Houghton, R.A., Davidson, E.A. and Woodwell, G.M.  1998.  Missing sinks, feedbacks, and understanding the role of terrestrial ecosystems in the global carbon balance.  Global Biogeochemical Cycles 12: 25-34.

What was done
Using a variety of measurement techniques, Goulden et al. determined the carbon balance of a 120-year-old black spruce forest in Manitoba, Canada.  Houghton et al. reviewed the scientific literature pertaining to global environmental effects on ecosystem carbon storage, analyzing four processes that have been advanced to explain the documented fact that "forest inventories and global carbon budgets based on atmospheric data confirm a terrestrial carbon sink in recent years."

What was learned
Goulden et al. discovered that from 1994 to 1997, the site they studied lost 0.3 ± 0.5 metric ton of carbon per hectare per year.  They also demonstrated that this loss was primarily caused by the thawing of deep soil carbon estimated by the carbon 14 dating method to be several hundred to several thousand years old.  They concluded that "changes in climate that promote thaw are therefore likely to cause a net loss of carbon from evergreen boreal forest ecosystems of the type studied."

Houghton et al. were much less confident about the global response to warming, noting that the existing data pertaining to temporal variability in the global carbon balance "are conflicting with respect to the question of whether increasing temperatures cause a release or storage of terrestrial carbon."  They did, however, suggest that forest regrowth following disturbance, CO2 fertilization, and nitrogen fertilization could each be contributing about one-third of the current 'missing sink' of carbon that we know is going somewhere into the terrestrial biosphere each year.

What it means
It is certainly possible that some ecosystems could experience a temporary net loss of carbon in the face of rising temperatures.  For the globe as a whole, however, this conclusion is far from certain.  Indeed, as the earth has emerged from the global chill of the "Little Ice Age" over approximately the last century, the terrestrial carbon sink has grown larger and larger, as has the CO2 content of the atmosphere.  The simplest explanation would seem to be that the aerial fertilization effect of this increase in the air's CO2 content, together with the fertilization provided by the aerial deposition of nitrogenous industrial effluents, has stimulated the terrestrial biosphere to sequester more and more carbon over this period, perhaps even helped a bit by the rise in temperature itself.  This past performance of the planet is probably the best guide we have for predicting the future; and it suggests that the biosphere should be applying an ever increasing brake on potential global warming as the air's CO2 content continues to rise.