No abstract

Age, distribution, and density of two populations of plains cottonwood (Populus deltoides var. occidentalis Rydb.) on the floodplain of the Milk River, southern Alberta and northern Montana, were studied in relation to historical river hydrology and sedimentation regimes. In Alberta, cottonwood recruitment leading to long-term survival on river meander lobes (point bars) correlates with years when daily maximum flows during the period of seed dispersal (June 1 to July 10) attain a stage equal to or greater than the 2-year return flood, based on the annual flood series. Such flood events during the seed dispersal period recur an average of once in 5 years. In Montana, on the floodplain for 25 km downstream of Fresno Dam, built in 1939, the densities of cottonwoods recruited since 1939 are significantly lower than on floodplain sites upstream, in Alberta. Results suggest that this is due to a marked reduction in flood magnitude and frequency, rates of sedimentation and meander migration. Based on this study, the prospects for cottonwood survival on floodplains downriver from dams in this and other prairie river valleys are not encouraging unless management measures are taken to reverse the trend.

Ages of channels of the anastomosing upper Columbia River, south‐eastern British Columbia, Canada, were investigated in a cross‐valley transect by 14C dating of subsurface floodplain organic material from beneath levees. The avulsion history within the transect was deduced from these data, and morphological stages in channel development were recognized. Additionally, floodplain sedimentation rates were established. The new data demonstrate that the upper Columbia River is a long‐lived, dynamic anastomosing system. Results show that anastomosis at the study location has persisted since at least 2700 cal. years BP, with avulsions occurring frequently. At least nine channels have formed in the studied cross‐valley transect within the past 3000 years. Channel lifetimes from formation to abandonment appear to be highly variable, ranging from approximately 800 to 3000 years. Log jams provoking avulsions and/or silting up of old channels are proposed as reasons for this variability. Long‐term average floodplain sedimentation rates appear to be significantly lower than previously proposed by Smith (1983, Int. Assoc. Sedimentol. Spec. Publ., 6, 155–168). A long‐term (4550 years) average of 1·75 mm year−1 (after compaction) was based on 14C dates, while a short‐term sedimentation rate of 0·8 mm was determined for a single, relatively small, seasonal flood in 1994 using sediment traps. However, short‐term sedimentation rates vary considerably over the floodplain, with levees aggrading up to four times faster than floodbasins. Channels of the upper Columbia River anastomosed reach follow a consistent pattern in their development, with each stage being characterized by different morphology and processes. Channel evolution comprises the following succession: (1) avulsion stage, in which a crevasse splay channel deepens by scour and levee sedimentation; (2) widening and deepening stage, in which bank slumping and bed scouring dominates; (3) infilling stage, in which either channel narrowing (bank accretion) or channel shallowing (bed accretion) takes place; and (4) abandonment stage, in which the residual (remnant) channel is filled exclusively by silt, clay and organic material. Vertical stacking (super‐ imposition) of active channels on recent channel‐fill sand bodies is a notable feature of the upper Columbia River, which suggests that reoccupation of residual channels is a common process.

No abstract

No abstract