If sand dunes are a metaphysical essence of desert, then palm oases seem their essential complement, the moist, shady exceptions that prove the rule of fiery desiccation. A desert movie with dunes but no palms is unimaginable. One of the archetypes of this artistic convention is an oasis that geologist Clarence King described amid the Coachella Valley’s sands during his 1866 desert crossing:
Under the palms we hastily threw off our saddles and allowed the parched brutes to drink their fill. We lay down in the grass, drank, bathed our faces, and played in the water like children. . . . Our oasis spread out its disk of delicate green, sharply defined upon the enamel-like desert which stretched away for leagues, simple, unbroken, pathetic. . . . With its isolation, its strange warm fountain, its charming vegetation varied with grasses, trailing water plants, bright parterres in which were minute flowers of turquoise blue, pale gold, mauve, and rose, and its two graceful palms, this oasis evoked a strange sentiment. I have never felt such a sense of absolute and remote seclusion; the hot, trackless plain and distant groups of mountain shut it away from all the world. Its humid and fragrant air hung over us in delicious contrast with the oven-breath through which we had ridden. Weary little birds alighted, panting, and drank and drank again, without showing the least fear of us.
Such accounts epitomize the desert’s mystique, but palm oases are not essential to it in the way that dunes are. When there is a dry seabed or lakebed in a desert, dunes inevitably form as the wind piles up exposed sand. Anyone who tries camping in the ethereal-looking dune world will experience this inexorable process unromantically as the afternoon wind picks up and drives tiny but hard-edged quartz crystals against every inch of skin and equipment. Oases need groundwater, which reaches the surface only in scattered places. Much of the groundwater in today’s deserts accumulated in the last pluvial period, and is sinking.
Despite their archetypal airs, palms are not essentially desert plants, even for oases. Most species live in tropical rain forests, and although many do live in drier habitats like savanna and thorn scrub, they aren’t adapted to aridity as honey mesquite or catclaw acacia are. Their big leaves transpire a lot of water, and their fibrous roots are shallow, so they need permanently moist soil. Far from being essential aspects of desert dunes, palm oases are rare in them. They are rare in desert overall.
Even where groundwater nears the surface in dunes, there is seldom enough for palms. Deep-rooted shrubs like mesquite dominate the less arid hollows of most California dunes. Sweltering on one of the state’s spectacular sand piles, the Kelso Dunes in the east Mojave, I was glad to find oases of a sort, but not of King’s romantic variety. On the lower slopes of the 600-foot-high dunes were little thickets of desert willow, Chilopsis, a riparian shrub related to the catalpas of suburban streets. Tracks showed where other animals, a fox, a rabbit, had sheltered in one thicket’s thin shade—the fox quite recently judging from the freshness of its scats. But the water that let Chilopsis grow was many feet underground.
The only two groves of California’s native palm, Washingtonia filifera, I’ve seen in dunes were at Thousand Palms Oasis in the Coachella Valley Preserve near Palm Springs. Even they didn’t seem very archetypical, although Cecil B. DeMille filmed Bible epics there. Now protected from fire, the palms crowd the water, making the groves resemble swamps more than King’s “disk of delicate green.” Bushes cover even the dunes, blurring picturesque contrasts between white sands and green shade. One grove contains the only standing water I’ve seen in a California palm oasis, a lovely pool full of topminnows, nervous red crayfish, and blue-spangled desert pupfish, but not a “strange warm fountain” such as enchanted King. Deep in the palms’ shade, it is tidily dammed, with a piped outlet. The pupfish were introduced from their native waters elsewhere in the Coachella Valley because exotic species are crowding them out.
I don’t know where King’s magical oasis was. Like John Van Dyke, he was prone to hyperbole so his transports may have been fanciful. His “warm fountains” and “bright parterres” emit odd echoes of William Bartram’s classic eighteenth-century descriptions of Florida sinkhole springs and wet prairies. But there are Coachella Valley groves with hot springs, and Joseph Chase described oases that might have been King’s minus the romance. One called Seven Palms, “partly hidden among dunes” on the windswept plain northwest of Palm Springs, had “a score or so of the trees scattered about . . . the charms of a patch of dingy salt grass, a pool of barely drinkable water, and unlimited quail, rabbits, snipe and duck.”
Washingtonia filifera, commonly called California fan palm, grows mainly in steep places with rock or clay substrates that belie little surface moisture. Even at the Coachella Valley Preserve most groves occur not on the dunes but in surrounding badlands. They are spectacular, with some of the biggest palms I’ve seen nestled beneath sheer mudstone cliffs. Many birds visit the streambeds’ trickles of surface water. Watching white-winged doves and lazuli buntings drink one fragrant morning, I saw what looked like a cloud of confetti above the rustling fronds, and then realized that a flock of white pelicans was circling far overhead. But there isn’t much in the way of “bright parterres.”
