It’s a sparkling 60-degree day in the Sonoran Desert west of Tucson, Arizona. From a rock ledge in the Waterman Mountains, visibility is better than 50 miles, and distant peaks frame the flat sweep of the Avra Valley. The azure sky is cloudless; a cool breeze recalls the freezing temperatures at dawn.
But Tom Van Devender didn’t climb up here for the view. At the moment he is visible only from the waist down, his other half having squeezed into a narrow fracture in the limestone wall. A hand appears, fumbling for the flashlight at his belt; a muffled voice speaks from inside the fissure.
Better check for blacktails when you come up.
Rattlesnakes, that is.
And watch your head.
Good advice--on entering this particular cave, an overhasty spelunker might well rap her noggin against a shoulder-high shelf of hardened sewage that coats most of the rock face inside. The lithic dung overlay looks to be at least six inches thick, and its dusty peat surface is broken by deep gouges that expose a dark, almost crystalline interior. It’s this mini-Mount Rushmore of manure that Van Devender, a naturalist with the Arizona-Sonora Desert Museum, has come to inspect--a natural wonder dating back at least to the last ice age, donated and unwittingly preserved by generations of the furry rodents known as pack rats.
Yes, pack rats. The same animals who’ve attained mythic stature as hoarders of multifarious junk are now gaining renown among scientists for their less-than-fastidious toilet habits. As the Waterman display amply demonstrates, pack rats do their business in their dens, and in arid regions the excretions can accumulate into rock-hard masses, called middens, that persist for thousands of years.
But middens are more than just superannuated scat. Whatever happens to be lying about the den also gets mixed in: grasses, twigs, leaves, spines, seeds, feathers, rocks, bugs, bones. The concoction is protected by the pack rat’s urine, which acts as a bactericide, so the organic matter doesn’t decompose. The urine also protects the offal from the ravages of termites and beetles--it carries insect-repelling chemicals produced by plants that were once eaten by the pack rats. Each new occupant of the den contributes to the stratigraphy, and the stinky structure outlasts them all. Hence pack rat middens provide a complex and continuous fossil record in parts of the world where fossils are scarce and continuous records unheard of.
That’s why naturalists like Van Devender venture into the bowels of mountains to hack away at cement compost heaps. He and a handful of fellow enthusiasts have analyzed more than 2,500 middens from caves and crevices in Idaho, Utah, Nevada, Colorado, Arizona, Texas, New Mexico, California, and northern Mexico, including the Baja Peninsula. The rich floral samples in middens have already revamped the history of desert vegetation in the Southwest. And midden analysis has helped revive debates about Native American settlements, tested theories of global warming, and shed light on the evolution of species.
Back in his office, Van Devender hands over a 14,000-year-old piece of the stuff. The midden looks like a misshapen chocolate bar chock- full of peanuts, caramel, and coconut. This particular specimen seems to be dark chocolate, but Van Devender says the color varies depending on the midden’s age and the circumstances of aging, such as the degree of weathering it has undergone. Younger middens look more like milk chocolate. There’s even a white-chocolate variety, dusted with chalk.
This came from a place in western Arizona, says Van Devender. It took a hammer and chisel to get it out. He points to the coconut. These are pine needles and juniper twigs. The caramel is matrix--mostly urine and dirt. Those things there--the peanuts--are all turds.
Fecal pellets are a primary constituent of middens, of course, but Van Devender’s interest is in the equally common food fragments that were dropped into the mix uneaten by the pack rats. A brick-size midden can yield as many as 100 different plants and animals: up to 60 shrubs, cacti, and grasses, 25 insect groups, and 15 different vertebrates. There’s just a lot of information packed into one of these things, he says. Not all the middens are that rich, but they’re all interesting because there’s no other fossil record.
