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


     

Published online 25 August 2005
Published in Soil Sci Soc Am J 69:1617-1626 (2005)
DOI: 10.2136/sssaj2004.0132
© 2005 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 Similar articles in ISI Web of Science
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 McGonigle, T. P.
Right arrow Articles by White, G. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by McGonigle, T. P.
Right arrow Articles by White, G. J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by McGonigle, T. P.
Right arrow Articles by White, G. J.
Related Collections
Right arrow Dryland Soils
Right arrow Range Soils
Right arrow Soil Organic Matter

Forest, Range & Wildland Soils

Enrichment over Time of Organic Carbon and Available Phosphorus in Semiarid Soil

Terence P. McGoniglea,c,*, M. Lala Chambersb and Gregory J. Whiteb

a Dep. of Biological Sciences, Idaho State Univ., Pocatello, ID 83209-8007
b Idaho National Lab., P.O. Box 1625, Idaho Falls, ID 83415-2203
c Currently at: Dep. of Botany, Brandon Univ., Brandon, MB R7A 6A9, Canada

* Corresponding author (mcgoniglet{at}brandonu.ca)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Rates of accumulation of organic C and associated changes in available P in surface layers are not well characterized, yet are important for development of subsoil exposed by disturbance. Subsoil was placed experimentally in 1993 into field trenches to simulate waste burial by soil caps in the semiarid western USA, planted, and sampled in 2001 to investigate rates of enrichment of surface soil with organic C and available P in relation to plant canopies. Various soil cap designs and irrigation regimens were studied. Under ambient moisture levels, organic C in the surface soil below Wyoming big sagebrush Artemisia tridentata wyomingensis Beetle & Young increased annually at an average rate of 0.5 g kg–1. We estimate 32 yr would be needed for exposed subsoil to increase to the level of soil surface organic C below sagebrush in undisturbed steppe. Soil surface bicarbonate-available P increased under ambient moisture inputs below sagebrush at a rate of 3.6 µg g–1 annually and had after 8 yr advanced more than half way from levels in exposed subsoil to those in established steppe. Irrigation stimulated enrichment of organic C and available P, and annual rates of increase across the experiment were 0.9 g kg–1 for C and 6 µg g–1 for P. Cap design effects were mostly absent. Enrichment of surface organic C and available P below shrub canopy, compared with intercanopy space, was evident in both the experimental and undisturbed plots. Corresponding increases in C and P were less pronounced under bunchgrass canopies.

Abbreviations: DOE, Department of Energy • EBR-1, Experimental Breeder Reactor-1 • IB, impermeable barrier • INEEL, Idaho National Engineering and Environmental Laboratory • PCBE, Protective Cap Biobarrier Experiment


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
PATTERNS SHOWING SOIL ORGANIC C decreasing with depth have been established for semiarid ecosystems among other biomes (Jobbagy and Jackson, 2000), but rates of C accumulation in surface layers are less well known. Burke et al. (1995) determined that the top 10 cm of abandoned fields in Colorado increased from 1.0 to 1.5 kg soil C m–2 over 50 yr. Insam and Domsch (1988) found that soil organic C in the top 15 cm of mineral soil increased in a 30-yr reclamation chronosequence from <5 g kg–1 initially to 8 g kg–1 in agricultural land and 17 g kg–1 under forest. However, consistent patterns of increase in A-horizon C were not found across a set of chaparral fire recovery chronosequences spanning 85 yr (Marion and Black, 1988). Radiocarbon systems date temperate O-horizon C in 10s of years, A-horizon organic C in terms of 10s or 100s of years, and gives ages in 100s and 1000s of years for that of B- and C-horizons, respectively (Trumbore, 2000; Rumpel et al., 2002). Like C, P is also more concentrated at the surface of undisturbed field soils (Jackson et al., 2000), but data are limited for rates of available or extractable P accumulation in upper soil layers.

The mechanisms by which organic C and available P accumulate at the soil surface are similar (Etherington, 1982). Inputs of litter are greatest at the surface, and these residues contribute organic C and associated P, which is mineralized to inorganic P. Litter inputs are higher at the surface not just for shoots, but also for roots. A consistent pattern of higher root production in surface layers occurs globally, although biomes differ in the percentage of shallow roots as a fraction of the total (Jackson et al., 1996). In addition to greater organic inputs to the surface, limited mobility of C and P in soil contributes to higher concentrations in upper soil layers (Lilienfein et al., 2004). A combination of insolubility and binding to soil mineral surfaces as inner-sphere complexes restricts P movement in soil (Sparks, 2002). Carbon is also of limited solubility and associates with clay minerals (Burke et al., 1989; Jobbagy and Jackson, 2000). Low or moderate mixing activities by soil fauna contribute to the establishment of higher concentrations near the surface (Brady and Weil, 2004).

