Climate controls on pedogenic carbonate formation and timing
Geological Society of America Bulletin, v. 1XX, no. XX/XX 1
Climate controls on spatial and temporal variations in the formation of pedogenic carbonate in the western Great Basin of North America
Erik J. Oerter† and Ronald AmundsonDepartment of Environmental Science, Policy and Management, 130 Mulford Hall, University of California, Berkeley, California 94720, USA
ABSTRACT
Pedogenic carbonate is common in arid and semiarid soils, and though its stable C and O isotope composition has been shown to reflect local environmental conditions, questions remain about the quantitative na-ture of this relationship and the implications for paleoclimate applications. To further address these questions, a climosequence of four Holocene soils in Fish Lake Valley, Nevada (western United States), was instru-mented and examined over the course of more than a year. The annual precipitation along the transect ranges from ~80 mm yr–1 to ~220 mm yr–1. Results show that the C and O stable isotope compositions of soil CO2, H2O, and carbonate change systematically with elevation and climate. However, there was considerable temporal variability in the conditions that affected carbonate isotope values. At the lowest elevation, CO2 δ13C and H2O δ18O values were similar to that in equi-librium with carbonate nearly year-round. At the midelevation sites, spring through summer CO2 δ13C and H2O δ18O values ap-peared to most closely match the δ13C and δ18O values of pedogenic carbonate. At the highest elevation, the C and O isotopes of carbonate did not reflect the soil CO2 δ13C and H2O δ18O values measured during the period of study, but they did appear to reflect soil respiration rates during the late spring. The transect shows that arid soil carbon-ate δ13C values strongly reflect variations in soil respiration rates and the resulting con-centrations of soil CO2. These results also indicate that depth profile sampling may be required to adequately constrain respiration rates, which affect the interpreted atmo-spheric CO2 concentrations.
INTRODUCTION
Climate is a primary driver of soil develop-ment (Jenny, 1941), and many soil properties strongly reflect variations in temperature and rainfall. Pedogenic carbonate accumulates in arid, semiarid, and seasonally dry climates, and its stable C and O isotopic composition provides powerful climatic information. Several decades ago, the framework for using the stable C and O isotope composition of carbonate as a climate proxy was developed (Cerling, 1984). However, detailed empirical studies of modern carbon-ate-forming environments have more recently evolved in concert with improvements in labora-tory and field instrumentation. The most critical question in relation to the use of pedogenic car-bonate as a paleoclimate proxy is: Does the sta-ble isotope composition of carbonate reflect the average annual soil environmental conditions, or does it instead reflect conditions character-istic of specific seasons (i.e., Breecker et al., 2009; Passey et al., 2010; Peters et al., 2013)? A related question is: How does the season of car-bonate precipitation vary geographically? These questions have significant relevance to the use and interpretation of pedogenic carbonate C and O stable isotope values in the paleorecord.
When carbonate (typically low-Mg calcite, CaCO3) precipitates in soil, its C isotopic com-position (d13C) is controlled by the d13C value of soil CO2 (Cerling, 1984), which is a function of vegetation metabolic pathways (Cerling et al., 1991) and soil respiration rates (Cerling, 1984). Soil water and temperature control the O isotopes (d18O) in soil carbonate at the time of mineral formation. Soil water is sourced from precipita-tion, which is a product of local hydroclimate and atmospheric circulation patterns (Dansgaard, 1964; Gat, 1996). Pedogenic carbonate pre-cipitates when the soil solution becomes super-saturated due to evapotranspiration, the soil CO2 concentration decreases, and/or soil temperature increases (Breecker et al., 2009). These conditions may occur during various times of the year and
may vary across sites in close proximity to each other, especially in intermontane environments where local climates are influenced dramatically by steep elevation gradients. While complexity might be the norm, there have been only a few studies that have acquired the data to systemati-cally investigate the issue. Pedogenic carbonate has been interpreted to form in the warm and dry conditions of late spring (Breecker et al., 2009), after dry-down from seasonal precipitation inputs (Peters et al., 2013), and even in winter (Quade et al., 1989), depending on the location and study. Breecker et al. (2009) evaluated the implications of a warm-season bias in paleosol-based estimates of paleoatmospheric CO2 concentrations and sug-gested that previous interpretations may have overestimated CO2 concentrations.
In order to build on existing research, we established a climosequence of four sites along a climate gradient in the Great Basin of North America, measuring climate and soil conditions over the course of a year. The data show strong correlations with elevation but are also sugges-tive of spatio-temporal variations in the condi-tions that promote carbonate precipitation. The opportunities and challenges for paleoclimate studies are examined in light of the growing body of modern soil carbonate studies.
STUDY LOCATION AND METHODS
Study Location
The soils studied are in the Trail Canyon area of Fish Lake Valley, Nevada, on the northeastern flank of the White Mountains (Fig. 1). Chiato-vich and Rock Creek(s) have formed a flight of alluvial fans and inset terraces with associated soils. A climosequence of four study sites was established at elevations from 1482 to 2602 m, along which mean annual precipitation (MAP) and mean annual air temperature (MAAT) vary from 78 mm yr–1 to 214 mm yr–1, and 12.2 °C to 7.0 °C, respectively (shown in Table 1 and dis-cussed more later herein). The ages of the soils
GSA Bulletin; Month/Month 2016; v. 128; no. X/X; p. 1–10; doi: 10.1130/B31367.1; 8 figures; 1 table; Data Repository item 2016043.; published online XX Month 2016.
†Present address: Department of Geology and Geo-physics, University of Utah, 115 S. 1460 E., Salt Lake City, Utah 84112-0102, USA; erikjoerter@ gmail .com.
2 Geological Society of America Bulletin, v. 1XX, no. XX/XX
at the lower three sites (A, B, and C) are esti-mated to be 3.8 thousand years (k.y.) old, based on radiocarbon dates on charcoal found in soils in the area and geomorphic correlations (Harden et al., 1991; Reheis and Block, 2007). The age of the highest-elevation site (D) is less well con-strained, but it is likely mid-Holocene, based on its position adjacent to the Rock Creek modern floodplain. The soils thus reflect soil-forming conditions since the mid-Holocene and should have integrated properties that reflect modern conditions. Field and laboratory methods are briefly summarized next, and they are described in more detail in the supplemental information.1
Field Methods
Soils were excavated to at least 1 m. This allowed us to carefully observe and sample soil horizons, as well as the alluvial parent material. Soil profile descriptions and sampling followed common pedological procedures (Soil Survey Staff, 1999; Schoeneberger et al., 2012). Bulk soil samples from each pedogenic horizon, as well as separate samples of the carbonate-bear-ing fine soil matrix material (and clasts with car-bonate rinds), were collected every 10 cm where carbonate was present. Duplicate soil samples (~25 g) for soil water d18O isotope analyses were placed in 30-mL borosilicate glass vials with Polyseal caps for subsequent vacuum distillation (as discussed in the following). These samples were collected immediately from excavated sur-faces to avoid evaporative isotope enrichment. Volumetric soil water and temperature sensors
were installed in soil excavation walls at depths of 10 cm, 25 cm, 50 cm, and 100 cm and were connected to data loggers set to record measure-ments every 0.5 h. Mineral oil was utilized inside precipitation collectors to prevent evaporation.
In order to conduct repeated nondestructive sampling of soil water for O isotopes, the O isotope composition of soil CO2 was used as a probe of soil water (Stern et al., 1999; Breecker and Sharp, 2008). Small-diameter (~6 mm outer diameter) stainless-steel gas wells were installed at depths similar to the soil moisture sensors. Soil air was sampled by syringe from each well (following purging) and injected into N2-filled leak-proof septum-capped vials for subsequent C and O isotope analysis.
During each sampling (May 2013, August 2013, November 2013, April 2014, May 2014), the following data or samples were collected: soil moisture and temperature data, air tem-perature and humidity data, accumulated pre-cipitation samples, plant stem samples, soil gas samples, atmospheric air samples, surface water samples from Rock Creek close to site D, and an intermittent groundwater spring close to site B.
