The use of composite fingerprints to quantify sediment sources in a wildfire impacted landscape, Alberta, Canada

摘要: There is increasing global concern regarding the impacts of large scale land disturbance by wildfire on a wide range water and related ecological services. This study explores impact 2003 Lost Creek in Crowsnest River basin, Alberta, Canada regional sediment sources using tracing approach. A composite geochemical fingerprinting procedure was used to apportion efflux among three key spatial sources: 1) unburned (reference) 2) burned 3) sub-basins that were subsequently salvage logged. Spatial characterized collecting time-integrated suspended samples passive devices during entire ice free periods 2009 2010. The combines Kruskal-Wallis H-test, principal component analysis genetic-algorithm driven discriminant function for source discrimination. Source apportionment based numerical mass balance model deployed within Monte Carlo framework incorporating both local optimization (genetic algorithm) optimization. mean relative frequency-weighted average median inputs from units estimated be 17% (inter-quartile uncertainty 0-32%) reference areas, 45% 25-65%) areas 38% 14-59%) burned-salvage logged areas. High representing 2 3, reflect lasting effects forest canopy floor organic matter including increased runoff availability high terrestrial erosion, streamside wasting river bank collapse. results demonstrate incremental pressures associated with logging catchment higher elevation Montane regions where growth vegetation recovery are relatively slow.