City Squirrels Show Genetic Split From Urban Life, Yet Their Survival Stays Untouched

In the sprawling metropolis of Berlin, one of Europe’s largest and most densely built-up cities, the Eurasian red squirrel has become an unexpected informant on how urbanisation reshapes wildlife populations from the inside out. A four-year study spanning three very different green spaces—a quiet woodland near the city’s edge, a historic cemetery hemmed in by residential blocks, and a heavily visited urban park—has revealed something both reassuring and subtly unsettling. Researchers found that the city has left a measurable genetic fingerprint on squirrel populations separated by only a few kilometres, yet this molecular divergence has so far produced no detectable consequences for the animals’ survival. The findings, published in Ecology and Evolution, offer one of the most integrated looks yet at how genetic and demographic responses to urban life can come apart, and why that disconnect matters for conservation planning in human-dominated landscapes.

Between autumn 2021 and spring 2025, the research team conducted an intensive capture-mark-recapture programme across the three sites, deploying ten live traps per location during eight seasonal project phases. Over the course of the study they recorded 540 capture events involving 141 individually identifiable squirrels, each tagged with a subcutaneous PIT-tag and sampled with a tiny ear tissue biopsy for genetic analysis. The three sites were deliberately chosen to represent a gradient of urbanisation intensity, quantified through a human activity index based on standardised counts of pedestrians, cyclists, dogs, and other human presence. The woodland, embedded in a larger contiguous forest near the Berlin-Brandenburg border, scored lowest at 0.26. The cemetery, enclosed by a stone wall and surrounded by residential neighbourhoods, registered a moderate 1.14. The urban park, bisected by major roads and bustling with visitors, reached 4.25—the highest level of anthropogenic pressure in the study.

The genetic results told a nuanced story. Across all three populations, standard measures of genetic diversity—total and mean allele numbers, allelic richness, effective number of alleles, and both observed and expected heterozygosity—were strikingly similar, and notably high compared with red squirrel populations elsewhere in Europe, including France, Great Britain, and Jersey. But finer-scale metrics revealed cracks beneath the surface. Each population harboured its own private alleles, with fifteen in the woodland, six in the cemetery, and eleven in the urban park, hinting at localised differentiation. The cemetery population showed a statistically significant deficit of heterozygotes, and standardised individual multilocus heterozygosity was significantly lower in the urban park than in the woodland. Low to moderate inbreeding coefficients relative to subpopulations suggested mild inbreeding and interrupted gene flow—the early stirrings of genetic segregation in the more urbanised sites, though far from advanced genetic erosion.

Bayesian clustering analysis using STRUCTURE software identified two moderately supported genetic clusters across the metapopulation, a result that surprised the researchers. Individuals from the woodland and the cemetery, despite being twenty kilometres apart, mostly belonged to the same cluster, while most squirrels from the urban park grouped in the second cluster—even though the cemetery lies less than four kilometres from the park. Geographic distance alone could not explain this pattern. To rule out the possibility that family groups were driving the clustering, the team removed thirty-five highly related individuals and reran the analysis; the two-cluster structure persisted, albeit more weakly, with the woodland population showing the greatest proportion of admixed animals. This robustness indicated genuine population-level structuring rather than kinship artefacts, contrasting with findings from Berlin’s urban hedgehogs, where apparent clustering dissolved once relatedness was accounted for.

The researchers interpret this pattern as evidence of early-stage genetic structuring shaped by landscape configuration rather than simple isolation by distance. The woodland’s even mix of both clusters is consistent with high connectivity and gene flow from the surrounding rural forest. In contrast, the urban park and cemetery each showed a predominance of one genetic background, a signature the authors attribute to founder effects followed by limited gene flow across the hostile urban matrix. Previous connectivity modelling based on citizen science data had identified the urban park as a potential corridor node linking green spaces across the city, but the genetic evidence suggests that potential pathways on a map do not automatically translate into unrestricted mating and movement. Major roads with four to eight lanes, railway corridors, and waterways appear to impose enough friction to subtly redirect gene flow, even in a species as mobile and behaviourally flexible as the red squirrel.

Crucially, however, the demographic side of the story remained largely flat. Using Cormack-Jolly-Seber models to estimate apparent six-month survival between trapping seasons, the team compared twenty candidate models incorporating site, sex, season, individual heterozygosity, inbreeding coefficient, and human activity. No single predictor clearly outperformed the alternatives, and the best-supported model—containing the inbreeding coefficient—was statistically indistinguishable from a null model of constant survival. Apparent survival estimates ranged narrowly from 0.69 to 0.75 across sites, with the urban park showing roughly six percent lower survival than the other two locations, a marginal tendency that could accumulate over years if real, but one wrapped in wide confidence intervals. Intriguingly, the inbreeding coefficient showed a weak, non-significant positive association with survival, running contrary to the classical expectation that more inbred individuals fare worse.

This absence of a heterozygosity-fitness relationship is itself scientifically meaningful. The authors emphasise that genetic structure can arise through founder effects, genetic drift, non-random recruitment, or local population turnover without any immediate demographic penalty. Genetic connectivity, as Lowe and Allendorf famously argued, does not equate to demographic connectivity—and the Berlin squirrels demonstrate this decoupling in real time. Comparable context-dependence appears across the animal kingdom: Apennine brown bears persist despite reduced genome-wide diversity and elevated inbreeding, while fragmented cactus wren populations show clear positive links between heterozygosity and survival. In Berlin, supplemental feeding, observed intensively at the cemetery, may further buffer survival by boosting local densities, though such feeding can also elevate disease transmission risks.

The study’s limitations are acknowledged candidly by its authors. The four-year window and moderate sample sizes constrain statistical power, and apparent survival cannot distinguish mortality from permanent emigration—a particularly thorny issue in urban landscapes where dispersal may be redirected or curtailed. Uneven recording of human activity among sites may have blunted the sensitivity of the human activity index, and unmeasured ecological variables such as predation pressure, disease, and resource availability likely contribute to the unexplained variation in survival. The woodland site, being a small window onto a vast continuous forest, presents a fundamentally different sampling context than the enclosed urban fragments, complicating direct comparisons of capture numbers and density estimates.

Nevertheless, the broader implications are clear and, in their way, hopeful. Berlin’s squirrels retain substantial genetic variation despite navigating a landscape laced with infrastructure barriers, suggesting that certain urban configurations—particularly those rich in green corridors and tree cover—can sustain genetically and demographically stable wildlife populations, at least over short timescales. Yet the subtle genetic signature already imprinted on the urban park population serves as an early warning: genetic structuring can precede demographic decline, meaning that today’s stable survival rates offer no guarantee against tomorrow’s vulnerability if connectivity continues to erode. For urban planners and conservationists, the message is that monitoring genetic structure alongside traditional demographic measures provides a far more sensitive early-detection system, one capable of flagging trouble long before population counts begin to fall. As cities worldwide continue to expand, the Berlin squirrels remind us that the molecular echoes of urbanisation arrive first—and that listening for them may be the wisest investment conservation can make.

Subject of Research: Population genetics and survival of Eurasian red squirrels along an urban-rural gradient in Berlin

Article Title: Urbanisation Leaves a Genetic Signature but No Detectable Demographic Effects in a Small Mammal Population

Article References: Drenske, S., Landgraf, C., Berger, A. T., Stemmer, A. D., Planillo, A., Eul, J. L., Wist, B., Röllig, K., Mikkelsen, A. J., Dammhahn, M., Fickel, J., & Kramer‐Schadt, S. (2026). Urbanisation Leaves a Genetic Signature but No Detectable Demographic Effects in a Small Mammal Population. Ecology and Evolution, 16(10), Article e74273. https://doi.org/10.1002/ece3.74273

Image Credits: AI Generated

DOI: 10.1002/ece3.74273

Keywords: urbanisation, Eurasian red squirrel, population genetics, genetic structure, survival analysis, Berlin, capture-mark-recapture, gene flow, urban ecology, heterozygosity, inbreeding, wildlife conservation