Curtis Monger

The Jornada Long-Term Ecological Research (JRN LTER) program consists of studies superimposed on three research entities, the Jornada Experimental Range, the Chihuahuan Desert Rangeland Research Center, and the Desert Soil-Geomorphology project (figure 17.1). The JRN site is in the northern part of the Chihuahuan Desert and represents, for the LTER network, the desert shrubland and desert grassland ecosystems of the southwestern United States. Climate data at the Jornada site and surrounding area span the last 110 years. Ecological data span the last 144 years.

Despite having over 100 years of data, researchers at the Jornada LTER have struggled to answer the focal question of this book: How have ecosystems responded to climatic variability? This is because, simultaneous with climate, another important factor has had a major impact on ecosystems—human land use. Cattle grazing, brush control, and habitat fractionation have merged with climate to produce external pressures on Jornada ecosystems (Schlesinger et al. 1990; Havstad et al. 2000). Even more uncertain is the cause-and-effect relationship between climate and ecosystems in prehistoric times. Here evidence is limited to indicators, such as former lake shorelines, plant fossils in packrat middens, fossil pollen, 13C/12C ratios in paleosols, and erosion rates. When some indicators are used by themselves, circularity arises if a conclusion about ecosystem response to climate change is based on an inference about climate change, which is based, in turn, on ecosystem change. For example, grasslands increased at the end of the middle Holocene as the result of increased rainfall, where the interpretation of increased rainfall is based on increased grass pollen in the middle Holocene sediments (Freeman 1972).

Although focusing on millennial-scale climate and ecosystem variability, this

Millennial Scale Climate Variability

Figure 17.1 Location of the Jornada LTER site in relation the Chihuahuan Desert as defined by Schmidt (1979). Jornada LTER is superimposed on the Jornada Experimental Range, the Chihuahuan Desert Rangeland Research Center, and the Desert Soil-Geomorphology Project area. Pluvial lakes are Lake Otero (L-O), Lake Estancia (L-E), Lake King (L-K), Lake Palomas (L-P), Lake Animas (L-A), Lake Cochise (L-C), Lake Goodsite (L-G), Lake San Agustin (L-SA), and Lake Trinity (L-T). The location of the nearest alpine glacier was Sierra Blanca Mountain (SB). Packrat data are from San Andres Mountains (SAM), Sacramento Mountain (SM), Hueco Mountain (HM), and Bishsop's Cap (BC). Pollen data are from piedmonts of the San Andres Mountains (P1) and Organ Mountains (P2). Carbon isotope samples are from Fort Bliss eolian basin floor (FB) and from the Organ Mountains piedmont (P2).

Figure 17.1 Location of the Jornada LTER site in relation the Chihuahuan Desert as defined by Schmidt (1979). Jornada LTER is superimposed on the Jornada Experimental Range, the Chihuahuan Desert Rangeland Research Center, and the Desert Soil-Geomorphology Project area. Pluvial lakes are Lake Otero (L-O), Lake Estancia (L-E), Lake King (L-K), Lake Palomas (L-P), Lake Animas (L-A), Lake Cochise (L-C), Lake Goodsite (L-G), Lake San Agustin (L-SA), and Lake Trinity (L-T). The location of the nearest alpine glacier was Sierra Blanca Mountain (SB). Packrat data are from San Andres Mountains (SAM), Sacramento Mountain (SM), Hueco Mountain (HM), and Bishsop's Cap (BC). Pollen data are from piedmonts of the San Andres Mountains (P1) and Organ Mountains (P2). Carbon isotope samples are from Fort Bliss eolian basin floor (FB) and from the Organ Mountains piedmont (P2).

Table 17.1 Assigned ages of Quaternary temporal terms used in this article

Time perioda



a.d. 1850 to present



Present to 4,000


4,000 to 8,000


8,000 to 10,000



10,000 to 20,000


20,000 to 125,000


125,000 to 790,000


790,000 to 1,650,000

a These periods are based on time intervals described by Bull (1991) and Gile et al. (1981).

b Radiocarbon ages in this paper are not calibrated to calendar years.

a These periods are based on time intervals described by Bull (1991) and Gile et al. (1981).

b Radiocarbon ages in this paper are not calibrated to calendar years.

chapter briefly discusses historic variability for comparison and as a means for describing the setting. The historic-prehistoric boundary for the Jornada area has been set at a.d. 1850 (table 17.1).

