Ground Sloth Dung And Packrat Middens

The greatest impediment to scientific innovation is usually a conceptual lock, not a factual lack.

Stephen Jay Gould, Wonderful Life:

The Burgess Shale and the Nature of History

In 1956, after two years on fellowships, it was time to look for a permanent teaching and research job. My top priority was finding a position that would give me an opportunity to recover a fossil pollen record next to radiocarbon-dated remains of extinct animals of the ice age. That combination of data should shed light on the causes of the extinctions by showing whether or not climatic changes correlated with them.

Yale conservation ecologist Paul Sears and his students had recently discovered important climatic changes in the pollen record from a long core recovered from lake sediments beneath Mexico City. In addition, in a long core taken in the San Agustin Plains of New Mexico, Sears and Katherine Clisby had found pollen evidence of a glacial-age spruce forest well below the elevations now occupied by spruce in the southern Rockies. Unlike pollen records inside the glacial ice margin of New England and the Great Lakes region, these cores yielded continuous records well beyond the latitudes covered by Quaternary glaciers.

From these and other findings I concluded that the southwestern United States was a promising place to look for fossil pollen deposits disclosing plant and animal range changes forced by climatic shifts in the ice age. The floodplains of Cochise County, Arizona, for example, offered abundant alluvium (sediment laid down by moving water) that invited coring and analysis. Some arroyos had yielded bones of mammoths and associated artifacts. In addition, the buried muds of the dry playas (basins) left by Quaternary lakes were rich in fossil pollen coinciding in time with the Quaternary megafauna. The 1,000-foot core Clisby and Sears took from the San Agustin Plains provided one of the first tests in the arid West of the wrongheaded interpretation that lower latitudes escaped the climatic influence of the last ice age. They discovered appreciable percentages of glacial-age spruce pollen in their core, collected in an environment dominated by juniper and pinyon, with ponderosa pine growing only on the most favorable soils. No longer was there any doubt that when glaciers advanced and retreated, plant and animal communities, spruce included, did as well. Radiocarbon dates demonstrated that all these changes were synchronized.

The arid Southwest also offered another crucial advantage, from my perspective: excellent preservation of datable fossils, especially in caves. Except in frozen ground, fossils buried in open sites are leeched of their organic content and are rarely datable by isotopic measurement. On the other hand, investigators have been aided by the remarkable preservation of bones in dry caves in arid regions, as well as in unglaciated boreal or subarctic regions with permafrost. In certain caves in the Grand Canyon, extinct bird and mammal bones, retaining organic residues such as collagen and in some cases with dried tissues still attached, were to prove ideal for radiocarbon dating. Also ideal are keratinous tissues, such as beaks, claws, hooves and horn sheaths, which decay rapidly in temperate and tropical environments but not in dry caves or rock shelters in arid regions. Skin, hair, cartilage, and even dung balls of long-extinct species can be remarkably well preserved. Recently such deposits have proved rich in well-preserved and identifiable DNA of the animals once inhabiting the caves.

All in all, the Southwest seemed like the very best place for me. Unfortunately, I could find no university positions open in Arizona or New Mexico, where I wanted to live. Then a friendly Arizona zoology student pointed out that universities might give research posts to faculty who came with their own grants, such as those from the newly established National Science Foundation (NSF). This proved to be excellent advice. The University of Arizona's preeminent anthropologist, Professor Emil Haury, director of the Arizona State Museum, put me in touch with Terah (Ted) Smiley, director of the university's new geo-chronology program. Although we had yet to meet, Ted and I collabo rated by mail on a grant proposal involving pollen analysis of archaeological sites with extinct animal remains. I suspect that Ed Deevey and/or Paul Sears reviewed our NSF proposal. In any case, we won an award that paid my salary and that of two research assistants for two years. In the fall of 1957 my wife Marian and I and our three small boys, Andy, Neil, and Tom, left our families on the East Coast and headed for Tucson.

I had not heard of the university's historic and world-famous Desert Laboratory on Tumamoc Hill. Built in 1903 by the Carnegie Institution of Washington, the laboratory was initially dedicated to desert botanical research (McGinnies 1981; Bowers 1988), such as studies on whether saguaros were disappearing from the vicinity of Tucson (they were not). Pioneer desert botanists, such as Forrest Shreve, who helped to found the Ecological Society of America, spent much of their careers here (Bowers 1988). In the 1930s, the Carnegie Institution of Washington abandoned several of its earlier programs in favor of more advanced "experimental" sciences. Under Ted Smiley's leadership, however, the Geochronology Laboratories carried on the tradition of research on arid regions, incorporating fossil pollen analysis, vertebrate paleontology, geomorphology, organic and inorganic geochemistry, and paleomagnetism. On Tumamoc Hill, Ted found space for offices, a conference room with library, a vertebrate fossil preparation lab, a hood and bench space for pollen extraction, and the magnetometer of a geophysicist.

