carbon stable isotope composition of modern ...

CARBON STABLE ISOTOPE COMPOSITION OF MODERN CALCAREOUS SOIL PROFILES IN CALIFORNIA: IMPLICATIONS FOR CO2 RECONSTRUCTIONS FROM CALCAREOUS PALEOSOLS NEIL J. TABOR AND TIMOTHY S. MYERS Roy M. Huffington Department of Earth Sciences, Southern Methodist University, Dallas, Texas 75275-0395, USA e-mail: [email protected] ERIK GULBRANSON Department of Geosciences, University of Wisconsin, Milwaukee, Wisconsin 53201, USA CRAIG RASMUSSEN Soil, Water and Environmental Science, University of Arizona, Tucson, Arizona 85721, USA AND

NATHAN D. SHELDON Department of Geological Sciences, University of Michigan, Ann Arbor, Michigan 48105-1005, USA ABSTRACT: Fourteen soil profiles from California were collected in order to measure the d13C of coexisting soil calcite and organic matter. Thirteen of the profiles contained a measurable amount of calcite ranging from 0.04 to 54.6 wt %. Soil calcite d13CPDB (d13C value vs. the calcite standard Peedee Belemnite) values range from 14.4 to 1.3ø, whereas organic matter d13CPDB values range from 24.0 to 27.7ø. The hydrology of these profiles is divided into two broad groups: (1) soils characterized by gravity-driven, piston-type vertical flow through the profile and (2) soils affected by groundwater within the profile at depths where calcite is present. The difference between soil calcite and organic matter d13CPDB values, D13Ccc-om, is smaller in profiles affected by groundwater saturation as well as most Vertisols and may be a product of waterlogging. The larger D13Ccc-om values in soils with gravity-driven flow are consistent with open-system mixing of tropospheric CO2 and CO2 derived from in situ oxidation of soil organic matter with mean soil PCO2 values potentially in excess of ;20,000 ppmV at the time of calcite crystallization. There is a correlation between estimates of soil PCO2 and a value termed ‘‘EPPT-U’’ (kJm2/yr) among the soil profiles characterized by gravity-driven flow. EPPT-U is the energy flux through the soil during periods of soil moisture utilization, and it is the product of water mass and temperature in the profile during the growing season. Thus, soils with high water-holding capacity/storage and/or low/high growing season temperature may form soil calcite in the presence of high soil PCO2, and vice versa. The results of this research have important implications for reconstructions of paleoclimate from stable carbon isotopes of calcareous paleosol profiles. KEY WORDS:

carbon isotope, oxygen isotope, soil, water balance, soil CO2, California

INTRODUCTION Calcite is an important authigenic soil mineral component in subhumid to arid climates (e.g., Buol et al. 2003). Calcite-bearing soils in ancient environments (paleosols) are also an abundant component of the terrestrial stratigraphic record (e.g., Retallack 1997, Alonso-Zarza and Tanner 2006). While much research has been done to understand their morphological development and potential relationship to factors of soil formation, climate in particular (Jenny 1941; Arkley 1963; Gile et al. 1981; Machette 1985; McFadden and Tinsley 1985; Retallack 1994, 2005; Royer 1999), stable carbon isotope analysis of soil and paleosol calcite has been increasingly used over the past three decades in order to define quantitative paleoclimate proxies. It has been recognized for some time that the d13CPDB (d13C value vs. the calcite standard Peedee Belemnite) value of soil calcite (CaCO3) typically represents a mass–balance mixture of at least two CO2 end members: CO2 derived from oxidation of organic matter in the soil and CO2 derived from the global troposphere (e.g., Cerling 1984, 1991; Yapp and Poths 1996; Breecker et al. 2009). This understanding has led to applications of stable carbon isotope values of calcite from paleosol profiles in order to reconstruct numerous aspects of paleoclimate and paleoenvironment, including temporal and spatial distributions of

photosynthetic pathways (e.g., C3 vs. C4) upon paleolandscapes and estimation of paleoatmospheric PCO2. However, because the activity of CO3 in calcite is fixed at unity there is no way to determine from this mineral the partial pressure of CO2 in the soil at the time of calcite crystallization. Therefore, in order to solve the two end-member mixing equations for the partial pressure of atmospheric PCO2 or d13C values of soil-derived CO2 (which permits estimation of C3 vs. C4 photosynthesizers undergoing oxidation in the soil profile), a partial pressure of CO2 in the soil at the time of calcite crystallization must be assumed. Most studies of paleosol calcite d13C values have assumed a uniform soil PCO2 of 5000 ppmV (Cerling 1991, Yapp and Poths 1996, Ekart et al. 1999, Tabor et al. 2004) because this is an approximate average of mean soil PCO2 measurements in modern soils characterized by pedogenic calcite accumulation. More recent studies of paleosol calcite d13C values have assumed different soil CO2 concentrations according to paleosol morphology (e.g., Montan˜ez et al. 2007; Gutierrez and Sheldon 2012; see description in Sheldon and Tabor [2009]) because there appears to be a systematic difference in the mean soil PCO2 among different soil orders, and these can be recognized in a straightforward manner through detailed study of paleosol morphology (e.g., Mack et al. 1993; Retallack 1993, 1994). Recently, Breecker and others (2009) reported carbon and oxygen

New Frontiers in Paleopedology and Terrestrial Paleoclimatology j DOI: 10.2110/sepmsp.104.07 SEPM Special Publication No. 104, Copyright Ó 2013 SEPM (Society for Sedimentary Geology), ISBN 978-1-56576-322-7, p. 17–34. This is an author e-print and is distributed by the authors of this article. Not for resale.

18

NEIL J. TABOR, TIMOTHY S. MYERS, ERIK GULBRANSON, CRAIG RASMUSSEN, AND NATHAN D. SHELDON

isotope measurements from soil calcite and soil CO2 of young profiles (,1000 years) from semiarid settings in north–central New Mexico that indicate (1) calcite forms only during the dry season as soil biota begin to undergo moisture stress, and (2) soil PCO2 during this interval of time is remarkably low compared to those values assumed in most previous studies of paleosol calcite. In spite of these sobering results, it remains a fact that we do not know the partial pressure of CO2 in the soil for most modern and ancient soil-forming systems. Presented herein are the d13C values of calcite and coexisting organic matter from US Department of Agriculture (USDA) official soil series profiles in California (western USA). These data are used to approximate the partial pressure of soil CO2 at the time of calcite crystallization. This data set constitutes a reasonable collection of soils with which to evaluate the typical range of, and the potential factors responsible for, soil PCO2 values during calcite crystallization for the following reasons: (1) Contemporary tropospheric CO2 is approximately 380 ppmV, with a d13C value of 8ø (Schlessinger 1996), but atmospheric PCO2 did not exceed 280 ppmV (Petit et al. 1999) and has maintained a d13C value of approximately 6.5ø (Lourantou et al. 2010) over the past 125,000 years. (2) Californian terrestrial ecosystems have been dominated by C3 photosynthesizers over the period of soil formation (Jacobs et al. 1999), and, therefore, d13C measurements of modern floras should provide reasonable estimates of soil-derived CO2 d13C values. (3) The timing and temperature of calcite crystallization may be reasonably constrained by synoptic weather data collected from climate and soil monitoring stations (e.g., Thorthwaite 1948). (4) The official soil series descriptions provide a virtually unique opportunity to evaluate whether the morphology or hydrology of these soil profiles facilitates crystallization of soil calcite during open-system exchange with the overlying atmosphere or if the calcite may be influenced by closed-system behavior or inheritance of some precursor carbonate. (5) The large range of soil temperatures and annual precipitation values among these soil profiles permits evaluation of basic climate factors against the variation of the soil PCO2 estimates.

BACKGROUND AND RATIONALE Samples Available for Analysis Samples in this work come from the California ‘‘Soil Series Pedolarium,’’ a systematic collection of soil material archived during the 1960s and 1970s as part of the USDA–Natural Resources Conservation Service (NRCS) official soils series profile descriptions (Soil Survey Staff 2010a), which is housed in the Department of Land, Air and Water Resources at the University of California, Davis campus. Initially, 79 different soil profiles represented by 436 soil horizons were subsampled from the archive by N.J.T. in 1999 as part of a separate study that involved nondestructive magnetic analysis. These subsamples were selected from bulk soil material and packed into 1-cm3 plastic containers. The subsamples were revisited by N.J.T. in 2008 for the current study. Among the initial 79 soil profiles subsampled from the archive, 14 profiles, represented by 42 different horizons, are reported to contain calcite (Fig. 1; Tables 1, 2). Information about these 14 profiles, named informally profiles ‘‘1’’ through ‘‘14’’ here, is given in Table 1. The carbonate among these soil subsamples is finely crystalline to microcrystalline calcite that occurs either as disseminated cement through the soil matrix or as filamentous concentrations lining pores and tubules (Table 1). Only a few of these samples included indurated, millimeter-sized carbonate nodules (Table 1). Note, however, that some of these soils include larger nodules and petrocalcic horizons, but as a result of the sampling of soil material in 1-cm3 plastic containers they were not included in this analysis. Macroscopic examples of soil organic matter, such as soil root matter, are relatively rare among these

samples. Nevertheless, fine rootlets from at least one carbonate-bearing soil horizon within each profile were isolated from bulk soil matrix and collected for measurement of stable carbon isotope composition.

Characteristics of the Soils The profiles reported to contain carbonate (Table 1) occupy a geographic domain across most of the state of California, from 117.458W to 121.878W and 33.958N to 41.548N (Fig. 1; Table 1). The 14 soil profiles reside within five different soil orders in the USDA Soil Taxonomy classification system (Soil Survey Staff 2010a, 2010b). Among these 14 profiles are six Vertisols, four Mollisols, two Aridisols, one Alfisol, and one Entisol (Table 1). Soil profile descriptions are taken from the Soil Survey Staff, NRCS, USDA Official soil series descriptions. They are available online at http://soils. usda.gov/technical/classification/osd/index.html (accessed July 6, 2010, through February 10, 2011). Most of these profiles are developed upon alluvium, with a mixed composition of sedimentary, igneous, and metamorphic rocks. However, profile 4 is developed upon mixed shale and sandstone strata of Mesozoic age; profile 10 is developed upon Paleozoic-age limestone strata; and profile 2 is developed upon Tertiary-age lacustrine sediment. The landscape positions of these profiles occur on flat- to low-angle (,88) slopes occurring at elevations ranging from approximately 100 to 2900 m above sea level. Elevation is strongly correlated with surface–atmospheric mean annual temperatures (MAT) ranging from 7 to 188 C and soil temperature regimes ranging from frigid to hyperthermic (Table 1). Mean annual precipitation among the profiles ranges from 100 to 432 mm/yr and is concentrated during winter and early spring. The soil moisture regime for most of these profiles is xeric or aridic/torric (Soil Survey Staff 2010a, 2010b). A xeric soil moisture regime indicates that soils are moist in some part of the profile on more than one-half of the days that the soil temperature is .58 C, dry in all parts for more than 45 consecutive days in the 4 months following summer solstice, and moist in all parts for more than 45 consecutive days in the 4 months following the winter solstice. ‘‘Aridic’’ or ‘‘Torric’’ soil moisture regimes indicate that profiles are dry more than half of the time when they are not frozen and are never moist for more than 90 consecutive days when soil temperatures are above 88 C in most years. Ground-covering plants and/or trees cover all profiles. Root systems occur from the surface to a minimum depth of 45 cm (profile 1) and a maximum depth of 185 cm (profile 11). The drainage class of the soil profiles ranges from somewhat poorly drained to somewhat excessively drained (Table 1). Profile 4 is characterized by redoximorphic Fe-segregations indicative of intermittent, but long-duration, periods of saturated and anoxic (‘‘aquic’’; Vepraskas 1994) soil moisture conditions. Profiles 8 and 9 became saturated by groundwater rise (‘‘endosaturation’’) to ,75 cm beneath the soil surface during spring months. As discussed below, the hydrology, soil moisture balance, morphology, and chemistry of profiles 4, 8, and 9 differ from those of other profiles in this study as a result of the effects of groundwater accumulation within the rooting zone. Subsequent discussion of profiles 4, 8, and 9 refers to them collectively as ‘‘groundwater affected.’’

