high start-up costs. These costs may be especially prohibitive where the lands themselves are difficult to work (for example, uplands, steep slopes, poorer soils, soils that have been badly degraded in the process of clearing, and areas overtaken by tenacious weed or grass species). While various sustainable systems and agricultural practices hold great promise in stabilizing and improving these lands, the immediate financial returns may be inadequate to attract farmers and investments. Reform will be particularly difficult when decreases, rather than increases, in the productivity of the land are required. In such cases, alternative employment opportunities are a most probable solution, but it may be necessary to provide direct subsidies to compensate landholders while they allow their properties to stabilize.

Policy devices that have encouraged deforestation in the past—tax abatement, credits, pricing policies, concessions, and subsidies —can be revised to induce small-scale farmers and other landholders to adopt sustainable agricultural practices. Optimally, national development agencies and international aid agencies would work together toward this goal. With a consortia of researchers, NGOs, and other institutional interests, they would identify the lands of greatest need, gauge local community conditions, coordinate appropriate land use and conservation measures, and help provide the financial backing for investment programs.

To attain the most efficient use of limited funds, it will be necessary to determine where natural regeneration is proceeding most acceptably and investments can be delayed or used most sparingly, and where human needs are more pressing and regenerative processes require “boosting.” As regeneration and economic development proceeds, the mix of land use inputs is likely to change and so too will the mix of appropriate incentives. Thus, for example, labor-intensive agroforestry systems that might be highly suitable in low-wage countries may be less financially viable in high-wage countries. Some degree of anticipation of the consequences of changing economic and agroecological conditions is prudent. The necessary initial steps, however, remain clear: provide local farmers and communities in the humid tropics with incentives to improve their current land use practices and restore degraded lands.

The incentives and investments just described will mainly affect lands that are already inhabited but in a degraded state. Special measures must also be taken to rehabilitate completely abandoned