Bark beetle disturbance and nitrogen cycling in conifer forests of Greater Yellowstone

Widespread bark beetle (Dendroctonae) outbreaks are currently affecting conifer forests throughout western North America, yet many ecosystem-level consequences of this disturbance are poorly understood. This study addressed three questions regarding beetle effects on forest nitrogen (N) cycling: (1) What are the decadal-scale changes to N cycling? (2) Does host forest type influence ecosystem response? (3) How does post-outbreak salvage logging change ecosystem response relative to beetle disturbance alone? Field studies were conducted in lodgepole pine (Pinus contorta) and Douglas-fir (Pseudotsuga menziesii) forest in the Greater Yellowstone Ecosystem (Wyoming, USA). Post-beetle outbreak patterns of N cycling in litter, soil and vegetation were consistent with the extensive tree mortality caused by bark beetles. Two to four years after disturbance, N content of both litter and soils approximately doubled. Foliar N of undisturbed vegetation also increased, and was positively correlated with soil N availability. By thirty years after outbreak, soil N availability returned to undisturbed levels though the canopy N pool remained depleted. Lodgepole pine and Douglas-fir forests had qualitatively similar responses to beetle disturbance in the litter-soil profile, though the response of undisturbed vegetation varied. Despite substantial differences in the pre-disturbance N dynamics, both forest types showed 25--50% increases in needle litter N content and soil N pools and fluxes. However, only in lodgepole pine did increased soil N availability lead to increased foliar N of unattacked canopy trees. Minimal increases in nitrification and soil NO3- pools, increased N concentration in undisturbed canopy and understory biomass, and declines in foliar Mn suggest multiple pathways for N retention in beetle-killed forests. Post-beetle outbreak salvage logging had few additional impacts on N cycling during the first year after harvest. Litter inputs, soil N cycling and lodgepole pine live sapling density were unaffected by salvage harvest. Residual tree sapling densities should be sufficient for stand replacement, but substantial differences in coarse wood inputs and understory composition may develop over longer time scales. This study is the first to describe soil, litter, and vegetation N dynamics in beetle-disturbed forests at a decadal time scale, as well as compare ecosystem responses among forest types and management treatments.