SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 12 March 2007
Published in Soil Sci Soc Am J 71:397-405 (2007)
DOI: 10.2136/sssaj2006.0061
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Eshel, G.
Right arrow Articles by Singer, M. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Eshel, G.
Right arrow Articles by Singer, M. J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Eshel, G.
Right arrow Articles by Singer, M. J.
Related Collections
Right arrow Water Quality
Right arrow Dryland Soils
Right arrow Carbon Sequestration

SOIL & WATER MANAGEMENT & CONSERVATION

Total Soil Carbon and Water Quality: An Implication for Carbon Sequestration

Gil Eshela,*, Pinchas Finea and Michael J. Singerb

a Inst. of Soil, Water and Environ. Sciences ARO, The Volcani Center, Bet Dagan 50250, Israel
b Dep. of Land, Air and Water Resources, Univ. of California, Davis, CA 95616

* Corresponding author (geshel{at}volcani.agri.gov.il).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Carbon sequestration in soil has been suggested as a means of reducing the rate of increase of atmospheric CO2. Most soil science research has been on soil organic carbon (SOC) sequestration but in arid and semiarid climates, soil inorganic carbon (SIC) may offer another option for C sequestration. A field study was conducted in Bakersfield, CA, to determine if irrigation water quality (fresh water [FW] vs. treated effluent [TE]) affected the distribution and amount of SIC and SOC in the upper 4 m of soil and parent material compared to a nonirrigated (NI) field. Significant carbonate depletions were found in the upper 2 m in both irrigated fields compared with the NI. Differences in carbonate content between irrigated fields were also related to soil texture. Total carbonate and clay-size carbonate were more abundant at the sites irrigated with TE than at the sites irrigated with FW, indicating that the TE had inhibited carbonate dissolution. Based on stable isotope analyses (13C and 18O) and radiocarbon dating, we estimated that irrigation for >75 yr sequestered about 7.15 kg m–2 (4 m)–1 of SIC under FW and between 0.9 and 2.4 kg m–2 (4 m)–1 under TE, if carbonate dissolution is C sequestration. Adding C loss due to SOC decomposition to the SIC sequestration, the fields may be a source for 8.8 and 17.4 to 15.9 kg m–2 (4 m)–1 of C under FW and TE, respectively. This study provides some of the first evidence of how water quality affects the C budget in an arid region.

Abbreviations: FW, freshwater • NI, nonirrigated • PSD, particle size distribution • SIC, soil inorganic carbon • SOC, soil organic carbon • TE, treated effluent


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soils play an important role in the global C cycle. They are the third largest active C pool (1550 Pg of organic C and 750 Pg inorganic C to a 1-m depth) after the hydrosphere (38 000 Pg) and the lithosphere (5000 Pg) (Lal, 2001a). The role of soil as either a source or a sink for greenhouse gases in general, and that of CO2 in particular, has been the focus of many studies (e.g., Bouwman, 1990; Houghton, 1995; Gerzabek et al., 1997; Schlesinger, 2000; Drees et al., 2001; Lal, 2001b). Cultivation is a practice that generates significant reductions in SOC (Stevenson, 1994; Lal, 2001c). The loss is related to the increase in soil aeration during cultivation and to the wetting and drying cycles caused by irrigation (Stevenson, 1994). On the other hand, De Clerck et al. (2003) reported a general trend of increasing total C content of a large number of California soils after 60 yr of cultivation. In southern California, they found a small but not statistically significant decrease in C content. Adoption of specific practices (e.g., organic farming, minimum tillage, no-tillage) in irrigated agriculture has produced a diversity of results with respect to net C sequestration (Schlesinger, 2000; Entry et al., 2002, 2004; Lee et al., 2006). Much less is known about the cycle of SIC or methods to enhance SIC sequestration.

The content of SOC in irrigated soils of semiarid and arid regions is low due to low formation rates and relatively high decomposition rates produced by high temperatures and sufficient water from irrigation. In these regions, net accumulation of SOC under irrigation is likely to be difficult (Lee et al., 2006), and C sequestration as inorganic forms may prove to be a viable alternative (Lal and Kimble, 2000a, 2000b; Lal, 2001a; Entry et al., 2004).

