Two methods for the establishment of a silvopastoral system in degraded pasture land

Margarida M Carvalho, Carlos Renato T Castro*, Luiz Carlos T Yamaguchi, Maurílio José Alvim*, Vicente P Freitas** and Deise Ferreira Xavier*

Embrapa Gado de Leite, Rua Eugênio do Nascimento, 610 - 36038-330, Juiz de Fora, MG. CNPq Fellow mmcarval@cnpgl.embrapa.br
*EMATER-Minas Gerais
**Embrapa Gado de Leite. Campo Experimental de Coronel Pacheco, Rodovia MG 133,
km 42. 36155-000, Coronel Pacheco, MG

Abstract

 

Two methods for the establishment of a silvopastoral system in a degraded pasture, located in south-eastern Brazil, were compared.

 

The silvopastoral system had the following characteristics: trees planted in 10 m wide strips, with four rows, across the slope and separated by 30 m wide strips without trees. The herbaceous forages were: Brachiaria decumbens and Stylosanthes guianensis var. vulgaris cv. Mineirão. Treatments were: Method 1) Simultaneous planting of trees and herbaceous forages; tree strip protection with wire fences, and grazing after the herbaceous forages establishment; Method 2) Planting of tree seedlings and of pigeon peas (Cajanus cajan) in the first rainy season; sowing of the herbaceous forages plus maize in the second rainy season; grazing after maize harvest and pasture establishment. In addition, tree species were planted in two different combinations: Combination A) Eucalyptus grandis and Acacia mangium; Combination B) Eucalyptus grandis, Acacia mangium, Mimosa artemisiana, Acacia angustissima and a hybrid Leucaena leucocephala x Leucaena diversifolia.

 

Tree growth was not affected by the methods of establishment, however, it differed significantly between the two combinations of species, the average tree height of Combination A being higher than that of Combination B. The silvopastoral system was established in 16 and 22 months for Methods 1 and 2, respectively, and the total budget necessary for the establishment was lower for Method 2. However, the final budget (discounting incomes with live weight gains and maize yield, for Methods 1 and 2 respectively) differed little between methods of establishment.

 

Keywords: Brachiaria decumbens, legume trees, pasture degradation, rehabilitation, 

 

Introduction

 

In the mountainous area of south-eastern Brazil, two major constraints on the development of the degraded pasture land are: i) the main soils are very acid and infertile; ii) soil mechanization, for pasture sowing, is difficult and may enhance erosion, with loss of the surface soil, which is the most fertile.

 

The low fertility of the main soils, particularly of the red-yellow latosol (oxisol), and the low carrying capacity of the naturalized molasses-grass (Melinis minutiflora) has led to its substitution by the more aggressive and higher yielding species of the genus Brachiaria. However, to be successful, the establishment of the Brachiaria species depends on the application of phosphorus fertilizers and on the use of a soil preparation method, not conducive to soil erosion (Carvalho 1998). In addition, the long term productivity of the sown grass pastures in these areas depends on the availability of a nitrogen source.

 

The silvo-pastoral system appears to be a viable option to develop the degraded pasture land which predominate in the region, due to its potential for soil fertility improvement and land stabilization, among other advantages to animal production systems (Carvalho 2001).

 

A great deal of information is still necessary for the utilization of silvo-pastoral systems in the region, among them, how to establish these systems efficiently. The protection of young trees against damage by grazing animals is a topic which deserves further studies. For the conditions of southern Brazil some information is available regarding the protection of tree seedlings introduced into existing grass pastures (Ribaski and Montoya 2001). When pasture and trees are to be established simultaneously, there is the possibility of delaying the herbaceous sowing, using annual crops while the trees grow beyond animal reach (Botero and Russo 1999).

 

Two methods for establishing a silvo-pastoral system designed to prevent erosion, improve soil fertility and rehabilitate degraded pasture land were compared in this study, emphasizing the time required for the establishment and the budget necessary for the establishment phase.