That’s practically, but not technically, correct. There are other sources of fossils in arid regions: sloth dung, for example, has been found in the same dry caves that harbor pack rat dens, and it serves the same hermetic function; the bones of large animals such as camels and horses sometimes turn up in these caves as well. At low altitudes, where lakes formed during wetter eras, ancient pollen survives in the now dry lake beds.
But none of those sources rival the specificity, continuity, diversity, and sheer quantity of pack rat middens. Compared with pack rats, sloths were rare even in their heyday, and they weren’t big collectors. Pollen analysis is relatively crude: it might tell you what genus of plant you’re dealing with but not which species. And while it’s nice to know that camels once roamed the American Southwest, those beasts aren’t particularly sensitive indicators of the contemporaneous ecology. What of the snakes, lizards, and mice; the beetles, flies, and mites; the grass, trees, and cacti? Until the early 1960s, when midden analysis was born, there were very few clues to the desert’s past.
Pack rat fever began in 1961, when a young plant ecologist named Phil Wells and a young mammalogist named Clive Jorgensen set out on a hike up Aysees Peak in Nevada’s Mojave Desert. The two were doing a survey of local plants and animals for the Department of Energy, which at the time was considering the area for a nuclear-weapons test site. On this particular day they were looking for evidence of juniper trees, which live at similar elevations in other regions.
The hikers weren’t having any luck, and daylight was waning; on their way down, the two stopped to rest and kvetch, when over his shoulder Jorgensen spied an extrusion of nasty, pitch brown midden beneath a rocky overhang about 30 yards away. (Fortunately, Jorgensen knew what he was looking at; less fortunate were miners of the gold rush, who a century earlier had mistaken middens for some kind of exotic natural snack food-- they admitted feeling a little troubled with nausea after their repast.) Jorgensen went over to the midden, broke off a chunk, and lo!--inside found juniper seeds and twigs.
Wells and Jorgensen toted a bundle of middens down the mountain and sent them to a laboratory at UCLA where a new technique called radiocarbon dating was being refined (the technique is based on the fairly predictable decay rate of carbon 14, a radioactive isotope that is absorbed by plants and passed on to animals). When the age of their samples came back at 9,320 years (plus or minus 300 years), the two knew they had found a rich new source of fossils. In the following months they collected more middens and in 1964 announced that the samples ranged in age from 7,800 to 40,000 years--the upper limit of the dating technique. Today scientists suspect that some middens might be even older, but their ages can’t be verified. Not to worry, though; 40,000 years of fossils can keep naturalists busy for decades to come.
Midden analysis has completely revised the paleoecological record of the American Southwest, says Van Devender. The image of the desert as a timeless, changeless community given to exotic life-forms and environmental extremes had to be tossed out. The new understanding--still unappreciated by many ecologists--reveals deserts as dynamic entities whose ecosystems are in constant flux, continuously exchanging plant and animal species with surrounding environments. Just 11,000 years ago, for example, the landscape near the Waterman Mountains was a woodland, home to pine and oak. The saguaro cactus that today seems to epitomize the Sonoran Desert has probably come into and gone out of Arizona from points south at least 15 times in the last 2 million years, like a yo-yo, says Van Devender.
Ironically, the pack rat itself is an import, and one that has been a poor adapter physiologically to desert life. That helps explain why its habits are unique among desert rodents and uniquely valuable to scientists. Pack rats, which vaguely resemble big gerbils--gray fur, black eyes, big ears--live all over North America, in nearly every climate zone. Compared with most desert denizens, they need a fair amount of water and don’t tolerate heat well. Yet while even die-hard desert rodents like pocket mice and kangaroo rats burrow below ground to escape the heat and dry air, pack rats don’t.
Instead they build dens against trees or cacti or in caves and rock shelters; any portable material within 150 feet of the den can wind up in the construction. Pack rats will also fortify their dens with menacing pieces of spiny cholla cactus to fend off coyotes and other malefactors. The finished product can stand a redoubtable five feet tall and measure the same or more in diameter.
