Abstract
Keystone species are disproportionately important for ecosystem functioning (1,2). However, while all species engage in multiple interaction types with other species, the importance of keystone species is often defined based on a single dimension of their Eltonian niche (3), that is, one type of interaction (e.g., keystone predator). Therefore, it remains unclear whether the importance of keystone species is unidimensional or if it extends across interaction types. We conducted a meta-analysis of tripartite interaction networks to examine whether species importance in one dimension of their niche is mirrored in other niche dimensions, and whether this is associated with interaction outcome, intimacy, or species richness. We show that keystone species importance is positively associated across multiple ecological niche dimensions, independently of species’ abundance, and find no evidence that multidimensionality of keystone species is influenced by the explanatory variables. We propose that the role of keystone species extends across multiple ecological niche dimensions, with important implications for ecosystem resilience and conservation.
Significance Statement Keystone species are often identified by focusing on a single type of interaction (e.g., predation, pollination, herbivory) which contrasts with the multiple roles that species play in biological communities. We conducted a meta-analysis of 18 tripartite interaction networks to explore if keystonness is correlated across the multiple dimensions of species Eltonian niches. Our results suggest that species importance tends to span across multiple interaction types, independently from abundance, which can be key to understand community resilience and collapse in face of multiple threats.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
Competing interests: Authors declare no competing interests.
Data accessibility statement: Datasets used in this work are available from figshare: https://doi.org/10.6084/m9.figshare.14333198.
R code used to perform the statistical analysis and generate plots is available from figshare: https://doi.org/10.6084/m9.figshare.14334038.
Formulas used in the methods moved to Supplementary Information. Effect of species richness on the results of the meta-analysis added to main text. Correlation between estimated "effective abundances" and estimated empirical abundances, for selected datasets, added in Suppplementary Information. Other minor changes to text, but not in the content itself.