Laboratory Methods
Soil color, particle size, and texture classifica-tions were made on the <2 mm fine soil fraction sieved from bulk samples. Carbonate C and O isotope analyses were performed on both the <2 mm fine material as well as carbonate coat-ings on clasts. Carbonate samples were analyzed on a GV IsoPrime dual-inlet mass spectrometer. Soil and plant xylem water was extracted by heating, distillation, and cryogenic collection in a vacuum line for 240 min to ensure complete water recovery. Extracted water samples were placed in glass vials and equilibrated with CO2,
and the O isotope composition of the CO2 was measured using a Thermo Delta Plus XL mass spectrometer in continuous-flow mode. Analy-ses of the C and O isotopic composition of CO2 in soil gas samples and atmospheric air samples were conducted on the same instrument.
Isotope values are reported in d notation: d = (Rsample/Rstandard – 1) × 1000, where Rsample and Rstandard are the 13C/12C or 18O/16O ratios for the sample and standard, respectively. Carbonate d13C and d18O values, and CO2 d13C values are reported in per mil (‰) referenced to the Vienna Pee Dee Belemnite (VPDB) standard, and CO2 and H2O d18O values are reported in per mil (‰) referenced to Vienna standard mean ocean water (VSMOW; Coplen, 1994).
The d18O values of soil water were calculated from soil CO2 d18O values using the O isotope fractionation relationship between CO2 and H2O (Brenninkmeijer et al., 1983) and average soil temperatures measured near each soil gas well for the 24-h interval prior to soil gas collection. The exact time for CO2-H2O O isotope equilibra-tion in these soils is unknown, and thus 24 h is the most reasonable interval to average soil tempera-tures and avoid bias introduced by variability in sampling times between sites and sampling dates. Further evaluation of 24-h temperature averaging in the calculation of H2O d18O values calculated from soil CO2 is discussed more in the section on Soil CO2-H2O Oxygen Isotope Equilibrium.
The d18O values of xylem water reflect soil water derived from different soil depths. There-fore, the fraction of soil water derived from differing soil depths was determined using a two-component linear mixing model (Phillips and Gregg, 2001): fractionupper = (dxylem – dlower)/(dupper – dlower), and fractionlower = 1 – fractionupper, where “xylem” is the plant water d18O value, “upper” is the d18O value of soil water most sim-ilar to xylem water above the inferred zone of soil water uptake, and “lower” is the d18O value of the soil water below. The estimated average depth of soil water uptake was then calculated as a depth-weighted average using the relative contributions of water from each depth: depth of soil water uptake = ([depthupper][ fractionupper]) + ([depthlower][ fractionlower]).
RESULTS
Climate, Vegetation, and Soil Properties versus Elevation
The average annual air temperature and pre-cipitation values for the study period are shown in Table 1. Average air temperature lapse rate with elevation was –5.1 °C km–1, and precipita-tion increased by 121 mm km–1. The measured seasonal distribution of precipitation is shown in
Fish LakeValley
WhiteMtns.
BC
DA
20kmSouthwestern North America
Salt LakeCity
LosAngeles
SanFrancisco
Fish LakeValley
125°W
118°20′W
110°W42°30′N 38°N
30°30′N 37°43′N118°W
Figure 1. (Left) Map of southwestern North America with location of Fish Lake Valley de-noted by star. (Right) Fish Lake Valley region with study sites A–D denoted by markers.
1GSA Data Repository item 2016043, description of field and laboratory methods and Tables S1–S6, is available at http:// www .geosociety .org /pubs /ft2016 .htm or by request to editing@ geosociety .org.
Climate controls on pedogenic carbonate formation and timing
Geological Society of America Bulletin, v. 1XX, no. XX/XX 3
Table 1. Spring season precipitation dominated the lowest- and highest-elevation sites (A and D), while the midelevations received more sum-mer precipitation (sites B and C). Winter pre-cipitation d18O values ranged from –14.25‰ (VSMOW) at 1482 m to –15.36‰ at 2602 m, and summer precipitation d18O values were –4.51‰ (VSMOW) at 1482 m and –8.08‰ at 2602 m (Table 1; supplementary Table S1 [see footnote 1]). Precipitation during the winter and spring is derived from storms originating in the North Pacific, while summer and fall precipita-tion is sourced from the subtropics and brought into Fish Lake Valley by northward-moving monsoon systems (Lechler and Galewsky, 2013). The precipitation O isotope lapse rates are –0.99‰ km–1 and –3.19‰ km–1 for winter and summer, respectively.
Vegetation communities varied predictably with elevation: saltbush scrubland at 1482 m elevation (site A), blackbrush shrubland at 1745 m (site B), sagebrush shrubland at 2140 m (site C), and piñon-sage woodland at 2602 m (site D; Table 1). These climatically constrained
biomes are common in the White Mountains (Hall, 1991) and throughout the Great Basin of North America (Grayson, 2011).
Soil properties are listed in Table 1 and sup-plementary Table S2 (see footnote 1). All of the soils were classified as carbonate-rich Aridisols (Haplocalcids), but the soil at the highest eleva-tion bordered on a Ustic (semiarid), rather than Aridic, moisture regime. Dry soil colors are yellow-red (10YR) and became consistently darker (lower value) with increasing elevation. A key change along the elevation gradient is the increasing particle size of the parent mate-rial. The lowest site (site A) was a sandy, nearly gravel-free soil formed in granitic and rhyolitic parent material from the White Mountains to the west, as well as eolian sand and dust. Site A soil had an incipient natric horizon (high salt con-tent) from 5 to 20 cm depth, though it was not thick enough to classify this soil as a Natrid. The midelevation soils at sites B and C were formed on gravelly rhyolite and andesite from higher in the Rock Creek watershed and nearby intrusive rocks. Soil at the highest elevation (site D) was
formed in coarse alluvial and colluvial gravels and cobbles derived from the Boundary Peak quartz monzonite upslope of the site, as well as from the adjacent Mustang Mountain metasedi-mentary slate. Soil horizon development, such as clay content and soil structure, generally increased with elevation and precipitation.
Pedogenic carbonate in the three lowest ele-vations was disseminated in the <2 mm grain size and also occurred as coatings on clasts. The highest elevation only had carbonate coatings on clasts. Isotope depth profiles (discussed later herein) indicated that the fine, disseminated car-bonate is pedogenic and not inherited from par-ent material or from eolian inputs. The carbonate coatings on clasts, i.e., pedothems (Oerter et al., 2016) were thin (≤1 mm thick; Stage I of Gile et al., 1966) and were only located on clast bot-toms, indicative of in situ formation. The depth to pedogenic carbonate increased with increas-ing elevation and MAP, with a relationship of: MAP [mm yr–1] = 3.47Depth [cm] + 80.8 (R2 = 0.87 on four measurements), which is similar to that previously observed in arid and semiarid
TABLE 1. GEOGRAPHIC, CLIMATIC, AND PRECIPITATION δ18O VALUES, SOIL AND VEGETATION PROPERTIES, AND SEASONAL PRECIPITATION DISTRIBUTION OF STUDY SITES
Note: Crassulacean acid metabolism (CAM)-type included with C4 group. Precipitation δ18O values are in VSMOW.*MAAT (mean annual air temperature); average air temperature during the period: 18 May 2013–18 May 2014.†MAP (mean annual precipitation); precipitation samples collected during the monitoring period (380 d.) and normalized to 365 d.§Summer precip collected 26 August 2013.#Winter precip collected 30 Nov 2013, 19 April 2014, 31 May 2014.
Oerter and Amundson
4 Geological Society of America Bulletin, v. 1XX, no. XX/XX
soils of California and Nevada ( Arkley, 1963), but more shallow (at a given MAP) than soil carbonate in the North American Great Plains (Jenny and Leonard, 1934; Royer, 1999).
