Soil carbon changes at a clear-cut forest site
Need for measurements of changes in soil carbon
● Repeated measurements have only rarely been used to detect changes in the
carbon stock of forest soils, probably because it is difficult to distinguish these
changes from the large spatial variability in soil carbon density.
● These measurements are however needed for validating various modelcalculated estimates and planning strategies to monitor changes in soil carbon
stocks.
Jari Liski and Elli Haapamäki
Finnish Environment Institute
Research Programme for Global Change
P.O.Box 140, FIN-00251 Helsinki, Finland
firstname.lastname@ymparisto.fi
Repeated sampling at a pine site
● We collected 125 soil cores from a 6 x 8 m area at a Scots pine site in southern
Finland in 1993 and 2005 to measure changes in the carbon stock of the soil
(Figs. 1 and 2).
● In 1995, the site was clear-felled and the soil was prepared lightly to aid natural
regeneration.
● To study the soil carbon changes at different scales, we collected the samples
from the meter scale (n=45), the dm scale (n=2x29) and the vicinity of trees
(n=3x9).
● We took the samples in 2005 15 cm away from the earlier samples and estimated
the soil carbon changes based on these pairs.
Loss of carbon from the organic layer, gain in the
mineral soil
FIG 1: Locations of the sampling points (m scale - black, dm
scale - red, tree points - blue; 1993 - dark, 2005 - light), trees
in 1993 (large dots) and tracks of soil preparation (organic
layer removed – crosses, organic layer deposited – dotted).
FIG 2: The study site in 1993 (above) and 2005 (below, the men standing on the stumps of
the earlier trees).
● The mean amount of carbon had decreased significantly in the organic soil layer
but increased in the 0-10 cm mineral soil layer (Fig. 3).
● These changes were rapid (on average 52 g m-2 year-1 in the organic layer and
17 g m-2 year-1 in the 0-10 cm mineral soil layer) and substantial compared to
the mean soil carbon values in 1993 (1.9 kg m-2 in the organic layer and 1.4 kg
m-2 in the 0-10 cm mineral soil layer).
● Deeper in mineral soil, the amount of carbon had decreased on average but
this was statistically significant in only two subsets of samples.
Soil preparation caused the lowest carbon values
● The organic soil layer had lost carbon especially there where the soil was
prepared and the organic layer was removed (Fig. 4). In these locations, the
0-10 cm mineral soil layer had gained less carbon than elsewhere or even lost
some of it.
● The locations of depositing the organic layer did not contain more carbon than
the untounched soil.
FIG 3: Change in amount of carbon (mean and 95 % confidence limits)
by soil layer and subsample (m scale - black, dm scale - red, tree points
– blue).
FIG 4: Soil preparation effects on the soil carbon changes (mean and
95 % confidence limits) in the organic and 0-10 cm mineral soil layers
(m scale – black, dm scale – red, untouched soil – dot, organic layer
removed - square, organic layer deposited - triangle).
Spatial variability in the soil carbon changes
● The organic soil layer had lost carbon especially under the former trees where
the carbon density used to be high (Fig. 5). The removals of the organic layer
caused by soil preparation are also seen on the map as low present soil carbon
values.
● In the 0-10 cm mineral soil layer, the carbon changes were patchy. However,
some of the largest carbon losses and the smallest gains were associated with
the tracks of soil preparation.
Conclusions
● The carbon stock of soil may change remarkably over a short period of time.
● Different soil layers must be sampled adequately to avoid biased results because the carbon stocks of the layers may change to the opposite directions.
● Soil preparation is an important cause for heterogeneity in soil carbon at recently
processed sites.
● The role of ground vegetation in the carbon balance of soil requires further
study.
FIG 5: Soil carbon density (kg m-2) in 1993 (left), 2005 (middle) and the change (kg m-2) in the soil carbon (right) in the organic layer (upper row) and the 010 cm mineral soil layer (lower row). The dots indicate the location of trees in 1993.
Acknowledgements
This study was financed by the European Commission through the Forest Focus pilot project ”Monitoring changes in the carbon stock of forest soil” (www.metla.fi/hanke/843002). We thank Hasse Hyvärinen for help in field
and laboratory and Hannu Rita for advice on the statisctics.