Along the windswept coast of southwest France, a row of hardy plant species has been quietly keeping a genetic diary of the past several centuries. A new study published in Heredity shows that eight members of the coastal sand dune communities of the Aquitaine shoreline did not drift through history independently. Instead, their effective population sizes rose and fell in striking synchrony, and the timing of their shared crash lines up with one of the most turbulent climatic episodes in recent European history: the Little Ice Age. The finding, based on the joint analysis of microsatellites and their flanking DNA sequences in more than 3,100 genotyped individuals, offers a rare community-wide view of how environmental upheaval is written into the genomes of the species that survive it.
The research team, led by Olivier Lepais of BIOGECO at the University of Bordeaux and INRAE, together with Maya Gonzalez of ISPA and Marie-Lise Benot, set out to answer a question that population geneticists have rarely tackled at the scale of an entire ecological community. Most studies of demographic history examine a single species at a time, inferring past population trajectories from patterns of genetic variation. But species that share a habitat are exposed to the same storms, the same shifting sands, and the same episodes of habitat loss. If past environmental changes were strong enough, the argument goes, they should leave comparable genetic fingerprints across many unrelated species at once. Testing that idea required a sampling effort of unusual breadth: the researchers developed species-specific microsatellite markers for eight dune plants and genotyped 3,116 individuals collected along the French Atlantic coast.
Microsatellites, also known as simple sequence repeats, are stretches of DNA in which a short motif of two to six base pairs is repeated over and over. Because the number of repeats changes relatively quickly through mutation, microsatellites are superb recorders of recent demographic events. Allelic richness, the number of distinct repeat-length variants circulating in a population, responds within generations to changes in effective population size, the genetically meaningful size of a population that determines how fast diversity is lost. The catch is that this fast mutation rate also means microsatellites saturate over longer timescales, blurring signals from the deeper past. To recover that deeper history, the team turned to the DNA sequences flanking each microsatellite, where point substitutions accumulate far more slowly. Heterozygosity at these flanking sites preserves a memory of ancient population sizes that the rapidly evolving repeat tracts have long since overwritten.
Combining the two marker types in a single inference framework is what gives the study its temporal reach. The researchers built a simple demographic model in which each population underwent a single change in size at some point in the past, and they used coalescent simulations to generate genetic data under a wide range of scenarios: different ancient and modern effective population sizes, different timings of the change, and different mutation rates. These simulated datasets were then compared with the observed genetic data using approximate Bayesian computation, or ABC, a family of methods that sidesteps the need for an analytical likelihood by simulating millions of datasets and retaining those that resemble the real one. In a further refinement, the team employed ABC random forests, a machine learning approach in which summary statistics computed from the genetic data act as predictors of the underlying demographic parameters, allowing the inference to weigh which combinations of statistics carry the most information.
The results were reassuringly clear on the quantities that matter most. Recent effective population sizes were well recovered, informed chiefly by allelic richness at the microsatellite repeats, while ancient effective population sizes were reliably estimated from heterozygosity in the flanking sequences. The timing of the demographic event proved harder to pin down for any single species, but the combined summary statistics, which mix microsatellite variation with flanking substitutions, sharpened the estimates considerably. This division of labor between fast and slowly mutating markers is the technical heart of the paper, and it demonstrates that sequencing-based microsatellite genotyping can do double duty: the same sequencing reads that yield repeat counts also yield the surrounding nucleotide variation needed to anchor the timeline.
What emerged from the analysis was a demographic signature shared across the community. Most of the eight species showed a strong and roughly synchronous decline in effective population size, with the timing of the crash estimated at between 135 and 450 years before the present. That window overlaps squarely with the Little Ice Age, the cold period lasting roughly from the fourteenth to the nineteenth century during which the North Atlantic storm belt intensified. For the Aquitaine coast, historical and geological records document exactly what intensified storminess means: waves of sand mobilization, known as sand drift, that buried vegetation and rolled inland across the coastal plain. Previous work on late Holocene sand invasion along this coast, including studies of the Médoc peninsula and the wider Aquitaine basin, has mapped these dune incursions and linked them to cold climate events. The genetic data now provide a biological mirror of that geomorphological story.
The mechanism the researchers propose is habitat loss. Open sand dune vegetation, the early-successional communities that colonize bare, mobile sand, depends on exactly the disturbance regime that storms once provided. But when storm-driven sand drift became too intense, it likely wiped out most of the open dune habitat rather than creating it, squeezing the specialist plants into a narrow strip along the shoreline where they remain today. A community-level contraction of habitat translates into a community-level contraction of effective population size, and that is precisely the synchronous decline the genomes record. Because the eight species are ecologically distinct, with different life histories and dispersal strategies, their shared demographic trajectory is best explained by a shared external driver rather than by any species-specific process.
The study also carries a cautionary note for conservation. Effective population size, not census count, determines a population’s capacity to maintain genetic diversity and adapt to changing conditions, and the classic 50/500 benchmarks for short- and long-term viability are measured in these terms. A historical crash that reduced effective population sizes across an entire community constrains the evolutionary options of every member species, even those that appear locally abundant today. The authors argue that this legacy of past habitat loss will limit how sand dune plants respond to future environmental change, including the ongoing transformation of European dunes by stabilization, afforestation, and development. Restoration efforts, other researchers have argued, must specifically consider species that require open and early-successional dune habitats, and the new genetic evidence underscores why: these species may already be running on reduced genetic capital accumulated through centuries of contraction.
Methodologically, the paper joins a growing movement in population genetics toward joint analysis of different mutation classes at the same loci. Earlier work on rear-edge oak populations by members of the same team showed that combining microsatellite repeat variation with flanking sequence substitutions could illuminate complex demographic histories involving gene flow and vicariance. The present study extends that approach from a single species to a whole community, and from a proof of concept to a comparative framework. The researchers also made their raw data available, depositing low-coverage whole genome sequences used for marker development in the European Nucleotide Archive, alongside extensive supplementary analyses that test how marker number, sample size, and mutation rate estimates affect the reliability of the demographic inference.
For ecologists and conservation biologists, the broader message is that genomes can serve as community-level archives. When multiple species sharing a landscape tell the same demographic story, with the same timing and the same direction of change, the case for a common environmental cause becomes compelling in a way that single-species studies cannot achieve. The dune plants of southwest France, it turns out, all recorded the same stormy centuries in their DNA. Reading that record required 3,116 genotyped individuals, hundreds of newly developed markers, and a simulation framework capable of extracting timing from a mixture of fast and slow mutations, but the payoff is a demonstration that past climate-driven habitat loss left a coherent, measurable signature across an entire plant community, one that continues to shape its capacity to face the changes still to come.
Subject of Research: Shared demographic history of coastal sand dune plant communities inferred from microsatellites and flanking sequence variation
Article Title: Joint analysis of microsatellites and flanking sequences shows shared demographic response of coastal sand dune plant communities to past environmental changes
Article References: Lepais, O., Gonzalez, M., & Benot, M.-L. (2026). Joint analysis of microsatellites and flanking sequences shows shared demographic response of coastal sand dune plant communities to past environmental changes. Heredity. https://doi.org/10.1038/s41437-026-00881-2
Image Credits: AI Generated
DOI: 10.1038/s41437-026-00881-2
Keywords: population genetics, microsatellites, flanking sequences, approximate Bayesian computation, effective population size, coastal sand dunes, Little Ice Age, habitat loss, community ecology, Aquitaine coast, plant conservation, demographic history
Cite Scienmag News
APA MLA Chicago
Juliet Wilcox. (September 23, 2026). Genetic Time Capsule Reveals Dune Plants Crashed Together During the Little Ice Age. Scienmag. https://scienmag.com/genetic-time-capsule-reveals-dune-plants-crashed-together-during-the-little-ice-age/
Juliet Wilcox. “Genetic Time Capsule Reveals Dune Plants Crashed Together During the Little Ice Age.” Scienmag, 23 September 2026, https://scienmag.com/genetic-time-capsule-reveals-dune-plants-crashed-together-during-the-little-ice-age/. Accessed 23 September 2026.
Juliet Wilcox. “Genetic Time Capsule Reveals Dune Plants Crashed Together During the Little Ice Age.” Scienmag. September 23, 2026. https://scienmag.com/genetic-time-capsule-reveals-dune-plants-crashed-together-during-the-little-ice-age/
Copy citation Download RIS
Tags: approximate Bayesian computationAquitaine coastbiodiversity and climate change in European dunescoastal dune plant geneticscoastal sand dunescommunity ecologycommunity-wide genetic analysis of dune ecosystemsdemographic historyecological consequences of Little Ice Ageeffective population sizeenvironmental upheaval and plant genomesflanking sequencesgenetic evidence of historical climate episodesgenetic signatures of past climate extremeshabitat lossLittle Ice AgeLittle Ice Age climate impact on plantsmicrosatellite DNA in plant population historymicrosatellitesplant conservationpopulation dynamics of sand dune speciespopulation geneticsshared demographic history of coastal plantsspecies co-migration during climate events
