Study area

This study was carried out at the Animal Biotechnology Laboratory, Faculty of Agricultural Sciences, Campus IV of the Universidad Autonoma de Chiapas, in the junction of coastal highway and Huehuetán, Chiapas (14° 54’ 29” N, 92° 15’ 38” O, 50 m altitude, rainfall of 2326 mm per year, average annual temperature of 28 °C and humid warm climate with summer rains).

Culture medium

One hundred milliliters of anaerobic culture medium (ACM) containing: 47.9 mL of distilled water, 30 mL of clarified rumen fluid (filtrated in a triple gauze and centrifuged at 20 817 g for 15 min at 4 °C and sterilized at 121 °C for 15 min at 15 psi in an autoclave; Felisa, FE-397, Mexico), 5 mL of mineral solution I (6 g K₂HPO₄ 1000 ML^-1), 5 mL of mineral solution II (6 g K₂HPO₄, 6 g (NH₄) ₂SO₄, 12 g NaCl, 2.45 g MgSO₄ and 1.6 g CaCl₂ H₂O in a 1000 mL of final solution), 5 mL of 8 % Na₂CO3 solution (8 g of Na₂CO₃ in 100 mL of distilled water), 5 mL of sodium acetate 1.5 % (1.5 g sodium acetate in 100 mL of distilled water), 2 mL of sulfide-cysteine solution (2.5 g L-cysteine dissolved in 15 mL of 2N NaOH+2.5 g of Na₂S-9H₂O dissolved in 100 mL H₂O), 0.1 mL 0.1 % resazurin (0.1 mL of resazurin in 100 mL final volume of heated until the indicator loses its color), 0.2 g of peptone trypticase and 0.10 g yeast extract (^{Ley de Coss et al., 2011}, ²⁰¹³).

BRUpj Isolation

The adaptation capacity and viability of the bacterial consortium from the rumen bacteria were evaluated in a culture medium with J. curcas pulp (first isolation). In 250 mL culture vials, containing 150 mL of ACM and 3 g of degreased J. curcas pulp without detoxification, 50 mL of FRF extracted from male bovines through a tube in the esophagus were added. The whole procedure was performed under CO₂ flow. The medium was kept 10 d at 38±0.05 °C. Then, in 10 tubes (13×100 mm PYREX^®, Mexico) with 4.5 mL of ACM, 0.1 g of the pulp were added. These were inoculated with 0.5 mL of the bacteria that survived the first isolation and were kept for 7 days at 38±0.05 °C. Produced gas was released every 24 h; afterwards 0.5 mL of this culture were transferred to 13×100 mm tubes (PYREX^®, Mexico) containing 4.5 mL of new sterile ACM with the same characteristics and then incubated for 24 h. The culture was transferred to vials containing 50 mL of culture medium for rumen bacteria utilizing pulp of J. curcas (BRUpj) and kept at 38±0.05 °C. This culture was used as the inoculum of potential BRUpj (T2).

Dry matter in vitro degradation (DMivD)

Test tubes (18×150 mm; PYREX^®, Mexico) containing 0.2 g of J. curcas pulp were sterilized for 15 min at 121 °C, 9 mL of culture media were added afterwards. The procedure took place under a CO₂ flow, following the technique reported by ^{Cobos and Yokoyama (1995)}. The test tubes were incubated at 38±0.05 °C for 3, 6, 12, 24, 48, 72 and 96 h after inoculation with 1.0 mL of inoculum (T1: FRF obtained from a bivine with a ruminal cannula; T2: RBUpj obtained from the isolation process). During each incubation period the non-degraded substrate concentration was measured. After each period the pH was measured with a potentiometer (OACTON, 43294, Singapore). In each treatment the mean values were obtained and adjusted to logarithmic values in order to obtain the average for the test with set values (^{Häubi, 2004}). The results were calculated with Microsoft Excel software.

Total bacterial count (TB)

With a Pasteur pipette, 0.5 mL of culture medium were removed from each treatment after the corresponding incubation period (3, 6, 12, 24, 48, 72 and 96 h). They were by capillarity placed in a Petroff-Hausser counting chamber (Hausser #3900, Electron Microscopy Sciences, USA) with a 0.0025 mm² area and 0.02 mm depth. For the counting, a phase contrast microscope was used (Biological BX51, Olympus, USA), with a 1000X magnification ^{Ley de Coss et al., 2013}). The bacterial count was calculated using the formula: amount of bacteria=(mean) (dilution factor, 2×10⁷). In addition, the generation rate was determined in order to stablish the type of growth among inoculums (FRF vs. RBUpj) using the formula: N=N ₀ 2 ⁿ (1), where N: final number of cells, N ₀ : initial number of cells and n: number of generations during the exponential growth period. Therefore, the generation time (gt) of the bacterial population was calculated as t/n, where t is time and it is stablished by the data from the exponential growth phase, n was calculated from the logarithmic transformation of equation 1: n=3.3 [log N-log N ₀ ] (^{Madigan et al., 2009}).

