LIVESTOCK, NUTRITIVE VALUE AND LOCAL KNOWLEDGE OF FODDER TREES IN FRAGMENT LANDSCAPES IN CHIAPAS, MEXICO

Guillermo Jiménez-Ferrer, Hernán Pérez-López, Lorena Soto-Pinto, José Nahed-Toral, Lorenzo Hernández-López and Jesús Carmona

José Guillermo Jiménez-Ferrer. Doctor, Universidad Autónoma de Yucatán, Mexico. Researcher, El Colegio de la Frontera Sur (ECOSUR), México. Address: ECOSUR, Carr. Panamericana y Per. Sur s/n, Barrio María Auxiliadora, 29290 San Cristóbal de las Casas, Chiapas, Mexico. e-mail: gferrer@sclc.ecosur.mx

Hernán Pérez-López. M.C., Universidad Autónoma de Chapingo, México. Agroforestal Consultant, ECOSUR, Mexico.

María Lorena Soto-Pinto. Doctor, Universidad Autónoma de México (UNAM). Researcher, ECOSUR, Mexico. e-mail: lsoto@sclc.ecosur.mx

José Nahed-Toral. Doctor, UNAM, Mexico. Researcher, ECOSUR, Mexico. e-mail: jnahed@sclc.ecosur.mx

Lorenzo Hernández-López. Anthropologist, Universidad de Chiapas, Mexico. Academic Technician, ECOSUR, Mexico.

Jesús Carmona de la Torre. QFB, Universidad Autónoma Metropolitana, Mexico. Laboratories Coordinator, ECOSUR, Mexico.

SUMMARY

The purpose of the present study was to characterize the livestock system and to identify nutritional potential and local knowledge of fodder trees and shrubs in two localities of Chiapas, Mexico. The study included participatory rural evaluation techniques, semi-structured interviews, laboratory analysis and a palatability essay with "Pelibuey" sheep. The latter was carried out through a random design. The livestock system was classified as agrosilvopastoral, characterized by its access to a variety of different ecological areas (crop, fallow lands and pine-oak forest). Thirteen multi-purpose woody fodder species were identified and ranked by farmers: Acacia angustissima, A. pennatula, Casimiroa edulis, Diphysa robinoides, Ficus cookii, Leucaena brachycarpa, Litsea glaucescens, Lysiloma auritum, Pistacia mexicana, Quercus crispipilis, Q. segoviensis, Rhus schiedeana and Tecoma stans. Ten uses and 43 products or services were identified for fuelwood (11), living fences (9), shade (9), medicinal purposes (7), tools (3), food (2), and condiment, charcoal, tanning and ornamental use (1 of each). The best ranked species by producers according to fodder availability during the dry season, palatability, livestock fattening and propagation were A. angustissima, C. edulis, L. brachycarpa, L. auritum y P. mexicana. Chemical analysis, in vitro digestibility and secondary compounds of 13 species showed significant variation among species. Crude protein content was 8.1-22.7% and total phenols 5.4-8.4%. Alkaloids showed low-medium values. The palatability test (p<0.05) showed that the most consumed species (g/ms/animal^-1) were A. angustissima (69.48), L. auritum (54.85) and A. pennatula (35.69).

GANADERÍA Y CONOCIMIENTO LOCAL DE ÁRBOLES FORRAJEROS EN PAISAJES FRAGMENTADOS DE CHIAPAS, MÉXICO

RESUMEN

El objetivo del presente estudio fue caracterizar el sistema ganadero e identificar el potencial nutricional y conocimiento local de árboles y arbustos forrajeros en dos comunidades de Chiapas, México. Se usaron técnicas de evaluación rural participativa, entrevistas semi-estructuradas, análisis de laboratorio y un ensayo de palatabilidad con borregos "pelibuey". Este último fue realizado mediante un diseño al azar. El sistema ganadero fue clasificado como agrosilvopastoril, caracterizado por acceso a diferentes áreas agroecológicas (tierras agrícolas, barbechos, y bosques de pino y encino). Trece especies de árboles y arbustos forrajeros fueron identificadas y calificadas por los productores: Acacia angustissima, A. pennatula, Casimiroa edulis, Diphysa robinoides, Ficus cookii, Leucaena brachycarpa, Litsea glaucescens, Lysiloma auritum, Pistacia mexicana, Quercus crispipilis, Q. segoviensis, Rhus schiedeana and Tecoma stans. Se identificaron 10 usos y 43 productos o servicios: leña (11), cercos (9), sombra (9), medicinales (7), herramientas (3), alimentos (2), y condimento, carbón, curtiduría y ornamento (1 de cada uno). Las especies con mejor calificación por los productores, considerando palatabilidad y cantidad de follaje en época seca, ganancia de peso y facilidad de propagación fueron A. angustissima, C. edulis, L. brachycarpa, L. auritum y P. mexicana. La composición química, digestibilidad in vitro y compuestos secundarios del follaje de las 13 especies mostraron una variación significativa entre especies. El contenido de proteína cruda fue 8,1-22,7% y de fenoles totales 5,4-8,4%. Se encontraron valores bajos y medios de alcaloides. Las especies más consumidas (g/ms/animal^-1) según la prueba de palatabilidad (p<0,05) fueron A. angustissima (69,48), L. auritum (54,85) y A. pennatula (35,69).