The university soon purchased the Desert Laboratory and its grounds plus leases, a total of 869 acres of the Sonoran Desert. The grounds harbor 350 species of vascular plants, 300 of them native (Bowers and Turner 1985), as well as numerous mammals, birds, and reptiles.* (Some of this area would make excellent camel habitat.) From the lab's roost on a bench at 2,400 feet, one can look out over Tucson into three of Arizona's "sky island" mountains, each close to or exceeding 9,000 feet in elevation and each supporting rich vegetation gradients. Thanks to seminal work by the Carnegie botanists, the lab features a rich legacy of long-term studies. The pleasure of working in such a place would be—well, I hoped that University of Arizona administrators would not subtract the considerable value added by a desk and lab space on Tumamoc Hill from my paycheck.

Not wasting time on formalities, on my first day on "the Hill" Ted

*For a state-of-the-art treatment of shrubs of the Sonoran Desert and the geomor-phology near Tucson, see McAuliff 1999.

waved me to a chair and handed over a large brown paper bag labeled "Rampart Cave." Inside I found many smaller bags of samples collected by a graduate student in anthropology, Dick Shutler Jr. Setting aside his pipe, Ted reached into one of the bags, handed me a fibrous brown object the size of a baseball, and asked if I had seen anything like it before.

Taken aback, I received the segmented bolus of tightly packed dry plant remains somewhat gingerly. Although it was too large and misshapen to be a "road apple," the dung ball of a horse or a mule, that was the closest I could come. Ted's eyes twinkled. He explained that it was ground sloth dung or coprolite (fossilized dung). (In polite company in those days we did not say "shit.") From a cave in the lower end of the Grand Canyon, Dick had collected the dung samples at 6-inch intervals from top to bottom of a 5-foot stratified section. Years later we could see where he had removed samples, one above the next, from the trench through the ground sloth dung. Shasta ground sloth (Nothrotheriops shastensis) bones were associated with it. (A virtually complete skeleton of a Shasta ground sloth, with a dung ball in the paunch area, had been discovered in the 1920s at the bottom of a fumarole near Aden Crater, New Mexico; this confirmed the identity of the Rampart Cave dung.) Although compact and hard, the specimen was not mineralized or calcified. I ventured a cautious sniff. Though it was almost odorless, it looked fresh, fresh enough to have come not from an extinct animal many thousands of years old, but from one that might be still alive!

Ted could not have had better bait to capture my interest. By radiocarbon assay on the dung we could determine when the ground sloths lived in Rampart Cave and, along with results obtained elsewhere, estimate the time of their extinction. Because the samples were stratified, they would differ in age, older at the bottom, younger on top. We could determine the rate of deposition.

In addition, I was certain the dung would contain fossil pollen. Finnish geologist Martti Salmi had successfully extracted pollen from a single dung ball of a South American ground sloth (Salmi 1955). Digestion does not destroy the morphologically distinctive walls of pollen grains or of fungal spores, which resist hydrofluoric acid, capable of dissolving silts, clays, and even quartz. With any luck we could do the same and recover the first profile from a stratified dung deposit of an extinct late-Quaternary herbivore. Through the fossil pollen and other plant remains, we could detect changes in vegetation, climate, and the diet of the ground sloths. Although stable isotopes may be of some help, the diets of extinct animals are impossible to determine in any detail from fossil bones. Unmineralized dung deposits are ideal but so unusual that most paleontologists never see them. This is one of the factors that made Rampart Cave such a uniquely valuable site.

Finally, we could use trace element geochemistry to look for essential mineral nutrients in the dung. Salmi had proposed that ground sloth extinction was driven by scarcity of essential trace elements, such as copper and cobalt. We could test that.

Ted said I could share his office space in the library, and I proceeded to move in. To varying degrees academics lead nomadic lives. For seven years my family and I had been on the move. I did not know it on that mellow, mild November day over 45 years ago, but we had come home. I soon began to teach classes at the Desert Lab, which was just far enough from the main campus to discourage the less-than-dedicated students. In general I found motivation a far better predictor of career and personal success than grades on a transcript. Those willing to travel some 20 minutes by auto (longer by bike or bus) from the main campus to take a course or do research were sure to be worthy.

Dick Shutler let me join his Rampart Cave ground sloth project, and our geochemist colleague Paul Damon suggested that we add his student Bruno Sabels to the team. Dick soon left to receive state-of-the-art training at a laboratory affiliated with Columbia University, one of the best radiocarbon laboratories in operation at the time. He took splits of his sloth dung samples with him to have them dated. As discussed further in chapter 4, his radiocarbon dates indicated that the youngest dung was roughly 10,000 years old.

Meanwhile, I extracted and counted pollen from the samples. The pollen profile reflected the diet of the Shasta ground sloth. There was an abundance of globe mallow, Sphaeralcea sp., which is in the hollyhock family. Globe mallows are not wind pollinated, so their pollen would not have blown into the cave; the sloth must have eaten flowering plants. In favorable years Sphaeralcea flowers heavily in spring, presumably the time the ground sloths occupied the cave.

