9484 - Laserfiche WebLink

Home Browse Search

December 2002 FLOW-SEDIMENT EFFECTS ON RIVERINE FISH METHODS Study design We used a stratified random-sampling approach to characterize physical and biotic attributes along the river and test for relationships between physical and biological metrics. The river was first stratified by dif- ferences in channel morphology and mean gradient. Four strata were omitted from study because they were either transitional, anomalous (Westwater Canyon), or similar to an adjacent stratum. Eleven study strata com- prised 335 km, or 90% of the study area. Using aerial photographs, we divided each stratum into reaches con- sisting of one meander sequence (1.0-3.1 km long) with the top of each demarcated by the upstream end of a riffle. From these, three study reaches were selected from each stratum using a random numbers table, for a total of 33 reaches. With three exceptions (studies of geomorphology, sediment dynamics, and Colorado pikeminnow), the study was conducted over a two-year period (1994- 1995), with sampling conducted in early spring (March and April) prior to runoff and in fall (September and October) following the summer thunderstorm, season. These represented two distinct seasonal periods when discharge levels were relatively stable. The first period (spring 1994) was largely a pilot effort and only strata 4, 6, 9, and 11 were sampled. All study strata were sampled in subsequent periods. Most analyses included only the three periods with complete data sets, although the spring 1994 samples were included when statisti- cally appropriate. Geomorphology To characterize channel morphology, we surveyed main-channel cross-sections during 1995-1997 at evenly spaced, 1.6-km intervals from rk 77 to rk 363 (except two reaches and a few sites too difficult to access or survey) using an electronic theodolite (total station) and a motorized rubber raft outfitted with a depth sounder. Mean river slopes were calculated from USGS 7.5-min topographic maps (1:24000 scale) and global-positioning surveys. Point counts of the surface bed material and bulk samples of the subsurface sed- iment were used to characterize substrate within each stratum. Sieves were used to size 100 or 200 randomly selected particles from the surface layer and 100-150 kg of underlying subsurface sediment. A total of 56 surface samples and 27 subsurface samples were taken from exposed bars at low flow. The median grain size (Dso) for each stratum was derived from composite samples obtained by grouping data from 3-10 individ- ual sites per stratum. Fine sediment is winnowed from the riverbed when flows reach sufficient magnitude to, dislodge coarse framework particles and move the surficial armor layer. Estimates of discharges required to transport bed ma- terial were made by combining several conventional 1723 flow and sediment transport equations, solved and cal- ibrated with the aid of flow and sediment data described above (see Pitlick and Van Steeter 1998). Assessment of coarse sediment transport frequency was derived from stream-flow records from four USGS gauging sta- tions (Fig. 1). Total depth of free space (DFS: absolute distance from the top of surficial rocks to the point where rocks are embedded in fine sediment) and relative DFS (free- space depth scaled to median rock size) was monitored over six years (1996-2001) in strata 9 and 8 to provide insight into the temporal dynamics of fine sediment intrusion in the armor layer. These strata were selected for monitoring because of high numbers of fish, rela- tively clean substrate, during the initial 1994-1995 study, and the immediate upstream and downstream proximity to a major inflow and potential sediment source (the Gunnison River). Twenty DFS measure- ments were taken per sampling effort at each of 16 permanent monitoring sites. In each stratum, four sites were in run habitats, four were in riffles. Each mea- surement consisted of laying one hand flat across the top of the substrate surface while using the extended fingers of the other hand to probe the substrate until the layer of embeddedness was reached. The distance from the finger tip of the probing hand to the perpen- dicular palm of the first hand was measured with a rule to the nearest 0.5 cm. Sampling was conducted 2-4 times annually during base flows of summer and fall (August through October). To derive relative DFS, in- dividual values of total DFS were divided by the me- dian rock width of the respective site (once annually during base flow, the b-axis of each of 100 rocks [?8 mm] sampled from an underwater transect at each site was measured to the nearest 1 mm with a vernier cal- iper). Other physical parameters A representative riffle and run were selected within each study reach as consistent sites for additional mea- surements; there were 66 such sites (11 strata x 3 reaches X 2 habitats). Within each, the following mea- surements were made at five locations parallel to shore at a standardized water depth of -30-45 cm, inde- pendent of substrate: total DFS, percentage of embed- dedness (portion of aerial surface area consisting of fines), interstitial void volume, and percentage of sur- face layer consisting of fines (<2 mm). A 0.134-mz Hess sampler (50 cm high) was used to circumscribe the five samples for total DFS, percentage embedded- ness, and void volume. Percentage embeddedness (sur- face area within the Hess sampler) was visually esti- mated before the Hess-enclosed substrate was dis- turbed. Total DFS within the Hess sampler was mea- sured once using the method previously described for DFS monitoring. Void volumes were then derived from water displacement: all rocks above the embeddedness layer within the Hess sampler were placed in a tub of