Using Cottonwood Dendrochronology to Reconstruct River Discharge and Floodplain Dynamics, Yellowstone River, Montana

Friday, 19 December 2014
Derek M Schook1, Jonathan M Friedman2 and Sara L Rathburn1, (1)Colorado State Univ, Geosciences, Fort Collins, CO, United States, (2)US Geological Survey, Boulder, CO, United States
Ecosystems and societies worldwide have evolved to depend upon the timing and magnitude of river discharge, and understanding past flows can help guide modern water management. We used tree rings of riparian plains cottonwoods (Populus deltoides subsp. monilifera) to reconstruct the history of flow variation and channel migration of a 20 km reach of the Yellowstone River in eastern Montana. Dendrochronological flow reconstructions commonly use upland trees, but our study highlights the improved resolution when floodplain trees are integrated into the data set . Our sample of 240 cottonwoods dating back to 1751 permits flow reconstruction of the Yellowstone to before the voyage of Lewis and Clark. Our tree ring series intercorrelation coefficient is 0.58, and the ring width index correlates to annual discharge at R = 0.67. Flow reconstruction indicates that the decades of highest (1820s, 1850s) and lowest (1830s, 1900s) flows all occurred prior to the instrumental record, revealing the value of an extended perspective. Cottonwood age distribution indicates that, like other western rivers, the rate of channel migration on the Yellowstone declined in the 20th century. However, the Yellowstone uniquely lacks mainstem dams and substantial water extractions, revealing the occurrence of hydrological and ecological change on a relatively natural river. Our study reach is the most geomorphically active of the entire 1100 km river between Yellowstone National Park and the Missouri River, but cottonwood age distribution reveals that trees that have established since the 1960s are underrepresented. The lack of younger cottonwood trees is likely caused by a decline in river migration rates, which may be attributed to i) climate change directly leading to a decline in fluvial processes driving river migration, ii) a decoupling in the timing of the snowmelt runoff receding limb and cottonwood seed release, or iii) both. Even on this relatively unmodified river, it appears that changes to the hydrologic regime are negatively affecting the floodplain ecosystem and associated habitats.