Land Management Change as Adaptation to Climate and Other Stressors: A Systematic Review of Decision Contexts Using Values-Rules-Knowledge

Abstract

Agricultural producers are already experiencing the adverse effects of climate change, highlighting the urgent need for adaptation. While incremental changes to cope with interannual variability are widely applied, there is limited understanding of the social contexts that inform, enable, or constrain more transformational adaptations in response to anticipated or actual climate change and other stressors. Systematic review methods are used to identify 31 empirical examples of land management change as an adaptation response by agricultural producers in developed countries. We then applied the values-rules-knowledge (vrk) framework to analyse interactions between societal values, institutional rules, and scientific and experiential knowledge. The vrk is a heuristic to help decision makers analyze how the social system shapes their decision context. Three propositions highlighting the relative influence of different values–rules, values–knowledge, and rules–knowledge relationships on agri-food and forestry land-management decisions are presented and discussed. We suggest that further testing of these propositions will provide evidence for decision makers about how decision contexts can be shifted to enable anticipatory transformative adaptation in the primary industries and support sustainable transitions towards more resilient futures.

1. Introduction

Primary industries such as pastoral farming, horticulture, viticulture, and forestry are exposed and sensitive to the adverse effects of climate change, with implications for rural value chains [1,2,3]. Climate variability and extremes, including higher temperatures, increased evapotranspiration, and more frequent and severe catastrophic weather events, will require farmers, foresters, and growers to adapt to maintain productive, functional systems. Adaptation involves changes in management practices, processes, and infrastructures to reduce risk and realize opportunities in response to actual or anticipated events [4,5]. Adaptation will involve a range of strategies at different scales, and within and across different land uses and sectors, to minimise collective risk and exposure [6,7]. Many of these changes are incremental, in response to inter- and intra-annual variability, however this may be insufficient and more transformational options—including land management change—may be needed [8].

For farmers and other land managers, adaptation occurs within a dynamic decision context [9,10,11], shaped by climatic and non-climatic shocks and stressors. For example, primary industries must simultaneously adapt to changing climatic conditions, regulatory settings, market risk, financial shocks, and input scarcity, all while maintaining a social licence to operate and public support [12,13,14]. Given this complexity and uncertainty, understanding how to enable adaptation and how to prioritise adaptation actions presents significant methodological and practical challenges [9,15].

Major, non-marginal change to farm systems or components of productive land use is crucial for the long-term sustainability of primary industries [16]. However, there are few empirical examples of transformational adaptation in agriculture [1,8]. While modelled scenarios are available to highlight, for example, potential for changes in future land use suitability, significant changes in land management practices now tend to be constrained by high uncertainty, investment requirements, and perceptions of future risk [8,17,18]. Furthermore, this does not account for the social contexts that inform farmer adaptation decisions, nor the structural factors that may enable or constrain land use change as a climate change adaptation strategy [19,20,21].

The aim of this paper is to identify, for the first time, how interactions between societal values, institutional rules, and scientific and experiential knowledge(s) enable or constrain changes in land management as an adaptation decision. Using systematic literature review methods, we identify 31 examples of land management change as an adaptation to shocks and stressors. The values–rules–knowledge (vrk) framework is then applied ex-post to analyse the context for land management decisions, and better understand how change is enabled or constrained [19,22].

The vrk framework is an emerging device used to examine the structural factors that influence individuals, groups, and organisations when making adaptation decisions [19,22,23,24,25]. The vrk framework helps identify specific values–rules, values–knowledge, and rules–knowledge interactions that impact on the decision context. Three propositions are presented regarding the relative influence of these vrk interactions in enabling or constraining land management change. We argue that by identifying and analysing these interactions, we can gain insight into different decision contexts that enable anticipatory transformative adaptation, and develop propositions that may serve as leverage points for future intervention [26] and that with further testing and empirical analysis of these propositions, decision makers will have greater confidence in the role they might play in enabling agricultural adaptation through purposeful shifting of the combinations thereof [23].

The paper is organised thusly. An overview of the vrk framework and systematic review methods are next, followed by results of the review. We then elaborate propositions regarding the relative influence of values, rules, and knowledge in enabling or constraining primary producers to change land management as an adaptation response, followed by a conclusion.

2. Review Methodology and Analysis

2.1. Values-Rules-Knowledge Framework

Rapidly accelerating changes in climate have significant implications for management and decision-making processes, most of which have long operated under the assumption of stationarity [27,28]. Stationarity assumes that natural systems fluctuate within a static envelope of variability, and therefore the likelihood of risks can be determined from observational records [28]. However, new research on cascading and compounding risks [29,30,31], telecoupling [32], and the challenges of modelling future climate, suggests that new decision processes are needed to address deep uncertainties and constantly changing risks. Therefore, a new approach to decision making is needed, one which moves beyond incremental, short-term, and reactive adjustments to ‘business as usual’, towards long-term, systemic, strategic, and transformational adaptations [33]. To support this shift towards transformational adaptation, greater consideration of the context for decision making, and insight into the ways in which societal processes affect problem framing and subsequent solution space are required [4,19].

