The same year that the two San Diego State biologists published their analysis of Washingtonia genetics, another biologist published yet another possible explanation for desert fan palm’s limited distribution—that its main dispersers are extinct: “It is likely that contemporary rare desert trees with very localized distributions and fleshy fruits (e.g. the desert palm, Washingtonia filifera, which occurs in tiny groves in the Sonoran desert and has canid dispersed seeds at present) could become very common if once again serviced by a wide-ranging megafaunal dispersal agent such as a camelid.” The biologist was Daniel Janzen, who became prominent in the 1980s for his studies of the relationships between vegetation and megafauna, the diverse assemblages of large animals that inhabited the Americas until ten thousand years ago. Although best known for his work on tropical dry forest in Central America, Janzen also studied deserts:
There are many books and general treatises on the vegetation and vegetation types of the deserts of north central Mexico and the southwestern United States. None give consideration to the role played by large herbivorous vertebrates in shaping individual plants or their arrays. . . . Papers on cactus ecology classically ignore the Pleistocene (and earlier) megafauna, as do studies of spacing of desert plants. . . . In like manner, detailed discussions of the evolution of the biologies of the Pleistocene and pre-Pleistocene megafauna rarely consider that virtually the entire flora of large desert plants must have been continually under selection for defenses against these mammals. The camel has the largest gape of any extant ruminant and eats very thorny vegetation. The relationship of such a pair of traits is probably not evolutionarily fortuitous, and if it is, it still needs consideration to understand ecological fitting of camels to deserts.
Finding Miocene and Pliocene camel, mastodon, and horse fossils at Red Rock Canyon and Barstow in 1915, John Merriam had surmised that southeast California must have been grassland or savanna to support so many large mammals. In a lengthy 1967 paper on the extinction of American megafauna, Daniel Axelrod blamed its disappearance partly on the replacement of woodlands and grasslands by regional deserts. But Janzen’s research on African deserts led him to believe that American ones could have supported the equivalent of giraffes, elephants, and diverse other herbivores as well as big carnivores such as the African lion, which also inhabited California: “There were four genera of camelids in the western half of North America at the close of the Pleistocene, and they probably treated the North American deserts just as the contemporary African camel does its deserts. Camels move long distances among local wet sites, eat fruits, and defecate seeds, and range from Kenyan thorn forests to the driest deserts.”
Likewise, Janzen’s research on large herbivore behavior in relation to desert plant features such as thorns and poisonous resins suggested that those features played a role in defending the plants, and thus had evolved in response to browsing more than other factors:
In the absence of contemporary wild medium-sized to large herbivores, it has been fashionable to try to understand the spininess of arid-land plants largely in the context of their interactions with the physical environment. Spines on desert plants “were probably developed in the first place as a response to the dry atmosphere. . . . Furthermore, thorniness is most highly developed in the most arid deserts, exactly where large grazing animals are rarest.” Such a comment needs to be paired with the alternative view that the scarcer the perennial vegetation, the better protected it must be to survive. . . . It is not hard to imagine how browsing megafauna could select for arborescent lilies or botanical hedgehogs.
Janzen cited a Mexican tree as a contemporary example: “Acacia farnesiana clearly has a memory of browsers that were—e.g. leave it alone and the thorns are short and the leaves long past them, browse it with a pair of clippers and the next branches to be produced have wicked long thorns and shorter leaves among them—something I saw happen also with some native acacia in Morocco being browsed by camels and goats—below the browse line, fierce thorns, above the browse line much reduced thorns.”
Janzen was particularly fascinated by nopal, the tall, branching prickly pear that, two centuries before, had given the Jesuit Miguel del Barco inklings of biotic change. Noting that, except for humans, coyotes, and various insects, few animals now feed on the sweet, juicy nopal fruits, Janzen wondered how they could have evolved if not in response to browsing by large mammals such as camels, mastodons, and ground sloths. What other creatures would have been strong and thick-skinned enough to plow into “nopaleras”—thorny thickets of cactus and other plants such as mesquite, acacia, and creosote bush—to reach the fruits? What others would have had mouths and stomachs tough enough to chew and digest the prickly pears? What other creatures would have broadcast the seeds widely enough in their dung to make Opuntia cactus so successful?
