Main text

In a recent study (Brun et al. 2022), we used an empirical screening approach to identify species in the French Alps and Switzerland whose abundance (1) contributed most to ecosystem productivity given environmental conditions and (2) was positively associated with productivity. We then demonstrated that the top five of these key species jointly contributed more to explaining productivity than common measures of functional community composition and diversity, and that key species were typically tall and showed high specific leaf area. However, the relationships we identified are highly dependent on environmental context, with marked deviations in low-productivity environments.
Bruun & Ejrnaes (2022) point out that our results may allow inferring how many species can go extinct before ecosystems collapse, and that we missed evaluating our findings in this light. They therefore combine data from our analysis with information on red-listed species in the study area, and demonstrate that red-listed species have significantly lower occurrence frequencies and local abundances than key species and would thus contribute less to ecosystem productivity. From these findings, they conclude that few key species may suffice to maintain productivity, and ecosystem processes in general, and that most species can go extinct without hampering the efficiency of ecosystem processes.
We believe that these conclusions overinterpret our results. Although the contrasting occurrence frequencies between key species and red-listed species are striking, they are not surprising. We focused on the most important species at regional scale and assessed to which degree their abundance serves as a proxy for productivity after accounting for environmental conditions, and such species need to be common to a minimum degree. In order to be red-listed, on the other hand, species need to be geographically restricted or locally scarce (UNEP-WCMC and IUCN 2016), so by definition key species and red-listed species hardly overlap in occurrence frequency. Does this mean red-listed species are irrelevant to productivity and ecosystem functioning in general? Most of the red-listed species are part of the 83% least frequent species that we did not investigate. That red-listed species’ contributions to productivity are smaller than those of key species can be expected, and our results are not even necessary to prove this. But this is just a ranking with regards to productivity that by no means implies that disappearing red-listed species hardly affect ecosystem processes in general, and especially on the long run.
A fair assessment of the contributions of rare and red-listed species should focus on per-abundance effects rather than per-species effects, and in this respect a large body of literature highlights that they often are disproportionately important. Geographically restricted and/or locally scarce species have been shown to be important drivers of grassland multifunctionality (Soliveres et al. 2016) and ecosystem stability (Säterberg et al. 2019). Moreover, rare and red-listed species disproportionately shape the functional structure of ecosystems (Leitão et al. 2016), as they often exhibit distinct combinations of traits (Mouillot et al. 2013; Loiseau et al. 2020). Such functionally distinct species further promote forest productivity in harsh environments (Delalandre et al. 2022) and ecosystem multifunctionality across biomes (Le Bagousse-Pinguet et al. 2021). Although per-species effects of rare species may currently be smaller than those of key species, relative abundances are dynamic (Loreau et al. 2003), and rare species today may become key in the changing world of tomorrow - as long as they manage to survive (Yachi & Loreau 1999).
Research on biodiversity and ecosystem functioning advances by better understanding the system, not by debating about the validity of established hypotheses. In our analysis, we identified high importance of key-species abundance to explain ecosystem productivity and we found support for the ‘mass ratio hypothesis’ (Grime 1998), but we also found that trait diversity has a positive association with productivity, in line with the ‘complementary resource use hypothesis’ (Naeem et al. 1994). A joint relevance of both hypotheses has been found repeatedly for productivity (Sonkoly et al. 2019; Gao et al. 2021) and other ecosystem processes (García‐Palacios et al.2017; Le Bagousse-Pinguet et al. 2021), and countless studies exist that support each of these two hypotheses individually. It therefore appears appropriate to view the ‘mass ratio hypothesis’ and ‘complementary resource use hypothesis’ as independent axioms in the context of multifunctionality: some ecosystem functions are driven more by averages of traits while others primarily respond to trait diversity or complementarity. What matters is our ability to understand ecosystem processes and how species are contributing to them, and this ability is unnecessarily limited by presuming that the main question is mass ratio or complementarity.