Effects of Different Land Use Types Managed By Smallholder Farmers on Soil Properties in Central Ethiopia

Background Aim: Land use change causes a remarkable change in soil properties. The nature of change depends on multiple factors such as soil type, type and intensity of land use, climate, and the like. This study investigated the variation in soil physicochemical properties across ve common land use practices i.e., enset system, farmland, and grazing-land (closed and open), and Eucalyptus woodlots practiced on originally same soil type and comparable topographic and climatic settings. Methods: A total of 105 disturbed and undisturbed soil samples [5 treatments (land use types) *7 replications (household)* 3 soil depth layers: 0–15cm, 15–30 cm, 30–45cm] were collected for selected soil chemical and physical analyses. Standard soil analytical procedures were followed in carrying out soil analysis. To meet the assumptions of normal distribution and homogeneity of variances, soil data on available phosphors were log-transformed before statistical analysis was undertaken and reported after back transformation. Two way analysis of variable were used to investigate the effects of land use and soil depth and their interaction on soil properties and when the analysis showed a signicant difference (p <0.05) among land use and soil depth men separation were made using Turkey’s pairwise comparisons. Results: There were signicant differences in physical and chemical properties of soil across land use and soil depth categories. Enset system had signicantly higher pH, available phosphorus (P), exchangeable potassium (K + ), soil organic carbon (SOC), and total nitrogen (TN) and their stocks than other land use types. Enset elds had higher SOC (78.4%) and soil TN (75%), and SOC and TN stocks of (66%) and (58%), respectively than cereal farmland. This study had also revealed a less expected nding of higher soil organic carbon and total nitrogen under Eucalyptus wood than farm land. Soil carbon and total nitrogen stocks showed a decreasing trend of enset system> closed grazing-land > eucalyptus woodlot > open grazing-land > farmland 0-45cm. Conclusion: Overall, some land use systems (e. g. enset agroforestry) improve the soil biophysical and chemical properties, while others such as cereal production degrade the soil. Hence appropriate land and soil management intervention should be promptly adapted to mitigating the continuous loss of nutrient from the dominantly practiced cereal farm land through maintaining crop residues, manure, crop rotation and scaling up agro-forestry system. over the age of 15–20 years old showed signicantly lower pH (5.68) and gravimetric soil moisture content (26.14%) than other land use types. The observed high organic carbon, total nitrogen and available P in the surface layer as compared to subsurface layer indicates that large amount of external inorganic fertilizers temporarily remain in the top surface soil compared deeper soil layer. Similar results were reported by (Wakene Negass and Heluf Gebrekidan, 2004; Alemayehu Kiu and Sheleme Beyene, 2013; Woldeamlak Bewket and Stroosnijder, 2003) who found higher available P in surface layers. and Starr (2015) for enset based agro-forestry systems for the 0–30 cm layer in southern Ethiopia. This may be attributed to the complexity and multi-strata nature of enset based agro-forestry system in their sites. The results of this study are in line with the ndings of other studies (Yimer and Abdulakdir, 2008; Gurmessa et al., 2016) who reported higher carbon and total nitrogen stocks under closed grazing land than other land uses in southern Ethiopia. However, the results of this study are different from the ndings of (Mekuria et al. (2014) in Ethiopia and Mureithi et al. (2014) in Kenya who have reported no signicant difference in carbon and total nitrogen stocks between closed and open grazing lands. This result agrees with by and found decreasing trend of and inorganic surface soil.


