Concern has been expressed regarding the longer-term ecological effects of forest clearance in tropical South America. Although indigenous exploitation in Amazonia is generally adaptive to the environment, modern agricultural and livestock developments are often less stable in character, resulting in low productivity exploitation systems and in land deterioration. The terrafirme zone is particularly vulnerable in this respect. In contrast, the vdrzea zone displays a higher potential for the exploitation of both terrestrial and aquatic resources and current land development should focus thereon. A spatial development model is proposed, emphasizing this ecological distinction. IN recent years, concern has been expressed regarding the scale and rapidity of forest clearance in tropical South America (Richards, 1973). Clearance is most widespread in Brazil where vast livestock and timber holdings are being established and where substantial peasant colonization is occurring; similar processes are apparent in lowland areas of Peru, Bolivia and Colombia. Budowski (1972, p. 384) has criticized 'the elimination of an extremely useful type of vegetation cover and its replacement by often terribly degraded lands'. Fosberg (1972, p. 388) suggests that 'there are other far better land-uses for at least some of these tropical forests than chopping them down and converting them into submarginal agricultural land'. Restraint has been urged; Denevan (1973, p. 133), for example, has concluded that 'road building and colonization in Amazonia probably should be restricted for now'. Similarly Sternberg (1973), while recognizing the difficulty of any 'hands-off' policy, urges that at least some extensive tracts of non-agricultural land be saved from degradation. Behind such views is the assumption that, in spite of the abundant unexploited land in Amazonia, existing development is ecologically, and hence economically, inappropriate; in the long run it will diminish regional resources and create land conditions which are difficult and costly to exploit. By implication, it is argued that greater agricultural benefit will derive from improving land of known and proven quality outside the rainforest. However, governments controlling Amazonia are proceeding apace with its development. They may acknowledge the ecological implications of their action (Anon., 1974), but are often thought inadequately responsive to them; as Sternberg (1 973, p. 262) caustically remarks, 'the rules of the game are that environmental concern should never impede development'. And this attitude is comprehensible. Settled areas of Brazil and the Andean states are experiencing serious socio-economic problems, which could be partly alleviated by forest development programmes; cogent political reasons also exist for the occupance of forest land (Rosenbaum & Tyler, 1971; Becker, 1974). In the short term, it is unlikely that the governments concerned will terminate or seriously curtail development programmes in the face of conservationist pressures. Consequently it is essential to ensure that forest development is sufficiently realistic from an ecological, as well as an economic, viewpoint to satisfy the long-term requirements of both conservationist and