CRIAÇÃO DE GADO E CONHECIMENTO LOCAL DE ÁRVORES FORRAGEIRAS EM PAISAGENS FRAGMENTADAS DE CHIAPAS, MÉXICO

RESUMO

O objetivo do presente estudo foi caracterizar o sistema de criação de gado e identificar o potencial nutricional e conhecimento local de árvores e arbustos forrageiros em duas comunidades de Chiapas, México. Usaram-se técnicas de avaliação rural participativa, entrevistas semi-estruturadas, análises de laboratório e um ensaio de palatabilidade com borregos "pelibuey". Este último foi realizado mediante um desenho aleatório. O sistema de criação de gado foi classificado como agrosilvopastoril, caracterizado por acesso a diferentes áreas agro ecológicas (terras agrícolas, barbechos, e bosques de pinheiros e encino). Treze espécies de árvores e arbustos forrageiros foram identificadas e qualificadas pelos produtores: Acacia angustissima, A. pennatula, Casimiroa edulis, Diphysa robinoides, Ficus cookii, Leucaena brachycarpa, Litsea glaucescens, Lysiloma auritum, Pistacia mexicana, Quercus crispipilis, Q. segoviensis, Rhus schiedeana and Tecoma stans. Identificaram-se 10 usos e 43 produtos ou serviços: lenha (11), cercas (9), sombra (9), medicinais (7), ferramentas (3), alimentos (2), e tempero, carvão, curtidora e ornamento (1 de cada um). As espécies com melhor qualificação pelos produtores, considerando palatabilidade e quantidade de folhagem em época seca, ganho de peso e facilidade de propagação foram A. angustissima, C. edulis, L. brachycarpa, L. auritum e P. mexicana. A composição química, digestibilidade in vitro e compostos secundários da folhagem das 13 espécies mostraram uma variação significativa entre espécies. O conteúdo de proteína crua foi 8,1-22,7% e de fenóis totais 5,4-8,4%. Encontraram-se valores baixos e médios de alcalóides. As espécies mais consumidas (g/ms/animal^-1) segundo a prova de palatabilidade (p<0,05) foram A. angustissima (69,48), L. auritum (54,85) e A. pennatula (35,69).

KEYWORDS / Agrosilvopastoral Systems / Chiapas / Fallow / Secondary Compounds /

Received: 05/12/2006. Modified: 02/07/2007. Accepted: 02/12/2007.

Introduction

In the last years, in academic and development forums on natural resources conservation, specialists have discussed the need for shifting from conventional livestock methods toward more sustainable production systems, capable of improving yields and quality (Mosquera et al., 2005). Furthermore, agricultural technical advisors, farmers and researchers have underlined the need for regarding local farmers’ knowledge and experience as a base from which to generate social and technological alternatives.

Fodder trees and shrubs have been recognized as valuable and accessible food sources, which may contribute to solving some nutritional constraints for livestock during the dry season. Also, this resource could contribute to the development of sustainable systems based on environmental services and multipurpose use (Sánchez and Rosales, 1999; Dagang and Nair, 2003). Previous studies have evaluated fodder tree species, local knowledge and potential for developing silvopastoral systems in several cultures and agroecological settings (Thapa et al., 1997; Thorne et al., 1999; Roothaert and Franzel, 2001).

Southern Mexico, particularly the Mayan Region, is an area of interest due to its impressive biodiversity (Steep, 2002). In this region, researchers have paid attention to biodiversity in forest fragments and cultivated areas managed by smallholders and indigenous communities, highlighting the importance of multipurpose fodder trees as an easily renewable resource for local communities (Soto-Pinto, 1990; Toledo et al., 1995; Nahed et al., 1997).

