Deliberative project design for understanding and working within complexity in agricultural systems

Highlights

• Two case studies are used to demonstrate how projects can be designed to build understandings of agri-food systems, while embracing complexity.
• Illustrates how projects can be developed along self-organising principles, with decentralised power and clear processes for participation and fractal co-learning.
• Case Study 1 explains how social data can be collected and analysed to build understanding of complex interlinking factors in a socio-agricultural system.
• Case Study 2 describes how transformational change was achieved through co-developed problem definition, diagnosis and evolving co-design of interventions.
• Distils lessons from cases to into attributes for deliberative project design that enable working within complexity in socio-agri-ecosystems.

1 Introduction

Producing food and fibre in the 21st Century involves complex industries operating in the context of an increasingly complex world (Kingwell, 2011). The major challenges of accelerated climate change frequently intersect with soil degradation on a massive scale (Dadzie et al., 2023), with major impacts on planetary boundaries now quantified (Li et al., 2026). The Club of Rome have labelled this interconnected web of global challenges “the Polycrisis” (Rakowski et al., 2025). Globally, and particularly in the recent Australian context, periods of drought and fire have rapidly see-sawed into periods of huge and sustained rainfall, to the extent that there is insufficient sunlight for crops to grow for months on end (BOM, 2023; Pers. Comm, Diana Fear, July 2022; Dyez et al., 2024; Ferrelli et al., 2024). At the same time, rising input costs from broken supply chains are linked with both the COVID-19 pandemic and increasing geopolitical instability (Chopra, 2023; Lin et al., 2023). As well as responding to these external pressures, global markets are changing to meet growing consumer demands for goods to be produced with higher levels of ethical and environmental responsibility (De Olde and Valentinov, 2019; Stimpson et al., 2019). In this context the lives of farmers are subject to change and they must choose how to respond, given the increasing pressures on their direct and indirect outputs.

New approaches, technologies and innovations have the potential to increase farm productivity and boost ecosystem health, however, conventional approaches to agricultural research, development and extension are often predicated on linear models and being able to reduce complexity and uncertainty (Allan et al., 2022). For example, agricultural extension continues to revere Roger’s extensively applied (Rogers, 1962) Diffusion of Innovations theory that describes knowledge-sharing as a linear process of expert, scientific knowledge diffusing among the farmers and ‘being adopted’ at various rates, depending on their receptivity. With capacity to influence being a key element of power, diffusion and similar technology transfer approaches place the scientist in a central position of power (Daouda and Bryant, 2016; Dannecker, 2023). This linear, reductionist background has led to widely accepted positioning of researchers as powerful knowledge-holders, ‘extending’ their knowledge to farmers, as essentially passive receivers, however, approaches to agricultural knowledge sharing are changing (e.g., Allan et al., 2022). While science continues to play a central role in maintaining agricultural productivity, scientists may not always be seen as a trusted provider of information in farming circles, with traditional perceived or actual power imbalances now increasing distrust (Montgomery, 2020).

For almost a century, a reductionist perspective has been central in approaches to both land-management and agricultural sciences, alongside the technological advances and higher agricultural yields that came with the rise of the Green Revolution around the world. These innovations, implemented on a global scale, created a sense of control and permanence underpinned by a strongly mechanistic view of farming practice (Schiere et al., 2012). This perspective sees life and nature as a machine; through this lens, soil may be considered as simply a structural medium for growing produce, to which the correct levels of fertiliser and pest control must be added, rather than a living system (Gomiero et al., 2006). The mechanistic view of natural resource management often includes the belief that resources will renew themselves endlessly, and that technology can overcome any challenges and obstacles (Meadows et al., 1972). Such a paradigm remains popular, evident in the prevalence of agri-tech investment and embodied in the first of the four future scenarios put forward in the CSIRO Ag2050 Scenarios Report (Clegg et al., 2024), which unlocks increased levels of productivity through technological advancements, focussed in few regional Australian cities.

The polycrisis of challenges contributes to the turbulent and sometimes volatile nature of farming across scales (Darnhofer, 2021). These interconnected factors create “wicked” (sensu Rittel and Webber, 1973) contemporary challenges that farmers and agricultural industries must negotiate (Chen, 2021, p179). In the midst of climate change, biodiversity loss and a booming human population, approaches and methods for research that engages and embraces the many complexities of farming, are desperately needed (Pretty et al., 2010). This approach is also embodied in the CSIRO Ag2050 Scenarios report in the second potential future scenario that emphasises the importance of connectivity and diversity in farming systems, where productivity and ecosystems support each other.

Numerous scholars advocate for a paradigm shift in our comprehension of challenges and opportunities involved in socio-ecological systems (see, for example, Luksha, 2023; Mancilla García et al., 2020), urging a departure from the tendency to perceive the world in terms of discrete entities and towards a better understanding of the role of dynamic processes, connectivity, and interdependency in driving complex socio-ecological phenomena. A new perspective, a more appropriate paradigm, is therefore required to understand the drivers of change and to build adaptive capacity within agriculture, beyond linear research and development paradigms (Bateson, 2020; Brown, 2023; Mancilla García et al., 2020; Zhang, et al., 2018). Further, Darnhofer et al. (2010) argue that engaging with a complexity perspective is a necessity for ensuring the viability of farming into the future. More recently, Conti et al. (2024) reinforce the urgency of supporting transformations in agri-food systems through the testing and refining of complexity principles in different contexts.

In the spirit of testing and refining, this paper presents research within Australian agri-food system contexts, posing the research question: What are the specific attributes of deliberative project design that enable understanding and working within complexity in socio-agri-ecosystems’?

In addressing this question, we first summarise the understandings and conceptualisations of complexity occurring in the literature bringing together a framework that is then used to discuss the emergent lessons from two current projects.

2 Literature review

‘Complexity’ describes the multilevel and non-linear interactions between the elements, aspects and dimensions of a situation of interest. A complexity perspective becomes useful when there are more factors and dynamics than can be explained from a mechanistic and reductive world view. From a complexity perspective, linear cause and effect logic gives way to knowing that causality is multidirectional and multidimensional, and there are always interrelated factors we cannot completely predict or control. This perspective includes a recognition that the natural world is prone to constant change, which can be unpredictable and sudden (Darnhofer et al., 2010). As a field of knowledge, complexity shifts the focus from parts of a “system,” to the multi-directional relationships among them (Morin, 2008). To grasp complexity, it is necessary to comprehend how interactions between elements contribute to the behaviour and emergence of the entire socio-ecological system in practical terms (Luke and Evensen, 2021). While not a focus of this paper, it is important to acknowledge that complexity thinking (Morin, 2008) has much in common with the holistic and relational perspectives of Indigenous thinking, such as that of Australian First-Nations Peoples (Yunkaporta, 2019).

