Achieving success with RISE: A widely implementable, iterative, structured process for mastering interdisciplinary team science collaborations

Abstract

Scientific experts from different disciplines often struggle to mesh their specialized perspectives into the shared mindset that is needed to address difficult and persistent environmental, ecological, and societal problems. Many traditional graduate programs provide excellent research and technical skill training. However, these programs often do not teach a systematic way to learn team skills, nor do they offer a protocol for identifying and tackling increasingly integrated interdisciplinary (among disciplines) and transdisciplinary (among researchers and stakeholders) questions. As a result, professionals trained in traditional graduate programs (e.g., current graduate students and employed practitioners) may not have all of the collaborative skills needed to advance solutions to difficult scientific problems. In the present article, we illustrate a tractable, widely implementable structured process called RISE that accelerates the development of these missing skills. The RISE process (Route to Identifying, learning, and practicing interdisciplinary and transdisciplinary team Skills to address difficult Environmental problems) can be used by diverse teams as a tool for research, professional interactions, or training. RISE helps professionals with different expertise learn from each other by repeatedly asking team-developed questions that are tested using an interactive quantitative tool (e.g., agent-based models, machine learning, case studies) applied to a shared problem framework and data set. Outputs from the quantitative tool are then discussed and interpreted as a team, considering all team members’ perspectives, disciplines, and expertise. After this synthesis, RISE is repeated with new questions that the team jointly identified in earlier data interpretation discussions. As a result, individual perspectives, originally informed by disciplinary training, are complemented by a shared understanding of team function and elevated interdisciplinary knowledge.

An unresolved challenge in the environmental sciences that has important implications for policymaking and resource management is how to transform individual disciplinary experts into effective members of interdisciplinary (among disciplines) and transdisciplinary (among researchers and stakeholders) teams (e.g., Cheruvelil et al. 2014, Caldas et al. 2019, Martin 2019, Wallen et al. 2019). Many difficult environmental problems require collaboration among professionals with diverse expertise (figure 1a). However, there are gaps in current university-based, research-driven graduate programs, which train students for specific environmental and related science professions (e.g., ecology, fish and wildlife biology, agriculture, forestry, geography, hydrology, sociology, economics, statistics, engineering; Alberti et al. 2011, Read et al. 2016, Oliver et al. 2018, Gosselin et al. 2020). Specifically, the curriculum of many master’s and PhD programs provide limited opportunities for training in teamwork skills; learning structured processes for collaboration; in-depth cross-discipline investigations; translation of concepts across disciplines; and learning how to conceptualize integrated interdisciplinary and transdisciplinary questions that will achieve team goals (figure 1b). In the present article, our purpose is to illustrate our widely implementable, structured process and practical guidelines. These actions will accelerate interdisciplinary and transdisciplinary team skill development through systematic interactions with a quantitative tool that enhances the team’s ability to formulate sophisticated questions related to difficult scientific problems (figure 1c). Our process can be used by diverse teams as a tool for research, professional interactions, or education. Achieving successful interdisciplinary and transdisciplinary collaborations is a multifaceted challenge, but, in the present article, our focus is only on illustrating our process.

Figure 1.

An overview of the ideas presented in this concept and question article in a nutshell. Included are (a) needs, (b) current barriers, and (c) bridges across barriers that the RISE process provides. Abbreviation: RISE, Route to Identifying, learning, and practicing interdisciplinary and transdisciplinary team Skills to address difficult Environmental problems.

Interdisciplinary and transdisciplinary science

Interdisciplinary and transdisciplinary science can provide insights into difficult and persistent environmental problems (Gropp 2017) that disciplinary approaches cannot (Clark 2007, Carpenter et al. 2009, Stuart et al. 2015, Pischke et al. 2017). Two common interdisciplinary and transdisciplinary conceptual frameworks are coupled human and natural systems (CHANS) and social–ecological systems (SES). The CHANS framework is focused on the dynamics among human and natural systems related to global environmental change (e.g., Liu et al. 2007, Shindler et al. 2017). The CHANS framework has been used to address environmental issues as diverse as agricultural nutrient inputs (Stuart et al. 2015), human and wildlife conflicts (Beck et al. 2019), and geographically distinct nature reserves (Liu 2017). The SES framework is conceptualized as a coevolutionary dynamic system of actors, institutions, and resources shaped by a social–ecological setting (e.g., Schlüter et al. 2014, Blair et al. 2017). The SES framework has been used to address environmental problems as varied as understanding the wildlife trade (Blair et al. 2017) and cooperation and resilience in the commons (Folke et al. 2010, Schlüter et al. 2016). Regardless of which conceptual framework is employed, progress in interdisciplinary and transdisciplinary collaborations is crucial for such complex issues and critically depends on team effectiveness.

RISE meets existing needs

In the present article, we illustrate our iterative, multistep, structured process, which we refer to hereafter as RISE (Route to Identifying, learning, and practicing interdisciplinary and transdisciplinary team Skills to address difficult Environmental problems). RISE is adaptable to both interdisciplinary and transdisciplinary problems. By using RISE, graduate students, early career scientists, and established professionals with different expertise learn from each other by repeatedly asking, quantitatively testing, interpreting, discussing, and synthesizing responses to jointly developed interdisciplinary and transdisciplinary questions. We illustrate the flexibility of RISE using three diverse teams: interdisciplinary scientists, environmental practitioners, and the participants in public meetings at which environmental professionals and the public interact.

