SSSAJ Grow Your Career with SSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 4 August 2005
Published in Soil Sci Soc Am J 69:1492-1502 (2005)
DOI: 10.2136/sssaj2004.0300
© 2005 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hanson, P. J.
Right arrow Articles by Trumbore, S. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Hanson, P. J.
Right arrow Articles by Trumbore, S. E.
Agricola
Right arrow Articles by Hanson, P. J.
Right arrow Articles by Trumbore, S. E.
Related Collections
Right arrow Forest Soils
Right arrow Biogeochemical Processes
Right arrow Isotopes

Forest, Range & Wildland Soils

Reconciling Change in Oi-Horizon Carbon-14 with Mass Loss for an Oak Forest

P. J. Hansona,*, C. W. Swanstonb, C. T. Garten, Jr.a, D. E. Todda and S. E. Trumborec

a Environmental Sciences Division, Oak Ridge National Lab., Oak Ridge, TN 37831-6422
b Center for Accelerator Mass Spectrometry, Lawrence Livermore National Lab., Livermore, CA 94550
c Univ. of California, Irvine, CA 92697

* Corresponding author (hansonpj{at}ornl.gov)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
First-year litter decomposition was estimated for an upland-oak (Quercus spp.) forest ecosystem using enrichment or dilution of the 14C-signature of the Oi-horizon. These isotopically based mass-loss estimates were contrasted with measured mass-loss rates from past litterbag studies. Mass-loss derived from changes in the 14C-signature of the Oi-horizon suggested mean mass loss over 9 mo of 45%, which was higher than the corresponding 9-mo rate extrapolated from litterbag studies (~35%). Greater mass loss was expected from the isotopic approach because litterbags are known to limit mass loss processes driven by soil macrofauna (e.g., fragmentation and comminution). Although the 14C-isotope approach offers the advantage of being a non-invasive method, it exhibited high variability that undermined its utility as an alternative to routine litterbag mass loss methods. However, the 14C approach measures the residence time of C in the leaf litter, rather than the time it takes for leaves to disappear; hence radiocarbon measures reflect C immobilization and recycling in the microbial pool, and do not necessarily replicate results from litterbag mass loss. The commonly applied two-compartment isotopic mixing model was appropriate for estimating decomposition from isotopic enrichment of near-background soils, but it produced divergent results for isotopic dilution of a multi-layered system with litter cohorts having independent 14C-signatures. This discrepancy suggests that cohort-based models are needed to adequately capture the complex processes involved in C transport associated with litter mass-loss. Such models will be crucial for predicting intra- and interannual differences in organic horizon decomposition driven by scenarios of climatic change.

Abbreviations: AMS, accelerator mass spectrometer • EBIS, Enriched Background Isotope Study • ORR, Oak Ridge Reservation • SE, standard error • SOM, soil organic matter • TVA, Tennessee Valley Authority


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
FUNDAMENTAL KNOWLEDGE of soil organic matter (SOM) dynamics is critical to understanding C sequestration, ecosystem C, and nutrient cycling processes and ultimately forest growth. Unfortunately, we still lack critical information regarding how SOM is formed and the mechanisms whereby it is distributed within the soil profile. Annual leaf-litter inputs are an obvious potential source of mineral soil C, but significant mass loss of annual leaf-litter cohorts occurs during their migration to the underlying mineral soil (Edwards et al., 1970). Litter mass-loss is a multifaceted process involving leaching of soluble chemicals, faunal facilitated comminution and fragmentation, and decomposition (catabolic degradation of litter constituents) by soil microorganisms (Edwards et al., 1970; Anderson, 1973; Singh and Gupta, 1977). Unambiguous estimates of the rate of accumulation of forest litter and/or the rate at which it transfers C to long-term storage pools are needed.

Litter decomposition bags, tethered leaves, or the isolation of litter cohorts in layered mesh are all methods that have been used to quantify the processes of litter decomposition (Anderson, 1973; Singh and Gupta, 1977; Jorgensen et al., 1980; Binkley, 2002). Disturbance to the litter microclimate, interruption of faunal access to the litter, and mass-accumulation from mineral contamination are, however, common problems confounding the interpretation of cohort mass loss measurements (Swanston and Myrold, 1997; Robertson and Paul, 2000; Bradford et al., 2002; Idol et al., 2002; Gartner and Cardon, 2004).

