U.S. Department of Defense


Date of this Version



Published in Final Report- Cottonwood Project (2010) 1-114


From 2007-2009, we studied the current (2006) and historic (1892, mid-1950s) extent, current age distribution, and plant species composition of plains cottonwood (Populus deltoides) and non-cottonwood riparian stands along eight study segments of the Missouri River between Fort Benton, Montana and Kansas City Missouri, covering 930 river miles (1500 km) or over 1/3 of the river's length. These segments included all of the unchannelized and unimpounded segments below Fort Benton, as well as portions of two impounded and one channelized segment. Based on GIS analyses of historic maps and aerial photography, the combined area of forests, woodlands, and shrublands in the historic floodplain declined 47% across all study segments from 1892 to 2006, with losses linked to clearing for human land uses (primarily agricultural cropland) or inundation by reservoirs. Most forest loss occurred between 1892 and the 1950s, while most shrubland loss has occurred from the 1950s to 2006. As of 2006, we estimate that a total of 75,600 hectares (186,900 acres) of shrubland, woodland, and forest occurred within the mapped area of the eight study segments, with 66,800 hectares (165,000 acres) or 88% of the total, composed of patch types which contain cottonwood as a significant component. Most (62%) of the cottonwood area is composed of stands >50 years old, and only 14% is from stands that have recruited in the last 25 years. These patterns, along with significant historic declines in shrubland and sandbar area on most segments, indicate that the fluvial geomorphic dynamics that drive cottonwood recruitment have been reduced, and cottonwood regeneration compromised, under the river management practices of at least the last 25-30 years. A significant proportion (24%) of forest in the 25-50 year age class, however, suggested that a temporary pulse in recruitment accompanied geomorphic adjustments in the channel during the first 2-3 decades after dam closure on some segments. We sampled plant species composition and vegetation structure across an age gradient of cottonwood and non-cottonwood riparian stands in each study segment, with a total of 332 stands sampled, of which 216 were relatively intact natural cottonwood stands. Indices of plant species richness, wetland affinity, and floristic quality differed among stands and study segments, providing possible metrics for evaluating stand- and segment-level differences in biotic integrity. Segments 4 (below Garrison Dam) and 10 (below Gavins Point Dam) had the highest stand-level species richness (~35 species per stand) and highest floristic quality (Coefficient of Conservatism values) of all segments, while segment 6 (below Oahe Dam) had the lowest of each. Stressors that may influence segment-level vegetation and landscape patterns include land management (e.g., grazing, vegetation management); forest clearing for agricultural cropland and urban or exurban expansion; channel incision and cessation of overbank flooding below dams; disruption of sediment supply and transport, with resultant declines in formation of alluvial surfaces needed for cottonwood recruitment; and aggradation with resultant water table rise adjacent to river-reservoir delta areas. In the report that follows, we provide summaries of current and historic land cover, cottonwood age distribution, and vegetation patterns for each of the eight study segments, as well as an overview across all segments.