A field study was conducted to analyze root throw and associated sediment transport in Hawk Creek Watershed, Canadian Rockies. A large crown fire in 2003 allowed the opportunity to study pre-fire and post-fire root throw. Based on field data, a significant relation was found between gradient and root plate volume, as well as individual root plate dimensions. Given that tree diameters increase as trees age and that a relation in the field data was found between tree diameter and root plate volumes, sediment transport due to root throw is expected to change in response to forest disturbance and stand age. Sediment disturbance, which is the amount of sediment upheaved during tree topple and does not take into account transport distance, shows higher values on steeper gradients. Sediment transport was notable for the steepest plots, with pre-fire values of 0·016 cm 3 cm -1 a -1 and post-fire values of 0·18 cm 3 cm -1 a -1 . A tree population dynamics model is then integrated with a root throw transport model calibrated for the Canadian Rockies to examine the temporal dynamics of sediment transport. Fire is incorporated as a disturbance that initiates development of a new forest, with the model cycling through generations of forest. Trees fall according to an exponential rate that is based on time since death, resulting in a time lag between tree mortality and sediment transport. When values of time-since-previous-fire are short, trees are generally <13 cm, and minimal sediment is upheaved during toppling. If trees reach a critical diameter at breast height (dbh) at time of fire, a pulse of sediment occurs in the immediate post-fire years due to falling of killed trees, with tree fall rates decreasing exponentially with time-since-fire. A second pulse of root throw begins at about 50 years after the previous fire, once new recruits reach a critical dbh and with initiation of competition-induced mortality.

Increased soil erosion in immediate post-wildfire years has been well documented in the literature, but many unanswered questions remain about the factors controlling erosional responses in different regional settings. The field site for the present study was located in a closed canopy, subalpine forest in Kootenay National Park, British Columbia that was subjected to a high-intensity crown fire in the summer of 2003. Low soil erosion values were documented at the study site in the years immediately following the 2003 wildfire, with estimates ranging from approximately 10 -1 up to 10 0 t ha -1 . Following the wildfire, notable duff coverage (the duff layer is the combined fermentation and humus soil organic layers) remained above the mineral soil. This finding supports earlier studies documenting only partial duff consumption by high-intensity wildfires in the boreal forest of Canada. It is postulated that remnant duff coverage after many high-intensity wildfires impacts the hydrological and soil erosional response to rainstorm events in post-wildfire years. In particular, duff provides detention storage for infiltrating rainfall and, therefore, may inhibit the generation of overland flow. Furthermore, duff also provides a physical barrier to soil erosion. The GreenAmpt model of rainfall infiltration is employed to better assess how interactions between rainfall duration/intensity and soil/duff properties affect hydrological response and the generation of overland flow. Model results show that duff provides an effective zone for detention storage and that duff accommodates all rainfall intensities to which it was subjected without the occurrence of surface ponding. In addition, the penetration of the wetting front is relatively slow in duff due to its high porosity and water storage potential.