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


     

Published online 29 March 2006
Published in Soil Sci Soc Am J 70:806-815 (2006)
DOI: 10.2136/sssaj2005.0333
© 2006 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 Blecker, S. W.
Right arrow Articles by Norvell, K. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Blecker, S. W.
Right arrow Articles by Norvell, K. L.
Agricola
Right arrow Articles by Blecker, S. W.
Right arrow Articles by Norvell, K. L.
Related Collections
Right arrow Ecosystem Management
Right arrow Biogeochemical Processes
Right arrow Pedology

Pedology

Phosphorus Fractions in Soils of Taylor Valley, Antarctica

S. W. Bleckera, J. A. Ippolitoa,*, J. E. Barrettb, D. H. Wallc, R. A. Virginiab and K. L. Norvella

a Dep. of Soil and Crop Sciences, Colorado State Univ., Fort Collins, CO 80523-1170
b The Environmental Studies Program, Dartmouth College, 6182 Steele Hall, Room 113, Hanover, NH 03755
c Natural Resource Ecology Lab., Colorado State Univ., Fort Collins, CO 80523-1499

* Corresponding author (ippolito{at}lamar.colostate.edu)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Phosphorus studies in the cold desert ecosystem of the Antarctic Dry Valleys have been largely confined to stream sediments and orthinogenic regions. Expanding P studies to soils may augment the understanding of P biogeochemistry and habitat suitability in this extreme environment. Our objectives were to examine P fractionation in Antarctic Dry Valley soils and sediments and compare their relationship to other soil biogeochemical data. Samples were obtained along transects perpendicular to the Harnish and Priscu streams in Lake Fryxell and Lake Bonney basins, respectively. We utilized a sequential inorganic P extraction procedure, analyzing for a series of labile through resistant P fractions. We further analyzed soils for labile organic P and biomass P. Results showed the amount of inorganic P increased from soluble to Ca-bound P at both sites, with greater weathering of P-bearing minerals at the Fryxell site inferred from the greater P levels found in most fractions as compared with the Bonney site. Fryxell site soluble P findings correlated positively with the Al-bound phase, possibly facilitating P availability to microfauna. The P fraction distribution at both the Fryxell and Bonney sites fits the general relationship between weathering intensity and P distribution of other arid ecosystems.

Abbreviations: DI, deionized water • EC, electrical conductivity • Po, organic phosphorus


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
STUDIES OF P BIOGEOCHEMISTRY in arid and semiarid ecosystems of temperate climates provide insight into the importance of abiotic and biotic controls of this essential ecosystem nutrient (Cross and Schlesinger, 2001; Titus et al., 2002; McCulley et al., 2004). Phosphorus studies in cold desert ecosystems, specifically those of the McMurdo Dry Valleys of Antarctica, have been largely confined to lentic and lotic systems (e.g., Dore and Priscu, 2001; McKnight et al., 2004) and orthinogenic regions (Arocena and Hall, 2003). Though inorganic fractions tend to dominate total soil-P in arid systems (Lajtha and Schlesinger, 1988; Cross and Schlesinger, 2001), biologically available pools of P are essential to soil organisms. In the dry valleys little is known about soil P dynamics. Gudding (2003) reported relatively high levels of acid-soluble P (presumably Ca bound) consistent with earlier reports of relatively high total P content (Campbell and Claridge, 1987), but it is unknown how much of this P is biologically available. Quantifying the distribution of soil P among inorganic and organic fractions will complement previous studies of C and N in these environments (e.g., Fritsen et al., 2000; Burkins et al., 2001; Barrett et al., 2002), and contribute to the understanding of biogeochemical controls over habitat suitability for organisms in this extreme environment.

