Silvopastoral systems with isolated timber trees within pastures in the Coffee region of Colombia

J C Camargo, A Feijoo, M C Zúñiga, H Cardona and J Gaviria

Facultad de Ciencias ambientales, Universidad Tecnológica de Pereira;
La Julita, A.A. 97, Pereira, Colombia
jupipe@utp.edu.co

Abstract

The abundance of isolated timber trees within pastures was evaluated in La Vieja river watershed, located in the coffee region of Colombia. Nine species were identified, being Nogal (Cordia alliodora) the most abundant. Principal Components and Discriminant Analyses showed that biophysical as well as socioeconomic variables influenced the presence of isolated trees within pastures. 

The most important variables showing effects on abundance of isolated timber trees were altitude, soil properties, weed control and controlled grazing. Three group sites defined by cluster analysis showed that the distribution of timber species changed according to ecological conditions and their abundance was limited by management practices within pastures. Farmers recognized the ecological and economic importance of isolated trees within pastures and they have defined uses for each species found. 

In this way, these species could be used to establish silvopastoral systems and contribute to convert cattle grazing to a more sustainable activity in the coffee region of Colombia

Key words: abundance, benefits, canonical variables, multivariate analysis, silvopastoral systems, use of timber trees

Introduction

In Latin America, more than 90 million ha of land are under pastures, most of them as a result of forest conversion to cattle ranching (FAO 1999). During the last years, timber production from cattle farming systems has increased due to an improvement of market possibilities (Howard 1995). In Central America, livestock farmers recognize the potential of valuable timber trees in their pastures to generate income to safeguard low animal product prices on the national and international markets and their contribution to recover degraded areas (Pomareda 2000). Thus the presence of timber species in pastures is not only an important economic alternative, but it is also a possibility to improve environmental conditions within livestock systems.

Agrosilvopastoral systems provide forage, shade for animals and other benefits. Valuable timber species such as Cordia alliodora, Cedrela odorata, Tabebuia rosea, Erythrina poeppigiana, Pithecellobium saman and Albizia sp. kept at low tree densities (4 to 15 trees ha -1), are common in the tropical lowland pastures (Kampen 1996; Barrios et al 1999; Camargo et al 2000, Souza et al 2000; Camargo et al 2004) and in the highland regions, where intensive dairying is practiced, there are good examples of planted and natural regeneration of Alnus acuminata in pastures (Russo 1990) and Cupressus lusitanica wind breakers (Harvey and Haber 1999). The presence of these trees has shown to be influenced by management of pastures especially with practices such as weed control (Camargo 1999). Also, it has been found that the presence and abundance of trees within pastures changes according to the cattle production system (dairy, meat or dual purpose) (Souza et al 2000).

In the coffee growing region of Colombia, problems concerning the drop of coffee bean prices in the international market lead farmers to change their land use mainly to pastures. In La Vieja river watershed most farmers have established intensive pasture systems, based on nitrogen fertilizer and use of herbicides which require large investments having potentially negative effects on the environment. However, it is possible to reduce the impacts, even increasing productivity by establishing or promoting trees within pastures.

The conversion of coffee plantations to pastures might mean a biodiversity loss, however if trees are retained and protected, it would be possible to consolidate silvopastoral systems with isolated timber trees using native species and this may also be a strategy for biodiversity conservation. Therefore, actions that promote the establishment of silvopastoral systems based on isolated trees within pastures should emphasize on strategies that guarantee the permanence and growth of these trees. Retaining and managing trees in pastures may improve farm productivity and sustainability by generating tree products and providing ecosystemic services (Ibrahim and Schlönvoigt 1999). Within fragmented forest landscapes, farm trees may also represent important habitats and biocorridors for plant and animal species that are important for maintaining local and regional biodiversity (Harvey and Haber 1999).

This paper shows the importance of isolated timber trees within pastures, determining biophysical and socioeconomic factors influencing their population. In addition, promising timber species that could be used to establish silvopastoral systems in the coffee region of Colombia are identified.

