Regime shifts in coupled socio-environmental systems: Review of modelling challenges and approaches☆
Introduction
Large-scale natural disasters, destruction of vital ecosystem services, colonisation by invasive species, and socio-economic crises are currently at the top of the international agenda. Such events interrupt the functioning of economic, ecological, or coupled socio-environmental systems (SES), and may lead to a persistent change in system structure. Even in the absence of external disturbances, in the contemporary highly interconnected world, coupled SES are more vulnerable than they would otherwise be (Helbing, 2013).
In various disciplines, regime shifts, critical transitions, non-marginal changes, and systemic shocks are closely-related terms used to denote a structural change, often with a perceived sense of abruptness. Specifically, in the resilience literature a ‘regime shift’ is a change from one system state to another, although this concept applies to cases where the transition occurs over any timescale, abrupt or otherwise (Walker and Meyers, 2004, Folke, 2006, Carpenter et al., 2011). The term is mainly used in ecology to describe significant, persistent changes in ecosystems – typically with vital consequences for socio-economic systems – which occur due to a switch in the dominant feedbacks that drive the system into a new regime (Biggs et al., 2009). The switch in the dominant feedbacks happens either as a results of a major external shock, or because the feedbacks dominating in the old regime are gradually eroding, passing a threshold after which new feedbacks prevail. As such, it is not unreasonable to apply the concept of regime shifts to socio-ecological or social systems (Schluter et al., 2012, Mueller et al., 2014, Lade et al., 2013), despite the fact that the latter has its own vocabulary to describe analogous phenomena. Specifically, the socio-economic literature uses the term ‘non-marginal change’, which is contrasted with gradual marginal change. Non-marginal change is a major change in the structure of an economy, shifting a socio-economic system onto a radically different trajectory, as opposed to its gradually moving along the same trend (Stern, 2008). Coupled SES are expected to experience major irreversible changes with non-marginal economic effects in a climate-changed world. Despite this, the majority of economic tools are designed to study exclusively marginal changes – i.e. small variations around a particular path. In economics ‘structural change’ refers to a long-term fundamental shift in the functioning of markets and economic structure, moving them into a different state. Abrupt structural change is often linked to macro-economic cycles, such as Kondratieff waves, which under a Schumpeterian interpretation could feature ‘creative destruction’ during downturns, and are accompanied by observed shifts in the time series of socio-economic data (Medhurst and Henry, 2011). The term ‘systemic shock’ is used in financial and environmental economics domain to refer to a major shift in a system state when normally uncorrelated markets and processes become correlated (OECD, 2003, Bhansali, 2008). Systemic shocks are global changes in the functioning of systems on which society depends. They may be driven either by micro-level gradual changes or external disturbances (e.g. natural hazards) (Filatova and Polhill, 2012). The resilience literature also uses the term ‘critical transitions’, which are fundamental shifts experienced by systems when they pass bifurcations (Scheffer et al., 2012). A critical transition to a contrasting system state occurs when a system is approaching a catastrophic bifurcation – a tipping point – around which even small perturbations lead to a large change in system level variables. Positive feedbacks play a vital role in such transitions as they trigger a self-propagating shift to a different state (Scheffer, 2009). Thus, a critical transition is a special type of regime shift, which may occur without any major external shocking event (Andersen et al., 2009).
In this paper we use the term ‘regime shift’ as it is the most all-encompassing concept to describe the phenomena in which we are interested. A regime shift may be driven either by a disturbance or a gradual change (Table 1). ‘Disturbance’ is an exogenous forcing in the form of a hazard event (e.g. hurricane, disease, fire) or in the form of an extreme change in an input variable (e.g. level of precipitation). After a disturbance, the system may either recover back to the same state (Table 1, I) or may shift to a new state1 (Table 1, III), depending on the magnitude, rate of change, duration and frequency of the disturbance as well as the resilience of the system itself. (Gunderson and Holling, 2002, Folke, 2006, Scheffer, 2009). Turner and Dale (1998) review the differences between large infrequent and small frequent disturbances. According to Lake (2000) a disturbance may be in the form of a pulse (short-term and sharp), a press (a sharply-arising and maintained disturbance), or a ramp (a disturbance steadily increasing over time and space without an endpoint). Collins et al. (2011) simplify these ideas to two important kinds of disturbance: long-term sustained press disturbances and discrete, rapid short-term pulse disturbances.
