Effects of the 2011 flood on riparian forests along the Missouri River, USA

Background/Question/Methods

Installment of six dams from 1933 to 1963 on the Missouri River has reduced the frequency, magnitude, and duration of flooding. In the spring of 2011 record rainfall in the upper basin coincided with snow melt to fill reservoirs to capacity, leading to flows exceeding 4,247 cubic meters per second and flood durations of over 60 days. Flood disturbance of this size had not occurred since 1952 due to flow regulation by dams. In 2012, 150 study sites (covering five forest age-classes: Sapling, <15 years; Pole, 15-30 years; Intermediate, 30-55 years; Mature, 55-119 years; and Old, >119 years) from a previous study were resampled on five remnant floodplain segments of the Missouri River from Montana to Nebraska. The same methods used for pre-flood surveys were used to characterize post flood structure and composition of riparian forest. Pre-  and post-flood data were compared to characterize 1) overall changes in mean density of live stems by age class for trees (>10cm dbh) and shrubs (<10cm dbh), 2) changes in mean shrub cover by age class, and 3) changes in mean stem density for select woody species for sites less than 55 years (post-dam establishment) and greater than 55 years (pre-dam establishment).

Results/Conclusions

Woody stem density declined significantly across study segments for all age classes except Old. Average density for Sapling sites decreased by 80%, Pole by 75%, Intermediate by 45%, and Mature by 40%; while Old increased by 12%. Shrub cover declined significantly in Sapling, Pole, and Intermediate sites but not in Mature or Old sites. Cottonwood density in post-dam (<55 year old) sites declined significantly for all study segments except near Pierre, South Dakota (segment 6). In pre-dam (>55 year old) sites, cottonwood density declined significantly in Montana (segment 2) and in South Dakota/Nebraska (segment 10) only. Russian olive density declined in all segments for post-dam sites and in segments 2, 4 (North Dakota), and 6 for pre-dam sites. Eastern red cedar declined in post-dam sites on segment 6, 8/9 (South Dakota/Nebraska), and 10. For the pre-dam sites, red cedar declined in all segments except segment 2 where it was not found. Overall, our data suggest that flood-related mortality has led to significant declines in tree and shrub densities, particularly in younger riparian forests.  The degree to which enhanced post-flood recruitment will compensate for flood-induced losses of early successional forests remains to be seen.