Groves in places like Anza-Borrego and Joshua Tree National Park tend to be even gaunter and can seem trick parodies of King’s archetype. Their dark green heads look miragelike protruding from some blazing ridge, and when I reach them, I find that the trees are real but the oasis is not. Instead of a strange, warm fountain, the palms preside over scratchy mesquite thickets or alkali-crusted mud and salt grass. If surface water exists, it usually looks unappetizing, scummed with dust and bacteria, laden with dead insects.
The army officer botanist W. H. Emory probably had a more typical experience than King’s when he discovered a grove in Anza-Borrego in 1846: “Here, on November 29, several scattered objects were seen projected against the cliffs and hailed by Florida campaigners, some of whom were along, as old friends. They were cabbage trees and marked the locale of a spring and a small patch of grass.” Emory was glad to get the water and grass for his stock, but he didn’t go on about it. Five decades later, Van Dyke was even less enthusiastic, perhaps envying King’s romantic panache: “These are the so-called oases in the waste that travelers have depicted as Gardens of Paradise, and poets have used for centuries as illustrations of happiness surrounded by despair. To tell the truth, they are wretched little mud-holes.”
Still, fan palms have charisma because they are the biggest, strangest plants in the Alta California desert. (Baja also has another species, W. robusta, but boojums trump even palms.) Palms are strange trees just to look at—the only monocotyledons, relatives of grasses and yuccas, that get so big. Washingtonia grows up to seventy-five feet tall with a trunk three feet in diameter. The origin of such great water-gulping plants in the continent’s driest desert has always been hard to explain.
Chase imagined Washingtonia waving like coconut palms by vanished beaches: “Some of the groups occur along the boundaries of the sea that anciently filled the great depression which is now partly occupied by the Salton Sea, and whose beach-mark is today startlingly plain at the base of the encircling hills. Such groups, probably, represent the indigenous growths. A number more are found at higher altitudes, but of these many are known to have been planted by the present or former Indian inhabitants of the region.” Other writers thought Spanish missionaries had planted groves, although this seems unlikely given their remoteness. In 1940, Edmund Jaeger declared the species “definitely known to be a native of the California desert,” but characteristically did not ask how this came about.
In his 1950 article on desert origins, Axelrod cited W. filifera as one of the subtropical woodland species that had “closely similar fossils in the Miocene and Pliocene floras of the Mohave region and its border areas.” And indeed, no tree in California seems a better candidate for the romantic role of relict from a verdant Madro-Tertiary past. Sitting in twilit palm groves, I’ve had a palpable sense of ghostly woodland hovering in gloom where only boulders and bushes now exist.
Of course, the same questions arise with fan palms as with other organisms that apparently once lived in nondesert but now live only in desert. James Cornett, a curator of the Palm Springs Desert Museum, observed that the present Mojave’s high desert lacks wild palms because of winter frosts, but he had trouble explaining another absence: “It is less obvious why desert fan palms do not occur in the coastal climates of southern California. Winter temperatures are relatively mild and locations where permanent moisture can be obtained are numerous.” Cornett guessed that “relatively cool summer temperatures and/or frequent fires” excluded palms from the coast by giving other tree species a competitive advantage around water. But this explanation seems tenuous. Inner coastal valleys get almost as hot as desert in summer, and fan palms are resistant to fire. In the desert they can benefit from fire, since it clears the soil for their seedlings.
In the late 1980s, Leroy R. J. McClenaghan and Arthur C. Beauchamp, two San Diego State biologists who wanted to study genetic variation in wild plants, chose fan palm groves in Anza-Borrego as a subject. Regarding Washingtonia filifera as a relict of Madro-Tertiary woodlands, they made “three a-priori predictions about the genetic structure of this species.” They expected the isolated palms to have less genetic variability than more common and widely distributed plants, since variability dwindles in relict populations as they become smaller and more scattered. They expected larger groves to have more variability than smaller ones, since less would have been lost. And they expected that each isolated grove would show a lot of variability in relation to other groves, since “genetic drift” would have occurred during their isolation.
Only the first prediction proved right. The palms did have less genetic variability than commoner plants. But big groves did not have more genetic variability than small ones, and the level of variability from grove to grove was not as high as anticipated. This was surprising for a species that supposedly had been widespread in the Madro-Tertiary Geoflora at least since the Miocene, and then had retreated to scattered oases as climate dried and cooled. It suggested that fan palms had come to occupy today’s desert in a different way:
A more plausible scenario would be one in which existing palm populations in the Colorado Desert are the products of seed dispersal from a source population having reduced genetic variability. It has been suggested that climatic changes may have completely eliminated fan palms from the Colorado Desert and restricted the species to small refugia populations in Baja California (J. Cornett, pers. comm.). The subsequent movement of seeds into southern California from a refugium population in northern Baja California would have resulted in these colonizing populations also having low variability and high genetic similarity because of their common ancestry. Gene flow among these colonizing populations would have reinforced their genetic similarity.