Materials and methods

Study area

The study area corresponds to the coffee region of Colombia and sites sampled are located within the municipality of Alcala, in an area of approximately 117.858 ha in La Vieja river basin. The altitude varies between 950 and 1600 m, the mean annual temperature is between 20°C and 27°C and the total average precipitation is between 950 and 2.500 mm per year. In the sample sites, most of the soils were classified as Andisols derived from volcanic ash, and since they are located near a river they show significant influences of alluvial sediments (Cardona and Gaviria 2003).

Identification of trees within pastures

Aerial photos that correspond to the study area were scanned and integrated to the geographic information system (GIS) through the software ArcView GIS 3.3. Then a systematic two stages sampling was conduced. In the first stage, primary units were defined by using a grid of 1Km*1 Km on aerial photographs. Within each primary unit, areas covered by pastures were defined and identified with GIS as pasture polygons. In the second stage, the central point of each pasture polygon was used as the reference to randomly select sixty potential sites for sampling. Afterwards, field work was carried out to identify the sites selected on aerial photographs and to become acquainted with the owners of farms where sites were located. Finally, thirty sites were selected for sampling. In these sites farmers accepted to participate and provided information about pasture management, pastures with cattle grazing and pastures that contained trees. The number of sites selected for sampling was based on available funds for this study.

Within each site a circular and temporary plot was marked in a nested sampling design and timber species were recorded in four categories according to different growth stages (Table 1). Thus, it was also possible to evaluate their natural regeneration. The importance value index (IVI), which is the sum of relative abundance, frequency and dominance (Lamprecht 1989), was calculated in order to determine the relevance of timber species.

Additional information collected from plots

In each plot, information about physical-chemical soils properties (0-20 cm), physiographic and vegetation composition was registered, including texture, apparent density (from soil core), compression or resistance to penetration (soil compaction) and effective soil depth. The soil chemical properties included pH, extractable acidity, P, Ca, Mg, K, acid saturation, base saturation, Ca/Mg, Mg/K, Ca/K, (Ca+Mg)/K, effective cationic exchange capacity (ECEC), cationic exchange capacity (CEC), organic carbon and total Nitrogen. The physiographic variables included altitude, slope, topographical position and climate conditions. Simultaneously, farmers were interviewed to evaluate socio-economic aspects such as land use dynamics and management practices such as weed control, fertilization, controlled grazing and use of nutritional supplements.

Factors that determined the presence of trees within pastures

Effect of site variables on abundance of isolated trees within pastures was evaluated first by correlation analyses trying to detect relationships of dependence between variables, selecting those with higher Pearson coefficient (>0.5). Then, to reduce the number of variables, most relevant variable were identified by Principal Components Analysis and these variables were used in a Cluster Analysis to identify sites with similar characteristics. For the separation of variables into clusters, a Discriminant Canonical Analysis was used. SPSS Release 8.0 for Windows was used to perform these analyses.

Results and discussion

General characteristics of the sites included in the sampling

Most the farms (70%) where sample sites were located presented areas with less than 50 ha (40 ha on average). More than 50% of the farms were totally covered by pastures and only in 12% of the total farms, pastures were of lesser proportion compared with other land-uses or covers. However, in the interview with farmers, cattle ranching represented the main economical activity. Intensive cattle management evaluated according to uses of nitrogen fertilizers and herbicides was reported by farmers in 40% of the sampled sites and in regards to nutritional supplements, 90% of farmers used some food source different to pastures. About 50% of them, used green fodder with species such as Gliricia sepium, Trichanthera gigantea and Leucaena leucocephala and 40% used concentrates. Regarding pasture management, a high number of sites have been fertilized (73%) and grazing was usually controlled in the evaluated sites (63%).

The mean, maximum and minimum values and standard deviations of the main biophysical variables evaluated within sites sampled are given in table 2.

Timber species found within pastures

Nine isolated timber tree species within pastures were identified, Nogal (Cordia alliodora) being the most abundant and dominant. Other species found were Laurel (Cinnamomum triplinerve), Písamo (Erythrina poeppigiana), Guacamayo (Croton draco), Doncel (Zanthoxylum rhoifolium), Dinde (Chlorophlora tinctoria), Samán (Samanea saman), Cedro (Cedrela odorata) and Mestizo (Guarea trychilioides). Values for number of individuals per ha according to the growth stage are given in table 3.

