Source ascription in bed sediments of a Mediterranean temporary stream after the first post-fire flush

Abstract

Purpose First flushes can be crucial to sediment transport in fluvial systems of drylands, as temporary streams are a characteristic feature of Mediterranean basins. After a wildfire, storm flows may enhance runoff delivery to channels, so increasing the first-flush effect. 137Cs and 210Pbex were used as tracers for recognizing the first post-fire flush effect on the source ascription of bed sediments temporarily stored in a Mediterranean temporary stream severely affected by a wildfire. Materials and methods Thirty sediment source samples were collected along one of the tributaries of a catchment (4.8 km2) located in Mallorca during a field campaign some weeks after the wildfire. The sample collection took into account effects of the wildfire and distinguished between soil surface and channel bank. To measure the source contribution temporarily stored with the bed sediment, five sediment samples¿deposited during the first storm, occurring 3 months after the wildfire¿were collected from the bed of the main channel. The 137Cs and 210Pbex concentrations were measured by gamma spectrometry. Then, a linear mixing model was used to establish the contribution of each source type to the bed sediments along the main stream. Results and discussion First post-fire flush effect was generated by a torrential event with a suspended sediment concentration peak of 33,618 mg L−1, although transmission losses under a very low runoff coefficient (1%) promoted sediment deposition. Significant differences were observed in fallout radionuclide concentrations between burned surface soil and burned channel bank samples (p < 0.05), as well as between burned and unburned sources in the downstream part of the catchment (p < 0.01). The radioactivity concentrations in bed sediment samples were statistically similar (p > 0.05). Source ascription in bed sediments upstream shows that 67% was generated from burned hillslopes, reaching 75% in the downstream part, because downstream propagation of the sediment derived from the burned area. Conclusions Bed sediments were mostly generated on burned surface soils because of the fire's effect on soil and sediment availability, high-intensity rainfall, and the limited contribution of channel banks, because these are fixed by dry-stone walls. This hydro-sedimentary response indicates an association between these factors driving erosion and sediment delivery, generating effective slope-to-channel sediment connectivity. The integration of the short-term response with the medium- and long-term analysis will allow for the analysis of the evolution of catchment sediment sources in future studies, determining if fire modifies the catchment sensitivity to that specific disturbance.