Complexity thinking encompasses the key characteristics of non-linearity, dynamics, self-organization, feedback loops, constraints, chaos, emergence, and intangibility (Cilliers, 2002). ‘Fractals’ are a feature of complexity that describe similar features that are expressed in different parts of a system, where smaller events can influence larger events, demonstrating where order can emerge without centralised control (Mandelbrot, 1998). While emphasising interconnectivity and the intangibility of some relationships, these authors also note the importance of defining boundaries as useful units of study as a conscious, socially constructed process that does not detract from an understanding of natural systems as inherently interconnected (Morin, 2008).

2.1 How research projects can support understandings of complexity

Continued, sustained viability and profitability are connected to innovation and environmentally regenerative practices for farmers (Darnhofer, 2021). With a need to support and drive innovation alongside improved farming and land management practices (Lambe et al., 2020); Darnhofer (2020) there is a need to move from a farming system view based on parts that interact, to a complexity perspective of co-evolution, entanglement and emergence, non-linear dynamics, chaos and change. These authors call for an interdisciplinary approach to building and sharing knowledge within which adaptation on-farm is a core goal, rather than the prevailing practice of researchers generating new knowledge and expecting its uptake by farmers. Transformational change can be supported through the co-evolution of ideas and practice, achieved through processes of ongoing, adaptive co-learning within a dynamic network of mutual influences. These influences can be both social and environmental, and may also be operating across multiple levels in an organisation or across organisations (Paniccia et al., 2025).

Lambe et al. (2020) explore transitions from one set of behaviours (or one paradigm) to what they view as a new stable set of behaviours (or a new paradigm), highlighting the importance of situating an intervention or change process within a broader societal context or social system. Datta and Mullainathan (2014) propose three essential stages in change processes: in defining the problem, in diagnosing the problem, and in designing the intervention. If key stakeholders (such as farmers and researchers) are involved in these three stages in a clear an explicit way, there is increased opportunity for transformational change and successful, long-term outcomes (Atmadja et al., 2021). The concept of co-design positions research end-users as co-creators in the research process, defined principally as “meaningful involvement of research users during the study planning phase of a research project” (Slattery et al., 2020, p.2.).

If we are to work with or overcome (numerous) challenges in agriculture, all factors need to be included, with a new strategy required to underpin our understandings of agricultural systems, as a form of complex socio-ecological system. Collaborative projects consisting of diverse autonomous and interconnecting individuals and groups, are themselves complex in nature (Scott et al., 2022). To engage with agriculture today, and to support farmers and their support systems to be adaptive and flexible in the face of challenges and change, collaborative research projects also need to operate through the dynamics of complexity. Complexity thinking provides a language and concepts to know and understand the dynamic interconnectivity of factors in an agricultural situation.

Working with a complexity lens can encourage farmers to see change as a potential opportunity for re-organisation of resources and farming practices; a time to identify the transition choices available and respond with adaptive flexibility rather than reaction (Darnhofer et al., 2010; Schiere et al., 2012). Adaptive approaches can support research and development being undertaken in turbulent and fast changing contexts (Allan and Stankey, 2009; Allan et al., 2011), with continual adaptation to change necessary for maintaining fitness at the scale of farms, and farming systems (Darnhofer et al., 2010). Allan et al. (2008), p. 168) explored how adaptive management paradigms can be used to navigate high levels of complexity and uncertainty, and to accelerate the rate at which managerially relevant knowledge can be acquired.

Learning about complexity through engaging with it in one’s own context can be effective (Yoon et al., 2018). Research projects that embrace collaboration and complexity in the context of education have supported the sector to work amongst the complex challenges of their field (Scott et al., 2020; Scott et al., 2022), finding that flexibility and adaptability are required within project design for maximum benefit when engaging with complex settings and problems. Important characteristics of projects that embrace complexity, tabled by Woolcott et al. (2021), include the relationality and interaction of many stakeholders; the effects of all interactions and actions into self-organising and emergent processes and outcomes; mutual learning amongst stakeholders; and the co-evolution of people and systems, which is stimulated by feedback loops, as well as conditions that are dynamic and sometimes turbulent. These characteristics can also be used to describe dynamic elements of farming systems.

Rather than the researcher being in a central, powerful role (e.g., Montgomery, 2020; Dannecker, 2023; Scott et al., 2022) identified that collaboration with a range of inter-institutional and farming groups can support innovation and learning by providing a distributed diversity of capacity and expertise. Strengthening social and professional networks is an established contributing factor in the resilience of agroecosystems (Altieri et al., 2015). Scott et al. (2022) call for a complexity-focused approach to project design and evaluation across fields, arguing that an explicit, complexity-focused perspective of organisation and governance produces sustainable outcomes. Diversity is a driver of coherence and viability within complex phenomena, combined with the non-linear and non-hierarchical relationality of diverse groups and perspectives within a collaborative research project (Scott et al., 2022). Further, it can support inter-connected groups working together with different forms and levels of governance, with this distributed power referred to by Biggs et al. (2015) as polycentric governance. Seeing and understanding collaborative research through a complexity lens is useful for managing project design, operations and evaluation, as well as providing a window to understanding their potential sustainability (Scott et al., 2022; Woolcott et al., 2021).

Conti et al. (2024) reinforce the importance of developing interventions in agri-food systems that embrace complexity, through working with multiple stakeholders and guiding them through a deliberative and participatory design and decision-making process. Further, transdisciplinarity is considered an essential feature of research in agri-food systems because of its capacity to provide deeper understandings of the complexities involved, as well as potentially uncovering “unconventional solutions” (Conti et al., 2024, p. 9). Deliberative design can empower participants and lead to just outcomes for natural resource management and complex environmental challenges (Chen, 2021; Luke et al., 2014). Deliberative design provides opportunity for participation, communication and empowered contribution, with different perspectives valued through a process that is fair and transparent. Both processes and information involved must have credibility and integrity, with the opportunity for participants to influence process (Luke et al., 2014). Aiming to improve collective decision-making, deliberative processes have been described as a means to address wicked problems, while playing a key role in shifting people from participants to co-learners and co-creators. These aims align closely with the concept of co-learning, which positions learning as a multidirectional, collaborative strategy for addressing complex challenges, defined as providing stakeholders with “opportunities to collaborate on initiatives addressing emergent challenges in their learning, living, and working environments.” (John et al., 2022, p.1). Co-learning should be supported by an environment that fosters safety and openness for shared reflection, inspired by activities that engage diverse perspectives and critical thinking, as a tool for building communities that promote autonomous action, with practices of feedback and self-assessment that strengthen learning for all participants (John et al., 2022).

The relevant conceptual processes and frameworks of co-design, co-learning and co-evolution can be central elements of deliberative project design. Such project design supports the application of a holistic approach for managing research and development projects adaptively, engaging participants and building trust through learning together in the search for understandings among complexity and uncertainty. Conceptual frameworks can also create a navigation pathway, so change is not perceived as “bumbling” (Stankey et al., 2006).