Concentration of volatile fatty acids (VFA)

In order to determine the VFA concentration, new ACM inoculated with FRF and RBUpj were kept (by triplicate) at 38.5 °C. Then, 2 mL of culture media, we added to 0.5 mL of 25 % metaphosphoric acid. The mixture was placed in plastic vials and centrifuged at 17 664 g for 10 min, were then stirred in a vortex (MED Science, MX-S, USA) and stored at 10 °C. For the analysis, 2 mL of the thawed sample, at 25 °C and centrifuged at 34 622 g for 20 minuntes were deposited in vials for chromatography. The VFA content was determined in a gas chromatograph (PerkinElmer^TM, Claurus 500; USA) with auto-sampler and capillarity column (ELITE-FFAP 15 m, Perkin-Elmer^TM), flame ionization detector (FID), N₂ carrier gas at 60 psi, H₂ and O₂ were used to generate the flame with a 45 and 450 mL min^-1 flow, according to the method described by ^{Cobos et al. (2011)}.

Ammonium nitrogen concentration

Two mL of the culture media were used to measure DMivD (by triplicate), were centrifuged at 1000 g for 10 min, 1.5 mL of the supernatant were then mixed with 0.375 mL of metaphosphoric acid (4:1). These were kept refrigerated and then centrifuged at 13 000 g for 25 min. The supernatant was then recovered in 2 mL vials and stored at -10 °C until their evaluation; 20 µL of this supernatant were mixed with 1 mL of phenol and 1 mL of sodium hypochlorite. The samples were incubated at 37 °C during 30 min in a water bath. 5 ml of distilled water were then added to dilute the samples and homogenized with a vortex (Science MED MX-S USA). The absorbance was measured at 630 nm in a UVVIS spectrophotometer (Perkin-Elmer^TM, Lambda-40, USA) calibrated with a method (r²=0.99) for ammonium nitrogen concentration according to ^{McCullough (1967)}.

Design and statistical analysis

The experimental design was completely random with four replications. Data of DM degradation, VFA, ammonium nitrogen concentration and pH of the culture media were analyzed using an GLM procedure, while the total bacteria concentration was analyzed by the Wilcoxon signed-rank test (^{SAS, 2009}). Mean values were compared with the Tukey test (p˂0.05).

DM in vitro degradation

The DMivD of the pulp was higher (p≤0.05) in T1 (68.66 %) than in T2 (58.00 %) only after 72 h (Table 1). In both treatments, degradation of the substrate took place from the 3^rd hour of incubation on. This indicates that the bacteria adapted to the culture media and degraded J. curcas pulp without detoxification. ^{Kasuya et al. (2013)} indicate that DMivD of the J. curcas pulp without detoxification was 54.9 % and increased to 77.9 % when modified with the Pleurotus ostreatus fungus for 45 d; besides, the concentration of phorbol esters in the pulp decreased 99 %. Martinez et al. (2005) report that the in vitro degradation of crude J. curcas pulp protein mixed with enzymes to simulate non-ruminant digestibility was 78.6 % and increased to 86 % when pre-heated. In our study the phorbol esters degradation was not evaluated, so it is not possible to assure that the bacteria will degrade it. The degradation of phorbol esters by Pleurotus ostreatus, B. adusta, Ganoderma recinaceum and P. rufa fungi has been documented (^{Dias et al., 2014}; ^{Barros et al., 2011}). Phorbol esters degradation occurs due to the synergetic action of the enzymes that catalyze the depolymerization of lignin (xylanases lignocellulosic activity, extracellular cellulases and manganese peroxidase). The phorbol esters degradation period is 15 to 30 d (^{Dias et al., 2014}; ^{Najjar et al., 2014}.). This enzymatic activity is present in the rumen bacteria (^{Hazlewood and Teather, 1988}; ^{Dehority, 2003}; ^{Cobos et al., 2011}); but the incubation period must be more of than 15 d in order to observe its effect.