Adaption decisions require: knowledge of options and their implications, values to assess the options, and rules that enable implementation. The vrk framework is used therefore to analyse how the decision context evolves as a result of ongoing interactions between values (v), rules (r), and knowledge (k) [19]. Developing insight into the decision context can help identify opportunities to shift structural factors that influence individual and collective decision-making processes, options, and choices (see Figure 1). For example, a decision context might constrain potentially useful adaptation options because decision makers lack knowledge of their efficacy or potential. Or a decision context could enable adaptation options if financial incentives, such as subsidies for land use conversion, were provided to farmers.

While the vrk framework was originally developed for assessing the decision context for climate change adaptation, it has subsequently been applied in diverse settings to understand how interacting values, rules, and knowledge(s) enable and constrain decision-making processes in multi-use woodlands landscapes [22], coral reef management [34], the protection of nature on private property [25], and ocean management [24]. These studies demonstrate that different combinations of values, rules, and knowledge can both constrain and enable adaptation action. In some cases, decision makers occupied multiple decision contexts while expressing, at times, conflicting values, rules, and knowledge(s) [25]. Here, we use vrk to analyse the decision context associated with changing land management as a response to shocks and stressors, including, but not limited to climate change. Retrospective analysis of case study examples can highlight key vrk interactions and help identify new pathways for re-imagining and re-ordering prevailing systems of values, rules, and knowledge currently in use, to identify and realise new opportunities.

2.2. Systematic Literature Review

Systematic literature reviews (SLRs) use repeated analytical methods to collect and analyse primary research studies. SLRs are the syntheses of primary research studies that use reproducible methods to identify and synthesise all relevant material on a specific topic (. SLRs differ from conventional literature reviews by defining a review strategy, making explicit inclusion and exclusion criteria, and by being peer reviewed and pre-published for transparency [35,36,37,38].

SLRs have been conducted in health research [39,40,41], in environmental and conservation research [35,42,43,44], and in climate change adaptation research [45,46,47,48,49,50]. This SLR followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [51,52,53]. Following PRISMA, we first defined the parameters of our search. In setting the parameters of our search, we created exclusion and inclusion criteria (see

Table 1. Exclusion and Inclusion Criteria.

Inclusion Criteria	Exclusion Criteria
Written in English	Not written in English
Published between 1 January 2005 and the 1 May 2020	Published either before the 1 January 2005 or after the 1 May 2020
Indexed on ISI Web of Science	Not indexed on ISI Web of Science
Descriptions of new land uses or changes to existing land management practices	Does not contain description of new land uses or changes to existing land management practices
Analysis of deliberate agricultural land use and/or land management change	Either does not analyse deliberate agricultural land use and/or land management change, or change was not deliberate
An empirical case study within a developed nation	An empirical case study within a non-developed nation

), as well as a template to collect metadata (see

Table 2. Criteria used in the data extraction, organisation, and analysis phase.

Category	Details
Bibliographic details	Author(s), title, publication date
Geography	Nation of origin, level of analysis (global, transnational river basin, national, province/state/regional/river basin, local/city/municipality/county, project, other).
Data source and type	Qualitative, quantitative, comparative, historic, observed, qualitative/quantitative, other.
Analytical focus of study	Outcome (e.g., focus of study is to determine the impacts and implications of exogenous factors to determine their effects, focus is on responding to and mitigating the effects of direct/indirect impacts). Contextual (e.g., focus of study is to understand the context of change by documenting the underlying social; economic; political; institutional; technological; cultural; and environmental conditions that influence capability and capacity for responding effectively and managing or coping with change).
Driver of change	Current climate variability and extremes; climate change (current and anticipated); market or financial shocks or market uncertainty; policy or legislation; land use change; urbanisation; water scarcity.
Land-use Change	Mixed farming; diversification; infrastructural sharing; diversified specialisation; intensified diversification; land sparing, land sharing; patchwork; industrial/land symbiosis.
Values	Principles or qualities that are desirable, which may include (but is not limited to) stated preferences, ethical systems, and principles that inform how people select actions and evaluate events. Values split into three types: (1) subjective intrinsic values (the inherent value of nature as an end in itself); (2) instrumental values (the value of nature for human utility); (3) relational values (concerns related to the meaningfulness of relationships, such as those between nature and people (e.g., sense of place, spirituality, social cohesion, responsibility to biodiversity).
Rules	Rules split into two types: (1) rules-in-use (norms, practices, taboos, habits, heuristics); (2) rules-in-form (regulations, legislation, subsidies, treaties and ordinances, formal rules)
Knowledge	Knowledge split into three types: (1) scientific and technical knowledge (Western knowledge(s)); (2) lay knowledge (experience, belief, worldviews); (3) local ecological knowledge (local and traditional knowledge).
Decision context	The circumstances that form the setting of a decision process; specifically, the interconnected systems of values, rules, and knowledges that form the way of viewing and framing the decision process.

).

