Structural controllability and management of cascading regime shifts

Juan C. Rocha, Anne-Sophie Crépin · 2025

Read Structural controllability and management of cascading regime shifts by Juan C. Rocha, Anne-Sophie Crépin (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-00206.txt.

Introduction

Juan Carlos Rocha and Anne-Sophie Crépin study how to manage cascading regime shifts in ecosystems, in a research paper from 2024. Ecosystems can flip abruptly to new states, and linked flips can cascade. Managers need to know where a push can stop the cascade or aid recovery, but coupled systems resist simple control.

Network controllability tools frame the problem. Linked shifts prove easier to manage when they share drivers. They get harder when coupling creates fresh feedback loops. Simulations show network shape and coupling strength both matter. The authors call for field tests of cascade claims. They warn that control of real cascades may be weaker than hoped.

Catalog record

CREATORJuan C. Rocha, Anne-Sophie Crépin
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-00206

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Structural controllability and management of cascading regime shifts Juan C. Rocha, Anne-Sophie Crépin Abstract Abrupt transitions in ecosystems can be interconnected, raising challenges for science and management in identifying sufficient interventions to prevent them or recover from undesirable shifts. Here we use principles of network controllability to explore how difficult it is to manage coupled regime shifts. We find that coupled regime shifts are easier to manage when they share drivers, but can become harder to manage if new feedbacks are formed when coupled. Simulation experiments showed that both network structure and coupling strength matter in our ability to manage interconnected systems. This theoretical observation calls for an empirical assessment of cascading regime shifts in ecosystems and warns about our limited ability to control cascading effects. Introduction Regime shifts are large, abrupt and persistent changes in the function and structure of systems 1,2 . They have been documented in a variety of social-ecological systems 3,4 , as well as climate 5 , finance 6 , or health 7 , to name a few examples. Changes in structure and function of ecosystems can diminish the benefits people get from nature such as food production, employment opportunities, climate regulation, or water purification 8 . For example, the North Atlantic cod collapse represented millions of economic loss and compromised more than 35000 jobs in 400 coastal communities 9 . The weakening or late arrival of the Indian summer monsoon can compromise food production in one of the most densely populated regions of the world. How management strategies avoid regime shifts or recover natural ecosystems to desirable regimes are key research questions and an active area of research. But one theoretical fact that complicates their study and management is that regime shifts can be interconnected 10 . The occurrence of one can impact the likelihood of another through sequential tipping, like a domino effect 10,11 ; or through two-way interactions where new feedbacks can amplify or dampen the probability of tipping over 10 . Conceptual work have looked at these connections in the context of committed risk 5,12 , and recent modeling work has studied at interactions between climate tipping elements to understand additional challenges of prediction 13 , or the role of minimal network configurations 14 . Besides climate, cascading effects have also been studied in the context of engineered systems 15,16 , social and economic systems 19 , or diseases 20 . The possibility of cascading regime shifts raises a few additional challenges for management. First, long distance coupling or teleconnections 21 between regime shifts implies that optimal local or regional managerial practices might not be enough to reach managerial objectives 10,22 . This not only illustrates the governance issue of with whom to coordinate for successful management of local resources; it also implies a tragedy of the commons with potential power asymmetries. Classical examples include transnational resource systems where pollution or exploitation of a resource on one side of the border impacts its quality on the other side, or downstream 23,24 . Second, and most importantly, connected regime shifts pose the challenge of identifying sufficient and appropriate points of intervention for managerial actions. In an interconnected world, how can we identify leverage points to manage regime shifts? Are there regime shifts – or their interactions - that are harder to manage? Figure 1: Structural controllability of individual regime shifts The number of nodes of the control set per regime shift (A), and their proportion with respect to the full causal network (B). Top controlling variables ranked by the number of regime shifts where they are part of the control set. Here we answer these questions in the light of controllability of complex systems. A system is controllable if one can drive it from any initial state to any desired final state in finite time 25–27 . Hence, controllability quantifies our ability to steer dynamical systems 27 . While control theory has been fairly well developed in the realm of linear systems, many open problems remain for nonlinear ones 26,28 . Strict global controllability of nonlinear systems remain elusive, but weaker notions of controllability can be described 26,29,30 . For example, open loop control has been used to steer chaotic systems to desired trajectories in the phase space. Here we focus on feedback vertex sets as a proxy of structural controllability 26,29 of regime shifts and their interactions. We use network topology to gain insights on the controllability of regime shift and the management challenges of their cascading effects. We further test our intuitions with minimal models of interconnected resources and pollution recipient systems. Methods We study structural controllability first in a static setting, and then using minimal simulation experiments to approximate the dynamics of interconnected systems. For the static setting we used causal diagrams derived from the regime shifts database, an open online repository of regime shift reviews 4 . The database offers causal graphs that summarize scientific hypotheses of how regime shifts work. Each directed graph (N=30) can be interpreted as a network where a link between two variables exists if a scientific paper has suggested a causal link to the occurrence of the regime shift 4 . We use these networks to study structural controllability (Fig S1 ). Feedback control: To steer a nonlinear system to a desired attractor, one needs to find the feedback vertex set of the system. When the system is described as a network, its state variables are represented by nodes (or vertices), and the dependence between variables as directed links (or arcs). A feedback in this context is a collection of nodes and links that form a circular pathway. Feedbacks are often responsible for the nonlinear dynamics of the system. If a network has feedbacks, it is called a directed cyclic graph (DCG), and if it lacks them a directed acyclic graph (DAG). The feedback vertex set is the set of nodes that if deleted would render a DAG. The feedback arc set is the set of links that if eliminated would render a DAG. Finding the feedback vertex set or the arc set are equivalent problems 31 , yet finding the minimum set is a nondeterministic polynomial-time hard (NP-hard) problem. Once the system is reduced to a DAG, linear control methods can be applied, such as finding unmatched nodes or the Kalman rank condition for linear systems 26,27 . Hence, the set of nodes that need to be controlled to gain leverage on the behaviour of a nonlinear system are the feedback vertex set plus the unmatched nodes of the remaining linear system 26,29 . Structural control on causal networks We compute the proportion of nodes that needs to be controlled for all causal networks from the regime shifts database (Fig S1 ), as well as their pair-wise combinations. The pair-wise combinations have been used to study plausible mechanisms that can couple different types of regime shifts known as cascading effects 10 . We then compare how structural controllability changes if ecological regime shifts are interconnected. Figure 2: Structural controllability of coupled regime shifts The proportion of nodes that need to be controlled to manage individual regime shifts (A) compared against the number of nodes (B) and ratio of nodes (C) that one

Read the complete text