Evaluation of soil fertility status in the Kyoga Basin of Uganda: A physio-chemical study in Buyende and Serere districts

Nutrients present in the soil are important elements required for the healthy growth of plants. This study, therefore


INTRODUCTION
Soil is the "soul" of infinite life and biodiversity and its quality affects nutrient cycling and human well-being (Kekane et al., 2015;Bogunovic et al., 2017).As a terrestrial ecosystem component, soil performs various functions, including storage of plant-available water, supply of adequate oxygen to roots, provision of favorable seedling establishment conditions, storage of nutrients, suppression of plant pathogens, and immobilization of contaminants, all of which are essential to plant growth (Khatoon, 2020).Previous studies have revealed that agricultural production was negatively impacted by soil degradation and poor soil nutrients (NPK), mainly where the agricultural intensification has been observed (Ebanyat et al., 2010;Giller et al., 2011;Mubiru et al., 2017).In addition, there are strong indications that nutrient balances and soil fertility status differ widely between farms in different wealth categories and between plots at different distances from homesteads (Zingore et al., 2007;Tittonell et al., 2010).
In Sub-Saharan Africa, most of the arable land is estimated to have low soil fertility, a situation which has been exacerbated by the ongoing decline in soil physiochemical properties and loss of nutrient stock (Hengl et al., 2015;Bamutaze et al., 2021).For example, a magnitude of studies has highlighted similar soil fertility status trends generally in Central and Eastern Uganda (Mubiru et al., 2017;Nimusiima et al., 2018;Bamutaze et al., 2021).This does not exclude Buyende and Serere districts from this scenario.Soil physio-chemical properties are key components of soil fertility and ecosystem functioning since they govern the efficacy of vertical and horizontal terrestrial processes, influence cumulative feedbacks, geochemical cycles, and overall stability of the earth system (Bamutaze et al., 2021).A plethora of studies e.g.(Moody and Phan, 2008;De Laurentiis et al., 2019) portrayed how different soil physio-chemical properties such as texture, pH, organic matter, total Nitrogen, extractable phosphorus, potassium, soil depth among others can be used to assess soil fertility status.
Soil fertility assessment is fundamental to suggest optimum conditions for plant growth (Yerima and Van Ranst, 2005).According to Khan et al. (2018), deficiency in soil nutrient parameters such as; nitrogen (N), phosphorus (P), and potassium (K), remains one of the major yields limiting factors for crops.Other vital nutrient parameters like soil organic matter (OM) and soil pH, provide macronutrient blocks for protein building in plants and also responsible for the absorption capabilities other nutrients (Minasny et al., 2016;Wood et al., 2018).The differences in fertilization, cropping systems and farming practices are the main factors influencing soil fertility at the field scale and therefore, the ability to produce sufficient crops for food security and income (Woniala and Nyomb, 2014).Understanding the variability of soil fertility status, its distribution, and the causes of the observed variability are important in improving sustainable crop production for districts.smallholder farmers in areas like Buyende and Serere Since soil is fundamental for sustainable agriculture production with prominent outcomes on food security due to increasing population and living standard (Mulumba and Lal, 2008), thorough knowledge of soils quality, and fertility status is required.Given the magnitude of key roles played by soil in many sub-Saharan African regions, such as in East Africa (Uganda), rapid population growth, poor farming methods, and an unfavorable developmental project have exerted great pressures on soil resources, interfering with its fertility status making farmers who cultivate on fragile environments to experience tremendous soil degradation and severe crop yield decline on their lands.Therefore, it is from this background that the present study attempts to assess the soil fertility status using physio-chemical soil properties for the suitability of different rain-fed crop cultivations in the districts of Buyende and Serere given their importance at the local and national scale.
Thus, the study intended to map spatial distribution of the physiochemical soil properties of the respective study districts.

Study area
This study was conducted in the districts of Buyende and Serere which are within the Lake Kyoga basin (Kyoga Plains Agro-Ecological Zone) of Uganda (Figure 1).The study area has mean annual rainfall that ranges from 800-1500mm/year with a mean elevation of about 1214 m.a.s.l (Lugoi et al., 2019).The two districts are associated with various agroecological characteristics including moist thicket supporting combretum in the South Eastern and Eastern part of Buyende district while terminalia woodland features are found in Serere district (Mubiru et al., 2017).

Soil sampling
Relevant information related to the registration, referencing and identification of the soil to be described was noted before the actual soil sampling such as the general site information including; location (district/sub-county and GPS coordinates), auger/sample number, farm owner, authors, and date.Two farms from each subcounty of Serere and Buyende districts (Figure 2) were purposely selected using a judgmental sampling technique (Weber et al., 2021).
This technique involves selecting sample points based on the knowledge held by the researcher and the field attendant, however, for this study, the farm owner guided in the selection of the sampling points on his/her farm and marked using a GPS.At each of the identified locations (Figure 2), soil samples were obtained *Corresponding author.E-mail: odongoronieplanks@gmail.com; Tel:+256753719913.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License   using a soil auger at a depth of (0-25 cm) to allow sufficient examination and description of the different soil properties (Nimusiima et al., 2018).This soil depth level was chosen because it has very little or no laterite (Kitutu et al., 2009), and soil at this depth show very little horizon differentiation apart from slight changes in firmness that fall within the agricultural layer (Majaliwa et al., 2010)..Soil sampling was done based on the guidelines for soil sampling (De Crop et al., 2017).Following the soil sampling protocols, 23 and 17 soil samples were collected from Serere and Buyende respectively.Soil sample information and data were recorded on a field data collection form for each site and sample (approximately (0.5-1 Kg).At each of these sites, the samples were securely packed in plastic bags and labeled appropriately with the date, time, and sample number.The samples were sent to the soil, plant, and water analytical laboratory at the College of Agricultural and Environmental Science, Makerere University for scientific analysis.