That spring I noticed a native globe mallow, with its bright orange-red flowers, growing around the Desert Laboratory. Emulating a ground sloth, I chewed some of the tops. The leaves were rather bland and slightly slimy, tasting like a young hollyhock, not bitter or resinous like many desert shrubs, such as creosote bush. With a hand lens I could detect indigestible star-shaped hairs known as trichomes on the foliage of the mallows. An abundance of the same type of hairs as well as pollen in the sloth dung supported the conclusion that the animals had been eating globe mallows.

Not all interpretations are as straightforward. In some extractions from the sloth dung, juniper exceeded 90 percent of the total pollen. But unlike the case of the mallows, this did not necessarily mean that the ground sloth ate juniper. Unlike globe mallow, juniper is wind pollinated. In a wet year juniper trees are loaded with ripe anthers discharging clouds of pollen, dusting all plants in the vicinity. Under those circumstances, any animals eating plants growing anywhere near juniper would unavoidably ingest an abundance of its pollen.

The presence of so much juniper pollen indicated the presence of junipers themselves and a change in climate over the last 30,000 years (Martin, Sabels, and Shutler 1961). From Carnegie botanical publications (Laudermilk and Munz 1934, 1938), verified by our own observations in later years, our field team established that junipers do not grow near Rampart Cave now, but the pollen record certainly suggested that they had in the past. (The same is true regarding Gypsum Cave, Nevada, where we later studied another dung deposit, as discussed in chapter 4.) Furthermore, given the season when juniper trees release their pollen, I concluded that the Shasta ground sloths visited Rampart Cave during winter or early spring.

Using mass spectrometry, Bruno Sabels found that the ground sloths at Rampart Cave had ample copper and other trace minerals in their diet. Salmi's suggestion that scarcity of trace elements caused the ground sloth extinctions had seemed questionable. According to Simpson's checklist, in the late Quaternary a dozen genera of ground sloths vanished from North and South America, along with or followed by dwarf ground sloths from the West Indies (see McKenna and Bell 1997 and White and Mac-Phee 2001 for updates; see Lyons, Smith, and Brown 2004 for a species list). It seemed unlikely that the entire New World had somehow run out of essential trace elements, even if southern Chile had done so. The soils in many parts of the Americas were not lacking in essential minerals. Copper, for example, is so abundant that Mexico, Arizona, New Mexico, Utah, and Montana, as well as the Andes in South America, host giant open-pit copper mines. Sabels's geochemical results proved that whether or nor Salmi's explanation for ground sloth extinction might apply in southern Chile, in Arizona it lacked support.

Not all of Dick Shutler's samples were of dung. The contents of one of his collections included fecal pellets of packrats and plant material brought into the cave by packrats. This might have revealed a new source of data for assessing climate and range changes, if we had only recognized its significance. In the early 1960s, ecologists Phil Wells and Clive

Jorgensen would report a surprisingly rich and previously unknown or underappreciated source of fossils: the fossilized middens of packrats (Neotoma), also known as wood rats (Wells and Jorgensen 1964). We had failed to recognize how abundant packrat middens can be in caves and rock shelters in most of the arid West.

Twenty-one species of packrats range from western Canada south to Nicaragua, with seven species in the Southwest of the United States (Vaughn 1990). True to their name, packrats throughout the ages have collected items of interest to them, often, but not always, items they could use for nest construction. Active packrat nests in open sites, often under prickly pear (Opuntia) or other cacti, are bushel-basket-sized piles of sticks, leaves, cactus pads, stones, and stored plant food accessible by hidden runways. Such middens are familiar to many hikers and naturalists, especially in arid America. In open sites, inactive middens do not last very long. However, middens in caves or rock shelters may harbor abandoned deposits that, saturated with packrat urine, can harden into resinous lumps as hard and durable as adobe (see plate 1). These fossil middens can last for thousands to tens of thousands of years, as long as they remain dry and are sheltered from precipitation, runoff, and incursions of termites (see various chapters in Betancourt, Van Devender, and Martin 1990). Similar deposits left by small mammals are reported in arid regions of South Africa, Australia, South America, the Near East, and Mongolia.

Each midden is a veritable time capsule, containing a sampling of whatever materials were available to packrats at the time the midden was active. Middens in caves are sometimes as small as a brick, sometimes a bit larger than a cement block, and contain leaves, stems, flowers, dry fruits, seeds, and pollen, along with an abundance of sticks, cactus pads or aureoles, packrat fecal pellets, and occasional bones or droppings of other animals, both living and extinct. Because plant fossils in middens are not mineralized, detailed anatomical analyses of leaves and other tissues can be made and remnant DNA identified.

The ancient rat middens proved to be a paleoecologist's bonanza. The contents of a single midden represent a very brief interval in near time. With accelerators one can directly date the remains of any species of woody plant found in a midden. Multiple radiocarbon dates can test the assumption of a midden's chronological integrity. Middens may be stratified, incorporating levels of different ages as determined by radiocarbon dating.

