B12B-01 INVITED
Progress, Potential and Pitfalls in the Use of Bomb 14C to Constrain Soil Carbon Dynamics
Forty four years have passed since atmospheric testing of thermonuclear weapons injected a major 14C spike into the atmosphere-biosphere-hydrosphere system. The use of bomb 14C, in combination with millennial decay of 14C, remains the most effective empirical tool for constraining rates of carbon (C) cycling in soils at timescales beyond experimental manipulations (>5 years). In the last 20 years, accelerator mass spectrometry has greatly increased the potential and throughput of soil 14C studies. At present, atmospheric Δ14C appears to be stabilizing at more constant values as a result of reinjection of bomb 14C from decadal storage in forests and soils. This means that current and future studies using bomb 14C have different sensitivities and uncertainties compared to those carried out during periods of rapid Δ14C decline such as the 1970s, 80s and 90s. Bomb 14C proves most effective when archived soil samples are available: simply using bulk Δ14C from samples collected at two or more times can surpass single time point Δ14C from soil fractions in providing robust C cycling rates. Of course, measurement of Δ14C in soil fractions from time series samples can significantly improve estimates of C cycling parameters. Samples collected between ca. 1965 and 1995 have now greatly surpassed pre-bomb samples in utility, although pre-bomb samples retain considerable usefulness for estimating the size of inert (millennial) C pools. Major pitfalls in the use of bomb 14C, particularly for single time point samples and fractions, are mainly associated with model assumptions. For example, calculated residence times can be highly sensitive to a minor component of old C (<10% of total C). Similarly, calculated residence times are also highly dependent upon rates of soil C accumulation or loss. A final key source of error is lag times between C fixation from atmospheric CO2 and incorporation in the measured soil C pool, either due to long-lived plant tissue, or residence times in other soil pools/horizons. All work using Δ14C should consider sensitivity and uncertainty related to these issues. Major potential exists in the use of Δ14C to constrain soil C dynamics as a function of soil depth, in relation to major unexplained losses of soil C, and to probe the mechanisms and rates of soil organic matter stabilization. These areas of major potential all lay outside conventional use of Δ14C to calculate simple residence times. http://www.gns.cri.nz/who/staff/2234.htm
B12B-02
Radiocarbon-based Turnover Time Estimates of Soil Organic Carbon in a Cool-temperate Deciduous Forest in Asian Monsoon Region
Significantly more carbon is stored in the soils than in present in the atmosphere. Although the potential for C storage rates may change in the future as climate change progresses, the dynamics of soil carbon is unknown enough. We separated two density fractions on the soil down to 75 cm depth and estimated turnover time of these SOC fractions for volcanic ash soils in a cool-temperate deciduous forest in Japan, at one of AsiaFlux monitoring sites. According to the eddy-covariance based and biometric based carbon flux measurements over 10 years long in this site, this ecosystem is storing C (net ecosystem exchange (NEE): -2.4 tC ha-1 year- 1). However, the partitioning of C storage among vegetation and soils at this site is unknown. Measurements of carbon and radiocarbon (14C) inventory were used to determine the turnover time of two fractions of SOM: humified low density material < 2 g/cc and high density or mineral-associated organic matter > 2 g/cc. Total SOC stocks down to the depth of 75 cm were 26.2 kg C m-2, with the majority of SOC (52 percent) in the AB horizon (20 - 50 cm). Storage of SOC in our site was larger and differed considerably from that in other temperate forests in North America and Europe. The major part of the SOC (74 percent) was carbon in low density fraction. In the AB horizon, carbon in low density fraction accounted for 75 percent of the total SOC. The age of this fraction in lower AB horizon (35 - 50 cm) was significantly old (2490 years) as well as high density fraction (2930 years), although this fraction seems to consist of labile carbon.@Turnover times in all fractions are investigated for some layers in the soil depth of 75cm, as well as fine roots, low density humified material and carbon associated with minerals. Turnover times in both fractions increased with soil depth and 11-2780 year for low density fractions and 610- 3740 year for high density fractions. The turnover times of SOC were relatively long (1760 | 3740 years) in lower AB and B horizons as compared with other temperate forests. The results show that this forest is capable of accumulating large amount of carbon in the belowground and might be sequestrating carbon as low density fractions semi-permanently.
B12B-03
A soil organic matter model based on measurable pools for use in incubation studies
The current generation of soil organic matter (SOM) models is based on conceptual pools. No analytical method exists to measure the amount of C stored in the separate pools of a real-world soil sample. We propose a new SOM model which is based on measurable SOM pools. These pools can be extracted using physical fractionation techniques such as wet sieving, micro-aggregate isolation, flotation, and dispersion. We distinguish the dynamics of SOM enclosed in macro-aggregates, micro-aggregates enclosed within macro-aggregates, free micro-aggregates and SOM outside of any aggregates, as demonstrated in Six et al., 2004. In addition, the dynamics of particulate and mineral associated organic matter are followed. Simultaneously, aggregate formation and breakdown are simulated. Aggregates dynamics are dependent on microbial activity. As a consequence, the feedback between soil structure and soil organic matter is explicitly integrated. The model concepts will be presented together with a calibration, and confrontation with measured data from incubation studies.
