field variability of carbon isotopes in soil organic carbon

4
Nuclear Instruments and Methods in Physics Research B 123 ( 1997) 45 I-454 ELSEVIER Beam Interactions with Materials6 Atoms Field variability of carbon isotopes in soil organic carbon S.W. Leavitt a* * , E.A. Paul b, E. Pendall a, P.J. Pinter, Jr. ‘, B.A. Kimball a Laboratory of Tree-Ring Research, The University of Arizona. 105 W. Stadium, Tucson. AZ 8572 1, USA b Department of Crop and Soil Sciences, Michigan State Universiry, East Lansing, MI 48824-132s. USA ’ U.S. Water Conservation Laboratory, USDA-ARS, 4331 E. Broadway Rd., Phoenix, AZ 85040, USA Abstract Free-air CO, enrichment (FACE) plant-growth experiments conducted in Arizona have allowed a spinoff isotope tracer experiment (tank CO, is i4C- and i3C-depleted) to follow the input of carbon into soil organic matter pools. Accurate assessment of the ‘“C and 13C inputs to soils by this pathway requires detailed knowledge of the isotopic composition of the soils before the experiments. We have examined the variability of S13C in soil organic matter in 8 experimental plots prior to the beginning of the 1995-96 FACE experiments with wheat. 613C variability was higher immediately after harvest of a previous crop in June, but the plots were much more homogeneous five months later immediately before planting. Intervening field management, including disking, plowing and installation of irrigation drip tape likely contributed to mixing the soils. 1. Introduction There is increasing interest in using the ‘“C and 13C at their natural abundance levels to resolve questions of input and turnover of carbon in soil organic carbon pools. This interest will further escalates as this technique is used in studies where C, plants have been replaced by C, plants (and vice versa) [l-6] , and in studies where future high CO, conditions are simulated using tank CO, whose iso- topic composition may impart a tracer to be taken up by the growing plants and transferred to the soil [7]. When the CO, enrichment experiments are done on a smaller scale (e.g., growth chamber or greenhouse, with plants grown in pots), there is an ample opportunity to homogenize the soils and precisely quantify their isotopic composition prior to experimentation. In field-scale experiments such as FACE (Free-Air Carbon dioxide Enrichment), pre-experi- ment soil homogenization would be very difficult (and expensive), and therefore detailed pre-experiment charac- terization of isotopic variability may be the most effective means of assessing shifts in soil isotopic composition as a result of these types of experiments. In 1992, the third and final year of a FACE experiment with cotton was conducted at the Maricopa Agricultural Center of The University of Arizona, ca. 50 km south of Phoenix. The tank CO, gas used in the experiment was petroleum derived ( 14C pmC = 0; 613C = - 37%0), hope- fully imparting an isotopic signal of sufficient strength to * Corresponding author. Fax: + I-520-621 -8229; email: sleav- [email protected]. plants grown at elevated CO, concentration (550 ppmv) that if they contributed further to a carbon build-up in the soil, a change in soil carbon isotopic composition could be detected. Unfortunately this isotopic component of the experimentation was only added in the third year, and no soil samples had been taken from the plots prior to the beginning of the first experiment to establish the back- ground baseline (plot positions were the same in all three years). Assuming that the “enriched” and “control” plots had the same initial soil organic carbon isotopic composi- tion, the isotopic composition at the end of the experiment was measured and by isotopic mass balance, inputs of “new” carbon derived from the experiment were calcu- lated. An isotopic shift in 613C of the FACE soils consis- tent with ca. 8- 10% input of freshly derived carbon during the experiment. Questionable confirmation was obtained when average TAMS-measured 14C activity of total or- ganic soil carbon from two plot pairs were used in the isotopic mass balance to calculate a 9% input (mean of 2 enriched plots = 88.0 pmC and control mean = 89.1). However, when the measurements from two plot pairs were considered separately (pair 1, enriched value = 87.6 pmC, control value = 91.6; pair 2, enriched value = 88.3 pmC, control value = 87.01, mass balance would produce widely divergent results. The first few 613C numbers also actually produced initial ambiguous results, but the weight of additional measurements produced a more confident estimation. We hypothesized that some combination of lab precision and field variability may have contributed to uncertainty with few initial data points. Soil sampling methodologies involving multiple soil cores and sample 0168-583X/97/$17.00 Published by Elsevier Science B.V. All rights reserved PII SOl68-583X(96)00707-0 XII. ENVlRONMENTAL/PALEOCLIMATIC STUDIES

