Influence of vegetation cover and other sources of variability on sediment and runoff response in a burned forest in South Korea

Post-fire field measurements of sediment and run off yield were undertaken in natural rainfall event-basis during five rainy months in Korea on a total of 15 small plots: four replica burned unseeded plots, six replica burned seeded plots, and five replica unburned plots. The main aim was to evaluate the effects of vegetation recovery and spatial distribution patterns on sediment and runoff response between and within the treatment replica erosion plots. Six-years after the wildfire, total sediment and runoff yield in the burned unseeded plots with 20%-30% vegetation cover was still 120.8 and 20.6 times higher than in the unburned treatment plots with 100% ground cover, 8.3 and 6.7 times higher than in the burned seeded plots with 70%-80% vegetation cover, while only 1.6 and 2.0 times higher than in the burned seeded plots with 50%-60% vegetation cover, respectively. The differences in sediment and runoff yield between the treatment plots was proportional to total vegetation cover, distance of bare soil to vegetation cover, magnitude of rainfall characteristics and changes in soil properties, but not slope gradient. Three out of the six within-treatment pairs of two replica plots showed large differences in sediment and runoff yield of up to 6.0 and 4.2 times and mean CV of up to 99.1% and 62.2%, respectively. This was due to differences in the spatial distribution patterns of surface cover features, including aggregation of vegetation and litter covers, the distance of bare soil exposed to vegetation cover closer to the plot sediment collector and micro topographic mounds and sinks between pairs of replica plots. Small differences in sediment and runoff of only 0.9-1.4 folds and mean CV of 8.6%-25% were observed where the within-treatment pairs of replica plots had similar slope, total surface cover components and comparable spatial distribution pattern of vegetation and bare soil exposed surface covers. The results indicated that post-fire hillslopes undergoing effective vegetation recovery have the potential to reduce sediment and runoff production nearer to unburned levels within 6-years after burning while wildfire impacts could last more than 6-years on burned unseeded ridge slopes undergoing slow vegetation recovery.