Timing and Extent of Late Quaternary Paleolakes in ... - Science Direct

QUATERNARY RESEARCH ARTICLE NO.

47, 306–315 (1997)

QR971896

Timing and Extent of Late Quaternary Paleolakes in the Trans-Pecos Closed Basin, West Texas and South-Central New Mexico David E. Wilkins and Donald R. Currey Limneotectonics Laboratory, Department of Geography, University of Utah, 270 Orson Spencer Hall, Salt Lake City, Utah 84112 Received May 2, 1996

The Trans-Pecos Closed Basin is a hydrographically closed region covering 20,000 km2 centered on Salt Basin, 160 km east of El Paso, Texas. Geomorphic and limnetic evidence have been used to identify four major highstands for Lake King during the last glacial maximum (LGM). Additional geomorphic features from a second, recently identified, paleolake, Lake Sacramento, have been found in the Beargrass subbasin, a nested subbasin approximately 75 km northwest of Salt Basin. Radiocarbon ages of the organic material in Lake King sediments date four abrupt climate changes and rapid lacustrine transgressions during the LGM with a quasiperiodicity of 2000 yr. Geomorphic evidence in the Beargrass subbasin identifies lake cycles contemporaneous with those in Lake King. The dates for these transgressions correlate with the dates of freshening events identified by researchers in paleolake basins elsewhere in New Mexico. The quasi-periodicity of the events approximates that of Dansgaard–Oeschger events identified from Greenland ice cores. The contemporaneity of the Trans-Pecos transgressions with transgressive events in other basins in the region suggests that paleolakes in the region were in phase with respect to abrupt climate changes during the latter stages of the LGM. q 1997 University of Washington.

INTRODUCTION

Reconstructions of late Quaternary climates in the intermountain region of western North America (WNA) are based on the interpretation of a spatial and temporal aggregation of proxy evidence and morphometric (e.g., hydrometric, limnometric, glaciometric) data. In the case of paleolakes in WNA, evidence for the presence of contemporaneous permanent or persistent (i.e., 102 to 103 yr duration) lakes over a wide range of latitudes and elevations (e.g., lakes Bonneville and Lahontan, and the Owens River lake system; Benson et al. (1990)) and under different basin configurations, i.e., with respect to tributaries or basin geometry (e.g., Currey, 1991; Enzel et al., 1992), allow for a generalization of climate conditions at a broad spatial and temporal resolution (e.g., COHMAP, 1988). However, this same evidence shows that the climate in the region was not homogeneous, but demonstrates a spatial heterogeneity influenced by local responses to broad climate forcing (Mock and Bartlein, 1995). The

QR 1896

/

a609$$$161

STUDY AREA

The Trans-Pecos Closed Basin is an internally drained, hydrographically closed region located between the Pecos River and the Rio Grande drainages in far west Texas and south central New Mexico (Fig. 1), representing the farthest eastward extent of the Basin and Range physiographic province (Fenneman, 1931). It encloses over 20,000 km2, with elevations ranging from 2835 m on Sacramento Peak to 1087 m at the deepest point in Salt Basin, a northwest–southeast trending half-graben located approximately 160 km east of El Paso. The study area extends from Marfa, Texas, on the south to Sacramento Peak near Alamagordo, New Mexico, on the north, and from the Delaware and Guadalupe mountains (Pecos River divide) on the east to the Hueco Mountains (Rio Grande divide) and the Tularosa Basin on the west. Similar to other basins in WNA, the Trans-Pecos Closed Basin contained permanent lakes at various times during the last glacial maximum (LGM). The basin terminus, Salt Basin, contains evidence from late Pleistocene lakes—this evidence consists of erosional remnants of breached lacustrine sediments (locally known as ‘‘islands’’) interspersed between evaporite-encrusted playas. Fragments of shorelines from Pleistocene lakes are preserved around the periphery of Salt Basin 8 to 14 m above the modern basin floor (average elevation 1100 m). Modern climates and biomes in the study area vary from the arid, upper Chihuahuan desertscrub at the lower elevations to the humid, mixed conifer forests (e.g., Pinus–Picea–Abies sp.) at the extreme upper elevations (Tuan et al., 1973; Van Devender et al., 1984). Average annual precipitation totals range from 280 mm at Salt Flat, Texas, near the floor of Salt Basin, to more than 760 mm at Sacramento Peak (Tuan et al., 1973). Runoff seldom reaches the floor

306

0033-5894/97 $25.00 Copyright q 1997 by the University of Washington. All rights of reproduction in any form reserved.