Upload: sw-leavitt

Post on 16-Sep-2016

217 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Field variability of carbon isotopes in soil organic carbon

Nuclear Instruments and Methods in Physics Research B 123 ( 1997) 45 I-454

ELSEVIER

Beam Interactions with Materials6 Atoms

Field variability of carbon isotopes in soil organic carbon

S.W. Leavitt a* * , E.A. Paul b, E. Pendall a, P.J. Pinter, Jr. ‘, B.A. Kimball ’

a Laboratory of Tree-Ring Research, The University of Arizona. 105 W. Stadium, Tucson. AZ 8572 1, USA b Department of Crop and Soil Sciences, Michigan State Universiry, East Lansing, MI 48824-132s. USA

’ U.S. Water Conservation Laboratory, USDA-ARS, 4331 E. Broadway Rd., Phoenix, AZ 85040, USA

Abstract Free-air CO, enrichment (FACE) plant-growth experiments conducted in Arizona have allowed a spinoff isotope tracer

experiment (tank CO, is i4C- and i3C-depleted) to follow the input of carbon into soil organic matter pools. Accurate

assessment of the ‘“C and 13C inputs to soils by this pathway requires detailed knowledge of the isotopic composition of the soils before the experiments. We have examined the variability of S13C in soil organic matter in 8 experimental plots prior to

the beginning of the 1995-96 FACE experiments with wheat. 613C variability was higher immediately after harvest of a previous crop in June, but the plots were much more homogeneous five months later immediately before planting. Intervening field management, including disking, plowing and installation of irrigation drip tape likely contributed to mixing

the soils.

1. Introduction

There is increasing interest in using the ‘“C and 13C at their natural abundance levels to resolve questions of input and turnover of carbon in soil organic carbon pools. This interest will further escalates as this technique is used in studies where C, plants have been replaced by C, plants (and vice versa) [l-6] , and in studies where future high CO, conditions are simulated using tank CO, whose iso- topic composition may impart a tracer to be taken up by the growing plants and transferred to the soil [7]. When the CO, enrichment experiments are done on a smaller scale (e.g., growth chamber or greenhouse, with plants grown in pots), there is an ample opportunity to homogenize the soils and precisely quantify their isotopic composition prior to experimentation. In field-scale experiments such as FACE (Free-Air Carbon dioxide Enrichment), pre-experi- ment soil homogenization would be very difficult (and

expensive), and therefore detailed pre-experiment charac- terization of isotopic variability may be the most effective means of assessing shifts in soil isotopic composition as a result of these types of experiments.

In 1992, the third and final year of a FACE experiment with cotton was conducted at the Maricopa Agricultural

Center of The University of Arizona, ca. 50 km south of Phoenix. The tank CO, gas used in the experiment was petroleum derived ( 14C pmC = 0; 613C = - 37%0), hope- fully imparting an isotopic signal of sufficient strength to

* Corresponding author. Fax: + I-520-621 -8229; email: sleav- [email protected].

plants grown at elevated CO, concentration (550 ppmv) that if they contributed further to a carbon build-up in the soil, a change in soil carbon isotopic composition could be detected. Unfortunately this isotopic component of the experimentation was only added in the third year, and no soil samples had been taken from the plots prior to the beginning of the first experiment to establish the back- ground baseline (plot positions were the same in all three years). Assuming that the “enriched” and “control” plots had the same initial soil organic carbon isotopic composi- tion, the isotopic composition at the end of the experiment was measured and by isotopic mass balance, inputs of “new” carbon derived from the experiment were calcu- lated. An isotopic shift in 613C of the FACE soils consis- tent with ca. 8- 10% input of freshly derived carbon during the experiment. Questionable confirmation was obtained when average TAMS-measured 14C activity of total or- ganic soil carbon from two plot pairs were used in the isotopic mass balance to calculate a 9% input (mean of 2 enriched plots = 88.0 pmC and control mean = 89.1). However, when the measurements from two plot pairs were considered separately (pair 1, enriched value = 87.6 pmC, control value = 91.6; pair 2, enriched value = 88.3

pmC, control value = 87.01, mass balance would produce widely divergent results. The first few 613C numbers also actually produced initial ambiguous results, but the weight of additional measurements produced a more confident estimation. We hypothesized that some combination of lab precision and field variability may have contributed to uncertainty with few initial data points. Soil sampling methodologies involving multiple soil cores and sample