Isotopic tracer studies using 13C or 14C have obvious potential for tracking litter mass loss without the disturbance artifacts of the manipulative approaches, and a number of papers have reported on the application of such methods to field measurements of the decomposition of crop/grassland litter (Nyhan, 1975; Wedin et al., 1995; Milchunas and Lauenroth, 1992), forest litter (Hobbie et al., 2004), peat (Domisch et al., 2000), and fine roots (Dahlman and Kucera, 1965; Balesdent and Balabane, 1996). Unfortunately, past isotope studies had to depend on short-term pulse labeling methods as a source for isotopically enriched plant tissues. Those methods may not produce uniformly labeled plant material (Robertson and Paul, 2000; Schimel, 1993), and therefore could lead to estimates of litter turnover inconsistent with true changes in bulk tissue mass loss or decomposition. Conversely, Bromand et al. (2001) have documented experimental methods for generating small quantities of uniformly labeled 13C-litter in wheat appropriate for lab and mesocosm studies (e.g., Lin et al., 1999; Hobbie et al., 2004).

Elevated levels of 14C-CO2 in the air and soil atmosphere were observed on the Oak Ridge Reservation (ORR) in east Tennessee during the summer of 1999 as a part of a latitudinal study of soil C cycling from Tennessee to Maine (Gaudinski and Trumbore, 2003). The rapid rise in the background 14C-signature of soil-derived CO2 was the result of a local release of 14C in the form of 14CO2 gas from a local industrial incinerator (Trumbore et al., 2002). Subsequent measurement of {Delta}14C in the cellulose from tree rings formed in 1999 showed that the localized growing-season release was unique in its magnitude, and additional surveys of 1999 tree wood and 2000 leaves showed that the release encompassed a significant portion of the ORR. The highest amounts of the 14C-enriched 2000 leaf tissues ({Delta}14C of 1000 to 2000{per thousand}) were found in vegetation from the western portion of the ORR (Trumbore et al., 2002). The ecosystem-scale enrichment of 14C represented a unique opportunity to study C cycling processes within undisturbed stands of the eastern deciduous hardwood forest.

Because the 14CO2 release occurred following canopy leaf production and maturation in 1999, foliar litterfall from 1999 was not expected to be highly enriched. However, the 2000 canopy leaves were enriched as a result of their growth from stored carbohydrates enriched by the 1999 14C-pulse. Therefore, the 2000 mixed-species litterfall represented a unique source of 14C enriched foliar tissues for application to a number of edaphic, physiological, and ecosystem-scale research questions. To take advantage of this opportunity, large quantities of leaf litter were collected from the west and east ends of the ORR during fall canopy senescence in 2000. The collections included enriched ({Delta}14C of ~1000{per thousand}) and near-background ({Delta}14C of ~220{per thousand}) litter for use in subsequent studies of soil C cycling as a part of the U.S. Department of Energy funded Enriched Background Isotope Study (EBIS) described further on the project's Web page (http://ebis.ornl.gov/).

As one component of the EBIS project, this paper reports on the use of the uniquely labeled forest litter for measuring annual rates of Oi-horizon litter decomposition in replicated and undisturbed plots under natural forest-floor conditions. It evaluates the utility of the isotopic approach for stable 14C-input versus transient conditions, and for 14C enrichment versus dilution of the Oi-layer. The isotopic estimates of litter turnover are also contrasted with traditional litter decomposition observations from past studies on the ORR.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Site Description
The EBIS project (Trumbore et al., 2002) was established in the autumn of 2000 on the ORR, in the U.S. Department of Energy's National Environmental Research Park near Oak Ridge, Tennessee (latitude N 35°58'; longitude W 84°16'). The mean annual precipitation is 1358 mm and mean annual temperature is 14°C (Johnson and Van Hook, 1989). All EBIS research plots are located on up-slope, ridge-top positions in the upland oak forest type (Quercus spp.; Acer spp.) with scattered pine (Pinus echinata Mill. and P. virginiana Mill.), mesophytic hardwoods (Liriodendron tulipifera L., Fagus grandifolia J.F. Ehrh.), and some hickory (Carya spp.). The ages of the overstory trees cover a broad range from about 40 to 150 yr, and the maximum canopy height is approximately 26 m. Maximum leaf area index is approximately 6 m2 m–2. The EBIS sites on the ORR encompass two soil types and two levels of 14C exposure in 1999. At all sites, replicated permanent plots were established for the manipulation of forest litter through reciprocal transplants of enriched versus near-background litter among sites.

The EBIS Experimental Design
Before leaf senescence in the fall of 2000, ~150 4.6 by 9 m plastic tarps were laid out on the forest floor along Pine Ridge for the collection of all forest leaf litter in the vicinity of the 1999 14C release. A similar number and distribution of tarps were laid out on Walker Branch Watershed at the east end of the ORR to represent near-background conditions. Litter was manually collected from the tarps weekly from September through mid-December 2000. Collected litter was transferred to ORNL greenhouses for air-drying and subsequent storage in large (1 x 1.5 m) vacuum bags. The dried litter was stored in rodent free trailers until needed for the experimental manipulations. Sufficient enriched (1005 ± 19{per thousand}) and near-background (221 ± 2{per thousand}) litter was collected in the fall of 2000 to conduct a multi-year, plot-level litter manipulation study (3 yr of litter additions). The near-background litter's 14C-signature was slightly elevated with respect to the expected northern hemisphere data (Levin and Hesshaimer, 2000) because the 1999 14C-release also exposed eastern portions of the ORR to a limited degree (Trumbore et al., 2002).

