GC33E-0580:
Recovery Time After a Late-Dry Season Fire: the Effect on Fluxes, Surface Properties and Vegetation Green-Up.

Wednesday, 17 December 2014
Michael Vijay Saha1, Paolo D'Odorico2 and Todd M Scanlon2, (1)University of Virginia Main Campus, Charlottesville, VA, United States, (2)Univ Virginia, Charlottesville, VA, United States
Abstract:
Large regions of Africa burn on an annual basis. These fires damage vegetation, change surface albedo and modify the hydrologic cycle. Quantifying the magnitude and persistence of these changes is key in understanding the complex ways in which fire affects ecosystem functioning at smaller scales and will inform ongoing modeling efforts. We report the results of a field study in a semi-arid savanna in northern Botswana during the transition from dry to wet season (Oct-Dec) in 2012 and 2013. The goals of this study were to: (1) characterize the multifaceted effect that late dry-season fires have on fluxes and radiative surface processes during green-up, and (2) describe the timescales over which these variables recover to non-burnt levels. Our study synthesizes a suite of data, including flux tower measurements, vegetation sampling, time-lapse photography and concurrent remotely sensed variables over plots with variable burn patterns.

Albedo decreased immediately after fire, converging on unburned values 10 days post-burn. The magnitude and direction of this response was comparable to the albedo change elicited by strong rainfall events. Soil temperature and soil heat flux were not significantly modified by fire. Carbon fluxes showed no discernible difference from an unburned control site immediately after fire. There was a small burst in ecosystem respiration at immediately following the first post-fire rainfall event, returning to baseline values after 3 days. Persistent CO2 release, which we attribute to soil respiration, occurred for 10 days after successive strong wetting events, confirming the centrality of available moisture in determining ecosystem function. Fire delayed the green-up in some plots, but this effect was variable and short-lived. One month after fire there was no evidence of a difference in ground observations of greenness between burnt and control plots or plots that differed in their time of burning. We attribute the relatively ephemeral effects of fire to the low intensity ground fires at our site. Preliminary results of a regional climate model incorporating our ground-based measurements are presented and the implications of these results on the understanding of the climatic role that fire plays in southern Africa are discussed.