Soils from semiarid and arid regions often contain Ca and Mg carbonates. Carbonate accumulation in soil profiles can result from all or part of the following: redistribution of detrital carbonate or redistribution of exogenous material (e.g., dust-borne carbonate) and in situ secondary carbonate formation due to primary mineral weathering. The redistribution can result from dissolution and reprecipitation of inherited carbonate minerals due to cycles of soil wetting and drying (Lippmann, 1973; Knight, 1991), or by translocation as colloidal suspensions with the soil solution (Dan, 1983; Baghernejad and Dalrymple, 1993; Neaman et al., 2000). Neither of these processes produces a net sequestration of atmospheric C. Irrigation of arid-zone soils is bound to alter the size, properties, and location of the SIC pool as influenced by irrigation amount and method, water quality, drainage and leaching, plowing, application of fertilizers, and other soil amendments (e.g., gypsum to Aridisols and liming material to Ultisols and Oxisols [Lal, 2001b]). Secondary precipitation of carbonates with divalent cations (e.g., Ca and Mg) is possible under limited leaching and dry conditions (Dan, 1983; Sposito, 1989; Nordt et al., 2000). The question remains whether irrigation management can increase net C sequestration.

The answer to the irrigation question is unclear, in part, because there is no agreement in the literature about the role of carbonate precipitation in sequestration. There is literature that argues that precipitation of carbonate solid should be considered as sequestration (e.g., Lal and Kimble, 2000a, 2000b; Eswaran et al., 2000; Scharpenseel et al., 2000; Ryskov et al., 2000; Mermut et al., 2000; Emmerich, 2003; Entry et al., 2004) and others argue that dissolution of solid carbonate should be considered sequestration (e.g., Suarez, 2000; Nordt et al., 2000; Schlesinger, 2000; Drees et al., 2001). Considering the overall reaction of calcite equilibrium, it is clear that while the acidity source is H2CO3, dissolving one mole of calcite will consume one mole of CO2 from the gas phase and precipitation of one mole of calcite will produce one mole of CO2:

Formula 1[1]
Because the partial pressure of CO2 (pCO2) in the soil is 10 to 100 times higher than in the atmosphere (Amundson and Davidson, 1990) and the pH of arid soils is typically above 7 under pCO2 = 3.5 kPa (Eshel, 2005), it is reasonable to assume that H2CO3 is the major acidity source involved in carbonate mineral dissolution in calcareous soils.

Several earlier studies have suggested that stable isotope ratios of C and O can be useful tools for identification of pedogenic carbonate in soils and to determine their proportion of the total carbonate (Magaritz and Amiel, 1980, 1981; Rabenhorst et al., 1984; Amundson and Lund, 1987; West et al., 1988; Mermut et al., 2000). Magaritz and Amiel (1981) used stable C analysis to estimate that about 500 Mg ha–1 of carbonate was removed from the upper 2.5 m of a calcareous soil profile containing 16 000 Mg ha–1 carbonate during 40 yr of irrigation with groundwater in Israel. They also estimated that about 800 Mg ha–1 was dissolved and reprecipitated in the upper 2.5 m of the soil profile. Amundson and Smith (1988) used stable O isotope analysis to estimate that about 56 Mg ha–1 carbonates were newly formed during 8 yr of well-water irrigation on calcareous soils in the San Joaquin Valley in California. McCaslin and Lee-Rodriquez (1979) compared two adjacent fields in New Mexico that had been irrigated for about 40 yr with TE and with FW and found more carbonate (1370 Mg ha–1) in the upper 1 m of the TE field but they did not discuss the carbonate origin (detrital vs. pedogenic). Nimkar et al. (1992) found an accumulation of 337 Mg ha–1 of carbonates in the upper 1 m of the soil profile following 7 yr of irrigation on calcareous soils with rather saline water from the Purna River in India.

The objectives of this study were (i) to study the impact of irrigation water quality on the soil C pool with emphasis on the SIC, and (ii) to estimate a total soil C budget under the two irrigation management systems.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The Study Sites
Two adjacent fields in Bakersfield, CA, mapped as Kimberlina fine sandy loam (coarse-loamy, mixed, superactive, calcareous, thermic Typic Torriorthent, Soil Survey Staff, 1988) were selected for the study. The Kimberlina soils are young, simple, A–C profiles formed from recent alluvium on young alluvial fans originating from local granitic and sedimentary rocks (Soil Survey Staff, 1988). The landscape and soils sampled appear to be young, probably <10 000 yr BP based on the absence of soil horizons. One field was irrigated with TE and the second one with FW. Both fields have been continuously cultivated since the early 1900s under conventional agriculture practices. The TE-irrigated field was a part of a 2000-ha farm that has been furrow irrigated with reclaimed municipal wastewater from the city of Bakersfield for >75 yr. Sludge from the wastewater treatment plant was also applied on this field. During this period of time, crops for nonhuman consumption including hay, cotton (Gossypium hirsutum L.), and alfalfa (Medicago sativa L.) have been grown. The FW field, which is located <1 km from the TE-irrigated field, had not received any reclaimed wastewater and had been furrow irrigated with groundwater. Samples were also collected from a "natural" site near the FW field that, to the best of our knowledge, had never been cultivated or irrigated (NI). The vegetation at the NI site consists of annual grasses and a large black walnut tree (Juglans nigra L.). An earlier soil-quality study conducted in these two fields by Wang et al. (2003) showed that the TE-irrigated field had slightly more soil compaction and less Mg in the upper 30 cm of the profile relative to the FW-irrigated field, but the two were otherwise very similar.

