The research team, led by investigators analyzing twenty of the world’s most consequential invasive wood-boring pests, assembled an unprecedented dual-scale dataset spanning national and municipal levels. The focal species ranged across four families, including jewel-toned buprestids such as the emerald ash borer, imposing longhorn beetles like the Asian longhorned beetle, diminutive bark and ambrosia beetles, and the notorious sirex woodwasp. Each species was confirmed as invasive using at least one of four authoritative sources: the EPPO Global Database, the CABI Compendium, the Global Invasive Species Database, and peer-reviewed scientific publications.

From official reports, peer-reviewed first-record publications, and authoritative databases including the Alien Species First Record Database, FAO, and the USDA Forest Service, the researchers compiled 129 official first species-country records from 49 countries. On the citizen science side, they turned to the Global Biodiversity Information Facility, or GBIF, applying a strict stepwise filtering protocol. Records labeled as Human Observations were only a starting point, since that category can include professional surveys and institutional monitoring. By screening source fields such as dataset names, institution codes, and collection codes, the team isolated records from seven genuine citizen science platforms, including iNaturalist, Artportalen, and BioDiversity4All. The final citizen science dataset contained 14,889 records from 70 countries and areas.

The core analytical question was deceptively simple: when a citizen scientist first photographs a pest in a given country, does that moment typically precede or follow the official first record filed by a government or regulatory authority? The researchers defined a time lag as the citizen science first record year minus the official first record year, and restricted their paired analysis to the 79 species-country combinations where both record types existed, involving 17 species and 38 countries.

The verdict was lopsided. Citizen science first records preceded official records in only two cases, or 2.5 percent of pairs, and occurred in the same year in just three cases, or 3.8 percent. In 74 cases, or 93.7 percent, the public record came later. The overall mean lag was a staggering 24.82 years, with a median of 15 years. Even when the analysis was restricted to the post-2010 era, when citizen science platforms had matured and participation surged, the mean lag dropped to 5.18 years but positive lags still dominated. A continuous analysis across all 79 pairs revealed a strong declining trend in the lag over time, with a Spearman correlation of -0.84 and a Theil-Sen robust slope of -0.92 years per year, but the authors caution that this decline likely reflects the growing adoption of citizen science platforms since around 2010 rather than a genuine improvement in detection capacity.

Why does the public miss these invaders? The answer lies partly in the biology of the insects themselves. Many wood-boring pests have cryptic life stages, with larvae tunneling hidden inside xylem or phloem tissue, and adults that are small, seasonal, or short-lived. Bark and ambrosia beetles such as Hylurgus ligniperda and Xylosandrus crassiusculus measure just a few millimeters and spend most of their lives concealed within host trees, making them nearly invisible to casual observers. Formal identification often requires subtle diagnostic characters, specialist knowledge, or even microscopic examination. In contrast, larger and more visually striking species such as the Asian longhorned beetle or the emerald ash borer are far more likely to be noticed, photographed, and uploaded.

The trait analysis confirmed this intuition quantitatively. Adult detectability, a composite variable combining adult body size and adult body coloration, emerged as the best-supported single predictor of citizen science recording intensity, with a Mann-Whitney U test yielding a p-value of 0.010. Adult body size alone was positively associated with recording intensity, with a Spearman correlation of 0.52 and a p-value of 0.022. Crucially, the researchers tested two potential confounders, range size and invasion history length, and found neither was significantly associated with recording intensity, supporting a genuine link between how visible an insect is and how often the public records it. Research attention showed no association, and public attention measured through Google Trends showed only a non-significant positive trend.

Recording intensity varied enormously among species, from just 15 records for the redbay ambrosia beetle to 4,174 records for the emerald ash borer. Geographically, the records were also wildly uneven, concentrated in the United States with 3,969 records, Sweden with 2,656, and China with 1,540, while many other countries had few or none. This spatial heterogeneity underscores a critical caveat: areas lacking citizen science records cannot be interpreted as areas free of invasion, and record abundance does not directly reflect pest occurrence. The patterns instead capture differences in public participation, platform activity, and monitoring infrastructure across countries.

