B14A-01 INVITED
Forest Soil Respiration: Identifying Sources and Controls
Most of the respiration in forests comes from the soil. This flux is composed of two components, autotrophic and heterotrophic respiration. In a strict sense the former should be plant belowground respiration only, but the term is used here to denote respiration by roots, their mycorrhizal fungal symbionts and other closely associated organisms dependent on recent photosynthate. Heterotrophs are organisms using organic matter, chiefly above- and belowground litters, as substrate (i.e. substrates of in general much higher ecosystem age). Because of the complexity of the plant-soil system, the component fluxes are difficult to study. I will discuss results of different approaches to partition soil respiratory components and to study their controls. The focus will be on northern boreal forests. In these generally strongly nitrogen-limited forests, the autotrophic respiration equals or exceeds the heterotrophic component. The large autotrophic component reflects high plant allocation of C to roots and mycorrhizal fungi in response to the low N supply. A physiological manipulation, girdling, which stops the flow of photosynthates to roots, showed that autotrophic respiration could account for as much as 70% in N-limited forests, but only 40% in fertilized forests. Also using girdling, we could show that a shift to lower summertime temperature leads to a decrease in heterotrophic but not in autotrophic activity, suggesting substrate (photosynthate) limitation of the latter. Physiological manipulations like girdling and trenching cannot be used to reveal the finer details of soil C dynamics. Natural abundance stable isotope (13C) and 14C approaches also have their limitations if a high resolution in terms of time, space and organism is required. A very high resolution can, of course, be obtained in studies of laboratory micro- or mesocosms, but the possibility to extend the interpretation of their results to the field may be questioned. In the CANIFLEX (CArbon NItrogen Forest Labelling Experiment) project, we use a short labelling period and highly enriched 13CO2 to produce a traceable pulse through components of the system directly in the field. Thus, we have labelled 50 m2 plots using 4-5 tall plastic chambers (i.e. 200-250 m3) in a young Scots pine forest. Results of a pilot study in 2006 have been published (Högberg et al. 2008 New Phytol.). The CANIFLEX project has shown that traceable quantities of labelled C peak within a few days in ectomycorrhizal roots, microbial cytoplasm and soil respiratory efflux. In the two latter, the C had a half-life of 280 h and 35 h, respectively. Using higher tracer levels than in the pilot study, we have now successfully conducted stable isotope probing of specific groups of soil organisms at this large scale, finding 13C tracer in PLFA (phospholipid fatty acid) biomarkers for fungi, but not in the majority of markers for bacteria. We are now into our second season of following the effects of additions of N on the belowground C dynamics using a combined 13C (photosynthate) and 15N (labelling of soil N) approach. The physiological and labelling approaches are complementary, but need to be linked through modelling. This is a major challenge.
B14A-02
How General is the Current Photosynthate Controls on the soil CO2 Flux Paradigm?
A variety of methods including girdling experiments and isotope labeling approaches have provided some evidence for a tight link between current C assimilation and soil CO2 flux. The results of these investigations have lead to the conclusion that autotrophs control soil CO2 flux. If the results from these investigations are general then our understanding of patterns and regulation of below ground C dynamics and the means by which ecosystem controls are studied and modeled must be reconsidered. While evidence for a coupling between current photosynthate and soil carbon dynamics has been conspicuous, data that may challenge this relationship have not been thoroughly considered. Results from foliar scorching treatments in longleaf pine (Pinus palustris) ecosystem that removed 95% of the foliage demonstrate that (i) mycorrhizal fungi production was not significantly reduced as a result of scorching, (ii) root mortality was not significantly affected because of disturbance of the carbon source, and (iii) total root non-structural carbohydrates were not significantly reduced after scorching. These results together with findings from other regions suggest that in some systems soil CO2 fluxes are less tightly linked to variations in C assimilation because stored C acts as a buffer. This stored C is a critical resource for rebuilding damaged foliage in many frequently burned ecosystems. We propose that plant adaptations to disturbance and recovery from disturbance may explain why some systems may be buffered from variation in C source strength and the link between above and belowground carbon dynamics is more diffuse.
B14A-03
An evaluation of methods for estimating and partitioning the C isotopic signature of soil respiration into its component fluxes
Soil respiration (or soil CO2 efflux) is a major component of ecosystem respiration and much research has been dedicated to evaluating its environmental and biological drivers. The isotopic composition of soil respiration (δ13CR) changes seasonally and spatially, but processes responsible for these variations have not been fully explored. The application of stable isotopes to understanding belowground processes will be strengthened by a more detailed analysis of the sources of variations in δ13CR. These variations may be due to ecological signals (e.g., changes in autotrophic or heterotrophic respiration rates), observation error, or analytical error. Keeling plots are typically used to estimate δ13CR, and various methods have been applied to collect samples of soil air for analysis, including static and dynamic closed chambers, open chambers, and gas wells buried at different depths. Our objectives here are to (i) evaluate methods for collecting soil CO2 (closed chambers vs. gas wells) in terms of their Keeling plot accuracy, (ii) estimate observational and analytical errors associated with each method, and (iii) partition the Keeling plot intercepts into autotrophic and heterotrophic contributions. We take advantage of two elevated CO2 experiments in which labeled CO2 produces an autotrophic signature for recently fixed C that is at least 12 permil lower than the heterotrophic end member, in contrast to ambient CO2 treatments, in which the difference between end-member signatures can be as low as 2-3 permil. We apply a Bayesian approach that enables us to fit a Keeling plot model to the observed δ13C and CO2 data, while simultaneously estimating and partitioning δ13CR into its different contributing fluxes by incorporating end-member δ13C data. This approach also provides explicit estimates of different sources of uncertainty such as those associated with unobservable ecological processes, observational error, and analytical error. This work develops rigorous statistical methods for analysis of Keeling plots, and will provide insight into the applicability of stable isotope partitioning of soil respiration.