Four research sites were established on the ORR (Fig. 1). Two ‘enriched’ sites on the west end of the ORR were established on Ultisol and Inceptisol soils of Pine Ridge and on Tennessee Valley Authority land on Chestnut Ridge (TVA), respectively. Two ‘near-background’ sites were established approximately 10 km further east away from the influence of the 1999 14C pulse. The near-background sites included a site with Ultisols within Walker Branch Watershed on Chestnut-Ridge, and a site with Inceptisols on Haw Ridge. The Ultisols are deep, highly weathered soils derived from dolomitic parent material. The Inceptisols are shallow, less weathered soils derived from limy shale/sandstone formations. According to the classification scheme of Green et al. (1993), the organic horizon for both of these soils would be classified as a Leptomoder (a transitional state between mor and mull forest humus). A detailed description of the hydrologic, physical, geochemical, and mineralogical properties of these soils can be found in previous studies (Wilson and Luxmoore, 1988; Jardine et al., 1988; Wilson et al., 1989, 1990).



View larger version (67K):
[in this window]
[in a new window]
  		Fig. 1. Map of the Oak Ridge Reservation near Oak Ridge, TN showing the location of the EBIS study sites. The Walker Branch (WB) and TVA sites are on Chestnut Ridge and have Ultisol soils. The Haw Ridge (HR) and Pine Ridge (PR) have Inceptisol soils.

 
At each of the four research sites on the ORR (Fig. 1), eight square 7 by 7 m plots were delineated with metal fence posts and plastic fencing (~61 cm tall). From late September through early December of 2000, the forest floor within each plot was covered with landscape cloth, and the ambient litterfall was periodically blown off the landscape cloth. After the landscape cloth was removed in mid-December, 14C enriched or near-background litter was added back to the respective treatment plots (500 g dry mass m–2) in May 2001. Treatment litter was not added to the experimental plots earlier in the annual cycle pending project funding decisions. Litter treatments were randomly applied to the eight plots at each of the four research sites. The following combination of replicated research plots was created by the experimental design:
  1. Plots with 14C-enriched root litter, 14C-enriched soil C, and applied 14C-enriched leaf litter (Pine Ridge and TVA western ORR)
  2. Plots with 14C-enriched root litter, 14C-enriched soil C, and near-background 2000 litter, (Pine Ridge and TVA, western ORR),
  3. Plots with background roots and soil C, and 14C-enriched 2000 litter (Walker Branch and Haw Ridge, eastern ORR), and
  4. Plots with background roots and soil C and near-background 2000 litter (Walker Branch and Haw Ridge, eastern ORR).

This paper deals only with first-year manipulations of the organic horizons of these plots.

Oi-Horizon Sampling
Time-zero and ‘first-year’ litter-layer and mineral soil core samples were collected in March of 2001 and January 2002, respectively. Time-zero O-horizon samples for a given plot within each research site represented a pooled sample of three randomly selected 0.25-m2 rings for a total sampled area of 0.75 m2. Each pooled sample was separated into Oi-horizon leaf litter greater than approximately 1 yr of age (i.e., the recognizable litter under the landscape cloth present before current year litterfall), and the Oe/Oa-horizon litter consisting mostly of humus with small amounts of decomposed leaf litter without distinguishable features indicative of the species of origin. One-year samples were collected similarly, but with smaller rings (0.062 m2) having a total collection area of 0.185 m2. Corresponding mineral soil samples were collected with a 0.1-m diam. circular coring device from the center of the sampled ring to a depth of 0.9 m, but are not discussed in this paper.

Litterbag Mass-Loss Observations
Although not a part of the EBIS project, comparable data for litter mass-loss from litterbag studies were available from a number of studies on the ORR. These studies included mass loss of litter from various tree species [Morus rubra L., Cercis canadensis L., Quercus alba L., Pinus taeda L.] from November 1960 to December 1961 (Witkamp, 1966), A. rubrum, Nyssa sylvatica L., and Q. alba litter from November 1970 to March 1972 (Kelly, 1973), Liriodendron tulipifera L. from November 1990 to November 1991 (O'Neill and Norby, 1996), and mixed-tree-species litter [Acer rubrum L., Cornus florida L., Quercus prinus L.] from March 1996 to October 2000 (Hanson et al., 2003b). All data were combined to generate a ‘mixed-species’ estimate of litter mass loss for closed-canopy, deciduous forests occupying the ORR. Temperature and rainfall data for the studies listed above were comparable, albeit not identical, to the weather during first-year of O-horizon decomposition described in this paper (Fig. 2). Abundant and seasonally uniform monthly precipitation in the years studied suggested that these data were appropriate for comparison against the changing 14C-signatures studied in this paper.



