It is hard to imagine even camels thriving in some of California’s present desert. Fortynine Palms Canyon on the north edge of Joshua Tree National Park is an example, a landscape of granite boulders like Cima Dome, but without Joshua trees, junipers, grass, or much else in the way of giant mammal fodder. Even spring wildflowers are sparse among the creosote bush and barrel cactus. The palm oasis at the top provides water for bighorns and deer, toads and tree frogs, but it is the usual Washingtonia grove, a crack in the rocks with a dusty trickle at the bottom. Still, the canyon does feature one relatively gigantic inhabitant. Despite its sparse vegetation, it has a highly visible chuckwalla population.
Perhaps this is because the granite there is particularly large grained and easy for scrambling about on, or just because the big lizards are used to hikers. Anyway, the canyon was the scene of the most dramatic lacertilian dalliance I’ve come across. While walking up the trail late one spring afternoon, I glimpsed something bluish black protruding above an upended boulder. When it began to bob up and down, I saw that it was the swollen, pop-eyed head of a male chuckwalla. Then a smaller greenish gray, wrinkled head appeared as if in response to the first’s push-ups—a female. They posed there in profile awhile, like pre-Columbian demigods on a stela.
Suddenly, the male moved up farther and, turning his body broadside to the female, displayed red and yellow spangles that looked iridescent as fire opal in the late sunlight. I found this spectacular for a usually drab species, and so did the female. When I climbed to a better vantage point, I could see that she was quite excited. She kept nosing at his tail, and twice got so worked up that she crawled on his back and rode him around briefly. She paid more attention to him than he did to her. After making his spectacular color display, he just kept doing push-ups, propped on a protruding bit of rock like a heraldic beast on a shield.
He eventually condescended to nose at her tail, but playing hard to get can backfire. She rebuffed him, scurrying away down the boulder’s side, and it appeared beneath his dignity to chase her. He dawdled there awhile as though trying not to seem disappointed. Then while the sun was setting he hurried in the opposite direction, probably toward the safety of his rock lair.
I could see why his tail had excited her. It was impressively long, plump and sinuous, colored a pristine creamy yellow that contrasted strikingly with the blue-black rest of him. It called to mind another Lewis Carroll nonsense verse:
How doth the little crocodile
Improve his shining tail,
And pour the waters of the Nile
On every golden scale!
That reminded me of something I’d noticed about the California desert fossil record. Crocodiles like those of today’s tropics have existed since the dinosaur age. California probably was too cold for them in the past few million years, but before that the climate was often warmer. At least three tropical dugong species left fossils in mid-Miocene marine deposits. If Madro-Tertiary California was as low lying and well watered as Daniel Axelrod maintained, it should have had many wide, slow rivers where fish were abundant and large herds came to drink—good habitat for crocodiles and for preserving crocodile bones. Their fossils abound elsewhere in North America. But I’ve seen no mention of them in California desert fossil deposits, which suggests that it was not such a low-lying, well-watered region.
Apparent crocodile absence is a minor anomaly in the Madro-Tertiary paradigm, one that Axelrod could have swept aside easily. After all, a few alligator-like fossils from just after the dinosaur age are known from the El Paso Mountains; perhaps later crocodiles simply haven’t turned up yet. Still, it is far from the only anomaly, and his reticence on desert evolution after the mid-1980s could have had another cause besides his paradigm’s ascendancy with biologists.
UC Davis botanist Michael Barbour made a good point in favor of his friend when he wrote, “During an era when most scientists became more and more specialized, Axelrod retained an ecosystem-level focus and curiosity. He asked, and answered, large questions.” But Axelrod’s virtuoso self-reliance had its downside. Rumblings of geological discontent had continued to grow, and a flurry of papers published in the 1990s may have left the volatile octogenarian speechless. Based on new techniques and concepts, they proposed ideas wildly at variance with his desert origin paradigm’s basic premise that California’s present high mountains began rising in the Pliocene epoch 5 million years ago.
Scientific paradigms undergo life cycles—from slow beginnings as skepticism and rivalry play out, to sudden ascendancy as the balance tilts in their favor, to surprise upsets as new scientific generations arise. The scientific generation that arose in the 1990s was impressively new. Their papers always had multiple authors and they covered subjects so technical and specialized that the authors sometimes seemed nonplussed by their own conclusions.