METHODS Samples of soil material that contain calcite were ground to a powder in a corundum mortar and pestle. Powders were weighed and then loaded into carbonate reaction vessels equipped with a sidearm for introduction of 100% H3PO4. Between 100 and 830 mg of sample powder was used depending upon the concentration of calcite in the soil material. After loading of the sample and H3PO4, atmospheric gases and sorbed water were evacuated from the reaction vessels using high-vacuum extraction lines, and calcite was then dissolved in vacuo

This is an author e-print and is distributed by the authors of this article. Not for resale.

STABLE CARBON ISOTOPE COMPOSITION OF CALCAREOUS SOILS

19

FIG. 1.—Location map of soil profiles analyzed in this study. The numbers on the map correspond to the numbers assigned to the soil profiles throughout the text, tables, and subsequent figures. by combining 100% H3PO4 with the soil material at 258 C for ;16 hours. The resulting CO2 samples were cryogenically purified, and CO2 yields were determined via mercury manometer with a precision of 60.2 lmol. CO2 samples were analyzed for carbon and oxygen isotope ratios using a Finigan MAT 252 isotope ratio mass spectrometer in the Huffington Department of Earth Sciences at Southern Methodist University. Individual rootlets collected from bulk soil material were treated overnight with 1 N HCl in order to remove any calcite that may have been present on the sample. Rootlets were then rinsed with deionized H2O until the rinse water remained at the original pH of the rinse water (;5.6) after at least 10 minutes of contact with the acid-leached sample. These organic-enriched residues were freeze-dried to remove sorbed water, and the carbon content and d13C values of these samples were determined from CO2 produced by closed-system combustion of the organic matter following the methods of Boutton (1991). Isotope values are reported in standard delta notation, as follows: d13 C ¼ ðRsample =Rstandard 1Þ 3 1000; where R is the ratio of heavy to light stable isotope (13C/12C) present in the sample or standard. d13C values are reported relative to the Peedee Belemnite standard (PDB; Craig 1957).

RESULTS The samples yielded calcite concentrations that ranged from 0.04 to 54.6 wt % of the bulk soil material (Table 2). The exception was profile 3 (Alo), which did not contain calcite despite the official soil series description indicating carbonate between 75 and 90 cm below the soil surface. Because calcite was not found in profile 3, it will not be considered in the subsequent ‘‘Discussion’’ section. Calcite d13CPDB values range from 14.4 to 1.3ø. Among the 10 soil profiles with multiple horizons containing calcite, carbon isotope variability within individual soil profiles ranges from 0.1 to 5.4ø. Profile 5 (Alturas; Fig. 2A) shows a strong correlation between increasing soil depth and more negative calcite d13C values from 2.5 to 120 cm. Two other profiles (profiles 6 and 14) exhibit little calcite d13C variation, ,1ø, with depth. Three other profiles (profiles 1, 2, and 14) record more positive calcite d13C values, up to 4.8ø, beneath the shallowest calcareous horizon. Four other profiles (profiles 8, 9, 9, and 10; there were two separate collections of profile 9; see Table 2) exhibit more complicated patterns with multiple positive and negative carbon isotope trends (e.g., Fig. 2B). With respect to soil taxonomic order, Entisols (profile 10) have the most positive d13C values (0.2 6 1.9ø, n ¼ 3), followed by Aridisols (3.1 6 3.3ø, n ¼ 4), Alfisols (7.1 6


Soil taxon

33.958N, 117.458W coarse-loamy, mixed, superactive, hyperthermic Typic Haplocalcid Ager 41.458N, 120.858W very fine, smectitic, mesic Chromic Haploxerert Alo 36.588N, 121.858W Fine, smectitic, thermic Aridic Haploxerert Altamont 37.758N, 121.708W Fine, smectitic, thermic Aridic Haploxerert Alturas 40.308N, 120.478W Fine, smectitic, mesic Typic Natrixeroll Amedee 40.308N, 120.208W Sandy, mixed, mesic duric Xeric Haplocalcid Anacapa 34.208N, 119.208N coarse-loamy, mixed, superactive, thermic Calcic Pachic Haploxeroll Beckwourth 39.758N, 120.298W coarse-loamy, mixed, superactive, mesic Oxyaquic Argixeroll Bellevista 39.758N, 120.188W fine-loamy, mixed, superactive, mesic Typic Durixeralf Beveridge 36.588N, 117.918W loamy-skeletal, carbonatic, frigid Lithic Torriorthent Bidwell 41.548N, 120.158W ashy, glassy, mesic Vitritorrandic Argixeroll Cropley 37.468N, 121.878W fine, smectitic, thermic Aridic Haploxerert Centerville 35.928N, 118.958W fine, smectitic, thermic Aridic Calcixerert Montague 41.838N, 122.428W fine, smectitic, mesic Petrocalcic Calcixerert

Aco

Location

Carbonate phase

Vertisol

Vertisol

Vertisol

Mollisol

Entisol

Alfisol filaments, nodules diffuse cement small nodules small nodules small nodules

filaments

Mollisol small nodules

Vertisol small nodules Vertisol diffuse cement Vertisol diffuse cement Mollisol diffuse cement Aridisol diffuse cement Mollisol diffuse cement

Aridisol diffuse cement

Soil order2

alluvium

alluvium

ashy alluvium alluvium

limestone

alluvium

alluvium

E

E

D to E

E

F

D

C

833

174

163

1450

2879

1439

1424

152

E

alluvium

1360

682

–

1364

100

1213

D to E

E

E

E

E to F

Drainage3

13

17

16

9

7

11

8

16

18

11

17

–

12

18

22

27

20

20

17

19

21

18

24

21

23

–

20

24

18 27

2 1

1

8

9

–

21

21

22

18

2

0

19

12

25

19

2 13

22

8

17

20

2 –

18

13

14

18

17

15

12

10

12

16

13

12

14

13

12

15

470

277

320

317

147

347

462

377

230

315

361

–

374

250

JJA DJF Elevation MAT temper- temperMAP (m) (8 C) ature ature Tmax4 TU5 (mm)

alluvium

lake sediment shale, sandstone shale, sandstone alluvium

alluvium

Parent material

MAT ¼ mean annual temperature; JJA ¼ June, July, August; DJF ¼ December, January, February; MAP ¼ mean annual precipitation. 1 Profile names are soil series terms for profiles that have been classified for the purposes of official US Department of Agriculture (USDA) Soil Survey and mapping. 2 Soil orders are from USDA Keys to Soil Taxonomy (Soil Survey Staff 2010b). 3 Soil drainage classes are as described in the USDA Keys to Soil Taxonomy (Soil Survey Staff 2010b). A ¼very poorly drained; B ¼ poorly drained; C ¼ somewhat poorly drained; D ¼ moderately well drained; E ¼ well drained; F ¼ somewhat excessively drained; G ¼ excessively drained. 4 Tmax is the maximum monthly mean surface air temperature over a soil. 5 TU values are mean surface air temperature over a soil during month(s) of soil moisture utilization, as determined from water balance calculations (see text and Table 3 for further explanation).

14

13

12

11

10

9

8

7

6

5

4

3

2

1

No.

Profile name1

TABLE 1.—Geographic, soil morphological, and climatic information about the calcite-bearing soil profiles used in this study.

20 NEIL J. TABOR, TIMOTHY S. MYERS, ERIK GULBRANSON, CRAIG RASMUSSEN, AND NATHAN D. SHELDON



21

TABLE 2.—d13C values of soil calcite and organic matter as well as estimates of soil PCO2. No.

Profile name

1

Aco

2

Ager

3 4 5

Alo Altamont Alturas

6

Amedee

7 8

Anacapa Beckwourth

9

Bellevista

9

Bellevista

10

Beveridge

11 12

Bidwell Cropley

13

Centerville

14

Montague

Depth (cm)1

Wt. %, CaCO3

d13CCC, øPDB

d13COM, øPDB

Soil2 PCO2

Soil3 PCO2

25.4 81.3 40.6 73.7 91.4 109.2 124.5

1.1 54.6 0.04 0.4 1.3 3.4 0.6

7.8 3.0 13.9 11.7 9.6 9.5 7.3

24.6

1050 460 N.R.4 N.R.4 2020 1920 950

1540 550 N.R.4 N.R.4 3760 3440 1210

83.8 2.5 15.4 34.3 52 66.0 66.0 91.4 91.4 117 137.2 30.5 58.4 91.4 71.1 71.1 104 137.2 14.0 36.8 55.9 55.9 83.8 11.4 38.1 55.9 55.9 83.8 4 16.5 38.1 127.0 96.5 96.5 50.8 66.0 68.6 68.6 83.8

0.8 3.7 2.2 4.9 0.2 1.5 2.9 6.2 1.9 3.2 0.07 8.6 12.0 1.0 2.5 0.2 0.04 0.04 2.5 3.6 6.6 13.9 4.5 2.7 4.1 4.9 5.6 4.7 1.2 1.0 17.5 0.1 0.5 0.3 0.5 2.8 17.5 8.3 3.5

14.4 5.1 6.8 8.3 9.1 8.5 9.0 9.1 8.7 9.4 8.8 0.5 1.1 7.4 9.8 9.9 11.0 5.6 6.3 7.6 6.9 6.8 7.8 6.1 7.5 6.9 6.9 8.2 1.3 2.3 0.4 9.0 14.3 14.3 14.1 11.4 9.0 8.9 9.1

N.R.4 680 960 1510 2180 1630 2060 2180 1780 2610 1870 350 370 1300 4440 5010 N.R.4 770 990 1510 1170 1140 1640 860 1260 1050 1050 1620 300 420 320 3780 N.R.4 N.R.4 5290 1330 2410 2570 2260

N.R.4 770 1160 2070 3570 2310 3280 3570 2620 4910 2810 390 420 1480 13,390 20,390 N.R.4 870 1320 2430 1670 1600 2790 1090 1810 1410 1410 2680 350 520 380 30,360 N.R.4 N.R.4 N.R.4 1970 6140 7350 5270

25.3 24.2 25.0

24.7

25.6 24.2 24.7

24.0

24.5

24.1 23.4 25.6 27.7 24.0

1

Depths are centimeters from the soil surface. Reported values reflect the midpoint of each horizon, as reported upon the sample vials housed in the Pedolarium at the University of California, Davis. 2 Soil PCO2 estimates calculated for maximum average monthly surface air temperature (Tmax values in Table 1). 3 Soil PCO2 estimates calculated for average surface air temperature during months of soil moisture utilization (TU values in Table 1). 4 N.R. ¼ not reasonable. See text for discussion.


22


FIG. 2.—Soil depth vs. d13C values for (A) profile 5 (Alturas soil series) and (B) profile 8 (Beckwourth soil series). Circles represent measured values of soil calcite d13C; error bars on the y-axis denote the thickness of the soil horizon from which the sample was taken (i.e., calcite d13C values are plotted in the middle of each soil horizon). The uncertainty (60.1ø) of the d13C value measurement plotted on the x-axis is less than the size of the diameter of the red circle. The thick gray vertical line is the predicted calcite d13C value if it formed only from CO2 derived from open-system oxidation of organic matter from that soil at temperatures equivalent to surface air during months of soil moisture utilization and deficit, whereas the thick black vertical line is the predicted calcite d13C value if formed from CO2 derived from open-system oxidation of organic matter from that soil at the maximum mean monthly surface air temperature. See text for discussion.

0.7ø, n ¼ 10), Mollisols (8.5 6 1.6ø, n ¼ 16), and Vertisols (11.3 6 2.6ø, n ¼ 15). Soil organic matter d13C values (n ¼ 14; Table 2) range from 23.4 to 27.7ø. With respect to soil taxonomic order, Entisols (profile 10) exhibit the most positive (24.1ø, n ¼ 1) organic matter d13C values, followed by Alfisols (24.3 6 0.4ø, n ¼ 2), Mollisols (24.3 6 0.6ø, n ¼ 4), Aridisols (25.1 6 0.7ø, n ¼ 2), and Vertisols (25.5 6 1.4ø, n ¼ 5). The difference between carbon isotope values of coexisting calcite and organic matter, D13Ccc-om, ranges from 10.6 to 25.1ø. With respect to soil taxonomic order, coexisting soil calcite and organic matter from the same horizon have the largest D13Ccc-om values in Entisols (24.5ø, n ¼ 1), followed by Aridisols (21.0 6 5.9ø, n ¼ 2), Alfisols (18.1 6 0.5ø, n ¼ 2), Mollisols (15.2 6 0.9ø, n ¼ 3), and Vertisols (13.1 6 1.9ø, n ¼ 5).

