HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Rasmussen, C. Search for Related Content PubMed Articles by Rasmussen, C. GeoRef GeoRef Citation Agricola Articles by Rasmussen, C. Related Collections Dryland Soils Soil Geomorphology and Geography Pedology

Pedology

Distribution of Soil Organic and Inorganic Carbon Pools by Biome and Soil Taxa in Arizona

Dep. of Soil, Water, and Environmental Sci., Univ. of Arizona, Tucson, AZ 85721-0038

^* Corresponding author (crasmuss{at}ag.arizona.edu)

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY REFERENCES

 
Arid systems represent an important component of the global soil C budget in that they cover 12% of the global land area and contain nearly 20% of global soil C stocks, both organic (SOC) and inorganic (SIC). The objectives of this study were to quantify SOC and SIC stocks in Arizona biomes, using Arizona as a model system for arid lands. Biome distribution was extracted from the Arizona Gap Analysis Project spatial vegetation dataset (GAP), while soil C data were extracted from the Arizona State Soil Geographic Dataset (STATSGO) at a scale of 1:250 000, and the western Yavapai County Soil Survey Geographic Dataset (SSURGO) at a scale of 1:24 000. Soil data were converted from a polygonal vector format to a raster format, and a raster-based method used to estimate SOC and SIC stocks by biome. Statewide, STATSGO soil C stocks indicate Arizona contains 0.5 and 1.5 Pg of SOC and SIC, respectively, with 27% of the SOC in pinyon-juniper biomes (PJ), and 34% of SIC in creosotebush-bursage biomes (CB). A comparison of soil C estimates between datasets indicates significantly greater estimates of biome SOC and SIC using SSURGO data relative to the STATSGO data. SSURGO soil C estimates varied considerably between the raster-based and soil taxa based method of data aggregation. Soil taxa data exhibited large intra-unit variation in each biome. In addition, soil C differed substantially between biomes by soil taxa (e.g., Haplargid SOC of 4.0 and 13.5 kg m^–2 in the paloverde-cacti (PC) and montane pine (MP) forest biomes, respectively). Raster based soil C estimations incorporate the spatial distribution and areal land cover of each soil type within a biome, providing a more accurate representation of soil C stocks.

Abbreviations: CB, creosote-bursage biome • GAP, Gap Analysis Project spatial vegetation dataset • MAP, mean annual precipitation • MAT, mean annual temperature • MP, montane pine forest biome • PC, paloverde-cacti biome • PJ, pinyon-juniper biome • PRISM, Parameter Regression Independent Slope Model • SIC, soil inorganic carbon • SOC, soil organic carbon • SSURGO, soil survey geographic database • STATSGO, state soil geographic database

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY REFERENCES

 
ACCURATE QUANTIFICATION of regional soil C stocks is a necessary component of atmospheric CO₂ and global climate change models. Soils represent a significant portion of the active C cycle, with estimates of organic C on the order of 1500 to 2000 Pg C, or roughly two-thirds of terrestrial organic C stocks (Anderson, 1995). In comparison, inorganic C stocks are estimated to range from 930 Pg (Schlesinger, 1997) to 1738 Pg (Eswaran et al., 1995). Arid systems (defined here as land area occupied by Aridisols) cover 12% of the global land area and contain roughly 20% of global soil C stocks, including both SOC and SIC (Schlesinger, 1982; Eswaran et al., 2000). Estimations of Eswaran et al. (2000) indicate that of the twelve soil orders, Aridisols contain the largest percentage of global soil C stocks because of their SIC content. Therefore, arid system soil C is an integral component of the global C budget.

Arizona presents an ideal area to characterize arid system SOC and SIC stocks because of its wide diversity of vegetation and soil types, and may be used as a model system for arid lands across the globe (Schlesinger, 1982; Hendricks, 1985). Further, climate change models predict significant alteration of precipitation and temperature patterns for a large portion of Arizona, (e.g., substantial replacement of desertscrub by semiarid grassland and woodland ecosystems) (National Assessment Synthesis Team NAST, 2000). This shift in biome distribution will undoubtedly affect soil C stocks and the magnitude of soil C storage in this region. Therefore, it is necessary to have an accurate account of current soil C stocks if future trajectories of soil C storage are to be estimated.