B12B-04 INVITED
Interactive effects of litter quality and soil mineralogy on temperate forest soil carbon response to temperature
Temperate forest soil organic carbon (C) represents a significant pool of terrestrial C that may be released to the atmosphere as CO2 with predicted climate change. To address feedbacks between climate change and terrestrial C turnover, we quantified forest soil C response to litter quality and temperature change as a function of soil parent material. We collected soils from three conifer forest-types dominated by ponderosa pine (PP), white fir (WF), and red fir (RF) from each of three parent materials, granite (GR), basalt (BS), and andesite (AN) in the Sierra Nevada of California. AN soils were dominated by short-range-order (SRO) minerals, GR soils by crystalline minerals, and BS soils by a mix of crystalline and SRO minerals. Field soils were incubated in the laboratory at their mean annual soil temperature (MAST), with addition of native 13C-labeled litter. Further, we incubated WF and RF soils at PP MAST with 13C-labeled PP litter; and RF soils at WF MAST with 13C-labeled WF litter to simulate a migration of MAST and vegetation type up-elevation in response to predicted climate warming. Results indicated that total CO2 and percent of CO2 derived from soil C varied significantly across forest-types, following the pattern of GR>BS>AN. Regression analyses indicated significant control of C mineralization and soil C priming by litter quality and SRO minerals. Addition of litter derived water-soluble compounds enabled priming of recalcitrant soil C in soils with high SRO mineral content, whereas water-soluble litter components did little to promote priming of extant soil C in soils of low SRO mineral content. Results further indicated a 10-300% increase in WF and RF forest-type soil C mineralization under climate change conditions that varied substantially between parent materials. Soils derived from andesite exhibited minimal change; whereas granite and basalt derived soils lost large amounts of soil C under climate change conditions. This study corroborates the varied response in soil C mineralization by parent material and highlights how the soil mineral assemblage may act to control conifer forest-type soil C response to climate change.
B12B-05
An Automated Chamber Network for Evaluation the Long-term Response and Feedback of Soil Carbon Dynamics to Global Change
Regional networks for measuring carbon sequestration or loss by terrestrial ecosystems on a year round basis have been in operation since the mid-1990s. However, continuous measurements of soil CO2 efflux, the largest component of ecosystem respiration have only been reported over similar time scales at a few of the sites. We have developed a multichannel automated chamber system that can be used for continuous measuring soil CO2 efflux. The system equips 8 to 24 large automated chambers (90*90*50 cm, L*W*H). Since 1997, we have installed the chamber systems in the tundra in west Siberia, boreal forest in Alaska, cool- temperate and temperate forests in Japan, Korea and China, tropical seasonal forest in Thailand, and tropical rainforest in Malaysia. Annual soil CO2 effluxes were estimated to be about 5-6 tC ha-1 y-1 in the boreal and cool-temperate forests, 10 tC ha-1 y-1 in the temperate forests, and 30 tC ha-1 y-1 in the tropical rainforests. Efflux showed significant seasonality in the boreal and temperate forest that corresponding with the seasonal soil temperature. However, the wavelike efflux rates in the tropical forests were correlated with the seasonality of soil moisture. From 2007, a big project that funded by Ministry of the Environment of Japan (MOE) has launched to evaluate the response and feedback of soil carbon dynamics of Japanese forest ecosystems to global change. We are installing another 6 chamber systems at the six of Japanese typical forests to conduct the soil warming experiments. For scaling-up the chamber experiments and understanding the mechanisms of soil organic matter (SOM) dynamics to global change, soil samples from about 100 forest ecosystems will be incubated for modeling development. Furthermore, the environmental (temperature and CO2) controlled large open-top chambers have been employed to investigate the balance of SOM (the input from litter falls and loss due to the decomposition) of forest ecosystems with global change.
