Stephen Garnett and Les Christidis argue for more governance in taxonomy to make species boundaries more stable (Nature 546, 25–27; 2017). There may be a better way to impose stability.

Taxonomy seeks natural groups by using different proxies, such as gene trees, morphology, ecology and behaviour, and testing them for strong correlations. As a result of these treatments, taxon delimitations are proposed and groups are given names according to the rules of nomenclature codes. The taxon names themselves are stable because they are defined solely by their type — such as a type specimen, a drawing of a species, or the genus that typifies a family. The application of names may, of course, be reassessed in light of new evidence.

We therefore suggest that, instead of listing and protecting species names, conservation should introduce the practice of referencing the taxonomic treatments used to apply the names to specific taxon concepts. This practice would provide stability for taxon boundaries in efforts to conserve biodiversity.

It would also boost incentives to provide and refine tools for recognizing and delimiting taxa and to deal with biodiversity beyond the category we designate as 'species'. Taxonomy would be rightly recognized as a fundamental science, and conservation would have a documented scientific basis for its work — without excessive bureaucracy (see P. M. Hollingsworth Nature 546, 600; 2017).