| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Box 6013, Bilby Research Center, Northern Arizona University, Flagstaff, AZ 86011
b Desert Research Institute, 2215 Raggio Parkway, Reno, NV 89512
c Dep. of Environmental Science, University of California, Riverside, CA 92521
* Corresponding author (kirk.anderson{at}nau.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRadiocarbon Age EstimatesMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: AMS, accelerated mass spectrometry • BP, before present • EC, electrical conductivity • MRT, mean residence time • TL, thermoluminescence
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRadiocarbon Age EstimatesMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Early models of pavement formation focused on the exhumation of clasts by deflation (Springer, 1958; Jessup, 1960; Sharon, 1962; Symmons and Hemming, 1968; Cooke, 1970; Cooke and Warren, 1973; Mabbutt, 1977) or the shrink and swell heave of clasts to the ground surface (Springer, 1958; Johnson and Hester, 1972; Mabbutt, 1977). More recently, the association of vesicular soils with desert pavements on many landforms in the Southwest has been shown to support an accretionary mechanism of pavement development (McFadden et al., 1986; McDonald, 1994). The shift from assuming that desert pavements are zones of erosion to the realization that they can be zones of deposition has caused studies of landscape development in the Mojave Desert to refocus on dust accumulation and soil formation (McFadden et al., 1987; McDonald et al., 1994; Reheis and Kihl, 1995). Examples of accretionary desert pavements on basalt flows of the Cima Volcanic Field include up to 2 m of relatively clast-free, quartz-rich eolian sands between basalt flows and the overlying desert pavements. The uppermost horizon, generally an Av, represents an actively accreting physically and chemically complex layer.
Vesicular horizons were recognized by Marbut (1935) to be common in arid environments. Recently, because of the unique properties and stratigraphic position of Av horizons at the ground surface, investigations are focusing on their effects on decreasing infiltration and increasing runoff, thereby influencing plant-water relationships, soil development, surficial processes, and landscape evolution (Musick, 1973; Wells et al., 1984, 1987; McDonald, 1994). Many researchers have investigated the processes of vesicle formation. Springer (1958) reproduced vesicles from sieved Av material after only five wet and dry cycles. The original Av properties of sieved soils return after 20 to 30 wetting and drying cycles; additional cycles increase the number, size, surface area, sphericity, and continuity of pores (Miller, 1971; Figueira and Stoops, 1983; Figueira, 1984). Evenari et al. (1974) determined that vesicle formation was influenced by the following: texture, wetting front displacement of soil air, surface sealing by crusts or clasts, air trapped when wet, and thermal expansion of air bubbles. Bouza et al. (1993) found that a high exchangeable Na percentage (>15) is an important vesicle formation factor for Patagonian soils. Vesicle stability is aided by CaCO3 (Evenari et al., 1971) and clay coatings (Sullivan and Koppi, 1991). Observations of Av horizons in the Cima Volcanic Field reveal strong columnar and platy soil structure, with individual peds exhibiting textural and chemical zonation from ped exterior to ped interior locations. Qualitative field properties include sandy, low-CaCO3 ped exteriors, and clayey CaCO3-rich ped interiors (McFadden et al., 1984). Using micromorphic, physical, and chemical analyses, Anderson et al. (1994)(this study) observed that vertical and horizontal translocation of fines and solute transport within the Av horizon increase the clay and CaCO3 content of ped interiors.
|
Radiocarbon Age Estimates |
|---|
|
TOPABSTRACTINTRODUCTIONRadiocarbon Age EstimatesMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Because of the importance of determining soil age in geomorphic, pedogenic, and archeological settings, radiocarbon dating of soils continues to be used by many researchers. Wang et al. (1996) obtained model ages based on the 14C activity and rates of younger C additions. Scharpenseel (1979) determined that careful consideration of sample selection and analysis of specific organic matter fractions (humic, fulvic, humin, humic-clay, etc.) could greatly improve interpretation of the radiocarbon age. However, because of the physical and chemical variability of soil processes, it remains unclear which humus fraction is the most representative of the initiation of soil genesis for any particular soil (Stout et al., 1981). Anderson and Wells (1997) reported consistent stratigraphic trends in accelerator mass spectrometry (AMS) radiocarbon ages of vesicular soils on an alluvial fan sequence in the Mojave Desert, suggesting that an age-dependent signature may be preserved in Av soils. Because of the uncertainties of AMS ages on soil organic matter, conservative interpretations using minimum mean residence times are useful within the context of supporting data, as discussed below.