Models of soil C response to climate change generally focus on SOC because of its potential rapid response to vegetation change and our ability to manage vegetation and land use practices to promote soil C sequestration. However, SIC represents a considerable pool of soil C relative to SOC in arid systems, and may play a large role in regional C budgets/dynamics. In Arizona, statewide SIC stocks are estimated to be four to six times greater than SOC (Schlesinger, 1982). The dominant source of SIC is still unclear, with some studies demonstrating that the majority of SIC is derived from playa derived eolian inputs, with little provenance from silicate weathering (Capo and Chadwick, 1999). In contrast, Naiman et al. (2000) suggest that a large portion of SIC represents a redistribution of exposed and eroded carbonate material derived from silicate weathering. Despite variation in SIC source, most estimations suggest that SIC is not a large active sink for atmospheric CO₂, with estimates of <10% of global SIC representing atmospheric CO₂ consumption (Eswaran et al., 2000; Mermut et al., 2000). Indeed, Schlesinger (1985) estimated the mean residence time of SIC to be on the order of 85 000 yr. However, it is unclear how SIC will respond to precipitation and temperature changes predicted under climate change scenarios, with the potential for considerable dissolution and degassing of CO₂ from SIC (Monger and Martinez-Rios, 2001).

Attempts to characterize regional soil C stocks include both ecosystem and soil taxa based approaches. The ecosystem approach involves averaging soil C data within a specific plant community or biome and multiplying the average soil C content by the estimated biome land area (Post et al., 1982; Schlesinger, 1997). This approach does not account for soil spatial heterogeneity, and results in large variability of soil C estimations within an ecosystem or biome. Coupled ecosystem-biogeochemical models generally use an ecosystem approach to estimate SOC budgets, using current vegetation distribution and a global average of soil C for each ecosystem to parameterize SOC stocks (Schimel et al., 1997; Bachelet et al., 2003).

The soil taxa approach has been described extensively in the soil science literature, and includes segregating the landscape by soil taxa (instead of biomes) and using the average taxa soil C and estimated land area to calculate soil C stocks (Franzmeier et al., 1985; Kimble et al., 1990; Davidson and Lefebvre, 1993; Kern, 1994; Homann et al., 1998; Eswaran et al., 2000; Monger and Martinez-Rios, 2001). The soil taxa approach presents a better framework for soil C estimation by reducing intra-unit soil C variability (Kern, 1994). However, intra-unit variability can still be substantial, even at detailed levels of soil classification such as the great group (Kern, 1994). Regional soil C budgets vary substantially depending on the method employed. For example, Schlesinger (1982) estimated Arizona SIC stocks of 8.2 and 5.9 Pg using a soil taxa and ecosystem approach, respectively. This variation arises from the averaging of soil C data within a unit (either biome or soil taxa) and applying this mean over a large land area.

An approach utilizing geographic information systems (GIS) allows for spatially explicit soil C stock quantification, such that errors introduced by averaging soil C data by biome or soil taxa may be reduced. Raster based GIS modeling, in particular, presents an improved approach to analyzing soil properties over geographic areas relative to vector-based approaches and ecosystem or soil taxa based methods. Raster modeling divides a geographic area into pixels of set size and assigns a numeric value to each pixel that represents "real-world" or soil-derived values (Bernhardsen, 1999). Use of raster data layers greatly improves computational time and facilitates partitioning of soil data by other data layers with simple calculations. For example, polygonal (or vector) soil data layers may be converted to a raster format, composed of pixels of known area with an associated soil C content (kg m^–2). Since the area of each pixel is known, soil C content may be converted to a mass basis for each pixel (kg). The individual pixels may then be summed according to vector or raster boundaries from other data layers without having to divide and add soil data from several different soil polygons. The raster based approach provides an estimate of soil C content that utilizes data from the entire area occupied by the separate data layer and ensures that regardless of how the geographic area is divided (either by biome, soil taxa, etc.), that the sum of soil C in that area remains constant and is not method dependent. In addition, biogeochemical models commonly use raster data to drive spatial models of ecosystem process (Bachelet et al., 2003), such that a raster-based estimation of soil C stocks provides a readily incorporated data layer.