Laboratory soil sample preparations
The samples collected (Figure 3) were prepared for laboratory analysis.The samples were air-dried for 3 days to eliminate moisture (Mubiru et al., 2017).Thereafter, they were grounded using a porcelain pestle and mortar, and sieved through a 2millimeter sieve to remove debris and other non-soil materials including stones and roots.The sieved soil particles were repackaged, relabeled, and analyzed (Ladwani et al., 2012).Several agronomical soil properties (soil pH, nitrogen (N), phosphorus (Av.P), organic matter (OM), exchangeable cations Potassium (K+) and soil texture (Proportions of Sand, Cay and Silt), were prepared for analysis.
Each of the above soil properties was specifically analyzed by a particular analytical method and procedure as described here; pH was measured using a pH meter with a soil/water solution at a ratio of 1:2.5 (Mahajan et al., 2022); N was determined calorimetrically by digestion using concentrated sulphuric acid and Selenium powder with Salicylic acid (Huang et al., 2022).P was determined spectrophotometrically at 882 nm wavelength using ammonium molybdate in the presence of ascorbic acid on Mehlich-l extract (Murphy and Riley, 1962).K+ was read on Mehlich-1 extract using a flame photometer; Texture analysis was performed using the hydrometer method (Mubiru et al., 2017) and organic matter was analyzed using the Walkley-Black method.All analysis procedures were performed using routine analytical procedures as outlined by Perkins et al. (2021) and other standard operating procedures that are internationally recommended.For each analysis, a reference sample and blanks were repeatedly included for quality control and assurance purposes.

Soil structural stability index
Structural Stability Index (SSI) is a physical parameter that determines the degree of degradability and erodibility of soil (Peng et al., 2023).According to Shen et al. (2022), SSI values provide a useful guide on the relative stability of the soil to externally imposed destructive forces and, hence, appropriate indicators of the relative susceptibility of the soils to detachment, runoff, and erosion.Therefore, in this study, equation ( 1) was adopted to compute SSI in R software for Buyende and Serere districts into the following categorical Indices; (that is [< 5%;5-7%;7-9%;>9%]) for structurally degraded soil, high risk of degradation, low risk of degradation and stable structure respectively. (1) Where is soil structural stability index, , and are mean values of soil organic matter, silt, and clay content respectively.

Statistical analysis
Field soil datasets were subjected to statistical analysis using R software.The descriptive statistics of minimum, maximum, range, mean, standard deviation (SD), and coefficient of variation (CV%) were computed for the selected physio-chemical properties.From this analysis, three variability classes as adopted from Asongwe et al. ( 2016)'s study were created (that is [< 15%; 15-35%; >35%]) for slightly variable, moderately variable, and highly variable respectively.

Spatial distribution of physio-chemical soil properties
Spatial Maps of Buyende and Serere for N, P, K, OM, Sand, Clay, and Silt were prepared in ArcGIS 10.2.2 by interpolating the measurements at the soil sampling sites using inverse distance weighting (IDW) (Tunçay et al., 2016;Mugume et al., 2018).IDW was selected because of its accurate consistency in interpolating point data using the mean squared error as the main criterion (Ganawa et al., 2011).

