Indicators may be used for ex-ante or ex-post analysis for the sustainability of land management practices, monitoring soils or erosion processes or impact assessment of various project activities.

• Indicators may be used for ex-ante analysis, e.g. in project planning. The impact of new land management practices on land quality and the sustainability of land use systems can be predicted and compared with the effects of current land use systems. Baseline data of the project region and research results are required for this purpose.

• Indicators are used for ex-post analysis, e.g. project evaluation. Two or more data sets, at the onset and the end of the interventions, are compared.

• Indicators may be used to monitor the performance of new land management practices or the impact of projects, innovative programs and agricultural policies supporting sustainable land use. Time series of data are needed for this purpose. An estimate of the time required to prove enhancement of sustainability is necessary prior to data collection. If it is unlikely that this can be achieved before the expiration of the project, precautions have to be taken that the counterpart organizations will pursue the necessary activities.

The indicators for the ex-ante or ex-post analysis do not have to be identical to the indicators used for monitoring after the project terminates. In the first case, the available information is a limiting factor, whereas additional information can be generated for monitoring. Thus, the indicators can be improved step by step.

Indicators provide assistance in decision making for farmers/pastoralists, project planners/ managers or political decision makers. Repeated collection and processing of data is costly and therefore only justified if the information is relevant for decision making processes. Even data of excellent quality is irrelevant if there is neither the interest nor the capacity to make use of them in decision making. Involvement of users in defining indicators and monitoring changes increases the interest in using indicators. Indicators are used either for baseline studies, e.g. project feasibility studies, project evaluation or for trend analyses, e.g. monitoring or impact assessment

Land use decisions are made on different levels: farm, community, district (corresponding to project, region), national and international. Farmers/pastoralists are the main decision makers, as it is they who decide whether innovations are applied or not. Information derived from long-term monitoring of benchmark sites assists land users in deciding whether or not to adopt innovative land management technologies. Rural development projects operate mostly at district level. Project managers/district governments need information at this level to assist decision making. This could be data on the extent of pasture degradation, deforestation, soil losses, or nutrient depletion under certain cropping systems.

Agricultural policies and programs such as soil conservation programs, fertilizer subsidies, or controlled input supply are formulated at national levels. However, increasingly national policies are being framed and somewhat controlled by international negotiations, as defined by agreements on trade, conventions and treaties. Also, central governments are gradually decentralizing control because of fiscal and other constraints, and this process of decentralization must be considered in developing indicators for practical application. This, also, is reflected by the tendency of bilateral donors to withdraw from central government institutions and to strengthen local institutions.

Often, even if indicators are available they are not applied due to the disproportional costs and effort for monitoring, processing, evaluating, and utilizing the information. When defining indicators and appropriate assessment methods the costs for the required well- trained staff, transport, equipment laboratory and data processing facilities are often neglected or underestimated. However, these costs can be substantial, and it is pertinent to assure that the counterpart organizations will be able to continue monitoring even after the project terminates. Therefore, relevant, easy to handle indicators, and cost efficient monitoring and evaluation methods (e.g. participation of farmers/pastoralists) are needed.

Quantitative indicators are preferred as they are perceived to be simple, clear, accurate and valid. However, reliable quantitative data are rare in most developing countries. Sampling, handling, analyzing and interpreting may be biased, due to inexperienced or poorly trained personal and lack of appropriate equipment and materials. Data collection only makes sense if a certain quality is guaranteed.

Qualitative indicators have the advantage of providing richness and intuitive understanding that numerical data cannot convey. But this assessment may be even more demanding than the assessment of quantitative indicators. In addition they are more difficult to present and appear therefore less accurate than quantitative data.

While the collection of baseline data requires surveys of a limited duration, the monitoring of soil quality or the extent of soil erosion requires continuous observations, analyses, and data processing. Thus monitoring is costly.

The (ecological and socioeconomic) sustainability of new management technologies can be proven only after a period of at least ten years, i.e. often after the expiration of the development project. Thus monitoring or impact assessment methods need to be tailored in a way that allows the counterpart organizations to continue even without the personal and financial assistance of the project. This also refers to data storage. Data needs to be stored in a way that allows access and use for prolonged periods for other interested institutions or donor agencies planning new development projects.

Monitoring land degradation or of the impact of improved land management practices can be done locally (e.g. on reference fields/farms). Participatory land use planning and technology development requires the use of local knowledge for monitoring and the participation of land users data collection.

Indicators based on local knowledge may differ from region to region, depending on the environment, farming systems and traditions. Consequently it may be difficult to compare indicators applied in different regions. Data aggregation from local or regional levels to national or international levels may become complicated or impossible to evaluate. Data aggregation requires standardized information, and often does not consider local peculiarities and local knowledge.

Agriculture is unlike other resource-based industries in that it is made up of millions of small-scale entrepreneurs who make individual decisions on the management of their (natural) resources and on investments of their capital. Although the land use decisions of any individual farmer may seem to be insignificant, these decisions are repeated over and over again in the landscape, and collectively can achieve major regional and even global significance. Agriculture is often cited as being part of the environmental problem, and it is recognised that agricultural land use systems are often significant contributors to non-point pollution and environmental degradation.