The pH in T1 was higher (p≤0.05) than in T2 at 72 h, without difference between treatments (p>0.05) in the other hours (Table 1). In both treatments pH was lower than 6.0 which indicate that bacteria producing VFA or organic acids, were active (^{Scandolo et al., 2007}) and adapted to the acid medium. ^{Dehority (2003)} indicate that the pH is modified by the organic acids produced during fermentation; nevertheless, bacterial activity is affected when the medium pH is lower than 6.0, mainly that of those bacteria using structural carbohydrates (^{Reynolds et al., 1993}). The pH in open systems, such as the rumen, may be less than 6.0 due to lactic acid accumulation. This occurs when the diet is rich in carbohydrates (which are easily fermented) or fiber deficient (^{Owens, 1998}). Nevertheless in in vitro studies, which are closed systems, the accumulation of organic acids in the media affects the pH because there is no flow to another system, which could cause the lower than 6 pH value obtained in our study. ^{Martinez et al. (2006)} point out that the starch content in J. curcas defatted pulp is 112 g kg^-1 DM, which is low compared to the content in corn. Thereon, ^{Alfaro et al. (2009)} report that starch content of 16 yellow corn hybrids ranges between 650 and 756 g kg^-1 DM (711 g kg^-1 mean DM).

Total bacterial count

The initial count of inoculated bacteria (FRF and RBUpj) to the medium was of 10⁴ mL^-1 of culture medium, which was obtained in the test to measure the DMivD of the pulp. The account of total bacteria at 24 h and 48 h was higher (p≤0.05) in T1 than in T2, but was higher at 96 h (p≤0.05) in T2 (4.06) (Table 2). The total bacterial count did not increase during the incubation period, and was lower than that reported in the rumen (10¹⁰ to 10¹² bacteria mL^-1) by ^{Dehority (2003)}. The lower than 6.0 pH value in both treatments may affect growth, since in F. succinogenes, R. flavefaciens and R. albus reduce their growing rate when their media pH is lower than 6.2 (^{Reynolds et al., 1993}; ^{Dehority, 2003}).

Although hetero-fermentation of the FRF and RBUpj bacteria was observed, the exponential growth was 3 to 3.5 h^-1. This is slow when compared to 1.5 h^-1 obtained by ^{Kunene et al. (2000)} and ^{Ley de Coss et al. (2013)}. They used a medium based on D-(+)-glucose and similar crop conditions; in addition, generation rate was 1.36 and 1.09 h with FRF and RBUpj, respectively. This is also slow compared to rumen bacteria, whose generation rate is between 15 and 30 min. The exponential growth and the generation rate could have been affected by the lower than 6 pH and the lowinitial concentration of bacteria in the inoculum (10⁴). The population of bacteria decreased from 72 h at T1 and increased in T2 at 96 h, indicating that the RBUpj consortium adapted better to the environment.

Volatile fatty acids concentration (VFA)

The concentration of propionic acid, butyric acid and total VFA had no difference (p>0.05) during the sampling hours (Table 3). The acetic acid concentration was higher (p≤0.05) in T1 at 3, 6 and 12 h of incubation. The normal VFA proportion produced in the rumen with optimal diets that contain less than 40 % soluble carbohydrates is: 60 to 65 % acetic, 25 % propionic and 10 to 15 % butyric (^{Cobos, 2007}). The proportion of VFA calculated in our research indicate that the acetic acid concentration was lower than the normal value, and the proportion of propionic acid and butyric acid is similar to that of a rumen under normal feeding conditions (^{Dehority, 2003}). This indicates that in both treatments there was heterofermentative activity.

Ammonium nitrogen concentration

The concentration of NH₄ in in vitro fermentations was higher (p≤0.05) in T2 at all incubation periods (Table 4). This indicates that the bacteria consortium developed in the selective medium with J. curcas pulp was better adapted to the conditions and showed higher proteolytic activity given by the synergy that may exist. In some cases up to 90 % of the variation in the proteolytic activity between strains was associated with the available substrate (^{Cotta and Hespell, 1986}); besides, cooperation between species has shown more activity. ^{Wallace (1985)} observed that Butyrivibrio fibrisolvens, Selenomonas ruminantium and Streptococcus bovis grew better in pairs when the culture medium had casein as the sole nitrogen source. This response indicated that bacteria with proteolytic activity can cooperate to its increase. According to ^{Dehority (2003)}, the rumen bacterial species with greater proteolytic activity are: Ruminobacter amylophilus, Prevotella ruminicola, S. ruminantium, B. fibrisolvens and S. bovis; where the primary products are amino acids and peptides, and P. ruminicola is the greatest producer of NH₃.