Rich in seeds and other identifiable macrofossils of many arid land trees and shrubs, fossil middens reveal the species composition of pre-

Plate 1. The author at a glacial-age packrat midden with extralocal juniper twigs, Montezuma Head, Organ Pipe Cactus National Monument, January 1976. Photo by Hal Coss.

historic plant communities, including some once inhabited by the extinct fauna of the late Quaternary. Added to fossil pollen records from desert playas, springs, spring mounds, and mountain lakes, the midden records permit the construction of maps showing the near-time ranges of many woody plant species and even some herbs (Betancourt, Van Devender, and Martin 1990; Brown and Lomolino 1998; Thompson 1990). Best of all, the middens were a new source of glacial-age plant fossils in dry climates, where stratified lake deposits were less abundant and often truncated. Though the combination of middens and fossil pollen data is especially informative, middens proved to be far more than a check on other approaches. They opened doors to new discoveries of biotic and climatic changes in near time.

The environments known to the extinct megafauna and to prehistoric people of the arid West could now be sampled in far more detail than had been possible previously. Stratified middens or those of different ages show changes in local ecology over time. They therefore provided a powerful new method of characterizing near-time environmental change and—when animal fossils were available—its relationship to faunal change, including late-Quaternary extinction. It was magical. We found ourselves time traveling.

Paleoecologists were by no means the first to appreciate fossil middens. A century and a half ago, about 60 miles northwest of Las Vegas, miners traveling west entered a canyon south of a dry playa known as Papoose Lake. They were out of rations and searching for food. Then, as one of them wrote: "Part way up we came to a high cliff and in its face were niches or cavities . . . in some of them, we found balls of a glistening substance looking like pieces of variegated candy stuck together. ... It was evidently food of some sort, and we found it sweet but sickish, and those who were hungry . . . making a good meal of it, were a little troubled with nausea afterwards" (Manly 1987). A little nausea seems a small price to pay for "a good meal" of what looked like candy, "glistening" with ancient rat urine. Intrepid as they are, I do not know of any recent investigators who claim to have emulated the hungry miners. By coincidence, Wells and Jorgensen made their discovery within roughly 30 miles of Papoose Lake, in a cave on Aysees Peak, a desert mountain in southern Nevada near the Atomic Energy Commission's nuclear test site.

Our own near miss in recognizing the significance of fossil middens was much less dramatic. Dick Shutler's sample from 36 inches at Rampart Cave contained aureoles of hedgehog cactus (Echinocereus) and a juniper (Juniperus) twig, both part of a packrat midden. Since juniper did not grow near the site, the stratigraphic location of the twig supported fossil pollen evidence of a major change in vegetation, and presumably climate, during the last continental glaciation. However, we did not register that the twig was in a fossil midden, so we missed its larger implications. I wonder how many other ecologists, archaeologists, and paleontologists did the same in their field work.

It turned out that second only to the sloth dung itself, most of the organic fill of Rampart Cave represented middens or scattered plant material, both deposits of prehistoric packrats. Packrats living in the cave would have collected all of their food and nesting material within roughly 50 to 100 feet of its mouth. The rats left sizeable middens beneath or above rock ledges; within the 36-inch level, apparently during a lengthy absence of the ground sloths; and on the surface of the dung deposit, evidently after the extinction of the ground sloth. A significant fraction of the middens contained juniper, in some cases mixed with fecal pellets and horn sheaths of extinct mountain goats and bones of these and other extinct mammals (Phillips 1984).

In the fossil middens from Aysees Peak, Wells and Jorgensen found abundant juniper twigs and leaves (needles). Juniper often occurs in desert mountains, but despite a thorough search, Wells and Jorgensen could find none growing on Aysees Peak. They sent the juniper remains to Libby's radiocarbon dating laboratory at UCLA and reported, "The result, 9320 ± 300 B.P., supported expectations. Ten other middens dated between [40,000 and 7,800 BP] indicated that juniper's lower limit in the region was then at least 600 m [2,000 feet] below where it is now" (Wells and Jorgensen 1964, 1173). They had nailed down the first of what would become an avalanche of new fossil discoveries. Midden analysis soon revealed glacial-age distributional shifts in a great variety of plants, including trees, shrubs, succulents, grasses, and herbs. Thanks to packrats we could determine, at least in general, the nature of the environment inhabited by the extinct animals of near time.

Full appreciation of these findings requires an understanding of vegetation gradients, formerly known as "life zones." A life zone is a geographic area in which a distinctive group of plants and animals is typically found. Of these, the trees and shrubs are generally the easiest to spot and to identify in the fossil middens. Each life zone typically occurs in a particular latitudinal range, where the appropriate climate prevails. The zones vary with altitude. In time the field ecologist will appreciate that what one sees is a continuous gradient rather than discrete zones. In the Northern Hemisphere, a higher elevation farther south may be as cool as, and thus support much the same vegetation as, a lower elevation farther north. In fact, it is often much easier to see the different zones by going 3,000 feet up or down a mountain than by driving several hundred miles north or south at the same elevation.