Setting of the Jornada LTER Site

Located at 32.5° N and 106.8° W, in New Mexico, USA, the Jornada LTER site is in the Basin and Range province (Peterson 1981), which is characterized by parallel mountain ranges separated by structural basins filled with Cenozoic sediments (Hawley 1986). Elevations at the Jornada range from 1,180 m (3,870 ft) in the Rio Grande floodplain to 2,749 m (9,012 ft) in the Organ Mountains.

The modern terrain is mainly the product of the Rio Grande Rift tectonic system that has been active since the Oligocene (Mack et al. 1998). Bedrock units of Meso-zoic volcanic and Paleozoic sedimentary rock have been tilted, creating blockfaulted mountain ranges whose eroded sediments fill neighboring basins. In the Jornada basin, the depth of these sediments exceeds 2134 m (7,000 ft) (Seager et al. 1987). Beginning about 5 million years ago, sediments from adjacent mountains were supplemented with sediments from the ancestral Rio Grande, which spilled into several basins of the Jornada region, filled the basins with river deposits, and alternately spilled into adjoining basins as those basins became topograpically lower (Mack et al. 1997). The Rio Grande floodplain as it exists today was formed after the river downcut through its previously deposited sediments. This downcut-ting probably began sometime between 780,000 years ago (the age of the Matuyama-Brunhes paleomagnetic boundary) and 760,000 years ago (the age of the Bishop volcanic ash) (Mack et al. 1993, 1996, 1998).

Vegetation at Jornada consists of shrublands, desert grasslands, and, to a minor extent in higher elevations, juniper savanna (Allred 1988, 1993). Shrublands and grasslands occur in elevations below about 1,678 m (5,500 ft), which roughly corresponds to 330 mm (13 in.) annual rainfall (USDA-NRCS 1999; Dick-Peddie

1993). Juniper savannas, depending on aspect, occur above this elevation-rainfall boundary. Shrublands are commonly dominated by mesquite (Prosopis glandulosa) on sandy soils of the basin floor, by creosotebush (Larrea tridentata) on many soil types of piedmont slopes, and by tarbush (Flourensia cernua) on clayey soils of lower piedmont slopes. Fourwing saltbush (Atriplex canescens) and broom snakeweed (Gutierrezia sarothrae) are components of shrublands on a variety of soils and landforms (Gardner 1951). Grasslands are commonly dominated by black grama (Bouteloua eriopoda), with lesser amounts of mesa dropseed (Sporobolus flexuosus), and red threeawn (Aristida spp.) on sandy and loamy soils of basin floors and piedmont slopes, and by tobosa (Pleuraphis mutica) and burrograss (Scleropogon brevifolius) on clayey soils of lower piedmont slopes and basin-floor depressions. Savanna vegetation is composed of red-berry juniper (Juniperus ery-throcarpa) and various grasses, shrubs, and Mexican pinyon pine (Pinus cem-broides) (Allred 1993) on mountain slopes and valleys.

Historical Climate Variability at the Jornada LTER Site Measured Climate Variability (1892-1993)

The climatic record at the Jornada Experimental Range began July 1915 (Ares 1974) and is summarized in figure 17.2. A statistical treatise of historical climate at the Jornada LTER site by Wainwright is forthcoming in a synthesis volume on the Jornada LTER site to be published by the Oxford University Press. The climatic record at New Mexico State University (NMSU) in Las Cruces (figure 17.1), 45 km south of the Jornada weather station, began January 1892 (Malm 1994). Based on the record at NMSU, Malm (1994) compiled the following statistics. The mean annual temperature is 16°C (60.3°F). The mean annual rainfall is 222 mm (8.73 in.), with greater than 50% falling in the summer monsoonal season of July, August, and September. The annual pan evaporation is 2393 mm (94.2 in.), which is more than ten times the rainfall amount. Highest pan evaporation occurs in June, with an average of 341 mm.

Differences between rainfall at New Mexico State University and rainfall at the Jornada Experimental Range are statistically significant (Conley et al. 1992). Though little differences occurred for winter rainfall (53 mm at NMSU vs. 54 mm at Jornada), larger differences occurred for summer rainfall (135 mm at NMSU vs. 151 mm at Jornada). Temperature differences between New Mexico State University and the Jornada Experimental Range are also statistically significant, with NMSU showing a slight and linear warming trend since 1892 in contrast to the Jornada which shows a warming trend until about 1950 after which it shows a slight cooling trend (Conley et al. 1992).