Seasonal Soil Volumetric Water Dynamics
Soil volumetric water contents averaged over each season at each site are shown in Figures 2A–2D and supplementary Table S3 (see foot-note 1). Seasons are defined as, spring = March, April, May; summer = June, July, August; fall = September, October, November; and winter = December, January, February. The lower-ele-vation (A–C) soils were wettest at depths up to ~25 cm during the spring (summer at site D), and dried out through the summer and fall, eventu-ally reaching their driest during the winter (Figs. 2A–2D). The seasonal change in soil moisture was more pronounced at depths down to 50 cm, and much less at 50 cm and deeper. In contrast to wetter conditions in the shallow soils during the spring, summer was the season of highest soil moisture, and winter had the lowest soil moisture at soil depths of 50 cm and 100 cm at all sites.
Carbonate C and O Stable Isotopes
Soil carbonate profiles of d13C and d18O values are shown in Figures 3A and 3B and reported in supplementary Table S4 (see foot-note 1). At sites with both disseminated carbon-ate and coatings on clasts, it appears that fine carbonates have been largely to wholly reset to pedogenic conditions (Figs. 3A and 3B; supple-mentary Table S4 [see footnote 1]). Previous work in Fish Lake Valley near site B by Pendall et al. (1994) suggests that the time required for this resetting of d13C and d18O values is between 1 and 3.8 k.y. In the following analyses and discus-sion, we focus specifically on the pedothem car-bonate coatings formed on clast bottoms, because (1) all soils in this study contained this form of the mineral, and (2) carbonate coatings on clast bottoms were undoubtedly formed in situ.
Soil CO2-H2O Oxygen Isotope Equilibrium
The use of soil CO2 d18O values to estimate soil H2O d18O values is based on O isotope equi-librium between soil CO2 and soil H2O (Stern et al., 1999; Amundson et al., 1998; Breecker and Sharp, 2008). In order to further test this assump-tion and calibrate the relationship between d18O values of soil CO2 and soil H2O at our sites, soil water d18O values were measured by two methods and compared. The d18O values of soil water obtained from vacuum distillation tended to be 1.5‰ lower than the calculated value of water obtained from the d18O values of soil CO2 (Fig. 4), with a slope of 0.93. Stern et al. (1999) suggested that the d18O value of soil CO2 is
~1‰ higher than that predicted by equilibrium between pure CO2 and H2O. The offset is likely due to rate limitations in isotope exchange, which result in partial diffusional enrichment of 18O in the CO2. Here, the observed enrichment appears to be ~1.5‰. Therefore, we subtracted a 1.5‰ correction factor from the soil water d18O values calculated from in-soil CO2-equilibrated d18O measurements. It should be noted that vacuum distillation of soil water can be subject to a vari-ety of errors during the distillation process (e.g., Araguás-Araguás et al., 1995), and the depth pro-files of vacuum-distilled soil water in this study do show variability at some depths that is not reflected in the soil CO2 measurements. However, the general shapes of the depth profiles of the two soil water measurement approaches are similar (Figs. 5A–5D). Evaluation of whether 24 h is a reasonable averaging time for the temperature used to calculate soil H2O d18O values from soil CO2 (Brenninkmeijer et al., 1983) can be con-ducted by comparing the depth profiles of cal-culated to vacuum-distilled soil H2O d18O values (Fig. 5). Deep soil temperatures do not vary over 24 h, while shallow soil temperatures vary diur-nally, especially during summer. If soil CO2-H2O O isotope equilibration is fast enough to be affected by transient soil temperatures at shallow depths, the calculated and vacuum- distilled soil H2O d18O values would differ at shallow depths. Instead, the depth profiles of each method match well (after the 1.5‰ correction factor is applied) throughout the soil depth profiles at each site (Fig. 5), indicating that 24 h is likely an appropri-ate time interval to average temperatures for cal-culations of soil H2O d18O values from soil CO2.
Seasonal Soil CO2 Concentration Dynamics
Soil CO2 concentrations with soil depth at each sampling event are reported in supplemen-tary Table S5 (see footnote 1). Soil CO2 concen-trations at all sites were lowest at the November sampling, suggesting dry soil and cold tem-peratures that inhibit soil biological activity. At the lower-elevation sites (sites A–C), soil CO2 concentrations were higher in April, reflecting warming soil temperature, in contrast to site D, where soil CO2 remained low along with soil temperature. At the lowest elevation (site A), soil CO2 concentrations in May were the highest of the study period, and soil CO2 concentrations were highest in August at the higher-elevation sites (sites B–D).
Soil CO2 and H2O Isotope Dynamics along the Transect
Soil CO2 d13C values decreased with increasing soil depth and elevation (Figs. 6A–6D; supplementary Table S5 [see foot-
0
20
40
60
80
100
0.02 0.04 0.06 0.08 0.10 0.12 0.14
Soil
Dept
h (c
m)
Site B
Summer 2013Fall 2013Winter 2014Spring 2014
B
0
20
40
60
80
100
0.02 0.04 0.06 0.08 0.10 0.12 0.14
Soil
Dept
h (c
m)
Site
Summer 2013Fall 2013Winter 2014Spring 2014
C
0
20
40
60
80
100
0.02 0.04 0.06 0.08 0.10 0.12 0.14
Soil
Dept
h (c
m)
Site D
Summer 2013Fall 2013Winter 2014Spring 2014
D
0
20
40
60
80
100
0.02 0.04 0.06 0.08 0.10 0.12 0.14
Soil
Dept
h (c
m)
Soil Moisture (m3/m3)
Site A
Summer 2013Fall 2013Winter 2014Spring 2014
A
Figure 2. Seasonal soil moisture depth pro-files; markers are average values of continu-ous measurements (0.5 h) for each season (months specified in section on Seasonal Soil Volumetric Water Dynamics); seasonal maximum and minimum values are listed in supplementary Table S3 (see text foot-note 1). Gray bands indicate average depth zones of plant water uptake determined by O isotope composition of plant xylem water. Arrows denote temporal evolution of depth profiles between seasons.
Climate controls on pedogenic carbonate formation and timing
Geological Society of America Bulletin, v. 1XX, no. XX/XX 5
note 1]). At the lowest elevation (site A), soil CO2 d13C values were lowest in Novem-ber and highest in August (Fig. 6A). In con-trast, at the mid- and highest-elevation sites (B–D), August soil CO2 d13C values were lowest, while November values were highest (Figs. 6B–6D).
Soil water d18O values generally decreased with increasing elevation and depth (Fig. 6E–6H; supplementary Table S5 [see footnote 1]). Beyond this trend, seasonal oscillations occurred near the surface due to both the addi-tion of precipitation (in the winter and spring) and to the evaporative removal of water (in the late spring to summer). These seasonal oscilla-tions tended to dissipate with depth, and sea-sonal oscillations were restricted to ≤4‰ at 100 cm depth in all soils.
Seasonal Plant Water Uptake Dynamics
The apparent depths of soil water uptake by plants are shown in Figure 2 and in supple-mentary Table S6 (see footnote 1). At site A, Atriplex spinifera utilized water from ~12 to 20 cm throughout the year, except during May 2014, when xylem water was most similar to soil water from 55 cm. At site B, xylem water from Coleogyne ramosissima was always more enriched in 18O than the soil water at any depth. At site C, Grayia spinosa xylem water was most similar to soil water from ~22 to 48 cm depth, except for August, where depths of ~18 cm were closest to plant values. At site D, Artemisia tri-dentata xylem water most closely correlated to soil water from ~28–47 cm depth, while Pinus monophylla xylem water had lower d18O val-ues than any measured soil water, indicative of water extraction from deeper within the alluvial sediments.
