Q.How do scientists extrapolate the total number of species on Earth?

Scientists face a fundamental challenge: we have identified only about 1.5 million species so far, yet the actual number on Earth remains unknown. The extrapolation methods rest on patterns observed in well-studied groups and regions.

One approach uses the ratio of species across taxonomic levels. For instance, if insects consistently show a certain proportion of described to estimated species in thoroughly surveyed areas, that ratio can be applied globally. Another method examines the rate of new species discovery over time—plotting how many species are found each year reveals whether we are approaching completeness or still scratching the surface.

Biodiversity hotspots and sampling intensity also guide estimates. Tropical rainforests and coral reefs, though covering small areas, harbor disproportionately high species counts. By comparing species density in exhaustively studied temperate zones with preliminary data from underexplored tropical regions, scientists scale up to global figures. …

Estimating the total number of species on Earth is one of ecology's grand challenges. We have formally described and named roughly 1.5 million species so far, yet most biologists believe the true figure is many times higher—perhaps 5 million, perhaps 30 million, or even more. The problem is straightforward: we cannot count what we have not yet discovered, and new species are described every year by the thousands. So scientists have developed indirect methods to extrapolate from what we know to what likely exists.

The most intuitive approach is intensive sampling in representative areas. Researchers select a well-defined habitat—a patch of rainforest, a coral reef, a square kilometre of soil—and catalogue every species they can find using systematic collection, traps, DNA barcoding, and expert identification. They then plot species accumulation curves: as sampling effort increases, the number of new species discovered rises steeply at first, then levels off. By fitting mathematical models to these curves, ecologists can estimate how many species remain undetected in that habitat. Scale this up across similar habitats worldwide, and you get a rough global estimate. The catch, of course, is that sampling effort is wildly uneven—tropical rainforests and deep-sea ecosystems are chronically under-sampled compared to temperate forests or European meadows.

A second method exploits taxonomic ratios. Some groups of organisms are far better studied than others. Birds and mammals, for instance, are relatively well known; most large species have been described, and discovery rates have slowed. Insects, fungi, and nematodes, by contrast, are poorly known, with new species emerging constantly. Scientists compare the ratio of described to estimated species in a well-studied group, then apply that ratio to poorly studied groups. For example, if we know there are roughly 10,000 bird species and assume birds represent a certain fraction of vertebrate diversity, we can infer total vertebrate richness. Extend this logic to insects—where every tree in a tropical forest might host dozens of unique beetle species—and the numbers explode. Terry Erwin's famous 1982 study, in which he fogged rainforest canopies with insecticide and collected falling arthropods, suggested there might be 30 million insect species globally based on host-plant specificity and canopy stratification. Later refinements have moderated that figure, but the principle remains: use well-known groups as benchmarks.

Separate the methods from the final number: sampling/taxonomic-ratio/expert-survey/molecular methods are what the question asks for; the chapter's own figures (1.5M described, ~7M …