AID

spatial and temporal resolution of late Quaternary regional climate models for WNA improves as the extent and timing of smaller paleolakes on the periphery of the region, e.g., Lake Cochise (Waters, 1989) and Lake Mojave (Enzel et al., 1992), becomes known.

04-21-97 15:41:22

qra

AP: QR

PALEOLAKE LEVELS IN TEXAS AND NEW MEXICO

307

FIG. 1. Map showing regional setting of Trans-Pecos Closed Basin.

of Salt Basin—low-recurrence rainfall events sometimes result in portions of the basin being covered by shallow lakes which seldom persist for more than a few weeks under high evaporation rates in the basin, estimated to average between 175 and 200 cm/yr (Bjorklund, 1957; Kohler et al., 1959). Evaporite deposits (e.g., CaSO4r2H2O) in the playa sediments provide evidence of the hydrographic closure. No perennial surface water reaches the floor of Salt Basin graben, but the graben does receive significant groundwater contributions from the Permian Bone Spring limestone underlying the Sacramento and Guadalupe mountains (Bjorklund, 1957; Boyd and Kreitler, 1986). Bjorklund (1957) states that the water table is very close to or at the surface of the playas. Surface expressions of groundwater infiltration are visible as minor depressions and sinks on the Diablo Plateau to the west and in the Sacramento Uplands to the north. The basin has been tectonically active during the late Quaternary (Muehlberger et al., 1978; Reilinger et al., 1980)— fault scarps can be observed on aerial photos and on the ground. The tectonics have divided the Salt Basin graben into a northern and a southern section. The northern section of the graben is wider, with numerous wide, very flat, modern playa surfaces (locally termed ‘‘lakes’’) connected around the islands. The southern section of the graben is more compact, with a single, deep, playa-floored depression. The average playa floor elevation in the north is around 1100

Geomorphic research in Salt Basin was initiated by King (1948). King proposed that beach ridges on the northeastern margin of Salt Basin were evidence of two late Pleistocene lake maxima. Dubbed ‘‘Lake King’’ by Miller (1981), these lakes (and remnants of other lakes or lake stages that Miller thought dated to the LGM) in Salt Basin have received little mention in the literature. Additional lacustrine features, recently identified by Hawley (1993), in the Beargrass subbasin nested against the base of the Sacramento Mountains to the northwest of Salt Basin are remnants of Lake Sacramento, thought to have been contemporaneous with Lake King. This nested subbasin was first correctly identified by Bjorklund (1957). Subsequent studies and maps of the region, however, overlooked it, indicating instead that the Sacramento River flowed directly into Salt Basin and Lake King. Most recently, Hawley (1993) correctly identified both the closed nature of the basin and its paleolake features. Previous paleoenvironmental research in the basin (i.e., Van Devender et al., 1984, 1979; Wells, 1978) used biological proxies including plant and animal macrofossils collected from middens (e.g., Neotoma sp.; Van Devender et al., 1984) in their interpretations of basin paleoclimates. These studies

AID

qra

QR 1896

/

a609$$$162

04-21-97 15:41:22

m, while the southern section, with the one central playa, has a floor elevation of 1087 m. PREVIOUS RESEARCH

AP: QR

308

WILKINS AND CURREY

provide floristic reconstructions of paleoenvironments during the LGM, reporting a lowering of life zones they attribute to a significantly cooler, more mesic climate in the region. OBJECTIVES AND METHODS

How the basin responded to LGM climates has been overlooked. This paper uses the geomorphic expression of responses in the basin’s lacustrine geomorphic subenvironments to reconstruct the timing, duration, and extent of late Quaternary climates, drawing on geomorphic and stratigraphic evidence to make inferences regarding the basin paleolimnology. The primary objective of this study is to explore evidence of hydrologic and climatic changes that occurred during the LGM. The aggregation of data from different areas and subenvironments in the basin provides the foundation for an evaluation of late Quaternary palaeoclimates for the region. The emerging paleoenvironmental history of this basin contributes to both a better understanding of the spatial heterogeneity of climates (e.g., Mock and Bartlein, 1995), while at the same time supporting the idea of coherency of the regional climate, in the western United States during the last glacial maximum. The primary focus of our research has been on features associated with lacustrine activity. Field work provided a collection of spatial (e.g., surveyed shoreline elevations) and geochronological (i.e., 14C dates) data of selected stratigraphic and geomorphic features that are used here to reconstruct late Quaternary paleoclimates.

FIG. 2. Lake King geochemical hydrographic model. Dolomite production occurred during periods of intense evaporation, resulting in shallowing of the lake. The lake surface elevation portrayed for dolomite production is arbitrarily placed between the elevations of the highest black mat and the lowest identified shoreline. The blackening of the dolomite occurred following abrupt lake level rise and formation of anoxic bottom conditions. Periods between highstand and lowstand in each cycle represent quasisteady state conditions and formation of gypsum and calcite laminae.