Several studies conducted in the Chiapas Highlands (Camacho et al., 1999; López et al., 2001) as well as in lower altitude areas (Guillén et al., 2001; López-Merlín et al., 2003) have shown that animals feed on trees and shrubs in fallow areas under different systems and management practices (Jiménez-Ferrer, 2000). Some examples of valuable fodder tree genera have been identified in Chiapas, such as Acacia, Alnus, Buddleia, Calliandra, Cordia, Diphisa, Erythirina, Gliricidia, Guazuma, Leucaena, Pithecellobium, Saurauia y Thitonia (Soto-Pinto, 1990; Jiménez-Ferrer, 2001; López et al., 2001).

In general, in Mexico there is scarce information concerning the design of robust cattle systems. This study was designed to describe the livestock system and determine the nutritive value and local knowledge about trees and shrubs in Trinitaria, Chiapas, Mexico. The research is part of the project "Livestock and Environment" of ECOSUR (El Colegio de la Frontera Sur) in Chiapas.

Study area

The study was carried out in two localities, Espatajamal and San Caralampio, in the municipality of Trinitaria, Chiapas (16º07’N; 92º03’W) at 1200masl, with an average rainfall of 970mm and average annual temperature of 19.3ºC. Local soils have been classified as lithosol, and vegetation as forest fragments dominated by pine and oak (INEGI, 2000). The dominant forest fragment species are Acacia milenaria, A. pennatula, Leucena brachycarpa, Pinus radiata, Pithecellubium dulce, Quercus segoviensis and Q. crispipilis.

Methods

Agrosilvopastoral system analyses and local knowledge of fodder trees

Participatory approaches for developing production systems with small-scale farmers have been widely used in past decades (IIED, 1994; GEA, 1993). In this study, participatory rural appraisal (PRA) techniques, including workshops, preference ranking exercises, semi-structured interviews and transects, were adopted to explore the livestock system and local knowledge of fodder trees and shrubs. Interviews and workshops were conduced from Jan to Apr, 2002. Twenty representative households were selected among two communities. Semi-structured interviews were carried out in order to explore information on household decision making, the role of livestock in the community, fodder tree preference and principal uses of these trees.

Through a researcher-facilitated workshop, each farmer ranked each species in a matrix, using one column for fodder trees named in the interviews and four columns with the following variables: fodder production during the dry season, palatability, livestock fattening and propagation capacity. These variables were previously explained to farmers, who classed each tree species using the following scores: 3 (high), 2 (intermediate) and 1 (poor; Roothaert and Franzel, 2001). In order to identify and collect species named in the workshops, with the help of farmers, a transect in fallows and forests was carried out. Each species was collected and identified in the ECOSUR Herbarium. Workshops and transects were conducted during Mar and Apr, 2002.

Chemical composition and in vitro digestibility

Based on the information from workshops and transects, consumable parts (i.e., leaves and twigs of about 1 year of age) of foliage from 13 trees and shrubs were randomly harvested by hand plucking from 3 to 5 trees of each species, during Apr and May, 2002 (dry season). Chemical analyses included determination (AOAC, 1990) of dry matter (DM) and crude protein (CP). Gross energy (GE) was measured using an adiabatic calorimeter. Concentrations of neutral detergent fiber (NDF) and acid detergent fiber (ADF) were also determined (Goering and Van Soest, 1970; Van Soest et al., 1991), and in-vitro dry matter digestibility (IVDMD) was evaluated according to Tilley and Terry (1963). Tannin content was determined using the Folin-Denis method, cited by Calvo and Morales (1988), and alkaloid and cyanogenic glucoside content was determined using qualitative methods (AOAC, 1990).

Palatability

Leaf palatability (PA) of selected species (Acacia angustissima, Acacia pennatula, Lysiloma auritum, Tecoma stans and Pistacia mexicana) was evaluated on-farm, using sheep. Tree foliage of each species was gathered from fallows. Six male Pelibuey Red wethers from the region were used, 1.2 ±0.6 years old and 21.7 ±2.0kg body weight. Each animal was kept in a traditional 4×4m enclosure with troughs containing the different feeds arranged around the perimeter, so that the animals had equal access to, and free choice among, five feeds. The feeds were randomly allocated among the troughs, and their positions were re-randomized daily to avoid the "habit reflex" (Kaitho et al., 1996). Tree foliage (200g fresh weight of leaves) from each species was offered 6h per day (8.00-14.00) for 8 consecutive days. In the afternoon (15.00–22.00h) sheep were fed fresh Taiwan grass (Pennisetum purpureum Schum) ad libitum. For 7 days before the trial started, the sheep were kept in pens during the same period each day, and all troughs were kept full in order to accustom the sheep to the pens and to the daily routine. The sheep were starved each night (22.00-8.00) and fed each morning. The leaves were harvested daily (5.00-7.00) and kept in the shade until given to the sheep. Water was freely provided using a plastic bucket. Refused tree fodder was collected and weighed at the end of the 6h period. Samples of offered and refused feeds (100g) were taken daily in order to determine dry weight (Ben Salem et al., 1994; Kaitho et al., 1996).

