Cholesterol coating arterial walls. Malignant cells replicating through the colon. Depressive thoughts assailing the mind.

A late diagnosis can mean cardiac arrest, inoperable cancer, suicide. But early detection offers time to potentially reverse those system-wrecking forces — potential that has spurred the medical field to invest decades and countless dollars into screening techniques.

The University of Nebraska–Lincoln’s Dirac Twidwell, Dan Uden and fellow ecologists are wary of waiting on their field to catch up, of watching 10,000-year-old land cover disappear across neighboring states, of wondering whether the Sandhills and Loess Canyons can rebuff the ecological equivalents of infection.

They know the perils of reacting late to the so-called regime shifts that shuffle ecosystem dynamics as invasive vegetation overruns native: animal species facing existential threats, ranchers losing crucial forage for livestock, unpredictable issues arising from wildfires and floods.

“There’s a tipping point at which your system is fundamentally altered,” said Twidwell, associate professor of agronomy and horticulture. “It’s either very difficult to recreate a healthy system, or it’s irreversible, and you’re then living with a new reality.”

By relying on supercomputers, unprecedented land-cover data and a healthy dose of ecological theory, the researchers have introduced an approach that could finally help conservationists and landowners treat early warning signs rather than entrenched symptoms.

“There’s a class of problems in medicine that you can’t solve if you wait until there are serious symptoms,” Twidwell said. “Those exist in nature. We should do the same to prevent those things from happening.

“Imagine going into a doctor’s office. They take the samples they need and then inform you, ‘Great, we’ll get back to you in four to five years.’ That’s been the speed of our profession, but technology is changing everything. Now we can actually get back to people in a week.”

Boundary conditions

The team’s approach owes much to the land-cover dataset developed by researchers at the University of Montana and accessible on the Rangeland Analysis Platform, which offers bird’s-eye views of virtually every 100-square-foot tract of land across the western United States. Using algorithms and other analytical tools from Google Earth Engine, the team examined five types of land cover — bare ground, annual grasses and flowering plants, perennial grasses and flowering plants, shrubs, and trees — that covered western U.S. rangelands between 2000 and 2017.

The platform allows its users to measure the coexistence of any two land-cover types on each 100-square-foot tract. One tract might feature 57% annual flowers and 43% shrubs, or 84% perennial grasses and 16% bare ground, or 100% trees.

With state-of-the-art supercomputers allowing Twidwell, Uden and their colleagues to analyze the interactions among those land-cover types over time, the team can then search for the telltales of emerging regime shifts. Those telltales? Notable breakdowns in coexistence patterns that have held for many years, where one cover type is now displacing another to spread consistently across multiple tracts of land.

If multiple adjacent tracts once featured equal shares of perennial grasses and bare ground but then became dominated by 90% bare ground during the next decade — if one rises in prominence as the other falls — the engine will effectively flag it as a warning sign.

“So we see the boundaries between different regimes, and we can watch those boundaries either stay stable over time or start moving, which is what we’d be concerned about,” said Uden, a postdoctoral researcher in agronomy and horticulture. “We don’t just look at what’s happening at a given location. We’re able to see if change is happening not only right where you are but in your immediate neighborhood and in your distant neighborhood. So we can use it to see farther out and then track that change over time.”

The researchers recently demonstrated their approach by analyzing areas at known risk of troublesome regime shifts. Two of those shifts loom over Nebraska. One — the displacement of native grasses by the invasive eastern redcedar tree — especially threatens the Loess Canyons near North Platte. The other — erosion that can lead to desertification — has historically troubled the Sandhills, which cover more than one-fourth of the state.

Twidwell cited a wildfire that struck the Sandhills in 2012, and his research group’s attempts to begin tracking any long-term consequences of that wildfire a year later, as evidence of the progress that the new screening approach represents. Following the wildfire, the group manually and annually sampled three-square-foot plots throughout the affected area.

“We were relying on a small subset of information and probably sampled less than 1% of the total area,” Twidwell said. “Over this last year, we’ve gone from that to where we can see every (100-square-foot) pixel.

“It would have taken an army of field workers, in a profession that doesn’t have it, lifetimes to do what is now possible.”

‘That’s where my heart is’

Putting the screening approach to proper use across Nebraska, and living up to the land-grant mission of the university, means working with those working on behalf of ranchers and wildlife, Twidwell said.

Shelly Kelly, program director of the nonprofit Sandhills Task Force, is a leading advocate for preserving the region and the livelihoods of the ranchers who depend on its mixed-grass prairie.

“It’s one of the largest intact, native grasslands left in the United States and one of the few left in the world,” said Kelly, who met Twidwell at a conference in the early 2010s and immediately recognized a kindred spirit.