Team science

The need to effectively work in groups is so widespread that a new research area has emerged to address how teams work: the science of team science (SciTS; Stokols et al. 2008a, National Research Council 2015). SciTS researchers seek a better understanding of what makes teams successful (e.g., Stokols et al. 2008b, Salazar et al. 2012, Begg et al. 2014, National Research Council 2014, 2015, Fiore et al. 2018, Bisbey et al. 2021, Zaggl and Pottbacker 2021). SciTS topics include the value of team science, team formation, team composition, team communication, team function, and institutional complexities (e.g., Pennington 2011, Hall et al. 2018, Gosselin et al. 2020, Rolland et al. 2021b). Team formation and composition research addresses participants’ identity, previous experiences, and motivations to join a team science project (Sievanen et al. 2012, Martin 2019). Studies of team processes include how the team learns to work together (Pennington 2011, Pennington et al. 2013). Team function research relates to psychological safety, awareness, exchange, self-correction, and adaptation (Morton et al. 2015, Bisbey et al. 2021). The institutional setting can influence all team activities from team formation to processes (Herz et al. 2020, Salazar and Claudel 2021). Conceptual models have been proposed for SciTS (for a review, see Hall et al. 2012). These models are valuable for guiding this emerging field, but they often lack the practical, process-oriented steps that would make them relevant to real world science problems. As examples of these conceptual frameworks, Pennington (2016) distinguished between the individual and the group, focusing on individual learning as team science progresses. Other SciTS researchers examine how integrative capacity can be built and refined through interactions (Salazar et al. 2012). Specifically, the CFIR (Consolidated Framework for Implementation Research) and the RE-AIM (Reach, Effectiveness, Adoption, Implementation and Maintenance) frameworks provide templates for implementation (Rolland et al. 2021a). RISE builds on the existing SciTS foundation and provides an avenue for the practical implementation of SciTS conceptual models.

What novel contribution does RISE make to the science of team science?

In addition to acquiring interpersonal team skills, environmental science professionals face a trifaceted challenge. Specifically, trained disciplinary experts need to understand the definitions, paradigms, and methods of other disciplines (Donovan et al. 2015); to translate the key issues in their own disciplines to other disciplines (Eigenbrode et al. 2007, Schnapp et al. 2012, Crowell 2023); and to identify questions and related goals that guide interdisciplinary progress. We provide two examples to illustrate how this additional trifaceted challenge adds complexity for environmental science teams. As a first example related to the difficulty of cross-discipline learning and translation (facets 1 and 2), no one is surprised that four individuals who speak different languages (e.g., English, French, Spanish, Russian) cannot communicate well, but we assume that a geomorphologist, an ecologist, a sociologist, and an economist (who use different definitions, terminology, and scientific paradigms) can understand each other without interdisciplinary and transdisciplinary translation. As a second example related to the difficulty that interdisciplinary scientists face in making progress (facet 3), the goals that guide progress for many teams are clear at the outset (e.g., increase sales, renovate a house, care for a patient), but a geomorphologist, ecologist, sociologist, and economist often cannot even set meaningful interdisciplinary and transdisciplinary goals unless they first have a shared conceptual foundation. RISE adds to the science of team science by providing a structured approach that helps individuals better understand other disciplines, by translating ideas across disciplines, and by asking then answering increasingly integrated questions that move the team toward interdisciplinary and transdisciplinary goals.

Our test system

We illustrate the RISE process using examples of biologically involved sustainability in agricultural watersheds. RISE can be used for many scientific team problems, but, in the present article, we focus on environmental issues. Environmental problems are ideal test beds for developing a structured process that advances collaboration (Kelly et al. 2019, Burr et al. 2021). Biologically involved sustainability links human and natural systems through allocation decisions related to land (natural, cultivated, municipal), water (quantity, quality, timing), cultivated organisms (crops, domesticated stock), and other nonhuman biota (native, stocked, invasive organisms). Biologically-involved sustainability problems are distinct from purely physical or engineering issues (e.g., waste and energy management). Relevant human stakeholders in biologically involved sustainability problems include farmers, ranchers, residents (rural or urban), and recreationists (consumptive and nonconsumptive). To address interdisciplinary and transdisciplinary problems related to biologically involved sustainability, our example teams used an enriched interdisciplinary data set that included existing social science (geography, demography, sociology, economics) and biological–physical (hydrology, ecology) data collected in the same geographic location. Our learn-by-doing process (RISE) provides opportunities for individuals to develop interdisciplinary and transdisciplinary team skills such as critical thinking; cross-disciplinary and shared understanding of issues, paradigms, and methods; quantitative skills; and formulating relevant, meaningful, and integrated research questions. Interpersonal skills gained include developing a shared vision, empathy and respect for differing views, and leadership abilities. Other skills targeted by our process include consensus building and effective interdisciplinary team communication (Caldas et al. 2019, Farrell et al. 2021). All of these skills are critical elements of team science. Providing opportunities for scientists to engage in interdisciplinary and transdisciplinary team science formally (i.e., graduate school seminar) or informally (workshops, online training) is essential (Hall et al. 2018, Gosselin et al. 2020, Zaggl and Pottbacker 2021). RISE facilitates these opportunities to learn how to collaborate.

An analogy that illustrates common interdisciplinary and transdisciplinary challenges

We develop the RISE process using a graphical mountain climbing analogy (figure 2) in which arrows and footprints represent the route a scientist will traverse to become a member of a successful interdisciplinary and transdisciplinary team. Completing a specialized graduate degree and obtaining professional employment in a specific discipline requires the mastery of discipline-specific paradigms, theory, and methods. Therefore, an early career priority is to successfully summit the disciplinary mountain range (footsteps 1–3; figure 2a; in this example, become experts in hydrology, ecology, or economics).

Figure 2.

An illustration of a route to creating successful interdisciplinary and transdisciplinary teams. This route uses a mountain climber analogy. The footsteps represent the route a scientist will traverse to become a member of a successful interdisciplinary team. The arrows indicate the direction of the footsteps. In panel (a), early career scientists become disciplinary experts in the disciplinary mountain range (footsteps 1–3; Mt. Hydrology, Mt. Ecology, or Mt. Economics). The stars and dotted line in panel (a) indicate the peaks of the disciplinary mountain range or the highest levels of knowledge gained through disciplinary science. In the transitional stage depicted in panel (b), scientists make the decision to become interdisciplinary, which may result in a period of lower productivity while new interdisciplinary knowledge and skills are learned (steps 4–5). (b1) If transitional scientists seek to gain interdisciplinary skills through a process of trial and error (an unproductive and overly complex venture), they may fail and become discouraged (steps 6–7). However, if they use a step-by-step, structured process such as RISE (b2), they will accumulate knowledge related to complex environmental problems systematically (steps 8–14). In panel (b), the stars (steps) are placed identically in panels (b1) and (b2), but in panel (b2; the RISE process), the six steps per cycle are linked systematically. In panel (c), disciplinary experts successfully summit Mt. Interdisciplinary (footsteps 15). The star in panel (c) indicates the peak of Mt. Interdisciplinary. The horizontal dotted line in panel (c) represents the highest level of knowledge achieved by submitting Mt. Interdisciplinary. We suggest that RISE—panel (b2)—provides faster and better results than trial and error—panel (b1). See figure 3 and the related text for a description of the diagram that represents the RISE process (figure 2b2). We use this mountain climber analogy to share our experiences of the challenges of interdisciplinary projects to professionals who may want to work across disciplines but find that the constraints of their jobs and training create resistance. When professionals (university, agency) are trained as disciplinary experts and are accustomed to the methods and accomplishments of a disciplinary science career, the altered productivity trajectory for interdisciplinary teams can come as an unpleasant surprise. Our experience is that, in the face of unrealistic expectations about the speed of progress, interdisciplinary teams can become frustrated and even quit. We hope to encourage effective interdisciplinary team work by using this mountain climber analogy to point out the value of interdisciplinary studies, the utility of a structured approach, and also to provide realistic expectations. Abbreviation: RISE, Route to Identifying, learning, and practicing interdisciplinary and transdisciplinary team Skills to address difficult Environmental problems.