The effects of tree foliages supply on intake and palatability were analyzed using the model

Y_ijk= µ + T_i + A_j +D_m+ e_ijk,

where Y_ijk: response variable measured in the i^th treatment (T, i= 1-5), µ: population mean, A_j: animal effect, D_m: intake day effect, and e_ijk: random error. The data were analyzed as a complete random experiment and subjected to analysis of variance using the PROC-ANOVA procedure of SAS (2001), following Steel and Torrie (1980). Differences among foliage species were determined using Duncan’s multiple-range test (Duncan, 1955).

Results and Discussion

Agrosilvopastoral system

Farmers are small householders producing mainly maize and livestock. Their main activity is staple crop agriculture. All economic and cultural activities are centered around the milpa (MBS), which refers to maize (Zea maiz) associated with beans (Phaseolus spp.) and squash (Cucurbita spp.).

The livestock system can be characterized as an agrosilvopastoral system, which requires access to a variety of different ecological areas, including crop and fallow lands, as well as forest areas (pine and oak forest). Its main goal is livestock production for local and regional meat markets, but the meat is also used for self-consumption, especially for traditional festivities. Livestock are also used for draught animals power. Ploughing, harrowing and carting are the main draught animal operations. This system is closely related to crop agriculture, particularly to the milpa, which provides a significant amount of residues which may be utilized for cattle, sheep, horses and mules.

Herds tend to include Zebu (Bos indicus) × Brown Swiss (Bos taurus) races. There is a tendency to increase the herd size by crossing beef cattle with dairy cattle, due to the easy management and rapid growth and fattening of dairy cattle. However, the farmers lack experience in dairy cattle management. Herd management is based upon rotating livestock among crop paddocks, small crop areas, fallows and forest areas. Cattle-feeding is based on natural and naturalized grasses, browsing of trees and shrubs, and a supply of cut grasses and cornstalk residues (Figure 1). Table I shows the main characteristics of livestock-land use and livestock performance indicators. Labor is mostly provided by the family; livestock-system practices are carried out by men, while women and children only participate in crop agriculture.

Annual crops used to be ploughed and harrowed in April, sowed in May (after the rainfall started), and harvested and carted in November and December. Previously, land was left fallow for many years. More recently, due to land scarcity, 95% of farmers leave land fallow for only 3-5 years. This land-use pressure has contributed to an increase in agrochemical use. In January the dry season begins, along with frosts and water and fodder constraints. During this season, farmers supply livestock with maize residues. Meanwhile, until the rainy season begins in June, the livestock depend on artificial water pools. From January to May, livestock feed off of fallow lands, browsing trees and shrubs, and consuming dry maize and residues of maize, beans and squash. In April, when farmers begin to plough and harrow to prepare for maize planting, livestock are taken to fallow lands. In June, livestock are taken to pine (Pinus radiata) and oak (Quercus segoviensis and Q. crispipilis) forests, where the animals (cattle and horses) graze mainly on natural and naturalized grasses. Once the maize has been fully harvested, in winter, livestock are taken back to agricultural fields and small paddocks.

Rotational patterns between agricultural and forest areas, as well as the stocking rate, fluctuate throughout the year. Normally, farmers have a small "estrella" grass paddock near the house, 1 to 2ha of fallow land and 1 to 2ha of maize. Also, they utilize communal forests. Livestock management in forest areas is carried out collectively. Use of fire is restricted to agricultural areas, but nowadays this practice of burning is dying out due to governmental policies.

Animals are kept in good condition by browsing in forest and fallow areas during the dry season. However, local farmers report increasing pressure on forest and fallow lands due to high demand for basic-crop lands, commercial crops (tomato and chilli), fuel-wood, and timber, as well as an increase in number of livestock per family. These factors have led to an imbalance in stocking rates on fallow and forest lands in the studied area, which is estimated to be 2-3.5 animal units per ha (UA·ha^-1). Also, farmers mention a considerable decline during the last 10 years in the abundance of species of Acacia pennatula, A. angusstisima, Erythrina mexicana, Leucaena brachicarpa, Tecoma stans, Lysiloma auritum and Litsea glaucescens, which grow wild in fallow and forest lands and constitute an important resource for fire-wood, timber and fodder.

At present, high land pressure has led to a reduced and fragmented forest. Actually, the main functions of the forest for local families are grazing and timber and firewood extraction. Although use of domestic gas for food preparation has increased during the last several years, fire-wood extraction, associated with cultural-gastronomic patterns is a permanent practice throughout the year. Extraction of forest products has not been managed in a sustainable manner. Consequently, this could be negatively affecting both natural regeneration and soil conservation.