Elevated concentrations of soil organic C (Bird et al., 2002) and soil macronutrients (Schlesinger et al., 1996) have been reported in semiarid regions below shrub canopies relative to spaces between shrubs. Proposed mechanisms for these increased concentrations of organic C and nutrients are litter deposition (Charley and West, 1975; West and Skujins, 1977), solutes in aqueous stem flow (Whitford et al., 1997), fauna activity (Schlesinger and Pilmanis, 1998), and soil movement directed under shrubs by rain splash (Parsons et al., 1992) and wind (Coppinger et al., 1991). Therefore, the influence of proximity to vegetation must be considered when evaluating C and nutrient enrichment of soil in semiarid environments.

Soil caps have been established over buried radioactive waste from the Department of Energy (DOE) at various sites across the western USA. The function of these caps is to return precipitation inputs to the atmosphere and prevent water flow through the waste (National Research Council, 2001). Unless wastes are scheduled to be relocated, soil caps must remain functional for multiple half-lives of the disposed radionuclides. The half-lives in years for 90Sr, 137Cs, and 239Pu, which are three of the more common radionuclides at DOE sites, are 29.1, 30.2, and 2.41 x 104, respectively. Consequently, the designs of caps must consider soil-forming processes, given that pedogenesis proceeds over similar periods (Palmer, 1998). Features of soil profile development can be expected to have an impact on vegetation composition and dynamics, and in this way modify cap functional properties. For example, nutrient elevation below shrubs is expected to encourage shrub persistence by positive feedback (Schlesinger and Pilmanis, 1998). Change in cap soils must therefore be evaluated to provide input for the development of long-term management of such sites.

This study examined soils of experimental soil caps assumed to have been initially homogeneous with depth or at least to have not varied systemically with depth at the outset of the experiment, because plots were established by subsoil delivery to the site for trench filling. Our aim was to investigate the changes in surface soil properties from a known point in time in comparison with similar soil of a nearby area that had remained undisturbed in the long-term. Plant canopy influences were considered in our study. We expected that this investigation would improve our ability to predict future change in the vegetation and broader ecosystem of soil caps over buried waste based on increases in organic C and bicarbonate-extractable P in the surface layers.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Field Sites
The Idaho National Engineering and Environmental Laboratory (INEEL) in southeastern Idaho, USA, has annual precipitation of 220 mm and average annual temperature 5.6°C (Anderson and Inouye, 2001). The vegetation is dominated by Wyoming big sagebrush with various bunchgrasses and herbs (Anderson et al., 1996). The Protective Cap Biobarrier Experiment (PCBE) was established in 1993 at the INEEL to test the ability of four protective cap designs (Fig. 1) to prevent water from reaching buried waste. Stony layer biobarriers were inserted into two cap designs at either of two depths to attempt to restrict animal burrowing, a third type of insertion was an impermeable barrier (IB), and the fourth cap design has no insertion. No waste was placed in the PCBE, but the soil caps ere engineered to be otherwise similar to DOE caps at other sites. The matrix material of the cap was uniform with depth in all treatments within the top 50 cm (Fig. 1) and comprised a subsoil fill material taken from the nearby Spreading Area B (Anderson and Forman, 2003). Soil of Spreading Area B is of the Coffee Series, classified as a coarse loamy mixed frigid xeric haplocalcid (Olson et al., 1995). Particle-size data for the surface layer of the undisturbed Spreading Area B site for 0- to 10-cm before soil collection for the PCBE were 39% sand, 44% silt, and 17% clay (McDaniel, 1991), which were almost identical for texture values for the Coffee Series as determined by survey for 0 to 8 cm: 36% sand, 48% silt, and 16% clay (Olson et al., 1995).



View larger version (45K):
[in this window]
[in a new window]
  		Fig. 1. Schematic diagram of cap designs at the Protective Cap Biobarrier Experiment. Designs were 2 m of soil without or with a gravel and cobble biobarrier below the top 0.5 or 1.0 m of soil, or instead with an impermeable barrier (IB) in the form of a plastic membrane below 1 m of soil and above 0.6 m of clay.