“Then why the bright colors of Opuntia fruits? The traditional view of mammal color vision is that it is restricted to primates, tree shrews, and ground squirrels. However, a series of color choice tests with wapiti (Cervus canadensis) show clearly that they can distinguish orange from a variety of other colors, and cones have been located in the retinas of white-tailed deer and wapiti. I view the brightly colored large fruits of Opuntia as circumstantial evidence that at least some of the large herbivorous megafauna used color vision in food location.”
Janzen thought desert plant assemblages such as nopaleras would have provided a megafaunal smorgasbord, offering not only prickly pears but a variety of other conspicuous fruits—yellow mesquite and acacia pods, banana-like yucca and agave fruits—plus seeds, leaves, stems, roots, and even the cactus pads themselves. He cited the presence in a Shasta ground sloth’s dung from an Arizona cave of roots, stems, seeds, flowers and fruits of ephedra, globe mallow (a shrubby okra relative), saltbush, mesquite, agave, yucca, and nopal.
Janzen also posited a megafaunal origin for the vegetatively reproducing jumping cholla cactus: “Jumping cholla may well be the nastiest of the world’s burrs. When cacti break up at the stem joints through rough (or not so rough) treatment by contemporary herbivores eating or trampling stems or fruits, they are probably displaying a response selected for earlier by much more brutal treatment. It is easy to imagine the early evolutionary stages of jumping cholla as simply the spines on cactus pads lodged in thick skin on large feet.”
In another article, Janzen theorized that megafauna also might have affected smaller animals: “During recent fieldwork in Kenya (dry season, February 2–9, 1974), I saw not a single lizard or snake, and only one turtle, in about 1,000 miles of rural roads and 4 days of close scrutiny of four national parks ranging from 3,000 to 10,000 ft elevation. . . . Covering similar terrain and vegetation during the dry season in Mexico, Costa Rica, Colombia, or Venezuela, I would have seen hundreds of lizards and some snakes, with the same type of searching.” Janzen then conducted an informal survey of African versus American big game and reptile abundance, estimating that Africa’s reptile biomass was roughly 10 to 15 percent of its large mammal herbivore biomass, while America’s large mammal herbivore biomass in warm deserts and southward was 10 to 15 percent of its reptile biomass.
Janzen thought megafauna might depress reptile biomass in two ways. First, availability of carrion even during times of reptile scarcity could maintain large populations of raptors, small canids, and other habitual predators on reptiles. Second, big game herds could degrade reptile breeding and feeding habitat by browsing, trampling, and otherwise disturbing vegetation.
He thought herbivorous reptiles would be particularly vulnerable, not simply because of their dependence on plant food, but because of their metabolic requirements: “The ease with which leaf eating should evolve in a lizard fauna should be decreased as predator intensity increases, since it appears that very long periods of basking are an integral part of the digestive behavior of leaf-eating lizards. . . . The total absence of a foliage-eating Iguana or Ctenosaura [ground iguana] analogue from the African tropics is very conspicuous.” Janzen noted that African leaf-eating lizards are scarce not only in forests and savannas but in deserts: a study in the Kalahari recorded 1.2 percent of lizard gut contents to be plant matter, compared to 8.3 percent in North America. And the Kalahari has more lizard species than North American desert, although many are small and nocturnal.
Are California’s abundant desert reptiles beneficiaries of the American megafaunal extinction? This might seem more likely if the megafauna had died out here ten million years ago instead of ten thousand. We know that diverse reptiles very like today’s desert ones coexisted with camels and mastodons in the Anza-Borrego region several million years ago, even though most of the region’s known plant fossils aren’t desert ones. Yet the fact of our desert reptiles’ success remains. Might there have been “earth-old” deserts wherein extreme dryness or other factors released distant ancestors of chuckwallas and desert iguanas from megafaunal pressures long enough for them to evolve their unusual vegetarian habits?