Given our criteria, we specified the inclusion of papers that involve deliberate agricultural land-management change [8,54]. As such, non-deliberate spontaneous changes to agricultural landscapes, such as the return of forests after rural retreat, were not included. Papers that analyse or measure the effects of agricultural or land management changes on non-human systems, such as soil or bird populations, were also excluded. To make the distinction between developed and non-developed nations, we followed Ford et al. [37] in defining developed countries as those who are Annex 1 partners of the United Nations Framework Convention on Climate Change. Our search was conducted using the ISI Web of Science platform. Web of Science provides access to multiple databases that contain comprehensive citation data for many different academic disciplines [55,56,57]. Comparative analyses of databases showed that Web of Science is suitable as a primary search system for systematic reviews, but other available databases like Google Scholar are inappropriate [58,59]. The researchers chose Web of Science over alternative appropriate search systems like Scopus or ScienceDirect due to lack of institutional licenses to access these databases.

We used two search chains to collect citations (see

Table 3. Search chains deployed in ISI Web of Science.

Search Chain 1	land use* chang* adapt* agricultur*	2286 Results
Search Chain 2	“land use*” AND “chang*” AND “adapt*” AND “agricultur*”	1612 results

). To ensure a manageable number of citations were collected, the search was limited to papers published between January 2005 and May 2020. Following a deletion of duplicate citations, we were left with 2288 citations.

Titles and abstracts were screened, and 2226 citations were deleted, leaving 62 potentially relevant studies. A review of the remaining 62 citations was completed by both researchers independently. After this process, a total of 31 studies were included in the final review. The process is illustrated in Appendix A. Details of the 31 papers that were included in the review can be found in Appendix B. Also included in the appendices are further statistical results from the systematic review, such as the publication years of the 31 outputs (Appendix C), the four most common journals to find these outputs in (Appendix D), the geographical location of the case studies (Appendix E), as well as the cited drivers for land use change (Appendix F).

To assist in the identification of different vrk interactions, Solomonsz and colleagues [24] proposed three archetypal interactions: values–rules, values–knowledge, and rules–knowledge. Values–rules (vr) interactions occur in the decision context when the favouring of values impact the implementation or interpretation of rules, and vice versa. For example, it could be when farmers reject state subsidies (r) to change land-use because traditional land-uses support cultural values (v). Values–knowledge (vk) interactions occur in the decision context when under- or over-representation of certain values restricts or enhances knowledge, and vice versa. Farmers may desire higher economic returns (v), for instance, which leads them to reject research findings (k) that highlight the environmental impacts of intensive farming. Finally, rules–knowledge (rk) interactions occur in the decision context when certain knowledge(s) is included or excluded from the development of rules systems, and vice versa. If climate change policy (r) excludes Indigenous knowledge systems (k), this may impede climate change adaptation on Indigenous-owned land for example.

Table 4. Values–rules, values–knowledge, and rules–knowledge interactions (adapted [24]).

Interaction	Description
Values-Rules (vr)	vr interactions occur when the favoring of particular values impacts the implementation or interpretation of rules, and vice versa.
Values-Knowledge (vk)	vk interactions occur when the under- or over-representation of certain values restricts or enhances knowledge, and vice versa.
Rules-Knowledge (rk)	rk interactions occur when certain knowledge(s) is included or excluded from the development of rules, and vice versa.

summarises these vrk interactions.

It is important to note the challenges and limitations of using ex-post analysis of case examples of land management change. First, individual case studies are not a priori about values, rules, and knowledge as variables in the decision context. As a result, there are different degrees of detail about the relative influence of vrk interactions across the reviewed literature. Second, some common influences on the decision context may be categorized in multiple ways. For example, we have categorized subsidies as rules-in-form, but subsidies that fund research and development could also be considered as scientific and technical knowledge. In response, the authors have ensured that similar variables, and combinations of variables, across different case studies have been categorized consistently. Third, our aim in this paper is to identify interactions between values, rules, and knowledge(s) to help decision makers enable land management change as an adaptation decision, particularly in the context of climate change; however, only seven of the papers collected ex-post cite climate change as an explicit driver of land management change. We anticipate in the future that more empirical evidence will be available on land management change as a direct adaptation to climate change. Gaining insight into the decision context while documenting and analysing interactions between values, rules, and knowledge(s) can help identify societal influences and institutional structures that facilitate long-term and strategic actions, and illuminate barriers to adaptation. In the following section we present results of the systematic review, followed by the application and analysis of empirical cases, using vrk.

3. Results

The 31 papers selected were all published between 2005 and 2020, with the most popular year for publication being 2016 with five papers. Papers were published in a broad range of journals with Regional Environmental Change, LAND, and the Journal of Rural Studies all publishing two papers. Five papers analysed case studies in the United States of America, four from France, and three each from the United Kingdom and Australia. Six papers had a multi-national focus. Non-climate adaptation drivers were identified in 18 papers, with climate drivers identified in seven papers and a mix of climate and non-climatic drivers in six papers.

Applying the vrk framework to the case studies revealed vr, vk, and rk interactions that enabled or constrained land management change as an adaptation decision. Enabling conditions are used here to refer to conditions or characteristics that can help facilitate an effective response. Conversely, constraints are mutable, ‘soft limits’ on adaptation, and include a wide range of socio-economic, cultural, and biophysical constraining factors that make it harder to plan for and implement adaptation decisions or actions [60].

