Dr. Keith D Shepherd

In my view, any policy for improved soil fertility management must have the below ingredients to ensure efficiency and reliable learning.

  1. A systematic programme to properly diagnose soil fertility constraints and their associated risk factors spatially at different scales, using statistically valid sampling schemes. We have the technology to do this cost-effectively now. Participatory diagnosis by land users/communities is important but not a substitute for scientifically sound objective assessments. There is need for interaction among both types of systems.
  2. A systematic programme for testing soil fertility management options using standardized protocols and linked to the baseline above (no. 1) to provide evidence-based recommendations. Again this is required to complement and inform farmers testing strategies.
  3. Baselines and monitoring of soil fertility in soil management/development projects so impacts of interventions can be reliably assessed. Again no.1 above provides a method for doing this.

This evidence base is needed to inform decision making at all levels: individual farmers, communities, stockists, fertilizer/seed companies, land resource managers, national research and extension, government planning and finance ministries, donors, development agencies, etc. We have the technology to do this – we just need good design and systematic application. The types of systems I am describing are surveillance systems similar to those used in the public health sector – which indeed primarily guide public policy and practice.

Dr. Keith D Shepherd, Principal Soil Scientist
World Agroforestry Centre (ICRAF)
k.shepherd@cgiar.org