Settlement Design, Forest Fragmentation, and Landscape Change in Rondônia, Amazônia

Deforestation and colonization in Amazônia have attracted substantial attention. This article focuses on an area of 3,000 km2 within the Brazilian State of Rondônia. Two adjacent settlements were compared to assess the role of their different designs in landscape change. Anari was planned following an orthogonal road network. Machadinho was designed with attention to topography in laying out roads and farm properties, while including communal reserves. Field research was undertaken in conjunction with multi-temporal classifications of remotely sensed data (1988, 1994, and 1998) and landscape ecology methods. The results indicate that large patches of communal reserves play an important role in maintaining lower levels of fragmentation. Analyses of landscape structure confirmed that forest patches in Machadinho are less fragmented, more complex, and preserve more interior habitat. By comparing the effects of different settlement designs on landscape change and forest fragmentation, this article contributes to the debate about colonization strategies in Amazônia. Introduction The interest of naturalists and ecologists in landscape spatial patterns is extensive. This approach, responsible for an ecological perspective about the geographic space, is today represented by landscape ecology (Forman and Godron, 1986; Urban et al., 1987; Turner et al., 1995). Landscapes became objects of scientific analysis and synthesis, being understood as a spatially heterogeneous mosaic to be studied from the reciprocal effects among spatial patterns and ecological processes (Turner, 1989; Pickett and Cadenasso, 1995; Forman, 1997). The human dimension underlying landscape changes has also been highlighted (Naveh and Lieberman, 1994). The study of these relationships confers a practical perspective to landscape ecology, through the establishment of scientific bases for planning, management, conservation, and development of territories (Leser and Rodd, 1991). The applicability of these concepts to spatially explicit ecological studies is clear. In a world where human-altered landscapes are increasingly created, processes of disturbance need to be spatially quantified and understood (Pickett and White, 1985; Baker, 1995). Several methods based on the concept of landscape structure have been developed to address processes of disturbance within landscapes. Landscape metrics have been widely used for this purpose (Baker and Cai, 1992; McGarigal and Marks, 1995; Riitters et al., 1995; Gustafson, 1998), and the integration of spatial data and methods in geographic information systems (GIS) has improved this approach (Turner and Gardner, 1991; Sample, 1994; Haines-Young et al., 1996; Frohn, 1998). One of the most frequent examples of landscape disturbance in the tropics is derived from land-use/land-cover (LULC) change of forested environments (Dale et al., 1993; Dale et al., 1994). Landscape heterogeneity can either increase or decrease, depending on the parameter and spatial scale examined (Krummel et al., 1987). In general, the disturbed landscape has more small forest patches and fewer large, matrix patches than the intact landscape (Mladenoff et al., 1993; Malcolm, 1994). Taking this assertion as a hypothesis, this research examines landscape structure in Rondônia, Brazilian Amazon, through derived LULC classifications. The adjacent settlements Machadinho d’Oeste (Machadinho) and Vale do Anari (Anari) are analyzed in terms of their composition and configuration on a multitemporal basis to assess the role of their different designs in forest fragmentation and landscape change. Anari was planned following an orthogonal road network, a scheme used to colonize the majority of Rondônia. Machadinho was designed with attention to topography in laying out roads and farm properties, and included communal forest reserves. Results for both sites (Figure 1) are compared, trends are described, and methodological issues are discussed. Study Area Rondônia has experienced the highest deforestation rates in the Brazilian Amazon during the past twenty years (INPE, 2002). Following the national strategy of regional occupation and development, colonization projects initiated by the Brazilian government in the 1970s played a major role in this process (Moran, 1984; Schmink and Wood, 1992). Most colonization projects in the state were designed to settle landless migrants. Previous analysis of the settlement strategies implemented in the region have shown the need for multitemporal LULC assessments to understand the history of occupation and the trends for the future (Batistella et al., 2000). Settlement began in this area in the mid-1980s and represents an important type of human impact on the Amazonian landscape. The climate in the study area is classified as equatorial hot and humid. The well-defined dry season lasts from June to August, and the annual average precipitation is PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING Ju l y 2003 805 Photogrammetric Engineering & Remote Sensing Vol. 69, No. 7, July 2003, pp. 805–812. 0099-1112/03/6907–805$3.00/0 © 2003 American Society for Photogrammetry and Remote Sensing M. Batistella and E. F. Moran are with the Anthropological Center for Training and Research on Global Environmental Change and the Center for the Study of Institutions, Population, and Environmental Change, Indiana University, Bloomington, IN 47408. M. Batistella is presently with EMBRAPA Satellite Monitoring, Av. Dr. Júlio Soares de Arruda, 803, 13088-300 Campinas, São Paulo, Brazil ([email protected]). S. Robeson is with the Department of Geography, Indiana University, Bloomington, IN 47405. 02-031.qxd 6/6/03 4:53 PM Page 805