B14A-04 INVITED
Rhizosphere Respiration Enriches Soil Respiration Flux in 13CO2 under Boxelder (Acer negundo) Trees
Root-free plots were established with 2-meter deep trenches under each of 6, 5-m tall, deciduous boxelder trees. Surface vegetation was kept clear in these and adjacent, untrenched plots. In each plot surface fluxes (Rs) and soil CO2 and δ13C profiles were measured biweekly to monthly over 1.5 years. The carbon isotope ratio of soil respiration (δR) was estimated from soil CO2 and δ13C profiles. In the second year two open chambers plumbed to a tunable diode laser mass spectrometer were run continuously for Rs and δR, alternating between two pairs of trenched/untrenched plots. Between growing seasons flux rates were similarly low (< 2 μmoles m-2 s-1) and isotope ratios were relatively depleted (-29 ‰) in both trenched and untrenched plot sets. During the transition to active growing seasons a larger increase in Rs in plots containing roots (to 9 μmoles m-2 s-1, vs. to 4 μmoles m-2 s-1 in trenched plots) was associated with a greater seasonal enrichment in δR (-25.5 ‰ vs. -27 ‰ in trenched plots). Open chamber and soil δ13CO2 profile measurements supported the conclusion that the presence of active roots corresponded with an enriched δR, a result unexpected based on published studies. A relatively enriched root respiration signal was also observed by comparison of respired CO2 from excised roots, leaves, and sieved soils.
B14A-05
Effects of Temperature and Plant Carbon Supply on Soil Respiration
Soil respiration by microbes and plants is the largest source of CO2 efflux from land to the atmosphere, and it is strongly controlled by temperature. Studies in northern temperate forests have reported a wide range in seasonal temperature sensitivity values (Q10s), and some evidence suggests that this variability may be related in part to the seasonality of plant carbon supply and whether trees are evergreen or deciduous. However, relatively little work attempts to separate the individual effects of temperature and plant carbon supply on soil respiration. In this study we examined the impacts of plant carbon supply on soil respiration at seasonal and daily timescales by measuring respiration rates in areas where roots were excluded as well as neighboring areas where roots were present. We were interested in 1) whether the presence of roots affected the temperature sensitivity of soil respiration, and 2) whether evidence for canopy controls on CO2 production were apparent in areas where roots were present. Toward this effort, we gathered 18 months of respiration data in a coniferous forest in the Central Oregon Cascades (H.J. Andrews LTER). To examine the effects of plant carbon inputs, we excluded plant roots from areas by pounding 25cm diameter PVC pipe to 50cm depth. Using survey respiration measurements taken every 2-4 weeks, we found that seasonal Q10s in bulk soil and inside of root exclosures did not differ significantly. Hence, temperature sensitivity did not differ between soil with roots and without roots over seasonal time scales at this Pacific Northwest Forest. Respiration was also measured using continuous LI-8100 data to better understand controls at shorter time steps. In contrast to our annual and seasonal data, soil respiration did not vary synchronously with temperature at the diel scale. To investigate other potential controls of diel soil respiration, we examined correlations between soil respiration and a driver of photosynthesis, photosynthetically active radiation (PAR). We found that in bulk soil, soil respiration correlated most strongly with PAR 3-days prior, whereas in soil where roots were excluded there were no strong positive correlations, suggesting different controls on heterotrophic respiration. The correlation between 3-day prior PAR and respiration in soils with roots suggests that plant physiological variables may be relevant in determining efflux rates on daily timescales.