Birds, coyotes, humans, and other consumers of the small but sweet (tasting like dates) palm fruits presumably would have carried out the gene flow.
Cornett originally had agreed with Axelrod about fan palm origins but had changed his mind. In 1989, he cited “four lines of evidence” against Washingtonia filifera’s supposed Madro-Tertiary relict status. First, he maintained that the species has never been as widely distributed as it is now:
There does not, in fact, appear to be any fossil evidence indicating that the species was once widespread in the Mojave Desert. Axelrod (pers. comm.) with one exception has retracted his earlier assertions that the fossils in question could be assigned to the genus Washingtonia. The one exception is a fossil collected near Wikieup, Arizona, and deposited in the collections of the Museum of Paleontology at Berkeley. Axelrod believes it can be classified as Washingtonia. However, the specimen could not be located at the Museum, and thus could not be examined by the author. An examination of additional Axelrod palm fossils by the author failed to reveal any specimens that displayed thorns on the petioles [leaf stems], an important characteristic of the genus, and therefore none could be classified as Washingtonia.
Second, instead of declining in numbers as might be expected of a relict species, Washingtonia was increasing, from approximately 17,700 individuals in 1961 to 23,266 in 1987. Third, instead of contracting in range like a declining relict species, Washingtonia was expanding, establishing new wild populations in Death Valley as well as in Nevada and Arizona. Fourth, as the two San Diego State biologists had determined, instead of having large genetic diversity among scattered populations, Washingtonia had little diversity, suggesting that the populations had diverged fairly recently. “These four lines of evidence,” wrote Cornett, “all point to a recently-evolved, invasive species, not a relict. It seems most likely that the genus Washingtonia first evolved in Baja California sometime after the peninsula broke away from mainland Mexico approximately 4.5 million years ago. Today, the two species in the genus, W. filifera and W. robusta, occur together only in Baja California, suggesting this is the geographic origin. Had the genus been present before the peninsula broke away, one would expect it to be represented on the mainland, which it is not.”
Cornett thought that Washingtonia might have evolved from Brahea, a related palm genus that occurs today in both mainland Mexico and Baja California. He speculated that “the glacial episodes of the Pleistocene” might have influenced the evolution of a “cold-tolerant palm” like Washingtonia, and that it might have “appeared within the present boundaries of the United States no earlier than the end of the Illinoian glacial episode.” In 1991, Cornett further speculated that, like opossums and armadillos, “the desert fan palm is another example of a species with tropical affinities that has extended its range northward during this period of global warming.”
Cornett did not speculate as to the significance for the desert’s pre-Pleistocene past if Washingtonia was not a Madro-Tertiary flora relict. Of course, the possibility that one species is not such a relict did not necessarily challenge Axelrod’s desert origin ideas. Whether or not Axelrod’s Madro-Tertiary fossil palms are Washingtonia or Sabal—the other genus Axelrod identified in desert deposits—they did exist along with the other fossil trees he found in today’s desert. A nonrelict Washingtonia might just be an exception to the Madro-Tertiary rule. Axelrod did not respond in print to Cornett’s ideas, suggesting that he found them unchallenging.
William Spencer, a naturalist, did respond to Cornett’s ideas, challenging most of them. Spencer argued that Axelrod’s fossils could indeed be Washingtonia, since fossil palms are hard to identify and many Washingtonia petioles lack spines. He argued that Washingtonia might not have increased since the 1950s, since early palm counts were haphazard. He argued that the low genetic variability of Anza-Borrego Washingtonia might not prove nonrelict status, since broadcasting of seeds by humans and wildlife could have homogenized the population. And he maintained that Brahea, the Baja fan palm genus, is too different from Washingtonia to be an ancestor.
Spencer didn’t emphasize Axelrod’s paradigm in his defense of Washingtonia’s ancient relict status. His opposition to Cornett’s ideas focused more on the assertion that Washingtonia had recently spread from Baja to Alta California than on Cornett’s implicit denial that Washingtonia grew in Miocene Madro-Tertiary woodlands. Spencer seemed more interested in affirming that Washingtonia had been in southeastern California “since the beginning” than in determining when that beginning might have been.
Ironically, one possible interpretation of Cornett’s ideas also implies that fan palms could have been part of the California desert flora virtually since the beginning. If Washingtonia did not evolve from a Madro-Tertiary woodland species, but arrived from the south, it could be a product, even if a fairly recent one, of Jerzy Rzedowski’s ancient Mexican desert.