 
The experiment had three statistical replicates, each with four plots of 16 by 24 m arranged in a single line with their long axes parallel. One plot in each block was assigned one of the four cap designs at random (Fig. 2) . The experiment was originally dug out as a 33 by 320 m trench to a depth of 1.8 to 2.7 m. After biobarriers and membranes were inserted (Fig. 1), the trench was filled with subsoil from Spreading Area B at the INEEL. Texture of this subsoil was recorded as 19% sand, 48% silt, and 33% clay, but was known to vary among loads (Anderson and Forman, 2003). Soil caps were carried to the site as soil in trucks, deposited in 0.2-m layers, and compacted to a bulk density of approximately 1.29 g cm–3. Three irrigation treatments were applied as 8 m strips arranged in a random order within each replicate and laid over and perpendicular to the long axis of each 16 x 24 m plot (Fig. 2). Irrigation treatments were ambient precipitation or 200 mm annual irrigation applied as four 50-mm increments in July or October in each year beginning in the fall of 1994. Water was delivered with a drip irrigation system with dry plumbing to ambient strips. Technical problems dictated that the October 1999 irrigation was instead applied in April 2000.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 2. A schematic diagram of the arrangement of treatments within one replicate. The four cap designs were applied at random to the four 16 x 24 m plots of the replicate, which are shown as four separate columns. The three irrigation treatments were applied at random as strips corresponding to the three rows in the diagram. Then, each of the twelve combinations of cap design and irrigation in the replicate were divided into two split plots, and to each pair the vegetation treatments native species mix (N) and grass (G) were allocated at random. An example of a randomization is shown. Each of the replicates was re-randomized. The smallest cells in the diagram each correspond to a split plot of 8 x 8 m.

 
The twelve combinations of cap design and irrigation treatment within a replicate were each divided into two 8 by 8 m split plots and planted in November 1993. At random, one of each pair of split plots (Fig. 2) was planted with a monoculture of desert wheatgrass Agropyron desertorum (Fisch ex Link) J.A. Schultes ‘Nordan’, and the other was planted with a mixture of five native shrub and five native grass native species (Table 1) collected from the surrounding desert and raised in pots. Desert wheatgrass plots were drill-seeded in rows spaced 0.36 m apart. Native species were planted in a grid pattern with alternating rows of growth forms. Transplants that died during summer 1993 were replaced in 1994. Boreal sweetvetch Hedysarum boreale Nutt. and blue flax Linum perenne L. ‘Apar’ were drill seeded in rows between the native species transplants in March 1994.


View this table:
[in this window]
[in a new window]
  		Table 1. Native species transplanted into the experiment in 1993.{dagger}

 
The second field site is an undisturbed A. t. wyomingensis community near the Experimental Breeder Reactor-1 (EBR-1) building at the INEEL. Soil at the EBR-1 site is within the same Coffee Series as that of Spreading Area B. Vegetation at the undisturbed EBR-1 plots is a mature plant community typical of the Snake River Plain (Anderson et al., 1996), dominated by A. t. wyomingensis and with shrub canopy interspersed with open spaces, herbs, and various bunchgrasses. No grazing and no fire history are recorded for this section of the INEEL, and abundant soil microbiotic crust testify to the lack of disturbance. Six contiguous replicate plots, each 30 m by 45 m, were laid out in a 3 x 2 arrangement in fall 2001 at the EBR-1 site.

Collection of Samples
We collected soil cores of 7-cm diameter and 15-cm length from the PCBE plots in October 2001 and from the EBR-1 undisturbed area in May 2002. Samples were taken using a corer with clear plastic liners (AMS, American Falls, ID) that were driven vertically into the soil in a steel outer guide and pounding device. Large organic debris such as twigs and newly fallen leaves were removed from the surface of each core, but litter in the process of decomposition was inextricable from mineral material and was retained. Cores were taken from all 72 split plots from the PCBE. From each split plot, one core was taken below plant canopy and one in the open adjacent to each canopy, giving 144 cores. The canopies studied were of A. t. wyomingensis in the native vegetation plots and of A. desertorum in the grass-seeded plots. Basin big sagebrush, although originally planted (Table 1), appeared to have died out from the plots. This outcome is in keeping with the reports that A. t. wyomingensis performs well under relatively low elevation and dry environments (Welch and Jacobson, 1988) such as the Snake River Plain, and that this species is associated with soils of abundant silt compared with the more sandy soils that favor other sagebrush subspecies (Shumar and Anderson, 1986).

