Agroforestry is the integration of trees and shrubs in agricultural areas, like farmland, to help protect the soil and in some cases to provide additional income.

Reduced soil erosion, biodiversity conservation and carbon sequestration are some of the major effects agroforestry systems have on the environment.

Yet many of the areas crucial to world ecology are in developing countries where it is a difficult sell to discourage maximum utilization of available land for crops.

In September, 130 researchers from 25 countries met at the 2nd International Symposium on Multistrata Agroforestry Systems with Perennial Crops, held in Costa Rica* to find ways to promote agroforestry and assess the research results obtained in the eight years since the first symposium. The participants also set out to propose appropriate mechanisms for ensuring that farmers who make the effort to adopt such ecofriendly production systems reap some reward.

A role to play in a rapidly changing environment

Multistrata agroforestry systems associating crops with trees, with the crops developing in the shade provided by the trees, have existed for some considerable time in many tropical countries. For instance, perennial crops such as coffee or cocoa can be grown in the shade of fruit and/or forest trees, making up at least three strata of vegetation.

However, with the advent of intensive agriculture, some 40-50 years ago, farmers often abandoned the practice in favour of monocultures, which generate profits sooner. However, the prospects for this type of intensive agriculture, which is reliant on massive external inputs, are now limited, due to its serious adverse effects on the environment and to various economic constraints.

Furthermore, times are changing, as is the climate, and public awareness is increasing. Consumers are increasingly buying "ethically", "ecologically", "organically" or "fairly", and are thus more likely to pay more to be sure of buying ecofriendly, socially responsible products. This is where the products of agroforestry systems come in.

Lower production counteracted by the environmental services rendered

The main obstacle to the large-scale adoption of such systems has been maintaining farmers' incomes. The environmental services rendered by agroforestry systems are now recognized, but shaded cropping can produce lower yields than intensive monocultures. This is often the case, for instance, with arabica coffee grown at more than 1000 m above sea level.

Given the importance of this constraint, researchers are insisting on the need to maintain incomes. Two main ways of achieving this have been suggested.

The first consists in paying farmers for the environmental services rendered by their crops. This could be done through ecolabels such as "organic" or "Rainforest Alliance", which already exist on the market, through a national system of payment for carbon sequestration or, more locally, through direct payment from users of services, for instance hydroelectric dam operators paying farmers to protect soils against erosion.

The other possibility is to diversify the crops grown and domesticate woody species, in the aim of making such operations profitable by creating or improving wood product marketing channels. For instance, timber serves to improve incomes among coffee producers when coffee prices fall in Central America. Likewise, palm wine plays a major role in ensuring the sustainability of cocoa-growing systems in the forests of Ivory Coast.

Certification: a strategy to be integrated into development programmes

At the end of the symposium, it was also recommended that the issue of biodiversity be associated with that of system productivity in studies and models. In effect, while it has been clearly established that agroforestry systems conserve and/or increase flora and fauna biodiversity, the impact of that biodiversity on the economic sustainability of such systems has yet to be determined.

The beneficial effects of these agroforestry systems on a landscape scale, for instance their role in the connectivity between protected areas, also need to be explained more clearly. As regards existing biophysical results, compiling them in meta-databases would make it possible to organize the available information better, communicate about methodologies and results, apply those results more effectively, and identify research requirements.

A modelling approach, perhaps coupled with an analysis of these metabases, would help to resolve the most crucial issues, such as the relations between crop yields and shading, or the impact of climate change on the functions of and the services rendered by the ecosystem.

Lastly, certification is a strategy that could be integrated into the very heart of development programmes. However, this would mean improving the scientific bases of the biophysical and social criteria and indicators applied when attributing ecolabels.

* The symposium was organized by the Centro Agronómico Tropical de Investigación y Enseñanza (CATIE), CIRAD, the International Center for Research in Agroforestry (ICRAF) and the University of Wales. It followed on from an initial symposium held in February 1999, also at CATIE, in Turrialba, Costa Rica.