The underlying spatial variability of soil properties in hot deserts (Schlesinger et al., 1996; Titus et al., 2002) is also apparent in the soil and landscapes of the dry valleys (Virginia and Wall, 1999; Barrett et al., 2004). Lacking the biologic influence of higher plants, physical processes dominate these soils to an even greater extent than their warm counterparts (Wall and Virginia, 1999), and parent material dominates the spatial distribution of soil properties (Campbell and Claridge, 1987). The complex geologic history of Taylor valley (the main site of the McMurdo Dry Valleys Long-Term Ecological Research Program) has contributed to the great spatial heterogeneity. For example, Paleozoic age granites and granodiorites, Jurassic age dolerite, and Cenozoic age volcanics overlie basement Precambrian to Cambrian metasediments including schists, argillites, quartzites, and marbles (Campbell and Claridge, 1987); Quaternary glacial, alluvial, and lacustrine deposits add to the complexity, exemplified by glacial Lake Washburn, which distributed lacustrine sediments over much of the valley floor up to 150- to 300-m elevation approximately 10 to 20 ky (Hall and Denton, 2000). On smaller scales, variation in moisture and solar radiation also impact spatial variability of soil biogeochemical properties. For example, ephemeral snow patches have significant impacts on local hydrology, and the resultant soil chemical, physical, and biologic properties (Gooseff et al., 2003). Barrett et al. (2004) noted that sand wedge polygons provide reliable scaling units for certain soil properties (nematode abundance, SOM), while other variables are more influenced by factors operating over larger spatial scales (pH, salinity, carbonates, invertebrate community composition), likely related to till composition and age.

Weathering intensity (Bockheim, 1997; Gooseff et al., 2002) and faunal diversity and abundance (Treonis et al., 1999; Wall and Virginia, 1999) can vary markedly between soils (the dominant landscape feature of the dry valleys) and the meltwater streams (less extensive in area, but dynamically important in terms of landscape linkages and nutrient cycling) of the dry valleys (Gooseff et al., 2002; Maurice et al., 2002). Where the soils undergo extremely slow development, hyporheic zone weathering can approach levels found in temperate systems (Nezat et al., 2001). Barrett et al. (2002) noted greater biologic control of N in the hyporheic zone compared with soils in a study of inorganic N activity. Previous research has shown how resource heterogeneity impacts habitat suitability and faunal population distribution in this extreme environment (Treonis et al., 1999; Wall and Virginia, 1999; Barrett et al., 2004; McKnight et al., 2004). Since plant influences on soil properties are absent in this environment, patchiness of other soil properties tend to define favorable soil habitats and influence invertebrate diversity. Though simple in structure, with nematodes dominating the metazoan food web (Freckman and Virginia, 1997), spatial variability of these communities is remarkably complex, driven in part by geographic and soil factors (Freckman and Virginia, 1998; Powers et al., 1998; Courtright et al., 2001) as well as more stochastic and less predictable factors such as eolian redistribution (Nkem et al., 2005). For example, Wall and Virginia (1999) suggested that C may be more limiting than N in most instances, as evidenced by a lower correlation between N and nematode abundance and diversity. Though abiotic controls are likely to dominate P fractions in these systems, phosphate activity can be quite high in the more biogeochemically active meltwater streams, where algal mats impact soluble P availability (McKnight et al., 2004), and may be linked to the P deficiencies and photosynthetic activity of some dry valley lakes (Priscu, 1995). Freckman and Virginia (1997) provide one of the few studies linking soil fauna and soil P, noting a positive relationship between nematode presence and soil phosphate. Thus further investigation of both organic and inorganic P fractions may provide greater insight into P biogeochemical relationships in dry valley soils.

Given the contrasting weathering regimes, differences in P behavior between the soils and stream sediments would be expected. Within Taylor Valley stream channels, high rates of chemical weathering are realized within the hyporheic zone due to high stream discharges during the austral summer (Nezat et al., 2001; Gooseff et al., 2002). In comparison, the vast majority of soils remains quite dry year round and is subject to minimal chemical weathering (Campbell and Claridge, 1987). In both soil and hydrologic systems the source of phosphate in Dry Valley ecosystems has been attributed to the weathering of soil and rock (Green et al., 1989, 2005). A look at P fractions across such a weathering gradient should provide insight to the range and variability of P, the potential relationships among organic and inorganic P fractions and other biologically mediated nutrients and invertebrate productivity. Therefore, our research objectives were two-fold: (i) to examine the soil-P fractions in soils of Taylor Valley; (ii) compare the relationship among the soil-P fractions to other soil biogeochemical data.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Site Description
Taylor Valley (77° S 163° E) is one of several valleys in the McMurdo Dry Valley region of Southern Victoria Land, and contains glaciers, ephemeral glacier-fed streams, permanently ice-covered lakes, and vast areas of arid, poorly developed soils. These valleys comprise the largest of the ice-free areas of Antarctica that overall account for <0.5% of the continental area (Claridge et al., 2000). A mean annual temperature of approximately –19°C (Doran et al., 2002), mean annual precipitation <10 cm (Clow et al., 1988), and desiccating katabatic winds exemplify the extreme cold and aridity of this region.