It is important to note, these reviewed decision contexts are not about climate change exclusively, rather they include responses to a range of shock, stress, and/or other drivers of change that can provide analogous insight. Consequently, enabling conditions for each case study were highly contextual, but in general, included normative aspirations for sustainability in the broad sense of realising desirable social, economic, or environmental outcomes; while barriers were mutable limits needing to be overcome at specific points in time.

3.1. Values–Rules Interactions

The first set of interactions we examined were values–rules (vr) interactions. Following Solomonsz et al. [24], we anticipated vr interactions would occur when precedence is placed on values, which has a subsequent impact on the implementation or interpretation of rules, and vice versa. We present the different vr interactions that enabled land management change as an adaptation response, before exploring the interactions that constrained adaptation.

3.1.1. Enablers

Values–rules interactions can enable land management change when government subsidies (r) are used to support farmers in shifting land use to protect non-commercial values (v), such as the conservation of endangered species. In one example from the Netherlands, agri-environmental schemes encouraged the conservation of endangered bird species, but this required farmers to adapt their land management practices through maintaining longer swards of grass and mowing them less frequently, using differential grazing and manure application to allow for a richer mix of grasses, delaying mowing on margins, mapping and protecting nests, as well as predator control (Swagemakers et al., 2019). State subsidies were provided to farmers, and payment of these subsidies was linked to farmer compliance with these practice changes. In this example, subsidies enabled farmers to make changes to land management that support the conservation of endangered birds while not threatening their income or their way of life. In concluding their paper, Swagemakers et al. [61] argued that farmers’ interactions with nature helped them value the subjective intrinsic qualities of nature, which set in motion changes to local rules and policy to support restorative farming practices. The alignment between farmers’ values and rules helped shift the decision context and the resulting change in land management practices because these practices were consistent with their values and enabled by the rules.

Values are rarely homogenous, however, and often farmers and land managers in a region will hold diverse values, while rules and policy may be designed to protect and promote specific values. For example, in Bernués et al. [62], the authors examined barriers to the wider adoption of low-input pasture-based livestock farming systems in the European Mediterranean. The authors detailed how farmers held diverse instrumental values and objectives related to economic production (v), but policy makers (r) wanted to conserve environmental values, which shifted the system towards the identification of synergies. This generated unique changes in management. For example, decision makers identified areas for wolf and bear conservation that were not suitable for agriculture, as well as identifying opportunities for greater synergy between tourism and agricultural activities.

3.1.2. Constraints

Results also revealed vr interactions that constrained change. In one example, local farming communities informally scrutinized (r) fellow farmers who adopted conservation practices that contrast with accepted farming norms, such as herbicide spraying (v) [63]. Atwell et al. [63] interviewed farmers who reported that they were influenced by their neighbors’ opinions of their practices. If conservation practices conflict with accepted farming norms and practices, and farmers scrutinized each other over these decisions, this can constrain land management change as an adaptation decision. This corresponds to work on social networks and transformational adaptation, in which significant changes in practice were associated with ‘weak ties’ and less concern with what others thought [17].

Values and rules act as constraints in other contexts as well. In O’Rourke [64], the author shows how attempts to lock farming systems into traditional practices effectively meant locking in local cultures as well. Thus, as values slowly shift towards protecting biodiversity (v), farmers will be unable to act on this shift in values unless changes in rules (r) permit them to do so.

In some cases, subsidies provided by supra-national authorities (r) such as the European Union’s Common Agricultural Policy (CAP) framework may threaten the cultural values established by traditional farming systems (v), ultimately constraining land management change. In Havet et al. [65], the authors examined the establishment of integrated crop-livestock systems in western France since the 1960s. CAP subsidies encouraged farmers in western France to modernize farm management practices and land uses, but cultural values represented by long-standing transhumance and integrated crop-livestock systems, remained stable. The resulting tension arose when some farmers resisted change. In this case, cultural values constrained the decision context and inhibited changes in farming practices, despite the financial incentives provided by the subsidies.

In exploring vr enablers, we demonstrated that when rules are designed to protect one value (e.g., conservation of natural resources), but the community holds diverse values, it can generate synergies that enable change. In another paper, we discovered that in a similar context, these interactions resulted in trade-offs that potentially constrain change. Grădinaru and colleagues [66], for example, described how urban residents and farmers hold conflicting values in peri-urban Bucharest, Romania. Shepherds’ deeply held values relating to personal and occupational identity, such as family tradition, and expectations that grazing land in peri-urban areas is a right (v), contrasted with urban residents’ instrumental values, such as a right to breathe clean air (v). Here, shepherds’ values conflicted with zoning and land use rules (r), which posed a threat to the future of pastoral farming. In response, rather than change practices, farmers have either abided by the rules and reduced activity, or they have developed new partnerships to negotiate access to grazing land.

3.2. Values-Knowledge Interactions

Following Solomonsz et al. [24], we anticipated that values–knowledge (vk) interactions occur when the under- or over-representation of certain values restricts or enhances knowledge, and vice versa. Here too, we find examples of vk interactions enabling and constraining the decision context.