DISCUSSION
The results indicate that soil H2O, CO2, and carbonate d13C and d18O values change sys-tematically with elevation and climate. While there is inherent difficulty in assigning specific intervals (seasons) of carbonate formation from field measurements taken over the course of only one year, the conditions observed during differ-ent seasons generally appear to explain the C and O isotope trends of the soil carbonate. The data also indicate that in the soils studied here, with distinctly seasonal rainfall, soil carbonate C and O depth profiles integrate the yearly and multi-year variability, reflecting not one discrete inter-val of carbonate formation, but that of multiple formation events. These data thus suggest that arid and semiarid soils with similar climatic con-ditions to those studied here may have similar variability in the record of environmental condi-tions recorded by their pedogenic carbonate. It is possible that, in some soils, carbonate may not form at all during some years (Breecker et al., 2009), or that some seasons are only occasion-ally represented in the soil record. Thus, carbon-ates in arid regions are integrators of local con-ditions and processes, and while their precision as detailed indicators of seasonal climate may be limited, we next review the climatic signal embedded within the carbonate record.
Soil H2O and Plant H2O Uptake Effects
The two main factors that control the shape of soil moisture depth profiles are the relative proportions of seasonal precipitation and the depth of plant water uptake. At site A, Atriplex spinifera appeared to utilize water from shallow soil levels (~12–20 cm), which is likely the rea-son for the minima in the soil moisture depth
profiles during November and April. May was the wettest time of the study year at site A, and Atriplex spinifera increased its apparent depth of soil water use to ~54 cm. Atriplex conferti-folia also grows at site A and has been shown to attain its highest water use and metabolic rates during spring and before the onset of sum-mer high temperatures (Summers et al., 2009), which may explain the high soil CO2 concen-trations at site A during May, in contrast to the other sites, which exhibited soil CO2 concentra-tion maxima during August.
Isotope-based analysis of soil water uptake at site B did not yield easily interpretable results. Previous work has shown that Coleogyne ramosissima is able to utilize summer rainfall efficiently, even when temperatures are high (Gebauer and Ehleringer, 2000; Summers et al., 2009). The high xylem water d18O values mea-sured in Coleogyne may be due to evaporative loss through stem tissue during high tempera-tures, or they may reflect a very shallow depth of soil water uptake. Alternatively, Coleogyne may have utilized soil water from ~50 cm depth, because this was consistently the depth of soil water minima (except during May).
At site C, Grayia spinosa utilized soil water from ~22–48 cm depth, except for August, when depths of ~18 cm most closely resembled plant water. Artemisia is also a dominant spe-cies at site C, and while its water use was not studied at site C, Artemisia withdrew water from 28 to 47 cm at site D (supplementary Table S6 [see footnote 1]). Soil moisture values were at a minima at site C at 50 cm depth, implying that the soils were probably driest year-round at this depth, and that these low soil water levels may be related to plant water extraction.
The highest elevation (site D) had a woody Pinus monophylla forest and Artemisia triden-
0
10
20
30
40
50
60
70
80
90
100
–15 –13 –11 –9 –7 –5 –3 –1
Soil
Dept
h (c
m)
Soil Carbonate δ18O (‰ VPDB)
Site A RindsSite A FinesSite B RindsSite B FinesSite C RindsSite C FinesSite D Rinds
B0
10
20
30
40
50
60
70
80
90
100
–9 –7 –5 –3 –1 1 3 5
Soil
Dept
h (c
m)
Soil Carbonate δ13C (‰ VPDB)
Site A RindsSite A FinesSite B RindsSite B FinesSite C RindsSite C FinesSite D Rinds
A
Figure 3. (A) Soil carbonate δ13C depth profiles and (B) soil carbonate δ18O depth profiles. VPDB—Vienna Pee Dee Belemnite.
y = 0.9255x + 1.4726R2 = 0.91754
–12
–9
–6
–3
0
3
6
–12 –9 –6 –3 0 3 6
δ18O
(‰) b
y Eq
uil.
in S
oil
δ18O (‰) by Vacuum Disllaon
Figure 4. Comparison of soil water δ18O val-ues measured by direct equilibration of soil CO2 with soil H2O (in situ) vs. δ18O values of vacuum-distilled soil water. Dashed line is 1:1.
Oerter and Amundson
6 Geological Society of America Bulletin, v. 1XX, no. XX/XX
tata understory, two plants that appeared to exhibit contrasting water-use strategies. Arte-misia utilized water from 28 to 47 cm depth, while Pinus appeared to use 18O-depleted water not represented in the soil profile we sampled. A likely source of this water is a shallow allu-vial aquifer associated with nearby Rock Creek, which had temporally invariant d18O values
of –16‰ (supplementary Table S1 [see foot-note 1]). Soil moisture at site D was lowest at 25 cm depth throughout the year, likely reflect-ing ongoing extraction by Artemisia.
Soil CO2 C Isotopes and their Relation to Soil Carbonate
The isotope composition of soil CO2 varies seasonally in many environments due to varia-tions in the proportion of biologically respired CO2 and atmospheric CO2 (Cerling, 1984; Amundson et al., 1998). At low soil respiration rates (due to low soil moisture, temperatures, or sparse plant cover), the relative contribution of soil-respired CO2 to atmospheric CO2 in bulk soil CO2 is decreased. Thus, soil carbonate may have isotopic values that are indicative of spe-cific seasons and associated conditions during those seasons.
Soil carbonate forms in isotopic equilibrium with soil CO2 (Cerling, 1984). In this way, the C isotope profiles of carbonate are indicators of the mean soil respiration rates that occurred during carbonate precipitation. At any soil respiration rate, the d13C values of soil CO2 (and pedogenic carbonate formed in C isotopic equilibrium with it) can be described with a one-dimensional pro-duction/diffusion model (Cerling, 1984). Here, the upper boundary in the model is constrained by atmospheric CO2 concentration and d13C values measured ~5 cm above the soil surface, and the lower boundary is a no flux boundary with 13C values estimated to be –14‰ at site A (reflecting 100% C4 vegetation) and –25‰ at sites B–D (reflecting vegetation surveys that indicate C3-dominated vegetation; e.g., Cerling, 1984; Cerling et al., 1991). Modeled respiration rates that fit the measured values of soil CO2 were calculated to estimate soil respiration rates at the various sampling times through the year (Fig. 7). The soil CO2 production/diffusion model was also fit to the soil carbonate d13C values to derive respiration rates during carbonate precipitation using the preindustrial atmospheric CO2 con-centration of 200 ppm and d13C value of –6.5‰ (VPDB) CO2 (e.g., Francey et al., 1999) and soil-respired CO2 d13C values of –14‰ at site A and –25‰ at sites B–D (Fig. 7). This model fitting allows another comparison between modern soil conditions at various times of the year and the con-ditions that are recorded by carbonate formation.
At site A, d13C values of soil carbonate match soil respiration rates of ~0.2 mmol m–2 h–1, simi-lar to observed soil CO2 concentrations during May or perhaps November (Fig. 7A). Site B soil carbonate d13C values are fit by soil respiration rates of ~0.07 mmol m–2 h–1, reflecting August conditions (Fig. 7B). Soil carbonate at site C matches soil respiration rates of ~0.185 mmol
m–2 h–1, similar to conditions in May (Fig. 7C). Site D soil carbonate suggests high respiration rates, similar to CO2 concentrations in May, which reflect respiration rates of ~0.46 mmol m–2 h–1 (Fig. 7D).
Site B has a unique carbonate d13C pro-file, with strongly decreasing d13C values with decreasing depth above 40 cm, reaching a mini-mum value of –6‰ on clast coatings (Fig. 3A). This profile is suggestive of a very near-surface source of biological CO2 that dominated the profile during at least a portion of its develop-ment. A possible source for this CO2 could be respiration by Coleogyne ramosissima at very shallow depths, a conclusion supported by the presence of medium- and fine-size roots in the 0–4 cm surface horizon at site B (supplementary Table S2 [see footnote 1]), and very high d18O values of Coleogyne xylem water indicative of an evaporatively influenced shallow soil water source. However, none of the measured CO2 concentrations (supplementary Table S6 [see footnote 1]) or d13C values supports shallow res-piration during the year of study. Alternatively, this unique carbonate depth profile may be due to either a highly transient set of carbonate-forming conditions, or kinetically driven isotope fractionation.
