The Doñana case

A two-decade social-ecological research program—synthesis

For close to two decades, Doñana has been the ground on which I have learned to think about social-ecological systems. An estuary-delta on the lower Guadalquivir in southwest Spain, it is one of Europe’s most celebrated protected landscapes and, at the same time, one of its most chronically conflictual—a paradox that first drew me in and never quite let go. The case taught me that a coastal wetland’s ecosystems, agricultural landscapes, water resources, and the local communities and institutions that govern them form a single, tightly coupled system, and that they cannot be understood apart from the others. 

The Doñana case was never conceived as a single study but as a cumulative research programme, designed to accumulate insight iteration by iteration under an adaptive inference protocol (Holling and Allen 2002). Each iteration carried forward what the last had learned, narrowing an initial field of competing explanations and descending progressively from the system level toward the fine grain of strategic behaviour, before abstracting back upward toward theory (Méndez et al. 2012, 2019, 2022, 2023a). What follows synthesises that arc through several interwoven streams, and closes with spin-off abstractions that carried the case's tools beyond Doñana altogether.

1.  The empirical ground: a history of Doñana

Doñana’s present is sedimented history. Over the twentieth century the floodplain marshlands of the right-bank Guadalquivir were progressively drained, diverted, canalised and turned over to irrigated and rice agriculture, all under a persistent “hydraulic mission” (Molle et al. 2009) that treated water as something to be engineered rather than lived with. Conservation arrived, paradoxically, through the same command-and-control logic (Holling and Meffe 1996): the creation of Doñana National Park in 1969 protected a remnant while the surrounding hydrology was rebuilt around it, leaving the marshes dependent on a heavily managed water regime—still fed by rainfall and groundwater, but so extensively diverted and regulated that it functions as an engineered rather than a natural system. We first reconstructed this history as a systems narrative in the program’s opening iteration (Méndez et al. 2012). The rigid regime that emerged proved remarkably durable, surviving crisis after crisis—the waterfowl “botulism” die-offs of the 1980s, the 1998 Los Frailes mine spill that poisoned the Guadiamar, the failure of successive regeneration plans, and, from 2000, a contested megaproject to deep-dredge the estuary for maritime access to Seville’s inland port (Méndez et al. 2023a). This long, conflict-laden history is the empirical spine of everything that follows.

2.  An iterative analytical apparatus: the IAD and its politicisation

To make this history legible we needed an apparatus that could grow as new questions emerged. We began with Elinor Ostrom’s Institutional Analysis and Development (IAD) framework (Kiser and Ostrom 1982, Ostrom 2005, McGinnis 2011), using it to select the relevant factors, relationships and levels of analysis. As the limits of a purely rational-institutional reading became clear, we moved to the politicised IAD (pIAD; Clement 2010, Clement and Amezaga 2013), which brought discourse and power into the frame—the recognition that discourses themselves carry power, conferring or draining legitimacy from institutions. A later iteration descended to the micro level, combining Networks of Action Situations (McGinnis 2011, Kimmich et al. 2022) with an agency-based polycentric power typology (Morrison et al. 2019) to reconstruct the strategic games among Doñana’s actors (Méndez et al. 2023a). Finally, the apparatus consolidated into a deliberately pluralist neo-institutionalist approach, funnelling rational, historical (North 2005) and discursive (Schmidt 2008) strands through the IAD and pIAD, and adding concepts such as discourse inertia to explain why certain framings persist long after they have ceased to serve (Méndez et al. 2022).