 

Materials and methods

 

The experiment was undertaken at the Coronel Pacheco Field Station, which belongs to Embrapa Dairy Cattle Research Center, and is located in the south-eastern region of Minas Gerais State, Brazil, at 21° 33' 22'' of latitude south and 43° 06' 15'' of longitude west. The climate of the region is of the type Cwa, characterized by rainy summers and dry winters, the average annual rainfall is 1,500 mm, with only about 10% of this occurring during the winter period, and the average annual temperature is 19.5 °C.

 

The site was a degraded molasses-grass (Melinis minutiflora) pasture, with declivity of 30%, and the soil a red-yellow latosol (oxisol), with the following chemical characteristics before sowing: pH in water, 4.61; available P (Mehlich) 4.47 mg/dm³; exchangeable bases (cmol_c/dm³): K, 0.11; Ca, 0.33; Mg, 0.12; Al, 0.79.

 

Characteristics of the proposed silvo-pastoral system were as follows: trees were planted in 10 m wide strips, with four rows, prepared across the slope and separated by 30 m wide strips without trees; the herbaceous forages are sown in the whole area. The treatments compared were two methods of establishment: Method 1) Simultaneous planting of trees and herbaceous forages; tree strip protection with wire fences, and grazing after the herbaceous forages establishment; Method 2) Planting of tree seedlings and of pigeon peas (Cajanus cajan) in the first rainy season; sowing of the herbaceous forages plus maize in the second rainy season; grazing after maize harvest and pasture establishment. In addition, tree species in the strips were planted in two different combinations: Combination A) Eucalyptus grandis and Acacia mangium, spaced 3 m x 4 m, total of 100 trees/ha of each species. Combination B) Eucalyptus grandis, Acacia mangium, spaced 3 m x 6 m, total of 68 trees/ha of each species; Mimosa artemisiana, Acacia angustissima and a hybrid Leucaena leucocephala x Leucaena diversifolia, which were planted alternatively between Eucalyptus grandis and Acacia mangium trees in a row, with a total of 42 trees/ha of each species, and a final spacing for tree combination "B" of 3 m x 3 m (Figure 1). Mimosa artemisiana is a fast growing native legume tree species. The main objective of the present study was to compare the two methods of establishment.

 

The four treatments (2 methods of establishment x 2 tree combinations) were allocated in 5,000 m² plots, with two replications. The experiment was planted from November to December 1997. In Method 2 of establishment, the tree seedlings were planted in 1997, and pigeon peas was sown in the 30 m wide strips, aiming at improving soil fertility. In November and December of 1998, maize plus the herbaceous forages were sown after soil incorporation of pigeon pea.

 

The herbaceous forage species used were: Brachiaria decumbens and Stylosanthes guianensis var. vulgaris cv. Mineirão. The latter was included in the study due to its outstanding adaptation to acid soil conditions, and capacity to remain green in the dry season. In the 30 m wide strip soil preparation for sowing of these species was done in strips, across the slope, according to Cruz Filho et al (1986). Liming and fertilizers used were based on previous research results (Carvalho 1997), using the following sources and rates: dolomitic limestone 1,000 kg/ha; Araxá rock phosphate 600 kg/ha; single superphosphate 250 kg/ha; KCl 100 kg/ha and fritted trace elements (FTE Br-16) 30 kg/ha. Liming and Araxá rock phosphate were applied prior to soil preparation, and the other fertilizers were surface applied prior to sowing of the herbaceous forage (Method 1) and pigeon peas (Method 2). These species were oversown separately. Seed density was 10.0, 3.5 and 0.5 kg/ha for B. decumbens, Cajanus cajan and Stylosanthes guianensis, respectively. Seeds of Stylosanthes guianensis and of the tree species were inoculated with appropriate rhizobia strains.

 

Tree seedlings, which were produced in a nursery, were planted in pits and fertilized with (g/pit): a) eucalypt: dolomitic limestone 50; ammonium sulfate 75; single superphosphate 225 and KCl 15. b) legumes: dolomitic limestone 50; Araxá rock phosphate 80; single superphosphate 100; KCl 25 and FTE (Br-16) 10. In January 1998 the tree seedlings were fertilized again with (g/plant): a) eucalypt: ammonium sulfate 30; single superphosphate 50 and KCl 15. b) legumes: single superphosphate 30 and KCl 10.