To address more difficult and pressing environmental questions (Kelly et al. 2019), disciplinary experts may choose to work across disciplines. Just as one cannot successfully climb Mt. Everest without substantial planning and specific skill training, a disciplinary scientist will also need to prepare and gather skills to transition into a member of a successful interdisciplinary and transdisciplinary team (Henson et al. 2020). Initial stages of interdisciplinary collaboration are often characterized by decreased productivity as new knowledge and skills are learned (steps 4–5; figure 2a). The short life cycle of a grant can also limit interdisciplinary and transdisciplinary collaboration if individuals prioritize publishing in disciplinary domains in which they have experience. To create an effective interdisciplinary and transdisciplinary team, individuals who possess the appropriate disciplinary skills will transition from lone experts that direct independent laboratories (atop Mt. Disciplinary) to members of interacting teams who are willing to invest time learning paradigms and methods from other disciplines. Neither mountain climbing nor learning interdisciplinary science team skills are one-and-done enterprises; they require immersive training and repetitive practice for which some attempts will succeed and others will not. The climb up Mt. Interdisciplinary can be a frustrating trajectory in which the knowledge gained does not necessarily increase linearly with the time invested (figure 2b1, 2b2). Interdisciplinary and transdisciplinary projects with a new team require a significant (and often time consuming) investment in learning a shared terminology, creating a joint vision, becoming comfortable working with a team, and developing relevant and meaningful interdisciplinary questions. If transitional scientists seek to gain interdisciplinary skills through a trial-and-error process, they may fail and become discouraged (steps 6–7; figure 2b1). Just as an established route serves to guide climbers up a challenging mountain face, the RISE process transforms a series of trial-and-error actions (figure 2b1) into a structured process (steps 8–14; figure 2b2), in which scientists can successfully summit Mt. Interdisciplinary (step 15, figure 2c).

RISE: The route to interdisciplinary and transdisciplinary success

In this section, we briefly illustrate the general model for the RISE process (figure 3). Additional details on all steps are provided in the sections that describe the specific teams.

Figure 3.

A detailed illustration of the RISE process that enhances learning, practicing, and polishing of interdisciplinary and transdisciplinary team skills. RISE is a series of (a) six steps that are repeated and combined into an (b) upward iterative spiral. Each loop includes the same six steps. Step 6 in a previous loop links to step 1 in the subsequent loop to ensure continuity and progress. These six steps and iterations (or levels) in the upward spiral are discussed in detail in the text. Abbreviation: RISE, Route to Identifying, learning, and practicing interdisciplinary and transdisciplinary team Skills to address difficult Environmental problems.

Multiple steps and iterative loops in the RISE process

Six steps form a single iteration (i.e., cycle loop, level; figure 3a) of the RISE process, and the same six steps are repeated in subsequent iterations (figure 3b). Connections between the last step of one iteration and the initial steps of the next iteration create an upward spiral of continual improvement in team skills, better communication, progressive interdisciplinary integration through increasingly sophisticated questions, and greater accumulated knowledge about the problem of interest to the team. A team leader initiates step 1 and oversees the completion of all steps. A technical expert oversees the construction and operation of the quantitative tool in step 3. Otherwise, all team members contribute to all steps in all iterations.

Step 1 is to assemble an appropriate team that includes members with different expertise who share a commitment to addressing a specific environmental problem (figure 3a). For instance, biologically involved sustainability teams working within agricultural watersheds often share an interest in implementing agricultural practices that balance landowner productivity, public health, and environmental protection (Stuart et al. 2015, Read et al. 2016).

Step 2 is to create a roadmap that identifies a desirable team direction (figure 3a). Identifying and implementing a common vision requires that all team members contribute to a dynamic, multidimensional, jigsaw puzzle frame (representing the complex problem being tackled) into which they fit relevant, modified pieces of their disciplinary expertise. Working across disciplines, as with other examples of scientific problem-solving, benefits from an interactive understanding of both the parts and the whole (Sterling et al. 2010). Examples of a focused team purpose for biologically involved sustainability problems include creating interdisciplinary and transdisciplinary science for sustainability decision-making (example team 1), advising agricultural stakeholders of the best sustainable farming practices in the face of climate change (example team 2), and discussing site-specific conservation issues at a public meeting in which scientists and nonscientists have different levels of expertise and varying interests (example team 3).

Step 3 is to develop and apply a quantitative tool to a shared interdisciplinary and transdisciplinary data set to allow the team to repeatedly ask relevant questions, then interactively discuss, and interpret rapidly generated results (figure 3a). The quantitative tool will dialog with the framework developed and allow for the team members to externalize their disciplinary approach (Pennington 2015, 2016). Interactivity in the RISE process is a specialized type of feedback (Gabelica et al. 2016) that forces all team members to focus on the same question, data, outcomes, and interpretations. Examples of interactive quantitative tools include agent-based models, digital twins (Bauer et al. 2021), machine learning (Reichstein et al. 2019), virtual test beds, other modeling techniques, other statistical approaches, and case studies. In the present article, we choose to focus on use of an agent-based model as one means to create a constantly updated, shared reality on which the team can focus their diverse perspectives as they transform from disciplinary scientists into interdisciplinary and transdisciplinary team members. By using an agent-based model, the team members can discuss and provide feedback on the system’s behavior and the agents’ interactions. We provide additional details of the agent-based model tool below (example team 1) and elsewhere (Granco et al. 2022).

