Earth Has At Least 14 Million Insect Species, More Than Twice What Scientists Thought

For decades, scientists settled on a rough consensus: about six million insect species exist on Earth. A major new study says that number is almost certainly wrong, and the real count could be two to three times higher.

Researchers analyzed genetic data from more than 1.6 million individual tropical insects collected in Costa Rica, combined that with deep sampling of one of the most species-rich groups of parasitic wasps on the planet, and ran the numbers through powerful statistical tools built to account for species that were present but never actually caught. Their conclusion: as a conservative lower-bound estimate, there are at least 14 to 20 million insect species on Earth, a range that reflects the different reference groups used in the scaling-up calculation. Authors stress that this is a floor, not a final tally.

A separate analysis, built on a statistical point estimate with a much wider margin of error, pushes the count toward nearly 30 million, but the team treats that figure as less reliable and leans on its lower-bound numbers instead.

That kind of upward revision is no minor correction. Insects are the backbone of most land ecosystems, pollinating plants, breaking down dead matter, and feeding countless other animals. If scientists have been working with an estimate less than half the true count, the scale of what remains unknown about life on Earth is enormous. It also raises the stakes for ongoing worries about insect population declines worldwide.

A Costa Rican Forest as a Window to Global Insect Species

An extraordinary foundation made the study possible. Fifteen insect traps, large mesh tent-like structures designed to intercept flying insects, were set up along a roughly 100-kilometer stretch of Costa Rica’s Área de Conservación Guanacaste, a protected region running from Pacific lowland dry forest up through cloud forest and into Caribbean rainforest. Those traps ran for a combined 69 trap-years, where one trap running for one full year equals a single trap-year.

Every insect caught had its DNA analyzed. That work yielded 1,633,855 individual specimens representing 53,945 species, identified not by physical description alone but by unique genetic signatures called DNA barcodes.

Even that enormous haul could not capture every species in the area. In tropical environments, rare species vastly outnumber common ones, and intensive sampling keeps turning up new ones. So the research team needed a way to estimate how many species they were missing.

The Wasp That Unlocked the Global Insect Species Estimate

To gauge the extent of the undercount, the researchers focused on one well-studied group: a subfamily of parasitic wasps called Microgastrinae. These wasps lay their eggs inside caterpillars and rank among the most species-rich groups of parasitic wasps on Earth. The team had unusually thorough data on them from three separate sources: the 15 main traps, 15 additional traps spread across the region, and 11,373 individual wasps raised from roughly 1,500 species of caterpillars collected by hand over more than 40 years.

Combining all three sources, the researchers counted 1,414 wasp species from this one subfamily. They then applied a statistical method first developed for tracking disease outbreaks, one that combines multiple detection methods to estimate how many cases were never caught by any single one, to produce a lower-bound estimate of the true number of wasp species in the region. That estimate came out to 2,394 species, meaning the 15 main traps alone had detected only about 388 of them, roughly 16 percent of the likely true total.

That ratio became the key to everything else. Rather than claim they had shown it, the researchers assumed, as a deliberately conservative step, that this same level of undersampling also applied to the rest of the insects caught in those traps. The paper names the point as an assumption rather than an observed fact. Applied to the 53,945 insect species observed in the core traps, the correction produced an estimate of nearly 333,000 insect species living in the Costa Rican study area alone.

Scaling Up to the Whole Planet

Getting from one protected area in Costa Rica to a global figure took one more step. Researchers compared the number of tree species in the study area to global estimates of tree species on Earth, then used that ratio to project how much more insect diversity the whole planet likely holds. They repeated the approach using mammals, amphibians, and a group of large silk moths as alternative reference points.

Across all four reference groups, global insect estimates ranged from about 14.2 million to 20.3 million species, averaging 17.3 million. The team’s preferred estimate, based on trees, came in at just over 20 million. An earlier estimate by other researchers, using a completely different approach, had landed on 21 million, two independent methods pointing to nearly the same territory.

Every major choice in the analysis was designed to run conservative, a lower bound rather than an upper one. Traps sat at ground level, missing the forest canopy, where a large share of insect species spend their entire lives. Prior research at other tropical sites found that ground-level traps missed roughly 38 to 51 percent of certain insect groups living in the canopy. Beetles, ants, and some bees are also poorly captured by the type of trap used here. Correcting for any of those gaps would only push the global estimate higher.

Four Decades of Caterpillar Rearing Made It Possible

A large portion of the study’s data came from one of the most patient long-term biological projects in modern science. Researchers and local field workers in Costa Rica have been collecting caterpillars by hand and raising them to adulthood since 1984, a program now spanning more than 40 years. In all, 531,453 caterpillars were reared through the effort, a scale the paper’s authors describe as unmatched by any other known study. Wasps that emerged from those caterpillars turned out to form a largely separate community from the wasps caught in the mechanical traps, so the two methods were not duplicating each other. Each revealed a different slice of the same hidden world.

Six million was already a number too large to fully picture. The evidence now points to at least two to three times more, most of it still unnamed, and many species may disappear before scientists have the chance to describe them.