Ivan Johnston presented his desert shrub paper at the Eighth American Scientific Congress in Washington, D.C., in May 1940, and it was influential. A 1944 textbook resoundingly titled Foundations of Plant Geography echoed its idea that very ancient South American deserts had been the source of ancient North American ones. But at least one member of the audience may have questioned Johnston’s version of the desert past.
Daniel Axelrod was working as a postdoctoral fellow in paleobotany at the National Museum and the Carnegie Institution when Johnston presented his paper. The subject interested him, albeit from a different angle. When Axelrod was getting a PhD in paleobotany at UC Berkeley during the Depression, most junior scientists couldn’t afford to do much larking off on fieldwork such as Johnston had enjoyed in the 1920s. But Axelrod was an unusually energetic paleobotanist—probably one of the most energetic ever—and in 1940 he’d already been investigating the California desert’s past for six years.
Energy ran in the family. Having emigrated from Russia to Brooklyn, where Daniel was born in 1910, the Axelrods hurried on to Guam, of all places. Then they moved to Hawaii, where young Daniel became a surfer, then to the San Francisco Bay Area, where he joined the Boy Scouts and took up hiking. This led to an interest in forestry, and Axelrod put himself through the UC Berkeley Botany Department by working on Forest Service plant surveys in the summer. But forestry and field botany didn’t offer enough intellectual stimulation for a student who was avidly absorbing geology, climatology, and evolutionary biology as well.
Displaying characteristic decisiveness, Axelrod recalled an encounter that did offer enough stimulation: “In my junior year, Harry D. MacGinitie, then an ‘old man’ of about 35–40 years, came to the class in Forest Botany carrying a tray with well-preserved fossil plants—the Trout Creek flora. I asked what he was doing with them and learned that he was going to the herbarium to identify the species. I told him I knew what several of them were, so we went to the herbarium and looked them up—I was correct. I then realized that this was the sort of thing I would like to do.” Axelrod didn’t specify whether the sort of thing he would like to do was identify fossils or be proved correct about them. Doubtless it was a bit of both. He did develop very strong opinions on fossils and evolution during a career that would last the rest of the century.
California desert fossils puzzled him for several years, however. His early work tended to support the field botany version developed by Frederic Clements, Forrest Shreve, and Ivan Johnston. Foundations of Plant Geography noted this: “Fossil evidence bearing directly on the problem of the origin of the desert climax is not very abundant, although recent Pliocene and Pleistocene discoveries by Axelrod indicate the Mexican origin of the vegetation and largely substantiate Clements’s account.” Axelrod’s first publication on a California fossil flora, from the several-million-year-old Pliocene Eden Beds in Riverside County, implied a respectable antiquity for at least some local desert vegetation. Although the fossils mainly consisted of plants typical of nondesert southern California today—manzanita, ceanothus, live oak, Coulter pine, and ghost pine—they apparently included two taxa that now live in Mexican desert:
The occurrence of a desert element comprising Platanus (cf. P. wrightii) and Sapindus is consistent with the southern occurrence of this fossil flora, and is indicative of more arid conditions during the Pliocene than exist at the fossil locality today. Sapindus and Platanus reach their best development in the desert under a rainfall of 5 to 10 inches annually, while the chaparral now growing at Beaumont exists under a rainfall of approximately 20 inches. . . . With increased rainfall in the late Pliocene and Pleistocene, the desert elements probably retreated southward along available lines of migration, since these species are no longer found in California.
But Platanus and Sapindus—sycamore and a small tree called soapberry—are riparian plants in desert: most sycamore and soapberry species occur in woodland. When Axelrod published a longer piece on the Eden Beds a few years later, the picture had become more complex. The sycamore in the Eden Beds had begun to seem more like the species in nondesert coastal California today, Platanus racemosa. And although fossils of modern desert plants like mesquite, ephedra, and a sunflower family shrub called scale broom (Lepidospartum) occurred in the Eden Beds, Axelrod didn’t find enough of them to confirm his previous impression: “True desert conditions such as are found in the present adjacent Colorado Desert may not have been existent in the basin during Mount Eden time.”