Soil Sampling and Analysis
Fifteen profiles were hand augured in the three fields. Seven profiles were sampled in the TE-irrigated field, six in the FW-irrigated field, and two in the NI site. Originally we planned to auger to 4 m but, in some cases, a perched water table restricted our sampling depth. Samples were collected every 20 to 50 cm.

Soil samples were air dried and lightly crushed to pass through a 2-mm sieve. Soil moisture content was determined by weight loss after 24 h at 105°C. Particle size distributions (PSD), as volume percentages, were determined with a laser particle size analyzer (Coulter LS-230) (Eshel et al., 2004). The PSD analyses were made before and after carbonate removal with 5% HCl. Total carbonate content was measured by the manometric method following the addition of dilute HCl to dissolve the carbonates (Nelson, 1982). The C and O stable isotopic analysis of the SIC was preformed on an isotope ratio mass spectrometer (Finnigan MAT 251) at the Stanley V. Margolis Stable Isotope Laboratory at the Department of Geology at the University of California, Davis. The isotope ratios for the SIC were expressed as {delta}13Ccarb and {delta}18Ocarb values:

Formula 2[2]

Formula 3[3]

The reference used was the international Pee Dee Belemnite (PDB) standard.

Selected samples were sent to Beta Analytic Inc. (Miami, FL) for SIC radiocarbon age dating. The total soil organic matter and the SOC stable isotope ratio ({delta}13Corg) signature were analyzed by isotope ratio mass spectrometer (Europa Hydra 20/20) at the University of California-Davis stable isotope facility after SIC removal from the sample by acid fumigation (Harris et al., 2001). To convert SIC and SOC content in grams per kilogram to kilograms per square meter, we used a bulk density of 1.13 Mg m–3 for the 0- to 20-cm depth, which is the mean value for 200 samples collected from the 0- to 15-cm depth by Wang et al. (2003). We measured bulk density for deeper layers on samples extracted with a hand auger (Table 1). The samples had a mean bulk density 9% greater than the values of 1.44 to 150 g cm–3 reported in the soil survey report (Soil Survey Staff, 1988). To deal with the inconsistency in the depth of the sampled profiles, we analyzed the effect of water quality in three different layers: 0 to 2 m, 2 to 4 m, and 0 to 4 m. The average SIC and SOC contents were calculated for each layer and one-way ANOVA was applied to test the null hypothesis that irrigation and the water quality have no effect on SIC and SOC content using JMP Version 5.1 (SAS Institute, Cary, NC).


View this table:
[in this window]
[in a new window]

  		Table 1. The bulk density of clods collected from the hand auger samples of soils irrigated with fresh water (FW) and treated effluent (TE).

 

   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Carbonate Content and its Distribution
Spatial variability of total carbonate (mass percentage) was large among the sampling sites (Fig. 1 ). In five of the seven profiles sampled in the TE field, the most carbonate was found between 2.5- and 3-m depth. Two profiles could not be sampled below 2 m because of saturated conditions at this depth. It appeared that the saturated zone was caused by an indurated layer. Five of the six FW sites had nearly identical carbonate distributions (Fig. 1). Only FW-6 had a distinct bimodal accumulation of carbonate near the surface and at 2 m. The two NI sample locations had similar carbonate distributions in the upper 2 m.


Figure 1
View larger version (31K):
[in this window]
[in a new window]

  		Fig. 1. Carbonate distribution in soils irrigated with fresh water (FW), treated effluent (TE), and nonirrigated (NI). Filled symbols represent samples that were dated by 14C analysis.