Step 4 is to run interactive virtual experiments. In our team model run examples, increasingly sophisticated interdisciplinary and transdisciplinary questions, deliberately chosen by the entire team, collaboratively identify model inputs and outputs (figure 3a). In initial model runs, teams typically ask simpler questions and work as a loosely linked aggregation of disciplinary experts. By repeatedly interacting with each other, teams identify the critical connections and influential feedback loops among disciplines that differentiate among disciplinary or multidisciplinary (most straightforward), first-generation interdisciplinary (more complex), and next-generation transdisciplinary (the most complex) questions (Salazar et al. 2012).

Step 5 is a structured group discussion of the model outputs (figure 3a). This structured discussion format offers team members the opportunity to reflect on previous iterations (Gabelica et al. 2014, von Wehrden et al. 2019), including how each model run links to goals and questions, what information was gained, and what is missing that could be tested in future model runs. Critical parts of this step are team reflection on previous iterations, feedback among team members, hypothesis testing, model validation, assessing robustness, and analyzing sensitivity (Gabelica et al. 2012, Gabelica et al. 2014, Gabelica et al. 2016, Gosselin et al. 2020).

Step 6 is to create an adaptive record that documents goals, progress, changes in direction, future questions, and next-iteration model runs (figure 3a). Ideally, each team will start each loop in the upward spiral (initial steps, iteration 2) with a review of the adaptive record compiled in the previous loop (final steps, iteration 1). The result is a constantly evolving, connected, always updated record of progress (figure 3b). The upward spiral is not infinite but continues until the team agrees that goals related to team training, knowledge accumulation, or other beneficial outcomes have been met.

RISE is a structured, integrated, and repetitive learn-by-doing process. As a team completes iterations of the six-step process and moves up the spiral, each team member has many opportunities to interact with the same team on the same complex problem. Practice within and across individual cycles in this upward spiral improves each individual's ability to interact with other members of the team, to translate concepts across disciplines, and then to learn to ask increasingly sophisticated interdisciplinary questions. A sequence of 6–12 iterations, undertaken by the same team at weekly or monthly meetings, may be required to master interdisciplinary and transdisciplinary skills needed for environmental problems. Repeating step 1 (team selection) and step 3 (model development) may not be needed in every iteration. A consistent and persistent shared purpose and set of goals (step 2) can emerge across iterations as the team establishes an esprit de corps, so this step may not be needed in every iteration either. However, steps 4–6 will be essential components of each iterative loop.

Examples of how different types of teams can benefit from RISE

To illustrate the generality and value of RISE, we show below how our process can be applied to three example teams. Example team 1 is a research collaboration that is often sited at universities (Alberti et al. 2011). Example team 2 is an applied professional collaboration for which experts employed by varied agencies (state and federal) and universities come together to address a specific environmental management concern (Stuart et al. 2015). Example team 3 is a transdisciplinary public meeting, sited outside academic institutions (Plummer et al. 2022), at which scientists, environmental practitioners, and nonscientific stakeholders articulate their views on a specific environmental concern or weigh in on a specific regulatory decision.

All of the examples reflect real-world teams and are compilations of real teams with whom the coauthors have interacted. Questions and data output for example team 1 mirror the authors’ experiences as part of an interdisciplinary project through the National Science Foundation’s Coupled Natural and Human Systems program (NSF CNH 1,313,815). Teams of mixed practitioners from universities and agencies (state and federal) on which the authors have served commonly assemble to address specific land, water, biota, and impact decisions within and across jurisdictions. The questions and outputs for example team 2 are based on issues in the state of Kansas, a strong agricultural region, where farmers and ranchers grapple with climate change. Public meetings frequently occur when scientists, agency professionals, and the public are assembled to address a specific natural resource decision or environmental regulation. Questions and iterations of example team 3 are based on public meetings we have attended. We describe ways that all example teams could use the same agent-based model (Granco et al. 2022) that was used by our CNH team (example team 1). To demonstrate the RISE process’ adaptability, different questions, inputs, outputs, and structured discussions are used for each team (steps 4 and 5) who are likely to draw different conclusions (step 6).

Example team 1: Interdisciplinary research team

In example team 1, an interdisciplinary team of university research scientists, with different expertise, willingly collaborated on a common environmental problem. As real-world examples, interdisciplinary research teams strive to understand human and natural ecosystem links (e.g., the National Science Foundation’s Dynamics of Socio-Ecological Systems), pursue place-based understanding and prediction (e.g., the National Science Foundation’s Long Term Ecological Research Program), and address general scientific concepts across scales (e.g., the National Science Foundation’s Macrosystems Biology Program).

Step 1: Assembling appropriate teams

For this first example, the interdisciplinary team consisted of ecologists, hydrologists, geomorphologists, economists, geographers, sociologists, and engineers with expertise in biodiversity (fish), water (hydrology), land use (agricultural landscapes), economic drivers (focused on agriculture and municipal interests), social science (attitudes and values of agricultural watershed stakeholders), and integrated systems modeling (agent-based simulation and optimization). An advantage of this type of team is that research scientists often see mutual benefits in collaborating and frequently are enthusiastic about expanding funding opportunities and real-world impact by undertaking interdisciplinary and transdisciplinary research. As a potential obstacle, specialized disciplinary silos can hinder the identification of connected interdisciplinary and transdisciplinary questions among disciplinary experts (Mather et al. 2021).

