1. The main controls of soil organic carbon in Russian cryolithozone O.V.Chestnykh Forest Ecology and Production Center of Russian Academy of Sciences, Moscow We used an original database "Soils of the Northern Eurasia", including descriptions of more then 1000 soil pits (Fig. 3). For each pit description the base includes information on 87 parameters, combined to next groups: descriptive, physical-chemical features, gross chemical content. We selected from database the correspondent information to cryolithozone of Russia, including all tundra and forest tundra subzones, northern taiga subzone of Siberia and middle taiga zone of the East Siberia. Geographical latitude and longitude, soil type, bedrock type, natural subzone and vegetation type were amongst the analyzed factors. The reserves of soil organic matter were calculated for each soil pit. Horizon thicknesses, bulk weights of horizons, soil organic matter content (%) in horizons were taken into account. All these data were summarized by horizons to characterize a particular soil pit. The reserves of soil organic carbon were estimated by humus content. Empirical data were grouped by selected factors and averaged with standard errors. Fig 2. Geographical distribution of C soil density in taiga zone of Russian cryolithozone. The most evident was a latitudinal trend (Fig. 4). Soil carbon reserves linearly decreased with latitude increase from 162.3 tC ha -1 at 56  N to 49.7 tC ha -1 at 76  N (the step in latitude was 4 o ). The longitudinal trend (Fig. 5) within the range of 60-150 o E and 30 o step wasn't found to be apparent. Fig. 4. Average C soil density vs. latitude ranges. Fig. 5. Average C soil density vs. longitude ranges. Amongst bedrocks (Fig. 6) the least carbon reserves were found on sands (106.8 tC ha -1 ), with intermediate reserves on loams and maximal on clay (193 tC ha -1 ). We found a decrease of carbon reserves in the sequence of the following soil types (Fig. 7): "sod – peat – brown – podzolic – gleic - grey" within the range from 184.8 tC ha -1 to 100.0 tC ha -1. Fig. 6. Average C soil density vs. bedrock types. Fig. 8. Average C soil density vs. geographical subzones. We consider the range of carbon storage variations as a characteristic of relative importance of a particular factor. These ranges in the order of descending importance were the following: latitude (113 tС ha -1 ), bedrocks type (86 tС ha -1 ), soil type (85 С ha -1 ), ecosystem type (53 tC ha -1 ), natural subzone (39 tC ha -1 ) and longitude (21 tC ha -1 ). Soil carbon reserves (Fig. 8) in tundra subzones (arctic, typical, south) varied in the range of 95-105 tC ha -1, whereas in forest tundra and taiga subzones at 141-163 tC ha -1. Soil carbon storage decreased within the following sequence of the main ecosystem types (Fig. 9): meadows, forests, bogs, tundra (165.8, 142.1, 130.2 and 112.6 tС ha -1, respectively). Fig. 9. Average C soil density vs. vegetation types. CONCLUSION It is widely accepted by soil scientists that for proper estimation of soil carbon reserves it is necessary and sufficient to consider soil or landscape type maps. Our basic point is to add into consideration other important factors and to create the specific map, summarizing information on bedrock and soil type with additional latitudinal subdivision. Fig. 3. The profile of tundra gleic soil of south tundra subzone (Vorkuta region). SELECTED PUBLICATION D.G. Zamolodchikov, D.V. Karelin, O.V. Chestnykh. Phytomass reserves and primary productivity profiles of the Russian Nopth from Cola Peninsula to Chukotka. In: Global Change and Arctic Terrestrial Ecosystem. ECSC - EC -EAEF, Brussels, Luxemburg, 1995. P. 191-200 O.V. Chestnykh, D.G.Zamolodchikov, A.I. Utkin, G.N. Korovin. Distribution of Organic Carbon Reserves in Soils of Russin Forest. Lesovedenie. 1999. No 2. P. 13-21 (in Russian). O.V. Chestnykh, D.G. Zamolodchikov, D.V. Karelin. Resources of organic Carbon in the Soils of Tundra and Forest-Tundra Ecosystems in Russia. Russian Journal of Ecology. 1999. V. 30. No 6. P. 392-398. D. G. Zamolodchikov, V.O. Lopes de Gerenu, D.V.Karelin, A.I.Ivashchenko, O.V.Chestnykh. Сarbon Emission by the Southern Tundra during Cold Seasons. Doclady Biologocal Sciences. 2000. V. 372. P. 312-314. Correct estimates of soil organic carbon reserves are very important in construction of the global and regional carbon budgets. The general methodical approach to receive above estimations can be called cartographic. The procedure is next: a selection of some geographical base; collecting of empirical data, corresponding to different polygons of map; averaging of data within polygons; multiplying of average values on polygon areas; summarizing of all polygon values. In the case of soil carbon estimation the maps of soil types usually are used as main geographical base. From the other side, the carbon storage in soil is extremely variable even in same soil types. In our work we used also other types of geographical bases, as map of landscapes (Fig. 1) or map of ecoregions (Fig. 2). The objective of present work was to find the factors, which determine maximal variability of soil carbon storage values in initial data set. It is evident, that these factors need to be taken into account in developing of appropriate geographical base for soil carbon estimations. Fig 1. Geographical distribution of C soil density in tundra zone of Russian cryolithozone. Fig. 7. Average C soil density vs. soil types.