Many researchers are interested in the variability of root-respired δ 13 CO 2 as an indication of linkages between belowground plant respiration and canopy processes. Most studies in this area have, however, relied upon the assumption that temporal variability of total soil respired δ 13 CO 2 reflects autotrophic soil processes, but in fact few supporting measurements of purely autotrophic soil respiration (partitioned from total soil respiration) are available. Here we use a combination of physical and isotopic partitioning methodologies to track the variability in autotrophic and heterotrophic soil δ 13 CO 2 at five sites in Eastern Canada during a very dry growing season. Three dimensional modeling of soil isotopic transport dynamics in the static sampling chambers allow us to constrain measurement bias and to eliminate non-steady-state effects as a potential driver of observed variability. We provide experimental results that support a pivotal assumption made in prior interpretations of soil δ 13 CO 2 dynamics: we observed minimal isotopic variability in soil microbial δ 13 CO 2 efflux, but appreciable temporal variability in root-respired δ 13 CO 2 at sites where near drought conditions were observed, suggesting that isotopic discrimination is likely linked to seasonal variations in transpirational demand.