Step 2: Creating a shared purpose and goals

The shared purpose of example team 1 was to understand and predict integrated biological–physical–social science patterns and processes related to decision-making. For example team 1, the overarching goal was to identify whether, when, and why stakeholders who reside within an agricultural watershed would pay for and adopt a specific conservation policy under different climate scenarios

Step 3: Quantitative tool

We parameterized an agent-based model (figure 4) with data relevant to biologically involved sustainability questions in an agricultural watershed (Granco et al. 2022). Agent-based models can integrate data across space and time for diverse human agents (Grimm et al. 2005). In our model, the natural system is divided into three submodels (hydrological system, ecological system, and land-use system), which are affected by current and future climate scenarios (figure 4). The human system is represented by two submodels (cultural system and land-use system) that combine demographics and social values. The dynamic connection among these submodels is the agents’ perceptions of environmental change, which influence the agent's decision and are mediated by a value–belief–norm framework (Granco et al. 2019, Granco et al. 2022). Model inputs of interest include land use, agricultural crop choice, ecological state (i.e., species richness of fish, plants, and birds), and hydrological conditions (i.e., stream flow). Model outputs of interest include physical resource status (i.e., summer stream flow), biological resource status (i.e., change of fish richness), human agent perception of environmental degradation (i.e., updated values–belief–norm), level of support for the environmental policy (i.e., individual vote and total votes for the policy), and policy costs. The structure of the agent-based model formulation is intricate (Granco et al. 2022); however, in RISE, the role of the model is straightforward. The agent-based model acts as a quantitative tool that rapidly integrates actual, cross-discipline data from a shared location and, in so doing, repeatedly allows the team to formulate shared questions, test their hypotheses with model runs, and interpret data-driven model output as a group.

Figure 4.

Agent-based model developed for investigating public support for a new environmental policy in the Smoky Hill River Watershed, in Kansas, in the United States. Inputs for the model are divided by their submodels, for instance, submodel climate has precipitation as an input. Inputs are processed by the agents at each cell representing the local environment at multiple locations of the watershed. There are 900 agents (N = 900) that represent voting stakeholders (farmers and nonfarmers), and each agent occupies one cell. Agents observe the environmental conditions at their local environment and act on their individual decision rule to decide their vote for policy (yes or no). Outputs are gathered at each iteration of the model at agent (i.e., individual vote) and system level (i.e., environmental degradation). Feedback occurs at an annual time step. The model runs for 30 time steps, representing the time from 1986 to 2015. The model, adapted from Granco and colleagues (2022), is discussed in detail in the text.

Steps 4–6: Questions, inputs, outputs, and structured discussions

Example team 1 members posed questions that increased in sophistication and connectivity with each iterative loop (table 1). Because this first example was modeled on team interactions that the authors experienced, we provide more detail in the question-driven iterations (i.e., six iterations for example team 1 rather than the three iterations provided for the two subsequent example teams). In the initial loops (iterations 1 and 2), example team 1 asked and answered simpler questions about specific disciplinary patterns. Disciplinary or multidisciplinary questions that example team 1 used in iteration 1 described patterns of fish, water, land use, and social science within the shared Smoky Hill River study area (iteration 1, table 1). The questions used in iteration 2 addressed which disciplinary patterns were of common interest for future interdisciplinary and transdisciplinary discussions (iteration 2, table 1). The resulting structured discussions (step 5) and adaptive records (step 6) for loops 1 and 2 identified the need for a quantitative tool that integrated data and ideas across disciplines. In the third loop, example team 1 asked first-generation interdisciplinary questions by modeling interactions among physical, biotic, and social science data (iteration 3, table 1). In this third loop, the structured discussion addressed the assumptions, realism, and utility of the model results (step 5) in a way that increased team member appreciation of other disciplines and advanced team skills (step 6). In advanced iterations (4–6), example team 1 asked integrated next-generation (next-gen) transdisciplinary questions related to stakeholder willingness to pay for a conservation policy (iteration 4, table 1), whether this willingness to support conservation tracked biophysical trends (iteration 5, table 1), and whether different stakeholder groups varied in their support for conservation (iteration 6, table 1).

Table 1.

Examples of multiple iterations of questions, model inputs and outputs, structured discussion topics, lessons learned, and next steps for example team 1: researchers.

Iteration	Stage	Questions	Model input	Model output	Structured discussion topics	Lessons learned	Next steps
1	Multidisciplinary (preparing the team)	What disciplinary patterns and drivers exist across the shared study area?	No project-specific model	Not applicable	What patterns and processes are of interest within disciplines?	Better understanding of other disciplines	Continued practice of the structured team process
2		What are the common patterns of interest across disciplines?	No project-specific model	Not applicable	How can we use an agent-based model to ask relevant what-if questions?	A need exists for a quantitative tool that integrates disciplines	Identify model structure, inputs, outputs
3	First-generation interdisciplinary	How does land use, climate, river flow, and biodiversity affect human perceptions of use?	Biophysical variables (land use, climate, river flow, biodiversity)	Stakeholder (agent) attitudes and values	What interactions among biophysical and social variables are important?	Disciplinary data are hard to integrate without questions relevant to the entire team	Continue to use this structured process to improve interdisciplinary communication
4	Next-generation transdisciplinary	Can stakeholder willingness to support conservation be predicted?	Biophysical and stakeholder characteristics	Willingness to support a conservation policy	Are stakeholders willing to pay for a conservation policy? Why or why not?	Disciplinary resistance decreases as teams work together	Alternate explanations can be reframed as testable hypotheses
5		Does willingness to support conservation track hydrological and biological data?	Flow and fish simulations.	Willingness to support a conservation policy.	Are stakeholders more willing to pay under extreme conditions (floods, droughts)?	The team learns more about how to interpret the model (what is believable, what is not).	When do biophysical realities and stakeholder perceptions match?
6		Do all stakeholders act the same? What do differences mean for conservation?	Attitudes and values of different stakeholder groups.	Willingness to support a conservation policy.	What causes differences among stakeholders? What are the implications for conservation?	A lesson learned is the strength of using an interactive (not static) model.	Unexpected results can lead to more sophisticated interdisciplinary questions.

Note: The iterations increase in complexity from 1 to 6. Because identifying relevant questions is difficult, we provide six iterations for example team 1, the team that reflects the authors’ experiences.

Example team 2: Teams of environmental practitioners

Example team 2 consisted of environmental practitioners with expert technical skills (e.g., water management, land use, fish and wildlife biology, agriculture, engineering, environmental protection, outreach, extension, and community relations). Environmental practitioner teams are very common for many applied ecological issues (e.g., managing fish, wildlife, land, water).

