The National Science Foundation (NSF) today acknowledged that it bit off more than it could chew when it agreed in 2010 to build a unique network of dozens of ecological stations across the United States. Facing cost overruns and construction delays, NSF officials have decided to reduce the scope of the troubled National Ecological Observatories Network (NEON) and eliminate a major aquatic research component.
NSF recently discovered that the $433 million project, which was scheduled to be completed next year, “was delayed and projected to be approximately $80 million over budget if it stayed on its current trajectory,” says James Olds, head of NSF’s biology directorate. After consulting with NEON officials and outside scientists, Olds says NSF “identified a descope option that will keep the project scientifically transformational and should bring it in on time and on budget.”
The move to shrink NEON follows years of complaints from scientists that NSF and project management have been inflexible and that the community has been shut out of the decision-making process. And all sides acknowledge that NEON was terra incognito. A report in February by a top-level advisory body noted that “it is important to remember that the ecological research community has no experience with a project of this scale.” The report also tried to revive the flagging spirits of researchers who may have lost interest in the project because of what it called the “long time period of design and construction without any data flow.”
NSF’s announcement is a blow to those who plan to use NEON. “It’s terrible news,” says ecologist Walter Dodds of Kansas State University in Manhattan, who championed STREON, NEON’s now-canceled aquatic experimentation component. But NEON Inc., the Boulder-based nonprofit that manages the project, vowed in a statement that “the project will remain positioned to meet the goal of transforming continental scale ecology,” adding that large NSF-funded projects “commonly require adjustments of scope and in this regard NEON is not exceptional.”
Ecology's big data moment
The changes are the latest twist for a project first proposed 15 years ago—not by the ecology community, but by then-NSF Director Rita Colwell. NEON was seen as ecology’s entry into the world of big data. In lieu of the traditional approach of having scientists monitor their own tiny slice of the world, using their own methods and instruments, NEON would standardize the process—and collect data on a massive, continental scale.
After several false starts, researchers and NSF coalesced around a final plan for NEON. For some 30 years, scientists would collect a continuous stream of information from towers and sensors installed at sites within 20 ecological domains, including tropical forests, wetlands, high desert prairies, and even urban ecosystems. Data from “core” terrestrial and aquatic sites within each domain would be supplemented by information from two “relocatable” sites, using equipment that could be moved every decade or so. Three planes would fly over the domains periodically during the growing season to record changing vegetation patterns. The data would be accessible to all, allowing scientists to assemble a continental-scale picture of climate change, land-use trends, and the movement of invasive species.
Construction began in 2011 and was supposed to be completed by the end of 2016. But the path for project managers was never smooth. Some of the problems were of their own making, including high staff turnover and conflicts caused by relegating scientists to what they saw as a secondary role. The permitting process has turned out to be a nightmare, and there were also persistent technical challenges. Establishing sites outside the contiguous United States has been especially problematic; Hawaii and Alaska pose unique environmental challenges, and NEON abandoned an urban site in Puerto Rico earlier this year after two guards were killed.
NSF recently concluded that NEON was running a year behind schedule, a time frame “that is not acceptable,” Olds says. To get back on track, Olds says NEON will keep core sites in all 20 domains, but reduce the number of relocatable sites. As a result, the planned initial cohort of 60 sites could be reduced to 50, says one scientist familiar with the project.
STREON's choppy waters
The biggest change is dropping STREON. Unlike NEON’s other components, which gather information about the environment, STREON was designed to intentionally alter stream ecosystems—by adding nutrients, simulating extreme weather conditions, and removing top predators—and then document how they responded. But obtaining the permits needed to conduct such experiments proved difficult. “To do a STREON experiment, you need a significant reach of stream, and access to it for 30 to 40 years,” Olds explains. “You’re asking for permission to put chemicals into the water for a very long time, and any single owner can veto it.”
Olds insists that NSF has not abandoned STREON. “We are very interested in seeing the experiment go forward,” he says. “It simply will go forward in a context other than the construction of NEON.”
Aquatic scientists are skeptical. It will be difficult to run STREON independently, they say, because it relies on NEON sites as controls. They also see NSF’s decision as part of a larger pattern of systemic neglect of NEON’s aquatic components. This past June, Dodds and 18 other researchers wrote to NEON and NSF, noting that construction of aquatic sites was lagging far behind terrestrial sites, and urging them to close the gap by shifting resources. The number of STREON sites had already been cut in half from the original 20, they noted.
NEON officials rejected the idea, stating “we cannot make one component of the observatory a higher priority than others.” They blamed “permitting, site science requirements, and procurements” for delays, noting that “for various reasons, these obstacles have presented greater challenges on the aquatics side than on the terrestrial side.”
Dodds says the decision to cast STREON adrift means that NEON will be disproportionately focused on terrestrial sites. He says it also runs counter to a 2003 report by the National Academies that helped NEON win congressional support. The report noted the importance of supplementing NEON’s observational data with experimental results. “That’s what the community felt was important,” Dodds says. “And now there won’t be any experimental elements.”
Details to come
NEON officials still need to work out descoping details, Olds says. But today’s move is designed to “strengthen NSF’s oversight” of NEON Inc., he noted, adding that the agency expects the group “to work robustly with stakeholders and ensure that the science will best serve the evolving needs of the research communities it is designed to serve."
In its statement, Neon Inc. said: "Science has been and will continue to be the foundation of NEON ... We remain unwavering in our commitment to scientific integrity and fulfilling the mission of NEON using the best available science. We are no less excited about NEON's potential to contribute to essential ecological research for decades to come."
The descoping, Olds noted, won’t affect NSF’s plans to spend $65 million a year over the life of the project to operate facilities and make data available to researchers. “We have every intention of maintaining that amount,” he says. “It’s in our budget.”