DISCUSSION Calcite Formation in Soils Breecker and others (2009; see also Cerling and Quade [1993]) make an explicit, but general, argument for the formation of soil calcite

that is based upon solution chemistry of the CaCO3–H2O–CO2 system (e.g., Drever 1973) and represented by the formula 4m3 Ca2þ K2 aCaCO3 ¼ ; ð1Þ PCO2 K1 Kcal KCO2 where aCaCO3 is the activity of calcite (assumed to be equivalent to concentration); m Ca2þ is the concentration of calcium ion in solution; PCO2 is the partial pressure of CO2 in soil gas; K1, K2, and Kcal are dissociation constants, respectively, for carbonic acid, bicarbonate, and calcite; and KCO2 is the hydration constant for aqueous CO2. When aCaCO3 ¼ 1 the system is saturated with respect to calcite, and crystallization may occur. All K values are temperature dependent such that higher T drives the system toward crystallization of calcite, as does an increase in concentration of Ca2þ or a decrease in the concentration of soil CO2. Note that the variables affecting the activity of calcite in Eq. 1 share no direct correlation with climate, but their variation can be strongly related to the hydrology and water balance in a soil profile. Therefore, an understanding of the hydrology of the soil profiles in this work is needed in order to constrain the timing and conditions of calcite crystallization. Soil hydrology is profile-specific, but the soils



considered in this work may nevertheless be encompassed by two end members of soil solution transport: (1) free drainage and (2) waterlogging. Soil solution transport in freely drained soils proceeds downward from the soil surface via gravity-driven piston flow (e.g., Allison et al. 1984). Assuming that runoff and fog condensate (dew) are negligible, the water balance of a vegetated soil may be reasonably well approximated by water balance formulas that consider (1) monthly mean precipitation (P; Table 3), (2) monthly mean potential evapotranspiration (ETp; Table 3), and (3) water-holding capacity (WHC; Table 3) of the soil material within the rooting zone (Thorthwaite 1948, Arkley 1963). Mean precipitation values provide estimates of water that are an input to the soil and are estimated from direct measurements taken at weather stations near the soil profile sites. Evapotranspiration values are based upon temperature and daylight hours and provide estimates of water that is lost from the soil through biological and physical processes (Thorthwaite 1948). Water-holding capacity values provide estimates of the maximum bioavailable water within the rooting zone of the soil; they are calculated from average WHCs of different soil material textures as determined by the United States Department of Agriculture (see: http://www.mo10.nrcs.usda. gov/references/guides/properties/awcrange.html) normalized to textural thickness of soil materials in the rooting zone of each soil profile. Water-holding capacity may be exceeded if precipitation is greater than ETp, resulting in surplus (Table 3) moisture and leaching of soil solution to depths beneath the rooting zone. When ETp exceeds precipitation, bioavailable water will be extracted from the inventory of stored moisture (Table 3) in the soil material. This is called ‘‘utilization,’’ and it results in progressively lower storage values. Utilization of moisture held within the rooting zone may proceed until WHC reaches a value of 0, which means that the soil water balance has moved into a deficit (Thorthwaite 1948). Periods of soil moisture utilization and deficit occur in all of the Californian soil profiles studied herein. Utilization begins as early as February (profile 6) and extends as late as June (profiles 1, 2, 5, and 11; Table 2). Deficit begins as early as March (profile 6) and extends as late as November (profile 6; Table 2). Surplus periods occur in four calcareous soils (profiles 1, 8, 9, and 14) beginning in December and extending through March (Table 2). According to the arguments surrounding Eq. 1 above, the soil chemical system will move toward calcite crystallization during periods of soil moisture utilization and/or deficit because the concentration of Ca2þ in soil solution increases, as water is lost from the rooting zone (e.g., Bouma et al. 1997, Bouma and Bryla 2000, Casals et al. 2000). Soil water balance periods of recharge and surplus lead to dissolution and removal, respectively, of Ca2þ from the rooting zone. In this respect, crystallization of calcite is possible during soil moisture utilization in early spring (March, April). Furthermore, calcite crystallization should cease during soil moisture deficit, and dissolution may be possible as soil water storage values rise above 0 in late autumn (November, December). Soil profiles that include periods of soil moisture surplus (profiles 1, 8, 9, and 14) pose an interesting problem for the development of pedogenic carbonate in these profiles. This reflects that surplus moisture leads to dissolution and leaching of Ca2þ from the soil. Leaching of Ca2þ from the soil will force the system away from crystallization of carbonate over time. Therefore, occurrence of calcite in these profiles is probably related to special circumstances, including the following: (A) calcite is out of equilibrium with the modern soil water balance and will be eventually dissolved and leached from the profile; (B) there is a sustained contribution of Ca2þ to the soil solution either from in situ weathering of Ca-bearing minerals (limestone, silicates) and/or delivery of new carbonate to the profile via eolian processes that is in excess of the ability of surplus moisture to leach from the profile (Machette 1985, McFadden and Tinsley 1985); or (C)

23

hydrology other than gravity-driven piston flow, such as waterlogging, that is conducive to calcite crystallization occurs in the profile. Profile 1 (Aco series) has substantial amounts of surplus precipitation and probable leaching of soil material due, in large part, to coarse-textured soil materials and low WHC (Table 2). However, profile 1 is located near the Transverse Ranges of southern California and experiences a high dust flux (10–25 g/m2/yr) that may include up to 4 wt % CaCO3 (Reheis 2003). Therefore, it is possible that the longterm eolian input of calcium to this soil has been greater than the current soil hydrology’s ability to dissolve and remove it, and the presence of calcite in profile 1 probably relates to special circumstance ‘‘B’’ mentioned above. Endosaturation occurs in profiles 8 (Beckwourth) and 9 (Bellevista). Endosaturation in these profiles means that the water table rises to ;75 cm beneath the soil surface, within the zone of calcite accumulation (Table 2), during the months of March, April, and May. It is likely that these profiles never become leached of Ca2þ because there is either a lack of cation export or the soil is seasonally replenished with cations via endosaturation. This reflects that most groundwater systems are saturated with respect to calcite (Clark and Fritz 1997), and, therefore, leaching of calcite from the soil by groundwater processes is not likely. However, seasonal dissolution and recrystallization in the groundwater-affected part of the profile is certainly possible. Therefore, the presence of calcite in profiles 8 and 9 probably relates to special circumstance ‘‘C’’ mentioned above. The presence of groundwater in these profiles makes it difficult to predict when and where the carbonate chemistry of the soil becomes saturated with respect to calcite, and the water balance calculations presented in Table 2 do not capture these dynamics. Nevertheless, soil moisture utilization and deficiency likely occur later in the year because of endosaturation, indicating that conditions may not be appropriate for calcite crystallization until June or July. Alternatively, carbonate may form upon the upper capillary fringe of the groundwater table as soon as the groundwater table begins to fall in April and/or May. The official soil series description for profile 14 (Montague) states that the soil is currently characterized by slow permeability but is nevertheless well drained. However, the official soil series description also states that this moderately deep Vertisol developed upon Tertiary volcanics and is also characterized by low amounts of extractable Fe (i.e., chelated Fe and Fe-oxides), which indicates a past history of poor drainage. It is not possible to determine the mechanism for the change in drainage conditions of profile 14. Nevertheless, it seems possible that calcite accumulated in the profile during its period of poor drainage (special circumstance ‘‘C’’ mentioned above) and may be now undergoing dissolution and leaching during months of surplus soil moisture. Therefore, the presence of calcite in profile 14 relates to circumstance ‘‘A,’’ mentioned above. If calcite in profile 14 is indeed a relict phase that is undergoing removal, the material may be undergoing dissolution and recrystallization, and therefore its light stable isotopic composition may provide meaningful information about modern soil processes, similar to the circumstances for soil calcite formation in profile 1 (Aco).

Calcite and Organic Matter d13C Values Soil organic matter d13C values range from 27.7 to 23.4ø. This relatively narrow range of organic matter d13C values is typical of bulk organic matter from vascular plants that utilize the C3 photosynthetic pathway (e.g., Cerling and Quade 1993, Arens et al. 2000). However, soil calcite d13C values range from 14.4 to 1.3ø, comprising ;3.7 times greater variability than that associated with coexisting soil organic matter (Fig. 3). This indicates that soil calcite d13C values are sensitive to sources of carbon beyond that of the carbon isotope composition of organic matter in the soil. The d13C value of soil calcite is related to the d13C value of CO2 and temperature in the soil profile at the time of crystallization (Romanek et


24


TABLE 3.—Water balance calculations for soil profiles considered in this study. Values for Temperature are degrees Celsius, whereas values for ETp, P, P-ETp, Storage, and Surplus are in centimeters. No. 1 Month Temp ETp2 P P-ETp Storage Surplus 2 Month Temp ETp2 P P-ETp Storage Surplus 3 Month Temp ETp2 P P-ETp Storage Surplus 4 Month Temp ETp2 P P-ETp Storage Surplus 5 Month Temp ETp2 P P-ETp Storage Surplus 6 Month Temp ETp2 P P-ETp Storage Surplus

WHC1

Name Aco Jan 12 2.7 5.4 2.7 3.1 0 Ager Jan 0 0.0 4.8 4.8 13.5 0.0 Alo Jan 11 3.5 11.3 7.9 13.5 0.0 Altamont Jan 8 1.8 7.6 5.8 11.0 0.0 Alturas Jan 2 0.0 4.0 4.0 10.0 0.0 Amedee Jan 12 2.6 3.6 1.0 1.8 0.0

Feb 13 3.0 5.5 2.5 5.5 0.1

Mar 14 4.2 4.2 0 5.5 0

Apr 16 5.8 1.9 3.9 1.6 0.0

May 19 8.5 0.6 7.9 0.0 0.0

Jun 22 11.2 0.2 11.0 0 0.0

Feb 1 0.4 3.9 3.4 16.9 0.0

Mar 4 1.9 4.0 2.1 19.0 0.0

Apr 7 3.9 3.1 0.8 18.2 0.0

May 11 6.6 3.4 3.3 14.9 0.0

Jun 15 9.7 2.4 7.3 7.6 0.0

Feb 12 3.7 8.4 4.8 16.5 1.8

Mar 12 4.7 8.1 3.4 16.5 3.4

Apr 13 5.4 3.6 1.8 14.7 0.0

May 13 6.6 1.3 5.3 9.4 0.0

Jun 15 7.5 0.5 7.0 2.4 0.0

Feb 10 2.6 6.3 3.7 14.8 0.0

Mar 12 4.0 5.4 1.5 16.2 0.0

Apr 14 5.4 2.5 2.9 13.3 0.0

May 17 8.1 1.1 7.0 6.4 0.0

Jun 20 10.5 0.3 10.2 0.0 0.0

Feb 1 0.3 3.2 3.0 13.0 0.0

Mar 4 1.8 3.4 1.6 14.6 0.0

Apr 7 3.8 2.7 1.1 13.5 0.0

May 11 6.7 3.3 3.5 10.0 0.0

Jun 15 9.5 2.2 7.2 2.8 0.0

Feb 13 2.9 2.4 0.5 1.3 0.0

Mar 14 4.2 1.9 2.3 0.0 0.0

Apr 16 5.9 1.1 4.9 0.0 0.0

May 19 8.8 2.0 6.8 0.0 0.0

Jun 22 11.7 2.1 9.6 0.0 0.0

5.5 cm Jul 25 14.9 0.1 14.8 0.0 0.0 27.8 cm Jul 20 12.4 0.6 11.8 0.0 0.0 16.5 cm Jul 15 8.1 0.2 8.0 0.0 0.0 25.4 cm3 Jul 22 12.6 0.1 12.6 0.0 0.0 16.2 cm Jul 19 12.2 0.8 11.4 0.0 0.0 4.6 cm Jul 25 15.6 0.8 14.7 0.0 0.0