A 1996 paper on West Coast mountain origins in Science, one of the two leading journals, had nineteen authors. In four densely documented pages, it briskly knocked Axelrod’s basic premise askew by suggesting that California’s major mountains arose long before the Pliocene and have been high ever since: “The Sierra may have maintained or even lost elevation in the late Cenozoic as paleofloral and geomorphic arguments for uplift of the western United States have recently been questioned.” Using tectonic theory and new chemical techniques for analyzing fossil leaves and igneous rocks, the authors proposed startling new estimates of ancient elevations. They estimated that the Sierra Nevada rivaled today’s Andes in the late dinosaur age, rising to 18,000 feet above sea level overall, and that, although it lost elevation during the past 65 million years because of erosion and continental crust thinning, it was still an average 9,000 feet high 5 million years ago.
A year earlier, an article in another prestigious journal, Geology, had suggested that the Mojave region and the rest of southeastern California had “collapsed” some 24 million years ago when the Pacific and North American tectonic plates moved apart, thus stretching and thinning the continental crust. This raised the possibility that today’s western landscape of arid interior basins in the rain shadow of coastal mountains might have been the landscape of 20 million years ago as well.
In 1997, another Science article whose four authors included Jack Wolfe—cited by geographer and author Jeffrey Schaffer as among Axelrod’s main critics—challenged his basic premises about fossil floras:
Terrestrial plants are generally regarded as highly responsive to environmental changes, and thus fossil plants offer one of the best methods for inferring paleoenvironmental parameters. In one approach, the environmental tolerances of a fossil species are assumed to be the same as those of its nearest living relative. Because in this method it is assumed that plants do not evolve by adapting to different environments, conclusions based on this method have been interpreted to indicate that most of the Basin and Range Province of Nevada was at low altitudes (< 1 km) [under 3,000 feet] until < 5 million years ago, when uplift was inferred to have started which resulted in the 1 to 1.5 km present-day mean altitudes of the basins (based on Axelrod’s publications in University of California Publications in Geology from 1958 to 1992).
The paper posited startlingly different fossil plant “parameters” than Axelrod’s. It estimated, from “multivariate analysis of leaf physiognomy” of twelve mid-Miocene floras, that the Great Basin region stood a staggering 12,000 feet above sea level in the Miocene epoch 16 million years ago, almost three times its present altitude: “Paleobotanical evidence supports the hypothesis that Mesozoic thrust faulting and crustal thickening built a high terrain in what is now the Basin and Range province. . . . Geophysical observations combined with theoretical considerations of the region to the south of our study area suggested high altitudes at 20 Ma [million years ago] and a subsequent collapse; an increasing body of data and interpretations argue for high altitudes during the Tertiary in much of western North America.”
Axelrod’s fossil classifications also came into question. According to Schaffer, “Howard Schorn, a very meticulous paleobotanist, reevaluated Axelrod’s fossil plant identifications, and found that most of the plants were misidentified. When they were properly identified, they are found to be plants of former highlands, not lowlands, for Nevada.”
Schorn also questioned fossil identifications in other states. Axelrod had dated a deposit in eastern Oregon, the Alvord Creek flora, in the Pliocene epoch, and he had identified leaf fossils there as belonging to species of rose, madrone, ceanothus, and serviceberry (Amelanchier). Since species of these genera occur on the California coast today, this implied a low-altitude landscape. In 1994, Schorn and Nancy L. Gooch dated the Alvord Creek flora to the late Oligocene or early Miocene, some 15 million years earlier, and attributed Axelrod’s rose, madrone, and ceanothus leaves all to serviceberry. Most western serviceberry species today grow at higher altitudes, suggesting that Axelrod had gotten the elevation as well as the date wrong.