In general, livestock system performance shows low production indices compared to intensive high-input management systems practiced in other areas of Chiapas, as well as in other parts of Mexico (Villafuerte et al., 1999; Nuncio et al., 2001). Nevertheless, cattle activity in the region under study is very similar to other low-input systems in Northern and Highland Chiapas (López et al., 2001). Birth interval is very long and calf mortality is considerable, principally during the dry season. In general, the breeding season is not actively managed by farmers, and births are concentrated during the dry season. Health practices and financial assistance are limited. However, farmers tend to use vaccines and vitamins, as there is a common belief that these products are essential to good health. Weight gains average around 100kg per year, and fattening for market takes 2.5-3 years. This system is slow, but assures gains with low investment of capital and other inputs. Moreover, farmers tend to shorten fattening time. Animals are sold in local and regional markets in southern Mexico through intermediaries ("coyotes"). Commercial animal weights range from 250 to 350kg. "Koct’an", a tradition of distributing livestock meat among community families for local consumption, is practiced in the region. Additionally, there is a market for draught animals, used for carrying and hauling.

The livestock system in this region constitutes a family strategy to overcome agricultural risks and a way of earning money for medical and educational needs. Farmers’ technical knowledge is based mainly on their experience with integral management of livestock and crop agriculture. However, low capital investment, unavailability of technical information and scarcity of training programs have contributed to incompetence in designing new technological options in order to construct a more sustainable system. Increasing population and demand for meat in regional markets have led to changes in production systems; thus, a greater interest in livestock has developed. However, this change may promote a greater pressure on communal forest and fallow areas, deriving on a loss of natural resources and diminished diversity.

Use of fodder trees

Farmers identified 13 native fodder tree species of the botanical families Leguminosae, Anacardiaceae, Fagaceae, Lauraceae, Moraceae, Rutaceae and Bignoniaceae (Table II). Ten uses were identified for all the trees, distributed as follows: eleven for fuel wood; nine for living fences and house construction, nine for shade; seven as medicines; three for tools; two for food and/or tanning and one for condiment, charcoal, or ornamental use. Q. segoviensis was identified as the best species for firewood, as it produces little smoke, is durable, makes good charcoal, and demands a high market price. For building–poles, D. robinoides, R. schiedeana, L. brachycarpa and L. auritum were the most adequate ones due to their hardness and durability.

Most farmers (90%) reported that fodder species were found mainly in fallow areas, except those species of the genus Quercus, which were found either in fallow or other middle-aged and mature forest areas. Tree presence in agricultural crop lands was less frequent in this region than in other areas, such as the Highlands or Northern Chiapas. Leucaena sp. and P. mexicana were the preferred trees for livestock shade and living fences.

Fodder trees and local knowledge

Table III presents farmer perceptions of local fodder trees and shrubs. All species had leaves during the dry season. Nonetheless, farmers especially appreciated the value of A. pennatula, F. cookii, P. mexicana and R. schiedeana for their high amount of foliage during the dry season. These trees, according to farmers, contribute significantly to livestock feeding during this critical period.

Most species were qualified as high or intermediate with respect to palatability, according to farmer perceptions. However, L. brachycarpa and A. pennatula were seen as the best species for fattening livestock. Farmer knowledge of tree propagation showed that most species are easily managed, with the exception of D. robinoides and F. cokii.

For the most part, farmer perceptions were neither correlated with palatability essay, nor with protein content and digestibility percentage (p>0.05). Local knowledge of the relationship between plants and livestock in this region is less complex than in other previously studied areas (Toledo et al., 1995; Thapa et al., 1997; Nahed et al., 1998, Camacho et al., 1999; Rothaert and Franzel, 2001). This could be due to the fact that the population in this area has been mainly devoted to crop agriculture, and has had less experience with livestock, which were introduced to this region ~50 years ago. Furthermore, the local population is mestizo (combination between indigenous and Hispanic) with significant urban influence. These aspects suggest the need for future studies on local knowledge of fodder trees, considering farmer age, gender and previous experience with livestock.

Several studies in indigenous African and Asian communities have shown diverse criteria for rating fodder trees. For instance, Thapa et al. (1997) in Nepal mentioned the concept "Obano", used by farmers to indicate the role of fodder in improving animal production and health. Also, Rothaert and Franzel (2001) in Kenya reported the criteria most commonly used by farmers for classifying fodder trees to be a) satisfying animal hunger, b) improving health, c) palatability, and d) promoting fattening and milk production.