The lacustrine sediments consist of finely laminated, centimeter- and subcentimeter-scale couplets of olive- and graycolored evaporite laminae. These are described by both Friedman (1966) and by Hussain et al. (1988), who report that the couplets consist of layers composed primarily of calcite, gypsum, and halite, Friedman noting that aragonite and dolomite may also be locally abundant. Hussain et al. (1988) identify the lighter (olive) laminae as being gypsumrich and the darker (gray) laminae as being micritic and

richer in organics. These pairs of laminae appear varvelike—that is, there is a uniformity in their occurrence that implies a regularity in their deposition. Spatial variability of modern groundwater chemistry provides an insight into the local controls operating on the production of these alternating laminae. Wells closer to the Sacramento Mountains, i.e., the primary recharge zone for the local aquifer, have levels of bicarbonate higher with respect to sulfate than wells sampled nearer the playas; Bjorklund (1957) reveals the latter to be sulfate-enriched (with respect to bicarbonate) by an average ratio of 2:1. Boyd and Kreitler (1986) report the groundwater in the modern playa subenvironment is bicarbonate-poor and sulfate-rich and conclude that modern groundwater evolves to a Na–Mg– SO4 –Cl brine in the center of the basin (i.e., saline pan; Hardie et al., 1978) as carbonates are precipitated in the playa periphery and gypsum in and on the playa surface. Using the variability of groundwater as a modern analog, the varve-like couplets are believed to have formed under perennial lake conditions with seasonal fluctuations of lake level and associated lake chemistry, such as described by Warren (1989, p. 111). In our model geochemical hydrograph of LGM Lake King (Fig. 2), seasonal runoff and groundwater recharge would have increased the levels of calcium and total alkalinity (i.e., HCO3 / CO3) in solution and flushed organic matter into the lake from the surrounding basin. The result was the formation and deposition of the organic-rich calcite laminae. As this early-season freshening tapered off and total alkalinity levels in the lake were de-

AID

qra

TRANS-PECOS PALEOLAKE GEOMORPHOLOGY AND STRATIGRAPHY

Evidence for late Quaternary paleolakes in the basin includes beach ridges and spits formed by open-water processes as well as badlands exposures of lacustrine sediments. The reconstruction of the timing and extent of paleolakes in the Trans-Pecos is based on both the radiocarbon ages of Lake King sediments sampled from basin-center pits and on our interpretation of the degree of geomorphic preservation of Lake King and Lake Sacramento shorelines. Stratigraphic and Sedimentary Evidence

QR 1896

/

a609$$$162

04-21-97 15:41:22

AP: QR

309


TABLE 1 Radiocarbon Ages of Black Mat Sediments Sample Black Black Black Black Black

Mat Mat Mat Mat Mat

1a 1b 2 3 4

Elevation (m)

Lab No.

1100 1100 1101 1102 1100

Beta-83415 Beta-83417 Beta-77547 Beta-78281 Beta-88597

Agea (14C yr B.P.) 24,670 22,570 19,090 17,180 15,940

{ { { { {

330 180 200 180 320

Material dated TIC (inorganic carbonate) TOC (algal layer) TOC (bulk organic) TOC (algal layer) TOC (algal layer)

pleted with respect to sulfate, deposition of calcite gave way to the deposition of gypsum. Distinctive layers of black sediments separate packages of these couplets in the lacustrine stratigraphic record. Boyd and Kreitler (1986) identify one of these layers which they called the ‘‘black mat,’’ but only in its context as a marker bed. These layers are also identified by Friedman (1966) as late Quaternary dolomite. His model for the production of the dolomite describes intense evaporation in the shallow basin margins resulting in dense brines enriched in magnesium with respect to calcium. Friedman supports his evaporation scenario by noting the enrichment in 18O in the dolomite-dominant layers compared to the dolomite subordinate layers. These brines subsequently sink and flow to the basin center where previously precipitated calcite is replaced by dolomite. Using Friedman’s model, the production of dolomite suggests a period of diminished effective precipitation in the region that resulted in a much shallower lake. The blackening of the dolomite corresponds to the abrupt termination of these drier periods, as a rapid increase in freshwater inflows resulted in stratification of the limnion. The interruption of normal thermohaline circulation by this freshwater infusion produced anoxic conditions in the hypolimnion (e.g., Aksu et al., 1995; Williams et al., 1978). Sulfides (e.g., H2S), produced through anaerobic bacterial transformation of gypsum, reduced iron oxides and blackened the dolomitic sediments. Friedman (1966) supports this scenario, noting that a negative relationship exists between the presence of gypsum and dolomite. The abrupt freshening events also marked a return to deep water conditions and cyclic production of the calcite–evaporite couplets. In contrast to previous workers’ findings (e.g., Miller, 1981) that report an absence of any limnofaunal record in the basin, we found ostracode fossils (i.e., Limnocythere staplini) in the black mats. The presence of ostracoda in these layers lends further support to a freshening-event hypothesis. Four black-mat samples were collected at three sites in the northern section of the basin. Radiocarbon ages obtained from the organic fraction (total organic carbon, or TOC) in the four samples are listed in Table 1. The two older samples, Black Mats 1 and 2, and the youngest sample, Black Mat 4, were collected from shallow pits in the basin floor, while Black Mat 3 was collected from