Soils in Taylor Valley are generally classified as Anhyorthels (evidence of cryoturbation), and Anhyhapels (lack of cryoturbation), reflecting their relatively high salinity and pH (Bockheim, 1997; Beyer et al., 1999). Textures are dominated by sands and loamy sands with generally high but variable coarse fragment contents and ice-cemented permafrost occurring at an average depth of 50 cm (Campbell and Claridge, 1987). Minimal chemical weathering is also evidenced by weak horizonation, with salt accumulation and minimal brunification, the most common pedologic indicators. Organic C (0.025% ± 0.019) and total N (0.0013% ± 0.0009) contents are some of the lowest measured in any soil environment (Burkins et al., 2001).

The study sites encompass soil to ephemeral stream transects in the Lake Bonney and Lake Fryxell basins (Fig. 1 ). Lake Bonney lies at approximately 200 m in elevation and 25 km from the coast near the terminus of Taylor glacier; Lake Fryxell lies at approximately 50 m in elevation and 5 km from the coast. Soils around Lake Bonney tend to be slightly drier and warmer than soils around Lake Fryxell (Fountain et al., 1999), and less biologically productive in terms of nematode abundance and CO2 respiration (Parsons et al., 2004). These larger scale differences in climate and biology, coupled with the smaller scale differences in weathering and soil habitat between the hyporheic zone and nearby soils allows for biogeochemical comparisons at two different spatial scales.


Figure 1
View larger version (48K):
[in this window]
[in a new window]
  		Fig. 1. Location of study area in the McMurdo Dry Valleys, Antarctica (adapted from Barrett et al., 2004).

 
Field Methods
Soil samples were collected during the 2000–2001 austral summer from two sites in Taylor Valley; adjacent to Harnish stream in Lake Fryxell basin and adjacent to Priscu stream in Lake Bonney basin. Transects in the Lake Fryxell basin began approximately 500 m upstream of Lake Fryxell along Harnish stream, and continued upstream every 100 m for a total of four transects. Transects in the Lake Bonney basin began approximately 500 m upstream of Lake Bonney along Priscu stream, also continuing upstream every 100 m for a total of four transects in this basin. The beginning of each transect was in the center of the stream and designated 0 m. Additional samples were taken along each transect at 8, 16, 24, and 32 m from the center of the stream channel for a total of 20 samples at each site. Samples at 0, 8, and 16 m were within the wetted sediments and/or hyporheic zone of the stream; samples at 24 and 32 m represented sites beyond the reach of the hyporheic zone. The hyporheic zone is the area of subsurface flow through sediments underlying and adjacent to the stream. No flowing water was present at the time the samples were taken. Using presterilized plastic scoops, approximately 1.5 kg of soil was sampled at each site from the 0- to 10-cm depth and stored in Whirl-pak bags (NASCO, Fort Atkinson, WI). Approximately 5 g of soil were collected from the top 1 cm and stored in amber plastic bottles for chlorophyll a content, which is proxy for algal biomass (Holm-Hansen et al., 1965).

