Insect decline has become one of the most discussed findings in ecology, and the evidence base is more limited and more contested than the coverage suggests.
The influential studies
A study measuring insect biomass in German nature reserves over decades reported a substantial decline, which received enormous attention.
Subsequent studies in other regions have reported declines in various measures, with considerable variation in magnitude.
Which established the concern and produced a literature examining it more carefully.
The methodological problems
Long-term insect monitoring is rare, since it requires consistent effort over decades that funding cycles do not support.
Which means most evidence comes from a small number of datasets, concentrated geographically in Europe and North America.
Coverage of tropical regions, where most insect diversity is, is very limited.
Methods differ between studies, which complicates comparison and aggregation.
Biomass against abundance against diversity
These are different measures that can move independently.
Total biomass can fall while species number holds, or the reverse.
Which means studies measuring different things reach different conclusions without contradicting each other, and coverage frequently conflates them.
The reanalyses
Several high-profile studies have been reanalysed, with critics identifying issues in site selection, statistical approach and the extrapolation from local to global.
Which has produced a more measured picture — declines are real in many monitored populations, the magnitude is uncertain, and global extrapolation from limited data is not supported.
Some analyses have found freshwater insects increasing while terrestrial ones decline, which complicates any single narrative.
What is reasonably established
Substantial declines in specific well-monitored groups in specific regions, particularly butterflies and moths in intensively farmed landscapes.
Habitat loss and agricultural intensification as major drivers, supported by consistent evidence.
Pesticide effects on non-target insects, with substantial evidence for certain compound classes and continued debate about field-realistic exposure.
Light pollution effects, which have accumulated evidence more recently.
And climate effects, varying by region and species.
Why it matters regardless
Insects perform functions that are difficult to replace.
Pollination of a large proportion of flowering plants and a substantial share of crops.
Decomposition and nutrient cycling.
Pest control by predatory and parasitic species.
And food for birds, bats, fish, amphibians and reptiles, several groups of which are also declining.
The monitoring gap
The most useful response to methodological criticism is better data.
Standardised monitoring schemes have been established in several countries, and automated methods — acoustic monitoring, image recognition, DNA metabarcoding — are reducing the cost substantially.
Which should produce far better evidence within a decade, and the decisions are being made now.
What individuals can do
Reducing pesticide use in gardens, which is a meaningful area in aggregate.
Providing flowering plants across the season and structural diversity including dead wood and undisturbed ground.
Reducing outdoor lighting, particularly at short wavelengths.
And contributing to recording schemes, since amateur naturalists have produced most of the long-term data that exists.
Pollinator specifics
The group with most attention and most policy response.
Managed honeybee numbers have generally increased globally, since they are livestock, which is frequently confused with wild pollinator status.
Wild bee declines are documented in several regions, and wild pollinators contribute substantially to crop pollination alongside managed bees.
Which means honeybee hive counts do not indicate pollinator health, and the two are routinely conflated in coverage.
Pesticide regulation
Certain compound classes have been restricted in several jurisdictions following evidence of effects on pollinators.
Which has been contested regarding field-realistic exposure levels and regarding the alternatives used instead.
Integrated pest management, reducing prophylactic application, has evidence supporting both reduced use and maintained yields in several systems.
Recording schemes
Long-running volunteer schemes provide most of the usable long-term data in the countries that have them.
Which makes participation genuinely valuable, and the identification barriers have fallen with photographic recording and automated assistance.
New monitoring methods
Automated camera systems identifying insects from images.
Acoustic monitoring for species that produce sound.
DNA metabarcoding, identifying many species from a bulk sample.
Radar for aerial insect movement, which has revealed migration on a scale that was not appreciated.
Which together should transform the evidence base over the next decade, at a cost far below what equivalent manual identification would require.
Habitat provision at scale
Field margins, flower strips and reduced-intensity areas within farmland have evidence for supporting insect populations.
Which is where the largest area is, and agricultural policy determines whether it happens.
Payment schemes for these measures exist in several regions with uptake depending on the payment level and administrative burden.
Gardens as habitat
The aggregate area of domestic gardens exceeds protected area coverage in several densely populated countries.
Which makes garden management collectively significant, and studies have found substantial insect diversity in gardens managed for it.
Native plants, flowering across the season, and structural variety including undisturbed areas are the measures with evidence.
Which means individual action and policy action operate on the same land, and the aggregate matters.