Soil Geomorphic Evidence

In 1957, the U.S. Department of Agriculture began a study of soil-geomorphic relationships in a 400-square-mile (1024 km2) area surrounding Las Cruces (Gile et al. 1981). This project, termed the Desert Project, includes an area that overlaps the southern portion of the Jornada Experimental Range and Chihuahuan Desert Rangeland Research Center (figure 17.1). A major objective of the Desert Project was to map soils and geomorphic surfaces. In addition to Aridisol, Entisol, and Mollisol soil types, geomorphic surfaces were identified for three physiographic units: the Rio Grande valley border, the piedmont slope, and the basin floor (table 17.2). Ages of the geomorphic surfaces are based on a combination of radiocarbon dates of charcoal and pedogenic carbonate (Gile et al. 1981), K-Ar dates of lava flows (Seager et al. 1984), Ar-Ar dates of volcanic ash and pumice (Mack et al. 1996), paleomagnetism dates (Mack et al. 1993), and megafauna fossils (Hawley et al. 1969; Tedford 1981; Morgan et al. 1998).

Geomorphic surfaces in the valley border astride the Rio Grande floodplain occur as stepped fan terraces. Each surface rises from the Rio Grande floodplain, which acts as the local base level. Progressively higher steps are progressively older and contain progressively more pedogenic carbonate (Gile et al. 1966; Machette 1985). The piedmont slope surfaces also occur as progressively older steps in some areas. In other areas, younger sediments bury older surfaces, resulting in stacked sequences of buried paleosols.

Erosion as an Indicator of Bioclimatic Variability. Based on modern measurements, erosion is greater in shrublands than in grasslands (Abrahams et al. 1995)

Alluvial Fan VII-.m: ! AI[uYisl Fan Fol:an i" tlilE Ciully T^nttSO-L >1 Kjiir.^ji, 7:cii

Alluvial Fan VII-.m: ! AI[uYisl Fan Fol:an i" tlilE Ciully T^nttSO-L >1 Kjiir.^ji, 7:cii

Figure 17.5 Comparison of S13C and S18O values of pedogenic carbonates across the Organ-Paleosol boundary on piedmont slope (i.e. Alluvial Fan) and basin floor (i.e. Eolian) sites. Isotopic compositions are expressed as per mil relative to the PDB standard. Percent C4 vegetation is based on a model by Cerling (1984). The 14C ages, expressed as years b.p. x 1000, are of carbonate unless stated otherwise. The ages of carbonates are less reliable than dates of charcoal (Gile et al. 1981) (after Buck and Monger 1999).

Figure 17.5 Comparison of S13C and S18O values of pedogenic carbonates across the Organ-Paleosol boundary on piedmont slope (i.e. Alluvial Fan) and basin floor (i.e. Eolian) sites. Isotopic compositions are expressed as per mil relative to the PDB standard. Percent C4 vegetation is based on a model by Cerling (1984). The 14C ages, expressed as years b.p. x 1000, are of carbonate unless stated otherwise. The ages of carbonates are less reliable than dates of charcoal (Gile et al. 1981) (after Buck and Monger 1999).

because more bare soil is exposed in shrublands. Though arid, there is enough rain at the Jornada LTER site to cause significant amounts of erosion, especially in the summer monsoon season when thunderstorms are common. Consequently, a working hypothesis to explain episodic base-level changes and buried paleosols was given by Ruhe (1962) and Gile and Hawley (1966). The hypothesis states that periods of erosion and sedimentation correspond to periods of aridity when shrubland dominate, in contrast to periods of landscape stability and soil formation when grasslands or woodlands dominate.

This hypothesis follows the Langbein-Schumm curve, which emphasizes the importance of both vegetation and rainfall on erosion (Langbein and Schumm 1958). In their curve, sediment yield is low in extremely arid lands (<100 mm rainfall), but rises steeply and reaches a maximum near the arid-semiarid boundary (at about 300 mm rainfall). At progressively greater amounts of rainfall, erosion and sediment yield decline. At the Jornada Experimental Range, the average annual rainfall is 247 mm (9.72 in), which is near the Langbein-Schumm maximum of 300 mm, though some years are less (e.g., 90 mm) and some years are more (e.g., 475 mm) (figure 17.2). Thus, applying the Langbein-Schumm model to the paleosol record at the Jornada piedmont slope site, periods of high erosion-sedimentation and burial of soils downslope may represent climatic conditions similar to today. Periods of landscape stability and soil formation probably represent climatic conditions wetter than today.

In addition to buried paleosols, arroyo cutting can provide insight about past climates. In southeast Arizona, for example, a major period of arroyo cutting occurred in the middle Holocene Altithermal period when temperatures were at their postglacial high and effective moisture at its postglacial low (Waters and Haynes 2001). This period of major arroyo cutting in Arizona corresponds to deposition of the Fillmore and Organ unit at the Jornada site (table 17.2). After about 4000 years ago, less pronounced, but more frequent, arroyo cutting occurred in Arizona, which Waters and Hayes (2001) attributed to the development of increased El Niño frequency and strength.

At a longer timescale, the geomorphic surfaces astride the Rio Grande have been linked to climate variability by the following cyclic-entrenchment model (Hawley 1975). (1) The Rio Grande downcut during glacial periods, when greater rainfall gave rise to more water in the river and denser vegetative cover on the land. During this interval, the river channel carried a lower sediment load and had a greater capacity to entrench. (2) The Rio Grande backfilled during waning glacial and early interglacial times, when reduced rainfall gave rise to less water in the river and less vegetative cover on land. During this interval, the river carried a greater sediment load, lost its ability to entrench, and aggraded. (3) Aggradation ceased and the base level stabilized during the latter parts of interglacial times, until (4) the cycle began again with renewed downcutting in response to the renewed waxing phase of the next glacial cycle. Although this model provides a general explanation of links between rainfall and erosion, base level changes are also a function of stream slope, sediment size, hydraulic roughness, stream discharge, and bed load (Bull 1991). In some areas water table depths are also important.