Laboratory Methods
Soil samples were stored at 4°C after transfer to McMurdo Station. Within 48 h after transfer to the station, gravimetric water content (weight/weight) was determining by drying soils at 105°C for 24 h. Soil pH and electrical conductivity (EC) were measured on a 1:2 and 1:5 (soil/water) mixture, respectively (Thomas, 1996; Rhoades, 1996). Textures were determined by hydrometer method (Gee and Bauder, 1986). Microbial C and N were determined by ethanol-free chloroform fumigation (Horwath and Paul, 1994). Briefly, fumigated and unfumigated soil samples were extracted with 0.5 M K2SO4 at a ratio of 5:1 (extractant/dry soil). Microbial C and N, as well as total soil N and C, were analyzed with a Carlo Erba 1500 elemental analyzer (Milan, Italy). Organic C was analyzed by taking a subsample, acidifying with 50% HCl (v/v) to remove carbonates, and analyzed with a Carlo Erba 1500 elemental analyzer. Inorganic C was determined by difference between total and organic C. Additional soil subsamples were kept frozen at –20°C, shipped frozen to Colorado State University, and remained at –20°C until P fraction analysis.

Phosphorus Fractionation
A five-step sequential inorganic P extraction procedure was utilized for soils based on procedures outlined by Kuo (1996). Briefly, the extraction procedure identifies soluble and loosely bound using 1 M NH4Cl, Al (hydr)oxide-surface bound using 0.5 M NH4F, Fe (hydr)oxide-surface bound using 0.1 M NaOH, occluded using 0.3 M Na3C3H6O7 + 1 M NaHCO3 + Na2S2O4 (within the matrices of retaining components/minerals [Evans and Syers, 1971]), and the Ca-bound mineral P fraction using 0.25 M H2SO4. Soils were washed and centrifuged twice with saturated NaCl between each step, with the NaCl solution added to the previous filtrate. Extracts were filtered through a 0.2-µm membrane before colorimetric P determination by spectrophotometer (882 nm wavelength), following a modified ascorbic acid procedure (Rodriguez et al., 1994). Modifications were made to the occluded P fraction due to insufficient color development. These modifications were based on previous research by Weaver (1974), whereby a 2.5-mL aliquot was transferred to a 25-mL volumetric flask. Then, 1.5 mL of 5% ammonium molybdate solution, 15 mL of deionized water (DI), and 2.5 mL of color developing reagent (Rodriguez et al., 1994) were added, and the solution brought to a final volume using DI. The final solution was allowed to stand for 30 min before P analysis. In addition, we identified two organic P fractions: microbial biomass P (i.e., P originating from lysed microbial cells) and labile organic P (easily mineralizable organic P; Po) following procedures outlined by Zhang and Kovar (2000). Microbial biomass P was calculated as total labile P + CHCl3–lysed microbial cells minus total labile P. Labile Po was calculated as total labile P minus labile inorganic P. Briefly, total labile P was determined using 0.5 M NaHCO3 + K2S2O8, total labile P + CHCl3 was determine using 2 mL CHCl3 + 0.5 M NaHCO3 + K2S2O8, and labile inorganic P was determined using only 0.5 M NaHCO3. Phosphorus concentrations were determined colorimetrically as previously described.

Statistical Analyses
The data were statistically analyzed using the proc mixed model in SAS version 9.1 (SAS Institute, 2002). Differences within each P fraction along transects were examined using ANOVA at probability level (P) < 0.05, with mean separation determined using Fisher's Protected LSD procedure. Pearson correlation coefficients were determined between all factors examined to assess the influence of soil properties on the separate soil P fractions, with significant differences identified at P < 0.05.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Fryxell Area Transects
Phosphorus fractionation data for the Lake Fryxell area stream to soil transects are presented in Fig. 2 . Correlations between P fractions and other measured soil properties are presented in Tables 1 and 2. Average soil characteristics for the stream-soil transects are presented in Table 3. Overall, soil inorganic P concentration increased from the soluble fraction through the Ca-bound fraction for all transect locations. The dominance of Ca-bound fractions is likely due to the minimal weathering of P-bearing minerals (apatite and basalt) in this cold desert environment. A similar trend was found in studies of hot deserts (Lajtha and Schlesinger, 1988; Cross and Schlesinger, 2001), where acid-extractable P fractions tended to dominate bulk soil P pools.