3.  Reading Doñana through resilience: rigidity, lock-in, and sustainability pathways

The system-level diagnosis came from resilience thinking—Holling’s adaptive cycle and the panarchy heuristic (Gunderson and Holling 2002). Read through that lens, Doñana is a textbook rigidity trap (Holling et al. 2002): a system whose institutions have become so rigid, so committed to command-and-control, that they smother innovation, squeeze out diversity and erode the very resilience they claim to protect, all while appearing stable. We later mapped Doñana onto a landscape of stability domains with three basins: the rigidity trap it currently occupies, a deeper lock-in trap it has so far avoided (Allison and Hobbs 2004), and a sustainability pathway it might yet reach (Méndez et al. 2023a). That framing let us read pivotal events counterfactually—the 1969 park creation as an event that arguably kept the system from tipping into lock-in, and the estuary dredging megaproject as a plausible push toward it. Traps, in this reading, are not accidents but emergent, historically produced outcomes (Méndez et al. 2019).

The Doñana rigidity trap's making can be traced across tiers and time. From the 1880s a hydraulic-irrigation regenerationist discourse took hold at the constitutional level, codified through water and irrigation laws that recast the marshes as unproductive land to be drained and “colonised” for agriculture. An increasing-returns phase (to 1937) signalled large payoffs downstream; then a self-reinforcement phase (1937–1980s)—massive increasing returns, the setup costs of the first irrigation works, the adaptive expectations of a rice sector entering global markets—locked collective-action actors onto a single trajectory, behaving not irrationally but “picking the right horse” given their incentives, while an interventionist state, its technocratic engineering corps, coercive property reform and the co-option of the emergent conservation regime drained every alternative's appeal. What settled the system into a rigidity trap rather than a deeper lock-in was a counteractive act of institutional entrepreneurship—the 1969 Doñana National Park—preserving ecological value even as it synchronised with the same command-and-control logic. Since then, discourse inertia has held the configuration in place: the hydraulic-mission discourse has spiralled on in contest with a subordinated protection-for-conservation discourse (Méndez et al. 2022), reproducing down to individual strategic behaviour—imprinted in actors' very sense of what counts as rational, still signalling increasing returns (hydraulic megaprojects' sunk costs) and stabilised by the pragmatic power of feigned ignorance, non-decision-making and the passivity of tolerant authorities (Méndez et al. 2023a).

4.  A multi-tiered design: system, meso, micro

These pieces cohere because the case was built as a multi-tiered design. At the system level, resilience thinking and path dependence account for the deep historical trajectory (Gunderson and Holling 2002). At the meso level, the IAD, pIAD and neo-institutionalist strands explain aggregated institutional and discursive mechanisms—institutional entrepreneurship, increasing returns, hegemonic discourse (Ostrom 2005, Clement 2010). At the micro level, action-situation and power analysis recover the strategic behaviour of individual actors (McGinnis 2011, Morrison et al. 2019). The program travelled across these tiers deliberately: it opened at the system level with the long historical evolution of governance, then scaled down toward the micro grain of strategy, holding a productive meso-micro tension throughout (Méndez et al. 2012, 2019, 2023a). Emergence runs both ways—micro interactions embedded in system trajectories, system-level structures shaping micro choices—and only a design that keeps all three tiers in view can follow that property.

5.  A parallel abstraction: the paradox of contingency

Working through Doñana forced open a first abstraction, and it began with how we handled path dependence itself. Rather than adopt it as a settled idea, we traced its conceptual phylogeny across disciplines—from the economic history of David (1985) and Arthur (1989, 1994), through the historical institutionalism of North (1990, 2005) and the historical sociology of Mahoney (2000), to political science (Pierson 2000) and organisation studies (Vergne and Durand 2010)—treating the concept as a lineage whose branches carry different, sometimes incompatible, assumptions about history, agency and equilibrium (Méndez et al. 2019, Méndez 2021). Objectifying path dependence this way exposed the problem at its core. Much of what the system did looks irrational against the efficiency baseline of neoclassical economics—rice grown in biophysically unsuitable conditions, hydraulic megaprojects pursued in the knowledge they were bound to fail. Path-dependence theory files such outcomes under “contingency”, yet, following Mahoney (2000), this creates an epistemological paradox: the lock-in is explained by the very neoclassical feedback mechanisms that the initial, “random” events contradict. Contingency, we came to see, is not one thing but a continuum stretching from pure randomness to full predictability (Eagle 2005). The task, then, is to modulate contingency—to treat an outcome that looks random under one framework as merely unpredictable, and then render it predictable under a richer one (Méndez et al. 2019). Used this way, contingency becomes a heuristic and an early warning that contextual or endogenous factors, invisible to the standard model, shaped choices deep in the past. The corollary is that a multiple-case design might reveal those factors to be less contingent than they appear—recurring across cases with enough regularity to count as structural rather than accidental.

