= 6,07 + 1,11X (R 2 = 0,60), napier grass was Y= -1,72 + 1,64X (R 2 = 0,87), and pangola grass was: Y= 0,74X + 34,05 (R 2 = 0.81), respectively (Figure 1). The equation using solvent P2 (saline) and P3 (artifi cial saliva) were shown in Figure 2 and Figure 3, respectively. The equation and regression coeffi cient value (R 2 ) as shown in  Figure 1, also gave indication that faecal solution could be used as inoculant for replacing rumen fl uid on determining in vitro by Dhanoa et al. (2004). There was a close correlation between inoculant of faecal solution (P1-P3) with rumen fl uid (RF). Therefore, based on the results of this research, the use of distilled water as solvent for faeces could be recommended to replace rumen fl uid for predicting in vitro digestibility of feeds. Based on the graph, the correlation of feed in vitro solvent with rumen fl uid, there were some interesting information, such as tendency of similarity of regression coeffi cient value (R 2 ) between napier and pangola grass, also between rice straw and corn stover. It was assumed to be correlated with the characteristic of the feeds. Napier grass had higher soluble nutrients than rice straw and corn stover. In contrary, both straw (rice straw and corn stover) had higher cell walls or neutral detergent fi ber/NDF (Table 1).

CONCLUSION
Distilled water could replace artifi cial saliva as solvent for faeces and the dilution could be used as inoculant to replace rumen fl uid to predict digestibility of tropical fi brous feeds (in vitro). However the faecal solvent as inoculants produce lower in vitro digestibility than that of rumen fl uid.