Aug 25 14.5 0.3 14.2 0.0 0.0

Sep 24 11.2 0.6 10.6 0.0 0.0

Oct 20 7.4 0.8 6.8 0.0 0.0

Nov 15 4.1 2.4 1.7 0.0 0.0

Dec 12 2.7 3.1 0.4 0.4 1.0

Aug 18 11.0 0.9 10.1 0.0 0.0

Sep 15 7.5 1.5 6.0 0.0 0.0

Oct 9 4.3 3.0 1.3 0.0 0.0

Nov 3 1.2 4.5 3.3 3.3 0.0

Dec 1 0.0 5.4 5.4 8.7 0.0

Aug 16 8.0 0.2 7.8 0.0 0.0

Sep 17 7.5 0.6 6.8 0.0 0.0

Oct 16 6.5 2.2 4.3 0.0 0.0

Nov 13 4.5 5.2 0.7 0.7 0.0

Dec 11 3.4 8.4 5.0 5.7 0.0

Aug 22 11.6 0.1 11.5 0.0 0.0

Sep 21 9.5 0.6 9.0 0.0 0.0

Oct 17 6.5 1.7 4.9 0.0 0.0

Nov 12 3.3 3.9 0.6 0.6 0.0

Dec 8 1.9 6.5 4.6 5.2 0.0

Aug 18 10.7 0.8 9.9 0.0 0.0

Sep 14 7.3 1.3 6.1 0.0 0.0

Oct 9 4.1 2.4 1.7 0.0 0.0

Nov 3 1.2 3.5 2.3 2.3 0.0

Dec 1 0.0 3.7 3.7 6.0 0.0

Aug 25 14.7 0.8 13.9 0.0 0.0

Sep 24 11.3 0.7 10.6 0.0 0.0

Oct 20 7.3 1.8 5.6 0.0 0.0

Nov 15 3.9 2.5 1.4 0.0 0.0

Dec 12 2.6 3.4 0.9 0.9 0.0



25

TABLE 3.—Continued. No. 7 Month Temp ETp2 P P-ETp Storage Surplus 8 Month Temp ETp2 P P-ETp Storage Surplus 9 Month Temp ETp2 P P-ETp Storage Surplus 10 Month Temp ETp2 P P-ETp Storage Surplus 11 Month Temp ETp2 P P-ETp Storage Surplus 12 Month Temp ETp2 P P-ETp Storage Surplus 13 Month Temp ETp2 P P-ETp Storage Surplus

WHC1

Name Anacapa Jan 13 3.7 8.0 4.3 6.8 0.0 Beckwourth Jan 1 0.0 9.4 9.4 7.3 2.1 Bellevista Jan 1 0.0 5.7 5.7 7.6 5.6 Beveridge Jan 5 0.7 2.6 1.9 3.5 0.0 Bidwell Jan 1 0.0 4.3 4.3 9.5 0.0 Cropley Jan 10 2.5 6.5 4.0 7.2 0.0 Centerville Jan 8.3 1.3 5.3 3.9 6.7 0.0

Feb 13 3.7 8.5 4.7 11.5 0.0

Mar 13 4.8 6.2 1.4 13.0 0.0

Apr 14 5.7 2.7 3.0 10.0 0.0

May 16 7.1 0.4 6.7 3.3 0.0

Jun 17 8.0 0.2 7.9 0.0 0.0

Feb 0 0.0 5.4 5.4 7.3 5.4

Mar 3 1.4 5.0 3.6 7.3 3.6

Apr 6 3.2 3.0 0.2 7.1 0.0

May 11 6.8 3.6 3.2 3.9 0.0

Jun 14 9.0 1.6 7.4 0.0 0.0

Feb 0 0.0 4.5 4.4 7.6 4.4

Mar 3 1.4 3.8 2.4 7.6 2.4

Apr 6 3.2 2.1 1.1 6.5 0.0

May 11 6.8 2.3 4.5 2.1 0.0

Jun 14 9.0 1.7 7.4 0.0 0.0

Feb 7 1.3 3.3 2.0 5.6 0.0

Mar 11 3.0 1.8 1.3 4.3 0.0

Apr 14 5.3 0.7 4.6 0.0 0.0

May 19 9.6 0.4 9.1 0.0 0.0

Jun 24 13.9 0.3 13.7 0.0 0.0

Feb 1 0.4 3.4 3.1 12.5 0.0

Mar 4 1.7 3.4 1.7 14.3 0.0

Apr 8 3.9 2.6 1.3 13.0 0.0

May 12 7.1 2.6 4.4 8.6 0.0

Jun 17 10.2 1.9 8.3 0.2 0.0

Feb 11 3.0 7.1 4.2 11.3 0.0

Mar 13 4.5 4.0 0.6 10.7 0.0

Apr 14 5.5 3.1 2.4 8.4 0.0

May 17 8.2 0.7 7.5 0.9 0.0

Jun 19 10.1 0.2 10.0 0.0 0.0

Feb 11.0 2.2 4.8 2.6 9.3 0.0

Mar 13.5 3.9 4.7 0.8 10.2 0.0

Apr 16.5 6.1 2.7 3.3 6.9 0.0

May 20.3 9.8 1.1 8.7 0.0 0.0

Jun 24.2 13.8 0.2 13.6 0.0 0.0

17.0 cm Jul 19 9.4 0.1 9.4 0.0 0.0 7.3 cm4 Jul 18 11.7 0.6 11.1 0.0 0.0 7.6 cm4 Jul 18 11.7 0.7 11.0 0.0 0.0 5.8 cm Jul 27 17.5 0.3 17.2 0.0 0.0 27.8 cm Jul 22 13.9 0.8 13.2 0.0 0.0 35.9 Jul 21 11.4 0.0 11.4 0.0 0.0 18.8 Jul 27.5 17.9 0.1 17.8 0.0 0.0

Aug 19 9.2 0.1 9.1 0.0 0.0

Sep 19 8.0 0.5 7.5 0.0 0.0

Oct 17 6.7 1.0 5.7 0.0 0.0

Nov 15 4.9 3.8 1.0 0.0 0.0

Dec 13 3.8 6.3 2.5 2.5 0.0

Aug 17 10.4 0.5 9.9 0.0 0.0

Sep 14 7.7 1.1 6.5 0.0 0.0

Oct 8 3.9 2.4 1.4 0.0 0.0

Nov 4 1.9 5.7 3.8 3.8 0.0

Dec 3 0.0 7.8 7.8 7.3 0.5

Aug 17 10.4 0.9 9.5 0.0 0.0

Sep 14 7.7 1.2 6.5 0.0 0.0

Oct 8 3.9 2.3 1.5 0.0 0.0

Nov 4 1.9 3.9 1.9 1.9 0.0

Dec 3 0.0 5.6 5.6 7.5 0.0

Aug 26 15.3 0.3 15.0 0.0 0.0

Sep 22 10.2 0.5 9.7 0.0 0.0

Oct 16 5.7 0.6 5.1 0.0 0.0

Nov 10 2.1 1.4 0.7 0.0 0.0

Dec 5 0.7 2.3 1.6 1.6 0.0

Aug 21 12.2 0.7 11.5 0.0 0.0

Sep 16 7.9 1.2 6.8 0.0 0.0

Oct 10 4.3 2.5 1.9 0.0 0.0

Nov 4 1.2 4.1 2.9 2.9 0.0

Dec 1 0.0 4.2 4.2 7.1 0.0

Aug 21 10.8 0.0 10.8 0.0 0.0

Sep 20 9.1 0.2 9.0 0.0 0.0

Oct 17 6.7 1.9 4.8 0.0 0.0

Nov 13 3.8 2.8 1.0 0.0 0.0

Dec 10 2.4 5.6 3.2 3.2 0.0

Aug 26.6 15.7 0.1 15.7 0.0 0.0

Sep 23.7 11.2 0.5 10.7 0.0 0.0

Oct 18.8 6.8 1.4 5.4 0.0 0.0

Nov 12.7 2.9 2.7 0.2 0.0 0.0

Dec 8.5 1.3 4.1 2.8 12.8 0.0


26


TABLE 3.—Continued. No. 14 Month Temp ETp2 P P-ETp Storage Surplus

WHC1

Name Montague Jan 1 0.3 8.1 7.9 12.2 7.9

Feb 4 1.0 5.5 4.5 12.2 4.5

Mar 6 2.4 4.1 1.7 12.2 1.7

Apr 9 4.3 2.5 1.8 10.4 0.0

May 14 7.6 2.6 5.0 5.4 0.0

Jun 18 10.6 2.0 8.5 0.0 0.0

12.9 Jul 22 13.9 1.0 12.8 0.0 0.0

Aug 21 12.4 1.1 11.3 0.0 0.0

Sep 17 8.4 1.4 6.9 0.0 0.0

Oct 12 4.6 3.1 1.5 0.0 0.0

Nov 5 1.5 6.3 4.8 4.8 0.0

Dec 2 0.4 8.9 8.6 12.2 1.2

WHC ¼ water-holding capacity; Temp ¼ temperature; ETp ¼ evapotranspiration; P ¼ precipitation. 1 WHC is calculated based upon the sum of estimated WHCs of soil material from the soil surface to the base of the depth of rooting, as described in each profile’s official soil series description. Soil material WHC estimates are based upon soil texture and are taken from USDA, as follows (in mm H2O/cm soil): coarse sand ¼ 0.4; fine sand ¼ 0.6; loamy sand ¼ 0.8; sandy loam ¼ 1.2; loam ¼ 1.5; sandy clay loam ¼ 1.5; clay loam ¼ 1.8; clay ¼ 1.5; and self-mulching clay ¼ 2.0. 2 Evapotranspiration (ETp) estimates follow calculations presented in Thornthwaite (1948). 3 Redoximorphic Fe-segregations indicate long-term saturation of the profile, rendering the accuracy of these water balance calculations suspect. 4 Endosaturation occurs in these profiles during the months of March, April, and May, rendering the accuracy of these water balance calculations suspect. al. 1992): e13 ccCO2

¼ 11:98 0:12 3 T8C ;

ð2Þ

where e13 ccCO2 is an enrichment factor and corresponds to the per-mil 13 difference between calcite and coexisting CO2 (e13 ccCO2 ¼ [(d Ccc þ 13 1000)/(d CO2 þ 1000) 1]). We consider two possible soil temperatures for calcite crystallization: (1) mean surface air temperature during months of soil moisture utilization (TU; Table 1) and (2) maximum monthly mean surface air temperature (Tmax; Table 1). Soil CO2 may be characterized by mixing of one, two, or three components that originate from different circumstances and which can result in significantly different d13C values of soil CO2 and pedogenic calcite (Cerling 1984; Hsieh and Yapp 1999; Yapp 2001, 2002; Tabor and Yapp 2005a, 2005c; Feng and Yapp 2009; Sheldon and Tabor 2009). The three CO2 components are derived from (1) oxidation of in situ organic matter, (2) tropospheric CO2, and (3) dissolution and/or incorporation of preexisting carbon-bearing minerals such as calcite in the soil (Rabenhorst et al. 1984, West et al. 1988, Hseih and Yapp 1999, Kraimer and Monger 2009, Sheldon and Tabor 2009, Sheldon and Tabor this volume). The following discussion considers these different CO2 mixing scenarios and their applicability to calcite from the modern soil profiles studied herein. One Soil CO2 Component: Soils characterized by one component of CO2 form under a chemically closed or semiclosed system (e.g., Whelan and Roberts 1973, Mintz et al. 2011) that arises in settings in which gaseous diffusion between the soil and the troposphere is limited, usually as a result of poorly drained, waterlogged conditions within the soil profile. Diffusion from the troposphere to the wet soil is limited, and aqueous O2 in the soil is quickly consumed by oxidation of organic matter (Feng and Hsieh 1998). Under these saturated conditions diffusive carbon isotope enrichment of CO2 derived from oxidizing organic matter can be small (,1ø; O’Leary 1984), and, therefore, the per-mil difference between organic matter and calcite that forms under these conditions will be primarily related to the temperature of crystallization according to Eq. 2 above (i.e., om ﬃ 13 13 cc; e13 ccCO2 ﬃ eomCO2 ). In this respect, the smallest D Ccc-om values, 13 and among the most negative calcite d C values, are expected in soils characterized by one component of CO2 (Fig. 3). Soils are most likely to be characterized by one component of CO2

during flooding and/or waterlogging. Therefore, soils that preserve morphological evidence of redoximorphic conditions or are observed to experience seasonal groundwater flooding are most likely to experience conditions that could lead to calcite formation in the presence of one component of soil CO2. Profile 4 (Altamont) is the only soil with redoximorphic features, and it contains the most negative calcite d13C as well as the smallest D13Ccc-om values (Table 2; Fig. 3). However, other groundwater-affected soils (profiles 8, 9, and 14) show a broad range of calcite d13C and D13Ccc-om that include positive and large values, respectively, that are mostly inconsistent with calcite crystallization in the presence of one component of CO2 derived solely from oxidation of soil organic matter. This indicates that in spite of seasonal or episodic groundwater flooding, the majority of calcite in these soils forms at times after endosaturation, when the soil profile is characterized by mixing of at least two components of soil CO2. Yet profiles 2 (Ager), 12 (Cropley), and 13 (Centerville), which have no observable evidence of redoximorphy or endosaturation, include calcite d13C and D13Ccc-om values that indicate formation in the presence of one component of soil CO2 (Table 2; Fig. 3). It is likely important that these particular soils with very negative calcite d13C values and small D13Ccc-om values and no morphological or observable indication of anoxia are all Vertisols. This may reflect that Vertisols are characterized by a high concentration of fine clay as well as episodic saturation of the soil profile, which includes volume expansion of the soil and constriction of diffusion pathways from the soil to the troposphere. This episodic saturation of the profile may lead to conditions that are permissive of calcite formation in the presence of CO2 (1) largely dominated by a single CO2 component derived from oxidation of organic matter that (2) undergoes minimal carbon isotope enrichment related to diffusion (e.g., Whelan and Roberts 1973, Mintz et al. 2011). Note, however, that profiles 2 and 13 also include some d13C and D13Ccc-om values that correspond to formation that is consistent with mixing of two components of CO2 between tropospheric CO2 and CO2 derived from oxidation of soil organic matter, indicating that calcite also forms at times when the soil is not diffusion-limited by the effects of soil saturation (e.g., Mintz et al. 2011). Two Soil CO2 Components: Soils characterized by two-component soil CO2 mixing include (1) CO2 from the open atmosphere (tropospheric CO2) and CO2 from in situ oxidation of biological