Differences in disturbance, substrate (bedrock and soil type), and aspect (solar exposure) introduce important complications in what actually grows where. Together with climate, elevation, and topography, we encounter a fistful of variables. (No wonder so many ecologists turn to modeling—it is the only way they can hope to keep up with nature's complexity.) These effects can be dramatic. For example, in the Northern Hemisphere, slopes facing north are relatively shady and thus cooler and less dry than south-facing slopes, which are fully exposed to the sun. Accordingly, a given plant will grow at different elevations on the different exposures. On the north-northeast side of a desert mountain, the lower limit of most plants is roughly 2,000 feet below their lower limit on the south-southwest side. That is, moving at a constant level from the south-southwest to the north-northeast side is ecologically equivalent to going up 2,000 feet in elevation.

The life zones form a magnificent gradient in which each species has its own unique habitat or niche. Different species may commonly be associated, but no two have identical distributions. For example, Engelmann spruce and alpine fir are often but not always associated in a spruce-fir community. Descending in elevation, these overlap with or give way to white fir, aspen, Douglas fir, limber pine, and ponderosa pine. Below, the ponderosas give way to pinyon-juniper woodland, typically with more pinyon at the upper elevations, while lower still, juniper may be accompanied not by pinyon but by single-leaf ash (Fraxinus anomala).

Superb examples of elevational gradients can be found in the Grand Canyon, which rises from 1,500 to 3,000 feet at the bottom to 9,000 feet on the Kaibab Plateau and 12,500 feet on the once-glaciated San Francisco Peaks outside Flagstaff. From Phantom Ranch on the Colorado River to the San Francisco Peaks, an elevational range of over 10,000 feet compressed into a distance of just under 80 miles, ecologist C. Hart Merriam in the late 1800s recognized the biological equivalent of a journey from the Mexican lowlands to Hudson's Bay in central Canada (Houk 1996).

By driving from Lees Ferry at 3,000 feet to the top of the Kaibab Plateau at 9,000 feet, one can see all but one of the life zones Merriam identified here. * At the bottom is the Lower Sonoran Zone, a desertscrub of blackbrush, shadscale, brickel-bush, and Mormon tea (Ephedra), with occasional sagebrush and other Great Basin shrubs and some Mojave Desert cacti. Going west, one climbs into open juniper woodland, then into denser pinyon-juniper woodland (Merriam's Upper Sonoran Zone). Jacob Lake, at 8,000 feet, lies in the Transition Zone, represented here by a magnificent ponderosa pine forest with patches of aspen and Gambel oak. Proceeding southward to the North Rim and the highest and coldest parts of the Kaibab Plateau, one enters the Canadian Zone, where ponderosa pine yields to Douglas fir mixed with aspen. Finally, a forest of spruce and fir grows above 8,500 feet, in the Hudsonian Zone. The uppermost zone in Merriam's system, the Arctic-Alpine, is restricted to elevations above 11,500 feet, which in Arizona occur only at the top of the San Francisco Peaks. Here hikers find endemic ground-hugging perennial herbs in meadows above the tree line.

Nearby, the overall gradient remains quite similar, but the specific associations vary somewhat. For example, within the Grand Canyon, hikers from the South Rim to Phantom Ranch descend through pinyon, juniper,

^According to the U.S. Board of Geographic Names, "Lees" has no apostrophe (letter from R. C. Euler, May 2001).

and banana yucca communities, but just below the rim, shady north-facing notches shelter occasional Douglas firs. Below, on the treeless Tonto Platform, they find a wide bench distinctively darkened by blackbrush— but if one knows where to look, I am told, occasional ponderosa pines can be found ensconced in shady north-facing breaks. On the slopes above the Colorado River and below the Great Unconformity, an incomprehensible time gap of 1.2 billion years, grow low shrubs of the sunflower family, such as aster, desert broom, and brickel-bush. Finally, riparian mesquites and catclaw acacia (Acacia greggii) grow next to the river, though catclaw may grow in moist sands up to 1,000 feet above it.

Similar gradients occur throughout Arizona and the West as plants respond to regional changes in climate and moisture resulting from differences in elevation, latitude, or longitude. The existence of such gradients over time is one thing that makes it possible to assess climate change. In cooler periods accompanying glaciation, Northern Hemisphere plants move southward in latitude, lower in elevation, or both to maintain a suitable environment. Similarly, in warmer periods, plants move northward or higher. Thus, for example, fossil evidence that a particular species formerly grew at a lower elevation than it does now suggests that the climate has warmed since that time. (Obviously, plants cannot literally pick up their roots and move; it is a question of where their dispersed seeds flourish.) Fossil pollen had been providing evidence of such climatic shifts; packrat middens were to provide still more, with less uncertainty about what plant species were involved.

