Grown forests are carbon stores, not carbon sinks: the carbon is locked into trees and released when the trees die (some of it might get sunk into the soil but IIRC it's a very small fraction). When the forest does not grow, its day-to-day carbon impact is pretty much neutral, growing trees are balanced by decaying dead trees.

There's limited "missed future CO2 absorption" to deforestation, but there's a huge "immediate" released of stored carbon.

I used to think old forests were carbon neutral but an argument prompted me to google, and I found this:

http://www.nature.com/nature/journal/v455/n7210/full/nature0...

http://www.scientificamerican.com/article/old-growth-forests...

http://www.sciencedaily.com/releases/2008/09/080910133934.ht...

Atmospheric carbon per area is about 1 kg/square meter. (x)

It's not inconceivable that trees could store roughly similar orders of magnitude of carbon per area.

  x: 0,3 * 5e18 * 400/1e6 / (7000000*7000000*4*3,14)

> but there's a huge "immediate" released of stored carbon.

Surely that depends on where the tree goes. If it's burned, or left rotting, sure. If it's harvested for building materials, might it not even lead to increase storage of carbon overall.

I think one of the biggest source of deforestation is "slash and burn" subsistence agriculture, so, definitely a CO2 source.

Building material harvesting very rarely the primary purpose of deforestation.

My recollection from school years is that the fraction that goes to the soil is significant, but I don't know about more recent data...