International workshop on
Development of Models and Forest Soil Surveys for Monitoring of Soil Carbon
April 5-8, 2006 at Koli, Finland
Carbon sequestration rates in Nordic forest soils - some results from three approaches.
Björn Berg1 *, Per Gundersen1, Cecilia Akselsson2, Maj-Britt Johansson3, Lars Vesterdal1 and Åke Nilsson3.
Carbon sequestration rates to the forest floor (O-horizon) are usually estimated by field sampling e.g. in forest inventories. However, such measurements are costly, time consuming and the results often uncertain. An alternative is to use the concept of calculable stable remains from decomposing litter and by direct measurements. Using Sweden as a case study we made an upscaling of carbon sequestration rates in the humus layer to the forested land covering 27*10 6 ha with mainly Scots pine and Norway spruce. Two different theoretical approaches based on (i) limit values for litter decomposition and (ii) N balance in soil give similar average carbon sequestration rates (180 and ca 100 kg C /ha/yr). The third approach is to directly measure the development of humus depth. Such measurements over 40 years, combined with C analyses give an average rate of 180 kg C ha/yr that is not different from that calculated using limit values.
For the upscaling (limit value approach), annual actual evapotranspiration (AET) was estimated for 17 000 grids of 5 *5km. For the forested area we obtained an annual sequestration of 4.8 *10 6 tons of C in SOM, with an average of 180 kg C ha/yr. The gradient in C sequestration rates follows AET and ranges from 410 kg C ha/yr in southwest to 40 kg C ha/yr in the north. Differences in litter fall and limit values among tree species give different sequestration rates. Norway spruce forests accumulate annually on average 200 kg C per hectare in SOM, Scots pine and the two dominant birch species have an average rate of 150 kg C ha/yr. For beech and oak, limited to south Sweden, the C-sequestration rates were around 400 kg ha/yr.
Using the N-balance method we used the same upscaling approach but used the assumption that the C/N ratio in the accumulating SOM has the same C/N ratio as the SOM layer, the approximate C sequestration rate could be estimated by multiplying the N accumulation by the C/N ratio in the SOM layer.
Sequestration rates based on field sampling (Swed Natl Forest Survey > 600 000 single measurements) were analysed within the same area as the limit value approach (above). Carbon sequestration rates ranged from ca 0 to a maximum of ca 700 kg C ha/yr with an average of 180 kg C ha/yr. Scots pine-dominated forests had a significantly higher sequestration rate (average 263 kg C ha/yr) than those dominated by Norway spruce (178 kg C ha/yr).
Although climate influences the sequestration rates, it appears to be just one of several factors. A Danish tree species trial on soils of different nutrient status showed that a soil factor may be needed for correct prediction using the limit value concept. For instance the sequestered amount of C was negatively related to concentrations of P and Ca in the soil. There was also a clear difference between tree species as regards the influence of the soil factor on C sequestration with species as Norway spruce and common beech being very sensitive, pedunculate oak less so and lodgepole pine being quite unaffected. We are currently analyzing the ability of the limit values concept to predict C sequestration in forest floors of different soil types.
Key words; Stable humus, recalcitrant, carbon sequestration, litter decomposition, limit value, N-balance, direct measurements, upscaling .