FTA Research  ·  Minimal Irrigation Project  ·  Madagascar
EN | FR
HomeOur Work › Our Methodological Approach
How We Work

Our Methodological Approach

A rigorous four-phase approach, grounded in the field and statistically validated across 10 sites.

How We Work

Our Methodological Approach

The MIP does not merely observe: it experiments, co-constructs, and measures. A rigorous four-phase approach ensures that every result is reliable, reproducible, directly transferable, and understandable to farmers..

01

Fundamental Research

Analysis of crop water requirements, physiological responses to water stress, and efficiency thresholds for each cultivated species.

02

Technological Development

Design of low-cost localized irrigation systems and development of simple decision-support tools.

03

Multi-site Experimentation

Comparative evaluation across various environments using a rigorous factorial design (randomized complete blocks, 3 to 4 or more replicates).

04

Participatory Action Research

Joint development of protocols with farmers, social validation through local research technicians, and knowledge transfer via extension sessions.

What we measure in the field
Data typeKey indicators & measurements
🌿 AgronomicCommercial yields (kg/ha), yield components, aerial and root biomass, phenology and plant growth
💧 HydricSoil moisture (capacitive sensors), AET, irrigation volumes, water use efficiency (Water Productivity)
🏔 PedologicalPhysico-chemical properties, infiltration rates, water retention capacity, soil fertility and organic matter
📊 Socio-economicProduction costs, gross and net margins, labour time, social acceptability and technology adoption rates
DailyWeather & irrigation monitoring
WeeklyPhenological monitoring
3–4Replicates per treatment
9 sitesGxE multi-site analysis
🎓

Innovative Outreach Tools

Digital database, demonstration videos in Malagasy, detailed technical specifications, illustrated technical sheets and guides in comic-strip format, and training modules for farmer field schools. Statistical analyses are based on mixed models that account for the site > block > plot hierarchy.