the base of a badlands erosional remnant. The stratigraphic position of the three oldest samples are consistent with their ages. However, Black Mat 4 was collected only 5 cm above and from the same pit as Black Mat 1; the stratigraphic proximity of these two layers seems to indicate an erosional event (i.e., Black Mats 2 and 3 are missing at this locality), perhaps as a result of the lake drying up sometime between Black Mat 3 and Black Mat 4. Hydroeolian erosion and deflation of the subaerially exposed lake sediments, similar to the processes operating in the modern playa environment, could account for removal of the missing stratigraphic record at that location. The basin’s paleolake cycles (Fig. 2) are recorded by the different sedimentary environments reflected in the lacustrine stratigraphy. Periods of lacustrine lowstands marked by the hydrographic ‘‘troughs,’’ and accompanied by increased evaporation, diminished surface area, and shallow margins, are marked by the presence of the dolomite layers. Abrupt climate changes marking the return to mesic LGM conditions were accompanied by increased inflows of fresh water and rapidly rising lake levels. These rapid changes led to the temporary formation of anoxic hypolimnia that provided the environment for bacterial decomposition of gypsum and the darkening of the dolomite layers. Once the lake reached its maximum elevation for each transgression, indicated by the ‘‘peaks’’ in the hydrograph, it maintained a quasi-steadystate condition of annually fluctuating lake levels and chemistry. The lowest lake levels in these hydrographics are recorded by dolomite layers.

AID

qra

QR 1896

/

a609$$$162

04-21-97 15:41:22

Geomorphic Evidence: Lake King Tectonic motion in the Salt Basin graben has created a depression that is both narrow and steeply sided along recently active faults (i.e., along the Sierra Diablo) and broad and gently sloped in other areas (e.g., along the southeastern section of the basin; Fig. 3). The geomorphic implication is that the best-preserved Lake King shorelines are along the southwestern margin of the basin where the lake was in contact with steep alluvial fans which provided coarse clastic material for construction of beach ridges. In other sectors of the basin, along the eastern lacustrine margin and to the westnorthwest, the shorelines formed on low-gradient slopes. The

AP: QR

310

WILKINS AND CURREY

finer sediments comprising these beach ridges have been, in many of these areas, degraded by subsequent slope-wash processes. Shoreline elevations were surveyed at three locations (Fig. 3): at Cienega School on the northern end of Salt Basin; at Jim Hill, 3 miles south of Salt Flat, Texas, along the west central margin of Salt Basin; and on the Gilmore Ranch at the foot of the Guadalupe Mountains on the eastern margin of Salt Basin. The beach ridges at the Cienega School and Gilmore Ranch localities reflect the influence of low-gradient slopes and low-energy environments. There, the beach

