I agree that we need to consider the diversity of species' functions, particularly those that are crucial for maintaining ecosystems (Nature 546, 22–24; 2017). However, functional diversity alone is not enough if we are to reach global conservation targets.

First, we have little idea of the functional traits of most insect species — by far the most species-rich taxonomic group. Second, we need to define what makes up a 'healthy' ecosystem (for example, it might simply reflect nature's value to humans). Third, we must not sacrifice biodiversity for functionality.

A degraded ecosystem can still have high functional value (see Nature 546, 7–8; 2017), as can a community of alien invasive species. Your example of three barnacles that may be functionally less useful than the trio of a starfish, an anemone and a seagrass is correct. But we might come to a different conclusion if one of the barnacles is threatened with extinction and the starfish and anemone are invasive.

We therefore need to include 'conservation value' as a fourth stool leg, in addition to functional, phylogenetic and species diversity. In my view, it is the most important one.