6. A second abstraction: ergodicity

Beneath contingency lay a deeper and rarely examined assumption: ergodicity. To call a system ergodic is to assume that the average of a single trajectory over time equals the average across all its possible states (Peters 2019)—that history, as sequence, can be set aside, and that the past reliably forecasts the future. Social-ecological systems are not like that. They are non-ergodic: shaped by endogenous novelty, irreversibility, and the fact that we cannot rewind and replay the tape of life. Our concern is that mainstream economic models—including those steering global responses to environmental change—often carry a hidden ergodic view of the world (Crona et al. 2021), one that can quietly serve business-as-usual interests in a world that does not behave ergodically. Making ergodicity explicit, in dialogue with the emerging fields of ergodicity and complexity economics (Arthur et al. 2020, Peters and Adamou 2022), seems to me one of the more consequential things social-ecological research can do (Méndez 2021).

7.  The IAD as a metatheoretical device

These threads converge in the program’s most abstract move: developing the IAD beyond a framework into a metatheoretical device. Metatheory means examining the philosophy behind our theories, frameworks and models (Overton 2013)—the assumptions that quietly decide which questions we can even ask. Distilling two decades of Doñana into a set of middle-range generalisations (Merton 1949), I am now working on a proposition for a diagnostic device that aggregates the case’s pluralistic insights into three dimensions—historical, rational and power—and pairs them with reflective questions and visual heuristics. Its purpose is not to add another framework but to prompt analysts to surface the assumptions about ergodicity, contingency, rationality and power that underlie whatever theory they are already using. There is a fitting symmetry here: the IAD was itself first offered as a metatheoretical synthesis of institutional approaches (Kiser and Ostrom 1982), and this is where I return it to that vocation—as a canonical case whose deepest lesson is that seemingly messy, irrational outcomes become legible once we make our own analytical foundations explicit.

8. A spin-off iteration: blue uncertainty

The case's tools eventually travelled beyond the case. When COVID-19 struck, I recognised in the pandemic the same epistemic predicament Doñana had trained me to see: a world facing radical uncertainty (Kay and King 2020)—what we simply do not know, and cannot describe probabilistically—being managed as if it were calculable risk. In a Perspective written amid the crisis (Méndez 2021), I exported two of the program's instruments to the problem of Anthropocene risk (Keys et al. 2019): the adaptive inference protocol (Holling and Allen 2002), repurposed as a learning device for policy under deep uncertainty, and modulating contingency (Méndez et al. 2019), as an epistemic guard against premature assessments of the unexpected. To these I added a third, philosophical component, coining the term blue uncertainty: the residue of randomness inherent in the world that even the best blue-sky science may never eliminate from our causal theories and models (based on Mahoney 2000)—an invitation to keep solving the puzzles that threaten social and ecological wellbeing while staying honest about our fallibility. Together the three components—philosophical (blue uncertainty), methodological (adaptive inference) and epistemic (modulating contingency)—form an intermediary approach designed to resonate with the prediction-and-control mindset of current policy making, and so ease complexity thinking into it. As a spin-off, it closed a loop: what began as a way of understanding one trapped estuary became a stance toward systemic risk at planetary scale.