Axelrod wrote his doctoral thesis on another fossil flora located just west of the present Mojave, in the mountains above Tehachapi Pass. The Tehachapi flora of fossil leaves, fruit, and wood that he dug from white ash deposits in volcanic beds was of Miocene age, millions of years older than the Eden flora. According to Johnston’s old desert paradigm, it might have been expected to contain many desert elements. It did have what Axelrod called “a typical desert scrub association” of desert peach (Prunus), mesquite (Prosopis), and a spiny buckthorn relative called abrojo (Condalia). But he thought the scrub had grown only on dry lower slopes. The overall vegetation was lusher, with subtropical trees like bursera and euphorbia bordering the scrub; ash, walnut, sabal palm, and cottonwood sharing riparian areas with sycamores; an oak savanna with ceanothus, buckthorn, and locust on uplands; and a woodland of pine, cypress, madrone, wax myrtle, and bay like that of today’s coastal California on hills.
Axelrod found no fossil evidence in the Tehachapi site that creosote bush, ocotillo, cactus, or turpentine broom had grown there. Rather, the site suggested to him a much less mountainous landscape than today’s, where twelve to twenty-five annual inches of rain watered woodland that had a dry season and semiarid patches but was otherwise unlike desert. He perceived similar conditions in other southern California fossil floras, including even older ones. A site from the Eocene epoch, twenty million years earlier than the Tehachapi flora, contained ancestors of walnut, wax myrtle, and soapberry.
Military service in the Pacific analyzing the aerial photos used to prepare for island invasions interrupted Axelrod’s career from 1942 to 1946, although he kept publishing on fossil floras. Then he taught in the geology department at UCLA, which probably frustrated him by limiting his research time. Although a good lecturer, he was impatient with academic chores like acting as a thesis advisor. He said that few students knew enough to make it worthwhile. Within a few years, nevertheless, he managed to publish a ringing challenge to the established desert origin paradigm:
A wide diversity of opinion exists with respect to the age and derivation of modern desert environments. According to one view, desert vegetation is an “earth-old feature.” A second theory, corollary to this, is that desert floras of essentially modern character have been in existence in their present positions since angiosperms first assumed dominance during the Cretaceous Period. Although both opinions have been expressed on several occasions during the past decade, no evidence supporting them has been presented. . . . Analysis of the available data shows clearly that there were no desert environments of wide extent during those times. Thus a third opinion, and the one to be elaborated here, is that the desert environments now characterizing wide subcontinental regions are a phenomenon of only the latest part of geological time.
Axelrod claimed Daniel MacDougal as an intellectual forebear, citing his declaration that “xerophilous types of vegetation are of comparatively recent origin” and his belief that California desert plants had evolved here. He ignored MacDougal’s ideas of sudden genetic mutation, however, and otherwise distanced himself from the previous scientific generation. Unconcerned with ideas of plant succession and ecosystem “evolution,” he dismissed Frederic Clements’s notion that a creosote bush and burroweed “desert climax” had invaded from Mexico recently, competitively replacing other climax vegetations. “Clements’s postulate of a dominant hardwood-deciduous forest in the region now desert during Eocene and early Oligocene” was “untenable” and his “belief” that grassland dominated the region in the mid-Tertiary was “also in error.”
Axelrod took longer to dispute Ivan Johnston’s idea of an ancient South American desert spreading north in the dinosaur age, but he did a more thorough job:
There appears to be no support from the geological record for the recently stated belief that “the desert floras may well have an age and continuity comparable with the floras of the wet tropics (Johnston, 1940).” In all cases known, their present areas were occupied by more mesic vegetation, not only in Tertiary time, but in Cretaceous as well . . . the present regional desert climates and flora seem to be no older than late Cenozoic anywhere in the world. Such environments are not “earth-old features.” They may more appropriately be called “climatic accidents” for they have been rare in earth history.
Axelrod’s recent desert had one particular advantage. It was more compatible than an “earth-old” one with the long-prevailing idea that West Coast mountains like the Sierra Nevada rose within the past few million years, causing a desert-forming rain shadow. Newer evidence seemed to support this. Francois Matthes, a geologist who studied the Sierra in the 1920s, decided that high plateaus there are ancient relics of low-lying, level landscapes that have been elevated along with the peaks since Pliocene times.
