Dynamics of forest carbon pools

The different forest carbon pools differ with regard to their dynamics. Usually three ecosystem carbon pools are considered: living biomass, deadwood and litter, and soil carbon (IPCC 2006).

Carbon stores in the soil pool are considered to be more stable and have slow dynamics (Liski, Perruchoud et al. 2002), compared to the carbon stores in the living biomass, but also with longer restoration times, if once disturbed. Some of the soil organic matter persists for millennia (Schmidt, Torn et al. 2011). Factors that may lead to decomposition and release of soil carbon include for example soil preparation and drainage (Sulman, Desai et al. 2013, Wiesmeier, Prietzel et al. 2013). In a longer term perspective, practices leading to a reduced yearly input of carbon, such as declining growth and intensified harvesting of the biomass, may also lead to decreased soil carbon pools (Liski, Perruchoud et al. 2002).

Factors that increase the input of soil organic matter, e.g. via litterfall, or decrease the decomposition of the soil organic matter can lead to accumulation of carbon in the soil. Fertilization, which increases growth, may for example increase the inputs of litter to the soil, while ceasing of drainage will usually slow down the decomposition rate. Soil sequestration rates in Swedish forest soils have been evaluated to be 40-410

56

kg ha^-1 yr^-1, -60-360 kg ha^-1 yr^-1, or -20-730 kg ha^-1 yr^-1, depending on the estimation method (Berg, Gundersen et al. 2007), while Lal (Lal 2008) reports rates from the southeast of the U.S.A. to be 1.2-6 kg ha

-1 yr^-1, and in Canada 12-17 kg ha^-1 yr^-1.

Even if the carbon sequestered in living biomass is, on average, shorter lived than soil carbon it may remain stored for centuries. The expected lifetime of the trees is much influenced by the species, the site conditions, the natural disturbance patterns, and in the case of managed forests, also the management.

The carbon in living biomass is sequestered at faster rates than in the soil. In Danish experiments biomass accumulation rates up to 14-16 tons of dry matter ha^-1 yr^-1 in aboveground biomass for conifers and 4-8 tons of dry matter ha^-1 yr^-1 for broadleaves have been achieved (unpublished data not covering a full rotation). This corresponds to an accumulation of approximately 7-8 and 2-4 tons of C ha^-1 yr^-1 (Graudal, Nielsen et al. 2013). Clonal eucalypt plantations in Brazil probably are close the limits of what wood biomass production systems can produce. Grown in very short rotations of 2 to 6 years, they have been reported to produce as much as 40-80 m³ ha^-1 yr^-1, with a mean annual increment of about 60 m³ ha^-1 yr^-1. Basic densities of vary from about 0.45 to 0.5 dry tonnes m^-3, which give a corresponding average biomass production of 27-30 tons ha^-1 yr^-1 and a carbon sequestration of 14-15 tons of C ha^-1 yr^-1. At the age of 24 months, the diameter at breast height is about 7.5 cm and the stand height about 13 m (Couto, Nicholas et al. 2011).

There is scientific consensus for a general model for carbon dynamics of living biomass in managed forest stands. After planting/regeneration/establishment carbon assimilation evolves exponentially with a slow start. After a period of time net accumulation rates levels off and converges towards zero at maturity (figure 27, upper panel). Normally the soil carbon pools are relatively stable, while there may be some variation in forest floor dynamics. Carbon dynamics of managed forest stands are influenced by location, soil type, water availability, tree species, and management regimes but follow the overall pattern illustrated in Figure 27.

57

Figure 27. Conceptual illustration of carbon dynamics of managed forests. Adopted from (Jandl, Lindner et al. 2007).

More debated are the finer details on how fast new stands grow, their maximum rate of assimilation, and when and if an equilibrium is reached. Some find that unmanaged forests will reach equilibrium, where the carbon assimilation balances the net carbon release from decomposition processes in the soil, leading to zero net assimilation. Others are less inclined to believe that an upper maximum to ecosystem carbon sequestration exist. Declining growth rates with increasing tree or stand age is attributed to a shift in the ratio between photosynthetic activity (gross primary production, GPP) and autotrophic respiration (the plants own respiration to build and maintain different plant tissues) leading the net primary production (NPP) to converge towards zero. Other factors mentioned are reduced photosynthetic efficiency or reduced area of the productive apparatus (green leaves) with increasing age. Finally factors not relating to the physiology of plants, influencing the carbon balance of a forest could be increased mortality in older forests leading to increased heterotrophic respiration (Framstad, Wit et al. 2013). Based on national forest inventories (Luyssaert, Schulze et al. 2008) for example found that even century old forests could act as carbon sinks. The findings are corroborated by other studies using different methods e.g. (D'Amato, Bradford et al. 2011) (permanent sample plots) and (Knohl, Schulze et al. 2003) (detailed carbon flux measurements, so-called eddy covariance studies). Opposing observations are reported by (Lippke, Oneil et al. 2011) from the Douglas Fir region of USA. Here no net assimilation was found in stand older than ~100 years. (Schmid, Thurig et al. 2006) found through multiple model comparisons that in the absence of large-scale disturbances the forest biomass and soil carbon can be increased. Forests can be used as carbon sinks.

58

The sink period was estimated to last for a maximum of 100 years. A recent literature review on carbon dynamics in northern forests confirm the general pattern of carbon dynamics in managed forests (Fig 26), but demonstrate that most northern forests most likely continue to work as carbon sinks far beyond normal rotation age (Framstad, Wit et al. 2013).

8.2.3. Potential contribution from forest-based bioenergy to climate change mitigation