Concluding remarks: toward a sustainability transformation in Doñana

There is, first, reason for cautious optimism. Doñana is caught in a rigidity trap, not the far deeper lock-in trap it has so far avoided—it still retains much of its structural and natural capital, and with it a high potential for change (Méndez et al. 2012). The scope for a managed transformation is real; the question is how to open it.

If the trap has a keystone, it is a power-counteraction mechanism that works as a double-edged sword: any actor moving first—to dredge, to build, to intervene—is swiftly checked by another wielding the same legal, discursive and institutional weapons (Méndez et al. 2023a). This has so far averted catastrophic interventions and a fall into lock-in, but it equally blocks the coordinated collective action a sustainability pathway would demand. Moving beyond the stalemate—toward actors genuinely “working with nature”—means accepting greater short-term risk: relinquishing power-grabbing and economic-maximisation goals in favour of mutually agreed development limits and biophysical thresholds.

Transformation can build on what already holds the line. Further deterioration is presently averted by a rich legal legacy, reactive judicial rulings, integrated knowledge co-production, and watchful local and international actors (Méndez et al. 2022)—assets to be strengthened, not assumed. Crises have repeatedly opened windows of opportunity, too: the 1998 Los Frailes spill let independent actors introduce participatory, learning-by-doing restoration, even if inertia has too often closed such windows again. And the region holds a deep reservoir of latent pluralism—decades of participatory and transdisciplinary initiatives and alternative visions, repeatedly cancelled out by dominant discourses. Reactivating it calls for shielded spaces of governance experimentation that build trust across fragmented, competing agencies, agreement on shared goals and functioning models, open decisions to public participation, and cultivate not only first-order learning in routines but the harder second-order learning that confronts entrenched values—while disrupting discourse inertia directly (Méndez et al. 2012, 2022).

Underlying all of this is a warning the programme sounds in unison: beware the panacea. Command-and-control and optimal, blueprint “fixes” are precisely what built the trap; a social-ecological system, understood as a complex adaptive system, is ill-served by final solutions blind to its uncertain, stealthy dynamics (Méndez et al. 2023a). What it needs instead are diagnostic, context-sensitive remedies reached through learning-by-doing—with analysts and actors alike treated, in the IAD's phrase, as fallible learners able to learn from error.

Finally, the case leaves an epistemic discipline that is also an ethical one. When we leave our assumptions about contingency, ergodicity and rationality implicit, and dismiss awkward outcomes as irrational or random, we misdiagnose—and flawed diagnoses breed institutional misfits and unjust policies that burden local communities with responsibilities and costs they did not create (Méndez et al. 2023b). Getting the analysis right is therefore a matter not only of rigour but of fairness. In the spirit of blue uncertainty (Méndez 2021), the task is to keep working the puzzle—surfacing the assumptions that quietly decide which futures we can imagine—while staying honest about the limits of what we can know. These lessons come from a single, densely documented case; testing how far they travel is the next step.

Limitations and next steps

Two limitations bound these claims. First, they are abstracted from a single, if densely documented, case: how far the lessons and the diagnostic device travel awaits a carefully designed multiple-case comparison that honours both cross-case comparability and the place-based peculiarities of other settings (Méndez et al. 2023b). Second, the device foregrounds history, rationality and power but leaves two dimensions for later work—a cultural one (culture as external attribute, socially embedded property and evolutionary process) and a cognitive one (the interplay of rationality, perception and cognition beyond biases and heuristics)—both of which the IAD's whole-systems, "holonic" design is well placed to accommodate. Pursuing these is the natural next step, and one this research is already taking up through comparative work within a long-term, multi-site research infrastructure.

A cumulative, shared endeavour 

Taken together, the streams and insights above are why I treat Doñana as a canonical case. It is a place, a history and a conflict; it is also an argument: that patient, pluralistic, multi-level institutional analysis can convert what looks like chaos into explanation, and that doing so responsibly obliges us to examine the assumptions we bring to the work.

