Regional soil organic carbon storage estimates for western Oregon by multiple approaches

Soil is an important factor in regional and global C budgets because it serves as a reservoir of large amounts of organic C. In our study, we compared six approaches of estimating soil organic C (kg C m(-2), not including the surface organic horizon, hereafter called soil C) and its spatial pattern in the mountainous, largely forested western Oregon region. The approaches were (i) USDA NRCS pedons, (ii) other pedons, (iii) the State Soil Geographic Data Base (STATSGO), (iv) the United Nations Soil Map of the World, (v) the National Soil Geographic Data Base (NATSGO), and (vi) an ecosystem-complex map. Agreement between approaches varied with scale. For the entire region (10(5) km(2)), estimates of average soil C varied from 4.3 to 6.8 kg C m-2 for the 0- to 20-cm depth and from 12.1 to 16.9 kg C m(-2) for the 0 to 100-cm depth. At the subregional scale (approximate to 10(4) km(2)), all approaches indicated higher soil C in the coastal area than in the inland southern area, but relative amounts in other subregions varied among the approaches. At the subsubregional scale (approximate to 10(3) km(2)), soil C was consistent between individual STATSGO map units and NRCS pedons within those map units, hut there was less agreement with other pedons. Rigorous testing of soil-C maps requires data from pedons that are located by objective criteria, in contrast to the subjectively located pedons now available. The uncertainty associated with regional soil-C amounts and spatial patterns should be considered when soil-C maps are integrated into regional or global assessments of physical and biotic processes because simulation-model outputs may be sensitive to soil C.