 
In the two irrigated fields, the average total carbonate (as CaCO3) content in the root zone (0–2 m) was significantly lower than the carbonate in the NI field (Table 2). No significant difference in carbonate content between the TE- and FW-irrigated fields was found within the 0- to 2-m depth. The depth at which the majority of the carbonate was found in the TE-irrigated field (2–4 m) was well below the zone of most active root growth. In the field irrigated with FW and in the nonirrigated field, less carbonate accumulated and most was at a shallower depth than the TE-irrigated field (Table 2).


View this table:
[in this window]
[in a new window]

  		Table 2. Mean carbonate (as CaCO3) content from sites irrigated with treated effluent (TE) or fresh water (FW) and nonirrigated (NI) sites at Bakersfield, CA.

 
The total carbonate in the 0- to 4-m depth was 59.8 kg m–2 greater under TE irrigation and 59.5 kg m–2 lower under FW irrigation than the NI profile. The carbonate removal we found under the FW irrigation is in the range of carbonate removal from the upper 2.5 m of a soil profile following 40 yr of irrigation with groundwater in Israel (Magaritz and Amiel, 1981). The carbonate amounts and vertical distribution in the NI sites are similar to what was found in the Kimberlina profile (nonirrigated, annual grass) described in the soil survey report (Soil Survey Staff, 1988). These data suggest that the irrigation water quality affects the amount and vertical distribution of SIC in the soil profile differently (Table 2, Fig. 1).

Redistribution is not sequestration, but, depending on one's viewpoint, additions or depletions of carbonate may be considered sequestration. To assess the sources of carbonate and the impact of irrigation, we measured some morphological characteristics, the isotopic composition of both the SIC and SOC, and the radiocarbon age of the soil carbonate.

Carbonate Morphology and Size
During the soil sampling, we observed that the carbonate appeared in three main morphological forms: (i) as a cemented hard pan at a depth of approximately 3 m in the TE-irrigated field, (ii) as sporadic, finger-shaped calcite concentrations in the FW-irrigated field and NI field at a depth of 2 m (Profiles FW-5 and NI-1, respectively), or (iii) as part of the matrix at depths of 0.5 m in all the fields and at depths of 2 to 3 m in the TE-irrigated field. Generally, the carbonates tended to concentrate in finer textured layers. We interpret the cemented layers to be pedogenic carbonate that may be redistributed or "new" carbonate. Carbonates appearing as part of the matrix may be either detrital carbonate inherited with the parent material or formed as part of pedogenesis.

We used the PSD before and after carbonate removal as a measure of pedogenic carbonate. We assumed that detrital carbonate was in coarser grains than pedogenic carbonate and was distributed mainly as individual grains. Because pedogenic carbonate precipitates in pores and on particle surfaces, we assumed that pedogenic carbonate would act mainly as a cementing agent of soil particles. Increases in the content of the fine fractions after carbonate dissolution indicates that carbonate is a cementing agent, Conversely, decreases in the fine fractions after carbonate removal suggests the presence of clay- and silt-size carbonate. We selected one profile from each site that had the most carbonate to enhance the sensitivity of the analysis. In the soil samples from FW irrigation, carbonate removal produced a significant decrease in the sand content, which was compensated by an increase in the silt fraction. Carbonate removal had no significant influence on the clay content along the entire profile (Fig. 2a and 2c ), suggesting that the carbonate cementing the silt-size particles is pedogenic.


Figure 2
View larger version (29K):
[in this window]
[in a new window]

  		Fig. 2. Changes in the particle size distribution (PSD) due to carbonate removal, determined by comparing the clay content (volume percentage) of (a) fresh water irrigated and (b) treated effluent irrigated soils before carbonate removal (filled symbols) and after carbonate removal (open symbols); and changes ({Delta}) in the clay, silt, and sand contents of (c) fresh water irrigated and (d) treated effluent irrigated soils due to carbonate removal.

 
A similar carbonate behavior was found in the upper 2 m of the TE profile. On the other hand, about 50% reduction in the clay content below a depth of 2 m was found after carbonate removal (Fig. 2b). Between 2.2 and 3.2 m, the silt and sand fractions increased (Fig. 2d), suggesting that a significant portion of the carbonate was clay size. This carbonate accumulation may be due to transport and deposition of colloidal CaCO3 (Dan, 1983; Baghernejad and Dalrymple, 1993; Neaman et al., 2000). It also can result from irrigation with TE, because the presence of dissolved organic matter has been shown to inhibit the rates of calcite dissolution and precipitation (Lebron and Suarez, 1996). McCaslin and Lee-Rodriquez (1979) reported a similar pattern. They found that the soil samples from a field irrigated with secondary effluent for about 40 yr contained more clay relative to soil samples from an adjacent FW-irrigated field. The difference in clay content disappeared after carbonate removal. If we assume that all Ca and Mg added by the effluent during 75 yr of irrigation precipitated as carbonate in the upper 4 m of the profile, it would account for 30% of the additional carbonate found in the TE field compared with the NI field (Table 2). All these observations suggest that the carbonates found in the TE field below 2 m are a mixture of pedogenic and detrital carbonates.