3.2.1. Enabler

The alignment between values and knowledge can open up or enlarge the decision context [67]. For example, when farmers value their local environment, and they believe their actions can improve that environment, the combination can enable land management change. Adelaide, Australia already experiences extremely hot days—above 35 °C. each year, and that is likely to increase in frequency [68], making adaptation a necessity. In Robinson et al. [6], the authors examined actions taken by farmers on the peri-urban fringes of the city to adapt to climate change. Their responses to a postal questionnaire confirmed that “actions taken by individuals to adapt to climate change in the peri-urban fringe were closely linked to the nature of the environmental values they hold (or their ecological worldview) and to place attachment. Individuals with a strong place attachment to the Adelaide Hills who possessed knowledge of and/or beliefs in climate change were most likely to take action” [6] (p. 9). For example, over 80% of farmers adopted moisture conservation practices in response to heat stress and drought, and many others planted different pasture species to combat shorter growing seasons. In some cases, there were complete shifts in farming systems from cropping to sheep livestock farming. Previous experience with wildfires also made farmers more likely to take actions to adapt to climate change, such as removing major trees from nearby houses, and installing irrigation systems.

3.2.2. Constraints

The combination of values and knowledge influences the decision context in other ways too. For example, attempts to raise awareness of environmental issues (k) may be unsuccessful if proposed interventions do not align with local values (v). In Armstrong and Stedman [69], the authors examined landowner attitudes to planting riparian margins in a rapidly urbanizing peri-urban neighbourhood. In this research, farmers were aware of the benefits of planting riparian margins (k), but residential citizens, despite attempts by local organisations to raise awareness and knowledge of environmental buffers, took few steps to implement changes on-the-ground. One interviewee reported that “people want a manicured lawn…they want to do what everybody else is doing…They’re just so used to doing what they do, but they’re doing the wrong thing” (v) [69] (p. 1199). In this example, riparian organisations that had experience working with farmers promoted best practice as a minimum of 30 feet of riparian vegetation. These best practices, however, did not align with new residential citizens’ expectations, and failed to account for these value differences that affected implementation.

In Everingham et al. [70], conflicting values and objectives (v) were difficult to resolve because the decision context was characterized by high uncertainty (k). The authors examined land management conflicts resulting from the development of open cut coal mines in rural southeast Queensland. Coal mining led to conflict between farmers, grazers, as well as mining and gas extraction industries. The decision context in this case involved a “multiplicity of stakeholders with inherently antagonistic objectives, interests and values” [70] (p. 72). In their view, the management strategies deployed by different actors such as local government, industry, infrastructure providers, and market actors to deal with these conflicts will not resolve the problem because “they tackle aspects of the problem in a fragmented way” [70] (p. 75). Furthermore, the solutions adopted “rely on inconclusive ‘science’—and cannot readily incorporate hard-to-quantify aspects such as values, emotional impacts and subjective considerations” [70] (p. 75). In this example, poor scientific knowledge, coupled with a decision-making framework that relies on an inadequate knowledge base, struggled to incorporate subjective values into the decision context. Failure to incorporate all values, and lack of knowledge, resulted in a decision context that was hard, if not impossible, to shift.

In another case, the decision context involved deeply held values but incomplete knowledge. Here, in Atwell et al. [71], the authors examined the conservation of ecosystem services in the U.S. corn belt. The decision context in this paper was characterized by high uncertainty, driven by an emerging biofuel economy that would shape landscapes over the next two decades, combined with incomplete scientific and technical knowledge (k). However, farmer values (v) remained strong, and they wanted to devise solutions that were sensitive to local values and practices. This combination of strong values and high uncertainty constrained the decision context, especially when external groups were pressuring farmers to conserve ecosystem services “while not providing adequate compensation for them to do so” [71] (p. 1086). Land management change was further constrained because conservation interests favored the regulation of ecosystem services whereas farmers favored market-based incentives.

3.3. Rules-Knowledge Interactions

The third set of interactions, rules–knowledge (rk), are used to describe how systems of rules are built on knowledge systems, and how this system includes or excludes knowledge, and vice versa [24]. For example, evidence-based policy making endeavors to be based on, or informed by, rigorously established objective evidence, and may exclude lay or traditional knowledge. First, we present the different rk-enabling interactions, before exploring those that served to constrain land management change.

3.3.1. Enablers

Results from the review revealed several examples of land management change that were enabled by the interaction between rules (r), in the form of subsidies, and knowledge (k), particularly traditional, environmental, or ecological knowledge. For example, Atwell et al. [71] reported that a farmer lived with a persistently wet paddock for 15 years because the cost of drainage was too high (k). The farmer eventually converted the paddock to a wetland only because of a neighbour’s comment that there were subsidies available to pay for the conversion (r). In this example, the farmer had direct experience and knowledge of drainage issues, as well as actions that could be taken to improve it, but needed support in the form of a government subsidy before the option of land management change would be considered, let alone realised.

Similarly, land management change can be enabled through the production of new knowledge, including technical or scientific research. For example, Swagemakers et al. [61] demonstrate that government funding (r) for research on the chemical properties of Vacche Rosse cow milk (k) increased local demand for this milk in Parmigiano cheese production. Higher demand for Vacche Rosse milk in turn enabled a shift from Holstein-Freisen cows to this breed. Furthermore, the shift created co-benefits: Vacche Rosse cows are better suited to low-quality grazing, enabling farmers to adopt traditional rather than intensive farming practices.