To probe whether citizen science could at least supplement spatial coverage at a finer scale, the team turned to China, compiling 36 official species-city records across 11 provinces for four invasive pests: Dendroctonus valens, Hylurgus ligniperda, Sirex noctilio, and Phoracantha semipunctata. Strict filtering of GBIF records retained 1,318 occurrences from nine focal wood-boring species in China, but these were dominated by three conspicuous longhorn beetles, Anoplophora chinensis, Aromia bungii, and Anoplophora glabripennis, which together accounted for 93.9 percent of all retained records. Spatially, records clustered in eastern China, with Beijing, Zhejiang, Shandong, Shanghai, and Jiangsu accounting for 57.5 percent of records with province-level information. Notably, of the four officially documented invasive species, only P. semipunctata and D. valens appeared in citizen science data, and no citizen science records existed at all for H. ligniperda or S. noctilio.

Yet the study did uncover a handful of compelling exceptions that hint at what citizen science can achieve under the right conditions. In China, the citizen science first record of P. semipunctata came in 2022, two years before its official first record in 2024, and public observations documented the species in both Guangdong and Yunnan, whereas official records listed only Guangdong. Similarly, the citizen science first record of Trichoferus campestris in Hungary preceded the official record by one year. These cases suggest that when a pest is sufficiently visible and public engagement is strong, citizen scientists can occasionally provide genuine early warning and valuable spatial supplementation. The authors also note that their analysis was restricted to GBIF-indexed platforms and did not capture public reports submitted directly to biosecurity agencies, meaning the true contribution of community-initiated detections is likely underestimated, as such reports are often absorbed into official records rather than cataloged as citizen science.

The broader implications are clear. For invasive forest wood-boring pests, the primary value of citizen science lies not in shortening the detection lag but in expanding spatial surveillance and tracking the subsequent spread of established populations. Prevention and early interception should continue to rely on phytosanitary measures aimed at reducing arrival risk, including treatment and inspection of wood packaging materials, strengthened quarantine protocols, professional surveillance networks, and expert-based identification. Future efforts should strengthen the links between citizen science platforms, forest-health agencies, quarantine authorities, and taxonomic experts, while providing targeted public guidance on host symptoms, boring damage, frass, and exit holes. The study also draws a useful contrast with systems like Mosquito Alert, which has successfully contributed to first detections of invasive mosquitoes at the municipal scale, demonstrating that citizen science can advance detection timing when the target organisms are frequently encountered, easy to photograph, and relatively easy to recognize. For the hidden borers gnawing silently inside the world’s trees, however, the eyes of the public are not yet a substitute for the trained gaze of the professional entomologist.

Subject of Research: The role of citizen science in early detection and spatial monitoring of invasive forest wood-boring pests

Article Title: Opportunistic citizen science expands spatial monitoring but rarely advances early detection of invasive forest wood-boring pests

Article References: Bao, Z., Pawson, S., Chang, X., Cheng, L., Xie, Q., Yang, W., & Shi, J. (2026). Opportunistic citizen science expands spatial monitoring but rarely advances early detection of invasive forest wood-boring pests. NeoBiota, 109, 107-128. https://doi.org/10.3897/neobiota.109.204328

Image Credits: AI Generated

DOI: 10.3897/neobiota.109.204328

Keywords: citizen science, invasive species, forest pests, wood-boring beetles, early detection, biosecurity, GBIF, emerald ash borer, bark beetles, surveillance, species detectability, China

The dominance of insects in these epicenters is worth pausing on. Insect herbivores differ from many pathogens in how they disperse, how quickly they can adapt to new hosts, and how strongly their populations respond to climate. Many forest insects are capable of long-distance flight or hitchhiking in timber, packaging materials, and live plant trade, which helps explain why temperate trade hubs with dense shipping networks accumulate so many established species. Warmer winters in temperate zones have also progressively removed a historical barrier: cold temperatures that once killed off overwintering larvae or eggs. As minimum temperatures rise, species that were previously limited by frost can survive and complete their life cycles further poleward, a mechanism that is consistent with the study’s finding that regions with stable temperatures and adequate precipitation support elevated pest richness.

The study’s central risk accounting distinguishes between forests that sit inside the native range of a given pest and forests that lie outside it. This distinction matters because the ecological consequences of pest attack are usually far more severe when a pest encounters host trees that share no evolutionary history with it. Native pests and hosts have typically co-evolved some degree of tolerance or resistance, whereas novel host-pest pairings can produce explosive mortality. The classic example is Dutch elm disease, a fungal pathogen vectored by bark beetles, which devastated elms across Europe and North America after being moved outside the region where tree and pathogen had coexisted. By quantifying both within-range and out-of-range exposure, the new work captures this asymmetry: potential risk zones where timber production, carbon-rich forests, or diverse tree communities overlap with high pest richness cover about 3.5 percent of global forest area when hosts are within a pest’s native range, and about 1.6 percent when they are outside it.