Each canopy was selected at random, and a core was taken beneath the canopy, midway between the center and edge of the canopy on the west side. A core corresponding to the open area associated with each canopy was collected from the first vegetation-free section of adjacent soil that was found in a sweep beginning on the west side of the canopy and proceeding counter clockwise. The top and bottom of each core were secured by tight fitting plastic covers, and the respective ends labeled. Similarly, in the EBR-1 undisturbed plots, cores were taken from the canopy and associated open space of Wyoming big sagebrush and Indian ricegrass Achnatherum hymenoides (Roemer & J.A. Schultes) Barkworth, which was the only bunchgrass of sufficient abundance at the undisturbed site. Six replicates, two vegetation types, and two locations gave a total of 24 cores from the undisturbed EBR-1 plots.

Cores were returned to the laboratory and stored for 72 h at 5°C. To prepare for cutting, each core was wetted by capillary action by drilling six 2-mm holes in the bottom cover and standing the core upright in a tray of shallow deionized water overnight. Each core was then placed in –20°C freezer for 2 h to further ensure coherency. Cores were then divided into depth segments making cuts with a tile saw at cumulative distances from the soil surface of 1.25, 2.5, 3.75, 5, 7.5, and 12.5 cm. This division of cores gave 864 and 144 soil samples from the PCBE and EBR-1 sites, respectively. Soil samples were then air dried and stored in polythene bags at room temperature.

Laboratory Analyses
Soils were passed through a 2-mm sieve and returned to storage bags before all analyses. Moisture content was determined for soil in each bag to adjust determinations from an air-dry basis to an oven-dry basis.

The method of Walkley and Black (1934) was used to measure soil organic C by oxidation with potassium dichromate using digestion-tubes and the heating-block modification of Heanes (1984). A sample of soil was ground to pass a 0.5-mm sieve and 0.5 g soil was transferred to a 50-mL heating-block tube. Soil was then kept for 30 min at 150°C in a preheated block with 5 mL 0.167 M potassium dichromate and 7.5 mL concentrated sulfuric acid. Tubes were intermittently removed and shaken by hand for 15 s. After heating, the tubes were removed to cool for 30 min. The contents of each tube were then transferred to a 125-mL Erlenmeyer flask with a single rinse of 2.5 mL concentrated sulfuric acid to carry over residue. Titration was with 0.2 M ferrous ammonium sulfate to the dark-red end point produced by twenty drops of Ferroin indicator. Calculation of organic C was based on the oxidation of 1.5 moles C by each mole of dichromate (Nelson and Sommers, 1996).

Soil P was extracted by the method of Olsen and Sommers (1982). Extracts were prepared from 2 g of soil shaken for 30 min with 40 mL of pH 8.5 adjusted 0.5 M NaHCO3 and filtered through Whatman No. 42 to remove soil. Filtrates were stored frozen in plastic vials. From the total P, this procedure extracts a part that is considered to be an index of that P available to plants. Concentration of P in extracts was determined using inductively coupled plasma atomic emission spectroscopy (ICP–AES) after passing extracts through a 0.45-µm membrane filter to remove fine particulate material. Strictly, only inorganic P in solution is available to plants, but the sum of inorganic and organic forms of P in the extracts as determined by ACP-AES was taken as our index of available P.

Statistical Analyses
The PCBE was evaluated in its entirety by analysis of variance with the twelve combinations of treatment factors for cap design and irrigation within a replicate serving as main plot factors in an otherwise hierarchical design. The split-plot factor was vegetation type, which was applied at two levels: desert wheatgrass and Wyoming big sagebrush. The split-split plot factor was location, which was applied at two levels: canopy and open. The split-split-split plot factor was depth, which was applied at six levels corresponding to the depth intervals previously described.

Data from the EBR-1 plots were analyzed by a split-split plot analysis of variance within six blocked replicates. Vegetation type was the main-plot factor at two levels: Wyoming big sagebrush and Indian ricegrass. The split-plot factor was location with two levels: canopy and open. The split-split plot factor was depth with six treatment levels corresponding to the same depth intervals as the PCBE.

All treatment effects were assessed by analysis of variance with separation of means using the Tukey system (Zar, 1999).