The objectives of this study were to quantify both SOC and SIC stocks in Arizona using a raster-based method. This approach provides a regional quantification of SOC and SIC that may be readily incorporated into regional biogeochemical models of climate change. In addition, soil C data from varying scales (e.g., 1:250 000 and 1:24 000) was compared to characterize the influence of scale on estimating soil C stocks.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY REFERENCES

 
Study Area Characteristics
The state of Arizona covers an area of roughly 295 000 km². Arizona has three primary physiographic provinces, the Basin and Range in the south and west and the Colorado Plateau in the north and east, with a Transitional province separating the two (Fig. 1 ). The Basin and Range province is characterized by extensive faulting and contains numerous mountain ranges that rise abruptly from broad valleys and basins (Hendricks, 1985). In contrast, the Colorado Plateau consists of gently sloping sedimentary rocks that have been eroded into plateaus and deep canyons, with most relief a function of plateau dissection (Hendricks, 1985).

View larger version (134K): [in this window] [in a new window]   Fig. 1. Major physiographic regions of Arizona. Physiographic boundaries taken from Hendricks, 1985. Map grid lines are projected in Universal Transverse Mercator (UTM), Zone 12.

Data Sources
Biome distribution was derived from the Arizona GAP dataset (Halvorson et al., 2001) (Fig. 2 ). GAP classification is based on remotely sensed data with vegetation units classified into biomes according to the dominant vegetative cover. Biomes are characterized by distinct vegetative physiognomy and evolutionary history, and persist together through space and time. GAP data are mapped at a scale of 1:100 000.

View larger version (97K): [in this window] [in a new window]   Fig. 2. Spatial distribution of the six dominant biomes in Arizona. Biome data are derived from the Arizona GAP dataset and blank areas represent other biomes. The outlined area in the center of the state is Yavapai County. Map grid lines are projected in Universal Transverse Mercator (UTM), Zone 12.

View larger version (72K): [in this window] [in a new window]   Fig. 3. Spatial distribution of the five dominant Arizona biomes in Yavapai County. Biome data are derived from the Arizona GAP dataset and blank areas represent other biomes. Map grid lines are projected in Universal Transverse Mercator (UTM), Zone 12.

Data Extraction and Calculation
Pedon-based C pools were calculated from both STATSGO and SSURGO data and expressed on a mass per area basis (kg m^–2) using Eq. [1] and [2]:

[1]

where SOC_i is the soil organic C content (kg m^–2) for horizon i, D_i is the horizon depth (cm), _bi is the oven dry bulk density (g cm^–3), OM_i is the organic matter weight percentage (%) divided by two (assuming organic matter is 50% C), RF_i is the volume percentage rock fragment (>2 mm) content (%), and 10 is a unit conversion factor (Tan et al., 2004). Weight percentage of RF was converted to volume percentage assuming a rock density of 2.6 g cm^–3. Soil organic C was calculated for each horizon in a pedon, and all horizons summed by pedon, regardless of pedon depth. Soil clay content was calculated using the same equation, replacing (OM/2) with clay content as a weight percentage (%). Soil inorganic C was calculated using Eq. [2]:

[2]

where SIC_i is soil inorganic C content (kg m^–2) for horizon i, CaCO_3i is the calcium carbonate equivalency in weight percentage (%) and 0.12 is a conversion factor converting CaCO₃ to C. All calculations utilized the average parameter value from the STATSGO and SSURGO datasets. Data extraction and calculations were performed using a combination of Microsoft Office Access 2003 (Microsoft Corp., Redmond, WA) and JMP IN 5.1 (SAS Institute, Cary, NC).

