David J. Will of Island Conservation in Santa Cruz, California, developed the framework around a distinction most eradication plans gloss over: the difference between whether an eradication can be implemented and whether the consequences of failure justify planning and resourcing for detection-triggered response. A project may be entirely feasible in that a best-practice baiting operation can be delivered effectively on schedule and on budget, while still lacking any credible plan for detecting and removing survivors. On high-risk islands, Will contends, that gap is where success or failure is increasingly decided.

The evidence motivating the framework comes partly from a workshop titled Improving success rates for rodent eradication on high-risk islands, convened at the 2026 Island Invasives Conference in Auckland, New Zealand. More than 100 practitioners, researchers, technical providers, policy makers, and funders examined why recent high-risk attempts have narrowly failed despite close adherence to accepted practice. Several rodent eradications on inhabited Pacific islands, including Lord Howe Island, Palmerston Atoll, Tetiaroa Atoll, and Wake Atoll, succeeded only after teams implemented targeted efforts to detect and remove remnant populations after the primary removal operation failed to eliminate them. High-profile tropical attempts on Floreana, Henderson, and Late reduced rodent populations to low densities yet still fell short of eradication.

The biological and operational constraints of tropical and inhabited systems explain why. Persistent year-round breeding, abundant alternative foods, bait competition from non-target species, rugged terrain, limited access permissions, and socio-political complexity all increase the likelihood that some individuals are never exposed to the primary removal method. In temperate, uninhabited islands where most eradication evidence was generated, those factors behave differently, so the transferability of past experience declines precisely where conservation and social stakes are highest. In many failed eradications, the causes remain unclear even after correlated risk factors are identified, requiring structured diagnosis to separate ecological from operational drivers.

Will’s framework sorts projects into three strategy archetypes. Single-event implementation relies on one primary removal method supported by strong comparable precedent, followed by one-time confirmation monitoring, with no predefined contingency beyond limited mop-up. Planned contingency response keeps the primary operation but adds survivor-focused monitoring and predefined corrective actions designed for specific, anticipated failure modes triggered by monitoring evidence. Adaptive multi-method campaigns go further, deploying multiple removal methods in sequence using decision thresholds based on evidence from survivor-focused monitoring, because survivorship mechanisms remain unresolved or the consequences of failure justify greater response readiness.

Six planning questions guide projects into the appropriate class. Teams first define what eradication success means locally and how much confidence is required before success can be declared, including their tolerance for failure risk. They then assess whether previous approaches on socio-ecologically similar islands provide a credible basis for expecting success, since factors influencing whether every rodent encounters the removal method on tropical islands may differ substantially from those that informed temperate-island strategies. Third, they identify poorly transferable or untested dependencies: where success rests on assumptions lacking precedent in comparable contexts, predefined detection and response pathways should be built in.

The remaining questions sharpen the risk calculus. Where previous attempts on the same island or comparable systems have failed and the cause of survivors remains unresolved, repeating the same strategy without additional response capacity is unlikely to reduce risk; unresolved prior failure generally pushes a project toward the adaptive campaign class. Teams then evaluate the consequences of failure, not only at the project level but across organizational, regional, and sector-wide impacts, since a high-profile failure can erode stakeholder confidence, future investment, regulatory support, and willingness to pursue similar interventions elsewhere. Finally, they judge whether a predefined correction is credible: if anticipated risks can be clearly bounded and addressed, contingency planning suffices, but where failure modes are poorly defined or no single correction is likely to resolve the risk, detection systems, response tools, governance triggers, and funding become core components of the strategy rather than optional backup.

A retrospective application to Wake Atoll illustrates the logic. The 700-hectare Pacific atoll saw a first eradication attempt remove one rat species but not the other. During the second attempt, survivors were detected following the initial removal operation, and an extended response phase was required to achieve eradication. At the design stage, comparable precedent was limited, prior failure was present, and the mechanisms of survivorship remained unresolved, while the consequences of a second unsuccessful attempt were high given the site’s conservation significance. Under these conditions, the framework would have favored planning the project as a multi-method campaign from the outset. In practice, the initial single-event implementation was followed by an extended response phase once survivors were detected, a sequence the framework aims to anticipate rather than improvise.

The framework carries practical weight for funders and communities as well as operators. It clarifies when some eradications responsibly require larger initial budgets, longer timelines, and flexible or staged funding to increase the likelihood of removing survivors before failure becomes entrenched. Although this raises upfront costs, it can substantially reduce the risk of total failure and avoid the far greater cost, delayed conservation benefits, and erosion of stakeholder confidence that follow when high-profile projects require repeated attempts. Not all contingencies provide meaningful risk reduction, however; their value depends on whether failure modes are understood, survivorship can be detected early enough to trigger a predefined response while success remains achievable, and the authority and funding to act are already in place. Adaptive management, in this view, is not simply the ability to adjust tactics after implementation but a structured, pre-planned process establishing decision points, response options, milestones, and learning loops in advance, co-designed with stakeholders.

