A synthesis of 423 studies and more than 222,000 data points, published in Nature Ecology & Evolution on 6 October, challenges the comforting idea that ecosystems have plenty of redundant species — and finds that the ocean’s capacity to store carbon may depend on biodiversity most of all.
Ecologists have long debated how many species an ecosystem really needs. One influential idea, functional redundancy, holds that many species do similar jobs, so that losing a few makes little difference: another pollinator, another decomposer, another predator steps in. If true, it would offer a degree of reassurance in an age of rapid biodiversity loss. A new study suggests that reassurance has been overstated.
Researchers led by King’s College London and Imperial College London, working with the Natural History Museum, combined data from 423 published studies — 222,829 data points across land, freshwater, marine and estuarine ecosystems — into what they describe as the largest database of its kind, more than twice the size of the largest previous effort. The paper, “Biodiversity safeguards ecosystem services and functions worldwide”, appeared in Nature Ecology & Evolution on 6 October 2026, with Emma R. Moffett as first author and William D. Pearse as senior author.
Why scale matters
Earlier evidence on this question came largely from individual field experiments — plots of grassland sown with one, two, four or sixteen species, for example — each limited to a single habitat and a handful of measurements. Such experiments were influential but left room for argument about whether their results applied to forests, rivers, coral reefs or farmland. Pooling hundreds of studies across very different ecosystems, as this team did, allows researchers to ask whether a pattern holds generally rather than in one carefully controlled setting. Meta-analyses have their own weaknesses: they inherit the biases of the studies they combine, and some ecosystems and regions are far better studied than others. The authors’ own caveat about ocean data illustrates the point. But the breadth of the dataset makes the central conclusion considerably harder to dismiss.
Steady gains, not a plateau
The team examined 23 categories of ecosystem services and functions — the benefits people receive from nature, such as pollination, water purification, carbon storage and natural pest control. If redundancy were strong, these benefits would rise quickly as the first few species were added and then level off. Instead, most rose steadily as biodiversity increased. In other words, every additional species tended to add something, and every loss tended to subtract something.
The finding matters because policy often implicitly assumes a plateau: that protecting a “representative” subset of species is enough to keep ecosystems functioning. The new analysis suggests that, for most services, there is no safe margin of species that can be lost without cost.
The ocean surprise
The strongest relationship in the entire dataset involved ocean carbon sequestration — the capture and storage of carbon dioxide in marine waters and sediments. Blue-carbon systems, from phytoplankton communities to the microbial food webs that transport carbon into the deep sea and the coastal habitats such as mangroves and salt marshes that bury it, appeared to depend on diverse marine communities more than any other service measured. “The ocean result stopped us in our tracks,” said Dr Moffett, noting that it is also one of the least studied areas the team found.
The authors are careful to flag a limitation: the ocean carbon result rests on a small number of datasets, and they identify it as an urgent knowledge gap rather than a settled conclusion. That caution is important, since blue carbon increasingly features in national climate plans and carbon-credit schemes.
Important exceptions
Not every service followed the pattern. Protection against hazards such as coastal flooding and erosion proved relatively insensitive to overall biodiversity, because it often depends on one or two foundational species — for example, the shrubs that stabilise sand dunes. The authors argue that conservation must therefore do two things at once: maintain overall diversity and safeguard irreplaceable foundational species.
Looking ahead
The team linked its database to biodiversity projections under the socioeconomic pathways used by the Intergovernmental Panel on Climate Change (IPCC). Their model predicts that biological pest control on farmland — a free service provided by the natural enemies of crop pests — declines under a fossil-fuel-intensive development path compared with lower-fossil-fuel paths, with the sharpest losses in countries with rapid population growth and lower levels of development. “We can now predict the benefits biodiversity provides anywhere,” said Dr Pearse. The full database and model forecasts have been made publicly available.
Policy context
The findings arrive as governments work towards the targets of the Kunming–Montreal Global Biodiversity Framework, adopted in 2022, which commits countries to protect 30 per cent of land and sea by 2030 and to restore degraded ecosystems. Much of that effort is measured in area protected. The new evidence suggests that what is protected — and how much of the species diversity within it survives — matters as much as the number of hectares.
Relevance for India
India is one of the world’s megadiverse countries and also one where agriculture depends heavily on free ecosystem services such as pollination and natural pest control. The projection that pest-control services decline most in fast-growing, lower-income countries is directly relevant to Indian farm policy, where pesticide use, monocropping and habitat loss interact. The blue-carbon finding is pertinent to India’s mangrove-restoration commitments, including along the Godavari and Krishna deltas in Andhra Pradesh. As a meta-analysis, the study identifies broad patterns rather than local prescriptions, but its open database gives Indian researchers and planners a new tool for estimating what is at stake in specific landscapes.
– Vamsi Priya Potharaju
Key facts
– Journal/date: Nature Ecology & Evolution, 6 October 2026 (DOI: 10.1038/s41559-026-03200-4)
– Data: 423 studies; 222,829 data points; 23 ecosystem-service categories
– Main finding: most services rise steadily with biodiversity; functional redundancy overestimated
– Strongest link: ocean carbon sequestration (based on few datasets — flagged as a knowledge gap)
– Exception: hazard protection depends on a few foundational species