Co-design processes can increase equity in research by flattening traditional research power structures and enabling shared responsibility between researchers, practitioners and communities during processes of project design and delivery (Slattery et al., 2020; Iniesto et al., 2022). However, multiple authors identify a dearth of detailed evaluations of co-design outcomes and impact (Slattery et al., 2020; Iniesto et al., 2022). Hence our search for the specific attributes of deliberative project design that enable understanding and working within complexity in socio-agri-ecosystems.

3 Methods

In this paper we reflect on how we developed understandings and improved knowledge-flows by considering two collaborative projects within the Cooperative Research Centre for High Performance Soils (Soil CRC). We demonstrate on our deliberate project design for building connections within farming contexts across scales to support improved adaptive capacity, and reflect on the outcomes. Each project works with complexity in a different way, yet both provide insights into how to build skills for engaging with complexity in an socio-agricultural context.

In applying a reflective, qualitative methodology, we are guided by Cilliers (2002), Woolcott et al. (2021), Scott et al. (2020) and Conti et al. (2024), condensed into six principles that should underpin the design and implementation of collaborative research projects:

• Defining boundaries: context, scope, constraints, goals and focus.
• Self-organisation: power, accountability, systems, process and hierarchies.
• Adaptability: capacity to continually learn and respond to the unexpected.
• Non-linearity: unorthodox pathways to outcomes/success
• Feedback loops: opportunity for reflection, co-learning and co-evolution.
• Emergence and the intangible: unexpected insights and outcomes.

These principles helped to identify emergent attributes of deliberative project design.

3.1 Data used for team reflections

This paper has been co-authored by many of the project team members, thus includes direct insights from the project team involved in each project. Both projects involved ongoing workshops, interviews and focus group discussions that took place over several years (Approved by Southern Cross University Human Ethics Research Committee, approval numbers 2022/052; 2022/094; ECN-19-062). From these ongoing (and documented) discussions among the project teams, noting substantial overlap between team members on both projects, expected and unexpected insights emerged. From these sources and from the broader insights of the project team members who have authored this paper, a deep reflection of the projects was undertaken (with evaluation methods detailed below), from which they were considered in terms of how they may have enacted the various attributes of complexity provided above.

3.2 Our case study context

The Soil CRC is a 10-year national program of collaborative research that includes Australian universities, industry groups, government departments and 20 rural (farming) groups. The Soil CRC was formed in 2017 to support the ongoing profitability and resilience of farms and the agriculture sector through generating and sharing new knowledge, tools, metrics and practices for improved soil health and productivity. Their support for a broad range of research is based on an understanding of both soil and the soil knowledge-network as complex, evolving and interconnected. The Soil CRC has a range of partner groups, which include farming groups, agricultural organisations, government departments and research agencies from a wide range of areas, disciplines, geographies, organisational structures and kinds of farming. Funding criteria require diverse groups and organisations to be actively involved in the design of projects prior to funding approval as well as in project implementation.

The two Soil CRC projects that are the focus of this paper include a national Social-Benchmarking Study of rural landholders (Case study 1), and a Knowledge-Sharing Project aimed at improving farmer engagement in and across Australian farming regions (Case study 2).

3.2.1 Case study 1: social benchmarking of rural landholders

The Soil CRC Social Benchmarking project used an established mail-out survey method to ask a range of questions of landholders across in selected regions that included North Central Victoria (Victoria), the Eyre Peninsula (South Australia), the North-West Wheatbelt (Western Australia), Central West New South Wales (New South Wales), Tasmania and the Wimmera of North-Western Victoria (Victoria) (Luke et al., 2023). First, in each region, local partnerships were established between universities and farming groups and/or natural resource management (NRM), catchment management authorities (CMAs) and/or other government agencies. Each region included a farming group partner and an NRM partner, with the exception of the two Victorian surveys, which engaged with the Catchment Management Authorities only. The local groups usually acted as a connector with local councils, who shared ratepayer lists with the University-based survey team. Local partners determined the spatial boundaries of the survey, within which either a random sample, or complete census was undertaken to properties more than 10 ha in size. Each contact address was connected to a spatial identifier number linked with cadastral data sets, enabling cross-referencing with biophysical data layers such as soil type and rainfall. The implementation process is a form of probability sampling, enabling responses to be extrapolated to the broader population surveyed (Kalton, 2023). A mail-out strategy was augmented with an identical online survey that could be accessed via QR code or the Soil CRC homepage.

Key topics were co-designed and refined during a sequence of workshops with local partners to ensure maximum relevance to the needs of the region, while core survey questions contributed towards a national data set. In this way, social benchmarking data was collected on a range of topics, with mostly quantitative questions on farmer values, attributes, issues and farm management styles built into the sixteen-page paper booklet (Alexanderson et al., 2023; Luke et al., 2023). Each survey contained items on soil management; economics; succession planning; practices implemented; knowledge related to those practices; practical influences on farm productivity; as well as landholder attitudes to farming practices and other issues such as climate change. The survey form evolved over the process of workshopping its design with the different partner groups. For example, open questions asking farmers about major challenges and profitability insights became standard to include. A survey focus identified in one region sometimes became core to the design and/or analysis, such as soil-testing practices; interest in regenerative agriculture; and questions on farmer-data management and capacity. Analysis by generation, which was first requested in the South Australian workshops, provided insight into landholder management styles, needs and aspirations by established aged group and became standard practice in subsequent reporting.

The survey took landholders up to an hour to complete, with some questions requiring a significant process of reflection on practices related to land-management and farming. In each region, a sample of one to two thousand surveys were mailed out, with response rates ranging from 24 to 35 percent. Brought together, the survey data presents the opportunity for detailed analysis of the complex drivers of on-farm decision-making, in line with best-practice survey methodology (Luke, 2025). This research was mixed-method in several ways: 1), being that key topics embedded in the survey arose from the regional workshops, and 2) because responses generated quantitative data, analysed using a range of statistical tools and showing broad insights, with open survey questions enabling deeper, qualitative insights into landholder responses. Each survey resulted in a promptly-shared four-page summary and a Rural Landholder Social Benchmarking Report, informing an interactive dashboard.

The impact of the Social Benchmarking surveys was evaluated through landholder responses and comments about the process provided. Evaluation of the broader project on regional farming activities and strategic direction of key organisations, took place during workshops with the regional organisational partners. These include discussions informing the co-design process, with post-survey implementation workshops providing feedback and opportunity for ground-truthing and reflection. End of project evaluations took place with the leaders of the regional partner organisations, including some who are co-authors of this paper.