 According to IVI values (120), Cordia alliodora was the most important species compared to other species identified. The abundance of this species might in part be explained by the promotional use of this tree in reforestation programs given by government institutions and that Cordia alliodora has been usually intercropped within coffee plantations. During the coffee crisis, farmers replaced coffee plantations with pastures and some mature Cordia alliodora trees remained. Cinnamomum triplinerve and Erythrina poeppigiana also had important IVI values (72 and 65, respectively) and both are important species usually found in remaining or secondary forests of the coffee growing region, located mainly as riparian forest (Table 4).

Cordia alliodora results important due to its abundance and possibilities of adaptation to ecological conditions of pastures, even tolerating land use changes, appearing as a promising species in this context. It has been also shown by other studies conducted within farms in low lands of Costa Rica, where 25 timber species were found with natural regeneration in pastures, being Cordia alliodora the most common (Camargo 1999). Souza et al (1999) reported a density of 11 trees per ha of Cordia alliodora within pastures in Costa Rica, also being the most abundant tree species encountered.

 The evaluation of total number of individuals in different growth stages or natural regeneration revealed three patterns. The first pattern consisted of species having individuals in at least three growth stages and a total density of 16 to 90 individuals per ha. Cordia alliodora was the most important and the only species with individuals in all growth stages. Other species with this pattern were Chlorophlora tinctoria, Croton draco and Samanea saman, which did not show young individuals (Figure 1).

There were two species in the second pattern, Cinnamomum triplinerve and Erythrina poeppigiana. These two presented a large number of seedlings per ha (191 and 93, respectively) but few individuals in other growth stages (5 and 6 mature trees per ha) and young trees were not found (Figure 2). The third pattern consisted of three species with a low density of individuals and it was doubtful if these had a natural regeneration, because for two species (Cedrela odorata and Guarea trychilioides) only mature trees were found and the density was very low (1 tree per ha). Another species was Zanthoxylum rhoifolium also showing few seedlings and mature trees and its young trees did not reach a density of 1 individual per ha on average (Figure 3).

Cordia alliodora was the only species found with individuals in all growth stages showing a significant potential for silvopastoral systems. According to its natural regeneration pattern, it is likely that most of the mature trees were planted, thus explaining the higher number of mature trees. Other species can not guarantee an entirely natural regeneration, due to the absence of young trees (Camargo 1999; Ibrahim and Camargo 2001). Natural regeneration also depends on weed control and the surviving trees that are the result of management and growing conditions over time (Concklin 1987). Unlike young trees, seedlings are less defoliated by cattle. However, they may be severely damaged by cattle treading (Hart and Norton 1988).Young trees are more stable and have fewer problems since they have survived ecological and human obstacles. However, sometimes they are defoliated by cattle and cut by farmers. Young trees are very important for ensuring natural regeneration, because if they can survive they will become mature trees (Archer 1995) and this shall serve to establish silvopastoral systems.

The use of natural regeneration might be a good method to establish silvopastoral systems. Nevertheless there are different critical factors such as low availability of propagation sources (lack of seedbed trees); poor efficacy of natural dispersion by cattle due to low proportion of feasible seeds and concentration of them in just the right places for animals to depredate on; high mortality of plants due to germination in periods with low humidity where seeds have high vulnerability (Janzen 1977); soil degradation; grass competence (Nepstad et al 1991); and damage caused by cattle (Hatheway and Baker 1970). It could be feasible to reduce the effect of critical factors through adequate pasture management. This implies knowledge about ecology of timber species within pastures, especially during early stages of growth and also about animal behavior. Thus conditions for establishing trees in pastures may be improved even in pastures under grazing.

On average 489 trees per ha of different species were found considering all growth stages (Figure 4), indicating the potential to develop silvopastoral systems based on management of natural regeneration. However the general tendency was the absence of young trees, an important factor because without young trees, their population can not increase and if farmers decide to harvest the timber, they might definitively eliminate a tree from the pastures.

