Location

Samples were obtained from the ‘La Posta’ Experimental Field belonging to the National Institute of Forestry, Agriculture and Livestock Research (INIFAP), located in Paso del Toro, Medellin, Veracruz, between parallels 18° 50’ and 19° 09’ north latitude and the meridians 96° 02’ and 96° 16’ west longitude, 50 to 55 masl, average annual temperature 24 to 28 °C, precipitation 1 100 to 1 600 mm, average relative humidity 31% (^{INEGI, 2009}).

Obtaining the samples

Plants (30 days old) were manually cut, bagged and labeled and dried at 55 °C in a forced air oven to constant weight. Subsequently, they were ground and identified, placed in plastic bags for analysis in the Animal Nutrition laboratory of the Department of Zootechnics of the Autonomous University Chapingo (UACH), state of Mexico.

Determined variables

The following variables were determined (^{AOAC, 1990}; ^{Van Soest et al., 1994}; ^{Giraldo et al., 2007}; ^{Navarro-Ortiz and Roa-Vega, 2018}), all in g (100 g)^-1 of sample: dry matter (DM), crude protein (CP), crude fiber (CF), ether extract (EE), ash (ASH), organic matter (OM), acid detergent fiber (ADF), neutral detergent fiber (NDF), crude lignin (CL), hemicellulose (HEM), cellulose (CEL), cellular content (CC), in vitro digestibility of dry matter (IVDDM) and in situ digestibility of DM (DISDM).

In vitro and in situ digestibility of dry matter

In a 100 cm³ plastic tube, 0.5034 ±0.002 g of sample was weighed, 20 cm³ of Mc Dougall’s saliva were added as buffer and 5 cm³ of rumen liquid. The mixture was incubated at 39 °C for 48 hr, with shaking every 12 hr. The residual NDF was determined and the following formula was applied to calculate the digestibility of the sample in g (100 g)^-1: IVDDM= 100-(residual NDF/g of sample)×100. Anaerobic conditions were achieved using a CO₂ stream. The ruminal fluid was obtained from a fistulated Holstein bovine (^{Giraldo et al., 2007}; ^{Navarro-Ortiz and Roa-Vega, 2018}), who was consuming an alfalfa-based diet. DISDM was determined similarly but 5 × 5 cm Ankom^® bags (Ankom Co., Fairport, NY, USA) were used instead of a plastic tube. 0.5402 ± 0.0472 g of sample were placed in the bags, the same fistulated Holstein bovine was used and the bags remained 48 h in the rumen. After this time, the bags were dried at 100 °C, the residual NDF was determined, and a formula similar to the above was used for the calculation of DISDM.

Experimental design and statistical analysis

A completely randomized experimental design was used with four treatments (four legumes), three repetitions for the variables DM, CP, CF, EE, ASH, MO, NDF, CL, CEL, HEM and CC and four repetitions for the IVDDM variables and DISDM. After performing the analysis of variance, the Tukey test was used for the separation of means (p< 0.05), using the Statistical Package for the Social Sciences (^{SPSS, 2011}) version 8.0.

Proximal analysis

The variables evaluated from this analysis (Table 1) were: crude protein (CP), ash (ASH), ethereal extract (EE), crude fiber (CF), organic matter (OM) and nitrogen-free extract (NFE). Only CP, ASH and OM were statistically different between legumes. A. pintoi 25.2 g (100 g)^-1, showed the highest value of CP, followed by C. ternatea 22.7 g (100 g)^-1; while S. guianensis 16.2 g (100 g)^-1 and M. atropurpureum 16.5 g (100 g)^-1, showed the lowest values. The ASH variable was (p< 0.05) the highest in A. pintoi 9 g (100 g)^-1, intermediate in S. guianensis 8.1 g (100 g)^-1 and the lowest in C. ternatea 6.8 g (100 g)^-1 and M. atropurpureum 6.4 g (100 g)^-1. These CP and ASH values were close to those of the consulted literature, with plant ages between 30 and 60 days.

^{García-Ferrer et al. (2015)} reported 21.8 g (100 g)^-1 of CP for A. pintoi, 18.3 g (100 g) ^-1 for C. ternatea and 14.5 g (100 g)^-1 for S. guianensis. ^{Kavana et al. (2005)}reported 15.3 g (100 g) ^-1 for C. ternatea, 13.4 g (100 g)^-1 for M. atropurpureum and 12.6 g (100 g) ^-1 for S. guianensis. That is, the order of CP found in the present study is conserved between species, with A. pintoi showing the highest values of CP. ^{Sotelo et al. (2018)} and ^{Oyekunle et al. (2018)} reported 12.0 g (100 g) ^-1 and 10.9 g (100 g) ^-1 of ASH for S. guianensis and A. pintoi, respectively (high values). ^{La O et al. (2006)} reported 8.3 g (100 g) ^-1 of ASH for M. atropurpureum and ^{Bugarin et al. (2009)} reported 6.7 g (100 g)^-1 of ASH for C. ternatea (low values). In this study also, A. pintoi and S. guianensis showed the highest ASH values and the other two legumes the lowest values.

All the legumes studied showed CP values greater than 16 g (100 g) ^-1. But A. pintoi and C. ternatea showed CP values greater than 22 g (100 g)^-1, which makes them good sources of protein for animals in the tropics. Probably due to the nutritional quality of A. pintoi and C. ternatea, ^{Jusoh and Nur-Hafifah (2018)} found a higher preference in rabbits for A. pintoi, followed by C. ternatea and less for S. guianensis.

