Much remains unknown, of course - over generations, it's still possible that the introduced insects could develop a taste for native species, or set off unforeseen, and effectively irreversible, cascades of consequences. Post-release studies of biocontrol insects are often spotty and inconsistent, adding to the uncertainty. But so far, these modern biocontrols have shown few, if any, side effects. In many efforts, the problem is not over-enthusiasm but underperformance: The insects simply die out, leaving the job undone. 

Biocontrol, however, has an older, uglier ancestry. In the late 1800s and early 1900s, humans transplanted cane toads, mongoose and other species to new countries, naively expecting the newcomers to control domestic pests. People - and native wildlife - are still dealing with the consequences, and the most spectacular screwups have become cautionary tales. 

"When people were doing things like introducing cane toads, people were making idiotic introductions all over the world," Hufbauer points out. "It wasn't just in the name of biological control - people were saying 'Oh, a boar in California! Wouldn't that be great?' People were doing things that we now think are insane. So we have to hope that today, we have a pretty good understanding of what's going on, and that we're not making mistakes that 50 years from now, we'll look back on and regret." 

DeLoach and his team assured themselves that tamarisk, the sole member of its botanical family in North America, was the only target for incoming Diorhabda. What they didn't expect was that the weed itself might have some mourners. 

Just as DeLoach was making final preparations to release Diorhabda in the mid-1990s, the southwestern willow flycatcher, a small songbird native to river corridors in Arizona and New Mexico, was added to the endangered species list. Biologists found that for some of the few hundred remaining pairs of nesting flycatchers, tamarisk could serve as a passable substitute for native willows. If the beetle worked as advertised, they worried, some flycatchers might be left with little to hold up their nests. 

After all, no one was sure how quickly the beetle would chew through its targets, or exactly how far it would spread. And with the Colorado and many other Southwestern rivers long deprived of the natural floods that support native vegetation, the resurgence of cottonwoods and willows wasn't guaranteed. (The federal recovery plan for the willow flycatcher recommends that exotic species be removed in stages, without the help of biocontrol, and only in places otherwise hospitable to native plants.) 

So followed several years of study and teeth-gnashing on all sides. Tom Dudley, then a researcher at the University of California at Berkeley, was a stream ecologist "sick of watching tamarisk take over rivers," he says, and he began collaborating with federal researchers on the ecological impacts of the beetle. When they realized that some varieties of Diorhabda only did well at certain latitudes, Dudley called his old school classmate, Dan Bean, who - after his adolescent culinary experiments in California - had become an expert in how beetles respond to different lengths of day and night. 

Bean, working with other researchers, showed that beetles collected from northwestern China could only reproduce well north of the 38th parallel, which runs through southern Colorado and Utah, skirting willow flycatcher habitat. These beetles were eventually approved for release, and in the spring of 2001, with permits firmly in hand, researchers released members of this northern population of Diorhabda at several sites, including one in the Great Basin near Lovelock, Nev. 

For more than a year, not much happened, and "biocontrollers were pretty well bummed out," remembers Bean. Then, in the summer of 2002, researchers in Nevada noticed small spots of browning tamarisk. By the following summer, about 400 acres of tamarisk along the Humboldt River had turned brown, and by the summer of 2004, Diorhabda had made its first pass through tens of thousands of acres of tamarisk-infested riverbank. 

Jack DeLoach - now in his 70s - repeatedly postponed retirement in order to see Diorhabda on the march. "It's wonderful news," he says of the beetle's progress. "I never lost hope, but it was ... difficult." 

Despite their impressive arrival in North America, the beetles work slowly. Tamarisk and Diorhabda, after all, are longtime adversaries in their native lands, and tamarisk plants bounce back from beetle attacks again ... and again ... and again. Diorhabda larvae, which do most of the tamarisk chewing, can defoliate a 15-foot-high tamarisk in four days, but it takes five or six attacks over multiple years to kill a plant. 

And the beetles don't always win. "Not every release has worked," says Bean. "People who have been around successful releases can't believe the beetles aren't taking over the world, but a lot of them just flat-out fail." 

On the Bighorn River in Wyoming, where a university researcher released beetles in 2001, large swaths of brown tamarisk weren't seen until 2006. Beetles at several sites on the West Coast were obliterated by ants. "Those poor beetles just got chopped up and handed out until there was nothing left of them," remembers Dudley. 

No biocontrol is a complete cure. No matter how successful its campaign, the beetle will never rid the West of tamarisk, for that would mean destroying its own food supply. In optimistic moments, researchers envision that Diorhabda will knock tamarisk populations back by 75 or 80 percent, then settle into an equilibrium with the plant, helping create what one researcher calls "ecological fair play" by giving native plants more breathing room. 

But the acres of brown tamarisk in Nevada were more than enough to encourage new releases. Beetles were released in western Utah, southern Colorado, and Dinosaur National Monument. In 2004, a county weed supervisor in southeastern Utah released some Diorhabda from Kazakhstan - neighbors of the Chinese beetles - along the Colorado River near Moab. They swarmed through the river corridor, colliding with tourists' T-shirts and windshields, eventually covering some 100 miles of the Colorado. By the end of 2006, they had reached the mouth of the Dolores, a tributary of the Colorado, and were poised to head toward the Colorado state line. This past July, says Bean, "they started flying upstream like crazy," defoliating about 25 miles of tamarisk on the Dolores, and by August, they had crossed the border into Colorado. A nervous editorial in the Moab Times-Independent reflected local ambivalence: "We're excited about the apparent success the beetle is having," the editors wrote. "But we're concerned about the unknowns ahead ... we hope Mother Nature won't punish us for our actions." 

When Dudley and Bean make their fall foray down the Dolores, most of the beetles, taking their cue from the shorter days, are already snuggled between the leaf litter and the soil, stilled for the winter. But just weeks ago, many of the tamarisk along the river were draped with beetles, and the results are striking. On the state line, marked here by a small yellow sign and a barbed-wire fence, the thick tamarisk stands are almost completely brown. They're not dead yet, but the siege is under way: When the plants green up again next spring, Diorhabda will be waiting.