27

FIG. 3.—Plot of D13Ccc-om vs. calcite d13C values from soil data presented in Table 2. Red dots correspond to soils with morphological or observational evidence of groundwater saturation within the zone of calcite accumulation. Black diamonds indicate data from profiles that are in the soil order Vertisols. Blue triangles indicate data from soil profiles with limestone parent material. Red squares indicate data from profiles that are in the soil order Aridisols. Open circles indicate data points that may correspond to calcite crystallization under closed-system conditions (i.e., they do not provide estimates of soil PCO2). Note that a data point may be represented by more than one symbol. For example, a black diamond with an open circle around it indicates a data point from a soil profile that is in the soil order Vertisols and for which calcite crystallization may have occurred under closed-system conditions. See text for further discussion.

carbon in the soil (also called ‘‘soil-respired CO2"; Cerling 1984, Sheldon and Tabor this volume). Steady-state solutions to the onedimensional Fickian diffusion equation yield reasonable representations of concentration and d13C values of soil CO2 with depth in modern soils characterized by mixing of two sources of CO2 (Cerling 1984). In these soils, PCO2 and the d13C value of Earth’s atmosphere is the upper boundary condition, and there is a depth-dependent CO2 production term that describes the oxidation of organic carbon in the soil. Above a characteristic CO2 production depth within the soil, both the concentration and d13C of soil CO2 will progressively approach the values of the Earth’s atmosphere. Contemporary tropospheric CO2 is approximately 380 ppmV, with a d13C value of 8ø (Schlessinger 1996), but for most of the past 600,000 years atmospheric PCO2 did not exceed 280 ppmV and maintained a d13C value of approximately 6.5ø (Petit et al. 1999). Note that soil CO2 derived from in situ oxidation of organic matter during two-component CO2 mixing will be about 4.4ø more positive than the d13C value of the organic matter from which it is derived as a result of diffusive enrichment (Jost 1960, Solomon and Cerling 1987). In this respect, it appears that pedogenic calcite d13C values in soils characterized by mixing of two soil CO2 components record a temperature-dependent (Eq. 2) mass balance between CO2 with relatively negative d13C values contributed from soil-respired CO2 and CO2 contributed from the troposphere with relatively positive d13C values.

A complete mixing profile of the two CO2 components is seldom preserved by soil calcite d13C values as a result of leaching of carbonate to depths beneath the zone characterized by large changes in soil CO2 d13C values over short vertical distances (e.g., Jenny 1941, Quade et al. 1989, Retallack 1994, Royer 1999). Nevertheless, profile 5 (Alturas; Fig. 2A) appears to preserve a calcite d13C depth trend that closely matches patterns that are calculated for Fickian-type, steadystate, soil diffusion models of CO2 (Cerling 1984, Quade et al. 1989, Cerling and Quade 1993, Breecker et al. 2009, Sheldon and Tabor 2009). Calcite d13C in profile 5 approaches a uniform value of approximately 9ø at ;50 cm below the soil surface (Fig. 2A), indicating this is well below the zone characterized by large changes in soil CO2 d13C values over short vertical distance (as well as the characteristic production depth) of this soil during calcite crystallization. The depth trend in calcite d13C values (Fig. 2A) is consistent with previously observed carbon isotope depth trends defined by calcite d13C values in other soils that exhibit invariant and most negative calcite d13C values beginning 30 to 50 cm beneath the soil surface (Quade et al. 1989, Cerling and Quade 1993, Breecker et al. 2009). These studies provide the principle modern evidence that has led to a recommendation for sampling of paleosol calcite d13C values at depths beneath 30 to 50 cm to ensure that calcite d13C values do not reflect crystallization in the zone characterized by large changes in soil CO2 d13C values over short vertical distance. However, the two separate sample sets from profile 9 (Bellevista) record more negative calcite


28


d13C values to 84 cm in depth, indicating that the zone characterized by large changes in soil CO2 d13C values over short vertical distance for this soil extends to at least 84 cm beneath the soil surface when the profile is not flooded by endosaturation. Bellevista soils also include petrocalcic horizons, or layers lithified by calcite and silica cements, beginning at 51 cm, and, therefore, steady-state diffusion could be different in this soil as a result of large differences in the permeability and porosity of soil media through the profile. Nevertheless, several other profiles (1, 12, and 13) consistent with Fickian-type, opensystem, two-component CO2 mixing preserve more negative calcite d13C values at shallower positions in the profile. This indicates that perhaps more than one steady-state CO2 mixing profile has been present through the history of calcite crystallization in these profiles. These results are important, as they indicate (1) that soil respiration during periods of soil calcite crystallization may occur at much greater depths than previously thought (Breecker et al. 2009), (2) that soil depth does not conform to a single soil CO2 mixing profile, and (3) that it is likely that many paleosol calcite carbon isotope data sets have been collected from depths well above the depths that are appropriate for estimates of atmospheric PCO2 (e.g., Ekart et al. 1999). Among the 46 paired calcite–organic matter d13C values presented herein, 40 are permissive of formation in a soil characterized by mixing of CO2 derived from tropospheric CO2 and CO2 derived from in situ oxidation of soil organic matter (Fig. 3). As will be discussed in a subsequent section, these 40 paired calcite–organic matter d13C values provide estimates of the concentration of CO2 in the soil at the time of calcite crystallization. Three Soil CO2 Components: On the basis of geological and chemical arguments, several studies have concluded that threecomponent CO2 mixing relations can exist in various soils (Hsieh and Yapp 1999; Yapp 2001, 2002; Tabor et al. 2004; Tabor and Yapp 2005b, 2005c). As in two-component CO2 mixing, three-component mixing assumes soil CO2 is represented by sources contributed from oxidation of organic matter (i.e., soil-respired CO2) and the global troposphere (e.g., Sheldon and Tabor [this volume]) but includes as well a third component of CO2 contributed from dissolution of preexisting carbonate in the soil profile. Although carbonate may have very negative d13C values in rare instances (e.g., one-component CO2 mixing; Yemane et al. 1989, Coney et al. 2007, Tabor et al. 2007), CO2 that is produced from dissolution of carbonate generally is enriched in 13 C relative to atmospheric CO2 or CO2 derived from oxidation of organic matter (e.g., Hoefs 1997, Criss 1999, Sharp 2007). As a result of these mixing relationships, pedogenic minerals that form in threecomponent soil CO2 mixing systems are characterized (and for the most part recognized) by d13C values that are more positive than those associated with one- or two-component soil CO2 mixing. Another possible scenario by which a third component may affect the measured d13C values of soil calcite is physical entrainment of detrital or parent-material carbonate grains within pedogenically formed calcite (e.g., Rabenhorst et al. 1984, West et al. 1988, Kraimer and Monger 2009). Marine limestone has relatively high calcite d13C values, ranging from approximately 2 to 6ø (Veizer et al. 1999), which are substantially greater than the calcite d13C values encountered in most soils characterized by mixing of two soil CO2 components (e.g., Salomons et al. 1978; Schlessinger 1982, 1985; Cerling 1984; Rabenhorst et al. 1984; Amundson et al. 1989, 1998; Quade et al. 1989; Cerling and Quade 1993; Wang et al. 1996; Monger et al. 1998; Nordt et al. 1998; Kraimer and Monger 2009). In this respect, incorporation of a third carbon component, whether it be as CO2 or physical entrainment of preexisting carbonate, is expected to result in generally more positive d13C values in soil calcite. As mentioned, profile 1 (Aco) receives a substantial amount of detrital carbonate through eolian deposition of silt-sized grains. However, this soil profile also experiences months of soil moisture surplus. It is impossible to determine with existing data whether the

following is correct, but we suppose that dissolution of eolian and pedogenic carbonate during months of soil moisture surplus permits equilibration of detrital-sourced carbonate with a two-component CO2 dominated by soil-respired CO2 and tropospheric CO2 during months of soil moisture utilization and deficit. To be clear, we assume this scenario applies to all other soils that receive small doses of finegrained detrital carbonate. Profile 10 (Beveridge) is developed upon limestone parent materials (Table 1) and therefore may include some detrital carbonate that is entrained within the Bk horizons of these soil profiles. Profile 10 has among the most positive calcite d13C values measured in this study (Fig. 3) and contains increasing concentrations of carbonate downward through the profile (Table 2), as may be expected for a soil weathering front upon limestone (Rabenhorst et al. 1991). Reflected light microscopic inspections of these samples provide no obvious indications (e.g., invertebrate fossils) of an entrained population of detrital carbonate grains. Petrographic microscopy is a superior means by which to evaluate the presence of detrital grains (West et al. 1988, Kraimer and Monger 2009), but none of these samples have been inspected petrographically as a result of concerns related to exhausting the small amount of sample available for geochemical analysis. Because we are not able absolutely to rule it out at this time, we consider samples from profile 10 (Beveridge) to incorporate a detrital source of carbonate, and we assume as well that measured calcite d13C values from this profile may be more positive than those from a purely pedogenic calcite.

Estimates of Soil PCO2 and Soil-Respired CO2 The following equation (after Yapp and Poths [1996]) relates measured soil calcite and organic matter d13C values to soil PCO2: CSðccÞ ¼ CAðccÞ 3

d13 CAðccÞ d13 COðccÞ d13 CCC d13 COðccÞ

ð4Þ

:

The subscript (cc) indicates calcite. d13CCC is the measured d13C value of pedogenic calcite (d13CCC in Table 2). The d13C values subscripted with ‘‘A’’ and ‘‘O’’ are predicted d13C values of calcite if formed solely in equilibrium with CO2 derived from the atmosphere and from oxidation of soil organic matter (i.e., respired CO2), respectively. CA refers to the concentration of CO2 in the soil if the only contribution to soil CO2 were from the atmosphere. CS is the actual concentration of CO2 that was present in the soil at the time of calcite crystallization; it is the sum of atmospheric and soil-respired CO2 (Yapp and Poths 1996). Both CA and CS are expressed in units of ppmV. d13CO(cc) values may be estimated from d13C values of coexisting organic matter within the soil profile (d13COM in Table 2). d13CO(cc) values include an ;4.4ø diffusive enrichment associated with oxidation of soil organic matter (Cerling et al. 1991), as well as a temperature-sensitive quantity of 13C enrichment associated with equilibrium isotope exchange between CO2 and calcite (Romanek et al. 1992): 13 d13 COðccÞ ¼ ðe13 ccCO2 þ 1Þ 3½ak 3 d COM þ ðð1000 3ðak 1ÞÞ þ 1000 1000;

where e13 ccCO2

ð5Þ 13

is equivalent to the quantity defined in Eq. 2 and d COM represents the d13C values of organic matter reported in Table 2. The variable ak is the ratio of the soil diffusion coefficients of the CO2 molecules with mass numbers 44 and 45 (44D/45D). The value of ak used here is 1.0044 (Cerling et al. 1991). The CA values at the time of calcite crystallization among these soils are not known precisely. Nevertheless, these soil profiles occupy geomorphic positions interpreted to be no older than late Pleistocene (125,000 years old). Atmospheric PCO2 has remained low, between