Figure 2
View larger version (27K):
[in this window]
[in a new window]
  		Fig. 2. Lake Fryxell average P fractionation concentrations: (a) soluble P; (b) Al-bound P; (c) Fe-bound P; (d) occluded P; (e) Ca-bound P; (f) biomass P; (g) labile organic P. Error bars represent one standard error of the mean.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Fryxell stream transect phosphorus fractionation correlation coefficients.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Bonney stream transect phosphorus fractionation correlation coefficients.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Fryxell and Bonney average stream transect soil characteristics.{dagger}

 
Given the aridity of the dry valleys and absence of significant leaching, ions (i.e., Al and Fe) released from rock minerals tend to remain in the surface soil (Campbell and Claridge, 1987). Thus, the weathering of P-containing parent materials can promote the accumulation of secondary mineral precipitates such as Al and Fe phosphates. Marine deposition and eolian translocation are unexamined though potential P sources in the dry valleys, as evidenced by studies of sediment flux in this system (e.g., Lancaster, 2002; Lyons et al., 2003). However, unlike other constituents (e.g., nitrate and sulfate), which can have a significant marine component (Keys and Williams, 1981), atmospheric P deposition is likely minor, given the lack of a significant gaseous component to the P cycle as well as the relatively low P content in sea water (Stumm and Morgan, 1981).

The soluble P fraction concentration was <1 mg kg–1 within the Harnish hyporheic zone and significantly increased outside the channel (Fig. 2a). The Al-bound and Fe-bound P concentrations were <2 and 14 mg kg–1 at all transect locations, respectively (Fig. 2b and 2c). The increase in soluble P observed between the hyporheic zone sediments and the nearby soil sites was positively correlated with labile Po while the Al- and Fe-bound P were negatively correlated with biomass P and labile Po, respectively. This suggests soluble P release from the Al- and Fe-bound phases may be indirectly influenced by soil biota. McKnight et al. (2004) found soluble reactive P to be lower in Taylor Valley streams with abundant algal mats (<0.54 µM) than in streams with sparse algal mats (1.57 µM) demonstrating the potential for biota to immobilize available P.

The Fe-bound P concentrations were variable across the transect, generally increasing with distance from the stream center (Fig. 2c), and correlated negatively with labile Po (Table 1). Both fractions contained P concentrations < 14 and < 8 mg kg–1, respectively. The mineralization of labile Po may be contributing to the accumulation of Fe-bound P. As labile Po is mineralized it releases soluble P to the soil solution, whereby free Fe ions complex solution P in Fe (hydr)oxide mineral phases. This mechanism could account for the negative correlation between Fe-bound P and labile Po, and the positive correlation between soluble P and labile Po.

Harnish stream transect soils contained between 30 and 55 mg kg–1 occluded P (Fig. 2d). This fraction represents P bound within iron oxide coatings. The occluded P fraction was lowest at the hyporheic zone boundary, but overall correlated negatively with sand content and positively with the silt and clay content (Table 1). Iron oxides in this environment are generally amorphous, ranging from hydrous iron oxides with a goethite-like structure to anhydrous hematite-like material (Campbell and Claridge, 1987), and can sorb greater quantities of anions as compared with their crystalline counterparts. For example, Nanzyo (1986) showed increased phosphate adsorption on an amorphous iron hydroxide gel as compared with crystalline goethite. McLaughlin et al. (1981) showed a 1-mo aged amorphous Al gel sorbed approximately 35 times more P than a crystalline gibbsite, hypothesizing that P surface adsorption of short-range order hydrous oxides in soils will behave similarly to that of aged Al gels. Both findings could explain the increased P in this fraction of the soils studied in the Fryxell basin.

There was no difference between Ca-bound P, the dominant P-containing fraction, at any transect location (Fig. 2e). This trend is not surprising, given the general aridity and minimal chemical weathering in this region, despite reports of greater chemical weathering within stream reaches of the Fryxell basin (Gooseff et al., 2002). Limited weathering in this environment also prevents accumulation of the other P fractions.

