Skid trail use influences soil carbon flux and nutrient pools in a temperate hardwood forest

Skid trails comprise up to 40% of managed forest areas and are subject to wide variations in disturbance intensity. Despite potential for accelerated carbon (C) and nutrient losses following use, they are underrepresented in studies of forest soil biogeochemistry. To address this, we assessed soil chemistry and respiration as CO2 efflux (FCO2) on primary and tertiary skid trails relative to adjacent cut forest (control) between May and October, following a winter partial-harvest in a northern hardwood forest. Primary trails were highly mixed and compacted (bulk density: ::approximate to 1.18 g cm(-3)) relative to forest controls (approximate to 0.53 g cm(-3)) and tertiary trails (approximate to 0.59 g cm(-3)). FCO2 rates corrected for higher trail temperatures (+2.5 C) were half that of controls. An absence of root respiration likely accounted for most of this difference, with precipitation event-based moisture saturation and labile C limitations for heterotrophs accounting for the remainder. A similar but weaker pattern was found for tertiary trails: after correcting for higher temperatures (+0.7 degrees C), trails produced 23% less FCO2 than controls in ruts only, suggesting compaction and/or root damage inhibited FCO2. Concentrations of soil C, N, and cations were 16-52% lower on frequently re-used primary skid trails and 13-30% lower on tertiary skid trails than adjacent forest controls, but 0-20 cm pool sizes were similar amongst treatments. All treatments, except for primary skid trails, expressed substantial inter correlated declines in soil C, N, and cations (-7% to 40%) by October that were also correlated to FCO2 rates (r(2) = 0.10, p < 0.05), suggesting losses were related to enhanced decomposition. Overall, both cut forest and tertiary trails may be areas of substantial short-term C loss, with greater potential net losses on denuded trails. Conversely, primary skid trails were likely small but persistent atmospheric C sources due to chronically inhibited regrowth, yet paradoxically provided greater SOC retention than tertiary trails and controls. Although better long-term measurements of C and nutrient flux are needed, skid trails comprise discrete management units that may respond predictably to use. Inclusion of these as components in models of forest biogeochemistry may improve the accuracy of flux predictions. Crown Copyright (C) 2017 Published by Elsevier B.V. All rights reserved.