The objectives of the research presented in this paper are to (i) identify the physical and chemical properties of Av peds underlying desert pavements and to (ii) determine the processes active within the Av horizon that may affect desert pavement formation. Detailed physical, chemical, and micromorphological analyses of Av peds enhance understanding of vesicular ped formation and the influence of Av horizon genesis on the development of accretionary desert pavements. Radiocarbon age estimates provide preliminary chronologic control in understanding the timing and rate of Av soil genesis for (i) comparisons with other dated soils, (ii) correlation with potentially important climatic periods, and (iii) determining whether Av horizons form during discrete periods of the geologic past, e.g., Pleistocene or Holocene.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRadiocarbon Age EstimatesMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The basalt flow is 700 m above sea level and grades <2° to the west. The flow surface lies 10 m above the local base level as an erosional remnant (Fig. 1). Vegetation includes creosote bush [Larrea tridentada (Sessé & Moc. ex DC.) Coville], brittle bush (Encelia farinosa), mormon tea (Ephedra trifurca), and joshua tree (Yucca brevifolia Engelm.). Mean annual precipitation is between 12 and 25 cm and the mean annual temperature is 16 to 18°C (National Oceanic and Atomspheiric Administration [NOAA], 1978, p. 570). The basalt flow overlies Pleistocene alluvial fan gravels. A mantle of desert dust overlies the basalt flow (Fig. 1, Tables 1 and 2). Overlying the soil is a tightly interlocking desert pavement mosaic with millimeter-scale distances between well-varnished basaltic boulders and cobbles (McFadden et al., 1989). The <5-mm interclast spaces consist of basaltic fragments derived from local salt weathering of larger clasts and loose quartz-rich sand, silt, and clay eolian deposits. In some locations a thin, cohesive crust has formed. Pebble to small cobble-size clasts generally rest on the surface, whereas larger cobbles can be embedded in the Av horizon up to several centimeters. Poorly preserved pavement fabric occurs in areas where plant growth and animal activity are high. Relief across the surface is generally low, except where bedrock highs are exposed or where drainages have incised.
|
|
In the laboratory, ped domains were defined according to macroscopic characteristics such as Munsell color, texture, carbonate nodules, and soil structure. Peds were subsampled according to the defined domains to quantify physical and chemical variations within peds (Fig. 2) . Based on variations in soil properties, the following seven distinct ped domains were identified and sampled: (1) material adhering to ped tops, (2) material adhering to ped sides, (3) ped top, (4) ped center, (5) ped bottom, (6) ped upper side, and (7) ped lower side (Fig. 2). Material adhering to ped surfaces (Domains 1 and 2) is generally <3 mm thick. Other sampled domains are generally between 1 and 3 cm thick, according to ped size. Subsampled ped domains from single peds provided too little material for subsequent laboratory analysis. Therefore, analyzed samples consisted of combined material from subsampled domains of several peds from each separate site. The exception to this was subsampling for the accelerator mass spectrometry radiocarbon analysis, which was performed on the center domain of two individual peds.
|
Bulk soil samples for AMS radiocarbon age determination were collected from ped centers at Sites I and II. Ped centers were sampled after trimming ped tops and sides until the center domain remained. Pretreatment for both samples included four steps: removal of macroscopic roots, acid washes in hot HCl to remove inorganic carbonates, alkali washes in NaOH to remove secondary organic acids, and final acid wash to neutralize the sample. These treatments were designed to remove the acid and alkali-soluble fraction, leaving the alkali-insoluble organic fraction for age determination. Pretreatment and analysis were performed by Beta Analytic, Inc. (Florida).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRadiocarbon Age EstimatesMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Channel pores are absent from exterior positions of ped tops and ped sides. Interior positions of ped tops, sides, and centers have interconnected channel pores, but they are commonly filled with sediment. In the lower positions of ped centers, ped lower sides, and ped bottoms, channel pores are common. Channel pores are often connected to the macropores between ped faces, forming a continuous conduit from the ground surface to interior ped domains. Channel pores in ped bottoms can be interconnected to a great extent, forming platy soil structure. Argillans and siltans commonly line the lower surfaces of platy structure, whereas the upper surfaces may have a coating of CaCO3, called calcitans (Fig. 3) .
|
|
|
Trends in particle-size distribution and CaCO3 content also occur along the topographic transect, from Site I to Site III. The silt + clay content for material adhering to ped tops increases from 34% at the upper site (Site I) to 40% at the lower site (Site III). Similarly, the silt + clay content of material adhering to ped faces increases from 40 to 45% and the clay content for ped centers increases from 27 to 40% between Site I and Site III. An increase in CaCO3 content in downslope positions is noted for some peds in Domains 1, 3, and 4.
