Insight, Biodiversity & Governance
A €40 million European research programme has opened on ecosystems so changed that returning them to their historical state may no longer be possible. That converts restoration from an ecological question into a choice about which future is acceptable, and choices have owners.
The Biodiversa+ partnership has launched a forty million euro research initiative on novel ecosystems: systems transformed so substantially by climate change, invasive species or land use that returning them to their historical state may no longer be realistic (as reported). The agenda explicitly includes socio-ecological consequences, future trajectories, risks, benefits and equity, rather than biodiversity measurement alone, and it covers terrestrial, freshwater, coastal and marine environments.
That framing is a larger conceptual move than it appears, and it lands at an awkward moment, three weeks after Member States submitted their first draft national restoration plans under a regulation whose entire architecture assumes a target state to restore towards.
Daniel Pauly named the underlying problem in a single page in 1995. Writing about fisheries, he described the shifting baseline syndrome: each generation of scientists takes the stock sizes present at the start of their own careers as the natural baseline, so a long decline is repeatedly re-based and the cumulative loss disappears from view.
The concept has travelled well beyond fisheries because the mechanism is general. What counts as normal for a landscape is anchored to living memory, and living memory is short.
Novel ecosystems research inverts the problem rather than solving it. Shifting baselines describe a baseline that slides downward unnoticed. A novel ecosystem is one where the historical baseline is known, is not in dispute, and is simply unreachable, because the temperature, rainfall, species assemblage or hydrology that supported it no longer exist.
Which removes the thing that made restoration look technical. A target drawn from history has the enormous political advantage of appearing to be a fact rather than a preference. Once the historical state is off the table, the target has to be chosen, and a chosen target is a policy with winners and losers rather than a scientific finding with a confidence interval.
A transformed landscape usually has several plausible trajectories, and the differences between them are distributional rather than ecological.
Maximise carbon storage. Favours fast-growing, often non-native, often monocultural systems. Attracts finance readily because the output is measurable and saleable.
Maximise biodiversity under new conditions. May mean accepting species that were never there, including some arriving from further south, and may mean tolerating the loss of a locally iconic species that can no longer survive. Ecologically defensible, politically hard.
Maximise a particular ecosystem service. Flood attenuation for a town, water quality for a catchment, cooling for a city. Benefits an identifiable group, which makes it fundable and contestable in equal measure.
Maximise continuity of livelihood. Keeps the grazing, the fishing, the tourism or the cultural practice going in some form, even if the ecology is not what an ecologist would choose.
Those four produce different landscapes. Nothing in the science selects between them, and the selection is currently being made inside research consortia and agency guidance rather than anywhere with a mandate.
We argued about the national restoration plans that costs are concentrated on identifiable landowners while benefits are diffuse, which predicts organised opposition and passive support. Novel ecosystems add a second difficulty on top of that one.
In conventional restoration, the target is contestable but the direction is not. Everybody is arguing about where and how much, not about what good looks like. A pre-drainage wetland is a shared reference point even for people who do not want one.
Remove that reference point and there is no shared object to argue about. The farmer, the ecologist, the municipality and the angling club are no longer disagreeing about the route to a common destination. They are proposing different destinations, and there is no technical procedure for adjudicating between them.
This is why the equity dimension in the research agenda is not a softening addition. It is the core analytical problem, and it will determine whether anything gets implemented.
The word novel does a great deal of work. Declaring an ecosystem novel is itself a consequential act, because it releases managers from an obligation to restore and opens the field to alternatives. There will be pressure to apply the label where restoration is merely expensive rather than genuinely impossible, and the threshold between those two is a judgement rather than a measurement. Somebody should be tracking who applies the label and on what evidence.
Irreversibility should be tested, not assumed. The claim that a historical state is unreachable is an empirical claim about hydrology, climate envelopes and species availability. It is sometimes right and it is convenient when it is wrong. It warrants the same scrutiny as any other finding that happens to reduce an obligation.
The people who will live with the outcome have knowledge the model does not. What a landscape was like forty years ago, which springs still run, where the fish went, which fields flooded before the drain was cut. That is the corrective to shifting baselines, it sits with older residents rather than in a dataset, and it is disappearing on a schedule.
Who participated in choosing the target, and who was present when the choice was made. Not consultation counts. Whether the trade-off between the four futures above was made explicit to anybody outside the project.
What was given up. Every chosen trajectory forecloses others. Naming the foreclosed options at the point of decision is cheap and almost never done, and it is what makes a decision reviewable later.
Whether the label held. A follow-up, ten years on, asking whether ecosystems declared novel were genuinely unreachable or whether the declaration reflected the budget available at the time.
Local knowledge, recorded before it goes. This is the discovery function that no results framework requests, and in this case it has an expiry date measured in a decade or two.
Restoration policy has been able to avoid the question of what nature ought to be by pointing at what it used to be. That escape route is closing, and the discipline that replaces it will have to be part ecology and part politics, conducted openly rather than inside a work package.
The Lab works on this in climate and ecosystems and through field research with the people who live in a landscape rather than model it.
If you are designing a restoration or novel ecosystem programme and want the choice of target treated as a decision rather than a finding, tell us what you need to know.
This is an independent insight piece by Transitions Lab. For the Lab's applied work, see Climate & Ecosystems. See also Whose Field Becomes a Wetland on who organises when restoration costs are concentrated, The Right That Matters Is to the Tree on which specific right predicts investment in land, and The Benefits Nobody Was Looking For on outcomes no framework asked for. To discuss a study, see Contact.