Characterizing differences in existing vegetation driven by natural variation versus disturbance legacies could become a critical component of applied forest management practice with important implications for monitoring ecologic succession and eco-hydrological interactions within the critical zone. Here we characterize variations in aerial LiDAR derived forest structure at individual tree scale in Arizona and New Mexico. Differences in structure result from both topographic and climatological variations and from natural and human related disturbances. We chose a priori undisturbed and disturbed sites that included preservation, development, logging and wildfire as exemplars. We compare two topographic indices, the topographic position index (TPI) and topographic wetness index (TWI), to two local indicators of spatial association (LISA): the Getis-Ord Gi and Anselin’s Moran I. We found TPI and TWI correlate well to positive z-scores (tall trees in tall neighborhoods) in undisturbed areas and that disturbed areas are clearly defined by negative z-scores, in some cases better than what is visible from traditional orthophotography and existing GIS maps. These LISA methods also serve as a robust technique for creating like-clustered stands, i.e. common stands used in forest inventory monitoring. This research provides a significant advancement in the ability to (1) quantity variation in forest structure across topographically complex landscapes, (2) identify and map previously unrecorded disturbance locations, and (3) quantify the different impacts of disturbance within the perimeter of a stand or event at ecologically relevant scale.
Swetnam T., Lynch A., Falk D., Yool S., Guertin D. (2014): Discriminating Natural Variation from Legacies of Disturbance in Semi-Arid Forests, Southwestern USA. Abstract GC33D-0543 presented at 2014 Fall Meeting, AGU, San Francisco, CA, 15-19 Dec..
This Paper/Book acknowledges NSF CZO grant support.