Another means of linking seasonal CO2 concentrations to the soil carbonate is through matching the isotope composition of the car-bonate directly to the CO2 at different times of the year, though both methods are weakened by the short-term nature of the CO2 measure-ments versus the long-term signal embedded in the carbonate. To allow a direct comparison to ca. 4 ka pedogenic carbonate, measured soil CO2 d13C values were corrected to preindustrial atmospheric C isotope values by adding 2‰ to the measured soil CO2 d13C values (Francey et al., 1999). These corrected soil CO2 d13C values were then used to calculate d13C values of carbonate (Romanek et al., 1992) with the average soil temperatures (Deines et al., 1974) measured during the 24-h period preceding soil gas sampling (“predicted”). The closeness of fit can be quantified using the relation (Breecker et al., 2009): D13Cmeas-predict (‰) = d13Cmeasured – d13Cpredicted.
Figure 8 shows D13Cmeas-predict values for mea-sured soil CO2 d13C values through the year at the depths closest to measured soil carbon-ate d13C values. The C isotope offset between measured CO2 (1 yr) and that of carbonate that has accumulated over millennia is likely only an approximate guide to the source and timing of carbonate precipitation. In particular, the mea-sured soil CO2 at the highest elevation was much more enriched in 13C than the soil carbonate, presumably reflecting less productive conditions
0
20
40
60
80
100
–15 –10 –5 0 5 10
Soil
Dept
h (c
m)
Vacuum Dis�lledSoil Gas (corrected)
Site C
0
20
40
60
80
100
–15 –10 –5 0
Soil
Dept
h (c
m)
Vacuum Dis�lledSoil Gas (corrected)
Site D
0
20
40
60
80
100
–10 –5 0 5 10 15
Soil
Dept
h (c
m)
Soil H2O δ18O (‰ VSMOW)
Vacuum Dis�lledSoil Gas (corrected)
Site A
0
20
40
60
80
100
–10 –5 0 5 10 15
Soil
Dept
h (c
m)
Vacuum Dis�lledSoil Gas (corrected)
Site B
Figure 5. Comparison of δ18O values of soil water by vacuum distillation and direct equili-bration with soil CO2 (with 1.5‰ correction applied; see text in section on Soil CO2-H2O Oxygen Isotope Equilibrium). VSMOW— Vienna standard mean ocean water.
Climate controls on pedogenic carbonate formation and timing
Geological Society of America Bulletin, v. 1XX, no. XX/XX 7
during 2013–2014 than the long-term average. The remaining sites suggest close agreement between carbonate and CO2 through the year, except during November at site C (Fig. 8). Thus, the C isotopes in the soil carbonate do broadly reflect overall C isotope conditions at the lower sites, with about a 2‰ difference between each elevation, equivalent to a d13C lapse rate of –7‰ km–1 (Figs. 6A–6D).
Soil H2O O Isotope Dynamics and their Relation to Soil Carbonate
After infiltration, soil water is lost through transpiration and evaporation, with the lat-ter process causing an enrichment of 18O in the remaining soil water. This process results in depth trends of soil water d18O values that commonly decline nonlinearly (Barnes and Allison, 1983, 1988). With sufficient depth, soil water d18O values approach that of the average rainfall (Wang et al., 1996). Super-imposed on these long-term processes are rain-fall events that “reset” the soil water d18O pro-file to reflect that of the precipitation, which were observed most notably at sites A and C (Figs. 6E and 6G). Subsequent periods of evap-oration drive soil water d18O to higher values in the upper portions of the soil, which happened in the spring at sites A–C and at sites A and D in the summer (Figs. 6E–6H).
Along this transect, the importance of evap-oration declines with increasing elevation and increasing plant cover (Table 1), as reflected by the shape of the soil water O isotope pro-files (Figs. 6E-6H). The deepest measured soil water d18O values also decrease with increas-ing elevation, at an isotopic lapse rate of –4.5‰ km–1. The deepest soil water sampled at sites A–C had d18O values most similar to sum-mer precipitation, while deep soil water at the highest elevation (site D) reflected the yearly
weighted average of precipitation d18O values (Fig. 6).
Measured soil H2O d18O values were used to calculate d18O values of carbonate (Kim and O’Neil, 1997) with the average soil tempera-tures measured during the 24-h period preceding soil gas sampling. These “predicted” carbonate d18O values can be used to guide interpretations of the timing of O isotope equilibrium between soil H2O and soil carbonate using (Breecker
et al., 2009): D18Omeas-predic (‰) = d18Omeasured – d18Opredicted.
Figure 8 shows D18Omeas-predic values for mea-sured soil H2O d18O values at the depths closest to measured soil carbonate d18O values at the study sites through the year. As with carbon, the predicted carbonate oxygen isotope values are based on just 1 yr of soil water sampling (and precipitation). At site A, carbonate d18O values at 25 cm depth had a D18Omeas-predic nearly equal
0
20
40
60
80
100
–16–14–12–10 –8 –6 –4 –2 0 2
Soil
Dept
h (c
m)
FSite B
0
20
40
60
80
100
–16–14–12–10 –8 –6 –4 –2 0 2
Soil
Dept
h (c
m)
G
Site C
0
20
40
60
80
100
–16–14–12–10 –8 –6 –4 –2 0 2
Soil
Dept
h (c
m)
August 2013November 2013April 2014May 2014Summer Precip.Winter Precip.Wt. Ave. Precip.
HSite D
0
20
40
60
80
100
–16–14–12–10 –8 –6 –4 –2 0 2
Soil
Dept
h (c
m)
Soil H2O δ18O (‰ VSMOW)E
Site A
0
20
40
60
80
100
–18 –16 –14 –12 –10 –8
Soil
Dept
h (c
m)
Soil CO2 δ13C (‰ VPDB)
Site A
August 2013November 2013April 2014May 2014
A
0
20
40
60
80
100
–18 –16 –14 –12 –10 –8
Soil
Dept
h (c
m)
Site B
August 2013November 2013April 2014May 2014
B
0
20
40
60
80
100
–18 –16 –14 –12 –10 –8
Soil
Dept
h (c
m)
Site C
August 2013November 2013April 2014May 2014
C
0
20
40
60
80
100
–18 –16 –14 –12 –10 –8
Soil
Dept
h (c
m)
Site D
August 2013November 2013April 2014May 2014
D
Figure 6. (Left) Measured soil CO2 δ13C val-ues. (Right) Soil H2O δ18O values measured by in situ equilibration with soil CO2 (with 1.5‰ correction applied; see text in section on Soil CO2-H2O Oxygen Isotope Equilib-rium). Arrows denote temporal evolution of depth profiles between seasons (omitted in site B [δ18O] panel for clarity). Solid gray bars indicate winter precipitation δ18O val-ues, short-dashed gray bars indicate summer precipitation δ18O values, and long-dashed gray bars indicate yearly volume-weighted average precipitation δ18O values. VPDB—Vienna Pee Dee Belemnite; VSMOW— Vienna standard mean ocean water.
Oerter and Amundson
8 Geological Society of America Bulletin, v. 1XX, no. XX/XX
to zero throughout the year (except for April 2014) indicating that carbonate may form year-round in this soil (Fig. 8). The other sites had lower carbonate d18O values than that which would form in equilibrium with the soil water in the 2013–2014 interval, except for site B, which had a D18Omeas-predic nearly equal to zero in August of 2013 and in April and May of 2014 (Fig. 8), indicating late spring and summer car-bonate formation.
