Soil Properties under Selected Homestead Grown Indigenous Tree and Shrub Species in the Highland Areas of Central Ethiopia

A study was conducted in Galessa-Jeldu areas to evaluate soil pH, organic C, total N, available P and exchangeable bases under four indigenous and one exotic tree species, and to examine the correlation between the different soil properties. The tree and shrub species included in the study were Senecio gigas Vatke, Hagenia abyssinica (Bruce) J.F. Gmel., Dombeya torrida (J.F. Gmel.) P. Bamps, Buddleja polystachya Fres. and Chamaecytisus palmensis (Christ) Bisby and K. The first four are indigenous, while the last one is an exotic N-fixing species. The soil pH values under H. abyssinica and S. gigas were above 6.34 as compared to the soil pH values under C. palmensis, D. torrida and B. polystachya. Total N was slightly higher under H. abyssinica although not statistically different with the N content under other species. Organic C was higher under H. abyssinica than under B. polystachya. The variation for exchangeable K was more at 0-15 cm soil depth than at the 15-30 and 30-50 cm soil depths. The correlation between soil pH and total N as well as pH and organic C was positive in almost all the horizontal positions. Hagenia abyssinica, S. gigas and C. palmensis are some of the potential species that can be evaluated in different land-use systems of the high altitude areas to protect the loss of plant nutrients.


Introduction
High human and livestock population, a decline of forest resources and soil fertility depletion are some of the features of the highlands of central Ethiopia (ICRAF, 1990;German et al., 2005;Amare et al., 2006). The natural forests in the highlands of Ethiopia in general are heavily exploited to fulfill the cash and wood demand of the growing population. Similarly, the areas previously covered by forests have declined as a result of the expansion of farming with annual crops. Homesteads are one of the most important niches in which farmers feel confident to plant and maintain tree and shrub species.
Homesteads in the high altitude (> 2900 m.a.s.l) areas of central Ethiopia have a better tree and shrub species composition than farmlands (Kindu et al., 2006a). The proportion and area coverage of indigenous species around the homesteads is also considerable as compared to exotic species. The increased proportion of indigenous species over the exotics is due to their adaptability to the local environmental conditions, resistance to pests and diseases, availability as sources of planting material and familiarity to the local farmers. Farmers grow tree and shrub species around homesteads to obtain various products (wood, food and feed) and services (live fence, shade and soil fertility improvement) (Berhane et al., 2006).
Farmers in the high altitude areas highly recognize Hagenia abyssinica, Dombeya torrida and Senecio gigas as important indigenous soil improving tree and shrub species (Kindu et al., 2006a). The contribution of some other indigenous tree species for soil fertility improvement in farmlands has been investigated in the intermediate altitude (1500-2400 m.a.s.l) areas of Ethiopia by Abebe et al. (2001); Yeshanew et al. (1999); Hailu et al. (2000); Poschen (1986) and Gindaba et al. (2005). Assessments of the soil properties under and away from the tree canopies were the focus of all of the studies. The report from most of the previous studies show a higher soil organic matter, total N and other soil attributes under the tree canopies than in open areas.
Studies on soil properties under farmers recognized indigenous tree and shrub species around homesteads are limited in the high altitude areas. Cognizant to the research gap, a study was conducted from 2004 to 2006 (a) to evaluate soil pH, organic C, total N, available P, exchangeable bases under four indigenous and an exotic tree species; and (b) to examine correlations between the different soil properties.

Study Site
The study was conducted from 2004 to 2006 in the upper plateaus of Galessa-Jeldu areas, Dendi and Jeldu districts, central Ethiopia (9 o 02' 47'' to 9 o 15' 00 N and 38 o 05' 00'' to 38 o 12' 16'' E). The altitude ranges from 2900 to 3200 m.a.s.l. The rainfall pattern is bimodal. The main rainy season is from June to September and the short rainy season from March to April with a mean annual rainfall of 1399 mm. Barley is the most dominant crop followed by potato and enset (Ensete ventricosum). The most common land-use systems in the study areas are crop, pasture and forest. Trees and shrubs are concentrated around the homestead and the nearby Chilmo state forest. The original vegetation in the area was mainly H. abyssinica and Juniperus procera. The exotic tree and shrub species introduced in the area include C. palmensis and Acacia decurrens. Cattle, sheep and horses are dominant livestock in the study area. The soil is characterized as Haplic Luvisols. The physical and chemical properties of the soil are presented in Table 1.