Once Wells and Jorgensen had opened the door, many investigators in the Southwest began focusing on middens. Our team was no exception. With a modest amount of funding from the National Geographic Society, the NSF, and other governmental agencies, an interdisciplinary "Corps of Discovery" dedicated to finding, analyzing, and interpreting these middens took shape at the Desert Lab.

Fossil middens are not always easy to find or identify, even if one knows how to look for them. Talented midden prospectors, however, have turned up at some unexpected times. One discovery particularly memorable to me occurred in early June 1969, on an Arizona Academy of Science research trip down the Colorado River through the Grand Canyon.

On our fourth night we camped at River Mile 108.5, just above Shin-umo Rapids. The trip's organizer, historian Marty Link, had scheduled "campfire seminars," and that evening it was my turn. I held forth on secrets of the past, including the treasures to be found in ancient packrat middens. How wonderful it would be, I added, to find a series of mid dens along the biotic gradient from the bottom to the top of the Grand Canyon. They might reveal what happened to plants—and thus potentially to animals—at different elevations as the climate changed during and after the last ice age. I invited my listeners to search for middens, noting that junipers, for example, no longer grew along this reach of the river, but that just possibly, there might be middens nearby that harbored juniper twigs from the last cold interval at the end of the Pleistocene.

Before sunup the next morning, before I could stir from my sleeping bag, I found three teenagers in my face. Since first light they had been crawling over boulders and shelving rocks, looking beneath ledges and into crannies for old middens. Now they had rushed back for breakfast and a chance to display their find. They thrust fist-sized chunks of some hardened lumps of plant material under my nose. The chunks smelled resinous and slightly fetid, the distinctive odor of old packrat middens. I thought I saw the fecal pellets of packrats. "See those twigs?" one of the boys demanded. "Don't they look just like juniper?" I fished out my hand lens, peered closely at a few dusty brown stems, and had to agree that they did.

After breakfast a small group of us followed the eager young prospectors of Pleistocene secrets to their claim. Not far from camp and within a few feet of a pin marking an archaeological site, the boys showed us dry plant material wedged in the rocks beneath a small overhang. On some twigs we could see the tiny but diagnostic awl-shaped needles of juniper. No junipers could be spotted growing in the vicinity, certainly none within packrat foraging range. We were looking at a late-Pleistocene displacement. It had taken the self-appointed field team less than an hour to come up with pay dirt. First-time prospectors are rarely so successful. This was my introduction to Jim (James I.) Mead, a high school student from Tucson. Son of Albert Mead, former head of the Zoology Department at the University of Arizona, Jim had grown up roaming the mountains of southern Arizona with his buddies, searching for land snails, his father's specialty. Jim is now a professor of geology at Northern Arizona University in Flagstaff.

The Shinumo discovery (Van Devender and Mead 1976) portended a flood of new records of plant displacements in the Grand Canyon at all elevations, some from collecting sites so difficult to access that the field team, Ken Cole and Geoff Spaulding, had their drinking water flown in by helicopter.

Not all who tried their hand at midden prospecting decided to join the club. On another outing, one very promising student who searched for juniper in a large rock shelter surrounded by teddy bear cholla (whose fiercely barbed joints are especially favored by packrats for protecting their nests) emerged terribly covered with spines. Although he had found a few scraps of juniper, previously unknown in the Tucson Mountains, he reverted to his first love, the excavation of Pliocene gomphothere bones in southeastern Arizona, safely removed from the unforgiving cholla spines in certain rat middens.

This was an exciting time. Investigators including Julio Betancourt, Ken Cole, Pat Fall, Jim King, Cynthia Lindquist, Jim Mead, Art Phillips, Geoff Spaulding, Bob Thompson, Tom Van Devender, Phil Wells, and Jeff Zauderer turned up rich midden records from various parts of Arizona, New Mexico, and Texas, or in the case of Fall and Lindquist, in hyrax middens from Jordan. Radiocarbon dating of the middens enabled us to generate Quaternary vegetation maps (Betancourt, Van Devender, and Martin 1990) superior to those based mainly on fossil pollen records and biogeography (Martin and Mehringer 1965). As Wells and Jorgensen had anticipated, however, the middens strengthened the findings of those using fossil pollen for paleoclimatic investigations. At the University of Arizona, professors Vera Markgraf and Owen K. Davis combined both approaches.

