Soil carbon dynamics in a subtropical mountainous region, south china: Results based on carbon isotopic tracing

Carbon cycling within the terrestrial ecosystems is predominant as the most uncertain component in the global carbon cycles (Houghton et al. 1998; Steffen et al. 1998), and is therefore critical in global carbon budgeting (Trumbore et al. 1996; Rosenzweig and Hillel 2000). A large portion of terrestrial carbon resides in soil organic carbon (Malhi et al. 1999; Garten et al. 2000), and carbon storage in soils can be increased by reforestation of agricultural land (Binkley and Resh 1999; Scott et al. 1999) and by the effective management of existing forests (Johnson and Curtis 2001). It is then pressing to decipher soil carbon dynamics for the soils in different climate regimes, due to the Kyoto Protocol (UNFCCC 1997). Soil contributes to a greater extent to total carbon storage than do above-ground vegetation in most forests (Johnson and Curtis 2001). The total amount of soil organic carbon (SOC) in the upper meter of soil is about 1500 × 10¹⁵ g C (Eswaran et al. 1993; Batjes 1996), and the global atmospheric pool of CO₂ is about 750 × 10 ¹⁵ g C (Harden et al. 1992). The CO₂ emission from soil into atmosphere is about 68.0-76.5 × 10¹⁵ g C per year, and this is more than 10 times the CO₂ released from fossil fuel combustion (Raich and Potter 1995). Variations in SOC pools and SOM turnover rates, therefore, exert substantial impacts on the carbon cycles of terrestrial ecosystems in terms of carbon sequestration in soil and CO₂ emission from soil. The distribution of SOC with depth is attributed mainly to continuous input and decomposition of soil organic matter (SOM), and correlates directly with soil development and SOM turnover (Chen et al. 2005). Regional, continental or global models are useful to understand SOM dynamics according to land use changes and management practices (Cole et al. 1996). These models require a thorough knowledge of the distribution of C in different soils and under different land uses practices (Paustian et al. 1997). Quantification of changes in soil carbon dynamics, including SOM turnover rate and distribution of SOC with depth, is therefore critical for determining carbon storage in soils and for modeling soil carbon cycling. The use of natural ¹³C abundance to determine SOM turnover associated with land management (Balesdent et al. 1988; Follett et al. 1997; Collins et al. 1999) and climate changes (Loiseau and Soussana 1999; Hobbie et al. 2002, 2004) is gaining popularity. δ¹³C analysis has become a valuable measure in the study of SOM dynamics (Bird et al. 1996), especially in the regions with records of vegetation shifts between C₃ and C₄ species (Gregorich et al. 1995; Ineson et al. 1996; Boutton et al. 1998; Collins et al. 1999, 2000). The changes in isotopic composition of soil with known and dated vegetation changes are directly related to SOM dynamics (Balesdent 1987, 1990; Martin et al. 1990; Garten et al. 2000). SOM δ ¹³C values correlate well with SOM sources, SOM composition and turnover processes during soil development (Balesdent et al. 1993; Chen et al. 2002a; Powers and Schlesinger 2002; Wynn et al. 2005). The changes in δ¹³C of SOM with depth have several possible explanations (Balesdent et al. 1990; Wynn et al. 2006). One popular explanation is the effect of carbon isotope fractionation due to preferential decomposition of SOM components with different isotopic composition (Benner et al. 1987; Wedin et al. 1995) and kinetic fractionation of carbon isotopes through microbial respiration of CO₂ during SOM decomposition (Mary et al. 1992; Macko and Estep 1984). The spatial and temporal variations of SOM δ¹³C in relation to SOM turnover are then effective proxies for deciphering SOM dynamics. Soil layers with positive SOM δ¹⁴C values contain ¹⁴C produced by nuclear weapon testing ("bomb C") from the 1950s to the 1960s, and the maximal depth that "bomb ¹⁴C" reaches is called "bomb ¹⁴C" penetrating depth (Shen et al. 2001). The ¹⁴C dating results measured with total soil organic carbon are usually prone to be younger, due to addition of new organic carbon during pedogenesis. This kind of ¹⁴C dating result is generally called to be SOM ¹⁴C apparent age (Shen et al. 2000). The SOM ¹⁴C apparent ages of the upper soil layers with SOM δ¹⁴C greater than 0, which can not be obtained directly from measurement, can now be calcu-lated based on SOM ¹⁴C budget model (Chen et al. 2002b). Little is known about the effect of leaching on distribution of SOM with depth, which is unfavorable for evaluating the potential capacity of soil to sequester carbon. Sporopollen (pollen and spores) are abundant in upper soils, and their vertical distributions are controlled substantially by leaching (Zheng et al. 2002). The distribution of sporopollen with depth may be a useful index of leaching potential. We intended to evaluate the effect of leaching on SOM vertical distribution, based on variations in SOC concentration and SOM ¹⁴C apparent age with depth. The distribution of sporopollen with depth can serve as a reference for our evaluation. Five soil profiles at different elevations with specific vegetation composition were selected at the Dinghushan Biosphere Reserve (DHSBR), South China, and soil samples were taken using the thin-layered method (Becker-Heidmann and Scharpenseel 1986). Our aims were to study the spatial and temporal variations of SOM along an altitudinal gradient at the DHSBR that may serve as a substitution of different climate zones, based on SOC concentrations, SOM ¹⁴C dating, SOM δ¹³C values and sporopollen abundance of the soil samples. Studies on SOM dynamics along an altitud-inal gradient in a mountainous region may present clues for deciphering soil carbon cycling in different climate regimes.