“He reached out just to make sure that the research he was thinking about was applicable,” Kelly said, estimating that the task force communicates with about 40 ranchers per year. “If a professor hasn’t been a rancher, they don’t always know how the information will be received or what parts are most useful. That’s the role that I feel like I play: to try to help answer those questions.

“And when we start talking about these ideas, more ideas always come up. It’s just a brainstorm to try to share information and kick ideas around. The neat thing about Dirac is (that) he takes that information back and acts on it.”

Uden and Twidwell described the group’s collaboration with agencies, land managers and landowners as a mutual learning process in which data can guide practice, which then provides essential context to that data, and so on.

“It really is a co-production,” Uden said. “We can see these general indicators of change, but then we go to them to say, ‘What happened here at this time?’ And they say, ‘Oh, that’s where this has been happening,’ or, ‘That’s where I did this, and my neighbor did that.’”

“It’s exciting from the science end,” Twidwell said, “but then we can respond rapidly to their needs so they can learn, as well. We often hear, ‘Huh, that’s what was preventing me from having the same issue as my neighbor.’”

The group also partners with the Nebraska Game and Parks Commission, whose wildlife division administrator, Alicia Hardin, helps oversee the management of both public and private land. Hardin said the screening approach should help the commission better prioritize limited funds by focusing on areas where the eastern redcedar has yet to encroach but is showing signs of invading. Preventing that encroachment, she said, is both easier and less expensive than removing the invasive species after it has laid down roots.

The potential of longer lead times and more precise guidance is already making an impression at the commission, Hardin said.

“It really opened the eyes of some of our biologists, and they had some really good conversations,” she said. “What can we do management-wise? Where should we really focus? How can we work with the private landowners around us to get some of that work done?

“I’m really excited, and I think it’s a great step forward for all of us.”

But preserving the Sandhills and Loess Canyons may also require a broader shift from reactive to proactive thinking on par with the emerging leaps in technology and ecology, Twidwell said. The fact that the consequences to ecosystems often lag behind the shifts themselves can muffle alarming signs, he said, belying their power to upend the status quo on issues as disparate as water management and school funding.

And because previous generations of Nebraskans could mostly afford not to understand, think about or prepare for regime shifts, abruptly adopting a proactive mentality can be a big ask.

“That becomes a real challenge; it’s the same as talking about why preventative medicine is very difficult,” Twidwell said. “It comes down to your choice and your cultural will to take care of yourself. And if it’s that hard with medicine, you can imagine how hard it is with managing natural resources.”

Kelly said she’s optimistic that communicating the generational importance of that mentality with those who ultimately have the power to implement it — the ranchers themselves — is giving the state a fighting chance.

“Dirac’s ability to speak with ranchers is top-notch, and that doesn’t come by accident,” she said. “He has taken the time to understand the culture and to understand how he can be most effective when communicating with them. And it does take time, and it takes a willingness to learn.

“He’s a linchpin to this war we’re waging, and people like him are what’s going to make the difference on if we win or lose that war. So many states to our south have lost it, and I would love to see Nebraska keep some of the grasslands intact — especially the Sandhills, because that’s where my heart is.”

Forecasting forecasts

Though the research group remains locally focused, it’s also thinking about how the screening approach could assist that fight elsewhere. The ability to zoom out from local to community, statewide and even national scales should make it useful in addressing the array of issues that stem from regime shifts, Twidwell said.

“We can provide people information on how the system is operating in ways that are maybe different than how we think about our system,” he said. “So: Here’s how the American burying beetle — an endangered species — is interpreting change in the system, versus a game species that’s important for hunting groups, versus how hydrology is functioning, versus how forage and the beef industry should be looking at this. All of these are associated with regime shifts, but they’re all responding at different levels of the signals. We haven’t been able to give people that kind of complexity before.”

That complexity contrasts with the relative simplicity of the screening approach itself, Uden said. Whereas prior tools for assessing the presence and threat of a regime shift have often demanded expertise in the specifics of a given ecosystem’s history, the team’s approach instead aims to start generally and home in on particulars from there.

“A lot of previous work is based on a deep knowledge of the system and its feedbacks — exactly how it’s functioning,” Uden said. “What we can see is just a signal of change — that something is happening — which we then pass on to the next stage of analysis, which would be a more traditional process of going out, talking to an expert and seeing what’s actually going on. But we can detect that signal without really any knowledge of what’s going on in that particular system. That’s a very fundamental difference.”

If Twidwell has his way, the science of forecasting regime shifts will eventually become as routine as meteorology is today.

“In the future, you’ll have these platforms where, in these native ecosystems, you can just track change like the weather,” he said. “We haven’t had those before, but it’s fundamentally the same thing.

“But making science truly useful is a heavy lift. If done right, the technology can so effectively deliver science and information into the hands of people that the scientific process often becomes taken for granted. Hopefully this research can become taken for granted one day.”

---

Ecology