NEARC Fall 2017

AUTHORS: Caroline A. Curtis*, Graduate Program in Organismic and Evolutionary Biology, University of Massachusetts Amherst; Valerie J. Pasquarella, Postdoctoral Fellow, Northeast Climate Science Center, University of Massachusetts Amherst; Bethany A. Bradley, Department of Environmental Conservation, University of Massachusetts Amherst

ABSTRACT: Non-native pines are established in plantations throughout the southern hemisphere and, in many areas, have escaped and spread into the surrounding ecosystem. These invasive populations alter ecosystem function and threaten native species. In Chile, some studies have quantified land cover change relative to non-native pines and have generally found a decrease in native forest and an increase in non-native pine since the 1980’s. However, in measuring land cover change, these studies fail to differentiate increasing plantation size and invaded areas, which is an important distinction for conservation and management of native ecosystems in Chile. The large spatial extent and high detectability of non-native pines provides a unique opportunity to apply remotely sensed data to quantify current land cover and model land cover change through time. I downloaded all high quality Landsat images for scenes in Chile in which invasion occurred. For each scene, I created reference data and identified invaded areas based on Google Earth images. I used time series models to quantify spatiotemporal patterns of land cover change and the Random Forest algorithm to classify images and create land cover maps. By separating plantations from invaded areas, I can better understand how non-native pines impact native ecosystems and how dispersal proceeds across the landscape.