There were no differences between biomass P with transect location, although this fraction tended to decrease with increasing distance from the stream channel (Fig. 2f). Biomass P correlated negatively with pH and EC, and positively with clay content (Tables 1 and 3), which corresponds to results shown by Treonis et al. (1999), where greater invertebrate diversity existed in sites with lower pH and EC. The positive relationship between biomass P and clay content suggests organisms associate with larger surface area particles for protection and because clays are more likely to bind various P fractions, possibly accounting for greater P (and other nutrient) availability. Other researchers have also found that cell numbers and microbial biomass were most concentrated in the smaller silt and clay fractions (Kanazawa and Filip, 1986; Jocteur et al., 1991; Kandeler et al., 2000). Conversely it has been suggested that low nutrient availability may have been responsible for reduced biodiversity in larger size fractions of soil (Sessitsch et al., 2001).

In general, labile Po decreased from the stream channel center to the hyporheic zone boundary (Fig. 2g). The higher concentration of labile Po evident at the 24-m transect location is difficult to interpret given our collected data and correlation comparisons. Such an interaction of physical and biological factors may relate to our labile Po findings.

Bonney Area Transects
Phosphorus fractionation data for the Priscu stream to soil transects in the Lake Bonney basin are presented in Fig. 3 . Correlations between P fractionation and other measured constituents are presented in Table 2, and average stream transect soil characteristics are presented in Table 3. As with Harnish transect data, inorganic P concentration increased from the soluble fraction through the Ca-bound fraction for all transect locations, with the increase likely related to the limited chemical weathering of P-bearing parent materials as previously suggested.


Figure 3
View larger version (23K):
[in this window]
[in a new window]
  		Fig. 3. Lake Bonney average P fractionation concentrations: (a) soluble P; (b) Al-bound P; (c) Fe-bound P; (d)occluded P; (e) Ca-bound P; (f) biomass P; (g) labile organic P. Error bars represent one standard error of the mean.

 
The soluble P fraction was <0.3 mg kg–1 and no differences existed across transects (Fig. 3a). In addition, the Al-bound, Fe-bound, and biomass P concentrations were almost always below detection limits (Fig. 3b, 3c, and 3f) and thus correlations for all four fractions were inconclusive.

Although no differences existed across transects (Fig. 3d), occluded P correlated negatively with sand content and positively with the silt and clay-sized fractions. A slight decrease in pH (Table 3) could favor the formation of Fe-coated P. As with the Fryxell site observations, these coatings are found primarily in the silt and clay-sized fraction, with greater surface area to promote coating nucleation and formation of the occluded fraction.

The Ca-bound P fraction contained the greatest amount of inorganic P, and differences were evident within transect locations (Fig. 3e). The Ca-bound P correlated positively with the sand-sized fraction and negatively with the silt and clay-sized fractions. The association of this phase, presumably the P-supplying parent material, with the larger size-fraction supports the contention that limited weathering is occurring at the Bonney site. However, differences in provenance among the size fractions, which has not been examined, could present another possible explanation.

No labile Po differences existed across transects (Fig. 3g), but a positive correlation existed with the silt-sized fraction (Table 2). As with the Fryxell site biomass P, this suggests that the labile Po phase may become immobilized within the smaller-sized fractions (i.e., silt), and with increased weathering and mineralization may facilitate precipitation of Fe-bound phases.

Comparisons between Basins and Similar Ecosystems
Phosphorus concentrations in the measured organic and inorganic fractions were greater in Fryxell basin compared with the Bonney basin, with the exception of the Ca-bound P phase (Fig. 2 and 3). Lower P content at Bonney may be due to the less intense weathering and generally lower biological activity as compared with Fryxell, or less retention of weathered P fractions in older tills (Fritsen et al., 2000; Courtright et al., 2001; Barrett et al., 2004). Given the high coefficient of variation found by Schlesinger et al. (1996) for phosphate in arid systems of the southwestern USA, and the high variability of ecosystem parameters measured in Taylor Valley (Barrett et al., 2004), the limited expression of significant P differences both within and between the basins is not surprising.