View larger version (61K):
[in this window]
[in a new window]
  		Fig. 2. Mean monthly air temperature and cumulative monthly rainfall for previous litterbag decomposition studies conducted on the Oak Ridge Reservation plotted against the long-term (54-yr) climate record.

 
Environmental Monitoring
A single solar-powered, weather station with data logger was placed at each research site for the collection of hourly measurements of air temperature, relative humidity, through-fall, litter layer water content, and other environmental variables not discussed here. Air (1.5-m aboveground) and Oi-horizon temperatures were logged at all long-term reference plots starting in 2001. Air temperature and relative humidity were measured with a shielded combination sensor (Rotronic Model MP101; Huntington, NY). Oi-horizon temperatures were obtained with independent sealed thermistor probes (Stowaway Tidbit, ONSET Corporation, Bourne, MA). Continuous direct measurements of forest litter water content were based on the electrical resistance characteristics of wet versus dry litter (Hanson et al., 2003a). All environmental data was logged as hourly means or sums and stored on a data logger (CR10; Campbell Scientific, Inc., Logan, UT).

Elemental Measurements
Subsamples of enriched and near-background litter and all O-horizon plot samples were analyzed for 14C, total C, and total N. Radiocarbon values were measured on the Van de Graaff FN accelerator mass spectrometer (AMS) at the Center for Accelerator Mass Spectrometry, Lawrence Livermore National Laboratory, Livermore, CA. In preparation for AMS analysis, samples were combusted in evacuated, sealed tubes in the presence of CuO and Ag, then reduced to graphite coating on iron powder in the presence of H2 (Vogel et al., 1984). Splits of combusted sample were taken for 13C analysis from each organic and mineral horizon for correction of mass-dependent fractionation in the reported radiocarbon values, and all radiocarbon values are presented as {Delta}14C ({per thousand}) according to Stuiver and Polach (1977).

Samples were analyzed for total C and N on a LECO CN-2000 (LECO Corporation, St. Joseph, MI) using secondary standards traceable to NIST reference materials. Soil C and N concentrations (g element g dry mass–1) in combination with average measurements of forest floor dry mass (g m–2) were used to calculate C and N stocks (g m–2) for O-horizon layers.

Data Analysis
A two-compartment stable-isotope mixing model was used to evaluate the percentage contribution of current litterfall to Oi-horizon litter following decomposition from May through January of 2002:

[1a]

[1b]

[1c]
where L% is the percentage contribution of new litter to the residual after a period of decomposition, S0 is the Oi-{Delta}14C signature at time zero, S1 is the Oi-{Delta}14C signature after one period of decomposition (i.e., 9 mo in this paper), L is the {Delta}14C signature of added litter cohorts, and fx is the fractional mass of the Oi-horizon or its components. For a 9-mo increment of time, radioactive decay (1/8267 yr) is negligible in this calculation and is ignored. The mean and 95% confidence intervals for L% were calculated according to Phillips and Gregg (2001) using their Web-accessible Excel spreadsheet program ISOERROR 1.04 (Phillips and Gregg, 2003). This model is appropriate for stable systems where two uniform materials are mixed. An 8-yr record of litter inputs to the Throughfall Displacement Experiment on Walker Branch Watershed (Hanson et al., 2003c) demonstrated that inter-annual variation in aboveground litter production was low on the ORR (maximum range was 480 to 520 g m–2 from 1994 through 1998) suggesting that stable interannual litter production (and perhaps standing Oi-horizon stocks) could be assumed in the calculations.

Measurements of litterbag mass loss from the studies described above (see litterbag mass loss observations) were fitted to the following exponential relationship of litterbag mass remaining (MRLB) versus time that allows for variable loss rates over time (Kelly and Beauchamp, 1987):

[2]
where MRLB is the percentage of mass remaining in an individual litter cohort over time, t is time in years, k is the mass loss coefficient, p is the parameter that allows the rate of mass loss to change with time, and error is the sampling error associated with measurements of mass loss. Estimates of k and p in Eq. [2] were obtained from nonlinear regression of the data in Fig. 3. For the ORR, values of MRLB following more than 3 yr of decomposition are assumed to approximate the ash content plus recalcitrant chemical matter of the original litter cohort.



View larger version (37K):
[in this window]
[in a new window]
  		Fig. 3. Published data for litter mass loss from mixed and species-specific litterbag studies conducted on the Oak Ridge Reservation (upper graph), and the fitted response (solid line) to the full data set with the data for Nyssa removed (lower graph). Dashed lines in the lower graph are the 95% confidence interval around the regression line. Error estimates associated with k and p are ±95% confidence intervals.