Step 1

As real-world examples, interagency environmental practitioner teams regularly collaborate on balancing use and conservation of water (Watershed Restoration and Protection Strategy, www.kswraps.org) and threatened or endangered species management (Endangered Species Act Recovery Teams, USFWS, https://fws.gov/program/endangered-species). Environmental practitioners have many years of experience in their technical area; are typically employed by local, state, or federal agencies or by nongovernmental organizations; and often cooperate with each other to provide technical assistance to important stakeholders (in the present article, farmers and ranchers). As positive attributes, environmental practitioners understand the complexity of the political, institutional, and economic landscape, stakeholder needs, and the problems for which solutions are needed. However, environmental practitioners may have blind spots—potential obstacles—related to their employer's institutional mission that make them unable to see new pathways. In addition, because environmental practitioners operate largely outside the university, they may not be current on the most recent scientific developments.

Step 2

The overarching purpose of example team 2 was to gain information and techniques to advise farmers on successful agricultural practices in a changing world. Specifically, this team's goal was to craft a strategy that helps farmers both advance their success in uncertain climate conditions and build bridges with nonagricultural interests.

Steps 3–6

Members of example team 2 used the agent-based model to examine what-if conditions related to agricultural practices (step 3). In the initial iterative loop (iteration 1, table 2), example team 2 asked simple disciplinary or multidisciplinary questions about productivity (outputs: yield, profitability) of familiar crops (inputs: corn, soybeans, wheat) under current market, weather, and environmental conditions (step 4). The related structured discussion of model output showed how the model worked (step 5). In iteration 2, example team 2 crafted first-generation interdisciplinary questions that explored how time frame influenced agricultural planning given the uncertainty associated with climate change (iteration 2, table 2). Specifically, iteration 2 tested how relevant outputs (yield and profitability for a range of crops) varied for a single year, a 10-year average, and multiple 20-year climate change projections (step 4). Comparing means and interannual variability for yield and profitability over 1, 10, and 20 years (step 5) promoted a structured discussion about how a portfolio (multiple crops for multiple years in crop rotations) compared with single year, single crop targets in variable environments (step 6). In the third iterative loop, example team 2 asked more sophisticated, next-gen transdisciplinary questions that explored from the farmer's perspective the interactions between agriculture and other parts of the ecosystem (iteration 3, table 2). Specifically, in this third iteration, example team 2 asked how agricultural success was affected by environmental factors other than crops planted, such as how riparian plants aid flood control and, therefore, benefit farmers (step 5). A lesson learned from this third iterative loop was that natural vegetation, cultivated crops, groundwater, stream flow, climate, and flooding are interconnected. Therefore, to thrive, farmers need to reach out and build connections with other stakeholders (step 6).

Table 2.

Examples of multiple iterations of questions, model input, model output, structured discussion topics, lessons learned, and next steps for example team 2: environmental practitioners.

Iteration	Stage	Questions	Model input	Model output	Structured discussion topics	Lessons learned	Next steps
1	Multidisciplinary (preparing the team)	What crops are most lucrative given present temperature and precipitation regimes?	Current temperature, precipitation	Crop yield	Under present conditions, how do different crops perform?	The model can be a useful tool for helping professionals and stakeholders think through what-if issues	Continue to develop buy-in for how the model works as a scientific and persuasion tool
2	First-generation interdisciplinary	What is the effect of yearly variation in weather conditions?	Climate change related temperature, precipitation over 1, 10, or 20 years	Crop yield	Should we maximize mean or minimize variance?	Variation over a long time period is important and useful to consider as a strategy in the face of uncertainty and change	To thrive in the face of uncertainty, investigate proactive responses to climate variation
3	Next-generation transdisciplinary	What other parts of the ecosystem affect agriculture?	Flooding.	Crop yield.	Ecological direct and indirect effects affect farmers and ranchers.	It is in the farmer's interests to have healthy soil, river, wildlife, and to include other interests in planning.	Consider even broader implications of agriculture and the need to build bridges across stakeholders.

Note: For readability, only three iterations are shown to illustrate a general increase in the complexity of questions, model input and output, and structured discussion topics as the team practices working together.

Example team 3: Public meetings

Our third example team simulated a public meeting in which technical experts (such as those described in example team 1 and 2) meet with diverse nonscientist stakeholder groups (e.g., farmers, members of businesses that support agriculture, anglers, hunters, recreationists, municipal residents). These public meetings typically discuss options and consequences for land, water, and fish–wildlife actions (regulatory and voluntary).

Step 1

Real-world examples of public meetings include state fish and wildlife commission meetings, public meetings about a resource use conflict (i.e., dam removal or water use restrictions), and community conservation partnerships (e.g., watershed association activities or riverfront park development). Such mixed interest groups are unlikely to agree, but we include them as a team example to show how the application of RISE could improve the function of public meetings. Recurrent public meetings, or meetings in which the same individuals (or organizations) are repeatedly present, would be ideal for deploying RISE as these repeated meetings would allow experience and knowledge to accumulate for teams of the same individuals (or stakeholder groups) across iterations.

Step 2

The general purpose of our public meeting team example was to identify how science can address environmental problems at the intersection of agriculture, climate change, and environmental protection (Dong et al. 2017). Often a shared outcome from public meetings fails to emerge because technical experts and nonscientists cannot engage in meaningful collaboration because of divergent perceptions of the problem (Bergtold et al. 2022). Therefore, the goal of example team 3 was to identify shared connections between economic activities, human values, and the environment for diverse stakeholders.