3.2.2 Case study 2: knowledge-sharing for improved soil stewardship

The Soil CRC Knowledge-Sharing Project was designed to set up a series of processes to support local farming groups in four regions. The project outcomes were aimed at providing guidance for groups and individuals seeking to improve their engagement processes with farmers.

This project brought together a team of researchers from four Australian universities (Murdoch University, Charles Sturt University, Federation University and Southern Cross University), working alongside four farming groups. In Australia, these farming groups are typically made up of local farmers and agronomists who wish to provide a focal point for research, development and extension/engagement in their region. These included the West Midlands Group (WMG) of the West Australian (WA) Wheatbelt, Birchip Cropping Group (BCG) of Western Victoria, the Central West Farming Systems Group (CWFS) of New South Wales (NSW), and Agriculture, Innovation and Research Eyre Peninsula (AIR EP) of South Australia. The team evolved over the 2-year project due to staffing changes across the eight organisations, leading to a total of eight researchers, ten farming group representatives and two PhD students involved over the life of the project. The farmer group representatives consisted of seven extension-agents, two of whom were farmers and one an agronomist, with Chief Executive Officers (CEOs) from three of the groups in regular attendance (one of whom was also a farmer).

Each of the project partners were engaged in the initial project development and funding application, most of whom are co-authors of this paper. The project application outlined key objectives and broad stages, but was not prescriptive within each stage, enabling maximum flexibility once funded. Each of the farmer-group partners were allocated budget for monitoring their engagement approaches, with the engagement activities to be trialled identified during the project. This was an unusual approach for the Soil CRC, which had previously required more specific details about practices across the span of proposed projects. Once funding was secured, four online workshops were undertaken to ensure engagement in the design process, as well as supporting connectivity with previous projects.

The Knowledge-Sharing Project initiation workshop involved the co-design of the project’s ‘Big Picture Plan’ (Figure 1). Key elements of the project proposal were shared and collectively refined, with a broad timeline agreed upon. The second workshop included a team activity that applied an empowerment evaluation process following the work of Wiseman et al. (2007) and Fetterman (2002), deliberatively fostering collaboration among those with diverse perspectives to achieve co-learning through co-designing clearly linked processes, outcomes and metrics. A seven-step process responded to the following guiding question:

How can we create a framework for engagement and evaluating engagement that is useful locally but that can generate a shared understanding? Seven steps were undertaken to achieve this, as follows:

• Identification of issues: What needs improvement/changing?
• Goal definition: What would success for this project look like to you?
• Prioritisation of actions: Which elements should we focus on?
• Mapping relationships and connectivity: What are related projects?
• Building relationships: Who else should we bring in?
• Developing an evaluation framework: How do we measure what we are learning?
• Communication planning: How do we share what we are learning?

The project framework developed through this process was shared at subsequent team meetings and revisited regularly. Issues raised, such as farmer engagement-exhaustion and engaging younger farmers had corresponding goals and priorities, informed the evaluation framework, with process diaries and digital metrics put forward as ways to monitor engagement more broadly, without increasing burden on farmer participants. While driving increased farmer knowledge about soil health was an obvious project goal, demonstrating best-practice engagement and stimulating co-learning among the project team about how to engage effectively was a central goal raised by the team. Rigorous self-monitoring was put in place to record co-learning outcomes, with discussions and revelations held in a shared folder in the form of mind-maps, detailed meeting/workshop notes, and interview transcriptions. The framework for a knowledge-sharing plan was co-developed at the third project-development workshop. The framework included consideration of what needed to change and why; identifying soil-health best-practices and specific stakeholders to target; and intended evaluation methods.

Each of the four farming groups then participated in a regional focus-development workshop using this framework, with support from the lead researcher. Each of their knowledge-sharing plans drew on existing local priorities to focus engagement efforts on issues that matter in their farming context. This process drew on data from other Soil CRC projects, which included survey data such as information sources preferred by younger farmers, as well as barriers to adoption identified for each region during a Soil CRC ‘adoption’ project. The regional Knowledge-Sharing Plans became the basis of monitoring for each region.

Monitoring included five components that were: Process Diaries, Event Evaluation Forms, monthly project meetings, digital engagement metrics and interviews between researchers and those undertaking the engagement activities within the farming groups. Process Diaries were co-designed at a fourth project design-workshop, intended for completion by those organising and coordinating engagement activities. These were designed to enable ongoing reflection by the farmer group engagement team, following each event. They were designed to gather quantitative and qualitative insights on elements of success and opportunities for improvement, prompting detailed critical reflection.

Standardised evaluation sheets were co-designed by the project team for completion by farmers and agronomists attending events facilitated by the farming groups. Continuous monitoring by each farming group occurred as diverse engagement strategies were tested. A flow of knowledge-sharing across groups and regions was enabled through the monthly, semi-structured online meetings and workshops, during which the outcomes of engagement, and learnings from monitoring were discussed, as recorded in the Process Diaries. These meetings were well attended for most of the project, ranging from 5 to 15 attendees. They had a general (co-developed) structure designed to encourage the farming groups to share their engagement initiatives, experiences and reflections, sometimes with a discussion focus, such as digital engagement or engaging younger farmers.

Through this rigorous evaluation of activities and synthesis of insights, the project culminated in the development of a multi-media Knowledge-Sharing Guide (see Luke et al., 2024) which also incorporated a digital engagement guide, for use by individuals and organisations seeking stakeholders in the agricultural and NRM sector. Project evaluation took place through ongoing monitoring using transcripts generated at the monthly project meetings. Semi-structured interviews were conducted with farming group CEOs at the beginning, middle and end of the project to uncover reflections, insights, and evaluation of the project’s effectiveness over time.

4 Findings and discussion

From our insider knowledge of the projects above, and our distillation of the literature on addressing complexity, we sought to identify specific attributes of deliberative project design that enable understanding and working within complexity in socio-agri-ecosystems. These attributes are presented and explored below.

4.1 Exploring complexity in Australian farming systems

This project design of Case Study 1: Social Benchmarking of Rural Landholders, demonstrates capacity to explore and draw together many attributes of the farming system through its embodiment of the following six elements, following the complexity principles presented in Section 3.

4.1.1 Defining boundaries

Through co-development of project goals and boundaries within each region surveyed, the project design ensured that survey questions were relevant to the specific needs of each region. Inside a broad project objective to implement surveys across six farming regions, the project’s local goals and focus were not fixed but could be adapted and modified as necessary. Constraining the survey implementation within a geographic boundary that is meaningful to the local partners ensures the survey’s validity and relevance within the local context. It also works within the bounds of the priority topics identified by local partners, building on the core questions.

The specific attribute here is that boundaries work best when set from within the context of the activity.