 
Using the calculated values of L% from Eq. [1a], an alternate estimate of the current-year litter cohort mass remaining after the 2001 growing season can be derived from the following equation:

[3]
where MR14C is the percentage mass of litter remaining estimated from isotopic data, M1 is the mass of the Oi-horizon remaining after one decomposition cycle and ML is the mass of the litter-cohort originally added to the Oi-horizon at time zero. In this paper, the numerator (L% x M1) is the estimated mass of the newly added litter cohort after the 9-mo period of decomposition.

One-way analysis of variance was used to test for time-zero differences in Oi-horizon mass and C/N ratio across research sites, and to test for within site differences in litter mass, {Delta}14C signature, and C/N ratio following litter addition treatments. The litterbag mass loss function was fitted using nonlinear regression. All statistical analyses were conducted using SPSS 11.0 for Macintosh (SPSS, Inc., Chicago, IL).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Litterbag Mass-loss Studies
Historical litterbag decomposition studies on the ORR, all conducted under similar weather conditions (Fig. 2), show similar rates of mass loss over time for species-mixtures and for Acer, Quercus, and Liriodendron species-specific studies (Fig. 3). Nyssa foliage (a common mid-canopy tree species) exhibited higher rates of mass loss over time and was not included in the fitted relationship (Fig. 3, lower graph). The Kelly and Beauchamp (1987) equation for residual litter mass with time (Eq. [2]) provided a significant fit to the multi-study data set (r2 = 0.86) showing effective losses of 42.5, 27.5, and 23.5% of remaining mass per year for a single litter cohort through three annual cycles of litter decomposition, by which time the initial litter cohort would have moved below the Oi-horizon. Notwithstanding small adjustments due to influx of fungal C and influx/leaching of nutrients, this fitted relationship suggests that approximately 20% of the initial litter cohort mass is resistant to decomposition. Application of the traditional litter mass loss equation [MR = exp(–kt); Jenny et al., 1949; Olson, 1963] would have produced an estimate of residual recalcitrant mass closer to zero after 5 yr with a correspondingly reduced r2.

Isotopic Enrichment or Dilution of the Oi-Horizon
Mean mixed-species litter quantity (454 g m–2) and the litter C/N ratio (92.2 g C g N–1) of the added 2000 litter cohort were not significantly different among sites, but initial mass of the Oi-horizon at time-zero did vary from site to site (Table 1). The west-end enriched sites (TVA and Pine Ridge) had the lowest starting Oi-horizon masses ranging from 312 to 379 g m–2, Walker Branch was intermediate from 414 to 477 g m–2, and Haw Ridge had the highest initial mass at 550 to 567 g m–2. No significant differences in initial Oi-horizon mass were found for the designated treatment plots (i.e., litter enrichment or near-background litter additions within each site). Site to site differences in Oi-horizon mass were no longer apparent after the first-year litter additions (Table 1). On average, Oi-horizon temperatures and leaf-litter water content did not differ among research sites or treatments, but brief periods of variable litter water content were occasionally observed in association with isolated summer precipitation events (data not shown).


View this table:
[in this window]
[in a new window]
  		Table 1. Mean dry mass (±SE) and the corresponding C/N ratio (C/N ± SE) of the Oi-horizon at time zero, 2000 litter cohort addition, and the Oi-horizon 9 mo after litter additions summarized by research site and litter addition type.

 
As expected by the nature of the 1999 14C-pulse, the litter {Delta}14C-signatures of extant Oi-horizons at time-zero (March 2001) differed between sites but not by plots within sites (Table 2). Haw Ridge and Walker Branch, located on the east end of the ORR, had initial {Delta}14C-signatures ranging from 190 to 260{per thousand}; near expected average values in the North Hemisphere for 2000 (Levin and Hesshaimer, 2000; Trumbore et al., 2002). Sites on the west-end of the ORR had higher {Delta}14C-signatures, with means ranging from 515{per thousand} (Pine Ridge) to 627{per thousand} (TVA). The initial differences between sites were driven by wind direction and proximity of a given site to the 14C-source in 1999 (Trumbore et al., 2002).


View this table:
[in this window]
[in a new window]
  		Table 2. Litter {Delta}14C-signature (±SE) of the Oi-litter at time-zero, the 2001 enriched or background litter cohorts, and the Oi-litter 9 mo after the addition of either enriched or near-background litter treatments. Also shown are the probabilities (P) for a test of significant changes in {Delta}14C-signature between sampling periods by treatment, and (if significant) the calculated percent contribution of new litter additions to the Oi-horizon after 9 mo of decomposition for background (Walker Branch and Haw Ridge) or recently enriched (Pine Ridge and TVA) research sites, respectively. NC = no calculation executed (see text).