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
The interaction of vegetation, location, and depth was significant (P = 0.015) for organic C in the PCBE. Organic C of 12.0 g kg–1 in the top 1.25 cm of soil under the sagebrush canopy was significantly greater than the values between 8 and 9 g kg–1 in the open or 8.5 g kg–1 under desert wheatgrass canopy at this same depth (Fig. 3a) . Corresponding values for the ambient plots alone were 9.1 g kg–1 at the surface under sagebrush and 7.1 to 7.8 g kg–1 at the surface under desert wheatgrass or in the open. Below the 5-cm depth, organic C was close to 5 g kg–1 in all combinations of vegetation and location (Fig. 3a). Treatment effects for available P were broadly similar to those of organic C, but more pronounced. The interaction of vegetation, location, and depth was also significant (P = 0.007) for P. This interaction was related to available P of 58 µg g–1 in the top 1.25 cm of soil under the sagebrush canopy compared with 32 µg g–1 in the open alongside sagebrush at this same depth (Fig. 3b). Unlike the case for organic C, a canopy effect was evident for P for desert wheatgrass in the top 1.25 cm, with values of 38 and 28 µg g–1 under the grass and in the adjacent open space, respectively (Fig. 3b). Corresponding values for the ambient plots alone were 39 µg g–1 at the surface under sagebrush and 22 to 28 µg g–1 at the surface whether under desert wheatgrass or in the open. Below the 5-cm depth, available P was 10 µg g–1 or less in all combinations of vegetation and location (Fig. 3b).



View larger version (43K):
[in this window]
[in a new window]
  		Fig. 3. Soil (a) organic C and (b) available P in the Protective Cap Biobarrier Experiment in relation to vegetation type, location, and depth. Means with different letters are significantly different at the 5% probability level; n = 36. Bars with no letters have abcd omitted for clarity.

 
Mass was not determined for soil samples following separation of cores into sections for different depths, and so conversion of the organic C and available P data from a mass basis to a volume basis is not possible for our study. There must therefore consider the extent to which changes in concentrations on a mass basis may have occurred with, we might suppose, any simultaneous but not documented reduction in bulk density, to correspond to no net change in organic C and available P per unit soil volume. However, there is no reason to think that bulk density changed substantially or in any systematic fashion among depths at the PCBE over the 8 yr leading up to 2001, because soils were compacted to give a bulk density of 1.29 g cm–3 while filling the trench during the establishment of the experiment (Anderson and Forman, 2003). Further, it is a reasonable expectation that increases in organic C and available P concentrations on a mass basis in our study were driven by inputs of organic material rather than by changes in bulk density or physical redistribution of material already within the soil, because no physical disturbance was applied to the soils over the duration of the study.

A significant (P = 0.016) interaction of irrigation and depth was found for organic C and was related to a greater accumulation of C in the top 1.25 cm of the irrigated plots compared with that of the ambient plots (Fig. 4a) . Irrigation effects on organic C were also seen by Entry et al. (2002) and are consistent with greater growth following water addition, in turn causing greater litter input. Irrigation appears to augment the transition to the level of C enrichment found in undisturbed sagebrush. Yet irrigation management is not likely to be implemented for soil caps over buried wastes, because there would be concerns that irrigation water might reach the waste under exceptional circumstances.



View larger version (35K):
[in this window]
[in a new window]
  		Fig. 4. Soil (a) organic C and (b) available P in the Protective Cap Biobarrier Experiment in relation to irrigation and depth. Means with different letters are significantly different at the 5% probability level; n = 48.

 
The greater the P in the top 1.25 cm in the irrigated plots compared with the ambient plots (Fig. 4b) gave a significant (P < 0.001) interaction of irrigation and depth for available P in the PCBE, as was seen for organic C (Fig. 4a). A significant (P < 0.001) interaction for irrigation with cap design and depth was found for available P and was related to the top 1.25-cm depth, where soil P in the IB design plots with July irrigation (mean = 52 µg g–1, n = 12) and in the soil only plots with October irrigation (mean = 61 µg g–1, n = 12) were higher compared with means for the remaining treatment combinations, which fell in the range 20 to 46 µg g–1. Long-term climate data for the INEEL site (Anderson and Forman, 2003) show that mean monthly precipitation in July and October are similar at approximately 10 mm, but that mean monthly temperatures are much higher at about 20°C in July compared with close to 6°C in October. The immediate demand for evapotranspiration placed on the irrigation water in July was, therefore, more intense than that for water delivered in October. Previous soil moisture readings taken by neutron probe indicated that, in general, the depth of penetration of irrigation water was correspondingly greater for the October compared with the July irrigation, but also that all added water had been removed from the plots by the end of each year (Anderson and Forman, 2003). In accordance with this apparent similarity in cumulative evapotranspiration among plots, there were similar extents of enrichment of both soil organic C and available P in the surface soil in the two irrigation treatments (Fig. 4a, b).