B12B-06
Seasonal controls on soil respiration fluxes and isotopic content in an African Savanna ecosystem: Implications for site to regional carbon flux estimates
The savannas of South Africa are characterized by a patchy mosaic vegetation structure comprised of C3 trees and C4 grasses and a strong seasonal precipitation regime. This patchy vegetation structure combined with the strong seasonal precipitation regime potentially leads to substantial variation in both the isotopic content and magnitude of soil respiration. From the perspective of tower-based carbon flux measurements and regional carbon flux studies, the spatial and temporal patterns in soil respiration could present challenges to understanding regional carbon flux controls. In this study, we carried out both low and high frequency measurements of soil respiration using collars and continuous-measurement soil CO2 probes to understand short and longterm soil respiration dynamics. On a short timescale the continuous measurements of CO2 flux show soil respiration increases substantially after rainfall; however, a lag time between precipitation and increased soil respiration suggests physical factors affect flux response. Also, following precipitation, soil respiration hysteresis curves occur while soil moisture remains elevated, suggesting substrate limitations may reduce soil CO2 flux. On an annual timescale fluxes from under tree canopies are 35% greater than interspace grass dominated areas, with the majority of the difference occurring during the wet season. Because of the isotopic SOM differences between under tree canopies (-21.21‰) and interspace grass dominated areas (-16.66‰), the seasonal shifts in flux contributions from these two settings also leads to seasonal variation in site level soil δ13CO2 fluxes. As a consequence, the isotopic content of respiration measured from an eddy covariance tower and interpretation of ecosystem C exchange could be affected by wind direction and plant patch cover heterogeneity.
B12B-07
The diel patterns of soil respiration in four arid California ecosystems: fluxes, sources and hypotheses
Automated measurements provide the high-resolution information that enables us to analyze potential causes for diel variability in soil respiration. These diel patterns are the complex result of biological and physical processes that determine the production and diffusion of CO2 through the soil. We examined the diel patterns of soil respiration from four arid California ecosystems: (1) a pinon-juniper woodland in at the Burns Pinon Ridge Reserve near Joshua Tree National Park, (2) a cold desert shrub community and (3) a perennial grassland near the city of Bishop in the Owens Valley, and (4) a mixed oak-pine forest at the James Reserve in the San Jacinto Mountains. In addition to automated chamber and environmental measurements at these sites, we used isotopic (14C) partitioning techniques to separate the plant and microbial sources contributing to soil respiration at certain time points. Here we present the diel cycles of soil respiration and environmental variables, the source partitioning results, and hypotheses about what processes determine these diel patterns that both span, and are specific to the studied ecosystems. In these systems dominated by Mediterranean or desert climates, we observed that factors like relative humidity can dominate the diel variations in soil respiration for sites with very dry surface litter. At other sites and times of year, diel variation in soil respiration reflects photosynthetic and VPD influence on root respiration. The combination of automated chamber measurements with isotopes provides information useful for separating the plant and heterotrophic control on diel and seasonal soil respiration fluxes.
B12B-08
Effects of experimental drought on soil CO2 efflux in a Mediterranean Quercus ilex forest
Mediterranean ecosystems are among the most heavily and early utilized by man, and as a result, are frequently severely impacted. Because of these anthropogenic perturbations as well as strong ecological constraints like long summer drought, Mediterranean ecosystems are predicted to be especially responsive to environmental changes. One scenario based on the output from four climate models suggests that temperatures could rise by over 4°C by 2100 over the Mediterranean Basin. Annual precipitation is projected to decline by 10 to 40% over much of Africa and South-eastern Spain, with smaller but significant changes elsewhere. The objectives of this research was (1) to better understand how such drought would influences soil carbon storage and soil respiration in forested ecosystems, (2) to test the interaction between soil water and soil temperature in controlling soil respiration, and (3) to examine the relationship between canopy photosynthesis and soil respiration in a seasonally dry evergreen ecosystem. To reach these objectives, drought was simulated since 2003 by continuously excluding 20% of throughfall using gutters below canopy in a Quercus ilex forest in south of France (MIND project). In 2005, soil respiration was continuously measured in a control and a dry plot using an automatic soil respiration system (Rayment, open system). Respiration showed a diurnal hysteresis: efflux was higher during the day than at night for the same temperature. This small hysteresis could not be explained by the progression of the heat front in the soil profile and was correlated with canopy photosynthesis derived from an Eddy-Covariance tower nearby (Carboeurope project). However, canopy photosynthesis only explained 17% of the diurnal variability of soil respiration, while temperature explained 80%. Seasonally, soil water content explained 80% of the variation in respiration in both treatments, temperature explaining 60% of the residuals. Soil water content and soil respiration were systematically lower in the dry plot. However, soil respiration in the dry plot was higher than in the control plot for the same soil water content, and soil respiration was the same in both plots at high soil water content. Thus, after 3 years of rain exclusion, soil respiration seems to have acclimated to dryer conditions. The apparent Q10 was not influenced by the throughfall exclusion. Soil respiration was modeled using soil water and soil temperature relationships calibrated in both plots. The model showed that respiration in the dry plot was similar than in the control plot in winter for high soil water content. Soil respiration was also almost equal in both plots in mid-summer, for very low soil water content. Only in spring, for medium soil water content, soil respiration was higher in the control plot. Thus annually, respiration was almost equal in both plots (5% lower in the dry plot). Consequences for soil carbon storage in Mediterranean forests in the future are discussed.