Little is known about how cropping systems influence soil quality and fertility in Uganda. Some cropping systems are more valued and as a result are given more nutrients and planted in certain soils, all of which leads to varying soil quality and fertility. This study compared soil quality (soil pH, cation exchange capacity (CEC), electric conductivity (EC), total N, and depth to restrictive layer (DRL)) and fertility (extractable P, K, Ca, Mg, and Na, and base saturation (BS) from five cropping systems (banana (Musa × paradisiaca L.)-dominant (B), coffee [Coffea robusta (L.) Linden]-dominant (C), banana-coffee (BC), annual with no crop rotation (ANR), and annual with crop rotation (AR); fertilized and unfertilized soils; and three soil types (black (Phaeozem), red (Ferralsol), and black-stony) in south-central Uganda. The analysis included farm assessments to establish management history of studied fields and soil sampling from 52 fields in Masaka District, Uganda. Main-effects ANOVA was employed to determine differences in means in soil under different cropping systems, soil types, and fertilizer use. Soil quality (pH at depths of 0 to 10 and 20 to 30 cm, CEC, and EC) and fertility (extractable Ca and Mg) varied by cropping system. The AR and B systems had higher soil quality and fertility compared to other cropping systems. Soil quality (pH at depths of 0 to 10 and 0 to 15 cm and DRL) and soil fertility (extractable P and K) varied by soil type. Black and black-stony soils had higher soil quality and fertility than red soils. Soil quality and fertility did not vary by fertilizer use. The results of this study indicate that both cropping system and soil type are associated with soil quality and fertility in south-central Uganda.
East African soils exhibit poor quality characteristics that are attributable to their geological age, climate, and land use. In Uganda, the most common soil type is Ferralsol (FAO, 2009), which is depleted in nutrients, highly weathered and comparatively infertile soil. The fertility of East African soils is further degraded by anthropogenic activities, primarily through agriculture. As the majority of Ugandan population depends on agriculture, looking at agricultural activities and their impact on soil is important. Agricultural use changes soil properties through cropping and soil management. In Uganda, cropping systems are characterized by intercropping annual and perennial crops. The two most common crops in south-central Uganda are Robusta coffee [Coffea robusta (L.) Linden] and banana (Musa × paradisiaca L.), both of which are perennial and often intercropped (Okonya et al., 2013). Soil management practices are insufficient to maintain or improve soil quality and can include application of organic and inorganic fertilizers; practicing rotations , or fallowing ; and installing trenches for soil and water conservation (Nkonya, 2002). The dynamics between cropping systems, soil management, and soil types need to be studied to better understand their influence on soil quality and fertility. Several studies
have found that cropping systems have an effect on soil quality. Intensive monocropping of banana disturbed soil’s biotic structure, negatively impacted microbial respiration and water content at field capacity of Andosols and Nitisols
in French West Indies (Clermont-Dauphin et al., 2004).) Annual crops
were predicted to have the greatest erosion rates (93 tons of soil ha-1
) followed by rangelands
(52 tons of soil ha-1
-coffee (47 tons of soil ha-1
), and banana alone (32 tons of soil ha-1
) in central Uganda (Lufafa et al., 2003).
Cropping system influenced culturable rhizosphere bacterial community structure irrespective of plant species in West African soils (Alvey et al., 2003). Agricultural inputs can positively impact soil properties, moreover, coffee agroforestry systems had greater soil organic carbon than coffee monocrops in Ferrallitic soils in Uganda (Tumwebaze and Byakagaba, 2016). Low-input subsistence farming caused serious N depletion in Kenya (De Jager et al., 2001). Agricultural inputs can positively impact soil properties, especially organic and synthetic fertilizers. Frequent fertilizer use increased concentrations of exchangeable K and P in fruit and rubber tree plantation compared to the plantations with no fertilizer use in China (Zhang and Zhang, 2005). Straw application in Niger led to an increase in base saturation and pH and a decrease in extractable Al (Kretzschmar et al., 1991). Green manuring improved organic matter and soil microbial activity in the tropics (Chander et al., 1997). Application of banana stalks, field crop residues, and cattle manure increased banana yields in central Uganda (Bekunda and Woomer, 1996). More research is needed on the effects of cropping systems on soil quality and fertility in south-central Uganda. Additionally, farmer practices need to be included in the analysis. This study looked at soil quality and fertility parameters and their variation by cropping system, fertilizer use, and soil type in south-central Uganda.
