Geomorphic and vegetative recovery processes along modified stream channels of West Tennessee

Abstract

Hundreds of miles of streams in West Tennessee have been channelized or otherwise modified since the turn of century. After all or parts of a stream are straightened, dredged, or cleared, systematic hydrologic, geomorphic, and ecologic processes collectively begin to reduce energy conditions towards the premodified state. One hundred and five sites along 15 streams were studied in the Obion, Forked Deer, Hatchie, and Wolf River basins. All studied streams, except the Hatchie River, have had major channel modification along all or parts of their courses. Bank material shear-strength properties were determined through drained borehole-shear testing (168 tests) and used to interpret present critical bank conditions and factors of safety, and to estimate future channel-bank stability. Mean values of cohesive strength and angle of internal friction were 1.26 pounds per square inch and 30.1 degrees, respectively. Dendrogeomorphic analyses were made using botanical evidence of channel-bank failures to determine rates of channel widening; buried riparian stems were analyzed to determine rates of bank accretion. Channel bed-level changes through time and space were represented by a power equation. Plant ecological analyses were made to infer relative bank stability, to identify indicator species of the stage of bank recovery, and to determine patterns of vegetation development through the course of channel evolution. Quantitative data on morphologic changes were used with previously developed six-stage models of channel evolution and bank-slope development to estimate trends of geomorphic and ecologic processes and forms through time. Immediately after channel modifications, a 10- to 1%yearperiod of channel-bed degradation ensues at and upstream from the most recent modifications (area of maximum disturbance). Channel-bed lowering by &gradation was as much as 20 feet along some stream reaches. Downstream from the area of maximum disturbance, the bed was aggraded by the deposition of sediment supplied by knickpoint migration upstream; aggradation also occurred in initially degraded sites with time. Additionally, if degradation caused an increase in bank height beyond the critical limits of the bank material, a period of channel widening by mass wasting followed. Degradation knickpoints migrated upstream at rates greater than 1 mile per year; the rates attenuated with distance above the area of maximum disturbance. Channel widening rates of up to 16 feet per year were documented along some severely degraded reaches. Planar failures were generally more frequent but rotational failures dominated the most rapidly widening reaches. Total volumes of bank erosion may represent 75percent or more of the total material eroded from the channel, but this material generally exits the drainage basin. Mean factors of safety vary with the stage of channel evolution with the lowest values for planar and rotational failures occurring during the threshold stage (stage IV) 1.00 and 1.15, respectively. As channel gradients decrease, degradation ceases and then a period of “secondary aggradation ” (at lesser rates than degradation) and bank accretion begins that may fill the channel to near floodplain level. This shift@ in process represents an oscillation in channel bed-level adjustment. Streams in basins underlain by loess may require an order of magnitude more time than sand-bed streams to stabilize due to a lack of coarse-grained material (sand) for aggradation. A systematic progression of riparian species that reflects the six-stage model of channel evolution has been identified. This progression can be used to infer ambient channel stability and hydrogeomorphic conditions. Woody vegetation establishes on low- and mid-bank surfaces (the slough line, initially) at about the same time that bank accretion begins. This slough line forms at a mean temporary stability angle of 24 degrees and expands upslope with time by the accretion of sediments. Species involved in this initial revegetation are hardy, fast growing, and can tolerate moderate amounts of slope instability and sedimentation; these species include river birch, black willow, boxelder, and silver maple. Vegetation appears to enhance bank stability, and with increasing stability, species such as bald cypress, tupelo gum, and various hydric oaks, which are more characteristic of stable, premodified riparian settings, begin to establish. Detrended-correspondence analysis indicated species assemblages associated with the six stages of channel evolution and bank-slope development. Ordination of site variables based on species data such as channel widening, bank accretion, and woody vegetative cover also reflects the temporal changes identifies by the models. Long-term channel geometry was estimated from a quantitative model of bed-level change, and from documented trends in channel widening. An idealized stable channel of a major sand-bed stream may have a width/depth ratio near 10 and bank slopes of about 24 degrees. This stable channel will ultimately undergo the development of point-bars and incipient meanders, characteristic of unmodified streams.