ridges are subtle, with broad, low crests of fine, gypsiferous sand spaced 50 to 200 m apart. The Jim Hill locality is an example of shorelines formed on tectonically active parts of the graben. The steeper slopes resulted in higher-energy depositional environments and beach ridges constructed of gravel- to cobble-sized clasts. From elevations surveyed at these three locations, six shorelines have been identified. At Cienega School two shorelines were surveyed at elevations of 1114 and 1111 m. These shorelines were also found at Jim Hill as well as an additional shoreline at an elevation of 1108 m. Four shorelines at the Gilmore Ranch location have elevations measuring 1105, 1106, 1108, and 1109 m, but the two highest shorelines (1114 and 1111 m) at the other locations are not evident at Gilmore Ranch, perhaps having been removed by slope wash. The best preserved shorelines, at Jim Hill, are identified as K1 (1109 m), K2 (1111 m), and K3 (1114 m; Fig. 4). The relative ages of these three shorelines are based on degree of geomorphic preservation and position. K3, a wellpreserved clastic barrier beach, can be identified from Jim Hill to the southern section of the basin (a distance of approximately 40 km), and represents the maximum extent of Lake King. The 1111-m shoreline (K2) is thought to represent the maximum elevation of a lake from an earlier cycle. K2 has been washed out in several places at the Jim Hill location by subaerial postdepositional processes and also is partially covered by the 1114-m shoreline (K3). The 1109-m shoreline (K1) also has been breached in several places by postlacustrine subaerial processes (e.g., Sack, 1995). The preservation of remaining segments suggests that K1 was deposited subsequent to K2 and K3. Black mats in Lake King sediments provide the sole radiocarbon ages for lacustrine events. These sediments, representing what are thought to be abrupt transgressive events resulting in an anoxic hypolimnion, are limited to sediments precipitated under deep-water conditions—smaller transgressive events resulted in lower lake elevations, without the formation of anoxic bottom water conditions, and were not recorded by black mats. Based on the assumptions and within the limitations of this model, the black mats represent lake cycles with the highest stages. Based on geomorphic preservation and position at Jim Hill, the radiocarbon ages obtained from Black Mat 1b, Black Mat 2, and Black Mat 3 are tentatively assigned to the three highest shorelines. The oldest sample, dated at ca. 22,600 14C yr B.P. (Table 1), is paired with the 1111-m shoreline, K2. The intermediate age, ca. 19,100 14C yr B.P., is assigned to the highest shoreline, K3; at this highest stage of 1114 m, Lake King had a surface area of approximately 900 km2, a volume of 10.8 km2, an average depth of 12 m, and a maximum depth of 26 m. The third date, ca. 17,200 14 C yr B.P., is assigned to shoreline K1, which has a well-

AID

qra

FIG. 3. Map of Trans-Pecos Closed Basin, showing extent and surface elevations at paleolake maxima in each basin. Lake King at its maximum extent covered 902 km2 and had an average depth of 11 m. Lake Sacramento at its maximum extent covered 90 km2 and had an average depth of 16 m.

QR 1896

/

a609$$$162

04-21-97 15:41:22

AP: QR


311

FIG. 4. Lake King, Jim Hill shoreline locality. The three highest shorelines surveyed in Salt Basin are identified in plan and shown in profile views.

AID

QR 1896

/

a609$$1896

04-21-97 15:41:22

qra

AP: QR

312

WILKINS AND CURREY

preserved segment along the western side of the basin. The youngest of the dates, ca. 15,900 14C yr B.P., is assigned to the 1108-m shoreline found at the Gilmore Ranch locality— its geomorphic expression at the Jim Hill locality is visible on air photos, but was not surveyed. Geomorphic Evidence: Lake Sacramento

immediately north of the pass indicate a wide, broad-crested baymouth barrier at that elevation. The elevation of the upper spit, 1340 m, is the same as the elevation of shoreline B1 at Lagunita; the lowest spit elevation measures 1337 m. The spits at the pass are well preserved and probably represent later lake cycles, the 1337m spit represents the last lake cycle to leave a geomorphic record. At this time, no direct age control (e.g., 14C) exists for the Lake Sacramento shoreline chronology, although the timing of paleolake stands in the Sacramento basin is assumed to be penecontemporaneous with Lake King highstands. Stage B2 is tentatively correlated with Lake King stage K2 based on geomorphic position and preservation. Stage B3, together with the B4 and B5 oscillations, is correlated with Lake King stage K3. The lowest stage, B1, thought to represent a later lake cycle, possibly correlates with Lake King stage K1.