280 and 180 ppmV, for all of that time up to the Industrial Revolution (;1800 C.E.; Etheridge et al. 1996, Petit et al. 1999), and thus the range of CA values over the entire interval of soil formation is reasonably well constrained. Considering that the 14C ages of pedogenic calcite indicate that even very dry and old geomorphic landscapes (.106 years old) undergo complete recrystallization in ,20,000 years (Deutz et al. 2002), it is reasonable to restrict CA values to the higher end (280 ppmV) of PCO2 values that have occurred in the past 125,000 years. Note that if CA values were in fact higher or lower than 280 ppmV the actual value of CS would be higher and lower, respectively, than the calculated value. For example, soil PCO2 estimates of 1000 ppmV based upon an assumed CA of 280 ppmV overestimate the actual value (600 ppmV) if it had formed in soil when CA was 180 ppmV. Based upon these assumed values for calcite crystallization temperature, CA, d13CA(cc) and d13CO(cc), estimates of soil PCO2 (CS in Eq. 3) range from 300 to 30,360 ppmV (Table 2). As discussed earlier, high soil PCO2 is not conducive to calcite crystallization (Eq. 1). Therefore, very high soil PCO2 estimates in this data set seem implausible. Yet, how high is too high for soil PCO2 to permit soil calcite crystallization is unknown. High soil PCO2 estimates may be in part related to arithmetic artifacts of the model and the analytical precision of measurements of d13CCC and d13COM. Calcite and organic matter d13C values, D13Cccom values specifically, are inversely related to estimates of soil PCO2. Furthermore, the precision of d13CCC and d13COM measurements is no better than 60.1ø. This precision corresponds to a small range of soil PCO2 estimates at large D13Ccc-om values, whereas the same precision corresponds to very large ranges of soil PCO2 estimates at small D13Ccc-om values. For example, profile 10 (Beveridge) is characterized by the largest D13Ccc-om of 25.5ø; an uncertainty of 60.1ø in this value corresponds to a range of soil PCO2 estimates of ,1 ppm. However, profile 11 (Bidwell) has the smallest D13Ccc-om value, 14.4ø, that is also permissive of formation in a soil characterized by open-system, two-component CO2 mixing; an uncertainty of 60.1ø provides a large enough range to move this soil from an open system with soil PCO2 of 30,000 ppmV to one that is best modeled by a closed system, one CO2 component (i.e., a solution for PCO2 that results in negative values). The extreme sensitivity of the model to small changes in the measured d13C values of calcite and organic matter at relatively low D13Ccc-om values indicates that soil PCO2 estimates above ;20,000 ppmV are not resolvable in a useful way. Therefore, soil PCO2 estimates .20,000 ppmV should be regarded in this way, and not, for example, as 30,360 ppmV (e.g., profile 11; Table 2). In addition, the temperature used to calculate soil PCO2 at the time of calcite crystallization can have a very large effect on the resulting range of soil PCO2 estimates. As mentioned, the temperatures of calcite crystallization used to estimate soil CO2 in these profiles are (1) average of mean monthly surface air temperatures during months of soil moisture utilization and (2) maximum mean monthly surface air temperature. The difference between these two temperatures varies, but (2) is always greater than (1) by 3 to 158 C (Table 1). As a result of the temperature-dependent behavior of carbon isotope fractionation, higher temperatures of crystallization correspond to lower estimates of soil PCO2 (CS; Table 2). Recent D47 measurements of modern soil calcites indicate that crystallization occurs during the dry seasons at temperatures near or slightly in excess of the dry season surface air temperature (Passey et al. 2010). Therefore, temperatures in categories (1) and (2) above provide a range of soil PCO2 estimates that likely encompass the actual value of soil PCO2 at the time of calcite crystallization. Profile 8 (Beckwourth), which includes morphological evidence for waterlogging and D13Ccc-om values that indicate calcite crystallization in a one-component CO2 system, also contains several D13Ccc-om values permissive of two-component soil CO2 mixing (Fig. 3) and

29

corresponds to soil PCO2 estimates ranging from 770 to .20,000 ppmV; mean profile soil PCO2 estimates range from 3410 to 11,550 ppmV depending upon the temperature of soil calcite crystallization (Table 2). Other soil profiles with evidence of groundwater influence have D13Ccc-om values that correspond to substantially lower soil PCO2 than does profile 8 (Table 2; Fig. 3). Profile 9 (Bellevista) soil PCO2 estimates range from 860 ppmV at shallow depths to 2790 ppmV at maximum depths that contain soil calcite (Table 2). Profile 14 ranges from 2260 to 7350 ppmV; mean profile values range from 2410 to 6250 ppmV depending upon the temperature of soil calcite crystallization. Soil PCO2 estimates from Vertisols (Tables 1, 2) with D13Ccc-om values that are permissive of open-system, Fickian two-component soil CO2 mixing (Fig. 3) correspond to substantially higher CS values than do other soil orders in this data set. Soil PCO2 estimates for profile 2 (Ager) range from 460 to 3760 ppmV; mean profile soil PCO2 estimates range from 1630 to 2800 ppmV depending upon the temperature of soil calcite crystallization (Table 2). Profile 13 (Centerville) soil PCO2 estimates range from 1410 to 5290 ppmV; mean profile soil PCO2 estimates range from 1970 to 3310 ppmV depending upon the temperature of soil calcite crystallization (Table 2). As mentioned previously, profile 14 ranges from 2260 to 7350 ppmV; mean profile values range from 2410 to 6250 ppmV depending upon the temperature of soil calcite crystallization. The generally small D13Ccc-om values and higher soil PCO2 estimates from the soil with redoximorphic morphology (profile 8; Beckwourth) and Vertisols that also have small D13Ccc-om values indicative of closed system formation of calcite (profiles 2, 4, and 12) may actually indicate high soil PCO2 at the time of calcite crystallization (Table 2). However, it is possible also that calcite formed by (1) crystallization in closed system, one CO2 component conditions and (2) crystallization in open-system, Fickian twocomponent CO2 mixing conditions that are co-mingled within bulk soil material that result in mixtures with d13C values that correspond to artificially high soil PCO2 estimates. This kind of physical mixing of different generations of soil calcite has been observed in other modern Vertisols (Mintz et al. 2011), and it does result in soil PCO2 estimates that are too high for the interval of open-system, Fickian twocomponent CO2 mixing. The results presented here, and those previously published by Mintz and others (2011), indicate that redoximorphic soils, and perhaps most Vertisols, have a complicated soil hydrology that may lead to multiple and rapidly changing CO2 diffusion states that make their utility for estimating soil and atmospheric PCO2 from modern and ancient Vertisol D13Ccc-om data especially challenging. However, profile 14 (Montague) is a Vertisol, and although D13Ccc-om values correspond to relatively high soil PCO2 (2390 to 13,030 ppmV), it preserves no indication of closed-system calcite formation. Soil profiles with no morphological indication of waterlogging and no geochemical indications of closed-system soil CO2 mixing (profiles 1, 5, 6, 7, 10, 11, and 14) have D13Ccc-om values (Fig. 3; Table 2) and temperatures that provide a range of soil PCO2 estimates from 300 to .20,000 ppmV. Most D13Ccc-om values provide soil PCO2 estimates that correspond to relatively low soil PCO2 values. In particular, profiles 1, 6, and 10 preserve D13Ccc-om values that correspond to soil PCO2 estimates ranging from 300 to 1540 ppmV depending upon the temperature of calcite crystallization. Profiles 1 (Aco) and 6 (Amedee) are Aridisols and are characterized by relatively low productivity due to high demands of soil moisture by soil biota that are limited by low annual precipitation and low WHC of soil material in the rooting zone. In this respect, low soil PCO2 estimates from Aridisols are somewhat expected and are consistent with measurements of soil PCO2 at the time of calcite crystallization in other soil profiles developed in arid climates and with low WHC (Breecker et al. 2009). Profile 10 (Beveridge) preserves D13Ccc-om values that correspond to


30


soil PCO2 estimates ranging from 300 to 520 ppmV, the lowest soil PCO2 estimates of any soil profile considered herein. Profile 10 (Beveridge) is an Entisol characterized by a thin solum and is developed upon limestone parent materials in a cold climate. Soils developed upon limestone are commonly nutrient-limited and therefore typically have low-productivity floras associated with them (e.g., Buol et al. 2003). Furthermore, cool and cold climates limit the potential productivity of soil floras and the resulting soil-respired CO2 that is a product of this production (e.g., Lieth 1975). Therefore, D13Ccc-om values that correspond to relatively low soil PCO2 at the time of calcite crystallization in profile 10 may be reasonable. Yet the large D13Ccc-om values in profile 10 may also be due, at least in part, to inheritance or physical inclusion of calcite from its Late Paleozoic limestone parent material. This possibility reflects the facts that Late Paleozoic limestone typically contains calcite d13C values that are similar to, or greater than, those measured in profile 10 (Table 2; Veizer et al. 1999) and that physical entrainment of limestone parent material in soil calcite has been shown to result in d13C values that are too positive for the conditions at the time of pedogenic calcite crystallization (Kraimer and Monger 2009). It is impossible to evaluate from the existing data whether inheritance of Late Paleozoic limestone in profile 10 has resulted in artificially large calcite d13C values. Nevertheless, we consider the very low soil PCO2 estimates from profile 10 to be reasonable because they are consistent with the ecology and climate of that soil. The highest soil PCO2 estimates occur in profile 11 (Bidwell), with a range of 4220 to .20,000 ppmV (Table 2) depending upon the temperature of calcite crystallization. The most convincing estimates of soil PCO2 in this data set come from profiles with D13Ccc-om values that appear to conform to a single steady-state, Fickian-type, twocomponent CO2 mixing gradient that includes a vertical trend, defined by multiple calcite d13C measurements, of most negative calcite d13C values at depth and progressively more positive d13C values toward the profile top. Profile 5 (Alturas) records D13Ccc-om values ranging from 15.3 to 19.6ø, corresponding to soil PCO2 (CS) values ranging from 680 to 4910 ppmV. However, soil PCO2 estimates range from 2060 to 4910 ppmV, with an average of 2040 6 320 and 3300 6 860 ppmV depending upon the temperature of calcite crystallization, at depths .50 cm beneath the soil surface (Table 2). Note that these soil PCO2 estimates (CS) correspond to the sum of CO2 contributed from the atmosphere (CA ¼ 280 ppmV) and soil-respired CO2 [S(z); Cerling 1984] that is produced by in situ oxidation of soil organic matter. Therefore, soil-respired CO2 in this profile ranged from 1760 6 320 to 3020 6 860 ppmV depending upon the temperature of calcite crystallization. The resulting estimates of CS (Table 2) from profiles 5 (Alturas), 7 (Anacapa), 11 (Bidwell), and 14 (Montague) indicate that it is possible for soil-respired CO2 to be substantially higher at the time of calcite crystallization than has been observed in previous studies of calcite-bearing soil profiles characterized by two-component soil CO2 mixing. For example, Breecker and others (2009) focused upon four different soil profiles from a semiarid (210–375 mm/yr precipitation) region with aridic to ustic soil moisture regimes and a mesic soil temperature regime in central New Mexico. That work deduced that soil calcite tends to form seasonally, likely in the late spring, and at relatively low soil CO2 concentrations ranging from ,1000 to 2000 ppmV. The parent material for all four of those soils is coarse gravelly alluvium. The conditions for soil calcite formation in soil profiles considered by Breecker and others (2009) are similar to those from profiles 1, 6, and 10 in this study in that they are characterized by low rainfall and low soil WHC. Note that low soil PCO2 indicates a small contribution of CO2 from soil respiration [S(z); Cerling 1984] and relatively low soil productivity at the time of calcite crystallization. Conversely, high soil PCO2 indicates a large contribution of CO2 from soil respiration (Cerling 1984) and relatively high soil productivity at the time of calcite