Steps 3–6

Researchers within example team 3 can modify the agent-based model as needed (step 3). Nonscientific stakeholders often benefit from an opportunity to see for themselves through hands-on practice that science and data can be part of a solution that supports their interests. Communication of results will be tailored to the expertise of the public meeting team. Scientists and nonscientists have different levels of knowledge and expertise. Although it is true that nonscientists may not know as much as scientists about technical issues, scientists working in different disciplines also will have gaps in their knowledge. Consequently, both all-scientist and scientist–nonscientist teams, require clear communication of results. For example team 3, the first iteration (iteration 1, table 3) fulfills this show-me function by examining how climate change affected multiple components of an agricultural watershed (McAfee et al. 2019). In the first-generation interdisciplinary iterative loop, questions, model output and structured discussions focus on how general environmental changes (e.g., flooding) hurt all stakeholders (steps 4 and 5, iteration 2, table 3). Often, a diverse group of nonscientific stakeholders is not aware of the breadth and connections in the larger coupled human–nature ecosystem. However, these stakeholders can offer critiques, insights, and lived experiences that are often missing in academics (Eigenbrode et al. 2018). The RISE process is designed to value the input of all members of the team in formulating questions, providing feedback, and reflections on the outputs; as such, nonscientists and scientists can both be influential in question generation. This second iteration identified common ground across stakeholders with different primary interests. For example, both farmers and nonfarmers benefit from maintaining a healthy ecosystem (step 6; iteration 2, table 2). For iteration 3, next-gen transdisciplinary questions, model output, and structured discussions examined how different stakeholders might achieve a beneficial, shared conservation goal through compromise (steps 4 and 5, iteration 3, table 3). For example, in a water scarce environment, if crop choice differentially affects the overall available water budget, farmers may inadvertently introduce competition for water among stakeholders and create negative feedback loops with these stakeholders that have adverse consequences for farmers (table 3). A lesson learned from this third iteration was that if farmers do not provide economic and recreational opportunities for other stakeholder groups, these nonagricultural interest groups may not support farmers and ranchers in future local land-use decisions (step 6). Although a team may never fully solve an issue, stopping points for the upward RISE iterations will likely emerge as the team answers specific questions, as funding ceases, or as priorities change.

Table 3.

Examples of multiple iterations of questions, model input, model output, structured discussion topics, lessons learned, and next steps for example team 3: a public meeting.

Iteration	Stage	Questions	Model input	Model output	Structured discussion topics	Lessons learned	Next steps
1	Multidisciplinary (preparing the team)	What is the effect of yearly variation in weather conditions?	Climate change related temperature, precipitation over 1, 10, or 20 years	Crop yield	Climate change relative to farming and other interests	Establish appreciation of the need for common ground	Explore strengths and weaknesses of the model including utility for multistakeholder understanding and for discussing diverse interests
2	First-generation interdisciplinary	What other parts of the ecosystem affect agriculture?	Flooding	Crop yield	Agriculture is affected by other things than just crops planted (e.g., riparian plants are needed for flood control)	Related to flooding, ground water, stream flow, and climate change are all connected to agriculture	Ultimately, diverse stakeholders need to promote their own interests, but still compromise so that other stakeholder interests are also preserved
3	Next-generation transdisciplinary	How can agriculture coexist with the whole ecosystem? Why should this be a priority for farmers?	Water use by agriculture	Watershed scale water budget that affects other nonfarming uses	How can farmers create allies that support agricultural policies?	If farmers provide economic and recreational opportunities for other groups, those groups will support the farmers	Discuss the value of balancing multiple interests in an agricultural watershed The model can explore how environmental regulations aid this balance of interests

Note: The questions, model input, and model output for iterations 1 and 2 of example team 3 are the same as the questions, model inputs, and model outputs for example team 2: iteration 2 and 3 as the public meeting team builds on the expertise of the environmental practitioner team. However, the structured discussion topic, lessons learned and next steps are quite different for the public meeting team.

Advantages and Future Use of RISE

RISE adds to existing meeting, team-building, and project implementation procedures in four ways. First, RISE provides many, regular opportunities to systematically learn and practice working with a team on the same focused problem. Second, RISE provides a deliberately structured six-step process for team learning and practice that systematically connects progress across project components. Third, in RISE, the formulation of increasingly sophisticated interdisciplinary questions by the team is the guiding force that drives activities and determines progress. Fourth, RISE uses interactivity with a quantitative tool to accelerate the potential for cohesion and productivity. The real-time, relatively rapid team interactivity with quantitative technologies (e.g., agent-based models, digital twins, machine learning, case studies) applied to a shared interdisciplinary data set, is a unique strength of RISE. In this real-time, relatively rapid interactivity, the team focuses on a shared interdisciplinary data set, formulates relevant interdisciplinary questions together, processes those questions through an interactive quantitative tool (in the present article, an agent-based model), interprets the output as a team, identifies the next set of questions together on the basis of what was learned from model output interpretation, then repeats the cycle. The questions are identified by the group and will change with each iteration as the knowledge and cohesion of the group increases. If only data collected by principal investigators during their field seasons are used, team interactions will only be practiced a few times during the project. Using RISE, many iterations can occur rapidly (weekly, monthly) and can therefore accelerate team learning, quicken group cohesion, and advance interdisciplinary learning. To our knowledge, other approaches use pieces of RISE but do not systematically and repeatedly apply all of the steps we describe in the present article.

Evidence for effectiveness

RISE emerged from coauthor team interactions related to an NSF CNH grant. For this multiyear grant, our interdisciplinary team started out with a monthly meeting (i.e., agenda, data presentations, group discussions, and minutes) that were useful for learning about other disciplines. We ceased whole-team meetings to make time for focused data collection. When we reconvened regular meetings, we again shared ideas and results through a regular data-review meeting format. Up to this point, our interactions were similar to other teams. Our interdisciplinary progress was accelerated, however, because the team started to build and use the agent-based model. As we met collaboratively to discuss inputs and outputs of model runs, our coauthor team saw the advantages of interactively asking and answering group-determined questions. Evaluating whether our structured approach provides better understanding, more rapid progress, higher-quality questions, or higher-impact publications would be a valuable (but daunting) enterprise, well beyond the scope of the framework that we provide in the present article. We encourage other team science professionals to pick up the gauntlet and implement such a test of different cross-team processes.

