1. Presumption of adaptive strategies

There must be some basis for a presumption that adaptive strategies exist that can be identified and described, and that will prove useful for an analysis of routes to sustainable development. All available documentation should be consulted.

2. Coverage of problem

This is the most important consideration. The project aims to cover as wide a range as possible of ecosystem conditions, vulnerabilities (pressures, change, challenges etc.) and adaptive strategies. Typologies and key variables may emerge during the initial scoping stage that can be presumed to be significant for understanding adaptive strategies. For example, in ASALs one of the most significant variables is the agricultural-pastoral continuum, and so it is important to select sites representing various points along that continuum, sufficiently far apart to show major distinctions. For the IISD project in the Borana area in Ethiopia a total of six villages or hamlets were selected, three from a more agricultural neighborhood, and three from a more pastoral one.

3. Representation

The communities selected for field work must be sufficiently typical of a wider situation that they can serve as a meaningful case study. Although every community has its specificities, there are some broad things in common which should be sought:

• internal features, including values, beliefs, practices and local institutions (i.e. the cultural equipment a community brings to bear in managing its ecosystems and livelihoods); demographic variables; local economies; degree of privatization or appropriation of resources; and
• external pressures which we will have to bundle broadly. They might include, for example, extent of market integration, alienation of resources, government regional policy (isolation/ repression/ integration/ investment).

These are just examples. It will be difficult in any one country to draw solid conclusions from two different cases where several possibly independent variables are involved. This is where the literature review and policy issues paper come in - to capture the larger picture.

4. Logistics

Rural research areas are often by nature difficult of access, and people may be widely dispersed. Some logistical challenge is a good thing and helps to offset the danger of "tarmac research". But too ambitious a logistical challenge involving many remote sites in one country, would reduces cost-effectiveness.

5. Social access

The ease with which the research team can gain the trust and confidence of the people is a consideration which affects the selection of the NGO as well as the site. The more generous the time allowed for site selection and overcoming suspicion or hostility, the less restrictive this criterion will become. Previous experience in the area is also helpful.