The importance of government funding to generate new knowledge to enable farm adaptation was noted in other papers too. In Knox et al. [72], the authors argued that new technology, innovation, and practices will enable agri-food systems to adapt. The authors stated the “main technological risks off-farm are related to insufficient R&D investment in agriculture, coupled with a lag in technological uptake compared with the UK’s European neighbours” [72] (p. 253). Redressing this lag required investment in research by government to enable agricultural adaptation through land management change.

In some cases, traditional ecological knowledge and long-standing social institutions can create an enabling environment for farming communities to adapt to future challenges. In Fernald et al. [73], traditional local knowledge (k) was cited as a key factor in the resilience of the acequia irrigation systems of New Mexico, USA. The acequia are community-based flood irrigation systems, owned and managed by self-organized farmers, and used to deliver the natural resource of water to sustain agriculture during scarce or uneven yearly rainfall [73]. These social institutions (r) have persisted for hundreds of years [73] (p. 302). In conclusion, the authors argued that “acequias are resilient because they are in step with the scope and scale of variability in the natural system…the roots of sustainability are the intricate linkages that have developed over generations, connecting human and hydrologic systems” [73] (p. 306). In this example, rules and knowledge were co-developed together over centuries, often under circumstances of severe water shortages that mimic anthropogenic climate change. Hence, in this community, land uses can shift and adapt to water shortages and meet both social and environmental outcomes.

3.3.2. Constraints

We noted above that traditional knowledge and persistent social institutions can enable changes in land management. However, in one example, change was constrained because policies designed for economic growth and conservation separated farmers from their traditional practices. In Gomez-Baggethun et al. [74], the authors investigated the drainage and development of the Doñana wetland in south-western Spain. This drainage was driven by development policies (r) in Spain’s transition to a market economy in the 20th century. As the environmental effects of this policy became clear, strict conservation policies (r) were implemented to protect the remaining undrained wetland. These conservation policies restricted traditional farming practices (k) that had developed in the area, ultimately constraining land management change.

4. Discussion

Applying the vrk framework to the 31 case study examples, we identified different interactions between values, rules, and knowledge that influenced the adaptation decision context. These interactions enabled land management change by enlarging the decision space, revealing new options, and catalyzing change. Conversely, in combination, they also inhibited decision making by generating tradeoffs, fostering uncertainty, and making it harder to reach a consensus on options. A summary of enabling (

Table 5. The values–rules, values–knowledge, and rules–knowledge interactions that enabled land management change as an adaptation response.

Enabling vr Interactions	Enabling vk Interactions	Enabling rk Interactions
Subsidies create scientific and technical knowledge that helps farmers to conserve land and change in ways they could not previously achieve.	When farmers value their local environment, and they have knowledge to improve that environment, this can enable land management change.	Subsidies provide farmers with the ability to conserve land and change in ways they could previously not afford.
When farmers hold diverse values, but policy is only designed to preserve one value, this can lead towards the identification of synergies.		Government-funded research can generate new knowledge, which enables land management change.
		Persistent social institutions and traditional knowledge helped farmers remain resilient to environmental pressures, like droughts and low water supply.

) and constraining (

Table 6. The values–rules, values–knowledge, and rules–knowledge interactions that constrained land management change as an adaptation response.

Constraining vr Interactions	Constraining vk Interactions	Constraining rk Interactions
Producers use informal norms to scrutinize each other’s practices, constraining land management change if changes conflict with traditional values.	Attempts to raise awareness of environmental issues are unsuccessful if proposed interventions do not align with local values.	Strict conservation policies can separate farmers from traditional practices, constraining their ability to change land management.
Traditional farming practices struggle to adapt to changing values, especially if rules remain stable.	Conflicting values are difficult to resolve because the decision context was characterized by high uncertainty.
Subsidies threaten cultural values established by traditional farming systems, leading farmers to reject subsidies and potential land management changes.
When communities hold conflicting values, but rules only protect one value, this can constrain land management change as an adaptation response.

) vrk interactions is shown below.

We found no geographical trends in the enabling or constraining vr, vk, and rk interactions. There are some instances where examples of enabling and constraining interactions were found in the same country. For example, Australian research highlighted how vk interactions could enable land management change through place attachment and knowledge of climate change impacts [6], and how vk interactions could inhibit land management change if there is a knowledge deficit and multiple conflicting values [70].

Furthermore, climate drivers were more common in research from the United Kingdom and Australia than the United States or European case studies. We also found that nations with multiple papers contained a variety of different land management responses. For example, papers from the United States identified urbanization, land sparing, industrial land-use symbiosis, and specialization of crops as an adaptation response to different drivers. Changes in land management typically involve major, non-marginal change in a productive system, including the reordering of inputs, and allocations of land and labour, and may extend beyond the farm gate to incorporate changes in markets, distribution, or value chains [54]. As a result, decisions are highly consequential, with these decisions characterized by complex interactions between values, rules, and knowledge.

Drawing on the vrk framework, and the typology of interactions described by Solomonsz et al. [24], we find several examples of values–rules interactions, in which favoring values impacted the implementation or interpretation of rules, and vice versa. However, these interactions did not simply enable land management change. Some values–rules interactions constrained change. For example, when the community held multiple values, but rules were designed to protect one value. In one example, this tension led to the identification of synergies that enabled land management change as an adaptation response [62], but in another example it ossified the decision context, constraining change ([66]).