Effect of biophysical and socioeconomic variables on timber species population

Altitude, soil compaction and apparent density were the variables that showed the highest correlation (P<0.05) with the abundance of trees within pastures, whereas among socioeconomic variables weed control and controlled grazing were the most important. The population of Cordia alliodora was significantly larger (P<0.05) on slightly acidic soils rich in organic matter and also showed a positive correlation with un-compacted soils and with low apparent density. Population of Erythrina poeppigiana, Samanea saman and Zanthoxylum rhoifolium, presented a better response to fertile soils but even under limited physical properties, whereas weed control had a negative effect. Chlorophlora tinctoria, Croton draco and Cinnamomum triplinerve showed significant (P<0.05) and positive correlations with fertilization of pastures, but negative with soil compaction and high apparent density.

Five principal components were needed to explain 80% of the variation between sites in all of the data set. The first two explained 57% and were defined by biophysical variables as altitude and soil chemical and physical properties. Third, fourth and fifth components basically showed higher influence of socio-economic variables of which weed control and controlled grazing had the greatest relevance (Table 5).

Trying to elucidate a relationship between the abundance of trees within pastures and other variables, values of coefficients within each component were used to identify tendencies. Cordia alliodora was positively influenced by biophysical variables such as altitude and physical soil properties which were important variables within Component 1. Erythrina poeppigiana and Samanea saman were more abundant in lowlands (low altitude) and on soils rich in bases (high base saturation) variables that showed higher coefficient in the Component 2. Other species were associated with Components 3, 4 and 5 where variables such as weed control and controlled grazing, that represent the pasture management, were the most relevant.

Definition of sites with similar tree abundance pattern

Three groups were defined by cluster analysis involving 109 variables for each of the 30 sites sampled. The first was conformed by 6 sites, the second by 9 and the third by 15. Using the Wills Lambda test (Bortz 1993) significant differences (P<0.0001) were found between them. Two canonical variables were obtained from a Discriminant Analysis which represented the total variability and both reported a high correlation. The canonical variable 1 (Function 1) was represented specially by biophysical variables, and explained 95% of the total variation, of which altitude and chemical soil properties were the most relevant. The canonical variable 2 (Function 2), showed higher influence in its structure by variables associated with management of pasture and explained 5% of the total variability (Table 6).

The relation between cluster and canonical variables showed that the first group corresponded to sites located at medium altitude, moderate to steep slopes and within big farms with extensive cattle husbandry with the incipient presence of trees in pastures (Figure 5). The second group involved sites located in the low lands, on small farms, slightly undulating slopes, fertile soils with neutral pH and high CEC but physically limited by high apparent density and compaction. In these sites Erythrina poeppigiana and Samanea saman had the greatest average of total trees (Figure 5).The third cluster involved sites located at high altitude, with more intensive cattle husbandry under controlled grazing and use of nutritional supplements; soils were slightly acidic with excellent physical properties and there, Cordia alliodora was the most abundant species (Figure 5). The association between a more intensive cattle husbandry and high abundance of trees might be explained because the farmers want to retain trees within pastures trying to reduce risk by diversifying farm production. A similar conclusion was obtained in a study in Costa Rica (Souza et al 1999).

The compatibility between trees and cattle depends on relations between practices of management conduced in the pasture and biotic elements involved in the cattle systems (Vegetation and animals) under specific ecological conditions (soils, climate, topography, altitude). Cattle grazing may cause damage on trees due to defoliation, treading and soil compaction. Cattle behavioral patterns have an effect on the composition of species and on the competition between them, due to the selection of food species and treading patterns changes on the space, therefore grazing is heterogeneous and distributed according to spatial diversity of the vegetation, topography, distance to watering or supplementation places and shade (Hart et al 1993).

Figure 5. Depict of discriminator functions according to cluster or groups of sites. Groups
of centroids are the mean discriminant scores of each of the dependent variable categories
for each of the discriminant functions. La Vieja river watershed, Valle del Cauca, Colombia.

Use of timber trees in the La Vieja river watershed

Farmers at the moment are not giving an intensive commercial use to timber trees and in most cases timber has a domestic destination. According to their importance, trees that are mostly used are: Cordia alliodora, Cinnamomum triplinerve, Croton draco, Chlorophlora tinctoria, Cedrela odorata and Guarea trychilioides. Timber is especially used for building houses, stables, fences and structure for drying coffee (Table 7). Although species like Erythrina poeppigiana and Samanea saman are not for commercial use, farmers recognize their ecological benefits in the protection and in the improvement of soils.