0168-583X/97/$17.00 Published by Elsevier Science B.V. All rights reserved

PII SOl68-583X(96)00707-0 XII. ENVlRONMENTAL/PALEOCLIMATIC STUDIES

Page 2: Field variability of carbon isotopes in soil organic carbon

452 S.W. Lead et al./Nucl. Instr. and Meth. in Phys. Res. B 123 (1997) 451-454

composites have long been known as necessary to obtain representative physical and chemical analyses [8], although requirements may be different depending on the element to be analyzed. This study examines the field variability of 813C prior to the beginning of a new set of FACE experi- ments in 1995.

2. Methods

In this study, soil was sampled from the location of eight new circular plots (24 m-diameter plots embedded in a ca. 9 ha farm field) prior to the beginning of 1995-96 FACE experiments with wheat. These new plots were in locations not previously occupied by enriched or control plots during preceding FACE cotton and wheat experi- ments. On June 9, 1995, we sampled each plot with at least four soil cores taken along equally-spaced radii ca. 6 m from the center and at depths of both O-30 cm and 30-60 cm; in the case of two of the plots, 8 cores were taken. We obtained additional soils samples (O-15 cm soil depth with four cores from each plot) from core sampling done in November 15, 1995, immediately prior to planting of the wheat.

Soil samples were first passed through a 1 mm sieve to remove large plant fragments and pebbles, and a subsam- ple of 50-100 g was acidified with 0.5-1N HCl to remove soil carbonates. All recognizable plant fragments were removed by skimming from this liquid surface, by skim- ming after subsequent immersion of the sample into a concentrated NaCl solution ( p = 1.2 g cm-3), and by Anal manual removal from the rinsed, dried, and ground soil using forceps under 20 X magnification.

Because of the expense of TAMS 14C analysis and the large number of samples, we have only analyzed 613C of the total organic carbon fraction of these soils. Organic carbon content and 813C were determined after combustion of 0.10-0.25 g subsamples of these soil samples in quartz tubes at 900°C in the presence of copper oxide, silver foil and copper turnings [9]. Carbon yield was determined manometrically from the CO, combustion product, and CO, was analyzed primarily with a Finnigan MAT Delta-S mass spectrometer to determine 813C. Multiple combustion

Table 1 .^

and analyses of the same soil sample indicate precision of ca. 0.30%0 (1 s). The carbon content reproducibility ( f 1 s) is estimated as ca. +0.05%.

3. Results and discussion

The SL3C and %C results for the O-30 cm and 30-60 cm depths collected on June 9 are depicted in Fig. 1. Each of the 8 plot locations is graphed separately (F = FACE [enriched], C = control), and the averages of all enriched and control plots are summarized in Table 1. The two least negative (‘3C-enriched) samples at O-30 cm were at the 3C and 3F locations in the southeastern comer of the field. Initially, we felt that this was evidence of a higher propor- tion of organic carbon derived from plants with the C, photosynthetic pathway at some point in the soil’s devel- opment. These were Sonoran Desert soils (clay loam, entisols) developed along the adjacent Santa Cruz Wash system prior to first cultivation ca. 30-40 years ago. The high isotopic variability in June 1995 suggests that were the experiment to start on June 6, the CO,-enriched plot O-30 cm soil organic carbon 813C values (--23.38%0) were already 13C-depleted relative to control plots l-22.06%0) even before any wheat plants would be ex- posed to the ‘“C- and 13C-depleted tank gas used in the experiment. The 813C isotopic variability was ca. 1%0 for both enriched and control plots. The soil organic content at O-30 cm seemed to be lowest in the southwestern part of the field where 4F and 4C were located, although the average for enriched and control plots are virtually indis- tinguishable (Table 1).