 

In October 98, a soil sampling was performed in the Method 2 plots, prior to soil incorporation of pigeon peas, to assist in establishing rates of lime and fertilizers application for maize cultivation. Some of the chemical characteristics were: pH in water, 4.93; exchangeable bases (cmol_c/dm³): K, 0.14; Ca, 1.11; Mg, 0.44; Al, 0.50 and organic matter, 2.94%. Maize was sown in 10 x 15 m plots, allocated in the 30 m wide strip, with eight replications, in furrows spaced by 1 meter and seed density of 20 kg/ha. Sources and rates of lime and fertilizers used before sowing were: dolomitic limestone 2,000 kg/ha; and 150 kg/ha of the commercial mixture 08-28-16 + 0.5% Zn, corresponding to rates of NPK + Zn, which was applied in furrows. For the herbaceous forages, which were sown in the whole 30 m strips including the maize plots, fertilizers sources and rates applied were similar to the ones used in Method 1, except for lime and Araxá rock phosphate which were applied in the first year.

 

For the evaluation of the establishment methods, the following parameters were considered: tree height, time required to complete the establishment of the silvo-pastoral system and budget necessary for the establishment phase. Tree height was measured every four months, in four rows per plot. The mean of a species in each row per plot was used as replication in the analysis of variance. The experimental design was in randomized blocks, allocated in a factorial (2 methods x 2 species combination), with two replications, and the species nested on Methods and Species Combination, with four replications. Prices of all items (seeds, tree seedlings, fertilizers, fences) and of labor used, were recorded to calculate the budget. The calculated income relative to live weight gains (in Method 1) and maize yield (in Method 2), were used to discount from the total budget necessary to each treatment. Live weight gains were estimated, based on the stocking rates observed in the current experiment and in animal gains obtained in a previous experiment carried out in a nearby area, in a Brachiaria decumbens pasture (Cóser et al 1995).

 

Results and Discussion

 

Tree growth and pasture establishment

 

Tree growth was not significantly affected (P>0.05) by the methods of establishment in any period of height measurement. However, tree height was always affected (P<0.05) by the combination of species, the average height of the two species (Combination A) being higher than that of the five species (Combination B). Significant differences among species were observed in tree combination B (Table 1). The species Mimosa artemisiana and the Leucaena hybrid reached lower height than the other three species. A separated analysis of variance, performed with data of Eucalyptus grandis and Acacia mangium, indicated that the average height of the two species did not differ significantly, but were higher (P<0.05) in Combination A than in Combination B in all periods of evaluation.

 

In Method 1, the establishment of the Brachiaria decumbens x Stylosanthes guianensis pasture was completed three months after sowing. In this treatment, grazing started in April 98, when trees were too small to support animal damage, thus requiring protection.

 

By October 98, 10 months after planting of the tree seedlings, the average height across the five species and treatments reached 1.94 m, and 14 months after planting (February/99), all species reached height over 2 m (Table 1). Thus, despite the high acidity and low soil fertility conditions prevailing, the fast growing legume trees used in the experiment, particularly Acacia mangium and Acacia angustissima, had a very good performance, as observed in a previous study (Carvalho et al 1999). In the present experiment, both the tree seedlings and the pasture establishment were favored by the favorable rainfall conditions which occurred from December 97, immediately after planting, throughout the first summer period, when total rainfall reached 971 mm in five months. According to information for eucalyptus (Silva et al 1996), the minimum height of trees by the time of the first grazing, to avoid significant damage to the trees is 2.0 m. Therefore, the wire fences could be removed by March-April/99 and grazing extended to the whole area.

In Method 2, maize was harvested in late April 99 and pasture establishment was delayed relative to Method 1. Grazing in this treatment could start by October 99, when tree height (Table 1) was sufficient to avoid serious damage by grazing animals.