Though Ca-bound P tends to dominate P fractions in dry systems, organic P pools are important because they are more readily available to microbes (Cross and Schlesinger, 2001). Though cross-site comparisons of P are confounded by differences in extraction methods and operational definitions, relative percentages of comparable P fractions were determined for comparison between the cold desert system of this study and the hot desert system studied by Cross and Schlesinger (2001). In both cases the top 10 cm of soil were analyzed. Subsequent values for relative P percentage represent averages for both the hot desert (grassland and shrubland) and cold desert (Fryxell basin and Bonney basin) sites. Predictably, Ca-bound phases dominated the P fraction in both systems, comprising 95 and 99% for the cold desert and approximately 80 and 88% for the hot desert. Differences in Al and Fe bound phases (approximately 0 and 3% for the cold desert, respectively, and 4 and 8% for the hot desert, respectively) may be related to the greater clay content and weathering intensity associated with the hot desert. The greatest difference between the sites occurred in the soluble inorganic P phase, with the cold deserts having approximately 1 and 0%, and the hot deserts 11.9 and 6.5%. This difference may be operational (i.e., a function of the different extractants used to define this fraction), real (i.e., a function of the greater impact of plant and microbe mediated P found in the hot desert systems), or some combination thereof. Apparent similarities exist in the comparable labile Po fraction (0.7–1.1% for the cold desert versus 0.9–1.2% for the hot desert), and even though this fraction is a relatively minor component of the total soil P pool, could be an important nutrient source in both systems given the potentially faster turnover compared to other P fractions.

The distribution of soil P among chemical fractions at both the Fryxell and Bonney sites fits the general relationship between weathering intensity and P distribution of other ecosystems, where inorganic P fractions tend to dominate systems of limited weathering. A literature review of P fractions across different soil orders found the lowest organic P and greatest HCl extractable P associated with Aridisols (Cross and Schlesinger, 1995). Syers et al. (1970) showed reductions in Ca-bound P and increases in occluded P concentrations along a weathering gradient of greywacke rock in a North Auckland quarry. Walker and Syers (1976) described increases in non-occluded (i.e., soluble, Al, and Fe-bound) and occluded P with increasing pedogenic development.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Our results showed that soil P content increased from soluble to Ca-bound P at both sites in Taylor Valley. The Fryxell and Bonney sites both had similar Ca-bound P concentrations, indicating apatite-containing parent material is the likely source of phosphate. Overall, the Lake Fryxell site had higher soil P content and higher concentrations of P in the various fractions compared with the Bonney site, suggesting that increased chemical weathering exists at the Fryxell sites, or greater retention of weathered fractions. However, differences in parent material cannot be discounted and further study would help clarify the role of weathering. Within the Fryxell study area, spatial variability in the soluble and occluded P fractions, and correlations among the Al-, Fe-bound, and labile fractions with physical soil properties suggests that there is active weathering and exchange of P among chemical and biological pools. Less spatial variability across the transects and low concentrations of P in Al bound, Fe bound and Po suggest more restricted P weathering in soils of the Bonney basin. Previous work has demonstrated that P can be biologically limiting in lake and stream ecosystems (Priscu, 1995; McKnight et al., 2004), and it is possible that these broad-scale differences in P availability between the sites in Taylor Valley contribute to the different levels of P deficiency observed in Lake Fryxell and Bonney and the higher populations of soil invertebrates in the Fryxell basin (Virginia and Wall, 1999). The turnover of Po, as well as other nutrients, is quite important in contributing to P availability, and thus future studies in this ecosystem should more closely model this fraction over time.


   ACKNOWLEDGMENTS
 
This work was supported by National Science Foundation grants OPP 9810219 and OPP 0096250 and is a contribution to the McMurdo Dry Valleys Long Term Ecological Research (LTER) Program. Raytheon Polar Services and Petroleum Helicopters Incorporated provided logistical support.

Received for publication October 5, 2005.


   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 Blecker, S. W.
Right arrow Articles by Norvell, K. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Blecker, S. W.
Right arrow Articles by Norvell, K. L.
Agricola
Right arrow Articles by Blecker, S. W.
Right arrow Articles by Norvell, K. L.
Related Collections
Right arrow Ecosystem Management
Right arrow Biogeochemical Processes
Right arrow Pedology

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