At the 30-60 cm depth (Fig. 1). the average 813C values of the enriched plots ( - 22.03%0) and control plots (- 22.01%0) are nearly identical, suggesting greater homo- geneity at depth. Likewise, in the individual plot pairs (i.e., IF-lC, 2F-2C, etc.) neither enriched nor control plots are consistently more 13C-enriched. Furthermore, the isotopic variability was only 0.3-0.4%0 for all control and enriched plots, which is substantially lower than in the surface soils. On average, the organic carbon content is about 40% of that in the O-30 cm layer, with similar variability.

The samples at the O-15 cm depth taken in November show a much more uniform isotopic composition across all

Average 8°C and total organic carbon content of the soil in all four enriched plots and four control plots prior to the beginning of the 1995-96 FACE experiment

Date Depth Plot S”C Is c (a) IS n

June 1995 O-30 cm enriched - 23.38%~ 0.9%0 0.743 0.061 20

control - 22.06 1.36 0.738 0.060 20 June 1995 30-60 cm enriched - 22.03 0.26 0.310 0.078 20

control - 22.01 0.39 0.296 0.085 20

Nov. 1995 O-15 cm enriched - 23.09 0.16 0.591 0.093 16

control - 22.75 0.40 0.595 0.042 16

Page 3: Field variability of carbon isotopes in soil organic carbon

S.W. Leavin et al./Nucl. Instr. and Meth. in Phys. Res. B 123 (1997) 451-454 453

-16

C-30 cm -2o-

1

2 -2v 1 t z

y -24. $ a

+ + .i., t

o -2e I .- -. t

-0.6

1 1

1

.A?

-0.7 I , g

-0.6 o

-0.5

1 LO.4 IC IF 2F 2C 3C 3F 4F 4C

_.

: 30-60 cm

-21 -I

I 10.1 IC IF 2F 2C 3C 3F 4F 4C

Plot

Fig. 1. The 613C ( H ) and %C ( + ) of soil organic matter from

surface soils (upper) and subsurface soils (lower) sampled in June 1995. The designations “lc”, “IF”, etc., refer to the plot pair

and whether its treatment will be CO,-enriched (“F”) or control (*‘C”). The seqence of plots on the x-axis follows the arrange- ment in the field: the first four are on a line from west to east, and

the second four are on a parallel line from east to west, and

directly south of the plots on the first line. Error bars are k 1 s.

plots (Fig. 2). On average, the enriched plots are ca. 0.34%0 lighter than the control plots, but this difference is not statistically significant, indicating that at the start of the experiment the field was homogenized in the upper layers. This may be a consequence of the removal of old irrigation drip-tapes, disking to a depth of ca. 50 cm, chisel-plowing, field leveling, and installation of new drip tapes, all of which took place between June and November 1995. The average total organic carbon contents of 0.59- 0.60% for the enriched and control plot locations represent a value intermediate between the June O-30 and 30-60 cm total organic carbon contents. This would be consistent with plowing to 50 cm resulting in mixing in the low-carbon

1 I,,, 1C IF 2F 2C 3C 3F 4F 4C

PIIS

Fig. 2. The 613C (B) and %C (+) of soil organic matter from

surface soils sampled in November 1995. Plots are arranged as

described in Fig. 1. Error bars are f 1 s.

deep soils with the high-carbon soils of the upper 30 cm. The unusually low carbon content for 2F can be attributed

to one soil core with 0.18% organic carbon compared to

the other three cores from 2F ranging from 0.53-0.56%. A study of 613C variability of soil organic matter (0- 10

cm) in Saskatchewan [lo] in a 1.44 ha faxm field found

variability of 0.64%0 (Is). Van Kessel et al. [ 101 attributed some of this variability to slight changes in microtopogra- phy which would affect water availability to crops, and

hence the 613C of plants. This variability was less than that

of our June results, but more than we found in November. During the actual FACE experiment, the pre-plant field

levelling should minimize microtopography effects, but

there may be some localization of water associated with drip irrigation tapes that may contribute to field 613C

variability of plants and possibly soils.

