Large parts of past atmospheric mercury (Hg) inputs and Hg pollution are estimated to currently reside in surface soil pools. Soils may serve as long-term sources of mercury to surface water bodies possibly leaching to streams, rivers and lakes for decades or even centuries to come. The strong sequestration potential of Hg in surface soils is attributed to a strong absorption capacity of organic matter for Hg2+, specifically to humic and fulvic acids which make soil organic matter the strongest sorbent for Hg in terrestrial environments. Plants also contain considerable amounts of Hg with a midpoint Hg concentration for plant foliage reported at 24 flg kg-1 and lower concentrations for stemwood (average 6.3 ug kg-I). There is overwhelming evidence that Hg allocated in above-ground compartments (i.e., leaves, branches, and stemwood) is of atmospheric origin while root Hg originates mainly from soil uptake.
This evidence is based on (i) reported limited root uptake and translocation of Hg (ii) limited xylem Hg transport (and (iii) combined air/soil exposure experiments where above-ground plant Hg levels are generally a function of applied air Hg levels independent of soil Hg exposures. Thus, plant litterfall Hg represents a significant atmospheric Hg input to ecosystems and Hg associated with above-ground plant biomass-live and dead tissue such as litter or soil organic carbon- represents a large pools of sequestered Hg of atmospheric origin.
The importance of biomass-related Hg pools and cycling in terrestrial systems becomes evident with a simple calculation: Global organic carbon pools are estimated at 455 PgC for plants and 2011 PgC for soils. Assuming an average biomass Hg concentration of 10 ug kg-1 and a 50% carbon content in biomass, the global amount of Hg associated with plant and soil carbon pools yields an impressive 50,000 metric tons of Hg. Since roughly 50% of soil and biomass carbon originates from above-ground production the total global plant and soil organic C pools may be a reservoir for an impressive 25,000 metric tons of atmospheric Hg. This pool is enormous compared to a total atmospheric Hg burden estimated at 2500 metric tons in the lower 5 km of the troposphere.
A similarly estimation can be made for annual atmospheric Hg uptake by plants accounting for 950 metric tons of atmospheric Hg. This represents a continuing sink for almost 40% of the total atmospheric Hg pool every year. Direct measurements of plant-derived Hg inputs of atmospheric Hg to terrestrial systems by means of litterfall have been reported to be considerable. exceeding other deposition fluxes. A review of six forest studies yields an average litterfall Hg input of 16.1 ug m-2 yr-1 compared to open-field wet deposition of 7.5 ug m-2 y-1 .
The large pools of atmospheric Hg currently sequestered in global biomass and organic s(jil carbon pools and associated with annual plant Hg uptake is highly sensitive to changes in global climate, land use change, and ecosystem disturbances. In the past 150 years, land use and climate change led to a total biomass and soil carbon loss to the atmosphere of 121 Pg C (ICPP 2003) and to a 20% decrease in global forest area. The ICPP (2003) lists a variety of scenarios how land-use changes, climate change, and increasing atmospheric C02 are likely to affect plant growth and I productivity, terrestrial carbon uptake, and soil carbon pools in the future.
Although models vary considerably there is general agreement that (1) plant productivity (NPP) will increase until the mid 21 th century due to a fertilization effect of rising atmospheric C02; (2) that increased plant growth will increase terrestrial carbon sequestration until mid 21 th century; and (3) that towards the end of the 21 th century terrestrial carbon sink will decline due to increased heterotrophic respiration in response to global warming; and that (4) the fire frequency will increase/decrease by XX %.
It is currently unclear how changes in biomass and soil carbon pools may have affected or will continue to affect Hg sequestered within these pools. For example, a loss of 20% of the world's forest within the last 150 years-today estimated at 359 Pg C accounts for 180 Pg of biomass and may account for 1,800 metric tons of Hg associated with this carbon pool. It is likely that changes in biomass productivity or standing biomass pools will lead to concurrent changes in associated Hg inputs and sequestered Hg pools, but it is unclear how these processes are related. Losses of biomass and soil carbon pools by fires would likely lead to a complete release of the associated Hg pool to the. However, the fate of Hg during organic carbon mineralization is much more unclear: will Hg remain sequestered in the remaining organic fraction, will Hg be mobilized and potentially leach into surface water bodies, or will Hg emit-or in case of atmospheric origin reemit back-to the atmosphere adding to increased atmospheric Hg pollution along with industrial Hg emissions? There is some indication that organic carbon mineralization increases Hg losses to the atmosphere in sediments in terrestrial soils indicating that carbon mineralization rates could indeed lead to increased volatilization rates with a direct relevance to the atmospheric Hg burden.