ayne Belnap, the world's foremost authority on a highly specialized, complicated, and esoterically named group of organisms, sits in the dirt outside southern Utah's Canyonlands National Park. Red dust powders her bare toes. She's wearing a baggy T-shirt, shorts, and rubber sandals. Her whitish blonde hair is cropped into a helmet shape, and her round cheeks have a pallor more commonly associated with researchers who labor indoors than with biologists who work in the desert southwest. That's because her covered head is usually tilted toward the ground.

Belnap dunks her study subject, which looks like a clump of desiccated moss, into and out of a simple device called a slake test kit. It resembles a plastic ice cube tray and is filled with water. "You see how it holds together?" Belnap says, beaming. "Now that's good crust." A free thinking, fast-talking, slightly androgynous-looking grandma, Belnap has lived in Moab, Utah, for 20 years, working first for the National Park Service and now for the U.S. Geological Survey. She spends her days talking to government agencies, ecologists, academics, and environmental ministers around the world about an ecological community that 15 years ago barely registered on most biologists' radar: cryptobiotic soil crusts.

A tightly knit assemblage of lichens, fungi, algae, mosses, and cyanobacteria (once known as blue-green algae), soil crusts go by many names -- cryptogamic, microbiotic, cryptobiotic, microphytic. Each of the crust's constituent organisms holds little biological sway on its own. But collectively, they do nothing less than hold together the desert floor.

Their ecological role may be impressive, but crusts, from a human perspective, don't look like much. Blackish brown and dry, they resemble something burnt onto the bottom of a pot. Assume a rabbit's viewpoint, though, and crusts achieve a level of sublimity. Deep and velvety, they undulate over the desert floor, sometimes 10 centimeters high. Staring into the fungal forest, it's easy to imagine miniature echoes of the region's sandstone architecture -- here a castle, there a hoodoo. In the dry season, the crusts smell only vaguely mossy; come winter, they'll swell with moisture and assume the musty redolence of fallen wet leaves.

"It's a red-brown smell," Belnap says, sitting cross-legged in the wide gash left by a caterpillar tread on the desert floor. Grown in culture, the organisms taste like seaweed. And if you want to hear "cryptobabble"- crusts talk all the time, she says -- "you have to be very quiet and put your ear right next to the ground. You'll hear this faint murmuring." Belnap smiles impishly, and I almost fall for this bit of fungal fancy.

"Let's see what we have here," she says, plucking a chunk of reddish crust from beneath a shrub. She squirts the dirt with water from a plastic bottle. "Cyanobacteria, no moss." She dunks it in the water, and sand grains plummet to the bottom of the tray. She squirts another chunk for my edification, this time a blackish brown sample, and brings it to eye level. "We've got moss and lichens here," she says, pleased. "You see how it traps sediment?" Indeed, the filamentous tangle holds sand in a death grip. "These guys are bombproof," she observes, then corrects herself. "They're impervious to wind and rain erosion, but one pass of a Jeep tire crushes them."

If there's irony in dirt, it lies here. Cryptobiotic crusts have adapted to one of the harshest environments in the world -- highly alkaline, with extreme aridity and extreme temperature fluctuations. Yet the slightest bit of squashing, which played no part in their evolutionary history, will do them in. (Yes, these crusts evolved beneath the pounding of prong-horn antelope hooves, but antelope tend to roam this landscape only when it's frozen or moist.) If damaged crust is left alone for about 10 years, Belnap says, it may recover to its immature state, a community of cyanobacteria. But a mature and far more stable community, which includes lichens, fungi, and mosses, may take several hundred years more.

Biological soil crusts occur in arid and semiarid landscapes all over the earth, covering about 35 percent of the globe's soil surface. But the crusts in and around Moab, Belnap says, are especially threatened: "We're destroying them at a stunning rate." Recovery is especially slow here, because of low precipitation and the consistency of the sandy soil. Unlike the Mojave or Sonoran deserts, this region has little wind protection -- no rocks and few surface-rooting plants -- to keep the soil in place. Belnap reaches into her backpack and pulls out a photograph of a dead juniper tree whose bare roots appear to hover several feet above the desert floor. What happened? I ask. Vehicles damaged the crust, she explains, the soil fell apart, and eventually windstorms blew everything away.

"Without this soil," she continues, "you can lose native plants and shrubs." Then things could go from bad to worse: Exotic species, like cheatgrass, leap to take advantage of the weakened native community. Animals that fed on the original species go elsewhere or die. Eventually the desert ecosystem becomes weedier and simpler, a pale imitation of its former complexity.