Janzen did not ask such questions, confining himself to the Pleistocene epoch and recent times. He thought that many desert plants might have reached North America after the Central American land bridge formed three million years ago: “Cacti are widely believed to be of South American origin (A. Gentry, personal communication). If so, their original evolutionary interactions would have been first with the independently evolving South American megafauna, rich in large animals such as ground sloths, glyptodonts, and toxodonts, and later with the North American Pleistocene and pre-Pleistocene megafauna as the cacti moved northward (probably as seeds in the guts of megafauna).”
Many North American desert plants such as ocotillo and nolina are endemic, however, and even cactus origins are uncertain. And while desert plant fossils remain scanty in North America, fossils of big mammals that are known to adapt well to deserts, such as equids and camels, are abundant. Some Miocene American camels had giraffelike, elongated front legs and necks that could have been used for browsing in the crowns of Washingtonia palms, Joshua trees, or tall cactuses.
Of course, a fossil camel’s long neck does not prove that it fed on such plants. Most desert plants are short. Even if they did browse on tall ocotillos and Joshua trees, pre-Pleistocene camels might have interacted with such plants not in large regional deserts like today’s but in the thorn scrub, savanna, and woodland mosaics of Axelrod’s Madro-Tertiary Geoflora. “The plant-megafaunal interactions within the nopalera cannot be viewed in isolation from other habitats,” Janzen wrote. “For example, it is well appreciated that during the Pleistocene glaciations a much more forest-like vegetation covered what is presently desert and semidesert in northern Mexico and the southwestern United States. . . . It is easy to visualize a herd of gomphotheres [mastodons] ranging into nopaleras to eat Opuntia fruits in the summer, moving into the oak forest to eat acorns in the fall, and then back into the nopalera to eat Opuntia pads in the winter.” Still, the idea that large mammals shaped desert plant biology implies lengthy coevolution during times of high stress from aridity. Large herbivores would have treated woodland and savanna plants as brutally as they did desert ones, but those well-watered plants did not evolve such extreme defenses.
When I asked Janzen about this, however, he replied that growing in full sunlight can influence plants as much as growing in dry climate: “Ocotillo and cacti, for example, don’t need the dryness near as much as they need the sun that comes with the dryness and ensuing non-other canopies shading them. This means that desert-old plants can grow very well in non-desert places (e.g., rocky ridges, beach cliffs) and have desert-old life forms and co-evolve with large mammals etc.” That seemed compatible with Axelrod’s idea that desert plants evolved in scattered dry nondesert sites. But Janzen was too busy with Costa Rican conservation emergencies in 2008 to further pursue his ideas of two decades earlier. Prehistoric megafaunal effects on desert evolution remain largely unstudied.
Once I got the idea of megafaunal interaction into my head, though, it was hard to see the desert as I had before. When I took a walk on Cima Dome in the Mojave National Preserve one windy afternoon, the extraordinarily dense and tall Joshua tree woodland there seemed to shout: “Camels! Giraffe-camels!” Each branch of spiky yucca leaves clutched its crown of soft white blossoms as though holding them as far as possible from some ghostly browser. It was easy to imagine fifteen-foot Miocene camels sauntering through and nipping a cluster here and there.
As Janzen suggested, such a scene might not have been in a regional desert like the present. Cima Dome is one of the “edaphically arid” environments that Axelrod elected for predesert evolution of desert plants—a granite dome with thin, fast-draining soils and many bare slabs and boulders. At over five thousand feet elevation, with bunchgrass and juniper as well as Joshua trees and cactuses, it is more like Pleistocene woodland of ten thousand years ago than today’s scanty creosote bush and burroweed scrub. Even so, the idea of giraffe-camels feeding on the Joshua trees seemed to imply an older, more obscure desert past than Axelrod’s. And the possibility of such a past loomed larger in the years after Janzen’s papers.