Data Manipulation and Statistics
STATSGO horizon soil C and clay content data were summed for each soil map unit component. Component data was weighted by the component percentage, then summed by STATSGO soil map unit, providing a single component weighted value for soil C and clay content for each soil map unit. STATSGO soil C and clay content data, GAP biome type, and PRISM data were converted to a raster format, with a pixel size of 4 by 4 km. SOC data include estimates of litter layer C from the Forest Inventory Analysis (FIA) (USFS-USDA, http://www.fia.fs.fed.us/), which added from 0.2 to 2.1 kg m^–2 of SOC to STATSGO SOC estimates. Soil C and clay content data were multiplied by the pixel area (16 000 000 m²) to obtain a mass (kg) of C and clay in each pixel. The overlay of these raster layers was exported as a database file (with each row representing a pixel) and imported to a statistical software package for analysis. Mean annual precipitation and MAT were averaged by biome type. Soil C and clay content were summed by biome type. The average soil C and clay content on a mass per area basis (kg m^–2) was then calculated by dividing this sum by the area of each respective biome. This type of calculation incorporates not only the component percentage variation within a pixel or soil map unit, but also includes the spatial distribution and areal extent of soil map units within a biome.

A similar approach was used to quantify soil C stocks by biome and soil taxa from the SSURGO data. SSURGO and GAP data were converted to raster format, with a pixel size relative to the SSURGO data (500 by 500 m). Soil C and clay content was summarized by biome as described above for the STATSGO data. Soil C data was also summarized by order and great group in each biome to quantify the distribution of soil C by soil type within a biome. In addition to the raster-based approach, the mean soil taxa soil C content (kg m^–2 basis) in each biome was calculated by averaging the pedon data within each soil taxa. A coefficient of variation (CV) was calculated for the soil taxa method soil C to quantify C variability within the soil taxonomic units.

Regression and correlation analyses were performed between MAT, MAP, SOC, SIC, and clay content for both STATSGO and SSURGO data. All data manipulation, calculation, and statistical analyses were performed using JMP IN 5.1 (SAS Institute, Cary, NC). All geographic data was manipulated in ArcGIS 9.0 (ESRI Inc., Redlands, CA).

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY REFERENCES

 
Biome Characteristics
Analysis of biome distribution indicates that six biomes account for nearly 80% of Arizona's land area; pinyon-juniper (PJ), paloverde-mixed cacti (PC), creosotebush-bursage (CB), shrub-grass disclimax (SG), montane pine forest (MP), and mixed grassland (MG) (Table 1). The PJ, SG and MP biomes are mainly distributed in the Colorado Plateau region, while the PC and CB biomes dominate the desert regions in the southwestern half of the state, and the MG biome the southeast portion of the state (Fig. 1). The PC and CB biomes are part of the Sonoran Desert and have also been termed the Arizona Upland and Lower Colorado River Valley complexes, respectively (MacMahon, 2000). PRISM data indicates significant variation in MAT and MAP between biomes (Table 1). As expected, the Sonoran Desert biomes are hot and dry with MAT near 20°C and precipitation <300 mm, particularly the CB biome, with only 180 mm MAP. The Colorado Plateau biomes are generally cooler, with moderate MAP, except for the MP biome that receives over 600 mm MAP.

View this table: [in this window] [in a new window]   Table 1. Arizona land area distribution, environmental parameters and soil carbon stocks for selected biomes.

STATSGO Soil Carbon Pools
STATSGO soil C data calculated using the raster method indicate that Arizona contains 0.5 and 1.5 Pg of SOC and SIC, respectively. The SOC values are roughly half and SIC three to four times less than those estimated by Schlesinger (1982), who used both an ecosystem and soil taxa approach to estimate SOC and SIC stocks (at scales of 1:3 168 000 and 1:5 000 000, respectively). The difference in soil C stock estimates likely arises from differences in soil C estimation method and variation in the scale of data used for biome and soil taxa land area. Previous studies have demonstrated substantial variation in soil C stock estimates when using data from varying scales, with up to a 23% difference in SOC stocks when comparing calculations from coarse and fine scale data (Davidson and Lefebvre, 1993; Homann et al., 1998). Galbraith et al. (2003) postulated that scale variation in soil C estimation might arise from scale associated map unit composition changes (e.g., soils that occupy extensive area dominate coarse scale soil C estimations, with small areal extent soils lost in data aggregation).

