The percentage depends entirely on which number you trust, and the range is enormous. Official tallies from the International Union for Conservation of Nature (IUCN) list fewer than 900 confirmed extinctions since the year 1500, which works out to roughly 0.04% of known species. But researchers who account for unrecorded losses, poorly studied groups, and prehistoric human impacts estimate the real figure could be somewhere between 7% and 13% of all species on Earth. That gap between the official count and the likely reality is itself one of the most important facts about the biodiversity crisis.
Why the Official Count Looks Deceptively Small
The IUCN Red List is the most widely cited authority on species status, and it puts confirmed modern extinctions at 882 species, about 0.04% of the roughly two million species that have been formally described and named. That number is strikingly low, and it gives some commentators the impression that the extinction crisis is overstated. But the number reflects a documentation problem more than an ecological reality. Declaring a species extinct requires meeting strict criteria: exhaustive surveys of its known habitat, years of failed search efforts, and expert consensus. For most of the world’s organisms, that level of scrutiny has never been applied. Invertebrates make up about 99% of animal biodiversity, yet the status of a negligible fraction of them has been formally assessed.1PubMed. Extinction in a hyperdiverse endemic Hawaiian land snail family and implications for the underestimation of invertebrate extinction The IUCN number is not wrong so much as radically incomplete. It tells you how many extinctions we can prove beyond reasonable doubt, not how many have actually occurred.
Nobody Agrees on How Many Species Exist
Before you can calculate a percentage of species lost, you need a denominator, and that denominator is itself a matter of active scientific debate. One widely cited 2011 analysis estimated roughly 8.7 million species of eukaryotes on Earth, with about 7.8 million of those being animals.2PLoS Biology. How Many Species Are There on Earth and in the Ocean? But other researchers using different methods have arrived at much lower figures. One approach based on the rate at which new species are being described predicted only about 1.8 to 2 million species total.3PubMed. Predicting total global species richness using rates of species description and estimates of taxonomic effort A recent review noted that projections span from millions to trillions, depending on whether you include microorganisms.4PubMed Central. How many species are there on Earth? Progress and problems.
This uncertainty matters because it directly affects any extinction percentage. If the true number of species is 2 million and 130,000 have gone extinct, that is 7%. If the true number is 8.7 million and the same 130,000 have gone extinct, it is 1.5%. When you see different researchers quoting different extinction percentages, they are often not disagreeing about how many species are gone. They are disagreeing about how many existed in the first place, or about how far you can reasonably extrapolate from the groups we know well to the groups we do not.
What Extrapolations from Well-Studied Groups Suggest
Because most of biodiversity has never been formally cataloged, researchers have tried to estimate the true scope of extinction by studying groups with better fossil or collection records and then scaling up. Two independent efforts have converged on similar ballpark figures, which lends them some credibility.
One team focused on land snails, using a random sample of species to estimate how many had disappeared. They concluded that roughly 7% of all species on Earth, about 130,000, may have already gone extinct. That figure is masked, the researchers argued, by conservation biology’s heavy focus on vertebrates, which make up a tiny fraction of total biodiversity.5PubMed Central. Mass extinction in poorly known taxa A separate analysis focused on molluscs as a whole, the second-largest animal phylum by known species count. By extrapolating from documented mollusc losses, the authors estimated that 7.5% to 13% of all roughly two million known species may have gone extinct since around 1500, which would mean between 150,000 and 260,000 species lost.6PubMed Central. The Sixth Mass Extinction: fact, fiction or speculation?
These are deliberately bold extrapolations, and the researchers behind them say so. The logic is that molluscs and snails are diverse enough and well-enough collected over centuries that their extinction patterns can serve as a rough proxy for life in general. Whether that proxy holds for insects, fungi, or marine organisms is genuinely uncertain. But even if the real figure turns out to be closer to the low end of these ranges, it would still mean extinctions orders of magnitude greater than the official IUCN tally.
Birds Tell the Story Most Clearly
Among vertebrates, birds offer the most complete extinction record because they are well studied, conspicuous, and found on every continent. Even here, the documented losses are known to be underestimates. The IUCN lists about 187 bird species as extinct or suspected extinct since 1500.7Nature Conservation. Assessment of the threat status of bird species from Vietnam – Implementation of the One Plan Approach to conservation But fossil and subfossil evidence tells a much bigger story. A 2023 study combining recorded extinctions with statistical modeling of the fossil record estimated that at least 1,300 to 1,500 bird species, about 12% of all birds, have gone extinct since the Late Pleistocene. More than half of those losses were never formally documented because the species vanished before Western science reached them or left no trace in the fossil record. The Pacific Islands alone accounted for roughly 61% of all bird extinctions.8PubMed Central. Undiscovered bird extinctions obscure the true magnitude of human-driven extinction waves
The pattern on Pacific islands is particularly stark. When humans first colonized remote Pacific islands, two-thirds of bird populations on those islands were wiped out before European contact even began. The colonization of the Pacific caused the global extinction of close to 1,000 species of nonpasserine landbirds alone, with seabird and songbird losses adding further to the total.9PubMed Central. Magnitude and variation of prehistoric bird extinctions in the Pacific These were not species driven to extinction by industrial pollution or climate change. They were hunted, their eggs collected, their forests cleared, and their island ecosystems transformed by the arrival of rats and dogs that came with human settlers.
