Discharge-Export Relationships in Headwater Streams: The Influence of Invertebrate Manipulations and Drought

Abstract

The role of physical (discharge) and biological (macroinvertebrate communities) factors in the control of coarse (>4 mm) and fine (≤4 mm > 0.5 μm) particulate organic matter (FPOM) transport was studied in three headwater streams of the southern Appalachian Mountains. The role of discharge was determined by relating two years of continuous measurements obtained over discrete (ca. 2-wk) time intervals. The role of macroinvertebrates was examined by treating one of the three streams, C 54, with an insecticide during Year 2 to reduce populations and alter community structure. Maximum discharge was the only discharge parameter which adequately predicted (linear regressions) FPOM export during a sampling interval (r² ≥ 0.70). These regressions were unique for each stream and were constant between years for the untreated streams, despite a record drought during the second year. Relationships between discharge and export of coarse particulate organic matter (CPOM) were not as strong nor as consistent as those for FPOM. CPOM export was very sensitive to timing of CPOM inputs and storms (e.g., 78 to 88% of CPOM export during Year 2 occurred during a single fall storm). Consequently, CPOM export-discharge relationships differed not only among streams but also between years and did not show treatment effects. Treating C 54 with methoxychlor during Year 2 resulted in massive invertebrate drift and drastically reduced populations of macroinvertebrates, practically eliminating shredders and collector-filterers. Maximum discharge continued to be a good predictor of FPOM export (r² = 0.83) for C 54 during the treatment year, but the yield of export per unit of maximum discharge (slope of the discharge-FPOM export regression) decreased by 65%. This contrasts with the untreated streams in which the slopes were constant between years despite the >50 year drought during Year 2. Most (75%) of the decrease in FPOM yield from C 54 is directly attributable to the reduction in macroin-vertebrates. Only 25% is attributable to the drought even though stream flows dropped by ca. 40% during Year 2. Relationships between export and discharge observed in this study are stronger than those reported elsewhere because measurements of export were made independent of discharge over long time intervals (ca. 2 wk). Collecting export over 2-wk intervals clearly shows the effects of storms while avoiding the problem of hysteresis commonly seen in relationships between discharge and export concentration made during single storms.