Editorial NexoScience
A blood sample from a tagged wolf. A feather dropped by a migrating songbird. A fin clip from a fish released back into a river. Multiply moments like these by tens of thousands, collected across three decades of fieldwork on every continent, and a pattern starts to show up in the numbers: many wild species are quietly losing the genetic diversity that keeps them able to adapt.
The loss is small on average, but it is real and it is global. An international team of conservation geneticists pulled together every usable temporal measurement of genetic diversity they could find in the scientific literature, published over more than 30 years. After filtering out studies that did not meet strict inclusion criteria, they were left with 3,983 individual measurements covering 622 species from 36 classes and 16 phyla, spanning 141 countries and nearly every terrestrial and marine region on Earth. Genetic diversity is essentially the raw material evolution has to work with. A population with more of it has a better shot at adapting to a new disease, a shifting climate, or a degraded habitat. Combining all of that data through a formal statistical model, the researchers found a small but statistically significant average decline in within-population genetic diversity over time, an effect too consistent across such a large dataset to be explained by chance.
Birds and mammals are losing the most. When the team broke the results down by taxonomic group, birds showed the sharpest average decline of any well-sampled class, followed by mammals. Three other groups with enough data to analyze separately, flowering plants, insects and ray-finned fish, showed no significant average change, meaning their genetic diversity looked broadly stable over the study periods. That does not mean those groups are safe. Losses can lag behind population declines by years or decades, especially in species with large populations or long lifespans, so a stable reading today does not rule out trouble already building underneath.
Threats are common. Active help for genetic health is rare. Roughly two out of three of the populations in the dataset had at least one documented threat during the period studied, most often harvesting or harassment, land use change, or other human-driven pressures. Fewer than half received any form of conservation management, and only about a third had both a recorded threat and a recorded intervention. Even populations with no reported disturbance at all still showed a modest average loss of genetic diversity, suggesting a kind of background erosion tied to human-dominated landscapes that shows up even where no single obvious threat has been flagged.
Only one strategy was reliably linked to a genuine increase. Among the conservation actions with enough data to test, supplementation, physically adding new individuals to a population through measures like restoring habitat connectivity or translocating animals between sites, was the only one statistically associated with a rise in genetic diversity, an effect that showed up most clearly in birds. Legal protection on its own, breeding programs and short-term support like supplementary feeding were all still associated with an average net loss. That is not necessarily a mark against them: those measures tend to be aimed at populations that are already declining, so even a management action that meaningfully slows the damage can still show up as a net loss in the numbers.
The researchers are careful to frame this as a correlational picture built from observational data, not a controlled experiment, so it cannot prove that any single action causes a given outcome in isolation. What it does offer is the broadest census yet of how genetic diversity is actually changing on the ground, and a case for treating it as something to monitor directly, alongside population counts and species range maps, rather than as an afterthought.
The study, “Global meta-analysis shows action is needed to halt genetic diversity loss,” was led by Robyn E. Shaw and colleagues, coordinated by Catherine E. Grueber, and published in Nature. DOI: 10.1038/s41586-024-08458-x.