Those percentages may sound modest, but they translate into enormous absolute areas given that forests cover roughly four billion hectares of the planet. Even the smaller out-of-range figure corresponds to tens of millions of hectares of forest that harbor high-value ecosystem services while facing pest communities with which they have never contended. Moreover, the exposure metric is a potential-risk measure, not a prediction of inevitable damage. Actual outcomes depend on pest establishment success, host susceptibility, management responses, and chance events, all of which can amplify or dampen the underlying exposure. What the metric does provide is a prioritization tool: it identifies where surveillance budgets, quarantine inspections, and breeding programs for resistant planting stock would deliver the greatest expected benefit per unit of investment.

The linkage the authors draw between timber imports and pest richness deserves particular attention from a policy standpoint. Global trade in logs, lumber, pulpwood, and wood packaging has long been recognized as the dominant pathway for forest pest introductions, a conclusion formalized in international phytosanitary standards such as the International Plant Protection Convention’s measures on wood packaging material. The new analysis adds a macroecological dimension to this understanding by showing that forests embedded in regions with high timber import volumes are precisely the forests that support high pest richness alongside high ecosystem-service value. In other words, the same connectivity that makes an economy efficient at sourcing wood also makes its forests efficient at accumulating pests. This creates an inherent tension for wood-security policy: restricting trade outright is neither feasible nor necessarily desirable, but the study’s spatial blueprint suggests that import-driven risk is spatially predictable and can therefore be targeted.

Fragmentation emerges from the analysis as a second, more controllable driver of vulnerability. The finding that forest fragmentation amplifies pest impacts aligns with a substantial body of ecological theory and evidence. Fragmented forests have proportionally more edge habitat, where microclimates are warmer, drier, and more variable than in forest interiors, conditions that often favor herbivorous insects. Edges also concentrate some pest natural enemies less effectively and can disrupt the population dynamics of the birds and parasitoid wasps that suppress outbreak insects. Fragmentation additionally impedes the dispersal of the predators and parasitoids that would otherwise track pest populations, while simultaneously making it easier for human-assisted pest movement to reach isolated forest patches via nearby roads and settlements. From a management perspective, this is arguably the most actionable result in the paper, because connectivity and patch size are directly influenced by land-use planning, whereas climate and trade volumes are not.

The carbon dimension of the study connects forest health to climate policy in a way that is increasingly hard to ignore. Forests currently remove a substantial fraction of anthropogenic carbon dioxide emissions from the atmosphere each year, and standing forest biomass represents one of the largest terrestrial carbon pools. When pest outbreaks kill trees, that stored carbon can be released through decomposition or, worse, through salvage logging and burning, while the affected area temporarily shifts from a carbon sink to a carbon source. Large-scale episodes illustrate the magnitude: bark beetle outbreaks in North American and European conifer forests have converted millions of hectares from net sinks to net sources for periods of years to decades. The study’s identification of carbon-rich forests within high-pest-richness zones therefore quantifies a financial as well as an ecological exposure, since carbon credits, national greenhouse-gas inventories, and offset schemes all implicitly assume that forest carbon will remain standing.

Biodiversity exposure adds a third layer of concern. The finding that diverse tree communities can coincide with high pest richness cuts against a widely held expectation that diversity buffers against damage. That expectation, often summarized as the diversity-disease hypothesis, holds that mixed forests dilute the impact of host-specific pests because specialized herbivores and pathogens encounter susceptible individuals less frequently. The new results do not necessarily contradict that mechanism at the stand scale; rather, they show that at the global scale, the same climatic and trade conditions that support many tree species also support many pest species. Regions that are favorable for plant growth are favorable for the insects and pathogens that feed on plants, so high-diversity forests can simultaneously enjoy the dilution effect locally and face high pest pressure regionally. Reconciling these two patterns is an important task for follow-up research, and it suggests that diversity-based planting strategies should be designed with the regional pest pool in mind rather than assumed to be protective by default.

