Introduction

Erythrina americana is an arboreal legume considered as a multipurpose tree (^{García et al., 2001}; ^{Villanueva et al., 2015}). It has been proposed as a forest restorer, especially those located in the tropical region with a dry season (^{Fehling and Ceccon, 2015}), the ethnomedical use of its flowers and seeds has been documented in Mexico (^{García et al., 2001}; ^{Pino et al., 2004}; ^{Sotelo et al., 2007}); in livestock farming it is used as part of the fence (live pole), which, additionally, provides shade and firewood, and it works ass as a windbreak curtain (^{Grande et al., 2013}; ^{Oliva et al., 2013}).

The content of crude protein and structural carbohydrates in the foliage of E. americana make its use attractive as a food supplement for small ruminants (^{Benavides, 1999}; ^{Grande, 2010}). Particularly, sheep easily consume the foliage of the species and thus, can be incorporated in their diet up to 30% (^{Pinto et al., 2003}, ^{Best et al., 2017}). This legume produces foliage all year round when it receives controlled pruning (^{Meléndez, 2003}), and it can be established during that period (^{Enríquez et al., 2011}).

E. americana grows in the states of Puebla, Veracruz, Tabasco, Chiapas and Yucatán (^{Llera and Mélendez, 1994}) and is often located in livestock production units (^{Reyes and Jiménez, 1998}; ^{Grande, 2010}), a situation that facilitates the acquisition of vegetative material for its propagation. Despite its wide distribution in the tropical region of Mexico, there are few studies on the yield of foliage per tree (^{Meléndez, 2003}, ^{Oliva-Hernández and López-Herrera, 2017}) and the efficiency of manual harvest of the leaves in trees that have not had previous pruning is unknown. Therefore, the objective of the present study was to determine the influence of the harvester-technitianon the efficiency in manual harvesting of E. americana foliage that is part of the live fence.

Materials and Methods

Results and Discussion

Table 1 shows the chemical composition of E. americana foliage. It is confirmed that the crude protein content was higher than that of tropical pastures (^{Enríquez et al., 2011}). However, the crude protein in the foliage was lower than that indicated by ^{Reyes and Jiménez (1998)} and ^{Hernández-Espinoza et al. (2017)}, who indicate values between 15.4 and 25.6%. Differences in crude protein between studies can be attributed to the time of harvest and age of regrowth of the foliage (^{Benavides, 1999}; ^{Hernández-Espinoza et al., 2017}).

In regard to structural carbohydrates, ^{Reyes and Jiménez (1998)} indicate figures between 58.3 and 60.9 % for neutral detergent fiber and 41.9 to 41.4 % for acid detergent fiber, which are higher than what was recorded in this study for the neutral detergent fiber, and similar for the case of acid detergent fiber. The increase in the value of the first reduces the feed intake in ruminants. Depends on the state of maturity of the plant, when it is greater, so is neutral detergent fiber, hence the importance of determining the factors that affect its content in the foliage of trees with potential to be used as food for ruminants (^{Pinto et al., 2002}; ^{García, 2003}).

The foliage of arboreal legumes is characterized by its attractive protein value (^{Pinto et al., 2002}, ²⁰⁰³). However, it is important to consider that in this type of foliage there are secondary metabolites that protect the plant and help maintain balance in the plant-herbivore interaction (^{Ramos et al., 1998}).

In the Erythrina genus, a large number of secondary metabolites have been isolated in different parts of the plant (seeds, flowers, bark, leaves, roots), highlighting alkaloids (38 %), flavonoids (38 %) and proteoids (9 %) as families of chemical compounds that are cited with greater frequency (^{García et al., 2001}; ^{Pino et al., 2004}; ^{Ibarra et al., 2011}).

It is in the leaves of Erythrina that there are the fewest quantitative and qualitative studies to determine the presence of secondary metabolites. However, it has been determined that alkaloids (64 %), flavonoids (27 %) and triterpenoids (9 %) are the chemical families that have a higher frequency of presentation (^{Pino et al., 2004}). In the foliage of Erythrina goldmanii Standley. ^{Pinto et al. (2002}, ²⁰⁰³) report the presence of total phenols (0.6 %) and condensed tannins (16.28 g kg^-1 MS). Further studies are required to determine the influence of the level of inclusion of E. americana in the diet of ruminants on the productive behavior and state of health.

A wide variation was observed in the amount of foliage and branches harvested; the greatest variation corresponded to the foliage in relation to the branches (Table 2). The harvester-technician did not affect (p> 0.05) any of the variables studied. The wide variation detected in the harvested material between and within technicians contributed to explain the absence of differences between the collectors (Table 3). In this regard, ^{García and Oliva (2012)} studied a cocoite plantation (Gliricidia sepium (Jacq.) Kunth ex Walp.), with no background in the management of pruning and foliage harvesting; the authors point out an efficiency in the harvest of foliage of this species of 7.0 ± 2.7 kg h^-1 on a humid basis and 2.3 ± 0.9 h^-1 on a dry basis, results that are similar to what was recorded in the present study (Table 2) . Although the branches of E. americana have thorns, their presence was not a limitation in the manual harvest process; nevertheless, it was necessary to use raw-leather gloves to avoid injuries in the hand of the harvester.

The foliage relationship branches^-1 was lower than the one indicated by ^{Oliva-Hernández and López-Herrera (2017)} who declared a foliage relationship branches^-1 between 2.0 and 3.2 in trees of E. americana, when the trees have been subjected to a process of pruning at 90 -day intervals. The lowest branch^-1 foliage ratio detected in the present study is attributed to the greater weight and size of the branches, because the trees used had not been pruned.

In the study by ^{García and Oliva (2012)} it was determined that the crude protein in the G. sepium foliage was 23.6 %, and if it is considered that a harvester reached 2.3 kg of foliage on a dry basis h^-1, 543 g result of crude protein from G. sepium, which was higher than that obtained with E. americana (Table 2). The different crude protein content in the foliages of G. sepium and E. americana explains the contrast between studies.

On the other hand, there is evidence that the foliage of E. indica Lam can be incorporated into the diet of sheep by 30% in substitution to the use of soybean paste (^{Best et al., 2017}). A harvester produces 1.7 kg of foliage on dry basis h^-1, equivalent to 247 g of crude protein, this amount of protein covers 30% of the crude protein ingested by seven Pelibuey sheep with 20 kg of live weight that receive a complete diet with 2.8 Mcal of metabolizable energy kg of MS^-1 and 19.8 % of crude protein (^{Pascual et al., 2009}).

The availability of E. americana foliage most of the year (^{Meléndez, 2003}), the facility for its vegetative macropropagation compared with the sexual reproduction that requires scarified seeds to germinate (^{Fehling and Ceccon, 2015}), its feasible presence as a live fence (^{Grande et al., 2013}; ^{Oliva-Hernández and López-Herrera, 2017}) and the amount of crude protein obtained from the foliage during a three-hour working day favors this food resource as an option to give sustainability to the production systems of sheep that use multipurpose trees as part of living fences.

Conclusions

The efficiency in harvesting foliage and branches in E. americana without previous pruning history is not affected by the technician-harvester, which indicates that specific training is not required to perform such activity, a situation that facilitates the collection of foliage for use as ruminant food. The content of crude protein and structural carbohydrates in the foliage of E. americana allows it to be considered an ingredient to complement the crude protein requirements of small grazing ruminants or to be part of integral diets.