4.1.2 Self-organisation

The Soil CRC Social Benchmarking Survey project team consisted of a membership with a broad range of perspectives, distributed authority, and shared accountability. Through an emphasis on ‘stepping in lightly’, and visiting regions multiple times to help build relationships and trust, the Soil CRC research team developed trusted relationships with local partners, including farming groups, NRM organizations, CMAs, government agencies and local councils. The collaborative development of the survey questionnaire with these local partners demonstrates collective decision-making and shared responsibility. Two key roles played by the local stakeholders were to access rate-payer lists enabling survey implementation, and media engagement to encourage survey completion. Both these tasks required substantial self-organisation and commitment, with many project related activities being undertaken without continual direction from the Project Lead, though this varied across regions.

The specific attribute here is that understanding strengths of partners and enabling their processes to support the project, builds stronger alliances. Underlying principles of humility and trust are essential to enable this attribute to function.

4.1.3 Adaptability

Over the course of the survey project, the team sought to modify their approach to support the needs of local partners. In addition to background pressures on agricultural communities, COVID-19 restrictions, catastrophic floods for some, and droughts for others, led to unforeseen impacts on the project team and partners. Adaptability was embraced by modification of the survey implementation methods to suit local contexts and relationships (see also Conti et al., 2024). These modifications (see Luke et al., 2023) enabled landholders to choose their preferred mode of participation, either through a traditional mail-out survey or an online survey accessed via QR code or the Soil CRC homepage. This flexibility enabled the project to respond to the preferences and needs of landholders, while the evolving survey design accommodated emergent trends, external challenges and changing policy priorities. The funding agreement from the Soil CRC was focused on learning outcomes, rather than on auditing activity, and this enabled and supported using different approaches and making improvements as required.

The specific attribute that emerged here is that there is willingness to identify different approaches and improvements, and to have institutional arrangements (such as funder evaluation criteria) that support flexibility of approach.

4.1.4 Feedback loops

A process of continual feedback and reflection was achieved by engaging with local partners and landholders at multiple points during the process. The co-designed survey questionnaire allowed for continuous learning and refinement of the survey instrument through workshops and a pre-test with farmers, with a feedback workshop used to ground-truth findings. Additionally, the rapid publication of project summaries, and a few months later, a detailed Rural Landholder Social Benchmarking Report, provided feedback to the surveyed landholders and stakeholders, reinforcing confidence in the survey implementation as a useful tool for supporting change in the region.

The detailed information about the attributes, challenges and aspirations of landholders helps to support strategic planning in the region, fostering a cyclical process of reflection/monitoring, learning, and improvement for partner organisations, as well as the research team. Often, metrics from the survey were built into regional strategic planning documents, such as percent participation and engagement in relevant initiatives. Over time, as Social Benchmarking Surveys were implemented across Australia, the project team reflected on the feedback received, incorporated new insights, and evolved their approach and the core elements of the questionnaire accordingly. This cyclical process of reflection and co-learning fosters continuous improvement and adaptation, including for landholders completing the survey, who are themselves required to critically reflect on their own practice in order to respond to the questions posed. An example of farmer feedback provided is, “I was slow to answer, found it interesting”.

The results of the Social Benchmarking Studies were used to support funding applications for the farming groups, attracting substantial additional investment and capacity building opportunities to their regions, reinforcing feedback loops and connections beyond the life of these projects.

The specific attribute here builds on the one above; building genuine feedback processes into the project that ensure/enable positive and useful flexibility.

4.1.5 Non-linearity

The project recognised the importance of working closely with partners in their local contexts while maintaining a broader perspective, gathering data at a regional level while also contributing to a national data set. By designing the survey questionnaire to capture both local and national perspectives, the project pursued multiple pathways to achieve both local outcomes and overarching project goals, while the surveys built upon and transformed one another over time.

The design of the survey questionnaire and statistical analysis allowed for non-linearity to be embraced in the data analysis phase. Mixed-method analysis of patterns within and across the surveys allowed for unexpected findings to emerge, identifying complex and non-linear associations influencing decisions driving farming practice. For example, it was identified that having family working on the farm was linked to improved succession planning, which is logical and linear (Luke et al., 2023). However, it was also found that having family working on the farm had a relationship with collective decision-making, which was in-turn linked with feeling better supported and coping better, and also to capacity for managing data and for farm planning. Further, having a succession-plan in place was linked to a longer time-frame for strategic planning, which was linked to higher rates of sustainable farming practices being implemented. These more nuanced relationships, supported by findings in the qualitative data, suggest that “family” involvement may have influence that complements other forms of team decision making, a more nuanced and complex finding. This has significant implications for farming families struggling with succession, and for agricultural/NRM organisations seeking to support this process.

The specific attribute here is that there is willingness to explore data and seek complex relationships using both quantitative and qualitative approaches, even if these may not be simple to ‘address'.

4.1.6 Emergence and the intangible

Emergence, and developing an understanding of what may appear on the surface to be intangible, can be identified in both project process and outcomes: Within the surveyed regions, emergence of themes and unexpected effects are observed through the collection of data on a wide range of topics, and the subsequent analysis of drivers of on-farm decision-making, often with unexpected links and associations occurring. Moreover, the survey itself is an active tool for farmers to reflect on their farming practice, their challenges and their farming goals, as well as less tangible things such as their values and attitudes, with the potential for the completion of the detailed survey to support the emergence of new ways of thinking on the farm. These intangible factors have been found to play a crucial role in understanding the complexities of farming systems and decision-making processes.

The survey design included over 300 choices embedded within the survey questions, enabling the use of a range of statistical tools and segmentation for analysis that can zoom out and zoom in on different relationships and drivers of on-farm management practices. By bringing together the diverse perspectives of all types of landholders and farmers, the project embraces the complexity of farming practices, enabling an improved understanding of the influences on farm productivity, and emergent patterns arising from the data. The longitudinal data set created by the project allows for the exploration of trends and emergent phenomena in farming systems over time.

The key attribute here is to be willing to explore data, and seek complex relationships, even if these are not simple to ‘address'.

In sum, the Soil CRC Social Benchmarking Survey of Rural Landholders project demonstrates complexity through its dynamic and interactive approach, incorporating co-development, feedback loops and adaptability. These elements collectively acknowledge the interconnectedness and uniqueness of agricultural systems; the many stakeholders and connectors within them; a need for ongoing learning and adaptation, and the recognition of diverse perspectives to achieve outcomes.

This review of the project’s practice identifies key attributes of deliberative project design that enable working within complexity in socio-agri-ecosystems, which are:

• Boundaries work best when set from within the context.
• Underlying principles of humility and trust are essential for understanding strengths of partners and building stronger alliances.
• A willingness to identify improvements, and to have institutional arrangement (such as funder evaluation criteria) that support flexibility of approach.
• Building genuine feedback processes into the project that ensure/enable positive and useful flexibility.
• There is willingness to explore data, and seek complex relationships, even if these are not simple to “address”.