There was a significant (P = 0.02) interaction for irrigation with cap design and depth for organic C that was related to higher C in the top 1.25 cm of soil in the IB design plots with July irrigation (mean = 14 g kg–1, n = 12) and in the shallow-biobarrier plots with October irrigation (mean = 12 g kg–1, n = 12). The means for the remaining treatment combinations fell in the range 7 to 10 g kg–1. A significant (P = 0.041) interaction was also found for irrigation, cap design, vegetation, and location, which was associated with a high available P for IB plots with July irrigation under sagebrush canopy (mean = 29 µg g–1, n = 18) compared with means for the other 48 treatment combinations that fell in the range 10 to 26 µg g–1. These interactions involving irrigation and cap designs were both related to greater C and P input for the IB cap design with irrigation. Previous neutron probe monitoring of soil moisture in these plots indicated that ambient precipitation and irrigation water accumulated at the IB, which probably contributed to greater litter input on this treatment (Anderson and Forman, 2003). This increase in litter can account for the interactive effects of involving the IB treatment. In contrast, all ambient precipitation and irrigation inputs to cap designs with soil only and biobarrier insertions were returned to the atmosphere each year, indicting that the soil of the caps was effective as a reservoir to store water inputs annually until they could be transpired by the vegetation cover (Anderson and Forman, 2003). Yet differences in C and P were rare among cap designs, which is probably related to the similarity among cap treatments for the upper 50 cm of the profile.

The undisturbed EBR-1 plots had a significant interaction of vegetation, location, and depth for both organic C (P < 0.001) and available P (P < 0.001). Organic C was highest at 21 g kg–1 under sagebrush in the top 1.25 cm compared with 8 g kg–1 in the open at this depth (Fig. 5a) . Organic C was higher under sagebrush compared with adjacent bare soil at the three uppermost depths (Fig. 5a). Although the Indian ryegrass canopy did not affect organic C, the trend for numerically higher C under the grass compared with that in adjacent soil was seen at all depths (Fig. 5a). The highest value for Indian ryegrass was 11 g kg–1 under the canopy in the top 1.25 cm. Below 5 cm, all means for the interaction of vegetation, location, and depth fell in the range 7 to 10 g kg–1 (Fig. 5a).



View larger version (44K):
[in this window]
[in a new window]
  		Fig. 5. Soil (a) organic C and (b) available P in the Experimental Breeder Reactor-1 plots in relation to vegetation type, location, and depth. Means with different letters are significantly different at the 5% probability level; n = 6. Bars with no letters have abcd omitted for clarity.

 
Available P was significantly higher at 59 µg g–1 in the top 1.25 cm of soil under the sagebrush canopy compared with 15 µg g–1 in the open alongside sagebrush at this same depth (Fig. 5b). Corresponding P values of 24 and 17 µg g–1 in the top 1.25 cm of soil under the Indian Ricegrass canopy and in adjacent soil were not significantly different (Fig. 5b). Below the 5-cm depth, available P was <8 µg g–1 for all treatment combinations in the interaction of vegetation, location, and depth (Fig. 5b).

