Sediment source apportionment following wildfire in an upland commercial forest catchment

Abstract

Purpose Wildfires can have major impacts on water scarcity and water quality linked to off-site transfer of polluting ash and nutrients. Understanding sediment sources in burnt landscapes can help to develop mitigation strategies, especially in catchments planted with introduced species that are prone to fire. We investigated sediment sources activated by post-fire rainfall in a small-forested catchment that was impacted by a severe wildfire. The aim was to use environmental radionuclides and elemental geochemistry as tracers to apportion sediment sources within burnt plantation systems.

Methods Surficial (0–2 cm) topsoil (n = 9), sub-surficial (2–4 cm) topsoil (i.e. below the burnt layer; n = 8) samples from burnt hillslopes and forest roads (n = 5) and stream banks (n = 5) soil samples were taken in the Quivolgo catchment, El Maule region, Chile. Sediment samples (n = 9) were collected from behind a v-notched weir on three dates after the fire: May 2017, July 2017 and October 2017. Soil and sediment samples were analysed by gamma spectrometry and wavelength-dispersive X-ray fluorescence (WD-XRF) used to obtain tracer properties. These were evaluated visually and statistically to identify potential non-conservative tracers. Sediment apportionment was undertaken using the MixSIAR mixing model.

Results The tracer selection procedure resulted in ten tracers being used for sediment apportionment. Tracer suitability was based on (i) weak and non-significant linear relationship between tracer concentrations and specific surface area (SSA) and soil organic matter (SOM), and (ii) conservative behaviour supported by the inclusion of sediment samples within source convex hull. Sediments from sub-surface layer (2–4 cm) were the dominant source during the first two periods contributing 55 ± 11 and 78 ± 10% respectively, whereas road contribution was only important in the last period (71 ± 14%). Apportionment showed a shift in sediment source (i.e. from forest roads to hillslopes) compared to a previous study in the same catchment before wildfire. The main driver of erosion was attributed to overland flow convergence and consequent rill erosion across burnt hillslopes.

Conclusion The study demonstrated combined use of environmental radionuclides with elemental geochemistry for sediment apportionment within burnt forest plantations and highlighted a switch in predominant source (e.g. sub-surface burnt soil) activated by post-fire rainfall events. The findings in this research will help forest companies to develop strategies to reduce off-site impacts of sediment release after wildfire in forest plantations.