Implementation options

RISE can be operationalized as a formal class or as a continuing education exercise for practitioners (e.g., American Institute of Biological Science's “Enabling Interdisciplinary and Team Science” workshop). Below, we show how Project RISE (P-RISE) could be developed into a graduate-level, semester-long class that is open to students from all disciplines as long as they are committed to stimulating interdisciplinary and transdisciplinary research applied to environmental problems. P-RISE would use a case-study puzzle approach in which each student and instructor would participate as experts on a class team (similar to example team 1). Sources for the case study can be actual research, student suggestions, published examples (e.g., BioScience, Case Studies of the Environment, or the SESYNC Case Study Collection https://sesync.org/resources/case-study-collection). The instructor can lead the model building in step 3 or can adapt an existing quantitative tool (see NetLogo User Community Models, https://ccl.northwestern.edu/netlogo/models/community, or the Network for Computational Modeling in the Social and Ecological Sciences, CoMSES Net, https://www.comses.net). For the first component of P-RISE (25% of the final grade), each student would take responsibility for one disciplinary perspective in the case study (i.e., describe a jigsaw puzzle piece) and would be assessed on how well they prepared and shared their expertise. For the second component of P-RISE (40% of the final grade), students would apply the RISE structured process to the case study across multiple iterations. In component 2, students would be evaluated on collegiality, inquiry or curiosity, efforts to make disciplinary assumptions’ explicit to all, participation in reflection and feedback, and future planning. Students could also hand in an extra-credit personal journal that reflects on their experiences. In the third component of P-RISE (35% of the final grade), students would write individual papers on team-identified sections of the final team output. The unified team would compile these individual papers into a final summary that they would present to the external community. In addition to learning, polishing, and practicing team skills and gaining interdisciplinary and transdisciplinary understanding, P-RISE would teach valuable lessons on assigning credit and navigating authorship.

This semester graduate course also could be condensed into a 3-day workshop attended by professionals. In the workshop format, the first day would be dedicated to both introducing the case study, and having participants apply their expertise to this data-based example. The second day would have workshop attendees participate in at least one (optimally two or more) iteration of the six-step structured process and reflect on progress or problems the next day in a group discussion. The final day would include additional team interactions (question generation, discussing model outputs, communication of results) and a final team wrap up that would include feedback, reflections, lessons learned, and next steps.

In addition to coursework or professional development, RISE could be a regularly used option in the environmental scientist's toolbox. Implementing RISE requires some preparation, but a time investment is required for any process applied to a difficult problem. A committed team interested in the same technical problem (e.g., climate change, biodiversity, water management, future visioning) would be assembled to gather varied expertise. This team would commit to regular structured meetings across an extended time period (at least 12 meetings at weekly or monthly intervals). An interdisciplinary data set, collected at a shared geographic location, can be assembled. For groups that might actually use RISE, an intensive on-site consultation can be arranged to facilitate the implementation of the RISE process. As the above steps illustrate, with a limited additional investment in personnel and budgets, RISE could be operationalized for widespread use in universities, agencies, and nongovernmental organizations.

Conclusions

Consistent and useful lessons learned can emerge from the RISE process. First and most importantly, individuals can practice some technical skills by themselves. However, to advance interdisciplinary and transdisciplinary understanding, professionals need to practice team skills with their team by working on an actual environmental problem with real data. This finding is supported by the science of team science literature. Second, we believe that systematically repeating all steps within and across iterations in the RISE process is where the learning, practicing, and polishing of team skills can occur. Third, RISE is designed to lead the team in the process of formulating increasingly sophisticated questions, and we expect that undertaking rapid analysis of existing data sets through team interactions with a quantitative tool will aid with dialogue and deliberation of more complex and novel pathways. Fourth, the RISE process is a general approach and can be adapted for diverse teams ranging from researchers to practitioners to mixed scientist–nonscientist groups. Fifth, because researchers cannot test all variables of potential interest, structured discussions that reflect on both the model output and feedback received along the RISE process can help teams prioritize where to look next. In summary, RISE aims to provide a valuable group thinking tool that can help foster the development of team skills and effective team function.

Acknowledgments

This work was supported in part by the National Science Foundation’s Dynamics of Coupled Natural and Human Systems Program, award no. 1313815. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. Time for manuscript preparation was provided by the Kansas Cooperative Fish and Wildlife Research Unit (Kansas State University, the US Geological Survey, US Fish and Wildlife Service, the Kansas Department of Wildlife and Parks, and the Wildlife Management Institute). Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the US government. During manuscript preparation, Gabriel Granco was partially supported by the USDA/AFRI grant no. 2022-67023-36150 and NSF grant no. 2142490. During manuscript preparation, Jason Bergtold and Aleksey Sheshukov were partially supported by National Institute of Food and Agriculture, U.S. Department of Agriculture (Hatch Multistate 801 Projects W-4133 and S-1089, Project#: 1024218). We thank David Haukos, Richard Lehrter II, Jungang Gao, Sarmistha Chatterjee, James Nifong, and Joseph Aistrup for collaborative interactions. Andrew Carlson and three anonymous reviewers provided useful comments. This is contribution number 24-088-J from the Kansas Agricultural Experiment Station.

Author contributions

Martha Mather (Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Visualization, Writing – original draft, Writing – review & editing), Gabriel Granco (Conceptualization, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing), Jason Bergtold (Conceptualization, Investigation, Visualization, Writing – original draft, Writing – review & editing), Marcellus Caldas (Conceptualization, Visualization, Writing – original draft, Writing – review & editing), Jessica Heier Stamm (Conceptualization, Visualization, Writing – review & editing), Aleksey Sheshukov (Conceptualization, Visualization, Writing – review & editing), Matthew Sanderson (Methodology, Writing – original draft, Writing – review & editing), and Melinda Daniels (Conceptualization, Methodology, Visualization, Writing – original draft, Writing – review & editing)

Author Biography

Martha E. Mather ([email protected]) is affiliated with the US Geological Survey's Kansas Cooperative Fish and Wildlife Research Unit, at Kansas State University, in Manhattan, Kansas, in the United States. Gabriel Granco is affiliated with the Department of Geography and Anthropology at California State Polytechnic University, in Pomona, California, in the United States. Jason S. Bergtold is affiliated with the Department of Agricultural Economics at Kansas State University, in Manhattan, Kansas, in the United States. Marcellus M. Caldas is affiliated with the Department of Geography and Geospatial Sciences at Kansas State University, in Manhattan, Kansas, in the United States. Jessica L. Heier Stamm is affiliated with the Department of Industrial and Manufacturing Systems Engineering at Kansas State University, in Manhattan, Kansas, in the United States. Aleksey Y. Sheshukov is affiliated with the Department of Biological and Agricultural Engineering at Kansas State University, in Manhattan, Kansas, in the United States. Matthew R. Sanderson is affiliated with the Department of Sociology, Anthropology, and Social Work and the Department of Geography and Geospatial Sciences at Kansas State University, in Manhattan, Kansas, in the United States. Melinda D. Daniels is affiliated with the Stroud Water Research Center, in Avondale, Pennsylvania, in the United States.