1991: where "1.5 million species" came from

In 1991 Mycological Research published David L. Hawksworth's paper "The fungal dimension of biodiversity: magnitude, significance, and conservation". It is the origin of the most frequently repeated number in all of mycology: roughly 1.5 million fungal species worldwide, against only a few tens of thousands described at the time.

The value was not a guess. It came from a ratio: in well-studied regions, above all the British Isles, the number of known fungal species per species of vascular plant turned out to be approximately constant. Applying that ratio to the global number of plants gave an order of magnitude for fungi. The method is simple, which makes it open to criticism, repetition and correction — and that is exactly what happened to it.

The number did something more than describe. It turned a vague "there are a great many fungi" into a specific claim that could be tested: if 1.5 million is right, then almost everything that grows remains undescribed and taxonomy has centuries of work ahead. Arguments about fungal conservation, about funding collections and about surveying habitats gained a reference point.

The following decades verified the figure rather than voiding it. Environmental sequencing showed that the British ratio had been conservative, and in 2017 Hawksworth, with Robert Lücking, gave a range of 2.2 to 3.8 million species. The 1991 estimate remains a model of what a number in science should look like: an explicit method, open to checking, and ready for somebody to improve it.

An estimate, not a species census

In 1991 Hawksworth contrasted roughly 69,000 species then known to science with a conservative estimate of 1.5 million, six times the earlier benchmark. The method did not count individual fruit bodies. Its core was an extrapolation from fungus-to-vascular-plant ratios in comparatively well-studied regions. Hawksworth explicitly listed reasons for treating the result as a lower bound: it used a lower estimate of global vascular-plant diversity, made no separate allowance for the potentially vast number of insect-associated fungi, drew ratios from regions that were still incompletely studied mycologically, and made no upward adjustment for tropical or polar ratios.

In 2017 Hawksworth and Lücking revised the working range to 2.2–3.8 million by combining plant–fungus ratios with environmental sequence evidence and studies that split semi-cryptic species complexes. The result is therefore a model-dependent interval, not a counter reporting the exact number of fungal species alive.