The unsaturated hydraulic conductivity of the soils can be estimated by the volumetric water content data, allowing an estimation of the time for a storm event’s water to infiltrate to various depths. The estimated unsaturated hydraulic conductivity for site B is ~3.95 m yr–1, which is in the range of literature values for unsaturated hydraulic conductivity in sandy soils (Freeze and Cherry, 1979). This estimate indicates that the time for water to travel from the surface to 40 cm depth (the upper depth of the highest d18O values of carbonate; Fig. 3B) is 37 d, and 65 d to reach 70 cm (the lower d18O values of carbonate). These estimates also suggest that the carbonate at site B between the surface and 40 cm depth is forming dur-ing the summer, but it is possibly recording the precipitation received during the late spring. While there is considerable uncertainty in this calculation, it is interesting that the timing of carbonate formation may not reflect the pre-cipitation conditions during that same season (Oerter et al., 2016). A key conclusion is that the O isotopes in the soil carbonate broadly reflect conditions throughout the year at the lowest-elevation site (site A), but only during the spring and summer at the midelevation sites (sites B and C).
The period of study during 2013–2014 was a time of decreased precipitation, with Palmer drought severity ratings of ≤–3 (NCDC-NOAA, 2015). While these below-average precipitation conditions likely do not reflect average climate conditions during the 3.8 k.y. of carbonate for-mation in these soils, the comparison of climate conditions between sites along the transect may still reflect relative differences and contrasts that are applicable and relevant even in higher-pre-cipitation years.
Implications for the Interpretation of Pedogenic Carbonate in the Paleorecord
This project was undertaken to better under-stand the conditions and seasons in which pedo genic carbonate forms in North American deserts, and when placed into context with pre-vious research, offers some insights into the inter pre tation of C and O stable isotopes of car-bonate found in paleosols. The following dis-
0
20
40
60
80
100
–15 –13 –11 –9 –7 –5
Soil
Dept
h (c
m)
δ13C (‰ VPDB)
0.15 mmol0.2 mmol0.25 mmolcarbonate
A
0
20
40
60
80
100
–15 –13 –11 –9 –7 –5
Soil
Dept
h (c
m)
δ13C (‰ VPDB)
0.058 mmol0.070 mmol0.086 mmolcarbonate
B
0
20
40
60
80
100
–20 –18 –16 –14 –12 –10 –8 –6
Soil
Dept
h (c
m)
δ13C (‰ VPDB)
0.120 mmol0.185 mmol0.250 mmolcarbonate
C
0
20
40
60
80
100
–21 –19 –17 –15 –13 –11 –9 –7 –5
Soil
Dept
h (c
m)
δ13C (‰ VPDB)
0.370 mmol0.460 mmol0.550 mmolcarbonate
D
0
0.1
0.2
0.3
0.4
0.5
Resp
irao
n Ra
te(m
mol
m–2
h–1
)
Aug 201
3Nov 2
013
Apr 201
4M
ay 20
14
Soil CO2 Soil Carbonate
May
2013
Site A
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
Resp
irao
n Ra
te(m
mol
m–2
h–1
)
Aug 201
3Nov 2
013
Apr 201
4M
ay 20
14
Soil CO2Soil Carbonate
May
2013
Site B
0
0.05
0.1
0.15
0.2
0.25
0.3
Resp
irao
n Ra
te(m
mol
m–2
h–1
)
Aug 201
3Nov 2
013
Apr 201
4M
ay 20
14
Soil CO2
Soil Carbonate
May
2013
Site C
0
0.1
0.2
0.3
0.4
0.5
0.6
Resp
irao
n Ra
te(m
mol
m–2
h–1
)
Aug 201
3Nov 2
013
Apr 201
4M
ay 20
14
Soil CO2
Soil Carbonate
May
2013
Site D
Figure 7. (Left) Modeled respiration rates (mmol m–2 h–1) compared to δ13C values of pedogenic carbonate. (Right) Respiration rates modeled to fit soil CO2 δ13C values (solid line is best-fit value; dashed lines are maximum and minimum values from left panels), compared to res-piration rates modeled for δ13C values from carbonate. VPDB—Vienna Pee Dee Belemnite.
Climate controls on pedogenic carbonate formation and timing
Geological Society of America Bulletin, v. 1XX, no. XX/XX 9
cussion is specifically geared toward paleosols that formed in comparable desert environments to those examined here, and is not as relevant for the more biologically active end members of carbonate-forming soil environments.
The most obvious conclusion, one first articu-lated by Quade et al. (1989), and elaborated by others (e.g., Amundson et al., 1989; Breecker et al., 2009), is that in arid environments, the d13C value of pedogenic carbonate at any soil depth is almost invariably not a direct indica-tor of the isotopic composition of the biologi-cal source of CO2 (and plant community com-position), but is instead a soil CO2 barometer, reflecting differing soil respiration rates and soil CO2 concentrations (S[z] values). Second, there is commonly considerable variation in d13C val-ues of soil CO2 and carbonate with depth (for example, see Figs. 3 and 6), and only in more biologically productive environments do soil C isotope values reach constant values at shallow depths—but even those constant values may not reflect the isotopic composition of the biological sources (Fig. 3). The most challenging (or entic-ing) issue in paleosol C isotope interpretation is that the carbonate represents unknown values of both (1) soil respiration rates and (2) atmo-spheric CO2 concentrations. Numerous methods have been devised to use paleosol chemistry, depth to carbonate layers, etc., as methods of
constraining paleo–soil CO2 concentrations (and hence soil respiration rates), allowing researchers to estimate the paleoatmospheric CO2 concentrations (e.g., Retallack, 2009).
As mentioned already, the d13C values of soil carbonate can be highly variable within a soil profile, and they seldom stabilize below 30–50 cm, as is sometimes assumed in paleocli-mate reconstructions (see Fig. 3). For example, the carbonate C isotope composition in some profiles examined here varied by up to 8‰, which for a given assumed S(z) value results in atmospheric CO2 concentration estimates that range over 1000 ppm. From growing research on modern soils in arid settings (e.g., this study; Quade et al., 1989, 2007; Peters et al., 2013), complete depth profiles of carbonate d13C values will be important in order to understand more about the C and O isotope history of the soil. In that vein, and given the ease and low cost of carbonate analyses, it seems that more advanced uses of paleosol data will rely on the acquisition of data that are comparable to those examined in modern soils. Utilization of carbonate samples from throughout a paleosol depth profile will result in two key benefits: (1) It will eliminate erroneous assumptions about depth-indepen-dent C isotope values (an error that may result in overestimating paleoatmospheric CO2 con-centrations if the assumed S[z] value is too high
at the sampled depth), and (2) the profile can then be subjected to curve fitting (via Cerling’s [1984] production/diffusion model) to better constrain the nature of the soil CO2 produc-tion rate, which is important paleoclimatic information.
Presently, there are few paleosol C isotope profiles in pre-Quaternary soils (e.g., Behrens-meyer et al., 2007). Most paleo-CO2 reconstruc-tions rely on one or a few samples per paleo-sol, and depths (below the presumed original soil surface) are rarely given. Certainly, many paleosols have been truncated, making depth assignments difficult. However, without that information, based on modern soils, it would seem difficult to use carbonate data to accu-rately constrain paleoatmospheric CO2 concen-tration calculations. While it is unlikely that all paleosol C isotope depth profiles are distinctive enough to use fitting algorithms to confidently assign constraints on respiration rates and atmo-spheric CO2, depth profiles will do much to enhance the validity of various interpretations. Detailed carbonate depth profiles will allow researchers to test whether assumptions about soil CO2 concentrations produce analytical isotope profiles that are consistent with the observed data.