 
Near-background (221{per thousand}) and 14C-enriched (1005{per thousand}) litter additions were made from pooled materials collected in the fall of 2000 (see methods) and their respective {Delta}14C-signatures did not differ between research sites. One growing season (9 mo) after the application of near-background versus enriched litter to the forest floor, significant changes in Oi-horizon {Delta}14C-signatures were observed at all sites (Table 2). Carbon-14-enriched treatments produced significant increases in {Delta}14C-signatures on the Walker Branch (+415{per thousand}), Haw Ridge (+406{per thousand}), and Pine Ridge (+181{per thousand}) sites. The magnitude of the difference was proportional to the difference between a site's initial condition and the 14C-signature of the enriched litter. Substantial within-site variability at the TVA site precluded detection of a significant {Delta}14C difference between time-zero and 9 mo of decomposition even though the mean {Delta}14C-signature of the enriched TVA plots was 173{per thousand} greater than the time zero values. Adding near-background litter to Walker Branch and Haw Ridge plots did not produce a significant change in Oi-horizon {Delta}14C-signatures, but the same additions resulted in a significant dilution of the {Delta}14C-signature of both the TVA (–317{per thousand}) and Pine Ridge (–201{per thousand}) Oi-horizons.

The percentage of the residual Oi-horizon contributed by the 2000 litter cohort (L%) was calculated only for those systems demonstrating a significant change in {Delta}14C-signatures (Table 2). Enrichment data for the Walker Branch and Haw Ridge sites yielded L% values of 55 and 50%, respectively. Enrichment data for Pine Ridge indicated a lower value around 38%, and the dilution of Pine Ridge and TVA plots from the addition of near-background litter suggested unexpectedly high L% values of 71 and 78%.

Residual Mass: Litterbags Versus Isotopic Change
Interpolation of the fitted equation for litterbag mass change over an 9-mo period (May 2001 through January 2002) indicated that the expected mass remaining (MRLB) from the previous year's cohort of litter would have been 65.4% plus or minus a rather narrow confidence interval (Fig. 3 and 4). Applying the significant L% values from Table 2 to the mass-dependent calculations of Eq. [3], the isotopic enrichment approach yielded MR14C values ranging from 45 to 59%, and the isotopic dilution value for the Pine Ridge site gave an MR14C of 63%. The calculated MR14C value for the dilution treatment at the TVA site yielded an unrealistic estimate of 105%. All estimate of MR14C were found to have substantial variation driven by spatial variability in Oi-horizon mass (Table 1) and {Delta}14C measurements (Table 2).



View larger version (27K):
[in this window]
[in a new window]
  		Fig. 4. Mean ±95% confidence intervals for litter mass remaining after 9 mo measured by litterbag mass loss studies (MRLB), or calculated from 14C litter-enrichment or 14C litter-dilution. The dashed line for comparison is the mean mass remaining based on MRLB. The "?" placed on the right-hand column indicates the uncertainty of this result (see Discussion) even though the calculated value of litter mass remaining was an acceptable result from the two-compartment model (Eq. [1a]).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Mean mass-loss estimates derived from past ORR litterbag studies provided a solid baseline from which to compare isotopically derived measurements of litter decomposition. The k value of 0.55 for the combined data set projects a first-year mass loss of 42.5%, which is within the range of similar values for deciduous hardwood forests (Hanson et al., 2003a). Gartner and Cardon (2004) reported that decomposition rates of mixed-species litter is often not scaleable from single-species experiments and reported that litter mixtures can produce ~20% greater rates of mass loss than observed in single species studies. The ORR data for Acer and Liriodendron litter in Fig. 3, however, coincide nicely with the Quercus and mixed-litter data sets for mass-loss on the ORR.

Mass loss in 2001 from the combined time-zero Oi-layer plus new litter additions ranged from 45 to 48% for the Walker Branch and Haw Ridge (Table 1) consistent with anticipated first-year mass loss from litterbags previously discussed. No statistically significant changes in mid-winter (i.e., January) dry mass of the Oi-horizon were observed for Walker Branch and Haw Ridge (Table 1) suggesting that the litter turnover rates were balanced by the 500 g m–2 experimental litter additions. Conversely, the Oi-horizon mass of the TVA and Pine Ridge sites showed significant accumulation between time zero and the January 2002 sampling (P-values of 0.01 and 0.07, respectively), and a correspondingly lower mean mass losses between 29 and 42% during 2001. Litter C/N ratios, an important variable associated with the rate of decomposition (Flanagan and Van Cleve, 1983; McClaugherty et al., 1985) were unchanged over time for Walker Branch and Haw Ridge sites, but were significantly greater for TVA and Pine Ridge sites (P-values of <0.001 and 0.035, respectively) confirming the accumulation of litter mass at the latter sites.