Although no samples were available from the time of experiment establishment, the approximate concentrations of 5 g kg–1 organic C and 10 µg P g–1 found in 2001 were stable with depth over the interval 5 to 12.5 cm in the PCBE and are the best estimates of the initial concentrations for organic C and P in these plots. Enrichment from 5 to 12.0 g kg–1 organic C and from 10 µg P g–1 initially to 58 µg P g–1 under the sagebrush canopy in the top 1.25 cm was seen to have occurred over 8 yr, or under ambient conditions to 9.1 g kg–1 organic C and 39 µg P g–1 at the same location and depth. The average annual rates of increase under ambient conditions were 0.5 g kg–1 organic C and 3.6 µg P g–1. Across the whole experiment, annual rates were 0.9 g kg–1 organic C and 6 µg P g–1. To place these increases in context, we can compare these concentrations to those of the undisturbed EBR-1 plots that we take as a model of the condition of the PCBE plots that is expected after many decades. If this assumption is valid, then the PCBE plots have in 8 yr progressed a significant proportion of the distance toward their ultimate level of organic C enrichment, measured here as 21 g kg–1 under the sagebrush canopy in the top 1.25 cm. Even more strikingly, the available P in the PCBE plots seems to have already reached its end point of almost 60 µg g–1 over this same time interval of 8 yr, as found similarly in the EBR-1 plots under the sagebrush canopy in the top 1.25 cm. In the plots under ambient conditions, available P was over one half of the way along at 39 µg g–1. The enrichment of organic C in the surface 1.25-cm layer of soil below sagebrush in the PCBE that was evident after 8 yr from a starting point of 5 g C kg–1 seems to form part of a pattern of change leading over approximately 32 yr without irrigation to a concentration of about 21 g C kg–1 as seen in a long-term and stable community dominated by this shrub. However, concentrations of organic C in the depth interval 1.25 to 12.5 cm under sagebrush would appear to need a longer and unknown length of time to reach levels of organic C in establish steppe in the range 8–15 g kg–1. The concentration of approximately 8 g kg–1 organic C that was stable over the depth interval 5 to 12.5 cm in the EBR-1 plots, irrespective of proximity to canopy and type of plant cover, was greater than the value of 5 g kg–1 in the PCBE corresponding to the same depth interval. This greater organic C in the long-term undisturbed plots probably reflects a gradual enrichment achieved over many decades. We expect that this slow increase in organic C below 1.25 cm will, until such levels are reached, leave the cap ecosystem more vulnerable than fully established steppe to soil erosion and subsequent invasion by weedy species. In contrast, available P decreased with depth to a low of approximately 5 µg P g–1 in the EBR-1 plots compared with the value of 10 µg P g–1 below the 5-cm depth in the PCBE. Patterns seen here were consistent with the generalization made across many studies that the P concentration is reduced with depth in the soil profile more abruptly than that for any other nutrient element, with C second to P in this regard (Jackson et al., 2000).

Concentrations of organic C and available P here were similar to those reported for semiarid systems elsewhere (Charley and West, 1975; Virginia and Jarrell, 1983; Turner et al., 2003). Organic C at the 1-cm depth was 17 and 33 g kg–1 at two A. tridentata sites but fell to 7 g kg–1 below 12 cm at both sites, whereas bicarbonate available P was 29 and 8 µg g–1 at these upper and lower depths, respectively (Charley and West, 1975). Apart from differences among ecosystems, use of inductively coupled plasma to simultaneously determine organic and inorganic forms of available P may have contributed to the higher values for available P at the uppermost depth in the present study compared with those of Charley and West (1975). For example, Turner et al. (2003) reported 31 µg g–1 inorganic P and 8 µg g–1 organic P in the upper 30 cm below A. tridentata. Our soils were air-dried before P extraction, which elevates available P (Turner and Haygarth, 2003), but this approach is common among most studies. Leaf P concentrations for A. t. wyomingensis are moderate at 2.7 mg g–1 (Barker and McKell, 1983).

Our data indicate that the discontinuity in soil concentrations of organic C and nutrients underneath compared with between shrubs can form swiftly, which is in keeping with Artemisia litter inputs of up to 100 g plant–1 annually (Allen, 1993) and rates of mass loss reported for A. tridentata litter of 23 to 59% in the first year (Amundson et al., 1989). Combined leaf and twig production for individual A. t. wyomingensis plants are even higher at 350 to 760 g yr–1 (McArthur and Welch, 1982). Standing crop of leaf biomass was estimated for this sagebrush subspecies as 560 g plant–1 (Sturges and Trlica, 1978) or 160 g m–2 (Frandsen, 1983), including space between canopies. Levels of organic C enrichment found here under canopies of A. t. wyomingensis and A. desertorum were similar to those reported for plants of the Chihuahuan Desert, New Mexico (Bird et al., 2002). In that ecosystem, organic C in the top 0.5 cm was 5 g kg–1 in the open but increased to 7 and 14 g kg–1 under black grama Bouteloua eriopoda (Torr.) Torr. grass and under honey mesquite Prosopis glandulosa Torr. shrubs, respectively (Bird et al., 2002). However, the elevation of available P under A. t. wyomingensis and A. desertorum canopies seen here relative to adjacent open space was much more pronounced than effects for available P noted in other systems. Whereas our data are for available P in both organic and inorganic forms combined as determined by ICP–AES, others (Cross and Schlesinger, 1999; Kramer and Green, 1999) have distinguished these forms. No canopy effect was seen in the top 10 cm for either form of available P for Bouteloua grassland or for available inorganic P under creosote bush Larrea tridentata (Sesse & Moc. ex DC) Coville canopies (Cross and Schlesinger, 1999). The only significant canopy response for available P in that study was an increase in available organic P from 2.8 in the open to 3.7 µg g–1 under the Larrea canopy (Cross and Schlesinger, 1999). Similarly, Kramer and Green (1999) found no response of inorganic or organic available soil P to oneseed juniper Juniperus monosperma (Engelm.) Sarg. canopy at one site, whereas at another site these forms of P taken together were higher in the open at 5.1 µg P g–1 compared with 3.9 µg P g–1 under the canopy.