The Lake Sacramento basin is a subbasin nested 200 m above and about 90 km northwest of the center of Salt Basin. Overall, the subbasin is smaller and deeper (i.e., from the threshold to the floor) than Salt Basin, with steep sides along the northern and eastern margins. The topographic threshold is 43 m above the basin floor at Beargrass Pass, at an elevation of approximately 1350 m. The Sacramento basin is the terminus of the ephemeral Sacramento River but, like Salt Basin, no perennial flows currently reaches the basin floor. Unlike Salt Basin, however, absence of evaporite deposits in the Sacramento basin suggests a lack of hydrographic closure, now and in the past. Transects with shoreline elevations were surveyed at Lagunita Madre (Lagunita), which consists of a barrier beach complex, and at Beargrass Pass, which has two spits and a cuspate barrier. The highest paleolake stage, 1347 m, created a lake surface of 90 km2, a volume of 2.1 km3, a maximum depth of 40 m, and an average depth of 16 m. No overflow channel was observed in the field or on aerial photographs, suggesting that the basin is leaky enough to prevent overtopping of its threshold. Karst features (i.e., sinks) adjacent to the basin allude to subsurface drainage, and groundwater studies identify the area as a major recharge zone for the Bone Spring aquifer (e.g., Bjorklund, 1957). At Lagunita, five shorelines were identified at elevations of 1340 m (B1), 1342 m (B2), 1344 m (B3), 1346 m (B4), and 1347 m (B5) (Fig. 5). The 1347- and 1346-m shorelines form bayhead barriers, with the three lower shorelines forming a complete baymouth barrier approximately 2 km long with a uniform elevation of 1344 m along the crest. The 1342-m shoreline (B2) is transgressive to the 1344-m shoreline (B3), which buries the lower shoreline. The 1346-m (B4) and 1347-m (B5) shorelines represent a transgressive oscillation from the 1344-m stage. Based on geomorphic preservation, B4 appears to be more recent than B5, the latter probably representing a short-lived stage. The B4 stage did include a partial construction of a baymouth barrier on top of the 1344-m shoreline. At Beargrass Pass, elevations were measured at the attachment points of spits and on the crest of the barrier; the barrier forms the highest shoreline (1346.6 m) at Beargrass Pass, within 1/2 m of the elevation of the B5 shoreline in Lagunita. This shoreline at Beargrass Pass probably represents the B5 and, later, the B4 stages. There are no geomorphic features at Beargrass Pass representing the 1344-m stage, but vegetation and drainage patterns seen on aerial photos of the area

Rapid, or abrupt, changes in climate, called Dansgaard– Oeschger events, have been identified in stable-isotope records of the Greenland ice sheet (Bond et al., 1993; Dansgaard, 1985), as well as in lacustrine sediments in WNA (e.g., Allen and Anderson, 1993; Phillips et al., 1994). Researchers (ibid) have used these rapid climate shifts (usually to a more mesic climate) to explain changes in the geochemical stratigraphic records of those lake basins. Dansgaard– Oeschger events have a quasi-periodicity that averages 2550 yr (Dansgaard, 1985), closely correlating with the black mats in Lake King sediments as well as periodicity reported in other lacustrine sedimentary records in the region (e.g., Lake San Augustin; Phillips et al., 1992). Phillips et al. (1992) reconstruct the changes in lake levels of Lake San Augustin, concluding that the lake experienced four major transgressive cycles from ca. 22,600 to ca. 15,000 14C yr B.P. The timing of those transgressions closely approximates (i.e., within 100 yr) the timing of highstands and deposition of black mats in Lake King. Gaps in the lacustrine stratigraphic sections suggest that Lake King oscillated several times during the LGM. At the location of Black Mat 2, the sediment layer was only 2 to 4 cm thick at a depth of 0.75 m. The overlying sediments consisted of over 100 evaporite couplets in the lower 0.5 m, with a zone of more recent, displacive evaporite growth in the upper 0.25 m of the pit. Where Black Mat 1 was collected, however, the sampled layer was thicker (É10 cm), but at a shallower depth of 0.3 m; Black Mat 4 was collected from the same pit, only 5 cm above Black Mat 1. The stratigraphic proximity of these two layers indicates an erosional unconformity, probably resulting from subaerial exposure and erosion of older sediments. From the lack of a continuous sedimentary record between freshening events, we infer

AID

qra

QR 1896

/

a609$$$163

04-21-97 15:41:22

Palaeolimnologic Reconstruction and Implications

AP: QR


313

FIG. 5. Lake Sacramento, Lagunita Madre (Lagunita in Fig. 3) locality. Five shorelines, representing three lake cycles, are identified in plan view.

a quasi-pluvial environment for the basin—that is, Lake King probably experienced one or more terminal regressions (i.e., ceased to exist) between various lake maxima. The physical configuration of the Lake Sacramento basin (i.e., compact and deep) would have been more conducive to supporting a permanent lake than Salt Basin (i.e., broad and shallow). The persistence of Lake King in a marginal,

quasi-pluvial environment must have relied on contributions of groundwater discharged from the Bone Spring aquifer. Mesic conditions during the LGM resulted in higher (i.e., with respect to the terminal basin floor) piezometric surfaces and influent groundwater discharge into Salt Basin. This groundwater discharge was sustained, in large part, by the effluent discharge from Lake Sacramento, reducing lake area