Prehistoric Extinctions and the Long Human Shadow
Most discussions of human-caused extinction focus on the period since 1500, when written records began to accumulate. But the human impact on global biodiversity started tens of thousands of years earlier. The end of the last ice age saw a wave of large-mammal extinctions across the Americas, Australia, and Eurasia that has long been debated. The “overkill hypothesis” argues that human predation was the primary driver, and the available evidence from paleontology and archaeology is largely consistent with that explanation. Climate-based explanations struggle to account for the timing, geographic extent, and the fact that large animals were hit hardest while smaller species survived.10Quaternary Research. The Overkill Hypothesis as a Plausible Explanation for the Extinctions of Late Wisconsin Megafauna
This matters for the percentage question because if you only count losses since 1500, you miss a huge portion of the damage humans have done. Woolly mammoths, giant ground sloths, most of Australia’s megafauna, the moa of New Zealand, the elephant birds of Madagascar: none of these show up in IUCN statistics, but their disappearance reshaped ecosystems across entire continents. Including these prehistoric losses pushes the overall percentage of human-caused extinctions higher, though exactly how much higher is hard to pin down because the baseline species count for the Late Pleistocene is itself uncertain.
Dark Extinctions and the Species We Never Knew
There is a category of loss that is especially hard to quantify: species that went extinct before they were ever described by science. Researchers call these “dark extinctions.” For birds, one analysis estimated that roughly 56 dark extinctions occurred during the “taxonomic period” from 1800 to the present, a span when ornithologists were actively describing new species. Another 180 or so dark extinctions are estimated for the earlier “pre-taxonomic period” from 1500 to 1800.11PubMed Central. Dark extinction: the problem of unknown historical extinctions These are species that existed, lived alongside humans, and vanished without ever making it into a scientific catalog.
For less charismatic groups like insects, fungi, and deep-soil invertebrates, dark extinctions could dwarf the documented ones by a wide margin. If a beetle species endemic to a single tropical ridge goes extinct when that ridge is deforested, and no entomologist ever collected a specimen, the extinction is invisible. It does not appear in any database. The species never had a name, never had a Red List assessment, and never will. This is why researchers emphasize that any global extinction percentage based solely on documented losses is necessarily a floor, not a ceiling.
Freshwater Species Are Hit Hardest
Extinction risk is not evenly distributed across ecosystems. Freshwater habitats, which cover a small fraction of Earth’s surface but harbor an outsized share of biodiversity, are among the most affected. About 30% of all freshwater mammals are listed as threatened, with population trends declining faster than for mammals as a whole.12PubMed. Extinction risk of the world’s freshwater mammals Among freshwater fish, which account for more than half of all fish diversity globally, nearly one-third of existing species face extinction risk.13PubMed Central. Global Patterns and Drivers of Freshwater Fish Extinctions: Can We Learn From Our Losses? Rivers, lakes, and wetlands are simultaneously dammed, polluted, drained for agriculture, and invaded by nonnative species, creating a layered set of threats that is difficult to untangle.
Marine species, by contrast, have experienced far fewer confirmed extinctions: only about 20 to 24 marine species are recorded as extinct over the past 500 years.14PubMed Central. A review of recent and future marine extinctions Per assessed species, extinction and threat rates in the ocean are several times lower than on land.15Current Biology. Global Patterns of Extinction Risk in Marine and Non-marine Systems Whether marine species are genuinely more resilient or simply harder to monitor to extinction is an open question. Ocean species tend to have larger ranges and more dispersive larvae, which may buffer them against the kind of localized habitat destruction that wipes out island or freshwater species. But only about 3% of described marine species have been formally assessed, so the picture is far from complete.