Similarly, Solomonsz and colleagues [24] suggest that values–knowledge interaction influence the decision context when the under- or over-representation of certain values restricts or enhances knowledge, and vice versa. Our findings challenge this proposition, as we discovered that these decision contexts consist primarily of stakeholders holding multiple values in decision contexts with high levels of technical or scientific uncertainty. So, these results suggest it is the multiplicity of values, rather than the over- or under-representation of values, that typifies these decision contexts. These findings are supported, for example, in work on collaborative governance and decision making, particularly for resource management [75,76].

Lastly, Solomonsz et al. [24] proposed that rules–knowledge interactions would influence the decision context when certain knowledge(s) is included or excluded from the development of rules, and vice versa. In one example, traditional farming knowledge was not considered in the development of conservation policies to protect the Doñana wetland, ultimately constraining land management change [74], supporting Solomonsz et al.’s [24] proposition. However, the enabling rules–knowledge interactions were not driven by the inclusion of new knowledge into rules, but rather through political institutions either issuing subsidies or funding research that enabled producers to change land management practices. In another example, the development of rules and knowledge had been intertwined over centuries, which enabled farmers to quickly adapt to changing climatic conditions [73].

These results demonstrate first the utility and value in applying the vrk framework. While we applied the framework retrospectively to a small set of case studies, which were not explicitly seeking information or insight into vrk interactions, we were, nonetheless, able to articulate dynamic influences on decisions. Using the vrk framework revealed important but subtle differences in the decision context, and features that affected capabilities and capacities for effectively coping with change.

Results also show vrk interactions are more complex than Solomonsz et al.’s [24] initial propositions, particularly with regard to the mechanisms through which adaptation and changes in land management are enabled or constrained. Shifting the decision context to enable anticipatory transformational adaptation in primary industries, therefore, will also be more complex than simply balancing the over- or under-representation of knowledge(s) in rules or incorporating different values in knowledge generation. Given this complexity, we advance three propositions, informed by our findings, that provide potential pathways to shifting the decision context to enable anticipatory and transformative land management change.

Proposition 1.

Communities typically hold multiple (and often conflicting) subjective intrinsic, instrumental, and relational values simultaneously. This can constrain land management change as an adaptation response. However, in these decision contexts, rules can be rewritten and/or new knowledge can be generated that enables land management change.

In our collected papers, there were several examples of communities holding multiple and often conflicting values and objectives (e.g., [66,69,70]). These value conflicts often contributed to vr and vk interactions that constrained the decision context.

However, there were also several examples where rules were changed and/or new knowledge generated that enabled land management change in a context where farmers held multiple values and objectives. In one example, subsidies were provided to farmers to enable them to change land management to protect endangered birds without losing income [61]. In another example, research on the chemical properties of a specific breed of cow milk created a new market for this milk, enabling farmers to shift to this less-intensive breed [61]. Lastly, diverse values led to the community seeking synergies, such as declaring land unsuitable for agriculture as wolf and bear conservation areas [62].

Hence, some vr and vk decision contexts will constrain land management change if there are multiple values and competing objectives. Instinctively, decision makers might seek consensus over values and objectives to enable change. In contrast, our finding suggest that decision makers should instead seek to change rules, or generate new knowledge, to purposively influence a decision context to enable changes in management.

Proposition 2.

When rules align with local values, land management change decisions are enabled. Traditional farming practices and knowledge are often maintained and preserved in conjunction with (and are supported by) local cultural values and social norms. Legacy practices are slow to change unless rules are sufficiently responsive to local need and values.

In our collected papers, there were several examples where farmers engaged in traditional farming practices that support (and are supported by) local cultural values and social norms (e.g., [64,65]). These examples contributed to vr interactions that constrained the decision context. In several examples, rules were also rewritten to try and enable adaptation in this context. We discovered that when new rules did not align with traditional farmer practices, and the cultural values associated with these practices, the rules were ineffective at encouraging change. For example, in one paper farmers rejected new regulation to conserve ecosystem services because they were not provided adequate compensation to do so, and because the farmers favored market-based incentives and grassroots solutions over regulation [71].

However, on other occasions rules were rewritten, or remained adaptive, which enabled farmers to change in a way that was consistent with their local knowledge, norms, and values. In one example, resilient irrigation systems had been developed during times of resource scarcity, which meant that rules and knowledge had co-evolved under circumstances that mimic climate change-fueled water scarcity [73]. In another, subsidies enabled farmers to protect subjective intrinsic environmental values that set in motion changes to local rules and policies to further support restorative farming practices [61]. So, we hypothesize that when rules are rewritten in a way that is adaptive to and/or aligns with traditional farming practices, cultural values, and social norms, they may help enable land management change as an adaptation option.

Proposition 3.

Land management change is constrained in contexts involving multiple values and high uncertainty. When communities hold multiple subjective intrinsic, instrumental, and relational values simultaneously, it is likely that at least some of these will conflict. When these conflicts combine with scientific or technical uncertainty, decision making is difficult.

