Quantification of Drought-Induced Tree Mortality in Texas: Fine scale to Regional Estimates

Amanda M Schwantes, Duke University, amanda.schwantes@duke.edu (Presenter)
Jennifer J Swenson, Duke University, jswenson@duke.edu
Daniel M Johnson, University of Idaho, danjohnson@uidaho.edu
Jean-Christophe Domec, Duke University & Bordeaux Sciences Agro, jc.domec@duke.edu
Robert B Jackson, Duke University & Stanford University, rob.jackson@stanford.edu

During 2011, Texas experienced a severe drought that killed millions of trees across the state. Drought-induced tree mortality can have significant ecological impacts and is expected to increase with climate change. We identify methods to quantify tree mortality in central Texas by using remotely sensed images from before and after the drought at multiple spatial resolutions. Fine-scale tree mortality maps were created by classifying 1-m orthophotos from the National Agriculture Imagery Program. Percent dead and live tree cover was estimated from 30-m Landsat Imagery, using a Random Forest model and the orthophoto classifications as training and testing data. These coarse-scale percent tree mortality maps were then used to identify ecological systems most impacted by the drought as well as climatic and edaphic factors that control spatial distributions of tree mortality across central Texas. Ground observations coupled with our remote sensing analyses revealed that the majority of the mortality was Juniperus ashei and Quercus stellata. From a physiological standpoint this is surprising, because J. ashei is a drought-resistant tree. However, over the last century, this species has recently encroached into many areas previously dominated by grassland, and thus tends to occupy landscape positions that are marginal habitat for the species. We found that temperature anomalies from historical averages best explained these tree mortality patterns; however, other landscape variables describing dry landscape positions (e.g. areas dominated by shallow soils, a low compound topographic index, and a high heat load index) were also significant. As increases in extreme drought events are predicted to occur with climate change, it will become more important to establish methods capable of detecting associated drought-induced tree mortality, to recognize vulnerable ecological systems, and to identify climatic & edaphic factors that control the spatial distributions of tree mortality across a landscape.

Presentation Type:  Poster

Session:  Theme 4: Human influence on global ecosystems   (Mon 4:30 PM)

Associated Project(s): 

• Related Activity: Related Activity or Previously Funded CC&E Activity not listed ...details

Poster Location ID: 88