Physio-chemical soil properties
Results of physio-chemical soil properties of both Serere and Buyende varied considerably (Tables 1 to 4).According to soil textural class, Serere soils were generally sandy-clay-loam (52.2%), sandy-loam (21.7%), loam (17.4%) and only 8.6% for both clay-loam and loamy-sand (Table 1).Additionally, the textural content was more of sandy (55.3%) compared to clay (21.7%) and silt (23%), although high variations were observed in the silt content (CV>35%) (Table 1).In Table 2 for example, one site possessed recommended ranges of total nitrogen which are moderately variable (CV=34.7%).While three sites registered recommended ranges of Phosphorous though this component exhibited higher levels of variations across the study sites with CV=264% much greater 35%.For Potassium, six sites had recommended value ranges though with high variations across other sites (CV=68.1%),only two sites have values that suited recommended ranges of soil organic matter, though this parameter was too highly variable across all the sites.Six sites, possessed recommended ranges of pH, however, across the all sampling sites, this parameter registered very low levels of variability (CV=9.9%).For Buyende district, the soil textural class were generally sandy-clay-loam (47.2%), sandy-loam (41.2%) and only 11.6% for both sandy clay and loamy-sand (Table 3).For textural content, Buyende soils contain more sand (67.1%) compared to clay (23%) and silt (9.9%), although high variations were observed in the clay content (CV>35%) (Table 4).One site possessed recommended ranges of total nitrogen which are highly variable (CV=85.8%).Five sites registered recommended ranges of Phosphorous that exhibited higher levels of variations across the study sites with CV=163.3%much greater 35%.For Potassium, seven sites had recommended value ranges though with high variations across other sites (CV=82%), only one sites have values that suited recommended ranges of soil organic matter, though this parameter was too highly variable across all the sites.Nine sites, possessed recommended ranges of pH, however, across the all sampling sites, this parameter registered very low levels of variability (CV=15.5%).Mengel et al. (1987) revealed that, areas with a high percentage of clay and silt, are recommended for agricultural practices because they provide good aeration and retention, and thus supply nutrients and water, with the above results, one can conclude that the soils of Buyende are not good enough as compared to their counterpart (Serere district) to support agricultural practices if practices like conservation agriculture among others are not implemented to improve on the levels of the required soil nutrients.This argument is supported by earlier studies (Mubiru et al., 2017;Nimusiima et al., 2018;Bamutaze et al., 2021) whose findings indicate that various land use system affects variations in surface soil physio-chemical properties and if done in a sustainable way, they can significantly improve soil fertility status.

Soil structural stability index
Table 5 and Figure 3 depict a greater percentage of the soils of both districts being structurally degraded across the sampling sites, Moreso in Buyende district compared to Serere district which showed some few spots of low risk to soil degradation.It should be noted that these scenarios are as a result of the deficiency in soil organic matter, with more spots of silt and clay content particularly in Buyende district.

Spatial distribution of physio-chemical soil parameters
Results from spatial analyses (Figures 4 and 5) indicate great variations in the different soil physio-chemical properties.It can be observed that much of total nitrogen content is found in the sub-counties of Kidera, Kasilo, Serere TC, Bugondo, Kyere and Bugaya, while Buyende TC and Kagulu have relative low values of total nitrogen content.For available Phosphorus, more was observed in the sub-counties of Bugaya, Labori and Kasiro TC and low values of available Phosphorus was found in Kidera, Kyerere and Kateta.Higher values of Potassium were found in Labori, Bugaya, Kagulu,Kasilo TC and Buyende TC; Kidera,Nkodo,Kateta and Atiira generally had lower values of Potassium content.Labori, Bugondo and Bugaya registered higher values of soil organic matter as compared to Kagulu, Nkondo and Buyende TC.High pH values were observed in Bugaya,Labori and Buyende TC.For clay content, more was found in Labori, Buyende TC, Bugaya, Kateta and Pingire.However, the analysis further indicated that Buyende district was sandier while Serere district was siltier.Generally, these results depict a mixed conclusion about the presence and absence of the different physio-chemical properties in the two districts (Wang et al., 2021), with high and low parameter values not dictated by one sub-county.

Conclusions
Soil physio-chemical results from the two districts (Buyende and Serere) varied, though the two districts are within the same major agro-ecological zone (Kyoga Plains).Sandy-clay-loam was identified as the major textural class across the two districts, followed by sandyloam.Variations of soil physiochemical property were more pronounced in Buyende district compared to Serere district.SSI results showed that soils across the two districts were structurally degraded though this scenario was more emphasized in Buyende district compared to Serere district which showed some few spots of low risk to soil degradation.Additionally, great variations in physio-chemical properties were spatially observed across the two study districts thus portraying a mixed conclusion about the presence and absence of the different physio-chemical parameters in the two districts with high and low parameter values not dictated by one sub-county.This means agricultural practices such as conservation agriculture among others must be implemented to improve on the levels of the required soil nutrients.
The recommendations are based on the status of the soils in both study districts.The study recommends the adoption of climate-smart agricultural practices including application of fertilizers like green composite and farm yard manure which are proven to be cheap and environmentally friendly; avoiding burning or removing crop residues from the fields after harvesting; burning of destructive weeds and testing the soil fertility status.In addition, the government should support farmers to growing crops that perform well such as sorghum and cassava in Serere and maize, beans, and cassava in Buyende districts respectively in terms of subsidized farm inputs.Lastly, long-term studies are needed to establish the effects of land use systems on levels of physiochemical soil parameters.

Figure 1 .
Figure 1.Shows the location of the two study districts in accordance with the agro-ecological zones.Source: Adopted from National Adaptation Plan for the Agricultural Sector, MAAIF, November 2018.

Figure 2 .
Figure 2. Map drawn in ArcGIS 10.2 showing soil sampled points in the sub-counties of Buyende and Serere districts.

Figure 3 .
Figure 3. Frequencies of soil structured stability index for Buyende and Serere.

Table 1 .
Observed soil physiochemical properties from Serere district.

Table 2 .
Descriptive statistics of the physio-chemical properties of soils in Serere district.

Table 3 .
Observed soil physiochemical properties from Buyende district.

Table 4 .
Descriptive statistics of the physio-chemical properties of soils in Buyende district.

Table 5 .
Soil structural stability index for Serere and Buyende.