Soil Organic Carbon Content
Selected samples from the profiles FW-2, FW-5, NI-1, TE-5, and TE-6 were also analyzed for SOC content (Fig. 3 ). These profiles were selected because they were completely sampled to 4 m and they also were rich in SIC. In both irrigation management systems, SOC content (Corg) was lower than the NI samples along the entire studied profile (0–4 m). The Corg contents reported by the soil survey report (Soil Survey Staff, 1988) are similar to values we found in the NI profile, which supports our contention that this site is a representative profile of the soil series. Comparing the NI site to the irrigated fields, 18.3 and 16.0 kg m–2 of Corg was lost in the TE and FW fields, respectively, during the cultivation period. The majority of the SOC lost was from the plow layer (Fig. 3, Table 3), as commonly found in soils under conventional agriculture practices (Stevenson, 1994). No significant difference was noted in the SOC content between the irrigation management systems (Table 3).


Figure 3
View larger version (20K):
[in this window]
[in a new window]

  		Fig. 3. The total organic C of selected samples from two profiles irrigated with fresh water (FW), two profiles irrigated with treated effluent (TE), and one nonirrigated profile (NI).

 

View this table:
[in this window]
[in a new window]

  		Table 3. Mean soil organic carbon (SOC) content from sites irrigated with treated effluent (TE) or fresh water (FW) and nonirrigated (NI) sites at Bakersfield, CA.

 
Isotopic Analysis
To refine our estimation of the effect of water quality on the SIC and SOC in the two studied fields, isotopic analyses were conducted. Decomposition of organic matter directly affects the soil profile CO2 isotopic values, the corresponding dissolved inorganic C isotopic values and the isotopic values of pedogenic carbonate that may have precipitated (Clark and Fritz, 1997). The common value of 13{delta}Ccarb for carbonates that originate from sedimentary rocks is 0 ± 1{per thousand} (Clark and Fritz, 1997). Pedogenic carbonates have more negative 13{delta}Ccarb values that can range between –0.6{per thousand} and –14.6{per thousand}, depending on the C source. Atmospheric CO2 13{delta}C was about 6{per thousand} under preindustrial conditions and it is about 8{per thousand} under modern atmospheric conditions. The C3 plants have an average 13{delta}Corg of –27.1{per thousand}, while C4 plants have an average 13{delta}Corg of –12.1{per thousand} (Mermut et al., 2000). Because the SOC of croplands is commonly a mixture of C4 and C3 plants, it is important to measure both the 13{delta}Ccarb and 13{delta}Corg.

Stable Isotopes
Soil Organic Carbon
The topsoil 13Corg from the TE irrigation field was less negative than samples from NI and FW irrigation samples ({delta}13Corg = –22, –25 and –23.5, respectively; Fig. 4 ). The {delta}13Corg of the soil under TE irrigation remained the same in the first 1 m of the profile ({delta}13Corg = –22 ± 0.5{per thousand}). Similar values were noted for an orchard irrigated for about 40 yr with secondary effluent in Israel (Affek et al., 1998). Deeper in the profile (2–4 m), the TE SOC is more enriched in 13C ({delta}13Corg = –21.5{per thousand} up to –18{per thousand}), which may be a result of relatively higher contributions from microbial degradation of organic C derived from the effluent and the biosolids (Table 4; O'Brien and Stout, 1978; Macko and Estep, 1984). A high concentration of substrate can explain possible microbial activity in deeper horizons. Fine et al. (2002) reported that the total organic C in the leachate from lysimeters irrigated with secondary effluent was about seven times higher than that from FW-irrigated lysimeters. The less negative {delta}13Corg may also reflect the pedogenic carbonate that precipitated at this depth (Fig. 4). We used this lower value in later calculations.


Figure 4
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 4. The stable C isotope ratio of soil organic C ({delta}13Corg) from selected samples from two profiles irrigated with fresh water (FW), two profiles irrigated with treated effluent (TE), and one nonirrigated profile (NI).