4. Conclusions

There was much more field variability of 613C in total

organic carbon of surface soils (O-30 cm) in June, 1995, six months prior to the beginning of the 1995-96 FACE experiment than there was just two weeks before the

experiment. The deeper soils (30-60 cm>, however, were much more homogeneous across the field in June. The June soils were taken just after a winter-spring crop of oats was harvested from the field, which may have intro- duced heterogeneities, and before the field was mechani- cally mixed and homogenized by removal and installation of drip tape, disking, plowing and levelling. The soils sampled in November had more uniform 613C, and their organic carbon content was intermediate between the June deep and shallow soils. Additional 15-30 cm samples from November 1995 are being analyzed to determine the

depth of the homogenization. Previous results [7,11] imply the inputs of isotopically

labelled “new” carbon from these CO, enrichment exper- iments may only be sufficient to alter the isotopic compo-

sition of the soil carbon by 1%0 or so. Consequently, there is sufficient variability of 613C and total organic carbon content in these cultivated soils to require multiple samples

to quantify the average composition before the experiment. Without this baseline information, it may be impossible to quantify the real carbon inputs to soil carbon pools.

Acknowledgements

Soil sampling was assisted by D. Kane of the Labora- tory of Tree-Ring Research (LTRR) and H.-Y. Cho and staff of the Maricopa Agricultural Center. We thank S.

Schorr, and D. Kane for laboratory pretreatment and com- bustion of soils, and B. McCaleb for 613C analysis of the CO, samples in the Laboratory of Isotope Geochemistry at the University of Arizona (A. Long, Director).

XII. ENVIRONMENTAL/PALEOCLIMATIC STUDIES

Page 4: Field variability of carbon isotopes in soil organic carbon

454 S.W. Leavitt et al./Nucl. Instr. and Meth. in Phys. Res. B 123 (1997) 451-454

This research was supported by the U.S. Department of Energy (Terrestrial Carbon Processes Research Program) under grant DE-FGO3-95ER to The University of Arizona (S. Leavitt, T. Thompson, A. Matthias, R. Rauschkolb and H.-Y. Cho, PI’s). Additional FACE project support is being provided by the Agricultural Research Service, U.S. Department of Agriculture, especially the U.S. Water Con- servation Laboratory, Phoenix, AZ. This work contributes to the Global Change Terrestrial Ecosystem (GCTE) Core Research Programme, which is part of the International Geosphere-Biosphere Programme (IGBP).

[31

[41

151

k51

[71

Bl

191

References

[I] R.S. Dzurec, T.W. Boutton, N.M. Caldwell and B.N. Smith,

Oecologia 66 (19851 17.

[2] J. Balesdent, G.H. Wagner A. and Mariotti, Soil Sci. Sot.

Am. J. 52 (1988) 118.

[lOI

[ill

J. Balesdent and M. Balabane, Soil Biol. B&hem. 24 (19921

97.

J. Balesdent and A. Mariotti, Soil Biol. Biochem. 19 (19871

25.

J. Balesdent, C. Girardin and A. Mariotti, Ecology 74 (1993)

1713.

A.R. Townsend, P.M. Vitousek and S.E. Trumbore, Ecology

76 (19951 721.

S.W. Leavitt, E.A. Paul, B.A. Kimball, G.R. Hendrey, J.R.

Mauney, R. Rauschkolb, H. Rogers, K.F. Lewin, J. Nagy,

P.J. Pinter and H.B. Johnson, Agric. For. Meteorol. 70

(1994) 87.

H.D. Chapman and P.F. Pratt, Methods of Analysis for Soils,

Plants and Waters (Univ. Calif., Div. Agric. Sci., 1961) p. 1.

T.W. Boutton, in: Carbon Isotope Techniques, eds. D.C.

Coleman and B. Fry (Academic Press, San Diego, 1991) p.

173.

C. Van Kessel, R.E. Farrell and D.J. Pennock, Soil Sci. Sot.

Am. J. 58 (1994) 382.

S.W. Leavitt, E.A. Paul, A. Galadima, F.S. Nakayama, S.R.

Danzer, H. Johnson and B.A. Kimball, Plant and Soil, in

press.