During this study farmers did not recognize the importance of trees to generate incomes; however they can avoid investments and reduce costs in different agricultural alternatives. They had also an interest in learning about silviculture of trees within pastures to obtain incomes from timber selling and some of them knew about timber markets and current possibilities in the coffee region.

Timber from trees within pastures may have an improved price due to the fact that the offer from natural forest is decreasing. Also factors such as acceptance of timber from pastures due to its quality, availability of low cost technologies to establish trees, and abundance of land resources could contribute to conversion and modernization of cattle systems (Barrios 1998).

Conclusions

• In the La Vieja watershed there are timber species which have remained in the pastures as isolated trees and that may be used to consolidate, design and establish silvopastoral systems. Identical or similar recommendations might be valid for others areas within the coffee growing region of Colombia.

• Cordia alliodora showed a high potential as a promising species for the development of silvopastoral systems due to its abundance, natural regeneration, adaptability to ecological conditions within pastures and also because its timber has an important commercial value.

• The presence or absence of species in the study area is largely determined by differences between sites and species requirements according to their growth stage. Therefore strategies to promote silvopastoral systems should be adjusted depending on the species, the site and the objective.

• The abundance and presence of isolated trees within pastures in the La Vieja watershed depend on interactions between ecological and socioeconomic factors. Their performance during the establishment shows a high influence of biophysical variables whereas their stability and development depended on practices of management within the pastures.

• Even though most trees had a domestic use at the moment of the study and the incomes received from timber selling were insignificant in the area, farmers have always given an important value and recognize the ecological benefits of some species.

• Multivariate methods resulted helpful to analyze the simultaneous influence of biophysical and socioeconomics variables on the presence of isolated timber trees within pastures. However, within new variables generated by multivariate analyses (principal components and canonical functions), always a few group of variables showed to be those with greater influence to define these new variables. Therefore, the nature of each variable included within an analysis and its individual effects upon the response variable should always be considered.

Acknowledgments

This research was funded by COLCIENCIAS, Convenio UTP-GTZ and the Universidad Tecnológica de Pereira, Colombia. Also, we thank farmers from La Vieja river watershed who provided us information and permitted us to work within their farms. Sampling and collecting information were carried out by Harold Cardona and Jhonny Gaviria, who additionally provided us information from their thesis research.

References

Archer S 1995 Herbivore mediation of grass-woody plants interactions. Harry Stobbs Memorial Lecture. Tropical Grasslands 29: 218-235.

 

Barrios C 1998 Pastoreo regulado y bostas del ganado como herramientas forestales para la protección de arbolitos en potreros. Tesis M.Sc. CATIE, Turrialba, Costa Rica. 93p.

 

Barrios C, Beer J y Ibrahim M 1999 Pastoreo regulado y bostas del ganado para la protección de plántulas de Pithecellobium saman en potreros. Agroforestería en las Américas, 6:63-68.

 

Bortz J 1993 Statistik für sozialwissenschftler. 4 Auflage. Springer-Lehrbuch. Berlin, Germany. 753p

 

Camargo J C 1999 Factores ecológicos y socioeconómicos que influyen en la regeneración natural de Cordia alliodora en sistemas silvopastoriles del trópico húmedo y subhúmedo de Costa Rica. Tesis M.Sc. CATIE. Turrialba, Costa Rica. 127p.

 

Camargo C,  Ibrahim M,  Somarriba E, Finegan B and Current D 2000 Factores ecológicos y socioeconómicos que influyen en la regeneración natural de laurel (Cordia alliodora) en sistemas silvopastoriles del trópico húmedo y sub-húmedo de Costa Rica. Agroforestería en las Américas, 7:, 46 – 52.

 

Camargo J C, Feijoo A, Zuñiga M C, Gaviria J and Cardona H 2004 Management of natural regeneration: A feasible strategy to establish silvopastoral systems in the Coffee region of Colombia. In: The Importance of Silvopastoral Systems in Rural Livelihoods Provide Ecosystems Services. Proceedings of the 2nd. International Symposium on Silvopastoral Systems, Merida, Yucatan, México. pp. 126 - 131.

