• Dendroclimatic Reconstruction from Bald Cypress in Southwestern Indiana

      Van De Veer, Robin Lyn (2011-09-20)
      In the United States, bald cypress (Taxodium distichum) is generally recognized as an important component of the forested wetlands found in the Southwestern Coastal Plain and the Mississippi River Valley (which extends to the southern Midwest). The lifespan of this deciduous species is important not only commercially, but also in an ecological capacity. This study focuses mainly upon the role the tree plays in its environment and how it can be used as an indicator of climate through drought/flood signals in the rings. Bald cypress is a long-lived tree that can be over 1000 years old and is sensitive to climate and ground water hydrology. Because of these factors it is a favorable choice for dendrochronological study in the region. According to the International Tree-Ring Database, a chronology of the species is not well defined for southwestern Indiana. This research provides this missing information and creates the northern most bald cypress chronology in the Midwest. The study site is located in the extreme southwest of Indiana around Hovey Lake (a backwater lake of the Ohio River) about 10 miles south of Mount Vernon, Indiana. Samples were taken from trees near the shore, both on land and in the water.This study dated some trees to 1855. Analysis of the tree rings, climate data, and river discharge data revealed that bald cypress are not declining in southwestern Indiana. The rate of tree ring growth increases as PDSI does and the rate of river discharge does not seem to affect growth much at all. Even though this is the northernmost bald cypress chronology in the midwest and therefore should be stressed according to the theory of ecological amplitude, this chronology does not fall in the category with the highest series intercorrelation or mean sensitivity. The construction of the dam in 1975 has overwhelmed the climate signal in these trees and the trees continue to be suppressed due to the current water level.
    • Tree-Ring Analysis of Outbreak Dynamics across an Insect’s Entire Range: The Pandora Moth System

      de Graauw, Kristen (2012-10-22)
      In montane forests of the western United States, pandora moth (Coloradia pandora Blake) defoliates local pines, primarily ponderosa pine (Pinus ponderosa Dougl. Ex Laws.) and Jeffrey pine (Pinus jeffreyi Grev. & Balf.). This defoliation stresses the tree, leaving a distinct outbreak signature in the tree-rings. The occurrence of pandora moth outbreaks has been recorded in ponderosa pine tree rings in Oregon as far back as 1500 years, however little is known of the outbreak history throughout the rest of the pandora moth range. To gain a better understanding of the spatiotemporal dynamics of pandora moth I have reconstructed outbreaks across the entire range of the insect using 121 tree-ring chronologies from the International Tree-Ring Databank (ITRDB) and 19 chronologies from sites sampled for this study using dendrochronological techniques. I then created a fine resolution habitat model for pandora moth and a gridded tree-ring network of non-host chronologies, which was used to validate statistically confirmed outbreak events across the entire range of the insect. Sites with confirmed outbreaks were then tested against the habitat model for accuracy of model parameters. ArcMap 10 was used to create an animated map of the spatial and temporal distribution of pandora moth across its range. One of my most notable findings is the record of outbreak events further north than previously documented, and the agreement between those northern sites with my habitat model. In my research I have demonstrated the use of dendrochronology to study outbreaks across an insect’s range. The methods I have used here can be applied to other range-wide analyses.
    • Urban Flash Flood Risk Assessment and Inundation Model Utilizing GIS for Terre Haute, Indiana

      Ishman, Zachary Scott (2015-01-07)
      Use of ArcGIS to examine flash flooding variables and produce a flash flood risk assessment and inundation model for Terre Haute, Indiana. Risk assessment, produced within ArcGIS, indicates that an increase in developed area leads to an increase in very high flash flood risk area and majority of very high risk area resides in developed areas of Terre Haute. Inundation model, produced using ArcGIS and Python, indicates that the proposed model can determine locations of flash flooding, but spatial extent of model predicted flooding is not reliable based on field validation.