Padrões históricos e atuais do uso do solo em bacias de drenagem e sua relação com a proximidade de lagos e reservatórios

Abstract

Land-use change is a major driver of terrestrial and aquatic ecosystem transformations, affecting carbon fluxes, biodiversity, and ecosystem services worldwide. Understanding catchment land-use dynamics is critical, as these areas are shaped by several socioenvironmental factors. However, whether and how the aquatic ecosystem type (natural lake or reservoir) and the distance from its shores influence the catchment land-use dynamics remains limited. This study investigates the spatiotemporal land-use dynamics in 98 tropical catchments, encompassing natural lakes and reservoirs in northeast Brazil. For this, we used the MapBiomas Land-Use/Land-Cover Collection (1985–2022). The land-use dynamics (LUd) were depicted by a composite variable derived from principal component analysis (PCA) scores, which incorporated the current relative extent of land use, the land-use change rate, and long-term land-use trends. We hypothesized that: i) LUd is greater in reservoir catchments, ii) LUd increases with greater distance from the shore in both lake and reservoir catchments, and iii) the distance-mediated increase in LUd is stronger in reservoir catchments. Using regression tree analysis (RTA), we explore how morphometric and geomorphological characteristics of aquatic environments and their catchments, as well as demographic and climatic characteristics, influence the spatiotemporal land-use index (SLUi), calculated as the effect sizes (Fisher’s z) of LUd as a function of distance from the shore. Our results reveal that LUd is more intense in lake catchments and differs significantly between lake and reservoir catchments: LUd increases with distance from lakes but decreases with distance from reservoirs. This suggests that areas closer to lakes are more effectively protected than those further away, whereas in reservoir catchments, headwater regions are less impacted than areas closer to the reservoir shores. Human population size, slope, longitude, waterbody area, and catchment area are the key landuse predictors in the studied catchments. These findings provide new insights into the mechanisms driving land-use dynamics in catchments of tropical ecosystems, highlighting the importance of considering ecosystem type as a predictor of catchment land-use dynamics.