A regime shift may also occur due to gradual changes in the system's components (Table 1, IV), which up to a critical point do not cause a shift in system state (Table 1, II). Regime shifts arising from gradual changes in explanatory variables (exogenous or endogenous drivers of response variables) have become especially apparent in a time of collapse of ecosystems, financial crises, housing bubbles, and climate change. In all these cases it is difficult to identify a single disturbance that caused a regime shift. Instead, it was gradual overfishing that led to the near-extinction of species and destruction of coral reefs (de Young et al., 2008); the slow accumulation of CO2 and other green-house gases that caused climate change and its adverse consequences (IPCC, 2007, Stern, 2008); economic agents one-by-one adopting seemingly rational rules that caused structural changes in financial markets and economy (Anand et al., 2011); and the gradual spread of expectations among individuals of receiving a dividend from housing asset investments as housing prices grow annually driven by an increasing demand that was itself caused by those expectations (Arce and Lopez-Salido, 2011). Often a regime shift occurs when a system is moved towards a threshold by a combination of gradual changes and the shift is precipitated by a disturbance that would otherwise not be as harmful (Biggs et al., 2009).
Moreover, a regime shift may arise not only from gradual changes in a single variable, but from the interactions among processes operating at different spatial and temporal scales. As Carpenter and Turner (2000) point out, the time periods of changes in ecosystems span several orders of magnitude. A further complication is that the emergence of regime shifts from the bottom up in complex SES is embedded in heterogeneous spatial landscapes. The initial spatial correlation of site conditions and domino-effect responses across neighbouring cells strongly affect the consequent evolving patterns of a dynamic adaptive system (Scheffer, 2009). The effects of interactions among different processes across several variables are captured by concept of the ‘perfect storm’. Here, the values of each of the variables taken individually might not be thought extraordinary, but collectively they form a highly unusual set of circumstances sufficient to cause a regime shift.
From a complex adaptive systems perspective, SES are seen as constantly changing, co-adapting, and perpetually out of equilibrium (Arthur et al., 1997, Folke, 2006). Marginal changes when a system gradually moves along a certain trend are quite “convenient” for decision-makers (and modellers), as prediction of future states can with a certain confidence rely on the historic trends and historic data. In other words, we know with a reasonable degree of certainty that with a unit change in driving variable(s) the response variable is likely to change in a predictable direction with a predictable extent. However, a growing body of literature suggests that it is common for complex SES to experience abrupt sudden shifts from one system state to another (Kinzig et al., 2006, Stern, 2008, Scheffer, 2009, Anand et al., 2011, Vespignani, 2012). A system experiencing a regime shift transforms into a system with new properties, structure, feedbacks, and underlying behaviour of components or agents. Macro variables of interest then do not change marginally with a gradual change in independent variables: there is a shift in the trend observed. These altered internal dynamics often prevent or impose a significant barrier to returning to the previous regime, and hence the possibility of regime shift occurring over relatively short timescales is of interest to decision-makers whose power and influence may be adversely affected. The number and diversity of regime shifts encouraged scholars to start collecting them to the Thresholds Database2 and the Regime Shifts Database.3
As these critical events continue to happen, policy-makers need to find effective ways of managing the circumstances in which regime shifts occur (mitigation), or of reducing any negative consequences of regime shifts that cannot be avoided (adaptation). As Polasky et al. (2011) note, the probability of a regime shift could be exogenous (e.g. if management actions have no effect on the likelihood of a regime shift), or endogenous (e.g. when the probability of a regime shift is a function of a resource management policy choice). While empirical evidence based on historic data analysis is growing, anticipating an upcoming regime shift is still a challenge. The discovery of a range of early warning signals, which seem to precede many of the regime shifts registered in the past and which are universally observed in Earth sciences, medicine, and economics (Biggs et al., 2009, Scheffer et al., 2009), is a major breakthrough in this direction. Modelling tools to explore existing and potential future regime shifts and consequences under which they are likely to occur in coupled SES are therefore in high demand. Yet, the design of models to explore system resilience and occurrences of regime shifts is a challenging domain (Schlueter et al., 2012). The development of statistical, equilibrium and dynamic simulation models, which help our understanding of the emergence of regime shifts triggered by exogenous or gradual endogenous processes could support the design of resilient policies to manage SES. Studies of the dynamics of a system undergoing a regime shift tend to use a single modelling approach. However, a systematic overview of various modelling approaches to studying regime shifts is missing.
This paper provides an overview of how various modelling paradigms approach specific modelling challenges that are relevant for exploring coupled SES experiencing regime shifts. There are two main questions that guide this research. First, what are the important modelling aspects to consider when studying regime shifts? Second, how do various modelling traditions approach studying regime shifts? We focus on four main modelling approaches: statistical analysis, system dynamics, equilibrium models, and agent-based models, the motivation for so doing being chiefly driven by the fact that these are the four most commonly applied to the study of regime shifts in coupled SES.