 

Budget for the establishment of the silvo-pastoral system

 

The budget for the establishment of the trees and for the pasture are described in Tables 2 and 3.

 

The budget for the establishment of the trees, which comprised planting the trees and their maintenance, and was common to the two methods of establishment, amounted to US$ 59.25 (Table 2). The budget for the pasture establishment varied between methods of establishment.

In Method 1, the budget for pasture establishment amounted to US$ 229 and was lower than the budget necessary for Method 2 (Table 3), which included the sowing of pigeon peas and amounted to US$ 309.

The total budget for the establishment of the silvopastoral system amounted to US$ 490.53 for Method 1 and US$ 368.69 for Method 2 (Table 4), as in the former there was the additional expenditure of US$ 201.99 used with preparation of wire fences to protect the tree strips.

It was considered that, in one hectare (100 x 100 m), the total length of wire fences to protect two strips of 100 m long each was 400 m. The final budget required for the establishment under the two methods may be reduced by discounting the incomes with products to be obtained during the establishment phase, namely liveweight gains in Method 1 and maize yield in Method 2, so that these budgets may decrease respectively to US$ 262.19 and US$ 217.08 (Table 4). The estimated live weight gains obtained during a dry season and two rainy seasons periods was 237 kg/ha (Table 5).

This figure was estimated by converting the number of cows which grazed the experimental plots (paddocks) into steers (Table 5), and then multiplying the number of grazing days by the average live weight gains obtained in a grazing experiment with Brachiaria decumbens which was undertaken in a nearby area with similar soil and relief characteristics (Cóser et al 1995). Total live weight gains indicated in Table 5 were multiplied by US$ 0.96, the market price of a kilogram of standing live weight to obtain the income of US$ 228.34 used to calculate the final budget for Method 1 (Table 4).

 

The calculated income with the expected maize commercialization amounted to US$ 151.61 (Table 4), given the yield of 4,000 kg obtained in the pasture strips, which means 80% of one hectare, after discounting the budget necessary for cultivation and harvesting (Table 6).

Thus, the final budget differed little between the two methods of establishment, as the estimated value for Method 1 was only US$ 45.11 higher than the calculated value for Method 2. In practice, the final budget for the establishment of the silvopastoral system under the two methods will vary, depending on factors such as variations in market prices of the two initial incomes (maize and standing liveweight gains), soil fertility conditions of the site and possibility of using electric fences for protecting tree strips in Method 1. In more fertile soils, farmers will be able to grow a maize (or other) crop in the first year too, decreasing further the final budget necessary for the silvo-pastoral system establishment. The electric fences have been used as a means of introducing trees into existing pastures, without interrupting the normal grazing management system of the farm (Baggio and Silva 1998; Febles et al 2001).

 

Time required for the establishment

 

Under the conditions of the present experiment, the time required for the establishment of the silvo-pastoral system when tree strips were protected with wire fences (Method 1) was 16 months, lower than that necessary for Method 2, which required 22 months. These rather short periods were made possible due to two main factors: i) the use of fast growing, adapted tree species; ii) the climatic conditions, which were favorable after tree seedlings planting in the field and throughout the first rainy season. In addition, the use of seed inoculation with appropriate rhizobia, and the application of adequate levels of fertilizers at sowing, also contributed to enhance the initial growth of the tree seedlings.

 

The longer time required for the establishment of the silvo-pastoral system under Method 2 is in part compensated by the total budget, which in the present study was US$ 121.84 lower than the one required in Method 1, as calculated from values in Table 4. Thus, the decision on what method of establishment to adopt will depend on the characteristics and needs of a particular property.
 

 

Conclusions

• The total budget necessary for the establishment of a silvopastoral system in a degraded pasture land was lower when trees were not protected with wire fences than when they were protected. However, in the latter case, grazing the whole area of pasture can be advanced by about six months relative to the other method.

• The commercialization of products which can be obtained during the establishment phase may decrease the difference in final budgets between methods of establishment.