Will is candid about the framework’s limits. It is qualitative, does not estimate eradication probability, rank projects by risk, or prescribe specific tools, and because it relies on practitioner judgement it remains vulnerable to overconfidence and misclassification, particularly where important ecological, operational, or social differences remain unresolved. Independent peer review of strategy classification should therefore be standard practice for high-risk projects, and important uncertainties persist around rodent behavior, bait uptake, detection probability, population dynamics, and environmental influences on bait effectiveness. Future use should focus on retrospective testing against completed eradications and prospective application during feasibility assessment. The framework assumes immigration has been prevented or managed, distinguishing survivorship from reinvasion, and it applies irrespective of delivery method, whether toxicants, trapping, hunting, or biological control. As eradications move into increasingly complex ecological and social contexts, the article concludes, success may depend as much on planning for how survivorship will be managed as on optimizing the primary removal operation itself.

Subject of Research: A decision framework for selecting eradication strategies that account for rodent survivorship risk on high-risk islands

Article Title: Selecting strategies for high-risk island rodent eradications

Article References: Will, D. J. (2026). Selecting strategies for high-risk island rodent eradications. Frontiers in Conservation Science, 7, Article 1957591. https://doi.org/10.3389/fcosc.2026.1957591

Image Credits: AI Generated

DOI: 10.3389/fcosc.2026.1957591

Keywords: island eradication, invasive rodents, survivorship, adaptive management, contingency planning, feasibility assessment, island restoration, tropical islands, detection-triggered response, conservation decision framework, Frontiers in Conservation Science, Selecting

The geographical setting is central to the story. Great Swan Island, roughly one by three kilometers with a maximum elevation of about 19 meters, and its rugged sister, Little Swan Island, separated by a channel only about 200 meters wide, together cover just 3.6 square kilometers. Today they host a small rotating Honduran naval garrison and no permanent civilian residents. But the archival record reveals layers of human presence stretching back at least to the seventeenth century. Shell middens containing burned top shells and crude pottery, reported by visitors in the 1930s, hint at possible pre-contact Indigenous visits, though the islands’ isolation and lack of permanent freshwater may have kept any occupation minimal. Christopher Columbus’s crew appears to have been the first European party to encounter the islands in July 1502 during his fourth voyage, and buccaneers including Henry Morgan and Captain Charles Swan—from whom the islands take their English name—used Great Swan Island as a staging base in the mid-to-late 1600s.

The authors conclude that from the pre-contact period through the early seventeenth century, the islands likely remained in a natural or near-natural state. In their pristine condition, the Swan Islands were probably completely cloaked in Caribbean coastal lowland dry forest, so dense that the naturalist Percy Lowe wrote in 1908 that trees came right down to the sea, allowing a visitor to step in five yards from the blinding glare of the beach into the gloom of the woods. Enormous seabird colonies of boobies, frigatebirds, and terns deposited the vast phosphate guano reserves that would later attract industrial exploitation. Green, hawksbill, and loggerhead turtles nested on the beaches in numbers unimaginable today, green iguanas were reportedly tame and abundant, and the reefs supported large predators including goliath groupers, sharks, snappers, and jacks, along with the grazing sea urchin Diadema antillarum and plentiful queen conch.

The pivotal rupture came in the 1850s. In 1857, phosphate-rich guano estimated at more than three million tons, yielding 40 to 80 percent bone phosphate of lime, was reported on the islands, triggering United States claims under the Guano Islands Act of 1856. From 1858, the Atlantic and Pacific Guano Company began clearing native forest on Great Swan Island to excavate the deposits, shipping the guano to Woods Hole, Massachusetts, where it was blended with menhaden waste into fertilizer. At the operation’s peak in 1887, some 300 men worked the island, supported by a settlement of laborer housing, drying sheds, and a small railroad from the diggings to a western pier. Mining continued under various companies until 1904, leaving pits that persist today as ephemeral ponds. The deforestation almost certainly increased soil erosion and sediment runoff into nearshore waters, initiating a slow chain of marine degradation that would unfold over the following century and a half.

The human and biological toll compounded rapidly. Goats introduced from Grand Cayman grazed down native vegetation, and ships servicing the guano trade and later enterprises delivered rats, mice, and feral cats—predators that devastated seabird colonies, reptiles, and invertebrates. Commercial turtle hunting on an industrial scale was underway by the 1880s; an 1884 Baltimore Sun report described a superintendent freeing more than 300 turtles from holding pens because no vessels had come to buy them. That same article contains what appears to be the first published record of the Swan Island hutia, Geocapromys thoracatus, a small rodent found nowhere else on Earth. After mining ended, leaseholders and the United Fruit Company cleared more forest for a 7,500-tree coconut plantation, fruit crops, and tobacco, while a wireless telegraph station, a US Weather Bureau hurricane station, and later a cattle quarantine station and CIA radio facilities further reshaped the landscape. By 1908, Lowe recorded seabirds nesting only on Little Swan Island, the rookeries of Great Swan having evidently already been extirpated.