A model, similar to the one for the Rio Grande, has been used to explain geo-

morphic surfaces on piedmont slopes (e.g., Hawley et al. 1976; Gile et al. 1981). (1) During periods of aridity that correspond to interglacial periods, vegetative density declined, bare ground increased, and erosion-sedimentation increased. (2) During periods of more effective precipitation during glacial periods, vegetative cover increased, bare ground decreased, and erosion-sedimentation decreased. (3) With the discontinuance of erosion-sedimentation, landscape stability ensued, and pedoge-nesis produced soil horizons.

Coppice Dune Formation as an Indicator of Vegetation Change. In historical times, the relationship between vegetation change and wind erosion caused the formation of coppice dunes. Coppice dunes are mounds of sand held together by shrubs, mainly mesquite. These dunes form when laterally blowing sand encounters and accumulates around mesquite plants. As the sand aggrades, the mesquite plant grows through the rising dune. During this process, mesquite stems send out roots and exploit the resources of the accreting sand.

Ecosystems on sandy soils are some of the most fragile ecosystems because sandy soils are more vulnerable to wind erosion than rocky or clayey soils. Land surveyor notes for 1857 relating to sandy sites in southern New Mexico lacked any mention of coppice dunes at section corners west of Las Cruces. When the same section corners were revisited in the 1960s, the level, smooth black grama land recorded for 1857 had changed to steep-sided coppice dunes (Gile 1966). The earliest aerial photography of the Jornada region was taken in 1936. Coppice dunes are present in these photographs, so the dunes must have formed between the late-1850s and 1936 (Gile 1966). Land survey notes indicate similar relations in the Jornada Experimental Range (Buffington and Herbel 1965).

Pedogenic Carbonate as an Indicator of Climate Variability. Climate, as one of the five soil-forming factors, exerts major control on the mineralogical nature of the soil profile (Dokuchaev 1883; Jenny 1941). This is particularly the case with horizons of pedogenic carbonate because (1) its presence is generally restricted to soils of dry climates and (2) its depth is a function of rainfall (Arkely 1963; Mc-Fadden and Tinsely 1985; Mayer et al. 1988). With respect to carbonate's presence in the soil, the boundary between carbonate-accumulating soils and non-carbonate-accumulating soils in the Western United States corresponds to an annual rainfall of about 500 mm (20 in) (Birkeland 1999). In the Midwest, however, carbonate can occur in limestone soils that receive up to 800-900 mm of mean annual rainfall (Jenny and Leonard 1934; Nordt et al. 1998).

Carbonate is present in modern soils and buried paleosols at the Jornada LTER site. Trenches and natural exposures reveal multiple buried paleosols with calcic horizons that date to at least 400,000 years b.p. based on correlation to the Jornada I surface (Gile et al. 1981). The presence of carbonate in these stratigraphic sections suggests, by deduction, that wet periods received less than 500 mm of annual rainfall during this 400,000-year period.

With respect to the depth of carbonate in soil, the zone of carbonate accumulation not only depends on the amount of rainfall, but also on texture, parent material, and erosion. Nevertheless, some trends have been observed for soils in general (Jenny and Leonard 1934) and for soils at the Jornada LTER site (Gile 1977). By tracing carbonate depth in soils of the Organ geomorphic surface from arid lower elevations to semiarid higher elevations, Gile (1977) documented a progressive deepening of carbonate at the Jornada site. At the lower elevations (220 mm of rainfall), carbonate depth was 25 cm below the surface. At the higher elevations (350 mm of rainfall), carbonate depth was 104 cm below the surface.

With respect to carbonate depth as a paleoclimatic indicator, carbonate in ancient soil profiles at the Jornada LTER probably reflect the wet-dry cycles of much of the late Quaternary. Such soils exist on stable geomorphic surfaces of middle Pleistocene age and older. These ancient polygenetic soils would have experienced multiple glacial-interglacial climates based on a 100,000-year glacial-interglacial periodicity (Morrison 1991). Vertical pipes crosscut many petrocalcic horizons in these ancient soils (figure 17.6). These pipes, according to the current working hypothesis, originated as a result of dissolution during wet glacial periods (Gile et al. 1981). In this sense, the pipes are small-scale versions of karst topography—that is, they are pedokarst. According to this hypothesis, greater rainfall would lead to deeper water penetration through the profile, movement across the laminar cap of the petrocalcic horizon, and down the pipes. As a test of this hypothesis, radiocarbon ages of both carbonate and organic matter in layers of the laminar cap have been measured (Gile et al. 1981, p. 122). Radiocarbon ages of these materials range from about 32,000 to 21,000 14C years b.p., near the time of glaciopluvial conditions based on lake studies.

Moreover, a shift from a wetter to a drier climate would cause an upward shift in the depth of wetting. This would lift the zone of carbonate formation from the laminar zone of the petrocalcic horizon to a higher zone in the soil. This shallower zone would extend horizontally across pipes (figure 17.6) and would be at a similar depth to carbonate accumulation in neighboring soils known to be middle Holocene based on radiocarbon dating. As a test of this hypothesis, radiocarbon ages of this shallow-zone carbonate have been measured and are middle Holocene in age, from about 4,940 to 1,600 14C years b.p. (Gile et al 1981, p. 149; Monger et al. 1998, p. 152).

0 0

Post a comment