The wettest year at NMSU received 498 mm (19.60 in.) in 1941, whereas the driest year received 87 mm (3.44 in.) in 1970. The period around 1900 was very wet, but was followed by a very dry period around 1910. The longest drought began in 1945 and continued through 1956. Temperature at NMSU ranged from a high of 43°C (109°F) on 8 July 1951 to a low of -23°C (-10°F) on 11 January 1962. On av-

Figure 17.2 Climate variables at the Jornada Experimental Range from 1915 to 1993 (USDA-ARS, 2000).

erage, 9 days per year reach 38°C (100°F), mostly in June and July. There have been 4 years between 1892 and 1991 when the temperature did not reach 38°C (100°F) (1927, 1938, 1941, 1988), and 3 years when more than 30 days were at or above 38°C (100°F) (1951, 32 days; 1978, 33 days; 1980, 32 days). Temperatures of -18°C (0°F) or lower are rare, with only eight occurrences between 1892 and 1991 (Malm 1994).

Other important climatic factors impacting the ecosystem are light intensity and wind speed. Light intensity has an annual average of 21.6 megajoules per square meter per day (MJ/m2/day) (516 langleys/day). Solar radiation is lowest in December with a monthly mean of 11.8 MJ/m2/day (282 langleys/day) and is highest in June with a monthly mean of 30.5 MJ/m2/day (729 langleys/day) (Malm 1994). Wind speeds are greatest in March, April, and May (figure 17.2), when occasional gusts can reach 145 km/hr (90 mi/hr).

Vegetation Change (1858-1963)

In 1857, the Jornada region was first surveyed by the United States General Land Office. The surveyors included in their notes observations about vegetation, soils, and topography. From these notes, the 1858 vegetation map of the Jornada Exper imental Range was constructed (Buffington and Herbel 1965). After establishment of the Jornada Experimental Range in 1912, subsequent vegetation maps were made in 1915, 1928, and 1963. Based on these maps, Buffington and Herbel (1965, p. 61) made the following conclusions about vegetation change:

In 1858 the Jornada Experimental Range was a great expanse of grass with only isolated spots of mesquite. On the higher areas along the mountains, [shrubs] brush was present; however, grass was also good in most places. A few tarbush plants were present in some of the lower lying areas. Since 1858 the grass cover has decreased tremendously, and the brush has increased to the point that it was present on the entire study area in 1963. Less than 25% of the study had a fair stand of grass in 1963.

Anecdotal Evidence of Vegetation Change (1598-1885)

Grazing of domestic livestock at the Jornada LTER site probably spans a longer time period than most places in the United States because the Jornada Basin is crossed by one of the oldest roads in the United States—El Camino Real or "Royal Road," which connected Mexico City with Santa Fe (Hallenbeck 1950). Don Juan de Onate's route from Zacatecas north to the land of the Pueblos in 1598 became the Camino Real. Onate's entourage in 1598 was approximately 4 miles long and consisted of 170 families, 129 soldiers, 80 wagons, and 7,000 head of livestock (Hallenbeck 1950).

After Onate's initial trip, caravans traveled the road about once every 3 years from Mexico bringing supplies to Santa Fe. In the 1700s and early 1800s, the road was increasingly used for trade between Chihuahua and Santa Fe, including livestock trade. When the Santa Fe Trail reached the northern terminus of the Camino Real in 1821, trading between the United States and Mexico increased. By the 1850s, however, traffic on the Camino Real diminished as shorter trade routes through Texas connected markets in the United States with Mexico. By 1882 the road ceased to be a major travel route when the railroad linked Chicago, El Paso, and Mexico City.

