By Patrick Musinguzi, Lincolin Abasize, Paul Tamale, Tom Matila, Peter Ebanyat, Jackline Bonabana, James Mutegi, and Thomas Oberthür
A farmer-led field study in western Uganda revealed varied rice yield gap reduction with nutrient application to major upland soil types and different responses to the impacts of climate variability due to moisture stress. Soil types are key factors for site-specific nutrient management and will influence the choice of climate-smart adaptation measures within this upland farming community.
In Uganda, the high demand for rice has been mainly met through lowland production domains. However, government policies are encouraging farmers to shift away from growing rice in its fragile lowlands, including wetlands, and instead are targeting more intensive rice production in upland areas (Matovu et al., 2020). Upland rice productivity is typically lower and more variable. Some of the constraints to upland rice production are associated with low soil fertility and moisture stress due to a changing and increasingly variable climate (Alou et al., 2018).
The spatial variability of upland soils complicates the on-farm decision-making process for selecting the right crop management interventions for enhanced rice grain yield and quality. There is a lack of information on rice yield responses across soil types amidst the changing climatic patterns faced by farmers. This uncertainty constrains farmers’ confidence during the selection of more innovative rice cropping strategies.
The development of agronomic and nutrient management information is critical to guide site-specific and climate-smart decisions based on soil type. This research explores rice yield responses, including grain quality, as influenced by improved nutrient applications under varying soil types, physiochemical properties, and moisture retention capacities amidst climate variability and climate change.
Farmer-led trial description
This three-year project was implemented in the Kikuube district located in western Uganda. The study was conducted on farmer’s fields in a sub-county dominated by a bimodal rainfall distribution and mixed cropping system. The experiments were conducted for four seasons within two years with phase 1 during August-December (long rain season) 2022B and March-July (short rain season) 2023A, and phase 2 during August-December 2023B and March to July 2024A.
The study assessed the combined effect of soil type and improved nutrient management under a Good Agronomic and Climate-Smart Management (GACLIM) package for upland rice yield. Three dominant soil types (Ferralsols, Plinthisols, Gleysols) were selected in the study area that represented the varied soil physical and chemical characteristics of rice soils. An improved rice cultivar (NAMCHE 5) was grown across the three soil types. Various rates of N, P and K were applied considering the 4R Nutrient Stewardship Principles (right source, rate, time and place)in combination with climate smart agronomic practices such as seed bed levelling, soil water conservation with bunds, timely planting by dry seeding, timely weeding, and good control of pests.
Initially, two treatment trials were laid out in a randomized complete block design under phase 1 (2022B and 2023A). The two plots measuring 20 m x 50 m included farmer practice (dominated by no fertilizer use) and a recommended nutrient application within a GACLIM package. A total of 60 kg N, 30 kg P, and 80 kg K per ha were split applied with diammonium phosphate (DAP) at planting, and later with urea and potassium chloride (MOP) at tillering and panicle initiation stages. Only P was applied at planting; N was applied in three splits at planting, tillering and panicle initiation in equal proportions (20 kg/ha) while K was applied twice in equal proportions at tillering and panicle initiation (40 kg/ha). The trials were established on 12 farm fields with soil type as the main plot and the two treatments as sub-plots for two seasons. Each of these treatments were replicated four times with farmers’ field applied as replicates.
In the second year (2023B and 2024A seasons), the experiment expanded to four treatments evaluated on 12 farmer fields located on the same three soil types. The recommended NPK application rates in phase 1 were again tested as a full rate, along with a half rate (30 kg N, 15 kg P, 40 kg K per ha); one-quarter rate (15 kg N, 7.5 kg P, 20 kg K per ha), and farmer practice. The proportions for split application as applied in year 1 were followed for N, P and K. This second phase was designed to adjust to concerns of affordability and resource constraints among participating farmers due to the costs associated with fertilizer used in year 1.
Total biomass and grain yield data was determined by using total field data and sub-samples oven dried to compute the moisture correction factor. The grains were later milled to compute the milling recovery for each treatment.
Soil conditions
The three soil types had varied soil chemical and physical properties. The Plinthisol sites had characteristically shallow soils dominated by gravel in the deep layers (>30 cm) while the Ferralsol sites had deeper soils with varied but inherently high P fixation. Gleysol sites were defined as poorly drained soils with some incidences of redoximorphic features [i.e., spots differing in colour from the surrounding soil caused by the oxidation and reduction of iron (Fe) and manganese (Mn) compounds] in the low-lying areas near or within wetlands.
Generally, soil fertility varied across fields (Table 1). All soils had moderate pH <6.0, very low extractable P (<15 ppm), and moderate levels of total N (some below the critical concentration <0.2%) and exchangeable K (<0.4 cmol/kg soil). Soil organic carbon was above 1.72%, a general indicator of good soil health, but varied within each field. Other parameters such as texture varied from loam to sandy loam. Exchangeable Ca, Mg and K varied across soil types and farmer fields, but on average these were above the critical levels.
