Geostatistical modeling and conservation implications for an endemic Ipomoea species in the Eastern Brazilian Amazon

摘要

The Canga ecosystem, in the Eastern Brazilian Amazon, is a biodiversity and endemism hotspot as well a potential resource for iron mining, raising conflicts between conservation and economic development. Here, we collected data from 60 randomly selected points in the Canga ecosystem in the Carajas National Forest, Brazil, to analyze the spatial patterns of Ipomoea cavalcantei occurrence in different phytophysiognomies. Through ordinary kriging and a GLM model, we identified that this endemic species occupies a very restricted range within the Canga ecosystem, specifically in areas of Open Rupestrian Fields and Shrub Rupestrian Fields where the iron crust restricts the soil depth to a few centimeters and the vegetation is mainly herbaceous or savannah shrub vegetation, with low sparse tree canopies. These conditions increased the passage of sunlight, revealing the importance of light conditions for the species. However, because the species occurs in suitable mining sites, there is continuing conflict between environmental and economic decisions. Spatial data showed that mining has increased in Carajns National Forest over the last 32 years, reducing the potential locations for I. cavalcantei to less than 50 % of its original range. Mining will potentially occupy the entire species range over the next three decades. Restoration outside its original habitat is difficult in the short term, because there is no knowledge of the species' capacity for growth, survival, and replication. Conservation initiatives for the Canga ecosystem should prioritize strategies that reduce regional dependence on mining and provide alternative economic activities, such as sustainable exploitation of bio-products and ecological tourism.