The study site was located in Masaka District, near Lake Victoria in south-central Uganda. The district covers an area of 1,603 km2
, half of which is wetlands, with an average altitude of 1,150 m above sea level. The
under a banana-coffee agroecological zone. Banana production has been on-going for 1000 to 1500 years (Lejju et al., 2006) while native Robusta coffee was developed as a plantation crop around 1900s (Thomas, 1947). A favorable equatorial climate with two rainy seasons per year has allowed intensified banana production without crop rotation for millennia (Lejju et al., 2006). However, due to population increase (152 to 248 people per km2
from 1999 to 2012) and consequent pressure on land resources, soil fertility has
been deteriorating (Sebukyu and Mosango, 2012). The banana-coffee cropping history of Masaka District and its declining soil fertility make
it a good area to study why and how soils are declining.
assessments, and soil sampling were conducted from June to September 2016 in Masaka District covering six sub-counties
(Bukakata, Mukungwe, Buwunga, Kabonera, Kyanamukaka, Kyesiga) and one division
(Katwe-Butego). In total, 52 smallholder farms were assessed representing 42
villages. Figure 1 shows the location of the sampled villages in Masaka District. The study was designed to examine field-level soil quality and fertility under annual and perennial cropping systems in Masaka District, Uganda. Multi-stage, purposive sampling method was used to identify farms. First, one to two villages
was randomly chosen from each of 26 parishes, making 42 villages in total. Then, one to two farms in each village were identified from either farmer training records kept by local extension services or by village leaders. One field was chosen per farm for assessment based on a cropping system. Farm
assessments included taking soil samples; interviewing farmers on soil management practices and history of the assessed field; and researcher observations of the soil, location, and crops grown. All farm assessments were performed by the same two people to ensure comparability of the results across fields. According to farmer recalls, field age ranged from one to 100 years of cropping with a mean of 28 years. The majority of fields (n=30)
has been in agricultural production between one and twenty years following removal of bush or native forest. The number of crops per field ranged from one to five with a mean of 2.6 crops
per field. Major crops included coffee
, banana, common bean (Phaseolus vulgaris
), and maize (Zea mays
L.). The majority of fields were intercropped while 13 fields were monocropped. All fields fell into one of the three major local soil types with black (Liddugavu, Phaeozems), red (Limyufumyufu, Ferralsols), and black-stony types (Luyinjayinja), representing 19, 25, and 8 fields, respectively.
Cropping and soil management
The study investigated five cropping systems: banana-dominant (B), coffee-dominant (C), banana-coffee (BC), annual with no crop rotation (ANR), and annual with crop rotation (AR). The B system had banana as the main crop, which was either mono-cropped or intercropped with one or more annual crops such as beans, maize, and cassava (Manihot esculenta Crantz). The C system had coffee as the main crop, which was either monocropped or intercropped with one or more annual crops such as beans, maize, and cassava. BC system had banana and coffee as two main crops, which could be intercropped with one or more annual crops such as beans, maize, and cassava. The ANR system had only annual crops such as maize, beans, and cassava, which could be mono-cropped or intercropped. The AR system also consisted of annual crops such as maize, beans, and cassava, which could be monocropped or intercropped. Farmers in this system, however, rotated crops from season to season. All farmers were interviewed on crops they grew during the time of the interview and in previous season, and crop rotation. Based on the responses to these questions and field observations, the researcher determined categorization of the cropping system. The analysis included a binary fertilizer use variable (no vs. yes). All farmers were interviewed on any nutrient application to the fields including organic (animal manure, mulch, agricultural residues, green manure, compost) and inorganic fertilizers (diammonium phosphate (DAP), calcium ammonium nitrate (CAN), urea).
The fertilizer use variable, therefore, did not differentiate between organic and inorganic fertilizers. Inorganic fertilizer application rates are too small (gross average rate of 1 kg ha-1) in Uganda to cause any significant changes in soil properties (Nkonya, 2002; Ronner and Giller, 2013). As a result, the fertilizer use variable combined organic (n=26) and inorganic (n=9) nutrient applications. The soil type variable included three levels: black, red, and black-stony. Farmers were asked to classify their soil and based on their responses; which were supplemented with field observations; each field was characterized as either under black, red, or black-stony soil. According to FAO-UNESCO soil legend, black soil corresponds to Phaeozems and is generally more fertile than other soil types (Goettsch et al., 2016). Red soil corresponds to Ferralsols (Goettsch et al., 2017) and is strongly weathered. Red soil forms more than 70% of the soil on which most of the farming is practiced in Uganda (Wortmann and Kaizzi, 1998). Black-stony soil is shallow, characterized by plinthitic and quartzitic stones, and is located on hilltops or outcrops (Mulumba, 2004).
Soil sampling and analysis
Fifteen soil properties were examined, including pH at different depths ( 0 to 10, 0 to 15, 10 to 20, 20 to 30, 30 to 50 cm), cation exchange capacity (CEC), electrical conductivity (EC), total N, extractable P, K, Na, Ca, Mg, base saturation (BS), and depth to restrictive layer (DRL).The CEC, EC, total N and extractable P, K, Na, Ca, and Mg were determined at depth of 0 to15 cm. Soil pH and EC were measured using the potentiometric method with soil to water ratio of 1:2. Soil CEC was estimated based on the quantities of Ca2+, Mg2+, and K+ extracted by the Mehlich-3 test (Ross and Kettering, 2011). Total N was measured by Kjeldahl digestion with sulphuric acid and selenium as a catalyst. Extractable P, K, Na, Ca, and Mg were measured by Mehlich-3 test (Mehlich, 1984). The BS was calculated based on the concentrations of Mg, K, Ca and Na. DRL was measured in the center of each field by digging vertically with a shovel until it was physically impossible to continue. Most often, the restrictive layer was characterized by parent material. All soil parameters were separated into two categories: soil quality and soil fertility. Soil quality included soil pH, CEC, EC, total N, and DRL. These parameters represent intrinsic soil properties that are generally slow to change. Soil fertility included extractableP, K, Ca, Mg, Na, and BS; Na is not a nutrient but it can indicate soil quality problems if high levels are found. These soil properties represent a dynamic state or health of a soil that reflects its condition under a specific management systems (Karlen et al., 1997).
