REV95427 - Laserfiche WebLink

Home Browse Search

species, nd overall watershed degradation (Stromberg 1993). Mining of stream channels and floo plains for aggregates can severely disrupt riparian ecos:~stems by altering channel processe and adjacent groundwater levels (Faber et al. 1989). In spite of the prevalence of mining a tivities in riparian zones, few attempts have been made to quantify impacts. We describe easured structural and growth responses of a mature plains cottonwood stand to a gradient f groundwater declute associated with sand mining along a Western Great Plains stream. STUDY SITE T e study site is a four mile reach of Coal Creek, an ephemeral, sandbed stream itt the high lairs of central Colorado (Figure 1). Coal Creek flows north from the Pahner Divide, hich separates the South Platte River and Arkansas River basins. Streams draining the Palm r Divide are subject to flash-floods produced by intense convectional storm activity associat with tJte divide (Hansen et al. 1978, Friedman 1993). The riparian zone of Coal Creek is ominated by mature cottonwoods with three diameter class centered around 80 cm, 50 cm a 1 cm. The trees range in age from a few to approxirr.ately 140 years; age disuibuti ns suggest that establishment of some individuals may tie associated with past floods or channel narrowing processes. Sand was mined from the active channel within a portion o the study reach beginning in June of 1992. METHODS F ur transects were established perpendicular to the channel and extended from valley wall to v lley wall. Two to four groundwater wells were established along the center line of each u ect by hand driving sand points or with the aid of a tnrck mounted drill. All wells have bee monitored at least monthly since August 1990. Chanruel cross-sections were surveyed at each uansect in 1991, 1992 and 1993. The physical status of cottonwoods was assessed y repetitive measurements of basal azea (stem cross-sec~:ional area) on all trees within a 50-meter wide belt along each uansect. A total of 760 trees were tagged and measu Each measurement consisted of three replicate readings at fixed positions above, at and be ow breast height. Also, a crown volume index value or." 0 - 100 was assigned to each tree annually based on a visual estimate of the percentage of potential crown volume occupied by live foliage and branches. Thus, a tree with no live foliage or branches would have a c wn volume value of 0, and a tree with no dead foliage or branches in the crown would ha e a value of 100. Tree mortality was recorded at the time of the basal azea ch transect was divided into a series of discrete surfaces of relatively constant elevation These surfaces were the basic experimental unit. Eaclt surface was assumed to have a u 'form depth to groundwater and an aggregate biological response. Active mining began in late June of 1992 anti consisted of the removal of channel bed material from the stream. ransect 2 was located within the area of mining. Transect 1 was located downstre m, and Transects 3 and 4 upstream of the mining.