Global Insect Rescue Plan Needs New Monitoring Technology to Measure Whether Conservation Targets Are Working
A Conservation Letters study finds that current biodiversity indicators cannot detect changes in insect populations, calling for AI-powered camera traps, weather radar, and automated identification systems to track whether 2030 conservation targets are protecting the planet's most numerous animals.
World governments agreed to 23 biodiversity targets under the Kunming-Montreal Global Biodiversity Framework, aiming to protect and restore nature by 2030. A study published in Conservation Letters on March 19 concludes that while these targets are well-designed and could help reverse falling insect numbers, the indicators currently used to measure progress are largely unable to detect any response by insects.
The research, led by Dr. Andrew Bladon of the University of Reading and Professor Lynn Dicks of the University of Cambridge, reports that insect populations decline by approximately one percent annually on average. Yet dragonflies and damselflies remain the only insect group to have been fully assessed for extinction risk globally, leaving scientists with little ability to determine whether conservation efforts are making a difference for the estimated one million named insect species and four million more awaiting discovery.
“Technology is opening up possibilities that simply did not exist a decade ago,” Bladon said, pointing to a new generation of monitoring tools that could close the measurement gap.
A Measurement Gap at Global Scale
The core problem is one of scale. Insects are the most species-rich group of animals on Earth, yet traditional monitoring methods, such as manual field surveys and light traps counted by hand, cannot cover enough ground to produce population-level data. The Conservation Letters study recommends four categories of technology to address this deficit: automated camera traps that use artificial intelligence to photograph and identify nocturnal insects; weather radar systems capable of tracking large-scale insect flight movements across entire regions; citizen science platforms where public observations are verified by expert and algorithmic review; and the consolidation of decades of historical population records to establish baseline measurements.
Professor Dicks described the stakes plainly: insects comprise the majority of Earth’s species, yet current evidence suggests their populations are falling. The study recommends that the United Nations establish a dedicated working group to develop insect-specific measurement systems aligned with the 2030 framework targets.
North American Data Underscores the Urgency
Separate research published in the Proceedings of the National Academy of Sciences provides a concrete illustration of the crisis that better monitoring could help address. The largest comprehensive assessment of pollinator extinction risk in mainland North America, led by Dr. Tara Cornelisse of NatureServe, found that 22.6 percent of nearly 1,600 assessed pollinator species face elevated extinction risk.
Bees are the most imperiled group, with 34.7 percent of 472 assessed native species at risk. All three pollinating bat species face elevated threat levels. Leafcutter and digger bees show the highest imperilment rates among bee families. The American Southwest concentrates the greatest number of at-risk species. Climate change, agricultural practices, habitat loss, and urban development are identified as the primary threats, with regional variation across the continent.
“Our study provides the most comprehensive picture yet of the pollinator crisis in North America,” Cornelisse said. The researchers recommend that policymakers integrate at-risk pollinator data into State Wildlife Action Plans and that land managers prioritize conservation of grasslands, shrublands, and woodlands. Pollinators contribute an estimated $15 billion annually to North American agriculture through crop pollination services.
Automated Pipelines Begin to Scale
Researchers are already building the kind of technology the Conservation Letters study calls for. A pipeline called BugNet, published in Frontiers in Ecology and Evolution by researchers at the University of Wisconsin-Madison and the University of Nevada Reno, demonstrates a scalable approach to automated insect identification from images. The system aggregates insect photographs from online databases, trains hierarchical classification models, and processes both internet-sourced and field images to classify insects across taxonomic levels.
BugNet’s annotation tool achieves approximately 1,160 bounding box annotations per hour and 6,000 categorical annotations per hour, rates that would be impossible with manual taxonomic work. A key innovation is the system’s treatment of missing taxonomic data: rather than requiring complete species-level labels, it assigns classifications at higher taxonomic levels when species identification is uncertain, avoiding misidentification of unknown taxa. This hierarchical approach is particularly relevant in diverse ecosystems where many species remain undescribed.
From Data Gap to Decision-Making
The three studies, taken together, outline both the problem and the emerging solution. North American pollinator data shows that extinction risk is high and unevenly distributed. The Conservation Letters analysis shows that global measurement systems are not yet capable of tracking whether conservation interventions are reducing that risk. And BugNet represents the class of automated tools that could begin to close the gap between what scientists know about insect decline and what policymakers need to act on it.
Whether governments will invest in deploying these technologies at the scale the crisis demands remains an open question. The 2030 deadline for the Global Biodiversity Framework targets is less than four years away, and for insects, the clock is ticking on measurement systems that do not yet exist at the necessary scale.