Battling desertification with decision support software

Dr Pierre Sibiry Traore

Sibiry Traore explains how decision support tools and GIS are being used to develop new, targeted agricultural techniques for Africa’s desert margins.

Far from being part of the ‘natural cycle’ of things, desertification is fundamentally about land degradation – the decline in the quality of soil, water or vegetation caused by human activities such as over-cultivation, and climate change.

Desertification affects over a third of the Earth’s surface and more than a billion people. In recent decades there has been a substantial decrease in precipitation in many parts of Africa, but particularly in countries on the margins of the Sahara known as the Sahel.

The soils of the region are sandy, coarse and retain very little water, so that very few plants can grow in them. Such poor, dry soils contain very little biomass and organic matter, and are highly susceptible to wind erosion. With the decreasing vegetation cover, there are fewer plant roots to hold the soil together. It’s a vicious cycle, made worse by the increasing population pressure and poor agricultural practices.

Desert Margins Programme

The Desert Margins Programme, initiated by the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), brings together national scientists from five Sahelian countries, as well as Botswana, Kenya, Namibia, South Africa and Zimbabwe. The programme is working to tackle land degradation in these countries by raising awareness of the problem and by organizing training in the use of decision support software or tools for national scientists.

At the desert margins, soil properties can vary enormously over a small area, requiring different agricultural techniques. For example, farmers need to apply different amounts of fertilizer to millet and other crops, or to sandy soil as compared with a clayey soil.

The interactions between various elements of farming systems are extremely complex. These are: the nutrients applied to soils, the differing rates at which these nutrients break down, the yields of different crops and crop varieties, livestock movements, and geography. Understanding these interactions requires the use of decision support tools. They provide information that is then coupled with a GIS, allowing the development of targeted techniques or cropping systems that can be rolled out in different situations and locations.


Two crop modelling systems that are applicable to West Africa are the Decision Support System for Agrotechnology Transfer (DSSAT) and APSIM.

DSSAT is a suite of programs for PCs that combines soil and weather databases with crop models and application programs that can simulate multi-year outcomes of crop management strategies. DSSAT allows users to ask ‘what if’ questions and produce in minutes results that would otherwise take years of real-world experimentation.

The researchers chose DSSAT rather than APSIM, a similar set of programs, even though APSIM is especially strong on the soil side, because it was produced by Australian research teams for Australian farmers. APSIM’s rigid intellectual property protection means that it would not have been able to access and change the source code.

If we run into a problem and we want to understand what’s not right in the model, we need to access the source code. Very few models have been produced for developing countries and so do not represent the local constraints we have here. Thus it is important to have access to the code to ensure that decision support is adapted to the conditions under which researchers are working. We have tried to convince the developers of APSIM to open up the source code, but we have little leverage. They have to make money out of it. This is a problem for international research institutes, which are mandated to produce non-proprietary work that is accessible to the largest number of people.

With DSSAT, however, we have flexibility in that we can change the code, make recommendations to the development team, who can then incorporate those suggestions into upcoming releases. With DSSAT we can do whatever we want. It’s not quite open source, but all parts of the model are accessible, and it’s cheap – less than US$200 – so even our national partners can afford it.

However, DSSAT is more of a research tool that we can use to develop new technologies or cropping methods. We have found that the target group for this tool is not the farmers, but national scientists. ICRISAT brings together scientists from the programme countries for workshops on DSSAT and other modelling tools. We also organize follow-up workshops, where the scientists work on real problems affecting their own regions. We then work with them to expand their skills in modelling appropriate farming systems, and to enhance capacities at the national level. The next step is to pass on to farmers this information and advice on cropping systems that will arrest or reverse land degradation.

The struggle against desertification requires new agricultural techniques that are applicable to conditions at the desert margins, as well as national capacities to use the decision support software that will help develop such techniques. Software tools need to be more appropriate to ACP countries, and developers need to be more flexible, because conditions in countries such as Niger are rarely comparable with those in Australia or the American Midwest.

Dr Pierre Sibiry Traore ( ) is the head of GIS at ICRISAT. For more information, visit . DSSAT software can be ordered from, .

11 March 2006

Copyright © 2016, CTA. Technical Centre for Agricultural and Rural Cooperation (ACP-EU)