Distribution, transport, and accumulation of pyrogenic black carbon (PyC) in post-wildfire watersheds

Wildfire is a powerful agent of change, capable of vastly altering the geomorphology, ecology, and hydrology of an area. Wildfire can also adversely affect human and ecological communities that depend upon forested watersheds for water supply. Pyrogenic black carbon (PyC) is generated in wildfires and its distribution, transport, and accumulation has implications for water quality, the carbon cycle, and contaminant transport. PyC is an encompassing term, including any material of fire-derived origin (soot, char and ash). The study area is in the Jemez mountains of northern New Mexico, which has a history of fire suppression and logging, resulting in overcrowded forests vulnerable to wildfire. We studied two recent fires, the Las Conchas Fire (LC) of 2011 and the Thompson Ridge (TR) fire of 2013. The results of this study added to our understanding of PyC accumulation and transport and to further development of a conceptual model.;A common method of measuring black carbon (BC) in soils is the chemo-thermal oxidation method (CTO-375). This method systematically eliminates forms of carbon (non-pyrogenic organic carbon (NPOC) and inorganic carbon (IC)) and ultimately distinguishes the remaining carbon within a sample as BC. Because the CTO-375 method narrowly quantifies the most condensed forms of BC, we adapted a few steps and details of the standard procedure to better quantify PyC (which includes less condensed forms of BC). If comparing TOC and PyC concentrations to other studies using different methods, we recommend using a conversion factor of 0.95 and 0.75 for TOC and PyC respectively.;For investigating the impact of fire on soil PyC concentrations, we did not use a conversion factor because we were able to draw conclusions by comparing PyC concentrations in data that we collected from pre- and post-fire soil samples. Pre-fire soil samples were collected in 2009 and 2011, and post-fire samples were collected at the same locations in 2014, 15 months following the TR fire. The postfire samples had greater and more variable PyC concentrations than the pre-fire samples. When comparing samples that were grouped by factor (including soil horizon, slope degree, slope aspect, watershed, burn severity, and geomorphic setting) the TR fire showed a distributing effect; with greater PyC concentrations in all post-fire samples, the fire appears to have homogenized PyC concentrations and evened out differences that existed before the fire.;The soil horizon, watershed, burn severity, and geomorphic setting proved to be the most influential factors on PyC concentration changes. The O horizon PyC concentrations were more variable and more responsive to fire than the A horizon PyC concentrations. While the O horizon may be a larger sink for PyC in the short term following fire, the A horizon appears to be a more stable sink for PyC in the longer term.;The Redondo watershed, which experienced a higher degree of burn severity and is comprised of steeper slopes than the other two watersheds, had greater post-fire PyC concentrations than pre-fire PyC concentrations. Therefore, in considering impacts of a low-severity burn such as the TR fire within short timescales, watershed characteristics are likely a reliable predictor of a watershed's response to the fire, and of its tendency to transport PyC. Burn severity appeared to impact soil PyC concentrations at the local sampling site and at the watershed scale. Hill-slope samples had greater post-fire PyC concentrations than pre-fire concentrations, which was not the case for riparian area samples. Hence, hillslopes appear to be an important storage and accumulation environment for PyC, at least within this short time frame.;We also measured erosion and deposition of sediment on four hillslopes: two hillslopes burned by the TR fire, one hillslope burned by the LC fire, and one control hillslope within the Rio Cebolla watershed. Using sediment fences and erosion pins to quantify and compare mass movement, we came to the following conclusions: 1) the burned TR hillslope was subject to greater erosion than the control hillslope, 2) the monsoon season proved more influential for erosion and deposition than did the snow melt season, 3) positioning on the hillslope is important for distinguishing erosion and depositional trends, and 4) variability in the erosion pin data points to small-scale heterogeneity.