Onate and other Spanish explorers mentioned plants cultivated or used by Native Americans, but made few comments on vegetation in general (Gardner 1951). Yet by the mid-1800s, descriptions of vegetation were more common, as compiled by Fredrickson et al. (1998, 196-197):

Writing about his passage through the Jornada del Muerto in 1846, Dr. Frederick Adolph Wislizenus . . . stated, 'The wide country through which we have to travel, in the elevation of from four to five thousand feet above the sea, [has] dry, hard soil, tolerable grass, and an abundance of mezuite and palmillas.' Two years later, journalist John Cremony . . . described the same area, '. . . is a large desert, well supplied with grama grass in some portion, but absolutely destitute of water or shade for ninety-six miles.' Beale . . . described the Jornada Plain as, '. . . a level plain, covered thickly with the most luxurious grass, and filled with wildflowers . . . Hundreds and hundreds of thousands of acres containing the greatest abundance of the finest grass in the world, and the richest soil are here lying vacant, and looked upon by the traveler with dread because of its want for water.' . . . Albert Fountain [1885] wrote that the Jornada Plain was treeless and waterless but covered with rich, nutritious grass.

Anecdotal descriptions of vegetation by travelers and military expeditions in the 1800s are also contained in Gardner (1951). This evidence, combined with interviews of long-term residents in southern New Mexico, led Gardner to conclude that there is little room for doubt that grass cover has markedly decreased and shrubs greatly increased during the past hundred years.

Causal Factors

Two major hypotheses have been put forth as causal factors in the shift from dom-inantly grassland vegetation to dominantly shrubland vegetation—(1) overgrazing by domestic livestock and (2) climate variability. The climate-variability hypothesis was tested by analyzing climatic data using several statistical methods (Conley et al. 1992). Based on these analyses, Conley et al. rejected the climate-variability hypothesis and accepted the overgrazing hypothesis as the cause of the vegetation shift.

To Buffington and Herbel (1965), however, the rapid increase of shrubs was caused by both climate variability and overgrazing. This combination involved the effects of seed dispersal, periodic droughts, and selective grazing of grass by livestock. The equilibrium between creosote bush and black grama, for example, was shifted in favor of the shrub with selective grazing and because black grama is sensitive to soil erosion and burial by sediments. Even if heavy grazing is discontinued, the return of the grass may require a very long time as compared to more humid grasslands because of the low rainfall, the loss of soil, the scarcity of propagules, and the presence of shrubs (Gardner 1951).

The invasion of grasslands is not limited to shrub expansion; there has been concurrent expansion of woodlands into grasslands (Dick-Peddie 1993). Both invasions cause a concentration of soil resources beneath the woody plants (Wright 1982; Schlesinger et al. 1990; Davenport et al. 1996). This redistribution of soil resources becomes an important factor for continued vegetation change, as do changes in fire regime, changes in vertebrate and invertebrate distribution, changes in atmospheric CO2, changes in water redistribution, and changes in competition by nonnative species (Gibbens et al. 1983; Ludwig 1987; Pieper and Beck 1990; Mc-Auliffe 1994; Herbel et al. 1994; Huenneke and Noble 1996; Whitford 1996; Herrick et al. 1997; McPherson and Weltzin 2000).

Millennial-Scale Climate Variability and Ecosystem Response

Conclusions about millennial-scale climate and ecosystem variability at the Jornada LTER site are based on physical, biotic, and soil-geomorphic evidence. Physical evidence includes features made by pluvial lakes, alpine glaciers, rock glaciers, and groundwater. Biotic evidence includes packrat middens, fossil pollen, and carbon isotopes. Soil-geomorphic evidence includes geomorphic surfaces and soil profile characteristics.

The most direct evidence concerning prehistoric climate are marks produced by physical entities, such as lake shorelines produced by Pleistocene lakes (figure

A. Physical Evidence causal factor causal factor

i______________________i inference about i______________________i inference about

B. Biotic Evidence causal factors causal factors

i__ _ _ ______i inference about i__ _ _ ______i inference about

C. Soil-Geomorphic Evidence causal factors

C. Soil-Geomorphic Evidence causal factors

Figure 17.3 Sources of proxy evidence for climate inferences based on physical, biotic, and soil-geomorphic entities and the marks and components produced by those entities. Solid lines show causal factor links. Dashed lines show uniformatarian inferences made about climate based on marks and components.

Figure 17.3 Sources of proxy evidence for climate inferences based on physical, biotic, and soil-geomorphic entities and the marks and components produced by those entities. Solid lines show causal factor links. Dashed lines show uniformatarian inferences made about climate based on marks and components.