AID

qra

QR 1896

/

a609$$$163

04-21-97 15:41:22

AP: QR

314

WILKINS AND CURREY

Aksu, A. E., Yasar, D., and Mudie, P. J. (1995). Paleoclimatic and paleooceanographic conditions leading to development of sapropel layer S1 in the Aegean Sea. Palaeogeography, Palaeoclimatology, Palaeoecology 116, 71–101. Allen, B. D. (1991). Effect of climatic change on Estancia Valley, New Mexico: Sedimentation and landscape evolution in a closed-drainage ba-

sin. In ‘‘Field Guide to Geologic Excursions in New Mexico and Adjacent Areas of Texas and Colorado (B. Julian and J. Zidek, Eds.). New Mexico Bureau of Mines and Mineral Resources Bulletin 137. Allen, B. D., and Anderson, R. Y. (1993). Evidence from western North America for rapid shifts in climate during the last Glacial Maximum. Science 260, 1920–1923. Benson, L. V., Currey, D. R., Rorn, R. I., Lajoie, K. R., Oviatt, C. G., Robinson, S. W., Smith, G. I., and Stine, S. (1990). Chronology of expansion and contraction of four Great Basin lake systems during the past 35,000 years. Palaeogeography, Palaeoclimatology, Palaeoecology 78, 241– 286. Bjorklund, L. J. (1957). ‘‘Reconnaissance of Ground-Water Conditions in the Crow Flats Area, Otero County, New Mexico.’’ New Mexico State Engineer Office, Technical Report 8. Bond, G., Broecker, W., Johnsen, S., McManus, J., Labeyrle, L., Jouzel, J., and Bonanl, G. (1993). Correlations between climate records from North Atlantic sediments and Greenland ice. Nature 365, 143–147. Boyd, F. M., and Kreitler, C. W. (1986). ‘‘Hydrogeology of a Gypsum Playa, Northern Salt Basin, Texas.’’ Texas Bureau of Economic Geology Report of Investigations 158. COHMAP members (1988). Climatic changes of the last 18,000 years: Observations and model simulations. Science 241, 1043–1052. Currey, D. R. (1991). ‘‘Hemiarid Lake Basins: Hydrographic and Geomorphic Patterns.’’ Univ. of Utah, Limneotectonics Laboratory Technical Report 91-2. Dansgaard, W. (1985). Greenland ice core studies. Palaeogeography, Palaeoclimatology, Palaeoecology 50, 185–187. Enzel, Y., Brown, W. J., Anderson, R. Y., McFadden, L. D., and Wells, S. G. (1992). Short-duration Holocene lakes in the Mojave River drainage basin, Southern California. Quaternary Research 38, 60–73. Fenneman, N. M. (1931). ‘‘Physiography of Western United States.’’ McGraw-Hill, New York. Friedman, G. M. (1966). Occurrence and origin of Quaternary Dolomite of Salt Flat, west Texas. Journal of Sedimentary Petrology 36, 263–267. Hardie, L. A., Smoot, J. P., and Eugster, H. P. (1978). Saline lakes and their deposits: A sedimentological approach. Special Publications of the International Association of Sedimentologists 2, 7–41. Hawley, J. W. (1993). ‘‘Geomorphic Setting and Late Quaternary History of Pluvial-Lake Basins in the Southern New Mexico Region.’’ New Mexico Bureau of Mines and Mineral Resources, Open File Report 391. Hussain, M., Rohr, David, M., and Warren, John K. (1988). ‘‘Depositional Environments and Facies in a Quaternary Continental Sabkha, West Texas.’’ West Texas Geological Society Publication 88-84, pp. 177–186. King, P. B. (1948). ‘‘Geology of the Southern Guadalupe Mountains, Texas.’’ U.S. Geological Survey Professional Paper 215. Kohler, M. A., Nordenson, T. J., and Baker, D. R. (1959). ‘‘Evaporation Maps for the United States.’’ Department of Commerce, Weather Bureau, Technical Paper 37. Langbein, W. B. (1949). ‘‘Annual Runoff in the United States.’’ U.S. Geological Survey Circular 52. Miller, R. R. (1981). Coevolution of deserts and pupfishes (genus Cyprinodon) in the American Southwest. In ‘‘Fishes in North American Deserts’’ (R. J. Naiman, Soltz, and David L., Ed.), pp. 39–94. Wiley, New York. Mock, C. J., and Bartlein, P. J. (1995). Spatial variability of late-Quaternary paleoclimates in the western United States. Quaternary Research 44, 425–433. Muehlberger, W. R., Belcher, R. C., and Goetz, L. K. (1978). Quaternary faulting in Trans-Pecos Texas. Geology 6, 337–340. Phillips, F. M., Campbell, A. R., Kruger, C., Johnson, P., Roberts, R., and Keyes, E. (1992). ‘‘A Reconstruction of the Water Balance in Western