What Is Driving the Losses
Habitat destruction stands above everything else as the dominant cause of species decline and extinction. A large-scale analysis of over 20,000 threatened species found that about 88% were affected by habitat destruction, and for more than 71% it was the single main factor pushing them toward extinction. Overexploitation was the dominant threat for about 7% of species, invasive species for about 7%, pollution for about 5%, and climate change for under 2%.16Conservation Science and Practice. The greatest threats to species Agriculture is the largest component of that habitat destruction.17PubMed Central. Food impacts on species extinction risks can vary by three orders of magnitude
The wildlife trade adds another layer. Unsustainable harvesting directly removes individuals faster than populations can replace them, and a perverse economic dynamic can make the problem worse: as a species becomes rarer, its market value can rise, making it profitable to hunt down the very last individuals.18Journal of Applied Ecology. Multiple co‐occurring bioeconomic drivers of overexploitation can accelerate rare species extinction risk Climate change, while currently responsible for a small share of completed extinctions, has been contributing to a growing proportion of observed extinctions since 1970 and is projected to threaten up to one in six species if current warming trajectories continue.19PubMed. Climate change extinctions 20PubMed. Accelerating extinction risk from climate change
The Current Rate Compared to Earth’s Baseline
One way to assess the severity of the crisis without fixating on a single percentage is to compare the current rate of extinction to the natural background rate, the pace at which species would disappear even without human interference. Using a conservative background rate of two mammal extinctions per 10,000 species per century, a 2015 analysis found that the average rate of vertebrate species loss over the past century was up to 100 times higher than that baseline. Species that have gone extinct in the last hundred years would have taken, depending on the group, between 800 and 10,000 years to disappear at background rates.21PubMed Central. Accelerated modern human-induced species losses: Entering the sixth mass extinction
Some researchers push back on using this comparison to declare a “sixth mass extinction,” arguing that spikes above background rates are not unusual in the fossil record and that past mass extinctions were far more catastrophic than anything yet observed. By this view, elevated extinction rates are concerning but do not yet meet the threshold that defines a true mass extinction.22Trends in Ecology & Evolution. Questioning the sixth mass extinction The disagreement is partly semantic: how exceptional does a spike in extinction need to be before you call it a mass extinction? But it also reflects genuine uncertainty about how the current crisis will unfold over the coming centuries. Even the skeptics acknowledge that extinction rates are abnormally high. What they dispute is whether we are at the beginning of a catastrophe comparable to the one that killed the dinosaurs, or in a severe but historically non-unique episode.
Extinctions That Have Not Happened Yet
One of the more unsettling findings in conservation biology is that the full consequences of habitat destruction are not felt immediately. When a forest is fragmented into smaller patches, many of the species living in those patches do not disappear right away. Their populations shrink, become isolated, and lose genetic diversity, but the final extinction can take decades or even centuries to play out. Researchers call this delayed loss “extinction debt.”
A pan-European study of 147 fragmented grassland remnants found that the number of specialist plant species alive today was better explained by the landscape as it existed in the past than by its current configuration. In other words, plants were still present that the current habitat could no longer sustain in the long run. These species were, in a sense, the living dead, persisting on borrowed time.23PubMed Central. Habitat fragmentation causes immediate and time-delayed biodiversity loss at different trophic levels A more recent study found that extinction debt varies by the role a species plays in its ecosystem: herbivores in small, isolated habitats carried the highest delayed-extinction burden, while species higher up the food chain showed different patterns depending on how connected their habitat patches were.24PubMed Central. Extinction debt of species and ecological interactions in a fragmented landscape
Extinction debt means that even if all deforestation and habitat destruction stopped tomorrow, some fraction of the species alive today are already committed to extinction. The percentage of species humans have driven extinct is not just a historical figure. It is still accumulating.
When One Extinction Triggers Another
Species do not exist in isolation, and when one disappears it can drag others down with it. Coextinction occurs when a species that depends on another, such as a parasite that can only live on one host, or a plant that can only be pollinated by one insect, vanishes because its partner disappears. Models suggest that coextinction may actually be the most common form of biodiversity loss, because dependent relationships are everywhere in nature and most of them involve species that have never been cataloged.25Proceedings of the Royal Society B. The sixth mass coextinction: are most endangered species parasites and mutualists? If a tropical tree goes extinct and it had a specialized fungal partner, a host-specific beetle, and a pollinating moth found nowhere else, you have lost four species but documented only one.
This cascading dynamic makes the true percentage of human-caused extinctions even harder to estimate. The extrapolation studies described earlier try to capture some of this hidden loss, but the web of dependencies in most ecosystems is so poorly mapped that any number should be treated as approximate.
What Gets Lost Besides a Species Count
Counting species lost treats all extinctions as equal, but they are not. Some extinct species were evolutionary one-offs, the sole survivors of ancient lineages with no close living relatives. Losing such a species erases a disproportionate chunk of evolutionary history. A 2024 study of bird extinctions found that the 610 species lost since the Late Pleistocene took with them roughly three billion years of unique evolutionary history and caused a disproportionate reduction in the total range of ecological roles birds fill.26PubMed. The global loss of avian functional and phylogenetic diversity from anthropogenic extinctions In other words, the species we have lost were not a random draw from the tree of life. They were disproportionately the unusual ones, the species that occupied ecological niches nobody else fills.
This means the ecological consequences of extinction are larger than the raw species count implies. Losing ten closely related sparrow species is a tragedy, but it removes a relatively small slice of functional diversity. Losing ten species from ten different ancient lineages reshapes entire ecosystems. How much evolutionary history is at risk from future extinctions varies among groups and regions, but the pattern is consistent: the species most vulnerable to human pressure tend to be those whose loss is hardest to replace.27PubMed. Predicting loss of evolutionary history: Where are we? Botanists have noted a similar blind spot for plants: plant extinctions are as severe as animal extinctions in proportional terms, but they are far more poorly documented, making it even harder to assess what has been lost.28Botanical Journal of the Linnean Society. Plant extinction in the Anthropocene