17.3A). Biotic fossils are also evidence of prehistoric climate because climate is a causal factor for the development of biological entities such as grassland. A climate-vegetation linkage, however, is confounded because fire, erosion, atmospheric CO2, seed dispersal by animals, successional stages, and nutrient limitations are also factors that affect vegetation (figure 17.3B). The least direct evidence of prehistoric climate is based on soil-geomorphic components because both climate and vegetation (as well as animals, topography, and parent material) are causal factors for soil and geomorphic entities (figure 17.3C). Nevertheless, a late-Pleistocene ge-omorphic surface on a terrace, for example, is a vestige of a landscape formed in the late Pleistocene with a higher base level (e.g., Hawley 1975). Moreover, dissolution pipes through a petrocalcic horizon are vestiges of soil formation in a wetter climate (Gile et al. 1981). As the complexity of these systems increases, uncertainty about cause-and-effect relationships also increases.

Physical Evidence

Lakes. Though dry now, many large depressions in the Chihuahuan Desert were filled with lakes at various times in the Quaternary. Lake Jornada (Gile 2002), for example, probably existed during much of the early and middle Pleistocene. More recent lakes, however, existed in the late Pleistocene and early Holocene, and they provide higher resolution evidence about climate change. Nine such pluvial lakes surrounding the Jornada LTER site (figure 17.1) have been reported in the literature: Lake Otero 40 km to the northeast (Hawley 1993), Lake Estancia 230 km to the northeast (Allen and Anderson 1993), Lake King 160 km to the southeast (Wilkins and Currey 1997), Lake Palomas 100 km to the south (Reeves 1969), Lake Animas 190 km to the southwest (Fleischhauer and Stone 1982), Lake Cochise 290 km to the southwest (Waters 1989), Lake Goodsite 35 km to the west (Hawley 1965), Lake San Agustin 200 km to the northwest (Markgraf et al. 1984), and Lake Trinity 110 km to the north (Neal et al. 1983).

Of these lakes, Lake Estancia, Lake King, Lake San Agustin, and Lake Cochise have the best chronologic information about water levels. High stands occurred at all four lakes during the latest Pleistocene. At Lake Estancia two major high stands occurred: one beginning at ca 19,700 14C years b.p. and another at ca 13,700 14C years b.p. (Allen and Anderson 1993). At Lake King, high stands occurred at ca 22,600, 19,100, and 17,200 14C years b.p. (Wilkins and Currey 1997). At Lake San Agustin, lacustrine ostracodes suggest a high stand from ca 22,000 to 19,000 years ago (Forester 1987). At Lake Cochise, a high stand occurred between 13,750 and 13,400 14C years b.p., after which the lake level dropped until about 9,000 14C years b.p. when it experienced a renewed rise (Waters 1989).

After the early-Holocene lake level rise at Lake Cochise (ca 9,000 years b.p.), the lake progressively dried. There is no evidence of lakes during the middle Holocene, although two small lakes probably formed shortly after the middle Holocene, but dates are uncertain (Waters 1989). At Lake San Agustin, ostracodes indicate that, at about 5,000 years ago, rainfall became so low the lake changed into its present dry playa (Forester 1987). Lake Estancia was dry in the middle Holo-cene based on deflation of lake sediments and dune formation during that period (Hawley 1993).

Alpine Glaciers. During the last full glacial period (marine-O-isotope 2, ca 25,000 to 10,000 b.p.) alpine glaciers formed on high mountain peaks of New Mexico (Richmond 1986). The nearest glaciated peak to the Jornada LTER site is 130 kilometers to the northeast at Sierra Blanca (elevation 3,660 m, 12,003 ft) (figure 17.1). This mountain is the southernmost late Pleistocene-age glaciated peak in the United States (Smith and Ray 1941). Glacial features found there include a cirque on the northeastern side of the mountain, a steeply sloping glaciated valley, well-defined moraines, and a protalus rampart within the cirque.

Rock Glaciers. Rock glaciers consist of poorly sorted angular boulders, fine material, and interstitial ice in permafrost areas of high mountains (Bates and Jackson 1987). The formation of interstitial ice is responsible for their downslope creep. In New Mexico, mountains located both northeast and northwest of the Jornada LTER site contain some 116 rock glaciers (Blagbrough 1994). They have been used as pa-leoclimatic indicators because, based on modern rock glaciers, they are most active a few hundred meters below the orographic snow line. Elevations of the New Mexico rock glaciers indicate that the mean annual temperature during the last full glacial period was approximately 7 to 8°C cooler than today (Blagbrough 1994).

