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

M Stone; A L Collins; U Silins; M B Emelko; Y S Zhang

doi:10.1016/j.scitotenv.2013.12.052

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

Sci Total Environ. 2014 Mar 1:473-474:642-50. doi: 10.1016/j.scitotenv.2013.12.052. Epub 2014 Jan 7.

Authors

M Stone¹, A L Collins², U Silins³, M B Emelko⁴, Y S Zhang⁵

Affiliations

¹ Department of Geography and Environmental Management, University of Waterloo, 200 University Avenue West, Waterloo, Ontario N2L 3G1, Canada.
² Sustainable Soils and Grassland Systems Department, Rothamsted Research-North Wyke, Okehampton EX20 2SB, UK; Geography and Environment, University of Southampton, Highfield, Southampton SO17 1BJ, UK. Electronic address: adrian.collins@rothamsted.ac.uk.
³ Department of Renewable Resources, University of Alberta, Edmonton, Alberta T6G 2H1, Canada.
⁴ Department of Civil and Environmental Engineering, University of Waterloo, Waterloo N2L 3G1, Canada.
⁵ Geography and Environment, University of Southampton, Highfield, Southampton SO17 1BJ, UK.

PMID: 24407033
DOI: 10.1016/j.scitotenv.2013.12.052

Abstract

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

Keywords: Genetic algorithm; Sediment source fingerprinting; Uncertainty analysis; Wildfire.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Alberta
Discriminant Analysis
Ecology
Environmental Monitoring / methods*
Fires*
Forestry
Geologic Sediments / analysis*
Geologic Sediments / chemistry
Trees