To be sure, the more geologists studied the landscape, the more complicated it seemed. By the 1950s, they knew that high mountains had existed in the Sierra Nevada’s present location during the mid to late dinosaur age. The granitic rocks of today’s Sierra originated as molten magma when those mountains arose over two hundred million years ago. Most thought that the ancient mountains, called the Nevadan orogeny, had eroded away into plains and low hills by the beginning of the Age of Mammals, although there wasn’t conclusive geological proof of this. Indeed, Axelrod regarded his nondesert plant fossils as some of the best available evidence that southeast California was more low lying, level, and moist than it is today for most of the past sixty-five million years.
Axelrod had to acknowledge Johnston’s point that some long-distance migrations of species like creosote bush must have taken place, but he didn’t think there had been enough of them to prove that the desert is old: “Such migrations account for only a small fraction of these floras, which are otherwise largely distinct.” He maintained that most California desert species evolved in the past three to five million years from plants that lived in the preceding woodlands and savannas but were able to survive and adapt in the climatic accident of desert.
Axelrod cited the desert’s most attractive feature, its spring wildflower displays, as one living proof of its youth:
Increasing topographic and climatic diversity of these areas during Cenozoic time has resulted in the differentiation of numerous species (and varieties) well adapted to these more narrowly defined environments of the desert and adjacent regions. Thus it is not surprising to find that many desert herbs are polyploids [a mutation involving multiples of the normal diploid chromosome number] whose nearest ancestors have equivalents in the forest, woodland, scrub, and grassland vegetation now marginal to the desert region. The desert floras must therefore have evolved since their more mesic ancestral species inhabited the lowlands of the area now desert.
It is striking how many desert spring wildflowers are related closely to ones in the much lusher landscape of the coast. Of the species I’ve mentioned in places like Red Rock Canyon and the Providence Mountains—fiddlenecks, poppies, larkspurs, lupines, phacelias, paintbrushes—almost all have congeners common in coastal grassland and woodland. Most of the desert’s wildflower species are endemic to it, but plants so much like coastal relatives might indeed have diverged from them recently.
Axelrod also cited desert shrub genera like Baccharis, Rhus, and Prunus that have close relatives (coyote bush, poison oak, and wild cherry) in coastal habitats. It was harder, however, for him to explain the majority of desert shrubs that have no close relations in coastal grasslands or woodlands. Many of those come from groups centered in the tropics. But he pointed out that many have more moisture-loving close relatives to the south in Mexico and Central America, and he maintained that they could have evolved their desert forms recently, like the wildflowers. His theory wasn’t perfect, Axelrod seemed to say, it was just better than the others:
The task of determining the origins of modern desert vegetation is by no means an easy one, as is amply evident from the preceding discussion. The problem is one whose solution rests largely with paleobotany, though it is clear that data from modern plant distribution can give important clues and that cytogenetic studies can aid greatly in determining centers of differentiation. The procedure followed above for interpreting the history of the Great Basin, Mohave, and Sonoran deserts appears sound. Furthermore, the general pattern of desert development elsewhere in the world seems to parallel closely that of western North America, with respect thus to age and origin.
Little published evidence exists of the botanical establishment’s reaction to Axelrod’s brusque dismissal of its ancient desert paradigm. That there was no reaction seems unlikely, especially given what he wrote about his opponents viewing desert as an “earth-old feature.” This purple epithet is too much like Ivan Johnston’s description of desert as an “old earth-feature” in his 1940 paper to be accidental. Axelrod may not have been mocking Johnston by reversing “old” and “earth.” The 1944 textbook, Foundations of Plant Geography, reversed them solemnly: “The Larrea-Fransera desert developed through . . . immigration of Mexican Sonoran species from the south where the desert climate and climax have long existed because they are an earth-old feature.” Still, such little things can prickle, like a cholla spine on a pants leg.
Johnston himself was increasingly hors de combat by the 1950s. Harvard’s administrative politics reduced him to a distracted outsider, and he published infrequently before his death of a heart attack in 1962. But Axelrod definitely nettled others, setting a lifelong tone. He particularly nettled geologists, some of whom looked askance at his use of plant fossils to dictate the landscape’s evolution over the past sixty-five million years. As a California colleague said four decades after Axelrod first challenged “earth-old” desert: “I thought by all means he should have been elected to the National Academy of Sciences, but he’s a very prickly person, and he raises a lot of hackles. There were two or three members of the geology section of the National Academy with whom he had particular quarrels, and it became clear that no matter how much we pushed from our side that there was opposition so he would never get in.”