The results of pH and electrical conductivity (EC) analysis indicate alkaline and saline conditions (Table 3); the pH values are 9.2 or greater. No noticeable pattern in pH values occurred within peds or along the topographic transect. The EC values indicate saline conditions (EC > 4 dS m-1) for many of the subsampled domains. Electrical conductivity of <2 dS m-1 occurs at all sites for material adhering to ped tops. The highest EC, 9.4 dS m-1, is in the ped center of the lowest topographic position, Site III.
Radiocarbon Ages
Accelerator mass spectrometry radiocarbon analysis of fine-grained bulk soil samples from two subsampled ped centers provides similar age determinations. Conventional, uncalibrated radiocarbon ages are 4530 ± 50 BP (Beta-93902) at Site I, and 4570 ± 50 BP (Beta-91527) at Site II, 25 m apart. The two sigma (95% probability) tree-ring calibrated radiocarbon age estimate from Site I ranges from 5315 to 4995 years before present (BP), whereas the age from Site II ranges from 5440 to 5045 years BP. The calibrated radiocarbon ages, interpreted as MRTs, are comparable with a 5000 ± 1000 BP thermoluminescence age obtained from an Av horizon from a similar-aged basalt flow nearby (McFadden et al., 1998). They are also useful for discussing relationships with regional climatic (Spaulding, 1991) and eolian chronologies (Clarke, 1994).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRadiocarbon Age EstimatesMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Peds contain two domains of low clay and CaCO3 content and five domains of higher clay and CaCO3. Domains 1 and 2 have consistently lower clay and CaCO3, and higher sand and silt content than the other domains. Ped centers have higher clay and CaCO3 contents than all other ped domains, with up to 15 times the CaCO3 concentration of exterior domains. Also, micromorphic investigations show argillans and calcitans lining horizontal channels and platy structural features in the lower portions of ped centers. Numerous irregular argillaceous and skeletal glaebules, and sand-filled granotubules in the upper portions of ped centers, indicate that this domain is relatively inactive with regard to the lateral translocation of sediment. Pedotubules filled with sandy material suggest that as upper ped centers become filled with sediment, the active zone of translocation shifts to lower ped domains, such as ped bottoms, where platy structures dominate.
Ped bottom domains have calcitans and layered argillans that represent distinct episodes of lateral clay translocation. Fining upward microbedding of silt and sand in ped bottoms indicates lateral sediment transport by suspension along the platy structure, which represents the active domain of lateral clay, silt, and sand translocation (see Fig. 3). Particle-size distribution, CaCO3 content, and micromorphology of Av peds indicate that soluble and easily translocated constituents are preferentially removed from the outer two domains, consequently accumulating in the ped center, upper-side, lower-side, and bottom domains.
From a landscape-scale perspective, the CaCO3 and clay data may provide insights into processes involved in the progressive reduction of relief on desert pavement surfaces. Incipient desert pavement landscapes have high constructional relief because of irregular topography inherited during initial emplacement, whereas mature pavement landscapes (middle to early Holocene and late Pleistocene) have reduced constructional relief (Wells et al., 1984; Bull, 1991). As pavements develop, topographic highs are reduced by salt weathering, rain splash, and colluviation, and topographic lows are filled with slopewash colluvial and eolian material. A measurable increase in both fines and CaCO3 occurs from Site I to Site III (see Fig. 1). Site III, at the topographically lowest position, has consistently higher silt and clay contents for many ped domains, and clayey, CaCO3-rich ped centers. Processes related to eolian deposition, surface transport of fines via overland flow, and subsurface transport of solutes to lower topographic positions may be responsible for the observed increases. The amount of clay and CaCO3 incorporated into the topographically lower soils may increase the volume of material in these positions at a greater rate than slightly higher positions, thereby aiding in relief reduction of increasingly older desert pavement surfaces.