This range is fairly close to the 5 to 7°C cooler-than-present range estimated by Phillips et al. (1986) and the 5.5°C cooler-than-present temperature estimated by Stute et al. (1995). Both were based on groundwater studies in northwestern New Mexico. This technique is based on the principle that solubility of noble gases in the atmosphere is a function of temperature. Thus, noble gases dissolved in the ground-water of a confined aquifer record the temperature at the water table (Stute et al. 1995).

Biotic Evidence

Packrat Middens. Packrat middens are clumps of plant materials gathered by Neotoma woodrats (Betancourt et al. 1990). The plant material is gathered by rats from a surrounding area of about 30 meters to construct dens and is cemented by dried urine. These nests, when in a protected shelter, such as limestone caves or ledges, can exist for tens of thousands of years and thereby provide a fossil record of changing plant communities (e.g., Spaulding 1991). Because the age of plant fossils can be determined with radiocarbon dating, a chronology can be assigned to bi-otic changes.

In the vicinity of the Jornada LTER site, packrat middens have been studied in the San Andres Mountains 60 km to the north, in the Sacramento Mountains 90 km to the east, in the Hueco Mountains 120 km to the southeast, and in the Franklin Mountains (Bishop's Cap) 60 km to the south (Van Devender 1990) (figure 17.1). One of the more complete records occurs at the Hueco Mountain site. Here, the fossil record suggests a four-step climatic-vegetative shift: (1) A pinyon-juniper-oak woodland grew in the area from 42,000 to 10,800 14C years b.p. (2) After this period, pinyon disappeared, but oaks and junipers persisted through the early Holocene until about 8000 14C years b.p. when oaks disappeared. (3) At this time, honey mesquite and prickly pear appeared in a gradual transition to desert grassland. (4) The shift to modern desert scrub conditions was completed before about 3650 14C years b.p., the time creosotebush appeared (Van Devender, 1990).

Although packrat middens provide firm evidence that certain plants were present at certain times, the limitations of using this method for broad paleoclimatic reconstruction have been reviewed by Hall (1997). First, woodrats gather vegetation only within a range of about 30 meters. Second, woodrats do not randomly collect plants from their 30-m home range, but select specific plants for food storage, nesting, and den construction. Third, rocky-escarpment vegetation may differ significantly from vegetation occupying broad piedmont slopes and basin floors. Fourth, field workers have been known to target middens containing exotic plants, therefore, plant species found in the modern flora could be bypassed.

Still, packrat middens provide important, well-dated information about biotic changes from which climatic inferences can be made. For the Chihuahuan Desert at large, four major climatic inferences have been made (Van Devender 1990). (1) The last full glacial period was a time of equable climates with cooler summers than today, greater rainfall, and mild winters with few freezes. (2) By early Holocene, winter rainfall continued to be greater than today, but summer temperatures increased beyond what they were previously. (3) By the middle Holocene, winter rainfall had been replaced by biseasonal rainfall that was dominated by the summer monsoon. Also, during the middle Holocene, summer temperatures dramatically increased over what they had been in the early Holocene and latest Pleistocene. (4) The late Holocene was a period when the modern climatic regime was established. This included fewer winter freezes, monsoonal summer rainfall, and increased droughts (Van Devender 1990, p. 126).

Fossil Pollen. The dominant shrubs at the Jornada LTER site—mesquite, cre-osotebush, and tarbush—are insect pollinated, and, therefore, do not release copious amounts of pollen into the wind. In contrast, plants that do release large amounts of pollen include the Gramineae and members of the genus Atriplex (Horowitz 1992). At the Jornada site, several grass types and Atriplex canescens (Fourwing saltbush) are common components of the modern flora. After pollen has fallen on the land surface, some of these silt-size particles illuviate into the upper soil horizons and some are transported laterally by runoff water into playas.

Although fossil pollen provides a picture of regional vegetation, it also has several limitations (Hall 1997), especially when the pollen record is from arid pale-osols. These limitations include imprecise chronologic control in which pollen ages can only be bracketed between strata having radiocarbon-dated charcoal. Sparse-ness of fossil pollen is also a limitation, especially for recently deposited eolian and alluvial sediments. Some plants, such as pines, produce proportionally more pollen than their abundance in the landscape, whereas insect pollinated plants produce proportionally less pollen than their abundance in the landscape (Horowitz 1992). Also, differential destruction of pollen grains in soils is another limitation (Moore et al. 1991). Poor preservation of pollen occurs in all arid-land depositional environments in which the sediments have been subject to weathering processes (Hall 1997). Pollen grains in arid soils are usually corroded, due either to oxidation or to bacterial or fungal activity. In addition to chemical decomposition, mechanical degradation occurs in cemented soil horizons (Horowitz 1992).