In addition, farmers mentioned that some multipurpose species commonly used for fodder are nowadays disappearing. One example is "cuajilote" (Parmentiera edulis), which used to be found in fallow lands, but today is only found in home gardens.

The majority of forage species mentioned in this study were reported to be used as forage for cattle but inappropriate for horses, donkeys, mules and chickens. Farmers did not mention evidence of foliage toxicity to animals.

Chemical composition and secondary compounds of tree foliage

Table IV shows the results of chemical and nutritional analyses. A wide variation in chemical composition between species was found. On the whole, PC content was higher (8.1-22%) than reported by another study (Norton, 1994). However, most species showed CP concentration, IVDMD and gross energy similar to values found for other fodder trees of Chiapas (Jiménez-Ferrer et al., 2001; López et al., 2001). NDF and ADF contents were moderate and low, respectively. All species had a gross energy content similar to other species in Highland Chiapas (Camacho et al., 1999). Acacia spp., D. robinoides and L. brachycarpa stood out from other species due to their high CP content and intermediate percentages of fiber fractions. These species had characteristics similar to other trees in the northern mountains of Chiapas, such as Erythrina mexicana and Guazuma ulmifolia (Ku-Vera et al., 2001). On the other hand, most species are higher in quality when compared with native and temperate pastures in the region (Nahed et al., 1998; López et al., 2001). According to IVDMD values, the trees may be classified in two groups. The first group with moderate to high digestibility (50-70%) includes species such as C. edulis, D. robinoides and A. angustisima. The second group, with digestibility values <50% includes L. auritum, F. cookii and L. glaucescens. Because most fodder tree species present higher CP contents and greater IVMD than those of maize stubble, some other cereal straws and native pastures of the region, they may be regarded as fodder resources during the dry season.

Results of the secondary compound show that tannic acid was low for most species. Tannic acid levels were lower than reported data from Africa and the Himalayan region (Makkar and Becker, 1998). Alkaloids and cyanogenic glucoside concentrations were moderate and low, similar to those reported by Kaitho et al. (1996) and Odenyo et al. (2003). On the other hand, the presence of alkaloids in species with high PC and moderate IVDMD contents such as A. angustissima, C. edulis, T. stans and D. robinoides could limit their use as fodder in intensive systems. However, low tannin levels have been shown to have beneficial effects in efficiency of microbial protein synthesis and ruminant ecosystems (Makkar et al., 1995).

Palatability of five fodder trees

Table V shows how dry matter intake and palatability changed over the course of the eight-day experiment. Results showed variability day to day, and among species. The most palaTable species were A. angustissima and L. auritum. Intake of A. angustissima increased from the initial days of evaluation (r²= 0.38, p<0.05). On the other hand, A. pennatula was found to have a high intake the first day, decreasing subsequently (r²=0.53, p<0.05). The other species did not show significant trends (Table V). Low palatability in A. pennatula, T. stans and P. mexicana could probably be attributed to the combination of moderate digestibility plus the presence of secondary compounds and physical characteristics of the foliage.

In this study, the increasing intake observed for A. angustissima is remarkable. Percentages of tannic acid and alkaloids were moderate for this species (Table IV). By contrast, in Africa, high toxic levels, low palatability and low digestibility (48g·kg^-1 after 24h) have been reported (Kaitho et al., 1996). Death of sheep which have fed off A. angustissima without having passed through an adaptation period has been also reported (Odenyo et al., 1997). Smith et al. (2001) stated that rats fed with 15-20% of A. angustissima suffered severe weight loss and death. All this data taken together may suggest a constraint in the use of this species as fodder, but the present results show potential use of this specie in Trinitaria, due to its moderate palatability and its abundance. Additionally, farmers in the study area did not report harm to animals fed with this species. Also, other results suggest the existence of specific ruminant micro-organisms able to decompose secondary compounds in local sheep (Odenyo et al., 2003) and physical immunity to toxic factors in local animals (Roothaert and Franzel, 2001). This leads to propose new hypotheses for further investigation.

Conclusion

Associating livestock with agricultural and fallow lands is a strategy that allows small holders to use resources integrally and to diversify products and services to improve their livelihood. Even though this system contributes to sustaining the family economy, there is evidence that the tendency of overgrazing forest and fallow areas causes high pressure on fodder resources and hinders land restoration. Fallow areas play a significant role in feeding animals, especially during the dry season where land and fodder are scarce.

Farmer perceptions of tree species, as well as variables suggested by researchers, contrast with laboratory and experimental results, especially with respect to palatability. This may indicate that farmers pay attention to visible variables, but have limited knowledge of non-visible factors such as nutritive value and secondary compounds. Local knowledge of fodder trees is relatively new in the study area. Livestock is a recent activity, established ~50 years ago, compared to maize cultivation, which is an ancient practice, at least 8000 years old. However, these results have shown that there is potential for using native fodder trees as livestock feed. This could help in the establishment of a sustainable agrosilvopastoral system combining local knowledge and research results.