Introduction
In developing countries like Ethiopia, smallholder farmers produce more than 90% of the total agricultural produce ( Agidew and Singh, 2018). Most of the smallholder farmers practice a mixed crop-livestock-tree farming system where they allocate portion of their holdings to crops, grazing lands and tree cultivations. Such practices serve two purposes: reduce risk and supply diverse products needed to sustain the households. Financial limitation to buy su cient external inputs enforces them to prioritize application of organic and inorganic fertilizers to important portion of their farm elds, which lead to formation of heterogeneous soil fertility between plots and within plots of land (Tittonel et al., 2005;Haileslassie et al., 2006;Duguma et al., 2009). The complex farming system practiced in densely populated southern highlands of Ethiopia displays a good example of such fertility gradient within and between plots.
Farmers in southern Ethiopia grow the stable food crop Enset ventricosum (Welw.) Cheesman) together with multipurpose trees close to their homestead and fertilizes it with manure, while out farms grown with annual crops such as maize and wheat, teff, and barely are fertilized with inorganic fertilizer ( Haileslassie et al., 2006;Mellisse, 2017). Management practices applied to the elds such as plowing frequency, level of crop residue removal and the like varies between these plots all of which are likely to in uence the physicochemical and biological properties of soils (Duguma et Mekuria et al., 2014). Such variations may emerge from effects of biophysical and socio-economic setting of smallholder farmers, and their experience in managing land and soil, which are quite different. Therefore, it is important to assess the variation in soil physical and chemical properties due to difference in land use types and their management.
In the current study area, farm households practice a complex farming system by allocating their small holdings into different land use systems, such as enset system, out farmland, grazing land, and woodlots. These different land use systems and their management practices obviously in uence soil properties of each system, which is worth investigating. The main objectives of this study were, therefore, to: (1) evaluate the impacts of different land use systems, including enset system, eucalypts woodlot, grazing (open and closed), and cereal farmland on physicochemical properties of the soils; (2) identify land and soil management practices favorable for nutrient accumulation and responsible for nutrient depletion under current land use and management practices in southern Ethiopia; and (3) examine biophysical and socioeconomic factors that in uence soil fertility management in the study area.
The rainfall distribution is bimodal with long (July to August) and short (March to September) rainy season (Fig. 2). The area receives the mean annual rainfall of 1167 mm and average monthly maximum and minimum temperatures were 27.3 0 C and 10 0 C, respectively ( Fig. 2).
The indigenous trees that are commonly found in the study area are Acacia abyssinica, Croton machrostachys, Cordia african, and economically important exotic trees, mainly Eucalyptus species. The study area is characterized by a subsistent mixed crop-livestock farming system. Maize, teff, and wheat are the commonly grown annual crops and enset from perennial crops. The annual cereal system covered 52.68% of the farmland, while the enset system, grazing land, and Eucalyptus woodlots covered 26.49%, 12.68% 8.17%, respectively (Haile et al., 2017).

Soil sampling and analysis
Within the Miskan woreda, three kebeles (the smallest administrative unit in Ethiopia) represented by the presence of different land use systems were purposely selected and a total of seven farm households were considered. In this study, the ve land use systems were considered as treatments and seven farm households as replications. The ve land use systems were esnet system, grazing systems (open and closed), annual cereal system, and eucalyptus wood lot (Fig. 3). Soil samples were taken from each land use system at a depth of 0-15cm, 15-30cm, and 30-45cm from adjacent land use types having similar soil type (i.e. soil textural class is clay loam) and climate (altitude) and slope (0-5%) to avoid confounding factors. For bulk soil destiny determination, soil samples were collected from same pits at the three soil depths (0-15 cm, 15-30 cm and 30-45 cm) from opposite sides of the pits using core sampler of 5 cm height by 3 cm diameter. From the ve land sue types of each farm household, soil samples were taken at three level of depths, totally 105 samples were taken for soil chemical and physical analysis. Representative soil samples were taken from each land use type of the different farm households and were analyzed at the Holetta soil and plant testing laboratory. The samples were mixed manually and subsamples were grounded to pass through a 2 mm sieve. Soil pH in water was determined at 1:2.5 (soil: water ratio), soil organic carbon (OC) by wet combustion method (Walkley and Black,134), and total nitrogen by Kjeldahl digestion, distillation and titration method (Anderson and Ingram, 1993). Soil available phosphorus was determined using Olsen method (Olsen, 1954). Soil exchangeable potassium (K + ) was determined by ammonium acetate extraction method using the ame photometer (Anderson and Ingram, 1993). The particle size analysis was made using hydrometer method and the soil textural classi cation was determined using USDA textural triangle. Soil bulk density was determined through volumetric method after the soil was oven dried at 105°C for 24 h (Hillel, 2004). Soil-moisture content was determined by gravimetric methods after the soil was oven dried at 105°C for 30 h (Anderson and Ingram, 1993). The carbon and nitrogen stocks ha -1 was calculated using the following formula by substituting N content at C content in the formula for nitrogen stock.

