Trees as bioindicators of hillslope degradation by debris flows and dangerous rockfalls along the Lefthand Canyon, Colorado Front Range

Abstract

AbstractTrees provide important ecosystem services such as hillslope stabilization from landslides and rockfall protection, and may also be useful indicators of likelihood of those events in the future. Anatomical changes of tree wood are caused by various physical and biological phenomena. There are several geomorphic processes that change tree‐ring characteristics, including hillslope movement, and if detectable, these changes may be used as bioindicators of terrain surface stability. This type of proxy data is vital to places that lack systematic monitoring infrastructure. In the present study, we have evaluated the effects of a heavy rainfall that happened on 11–13 September 2013 in the Colorado Front Range (CFR), which triggered more than 1300 debris flows. We intensely studied one large debris flow that was triggered in the region and the surrounding trees. We explored potential tree‐based records of hillslope instability including compression wood and the method of eccentricity index of tree rings prior to the 2013 event. In addition, the response index was correlated with daily precipitation totals of the previous and current year in a moving window mode. We also evaluated rockfall patterns to explore the frequency and spatial distribution of tree‐scar‐based records. Our primarily goal was to test the method in dry climate conditions and to reconstruct possible soil instability episodes and rockfalls that led to hillslope degradation and tree mortality. In this study, we aimed to fill a significant knowledge gap on debris flows and rockfalls frequency in a dry montane forest system. We were able to distinguish 5 years of increased response index (>25% of trees recording slope instability): 1955 (30%), 1982 (27%), 2002 (27%), 2006 (31%), and 2009 (26%). A spring and early summer season daily precipitation totals of the previous year and summertime daily precipitation totals of the current year explained part of the response index variability. Significant correlation coefficients were up to 0.27. Almost 30 historical rockfall events were dated but only one coincided with an earthquake. In total 138 scars visible on growing trees were mapped. Some of them were found even 3.1 m above the ground level. We conclude that trees can be effective bioindicators of hillslope instability related to debris flows and rock falling. However, extreme climate events such as heavy rainfall causing a high number of debris flows can be difficult to predict based on the method adopted in the present study.