The thesis had come off a Lied Library shelf, bound in black, covered in dust, 232 pages long, and bearing the scintillating title, "A Floristic and Vegetation Analysis of the Newberry Mountains." No one had looked at it in the last decade. But what it contained was utterly remarkable to Scott Abella.

The Newberries are border country, not merely between Arizona and Nevada but between the inhospitable Sonoran Desert and the almost extraterrestrial Mojave. It's a place where two ecosystems meet and compete, a conflict that could be influenced by climate change and invasive species. The Newberries are also part of one of the world's least studied ecologies. All in all, an ideal spot to expand the boundaries of human knowledge, venomous snakes aside.

So in that thesis Abella first held two years ago, the environmental studies professor saw a baseline record of the past, one that you could compare to the present to puzzle out the changes in Southern Nevada's landscape.

Unfortunately, there wasn't quite enough in the thesis; it didn't include raw data or an especially detailed map. If only there were some way to recover the missing information, Abella thought. But surely there wasn't, not after 30 years.

The only clue was that the long-gone graduate student had worked with the National Park Service. Not long afterward, Abella was out at the Park Service's office at Lake Mead talking to Alice Newton, a vegetation management specialist who coordinates research with UNLV. Abella asked if there might be any archived files or longtime employees who could point the way to the thesis's author, a guy named Jim Holland.

Yeah, Newton said, he's right there, about 100 feet down the hall.

And the research he conducted 30 years ago?

He still had it: 150 pages of handwritten data in two cardboard-covered three-ring notebooks, plus taped-together pieces of graph paper and meticulous maps. They were yellowing in his garage. At lunch, he went home and got them.

"It was pretty exciting because I'm the type of guy who still has all of that stuff," Holland says.

Maybe that thesis held sentimental value. The research launched Holland's career and convinced him to apply for a Park Service job when he was awarded his master's degree in biology in 1982. After postings in North Carolina, the Everglades, Denver, and Glen Canyon, Holland came back to Nevada and Lake Mead National Recreation Area. He is in charge of planning and recreational development of the park's 1.5 million acres, which includes part of the Newberries.

Abella told Holland he wanted to send another grad student out to replicate the research, only this time with GPS coordinates, photographs, databases, and other tools of 21st-century science. Thrilled to see his early work pass on to a new researcher, Holland also saw how the data could be useful to Park Service land managers.

"Our agency is dedicated to preserving the natural history for the public to enjoy," he says. "So as these systems are changing, we're trying to understand how our long-term goals should change as well. It might not significantly affect how we manage the lands but it certainly will change how we interpret things for the public."

Holland had collected his data in 1979 from 112 swaths of land. Each plot is 100 meters long and three meters wide, and every plant therein is counted by species and recorded. The plots are scattered over roughly 200 square miles of difficult terrain that is home to three species of rattlesnakes. They start at 400 feet above sea level and rise to over a mile, from the comparatively lush area around Davis Dam to the bright austerity of Spirit Mountain, which is the central site of human creation in the traditions of Yuman-speaking tribes and is significant to Hopi Native Americans as well.

It took Holland three months to sample all of his sites. It would take his successor, a graduate researcher named Chris Roberts, six months to locate and re-sample the 103 remaining intact sites and count 39,000 plants.

"I think [Holland] was significantly younger then than I am now," Roberts deadpans.

Roberts is lean and in his late 30s with dark hair. He's much paler than you would expect of a man raised in Texas, schooled in New Mexico, and fresh off six months of desert research. He credits his library-grade tan to hats, long sleeves, lots of water, and SPF 60 sunblock.

It is more difficult to rediscover sites from the pre-GPS era than it is to pick wholly new sites. Moreover, Roberts was collecting more data than Holland did - including information about sunlight received and soil types.

Another challenge was simply that the Newberries are a steep and rugged solitude, unfriendly to humans, and punctuated by granite batholiths rising from the eroded terrain like the bones of extinct monsters. "It kind of looks like some Fortress of Doom," Roberts says.

It is, he adds, quite beautiful.

Despite their unyielding appearances, deserts are surprisingly fragile landscapes. They are easily disturbed and slow to recover. For example, during World War II the Army trained tank divisions in the southern Mojave. The tread marks are still visible almost 70 years later.

Roberts wants to piece together a natural history of the Newberries for the last century or so and beyond. He has scoured the archives of the Boulder City Historical Society and the Searchlight Historical Society, which is pretty much just one woman, Jane Overy, and one room - "It's like a closet," Roberts says.

But that closet contains treasures - historical letters and photos showing areas like Christmas Tree Pass as they looked in the 1920s. Roberts has studied the unlikely history of cattle ranching in this corner of the Mojave. He's also studying others' analysis of pack rat middens - nests made of plant material and animal dung sometimes thousands of years ago and preserved in crystallized rat urine.

That work doesn't yield hard data that he can compare directly to Holland's thesis and Roberts' own re-sampling, but it can help put such data in context.

"The real skill isn't to gather the data, it's to tell a story here," Roberts says. "I can stare out at a landscape and start to put together a history."

And history is a way to look to the future.

As always, good science means collecting more data. The more data, the better the predictions. Roberts' new data, along with Holland's old data, are being added to the Mojave Inventory and Monitoring Network, a database that makes it easier for scientists to draw upon and follow up on past research.

Already, Roberts can tell you some things by comparing Holland's snapshot of plant life in the 1970s to today. For instance, there are more barrel cacti and at higher altitudes now. Other species have also proliferated, like invasive grasses. "I cannot, in good conscience, tell you that, for instance, barrel cacti are going to increase, but someone who has five snapshots could tell you. Right now, we have two," Roberts says.

Because some plants in the desert live for a long time and change slowly, it's hard to say for sure how big of an impact climate change is having. These species may be fine or they may be among the living dead, a species that could die out in the coming decades.

Consider the uncertain fate of the creosote bush. It is gray-barked, thin-leaved and seemingly unremarkable. It lives for about 150 years and reproduces rarely and only under exceptionally favorable conditions: plentiful rain, at the right time of the year, for three years in a row.

Invasive brome grasses imperil the native desert ecology because they reproduce yearly, easily, and widely. And they have a nasty habit of burning, which makes the other three traits all the more deadly.

Wildfires once were rare in the desert simply because native plants dotted the landscape rather than filling it. A lightning strike generally consumed a small patch of plants before burning itself out. The grasses surround native plants, giving fire a path over many acres. Then, after a fire, the grasses quickly repopulate the area, including the parts that used to have perfectly nice creosote bushes. A drier climate might make it even easier for brome grasses and harder for creosote. Were a creosote capable of feelings, it would be an indignant bush.

Why might humans care about the fate of creosote bushes? After all, there were bushes and now there are grasses. Big whoop.

There are practical considerations; a landscape of highly flammable grass is more likely to burn down your house. Land managers need to know that. But also, the bushes are part of a whole ecosystem of balanced plants and animals, including the charismatic ones like tortoises and bighorn sheep. Maybe if bushes are eliminated, the rest of these species will survive, but maybe not.

As Abella puts it: "Think of a brick wall. You can pull out a couple of bricks and the wall will still stand. But how many bricks can you pull out?"