In 1998, the year Axelrod died of heart failure, a paper in the other leading journal, Nature, estimated ancient elevations similarly to the Science papers:
We conclude that the Sierra developed an Andean-scale topography between 185 Myr (the youngest marine strata preserved in batholith wallrocks) and 70–80 Myr (the minimum age of deep incision) coincidental with crustal thickening via arc magmatism [massive intrusions of molten rock caused when crustal plate motion shoves volcanic island arcs against continents] in the Sierra and contractional deformation to the east. By 70–80 Myr ago, both the San Joaquin and Kings drainages had deeply incised the western flank of the range, after which we infer gradual reduction of mean crestal elevation from 4.5 km to 2.8 km.
Again, far from rising abruptly to its present overall height of 9,000 feet in the past 3 million years, the Sierra Nevada shrank slowly from a mean height of 15,000 feet in the past 65 million years.
In response to a laudatory obituary by Michael Barbour, Schaffer somewhat severely summed up the case against Axelrod in 1999. Referring to his work in identifying fossil pollen, Schaffer wrote: “Because he believed that the Sierra Nevada and Great Basin were low landscapes until late Cenozoic time, he identified all his specimens as low-elevation analogs. Professor Roger Byrnes (among others) once told me that about ninety percent of Axelrod’s pollen identifications were wrong. Unfortunately, most botanists don’t realize this and they repeat his errors. Ax’s geology was atrocious. . . . It will take decades or longer to purge his voluminous imaginary species, imaginary landscapes, and imaginary climates from the field of paleoecology.”
The new scientific generation continued to hammer Axelrod’s paradigm in the new century. A 2002 article estimated ancient elevations by analyzing silicate isotopes in Miocene and Pliocene volcanic ash:
Volcanic ashes currently exposed in the rain shadow of the modern Sierra Nevada of California show no indication of large scale late Cenozoic surface uplift of the Sierra and corresponding regional rain shadow development. Rather smectite isotope data tentatively suggest that elevations may have decreased over this time by as much as 2000 m toward the southern end of the range and 700 m in regions farther north. This suggests that the modern rain shadow cast over the western Basin and Range has been in existence since pre-Middle Miocene and that the Sierra Nevada has been a prominent orographic barrier since before this time.
If the southern Sierra Nevada stood some 6,000 feet higher 20 million years ago than it does now, that would have created an impressive rain shadow indeed. And what would have grown in that shadow? Presumably not the oak and pine woodlands whose fossils Axelrod dug up in the Mojave. According to his critics, those woodlands would have been growing in mountains at higher elevations. Presumably an aridadapted vegetation would have grown in the rain shadow, perhaps one not so different from today’s Mojave Desert.
Yet another multiauthored paper based on new techniques suggested something of the kind in 2004, although it was about Oregon’s desert instead of California’s. The paper compared 30-million-year-old Oligocene epoch “paleosols,” fossil soils, with ones from the last ice age, and found them “surprisingly similar.” In successive strata of ice age paleosols, trace fossils of earthworms alternated with trace fossils of cicadas, implying shifts between grassland and sagebrush scrub. (Earthworms are rare in dry scrubland soils: cicadas feed on woody plants.) In the 30-million-year-old paleosols, earthworm traces also alternated with cicada traces. The paper also mentioned possible traces of desert succulents (Cactaceae, Euphorbiacea) and shrubs (such as saltbush, Atriplex) in the Oligocene soils.
A 2006 Science article estimated the northern Sierra’s early Age of Mammals elevation by analyzing hydrogen isotopes in Eocene river clays exposed by gold mining on the Yuba River. The estimated elevation was not as high as the southern Sierra’s, but it was still pretty high, around 6,500 feet: “The data, compared with modern isotopic composition of precipitation, show that about 40 to 50 million years ago the Sierra Nevada stood tall (plus or minus 2200 meters) a result in conflict with proposed young surface uplift by tectonic and climatic forcing but consistent with the Sierra Nevada representing the edge of a pre-Eocene continental plateau.”
The new “old desert” paradigm is not confined to North America. Another Science article published in 2006 reported dune deposits in the Sahara Desert from the late Miocene epoch, 7 million years ago. An article in the same issue estimated the rise of the Tibetan Plateau, which isolated central Asia from rainfall, at about 35 million years ago, tens of millions of years earlier than was previously thought. Other studies have pushed back the ages of South American deserts like the Atacama almost as far. Global desert may have been less of a climatic accident and more of an old earth feature than Axelrod thought.