© 2015, Springer International Publishing Switzerland. Aims: We tested the hypothesis that vegetation composition and phenology drive both rhizospheric and heterotrophic soil processes, as an alternative to experimental approaches to partition these soil respiration sources. Methods: We compared surface CO2 efflux, depth profiles of CO2 production, extractable SOC and MBC, and soil temperature and moisture along transects between contrasting vegetation types (deciduous riparian trees and adjacent meadow) over the snow-free growing seasons of 2005 and 2006. Results: A dense flush of the nitrogen-fixing forb Melilotus officinalis dominated the meadow in 2005, corresponding with very high rates of CO2 efflux (56 % higher than under trees and 82 % higher than in the diverse meadow community present in 2006). In 2006, proximity to trees was associated with greater surface CO2 efflux and CO2 production to 50 cm depth. SOC was higher under trees than in the meadow (p < 0.0001). MBC was not consistently affected by position relative to trees, but seasonality of MBC was altered by trees as an interactive effect of transect position and sample date (p = 0.01). Conclusion: Community and phenology effects on soil carbon cycling demonstrate the interactions between plant traits and rhizospheric and heterotrophic soil respiration sources. This study highlights a unique case of a highly productive, nitrogen-fixing monoculture exhibiting vastly higher soil respiration rates than a diverse grass-forb community within a single, unmanaged site.