Key points are not available for this paper at this time.
Abstract Archived soils can provide valuable information about changes in the carbon and carbon isotope content of soils during the past century. We characterized soil carbon dynamics in a Russian steppe preserve using a 100‐year‐old‐soil archive and modern samples collected from the same site. The site has been protected since 1885 to the present, during which time the region has experienced widespread conversion to cultivation, a decrease in fire frequency, and a trend of increasing precipitation. In the preserve, the amount of organic carbon did not change appreciably between the 1900 and 1997 sampling dates, with 32 kg C/m 2 in the top meter and a third of that in the top 20 cm. Carbon and nitrogen stocks varied by less than 6% between two replicate modern soil pits or between the modern sites and the archive. Radiocarbon content decreased with depth in all sites and the modern SOM had positive Δ values near the surface due to nuclear weapons testing in the early 1960s. In the upper 10 cm, most of the SOM had a turnover time of 6–10 years, according to a model fit to the radiocarbon content. Below about 10 cm, the organic matter was almost all passive material with long (millennial) turnover times. Soil respiration Δ 14 CO 2 on a summer day was 106–109‰, an isotopic disequilibrium of about 9‰ relative to atmospheric 14 CO 2 . In both the modern and archive soil, the relative abundance of 13 C in organic matter increased with depth by 2‰ in the upper meter from δ 13 C = ‐‐26‰ at 5 cm to ‐‐24‰ below a meter. In addition, the slope of δ 13 C vs. depth below 5 cm was the same for both soils. Given the age of the soil archive, these results give clear evidence that the depth gradients are not due to depletion of atmospheric 13 CO 2 by fossil fuel emissions but must instead be caused by isotopic fractionation between plant litter inputs and preservation of SOM. Overall, the data show that these soils have a large reservoir of recalcitrant C and stocks had not changed between sampling dates 100 years apart.
Torn et al. (Fri,) studied this question.