Ages of Av Peds
The processes of soil organic matter turnover and the subsequent addition of younger C to the soil system do not conform to the basic assumption that "carbon isotopic ratios have not been altered except by 14C decay since the sampled material ceased to be an active part of the carbon reservoir" (Taylor, 1987). Because surface soils are an active part of the C reservoir, that is, younger C continues to be added to the soil during pedogenesis, they are very difficult to interpret. Radiocarbon ages on soils result from a mixture of material being added daily to fractions already in the soil, perhaps for thousands of years (Birkeland, 1999). Ages determined by bulk soil radiocarbon analysis are the average ages of all organic constituents in the soil, excluding those removed through pretreatment processes. In this context, MRTs do not represent steady-state systems because (i) the rates of additions and removals from the soil system may exhibit considerable temporal variation, and (ii) organic matter steady states are not attained in desert soils for a very long time (perhaps 300000 yr) because of the relatively slow rates of organic matter additions and removals (Wang et al., 1996). Although such age determinations can be problematic (Stout et al., 1981), limited interpretations can be made using MRTs as minimum ages (Martel and Paul, 1974; Goh et al., 1977; Scharpenseel, 1979). Some researchers think that specific organic fractions can provide reliable oldest ages, and therefore the closest to the inception of soil genesis (Goh et al., 1977; Holliday et al., 1985), while others have determined that results are neither repeatable nor predictable from one soil to another (Gilet-Blein et al., 1980).
Most studies of soil radiocarbon ages identify the potential source of young organic C as continued growth and decay of in situ organic matter; few studies have attempted to account for organic matter additions from eolian sources, even though Reheis and Kihl (1995) measured up to 23.5% organic matter from dust traps in the Cima Volcanic Field and 45% in other portions of the Mojave Desert. Given that Av soils in the Cima Volcanic Field have eolian origins (McFadden et al., 1984, 1987, 1998), and that >20% of desert dust consists of organic matter, it must be assumed that a significant portion of organic matter added to desert soils is from eolian sources. Therefore, radiocarbon ages determined on any soil fraction must take this into account. Although the origin of organic C for bulk AMS analysis of the sampled Av peds remains unclear and only speculative, the possible sources, and interpretations, must attempt to explain the contributions from both in situ organic matter production and eolian additions.
A conservative interpretation of the AMS ages of Av peds can be made by comparing them with (i) an independent thermoluminescence (TL) age on Av soils, (ii) regional eolian activity, and (iii) periods of regional aridity. Preliminary inferences can thus be made regarding increased aridity, eolian activity, and enhanced vesicular soil formation during the middle Holocene.
Because the radiocarbon ages are interpreted as minimum age estimates, Av soil genesis probably began sometime prior to the 5440 to 4995 BP period. The 5000 ± 1000 BP TL age on quartz sands from an Av soil ped on a nearby 560 ka basalt flow (McFadden et al., 1998) supports these ages. The TL age on eolian quartz sand suggests an influx of eolian material onto the basalt flows in the Cima Volcanic Field during the middle Holocene, influencing vesicular soil formation by adding eolian fines to the developing Av horizon. Considering the organic matter data from Reheis et al. (1995) discussed above, it is possible that increased amounts of eolian organic matter were also added to the soil system during this time. This might explain the correspondence between AMS and TL dating methods. Resulting ages from both methods could be interpreted as MRTs, given the dynamic processes active in the Av horizons, where organic matter and quartz grains are introduced to ped centers by the processes described earlier. Perhaps enough eolian organic matter was added during this time to lower the MRT age towards the 
5000-BP period.
The AMS and TL ages correspond to middle Holocene eolian activity at Silver Lake playa, 20 km upwind from the Cima Volcanic Field, where an eolian deposit is bracketed by 8350- and 3620-BP radiocarbon ages (Wells et al., 1989). Eolian deposits at the Kelso sand dunes, 10 km south of the project area, have been dated using optically stimulated luminescence to between 8420 and 3500 BP (Clarke, 1994). Spaulding (1991), using packrat midden data, suggested that a period of increased aridity occurred between 8000 and 4000 BP in portions of the Mojave Desert.
