Identifying typical and early warning species by the combination of functional-based diagnostic species and dark diversity

Abstract

The European Union Habitats Directive requires the identification of typical species that reflect the structure and functions of habitat types, as well as early changes in the habitat condition, but no common methods are available for their selection. Diagnostic species with high fidelity to a specific group of plots are identified by traditional methods, but their value as typical species is still debated. We designed a protocol for the identification of typical plant species based on a recently proposed method to detect diagnostic species by combining abundances and functional traits. We tested the method on a set of alpine habitat subtypes, comparing diagnostic species based on traits or Grime’s CSR strategies (competitive, stress-tolerant, ruderal) with those based on presence/absence or abundance only, and then we calculated for each species the dark diversity probability—i.e. probability of being absent from a habitat type with suitable ecological conditions. Functional-based methods allowed to recognize larger sets of exclusive species, adding dominant species linked to the structure and functions of habitat subtypes (i.e. to the functional centroid). Dark diversity probability was equally distributed between diagnostic and non-diagnostic species identified by functional-based methods. Species with higher dark diversity probability among those associated with the functional centroid can be considered as early warning indicators of changes in habitat conditions. The protocol proposed here enables species ranking on measurable variables (functional association, dark diversity probability) and can be applied as a standardized tool for the identification of typical plant species for habitat types dominated by plants.