3.4.2 Effects of leakage of CO2 in the subsurface (soils)

The study of natural analogues for CO2 leakage in the near surface (in soils) revealed that there is a strong negative correlation between O2 and CO2 contents. CO2 progressively replaces O2 and creates an almost anoxic environment (Gal et al., 2011). Therefore, the primary impacts of a CO2 leakage in the near surface would be soil acidification and toxicity. This can furthermore lead to plant stress or death and disturbances of animal or microbial activity in the sub-surface. Although plants have a higher tolerance for CO2 than animals, prolonged leaks could suppress respiration in the root zone (Damen et al., 2005). It is estimated that soil CO2 levels above 10-20% inhibit root development and decrease water and nutrient uptake (IPCC, 2005). The degree of impact depends though very much on the sensitivity of species, the rate and geometry of leakage. Some species could adapt and recover in time or they could be more tolerant to soil chemistry changes (Kirk, 2011; Al-Traboulsi et al., 2012).

A complex two seasons study made at a naturally gas vent within Latera geothermal field emphasised the significant impact of CO2 leakage in the soil. In the gas vent area (6 m diameter) the vegetation is absent (Fig. 3-3), pH is very low (3.5) and there are changes in mineralogy and bulk chemistry (Beaubien et al., 2008). In the area with a soil CO2 concentration varying from 5 to 40 % at 10 cm depth, only grasses were found growing, demonstrating their increased tolerance for CO2.

O. Fig. 3-3

Fig. 3-3: Location map (a) and photograph of the studied gas vent (b) (Beaubien et al., 2008).

Approximately the same results for the botanical survey were found also at the Laacher See site (Germany) in 2008. Although dicotyledonous plants generally do not seem to tolerate high concentrations of CO2 in soil, plants from this group (Polygonatum arenastrum) were observed at this site on transects where CO2 concentration was 10-35% at 15 cm depth and 35-90 % at 60 cm depth and actually absent in areas of decreased CO2 concentrations (Krüger et al., 2011). Therefore, Krüger et al. (2011) concluded that this plant could be used as a bio indicator for high CO2 concentrations in soil at this particular site and that the botanical impact of CO2 leakage is site specific and depends on factors as soil moisture and pH as well as plant species.

Another example of the botanical effects of the CO2 leakage in the near surface is offered by the case of Mammoth Mountain, California, USA where a large area of coniferous forest was killed due to a diffuse magmatic CO2 emissions (Farrar et al., 1999) that started prior to 1990 based on radiocarbon measurements of tree rings (Cook et al., 2001). The total amount of magmatic CO2 emitted in 1996 was estimated to around 530 tonnes per day (Farrar et al., 1999). Soil gas readings showed CO2 levels up to 95% in 1994 and levels of 15-90% in 2001 (IPCC, 2005). Average CO2 flux rates in the affected areas were around 300 tonnes per day in 1996 (Cook et al., 2001) and 90-100 metric tonnes per day in 2001 (in the largest affected area, Horseshoe Lake) (IPCC, 2005). In addition to soil gas surveys, airborne remote sensing is also used to map tree health in this region.

A less studied effect of a CO2 leakage in the near surface is the one associated with microorganisms. The survey made at a Laacher See vent in 2008 showed differences in the microbial activity and microorganism numbers on soil samples collected from CO2 rich areas and control points with background CO2 concentrations (Krüger et al., 2011). A decrease in the number of bacteria towards the centre of the vent and an increase of archaea number in the soils with higher CO2 concentration was observed. The authors concluded that the rise of CO2 concentration in soil led first to microaerobic and eventually to anaerobic conditions favouring the development of methane producing or sulphate reducing bacteria communities, thus the ecosystem adapted to CO2enrichment by substitution of species (Krüger et al., 2011).