Further, the above attributes emerge from the application of Datta and Mullainathan (2014) three essential stages for supporting change. Regional partners have been active participants in defining local issues, in working together to interpret the data (diagnosing the problem), and in designing the survey tool, as well as building it’s outcomes into strategic planning (designing the intervention).

4.2 Enacting complexity in the knowledge sharing project

This project was designed to be flexible and adaptable, to empower the group’s members, and support building of connections and learning, applying the six principles for embracing complexity. The project provided indirect complexity knowledge through the ways in which it was organised and conducted. An overview of the initial project design process is shown in Figure 1, demonstrating how the project was co-designed with all project partners and in relation to other projects.

4.2.1 Defining boundaries

By involving multiple stakeholders from the outset of project development and funding application, which noted that ‘The local partners and extension-practitioners will play a key role in driving and documenting the process of action-plan development and implementation’, the goals and focus of the project were co-developed over several months, and through various iterations, prior to application for funding. The project application emphasised the importance of “enabling multiple voices to learn together”. Over the 2-year lifespan of the project, this collaborative approach allowed for adaptability and evolution of the project goals to align with changing needs and priorities of the participating groups. By undertaking co-design working with the constraints and strategic plans of the partner group, the project tailored its knowledge-sharing efforts to meet specific needs and objectives while remaining aligned with the overarching goals to improve knowledge-sharing for soil stewardship (Conti et al., 2024). The project purpose and goals were re-visited regularly and refined at the mid-point of the project, stimulating renewed enthusiasm among the team.

Creating smaller boundaries within a bigger system can be an important attribute of successful project design, and this appeared to be easier if different partners take ownership of specific components. For example, the Federation University team members came from a digital-focussed research group. They responded to the project goal to better engage younger farmers and to broaden engagement by focussing on digital engagement strategies. This enabled them to take ownership of this part of the project, while working in a complementary way with the other project components and sub-groups, with the Southern Cross University team focussed on face-to face engagement, and the Murdoch University team focussed on Policy intersections.

A suggested key attribute of co-design from the social benchmarking survey is that boundaries work best when set from within the context. Lessons from this project reinforce this suggestion, while also highlighting that boundaries to be considered are not only physical, but also operational and relate to people and organisations, as well as lines on maps. A unified sense of purpose combined with a level of specialisation helps the project team to maintain a clear view of the project boundaries and their own role within that.

4.2.2 Self-organisation

Through bringing together people from different types of organisations across four Australian farming regions, this project was designed to create a self-organizing system, where different people, with a range of different roles, skills and perspectives, contribute their expertise and experiences to achieve common goals. The empowerment evaluation process and subsequent steps undertaken were designed to help the project team to internalise evaluation as part of their program planning and management (Fetterman, 2002). While the structure of the knowledge-sharing plans were designed by the full project team, the local plans were developed by the local partners who then used them to inform engagement activities over a 2-year period. This process was very de-centralised, with specific learnings shared with the project team.

Multiple individuals from each partner group actively participated in the project’s development and funding application. Self-organization in decision-making and coordination through polycentric governance across hierarchies was evident throughout the project (Biggs et al., 2015). The regular project meetings brought together partners with different expertise, skill sets and areas of focussed interest to support collective decision-making, as well as the building of relationships and trust among the group, thereby enabling a variety of experiences to be shared with insights that cross-organisational relevance. The diverse membership contributed to a range of perspectives, distributed authority, and shared accountability through the sharing of ideas, successes, failures and observations. Topics for shared discussion were identified by the project team, including learnings associated with different engagement strategies and approaches. Facilitation of the monthly farming group discussions was rotated among the participants to support the drawing out of different areas of interest and focus, supporting a continued emphasis on distributed authority and building resilience through normalising shared leadership.

The group cohesion and capacity for self-organisation was helpful for maintaining project momentum. For example, mid-way through the project and after the Christmas break, a leader of one of the partner organisations noticed that engagement had slowed at the monthly meetings. He responded by suggesting that a project re-focus session be facilitated, reminding the team of the end-goal, while re-defining how to get there in a meaningful way for the team. The project leader suggested that he facilitate these sessions, which he did. These re-focus discussions informed adaptations to the study frameworks, including improvements to the Process Diaries and recognition of their increased usefulness in documenting the effectiveness of events, while clarifying a lack of willingness to use the evaluation forms. Review of this project again reinforces the lesson from the social benchmarking survey that understanding team weaknesses and encouraging the strengths of partners builds stronger alliances, but also highlights that opportunities for new knowledge creation are enhanced through this process. The principles of humility and trust that enable this were very apparent in the Knowledge-Sharing Project.

4.2.3 Adaptive capacity and learning

As mentioned previously, local and national project priorities evolved over time, with local engagement strategies adapted significantly due to a range of factors. External influences included the re-strengthening, then easing of COVID-19 restrictions, which moved workshops online and then back into the field. The excessive rainfall moved more engagement events online, enabling increased digital monitoring to take place. Halfway through the project, three of the partner regions experienced catastrophic floods, with the Central West Farming Systems group seeing this as a particularly critical time to foster a deeper connection between affected farmers, helping to actively build resilience and adaptive capacity right in the heart of extremely challenging times. This was an important insight for the project team and was discussed at length. Using the Process-Diaries immediately following events consistently allowed for deeper reflection of learnings to take place, and be shared with the project team.

At a similar time in another region, the heavy rains led to an opportunistic field day that pulled together hundreds of farmers, providing them with knowledge to help them deal with bogged machinery in waterlogged soils. Exploring one way to respond to the perennial question of how to engage the generally unengaged, the project team deepened discussions on how these sorts of events could be adapted to also include or promote some other key areas of strategic priority. This is a solid example of how practice informed discussions by the project team (Figure 2), building shared understandings that were later published in the Knowledge-Sharing Guide, thus influencing future practice.

That willingness to quickly identify improvements and successes, to adapt them for your context and further team learnings, is an important attribute identified above that is also emphasised in this project. That suitable institutional arrangements can be built into a project to support that flexibility of approach is demonstrated very clearly in the Knowledge-Sharing Project.

4.2.4 Feedback loops

Feedback loops are inherent in the project’s design and evaluation processes, with cyclical processes of reflection, co-learning and co-evolution generated throughout this project. Frameworks for ongoing critical reflection in the form of co-developed Process Diaries were completed by organizers and coordinators involved in engagement activities. These diaries facilitated regular evaluation and feedback loops that were then discussed in regular project meetings, enabling the project team to learn from the experiences of themselves and others, to identify areas for improvement, and co-evolve their engagement strategies. Co-designed evaluation sheets for farmers and agronomists attending local events promoted co-learning by capturing their perspectives and insights, which could be drawn on and linked to in the Process Diaries, and in the regular farmer group discussions. However, the evaluation sheets were not well embraced by the farming groups, either clashing with existing evaluations, or contributing to over-engagement. These sort of feedback loops supports the project’s ability to learn from experiences and enhance its effectiveness over time in a responsive and adaptable manner. The key attribute of building in genuine feedback processes–accompanied by mutual trust - to ensure/enable positive and useful flexibility also emerges clearly in this project, where learning from shared experience became a central operating principle for this project and related farming group activities.

