A paradox runs through contemporary environmental policy: biodiversity is invoked everywhere, yet rarely fully understood.
In recent years, its centrality has grown significantly, both in research and in policy frameworks. And yet, as Zhou et al. (2026) observe, it remains “the least understood and least integrated into decision-making processes.” It is a constant presence in public discourse, but still elusive when it comes to translating it into concrete action.
At this point, the question becomes unavoidable: how can biodiversity be measured in a way that is useful for decision-making?
A first answer emerges from a highly applied line of research closely linked to the evolution of European policies. In particular, work on farmland biodiversity indicators stems from a specific need: to make biodiversity an integral component of the Common Agricultural Policy, no longer as an abstract principle but as an operational variable.
A recent study published in Ecological Indicators (2026) moves precisely in this direction. Through an extensive literature review, the authors identify and assess over one hundred indicators, with a clear objective: to determine which can effectively function within policy frameworks.
The key step is methodological. Indicators must be both ecologically sound and politically usable. To achieve this, the study bridges two domains that rarely fully intersect: on one side, Essential Biodiversity Variables, which ensure scientific coherence; on the other, the European Commission’s RACER framework, which evaluates an indicator’s relevance, reliability, and applicability.
From this intersection emerges a set of tools aimed at translating biodiversity into something observable and comparable: satellite-based vegetation data, soil DNA analysis, acoustic monitoring of birds, and metabarcoding techniques for insects. This represents a significant shift, making it possible to integrate biodiversity into public policy evaluation mechanisms in ways that were previously difficult.
In other words, biodiversity begins to take a form that can be measured — and therefore governed.
And yet, it is precisely here that a first critical tension appears.
Every indicator is, by definition, a simplification. And every simplification entails a loss. Tools that are easier to implement, such as satellite-based indicators, allow for broad and continuous coverage, but often function as proxies: they do not measure biodiversity directly, but capture only certain dimensions of it. By contrast, more sophisticated methods, such as environmental DNA analysis, provide far deeper insights, but are costly, complex, and difficult to scale.
This imbalance is not a minor technical issue, but a structural feature of the problem. The study itself acknowledges this clearly: there is no perfect indicator, only possible combinations, each with its own limitations.
Here, a deeper tension becomes visible: biodiversity is not difficult to protect only because it is complex, but because it is difficult to represent reliably.
While this first line of research seeks to develop increasingly effective tools, a second, more recent and more radical perspective calls the entire framework into question.
In Towards a Unified Framework for Biodiversity Action in the Triple Planetary Crisis (Zhou et al., 2026), the problem is reframed at a different level. The issue is no longer simply improving indicators, but interrogating the system within which they are meant to operate.
According to the authors, biodiversity monitoring suffers from deep fragmentation. Data on species and data on environmental pressures—such as pollution, climate change, and land use—are often collected separately, across different spatial and temporal scales. Scientific disciplines continue to operate largely in isolation. Most importantly, existing models struggle to capture the interactions among these factors.
This point is decisive. Biodiversity is not the outcome of a single driver, but of a network of pressures that overlap and reinforce one another. Reducing it to a set of indicators risks losing precisely what defines it: its systemic nature.
At this stage, the paradox becomes more apparent.
On the one hand, research on indicators demonstrates real and necessary progress. On the other, there is a growing awareness that biodiversity may not be fully reducible to simple metrics.
Zhou et al. highlight three recurring limitations: the misalignment between biological and environmental data, the lack of truly representative shared metrics, and—perhaps most critically—the limited predictive capacity of current models. It is not only that we struggle to understand what is happening; we also struggle to anticipate what will happen next.
This shifts the problem. It is no longer simply a matter of measuring better, but of understanding better.
Within this perspective, the very definition of biodiversity begins to change.
It is no longer only a matter of quantity—how many species, how much abundance—but of relationships: interactions among species, ecosystem stability, and adaptive capacity. In other words, resilience.
This implies a shift from a static to a dynamic perspective. It is not enough to capture a snapshot of an ecosystem; we need to understand how it evolves over time and how it responds to pressures.
It is here that new tools come into play, attempting to integrate diverse datasets and interpret biodiversity as a complex system. Rather than isolated indicators, the focus shifts toward networks, models, and correlations.
One of the most promising developments is the integration of traditional ecological models with artificial intelligence techniques. The idea is to combine the theoretical robustness of the former with the ability of the latter to detect complex patterns in large datasets.
This hybrid approach opens up a new possibility: not only describing the present, but anticipating the future. Biodiversity thus becomes not only something to monitor, but something to forecast.
This is a significant shift, moving the focus from reaction to anticipation.
At this point, the issue inevitably returns to the realm of policy.
Integrating biodiversity into decision-making systems means embedding it within economic models, reporting frameworks, and corporate strategies. In essence, it means making biodiversity part of the language through which decisions are made.
Yet this transition is delicate. If indicators are fragile or incomplete, the risk is that decisions will be built on partial foundations. In the worst cases, this may lead to forms of greenwashing, where measurement becomes more a tool of legitimization than of understanding.
Bringing these two perspectives together reveals a tension that cannot be resolved.
Without indicators, biodiversity remains outside decision-making processes. But every indicator, inevitably, simplifies.
The challenge, therefore, is not to choose between complexity and simplicity, but to develop tools capable of operating between the two—tools that allow action without claiming to exhaust the reality they describe.
Biodiversity ultimately challenges our way of knowing.
It forces us to confront something that must be measured in order to be governed, yet cannot be fully reduced to measurement. The issue, then, is perhaps not to find the perfect indicator, but to learn how to use indicators without confusing them with reality.
Not as instruments that fully capture complexity, but as tools that help us navigate within it. In this sense, other attempts to quantify environmental limits—such as the calculation of Earth Overshoot Day—face similar approximations, yet still retain their ability to provide meaningful guidance.
It is within this space—between what we can measure and what inevitably escapes us—that something deeper is at stake: the way we choose to inhabit the relationship between economy, nature, and the future.
References