Boreal forests often have a dense understory of ericaceous dwarf shrubs with ecological adaptations that contrast those of the canopy-forming trees. It is therefore important to quantify contributions by understory shrubs to ecosystem processes and disentangle shrub- and tree-driven responses, and their interactions, to climatic factors. We quantified soil respiration driven by the pine canopy and the ericaceous shrub understory over three years, using a factorial pine root-exclusion and shrub-removal experiment in a mature Pinus sylvestris forest. Soil temperature and moisture-related responses of respiration attributed to autotrophic (shrubs, pine roots) and heterotrophs were compared. Additionally, we assessed effects of interactions between these functional groups on soil nitrogen availability and respiration. Understory shrubs accounted for 22 ± 10% of total autotrophic respiration, reflecting the ericaceous proportion of fine root production in the ecosystem. Heterotrophic respiration constituted about half of total soil respiration. Shrub-driven respiration was more susceptible to drought than heterotrophic- and pine-driven autotrophic respiration. While the respiration attributed to canopy and understory remained additive, indicating no competitive release, the plant guilds competed for inorganic N. Ericaceous understory shrubs accounted for a small, yet significant, share of total growing season soil respiration. Overlooking understory respiration may lead to erroneous partitioning and modelling of soil respiration mediated by functional guilds with contrasting responses to soil temperature and moisture. A higher share of activity of both heterotrophs and pine roots, under drought conditions, could have important implications for soil organic matter accumulation and decomposition as the climate changes.

At the end of November 2016, we initiated a fully factorial, randomized-block experiment to distinguish respiration fluxes driven by dwarf shrubs, pine trees and free-living decomposer. The experiment included factorial combinations of pine root exclusion by trenching and ericaceous shrub removal. Eight replicate blocks with five treatment plots in each (1.2 m x 1.2 m, spaced by 5-10 m) were set up, giving a total of 40 plots. The five treatments encompassed a control (C, representing the sum of all soil respiration fluxes; i.e. pine trees, dwarf shrubs and decomposers), shrub removal (E), pine root exclusion (T), combined pine root exclusion and shrub removal (TE, represented as ‘heterotrophic’ respiration), and a disturbed control (DC) in which pine roots and hyphal networks were severed but allowed to re-establish in the plot from the edges.