[Interannual changes in PAR and soil moisture during the warm season may be more important for directing of annual carbon balance in tundra than temperature fluctuations].

A lot of studies on the impact of global climate changes on natural communities deal with cryogenic ecosystems, tundra in particular, since they are delimited by low air temperature and permafrost, thus being extremely sensitive to long-term climate fluctuations. Continuous warming in Northern Hemisphere is unmasking all the more details concerning complex system of direct relationships, feedbacks, and interactions of carbon balance factors as the main response function. While the set of such factors may be viewed as more or less complete, their relative contribution to C-balance, as is becoming clear with accumulating results of field observations, directly depends on temporal scale of observations and is not constant. As the results of field observations and modeling of tundra ecosystems show, any one of significant factors can become the leading one within the boundaries determined by the given scale of observations. Even the least significant factor can become the determining one for direction of carbon annual net flux in an ecosystem, if contributions of more significant factors canceled each other during the period of observations. In the most general situation, the greater is the variation of a significant factor during the period of observations, the larger is its partial contribution. The complete set of independent variables of C-balance is not limited by abiotic factors but should include such an important factor as a stock of plants living top mass, which can be treated as not only the natural product of C-balance but also as its independent parameter.

[Geoinformational model of the carbon reserve of the Russian tundra zone]. / Geoinformatsionnaia model' biudzheta ugleroda tundrovoi zony Rossii.

We present a summary of our own gas measurement data on biogenic fluxes of CO2 in various ecosystems of the Russian tundra zone, using seasonal and geographical extrapolations based on mathematical simulation modeling. The model follows construction principles of geoinformational systems and consists of (1) a computer map of tundra landscape boundaries; (2) a meterological database; (3) a model of changes of the phytomass; and (4) the actual model of carbon fluxes. The model, which uses a 10-day step, allows estimates of regional and landscape-associated macroscopic carbon fluxes and predicts the response of the tundras to climatic changes. The data obtained are in good agreement with independently obtained estimates of the major characteristics of tundra carbon balance.