The Doñana case also informs the Doñana LTSER Platform within the eLTER Research Infrastructure. Two decades of work on the region's governance, institutions and power supply the long-term, place-based social-ecological knowledge such a platform is meant to steward — complementing its biophysical monitoring, demonstrating a transdisciplinary method, and grounding research in the actors and conflicts that make it policy-relevant. None of this stands alone.

This synthesis is my own high-level reading of a body of work built collaboratively over many years. All the Doñana case's iterations were developed with colleagues and rest on the broader Doñana long-term social-ecological research community. Where I have compressed nuance to hold the arc together, the original publications carry the full detail. This work is cumulative and informed by the work of many others (social-ecological scientists, ecological economists, geographers, anthropologists, historians...) who have studied Doñana's dense, complex history and its human-nature interactions—whose scholarship this research draws on and remains indebted to.

References

Allison, H. E., and R. J. Hobbs. 2004. Resilience, adaptive capacity, and the “lock-in trap” of the Western Australian agricultural region. Ecology and Society 9(1):3.

Arthur, W. B. 1989. Competing technologies, increasing returns, and lock-in by historical events. The Economic Journal 99(394):116–131.

Arthur, W. B. 1994. Increasing returns and path dependence in the economy. University of Michigan Press, Ann Arbor, Michigan, USA.

Arthur, W. B., E. D. Beinhocker, and A. Stanger, editors. 2020. Complexity economics. Santa Fe Institute Press, Santa Fe, New Mexico, USA.

Clement, F. 2010. Analysing decentralised natural resource governance: proposition for a “politicised” Institutional Analysis and Development framework. Policy Sciences 43:129–156.

Clement, F., and J. M. Amezaga. 2013. Conceptualising context in institutional reforms of land and natural resource management: the case of Vietnam. International Journal of the Commons 7(1):140–163.

Crona, B., C. Folke, and V. Galaz. 2021. The Anthropocene reality of financial risk. One Earth 4(5):618–628.

David, P. A. 1985. Clio and the economics of QWERTY. The American Economic Review 75(2):332–337.

David, P. A. 1994. Why are institutions the “carriers of history”? Path dependence and the evolution of conventions, organizations and institutions. Structural Change and Economic Dynamics 5(2):205–220.

Eagle, A. 2005. Randomness is unpredictability. The British Journal for the Philosophy of Science 56(4):749–790.

Gunderson, L. H., and C. S. Holling, editors. 2002. Panarchy: understanding transformations in human and natural systems. Island Press, Washington, D.C., USA.

Holling, C. S., and C. R. Allen. 2002. Adaptive inference for distinguishing credible from incredible patterns in nature. Ecosystems 5(4):319–328.

Holling, C. S., L. H. Gunderson, and G. D. Peterson. 2002. Sustainability and panarchies. Pages 63–102 in L. H. Gunderson and C. S. Holling, editors. Panarchy. Island Press, Washington, D.C., USA.

Holling, C. S., and G. K. Meffe. 1996. Command and control and the pathology of natural resource management. Conservation Biology 10(2):328–337.

Kay, J., and M. King. 2020. Radical uncertainty: decision-making beyond the numbers. W. W. Norton, New York, New York, USA.

Keys, P. W., V. Galaz, M. Dyer, N. Matthews, C. Folke, M. Nyström, and S. E. Cornell. 2019. Anthropocene risk. Nature Sustainability 2:667–673.

Kimmich, C., E. Baldwin, E. Kellner, C. Oberlack, and S. Villamayor-Tomas. 2022. Networks of action situations: a systematic review of empirical research. Sustainability Science.

Kiser, L., and E. Ostrom. 1982. The three worlds of action: a metatheoretical synthesis of institutional approaches. Pages 179–222 in E. Ostrom, editor. Strategies of political inquiry. Sage, Beverly Hills, California, USA.

Mahoney, J. 2000. Path dependence in historical sociology. Theory and Society 29(4):507–548.