Because isotopic methods allow litter mass-loss to proceed normally without experimental artifacts, we might have expected the MR14C values to be lower than the values for litterbags. Litterbags are known to limit faunal facilitated comminution and fragmentation of organic debris, and may modify environmental conditions driving microbial decomposition (Anderson, 1973; De Santo et al., 1993; Gartner and Cardon, 2004). Observed MR14C values were lower for four of five estimates (Fig. 4), but the substantial variability around all estimates of MR14C suggests the need to consider potential sources of error involved in the measurement and calculation of MR14C.

Sources of Error
Observed variation in Oi-horizon mass is not consistent with the assumptions of the stable isotopic mixing model (Eq. [1a]). Changing mass of the Oi-horizon from time-zero to January 2002 at both Pine Ridge and TVA is partly responsible for the disproportionate decrease and increases in L% for the enriched and near-background litter addition treatments, respectively, at those sites. Random variation in thickness of the Oi-horizon within sampling plots also contributes to variability in the interpretation of the isotopic data. Furthermore, because the manual separation of the Oi- from the Oe/Oa-horizons is rather subjective (Federer, 1982), errors associated with the collection of the Oi-layer are compounded. These particular sources of error seem to have been minimal in the near-ambient Walker Branch and Haw Ridge plots (Table 1; SE ranging from 33 to 51 g m–2), but greater in the west-end TVA and Pine Ridge plots (Table 1; SE ranging from 34 to 114 g m–2). Interannual variation in weather conditions may have also contributed to inconsistent Oi-horizon mass, but the uniformity of temperature and moisture at all four research sites suggests that this was not the cause of differences between eastern and western sites on the ORR.

Dilution of isotopic signals from unmeasured C inputs having variable {Delta}14C-signatures is another potential confounding influence. Leake et al. (2001) reported significant changes in isotopic signatures for Pinus sylvestris litter from its exploitation by ectomycorrhizal fungal mycelium. Inputs of reproductive tissues (e.g., bracts, flowers, seeds), especially from Acer and Liriodendron species, were not quantified and would also result in the addition of near-background 14C-signatures onto the 1999-enriched TVA and Pine Ridge plots.

Finally, unequal losses of 14C-signatures from leaching, litter comminution, and decomposition processes represents yet another possible way for the assumptions of the two-compartment mixing model to be compromised. Anderson (1973) suggested that as much as 30% of the organic mass of a fresh litter cohort was soluble material with the potential to leach from litter and percolate out of the Oi-horizon. Experiments on additional samples of archived 2000 litter are underway to evaluate the importance of this source of error.

Evaluating Complex Mixtures
While it is not possible to derive explicit solutions for equations with multiple unknowns, data from multiple sources can be used to bracket the range of possible isotopic signatures and cohort masses in an attempt to explain the very high L% and MR14C for the TVA site. Because the Oi-horizon sampled after an annual cycle is likely to contain layered-litter from at least 2 annual cohorts, a three-compartment model can also be considered:

[4a]

[4b]
where S–1 is the {Delta}14C-signature of residual C from a litter cohort laid down approximately one year before time-zero; and L, S0 and S1 are as defined in Eq. [1]. The three-compartment mass balance equation can then be obtained by substituting litter cohort mass values for the mass fractions (fx) of Eq. [4a] and the conversion of {Delta}14C-signature data to appropriate ‘relative concentrations’ ([14C] = {Delta}14C/1000 + 1):


[5]
where MR1 and [14C1] are, respectively the Oi-horizon mass and [14C] remaining after 9 mo, and Mx and [14Cx] are the initial mass and [14C] contributions of the 2000 litter (x = L) and previous litter cohorts (x = 0 or –1). Dx is the periodic decomposition expected or measured from the respective cohorts (x = L, 0, –1). Carbon lost from the Oi-horizon is based on mass loss estimates, and does not necessarily represent the differential losses associated with comminution and leaching processes.

Equation [5] can be rearranged with assumptions about time-dependent cohort masses derived from litterbag mass-loss information in Fig. 3 to yield unique solutions that can be contrasted with the 14C-signatures of measured litter and Oi-horizon mass and {Delta}14C. Figure 5 is a diagrammatic representation of this approach showing approximations of mass loss estimates and multiple cohort contributions for examples of Oi-horizon 14C enrichment and dilution cases: Walker Branch and TVA. We assume that litter cohorts before 1999 had nearly uniform 14C-signatures consistent with small changes observed for the northern hemisphere (Levin and Hesshaimer, 2000).



View larger version (41K):
[in this window]
[in a new window]
  		Fig. 5. Diagram of the hypothetical path of 14C-isotopes through a well-mixed Oi-horizon with stable 14C-signatures (A. Walker Branch site), and through a manipulated Oi-horizon having distinct {Delta}14C signatures in all measurable litter cohorts (B. TVA site). Values contained within bold boxes are measured data from Tables 1 and 2. All other data are estimates of cohort-specific components of the time-zero and 9-mo litter Oi-horizons used to estimate one year's worth of change. Mass estimates (g dry matter m–2) and decomposition losses (g dry matter m–2 y–1) were derived from the fitted litterbag relationship in Fig. 3. The carbon content of newly added Oi-cohorts and older Oi litter was similar (46–47%).