All pack rats build dens, says Yar Petryszyn, a mammalogist at the University of Arizona. But in the Southwest the structures tend to be particularly elaborate--not only to keep predators from getting in but also to modify the animal’s habitat. The thicker the wall of the den is, the cooler the interior is, and the more moisture it traps. The inside of a pack rat den is typically 20 degrees cooler than the desert air, and the relative humidity is twice as high.
All of which makes the pack rat’s predilection for hoarding seem less like neurosis and more like common sense. But why do pack rats relieve themselves indoors? Why does an animal with such finely honed standards of domesticity make its outhouse in-house?
Ever go camping? asks Petryszyn. You don’t want to wander very far to take a whiz in the middle of the night. If you’re a pack rat, every excursion you undertake in the service of your digestive tract could land you in somebody else’s. The real question is why all rodents don’t follow the pack rat’s excretory example; the answer probably has to do with the space constraints and poor circulation of underground housing. Besides which, notes Petryszyn, other desert rodents just don’t pee as much as pack rats. Their kidneys have evolved to conserve water, not to waste it making middens.
Fortunately, pack rats do waste their water making middens. And when one pack rat passes on, another moves in, leaving regular deposits where the previous occupant left off. You can have generation after generation of pack rats using the same prime spot, says Petryszyn. If the den is on a flat, the middens are soon degraded by rainfall--even a light sprinkling will undo preservation. But the prime spots for pack rat habitation are in protected areas--dry crevices and caves--where middens remain intact from posterior to posterity.
Posterity for middens usually arrives winded and perspiring, wielding rock hammers and Ziploc bags. It takes minutes to collect the midden samples and months to sort through them. Back in the lab the samples are soaked in buckets of water until the crystallized urine dissolves. The water is poured through screens that trap the loose debris, which is then dried in an oven. A few pack rat pellets are sent out for radiocarbon dating. The rest of the debris--the thousands of plant and animal remains liberated from each midden--has to be identified and enumerated. The remains are mostly impossibly small, identically beige bits of fodder that must be sorted under a microscope with a pair of tweezers. It’s the taxonomic equivalent of Name That Tune--I can name that species with one gram!--and there are probably fewer than a dozen people in the United States who can play it.
Botanists aren’t trained to look at fragments this small, says Kate Rylander, a paleobotanist with the U.S. Geological Survey in Tucson. So it’s up to me to learn to recognize the characteristics that distinguish one species from another. Until you get used to it, it can drive you crazy.
Rylander is used to it. Her boss is Julio Betancourt, a physical scientist and paleobotanist whose midden investigations are among the most creative. With his first foray into midden analysis in the late 1970s, Betancourt presented a theory of the downfall of Chaco Canyon, a major Native American settlement in northern New Mexico that was built and then inexplicably abandoned by the Anasazi 800 years ago. Middens from the site showed that the settlement’s treeless environs were once covered with juniper and piñon. The trees thinned out soon after the Anasazi began building their pueblos, and the Anasazi disappeared soon after the trees did.
Our computer simulations showed that with reasonable per capita fuel use and modest population growth, the people living at Chaco Canyon could easily have wiped out the surrounding woodland, says Betancourt. The Anasazi apparently learned about deforestation the hard way.
Betancourt’s latest efforts link midden analysis with carbon dioxide cycles and the global warming debate. In midden plant material, he and graduate student Pete Van de Water expect to find further evidence for the theory developed in the early 1980s that what brought the last ice age to a close 11,000 years ago was a 40 percent increase in atmospheric carbon dioxide. That theory was based on CO2 measurements taken from ice-core samples in Antarctica and Greenland. If it’s correct, then this increase should also be reflected in the changing density of midden plant stomata, the pores on leaves where gas and water exchange occurs. When CO2 concentrations are higher, the exchange is more efficient, and consequently plants need fewer stomata.
