Identifying recommendation domains for scaling improved crop varieties in Tanzania Dr. Francis Kamau Muthoni Dr. Haroon Sseguya Prof. Bekunda Mateete Dr. Irmgard Hoeschle-Zeledon
Outline Introduction Background information on IITA & Africa RISING program What are recommendation domains? GIS method for generating sustainable recommendation domains (SRDs) Benefits of the generated SRDs Limitations/challenges Going forward
Background of IITA One of world's leading not-for-profit research partners in finding solutions for hunger, malnutrition and poverty
Africa Research In Sustainable Intensification for the Next Generation (Africa RISING) Goal: Create opportunities for smallholder HHs to move out of hunger and poverty through sustainably intensified farming systems that improve food, nutrition and income security especially for women and children Sustainable intensification (SI) = Increased productivity + conserve or enhance environment
Africa RISING Program Area [Africa map] [ESA Demos]
CBOs consortium We know how to work with farmers! A baby is born: Africa RISING-NAFAKA Partnership Our research has generated SI technologies to transform agriculture! CGIAR Centres ARIs I’ll fund you to partner in scaling best-bet SI technologies to 47K HHs, 58K hectares & increase yields by 50% in 3 years! USAID - FtF
Africa RISING-NAFAKA Partnership Project Goal: Accelerate the scaling and delivery of sustainable agricultural intensification technologies to improve smallholder maize and rice farming systems, household nutrition and dietary practices in Tanzania
Africa RISING-NAFAKA Partnership Project Objectives Improved crop varieties Good agronomic practices Natural resources management Household nutrition Post harvest Management
Africa RISING-NAFAKA Partnership Project Targets: 47,000 HHs access technologies to diversify and increase food supply, income sources & improve degrading smallholder croplands 58,000 ha increase in area under improved rice production technologies 50% increase in yields for maize as a result of adopting SI technologies
Africa RISING-NAFAKA Partnership project area
Where do I scale technology a, b, c, …? CGIAR Centres ARIs CBOs What & Why recommendation domains? ? ? ? ? ? ? ? ? ?
Recommendation domains for scaling technologies Recommendation domains are homogeneous areas with similar biophysical and socio-economic characteristics Targeting sites with similar characteristics, and for which a technology is suitable, increases probability of adoption Big Q: How to identify RDs in data limited situations?
Overview of GIS method for generating RDs
Data Collected
Selected Raster data for Tanzania
PCA analysis to reduce dimensionality in data Precipitation- Soil gradient Market- access gradient Elevation- Temperature gradient Stacked gradients
Cluster analysis to delineate RDs All 3 validation indices should be minimized Kmeans best for all indices Optimal clusters = 19 – 20 Clustering method & optimal number of clusters selected
Full Recommendation Domains Map All land cover considered suitable for agriculture
Validation of generated RDS Ideal clusters should be compact, well-separated and stable
The sustainability factor Critical ecosystems masked to generate SRDs
Impact Based Spatial Targeting Index - IBSTI SRD Area (Km2)Population Access -ibilityIBSTI Rank IBGTI FRDsSRDs Cultivated LandTotal Below Poverty Children <5WOCBA SRD (19) SRD (37) SRD (21) SRD (34) SRD (8) SRD (5) SRD (38) SRD (44) SRD (30) SRD (12) SRD (23) SRD (7) SRD (39) SRD (6) SRD (24) SRD (52) SRD (8) SRD (15) SRD (19) SRD (0.1) Total (24) The lower the IBSTI the higher the potential impact SRDs 7, 16 & 8 prioritized for targeting to maximize impact
Potential benefits of generated SRDs Guide formulation of evidence based policies for scaling SI technologies Matching technologies to socio-economic environments to maximize adoption given limited resources Data driven approach is replicable in different ecologies, situations or technologies Reduce subjective judgement in delineation Assessing scenarios under changing situations
Limitations of technology Input rasters generated from other spatial models may introduce error propagation Reasonable spatial/temporal resolutions need to be adopted based on data availability Socio-economic data most limiting though mobile call-logs data improving their generation Facebook to release 5 m resolution population layer Technical knowhow on geospatial analysis required Programming skills to automate the workflow
Going forward the next step Refine SRDs with higher resolution gridded data Field verification of suitability of candidate SI technology packages in ‘unreached’ sections of SRDs
Enhancing partnership among Africa RISING, NAFAKA and TUBORESHE CHAKULA Programs for fast tracking delivery and scaling of agricultural technologies in Tanzania
Africa Research in Sustainable Intensification for the Next Generation africa-rising.net This presentation is licensed for use under the Creative Commons Attribution 4.0 International Licence. Thank You