 

Cardona H y Gaviria J 2003. Estrategias de manejo para promover la regeneración natural de árboles maderables en pastizales de la cuenca del río la Vieja. Trabajo de grado, facultad de Ciencias Ambientales. Universidad Tecnológica de Pereira. 99p.

 

Concklin N 1987 The potential nutritional value to cattle of some tropical browse species for Guanacaste, Costa Rica. Thesis Ph.D. Cornell University. 124p.

 

FAO Food and Agriculture Organisation of the United Nations 1999 Production yearbook, FAO, Rome.

 

Hart R and Norton B 1988 Grazing management and vegetation response. In: Handbook of vegetation science. Vegetation science application for rangelands analysis and management. In: Paul Tueller (ed). Kluwer Academic Publishers. Dordrecht, Netherlands. pp 493-525.

 

Hart R, Bissio H, Samuel M and Waggoner J 1993 Grazing systems, pasture size and cattle grazing behavior, distribution and gains. Journal for Range Management, 46 (1).

 

Harvey C and Haber W 1999 Remnant trees and the conservation of biodiversity in Costa Rican pastures. Agroforestry Systems, 44: 37 - 68.

 

Hatheway W and Baker H 1970 Reproductive strategies in Pithecellobium and Enterolobium. Further information. Evolution 24: 253-254.

 

Howard A 1995 Price trends for stumpage and selected agricultural products in Costa Rica. Forest Ecology and Management 26: 101-110.

 

Ibrahim M and Schlönvoigt A 1999 Silvopastoral systems for degraded lands in the humid tropics. Environmentally friendly silvopastoral alternatives for optimizing productivity of livestock farms: CATIE's experience. In: Actas IV Semana Científica del CATIE. CATIE, Turrialba, Costa Rica. 6-9 de abril, 1999. pp. 277-282. Semana Científica (CATIE).

 

Ibrahim M y Camargo J 2001 ¿Cómo aumentar la regeneración de árboles maderables en potreros?. Agroforestería en las Américas, 8: 32, 35 - 41.

 

Janzen D 1977 Intensity of predation on Pithecellobium saman (leguminoseae) seed by Merobruchus colombinus and Stator limbatus (Bruchidae) in Costa Rican deciduos forest. Tropical Ecology 18: 162-176.

 

Kampen P 1996 Trees in grassland. The influence of the trees on grass production within silvopastoral systems of the Atlantic zone of Costa Rica. Proyecto REPOSA (CATIE, MAG Wageningen University). Report No. 104. Field report No. 149. 33p

 

Lamprecht H 1989. Silviculture in the tropics: tropical forest ecosystems and their tree species; possibilities and methods for their long-term utilization. GTZ, Eschborn, Germany. 296p.

 

Nepstad D, Uhl C and Serrao E 1991 Recuperation of degraded amazonian landscape: forest recovery and agricultural restoration. Ambio 6:248-255. 

 

Pomareda C 2000 Perspectivas en los mercados y oportunidades para la inversión en ganadería. In: Pomareda C and Steinfeld H (editores) Intensificación de la Ganadería en Centroamérica: Beneficios Económicos y Ambientales. pp 55-72. CATIE-FAO-SIDE, Costa Rica. http://www.fao.org/WAIRDOCS/LEAD/x6366s/x6366s04.htm  

 

Russo R O 1990 Evaluating Alnus acuminata as a component in agroforestry systems. Agroforestry Systems 10: 241-252

 

Souza de Abreu M, Ibrahim M y Silva de Sales J 1999 Árboles en pastizales y su influencia en la producción de pasto y leche. In: Memorias de VI Seminario internacional sobre sistemas agropecuarios sostenibles, CIPAV, Cali, Colombia. 

 

Souza de Abreu M, Ibrahim M, Harvey C y Jiménez F 2000 Caracterización del componente arbóreo en los sistemas ganaderos de La Fortuna de San Carlos, Costa Rica. Agroforestería en las Américas, 7 : 53 - 56.