Analysis of variance (ANOVA) was performed in R to examine the main effects of cropping system, fertilizer use, soil type on soil quality and fertility (RStudio Team, 2015). Following significant F- test, means were compared using Tukey’s Studentized Range Test at P ≤ 0.1. Analyses was performed on natural log-transformed EC, P, K, Ca, and Mg concentrations, which were back transformed for presentation to readers. Pearson correlations and simple linear regressions (RStudio Team, 2015) were included for better understanding of the relationships among soil parameters.
Cropping systems and soil quality and fertility
Soil quality and fertility varied by cropping system. Soil pH at depths 0 to 10 cm cm, CEC and 20 to 30 EC are varied by cropping system (Table 1). Soil pH at depths of 0 to 10 cm and 20 to 30 cm, CEC, and EC varied by cropping system. The AR had significantly greater soil pH at depth 0 to 10 cm compared to ANR and C systems. Soil pH at depth 20 to 30 cm was the greatest in B systems followed by AR, BC, ANR, and C systems. The AR, B, and BC had significantly greater soil pH at 20 to 30 cm depth compared to the C. The CEC level was the greatest in AR systems followed by B, ANR, BC, and C systems. The AR had significantly higher CEC concentration compared to the rest of the systems. The B had significantly higher CEC concentration compared to C systems. EC level was the greatest in AR systems followed by B, BC, ANR, and C systems. The AR, B, and BC systems had significantly higher EC concentration compared to the C. The C system had the lowest pH at depths 0 to 15, 10 to 20, and 30 to 50 cm. All systems had similar total N concentrations, ranging from 0.14 and 0.15 mg kg-1. The C system had the greatest depth to restrictive layer followed by BC, B, ANR, and AR systems. Soil fertility varied by cropping systems. Such soil fertility parameters as extractable Ca and Mg were significant (Table 2). The extractable Ca concentration was the greatest in AR systems followed by B, BC, ANR, and C systems. The Mehlich-3 Mg concentration was the greatest in AR systems followed by B, BC, ANR, and C systems. The AR systems had the greatest extractable P concentrations followed by B, BC, C and ANR systems. BC systems had the greatest extractable K followed by B, ANR, AR, and C while ANR and AR systems had higher Na concentration compared to B, C, and BC systems. The AR system had highest BS followed by B, BC, ANR, and C systems.
Soil types, quality and fertility
Soil quality and fertility varied by soil type. Such soil quality parameters as pH at depths 0 to 10 and 0 to 15 cm, and DRL were significant (Table 1). Black-stony soil had significantly greater pH at depth of 0 to 10 cm compared to red. Black soil also had significantly higher pH at depth 0 to 15 cm compared to red soil. Black-story and black soils had similar and higher pH at all depths compared to the red soil type. Black-stony soils had the shortest depth to restrictive layer (57 cm) followed by red (65 cm) and black soil types (70 cm). Black-stony and black soils also had higher and similar CEC concentrations compared to the red soil type. Black-stony soil had higher total N concentration compared to red and black soil types. Soil fertility varied by soil type. Such soil fertility parameters such as extractable P and K were significant. Black soil had the greatest concentration of P followed by black-stony and red. Black-stony soil had the greatest concentration of K followed by black and red soils with red soil having significantly lower K concentration compared to black-stony. Black and black-stony soil types had the greatest and similar concentration of Ca, Mg, and BS compared to red soil type. Black and red soils had similar and lower Na concentration compared to black-stony soil. The red soil type was the most frequent and it was primarily under BC and C cropping systems, the black soil type was the second most frequent and it was primarily under BC, B, and AR systems while the black-stony soil type was the least frequent and was under BC and ANR systems.
Fertilizer use and soil quality and fertility
Soil quality and fertility did not vary by fertilizer use. Out of a total of 52 fields, 17 received no fertilizer of any type. Fertilized soil, however, had higher soil pH, CEC, EC, and total N compared to the unfertilized soils (Table 1). Fertilized soils also had highest BS and nutrient concentrations while Na was not different between fertilized and unfertilized soils (Table 2).
Correlations among soil chemical properties
Almost all soil parameters were either highly (r > 0.8) or moderately (r of 0.5 to 0.8) positively correlated with each other at the significance level of P ≤ 0.01 (Table 3). Soil pH was correlated with almost all of the measured soil properties except for P, Na, and soil DRL. This indicates that soil pH is dependent on Ca, Mg, and K. Soil DRL was not correlated with any soil parameter, which could mean that it is influenced by either soil forming processes, landscape position, or erosional-depositional processes (Figure 2).