AID

qra

variability for Lake King while increasing lake persistence between highstands. Lake Sacramento essentially acted as a reservoir for Lake King, collecting and storing surface water runoff from the nearby Sacramento Mountains; the water eventually leaked into the surrounding limestone and thence into the overlying fill of Salt Basin. Water transferred in this fashion would have been protected from evapotranspiration, effectively increasing the contribution of the Sacramento River to Lake King (e.g., Langbein, 1949). SUMMARY

Our primary objective was to explore evidence for hydrologic and climatic changes which occurred during the LGM in the Trans-Pecos Closed Basin of west Texas and southcentral New Mexico, and to determine the timing and extent of these changes as recorded by the responses of paleolakes in the basin. The study aggregates the spatial and temporal data from marginal lacustrine (i.e., geomorphic features) and limnetic (i.e., offshore stratigraphy) records of a nested-basin-fed terminal-basin paleolake system. Geomorphic evidence provides spatial control of the extent of three major highstands for Lake King. Stratigraphic evidence from Lake King provides temporal control on the timing of these maxima during the LGM, with transgressions at 22,570, 19,090, 17,180, and 15,940 14C yr B.P. The timing of these events correlates with highstands in the Estancia (Allen, 1991) and San Augustin (Phillips et al., 1992) basins to the north and northwest. The quasi-periodicity of these transgressive events may provide terrestrial evidence of Dansgaard–Oeschger events as identified in the Greenland ice cores. The lack of continuous stratigraphic records suggests at least one terminal regression between maxima, leading us to infer that the climate and hydrology of the basin during the LGM resulted in a quasi-pluvial lake environment. ACKNOWLEDGMENTS This research was partially funded by NSF Dissertation Research Improvement Grant SBR9405427. Additional support was provided by a University of Utah Graduate Research Fellowship and a Desert Research Institute Jonathan O. Davis Scholarship. We also thank Superintendant Larry Henderson and Fred Armstrong of Guadalupe Mountains National Park, the Texas Nature Conservancy, Jim and Jack Lynch of the CL Ranch, John Gilmore and the Gilmore Ranch, and Chris Stahl of the Wilson Ranch for their cooperation. The comments provided by C. C. Reeves and an anonymous reviewer were very helpful and greatly appreciated.

REFERENCES

QR 1896

/

a609$$$163

04-21-97 15:41:22

AP: QR


315

United States Lake Basins to Climatic Change.’’ New Mexico Water Resources Research Institute Report 269. Phillips, F. M., Campbell, A. R., Smith, G. I., and Bischoff, J. L. (1994). Interstadial climatic cycles: A link between western North America and Greenland? Geology 22, 1115–1118. Reilinger, R., Brown, L., and Powers, D. (1980). New evidence for tectonic uplift in the Diablo Plateau region, west Texas. Geophysical Research Letters 7, 181–184. Sack, D. (1995). The shoreline preservation index as a relative-age dating tool for late Pleistocene shorelines: An example from the Bonneville Basin, U.S.A. Earth Surface Processes and Landforms 20, 363–377. Tuan, Y. F., Everard, C. E., Widdison, J. G., and Bennett, I. (1973). ‘‘The Climate of New Mexico.’’ New Mexico State Planning Office. Van Devender, T. R., Betancourt, J., and Wimberly, M. (1984). Biogeographic implications of a packrat midden sequence from the Sacramento Mountains, south-central New Mexico. Quaternary Research 22, 344– 360.

Waters, M. R. (1989). Late Quaternary lacustrine history and paleoclimatic significance of pluvial Lake Cochise, southeastern Arizona. Quaternary Research 32, 1–11.

AID

qra

QR 1896

/

a609$$$164

04-21-97 15:41:22

Van Devender, T. R., Spaulding, W. G., and Phillips, A. M. I. (1979). Late Pleistocene plant communities in the Guadalupe Mountains, Culberson County, Texas. In ‘‘Biological Investigations in the Guadalupe Mountains National Park, Texas’’ (H. H. Genoways and R. J. Baker, Eds.). National Park Service, Proceedings and Transactions Series 4. Warren, J. K. (1989). ‘‘Evaporite Sedimentology: Importance in Hydrocarbon Accumulation.’’ Prentice Hall, Englewood Cliffs, NJ.

Wells, P. V. (1978). An equable glaciopluvial in the West: Pleniglacial evidence of increased precipitation on a gradient from the Great Basin to the Sonoran and Chihuahuan deserts. Quaternary Research 12, 311– 325. Williams, D. F., Thunell, R. C., and Kennett, J. P. (1978). Periodic freshwater flooding and stagnation of the eastern Mediterranean Sea during the late Quaternary. Science 201, 252–254.

AP: QR