The rest of the paper is organized as follows. First, based on a review of the resilience literature we identify four groups of modelling challenges that are essential to reflect on when designing and describing a model for exploring regime shifts (Section 2). Based on specific modelling examples, Section 3 describes the manner in which different modelling methods approach the study of regime shifts. These examples are reviewed using the four challenges (2.1 Feedbacks between social and environmental systems in coupled SES, 2.2 Sources of regime shifts, 2.3 Complexity aspects, 2.4 Regime shift identification) as evaluation criteria, with details summarized in Table 4, Table 5, Table 6, Table 7. Section 4 discusses how different modelling approaches address the four aspects, and reflects on their strengths and weaknesses when studying regime shifts. We conclude with a discussion of challenges to future modelling work and reflect on the use of modelling for the design of policies to mitigate or adapt to regime shifts.
Section snippets
Regime shifts and challenges to modelling
When reviewing the empirical regime shifts literature, we came across several themes that were discussed to varying levels of detail but consistently in almost every paper. In this section, we group these themes into four broad categories and discuss their implications for modelling in more detail. The categories are then used as points of reflection when reviewing the modelling approaches in the next sections:
- 1.
Feedbacks between social and environmental systems in coupled SES. Links in models of
Modelling approaches to study regime shifts
A variety of modelling approaches have been applied to studying the dynamics of coupled SES. These include analytical and statistical approaches, cellular automata, micro-simulation, computational general equilibrium, partial equilibrium, system dynamics and agent-based modelling. For the purposes of this review, we focused on those that present published examples with applications to study regime shifts in SES. We also tried to collect at least four examples per approach, pursuing the search
Discussion
As Table 4, Table 5, Table 6, Table 7 show and Section 3 discusses, various modelling approaches are applied to the exploration of regime shifts. They give different levels of importance to the key modelling aspects that were identified in Section 2 based on the attention given to them in the empirical cases of registered regime shifts. It should be noted that the modelling approaches are at different stages of development. Perhaps because of this, they also differ in the number of instances of
Conclusions
One of the main aims when studying regime shifts in coupled SES is to understand their nature and thereby to find effective ways to manage circumstances in which regime shifts occur (mitigation), or reduce their negative consequences that cannot be avoided (adaptation). This, however, is an enormous challenge. Designed policies, particularly where they are accountable to an electorate, need to be simple to comprehend and communicate, and be perceived as fair. They thereby run the risk of
Acknowledgements
Funding from the Netherlands Organisation for Scientific Research (NWO) VENI grant 451-11-033, EU FP7 COMPLEX project 308601, and the Scottish Government Rural Affairs and the Environment Portfolio Strategic Research Theme 1 (Ecosystem Services) is gratefully acknowledged. We are also thankful to the Guest Editor Dr. M. Schlüter and to the four anonymous reviewers for their comments and feedback, which helped us improve the paper.
References (116)
Investigating agricultural sustainability and strategies in northern Mesopotamia: results produced using a socio-ecological modeling approach
J. Archaeol. Sci.
(2008)Modeling human decisions in coupled human and natural systems: review of agent-based models
Ecol. Model.
(2012)- et al.
Ecological thresholds and regime shifts: approaches to identification
Trends Ecol. Evol.
(2009) - et al.
Properties of land mosaics: implications for nature conservation in agricultural environments
Biol. Conserv.
(2006) - et al.
Computable general equilibrium models for sustainability impact assessment: status quo and prospects
Ecol. Econ.
(2006) - et al.
Evaluating policy interventions with general equilibrium externalities
J. Public Econ.
(2008) - et al.
Agrarian transition and lowland-upland interactions in mountain areas in northern Vietnam: application of a multi-agent simulation model
Agric. Syst.
(2005) - et al.
Regime shifts: can ecological theory illuminate the mechanisms?
Prog. Oceanogr.
(2004) - et al.
Regime-switching and interest rates in the European monetary system
J. Int. Econ.
(2000) - et al.
Regime shifts in marine ecosystems: detection, prediction and management
Trends Ecol. Evol.
(2008)
Assessing the transition from deforestation to forest regrowth with an agent-based model of land cover change for south-central Indiana (USA)
Geoforum
Spatial agent-based models for socio-ecological systems: challenges and prospects
Environ. Model. Softw.
Resilience: the emergence of a perspective for social-ecological systems analyses
Glob. Environ. Change Hum. Policy Dimens.
Does structure matter? The impact of switching the agricultural policy regime on farm structures
J. Econ. Behav. Organ.
Further towards a taxonomy of agent-based simulation models in environmental management
Math. Comput. Simul.
The concept and utility of 'ecological thresholds' in biodiversity conservation
Biol. Conserv.
