In Mexico, the Fabaceae family is represented by 92 genera and three sub-families: Leguminosoideae, Papilionoideae, and Caesalpinoideae, in Campeche this family has a high number of species (^{Zamora et al., 2011}) which might be used as animal feed. To be viable as a foraging resource, the legumes need to have nutritional characteristics that equal, or surpass protein content, and digestibility of other plants traditionally used as forage. The objective of this work was to determine the chemical composition, antinutritional factors and in situ digestibility of seven tree legumes of Campeche, Mexico.

Plant material was collected in the state of Campeche, in the municipalities of Edzna, China, Nueva Esperanza, and Campeche. The collection time of the material was the same for all species: rainy season (June-October) and dry season (March-May).

The studied species included: Lysiloma latisiliquum (L.) Benth., Bauhinia divaricata L., Senna racemosa (P. Mill.) Irwin et Barneby, S. pendula (Willd.), Albizia lebbeck (L.) Benth, Piscidia piscipula (L.) Sarg. and Lonchocarpus rugosus Benth. subsp. rugosus. Crude protein (CP), ash (^{AOAC, 2005}), neutral detergent fiber (NDF), acid detergent fiber (ADF), and lignin (^{Van Soest et al., 1991}), were determined. The following antinutritional factors were also analyzed: trypsin inhibitor (^{Kakade et al., 1974}), saponins (^{Monroe et al., 1952}), hemagglutinins, through the dilution technique (^{Jaffe et al., 1974}), and ureasic activity (^{AOAC, 2002}).

In situ digestibility of dry matter was evaluated (^{Orskov and McDonald, 1979}), using nylon bags (10 x 20 cm) with porosity 52 ± 10 mm, in which were placed 5 g of the sample (DM). Three cannulated Holstein cows, mean body weight of 450 kg, were used. Incubation times were 0, 4, 8, 16, 24, 48, and 72 h. The bags were later rinsed five times with running water for one minute until water was clear and dried in an air stove at 55 °C. In situ digestibility was calculated as the amount of disappeared DM.

Data were analyzed as repeated measurements using the PROC MIXED procedure of SAS (^{SAS, 2002}) for a completely randomized design. Means were compared by Tukey test (p< 0.05).

Differences (p ˂ 0.05) were seen among the studied legumes regarding ash content in the dry season. The greatest values were observed in S. pendula and P. piscipula (14.2 and 11.9%); these species also showed the highest concentrations in the rainy season (Table 1). A. lebbeck and S. pendula showed the highest percentages of CP (p˂ 0.05) in the dry and rainy seasons. The highest NDF value in the dry season was for A. lebbeck (66.8%), followed by B. divaricate (60.9%), S. racemosa (58.8%), and S. pendula (58.3%). In the rainy season, the highest was for L. rugosus (75.4%). On the other hand, the highest ADF value in the dry season was for A. lebbeck (46.2%) followed by L. rugosus (43.5%) and P. piscipula (40.7%). In the rainy season, L. rugosus and L. latisiliquum showed the highest values (63.2 and 57.9%, respectively), while S. racemosa had the lowest value (27.3%).

Table 1 Composition of tree legumes collected in Campeche, Mexico during the dry and rainy season in DM basis. 

S= estación seca; LL= estación lluviosa. Los valores en la misma columna con diferente letra son las diferencias entre ellos (p< 0.05).

The highest lignin content (p˂ 0.05) in the dry season was for L. rugosus (25.4%), while in the rainy season was for L. latisiliquum (12.8%). The results in this study for A. lebbeck are similar to those obtained by ^{Rodriguez et al. (2005)} for ash, ADF and NDF, with slight variations, possibly because of the different re-budding ages used. ^{López et al. (2008)} have reported the following values for P. piscipula harvested in Quintana Roo: 12.6% CP; 50% FDN; 34.6% FDA and 17.2 for lignin. With the exception of the lower concentration of CP observed by these researchers, the values are similar to those found in the present study. The lignin content of the plants in this study is within the normal values for tropical shrub legume species, as mentioned by ^{Camero et al. (2001)}.