Methodologically, the study reflects a broader shift in macroecology toward synthesizing massive occurrence databases, such as digitized museum records and citizen-science observations, into risk frameworks. This approach brings real power but also real caveats. Occurrence records are spatially biased toward wealthy regions with active sampling, toward roads and research institutions, and toward species that are conspicuous or economically important. The authors’ finding that northwestern Europe and eastern North America emerge as epicenters may partly reflect this sampling intensity, although both regions also genuinely host enormous non-native pest loads accumulated over two centuries of intensive trade. Readers should therefore interpret the mapped epicenters as a combination of true biological pattern and observation effort, and treat under-sampled tropical regions with appropriate caution, since the tropics host the majority of tree species and may harbor substantial undocumented pest diversity.

For forest managers and policymakers, several practical implications follow from the work. First, surveillance should be spatially targeted: the 3.5 percent and 1.6 percent risk zones provide a defensible starting point for allocating inspection capacity at ports, detection surveys in vulnerable forests, and pre-emptive breeding of resistant stock. Second, biosecurity investment at the point of entry remains the most cost-effective intervention, because preventing a pest establishment is orders of magnitude cheaper than managing an outbreak, as the cumulative costs of invasive forest insects, estimated in the hundreds of billions of dollars globally over recent decades, attest. Third, the fragmentation result argues for maintaining and restoring interior forest conditions, consolidating reserves, and reducing edge density in managed landscapes, measures that build general resilience regardless of which specific pest arrives next. Fourth, the climate signal implies that adaptation planning should anticipate expanding pest ranges rather than assume historical boundaries, meaning that planting choices made today for timber or restoration should consider the pest communities trees are likely to face over their lifetimes.

The study also highlights a gap in how forest ecosystem services are typically valued. Timber yield models, carbon accounting frameworks, and biodiversity assessments are usually conducted in isolation from one another, yet the new analysis shows that all three are exposed to the same underlying threat gradient. Integrating pest exposure into each of these accounting systems would give governments and companies a more honest picture of the risks embedded in supply chains, offset portfolios, and conservation commitments. Certification schemes for sustainable forestry, for example, could incorporate pest-risk zoning, and national forest inventories could track pest-related mortality as a standard variable rather than an afterthought following major outbreaks.

Finally, the work underscores that forest health is a genuinely global commons problem. A pest introduced through one country’s ports can devastate forests across a continent, as history with chestnut blight, Dutch elm disease, and emerald ash borer demonstrates, yet biosecurity budgets remain nationally fragmented and often small relative to the value at stake. The spatial blueprint offered by this analysis gives the international community a common map on which to coordinate: shared risk zones, shared trade pathways, and shared climate drivers all point toward the value of collective action, from harmonized wood-treatment standards to joint early-warning networks. As global trade continues to move living material faster and farther than ever, and as warming continues to relax the climatic constraints on pest survival, the window for cost-effective prevention narrows. The study’s message is ultimately one of timing: the exposure is mapped, the drivers are identified, and the remaining question is whether management and policy will act on that knowledge before the risk zones convert into realized losses.

Subject of Research: Potential exposure risk of global timber production, carbon stocking, and biodiversity to forest pests

Article Title: Potential exposure risk of global timber production, carbon stocking, and biodiversity to forest pests

Article References: Qiao, Y., Liu, J., & Edwards, D. P. (2026). Potential exposure risk of global timber production, carbon stocking, and biodiversity to forest pests. Nature Communications. https://doi.org/10.1038/s41467-026-78728-x

Image Credits: AI Generated

DOI: 10.1038/s41467-026-78728-x

Keywords: forest pests, biodiversity, timber production, carbon storage, Nature Communications, forest health, invasive species, Potential, exposure, risk, global, timber

The great spruce bark beetle is an atypical member of its subfamily. Phylogeographic analyses indicate that it is native to the Russian Far East, and its closest relative is the North American Dendroctonus punctatus, on the far side of the Bering Strait. From its Siberian origins the species has spread westward and is now established in most of the spruce forests of Eurasia. Unlike tree-killing bark beetles that mass-attack and overwhelm their hosts, D. micans behaves almost like a parasite: each mated female, 90 percent of which have already mated with their brothers under the bark before emergence, bores a solitary egg gallery into a living tree. The beetle tolerates the resin and chemical defences that repel other species, so it does not need to kill its host. Each female lays up to 250 eggs, often in a single batch, and the larvae feed gregariously along a feeding front in the phloem, packing their frass behind them. Development is slow, taking one to three years depending on climate, which means the insect’s brood chambers remain embedded in living trees for long periods.

