Mapping spatial pattern and mortality of Eastern and Carolina hemlocks to estimate effects and potential spreading of hemlock woolly adelgid infestations in the Southern appalachians

Sunday, November 10, 2013: 4:03 PM
Meeting Room 9 C (Austin Convention Center)
Tuula Kantola , Knowledge Engineering Laboratory, Texas A&M University, College Station, TX
Maria Tchakerian , Knowledge Engineering Laboratory, Texas A&M University, College Station, TX
Päivi Lyytikäinen-Saarenmaa , Department of Forest Sciences, University of Helsinki, Helsinki, Finland
Robert Coulson , Knowledge Engineering Laboratory, Texas A&M University, College Station, TX
Douglas Streett , Alexandria Forestry Center, USDA Forest Service, Pineville, LA
Outbreaks of different invasive species are massive threats to North American forests. Disturbances by invasive, non-native species can result in major ecological, social, and economic losses. Hemlock woolly adelgid (Adelges tsugae Annand, HWA) outbreaks are posing one of biggest threats to Eastern hemlock (T. canadensis L. Carr.) and Carolina hemlock (T. caroliniana Engelm.) in forested landscapes in the eastern United States. These hemlock species play an important role in forest ecosystems e.g., cooling mountain streams and providing habitat for many vertebrate and invertebrate species. No effective chemical- or biological controls against the HWA populations have been proved to be effective. In addition, there is no other tree species having a similar ecological niche and functions to be replaced in areas suffering of hemlock declining by the HWA. The Study area of 50 km2 is located in Linville gorge area, Grandfather Ranger District, Southern Appalachian Mountains, NC (N 35°56’, W 81°55’), where both Eastern and Carolina hemlocks serve as a foundation species.  The objectives of this study is to 1) detect Eastern and Carolina hemlock patches via high resolution aerial imagery and 2) study spatial distribution and pattern of hemlocks and hemlock mortality by HWA. A two stage classification method including classification trees and support vector machine is going to be applied in land cover and tree species classification and linearized Ripley’s K function to reveal spatial patterns of living and dead hemlocks patches. The results are important findings that could be used to project future distribution and extend of HWA infestations and to study potential of invasive plant species, e.g. Tree of heaven (Ailanthus altissima Mill.) or Japanese stilt grass (Microstegium vimineum (Trim. A. Camus.), to be established in future forest gaps created after hemlock elimination by HWA.