How Many Different Species Are There on Earth?

The most widely cited scientific estimate puts the number of eukaryotic species on Earth at roughly 8.7 million, but that figure only covers organisms with complex cells (animals, plants, fungi, protists) and leaves out the microbial world entirely. When bacteria, archaea, and viruses enter the picture, credible estimates range from tens of millions to as high as a trillion. What makes the question so slippery is that most life on this planet has never been formally described: according to one landmark analysis, about 86% of land species and 91% of ocean species have yet to receive a scientific name.

Where the 8.7 Million Figure Comes From

In 2011, a team led by Camilo Mora used the consistent mathematical pattern in how higher taxonomic groups (phyla, classes, orders, families) relate to the number of species within them to predict total species counts. Their model yielded approximately 8.7 million eukaryotic species, give or take about 1.3 million. At the time, just over 1.2 million species had been catalogued in central databases, which meant the vast majority were still waiting to be found and named.¹ That gap has narrowed only slightly in the years since. Description rates have recently accelerated, with the largest number of new species per year concentrated in the past two decades, but the backlog remains enormous.²

It is worth stressing what this number includes and what it leaves out. The 8.7 million figure covers animals, plants, fungi, and various protist lineages. It does not attempt to count bacteria, archaea, or viruses. If you want a single number for “all life,” you need to add those groups, and that is where the uncertainty explodes.

Insects and Arthropods Dominate the Count

No group contributes more to global species richness than the arthropods, and among them, insects are king. Early estimates from the 1980s, based on the number of beetle species found in tropical tree canopies, extrapolated to as many as 30 million arthropod species worldwide. That figure turned out to rely on an assumption of high host specificity among herbivorous insects that later fieldwork did not support. A study of tropical herbivore-host relationships found that host specificity was much lower than assumed, which brought the global arthropod estimate down to a range of about 4 to 6 million species.³ More recent analyses using multiple independent approaches have converged on similar figures. Even at the conservative end, insects alone likely number several million species, with beetles, flies, wasps, and moths making up the bulk. What is remarkable is that despite being the most species-rich class on the planet, insects remain among the most undersampled in biodiversity databases, as we will see later.

The Hidden Kingdom of Fungi

Fungi are one of the biggest wildcards in global species estimates. A 1991 landmark estimate placed the total at 1.5 million species, based on ratios of known fungi to known plants in well-studied regions like the United Kingdom. At that time, only about 70,000 fungal species had been formally described. Since then, the estimate has climbed steeply. Molecular sequencing from environmental samples (DNA extracted from soil, water, and decaying material rather than from cultured organisms) has revealed enormous hidden diversity. One analysis, incorporating these high-throughput sequencing results, suggested the true number could be as high as 5.1 million species.⁴

A separate review arrived at a range of 2.2 to 3.8 million fungal species by revisiting the fungi-to-plant ratio. The original 1991 estimate used a ratio of about 6 fungi for every plant species, but newer field surveys and molecular studies have found ratios as high as 19 to 1 in some sites. The mean across studies is closer to 10 to 1. Applying even a conservative version of that updated ratio to the roughly 380,000 known vascular plant species pushes the prediction near 3.8 million fungal species.⁵ A huge share of these unknown fungi likely live in soil, inside the tissues of plants and insects, or in deep-sea sediments, places that have barely been sampled with modern molecular tools.

Microbes Change Everything

If counting eukaryotes is hard, counting microbes is close to impossible with current tools. Bacteria, archaea, and viruses reproduce fast, swap genes laterally, and often cannot be grown in a lab, which makes traditional species concepts a poor fit. Still, researchers have tried to put numbers on it, and the results vary wildly depending on methodology.

One census-based approach, which compiled DNA sequence data from thousands of environmental studies, estimated roughly 1 to 2 million distinct bacterial and archaeal groupings on Earth.⁶ That number, while large, is far more conservative than some alternatives. A separate study using ecological scaling laws (the mathematical relationship between community size and diversity) predicted that Earth could harbor upward of a trillion microbial species.⁷ The difference between “a couple of million” and “a trillion” is not a rounding error. It reflects genuine disagreement about how to define a microbial species, how to account for extremely rare lineages, and how far you can extrapolate from local samples to the entire planet.

Even at the scale of a handful of dirt, the numbers are staggering. An analysis of DNA from just 30 grams of forest topsoil in Norway estimated roughly a million bacterial species in that single sample, based on the complexity of reassociating the extracted DNA and the expected ratio of rare to common species.⁸ If a clump of soil the size of a golf ball can contain that much diversity, extrapolating to every habitat on the planet, from ocean sediments to the atmosphere to hot springs, produces numbers that dwarf all of macroscopic biology.