The correspondence between AMS and TL dated Av horizons, with periods of increased eolian activity and regional aridity, suggests that the Av horizons investigated in this study may, in part, be influenced by a middle Holocene increase in eolian sediment and organic matter mobilization, deposition, and incorporation into surface Av soils. Wells et al. (1984)(1987) and McFadden et al. (1984)( 1986) have suggested that desiccation of Silver Lake at the end of the Pleistocene dramatically increased the supply of dust to the Cima Volcanic Field, greatly affecting soil genesis and clast fracturing by increasing the amount of salt-rich eolian fines. They postulated that Av horizon development on the basalt flows in the Cima Volcanic Field may have been initiated or enhanced during the Pleistocene to Holocene transition and its increased eolian activity. The results from chemical, physical, and chronologic analyses presented above suggest that the middle Holocene may also have been a time of increased eolian deposition, affecting pedogenic development of the already formed Av horizon, with a corresponding influence on desert pavement formation.
Implications for Desert Pavement Formation
The processes of sediment incorporation below desert pavement on well-developed soils include clay and CaCO3 accumulation in the upper several centimeters of the vesicular soil, causing horizon thickening and vertical accretion. Pavement clasts above this accreting Av horizon ride upward as material below continues to accumulate. Translocation and subsequent storage of eolian fines and salts in ped interiors provides a mechanism for cumulic soil development below pavement clasts, whereby material is added to the soil via lateral macropore conduits.
The initial phase of the process of eolian accumulation and pavement development remains speculative, but is probably influenced by adhesion and surface tension forces as water and sediment act to form laminar surface crusts that adhere to clast bottoms (Fig. 4) . A second phase includes continued eolian influx and CaCO3 accumulation, allowing formation of incipient vesicles and pedogenic alteration. When soil properties reach certain physical and chemical thresholds—based largely on clay type and amount, and salt and Na concentration—shrinking and swelling processes may enhance clast lifting. A critical third phase of development is related to the formation of well-defined soil structure by shrink-swell activity, where vertical pores between columnar peds are connected to platy structure at ped bottoms. When this phase is attained, further eolian material can be added to both the ped tops and ped interior positions.
An increase in clay and CaCO3 in Av horizons in topographic lows suggests that rates of cumulic soil development and associated pavement-clast lifting are accelerated in lower topographic positions, increasing the rate of relief reduction. In addition, as Av horizons continue to accumulate material, underlying soil horizons continue to develop as well, although the relations between Av genesis and subsoil development remain poorly understood. McFadden et al. (1987) found evidence that subsoil Btvk morphologies support the idea of the transformation of Av's to B horizons. However, increasing Av thickness with time on alluvial fan chronosequences seem to indicate rather long periods of Av stability and continued development (Anderson, 1999; McDonald, 1994). McDonald (1994) finds evidence that the Av horizon may, in fact, also rise vertically as material accumulates below both the pavement clasts and the existing Av layer.
Organic matter is continuously added to the soil by both in situ processes and eolian deposition. Preliminary interpretations of radiocarbon age estimates, and correlation with regional chronologies, suggest that the middle Holocene was a period of increased additions of datable material to the soil, notably quartz sand and organic C. Such an increase may provide enough material to cause a decrease of the mean residence time AMS and TL age-estimates, moving towards concordant chronologies. These concordant AMS and TL ages on Av peds correspond to a period of increased aridity and eolian activity throughout the Mojave Desert during the middle Holocene, with important influences on Av soil genesis and desert pavement development.
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 9, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRadiocarbon Age EstimatesMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. C. Graham, D. R. Hirmas, Y. A. Wood, and C. Amrhein Large near-surface nitrate pools in soils capped by desert pavement in the Mojave Desert, California Geology, March 1, 2008; 36(3): 259 - 262. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. R. Muhs, J. R. Budahn, D. L. Johnson, M. Reheis, J. Beann, G. Skipp, E. Fisher, and J. A. Jones Geochemical evidence for airborne dust additions to soils in Channel Islands National Park, California GSA Bulletin, January 1, 2008; 120(1-2): 106 - 126. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. A. Valentine and C. D. Harrington Clast size controls and longevity of Pleistocene desert pavements at Lathrop Wells and Red Cone volcanoes, southern Nevada Geology, July 1, 2006; 34(7): 533 - 536. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. W. Briggs and S. G. Wesnousky Late Pleistocene and Holocene Paleoearthquake Activity of the Olinghouse Fault Zone, Nevada Bulletin of the Seismological Society of America, August 1, 2005; 95(4): 1301 - 1313. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Bullard Arid geomorphology Progress in Physical Geography, March 1, 2004; 28(1): 130 - 144. [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
