3

The Need for a Dynamic Soil Information System

A panel discussion on the question “Why do we need a dynamic soil information system?” followed the keynote presentations. Moderated by Bruno Basso from Michigan State University, the panel consisted of David Babson from the U.S. Department of Energy’s (DOE’s) Advanced Research Projects Agency–Energy (ARPA–E), Jim Dobrowolski from the National Institute of Food and Agriculture (NIFA) at the U.S. Department of Agriculture (USDA), Matt Kane from the National Science Foundation (NSF), David Lindbo from USDA’s Natural Resources Conservation Service (NRCS), John Mesko from the National Corn Growers Association, and Stephen Wood from The Nature Conservancy. Each panelist offered a short statement and then the panel discussed the issue, taking questions from the moderator, each other, and audience members.

PRESENTATIONS

David Babson
Advanced Research Projects Agency–Energy

Babson began his presentation by explaining that ARPA–E is DOE’s advanced projects funding agency. “We’re kind of our moonshot research funding agency,” he said. “We tend to fund really high-risk, high-reward types of endeavors.” Funded projects include those on resilient energy infrastructure, affordable sustainable energy, U.S. economic development, leadership in science and technology, and—of particular relevance to the workshop—efforts to address climate change through mitigation.

The agency is pursuing programs related to carbon farming and using soils for carbon management because “all of the paths to 2 degrees of warming go through zero.” That is, to avoid the worst consequences of climate change, the increase in average global temperatures must not exceed 2°C, he said, but “our ability to just reduce emissions even down to zero and still avoid more than 2 degrees of temperature rise passed us by in the ’90s. We did not do enough.” Thus, at this point, achieving this goal will require reducing net global

emissions of carbon dioxide below zero (see Figure 3-1). “So we need to build a very large negative-emissions industry very quickly to be able to give us options to maintain our path to 2 degrees.”

Agricultural ecosystems can play a large role in carbon dioxide removal, Babson said, with significant reductions through the implementation of best practices and the potential to become net negative in emissions through broad implementation of cutting-edge technologies. Thus, ARPA−E is funding new technologies that can help make the agricultural sector as carbon-negative as possible, and it is actively exploring new opportunities for enhancing terrestrial carbon reduction potential. However, many of these opportunities are rendered ineffective without accurate and accessible soil data.

One example is the Systems for Monitoring and Analytics for Renewable Transportation Fuels from Agricultural Resources and Management (SMARTFARM) program. SMARTFARM aims to develop highly scalable soil measurement systems that can be used to inform market incentives for improved carbon management, with an emphasis on nitrous oxide as the primary driver of net positive emissions and soil carbon as the potential driver of net negative emissions. The research team has received phase 1 funding to capture high-resolution soil data and make the data available to the public. “As such,” Babson said, “ARPA−E is keenly interested in the discussions and outputs of this workshop, so that we can help to ensure the phase 1 data are of the greatest utility to the R&D [research and development] community and private-sector stakeholders.”

In phase 2 of the SMARTFARM program, researchers will develop the next generation of data collection and analysis tools for agricultural carbon accounting; the planned approaches will include perimeter and drone-based nitrous oxide monitoring, in situ sensors for soil carbon, and highly scalable model-based approaches to derive a net emissions estimate. For these researchers, Babson said, a harmonized soil information system would

provide a framework within which to capture the data and would make those data available for use by a broader audience.

Finally, he said, being able to obtain better measurements of various soil parameters will open up new possibilities for soil technologies, such as new methods for fixing carbon and nitrogen in soil and novel, scalable means of measurements. “We are interested in funding even more technologies in this space to really open up the possibilities for new strategies to do carbon management ecosystem services,” he said, noting that ARPA−E is soliciting proposals for “all kinds of new technologies that would help us achieve our aims in this space.”

Jim Dobrowolski
U.S. Department of Agriculture’s National Institute of Food and Agriculture

Describing his organization, Dobrowolski said that “We’re a small agency with a big budget.” Currently, NIFA has 248 employees and a budget of approximately $1.7 billion. About 40 percent of that amount is directed to land grant institutions to support risky and long-term research as well as extension, outreach, and education. The remainder, the competitive section, funds discovery and applied research focused on agricultural production, quality, and sustainability.

The study of soils remains an important part of the NIFA portfolio, Dobrowolski said. NIFA has funded more than 1,183 soil sustainability projects focused on soil erosion, nutrient management, and microbial activity, and it will continue to fund soil research, particularly through the Signals in the Soil program, a collaboration with NSF.

NIFA awards come with a requirement for data management, with an emphasis on connecting with existing inventories or networks. Efforts to improve connectivity include training the next generation of scientists in soil science and management, developing minimum standards and methods for data collection and integration of datasets, and creating plans for long-term data management, storage, and sharing. Funding also supports linkages with publicly accessible databases for collection information, tool development sampling methods, and data curation plans.

However, NIFA has often struggled with directing researchers to the right data repository, he said. The potential first stop is the scientific data services of USDA’s National Agricultural Library, which offers data management policy and planning, repository management, data and metadata curation and consultation, and preservation, and is serviced by the Ag Data Commons. Researchers can submit or link their data to the Ag Data Commons in order to meet the findability, accessibility, interoperability, and reusability data requirements of journals or the public access requirements of funders such as NIFA. However, at the moment, Dobrowolski added, domain and related informatics expertise is limited to the biological sciences, geospatial and biophysical sciences, genomics, federal open data policy, and life cycle assessment.

In January 2020, the White House Office of Science and Technology Policy issued a call for public comments on a draft set of desirable characteristics for data repositories used to locate, manage, share, and use data that result from federally funded research. This call offers an opportunity for workshop participants to make a difference, he said. “By participating fully, we can improve the current systems in place for widely monitoring soils—physical, chemical, and biological—and we can better understand, document, and manage the effects of land use and cover changes on soils.”

In closing, Dobrowolski said that creating a dynamic soil information system will be challenging. Developing and implementing a dataset that encompasses the chemical, physi-

cal, and biological attributes of soil within the context of environmental and land use conditions with a network of suppliers is a big lift. This work will require careful identification of known and innovative sampling methods with suitable metrics and attributes that are focused on the appropriate users. It will also require locating and assembling as much of the available data as possible. “NIFA has funded the collection of millions of soils-related data points,” he said, “but where are they now? I can’t answer that.” NIFA’s data are spread out over many different data repositories, and it needs to determine what data already exist and where they are housed in order to identify new national and regional priorities, synthesize research, and decrease the potential for duplicate studies. “This is the reason that we need the dynamic soil information system.”

Matt Kane
National Science Foundation

Kane began his presentation with a brief description of NSF, noting that it differs from other agencies because it does not have intramural laboratories. Instead, its main function is to review and fund science and engineering proposals. Concerning research into soils, he said, several NSF directorates, particularly the biological sciences and geosciences directorates, fund such research and would benefit from a soil information system. Other beneficiaries would be programs in environmental engineering and various cross-directorate activities that include the social sciences or computational science and engineering.

Much of NSF’s funding goes to programs that align with 1 of its “10 big ideas” developed under previous director France Córdova. Four are related to the idea of a dynamic soil information system: harnessing data for 21st-century science and engineering, understanding the rules of life, navigating the new Arctic, and growing convergent research. Each of these activities is an “umbrella activity” that involves several directorates and program areas.

In additional, several NSF core and special program areas would benefit from a soil information system, Kane said (see Figure 3-2). One example is the Signals in the Soil program, which is headquartered in the engineering directorate but also involves program officers in other directorates. Other relevant programs are in the geosciences directorate, such as geobiology, low-temperature geochemistry, and geomorphology and land use dynamics, and in the biology directorate, such as plant genome research and ecosystem science.

However, perhaps the most relevant, he said, are the more long-term programs, such as long-term ecological research sites, macro system biology, and particularly the National Ecological Observatory Network (NEON). Completed 2 years ago, NEON is the largest single investment ever made by the biology directorate. With a goal of enabling regional- to continental-scale research in biology and other areas, it produces 180 different data products from about 80 sites around the country and 21 different geographical regions. The data are freely open and available, because one of NEON’s goals is to democratize and standardize ecological research.

Other NSF programs provide open environmental data as well, Kane said, including the Long-Term Ecological Research network, the Critical Zone Observatories program, and Integrated Digitized Biocollections. “A major interest of NSF going forward,” he said, “is enabling the scientific community to make full use of all of this open environmental data to address questions, to identify questions, to really understand the earth and its biota as never before.”

Of relevance to dynamic soil information systems is a currently open competition to create a Center for Advancement and Synthesis of Open Environmental Data and Sciences.

The initiative rests on four pillars, Kane said. The first pillar is that the center should be an incubator for team science that analyzes and synthesizes open environmental data, which is where a dynamic soil information system would play a role as a form of open environmental data. The other three pillars are a creative and innovative cyber infrastructure, inclusive community engagement, and data science training. The total 5-year budget for the center will be $20 million. These pillars reflect NSF’s interests going forward, Kane said, and a dynamic soil information system could be one part of this overall collection of open environmental data that will support future science and problem solving.

David Lindbo
U.S. Department of Agriculture’s Natural Resources Conservation Service

Lindbo opened his talk with a vision for the future. “Wouldn’t it be great,” he asked, “if we could get soil information on a 10-meter grid with properties, interpret it for land use, use real-time water and climate information—soil moisture, water table depth, irrigation needs—forecast the effects of conservation practices, … and then determine the effective practices for desired land management goals … and if we could have all of that in one place?”

One role of NRCS is to provide producers, landowners, and land managers with the information they need to manage their land more effectively, Lindbo said, and to that end, NRCS is developing a dynamic soil survey with five parts. The first part is a collection of soil maps, which will include NRCS’s traditional vector-based products as well as raster-based products with information on soil properties at varying depths. The second part is connection of the soil data to ecological sites where NRCS has collected data about plants,

the ecological community, and movement from one ecological state to another in a given landscape. The third part is information on dynamic soil properties, which vary over space, time, and management. “These are the soil health properties,” Lindbo said, “but they could be more than that as well, as we start to think about urban effects and other land use effects.” The fourth part is climate and hydrological data. These four parts are combined, using both static data and temporal data, to produce a dynamic soil survey. The fifth part, he said, is the people who collect, synthesize, and curate the information that populates the survey so that “it continues to grow [and] it continues to produce information that is of value to more and more people.”

To produce this dynamic soil survey, Lindbo said, NRCS will need assistance: “We need field data, we need scientists, and we need your information.” A lot of data that would be useful in the survey have already been collected, he said, but NRCS does not always have access to those data. Even with access, the best way to incorporate those data into a unified system that will be the dynamic soil survey is not always clear. Thus, NRCS welcomes input from the workshop participants.

In conclusion, Lindbo said that the dynamic soil survey represents the future for NRCS. It plans to use that survey in multiple ways—for resource management, soil health, conservation planning, and various initiatives. “Whether they be climate related, water quality related, or urban agriculture related, they’re all going to need that soil information so the decisions can be made properly,” he said, “and from this workshop we hope to be able to see what’s out there, start talking to people, and move us forward to the future.”

John Mesko
National Corn Growers Association

The Soil Health Partnership, Mesko said, is a program of the National Corn Growers Association. It operates in 16 states and has more than 200 sites on working farms where it tests and evaluates the effects of various soil health−changing practices on the soils themselves and on the economics of the farm. It is unusual in the combination of information that it offers, he said, including soil data, management data, and economic data. The partnership receives less than 10 percent of its funding from the National Corn Growers Association, with the rest coming from various types of agencies.

One of the partnership’s goals is to help mitigate climate change, he said, and its approach is based on the understanding that, if agriculture is to play a role in achieving that goal, farmers must change the way they farm. The few farmers that are already making some of these changes—planting cover crops, reducing tillage, and managing nutrients—are early adopters, Mesko said: “These folks are probably going to do it with or without the involvement of a program like ours.” They generally do not require incentive payments to change their practices, although they will certainly accept one if offered; rather, they are motivated by an interest in exploring new technologies and ways to improve their farming.

Thus, the Soil Health Partnership has been trying to reach a larger group of farmers whom Mesko referred to as “middle adopters.” These farmers are aware of new practices—either from speaking with other farmers or reading about them in magazines or brochures—but they have not yet adopted some of the practices that could have the greatest effects on climate change, such as planting cover crops or reducing tillage. “We have 170 million acres of commodity crops produced in the United States, and, depending on who you ask, we’re somewhere around 10 to 14 million acres of cover crops currently,” he said, “even though cover crops are probably the hottest topic in agriculture right now.”

The partnership has developed various tools to help farmers understand the value of changing their practices. First, it provides data. For example, it has 7 years of data to help answer the question “How does planting a cover crop affect soil indicators and yield?” The partnership also helps with peer-to-peer networking so that farmers can talk with many other farmers about their experiences. This interaction is very important, Mesko said, because the single greatest barrier to adoption of new practices is the uncertainty that comes from variability. “Each farm is different, each farmer is different, they have different sets of equipment, [and] they have different access to, for example, cover crops seed or fertilizer, whatever the case may be.” Therefore, it often takes time to determine exactly which changes will be most effective. “Our data show that early in the change cycle for farmers there is typically either an increase in costs or, in some cases, a decrease in yield while farmers are going through this learning curve,” he said. “If a farmer has never planted a cover crop before, the first time out it often doesn’t go well. It has a lot to do with timing and seed application rate and the technology that they’re using, and does it fit into what they’re already doing on a year-in, year-out basis.”

Thus, he said, the Soil Health Partnership would appreciate access to a dynamic soil information database. However, he added, such a database will not be enough to convince large numbers of farmers to change their practices. “We have to be able to match that up with the management data that that farmer is using, whether it’s planting date or the amount that they spend on their cover crops seed or the change in equipment that they are using as they change their practices so that we can help them understand what the impact, up or down, is on their bottom line, as they make these changes.”

In closing, Mesko explained that adoption of new practices often represents a very large change for farmers—even one as seemingly simple as adding a cover crop to the crop rotation. Therefore, the Soil Health Partnership is focused on data not only about soils but also about farm management—the information that will help farmers make good decisions for both the environment and their bottom line. Unfortunately, he said, it has proved difficult to assemble that sort of information because the farmers believe that information belongs to them. “It’s their business choices, it’s their purchasing choices that they make, and without a strong connection with an organization, they’re not that interested in giving that up.”

Stephen Wood
The Nature Conservancy

The Nature Conservancy, Wood noted, is best known for its work in land management and land protection, particularly in the United States, where its chapters in all 50 states acquire and protect properties. “And that is still absolutely central to who we are as an organization,” he said. However, he added, in the years since its founding in the mid-1950s, the organization has increasingly realized that the largest environmental challenges are not only local, but also regional, national, and global, and therefore cannot be solved simply by protecting individual properties. This led The Nature Conservancy to integrate working lands into its vision of conservation and protection and to prioritize soil as a core part of its working land strategies, which in turn led to new approaches to its work.

Wood described The Nature Conservancy’s strategy for its soil and agriculture work in North America as an “impact strategy;” it integrates its state-by-state and chapter-by-chapter work into a broader vision with the goal of achieving a broader impact than originally envisioned. When the organization formalized this strategy 5 or 6 years ago, it set an ambitious target of having 50 percent of the corn–soy acres in the United States in soil health practices

by 2025. The target likely will not be achieved, he said, “but the purpose of setting such an ambitious target was to get us to think about what types of new partnerships need to exist and come together in order to have impact at that scale.” The Nature Conservancy decided on a combination of approaches, which includes collaborating with the private sector, influencing public policy, identifying and demonstrating methods that can scale the adoption of useful practices, working with farm advisors, and offering new financial incentives.

Interestingly, Wood said, The Nature Conservancy’s strategy excludes an element that was central to much of the workshop’s discussions—empirical soil measurements. The rationale is that The Nature Conservancy does not regularly take direct soil samples to determine, for example, whether soil carbon is changing over time as a result of its activities. “Part of that is because the scale that we’re working at is quite broad and it’s not a field-by-field type strategy,” he said, “and so that has put us in a situation where what we need are systems that allow us to assess the adoption of practices at broad scales and to assess and evaluate how soil might be changing at broad scales over time.” In short, he said, The Nature Conservancy could benefit from the type of dynamic soil survey that Lindbo described in an earlier presentation about NRCS.

In concluding, Wood said that equally important to better understanding changes in soil properties over time will be converting that knowledge into greater insights about agronomic and environmental outcomes and, in turn, developing more effective strategies for promoting desirable regional, national, and even global outcomes.

DISCUSSION

Basso opened the discussion session with a question for Mesko, via Slack, from Jenny Soong from Granular: are the data from the Soil Health Partnership available in a data repository, and, if not, when will they be available? Mesko answered that the partnership is working with the University of Minnesota’s G.E.M.S.™ (Genetic, Environmental, Management, and Socioeconomic data) platform to house the data in one place and make them available for sharing with other partnering organizations. The data are actually shareable now, he said, but they are not easily available. Making the data easily accessible is one of the partnership’s major areas of emphasis, he said.

Alfred Hartemink from the University of Wisconsin–Madison asked whether the multiple parallel efforts to create soil information systems present a problem. Lindbo responded that parallel efforts are a good thing. “[For] most of the parallel efforts that I’m aware of, we talk to each other, and we share information,” he said. Information can always be tweaked in different ways and used for new purposes, so the different efforts bring different things to the table. Mesko added that the different efforts typically have different goals that lead them to use the data in different ways. When the Soil Health Partnership started 7 years ago, he said, it received significant investment from the supply chain community, which was interested in understanding the effects of changes in practices on soils and on agriculture. Since then, however, many of those supply chain partners have developed their own versions of what the partnership is doing, which may not be as robust as the partnership’s database but meet their needs. Charles Rice added that these parallel efforts often collect data at different spatial and temporal scales—farmers may be interested in data at the scale of acres, while NRCS may be more interested in a national scale—which can complicate integration of data into a dynamic information system. The opportunity, he said, is to combine these different sources of information at different scales into a more dynamic information system—dynamic not only in time, but also in the sense of data coming in at different times of year or different decades. Combining data from different sources, Rice said, connected to a question

on Slack about citizen science, which presents another opportunity to improve the datasets’ robustness as long as the data quality can be ensured.

In response to a Slack question from Julian Kremers from CQuest about the role of standardization in sampling and in compiling the data in databases and soil information systems, Lindbo said that standardization is essential for ensuring meaningful comparisons between samples. “But the critical part,” he added, “is that [since] methodologies change over time, we have to be able to look backwards and make those comparisons as well.” Fenny van Egmond from the International Soil Reference and Information Centre added via Slack that the Food and Agriculture Organization’s Global Soil Partnership is developing standards for soil data exchange, working to standardize sampling practices, and building GLOSIS (Global Soil Information System) using Open Geospatial Consortium Standards.

Kathe Todd-Brown commented that standards are not static and must be regularly updated. “I think oftentimes there’s the tendency to just do a one and done,” she said, “but as new methods evolve and new measurements start coming online, we need to have some way to extend and revise the standards.” Matt Kane added that standardization was one of the reasons why NSF developed NEON. “We have 81 sites that are collecting soil samples in an identical way under very strict quality control, storing them in a single biorepository, now available for a variety of different analyses,” he said. As a result, measurements can be standardized over a continental scale.

Stephen Ogle from Colorado State University asked about challenges beyond collecting data. Specifically, he asked what some of the challenges are to using data to help farmers make decisions about their practices. Mesko replied that farmers consider many factors. They live and work in a community of people, and a major change in practice might require a change in business partners or retailers. The issue is much more complex than simply telling a farmer that by planting a cover crop in 3 years they will save X dollars and get Y greater yield. “If you really want to change the hearts and minds of farmers, we have to do all of this data, we have to do the incentive payments, we have to do the peer-to-peer network, but we also have to make it very cool to be a farmer that farms in this way,” he said. Right now, farmers remain focused on practices that maximize yields. However, he added, the next generation of farmers will likely have studied topics such as environmental science in school and be more open to studying the research and changing their practices.

Babson added that action likely to motivate farmers to adopt more sustainable practices and new technologies is connecting those farm practices to new types of carbon markets. A farmer who is interested in earning additional money from these markets will need highly accurate data and models in order to understand which practices to adopt. Referencing the earlier question about parallel efforts, he noted that carbon markets are different sizes:

Referring to Mesko’s comment about the importance of making sustainable farming seem “cool” to farmers, Babson said that if sustainable farming is shown to be profitable, then farmers will adopt it.

Erika Foster from Purdue University asked the panelists via Slack what role the training of students and early career scientists plays in the establishment of a dynamic soil informa-

tion system. Dobrowolski responded that students should be provided with some insights into the best ways to translate information from a soil information system to farmers and other users on the ground. “It’s not the same as it used to be, with delivering pamphlets or having traditional field days or get-togethers where you have a large group of farmers who are coming to listen to you speak about the newest technologies,” he said. “We need to focus on sociology and psychology and trust building so that we can promote behavior change and adoption.” In short, the individuals who are working with farmers must be not only familiar with and able to explain the science, but also able to communicate effectively and build relationships because many farmers only listen to people they trust. In addition, the training needed to communicate and build trust should start early in researchers’ careers. Dobrowolski said that often graduates later become extension specialists or extension faculty, but they have had little training in translating scientific information for people on the ground. Lindbo added that students earning bachelor’s degrees in soils, agronomy, or environmental science should also learn more about computers, programming, and geographical information systems.

Phil Robertson from Michigan State University asked whether farmers’ attitudes about healthy soils have been changing, particularly those who are not the early adopters, and about the role of retailers in affecting this change (or lack of change). Mesko responded that retailers are in business to sell products, so they have no interest, for instance, in promoting a practice that will require fewer chemicals. Nor are retailers motivated or equipped with the personnel to help farmers adopt a new practice unless it involves the farmer buying a product from them, “and, frankly, there’s not a lot of retailers who sell cover crop seed or strip till equipment,” he said. Mesko has observed that, because conservation practices are not part of their business models, retailers are seeking partnerships with entities such as NRCS or soil and water conservation districts that can provide technical assistance related to environmentally sustainable practices that farmers are interested in. However, he believes that awareness of soil health and the practices that promote it has not reached the tipping point. Without the help from retailers, whom farmers look to for help in making farming decisions, “it’s really hard to get over that hump,” he said.

Alfred Hartemink asked Stephen Wood to describe the benefits of a dynamic soil information system for The Nature Conservancy. Wood responded that the organization would benefit from the insights gained about soil knowledge and from projections of regional changes in aspects such as water quality, water availability, and greenhouse gas cycling.

Brian Darby from the University of North Dakota asked the panelists via Slack whether any data repository or information system outside of soil science serves as a successful model worth imitating. Matt Kane replied that the ocean sciences has a potential model, a centralized database called the Biological & Chemical Oceanography Data Management Office (BCO-DMO) with biological, chemical, and physical data for ocean science. However, he was not aware of any model, whether it be BCO-DMO or GenBank (the data repository for genetic sequences), that a soil information system should follow. Data related to soils are heterogeneous, and “it’s a huge challenge in understanding the soil environment,” he said. Dobrowolski agreed, pointing to the Consortium of Universities for the Advancement of Hydrologic Science (known by the acronym CUAHSI) as another example of a database for, in this case, hydrology data, but one that is limited relative to the number and types of data that it can receive. Via Slack, other participants suggested Resource Watch, Global Biodiversity Information Facility, Living Atlas of Australia, and Avian Knowledge Network.