crystallization. Thus, soil productivity is limited in some way for profiles with low soil PCO2 estimates compared to those with high soil PCO2 estimates. Soil productivity is affected by many things (e.g., Schlessinger 1996 [and references therein]) but it is nevertheless a function of two primary factors: (1) moisture availability to facilitate transpiration, respiration, and carbon-fixation reactions and (2) temperature that directly affects the rates at which these reactions proceed. Therefore, any solution that seeks to explain the variation of soil PCO2, and soil-respired CO2 in particular, must somehow incorporate both of these factors. Significantly, Californian soil profiles with relatively high soil PCO2 estimates (profiles 5, 7, 11, and 14; Table 2) have much higher WHC and greater amounts of soil moisture utilization (Table 3) that progress later into the year and during warmer growing conditions than do profiles with lower soil PCO2 estimates (profiles 1, 6, and 10). This may indicate that greater productivity and soil PCO2 occur at the time of calcite crystallization in soils that have greater amounts of soil moisture storage available for utilization by soil flora later into the growing season. In order to evaluate this possibility, we use average soil PCO2 estimates based upon D13Ccc-om values and compare them with the estimated heat content of liquid water during the period of soil moisture utilization for those soil profiles with hydrology characterized by gravity-driven piston flow and that are evidently not affected by groundwater or waterlogging (i.e., we use profiles 1, 5, 6, 7, 10, 11, and 14). Soil PCO2 values used here are the average of soil PCO2 estimates from all horizons either from (1) depths beneath the zone characterized by large changes in calcite (and soil CO2) d13C values over short vertical distances (e.g., 52 cm; profile 5) or (2) below 30 cm in depth (profiles 1, 6, 7, 10, 11, and 14) because that is the typically recommended upper limit for sampling in most other calcite-forming soil profiles (Quade et al. 1989, Cerling and Quade 1993). The average soil PCO2 values range from 360 to 6250 ppmV (Fig. 4; Table 2). The energy flux (referred to herein as EPPT-U) associated with periods of soil moisture utilization is a calculation of heat content of the soil moisture volume that is utilized for ETp during periods of soil moisture utilization (Table 3). The specific heat of H2O (4.18 kJ/kg/8K) is used to convert this utilized volume of water into a chemical energy flux by assuming that 1 cm3 of water is equivalent to 1 g of water and that EPPT2 U (kJ/m /yr) is represented by the difference between surface air temperature during months of soil moisture utilization and the freezing point of water (Rasmussen and Tabor 2007). EPPT-U values for each soil profile range from 159 to 1211 kJ/m2/yr (Fig. 4). The energy flux values were plotted against soil CO2 estimates calculated from (1) maximum monthly mean surface air temperatures (Fig. 4A) and (2) average surface air temperature during periods of soil moisture utilization (Fig. 4B). The energy flux value EPPT-U shares a strong correlation with the inverse of soil PCO2 estimates from calcitebearing profiles characterized by two-component soil CO2 mixing (Fig. 4), whether the PCO2 estimates are calculated from maximum monthly mean surface air temperatures (Fig. 4A; R2 ¼ 0.96) or average surface air temperature during periods of soil moisture utilization (Fig. 4B; R2 ¼ 0.97). That is, low EPPT-U flux to the soil corresponds to low soil PCO2, whereas high EPPT-U flux to the soil corresponds to high soil PCO2 at the time of calcite crystallization. The very strong correlation between soil PCO2 estimates and EPPT-U values indicates that soil respiration is primarily a product of soil moisture availability and temperature during periods of soil moisture utilization and calcite crystallization. This data set indicates that soils composed of materials that can hold a large volume of soil moisture at the onset of soil moisture utilization (Table 3) and that continue to receive rainfall during the months of moisture utilization may be capable of maintaining higher rates of water transpiration and soil respiration later into the growing season. The data set provided here further indicates that soil PCO2



31

FIG. 4.—Plot of the inverse of average soil PCO2 estimates from .30 cm in depth in the profile for soils characterized by gravity-driven piston flow hydrology vs. EPPT-U calculated for (A) maximum monthly temperature and (B) average temperature during soil moisture utilization. EPPT-U is the sum of energy transferred through the soil during months of soil moisture utilization. See text for further discussion.

during soil moisture utilization and calcite crystallization may be in excess of 5000 ppmV (Table 2; Fig. 4). Note that results from profile 11 (Bidwell) indicate a relatively high soil PCO2, but this soil also only stores about half of its potential maximum WHC (WHC ¼ 27.8 cm, maximum storage ¼ 14.3 cm; Table 3) and receives relatively little precipitation (P ¼ 7.1 cm) compared to ETp (ETp ¼ 21.2 cm) through the period of soil moisture utilization. In this respect, it is possible that a calcareous soil such as this could achieve much

greater energy throughput (EPPT-U values) and a longer period of soil moisture utilization that extends into the warmest times of year if it formed in a climate characterized by higher winter rainfall to increase the soil moisture storage volumes and greater rainfall during the growing season (such as a udic soil moisture regime). We expect that such a postulated soil profile is capable of forming soil calcite at substantially higher soil PCO2 values than those estimated from the Californian soil profiles studied herein.


32


SUMMARY AND CONCLUSIONS This work presents the results from a study of 14 different calcareous soil profiles from California and emphasizes the importance of soil hydrology—and variations in the WHC of soil material in the rooting zone in particular—in order to explain the occurrence and timing of crystallization as well as variations in carbon isotope compositions of calcite among the different soil profiles. There appear to be two general scenarios for carbonate accumulation among the 14 different soil profiles. Scenario 1 involves seasonal flooding and stagnant or lateral flow of moisture in the zone of calcite formation, whereas scenario 2 involves well-drained profiles characterized by gravity-driven vertical flow and downward leaching of Ca2þ to the depth of wetting in the profile. The stable carbon isotope geochemistry of calcite varies over a range of 15.7ø, whereas the d13C values of plant organic matter vary over a narrow range of 4.3ø among all soil profiles. Calcite d13C values tend to be more negative in soils that experience groundwater flooding in the zone of calcite accumulation than in well-drained soils characterized by vertical downward flow of soil moisture. We suspect that this tendency derives from differences is gas diffusion, whereby poorly drained soils may experience non–Fickian-type diffusion that results in a smaller isotopic enrichment of 13CO2 derived from oxidation of organic matter and that also limits diffusion of isotopically more positive tropospheric CO2 to sites of calcite crystallization. Soil PCO2 estimates that result from d13C values of coexisting calcite and organic matter in these poorly drained soil profiles are either unrealistically high or offer inconclusive results (i.e., negative values of soil PCO2; Table 2). Studies of modern soil profiles have begun to look for indicators of complex soil drainage that have the potential to be preserved in the paleosol record, such as cathode-luminescence petrography (e.g., Mintz et al. 2011), and future work may help to further our collective ability to recognize such complicated profiles. However, we consider that the most practical approach involves avoiding the use of carbon isotope data for estimates of paleoatmospheric PCO2 that is collected from paleosols with morphological evidence of variably or poorly drained conditions (Ekart et al. 1999, Sheldon and Tabor 2009). Considering the abundance of paleosol profiles with evidence for variable and poor drainage in Upper Paleozoic and Mesozoic sedimentary basins, our recommended approach will likely make it challenging, or perhaps impossible, to develop high-resolution (e.g., 103–104) atmospheric PCO2 reconstructions from paleosol carbon isotope data in particular parts of the geological record. Finally, soil PCO2 estimates from coexisting calcite and organic matter in well-drained profiles characterized by vertical, gravity-driven flow indicate substantial variation in PCO2 at the time of calcite crystallization. Furthermore, soil PCO2 estimates show a strong inverse correlation with EPPT-U values, a metric that combines (a) moisture available for soil transpiration and (b) temperature to facilitate soil respiration. These results indicate that more can be done to reconstruct paleoatmospheric pCO2 than is achieved by (1) assuming soil PCO2 at the time of calcite crystallization (e.g., Tabor et al. 2004) or (2) using a single soil PCO2 for all paleosol profiles since Silurian time. Therefore, we propose that future work should seek to expand this study of modern well-drained soils with more analyses of the carbon isotope geochemistry of coexisting soil calcite in California, as well as to other places characterized by different seasonality, to identify (1) a probable maximum range of soil PCO2 at the time of soil calcite crystallization and (2) more soils that appear to preserve a complete CO2 mixing profile based upon depth trends of calcite d13C that are similar in form to profile 5 (Alturas: Table 2; Fig. 3). A selection of representative soils could then be analyzed for short-term changes in the concentration and stable carbon isotope composition of the soil gas phase, in a manner similar to that developed by Breecker and others (2009), to evaluate whether coexisting calcite and organic matter d13C values provide

estimates of soil PCO2 that are consistent with actual concentrations in those soils. If there still exists a strong relationship between soil PCO2 estimates and EPPT-U values after all this additional work, then it may be possible to use paleoclimate data from general circulation models (e.g., Peyser and Poulsen 2008), in conjunction with lithological data from individual paleosol profiles, to estimate growing-season EPPT-U values and to thereby place a reasonable range of probable soil PCO2 values at the time of calcite crystallization in the paleosol profile. No single laboratory or combination of laboratories could complete this kind of research in a practical amount of time. Nevertheless, our ability to provide constrained estimates of paleoatmospheric PCO2 from carbon isotope studies of paleosol profiles is certainly one of the most important contributions that paleopedology has provided to the discipline of earth sciences. In this respect, a community-wide and coordinated effort to resolve the issues surrounding soil PCO2 for the purposes of paleoatmospheric PCO2 reconstruction from paleosols may be a useful rallying point for the discipline of paleopedology.

ACKNOWLEDGMENTS Thanks to Professor Randal J. Southard for introducing N.J.T. to the University of California, Davis, Pedolarium and for permitting sampling of soils from the soil archives. Thanks also to Daniel O. Breecker and Thure E. Cerling for their reviews of this manuscript. N.J.T. is supported by National Science Foundation (NSF) grant EAR 0617250, NSF-EAR 0545654, and NSF-EAR 0447381. N.D.S. is supported by NSF-EAR 1024535.

REFERENCES Allison GB, Barnes CJ, Hughes MW, Leaney FWJ. 1984. Effect of climate and vegetation on oxygen-18 and deuterium profiles in soils. In Isotope Hydrology 1983, IAEA Symposium 270; September 1983; Vienna, Austria. p. 105–123. Alonso-Zarza AM, Tanner LH (Editors). 2006. Paleoenvironmental Record and Applications of Calcretes and Pallustrine Carbonates, Special Paper 416: Geological Society of America, Boulder, Colorado. 239 p. Arens NC, Jahren AH, Amundson R. 2000. Can C3 plants faithfully record the carbon isotopic composition of atmospheric carbon dioxide? Paleobiology 26:134–164. Arkley RJ. 1963. Calculation of carbonate and water movement in soil from climate data. Soil Science 96:239–248. Bouma TJ, Bryla DR. 2000. On the assessment of root and soil respiration for soils of different textures: interactions with soil moisture contents and soil CO2 concentration. Plant and Soil 227:215–221. Bouma TJ, Nielsen KL, Eissenstat DM, Lynch JP. 1997. Estimating respiration of roots in soil: interactions with soil CO2, soil temperature and soil water content. Plant and Soil 195:221–232. Boutton TW. 1991. Stable carbon isotope ratios of natural materials: II. Atmospheric, terrestrial, marine and freshwater environments. In Coleman DC, Fry B (Editors). Carbon Isotope Techniques: Academic Press, San Diego, California. p. 155–170. Breecker DO, Sharp ZD, McFadden LD. 2009. Seasonal bias in the formation and stable isotopic composition of pedogenic carbonate in modern soils from central New Mexico, USA. Geological Society of America Bulletin 121:630– 640. Buol SW, Southard RJ, Graham RC, McDaniel PA. 2003. Soil Genesis and Classification: Iowa State University Press, Ames. 494 p. Casals P, Romanya J, Cortina J, Bottner P, Couteaux M-M, Ramon Vallejo V. 2000. CO2 efflux from a Mediterranean semi-arid forest soil. I. Seasonality and effects of stoniness. Biogeochemistry 48:261–281. Cerling TE. 1984. The stable isotopic composition of modern soil carbonate and its relationship to climate. Earth and Planetary Science Letters 71:229–240. Cerling TE. 1991. Carbon dioxide in the atmosphere: evidence from Cenozoic and Mesozoic paleosols. American Journal of Science 291:377–400. Cerling TE, Quade J. 1993. Stable carbon and oxygen isotopes in soil carbonates. In Swert PK, Lohmann KC, McKenzie JA, Savin S (Editors).