 

View this table:
[in this window]
[in a new window]

  		Table 4. Typical organic C stable isotope ratio (13{delta}Corg) values for treated effluent and biosolids.

 
Soil Inorganic Carbon
We found similar {delta}13Ccarb values (–5 to –8{per thousand}) in the carbonate from FW- and TE-irrigated fields at the 0- to 2-m depth. At the NI site, two of the five {delta}13Ccarb values were more negative (–6 to –11{per thousand}; Fig. 5a ). At the 2- to 4-m depth, the {delta}13Ccarb values for samples from the TE-irrigated field were less negative (–1 to –3{per thousand}) relative to samples from the FW-irrigation field and the nonirrigated site (–6 to –7{per thousand}). The {delta}18Ocarb values were more negative with depth for the 0- to 2-m depth at the three sites (Fig. 5b). This trend was also noted for the TE-irrigated field at the 2- to 4-m depth.


Figure 5
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 5. The depth distribution of the (a) stable C isotope ratio of soil inorganic C ({delta}13Ccarb) and (b) the stable O isotope ratio ({delta}18Ocarb) of selected samples from two profiles irrigated with fresh water (FW), two profiles irrigated with treated effluent (TE), and one nonirrigated profile (NI).

 
Age of Soil Carbonate
If C is being sequestered as SIC and is being influenced by the irrigation water quality, we would expect to find "young" C in the carbonate compared with "older" carbonate that was inherited with the deposition of the soil parent material. The radiocarbon analyses suggest that carbonate ages ranged from 1400 to 7030 ± 120 yr for samples from the irrigated sites (Table 5). The age of the bulk SIC in the FW-irrigated field is younger than the SIC in the TE-irrigated field. Also, the trend of the radiocarbon ages of samples from the TE-irrigated field indicated that the deeper the sample, the older the carbonate (Table 5). The carbonate ages, however, are not consistent with either purely detrital carbonate (the last glaciation) or recently precipitated carbonate due to irrigation.


View this table:
[in this window]
[in a new window]

  		Table 5. Radiocarbon ages of selected samples from two irrigation water sources in Bakersfield, CA.

 
The relatively young age of the carbonate from the field irrigated with TE may indicate a dilution of older material with young carbonate. This is consistent with the PSD analyses and the mass balances of carbonate considering that the NI site represents the situation before commencing irrigation.

Interpretation of the Isotopic Analysis and Total Carbon Mass Balance
Soil Organic Carbon
In the soils irrigated with FW and in the nonirrigated soils, the depletion of the 13Corg found in the first 1 m of the profile is commonly found in soil (O'Brien and Stout, 1978; Balesdent et al., 1993). This depletion may be related to 13Corg fractionation by soil microorganisms during decomposition of the soil organic matter and the addition of 13Corg-enriched microbial biomass to the organic C signature (O'Brien and Stout, 1978; Balesdent and Mariotti, 1996) or to the fact that easily metabolized components of the organic matter (e.g., organic acids) are richer in 13C (Six et al., 2000). The more negative {delta}13Corg values in the next 3 m can be explained by a lower contribution of the microbial biomass to the organic C signature at this depth, or in other words, an enriched fraction of organic matter that percolated to that depth. On the other hand, the {delta}13Corg signature in the TE irrigation soils from 2- to 4-m depth was found to be more positive than the C from the first 1 m (Fig. 4), which may indicate that microbial activity occurred in that deep layer, or that products of microbial decomposition have been translocated to this depth.

Soil Inorganic Carbon
We applied our {delta}13Ccarb and {delta}18Ocarb results to the general model proposed by Cerling (1984) (Fig. 6 ). The model suggests a relationship between climate, {delta}18Ocarb, and {delta}13Ccarb in native modern carbonate. The model considers the variation in vegetation type, gas exchange with the atmosphere, and the rain isotopic composition. Also, it assumes that all carbonate in the soils is pedogenic and that no detrital carbonate remains. It can be clearly observed that the nonirrigation and the FW-irrigation carbonate samples fit rather well within the range of "normal" carbonate in the model. On the other hand, the majority of the carbonate analyses from the effluent irrigation field are above the 30% threshold for atmospheric CO2 contribution to the soil atmosphere, which is the upper value for "normal" soil carbonates (Cerling, 1984). An unlikely explanation is that the carbonate precipitated at this depth has a larger atmospheric C contribution than carbonate precipitated above it, as the model suggests. A better explanation is that the TE contribution to the {delta}13Corg value (Fig. 4) is responsible or that the {delta}13Ccarb represents a mixture of old pedogenic and old detrital carbonate. Two samples from the top layers of the profiles were found to behave differently. They appear to be below the bottom line (0% atmospheric contribution), which was characterized by Cerling (1984) as "coastal" region (carbonate from this region is distinguished by the contribution of enriched 18Ocarb meteoric water). Since no high-meteoric water exists in this region, we speculate that they have a significant eolian contribution.