• Using fast growing and adapted trees, the silvo-pastoral system will be established in 16 to 22 months depending on the method of establishment adopted.
   

Acknowledgements

 

To the State Institute of Forestry, Minas Gerais (IEF) for supplying the eucalyptus seedlings and for technical advice regarding forest management. This study was partially supported by Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG)

 

References

 

Baggio A J e Silva V P 1998 Métodos de implantação de sistemas silvipastoris na região do Arenito Caiuá. In: Congresso Brasileiro Em Sistemas Agroflorestais, 2., Belém, 1998. No contexto da qualidade ambiental e competitividade: resumos expandidos. Belém: Embrapa-CPATU. p. 189-191.

 

Botero R y Russo R O 1999 Utilización de arboles y arbustos fijadores de nitrógeno en sistemas sostenibles de producción animal en suelos ácidos tropicales. In: Sanchez, M.D.; Rosales, M.M. (Editores) Agroforestería para producción animal en América Latina. Roma: FAO. Estudio FAO Producción y Sanidad Animal, 143. p. 171-192.

 

Carvalho M M 1997 Manejo do solo para o desenvolvimento de pastagens em áreas montanhosas da Região Sudeste. In: Passos, L.P.; Carvalho, M.M.; Campos, O.F. (Editores) Embrapa Gado de Leite: 20 anos de pesquisa. Juiz de Fora: Embrapa-CNPGL. 359p. p. 35-44.

 

Carvalho M M 1998 Recuperação de pastagens degradadas em áreas de relevo acidentado. In: Dias, L.E.; Mello, J.W.V. (Editores) Recuperação de áreas degradadas. Viçosa: UFV, Departamento de Solos; Sociedade Brasileira de Recuperação de Áreas Degradadas. 251p. p. 149-161.

 

Carvalho M M 2001 Contribuição dos sistemas silvipastoris para a sustentabilidade da atividade leiteira. In: MINAS LEITE, 3., 2001, Juiz de Fora. Sustentabilidade de sistemas de produção de leite a pasto e em confinamento: anais .... Juiz de Fora: Embrapa Gado de Leite. p. 85-107.

 

Carvalho M M, Freitas V P y Xavier D F 1999 Comportamento de cinco leguminosas arbóreas exóticas em pastagem formada em Latossolo Vermelho-Amarelo de baixa fertilidade. Revista Árvore, Viçosa, v. 23, p. 187-192.

 

Cóser A C, Cruz Filho AB, Martins C E e Alvim M J 1995 Efeito de diferentes cargas animais em pastagens de capim-gordura e braquiária. Pasturas Tropicales, Cali. v. 17, n. p. 37-40.

 

Cruz Filho A B, Cóser A C e Novelly P E 1986 Comparação entre métodos de plantio de Brachiaria decumbens em pastagens de capim-gordura em áreas montanhosas. Revista da Sociedade Brasileira de Zootecnia, Viçosa, v. 15, n. 4, p. 297-306.

 

Febles G, Ruiz T E, Alonso J and Gutierrez J C 2001 Introduction of trees in tropical pasture areas without altering the management of the livestock unit. In: International Symposium On Silvopastoral Systems, 2001, San José. Memórias. San José: CATIE. p. 187-189.

 

Ribaski J e Montoya L J 2001 Sistemas silvipastoris desenvolvidos na Região Sul do Brasil: a experiência da Embrapa Florestas. In: Carvalho, M.M.; Alvim, M.J.; Carneiro, J.C. (Editores) Sistemas agroflorestais pecuários: opções de sustentabilidade para áreas tropicais e subtropicais. Juiz de Fora: Embrapa Gado de Leite; Brasília: FAO. p. 205-233.

 

Silva J L S, Garcia R e Saibro J C 1996 Desempenho de bovinos e seus efeitos sobre as árvores em florestas de eucalipto (Eucalyptus saligna) na região fisiográfica da Depressão Central no RS. In: Simpósio Internacional Sobre Ecossistemas Florestais, 4., 1996, Belo Horizonte. Belo Horizonte: Biosfera. p. 342-345.