Biome SOC stocks (Tg) follow land area distribution, where those biomes with the greatest land area account for the majority of SOC in the state (Table 1; Table 1 presents data for the major biomes in Arizona by land area and C stocks, but is not inclusive of all biomes and soil C calculated for the state). The PJ biome, which occupies 20% of the state's land area, contains the largest overall SOC stocks (141 Tg), roughly 27% of the state's total SOC, indicating a significant role for this biome in the SOC budget of Arizona. Further, STATSGO SOC content indicates substantial variation between biomes on a kg m^–2 basis, ranging from 1.5 to 2.8 kg m^–2 (Table 1). Montane pine forest biomes contain the greatest SOC content, although Arizona MP SOC content is relatively low compared with conifer biomes in other areas of the western USA. For example, Homann et al. (1998) estimated SOC content on the order of 13 kg m^–2 for the upper 100 cm of Oregon conifer biomes using STATSGO data. The low MP SOC content likely results from moisture limitation and reduced net primary production in the relatively arid Arizona systems. The influence of climate on Arizona SOC stocks is demonstrated by regression analysis of the average biome SOC (kg m^–2) by average biome MAT and MAP that suggests a significant positive relationship between SOC and MAP (r² = 0.89; P = 0.02) and a negative relationship between SOC and MAT (r² = 0.59; P = 0.13).

Unlike SOC, SIC stocks (Tg) do not follow biome land area (Table 1). In particular, the CB biome accounts for 34% of the states total SIC, but only accounts for 13% of the land area, indicating this biome plays an overwhelming role in Arizona SIC storage. The desert biomes receive little effective precipitation for leaching of carbonates from the soil system, enabling accumulation of carbonate minerals in the soil profile (Gile et al., 1966; Lal and Kimble, 2000; Buol et al., 2003). This is exemplified by a strong positive relationship between the average biome SIC and MAT (r² = 0.88; P = 0.02) and a strong negative relationship to MAP (r² = 0.76; P = 0.06). In terms of total soil C stocks, SIC represents a much larger portion of total soil C (Tg) relative to SOC, except for the MP biome (Table 1). Indeed, the CB biome contains the largest overall soil C (577 Tg) because of its substantial SIC stocks.

Correlation analysis between soil C stocks, climate parameters and soil texture suggest significant correlation variation between biomes (Table 2). In general, biomes exhibit a significant positive correlation between SOC and MAP and a significant negative relationship between SOC and MAT, although most correlations are very weak. The exception is the PJ biome where MAP shows a significant negative relationship to SOC, although the correlation is very weak (r = –0.07). The climate parameter relationship to SOC is strongest in the desert biomes, suggesting a greater response of SOC to MAT and MAP in moisture stressed, biomass production limited biomes. The PC, CB, and SB biomes also exhibit significant correlations between SOC and soil clay content, suggesting that either these variables covary as a function of soil development, or a possible control of clay content on SOC stocks. Current models of SOC dynamics commonly rely on soil clay content as a means to partition SOC into fast, slow, and passive pools (Parton et al., 1987). However, the results from this study imply varying control of SOC stocks by soil clay in each biome, suggesting each biome may require a biome specific model or biome specific parameters for accurate estimation of SOC cycling.

View this table: [in this window] [in a new window]   Table 2. Correlation matrices for STATSGO soil C in each biome.