Figure 6
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 6. The relationship between the stable O isotope ratio and the stable C isotope ratio of soil inorganic C ({delta}18Ocarb and {delta}13Ccarb) from Bakersfield soils irrigated with fresh water (circles), treated effluent (triangles), and nonirrigated (squares). The solid lines represent a "normal" carbonate precipitate in arid and subhumid climates with 0, 10, and 30% atmospheric CO2 contribution. The dashed lines represent the range of vegetation type (100% C3 plants vs. 100% C4 plants), as suggested by Cerling (1984).

 
We estimated the fraction of the total carbonate that was pedogenic carbonate using the following equation adopted from Magaritz and Amiel (1980):

Formula 4[4]
In our calculation, we assumed {delta}13Cparent = –1.0 ± 1{per thousand}, which is a common value for parent material calcite, and {delta}13Csoil = –11.2{per thousand}, which is the lowest value in the nonirrigation sites. The last is in close agreement with the value of –11.0 found in a site near ours by Amundson and Lund (1987). Because we found the {delta}13Corg under TE irrigation between 2 and 4 m to be less negative by 2{per thousand}, we also calculated the pedogenic carbonate content using {delta}13Csoil = –9.2{per thousand} (instead of –11.2{per thousand}).

With the knowledge of the relatively high {delta}13Ccarb values for the TE-irrigated samples, it was not surprising to find that the percentage of pedogenic carbonate in the majority of these samples was low compared with that in the NI and FW-irrigation samples (Fig. 7a ). When we calculated the mass of pedogenic carbonate by multiplying the fraction of pedogenic carbonate by the total amount of carbonate, however, the picture looks different (Fig. 7b, Table 6). Under FW irrigation, about 59.5 kg m2 (4 m)–1 of carbonate (as CaCO3) has been lost (Table 3). If we assume that the sources of acidity needed for CaCO3 dissolution have come mainly from H2CO3, this depletion of carbonate can be considered as 7.2 kg m2 (4 m)–1 SIC (as C) sequestration during 75 yr of irrigation (Table 6). Under TE irrigation, the picture is more complicated. The depth of carbonate accumulation is well below the zone of most active root growth. This may indicate that processes in addition to evapotranspiration have had a role in carbonate accumulation in TE-irrigated fields. The high concentrations of organic C (total organic C = 40–250 mg L–1) below the root zone in TE irrigation (Fine et al., 2002) inhibit carbonate dissolution (Lebron and Suarez, 1996) and, in combination with generous irrigation, support colloidal transport of CaCO3 (Baghernejad and Dalrymple, 1993) to this depth. In addition, microorganisms are considered to have an important function in carbonate precipitation in arid soils (Phillips et al., 1987; Monger et al., 1991; Monger and Galleos, 1999). The presence of substrate may enhance microbial activity and precipitation of carbonate at this depth as well. At 0 to 2 m, the TE-irrigated field lost about 4.1 kg m–2 (4 m)–1 SIC. On the other hand, at the depth of 2 to 4 m, the TE soil has 1.7 to 2.4 kg m–2 (4 m)–1 more pedogenic carbonate (depending on which {delta}13Csoil was used) and 8.8 kg m–2 (4 m)–1 more detrital carbonate, both relative to NI. This pedogenic carbonate can be explained by reallocation of part of the 4.1 kg m–2 (4 m)–1 of carbonate that is "missing" at 0 to 2 m. The additional 8.8 kg m–2 (4 m)–1 detrital carbonate found in the TE soil may be explained by redistribution of CaCO3 from the upper layer of the profile and from exogenous material (e.g., dust-borne carbonate and the TE). It is also probable that the fields we selected for the study were not totally identical in carbonate content in the deeper layers at time zero (before irrigation started). The sum of differences between NI and TE and FW irrigation are summarized in the last column in Table 6 and give the range of difference in SIC content depending on the scenario used for calculation. These results suggest that irrigation with TE moderated the carbonate leaching from the 4-m studied depth relative to irrigation with FW.