Yield responses to farmers’ practice and recommended nutrient treatments (Phase I)
The recommended nutrient applications under the GACLIM package resulted in significantly higher yields in seasons 2022B and 2023A (Fig. 1).
Gleysol and Ferralsol sites had higher grain yield compared to Plinthisol sites. Unmilled rice yield, tiller numbers, plant height, panicle number, biomass, and grain weight were strongly influenced by soil type and treatments applied (P<0.001; Fig. 1). Ferralsols sites resulted in highest mean unmilled grain yield (2,972 kg/ha) while Plinthisol sites had the lowest yield (1,986 kg/ha). Ferralsol sites recorded the highest tiller count (6.9 per plant) while Plinthisol sites had the lowest (4.8 per plant).
Rice grain quality and milling recovery
The recommended nutrient application delivered by the GLACLIM package resulted in better grain quality and high grain nutrient content (Table 2). Soil type and nutrient application affected the grain N, grain P, and grain K. There was a notable grain weight increase, and the rice milling recovery was also strongly influenced. The highest milling recovery percentage (59%) was obtained from Ferralsol sites while the lowest (47%) was obtained at Gleysol sites.
Rice response to farmer practice and varied nutrient application rates (Phase II)
In seasons 2023B and 2024A, the application of the GACLIM package with varied nutrient applications rates (full recommendation, half rate, one-quarter rate) resulted in significantly different yields compared to the farmer practice (Fig. 2). The increase in grain yield was notable for full and half rate applications across all the three soil types (p<0.001), but not for the one-quarter rates or farmer practice. The full rates recorded the highest grain yield (2,809 kg/ha) and number of panicles (4.72 per plant) whereas farmer practice generated the lowest yield (2,039 kg/ha) and number of panicles (3.49 per plant). Among soil types, Gleysol sites recorded the highest (2,535 kg/ha) grain yield while Plinthosol sites recorded the lowest (2,372 kg/ha).
Application of the full GACLIM nutrient recommendation resulted in a higher recovery of grain (69.1%) after milling compared to the half (66.5%) or one-quarter (61.4%) rates. Farmer practice resulted in the lowest grain recovery (56.8%). Soil type did not significantly influence the milling recovery of grain.
Seasonal effects and yield response to climate-smart management
The use of climate-smart practices provided an added advantage during both the phase 1 and 2 short rain seasons (March to July 2023) and (March to July 2024). There was a notable impact of soil type on yield performance as Gleysol produced higher yields compared to Plinthisol or Ferralsols with no management, but crop performance improved with better management (Fig. 3). For first short season of 2024 (Season 202A), rice growth was registered with variations across soil types but did not reach the physiological maturity due to severe drought. For 2023A, low grain yield was also registered due to limited rainfall distribution. Overall, the short seasons (first season of a year) did not favor the introduced NAMCHE 5 variety, but early maturing rice variety could be an option for such short seasons in the area.
Conclusion
Upland rice production is significantly affected by soil type and climate-smart nutrient management regimes. The Ferralsol sites were the most consistently productive between 2022 and 2024 with the highest yield and grain recovery after milling. Interventions with lower nutrient input registered yields that were comparable to the full nutrient application package suggesting that lower nutrient rates could be economically viable and affordable for farmers. There is potential for targeting soil-specific nutrient management in upland rice production for high milling recoveries and increased climate change adaptation.
Acknowledgement
This research was made possible with funding from APNI from the African Plant Nutrition Research Fund (APNRF). The rice project aimed at improving crop productivity through proper rice nutrition, sound agronomic and climate-smart agricultural options with a market responsive business strategy. Special thanks to the field teams that have supported field experiments, data collection, laboratory analyses and statistical analysis.
Dr. Musinguzi (e-mail: musipato7@gmail.com), Ms. Abasize, Mr. Tamale, Mr. Matila, Dr. Ebanyat are with the Department of Soil Science and Land Use Management, College of Agricultural and Environmental Sciences, Makerere University, Kampala, Uganda. Dr. Mutegi is APNI Senior Scientist, Nairobi, Kenya. Dr. Oberthür, APNI Director Business & Partnerships, Benguérir, Morocco.
Cite this article
Musinguzi, P., Abasize, L., Tamale, P., Matila, T., Ebanyat, P., Bonabana J., Mutegi, J., Oberthür, T. 2025. Lessons from On-farm Experimentation on Climate-Smart Management for Upland Rice in Uganda. Growing Africa 4(1):5-9. https://doi.org/10.55693/ga41.WKDW5707
REFERENCES
Alou, I.N., et al. 2018. Growth, phenological, and yield response of upland rice (Oryza sativa L. cv. Nerica 4®) to water stress during different growth stages. Agric. Water Manage. 198:39-52.
Kaizzi, K.C., et al. 2014. Upland rice response to nutrient application in Uganda. Afr. J. Plant Sci. 89:416–425.
Matovu, B. et al. 2024. I. Navigating through Complexity by Profiling the Main Threats to Sustainable Tropical Wetlands Management and Governance: A Case Study of Mityana District, Uganda. Discov. Environ. 2, 18.