B14A-06
Continuous, in situ Isotopic Carbon Dioxide Measurements of Ecosystem Exchange Processes and Soil Respiration Using a WS-CRDS Analyzer
Understanding the interdependencies of sources and sinks within ecosystems and validating models of such systems greatly benefits from fast, continuous, in situ measurements of not only CO2 concentration, but also isotopic carbon abundances in CO2. Such high frequency (<5 minute) isotopic measurements help validate carbon transport and diffusion models of terrestrial ecosystems and are key to developing an overall understanding of the dynamics influencing global atmospheric carbon budget. By utilizing high time resolution instrumentation based on Wavelength-Scanned Cavity Ring Down Spectroscopy (WS-CRDS) the biosphere-atmosphere CO2 exchange mechanisms can be more carefully examined. This measurement technique achieves precisions of approximately 0.3‰ with measurement drift that is sufficiently low so as to avoid frequent calibration and can be deployed in remote, unattended locations for long-term, continuous measurements, enabling the observation of diurnal and seasonal trends in the CO2 exchange processes. We present two sets of data from a recently commercialized WS-CRDS analyzer. The first is from Wind River Canopy Crane, WA, analyzing air within the canopy of an old growth forest. By combining this high-resolution isotopic CO2 data with existing models of the ecosystem carbon budget, these models can be further examined to test their sensitivities to currently held assumptions about the effect of environmental forcings (e.g. temperature, precipitation, VPD/RH, soil moisture, incident radiaition) on heterotrophic respiration. The second set of data is from a soil chamber in which δ13CO2 is continuously measured. Such monitoring of soil has been instrumental in validating recent modeling efforts that describe isotope dynamics in diffusive environments where equilibrium has not been established. We see that true isotopic equilibrium is likely rare because of the time it takes for all isotopologues to equilibrate. This results in a range of dynamic fractionations (frequently several permil) that in soil environments are a result of changes in the CO2 production rate, gas diffusivity, and air-filled porosity. This effect is seen to some extent in most natural and disturbed environments and also as a direct consequence of sampling. Researchers have not previously been sensitive to these transient fractionation effects which can lead to misinterpretation of data. The real-time data presented here has allowed verification of previously presented models of these systems.
B14A-07
An Open Chamber System Coupled With a Tunable Diode Laser for Measurement of Delta 13C, Delta 18O, and Efflux Rate of Soil Respired CO2
High frequency observations of the stable isotopic composition of CO2 effluxes from soil have been sparse due in part to measurement challenges. We developed an open-system method that utilizes a flow- through chamber coupled to a tunable diode laser (TDL) to quantify the rate of soil CO2 efflux and its δ13C and δ18O signatures. We tested the method first in the laboratory using an artificial soil test column and then in a semi-arid woodland. We found that CO2 efflux rates of 1.2 to 7.3 μmol m-2 s-1 measured by the chamber-TDL system gave similar results to measurements made using the chamber and an infrared gas analyzer (IRGA) (R2=0.99) and compared well to efflux rates generated from the artificial test column (R2=0.94). Measured δ13C and δ18O of CO2 efflux from the test column were not significantly different from measurements of gas inside of the test column across all efflux rates (p>0.05) after accounting for diffusive enrichment. Isotopic differences between chamber measurements and values from CO2 gas introduced into the test column resulted primarily from diffusion of atmospheric CO2 into the test column and pressure artifacts from the chamber. Field measurements during drought demonstrated a strong dependency of CO2 efflux and isotopic composition on soil water content. Addition of water to the soil beneath the chamber resulted in average changes of +6.9 μmol m-2 s-1, -5.0‰, and –55.0‰ for soil CO2 efflux, δ13C and δ18O, respectively, within 25 minutes of water addition. The soil chamber coupled with the TDL was found to be an effective method for capturing high-resolution soil CO2 efflux and its stable isotopic composition.
B14A-08
Assessing Vertical CO2 Production Rates and Surface Fluxes Using Automated Diffusion Chambers
In recent years soil CO2 emissions has been the subject of intense investigation because (i) its potential role in amplifying global warming; and (ii) gaseous compounds formed in the soil environment are, in general, good indicator of soil biology and biochemistry. Accurate techniques used to monitor soil CO2 profile concentrations offers the opportunity to identify localized carbon dioxide sources and potential sinks in the soil, and to understand the processes that control CO2 production and emission. In this study, we developed a method to continuously monitor soil CO2 concentration, by using a new type of soil diffusion chamber. We estimated soil CO2 efflux using a new model to determine the vertical CO2 gradient across the soil profile up to 80 cm depth, in conjunction with models to determine the soil CO2 diffusion coefficient. Furthermore, we assessed vertical CO2 production rates within the soil profile. Daily mean value of CO2 concentration had a significant variation correlated to soil temperature. Moreover, the vertical soil CO2 concentration showed similar temporal variation at all depths. From January to August 2008, seasonal mean values of soil CO2 production varied between 1.97 to 6.84 gC/m2/day across the soil layer 0-10 cm. Between 10 and 20 cm depth soil CO2 production varied between 0.67 and 2.68 gC/m2/day, and across the soil layer between 20 to 40 cm depth the CO2 production varied between 0 and 0.02 gC/m2/day. Over the same period, seasonal mean values of modelled soil efflux ranged between 3.12 and 12.96 gC/m2/day. These values correlated well with soil temperature and flux values measured using automated soil surface chamber. We present a simple technique to measure continuously soil CO2 profile by burying small CO2 diffusion chambers. Overall this experiment points out the ability to measure continuously, and for prolonged periods of time, CO2 concentration across a soil profile.