In addition, Betancourt says, the amount of that warming, estimated to have been between 5 and 9 degrees, is about the same as that predicted to take place due to the greenhouse effect--so the changes in vegetation from the last ice age can serve as a guide to the magnitude and rapidity of what may happen in the future.
While middens are proving valuable in climatology and archeology, Betancourt believes their real power will ultimately be felt in evolutionary biology. With the wealth of fossilized material in middens, scientists can finally put theories of speciation and adaptation to the test of history. Imagine having thousands of fossils that are already dated and identified, with a resolution of plus or minus 100 years in time, and plus or minus 50 meters in space, he says. That’s what middens represent. And since you can extract DNA from the material, you actually have the ability to do population genetics in the fossil record.
We can look at migrations that occur over hundreds of kilometers and thousands of years and ask, ‘Are these populations changing genetically as they migrate?’ We can take a living population and track it back in the fossil record. We can ask, ‘Where did this population come from? How did it get here?’
Midden analysis may even provide a test of punctuated equilibrium, a much-disputed theory according to which evolution takes place in bursts separated by long periods of stasis. To challenge that theory empirically, you’d need to look at population changes over a span of 500, 2,000, 10,000 years, says Jeffrey Mitton, a geneticist at the University of Colorado in Boulder. And you’re more likely to get that information from pack rat middens than from any other source.
There are other, and older, sources of DNA in the fossil record, Mitton says. The most recent example is a 135-million-year-old weevil encased in amber. But reconstructing the fate of a species from the genes of one individual is as risky as learning English from Bart Simpson. There’s no telling how eccentric your source is--was it the exception or the rule?
People go through the fossil record and they find a tooth here, a femur there, says Felisa Smith, an ecologist at the University of New Mexico. So they come up with these theories based on two or three single individuals. Maybe they’re 10,000 years apart. Maybe they’re different species. That’s what we’re stuck with.
Not anymore. With the aid of midden analysis, Smith is planning a rigorous fossil test of Bergmann’s rule, a pioneering evolutionary concept concocted in the nineteenth century by German physiologist Carl Bergmann without the encumbrance of hard evidence. He supposed that the body size of a mammal should be inversely proportional to average temperatures in the animal’s environment. In other words, mammals in hot climates like deserts should be smaller than mammals in cold climates like the tundra.
Elephants and giraffes do spring to mind as immediate contradictions, but most interpretations of the rule assume that Bergmann meant to compare races within a single species, or species within a genus-- not the wholesale fauna of a given temperature zone. And his reasoning is solid: owing to their high surface-to-volume ratios, small animals lose heat much more readily than large ones, so if they find themselves in a hot place, it behooves them to be small.
The best test of Bergmann’s hypothesis would show whether a population of mammals became smaller over time in response to an increase in prevailing temperatures. The mammals Smith is working with are pack rats themselves. By extracting DNA from the hair trapped in middens, she can trace the evolving genetics of a population to ensure she is working with a single species. Because she knows that average temperatures in the Southwest increased about 9 degrees at the end of the last ice age, all she has to do is document the size of pack rats before and after the thaw. And she can deduce body size from, well, fecal pellet width. There’s a correlation with length, but it’s much more robust with width. For obvious reasons. Smith had noticed the correlation between pack rat body size and pellet width in extant populations, but it was the kind of information she never knew quite what to do with. It didn’t occur to me that the pellets in middens were intact. When she found out they were, she knew she had herself a project.
My preliminary data already show a dramatic size decrease right when there’s an increase in warming, says Smith. About 11,500 years ago, the animals’ average body size was about 450 grams. Around 10,500 years, they’re about 325 grams. And at roughly 7,500 years, they’re about 275 grams, which is where they level off.
Bergmann wasn’t kidding. Apparently the rodents’ body mass has shrunk by some 40 percent since the ice melted, while their middens kept right on growing. Perhaps it’s the destiny of pack rats to be forever upstaged by their refuse.