CONCLUSIONS
This climosequence in the western Great Basin indicates that the C and O stable isotope compositions of soil CO2, H2O, and pedogenic carbonate change systematically with elevation and climate. At the valley floor (1482 m above sea level [masl]), soil carbonate d13C and d18O values reflect measured soil CO2 d13C and H2O d18O values, respectively, nearly year-round, especially in the 25–50 cm soil depth interval. At the midelevation sites (1745 and 2140 masl), spring through summer conditions most closely mirrored the d13C and d18O values of pedogenic carbonate. At the highest site (2602 masl), the C and O isotopes of carbonate never reflected the measured soil CO2 and H2O conditions dur-ing the sampling period, though modeled soil respiration rates suggest that spring season conditions were most conducive to carbonate formation. This relatively diverse variability in the timing of carbonate formation is indica-tive of the complexity of the soil C and O iso-tope systems in arid environments. Despite the potential differences in seasonality of carbonate precipitation, each elevation has unique C and O isotope patterns that track changing climate conditions with elevation.
In this location, the warmest times of the year (late spring through summer) did not have the driest soil moisture conditions and did not always coincide with the lowest soil CO2 con-
–6–4–2
024
Δ13 C
m-p
(‰) 25cm 50cm 100cm
–6–4–2
024
Δ13 C
m-p
(‰)
Site B
Site A
–6–2–4
224
Δ13 C
m-p
(‰)
Site C
–10–8–6–4–2
02
Δ13 C
m-p
(‰) Site D
May 2013
Aug. 2013
Nov. 2013
Apr. 2014
May 2014
–8–4
048
Δ18 O
m-p
(‰) 25cm 50cm 100cm
–8–4
048
Δ18 O
m-p
(‰)
Site B
–8
–4
0
Δ18 O
m-p
(‰)
Site C
–10
–5
0
Δ18 O
m-p
(‰) Site D
May 2013
Aug. 2013
Nov. 2013
Apr. 2014
May 2014
Site A
Figure 8. (Left) Comparison of the difference between measured (“m”) carbonate δ13C values and δ13C values calculated (predicted: “p”) to be in C isotope equilibrium with soil CO2. (Right) Similar comparison of measured (“m”) carbonate δ18O values compared to δ18O values calcu-lated (predicted: “p”) to be in O isotope equilibrium with soil H2O. Values are for soil depths listed at top, except: site C triangles are 75 cm depth, and site D squares are 37 cm depth.
Oerter and Amundson
10 Geological Society of America Bulletin, v. 1XX, no. XX/XX
centrations. Soil CO2 levels during the most likely times of carbonate formation at the lower-elevation sites were between ~450 and ~1800 ppm, reaffirming the conclusion that S(z) values of 5000 ppm are too high for paleo-CO2 esti-mates (Breecker et al., 2009). However, while the abundance of aridic paleosols in the geo-logical record is poorly known, it seems that detailed C isotope depth profiles of paleosol carbonate will provide improved constraints on the paleoclimate information contained therein.
ACKNOWLEDGMENTS
We gratefully acknowledge Marco Pfeiffer for his contributions in the field, Todd Dawson’s keen inter-est and support, and Paul Brooks and Wenbo Yang for their expertise in the laboratory. The thoughtful reviews by Lee Nordt and Dan Breecker, and the edi-torial input of Ken MacLeod, improved this paper. A Geological Society of America Graduate Student Re-search Grant helped fund this work.
REFERENCES CITED
Amundson, R., Doner, H.E., Chadwick, O.A., and Sowers, J.M., 1989, The stable isotope chemistry of pedogenic carbonates at Kyle Canyon, Nevada: Soil Science Soci-ety of America Journal, v. 53, no. 1, p. 201–210, doi: 10 .2136 /sssaj1989 .03615995005300010037x .
Amundson, R., Stern, L., Baisden, T., and Wang, Y., 1998, The isotopic composition of soil and soil-respired CO2: Geoderma, v. 82, no. 1–3, p. 83–114, doi: 10 .1016 /S0016 -7061 (97)00098 -0 .
Araguás-Araguás, L., Rozanski, K., Gonfiantini, R., and Louvat, D., 1995, Isotope effects accompanying vacuum extraction of soil water for stable isotope analyses: Journal of Hydrology (Amsterdam), v. 168, p. 159–171, doi: 10 .1016 /0022 -1694 (94)02636 -P .
Arkley, R.J., 1963, Calculation of carbonate and water movement in soil from climatic data: Soil Science, v. 96, no. 4, p. 239–248, doi: 10 .1097 /00010694 -196310000 -00003 .
Barnes, C.J., and Allison, G.B., 1983, The distribution of deuterium and O18 in dry soils. 1: Theory: Journal of Hydrology, v. 60, no. 1–4, p. 141–156.
Barnes, C.J., and Allison, G.B., 1988, Tracing of water-movement in the unsaturated zone using stable iso-topes of hydrogen and oxygen: Journal of Hydrology (Amsterdam), v. 100, no. 1–3, p. 143–176, doi: 10 .1016 /0022 -1694 (88)90184 -9 .
Behrensmeyer, A.K., Quade, J., Cerling, T.E., Kappelman, J., Khan, I.A., Copeland, P., Roe, L., Hicks, J., Stubble-field, P., and Willis, B.J., 2007, The structure and rate of late Miocene expansion of C4 plants: Evidence from lateral variation in stable isotopes in paleosols of the Siwalik Group, northern Pakistan: Geological Society of America Bulletin, v. 119, no. 11–12, p. 1486–1505, doi: 10 .1130 /B26064 .1 .
Breecker, D., and Sharp, Z.D., 2008, A field and laboratory method for monitoring the concentration and isotopic composition of soil CO2: Rapid Communications in Mass Spectrometry, v. 22, no. 4, p. 449–454, doi: 10 .1002 /rcm .3382 .
Breecker, D.O., Sharp, Z.D., and McFadden, L.D., 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, v. 121, no. 3–4, p. 630–640, doi: 10 .1130 /B26413 .1 .
Brenninkmeijer, C.A.M., Kraft, P., and Mook, W.G., 1983, Oxygen isotope fractionation between CO2 and H2O: Chemical Geology, v. 41, p. 181–190, doi: 10 .1016 /S0009 -2541 (83)80015 -1 .
Cerling, T.E., 1984, The stable isotopic composition of mod-ern soil carbonate and its relationship to climate: Earth
and Planetary Science Letters, v. 71, no. 2, p. 229–240, doi: 10 .1016 /0012 -821X (84)90089 -X .
Cerling, T., Solomon, D.K., Quade, J., and Bowman, J., 1991, On the isotopic composition of carbon in soil carbon dioxide: Geochimica et Cosmochimica Acta, v. 55, p. 3403–3405.
Coplen, T.B., 1994, Reporting of stable hydrogen, carbon, and oxygen isotopic abundances (technical report): Pure and Applied Chemistry, v. 66, no. 2, p. 273–276, doi: 10 .1351 /pac199466020273 .
Dansgaard, W., 1964, Stable isotopes in precipitation: Tellus, v. 16, no. 4, p. 436–468, doi: 10 .1111 /j .2153 -3490 .1964 .tb00181 .x .
Deines, P., Langmuir, D., and Harmon, R.S., 1974, Stable carbon isotope ratios and the existence of a gas phase in the evolution of carbonate ground waters: Geo-chimica et Cosmochimica Acta, v. 38, p. 1147–1164, doi: 10 .1016 /0016 -7037 (74)90010 -6 .
Francey, R., Allison, C., Etheridge, D., Trudinger, C., Enting, I., Leuenberger, M., Langenfelds, R., Michel, E., and Steele, L., 1999, A 1000-year high precision record of d13C in atmospheric CO2: Tellus, ser. B, Chemical and Physical Meteorology, v. 51, no. 2, p. 170–193, doi: 10 .1034 /j .1600 -0889 .1999 .t01 -1 -00005 .x .
Freeze, R.A., and Cherry, J., 1979, Groundwater: Engle-wood Cliffs, New Jersey, Prentice-Hall, 604 p.