Socio-ecological regime transitions in Austria and the United Kingdom
Ecol. Econ.
Feedback loops and types of adaptation in the modelling of land-use decisions in an agent-based simulation
Environ. Model. Softw.
Spatial-network, general-equilibrium model with a stylized application
Reg. Sci. Urban Econ.
Ecological thresholds: a survey
Ecol. Econ.
Complexity, land-use modeling, and the human dimension: fundamental challenges for mapping unknown outcome spaces
Geoforum
Optimal management with potential regime shifts
J. Environ. Econ. Manag.
Evolutionary modelling of the macro-economic impacts of catastrophic flood events
Ecol. Econ.
A framework to structure agent-based modeling data for social-ecological systems
Struct. Dyn.
Epidemics of rules, rational negligence and market crashes
Eur. J. Finance
Adaptation, spatial heterogeneity, and the vulnerability of agricultural systems to climate change and CO2 fertilization: an integrated assessment approach
Clim. Change
Housing bubbles
Am. Econ. J. Macroecon.
Complexity and the economy
Science
The economy as an evolving complex system II
The Complexity of Cooperation: Agent-based Models of Competition and Collaboration
Population growth and collapse in a multiagent model of the Kayenta Anasazi in Long House Valley
Proc. Natl. Acad. Sci. U. S. A.
Tail risk management
J. Portf. Manag.
Turning back from the brink: detecting an impending regime shift in time to avert it
Proc. Natl. Acad. Sci. U. S. A.
Catastrophic collapse can occur without early warning: examples of silent catastrophes in structured ecological models
Plos One
Applied general equilibrium models
OECD Economic Studies (7)
Early warnings of regime shifts: a whole-ecosystem experiment
Science
Spatial complexity, resilience, and policy diversity: fishing on lake-rich landscapes
Ecol. Soc.
Rising variance: a leading indicator of ecological transition
Ecol. Lett.
Hares and tortoises: interactions of fast and slow variables in ecosystems
Ecosystems
An integrated conceptual framework for long-term social-ecological research
Front. Ecol. Environ.
Indicators of regime shifts in ecological systems: what do we need to know and when do we need to know it
Ecol. Appl.
Scale mismatches in social-ecological systems: causes, consequences, and solutions
Ecol. Soc.
Slowing down as an early warning signal for abrupt climate change
Proc. Nat. Acad. Sci. U. S. A.
Trophic cascades triggered by overfishing reveal possible mechanisms of ecosystem regime shifts
Proc. Natl. Acad. Sci.
Comparing extremism propagation patterns in continuous opinion models
J. Artif. Soc. Soc. Simul.
How can extremism prevail? A study based on the relative agreement interaction model
J. Artif. Soc. Soc. Simul.
Logistic function and arcus tangent
Math. Econ.
Climate-change feedback on economic growth: explorations with a dynamic general equilibrium model
Environ. Dev. Econ.
Assessing the safety of (numerical) representation in social simulation
Economic growth under catastrophes
Integrated Catastrophe Risk Modeling
Cited by (117)
Agent decision-making: The Elephant in the Room - Enabling the justification of decision model fit in social-ecological models
2023, Environmental Modelling and SoftwareIdentifying robust adaptive irrigation operating policies to balance deeply uncertain economic food production and groundwater sustainability trade-offs
2023, Journal of Environmental ManagementA sociotechnical framework to characterize tipping points in water supply systems
2023, Sustainable Cities and SocietyPreparing for an uncertain future: Merging the strategic foresight toolkit with landscape modeling in northeast Minnesota's forests
2023, Landscape and Urban PlanningHow to keep it adequate: A protocol for ensuring validity in agent-based simulation
2023, Environmental Modelling and SoftwareCitation Excerpt :Non-representativity occurs due to a biased sample, which can be caused by different, sometimes subtle reasons, including attrition, self-selection, survivorship or failure bias, observer bias, and unobserved heterogeneity (Vandecasteele and Debels 2007; Gangl 2010; Gormley and Matsa 2014; Jager et al., 2020; Smith 2020). While some minor biases may be corrected by statistical means, structural breaks, non-stationarity or regime shifts – such as climate change – substantially alter statistical x-y relationships causing extreme sample bias: Observed and target situations are so fundamentally different that they must be considered different statistical (sub)populations (non-representative sample setting) and direct generalisation is not possible (Perron 2006; Andersen et al., 2009; Leamer 2010; Filatova et al., 2016; Verstegen et al., 2016). In non-representative sample settings, anticipation of system behaviour for unobserved situations has to rely on structural knowledge about internal system processes (see next section).
- ☆
Special Issue on Modelling Systemic Change.