Van Soest’s analysis

The variables evaluated from this analysis (Table 2) were: neutral detergent fiber (NDF), acid detergent fiber (ADF), crude lignin (CL), hemicellulose (HEM), cellulose (CEL) and cellular content (CC).

Cell content

Legume A. pintoi produced the highest values 67.1 g (100 g)^-1 of CC (non-fibrous compounds) and M. atropurpureum, showed the lowest values 49 g (100 g)^-1. S. guianensis was intermediate and C. ternatea was low-intermediate in this variable that represents the non-fibrous or soluble compounds (Table 2). The CC values shown by A. pintoi were higher than those reported by ^{Castaño and Cardona (2015)} who found 47.7 g (100 g)^-1. At different ages (20 to 80 days old) and times of the year, ^{García Ferrer et al. (2015)} found CC values from 32.8 g (100 g)^-1 to 55.2 g (100 g)^-1 in A. pintoi; likewise, in S. guianensis they found values from 25.9 g (100 g)^-1 to 42.9 g (100 g)^-1. That is, according to these authors, A. pintoi tends to be better than S. guianensis in this variable.

Fiber variables

Contrary to cellular content, A. pintoi produced the lowest values of all the fiber variables: NDF, ADF, CL, HEM, and CEL and M. atropurpureum showed the highest values. A. pintoi together with S. guianensis showed the lowest CL values (Table 2). ^{García-Ferrer et al. (2015)} found NDF values of 54.8 g (100 g)^-1 for A. pintoi, 57.1 g (100 g) ^-1, for S. guianensis and 57 g (100 g) ^-1 for C. ternatea. ^{Lagunes-Rivera et al. (2019)} found ADF values of 35 g (100 g)^-1 for A. pintoi and 44 g (100 g) ^-1 for S. guianensis.

These values are greater than those of the present study, probably because the intervals between cuts were greater than 30 days in the works of these authors. But, although these values of the fiber variables are high compared to those of the present study, they maintain the relationship of low values of A. pintoi and high values of the other legumes. In a study by ^{Lagunes-Rivera et al. (2019)}, the species A. pintoi had lower fiber content, contrasting with the other herbaceous legumes that they studied.

In vitro and in situ digestibility of dry matter

Legume A. pintoi showed the highest values of IVDDM and DISDM, and M. atropurpureum showed the lowest values of these variables. S. guianensis and C. ternatea showed intermediate values (Table 3). The IVDDM values for A, pintoi, S. guianensis, C. ternatea and M. atropurpureum were: 84.5 g (100) ^-1, 74 g (100) ^-1, 68.9 g (100) ^-1 and 63 g (100) ^-1, respectively. The DISDM values, in the same order of these legumes, were: 88 g (100) ^-1, 73.7 g (100) ^-1, 69.9 g (100) ^-1 and 56.6 g (100) ^-1, respectively. That is, A, pintoi showed the highest digestibility and M. atropurpureum showed the lowest value of this variable.

^{García-Ferrer et al. (2015)} found, at 21 days of regrowth, DISDM values of 80.2 g (100), 74 g (100)^-1 and 76 g (100) ^-1 for A. pintoi, S. guianensis and C. ternatea, respectively. Values similar to those of the present study (Table 3). ^{Lagunes-Rivera et al. (2019)} found that A. pintoi showed the highest IVDDM values compared to three other forage legumes from the tropical region of the state of Puebla, Mexico.

The IVDDM and DISDM values of A. pintoi were very good, which implies that, if the cellular content is digested by 99 g (100) ^-1 (^{NRC, 2001}), then the in vitro and in situ digestibilities of the NDF it will be 51.1 g (100) ^-1 and 65.8 g (100) ^-1, respectively.

Correlations between the variables studied

IVDDM and DISDM showed a positive correlation with ASH (p< 0.05). On the other hand, ASH showed a positive correlation (p< 0.05) with CC. Consequently, CC had (p< 0.05) positive correlation with IVDDM and DISDM (Table 4). This correlation was also deduced from the positive correlation (p< 0.05) between CC and CP. In other words, the greater the cellular content, the greater the digestibility.

CP was positively correlated with CC and this variable had a positive correlation with IVDDM and DISDM (Table 4), therefore, CP had a positive correlation with digestibilities (p< 0.05). In other words, the greater the protein, the greater the digestibility.

As expected, the ADF was positively correlated (p< 0.05) with CEL (Table 4). This is because cellulose is a component of the cell wall (NDF). The NDF was negatively correlated (p< 0.05) with CP and the digestibilities (p< 0.05). In agreement, ^{García-Ferrer et al. (2015)} found a negative correlation (p< 0.01) between NDF and DISDM. Likewise, in this study, HEM (component of the NDF) was (p< 0.05) negatively correlated with DISDM (Table 4). That is, the higher the content of the fiber variables, the lower the protein and the lower the digestibility.

These negative correlations indicate that as the fibrous fraction of the plant is high, either due to its older age, due to environmental differences (^{Castaño and Cardona, 2015}) or for genetic reasons, its fibrous components increase and its cellular contents decrease. The first two factors are excluded, since the four legumes were the same cutting age and were affected by the same environmental factors. As the fiber is partially soluble and the cellular content is highly soluble (^{NRC, 2001}), it follows that the digestibility decreases because the fibrous fraction of the legume increases.