The twentieth century catalogued a cascade of extinctions. The endemic whiptail lizard Ameiva fuliginosa, collected in 1887, vanished by 1912, apparently a casualty of cat predation. The islands’ remarkable land snail fauna—23 described species, at least nine endemic—was abundant as late as 1937 but apparently collapsed completely soon afterward, most likely under sustained rat predation. The Swan Island hutia, abundant on Little Swan Island into the early 1950s, went extinct by 1960, a demise researchers attribute to feral cat predation compounded by the catastrophic Category 5 Hurricane Janet in 1955, whose deluge likely drowned animals sheltering in limestone crevices. The endemic Little Swan Island racer snake Cubophis brooksi has not been reliably seen since the late 1990s. Only the green iguanas persisted in force; a 2024 survey estimated 18,000 to 20,000 animals on Great Swan Island, though unlike the tame populations Lowe described, they now flee humans, an apparent behavioral legacy of decades of hunting.

The marine story follows a parallel, delayed trajectory. As late as 2000, the reefs surrounding the islands were still reported as reasonably healthy, with large predatory fish and four shark species present—remarkably late by Caribbean standards, where reef decline generally began decades earlier. Yet the archival synthesis reveals that between 2007 and 2011/13, the reefs underwent a dramatic phase shift. The first quantitative surveys, conducted in 2011 and 2013, found most coral reduced to rubble, with average live coral cover of only about 13 percent in 2011 and 10 percent in 2013, while macroalgae blanketed roughly a third of the reef surface. Predatory fish and sharks had become very rare. Surveys in 2024 showed little recovery, with live coral dominated by weedy, disturbance-tolerant species such as Siderastrea siderea and Porites astreoides, and only three small, bleaching colonies of staghorn coral remaining.

The researchers identify a convergence of stressors that likely pushed the reefs past a tipping point after 2007. These include more than a century and a half of land-use change and sedimentation; the collapse of seabird populations, which historically transported marine-derived nutrients from ocean to island and back into nearshore waters, boosting reef productivity and resilience; the loss of nesting sea turtles, which function as ecosystem engineers grazing seagrass and consuming reef sponges; sustained overfishing of herbivorous and predatory fish; the regional spread of white-band disease; the catastrophic 1983–84 mass mortality of the grazing urchin Diadema antillarum, which removed up to 85 to 90 percent of Caribbean populations; the arrival of invasive lionfish in the early 2000s; and warming-driven coral bleaching. Critically, NOAA hurricane records show that four intense Category 4 and 5 storms—Mitch in 1998, Iris in 2001, Wilma in 2005, and Dean in 2007—struck within a nine-year window, with return intervals of three, four, and less than two years, far too short for reefs already weakened by chronic stress to recover.

The study’s implications extend beyond historical curiosity. The Swan Islands sit within one of the Caribbean’s largest marine protected areas, declared a national marine park in 1991 and encompassing roughly 5,672 square kilometers of ocean, yet the reserve remains largely unmanaged. The authors argue that historical ecological knowledge is essential both for dispelling the persistent myth that the islands are pristine—as recent tourism and promotional materials still claim—and for setting realistic restoration benchmarks. They advocate a holistic ridge-to-reef strategy, noting that coral restoration efforts are likely to fail unless paired with terrestrial management, invasive species control, and the recovery of seabird and turtle populations that sustain land-sea nutrient connections. Recent development proposals, including a planned maximum-security prison on Great Swan Island that was cancelled in 2023 amid financial and environmental concerns, underscore the stakes. The Swan Islands, the authors conclude, now stand as a powerful model system for understanding how remoteness delays but does not prevent ecological collapse, and for designing restoration strategies grounded in the deep history of what these islands once were.

Subject of Research: Long-term anthropogenic ecological change on the Swan Islands in the western Caribbean

Article Title: Historical ecology of the Swan Islands (Islas del Cisne) chronicles long-term anthropogenic changes in the Western Caribbean

Article References: Mladenov, P. V., & Fitzpatrick, S. M. (2026). Historical ecology of the Swan Islands (Islas del Cisne) chronicles long-term anthropogenic changes in the Western Caribbean. Discover Ecology, 2(1), Article 23. https://doi.org/10.1007/s44396-026-00036-x

Image Credits: AI Generated

DOI: 10.1007/s44396-026-00036-x

Keywords: Swan Islands, historical ecology, coral reefs, invasive species, guano mining, seabirds, sea turtles, hurricanes, Caribbean, biodiversity loss, island restoration, Honduras