Statistical analysis
The results of the physical and chemical properties, and carbon and N stocks were subjected to GLM test with two-way analysis of variance using SPSS version 16 to test the effects of land use, soil depth and their interaction on soil chemical and physical properties. The means for treatments that showed signi cant differences by F-test were separated by Tukey's honestly signi cant difference test (Tukey-HSD test) and a signi cance level was declared with P < 0.05. Prior to statistical analysis we preformed Levene's test to check for the homogeneity of variances (homogenous in all cases). To meet the assumptions of normal distribution and homogeneity of variances, data on available P was log transformed before statistical analysis was undertaken and reported after back transformed.

Impacts of land uses on soil physical properties
Land use had signi cant effect on soil physical properties mainly textural fraction, soil moisture and bulk density ( Table 2 ). Clay fraction was the only particle size fraction (%) that signi cantly (p < 0.005) affected by land use but not affected by soil depth and by interaction effects (Table 2)but the sand and silt fraction did not vary between the different land use systems, soil depth and by their interaction ( Table 2). Signi cantly higher clay proportions of 41.47 ± 1.25% and 40.9 ± 1.96 were observed in farmland and closed grazing land, respectively while the lowest clay proportion of 32.9 ± 1.85% was recorded from enset eld ( Table 3). Although the clay fraction varied between land use types, the soil textural class in the study area is classi ed as clay loam as per USDA classi cation.
Land use types and soil had a signi cant (p = 0.009) effect on gravimetric soil moisture content ( Table 2). Similar to clay fraction, cereal farmland and closed grazing land had signi cantly higher gravimetric soil moisture contents of 32.41 ± 1.19% and 34.79 ± 1.19%, respectively, while the adjacent soils under woodlots had low a moisture content of 26.14 ± 1.19 %) ( Table 3). Surface soil holds signi cantly higher gravimetric soil 28.63 ± 0.9 (Table 3). Table 1 Description of the studied land use types in the central highlands of Ethiopia.

Enset system
Enset system is eld located close to home and used for growing enset plant and often fertilized with manure and other household waste for more than 50 years. Enset is a perennial herbaceous species of owering plant in the banana family, Musaceae and one of the staple crops which has been feeding about 20 million peoples in southern Ethiopia. It also hosts diverse multipurpose tree, shrub as well as crops and grass.

Closed grazing land
Closed grazing land is privately owned pasture land located far from home. Served as pasture land for more than 50 years. It is used as closed grazing land in wet season (June-November) and open grazing land in dry season. It is located down the slopes; hence it is subjected to waterlogging during wet season.

Open grazing land
Open grazing land is privately owned pastureland, located close to home and used as open grazing land for more than 25 years. It is established on the degradedfarm land and often few park land trees are retained where the cattle use as shade during the day time.

Eucalyptus Woodlots
Woodlot is small-scale plantation of Eucalyptus camaldulensis. It is established on the degraded cropped land where farmers abandoned because of their poor soil fertility some 15-17 years ago.

Cereal farmland
Cereal farmland is a low input cereal cropping system, located far from home and dominantly used for growing annual cereal crops such as wheat, teff, and maize. The eld has been used for cereal farming for more than 50 years and fertilized with inorganic fertilizers (DAP and Urea). It is intensively and frequently plowed, and crop residues are often removed for feed and fuel. Values followed by the same letters in a row are not signi cantly different at P < 0.05; or 0.01* signi cantly different at p < 0.05; ** signi cantly different at (p = 0.01) ns denotes not signi cantly different. Values followed by the same letters in a row are not signi cantly deferent at P < .05; or 0.001* signi cantly different at p < 0.05; ** signi cantly different at p = 0.001 ns denotes not signi cantly different.
On contrast, bulk density (g/cm3) signi cantly affected by soil depth but land use did not signi cantly affect it ( Table 2). Soil bulk density was signi cantly higher (1.25 ± 0.02g cm-3) at 30-45cm (Table 3). Though, land use did not affect soil bulk density in terms of absolute values cereal land use had the higher bulk density of 1.23 ± 0.09a (Table 3).