Figure 7
View larger version (18K):
[in this window]
[in a new window]

  		Fig. 7. The estimated pedogenic carbonate calculated by Eq. [3] as a function of depth: (a) the fraction (percentage) of the pedogenic carbonate from the total carbonate and (b) expressed as mass of carbonate per mass of soil. Soils were irrigated by fresh water (FW), treated effluent (TE), or nonirrigated (NI). The error bars represent the range of calculated values applying ±1 in the 13Cparent.

 

View this table:
[in this window]
[in a new window]

  		Table 6. The mean pedogenic and detrital carbonate (as soil inorganic carbon [SIC]{dagger}) contents as estimated by using Eq. [4] and Table 2 for soils irrigated with treated effluent (TE) or fresh water (FW) or nonirrigated (NI).

 
The Soil Survey Staff (1988) suggested that the Kimberlina series is formed from recent alluvium on young alluvial fans. When we plotted the radiocarbon age as a function of estimated pedogenic carbonate contribution (as a percentage of the total carbonate) (Fig. 8 ), we found that the y intercept is in good agreement with the estimated age of the soil. If we consider that the detrital carbonate was completely reset to modern at the time of fan deposition (Pendall et al., 1994), the age of 11 000 yr BP for the detrital carbonate is in good agreement with the estimated age of the parent material deposition in the Kimberlina series. This unique relationship supports our estimation of the contributions of detrital and pedogenic SIC fractions.


Figure 8
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 8. The 14C age (from Table 2) as a function of the percentage of carbonate for soils irrigated with fresh water (circle) or treated effluent (triangles) and the estimated value of 90 yr ago (the time before irrigation started) with 100% pedogenic carbonate (X).

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
This study provides some of the first evidence of how water quality affects SIC and SOC pools. It also demonstrates the importance of sampling to greater depths to get a better total C budget in irrigated soils. The SIC pool is affected by several dynamic processes and by the sources of carbonate. Conclusions based on differences in carbonate content alone are likely to be misleading. The combination of analyses we used helped to improve our estimate of SIC sequestration. In particular, the stable isotope analysis helped to clarify the picture even though we found high variability in the SIC distribution within and between the studied fields.

We compared the SIC composition in two irrigated fields to a reference site that was not irrigated. A key for the comparison is an assumption that the fields had an identical carbonate content and distribution before irrigation started. The interpretation of stable C isotopic analysis suggests that it is not the case because the large amount of detrial carbonate in the TE-irrigated field below 2 m was not found at the other sites. With all this uncertainty, our calculation suggests that under both irrigation management types, carbonate was depleted. If we assume that the major source for this SIC depletion was H2CO3, we might be able to account for 7.2 kg m–2 (4 m)–1 SIC sequestration under FW irrigation and for 0.9 to 2.4 kg m–2 SIC sequestration under TE irrigation (Table 6). If carbonate accumulation is counted as C sequestration, the two fields were a source for the same amount of C as mentioned above from the SIC pool.

We also were able to estimate the date of deposition of the detrital carbonate as 11 000 yr BP by extrapolating the relationship between percentages of pedogenic carbonate with the radiocarbon age of this carbonate. This estimation is in good agreement with earlier estimations of the age of the formation of the soil based on morphology and geomorphology.

Under both irrigation management systems, there were large losses of Corg during the cultivation period (18.3 and 16.0 kg m–2 [4 m]–1 for TE and FW irrigation, respectively), which is not surprising because both of them were under conventional agricultural practices. The total C budget for >75 yr of FW and TE irrigation accounts for the loss of 8.8 and 17.4 to 15.9 kg C m–2 (4 m)–1, respectively, if carbonate dissolution counts as C sequestration. One can also add to these estimates an additional 1 to 2 kg m–2 C for both fields as a result of energy (water pumping and tractors) and fertilizer applied to the field during the cultivation period (Schlesinger, 2000; Entry et al., 2002, 2004).


   ACKNOWLEDGMENTS
 
The work reported here was supported by the Kearney Foundation of Soil Science Project no. 200110. We wish to thank Blake Sanden, Kern County irrigation and agronomy farm advisor, who gave us access to sites and helped us to coordinate the soil sampling. We also appreciate the invaluable assistance of Dr. Guy Levy and Dr. Uri Mingelgrin.

Received for publication February 14, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Eshel, G.
Right arrow Articles by Singer, M. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Eshel, G.
Right arrow Articles by Singer, M. J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Eshel, G.
Right arrow Articles by Singer, M. J.
Related Collections
Right arrow Water Quality
Right arrow Dryland Soils
Right arrow Carbon Sequestration

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome