3

Framing the Value of Collaborative and Multidisciplinary Research

As the Panel examined the study Statement of Task, two phrases—“enhance the success and impact of inter-institutional activities” and “on national needs and global food security”—stood out as a higher-level context for the goals of study. Ronnie Green, a member of the Panel, offered the view that “Congress wants us [the land-grants] to tackle the really big questions.”

The Panel recognizes that having greater impact—the power to create change—depends on being able to use cutting-edge capabilities from across the scientific and engineering disciplines, but that assembling the intellectual and organizational capacity to exploit those tools for real-world applications requires teamwork, often at a systems-oriented, cross-sectoral level. Funding is also necessary to support such work. This chapter discusses these three ingredients in the context of collaboration in the land-grant system.

SCIENTIFIC CONVERGENCE

The land-grant system has an opportunity to embrace and capitalize on the advances and advantages made possible by multidisciplinary approaches to problem solving, as described in a 2009 report from the National Research Council called A New Biology for the 21st Century (NRC, 2009). The A New Biology report was prepared in response to a request to examine the current state of biological research in the United States and recommend “how best to capitalize on recent technological and scientific advances that have allowed biologists to integrate biological research findings, collect and interpret vastly increased amounts of data, and predict the behavior of complex biological systems” (NRC, 2009). The report was among the first to recognize that the “essence” of the “new biology” is the reintegration of the many subdisciplines of biology as well as the integration of biology with physics, chemistry, computer science, engineering, and mathematics. This integration would enable the creation of a research community with the capability to tackle a broad range of scientific and societal problems, including those associated with food, environment, energy, and health. However, the report’s authoring committee also pointed out that fundamental questions in all areas of biology, from understanding the brain to carbon cycling in the ocean, would become more tractable as the new biology grows into a reality that can contribute to advances across the life sciences. The report received considerable attention and was part of the foundation for a new vision of biological research.

However, although the idea of integrating biology with other disciplines had great appeal across the various scientific communities, it was not necessarily straightforward to adopt this new approach. In 2014, the A New Biology report was followed by another National Research Council report called Convergence: Facilitating Transdisciplinary Integration of Life Sciences, Physical Sciences, Engineering, and Beyond. The report defines convergence as

The purpose of this second study was to address the challenges to creating and sustaining environments that foster convergence, to better understand these challenges, and to explore examples of ongoing successful convergence programs to inform investigators and organizations interested in expanding or establishing their own efforts. The chair of the authoring committee, Joseph DeSimone, noted,

APPLICATIONS IN AGRICULTURE

How do these ideas and developments relate to the land-grant system? The nature of key questions in food and agricultural science is evolving, and the scientific approaches to address them are increasingly at the convergence of multiple disciplines, use information collected dynamically across multiple scales or geographies, and require advanced data science capability. Research, teaching, and extension that use a systems perspective supported with data science expertise and capability are needed to address the multifaceted problems now facing the agricultural and food systems. For example, while traditional research and extension may have focused primarily on improving crop yields, current questions are more broadly framed, for example, on how to improve crop yields and nutritional benefits in a changing climate and/or without environmental degradation.

Science Breakthroughs to Advance Food and Agricultural Research by 2030 (NASEM, 2019) notes that agriculture faces serious problems today that are unlike those in the past. The natural systems on which agriculture depends are showing serious signs of stress, including water scarcity, increasingly variable weather, floods, and droughts brought on by climate change, land use change, population growth impacts, and other factors. Productivity growth for staple crops is stagnating or declining worldwide, a “warning sign” that new methods will be required to address the continuing need for increased productivity. Neither is the U.S. food supply immune to the impacts of pandemics, conflicts and wars, foodborne illness and the threat of pests and pathogens to crops, livestock, and poultry. U.S. farmers and producers need more tools to manage the pressures they face.

The Science Breakthroughs report defined the major goals for food and agricultural research in the next decade to include “(1) improving the efficiency of food and agricultural systems, (2) increasing the sustainability of agriculture, and (3) increasing the resiliency of agricultural systems to adapt to rapid changes and extreme conditions” (NASEM, 2019, p. 2).

Underpinning these goals are major research challenges. How can we improve nutrient use efficiency in crop production systems? How will we reduce soil loss and degradation, mobilize genetic diversity for crop improvement, optimize water use in agriculture, improve food animal genetics, develop precision livestock production systems, improve nutritional value of agricultural products, detect and prevent plant and animal diseases and foodborne pathogens, and reduce food loss and waste throughout the supply chain?

The Science Breakthroughs report suggests that some answers lie in finding ways to employ recent scientific and technical advances that offer new opportunities to carry out systems-level approaches in food and agriculture research. Chapter 9 of the Science Breakthroughs report, Strategy for 2030, describes how developments in computing, information science, machine learning, materials science and electronics, genomics and gene editing, and behavioral and cognitive science have broad application in agriculture. (Box 3-1 lists these crosscutting “breakthrough” tools.) For example, advances in molecular biology now allow for the development of new food sources and traits to increase resistance to insect pests and diseases; increase nutritional components; increase yields, and nutrient and water use efficiencies; and increase the ability to withstand weather extremes.

Microelectronics and nanotechnology provide improved capabilities to sense and monitor physical, chemical, or biological properties and processes to improve food production sustainability; enable and automate controlled environments for production; control microbes to improve food safety and minimize food waste; and create new materials to monitor and improve animal health. Breakthroughs in robotics and drones do not just enable research but provide producers with tools with which to assess and address field conditions and apply pesticides more precisely, thus reducing environmental impacts.

Artificial Intelligence Research Institutes

Data science, including machine learning and artificial intelligence (AI), is an essential capability for successfully addressing the most important opportunities and challenges in food and agriculture. The Science Breakthroughs report explicitly noted the pivotal role of advanced data analytics. In 2020, the National Institute of Food and Agriculture (NIFA) and the National Science Foundation contributed funds to establish seven new collaborative AI institutes to accelerate research, expand America’s workforce, and transform society in the future in terms of food and agriculture.¹ In a mini-workshop (see Appendix B) organized by the Panel, participants from two of the institutes explained their work.

One is the Artificial Intelligence for Future Agricultural Resilience, Management, and Sustainability Institute, led by the University of Illinois at Urbana-Champaign with partners at Tuskegee University, Michigan State University, the Donald Danforth Plant Science Center, the U.S. Department of Agriculture’s (USDA’s) Agricultural Research Service, and the Argonne National Laboratory. The partners are using shared AI tools to explore projects that range from the use of robotics for small-scale producers to “farm of the future” modeling for resilience and sustainability. The second, the AI Institute for Food Systems, based at the University of California (UC), Davis, with partners from UC Berkeley, Cornell University, and the University of Illinois, is pursuing an ambitious array of AI-supported projects around agricultural production, nutrition, food processing, molecular breeding, and importantly, the development of AI specific to the challenges of food and agriculture. Both projects have strong teaching, research, and extension elements.

___________________

¹ See https://www.nifa.usda.gov/about-nifa/press-releases/usda-nifa-nsf-invest-220m-artificial-intelligenceresearch-institutes, accessed September 20, 2022.

The keynote presenter at the workshop, Tom Andriola of UC Irvine, focused on the flexibility and power of these collaborative platforms. In addition to creating an opportunity to provide computing and data science tools for use by all participants, the investment in these AI institutes can serve as a catalyst to design and implement a scaling strategy and plan for data science that is inclusive, and maximizes the full potential of all land-grant colleges and universities in terms of enhanced collaboration. However, data science will only fulfill its potential if all institutions have an opportunity to contribute and have access to well-curated and integrated data assets on the food and agricultural system (see Box 3-2).

CREATING SUCCESSFUL COLLABORATIVE TEAMS

The land-grant system, with its nationwide remit, seems an ideal place to activate the recommendations in the Science Breakthroughs report. Many of the approaches focused on systems-level discovery will require regional or national cooperation and planning in addition to multidisciplinary efforts. If successful, systems approaches will produce essential information that will be the basis of new knowledge to inform decision making at various scales, along with new tools to implement those decisions. The broader impacts of collaboration are realized at the point when science catalyzes real change. However, getting to that point involves the complex endeavor of many people working together toward a common goal in a sustained way.

One of the respondents to the Panel’s preliminary observations remarked on a research group at Colorado State University where investigators study interdisciplinary scientific teams, in real time, to advance a predictive theory of what makes teams successful. Subsequently, the Panel asked Jennifer Cross, a leader in this field, to give a presentation on the “Science of Team Science.” Cross explained that she and her colleagues shadow a team and evaluate the connections between its members over time, creating graphic maps illustrating the frequency and quality of contacts between team members. Periodic surveys of participants capture the nature of relationships by asking participants such questions as which colleague they met

with to discuss the project, who they sought out for advice, who they had a beer with, and who they trust. The maps show who connects most frequently.

Cross said that the science of team science shows that individuals destined to lead projects need to have—or develop—a collaborative mindset, openness to learning from others, the capacity to consider divergent perspectives, and interpersonal relationship and communication skills. These values, mindset, and interpersonal skills are the foundation needed for the more complex tasks of team science, particularly as it grows to engage partners in other sectors (Cross et al., 2022; Hall et al., 2018, 2019).

In addition to the capabilities and characteristics of individuals, the success of a team depends on building the capabilities of the team. Team science competencies include several domains: trusted and genuine relationships, communication, collaborative knowledge generation, collective problem solving, team management, and team co-leadership. Doing participatory team science requires commitment and patience because it takes time to build relationships and trust, create a shared vision, and overcome the many barriers to participatory research. According to Cross, to make progress in a collaboration, there is no shortcut around building trust and a shared vision, and she has the data to prove it. Cross and her colleagues monitor the growth and dynamics of teams as new people join and leave and assess the coherence of the members of the collaborations by using an individual’s stated satisfaction with the project, including the feeling that he/she is contributing and the sense that members share a common vision.

Food systems research includes a diverse array of scientific fields—soil and crop science, animal and meat science, economics, public health, nutrition, engineering, ecology, and rural sociology—and stakeholders who will require the team to fully understand and appreciate the context of their situation. This diversity in scientists and stakeholders requires that some team members cultivate the ability to be “boundary crossers” who help translate science to community members and scientists from other fields as well as understand the diverse perspectives of others. That may not be the most scientifically productive individual but the one on which the success of the collaboration depends. One of the primary tensions to be resolved is that between scientists seeking the gold standard of scientific research versus team members who see the value of feasible research that can meet stakeholder expectations and timelines. Researchers who strive to use perfect data and frontier scientific methods may not be a good fit at the center of a team and may be best engaged as peripheral members.

Cross found that large transdisciplinary teams benefit from participating in evaluation and training activities designed to enhance their capacity to address challenges and foster self-awareness about best practices for working in teams. “Developmental evaluation” is an approach to team development and assessment focused on continuous improvement by supporting team adaptation under dynamic and evolving conditions in real time. Her work is popular at Colorado State University, where science teams actively seek resources to support her work with them.

The Panel is aware that participants face barriers that affect their ability to fully participate in collaborations, and Cross and her colleagues suggest that the results of those stresses can be captured by their research method. This scientific approach to building self-aware multidisciplinary teams is helpful in illuminating the way forward.

CAPACITY FUNDING OF LAND-GRANT COLLEGES AND UNIVERSITIES

Funding is, and will be, essential to conducting inter-institutional collaboration in the land-grant system. The federal government provides money to the land-grant colleges and universities in the states according to a complex set of legislative authorities and formulas to support the system’s tripartite mission of agricultural research, education, and extension. NIFA distributes funds to the states as capacity grants, sometimes called formula funds, in amounts based on the proportion of rural and farm population in the state or, relative to the 1994s, based on numbers of Native American students.

Separate from the capacity grants, NIFA also awards funds on a competitive basis through the Agriculture and Food Research Initiative (AFRI) to the highest-quality proposals submitted by a large pool of eligible institutions and organizations that include many non-land-grant entities. AFRI grants cover foundational and applied projects, including small and large collaborative projects, such as the Coordinated

Agricultural Projects. Congress has approved funding of $445 million for AFRI for fiscal year (FY) 2022 (CRS, 2022a). Success rates for obtaining an AFRI grant vary with each specific program and range from 10 to 40 percent (NIFA, personal communication, September 15, 2022).

The legislative authorities that provide funds to the land-grant system are specific to each category of land-grant institution. The Hatch Act of 1887 and the Smith-Lever Act of 1914 fund research and extension, respectively, at the 1862 institutions.² In addition to other provisions, these funds must be matched 1:1 by the states with nonfederal funds, and 25 percent of the capacity portion of each of the funding lines must be used for collaboration with one or more states.

The Evans-Allen Act of 1977 and the National Agricultural Research, Extension, and Teaching Policy Act (NARETPA) of 1977 fund research and extension activities, respectively, at the 1890 institutions. Unlike the Hatch Act and Smith-Lever Act funds, there is no required set-aside for collaboration. However, these funds must also be matched 1:1 with nonfederal funds. If 100 percent of matching funds is not found, the 1890 institutions can apply to NIFA for a waiver of up to 50 percent of the matching requirement. Although a waiver allows the university to receive the federal funds, it reduces the overall amount it would have received had the full match been obtained. The 1890s have had to leave money on the table every year since Congress began to require 100 percent matching funds in 2007. In 2020, 10 of the 19 1890 institutions did not meet the 100 percent match; as a result, they collectively “lost” $21 million that otherwise would have been realized (CRS, 2022b; see Table 3-1).

The fact that, in the same state, 1862 institutions with primarily White students have been able to obtain matching funds from state legislatures while 1890 institutions with primarily Black students have not is a striking and disturbing inequity. Tennessee withheld nearly $544 million in matching payments over several decades to its sole 1890 institution, Tennessee State University, a situation the legislature is now attempting to redress. In 2021, Maryland settled a lawsuit for $577 million with the state’s four Historically Black Colleges and Universities (HBCUs), one of which is an 1890 institution, claiming the state “persistently underfunded” the HBCUs while supporting programs at other institutions, which had the effect of pulling students away from the HBCUs (NEA, 2022).

The consequences of this inequity are apparent in many ways. For example, in a 2018 study, the U.S. Government Accountability Office (GAO) examined capital project needs at HBCUs. The HBCUs responding to GAO’s survey reported that 46 percent of their building space, on average, needs repair or replacement; GAO identified capital project needs at nine HBCUs that it visited in the areas of deferred maintenance, facilities modernization, and preservation of historic buildings (GAO, 2018). The U.S. Department

___________________

² A portion of Hatch funds supports the State Agricultural Experiment Stations associated with the 1862 institutions. Smith-Lever funds support the partnership of the 1862s in Cooperative Extension, a partnership of the land-grant system with federal, state, and local agencies to bring knowledge to practitioners.

of Education’s HBCU Capital Financing Program provides access to needed funding designed to help address capital project needs for some HBCUs and has helped modernize their facilities to improve student recruitment, but fewer than half of HBCUs have used the program, according to U.S. Department of Education data.

The situation is even worse for the 1994 institutions. Slightly more than 70 percent of Tribal College and University (TCU) funding comes from the federal government. Because of the government-to-government relationship between the tribes and the federal government, states have no obligation to provide funding to the TCUs and most do not (Nelson and Frye, 2016). In addition, under the Tribally Controlled Colleges and Universities Assistance Act of 1978, the federal government can authorize funding streams for the TCUs, but what has actually been appropriated has been only about half of authorized amounts.

The Tribal College Endowment Program provides 1994 institutions with discretionary funds to support agriculture and the mechanical arts with interest on the 1994 Institutions Endowment Fund, distributed annually according to a formula related to the number of Native American students. The Tribal College Research Grant Program provides grants for research in cooperation with specified types of institutions (including 1862 and 1890 institutions). The Tribal College Extension Grants Program supports 1994 institution extension activities.

Another perspective on funding across the land-grant system is to view it relative to the number of institutions in each category and the number of students they serve. Table 3-2 provides this information, showing that on a per institution basis, the 1862 institutions receive more capacity funds for research than the 1890 institutions and vastly more than the 1994 institutions. However, they also collectively serve many more students.

It is also worth noting that because the allocation of Hatch and Evans-Allen capacity funds are based on the farming and rural population of states, smaller 1862 institutions, such as those in the northeastern United States where the number of farms is lower, also find funding levels insufficient to support their research and extension activities (per comment on the Panel’s preliminary observations).

THE LARGER FUNDING PICTURE

Inequities notwithstanding, there is a larger backdrop to the picture of resources for the work of the land-grant system and its potential to engage in greater collaboration, and that is the fact that public-sector (federal and state) funding for agricultural research (adjusted for inflation) has been in decline for two decades. According to researchers at USDA’s Economic Research Service, expenditures peaked in 2002 at $7.64 billion (in 2019 dollars) and by 2019 were $5.15 billion, about the same level as in 1970 (Nelson and Fuglie, 2022). This downward trend puts pressure on the entire land-grant system, in spite of private-sector investments in agriculture and food of more than double that of the public. It has been acknowledged that private-sector funds are not a substitute for public support as they focus on profit generating activities, whereas public-sector funds address public goods that arise from food and agriculture, including economic returns (CRS, 2022a). Numerous economists have estimated that public funding of agricultural research provides a minimum return of 20:1 for each dollar invested (Nelson and Fuglie, 2022).