Changes in plant-soil interactions associated with shifts in vegetation composition and climate change may have a range of effects on soil microbial activity, including increases (positive priming), inhibition (negative priming), or no net change in organic matter decomposition. The carbon-rich soils, including peats, of high latitude ecosystems, in particular, are at risk of large carbon losses linked to ongoing vegetation shifts, yet their vulnerability to priming <em>in situ</em> remains unresolved. Here we deploy a field-based technique, which harnesses the contemporary atmosphere as a radiocarbon (¹⁴C) &lsquo;label&rsquo; together with a ¹⁴C-depleted (&lsquo;ancient&rsquo;) peat substrate, to quantify soil organic matter (SOM) decomposition in the presence or absence of roots and rhizosphere processes in subarctic Sweden. Collars encased with different mesh sizes were placed in control and girdled (in which belowground carbon transport from the plant canopy was disrupted) mountain birch forest and willow shrub stands to test the hypothesis that the presence of ectomycorrhizal roots and extra-radical mycorrhizal mycelium increases SOM decomposition through positive priming. As expected, carbon dioxide (CO₂) and dissolved organic carbon (DOC) from root ingrowth cores were significantly enriched in ¹⁴C (contemporary carbon) compared to CO₂ and DOC from root exclusion cores (peat carbon), allowing partitioning of carbon mobilisation between heterotrophic (peat substrate) and recent autotrophic (plant) sources. Neither vegetation community (birch or willow), nor girdling treatment, were statistically significant as main effects, but there was a significant rhizosphere priming effect ratio of 1.36 across all groups; thus, the ancient peat-derived CO₂ flux was 36 % higher in the presence of a rhizosphere than when it was absent. The lack of a significant girdling effect did not support our specific hypothesis that the presence of ectomycorrhizal roots and their associated mycelium increases SOM decomposition, but the substantial variability of modelled ancient CO₂ efflux and DOC concentration during the peak growing season is consistent with the existence of &lsquo;hot-spots&rsquo; of microbial activity. Our study provides a potential alternative to artificial substrate (e.g. glucose) additions, or ¹³C labelling (e.g. pulse-chase), to estimate priming in ecosystems. Furthermore, the study, undertaken <em>in situ</em> in the subarctic, emphasizes that increased primary productivity and associated rhizosphere processes, associated with shifts in vegetation composition and climate change, may not translate simply into increased C sequestration at whole-ecosystem scale.