Pollen analysis is further handicapped at detecting change within grasslands by taxonomic limitations, in which case phytoliths have been useful (e.g., Fredlund and Tieszen 1997). Phytolith studies at the Jornada LTER are rare, but they offer opportunities for future research because phytoliths have been identified for the major grasses, yucca leaves, and conifer needles in the Jornada region (Pease 1967).

Given these limitations, comparative changes in pollen have, nevertheless, provided insight into biotic changes. Most of the fossil pollen studies in the vicinity of Jornada have been collected from buried paleosols on piedmont slopes. At the Jor nada LTER site, fossil pollen was studied on the piedmont slope of the San Andres Mountains 5 km east of the Jornada Experimental Range (Freeman 1972) and on the piedmont slope of the Organ Mountains 65 km to the southeast (Monger et al. 1998) (figure 17.1). For making inferences about bioclimatic changes, the main pollen categories of these studies were the Cheno-Am and the Gramineae taxo-nomic units. The Cheno-Am taxon includes all members of the Chenopodiaceae (except Sarcobatus, greasewood) and Amaranthus (in the Amaranthaceae plant family). A prominent member of the Cheno-Am taxon is fourwing saltbush (Atri-plex canescens). The Gramineae family includes all grasses.

Two major inferences about bioclimatic change were made from the piedmont slope site of San Andres Mountains: (1) Desert scrub vegetation and aridity was greatest in the middle Holocene. (2) The middle Holocene desert scrub and aridity gave way to increased grasslands and a more mesic climate by the late Holocene (Freeman 1972). Three bioclimatic inferences were made from the piedmont slope site of the Organ Mountains: (1) Grasslands were prominent in the early Holocene and latest Pleistocene; (2) by middle Holocene time, there was increased desert scrub vegetation; and (3) as with the San Andres site, desert scrub and aridity gave way to increased grasslands and the more mesic climate of the late Holocene.

Bioclimatic inferences based on a synthesis of fossil pollen data in the Southwest as a whole were made by Hall (1997, p. 36): "By 14,000 years ago, as the climate warmed, pinyon dropped out of the low-elevation terrain and were replaced by a drier Chenopodiaceae-Asteraceae shrub grassland. Throughout the Holocene, low-elevation desert regions were dominated by desert-shrub grassland vegetation. The mid-Holocene was characterized by a decrease in grasses and an increase in shrubs due to hot, dry climate during that period."

Ratios of 13C/12C in Soil Organic Matter and Pedogenic Carbonate. At the Jornada LTER site, shrub species use the C3 photosynthetic pathway—with the exception of fourwing saltbush, which uses the C4 pathway (Syvertsen et al. 1976). Grass species use the C4 photosynthetic pathway. Cacti, which are a minor component of both shrublands and grasslands at the Jornada site, use the CAM photosyn-thetic pathway.

Atmospheric CO2 has 13C values that range from -7.8 to -12 %o (Boutton 1991). Because C4 plants incorporate more atmospheric 13CO2 into their biomass than C3 plants, C4 plants have higher 513C values. The range for C4 plants is from -7 to -15 %o, in contrast to C3 plants that range from -20 to -35 %o (Cerling and Quade 1993). Of the plants measured at the Jornada LTER site, mesquite, creosotebush, tarbush, and soaptree yucca fall into the C3 range, whereas tobosa grass, black grama grass, prickly pear cactus, and fourwing saltbush fall into the C4 range (figure 17.4).

Because soil organic matter maintains the same S13C signature as the vegetation from which it was derived, S13C values of soil organic matter have been used to make inferences about vegetation growing on landscapes of the past (e.g., Kelly et al. 1993). At the Jornada LTER site, S13C values of soil organic matter reflect the invasion of desert shrubs into grasslands during the middle Holocene (Cole and Monger 1994; Monger et al. 1998) and the last 150 years (Connin et al. 1997a,b).