Viruses add yet another layer. Because viruses infect every cellular organism, their diversity tracks and likely exceeds that of their hosts. Combining estimates of prokaryote species with what is known about the number of viral types per host, researchers have produced back-of-the-envelope figures ranging from about 10 million to a billion distinct virus species globally.⁹

Why the Ocean Has Fewer Named Species Than You Would Expect

The oceans cover about 70% of Earth’s surface, have a far longer evolutionary history than terrestrial habitats, and are home to more high-level lineages (phyla and classes) than land and freshwater combined. You might expect the sea to house the lion’s share of species. It does not. Only about 16% of all named species are marine.¹⁰ Terrestrial environments, which occupy roughly 30% of the planet’s surface, contain an estimated 80% of all species.¹¹

Part of the explanation is biological: life on land diversifies faster. The physical fragmentation of terrestrial habitats by mountains, rivers, deserts, and climate zones creates more opportunities for populations to become isolated and diverge into new species. The ocean, by contrast, is comparatively well connected, which slows that process. An analysis of diversification rates across animal phyla found strong support for the idea that land-dwelling lineages split into new species at faster rates than their marine relatives.¹¹ Insects, which are almost entirely terrestrial, account for a huge fraction of global species and have no marine equivalent in terms of sheer richness.

But the gap is also partly a sampling artifact. The deep sea, which makes up the largest habitat on Earth, is barely explored. A survey of the lower bathyal and abyssal seafloor off eastern Australia collected over 25,000 specimens and identified them to more than 1,000 taxonomic groups, but roughly 58% of the species-level specimens could not be matched to any described species.¹² Biodiversity on the deep-sea floor is among the highest on the planet, composed mostly of small invertebrates that have never been seen by a taxonomist.¹³ As deep-sea sampling improves, the apparent dominance of land over ocean in species counts will likely narrow, though probably not disappear.

Cryptic Species and What DNA Has Revealed

One of the most consequential developments in biodiversity science over the past two decades is the routine use of DNA barcoding to identify species. Before molecular tools, taxonomists relied on physical differences (shape, size, color, anatomy) to distinguish one species from another. That approach works reasonably well for birds and mammals, where species tend to look distinct, but it misses organisms that are physically identical yet genetically and ecologically separate. These are called cryptic species, and they turn out to be everywhere.

A now-classic study of a single skipper butterfly species in Costa Rica, Astraptes fulgerator, used DNA barcodes combined with 25 years of natural history data to show that what had been treated as one species was actually a complex of at least 10.¹⁴ A similar result came from another neotropical skipper, Udranomia kikkawai, where complete nuclear genome sequencing confirmed that what looked like one widespread species was actually three, each with distinct ecology and subtle differences in wing pattern that had been overlooked.¹⁵ The phenomenon is not restricted to butterflies or to the tropics. Cryptic species have been uncovered in freshwater fish from the Qinghai-Tibet Plateau,¹⁶ in marine invertebrates, in nematodes, and across nearly every group that has been examined with molecular tools.

The implication for global totals is straightforward: if one “species” routinely turns out to be several, then every existing species count is an underestimate, potentially a serious one. Large-scale barcoding inventories of tropical insects have led researchers to conclude that cryptic species are a widespread phenomenon and that further molecular studies will substantially increase current estimates of insect species richness.¹⁵

What Counts as a Species Depends on Who You Ask

Underlying all of these estimates is a surprisingly contentious question: what is a species? The definition matters enormously for the final tally. When researchers compared biodiversity surveys that used different species concepts, surveys based on a phylogenetic approach (which defines species by unique evolutionary lineages in a family tree) consistently found about 48% more species than those using other frameworks, along with corresponding decreases in each species’ population size and geographic range.¹⁷ That is not a trivial difference. Simply changing the yardstick by which you measure species can inflate or deflate biodiversity estimates by nearly half. For microbes, the problem is even worse, since bacteria do not reproduce sexually and thus cannot be defined by the traditional test of whether two organisms can interbreed.