To my knowledge, Pete Mehringer (who, like Jim Mead, had spent many hours of his youth exploring the desert) and Wes Ferguson, a faculty member at the Tree Ring Laboratory of the University of Arizona, preceded others in making a fundamental find. At an elevation of 6,000 to 7,000 feet in the Clark Mountains of California, near Las Vegas, they found, in what is now pinyon-juniper woodland, fossil middens that contained the preserved remains of bristlecone pine, limber pine, and white fir (Abies concolor) (Mehringer and Ferguson 1969). In Nevada, neither bristlecone nor limber pine now grows as far south or as low in elevation as the Clark Mountains. Radiocarbon dates verified that both occupied the mountains in the cooler climates of the late Quaternary; the white fir, too, had descended in elevation in glacial times. The fossil pollen record of spruce in cores from the Willcox Playa had given Pete and me indications of a similar descent, but the middens furnished the first macrofossils, evidence more secure than fossil pollen, which is vulnerable to long-distance transport, that in the last glacial age not only juniper but also other montane trees in the arid West grew at lower elevations than they occupy at present.

Like fossil pollen from sediment cores, fossil packrat middens eventually showed that during the last glacial episode, most woody plants throughout the West descended by at least 2,000 to 3,000 feet, significantly below the elevations that they occupy now. In the process they might migrate south by hundreds of miles. Any fossil deposit that was over 8,000 radiocarbon years old could be expected to contain displaced species. In addition to juniper and bristlecone pine, examples included spruce in West Texas (Van Devender and others 1977) and in the Grand Canyon (Cole 1990; Coats 1997); and fir, Douglas fir, limber pine, and sagebrush (various chapters in Betancourt, Van Devender, and Martin 1990). Fossil pollen from sinkhole lakes revealed that alpine sedges and grasses had replaced boreal trees, such as spruce and fir, on the Kaibab Plateau (Weng and Jackson 1999). In southern Arizona's Organ Pipe Cactus National Monument, whose low mountains now support a Sonoran thornscrub community of organ pipe cactus, saguaro, and foothills paloverde (Cercidium microphyllum), Tom Van Devender found middens over 8,000 years old yielding Mojave desertscrub species: sagebrush, Ajo Mountain scrub oak (Quercus ajoensis), Joshua tree (Yucca brevifolia), and on Montezuma Head at 1,100 feet, one-needle pinyon (Pinus mono-phylla) and juniper (Van Devender 1990). This is a community much more like that found in Joshua Tree National Park, which is to the northwest, in California, and higher in elevation.

Investigators have interpreted these range changes in Arizona and adjacent states as the result of cooler climate and/or increased precipitation. Later, in the warming postglacial climate, spruce forest replaced arctic alpine tundra above 8,500 feet, while at lower elevations pinyon-juniper replaced spruce-mixed conifer woodland and Mojave desertscrub replaced pinyon-juniper-single-leaf ash woodland.

Just to make the record more interesting, various plant species did not simply descend in lockstep in glacial times. The ranges of some plants expanded or shrank beyond what one might have expected, or switched direction. For example, shadscale, a common Great Basin shrub of arid climate, presently grows near Lees Ferry on the eastern edge of Grand Canyon National Park, and, sporadically, in the western end of the park near Rampart Cave. Fossil middens reveal that in the late-glacial climate, shadscale expanded its range in many directions, even upward (Spauld-ing 1990), seemingly against the tide of species coming down. No other desert shrubs are known to do this. A change in soil pH as well as climate may have been involved (Martin 1999).

Interestingly, Pete and Wes found no indications that ponderosa pine (Pinus ponderosa), the dominant species of Merriam's Transition Zone, had occupied the Clark Mountains with the other montane conifers. Ponderosa is widespread today, ranging from the Sierra Madre Occidental of northern Mexico through the Southwest and the lower elevations of the Rocky Mountains into western Canada. In the last late or full glacial, it might well be expected to have occupied lower elevations in many parts of the West. Instead, the middens suggest that it was absent from much of its modern range. Fossils of ponderosa pines more than 10,000 years old have yet to turn up in the midden record in the Grand Canyon (Cole 1990) or in glacial-age middens outside of Arizona. In the United States, ice age ponderosa appears to be limited to southern Arizona and New Mexico. (I am well aware that in making a sweeping statement about what has not been found, I am tempting the fates to deliver contrary examples. So be it.) Julio Betancourt, Tom Van Devender, and I are keenly aware of the "conceptual lock" rather than "factual lack" against our historical revision, which finds cultural, not climatic, history of critical importance in this case (Betancourt, Van Devender, and Martin 1990, 2).

Ponderosa continues to expand in the region, as shown by repeat photos by Ray Turner of the U.S. Geological Survey and others of originals taken over the last 100 years. The postglacial spread of ponderosa is so extraordinary that it challenges us to consider forcing functions beyond climatic change. That is, did something happen to favor ponderosa after the last ice age, something in addition to the climatic warming that led most forest and woodland species to ascend in elevation? A major change in fire history, including season and intensity of firing, is one possibility. Ponderosas are fire-adapted. And with the arrival of people in the New World around the end of the last ice age, a change in wildfire frequency could be expected. Julio Betancourt (personal communication, December 2001) suggests that ponderosa pine benefited by fires set by Native Americans, artificial ignitions of relatively light intensity, set well in advance of the normal summer lightning strikes and ignitions. By removing excess fuel in advance of the season when firestorms are likely to develop, cool fires could have favored the ponderosas.

