A global view of what nature provides
A major international synthesis has found that biodiversity supports a wide range of ecosystem functions and services, from pollination and water quality to carbon storage and food production. The research brings together evidence from 423 studies to examine not only whether diversity matters, but how strongly different services depend on it.
Published in Nature Ecology & Evolution on 6 October 2026, the analysis offers a broad quantitative view of the links between the variety of living organisms and the benefits ecosystems provide to people. It also reveals that the strength of those relationships differs substantially across ecological processes.
The findings are particularly relevant to conservation policy because protecting biodiversity is sometimes discussed separately from economic or social priorities. In reality, many services that support livelihoods depend on functioning communities of organisms.
Yet the authors caution against treating all services as interchangeable. Some show consistent gains as biodiversity rises, while others depend more heavily on the survival of a relatively small number of distinctive species.
How researchers assembled the evidence
Emma R. Moffett and colleagues searched the scientific literature for studies directly measuring relationships between biodiversity and ecosystem functions or services. Their initial database search returned 312,291 publications.
Using a stratified approach to identify relevant work across ecosystems and regions, they screened a more manageable selection of papers and extracted quantitative measurements where available.
The final analytical dataset contained 222,829 data points from 1,959 datasets drawn from 423 unique papers. This distinction matters: the researchers did not conduct 222,829 separate experiments, and several datasets could come from the same published study.
The synthesis combined observational and experimental evidence from terrestrial, freshwater, marine and estuarine ecosystems. It also included different ways of measuring biodiversity, such as species richness and functional or taxonomic group diversity.
Data were standardised before the team compared the forms and strengths of biodiversity relationships across ecosystem-service categories. This allowed studies using different units and measurements to contribute to a common analytical framework.
The researchers assessed 23 categories of ecosystem functions and services, explored how relationships varied by ecosystem and study design, and developed a scenario-based approach to investigating potential future changes in pest regulation.
Importantly, a global synthesis is not the same as a single controlled experiment. It draws strength from combining many settings but must account for differences in methods, locations and measurement quality.
What biodiversity means for ecosystem services
Ecosystem functions are ecological processes, such as plant production or nutrient cycling. Ecosystem services are the benefits people receive from those processes, including clean water, pollination and food provision.
The boundary between the two is not always sharp. Plant production, for example, can be considered a function when studied as an ecological process, or a service when linked to food or climate regulation.
Researchers distinguish these concepts to avoid implying that every ecological measurement directly translates into an immediate human benefit. A change in a measured function may affect people only through a chain of ecological and economic relationships.
Biodiversity can influence services through several mechanisms. Different species may use resources in complementary ways, maintain activity under different conditions, or provide insurance when one species declines.
Some organisms are also unusually important because they perform a function that few others can replace. The study found evidence consistent with both broad diversity effects and the importance of distinctive foundational species.
Linear relationships were surprisingly common
The analysis compared different mathematical shapes linking biodiversity with ecosystem performance. A linear relationship implies that increases in diversity continue to be associated with gains across the observed range.
A saturating relationship suggests that benefits increase initially but eventually level off, potentially because additional species perform roles already represented in the community.
Across 23 categories, quadratic relationships were most prevalent in 11 categories and linear relationships in nine. Only two categories were best described by a logarithmic model.
These counts indicate that a simple assumption of rapid saturation does not describe all ecosystem services. For several important processes, losing additional species may continue to reduce performance rather than having negligible effects after a particular threshold.
Regulating services such as water quality and pollination frequently showed linear relationships. This suggests that diversity may contribute incrementally to those functions across a broad range of conditions.
The pattern does not mean that every species has precisely the same ecological value or that a fixed number of species will deliver an identical gain everywhere. The researchers examined statistical patterns across heterogeneous studies.
Some services were better described by curves showing diminishing gains. Understanding these differences can help conservation planning move beyond a single rule about how many species are enough.
The strongest reported effect involved ocean carbon
The researchers also estimated standardised effect sizes, allowing them to compare the strength of biodiversity relationships across categories even when the underlying mathematical forms differed.
Most ecosystem functions and services were positively associated with greater biodiversity. However, the size of those relationships varied markedly.
Oceanic carbon sequestration showed a particularly strong positive association, with a reported Fisher’s z-transformed correlation effect size of 1.48. By comparison, the effect for regulation of air quality was approximately minus 0.03, indicating a very weak overall relationship in that category.
These effect sizes are statistical measures, not percentages. A value of 1.48 does not mean that biodiversity increased carbon storage by 148 percent.
The marine result is potentially important because ocean organisms contribute to the biological processes that move carbon through marine food webs and into deeper waters.
However, the authors explicitly warned that oceanic carbon sequestration was represented by only seven datasets, compared with 154 for terrestrial carbon sequestration. That small marine evidence base makes the unusually large estimate uncertain.
Consequently, the study supports closer attention to marine biodiversity, but it does not justify a precise global prediction of carbon loss based on the effect size alone.
Some services depend on particular species
Hazard regulation showed comparatively weak overall dependence on broad biodiversity measures. The authors suggested that some protective functions may rely disproportionately on a small number of functionally distinctive organisms.
For example, the structural role of a habitat-forming species may matter more for a particular local protective function than the total number of species counted in a survey.
That does not mean surrounding biodiversity is unimportant. Foundational species often depend on other organisms through food webs, reproduction and habitat interactions.
The finding instead points to a dual conservation objective: protect overall diversity while also identifying species whose ecological roles are especially difficult to replace.
Policies based only on species counts could overlook those functional differences. Equally, policies focused exclusively on a few prominent species may neglect the wider communities needed to sustain them.
What the evidence says about study design
The dataset included 1,079 observational datasets and 879 experimental datasets. These approaches answer related but different questions.
Experiments can isolate mechanisms by manipulating biodiversity under controlled conditions. Observational studies examine existing communities and may capture broader ecological complexity and longer histories.
The researchers found a modest overall difference between study types, with observational studies showing slightly stronger biodiversity effects than experiments. The reported coefficient was 0.03, with a P value of 0.005.
Although statistically detectable, the overall difference was small. It should not be interpreted as evidence that observational studies are universally better or that experiments are unreliable.
Results also varied across services. In observational comparisons, stronger relative effects appeared for multifunctionality and pollination, while some other functions showed weaker patterns.
These differences reinforce the need to consider what was measured, at what scale and under which environmental conditions before applying a single global estimate to a local ecosystem.
Why the findings matter for food security
The synthesis also explored how biodiversity loss could influence future ecosystem services under socioeconomic and land-use scenarios. One application involved pest regulation, a service relevant to agriculture and food production.
Predictions indicated that biodiversity losses may undermine pest regulation in countries experiencing population growth and intensive development.
Such modelling can help identify potential risks before ecological change becomes visible in agricultural output. It does not establish the exact amount of crop loss that any particular country will experience.
The study’s broad message is that conserving biodiversity can contribute to maintaining ecological processes that support food systems. Yet the strength of the relationship depends on the service, region and composition of local communities.
For decision-makers, this argues for combining biodiversity indicators with direct measurements of water quality, pollination, pest pressure and other outcomes relevant to people.
Restoration programmes should also evaluate whether returning species actually improves the functions they are intended to support, rather than assuming that higher richness alone guarantees success.
Limitations of the global evidence
Although 423 studies form a substantial evidence base, the research literature is not evenly distributed around the world. The authors reported a strong concentration of terrestrial studies and research from developed countries.
The final database included 1,759 terrestrial datasets, compared with 183 freshwater, 153 marine and 23 estuarine datasets. These imbalances limit the precision of conclusions for less-studied systems.
The typical spatial scale was also small, with a median measurement grain of one square metre. Relationships measured in plots may not translate directly to landscapes, river basins or national economies.
The literature survey was stratified rather than exhaustive, so some eligible work may not have been captured. Differences in how original researchers defined and measured biodiversity also introduce uncertainty.
The oceanic carbon result is an especially clear example of why sample size matters within a synthesis. A striking effect can be important while still requiring confirmation from more studies.
Modelled future outcomes depend on assumptions about development, land use and biodiversity change. They should be read as scenario-based estimates, not guaranteed forecasts.
Finally, a positive biodiversity-service relationship does not establish that adding species will always restore a damaged service immediately. Ecological recovery can depend on soil conditions, climate, habitat connectivity and the identity of the species involved.
A clearer case for protecting ecosystems
The synthesis provides unusually broad evidence that biodiversity is tied to ecological performance. Its 1,959 datasets and more than 222,000 observations support the conclusion that diversity contributes positively to many functions and services.
At the same time, the study argues against one-size-fits-all conservation targets. Water quality, pollination, carbon storage and hazard regulation do not all respond in the same way to changes in biodiversity.
Protecting nature therefore requires attention to both the number of species and the ecological roles they perform. The strongest policy decisions will be informed by local monitoring and the particular services communities depend upon.
The central conclusion is straightforward: biodiversity is not merely an inventory of plants and animals. It is an important component of the ecological systems that support human well-being, although the benefits and uncertainties vary across environments.
Source Information
Study: Biodiversity safeguards ecosystem services and functions worldwide
Authors: Emma R. Moffett, Joel H. Gayford, Lexi Chen, Olivia F. Morris, Yuming Shi, Lucy Somekh, Nicole Stasik, Andy Purvis, Guy Woodward and William D. Pearse.
Journal: Nature Ecology & Evolution
Published: 6 October 2026
Methods: Global synthesis of 423 papers, 1,959 datasets and 222,829 data points, incorporating experimental and observational research.