Gat, J.R., 1996, Oxygen and hydrogen isotopes in the hydro-logic cycle: Annual Review of Earth and Planetary Sci-ences, v. 24, p. 225–262, doi: 10 .1146 /annurev .earth .24 .1 .225 .
Gebauer, R.L., and Ehleringer, J.R., 2000, Water and nitro-gen uptake patterns following moisture pulses in a cold desert community: Ecology, v. 81, no. 5, p. 1415–1424, doi: 10 .1890 /0012 -9658 (2000)081 [1415: WANUPF]2 .0 .CO;2 .
Gile, L.H., Peterson, F.F., and Grossman, R.B., 1966, Morphological and genetic sequences of carbonate accumulation in desert soils: Soil Science, v. 101, no. 5, p. 347–360, doi: 10 .1097 /00010694 -196605000 -00001 .
Grayson, D.K., 2011, The Great Basin: A Natural Prehis-tory: Berkeley, California, University of California Press, 432 p.
Hall, C.A., 1991, Natural History of the White-Inyo Range, Eastern California: Berkeley, California, University of California Press, 560 p.
Harden, J.W., Slate, J.L., Lamothe, P., Chadwick, O.A., Pendall, E., and Gillespie, A.R., 1991, Soil Formation on the Trail Canyon Alluvial Fan, Fish Lake Valley, Nevada: U.S. Geological Survey Open-File Report 91–291, 29 p.
Jenny, H., 1941, Factors of Soil Formation: A System of Quantitative Pedology: New York, Courier Dover Pub-lications, 191 p.
Jenny, H., and Leonard, C., 1934, Functional relation-ships between soil properties and rainfall: Soil Sci-ence, v. 38, no. 5, p. 363–382, doi: 10 .1097 /00010694 -193411000 -00004 .
Kim, S.T., and O’Neil, J.R., 1997, Equilibrium and nonequi-librium oxygen isotope effects in synthetic carbonates: Geochimica et Cosmochimica Acta, v. 61, no. 16, p. 3461–3475, doi: 10 .1016 /S0016 -7037 (97)00169 -5 .
Lechler, A.R., and Galewsky, J., 2013, Refining paleoaltim-etry reconstructions of the Sierra Nevada, using air par-cel trajectories: Geology, v. 41, no. 2, p. 259–262, doi: 10 .1130 /G33553 .1 .
National Climatic Data Center–National Oceanic and Atmo spheric Administration (NCDC-NOAA), 2015: http:// www .ncdc .noaa .gov /sotc /drought (last accessed 1 April 2016).
Oerter, E.J., Sharp, W.D., Oster, J.L., Ebeling, A., Valley, J.W., Kozdon, R., Orland, I.J., Hellstrom, J., Wood-head, J.D., Hergt, J.M., Chadwick, O.A., and Amund-son, R., 2016, Pedothem carbonates reveal anomalous North American atmospheric circulation 70,000–55,000 years ago: Proceedings of the National Acad-emy of Sciences, January 11, 2016, doi: 10 .1073 /pnas .1515478113 .
Passey, B.H., Levin, N.E., Cerling, T.E., Brown, F.H., and Eiler, J.M., 2010, High-temperature environments of human evolution in East Africa based on bond order-ing in paleosol carbonates: Proceedings of the National
Academy of Sciences of the United States of America, v. 107, no. 25, p. 11,245–11,249, doi: 10 .1073 /pnas .1001824107 .
Pendall, E.G., Harden, J.W., Trubmore, S.E., and Chadwick, O.A., 1994, Isotopic approach to soil carbonate dy-namics and implications for paleoclimatic interpreta-tions: Quaternary Research, v. 42, no. 1, p. 60–71, doi: 10 .1006 /qres .1994 .1054 .
Peters, N.A., Huntington, K.W., and Hoke, G.D., 2013, Hot or not? Impact of seasonally variable soil carbonate formation on paleotemperature and O-isotope records from clumped isotope thermometry: Earth and Plan-etary Science Letters, v. 361, p. 208–218, doi: 10 .1016 /j .epsl .2012 .10 .024 .
Phillips, D.L., and Gregg, J.W., 2001, Uncertainty in source partitioning using stable isotopes: Oecologia, v. 127, no. 2, p. 171–179, doi: 10 .1007 /s004420000578 .
Quade, J., Cerling, T.E., and Bowman, J.R., 1989, Sys-tematic variations in the carbon and oxygen isotopic composition of pedogenic carbonate along elevation transects in the southern Great Basin, United States: Geological Society of America Bulletin, v. 101, no. 4, p. 464–475, doi: 10 .1130 /0016 -7606 (1989)101 <0464: SVITCA>2 .3 .CO;2 .
Quade, J., Rech, J.A., Latorre, C., Betancourt, J.L., Glee-son, E., and Kalin, M.T.K., 2007, Soils at the hyperarid margin: The isotopic composition of soil carbonate from the Atacama Desert, northern Chile: Geochimica et Cosmochimica Acta, v. 71, p. 3772–3795, doi: 10 .1016 /j .gca .2007 .02 .016 .
Reheis, M. C., and Block, D., 2007, Surficial Geologic Map and Geochronologic Database, Fish Lake Valley, Es-meralda County, Nevada, and Mono County, Califor-nia: U.S. Geological Survey Data Series 277, 1 sheet.
Retallack, G.J., 2009, Refining a pedogenic carbonate CO2 paleobarometer to quantify a middle Miocene green-house spike: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 281, no. 1, p. 57–65, doi: 10 .1016 /j .palaeo .2009 .07 .011 .
Romanek, C.S., Grossman, E.L., and Morse, J.W., 1992, Carbon isotopic fractionation in synthetic aragonite and calcite: Effects of temperature and precipitation rate: Geochimica et Cosmochimica Acta, v. 56, no. 1, p. 419–430, doi: 10 .1016 /0016 -7037 (92)90142 -6 .
Royer, D.L., 1999, Depth to pedogenic carbonate horizon as a paleoprecipitation indicator?: Geology, v. 27, no. 12, p. 1123–1126, doi: 10 .1130 /0091 -7613 (1999)027 <1123: DTPCHA>2 .3 .CO;2 .
Schoeneberger, P.J., Wysocki, D.A., Benham, E.C., and Soil Survey Staff, 2012, Field Book for Describing and Sampling Soils, Version 3.0: Lincoln, Nebraska, Na-tional Soil Survey Center, Natural Resources Conser-vation Service, 100 p.
Soil Survey Staff, 1999, Soil Taxonomy: A Basic System of Soil Classification for Making and Interpreting Soil Surveys: Washington, D.C., Natural Resources Conser-vation Service, U.S. Department of Agriculture, 871 p.
Stern, L., Baisden, W.T., and Amundson, R., 1999, Pro-cesses controlling the oxygen isotope ratio of soil CO2: Analytic and numerical modeling: Geochimica et Cos-mochimica Acta, v. 63, no. 6, p. 799–814, doi: 10 .1016 /S0016 -7037 (98)00293 -2 .
Summers, H., Smith, B., and Hansen, L., 2009, Compari-son of respiratory and growth characteristics of two co-occurring shrubs from a cold desert, Coleogyne ramosissima (blackbrush) and Atriplex confertifolia (shadscale): Journal of Arid Environments, v. 73, no. 1, p. 1–6, doi: 10 .1016 /j .jaridenv .2008 .09 .010 .
Wang, Y., McDonald, E., Amundson, R., McFadden, L., and Chadwick, O., 1996, An isotopic study of soils in chronological sequences of alluvial deposits, Provi-dence Mountains, California: Geological Society of America Bulletin, v. 108, no. 4, p. 379–391, doi: 10 .1130 /0016 -7606 (1996)108 <0379: AISOSI>2 .3 .CO;2 .
Science editor: chriStian Koeberl aSSociate editor: Kenneth G. Macleod
ManuScript received 18 June 2015 reviSed ManuScript received 26 deceMber 2015 ManuScript accepted 13 January 2016