Acknowledgements

The authors acknowledge the Communities of Espatajamal and San Caralampio for their participation in this study and the Animal Nutrition Department, Instituto Nacional de Ciencias Médicas y Nutrición "Salvador Zubirán", México.

References

1. AOAC (1990) Official Methods of Analysis of the Association of Official Analytical Chemists. 15^th ed. Washington, DC, USA. 30 pp.        [ Links ]

2. Ben Salem H, Nefzaoui A, Abdouli H (1994) Palatability of shrubs and fodder trees measured on sheep and dromedarie s: I. Methodological approach. Animal Feed Sci. Technol. 46: 143-153.        [ Links ]

3. Calvo CC, de Morales JL (Eds.) (1988) Manual de técnicas de Laboratorio para el análisis de alimentos. Publicación L-63. Instituto Nacional de la Nutrición "Salvador Zubirán". México. 258 pp.        [ Links ]

4. Dagang AB, Nair PK (2003) Silvopastoral research and adoption in Central America: recent findings and recommendations for future directions. Agrofor. Syst. 59: 149-155.        [ Links ]

5. Duncan DB (1955) Multiple ranges and multiple F-test. Biometrics, 11: 1-42.        [ Links ]

6. GEA (1993) El proceso de evaluación rural participativa: una propuesta metodológica. Grupo de Estudios Ambientales. México. 103 pp.        [ Links ]

7. Goering HK, Van Soest PJ (1970) Forage Fiber Analyses. Agriculture Research Handbook Nº 379. ARS-USDA. Washington, DC, USA. 20 pp.        [ Links ]

8. Guillén J, Jiménez-Ferrer G, Nahed TJ, Soto-Pinto L (2001) Ganadería Indígena en el norte de Chiapas. In Hernández L. (Ed.) Historia ambiental de la ganadería en México. Instituto de Ecología. Xalapa, Ver., Mexico. 276 pp.        [ Links ]

9. IIED (1994) Rapid Rural Appraisal (RRA). Special issue on livestock. Notes Nº 20. International Institute for Environment and Development. London, UK. 173 pp.        [ Links ]

10. INEGI (2000) Anuario estadístico de Chiapas. Instituto Nacional de Geografía, Estadística e Informática. Gobierno del Estado de Chiapas. México. 269 pp.        [ Links ]

11. Jiménez-Ferrer G (2000) Árboles y arbustos con potencial forrajero en el Norte de Chiapas México. Thesis. Universidad Autónoma de Yucatán. México. 256 pp.        [ Links ]

12. Jiménez-Ferrer G, Ramírez-Avilés L, Kú-Vera JC, Soto-Pinto L, Villanueva-López G (2001) Nutritive value of fodder trees in the mountains of Northern Chiapas, Mexico. Mem. 2^nd Cong. on Agroforestry and Livestock Production in Latin America. CATIE-FAO. Costa Rica. pp. 336-340.        [ Links ]

13. Kaitho RJ, Umunna NN, Nsahlai IV, Tamminga S, Bruchem JV, Hanson J, Van De Wouw M (1996) Palatability of multipurpose tree species: effect of species and length of study on intake and relative palatability by sheep. Agrofor. Syst. 33: 249-261.        [ Links ]

14. Ku-Vera JC, Ramírez L, Jiménez-Ferrer G, Alayón L, Ramírez C ( 2001) Árboles y arbustos para la producción animal en el trópico Mexicano. Mem. 1ª Conf. Electr. "Agroforestería para la Producción Animal en Latinoamérica". FAO. Rome, Italy. pp. 67-79.        [ Links ]

15. López CM, Jiménez-Ferrer G, De Jong B, Ochoa S, Nahed TJ (2001) El sistema ganadero de montaña y especies leñosas forrajeras promisorias en la Región Norte-Tzotzil de Chiapas, México. Rev. Vet. (México) 32: 93-102.        [ Links ]

16. López-Merlín D, Soto-Pinto L, Jiménez-Ferrer G, Hernández-Daumás S (2003) Relaciones alométricas para la predicción de biomasa forrajera y leña de Acacia pennatula y Guazuma ulmifolia en dos comunidades del Norte de Chiapas, México. Interciencia, 28: 334-339.        [ Links ]

17. Makkar HPS, Becker K (1998) Do tannins in leaves of trees and shrubs from African and Himalayan regions differ in level and activity?. Agrofor. Syst. 40: 59-68.        [ Links ]