In the rainy season, most of the evaluated species showed a variation in the concentration of the trypsin inhibitor, with high values in L. rugosus, S. pendula and B. divaricata, and the lowest in S. racemosa. In the dry season, they were surpassed by L. latisiliquum, L. rugosus, and S. pendula; on the other hand, the lowest values were observed in A. lebbeck and S. racemosa (Table 2). This compound results in the decrease of the trypsin activity, causing growth problems and pancreas hypertrophy. However, ruminants are less affected than non-ruminants because trypsin inhibitors are degraded slowly in the rumen (^{Fuller, 2004}).

Table 2 Antinutritional factors in tree legumes collected in Campeche, Mexico. 

S= estación seca; LL= estación lluviosa. TIU= unidades de inhibidor de tripsina, a= maxima dilución que produce aglutinación en 1 h, b= aumento en unidades de pH.

Saponin was not present in the legumes studied, with the exception of L. latisiliquum which content was low in the dry season however this concentration is not harmful for ruminants. Hemagglutinins were present in L. latisiliquum in both rainy and dry season and in L. rugosus in the dry season. The ureasic activity can be harmful, since it produces ammonia when urea is used as a supplement values of 0.11 units or less are considered aceptable (^{Arelovich, 2008}). The values observed in this study were lower to this amount, except for L. latisiliquum and A. leebeck in the dry season.

In situ digestibility of the forage sampled in the rainy season showed significant differences (p< 0.05) among species in the first four hours of incubation, with the highest values for S. pendula (59.1%) and S. racemosa (46.2%), while L. rugosus showed the lowest value (32.1%; Figure 1).

Figure 1 In situ DM digestibility of tree legumes collected in the rainy season in Campeche, Mexico. 

S.pendulaandS.racemosabehavedsimilarlyduringincubation, with higher values at 72 h (89.1 and 88%, respectively), followed by B. divaricata and P. piscipula (67.1 and 60.2%, respectively). The specie with the lowest digestibility was L. rugosus with 32.0 and 40.1% at 4 and 72 h incubation.

Forage collected in the dry season showed differences (p< 0.05) among species, being S. racemosa, B. divaricata, and S. pendula statistically similar with 51.1% at 4 h incubation, but different to A. lebbeck and P. piscipula with 34% and these last ones similar to L. rugosus and L. latisiliquum (Figure 2). As incubation time passed, the first three species kept the highest digestibility, up to 72 h. In both the dry and rainy seasons, the lowest value was for L. rugosus, with 48.1% digestibility of DM at 72 h.

Figure 2 In situ DM digestibility of tree legumes collected in the dry season in Campeche, Mexico. 

In situ digestibility of guacimo (Guazuma ulmifolia Lam.) has been reported as 77.7% (^{Pinto et al., 2004}). It is considered a forage species because of both its acceptability and its high digestibility. Its digestibility value is near to those found in this study for S. pendula and B. divaricata at 24 h incubation. Leucaena leucocephala (Lam.) de Wit is a legume used in animal production systems, which has shown, in the rainy season, a digestibility at 48 h of 66.2% and at 72 h, 68.0% (^{Sánchez et al., 2008}). These values are similar to those obtained for B. divaricata at 48 and 72 h, as well as for P. piscipula at 72 h.

Values above 60% after 20 h of incubation have been reported for Leucaena leucocephala (^{Razz et al., 2004}). This value is comparable with the results obtained for B. divaricata and S. pendula. The digestibility reported for Gliricidia sepium (^{Olivares et al., 2005}) was 36, 50, and 70% at 12, 24, and 48 h, similar to L. rugosus, mainly at 16 and 24 h (36.1 and 43% digestibility), respectively.