Correlation analyses by biome indicate significant positive correlations between SOC and soil clay content (Table 4). It is interesting to note that SIC also exhibits a significant positive correlation with clay content, particularly in the desert biomes (PC and CB). The positive correlation between SIC and clay content suggests these two properties covary as a function of soil development and may indicate an accumulation of eolian material. Older landscapes in PC and CB biomes commonly exhibit a significant degree of pedogenesis, in terms of clay content and the formation of calcic or petrocalcic horizons in the subsurface, despite the present day arid climate. It has been suggested that the clay and SIC enrichment of older Southwest desert landscapes are relict from a cool, wet Pleistocene paleoclimate with greater rates of weathering and translocation [(Gile et al., 1966; Marion et al., 1985; Schlesinger, 1985; Southard, 2000)]. The combined positive relationship between SIC, clay, and SOC suggest that SOC accumulation is either a function of greater time of soil formation and accumulation of SOC over time, or that older, clay enriched soils have greater potential for SOC stabilization.

View this table: [in this window] [in a new window]   Table 4. Correlation matrices for Western Yavapai County SSURGO data by biome.

Soil C stocks exhibit appreciable variation within a soil order as indicated by relatively high CV's for both SOC and SIC. Soil inorganic C CV values in particular are very high (all > 60%), suggesting that soil taxa based estimates of SIC possess significant inherent error. Entisols tend to exhibit the greatest variation in SOC in each biome. Entisols encompass a wide range of weakly developed soils, including recent alluvial deposits, shallow soils on residuum in mountainous regions, and highly eroded soils. The large variation in Entisol soil C stocks is likely a function of variability in soil depth, parent material, and soil forming environment. Kern (1994) also found large SOC CV by soil order and suggested averaging SOC data by great group to reduce variability.

Great Group
Biome soil C stocks were further examined by great group (Table 6). The PJ biome is dominated by Argiustolls and Haplustolls, with moderate composition of Haplusterts and Haplocalcids. Sonoran Desert biomes exhibit a similar distribution of great groups with the PC biome dominated by Haplargids and Torriorthents, while the CB biome is predominantly Haplargids, and Haplocalcids. The SG biome is also dominated by Haplargids, with a mix of Haplustolls and Argiustolls, while the MP biome is predominantly Ustorthents. As with the soil order distribution, biome great group distribution suggests significant variation in soil forming environments between biomes.

Soil C variability as estimated by the soil taxa method tended to be lower within great groups relative to their respective order (e.g., Haplargid and Haplocalcid SOC CV's of 63 and 59%, respectively, relative to an overall Aridisol SOC CV of 71% in the PJ biome) (Table 6). In particular, the great group segregation of Entisols highlighted the variation of SOC within this order. Torrifluvent SOC is substantially greater than Torriorthent SOC in the PC and CB biomes, indicating the young, alluvial Torrifluvents have a greater capacity for SOC storage than the Torriorthents (which in the Southwest commonly represent soils occupying erosive landscapes where diagnostic horizons have been stripped away) and explains the high CV's observed in the Entisol soil order. Despite reduction in soil C CV using a finer level of soil classification, soil C values still vary considerably within a taxonomic unit. Large intra-unit variation, coupled with the potential misrepresentation of soil C content across the landscape, represent severe limitations to using the soil taxa method to accurately assess biome soil C stocks.

	SUMMARY

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY REFERENCES

 
Analysis of Arizona soil C stocks by the raster method indicates that total SOC content generally follows biome land area, while SIC stocks are overwhelming located in the desert biomes in the southwest portion of the state. At the STATSGO scale, regression analyses indicate SOC and SIC are strongly related to climate parameters. In addition, SOC tended to show significant positive correlations to soil clay content in each biome. The strength of the SOC correlation to climate and clay vary by biome, with the Sonoran Desert biomes particularly sensitive to MAP and MAT relative to the other biomes, suggesting the need for biome specific parameters to accurately model soil C dynamics in these systems.

SSURGO SOC and SIC data were substantially greater than STATSGO estimates, indicating an influence of scale on regional soil C stock estimations. A comparison of the raster method to the soil taxa method indicates significant variation in estimated soil C content (on a kg m^–2) basis between the two methods. Soil taxa data, both at the order and great group level of classification, exhibited large intra-unit variation and taxonomic unit soil C differed substantially between biomes. For these reasons, a soil taxa based soil C estimation may greatly misrepresent soil C content. In contrast, the raster method incorporates the spatial distribution and inherent land cover of each soil type into its soil C estimations, possibly providing a more accurate representation of soil C stocks. Analysis of SSURGO data using a raster based approach also indicates significant variation in soil C stocks within a soil order or great group between biomes, suggesting landscape segregation based both on biome and soil taxonomy may provide a more reliable estimate of regional soil C stocks.

