Closing the maize yield gap in Africa: The TAMASA-Nigeria experience

Maize is cultivated by approximately 55 million smallholder farmers in sub-Saharan Africa. Farmers’ current maize yields are 50 to 75% lower than attainable yields. The persisting yield gap has been attributed to many biophysical and socioeconomic factors, and are exacerbated by extant weak support systems for wide technology adoption among farmers.

Maize is cultivated by approximately 55 million smallholder farmers in sub-Saharan Africa. Farmers’ current maize yields are 50 to 75% lower than attainable yields. The persisting yield gap has been attributed to many biophysical and socioeconomic factors, and are exacerbated by extant weak support systems for wide technology adoption among farmers. Recent advancements in agronomic research have resulted in the development and use of modern tools and analytics to collect and aggregate large volumes of spatially referenced agronomic, socioeconomic, and biophysical data.

This provides a significant opportunity to transform research into widely useable and scalable knowledge products and services, including Decision Support Tools (DST) for farmers, service providers, industry, policy makers, and investors. Information generated through these tools has immense potential to improve existing practices such as blanket recommendations for fertilizer use or varieties across a diverse range of environments.

The Taking Maize Agronomy to Scale in Africa (TAMASA) project, funded by the Bill & Melinda Gates Foundation, was designed to engage service providers (i.e., input suppliers, government and private research and extension services, agro-dealers, and others) in a co-development process to address maize yield constraints in sub-Saharan Africa. TAMASA leverages on innovative technologies and data analytics to adapt and co-develop scalable DSTs that can be used to assess maize yield and reduce yield gap at scale. The project, which is being implemented in Ethiopia, Nigeria, and Tanzania, seeks to deliver agronomic intervention to some 600,000 maize farming households.

IITA leads the project in Nigeria, working together with the International Maize and Wheat Center (CIMMYT), the International Plant Nutrition Institute (IPNI), African Soil Information Systems (AfSIS), and other stakeholders. TAMASA is being implemented in Kano, Kaduna, and Katsina states. Delineation of target geography was based on geospatially explicit analyses that incorporated critical factors for maize intensification such as existence of major maize production systems, population density, distance to markets, and presence of scaling partners. The project’s activities in the country targets a focal area of about 1.7 million hectares of farmland, within a larger area of interest (AOI) of approximately 14 million hectares.

Assessing stakeholders’ needs and co-developing analysis tools

At the onset of TAMASA, IITA convened several roundtable stakeholders’ needs assessment and follow-up consultations to identify the most critical innovations/tools that are relevant for near real-time improvement of maize agronomy and yield. Stakeholders included input suppliers such as SeedCo and Notore Fertilizer Company, extension service providers such as Sasakawa Global Association (SG2000), State Agricultural Development Programs (ADPs), and input service providers such as Doreo Partners (also Babban Gona) and Propcom–Maikarfi. Through this extensive engagement process, three tools were identified: (1) Nutrient Expert (NE) for Maize for site-specific nutrient management and fertilizer recommendations; (2) Maize Variety Selector (MVS) for location-specific targeting of appropriate maize varieties; and (3) Fertilizer blending for guiding the fertilizer industry on appropriate elements that should be included in fertilizer blends, based on spatially delineated soil conditions.

Strategic partnership for delivery and strengthening of institutions is a key component of TAMASA. Partners are important in the generation of scientific knowledge through their support in establishment of research field trials, co-development of tools, and capacity building of relevant institutions. In Nigeria, the project partnered with the Center for Dryland Agriculture at Bayero University Kano (CDA-BUK), the Institute for Agricultural Research (IAR), and ADPs of focal states.

Establishing baseline yields through surveys

A comprehensive baseline survey was completed in 2015 with a followon Agronomy Panel Survey involving 805 households in 2015 and 2016. The surveys generated uniquely robust and essential spatial data on soil properties, maize yield, and socioeconomic conditions of smallholder farmers within the project’s focal AOI, and are currently being translated into knowledge products. By revisiting the same respondent farming households annually until 2018, TAMASA is generating unprecedented volumes of spatio-temporal data on maize smallholder farm and household characteristics. This information is pivotal for ex-ante and ex-post analysis of technology uptake and impact at scale.

Based on the increasing demand for innovative approaches for yield prediction, yield gap assessment, and evaluation of constraints from local to regional scale, TAMASA is pioneering the innovative application of drones (or unmanned aerial vehicles, UAVs) for in-season yield prediction at five locations within the project’s focal area. By acquiring high-resolution geospatial imageries (12 cm pixel size) with a multispectral-4C sensor carried aboard the UAVs at low altitude (100 m), TAMASA has successfully demonstrated the possibility of linking plot-level agronomic data (including yield) to remotely sensed vegetation characteristics at field scale.

The initial results obtained from multi-level processing of acquired data offer a promising outlook for in-season prediction of maize yield in farmers’ fields. For instance, UAV-derived grain yield predictions, based on a remotely sensed vegetation index, were very close to the measured yield in farmers’ fields at Bunkure, Kano (r2 = 0.87, Fig. 1). Further, the results suggest that innovative application of UAVs in maize-based systems may complement or replace cost-intensive, ground-based, spatiotemporal yield assessment, which are quite impractical to conduct at scale.