McGinnis, M. D. 2011. Networks of adjacent action situations in polycentric governance. Policy Studies Journal 39(1):51–78.

Méndez, P. F., N. Isendahl, J. M. Amezaga, and L. Santamaría. 2012. Facilitating transitional processes in rigid institutional regimes for water management and wetland conservation: experience from the Guadalquivir estuary. Ecology and Society 17(1):26.

Méndez, P. F., J. M. Amezaga, and L. Santamaría. 2019. Explaining path-dependent rigidity traps: increasing returns, power, discourses, and entrepreneurship intertwined in social-ecological systems. Ecology and Society 24(2):30.

Méndez, P. F. 2021. Blue uncertainty: warding off systemic risks in the Anthropocene—lessons from COVID-19. Earth System Governance 7:100101.

Méndez, P. F., D. Fajardo-Ortiz, and J. M. Holzer. 2022. Disrupting the governance of social-ecological rigidity traps: can pluralism foster change towards sustainability? Pages 243–291 in J. M. Holzer, J. Baird, and G. M. Hickey, editors. Pluralism in ecosystem governance. Advances in Ecological Research, Volume 66.

Méndez, P. F., F. Clement, G. Palau-Salvador, R. Díaz-Delgado, and S. Villamayor-Tomas. 2023a. Understanding the governance of sustainability pathways: hydraulic megaprojects, social-ecological traps, and power in networks of action situations. Sustainability Science 18:303–321.

Méndez, P. F., M. Adamescu, B. Bertsch-Hörmann, F. Clement, M. Cohen, R. Díaz-Delgado, J. Dick, D. Fajardo-Ortiz, S. Gaba, V. Gaube, L. Halada, J. M. Holzer, Z. Izakovičová, D. E. Orenstein, E. Oteros-Rozas, G. Palau-Salvador, I. T. Rosário, and S. Villamayor-Tomas. 2023b. Averting lock-in risks in social-ecological systems: a roadmap for pluralistic research and governance. SocArXiv preprint.

Merton, R. K. 1949. Social theory and social structure. The Free Press, New York, New York, USA.

Molle, F., P. P. Mollinga, and P. Wester. 2009. Hydraulic bureaucracies and the hydraulic mission: flows of water, flows of power. Water Alternatives 2(3):328–349.

Morrison, T. H., W. N. Adger, K. Brown, M. C. Lemos, D. Huitema, J. Phelps, L. Evans, P. Cohen, A. M. Song, R. Turner, T. Daw, and R. L. Hughes. 2019. The black box of power in polycentric environmental governance. Global Environmental Change 57:101934.

North, D. C. 1990. Institutions, institutional change and economic performance. Cambridge University Press, Cambridge, UK.

North, D. C. 2005. Understanding the process of economic change. Princeton University Press, Princeton, New Jersey, USA.

Ostrom, E. 2005. Understanding institutional diversity. Princeton University Press, Princeton, New Jersey, USA.

Overton, W. F. 2013. Relationism and relational developmental systems: a paradigm for developmental science in the post-Cartesian era. Pages 21–64 in R. M. Lerner and J. B. Benson, editors. Advances in Child Development and Behavior, Volume 44.

Peters, O. 2019. The ergodicity problem in economics. Nature Physics 15:1216–1221.

Peters, O., and A. Adamou. 2022. The ergodicity solution of the cooperation puzzle. Philosophical Transactions of the Royal Society A 380:20200425.

Pierson, P. 2000. Increasing returns, path dependence, and the study of politics. American Political Science Review 94(2):251–267.

Schmidt, V. A. 2008. Discursive institutionalism: the explanatory power of ideas and discourse. Annual Review of Political Science 11:303–326.

Vergne, J.-P., and R. Durand. 2010. The missing link between the theory and empirics of path dependence: conceptual clarification, testability issue, and methodological implications. Journal of Management Studies 47(4):736–759.

lifelong learning = flow + emotion + epistemic humility