- A recent horizon scan led by William J. Sutherland shifts conservation thinking away from visible damage toward emerging developments that could shape biodiversity outcomes over the next decade, even if they have not yet hardened into crises.
- The fifteen issues identified span technology, climate, biology, and finance, with a particular emphasis on computational advances that could expand monitoring and modeling while also narrowing what can later be revisited or challenged.
- Alongside technological change, the scan highlights physical, institutional, and biophysical pressures, from drone-related plastic pollution and new forest finance mechanisms to drying soils, darkening oceans, and abrupt shifts in the Southern Ocean.
- The authors also situate these risks against two background constraints already underway—eroding environmental data systems and tightening conservation finance—and, looking back ten years, argue that the value of horizon scanning lies less in prediction than in improving preparedness before change becomes costly.

Conservation debates are usually framed by damage already visible. Forests are cleared, fisheries decline, protected areas invaded, and budgets cut. Developments that have not yet hardened into crises tend to escape sustained attention, partly because they are unfamiliar and partly because they fall between established fields. A recent horizon scan led by William J. Sutherland of Cambridge University and published in Trends in Ecology & Evolution sets out to correct that imbalance by asking a strategic question: which emerging changes, still poorly understood, are most likely to shape biodiversity outcomes over the next decade?
The exercise is not an attempt at prediction. It functions more like a risk register for those who prefer early warning to surprise. The fifteen issues identified span technology, climate dynamics, biology, and finance. None is certain; all carry consequences at scale.
Several of the most consequential developments stem from advances in computation. Tiny machine-learning systems, designed to run on minimal power without internet access, promise to extend ecological monitoring into places long beyond the reach of conventional data collection. That matters for conservation, which still relies heavily on observation in remote and underfunded regions. That efficiency is not without cost. When models process information on site and discard what they do not classify as relevant, the opportunity for later reanalysis is lost. Intelligence becomes cheaper, while opportunities for later scrutiny diminish.
Related gains are emerging from optical AI chips that use light rather than electricity. These technologies could sharply reduce the energy and water demands of data processing, easing one constraint on the expansion of conservation analytics. History suggests efficiency tends to widen use, not contain it. Faster, cheaper computation may enable better monitoring while also encouraging broader deployment of energy-intensive digital infrastructure elsewhere.
The same tension runs through the rise of digital twins, high-resolution simulations of ecosystems, cities, and even the climate itself. These models promise tighter integration of disparate data streams and more sophisticated scenario testing. They also risk concentrating authority in systems that few users fully understand. When decisions depend on opaque models, failure can delay action rather than inform it.
Some emerging threats are more physical. Fiber-optic drones, developed rapidly in recent conflicts, trail kilometers of plastic cable that is often abandoned once missions end. In landscapes already destabilized by war or weak governance, these strands accumulate, entangling wildlife and fragmenting habitat. The technology is spreading faster than any serious effort to mitigate its environmental impacts.
Other shifts are institutional. Brazil’s proposed Tropical Forests Forever Facility aims to provide long-term, performance-based finance to countries that keep deforestation low. It could stabilize forest protection where donor funding has proved intermittent. A poor outcome may reinforce the idea that forests matter only when they can be narrowly treated as financial assets.

Unexpected feedbacks between human health and land use also feature. The rapid uptake of appetite-suppressing drugs has begun to alter food consumption patterns, particularly demand for beef and highly processed foods. Sustained shifts of this kind could ease pressure on pasture expansion and cropland at the margins. Whether that translates into conservation gains will depend less on individual diets than on how global commodity markets respond.
Climate-driven changes recur across the scan. New analyses suggest global soils are drying faster than previously recognized, with implications for ecosystems, agriculture, and even sea-level rise. At sea, declining light penetration across large areas of the ocean may be altering primary productivity in ways that are only beginning to be measured. Meanwhile the Southern Ocean, long treated as relatively stable, is showing signs of abrupt change that could ripple through global circulation and marine ecosystems.
Biological interventions are advancing as well. Chemicals that can delay or reset flowering times may help crops and rare plants cope with erratic seasons, though ecological side effects remain uncertain. Commercial soil inoculants marketed as beneficial fungi are spreading rapidly despite weak evidence that they deliver promised gains. Once applied at scale, their influence on soil communities may be difficult to reverse.
A few proposals that would once have been dismissed outright are no longer implausible. Microbes that convert plastic waste into edible protein could reduce pollution and pressure on farmland if they can be deployed safely. At the same time, synthetic “mirror” biomolecules, designed to resist degradation, raise uncomfortable questions about biosecurity should they ever escape controlled settings.
The analysis is shaped by two background pressures the authors deliberately set aside because they are already underway. One is the erosion of environmental data infrastructure, from satellite programs to field monitoring, which risks making emerging changes harder to detect just as systems grow more complex. The other is geopolitical strain on conservation finance, as aid budgets contract and priorities shift. These forces do not qualify as emerging issues precisely because they are no longer speculative. They nonetheless shape the fate of every issue identified.
The paper’s value is sharpened by a retrospective glance. Looking back ten years to the issues flagged in 2016, the authors note that some developments once considered remote have arrived faster than expected. Artificial intelligence, then a peripheral concern for conservation, is now embedded in routine monitoring and analysis. Other fears, such as unregulated fishing in the central Arctic, were partially averted through international agreements. Preparedness, rather than accuracy, is what horizon scanning is meant to improve.
That modest ambition may be its strength. Conservation has often been forced into reactive mode, responding to damage once it is visible and costly. By drawing attention to plausible but under-discussed shifts, the scan narrows the gap between novelty and response. In a field where delay is usually expensive, that alone may justify the exercise.
Citation:
- Sutherland, William J. et al. A horizon scan of biological conservation issues for 2026. Trends in Ecology & Evolution. Published online December 2, 2025. DOI: 10.1016/j.tree.2025.10.016