Climate Change in Continental Isotopic Records, Geophysics Monograph 78: American Geophysical Union, Washington, DC. p. 217–231. Cerling TE, Solomon DK, Quade J, Bowman JR. 1991. On the isotopic composition of carbon in soil carbon dioxide. Geochimica et Cosmochimica Acta 55:3403–3405. Clark I, Fritz P. 1997. Environmental Isotopes in Hydrogeology: Lewis Publishers, New York. 328 p. Coney L, Reimold WU, Hancox PJ, Mader D, Koeberl C, McDonald I, Struck U, Vajda V, Kamo SL. 2007. Geochemical and mineralogical investigation of the Permian–Triassic boundary in the continental realm of the southern Karoo basin, South Africa. Palaeoworld 16:67–104. Craig H. 1957. Isotopic standards for carbon and oxygen and correction factors for mass-spectrometric analysis of carbon dioxide. Geochimica et Cosmochimica Acta 12:133–149. Criss RE. 1999. Principles of Stable Isotope Distribution: Oxford University Press, New York. 254 p. Deutz P, Montanez IP, Monger HC. 2002. Morphology and stable and radiogenic isotope composition of pedogenic carbonates in late Quaternary relict soils, New Mexico, USA; an integrated record of pedogenic overprinting. Journal of Sedimentary Research 72:809–822. Drever JI. 1973. The preparation of oriented clay mineral specimens for X-ray diffraction analysis by a filter–membrane peel technique. American Mineralogist 58:553–554. Ekart DD, Cerling TE, Montanez IP, Tabor NJ. 1999. A 400 million year carbon isotope record of pedogenic carbonate: implications for paleoatmospheric carbon dioxide. American Journal of Science 299:805–827. Feng J, Hsieh YP. 1998. Wetlands and aquatic processes; sulfate reduction in freshwater wetland soils and the effects of sulfate and substrate loading. Journal of Environmental Quality 27:968–972. Feng W, Yapp CJ. 2009. Paleoenvironmental implications of concentration and 13C/12C ratios of Fe(CO3)OH in goethite from a mid-latitude Cenomanian laterite in southwestern Minnesota. Geochimica et Cosmochimica Acta 73:2559–2580. Gile LH, Hawley JW, Grossman RB. 1981. Soils and Geomorphology in the Basin and Range Area of Southern New Mexico––Guidebook to the Desert Project, Memoir 39: New Mexico Bureau of Mines and Mineral Resources, Socorro, New Mexico. p. 222. Gutierrez KA, Sheldon ND. 2012. Paleoenvironmental reconstruction of Jurassic dinosaur habitats of the Vega Formation, Asturias, Spain. Geological Society of America Bulletin 124:596–610. Hoefs J. 1997. Stable Isotope Geochemistry: Springer, Berlin. 201 p. Hsieh JCC, Yapp CJ. 1999. Stable carbon isotope budget of CO2 in a wet, modern soil as inferred from Fe(CO3)OH in pedogenic goethite: possible role of calcite dissolution. Geochimica et Cosmochimica Acta 63:767–783. Jacobs BF, Kingston JD, Jacobs LL. 1999. The origin of grass-dominated ecosystems. Annals of the Missouri Botanical Garden 86:590–643. Jenny HJ. 1941. Factors of Soil Formation: McGraw-Hill, New York. 281 p. Jost W. 1960. Diffusion in Solids, Liquids, Gases: Academic Press, New York. 558 p. Kraimer RA, Monger HC. 2009. Carbon isotopic subsets of soil carbonate—a particle size comparison of limestone and igneous parent materials. Geoderma 150:1–9. Lieth H. 1975. Modeling the primary productivity of the world. In Lieth H, Whittaker RH (Editors). Primary Productivity of the Biosphere: Springer, New York. p. 237–263. Lourantou A, Chappellaz J, Barnola J-M, Masson-Delmotte V, Raynaud D. 2010. Changes in atmospheric CO2 and its carbon isotopic ratio during the penultimate deglaciation. Quaternary Science Reviews 29:1983–1992. Machette MN. 1985. Calcic soils of the southwestern United States. In Wiede DL (Editor). Soils and Quaternary Geology of the Southwestern United States, Special Paper 203: Geological Society of America, Boulder, Colorado. p. 1–22. Mack GH, James WC, Monger HC. 1993. Classification of paleosols. Geological Society of America Bulletin 105:129–136. McFadden LD, Tinsley JC. 1985. Rate and depth of pedogenic–carbonate accumulation in soils: formulation and testing of a compartment model. In Wiede DL (Editor). Soils and Quaternary Geology of the Southwestern

33

United States, Special Paper 203: Geological Society of America, Boulder, Colorado. p. 23–41. Mintz JS, Driese SG, Breecker DO, Ludvigson GA. 2011. Influence of changing hydrology on pedogenic calcite precipitation in Vertisols, Dance Bayou, Brazoria County, TX: implications for estimating paleoatmospheric pCO2. Journal of Sedimentary Research 81:394–400. Monger HC, Cole DR, Gish JW, Giordano TH. 1998. Stable carbon and oxygen isotopes in Quaternary soil carbonates as indicators of ecogeomorphic change in the northern Chihuahuan Desert, USA. Geoderma 82:137–172. Montan˜ez IP, Tabor NJ, Niemeier D, DiMichele WA, Frank TD, Fielding CR, Isbell JL. 2007. CO2-forced climate and vegetation instability during Late Paleozoic deglaciation. Science 315:87–91. Nordt LC, Hallmark CT, Wilding LP, Boutton TW. 1998. Quantifying pedogenic carbonate accumulations using stable carbon isotopes. Geoderma 82:115– 136. O’Leary MH. 1984. Measurement of the isotopes fractionation associated with diffusion of carbon dioxide in aqueous solution. Journal of Physical Chemistry 88:823–825. Passey BH, Levin NE, Cerling TE, Brown FH, Eiler JM. 2010. Hightemperature environments of human evolution in East Africa based on bond ordering in paleosol carbonates. Proceedings of the National Academies of Science of the United States of America 107:11245–11249. Petit JR, Jouzel J, Raynaud D, Barkov NI, Barnola J-M, Basile I, Chappellaz J, Davis M, Delaygue G, Delmotte M, Kotlyakov VM, Legrand M, Lipenkov VY, Lorius C, Papin L, Ritz C, Saltzman E, Stievenard M. 1999. Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399:429–436. Peyser CE, Poulsen CJ. 2008. Controls on Permo-Carboniferous precipitation over tropical Panaea: a GCM sensitivity study. Palaeogeography, Palaeoclimatology, Palaeoecology 268:181–192. Quade J, Cerling TE, Bowman JR. 1989. Development of Asian monsoon revealed by marked ecological shift during the latest Miocene in northern Pakistan. Nature 342:162–166. Rabenhorst MC, West LT, Wilding LP. 1984. Identification of pedogenic carbonates using stable carbon isotope and microfabric analysis. Soil Science Society of America Journal 48:125–132. Rabenhorst MC, West LT, Wilding LP. 1991. Genesis of calcic and petrocalcic horizons in soils over carbonate rocks. In Nettleton WD (Editor). Occurrence, Characteristics and Genesis of Carbonate, Gypsum and Silica Accumulations in Soils, Special Publication 26: Soil Science Society of America, Madison, Wisconsin. p. 61–74. Rasmussen C, Tabor NJ. 2007. Applying a quantitative pedogenic energy model across a range of environmental gradients. Soil Science Society of America Journal 71:1719–1729. Reheis MC. 2003. Dust Deposition in Nevada and Utah, 1984–2002: United States Geological Survey Open File Report 03–138, Washington, DC. p. 1– 11. Retallack GJ. 1993. Classification of paleosols: discussion and reply. Geological Society of America Bulletin 105:1635–1636. Retallack GJ. 1994. The environmental factor approach to the interpretation of paleosols. In Amundson R (Editor). Factors in Soils Formation: A Fiftieth Anniversary Retrospective, Special Publication 33: Soil Science Society of America, Madison, Wisconsin. p. 31–64. Retallack GJ. 1997. Neogene expansion of the North American prairie. Palaios 12:380–390. Retallack GJ. 2005. Pedogenic carbonate proxies for amount and seasonality of precipitation in paleosols. Geology 33:333–336. Romanek CS, Grossman EL, Morse JW. 1992. Carbon isotopic fractionation in synthetic aragonite and calcite: effects of temperature and precipitation rate. Geochimica et Cosmochimica Acta 56:419–430. Royer DL. 1999. Depth to pedogenic carbonate horizon as a paleoprecipitation indicator? Geology 27:1123–1126. Salomons W, Goudie AS, Mook WG. 1978. Isotopic composition of calcrete deposits from Europe, Africa and India. Earth Surface Processes 3:43–47. Schlessinger WH. 1982. Carbon storage in the caliche of arid soils: a case study from Arizona. Soil Science 133:247–255. Schlessinger WH. 1985. The formation of caliche in soils of the Mojave Desert, California. Geochimica et Cosmochimica Acta 49:57–66.


34


Schlessinger WH. 1996. Biogeochemistry—An Analysis of Global Change: Academic Press, London. 588 p. Sharp Z. 2007. Principles of Stable Isotope Geochemistry: Prentice Hall, New York. 360 p. Sheldon ND, Tabor NJ. 2009. Quantitative paleoenvironmental and paleoclimatic reconstruction using paleosols. Earth-Science Reviews 95:1–52. Soil Survey Staff. 2010a. Natural Resources Conservation Service: United States Department of Agriculture Official Soil Series Descriptions. http:// soils.usda.gov/technical/classification/osd/index.html. Accessed February 10, 2010. Soil Survey Staff. 2010b. Keys to Soil Taxonomy, 11th ed.:USDA-Natural Resources Conservation Service, Washington, DC. 338 p. Solomon DK, Cerling TE. 1987. The annual carbon dioxide cycle in a montane soil: observations, modeling, and implications for weathering. Water Resources Research 23:2257–2265. Tabor NJ, Montanez IP, Steiner M, Schwindt D. 2007. The d13C values of Permo–Triassic carbonates from South Africa reflect a stinking, sulfurous swamp, not atmospheric conditions. Palaeogeography, Palaeoclimatology, Palaeoecology 225:370–381. Tabor NJ, Yapp CJ. 2005a. Coexisting goethite and gibbsite from a high– paleolatitude (558N) Late Paleocene laterite: concentration and 13C/12C ratios of occluded CO2 and associated organic matter. Geochimica et Cosmochimica Acta 69:5495–5510. Tabor NJ, Yapp CJ. 2005b. Incremental vacuum dehydration–decarbonation experiments on a natural gibbsite (a–Al(OH)3): CO2 abundance and d13C values. Geochimica et Cosmochimica Acta 69:519–527. Tabor NJ, Yapp CJ. 2005c. Juxtaposed Permian and Pleistocene isotopic archives: surficial environments recorded in calcite and goethite from the Wichita Mountains, Oklahoma. Geological Society of America Special Paper 395:55–70. Tabor NJ, Yapp CJ, Montan˜ez IP. 2004. Goethite, calcite, and organic matter from Permian and Triassic soils: carbon isotopes and CO2 concentrations. Geochimica et Cosmochimica Acta 68:1503–1517. Thornthwaite CW. 1948. An approach toward a rational classification of climate. Geographical Review 38:55–94. Veizer J, Ala D, Azmy K, Brluckshen P, Buhl D, Bruhn G, Carden GAF, Diener A, Ebneth S, Godderis Y, Jasper T, Dorte C, Paweliek F, Podlatha OG, Strauss H. 1999. 87Sr/86Sr d13C and d18O evolution of Phanerozoic seawater. Chemical Geology 161:59–88. Vepraskas MJ. 1994. Redoximorphic Features for Identifying Aquic Conditions, Technical Bulletin 301: North Carolina Agricultural Research Service, Raleigh, North Carolina. 33 p. Wang Y, McDonald E, Amundson R, McFadden L, Chadwick O. 1996. An isotopic study of soils in chronological sequences of alluvial deposits, Providence Mountains, California. Geological Society of America Bulletin 108:379–391. West LT, Drees LR, Wilding LP, Rabenhorst MC. 1988. Differentiation of pedogenic and lithogenic carbonate forms in Texas. Geoderma 43:271–287. Whelan T III, Roberts HH. 1973. Carbon isotope composition of diagenetic carbonate nodules from freshwater swamp deposits. Journal of Sedimentary Petrology 43:54–58. Yapp CJ. 2001. Mixing of CO2 in surficial environments as recorded by the concentration and d13C values of the Fe(CO3)OH component in goethite. Geochimica et Cosmochimica Acta 65:4115–4130. Yapp CJ. 2002. Erratum. Geochimica et Cosmochimica Acta 66:1497. Yapp CJ, Poths H. 1996. Carbon isotopes in continental weathering environments and variations in ancient atmospheric CO2 pressure. Earth Planetary Science Letters 137:71–82. Yemane K, Siegenthaler C, Kelts K. 1989. Lacustrine environment during Lower Beaufort (Upper Permian) Karoo deposition in northern Malawi. Palaeogeography, Palaeoclimatology, Palaeoecology 70:165–178.
