The Utility of Aerial Photography in Detecting Beech Bark Disease Occurrence and Severity in Wisconsin and Michigan
File(s)
Date
2014-05Author
Karschnik, Travis M.
Publisher
University of Wisconsin-Stevens Point, College of Natural Resources
Metadata
Show full item recordAbstract
Beech bark disease (BBD) is an insect mediated, fungal disease complex that threatens the current health and future of Wisconsin’s American beech forests. The insect component of BBD was first identified in Wisconsin in September, 2009, but the extent and severity of the infestation has not been adequately studied. Michigan forests have experienced high mortality of American beech due to BBD and the complex has spread quickly throughout the state since its identification in 2000. Forest Inventory and Analysis plot data, provided by the USDA Forest Service, has identified significant amounts of American beech in Wisconsin’s forests concentrated on the east coast along Lake Michigan and on the Menominee Indian Reservation. BBD has the potential to have a significant impact on Wisconsin’s beech forests if early detection and management is not implemented.
The two goals of this study were, 1) to measure the biological response of American beech to BBD, and 2) to determine if BBD could be detected by using natural color, 1 and 2 meter resolution aerial orthophotography. The study relied on ground data collected using the Beech Bark Disease Monitoring Impact and Analysis System (BBDMIAS) methodology and aerial imagery provided by the National Agriculture Imagery Program (NAIP).
We found, through a number of non-parametric analyses, that trees with BBD were less healthy than those without BBD. Beech trees with BBD had significantly higher crown transparency and dieback, lower vigor, and more damages compared to trees without BBD. A repeated measures ANOVA indicated no significant difference in red-band reflectance within disease free sites in Wisconsin, whereas a 2-sample t-test
comparing the mean reflectance values of diseased sites to the reflectance values of non-diseased sites indicated a significant increase in mean reflectance in the red-band. Biological ground data, in combination with image classification techniques, were not able to accurately detect tree mortality. Aerial photography analysis methods and BBDMIAS were used complementarily to provide a more comprehensive knowledge of the detectability of BBD in Wisconsin and Michigan forests. Further research incorporating additional spectral data, e.g., color infrared, over longer time scales has great potential to yield more detailed information that land managers and researchers could use to measure and track BBD over wide spatial and temporal scales.