Emerging databases for soil profiles offer an approach for exploring changes in belowground C that accompany ecosystem shifts. We used steady state models, soil data for bulk and fractionated soil stocks, and radiocarbon data to calculate changes in soil C based on measurements from detrital (free light), aggregate-bound (occluded) and mineral associated (complexed or chemically bound) carbon pools and for bulk soil at incremental soil depths. We explored a space-for-time sequence of increasingly warmer soils from Alaskan Black Spruce permafrost (Gelisols; Mean Annual Temperature at 50cm depth -1.5C), Alaskan Black Spruce non permafrost (Inceptisols; MAT at 50cm +3C ), and Iowa Prairie (Mollisols; MAT at 50cm +9C) developed on similar geologic substrates (loess). These temperature ranges were also representative of 50cm temperatures from model output by CLM for Yr 2014 , Yr 2100, and Yr 2300 for Interior Alaska. The space-for-time concept assumes that the soil will step instantaneously from one ecosystem state to another, with adjustments of input and loss captured by steady state models. Fitting an exponential equation to depth trends in soil C within 2m depths and exploring model output of these fits, we found that e-folding depths were related to depths of rooting and changes in bulk density. The direction and magnitude of the C loss or gain was dictated by the C stocks of initial and final ecosystems. The timing of the loss/gain was dictated by the distribution and turnover time of C in the fractions. Thawing from Gelisol to Inceptisol in loess parent materials resulted in only very small net changes to soil C over 100 years, reflecting both loss of detrital and gain into occluded C forms. Further warming and shifts to the Mollisol resulted in net increases (CO2 sink) , mainly to the detrital C pool from deep roots. These methods enable analysis of large datasets where depth profiles of soil report bulk density, organic C, and have some data or proxy information on C fractions, their turnover times. Without fraction data, these methods provide constraints for changes in belowground net C stocks upon ecosystem shifts, but the timing of changes must be constrained by other methods or assumptions.