 
In a stable natural system (approximated by Walker Branch and Haw Ridge), pretreatment litter cohorts making up the layers of the Oi-horizon were assumed to have similar {Delta}14C-signatures (~245{per thousand}). Given this assumption, we see that the application of litterbag decomposition estimates for one season to the measured data allows Oi-horizon {Delta}14C estimates to range from 617 to 696{per thousand} overlapping the measured value of 660{per thousand} (Fig. 5a). For the divergent case of the TVA Oi-horizon diluted by near-background litter (Fig. 5b), we need to consider that the original 1999–14C pulse may have incorporated a disproportionately enriched 1999 litter cohort (~674{per thousand}) into the Oi-horizon. Including this assumption in the three-compartment model allows us to evaluate the Oi-horizon remaining after one season of decomposition to be a variable combination of 14C-signatures associated with independent litter cohorts. Unfortunately, the results (397–406{per thousand}) still show limited agreement with the measured Oi-horizon data (308{per thousand}) after one season of decomposition, suggesting the need to include explicit measurements and parameterization of {Delta}14C-signatures and C losses associated with faunal comminution and leaching processes in our mass balance calculations. Such work is underway.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
This study successfully used mixed-species oak-forest litter enriched with 14C to evaluate turnover rates of litter within the Oi-horizon, an approach that avoids potential artifacts associated with litter containment systems (e.g., litterbags). Normal quantities of leaf litter were applied to large replicated field plots. On average, measured changes in the 14C-signature of the Oi-horizon of an upland-oak forest produced calculated estimates of litter mass remaining after one season that were lower than expected based on traditional litter bag studies. The difference between the two approaches was consistent with the exclusion of soil macrofauna by litterbags.

Even though isotopic enrichment or dilution techniques offer the advantage of being a non-invasive method, a high degree of variability for some of these sites undermined its utility as a substitute for litterbag mass loss methods. The 14C approach measures the residence time of C in the leaf litter, rather than the time it takes for leaves to disappear; hence radiocarbon measures are subject to C immobilization and recycling in the microbial pool, and do not necessarily reflect results from litterbag mass loss. The use of cohort-based 14C-signatures for tracking C movement and cycling beyond the litter layer via leaching or macrobiotic transport might improve the agreement between isotopic and mass-loss based estimates of litter decomposition.

While the two-compartment isotopic mixing model appears appropriate for systems with stable 14C inputs, it was insufficiently robust for uniform application within the current experiment. A more detailed model, as exemplified by the analysis of Fig. 5, is needed to adequately capture the complex processes involved in litter mass-loss and to handle interannual differences in the isotopic signature of litter cohort additions. Future analyses of additional data from the EBIS project will attempt to reconcile the patterns of 14C movement throughout the experimentally manipulated Ultisols and Inceptisols using sophisticated, cohort-based models of the organic horizons (e.g., Berg and Matzner, 1997; Berg, 2000).


   ACKNOWLEDGMENTS
 
Funding for the EBIS project was provided by the U.S. Department of Energy (DOE), Office of Science, Biological and Environmental Research (BER), as a part of the Terrestrial Carbon Processes (TCP) Program. Oak Ridge National Laboratory (ORNL) is managed by UT-Battelle, LLC, for the DOE under contract DE-AC05-00OR22725. EBIS project participants appreciate access and use of Tennessee Valley Authority (TVA) land on Chestnut Ridge near the Oak Ridge Reservation allowed under Contract No. 105906 between TVA and ORNL. Work at Lawrence Livermore National Laboratory (LLNL) was performed under the auspices of the U.S. Department of Energy by University of California, LLNL under Contract No. W-7405-Eng-48. The authors would like to thank Steve DiFazio and Stan Wullschleger for their helpful comments on early drafts of this paper.

Received for publication September 7, 2004.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
M. Froberg, P. M. Jardine, P. J. Hanson, C. W. Swanston, D. E. Todd, J. R. Tarver, and C. T. Garten Jr.
Low Dissolved Organic Carbon Input from Fresh Litter to Deep Mineral Soils
Soil Sci. Soc. Am. J., March 12, 2007; 71(2): 347 - 354.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hanson, P. J.
Right arrow Articles by Trumbore, S. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Hanson, P. J.
Right arrow Articles by Trumbore, S. E.
Agricola
Right arrow Articles by Hanson, P. J.
Right arrow Articles by Trumbore, S. E.
Related Collections
Right arrow Forest Soils
Right arrow Biogeochemical Processes
Right arrow Isotopes

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS