Lord Howe Island is a 1455-hectare crescent of volcanic land in the Tasman Sea, roughly 600 kilometres east of the Australian mainland, and a World Heritage-listed hotspot of endemic biodiversity. Black rats arrived there in June 1918, when the steamboat SS Makambo ran aground for nine days and rats escaped aboard the vessel onto an island that had never known rodents. Within two years, rats reached plague proportions. They are implicated in the extinction of five endemic bird taxa, along with the decline or disappearance of reptiles, land snails, insects, and plants found nowhere else on Earth. Predation by black rats on the island was eventually listed as a Key Threatening Process under both New South Wales and Australian federal law, making a response unavoidable for anyone hoping to preserve the island’s ecological character.

The Rodent Eradication Programme that followed was, by any measure, a logistical feat. Because Lord Howe Island is permanently settled, with around 350 residents, the campaign could not simply blanket the island from the air without careful community negotiation. Beginning in early 2019, the team deployed the rodenticide brodifacoum through a combination of ground-based bait stations and hand-broadcast applications across 300 hectares of the settlement area, alongside aerial baiting of a further 1200 hectares of forested terrain. The last rat was detected on 9 October 2019, and the programme was declared successful. Then, in April 2021, monitoring devices in the settlement area registered rat activity once again, and the central mystery of this study was born.

The two possible explanations carried very different management implications. Remnant survivors would suggest weaknesses in the eradication method itself, particularly in the complex human settlement zone, and would argue for intensified local surveillance. A new incursion, by contrast, would point instead to biosecurity failures along the transport routes connecting the island to the mainland, including regular flights from Sydney and Port Macquarie and a fortnightly barge service from the NSW coast. Telling those scenarios apart required genetics, and fortunately the researchers had a crucial resource: archived rat specimens from the Australian Museum, collected on Lord Howe Island in 2003 and 2017 as part of unrelated research, which preserved the genetic fingerprint of the pre-eradication population.

The team assembled a dataset of 130 rats: 38 collected on the island in 2021 after the eradication, 27 representing the pre-eradication island population, 60 from mainland New South Wales, and 5 from southwest Pacific islands including Norfolk Island, Papua New Guinea, and Vanuatu. They amplified a 307-base-pair region of the cytochrome B mitochondrial gene in 125 individuals, and a longer 927-base-pair fragment in a subset of 66, then genotyped 12 polymorphic microsatellite loci across the genome. Mitochondrial DNA, inherited only through the maternal line, placed each population in a global context, while microsatellites, scattered across the nuclear genome, provided the fine resolution needed to distinguish closely related populations.

The mitochondrial results were unambiguous. Every rat sampled on Lord Howe Island, both before and after the eradication, carried the exact same cytochrome B haplotype, a Lineage I sequence that has otherwise been recorded only in widely scattered locations such as Hokkaido in Japan and Texas in the United States. This haplotype has never been detected in mainland Australia, Melanesia, or the southwest Pacific. The mainland NSW rats, by contrast, mostly carried different but similar Lineage I haplotypes, and ten individuals from the Port Macquarie area carried a divergent Lineage II haplotype associated with southeast Asia. The presence of both lineages in eastern Australia points to multiple historical introductions of black rats since European settlement, but the single island haplotype stood apart from all of them.

The nuclear DNA told a consistent and even more detailed story. Microsatellite diversity was highest in the large mainland NSW population, moderate in the pre-eradication island population, and dramatically lower in the 2021 rats, whose allelic richness had fallen by roughly half relative to their pre-eradication predecessors. Five of the twelve loci were monomorphic in the post-eradication sample, and the animals were highly interrelated, with mean relatedness values indicating that most individuals were first-degree relatives. That combination of low diversity and high relatedness is the classic genetic signature of a population founded by a very small number of individuals, possibly as few as two, and held at low numbers for multiple generations. Bayesian clustering and principal coordinate analyses both grouped the 2021 rats firmly with the pre-eradication island population rather than with the mainland.

Just as importantly, the study confirmed that the island and mainland populations were clearly genetically distinct in the first place, which is what made the forensic assignment possible. The fixation index between them was 0.10, and each population carried a substantial share of unique alleles. Despite concentrated sampling at the very ports involved in island resupply, the researchers found little evidence of gene flow in either direction, a pattern consistent with the incumbent advantage enjoyed by a long-established population that repels or fails to accommodate new migrants. Intriguingly, the island’s mitochondrial haplotype does not match the Australian east coast, suggesting the 1918 founders may have arrived from somewhere else entirely, a question that only more comprehensive global sampling of black rat diversity can resolve.

The eighteen-month gap between the last official detection and the reappearance of rats remains the hardest part of the story to explain biologically. Rats are famously fecund, with short generation times and explosive reproductive potential under favourable conditions, so a remnant population persisting undetected for that long, even under intensive monitoring, implies that a handful of animals endured at extremely low density for multiple generations, likely within the shelter and resources of the settlement area. A strikingly similar lag was documented on the Italian island of Ventotene, another inhabited island where remnant rats resurfaced long after eradication was declared complete. Both cases reinforce a sobering lesson: the presence of people, buildings, and infrastructure creates refugia that poison campaigns struggle to reach, and inhabited islands demand extraordinary vigilance even after apparent success.

The good news is that Lord Howe Island responded decisively. Once the 2021 rats were detected, a rapid and comprehensive response eliminated the remnants, with the last rat detected in August 2021, and an upgraded eradication and biosecurity programme was installed using chew cards, wax tags, tracking tunnels, kill traps, detection dogs, and camera traps at high-risk reinvasion points. Five years on, the island’s native fauna and flora are showing significant signs of recovery, from rebounding birds to recovering invertebrates and plants. The broader message for conservation science is procedural as much as biological: genetic monitoring, including the collection and archiving of samples before eradication begins, should be a standard component of rodent eradication programmes. In this case, museum specimens collected years earlier for unrelated research proved decisive, transforming an unsettling reappearance into a solvable puzzle and, ultimately, a vindication of the eradication itself.

Subject of Research: Genetic assessment of whether post-eradication black rats on Lord Howe Island were remnant survivors or new invaders

Article Title: Genetic assessment of the introduced rats found on Lord Howe Island post rodent eradication, new invaders or remnant survivors?

Article References: Frankham, G. J., Helgen, K. M., & Eldridge, M. D. B. (2026). Genetic assessment of the introduced rats found on Lord Howe Island post rodent eradication, new invaders or remnant survivors?. NeoBiota, 105, 25-45. https://doi.org/10.3897/neobiota.105.163728

Image Credits: AI Generated

DOI: 10.3897/neobiota.105.163728

Keywords: Lord Howe Island, black rat, Rattus rattus, invasive species, rodent eradication, mitochondrial DNA, microsatellites, population genetics, island conservation, biosecurity, bottleneck, NeoBiota

The study, led by Carina Lee Terry of National Taiwan University together with colleagues including senior authors Hurng-Yi Wang and Cheng-Ruei Lee, set out to answer a deceptively simple question: when two closely related species share the same island, do they show the same patterns of genetic structure within their populations? Geographic isolation and environmental differences are both well-known drivers of genetic differentiation, and decades of research have examined their relative importance in individual species. But comparing closely related species living side by side on the same landmass offers a rare natural experiment, because it holds much of the shared history constant while allowing differences in ecology, in this case elevational range, to reveal their influence.

The two subjects of the investigation were Petaurista lena and Petaurista grandis, two giant flying squirrels endemic to Taiwan. Although they have similar life histories, the species occupy different elevational bands in the island’s mountainous terrain, which rises steeply from coastal plains to peaks approaching four thousand meters. Taiwan’s dramatic topography, sculpted by tectonic uplift and carved by Pleistocene glaciations, has long been recognized as a cradle of endemism, and small mammals on the island have repeatedly shown phylogeographic breaks that hint at past isolation. The researchers wanted to know whether these two squirrels, gliding through different altitude zones, had responded to that shared landscape in the same way.

To find out, the team turned to population genomics. They generated genome-wide single nucleotide polymorphism data from tissue samples, which came from the Taiwan Biodiversity Research Institute, the National Museum of Natural Science, and lawfully harvested animals provided by Truku Indigenous hunters in the Hualien region under permits issued by Taiwan’s Forestry Bureau. With thousands of genetic markers in hand, they applied a battery of analytical tools: ADMIXTURE and principal coordinates analyses to visualize population structure, spatial genetic analyses to detect how relatedness decays with distance, and a framework for disentangling the effects of geographic distance from those of environmental variables on genetic differentiation.

The first major result was a striking parallel. Both species showed a clear north-south genetic split, with individuals from the northern part of their ranges genetically distinct from those in the south. This shared pattern suggests that the fundamental geography of Taiwan, its elongated north-south axis and the mountain ridges that run down its spine, has imposed a common signature of isolation on both squirrels. When the researchers statistically accounted for neutral factors such as geographic distance, they found little association between genetic variation and environmental characteristics. In other words, the climate and habitat differences across the island did not appear to be sculpting the squirrels’ genomes in any strong, detectable way; the landscape’s shape was doing most of the work.

But beneath that shared surface pattern, the demographic histories of the two species diverged dramatically, and this is where the study delivers its most surprising findings. Using fastsimcoal2, a coalescent-based program that models complex evolutionary scenarios and estimates the demographic parameters that best explain observed genetic data, the team reconstructed each species’ past. Petaurista lena, the species associated with higher elevational ranges, experienced an early north-south divergence dated to roughly 163 thousand years ago, deep in the Pleistocene. After that split, gene flow between the northern and southern populations remained comparatively high, and the species maintained a larger overall population size with correspondingly higher genetic variation.

Petaurista grandis told a very different story. Its north-south differentiation is far more recent, dated to approximately 46 thousand years ago, and its population size is smaller. Most dramatically, the northern population of P. grandis appears to have passed through a strong bottleneck, a severe reduction in numbers that would have stripped away genetic diversity and left the surviving population with a compressed, less varied gene pool. The team corroborated their coalescent modeling with Stairway Plot v2, an complementary method that infers changes in population size through time from the folded SNP frequency spectrum, and plotted the inferred trajectories against historical temperature records to see how the squirrels’ fortunes tracked the climate’s swings.

Why would two species with such similar ecologies end up with such different demographic pasts? The researchers turned to niche suitability modeling with Maxent, which uses occurrence records and climate layers to estimate where conditions favorable to each species existed in the past. The answer that emerged is elegantly simple: area. Across historical climate scenarios, the suitable habitat for P. lena was generally larger than that available to P. grandis. A species with more room to live can sustain larger populations, retain more genetic variation, and maintain enough connectivity between regions to soften the effects of any single isolation event. P. lena’s earlier divergence with sustained gene flow fits that picture of a species with a broad, stable footprint across the island through the Pleistocene.

P. grandis, by contrast, appears to have been squeezed into a smaller historical suitable area, which would have kept its numbers lower and made its populations more vulnerable to the climate oscillations of the ice ages. The authors suggest that its relatively recent north-south divergence and reduced genetic variation may be tied to that constrained historical range. They also raise the possibility of competitive exclusion between the two species as a contributing factor, a reminder that the demographic history written in a genome reflects not just climate and geography but also the ecological push and pull among species that share a landscape. The elevational ranges that separate the two squirrels today may thus be both a cause and a consequence of their divergent histories.

The broader lesson of the study resonates far beyond Taiwan’s forests. Species with similar life histories on the same island can indeed converge on similar patterns of within-species divergence, in this case a north-south split that mirrors the island’s geometry. But the depth of that divergence, the amount of gene flow, and the size and resilience of populations can differ enormously depending on each species’ elevational range, its response to past climate change, and its interactions with close relatives. For conservation, the implications are concrete: the northern population of P. grandis, with its bottlenecked, low-diversity gene pool, may warrant particular attention as the island’s climate continues to change. And for evolutionary biologists, the two squirrels of Taiwan offer a vivid demonstration that the same island can tell two species very different stories, and that genomes remember which story each one lived.

Subject of Research: Comparative population genomics and demographic history of two sympatric giant flying squirrel species with different elevational ranges in Taiwan

Article Title: Distinct demographic histories associated with elevational ranges of two giant flying squirrel species in Taiwan

Article References: Terry, C. L., Chang, S.-W., Pei, K. J.-C., Chen, H.-W., Chen, Y.-J., Wang, H.-Y., & Lee, C.-R. (2026). Distinct demographic histories associated with elevational ranges of two giant flying squirrel species in Taiwan. Heredity. https://doi.org/10.1038/s41437-026-00891-0

Image Credits: AI Generated

DOI: 10.1038/s41437-026-00891-0

Keywords: giant flying squirrels, Petaurista, Taiwan, population genomics, phylogeography, demographic history, elevational range, Pleistocene climate, niche modeling, genetic divergence, bottleneck, island biogeography