Impact of land use types on soil chemical properties
Similar to soil physical properties, land use had a signi cant effect on soil chemical properties (Table 4). Soil organic carbon (%) and total nitrogen (%) showed signi cant difference among land use types (p = 0.000), soil depths (p = 000) and by interaction effect (p = 0.001) ( Table 4). Enset elds had signi cantly higher soil organic carbon and total nitrogen than other land-use types, while the adjacent the cereal farmland had the lowest content of soil OC and total N ( Table 5). Signi cantly higher organic carbon of 2.48 ± 0.17 and total nitrogen of 0.21 ± 0.08) and were measured in enset eld (Table 5). Soil organic carbon and total nitrogen were signi cantly affected across soil depth. Signi cantly higher organic carbon of 2.46 ± 0.08% and total nitrogen of 0.21 ± 0.06% were measured from surface (0-15cm) respectively (Table 5). Available phosphorus (ppm) showed statistically signi cant difference among land use but it did not vary with soil depth and by the interaction effect (Table 4). Enset soil contained signi cantly higher available phosphorus of (213.63 ± 19.8 mg kg -1 ) than other land uses while the adjacent soils under closed grazing land had the lowest available P (18.12 ± 1.5 mg kg -1 ) ( Table 5). Likewise, exchangeable K + has shown signi cant (P < 0.05) difference among land use types ( Table 4). The soil exchangeable K of Enset eld was about ve times higher than the closed grazing land, and it was 2.8, 2.4 and 1.9 times higher than the farmland, eucalyptus woodlots and open grazing land, respectively ( Table 5).
The soil OC and total N stocks showed a decreasing trend with increasing soil depths regardless of the land use types ( Table 5). The upper soil depth had higher SOC and total N for the ve land use types. Enset eld which received manure as input had higher soil OC and total N and their stocks (Table 5). Table 5. Effects of land use types on soil organic carbon (SOC) and total nitrogen (TN) Stocks in central Ethiopia.  (Table 7). The total N stock also showed a similar trend, which was higher in enset farm and closed grazing land, but it was the lowest in the cereal farmland (Table 7).  (Table 7).
Surface soil that received higher inputs had the highest soil organic carbon stock of (40.90 ± 1.4 Mg C ha-1) and TN stock of (2.5 ± 0.1 Mg ha-1) respectively than subsurface soil (Table 7).
OC and TN at correlation coe cient of (0.983**) (Table 8) and OC and TN were positively and strongly correlated with available P at r (0.421**) and (0.425**), respectively while they were strongly and negatively correlated with BD at r (-0.356**) and (0.339**) (Table 8) 4. Discussion In addition, the frequent and long year intensive tillage in farmland may rise clay fraction through enhancing weathering process as it shears and pulverizes the soil and changes moisture, aeration, and temperature regime (Yimer et al., 2008;Awdenegest et al., 2013). In contrary to this study, the insigni cant effect of land use on soil textural fraction was reported in Kenya (Shepherd et al., 2000). The signi cantly lower soil moisture observed in the soils of woodlots may be due the voracious moisture uptakes of eucalyptus tree given its deep rooted and fast-growing nature (Sanginga and Swift, 1992). Other studies conducted in other parts of Ethiopia reported lower soil moisture content under eucalyptus plantation compared to other land uses types (Getachew et al., 2013;Chanie et al., 2013). Almost all farmers in the study area perceived that eucalyptus tree species can deplete major soil nutrients, such as N, P and K, and moisture content due to the rapid growing nature and consequent high uptake of water and nutrients. The