That slow, cryptic, living-host lifestyle is precisely what made the pest vulnerable to a very particular enemy. Rhizophagus grandis, a beetle of the family Monotomidae measuring 4.5 to 5.5 millimetres, has followed D. micans across its territorial expansion, though with temporal and spatial lags caused by long-distance passive transport of the pest with timber trade and by habitat fragmentation. More than a hundred species of insects, mites and nematodes have been found in the brood systems of D. micans, but almost all are inquilines, mycetophages or saprophages; only R. grandis is consistently associated with the pest. The predator’s adults and larvae consume the eggs, larvae, pupae and young adults of their prey, and both life stages show high resistance to the host tree’s monoterpenes. This tolerance grants them access to a rewarding niche with ample resources and virtually no interspecific competitors.

The predator’s prey-finding mechanism is extraordinarily efficient. Adults locate their prey’s galleries by means of specific chemical cues, or kairomones, primarily oxygenated monoterpenes derived from the larval frass of D. micans. In Belgium, where brood chamber densities can be as low as one to four per hectare, more than 90 percent of the chambers are gradually colonised by R. grandis. Once inside a brood chamber, the adults feed on the prey’s immature stages and lay 50 to 100 eggs under laboratory conditions. Females assess the availability of prey for their offspring through oviposition-stimulating cues present in the larval frass and can apportion their egg load accordingly. The predator larvae are voracious and display a coordinated aggregative behaviour mediated by chemical signals, consuming the equivalent of one fully grown D. micans larva during their entire development. At the end of the larval phase, the pre-pupae leave the bark and pupate in the litter at the base of the tree.

The historical record shows how the pest outran its enemy. Digitised old literature reveals that D. micans was already present, as a rare endemic species, in central European mountain ranges by the end of the eighteenth century, with early records from the Bohemian Massif and the Harz Mountains, and scattered finds in Sweden, the Italian Alps, Norway and Lapland. From the mid-nineteenth century onward, large outbreaks occurred and the pest spread progressively across Europe, reaching the Vosges in France by 1862, Belgium, the Carpathians and the Dinaric Alps. The invasion finally reached the Atlantic when the beetle arrived in Brittany in 2009 and moved to the western tip of Finistère in 2010. In newly colonised territories, where R. grandis had not followed, the pest became notably harmful. Georgia first detected it in 1955 in the Borjomi gorge, where the number of infested oriental spruce trees grew from about 200 to 700 within a single year, and by the late 1960s the infestation spanned a radius of over 300 kilometres. Türkiye recorded its first attacks in 1966 near the Georgian border, and by 2011 the beetle occupied the entire 452,693-hectare oriental spruce area of the Turkish Black Sea Region. Great Britain, an island with a relatively impoverished forest fauna, found the pest breeding in Sitka spruce in 1982, and dendrochronological analysis of occluded galleries showed that the initial attacks dated from 1973 or earlier, near a sawmill at Ludlow in Shropshire that imported wood from the continent.

The biocontrol campaigns began in Georgia in 1963, when the entomologist A.L. Kobakhidze released 23 adults and 36 larvae obtained from Czechoslovakia into infested trees near Tsagveri. Subsequent introductions from Czechoslovakia, the Latvian SSR and multiple Soviet republics progressively increased the founder stock, and the diversity of this original genetic material is thought to have enhanced acclimatisation. Seventeen small production laboratories were established at state forestry enterprises across the republic, using log-rearing: living prey larvae or mature females are introduced into fresh spruce logs, pairs of predators are inserted through slits in the bark where a hollow noise indicates a brood chamber beneath, and the pre-pupae emerging from the logs are collected in moist sand. Over the lifetime of the programme, approximately four million R. grandis were bred and released into Georgian forests. France followed in 1978, Great Britain in 1983 and Türkiye in 1985, ultimately releasing 800,000, 250,000 and 10,880,000 predators respectively.

Production technology evolved considerably. Log-rearing was labour-intensive, spatially demanding, produced unpredictable yields of 23 to 80 pre-pupae per female, under-exploited the available prey, and was prone to devastating epizootics caused by the entomopathogenic fungus Beauveria bassiana, which could destroy 80 to 100 percent of production. From 1984 onwards, a semi-artificial method was devised and progressively refined, operating at a constant temperature of 19 to 23 degrees Celsius. Spruce bolts are infested with prey females, and after about 55 days the fourth- or fifth-instar larvae are collected. Small flat polystyrene boxes are prepared with alternating layers of re-moistened spruce bark powder and fresh bark, into which 20 to 25 prey larvae and one pair of predators are introduced. Average yields of 50 to 100 eggs and young larvae per box can be obtained, with an overall average of approximately 44 young adults per female. A key advantage is that Beauveria mycelium can be seen through the transparent containers, allowing preventative destruction of infected boxes and rigorous prophylaxis between rearing cycles. An artificial diet based on blue bottle fly larvae and cat food croquettes has even been developed for Türkiye’s massive programme.