Received for publication April 8, 2005.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY REFERENCES

 

• Anderson, D.W. (1995) The role of nonliving organic matter in soils. In R.G. Zepp and C.H. Sontagg (ed.) The Role of Nonliving Organic Matter in the Earth's Carbon Cycle. p. 81–92. John Wiley & Sons, New York.
• Bachelet, D., R.P. Neilson, T. Hickler, R.J. Drapek, J.M. Lenihan, M.T. Sykes, B. Smith, S. Sitch, and K. Thonicke. 2003. Simulating past and future dynamics of natural ecosystems in the United States. Global Biogeochem. Cycles 17:1045–1066.[CrossRef]
• Bennet, P.S., M.R. Kunzman, and L.A. Graham. 1999. Descriptions of Arizona vegetation represented on the GAP vegetation map. Unpublished report associated with the Arizona GAP dataset. Biological Resources Division, USGS.
• Bernhardsen, T. 1999. Geographic information systems: An introduction. p. 51–82. John Wiley & Sons, New York.
• Buol, S.W., R.J. Southard, R.C. Graham, and P.A. McDaniel. 2003. Soil genesis and classification 5th ed. Iowa State Press, Ames.
• Capo, R., and O.A. Chadwick. 1999. Sources of strontium and calcium in desert soil and calcrete. Earth Planetary Sci. Lett. 170:61–72.
• Daly, C., R.P. Neilson, and D.L. Phillips. 1994. A statistical-topographic model for mapping climatological precipitation over mountainous terrain. J. Appl. Meteorol. 33:140–158.[CrossRef]
• Davidson, E.A., and P.A. Lefebvre. 1993. Estimating regional carbon stocks and spatially covarying edaphic factors using soil maps at three scales. Biogeochemistry 22:107–131.[CrossRef]
• Eswaran, H.E., E. Van den Bergh, P. Reich, and J. Kimble. 1995. Global soil carbon resources. p. 27–44. In R. Lal et al. (ed.) Soils and global change. Adv. Soil Sci. Lewis Publ., New York.
• Eswaran, H., P.F. Reich, J.M. Kimble, F.H. Beinroth, E. Padmanabhan, and P. Moncharoen. 2000. Global carbon stocks. p. 15–26. In R. Lal et al. (ed.) Global climate change and pedogenic carbonates. Lewis Publishers, Boca Raton, FL.
• Franzmeier, D.P., G.D. Lemme, and R.J. Miles. 1985. Organic carbon in soils of the North Central United States. Soil Sci. Soc. Am. J. 49:702–708.[Abstract/Free Full Text]
• Galbraith, J.M., P.J.A. Kleinman, and R.B. Bryant. 2003. Sources of uncertainty affecting soil organic carbon estimates in Northern New York. Soil Sci. Soc. Am. J. 67:1206–1212.[Abstract/Free Full Text]
• Gile, L.H., F.F. Peterson, and R.B. Grossman. 1966. Morphological and genetic sequences of carbonate accumulation in desert soils. Soil Sci. 101:347–360.
• Halvorson, W., K. Thomas, and L. Graham. 2001. The Arizona GAP project final report. Spec. Tech. Rep. USGS Sonoran Desert Field Station, University of Arizona, Tucson.
• Hendricks, D.M. 1985. Arizona soils. University of Arizona, Tucson.
• Homann, P.S., P. Sollins, M. Fiorella, T. Thorson, and J.S. Kern. 1998. Regional soil organic carbon storage estimates for Western Oregon by multiple approaches. Soil Sci. Soc. Am. J. 62:789–796.[Abstract/Free Full Text]
• Kern, J.S. 1994. Spatial patterns of soil organic carbon in the contiguous United States. Soil Sci. Soc. Am. J. 58:439–455.[Abstract/Free Full Text]