Soil chemical properties
Evidence has shown that land use differences signi cantly affect soil pH as soil in Enset system had higher soil pH (6.81) than soil in Eucalyptus woodlots pH (5.68). The observed higher soil pH in enset eld, which received manure, household waste and ash as input may be due to the soil's increasing recovery of natural buffering capacity in response to increased organic matter in the soil. perceived better soil fertility in elds close to home than those situated far from home. Farmers in southern Ethiopia including in the study area grow the perennial food crop Ensete ventricosum (Welw.) Cheesman) close to the homestead and fertilize it with organic fertilizers such as manure, household waste (ash and crop residues), but they grow annual eld crops further away from the homestead and fertilize them with inorganic fertilizers, mainly N and P fertilizers ( Haileslassie et al., 2006;Haile et al., 2017;Mellisse, 2017). In the study area, farmers apply large amount of organic inputs such as household wastes, livestock manure and urine, wood and/or dung ash, crop residues including pruned lateral branches of enset plant, which could lead to increased soil nutrient and carbon content over a long period of time (Fig. 4). For instance, the majority of the interviewed farmers often applied organic inputs such as manure, crop reside or green manure and household waste including ash and cattle urine mainly in homestead or enset eld.
The improved soil fertility in the enset system might also be associated with the co-existence of multipurpose trees, shrubs, herbs, and grasses that could enhance internal organic inputs ( The current study has shown a less expected nding of extremely high available P of 213 mg kg -1 under enset system which seems much higher than reported in the literatures. The possible reasons for observed extremely high available P in enset eld could be due to the long-term application of household wastes in the form of wood and dung ash. Next to the enset eld, the closed grazing land could be the second alternative land use system for storing soil nutrient and carbon stocks. The observed high SOC and TN in closed grazing land could be attributed to the protection of soil nutrient from wind and soil erosion due to the permanent grass cover, which agrees with the nding of Gurmessa et al. (2016) who reported higher OC and TN contents and their stocks in closed pastureland than other land use types in southern Ethiopia. Moreover, the improvement in soil OC and TN under closed grazing land may be due to the high organic input added into the system and high turnover rate of ne grass root biomass especially in the surface layer (Rhodes et al., 2000). In contrast, the observed low available P and exchangeable K in soil under closed grazing land could be due to reduced pH and absence of grazing inputs such as urine and dung because of enclosure of closed grazing land during wet season and collection of dung during dry seasons by the farmers for fuel. Another possible explanation for the observed low available P could be associated with waterlogged conditions of the sites. For instance, we found high clay contents of 40.9% in adjacent closed grazing land and 41.96% in cereal farmland, respectively ( The current study has also shown a less expected nding of higher OC and TN in woodlots than the adjacent farmland. For instance, among seven cases considered in this study there was one case where woodlots had higher OC and TN contents of 2.75% and 0.25%, respectively ( Table 5 Other biophysical conditions such as land fragmentation (increased distance of cereal eld from home) (Tittonell et al., 2005)  In general, integrating perennial food crops such as enset with multipurpose trees close to the homestead and fertilizing it with manure, compost and household waste as well as protecting grazing lands from free grazing through enclosure has not only improved soil nutrient status but also enhance soil carbon stocks and mitigate climate changes at watershed level (Haileslassie et al.,2006) and at farm level (Negash and Starr, 2015). However, the dominantly practiced cereal farming has been degrading the quality of soils.

Conclusions
The results of this study showed that land use types and soil depth were signi cantly in uenced physicochemical properties of the soils. Soils under enset system and surface soil had signi cantly high pH, available P, exchangeable K, OC, TN and their stocks than other land use and depth categories. Moreover, closed grazing land had higher OC and TN except for available P and exchangeable K.
Despite the popular notion of environmentally alleged effects of eucalyptus tree on soil properties we only found speci c negative effects on soil properties (e.g., low pH and moisture content) from woodlots established some 15-20 years ago on degraded crop lands where farmers abandoned because of soil degradation. The improvement of soil OC and TN under woodlots indicates that planting of fast-growing eucalypts tree species on degraded crop lands will enhance their soil quality while improving the livelihoods of rural communities and could reduce the pressure on remnant native forests for various wood products.
The OC and TN stocks often served as a proxy for soil quality indicators which followed a decreasing trend of enset farm > closed grazing land > eucalyptus woodlots > open grazing land > cereal farmland at 0-45 cm soil depth. Other soil nutrients such as available P and K + , with the exception of available P in the closed grazing land were also measured in the same trend. The overall results of the study clearly indicated that among the ve land use types considered in this study enset and closed grazing land-use types will have higher potential for improving soil quality and mitigating climate change while the dominantly practiced cereal cropping is responsible for the continuous loss of soil organic carbon and nutrients in the study area. Hence, future soil management strategies should focus on improving soil nutrient management through the retention of crop residues, practice of crop rotation and promotion of agro-forestry system to reduce the continuously loss of soil nutrients from the dominantly nutrient depleting cereal cropping system.

Declarations
Ethics approval and consent to participate Location map of the study area. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 2
Climatic condition of the study area (Source: NMSA, 2015). Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.  Enset system receives large amount of organic inputs Enset and closed grazing land are located close to home while closed grazing land and farm(cereal land far from it Figure 6