Taxonomic Bias in What Gets Studied

Not all organisms are equally likely to be discovered, described, or monitored. There is a dramatic skew in biodiversity data toward large, charismatic, or easy-to-observe species. An analysis of records in the Global Biodiversity Information Facility (GBIF) found that birds accounted for more than half of all occurrence records, despite representing only about 1% of the species catalogued. The median number of occurrence records per bird species was 371. By contrast, arachnids, which are about three times more species-rich than birds, had a median of just 3 records per species. Insects, the most diverse class on the planet, were among the most underrepresented in the database relative to their known richness.¹⁸

This bias matters for conservation and for the accuracy of species estimates alike. If the groups with the most undiscovered species are also the groups receiving the least attention, then the gap between known and total biodiversity is wider than any single number suggests. Fungi, soil invertebrates, and marine microorganisms all fall into this neglected category. The imbalance also means that extinction risk assessments are heavily skewed toward vertebrates, leaving the vast majority of life on Earth without any formal conservation evaluation.

Dark Extinctions and the Race Against Time

A particularly sobering dimension of the species-counting problem is that some species go extinct before they are ever discovered. Researchers call this “dark extinction.” Intensive fieldwork in biodiversity hotspots sometimes turns up species so rare and threatened that they vanish within years of their first description. In Brazil’s Atlantic coastal forests, recent bird extinctions suggest that some species disappeared before anyone had a chance to formally recognize them.¹⁹

The problem is worse for the period before systematic scientific recording began. The arrival of Europeans on oceanic islands, and the rats, goats, and cats they brought along, triggered waves of extinctions centuries before naturalists started cataloguing local species. Statistical methods applied to world birds estimate roughly 56 dark extinctions during the taxonomic period (from about 1800 to the present) and about 180 more in the pre-taxonomic period (1500 to 1800).²⁰ Birds are among the best-documented groups on the planet, so the numbers for less-studied organisms, like insects, fungi, and marine invertebrates, are almost certainly higher in proportion.

Parasites as a Diversity Multiplier

One group that rarely gets its due in popular discussions of biodiversity is the parasites. Parasitism is the most common lifestyle among animal species, and its implications for total species counts are enormous. Several local-scale studies have found that every free-living animal species hosts at least one parasite species unique to it. If that pattern holds globally, then roughly half of all animal species could be parasites.²¹ Parasitic worms, parasitic wasps, mites, fleas, protists, and fungi that live inside or on other organisms represent a massive layer of diversity that is systematically undercounted because parasites are small, hard to find, and often can only be identified by dissecting or molecularly sampling their hosts.

The discovery rate for parasites lags far behind that of their hosts, which means the true ratio of parasites to free-living species is still unknown. But even conservative assumptions suggest that leaving parasites out of the conversation gives a seriously incomplete picture of life on Earth.

Life Underfoot

Soil is one of the most species-dense habitats on the planet, yet it remains poorly characterized compared with forests, coral reefs, or grasslands. Nematodes (roundworms) are a useful window into soil diversity because they are found in virtually every terrestrial habitat and occupy a wide range of ecological roles, from bacterial feeders to plant parasites to predators. A molecular survey of nematode communities in a Costa Rican lowland rainforest estimated between 464 and 502 nematode species within four small sampling plots, each just 40 by 40 meters.²² Extrapolating that kind of local richness across all soil habitats worldwide produces numbers that are difficult to fathom, and nematodes are only one slice of the soil food web. Mites, springtails, tardigrades, fungi, and bacteria all occupy the same handful of dirt, each group contributing its own layer of diversity.

Global patterns of soil nematode assemblages show wide variation in family richness, from as few as 1 to as many as 30 families per site, driven largely by soil moisture and bulk density rather than by any single climatic variable.²³ The lack of a strong, simple predictor of soil biodiversity is itself revealing: it suggests that local conditions matter more than broad geography, making it very hard to estimate total soil species numbers from satellite images or climate maps alone.

Building a Global Checklist

Given the scale of the unknown, simply keeping track of what has been described is a major logistical challenge. The Catalogue of Life, developed jointly with the Global Biodiversity Information Facility, is the most ambitious attempt at a comprehensive species checklist. It currently integrates over 160 supporting data sources and relies on a mix of expert-curated taxonomic databases, regional checklists, and informatics tools to harvest and quality-control species names.²⁴ Even so, coverage is uneven. An analysis of GBIF’s holdings for a single well-known plant family, the legumes, found that only about 31% of recognized legume species had even one georeferenced record in the database.²⁵ If one of the most economically important plant families on Earth is only a third documented in the world’s largest biodiversity database, the coverage for less conspicuous groups is likely far worse.

The practical consequence is that any map or model that relies on existing data to predict where biodiversity is highest will be biased toward well-studied regions and well-studied taxa. Tropical forests, deep-sea floors, and soil ecosystems are all likely to be underweighted, which in turn affects how conservation priorities are set.