Release strategies were refined in parallel. In France, predators were introduced as a priority at the leading edge of the pest’s expanding range, in stands undergoing incipient infestation at low prey density, on the rationale that even limited introductions could establish an effective biocontrol zone ahead of the advancing front—forestalling by 15 to 20 years the natural arrival of the predator and sparing considerable timber volumes. The alternative approach bridged the demographic gap in long-established outbreak areas, requiring large numbers from the centralised mass-rearing facility. In Great Britain, low-density inoculative releases treated as many infested locations as possible, with sites receiving 10 to 50 pairs depending on the number of attacked trees recorded during intensive surveys. Mark-release-recapture experiments with 2,500 fluorescent-dusted adults in an isolated spruce block showed that predators reached infested trees up to 200 metres from a central release point, and that the attack rate by R. grandis scaled linearly with the number of prey immature stages present—evidence that the predator can detect how much prey is available and adjust its response accordingly.

The impact was measured with a simple but elegant technique. Because most attacked trees remain alive, identified cohorts of trees can be monitored for years along calibrated

Subject of Research: Classical biological control of the great spruce bark beetle Dendroctonus micans using the specific predator Rhizophagus grandis in Georgia, France, Great Britain and Türkiye

Article Title: Successful bark-beetle biocontrol. A review of the only case in history Releasing Rhizophagus grandis in Georgia, France, Britain and Türkiye against Dendroctonus micans

Article References: Grégoire, J.-C., Alkan Akıncı, H., Burjanadze, M., Evans, H. F., Fielding, N., & Mayer, F. (2026). Successful bark-beetle biocontrol. A review of the only case in history Releasing Rhizophagus grandis in Georgia, France, Britain and Türkiye against Dendroctonus micans. NeoBiota, 109, 271-313. https://doi.org/10.3897/neobiota.109.205672

Image Credits: AI Generated

DOI: 10.3897/neobiota.109.205672

Keywords: bark beetles, biological control, Rhizophagus grandis, Dendroctonus micans, forest pests, invasive species, Successful, bark-beetle, biocontrol, review, only, case

The factsheet, published by Pensoft Publishers on behalf of the three consortia, opens with a headline that captures the shared ambition of all three projects: technology-driven early detection of biotic stressors threatening European plant health. It is more than a summary document. By presenting each project’s core objective side by side, and by highlighting the thematic lines, technological solutions and methodological approaches the projects have in common, the factsheet offers policymakers, researchers and stakeholders a concise map of where Europe’s digital plant-protection effort stands and where it is heading. The shared vision uniting the three initiatives is that artificial intelligence, remote sensing and citizen engagement can transform how Europe identifies and responds to pest outbreaks.

Each project brings a distinct focus to the partnership. FORSAID concentrates on Europe’s forests, developing a comprehensive framework for the detection, monitoring and management of ten EU-regulated biotic stressor species. The four-year project brings together seventeen partner organisations from ten countries, and it places artificial intelligence and digital technologies at the core of its strategy for mapping the future of forest pest control. Forest ecosystems are particularly vulnerable to regulated pests because once an invasive organism establishes itself in a woodland landscape, eradication is often technically impossible and containment becomes the only realistic option. Early detection is therefore the difference between a manageable incursion and a landscape-scale catastrophe.

CERBERUS addresses the other side of the European agricultural spectrum: Mediterranean crops. The project is building a multi-layer early detection and monitoring system that integrates remote and proximal sensing, ground-based automation and artificial intelligence. Its data streams come from three distinct sources — space, the field and the public — and they feed into the management of both quarantine pests and commonly managed pests across vineyards, olive groves and citrus orchards. The system’s outputs include AI-driven risk maps and spraying recommendations, a combination designed to reduce pesticide use while maintaining effective crop protection. That dual goal reflects a broader pressure on European agriculture, where regulatory targets for pesticide reduction must be reconciled with the economic realities of farming in regions already stressed by climate change.