In addition to soil organic matter, soil carbonate (CaCO3) carries a S13C signa-

Figure 17.4 Ranges of carbon isotope (S13C) values at the Jornada LTER site. Typical S13C ranges are from Boutton (1991) and Cerling and Quade (1993). Ranges for plant types are for species gathered at the Jornada LTER site (Gallegos 1999).

ture of vegetation (Cerling 1984; Wang et al. 1996). Inferences made about vegetation from soil carbonate involve the presumption that S13C values of carbonate are enriched 14 to 16 %o with respect to the vegetation controlling soil CO2 (Cerling et al. 1989; Amundson et al. 1998). In theory, if carbonate precipitates in iso-topic equilibrium with CO2 respired by a pure C4 grassland, carbonate should have a S13C value of +2 %o. In contrast, if carbonate precipitates in isotopic equilibrium with CO2 respired by a pure C3 shrubland, carbonate should have a S13C value of -12 %o (Boutton 1991).

At the Jornada LTER site, a record of S13C values in pedogenic carbonate is contained in buried paleosols and overlying modern soils. On piedmont slopes, buried paleosols are common in alluvial deposits. On basin floors, buried paleosols are common in eolian deposits. The Organ allostratigraphic unit, a major deposit of middle Holocene age, occurs on both the piedmont slope and basin floor landforms (table 17.2). This unit began to be laid down around 7,500 14C years b.p. based on dated charcoal (Gile 1975). Because it disconformably buries older soils (pale-osols), carbon isotopes in paleosols are considered to be a component of the vegetation growing on those paleosols until they were buried by Organ sediments.

Table 17.2 Geomorphic surfaces at the Jornada LTER site and surrounding area that occur on areas astride the Rio Grande floodplain (i.e., valley border), the piedmont slope, and the basin floor landformsa

Geomorphic surface

Carbonate stage

Piedmont slope

Basin floor

Es timated soil age

Valley border





(years b.p. or epoch)

Coppice dunes

Coppice dunes


Historical (since a.d. 1850)

Lake Tank

Present to 150,000




0, I


Middle and late Holocene (100 to 7,000)




(100(?) to 1,000)




1,100 to 2,100




2,200 to 7,000


Isaacks' Ranch



Latest Pleistocene (10,000-15,000)

Late Picacho

Late Jornada II



Latest to late Pleistocene (15,000-75,000)


Jornada II

Petts Tank



Late to middle Pleistocene (75,000-150,000)




Late to middle Pleistocene (150,000-250,000)

Jornada I

Jornada I

Jornada I



Middle Pleistocene (250,000-400,000)

Dona Ana


Middle Pleistocene (>400,000)

Lower La Mesa


Middle to early Pleistocene (780,000)

JER La Mesa


Early Pleistocene (780,000-1,600,000)

Upper La Mes


Late Pliocene (2,000,000-2,500,000)

a After Gile 2002.

a After Gile 2002.

On the piedmont slope, a major isotopic shift was observed across the contact between paleosols and the overlying Organ unit (figure 17.5). This shift ranged from -2 %0 in the paleosols to -8 %0 in the younger Organ unit. The S13C values accorded well with an increase of Cheno-Am pollen, and erosion that indicated a change from a C4 grassland to a C3 shrubland in the middle Holocene (Cole and Monger 1994; Monger et al. 1998).

On the basin floor, a similar isotopic shift was observed in Organ eolian sediments (figure 17.5). This shift ranged from -1 %0 in the paleosols to -7 %0 in the Organ unit and was also interpreted as indicating a change from a C4 grassland to a C3 shrubland (Buck and Monger 1999). Unlike sites on the piedmont slope, however, the upper strata of two of the eolian sites (figure 17.5D and 17.5G) suggest a gradual return of grasses in that landform. In four of the profiles (figures 17.5A, B, D, E), the lower and older strata may indicate increased C3, possibly a juniper savanna at about the last full glacial period.

Oxygen isotopes (18O/16O) in soil carbonates have been used to make inferences about paleotemperatures (Cerling 1984; Cerling and Quade 1993) and rainwater sources (Amundson et al. 1996; Liu et al. 1996). However, analysis of S18O values in the vicinity of the Jornada LTER site shows no consistent trend. In some cases, the S18O values change little despite major shifts in carbon isotopes (figures 17.5A, B, C). In other cases, the S18O values have trends similar to those of carbon isotopes (figures 17.5D, E, H).

0 0

Post a comment