• Kimble, J.M., H. Eswaran, and T. Cook. 1990. Organic carbon on a volume basis in tropical and temperate soils. p. V248–V252. In Trans. Int. Congr. Soil Sci. 14th, Kyoto, Japan.
• Lal, R., and J.M. Kimble. 2000. Pedogenic carbonates and the global carbon cycle. p. 1–14. In R. Lal et al. (ed.) Global climate change and pedogenic carbonates. Lewis Publishers, Boca Raton, FL.
• MacMahon, J.A. 2000. Warm deserts. p. 285–322. In M.G. Barbour and W.D. Billings (ed.) North American Terrestrial Vegetation. 2nd ed. Cambridge Univ. Press, New York.
• Marion, G.M., W.H. Schlesinger, and P.J. Fonteyn. 1985. CALDEP: A regional model for soil CaCO3 (Caliche) deposition in southwestern deserts. Soil Sci. 139:468–481.
• Mermut, A.R., R. Amundson, and T.E. Cerling. 2000. The use of stable isotopes in studying carbonate dynamics in soils. p. 65–86. In R. Lal et al. (ed.) Global climate change and pedogenic carbonates. Lewis Publishers, Boca Raton, FL.
• Monger, C., and J.J. Martinez-Rios. 2001. Inorganic carbon sequestration in grazing lands. p. 87–118. In R.F. Follet et al. (ed.) The potential of U.S. grazing lands to sequester carbon and mitigate the greenhouse effect. Lewis Publishers, Boca Raton, FL.
• Naiman, Z., J. Quade, and P.J. Patchett. 2000. Isotopic evidence for eolian recycling of pedogenic carbonate and variations in carbonate dust sources throughout the southwest United States. Geochim. Cosmochim. Acta 64:3099–3109.[CrossRef][ISI]
• [NAST] National Assessment Synthesis Team (2000) Climate Change Impact on the United States: The potential consequences of climate variability and change. Cambridge Univ. Press, New York.
• Parton, W.J., D.S. Schimel, C.V. Cole, and D.S. Ojima. 1987. Analysis of factors controlling soil organic matter levels in Great Plains grasslands. Soil Sci. Soc. Am. J. 51:1173–1179.[Abstract/Free Full Text]
• Post, W.M., W.R. Emanuel, P.J. Zinke, and A.G. Strangenberger. 1982. Soil carbon pools and world life zones. Nature (London) 298:156–159.[CrossRef]
• Schimel, D., J. Melillo, H. Tian, A.D. McGuire, D. Kicklighter, T. Kittel, N. Rosenbloom, S. Running, P. Thornton, D. Ojima, W. Parton, R. Kelly, M. Sykes, R. Neilson, and B. Rizzo. 1997. Contribution of increasing CO₂ and climate to carbon storage by ecosystems in the United States. Science 287:2004–2006.
• Schlesinger, W.H. 1982. Carbon storage in the caliche of arid soils: A case study from Arizona. Soil Sci. 133:247–255.
• Schlesinger, W.H. 1985. The formation of caliche in soils of the Mojave Desert, California. Geochim. Cosmochim. Acta 49:57–66.
• Schlesinger, W.H. 1997. Biogeochemistry: An analysis of global change. Second ed. Academic Press, San Diego.
• Southard, R.J. 2000. Aridisols. p. E-321–E3-328. In M.E. Sumner (ed.) Handbook of soil science. CRC Press, Boca Raton, FL.
• Tan, Z., R. Lal, N. Smeck, F.G. Calhoun, B.K. Slater, B. Parkinson, and R.M. Gehring. 2004. Taxonomic and geographic distribution of soil organic carbon pools in Ohio. Soil Sci. Soc. Am. J. 68:1896–1904.[Abstract/Free Full Text]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Rasmussen, C. Search for Related Content PubMed Articles by Rasmussen, C. GeoRef GeoRef Citation Agricola Articles by Rasmussen, C. Related Collections Dryland Soils Soil Geomorphology and Geography Pedology