Selection of the Tree and Shrub Species
A total of 150 households were interviewed to find out indigenous tree and shrub species that are traditionally considered by farmers as important soil fertilizers. Subsequently the farmers were asked to describe the type of tree and shrub species that they consider important. The most important tree and shrub species for the farmers were species that regularly shed their leaves, have fast decomposing leaves, grow fast, propagate easily, produce high biomass, and protect soil erosion. Finally, farmers selected Senecio gigas Vatke, Hagenia abyssinica 10 (Bruce) J.F. Gmel, Dombeya torrida (J.F. Gmel.) P. Bamps and Buddleja polystachya Fres (Kindu et al., 2006a). Three villages that had all the required indigenous species were identified. Tree and shrub species free from addition of farm weeded material; manure or house wastes and inorganic fertilizers were selected and demarcated in each village. The three villages were considered as replications.
Most tree and shrub species in the village exist in hedges. A total of three trees per indigenous and exotic species (Chamaecytisus palmensis (Christ) Bisby and K. Nicholls) were included in the present study. The exotic tree species was used for comparison purposes. Detailed description of the species is presented in Table 2.

Soil Sampling and Analysis
A radial pattern of soil sampling was used. Sampling locations were 75 cm (hereafter referred to as closest), 150 cm (hereafter referred to as midst), and 225 cm (hereafter referred to as distant) positions at both sides from the base of each marked tree (Hailu et al., 2000;Wezel, 2000;Power et al., 2003). Sampling depths were 0-15, 15-30 and 30-50 cm (Kindu et al., 1997). Soil samples collected under each species and village from similar depths and positions were thoroughly mixed to obtain composite samples. The total number of composite samples was 135 (Soil improving tree and shrub species (5) * depths of sampling (3) * distances of sampling from the base of the tree (3) * replications (3)).

Statistical Analysis
A one-way analysis of variance (ANOVA) was conducted on soil pH, OC, N and exchangeable bases. The effect of depths and positions, and their interactions on soil properties was conducted following the procedures of liner models ANOVA. Both analysis were run using SAS (SAS institute, 1999). The significance between means was tested using the least significance difference (LSD). The following model was considered while running the ANOVA: where is the overall mean, i the i th treatment (tree species) effect, j the j th block (site) effect and e ij is the random error associated with Y ij .
Correlation analysis was performed to understand the relation between OC vs N; pH vs OC, N, Al, Mn, BS and CEC. Levels of P < 0.05, P < 0.01 and P < 0.001 were chosen to test significance differences.

Soil pH, Organic C, Total N and Available P
Soil pH at the depth of 0-15 cm varied horizontally from 6.01 to 6.8, 5.9 to 6.7 and 5.85 to 6.95 in the closest, midst and distant positions, respectively (Table 3). The soil pH values under H. abyssinica and S. gigas were above 6.34 as compared to the soil pH values under C. palmensis, D. torrida and B. polystachya. The pH values under the five tree and shrub species varied significantly among the soil depths (Table 3 and 4). Higher pH values were found at the topsoil than that of lower depths. Similarly, the soil pH varied along the three horizontal positions. Higher pH values were noticed in the closest position than in the midst and the distant positions.
The soil OC and N content under H. abyssinica, S. gigas, C. palmensis and D. torrida were comparable in the top 0-15 cm depth (Table 3). On the other hand, the soil OC content under H. abyssinica was higher by 23.25, 24.53 and 21.03 mg g -1 than under B. polystachya in the closest, midst and distant positions, respectively. Total N was slightly higher under H. abyssinica although not statistically different with the N content under other species. The content of soil P had the following sequence in the top 0-15 cm soil depth of the closest and midst horizontal positions: H. abyssinica > S. gigas > C. palmensis > D. torrida > B. polystachya (Figure 1). Organic C, N and P showed significant differences among the soil depths, as well as the horizontal positions (Table 4). The content of OC, N and P depicts a decreasing pattern from the 0-15 to the 30-50 cm soil depths and from the closest to the midst and distant positions.

Exchangeable Bases
The variation for exchangeable K and Ca was high at 0-15 cm soil depth than at the 15-30 and 30-50 cm soil depths ( Table 5)

Correlation between Soil Properties
The correlation between OC and N was positive and statistically significant at all the soil depths of the three horizontal positions (Table 6). Soil pH was positively and significantly correlated with CEC and BS, and negatively and significantly correlated with Mn. The correlation between soil pH and N as well as pH and OC was positive in almost all the cases. No significant correlation was found between pH and Al; pH and OC; pH and N, except at the 0-15 cm soil depth and in the closest position of the five tree species.