4.2.5 Non-linearity

Non-linearity and dynamic processes are evident in the project’s design and implementation. The continual engagement between farming groups and researchers involves ongoing interactions and adjustments, allowing the project to adapt and respond to changing circumstances and needs. The project’s focus-development sessions and regular online meetings and workshops fostered dynamic processes and relationships, promoting the emergence of new insights and connections. The project built on local knowledge from previous Soil CRC projects, utilising broader knowledge from elsewhere to generate local insights, that could then be scaled-up or trialled by other partner groups in the project, and monitored for their effectiveness in their local context. One example of this is insights from farmer group engagement with women, which informed the research approach of one of the project PhD students who was focussed on women’s knowledge in farming systems. The Knowledge-Sharing Guide developed in the project then allows for the amplification of those insights, developing a capacity for more broad sharing beyond the project boundaries and for uptake by a wider audience.

4.2.6 Emergence and the intangible

The regular structured online meetings and workshops facilitated a flow of diverse ideas, practice and reflective learning across groups and regions, allowing for a broad range of insights and connections to emerge in regards to their in person events and digital engagement strategies. The project process facilitated conversations about how best to undertake engagement, based around ideas that were topical for the groups at that time, such as podcasting, or paddock walks. As the farming group members shared their challenges, opportunities and successes, new ideas and strategies emerged from the group as a whole. Discussions focussed on the complexities associated with engaging and communicating within the agriculture space, and of catching people’s attention through various combined means. Importantly, their iterative discussions identified strategies that were effective, and ultimately, helped develop confidence both in continuing with the approaches that were successful, and in trying new ways to engage farmers and agronomists on topics related to strategic priorities for improving or changing farming practices. Insights were shared about the use and effectiveness of various digital methods for communication, as the groups posed questions of each other and shared tips and tricks for successful digital communication. Regular discussions about online initiatives evolved into new ideas, such as running sequential polls on Twitter that could be used to gauge interest and knowledge-levels on a topic, as well as be used as a lever to promote the next event.

The project team learned the importance of providing clear information about upcoming events using multiple methods of engagement that comprised both digital and non-digital approaches. Knowledge (documented in the Process Diaries) was shared with the project team about successful engagement processes, such as how to best use text messaging services to facilitate information about upcoming events. This sort of knowledge could then be rapidly adopted and trialled by other groups, an example of what we call fractal learning, where a successful activity trialled by one sub-group is shared with the wider team and rapidly adopted by another sub-group, who in-turn provide feedback to the project team about how it worked for them.

Failures were also discussed, such as webinars with low attendance, Twitter feeds with low uptake, and field days that missed the mark. Important insights emerged about how to embrace new approaches and avoid failures in different contexts, such as trying to keep everyone in a large group happy, versus sharing targeted knowledge. They also learned how effective knowledge-sharing efforts can be when farming groups strategically facilitate well-informed and well-respected farmers to share information with their neighbours and other farmers.

A willingness to explore data, and seek complex relationships, even if these are not simple to “address”, is a key attribute of deliberative project design that enable understanding and working within complexity in socio-agri-ecosystems. Emergent outcomes of the self-organising engagement processes by a farming group, shared with the project team and then rapidly trialled by other farming groups, who then reflected upon and shared their insights back with the group, demonstrates complexity by showing how knowledge can flow in non-linear ways and lead to fractal learning. Another specific attribute we identify as important is a regular return and refinement of overall purpose, which we found to help build trust and willingness to adapt, a central learning for research projects working within complex systems, to help them to best address their identified challenges.

Continual team discussions and reflections on engagement activities provided feedback loops that enabled the recognition of constraints, the sharing of learnings and the incorporation of new insights gained from engagement processes in other regions. These elements fostered a collaborative, adaptive and dynamic approach to knowledge-sharing and co-learning, enabling the project and farming groups to effectively engage with different stakeholders; adapt to changing circumstances; and facilitate the emergence of valuable insights and connections (Datta and Mullainathan, 2014).

The reflective process enabled by the Process Diaries was scaled up and widely applied by the project team. The Process Diaries have since been integrated into the standard procedures of at least two of the farming groups involved in the projects, and have been embedded in two subsequent, farmer group led research projects. This demonstrates that this tool has a life beyond the project, and can continue to improve the adaptive capacity and resilience of the groups as they seek to engage further with farmers on key land management issues.

In summary, the Soil CRC Knowledge-Sharing Project enacted complexity through its diversity of members, co-developed goals and cyclical processes of reflection and co-learning, building trust for emergent findings to be embraced by others, offering multiple pathways to local and scaled-up outcomes. These attributes enabled the project to navigate and engage with the complex dynamics of the socio-agricultural landscape, fostering knowledge-sharing and collaboration for regenerative outcomes for their organisations and within the farming-system network.

4.3 Summary

While innovation and the trialling of new approaches to farming becomes essential in increasingly turbulent times, Darnhofer (2020) reminds us that farmers choices and decisions are influenced and constrained by many factors. This includes the information they have access to; their support structures, trust and relationships; their financial capacity and predisposition towards risk, as well as many local and global factors entirely beyond their control. The ways in which we build understandings of farming systems are strongly influenced by the ways in which we learn, access knowledge and work together - even the ways in which we think. We follow and build upon the work of Conti et al. (2024) and Scott et al. (2022). Our case studies provide examples of how researchers and project managers can create an environment where complex factors can come together in relevant contexts while bridging disciplines and agencies, uncovering novel insights, options, and ideas, as well as new ways of working together effectively to build resilient projects in an agricultural context (Sellberg et al., 2021). In both of these projects, complexity was considered at all three essential stages in change processes, including problem definition, diagnosis and in building solutions (Datta and Mullainathan, 2014).

We sought to suggest some specific attributes of deliberative project design that enable understanding and working within complexity in socio-agri-ecosystems, and have listed these as we understand them, relate to each project. In Table 1, below, these attributes, and a summary of how they might be expressed (or not), in traditional project design compared with how they have been enacted in our case study projects. Six principles for embracing complexity in project design are outlined for those seeking to enact and work deliberatively with complexity in future projects.

TABLE 1. Traditional project design compared with deliberative project design embracing complexity.

