delta C-13 values of soil organic carbon and their use in documenting vegetation change in a subtropical savanna ecosystem
Academic Article
Overview
Research
Identity
Additional Document Info
Other
View All
Overview
abstract
Plants with C3, C,4, and CAM photosynthesis have unique 13C values which are not altered significantly during decomposition and soil organic matter formation. Consequently, 13C values of soil organic carbon reflect the relative contribution of plant species with C3, C4, and CAM photosynthetic pathways to community net primary productivity, and have been utilized to document vegetation change, to quantify soil organic matter turnover, and to refine our understanding of earth-atmosphere-biosphere interactions. Here, we review the basis of this methodology, and illustrate its use as a tool for studying grass-woody plant dynamics in a savanna ecosystem. In the Rio Grande Plains of southern Texas, C4 grasslands and savannas have been largely replaced by C3 subtropical thorn woodlands dominated by Prosopis glandulosa. We used 13C values of soil organic matter, above- and belowground plant biomass, and litter in conjunction with radiocarbon dating and dendrochronology to test the hypotheses that: (1) C3 Prosopis groves in uplands and C3 Prosopis woodlands in low-lying drainages have been long-term components of the landscape; and (2) Prosopis woodlands of low-lying drainages have expanded up-slope since Anglo-European settlement. Current organic matter inputs were not in isotopic equilibrium with soil organic carbon in any of the patch types sampled. In upland grasslands, 13C values of vegetation (-20) were lower than those of soil organic matter (-17), suggesting increased C3 forb abundance in response to long-term, heavy grazing (herbaceous retrogression). In wooded landscape elements, 13C values of current organic matter inputs were characteristic of C3 plants (- 28 to - 25), while those of the associated soil organic matter were typically -20 to -15. These 13C values indicate that woodlands, groves, and shrub clusters dominated almost exclusively by C3 plants now occupy sites once dominated by C4 grasses. A particularly strong memory of the C4 grasslands that once occupied these sites was recorded in the 13C values of organic carbon associated with fine and coarse clay fractions (- 18 to -14), probably a consequence of the slow organic carbon turnover rates in those soil fractions. When 13C values of soil organic carbon were evaluated in conjunction with radiocarbon measurements of that same carbon, it appeared that herbaceous retrogression and a shift from C4 grassland to C3 woodland occurred recently, probably within the last 50-100 years. Demographic characteristics of the dominant tree species corroborated the 13C and 14C evidence, and indicated widespread establishment of P. glandulosa and associated shrubs over the past 100 years. Together, these data provide direct, spatially explicit evidence that vegetation change has occurred recently across the entire landscape at this site. Environmental conditions where C3, C4, and CAM plants coexist (e.g., dry, alkaline soils) generally do not favor the preservation of pollen and phytoliths, and these same areas usually lack historical records of vegetation change. Consequently, vegetation dynamics have been difficult to quantify in grasslands, savannas, and woodlands. However, our results demonstrate clearly that 13C values of soil organic matter afford a direct and powerful technique for reconstructing vegetation change in these areas.