Discussion
Tree and shrub species can differ in their effect on soil properties through various mechanisms, including rates of nutrient inputs, outputs, and cycling. The high content of OC and N under H. abyssinica as compared to B. polystachya could be due to the addition of higher organic resources of the former than the later. Hagenia abyssinica constantly sheds high amount of leaves and provides mulch and green manure to the soil within its vicinity. Kindu et al. (2006b) reported the presence of high amount of litter deposition under 64 months old H. abyssinica and Grevillea robusta on Nitisols of central Ethiopia. Dombeya torrida, S. gigas and C. palmensis shed substantial amount of leaves even though their leaf shedding pattern is not as regular as that of H. abyssinica.
Farmers in the highlands of central Ethiopia rake the excess foliage litter under H. abyssinica, D. torrida and S. gigas trees and use them as organic fertilizer source for the nearby crop fields. However, H. abyssinica, D. torrida and S. gigas as non N-fixing tree species only cycling the N present in the soil, not adding N inputs to the system, as happens through biological nitrogen fixation (BNF) in C. palmensis. Non N-fixing tree and shrubs species in general obtain their N and other nutrients through effective retrieval from the soil (Jama et al., 2000).
The contents of P, K, Ca and Mg in the soil under H. abyssinica, S. gigas and C. palmensis were relatively high as compared to the other species considered for the present study (Table 5 and Figure 2). Enrichment of these nutrients under the three species could be associated to the rooting system and efficient nutrient cycling power of the trees. Deep-rooted trees and shrubs often act as nutrient pumps , taking nutrients from deep subsoil horizons into their root systems, translocating it to their leaves, and recycling it back to the surface of the soil via leaf fall and leaching (Kindu et al., 1997(Kindu et al., , 1999Berger et al., 2006). Hence, the cycling of essential nutrients maintains an abundant supply of nutrients for incorporation into new biomass while at the same time limiting nutrient losses from the soil profile.
The soil under B. polystachya had the lowest nutrient content as compared to all other indigenous and exotic species included in the present study. The low soil nutrient content under B. polystachya can be related to the very low leaf shedding characteristics of the species. As a result of low leaf shedding, the nutrient gain of the soil system under B. polystachya through litter fall can be inadequate. It is likely to notice the positive impact of the B. polystachya through chopping and incorporating the green biomass of the tree into the soil system. Buddleja polystachya readily coppice and provide substantial amount of green biomass.
Organic C, N, P, K, Ca and Mg progressively declined with depth and horizontal distance from the base of the tree and shrub species. The presence of the tree and shrub species and associated soil enrichment would therefore seem to be restricted to the near surface layers. The effect of the trees on the bulk of the soil in the lower depths was very minimal (Table 3, 5 and Figure 2). The minimal effect for the five tree and shrub species on the soil properties in lower as compared to topsoil depths could be related to young age of the trees. The age of the tree and shrub species included in the present study was between four and nine years (Table 2). Trees can positively influence soil properties in lower soil depths during many years of their below and aboveground growth (Pandey et al., 2000;Chang et al., 2002). Improvement of soil nutrients in the upper soil depths and close to the tree stems has been reported to various tree and shrub species (Yeshanew et al., 1999;Tadesse et al., 2000;Abebe et al., 2001;Gindaba et al., 2005).
The soil pH under H. abyssinica and S. gigas was high as compared to the soil pH under other species (Table 3). The elevated soil pH under the two species could be a result from increased base cation cycling and subsequent enrichment of the base status of the underneath soil. The presences of higher level of exchangeable base forming cations contribute to the amelioration of soil acidity (Brady, 1990). Our findings on the levels of K, Ca, Mg and Na under H. abyssinica and S. gigas are in accordance with the investigation of Sae-Lee et al. (1992) and Tadesse et al. (2000).
The level of CEC under the five species was slightly higher in the top 0-15 cm soil depth than in the lower 15-30 and 30-50 cm soil depths (Figure 2). Likewise, the soil under H. abyssinica and S. gigas depicted relatively more CEC than under other species. Like the contents of base cations, higher level of CEC under the two species could be due to greater amount of litter deposition. The level of CEC varies with changes in soil pH, organic matter, and clay content. Cation exchange capacity provides a reservoir of nutrients to replenish nutrients that can be removed from the soil water by plant uptake and leaching (Camberato, 2001).
Total soil N was positively and strongly correlated with OC under the five tree and shrub species. Such a correlation of N with OC was expected as the amount of the former parallels with soil organic matter (Schlesinger, 1997;Brady and Weil, 2002). Similarly, soil pH was positively related with BS and CEC while negatively related with Al and Mn. Soil pH drops as acidic H and Al cation in the exchange sites increases and basic Ca, Mg and K cation decreases (Brady and Weil, 2002).

Conclusions
The soil under H. abysinica, S. gigas and C. palmensis retained a substantial amount of plant nutrients. This is an indication of the species potential to protect the loss of nutrients and thereby contribute to long-term ecosystem sustainability. Hence, the planting efforts of H. abyssinica, S. gigas and C. palmensis should be strengthened in different land-use systems of the high altitude areas.

Acknowledgements
The authors would like to thank the Austrian Development Cooperation, the Commission for Development Studies at the Austrian Academy of Sciences (KEF), the African Forestry Research Network (AFORNET) and the Ethiopian Institute of Agricultural Research (EIAR) for the financial assistance that has supported this work. We would also like to express our gratitude to the Institute of Forest Ecology and Institute of Applied Geology at UNI-BOKU in Austria for the laboratory analysis of soil samples. The support from staffs of forestry, soil and water management research divisions at Holetta, and farmers and development agents in the Galessa-Jeldu areas are highly appreciated. We thank anonymous reviewers for their constructive comments.