STELLA completes the triad with a holistic digital platform known as the STELLA Pest Surveillance System, or PSS. The system is designed to support the early warning and early detection of regulated pests, drawing on AI-supported innovations in satellite and UAV-based sensing, in-situ sensor technologies and crowdsourcing. What distinguishes STELLA’s approach is the pairing of detection capability with a response strategy: the project explicitly aims to help stop outbreaks before they can spread, rather than simply documenting them after the fact. In plant health, that distinction matters enormously, because the window between the arrival of a pest and its irreversible establishment can be measured in weeks.

Although FORSAID concentrates on forest pest control while CERBERUS and STELLA focus on agricultural crops, the factsheet demonstrates that the three projects have far more in common than a shared funding call. All three are tackling comparable challenges with overlapping tools. AI-driven risk modelling, remote sensing from satellites and drones, and citizen science monitoring appear in all three portfolios, adapted to different crops and ecosystems. This convergence is not accidental. The Horizon Europe call that funded the projects was designed to build a portfolio of complementary capabilities, and the factsheet makes visible how those capabilities interlock — from the algorithms that interpret sensor data to the networks of observers who report suspicious symptoms in the field.

The collaboration is not confined to documents. The common ground between the projects has already brought their younger generation of researchers together. In June 2025, PhD students and postdoctoral researchers from all three consortia met for a first joint Young Scientist Meeting, where they discussed forming dedicated working groups on citizen science, communication, remote sensing and pest surveillance. For early-career researchers, whose careers often unfold within the boundaries of a single project, such cross-consortia meetings create the professional networks and shared methodological vocabulary that outlast any individual grant. The working groups under discussion signal an intent to institutionalise that exchange rather than leave it to chance encounters at conferences.

Since that meeting, the partnership between FORSAID and STELLA has moved from discussion into the field. In May 2026, the two projects carried out joint fieldwork on the Greek island of Euboea, deploying drone-based remote sensing to evaluate tree health in forests affected by canker stain disease. The campaign illustrates the practical value of the sister-project model: STELLA’s expertise in UAV-based sensing and FORSAID’s forest-focused analytical framework could be combined on a single site, testing how detection pipelines developed for different regulated pests perform when applied to a shared, real-world outbreak. Euboea’s canker-stain-affected forests provided exactly that test case, and the results are expected to inform both projects’ ongoing methodological development.

The technological core shared by all three projects deserves a closer look, because it represents a genuine shift in how plant health surveillance is conceived. Traditional pest monitoring relies heavily on physical inspection by trained surveyors, a labour-intensive process that inevitably leaves vast areas unsurveyed between visits. The approach championed by FORSAID, CERBERUS and STELLA layers multiple observation systems on top of one another. Satellites provide broad, repeated coverage of vegetation condition across entire regions. Drones and UAVs add high-resolution imagery over specific sites flagged by the satellite layer. Ground sensors and automated field platforms capture micro-level signals that aerial systems cannot resolve. Finally, crowdsourcing and citizen science enlist farmers, foresters and the general public as distributed observers, extending the human sensing network far beyond what any institutional survey programme could staff. Artificial intelligence sits at the centre of this stack, fusing heterogeneous data streams into risk maps and alerts that decision-makers can act on.

As all three projects move deeper into their implementation phases, this spirit of collaboration is expected to grow further. The joint factsheet serves both as an introduction to the combined work of FORSAID, CERBERUS and STELLA and as an invitation to explore what a united, technology-driven plant health protection response can achieve across Europe. For a continent whose forests and farms are under increasing pressure from invasive pests and pathogens, the message of the factsheet is straightforward: no single project, technology or institution can provide the early detection Europe needs, but a coordinated portfolio of AI, remote sensing and citizen engagement — spanning forests and farmland alike — can build a surveillance fabric dense enough to catch outbreaks while they are still small enough to stop.

Subject of Research: Technology-driven early detection and monitoring of plant pests and pathogens in European forests and Mediterranean agriculture

Article Title: Safeguarding European plant health: New joint factsheet outlines synergy between Horizon Europe projects FORSAID, CERBERUS and STELLA

Article References: Safeguarding European plant health: New joint factsheet outlines synergy between Horizon Europe projects FORSAID, CERBERUS and STELLA. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: plant health, pest detection, FORSAID, CERBERUS, STELLA, Horizon Europe, artificial intelligence, remote sensing, citizen science, forest pests, Mediterranean agriculture, early warning systems