Bridging weather data gap

Implementing agronomic intervention at scale requires near real-time knowledge of weather and weather variability. In partnership with Bayero University, Kano and Kukua. b.V. (a Dutch weather service company), IITA 32 successfully facilitated the installation of 40 weather stations across the focal states. These weather stations are providing real-time and remotely accessible weather data to support research-based predictive agronomy and tool co-development. The data can support national-level weather forecasts by the designated national agency (Nigeria Meteorological Service, NiMET), and become a resource for index-based insurers and financial institutions who require hyper-local weather data to determine the type and magnitude of risks that their prospective and existing clients may be exposed to, including farmers and service providers.

Figure 2 Photo
Figure 1. UAV-based assessment of grain yield in maize-based smallholder farmers’ field at Bunkure, Kano.

Developing nutrient expert and maize variety selector tools

TAMASA has engaged relevant stakeholders to co-develop both NE and MVS, which is expected to enhance tool ownership, application, and further improvement beyond the lifespan of the project. The first version of the NE (NE v0) was developed based on experimental data obtained from ~200 nutrient omission trials (NOTs) established in 2015, while the second version (NE v1) is a recalibrated version that incorporated Figure 1. UAV-based assessment of grain yield in maize-based smallholder farmers’ field at Bunkure, Kano. 33 observations from over 100 NOTs experiments which were conducted in 2016 (Fig. 2a and b). As a critical step for rigorous testing of NE value addition, about 200 Performance Trials were established within farmers’ fields across 13 LGAs in 2016. The nutrient recommendations generated from NE were compared with soil-test based nutrient recommendations, current regional fertilizer recommendations for maize, and zero-fertilizer control plots. Results strongly point to higher returns on investment accruing to NE tool application and fertilizer recommendations based on the soil test. NE recommended lower amounts of phosphorus (19 kg/ ha) and potassium (23 kg/ha) fertilizers, yet maize yields were close to those of soil test-based and regional fertilizer recommendations. Using the available blends in Nigeria, this meant a decrease in the use of NPK fertilizer, resulting in an investment saving of about $80 per hectare.

Matching fertilizer blends to local conditions

Soil properties vary widely within the maize production areas of sub-Saharan Africa, yet farmers are mostly offered singular fertilizer blends (such as NPK 15:15:15), which may be inappropriate to local soil fertility conditions and usually translate to huge financial loss to many farmers. Proper soil fertility management for optimal yield requires appropriate fertilizer blends at the right dosage and applied at the right time and at the right place. To this end and through partnership with OCP, a Morocco-based fertilizer company, the project is evaluating maize response to different fertilizer blends across maize ecologies/fertility domains in Nigeria.

Figure 2. Maize grain yield, t/ha (a) and cost of fertilizer input, US$/ha (b) for fertilizer rates generated with nutrient expert (NE), soil-test based (ST) and regional (RR) recommendation methods. CR (control) is the yield for the no fertilizer input plots. Sample size (N) = 177. Note: exchange rate used – US$1 = 500 Naira.

In 2016, the project commenced fertilizer blending activities to facilitate the development of new fertilizer blends appropriate to the Figure 3. OCP trial locations and underlying spatial variability of soil pH and soil carbon within the maize target region of Nigeria. soil conditions of the maize belt of Nigeria. This is expected to result in more efficient and affordable products for increased yields and profit of farmers. It is a collaborative project between OCP-Africa, AfSIS, IITA, and national partners in Nigeria. The activities include soil characterization with an emphasis on identifying soil nutrient limitations, formulation and production of new fertilizer products, and regional scale testing of these new products. The maize belt of Nigeria covers an area of ~24 million ha, and 3000 locations have been surveyed and

Figure 3. OCP trial locations and underlying spatial variability of soil pH and soil carbon within the maize target region of Nigeria.
Researcher inspecting maize crop and cob.

sampled for soil characterization. By design, the sampling locations are spread across 60 sentinel sites, with 10 cluster-grids within each sentinel site, and 5 sample locations per cluster (Fig. 3). The survey also provided unprecedented science-based information on the distribution of crop coverage in the area. Top- and subsoil samples have been analyzed for nutrient concentration, organic carbon, total nitrogen, pH, cation exchange capacity (CEC), texture, mineral composition, and elemental composition. This information provides rich insight into inherent soil characteristics and variability at scale within the maize target region (Fig. 3).

Based on the soil characterization and nutrient need assessments, new fertilizer blends are being proposed to the industry. These blends have higher phosphate (P2 O5 ) content compared to the commonly available fertilizers, and include additions of sulfur, zinc, and boron to address the inadequacy of these nutrients in the soil. Limitations of available potassium seem to be more localized, therefore, composition of K2 O was increased by 22% in one of the new blends. However, potassium was not included in the second blend. The proposed new blends will be tested during the 2017 cropping season at 1500 locations that have been previously sampled and surveyed. Yield response will be compared with the application of commonly used fertilizer blend in Nigeria (i.e., NPK 15-15-15).

Consequently, testing of the new fertilizer blends in larger areas will provide more insight into their performance and generate more information about the variability in yield response, a critical component in farmer adoption. The knowledge products generated from the extensive soil characterization will be shared with other fertilizer companies such as Notore and Indorama to influence them to produce appropriate (region-specific) fertilizer blends that can help improve return on investment for maize smallholder farmers in the region, particularly Nigeria.

ORCID: 0000-0001-5199-5528, 0000-0001-6016-6027, 0000-0002-1844-2574, 0000-0003-4831-2823, 0000-0002-5567-5289

Leave a comment