Finally, in our collected papers, there were several examples where farming communities held multiple values and objectives and the decision context was characterized by a high level of scientific and technical uncertainty (e.g., [70,71]). In these examples, vk interactions constrained land management change as an adaptation response. For example, in Everingham et al. [70], the authors described how inconclusive science coupled with a decision-making framework that relied on an inadequate knowledge base contributed to a decision context that was hard to shift. In Armstrong and Stedman [69], lack of knowledge about riparian margins among urban residents, and failure to account for urban residents’ values when promoting the benefits of riparian margins, constrained land management change.

Unlike the previous two propositions, we did not find any examples from our collected paper of these decision contexts being shifted to enable land management change. However, like our first proposition, we suggest that decision makers do not try to reach a consensus over values despite value conflicts constraining the decision context. Rather, we suggest it will be more efficacious for decision makers to fund salient, legitimate, and credible research that provides farmers with information to enable future changes. By funding new research, decision makers can help reduce technical and scientific uncertainty to catalyse change.

The results of the analysis and these propositions have the potential to provide decision makers with greater certainty about how to best shift the decision context to enable anticipatory transformative adaptation. While further study, including experimental policy design, and review of empirical examples is needed, our results show there are a variety of vr, vk, and rk interactions that constrain the decision context. However, in many examples, practical actions were taken that enabled change in decision contexts that were previously typified by path dependency.

The review findings also contribute to refining the vrk framework. We note that values, rather than rules or knowledge, appear to be the main factor constraining the adaptation decision context. In our opinion, vrk scholarship has currently undertheorized values. In this review, we attempted to redress this by identifying the differences between subjective intrinsic, instrumental, and relational environmental values. However, there could be a benefit of introducing broader conceptualizations of values, such as those provided by Schwartz [77]. We further request that scholars using the vrk framework focus equal attention on the collection and analysis of data on values as they have on rules and knowledge.

Although our findings suggest that values are the main factor constraining the adaptation context, we propose, perhaps counterintuitively, that decision makers are best able to shift the decision context through rewriting rules and generating new knowledge. In our collected examples, there were several occasions where value conflicts were surmounted through enabling rules and new knowledge; however, there were no examples where value conflicts were resolved through achieving consensus.

Our review also highlights some of the difficulties of applying vrk analysis ex-post. Gorddard et al. [19] describe vrk and the adaptation decision context as a perspective that addresses the influence of societal structures on individual decision-making processes. The papers collected in this review typically aggregated multiple farm decision making processes rather than providing a detailed explanation of individual decisions made by individual farmers. Thus, although vrk promises a balanced appraisal of structure and agency in adaptation decision making, ex-post analysis, which is sensitive to both agency and structure, is difficult due to papers reporting multiple, rather than individual, farm adaptation decisions.

Finally, it is important to highlight some of the current limitations of the vrk framework. First, the vrk framework was developed by ecologists who, in creating the framework, did not engage with existing conceptual and theoretical frameworks that explain change at various scales. For example, vrk could be strengthened by adopting insights into change from the multi-level perspective [78,79], agricultural innovation systems [80,81], boundary work [82,83], diffusion of innovation [84], and extension [85]. Relatedly, the second issue is that the core concepts of ‘values’, ‘rules’, and ‘knowledge’ remain undertheorised in vrk scholarship. We argue integrating lessons and concepts from alternative theories of change will help vrk to better theorise the connections between values, rules, and knowledge. Despite these issues, vrk remains a useful framework for researchers who want to understand and encourage anticipatory transformational adaptation in primary industries.

5. Conclusions

Adaptation to climate change is emerging as a priority scientific need, and there is growing awareness of the imperative to move beyond small-scale, incremental change, to realise major, non-marginal change to reduce risks and realise opportunities. Changing land management is a consequential decision, influenced by a range of social, economic, and environmental considerations. Understanding the dynamics of decisions to maintain or change in response to external shocks or stresses, can provide insight into successfully shifting decision contexts for adaptation elsewhere.

One of the aims of the vrk framework is to analyse adaptation decision contexts to enable anticipatory transformative adaptation. In this paper, we conducted a systematic review of peer-reviewed literature in developed countries where purposeful agricultural land management change was conducted as an adaptation response. Applying vrk, we were able to highlight how different combinations of values–rules, values–knowledge, and rules–knowledge interactions can both constrain and enable adaptation in different situations. On this basis, we were able to derive three propositions on how values, rules, and knowledge constrain the decision context, and how they can be reworked and reimagined to enable land management change as an adaptation response.

As climate change increases with severity and frequency, we anticipate that more farmers will be prompted to adapt. We also expect more research will be published on farmer adaptation to climate change, providing us with additional empirical evidence to test our propositions. We believe that further testing of the propositions produced by our review will validate our findings and will help provide firm guidance for decision makers wanting to shift the decision context to enable anticipatory transformative adaptation in primary industries. Documenting, analysing, and overcoming barriers to action, enhancing social learning for understanding, and motivating action towards climate change adaptation by understanding primary influences on decision making, provide a valuable complement to model-based studies of future climate. However, there is an ongoing need for methodological refinement and application in other settings so that vrk insights can be easily adopted, and to enable sharing of lessons and experiences to inform future policy.