Exposed tree root analysis as a dendrogeomorphic approach to estimating bank retreat at the South River, Virginia

We use a biometric approach based on anatomical changes in the wood of exposed tree roots to quantify riverbank erosion along South River, Virginia, a site where commonly applied techniques for determining bank erosion rates are either not appropriate because of the required spatial scale of analysis (i.e., erosion pins, traditional surveys, LiDAR analysis) or have failed to detect obvious erosion (i.e., photogrammetric techniques). We sampled 73 exposed roots from 22 study reaches and identified the year of exposure macroscopically (2 to 20 times magnification) and microscopically (20 to 100 times magnification), comparing the estimated erosion rates between levels of magnification and to those obtained with photogrammetric techniques. We found no statistical differences between the results of macroscopic and microscopic analyses (t-test, α = 0.01) but encountered difficulty in identifying the year of root exhumation in some samples. When comparing exposed root analysis to photogrammetric techniques, the results indicate that the exposed root approach is a feasible and effective method for estimating annual- to decadal-scale bank erosion. In addition to producing erosion rates statistically indistinguishable from photogrammetric techniques at sites with erosion rates large enough for detection using historical aerial photographs (regression analysis and t-test, α = 0.01), exposed root analysis was able to estimate erosion rates at sites where photogrammetric techniques failed. We also identify deciduous species well suited for this approach (Fraxinus pennsylvanica) and others that prove more problematic (e.g., Acer negundo, Celtis occidentalis, Acer saccharinum). This study is significant because it describes a robust tool that provides insights into annual- to decadal-scale erosion where other commonly applied techniques may not be appropriate or easily applied.