The elevation of organic C and available P was considerably stronger under sagebrush than under the bunchgrasses studied. This difference between shrub and grass is likely related in part to differences between litter inputs of these plants above- and belowground. Although desert bunchgrasses are similar to A. t. wyomingensis for leaf mass per unit land area (West and Hassan, 1985), the latter is drought deciduous in that it drops ephemeral leaves during the dry part of summer. This character likely contributes to the stronger organic C accumulation below the shrub compared with beneath the grass as seen in this study. Root investment by A. t. wyomingensis is especially high in spring (Fernandez and Caldwell, 1975) and occurs at the expense of shoot growth rate for this subspecies (Booth et al., 1990). The root system dry mass of A. t. wyomingensis constitutes 225 g plant–1 (Sturges and Trlica, 1978). Stronger nutrient accumulation below shrubs compared with grasses could also relate to the physical canopy structure insofar as the shrubs have aqueous stem flow (Whitford et al., 1997). Organic C inputs are from leaf litter and rhizodeposition, with 25% of total assimilate allocated to root production and a further 25% for three sinks combined: root respiration, transfer to microbial biomass, and microbe respiration (Kuzyakov and Domanski, 2000). Root growth for Artemisia is high throughout the top 30 cm (Fernandez and Caldwell, 1975; Caldwell, 1985), and on average grasses are shallower rooting than shrubs (Jackson et al., 1996). Therefore, leaf litter fall is the likely explanation for the majority of the shallow C enrichment seen here in the top 1.25 cm below A. t. wyomingensis shrubs. Indeed, lipid signatures have shown for forest systems that leaf contributes much of the input to the uppermost organic C in soil, whereas the importance of root inputs are greater with depth (Naafs et al., 2004).


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The average rate of increase over 8 yr for organic C at the soil surface below sagebrush canopy following subsoil emplacement was estimated here as approximately 0.5 g kg–1 annually, rising from 5 to 9 g kg–1 under ambient moisture inputs. The ecosystem would take a total of 32 yr to reach the level of 21 g kg–1 organic C found at the surface below sagebrush in a long-term undisturbed community. From a starting point of 10 µg P g–1 the surface soil below sagebrush had reached a level of available P of 39 µg P g–1 under ambient conditions over the same 8-yr period, corresponding to an average rate of increase of 3.6 µg P g–1 annually. Over the whole experiment and including irrigated plots, annual rates of increase were higher at 0.9 g kg–1 organic C and 6 µg P g–1 available P. Increases in surface soil organic C and available P were much less pronounced under grass canopies and in the open compared with under sage canopy. Canopy effects as seen in the long-term undisturbed steppe were developed over 8 yr for organic C and available P below sagebrush compared with adjacent soil. A canopy effect for available P was also developed over 8 yr under desert wheatgrass compared with the adjacent open space.


   ACKNOWLEDGMENTS
 
The PCBE plots were established by the late Dr. J.E. Anderson. ESRA Barriers Project 36511 of the DOE supported this research. Technical assistance was contributed by Stefanie Cook, Cindy Holst, Todd Perkins, Brian Miller, and Canary Tennison. Thanks to S.M. Stoller Corporation for permission to sample the PCBE plots and to Dr. Mike McCurry for guidance with inductively coupled plasma spectroscopy.

Received for publication April 2, 2004.


   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 Similar articles in ISI Web of Science
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 McGonigle, T. P.
Right arrow Articles by White, G. J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by McGonigle, T. P.
Right arrow Articles by White, G. J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by McGonigle, T. P.
Right arrow Articles by White, G. J.
Related Collections
Right arrow Dryland Soils
Right arrow Range Soils
Right arrow Soil Organic Matter

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