| Abstract/Description or Keywords |
The impact of various land use activities (forestry, mining, and agriculture) on the quality of fine-grained sediment (<63 _m) was investigated in the Quesnel River Basin (QRB) (approx. 12,000 km2) in British Columbia, Canada. Samples of fine-grained sediment were collected monthly during the snow-free season in 2008 using time-integrated samplers at sites representative of forestry, mining, and agricultural activities in the basin. Samples were also collected from replicate control sites that had undergone limited or no disturbance in recent years, and also from the main stem of the Quesnel River. Generally, metal and nutrient concentrations for “impacted” sites were greater than for control sites. Concentrations of As (mining sites), Cu (forestry sites) and Zn (forestry sites) were close to or exceeded upper Sediment Quality Guideline (SQG) thresholds, while Se concentrations for mining sites were elevated and within the range cited for contaminated environments. Phosphorus values were generally <1000 _g g-1 for all land use activities and below available SQGs. Values for individual samples were, however, greater than upper SQG levels, such as 22.7 _g g-1 (As), 801 _g g-1 (Cu), 5.0 _g g-1 (Se) and 2192 _g g-1 (P). These results suggest that metal mining and forest harvesting are having a greater influence on the concentration of sediment-associated metals and nutrients in the Quesnel basin, than agricultural activities. Temporal and spatial differences in the metal and P content of fine suspended sediment within the QRB during the 2008 field season were analyzed using rank sum tests and comparisons to discharge (Q) and precipitation (PPT) values. Temporal results suggest the overall mining signature was often a function of changes in source inputs and delivery, while forestry and agriculture were influenced by variations in transport conditions (e.g. PPT and Q). Spatial variation was predominantly due to differences between mining and control geochemical concentrations. Forestry and agriculture differed for select elements, but played a lesser role than mining. To further characterize land use variation the fingerprinting method was used, involving stepwise discriminant function analysis (DFA). This resulted in a composite signature capable of differentiating correctly 100% of the source geochemical contributions from each land use type. The composite signature was used to inform basin-scale geochemical signature prediction using a multivariate mixing model. This determined that agriculture was the highest overall contributor to the outlet signature at Quesnel. Additionally, the control influence was strongest earlier in the sampling campaign while mining contributed most in the latter sampling periods. |
