The influence of fire history on soil nutrients and vegetation cover in mixed-severity fire regime forests of the eastern Olympic Peninsula, Washington, USA

The rain shadow forests of the Olympic Peninsula exemplify a mixed-severity fire regime class in the midst of a highly productive landscape where spatial heterogeneity of fire severity may have significant implications for below and aboveground post-fire recovery. The purpose of this study was to quantify the impacts of wildfire on forest soil carbon (C) and nitrogen (N) pools and assess the relationship of pyrogenic carbon (PyC) to soil processes in this mixed-severity ecosystem. We established a 112-year fire chronosequence with nine similar forest stands ranging in time since lastfire (TSF) from 3 to 115 years prior to site establishment. At each site, we measured understory vegetation cover, overstory composition, physical and chemical attributes of surface mineral soils to a depth of 10 cm, and forest floor organic matter. Additionally, non-ionic resin lysimeters were buried over the winter and spring (7–8 months) at the interface of organic and mineral soil to collect O-horizon leached DOC that would potentially contact PyC particles on the forest floor. Nitrogen transformations were also monitored in laboratory soil incubations for a subset of sites. The TSF gradient was significantly correlated with PyC mass in the O-horizon (r = −0.4), O-horizon C (r = 0.4), total phenol content in both O-horizon (r = 0.4) and mineral soils (r = 0.2), and potentially mineralizable N (PMN) (r = 0.4). Recent fire sites contained higher mineral soil total N and inorganic available N, but were not correlated with TSF. Total DOC that accumulated on the non-ionic resins averaged 1.14 (SE ± 0.54) g DOC m−2 year−1 and increased with TSF (r = 0.52; p < 0.0001). Over time, soils appeared to shift toward a more phenolic-rich organic and surface mineral soil, a higher PMN index, denser moss cover, and higher cover of Mahonia nervosa and Rosa gymnocarpa. Multivariate, non-parametric analysis of soil and vegetation factors showed a significant relationship with the TSF gradient between sites (p < 0.001), but not within sites. Soil characteristics were found to be less sensitive to wildfire disturbances than aboveground vegetation composition.