The analysis of these two Soil CRC case studies provides novel insights about different ways in which projects can be developed deliberatively, allowing complex relationships, interactions and intersections of knowledge and perspectives to come together, enabling emergent insights for improved adaptability, resilience and practical application within projects. The Social Benchmarking Project demonstrates this most fully through the decentralised power arrangements that led to locally-grounded questions that fed into a complex dataset that served both local and national scales, while the Knowledge-Sharing Project demonstrates complexity attributes through the way it was designed, led, and run by the team as it evolved over time, embracing multiple potential and emerging pathways to local outcomes and success, while retaining the whole project goals in mind. It also embraced the fractal nature of complexity by recognising successes in one area and repeating or scaling them up elsewhere.

The project team followed an empowerment evaluation process to work collaboratively to draw on existing data to properly define the challenges they wished to address at both local and project scale, to design the interventions they wished to trial, and to design the evaluation processes they would undertake to demonstrate their effectiveness. This level of engagement is what Datta and Mullainathan (2014) view as key to supporting transformational change that can lead to successful, long-term outcomes. This has been practically demonstrated by many of the project tools and data being used well beyond the life of the project, such as the Process Diaries, and the Social Benchmarking data. We have demonstrated how these efforts can contribute towards improved engagement, knowledge-sharing and ultimately, improved stewardship of land and soil, to support transformative change processes in our farming systems (Darnhofer, 2021).

Multiple pathways to local outcomes/success were discussed during the projects, while keeping overarching project goals in mind. Specifically, positive localised and scaled-up outcomes occurred through initiatives emerging from the nation-wide landholder survey project, and also from the Knowledge Sharing Project. As a result of interaction between diverse groups and people, co-learning and co-evolution and, to varying extents improved adaptive capacity, have occurred both generally at a project level, and for the partner organisations who are co-authors on this paper.

A key lesson from this project is that deliberative approaches to working in complex socio-agri-ecosystems require more than simply adopting participatory methods; they require deliberate attention to how boundaries, relationships, and learning processes are designed and maintained. One mistake encountered was initially underestimating how strongly externally imposed project boundaries can constrain collaboration and limit the relevance of outcomes. In practice, more effective progress occurred when boundaries were negotiated from within the project context, allowing partners to shape the scope according to lived realities. Another important insight was that humility and trust are not abstract values but operational conditions: where these were actively cultivated, partners’ strengths became clearer and alliances deepened; where they were assumed rather than nurtured, collaboration stalled. The project also revealed that flexibility must be institutionally supported. Without evaluation frameworks or funding arrangements that allow adaptation, teams may recognise opportunities for improvement but feel unable to act on them. Finally, genuine feedback loops proved essential but difficult to maintain; early stages of the project treated feedback too informally, which limited its usefulness. Embedding structured mechanisms for reflection and openness to exploring complex, non-linear relationships in the data ultimately strengthened the work. Future projects seeking to replicate or extend these methods should therefore prioritise internally grounded boundary setting, invest deliberately in trust-building, design institutional space for adaptive practice, and create explicit feedback processes that allow teams to engage productively with complexity rather than simplify it prematurely.

It is important to acknowledge the role of the Soil CRC in its deliberative design and capacity to bring together collectives of researchers, agronomists, extension agents, government departments and farming groups over many years, with the Soil CRC 7 years into a 10-year tenure at time of writing. With a tagline of ‘performance through collaboration’, the Soil CRC itself can be argued to demonstrate complexity through an emphasis on building partnerships for the sharing of skills and knowledge between a range of participants with different skill-sets and organisational capabilities. Their emphasis on participatory project design supports capability and capacity building, partnerships and leadership, for improved management that can lead to higher buy-in to project aims, goals and outcomes (Parks et al., 2021). Rather than engaging through the usual knowledge and power hierarchies which are present in many agricultural research projects (Garcia et al., 2024), projects often involve a decentralised arrangement of people with various roles across farming regions, associated institutions and universities, with each group engaging and contributing according to capacity and attributes. Diversity has been consistently identified as a driver of resilience in situations involving complexity (Bateson, 2020; Sellberg et al., 2021; Conti et al., 2024), and this approach to project management brings in a diversity of perspectives, ideas and skills from across multiple organisations.

The higher-level management of the Soil CRC allows for another layer of enabling relationships to develop and strengthen across projects and time, connecting, extending and strengthening broader agricultural knowledge networks across Australia. Arguably, the process of writing this paper, engaging with and reflecting on the projects using language from the complexity literature, has built and strengthened complexity literacy and capacity among the project teams. Our teams’ reflections provide pathways for current and future cooperative research centres (CRCs) to further integrate deliberative design that supports collaboration and capacity to work within complex situations and contexts (Conti et al., 2024).

5 Conclusion

This paper demonstrates how complexity has been enacted in two case study projects by taking a wide range of factors and influences into account in the project design, and through deliberate linkage with interconnected projects and initiatives. The projects discussed in this paper both aimed to support local grower groups in improving their engagement with farmers and key stakeholders in their respective regions. Through bringing together diverse stakeholders with strong links to local, regional and national agricultural networks, linked together through inter-organisational partnerships within multiple projects over multiple years, and through deliberatively considering complexity in their design, this paper provides evidence that these Soil CRC projects have enabled increased complexity knowledge and awareness for the collaborators involved in these agricultural research projects.

Agricultural industries are increasingly dealing with multiple, and sometimes what may appear to be conflicting expectations, as they are required to produce our food and fibre while also preserving and regenerating the health of ecosystems as well as rural communities. As a complex nexus of social, economic, cultural, political and environmental factors come together for each farmer, it may be considered not just an ideal, but a responsibility of supporting organisations to help them to build their capacity for adaptation through understanding and dealing with complexity, as well as for local grower groups. The networks of information flow can be viewed as an attribute of complex communication and knowledge-emergence patterns. We provide six attributes for attributes of deliberative project design that enable understanding and working within complexity in socio-agri-ecosystems that are linked to We provide six principles for deliberative project design embracing complexity, offering here that complexity thinking, and projects that embrace it, provide a useful epistemological lens for studying and supporting change in socio-agricultural systems.

This study provides examples of how channels of communication and design have been opened up for transdisciplinary and collaborative agricultural research. This sort of work also requires a rethinking of the role and expectations of leading researchers when conducting and facilitating collaborative projects. Transdisciplinary and collaborative research requires well-defined thinking around how to design projects that embrace different knowledges, while also framing power imbalances, professional constraints and group dynamics (Thaler, 2022). The Principles for Enacting Complexity arising from these case studies demonstrate clear methods for addressing some of these, yet there is further work to be done in relation to a closer examination of power dynamics in collaborative research and the extent to which empowerment and engagement is realised for all project team members.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Rosa M. Badia, Barcelona Supercomputing Center, Spain

Reviewed by: Douglas Jackson-Smith, The Ohio State University, United States

Lucy Barkley, University of Gloucestershire, United Kingdom