18. Makkar HPS, Blummel M, Becker K (1995) In vitro effects and interactions of tannins and saponins and fate of tannins of rumen. J. Food Agric. Sci. 69: 481-493.        [ Links ]

19. Mosquera-Losada MR, McAdam J, Rigueiro A (2005) Silvopastorlism and sustainable land management. CABI Publishing. Lugo, Spain. 429 pp.        [ Links ]

20. Nahed J, Villafuerte L, Grande D, Pérez G, Alemán T, Carmona J (1997) Fodder shrub and tree species in the highlands of southern México. Animal Feed Sci. Technol. 68: 213-223.        [ Links ]

21. Nahed TJ, Sánchez A, Grande D, Pérez-Gil F (1998) Evaluation of promissory tree species for sheep feeding in The Highlands of Chiapas, Mexico. Animal Feed Sci. Technol. 73: 59-69.        [ Links ]

22. Norton BW (1994) The nutritive value of tree legumes. In Gutteridge RC, Shelton M (Eds.) Forage Tree Legumes in Tropical Agriculture. CABI. Queensland, Australia. pp. 177-191.        [ Links ]

23. Nuncio-Ochoa G, Nahed J, Díaz H, Escobedo AF, Salvatierra-Izaba B (2001) Caracterización de los sistemas de producción ovina en el estado de Tabasco. Agrociencia, 35: 469-477.        [ Links ]

24. Odenyo AA, Osuji PO, Karanfil O, Adinew K (1997) Microbiological evaluation of Acacia angustissima as a protein supplement for sheep. Animal Feed Sci. Technol. 69: 99-112.        [ Links ]

25. Odenyo AA, Osuji PO, Reed JD, Smith AH, Mackie RI, McSweeney CS, Hanson J (2003) Acacia angustissima: Its anti-nutrients constituents, toxicity and possible mechanisms to alleviate the toxicity - a short review. Agrofor. Syst. 59: 141-147.        [ Links ]

26. Roothaert RL, Franzel S ( 2001) Farmers’ preferences and use of local fodder trees and shrubs in Kenya. Agrofor. Syst. 52: 239-252.        [ Links ]

27. Sánchez MD, Rosales MM (1999) Agroforestería para la producción animal en América Latina. Estudio Producción y Sanidad Animal 143. FAO. Rome, Italy. 515 pp.        [ Links ]

28. SAS (2001) SAS/STAT Program SAS Institute Inc, Cary, NC, USA.         [ Links ]

29. Smith AH, Odenyo AA, Osuji PO, Wallig MA, Kandil FE, Seigler DS, Mackie RI (2001) Evaluation of Toxicity of Acacia angustissima in a rat bioassay. Animal Feed Sci. Technol. 91: 41-57.        [ Links ]

30. Soto-Pinto ML (1990) Plantas útiles de cuatro comunidades de Chiapas: perspectivas en el uso sostenible de la tierra. Rev. Fitotec. Mex. 13: 149-168.        [ Links ]

31. Steel RGD, Torrie JH (1980) Principles and Procedures of Statistics. 2^nd ed. McGraw-Hill. New York, NY, USA. 633 pp.        [ Links ]

32. Steep RS (2002) Mountain ethnobiology and development in highland Chiapas, Mexico. Mountain Res. Devel. 20: 218-219.        [ Links ]

33. Thapa B, Walker DH, Sinclair FL (1997) Indigenous knowledge of the feeding value of tree fodder. Animal Feed Sci. Technol. 67: 97-114.        [ Links ]

34. Thorne PJ, Subba DB, Walker DH, Thapa B, Wood CD, Sinclair FL (1999) The bases of indigenous knowledge of tree fodder quality and its implications for improving the use of tree fodder in developing countries. Animal Feed Sci. Technol. 81: 119-131.        [ Links ]

35. Tilley JMA, Terry RA (1963) A two stage technique for the in vitro digestion of forage crops. J. Br. Soc. Grassland 18: 104-111.        [ Links ]

36. Toledo VM, Batiz A, Becerra R, Martínez E, Ramos HC (1995) La selva útil: Etnobotánica cuantitativa de los grupos indígenas del trópico húmedo de México. Interciencia, 20: 177-187.        [ Links ]

37. Van Soest PJ, Robertson JB, Lewis BA (1991) Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74: 3583-3597.        [ Links ]

38. Villafuerte D, García MC, Meza S (1999) La cuestión ganadera y la deforestación. Universidad de Ciencias y Artes del Estado de Chiapas. Centro de Estudios Superiores de México-Centroamérica. Tuxtla Gutiérrez, Chiapas, Mexico. 215 pp.        [ Links ]