The fossil record of Colorado pinyon (Pinus edulis), too, is unusual. Colorado pinyon is presently widespread in the Grand Canyon and Colorado Plateau at elevations just below those of ponderosa pine. In the late Quaternary, it did not simply descend a few thousand feet like spruce, Douglas fir, and Utah juniper. Instead, it almost vanished from Grand Canyon National Park (Cole 1990). Then, along with ponderosa pine, it expanded in range in the postglacial, perhaps as a result of human ignitions. In Chaco Canyon National Historical Park, the record indicated prehistoric human as well as climatic impacts on pinyon distribution.

The midden harvest was rich indeed. A number of species of plants appeared in the fossil record for the first time, greatly increasing our knowledge of their temporal distribution. And the dynamic changes implied by the past distribution of desert trees and shrubs laid to rest any thought that the arid Southwest might have escaped the climatic changes of the late Quaternary. A century ago, many botanists believed that the southwestern deserts were too low in latitude and elevation to have undergone significant climatic change in step with higher-latitude glaciations. However, both fossil pollen and macrofossils of the common plants that attracted foraging packrats showed that the dramatic shifts in eastern plant communities uncovered by Ed Deevey's pollen lab and many others since the 1950s had their equivalents in the West. The information needed to understand global climate change in the arid West in radiocarbon time comes to us courtesy of long-dead packrats.

As the midden research proceeded over the years, I kept one eye open for any unprecedented change in climate or environment that might help to explain the megafaunal extinctions. The extinct animals certainly lived during a time of dramatic vegetation change in the last cold stage, beginning 23,000 years ago and ending 8,000 to 10,000 years ago. However, nothing that I could detect in the new paleoecological and geo-chemical data from either the West or the Andes, under study by Julio Betancourt, Jay Quade, and their students, suggested a unique climatic crisis that would account for a unique extinction episode.

In contrast, midden analysis did support an important argument in favor of the overkill theory. As the list of tree and shrub species known from middens accumulated, it became increasingly apparent that plants, unlike large terrestrial animals, had not experienced a wave of extinctions in the late Quaternary. The near-time fossil record of plants in the arid West is similar to that of beetles (Coope 1995). Those records show that both beetles and vascular plants are sensitive to climatic change, but that neither suffered appreciable near-time extinction. Presumably, then, neither group was vulnerable to whatever wiped out the large mammals. This is in accord with the concept that early hunters, not climate change, caused the extinctions of megafauna.

Recently Jackson and Weng (1999) challenged the view that all trees escaped extinction in near time. They reported an extinct species of spruce (Picea critchfieldi), characterized by a seed cone of unusual size. Its late-Quaternary fossils are found in the southeastern states. Jackson and Weng have attributed its extinction to late-glacial climatic change, and those who believe that the same climatic impact accounts for megafaunal extinction (Grayson 2001) quickly picked up on this exception to the claim that plants suffered no extinction in the late Quaternary.

The simplest response to this argument is that the extinction of a single plant species presumably (but not clearly) coeval with large animal extinction is hardly enough to sweep away the overkill model. There is no reason to assume that climate changes generally tolerable to mammals would also have been tolerable to every species of plant in existence at the time. During the rapid climatic fluctuations of the late glacial and early Holocene, a temperate spruce species of limited range may have lagged behind its habitat, that is, dispersed more slowly than the climate changed.

On the other hand, might Critchfield's spruce have been eliminated by human activity? It should not surprise us if an occasional plant species were drawn into the overkill vortex. For example, when subjected to climatic stress, including drought, Critchfield's spruce would have been especially vulnerable to out-of-season anthropogenic ignitions, that is, the fires of spring. If so, one hypothesis would be that this temperate conifer fell victim to fire drives by the Clovis hunters—the same drives that may have favored fire-tolerant trees, such as ponderosa pine, as noted earlier. The conifer Torreya, a relict tree on the banks of the Apalachicola River in northern Florida and southern Georgia, may have had a history similar to that of Critchfield's spruce and barely escaped extinction (which may yet be its fate). When trees vulnerable to or favored by fire show striking changes in the fossil record around the time that people arrive in new lands, the possibility of an anthropogenic agency initiating, or at least furthering, the changes should be entertained. With the arrival of humans and the extinction of the megafauna, anthropogenic fire replaced herbivory by large mammals as the probable consumer of most above-ground savanna and grassland biomass. Fires can trigger the growth of human food plants in such environments and may have been set for this purpose.

As the 1960s drew to a close, we were equipped with several crucial new tools for exploring the late Quaternary and near time. Using fossil pollen analysis, packrat midden analysis, and radiocarbon dating, field teams from the Desert Laboratory proceeded to sample rock shelters and caves in the Grand Canyon and adjacent canyon lands. This provided rich opportunities to see what some of the large-animal extinctions looked like up close.

Was this article helpful?

0 0

Post a comment