Indices of phylogenetic and functional diversity: Understanding their applicability in ecology and conservation science

Background/Question/Methods

Diversity is often measured as the number of species in a community or region, also known as species richness. The term ‘taxonomic diversity’ (TD) also include diversity indices such as Shannon and Simpson, which factor in species’ relative abundances. However an increasing number of studies nowadays include other facets of diversity, namely phylogenetic diversity (PD) (i.e. the amount of evolutionary history accumulated in a community) and functional diversity (FD) (i.e the diversity of life history or functional traits in a community). This comes from the realization that all species are not equal, and that some are more distinct evolutionary or functionally than others. With this surge in interest, a large variety of indices have been proposed. Two challenges have been prevalent across this literature: (1) TD, PD and FD are usually highly correlated, making it difficult to understand whether and how much of the empirical patterns are independent of each other; and (2) most of the indices show counter-intuitive properties under different circumstances, making interpretation and application particularly challenging. Here I take a virtual ecologist approach to compare the most common multi-facetted diversity indices. I distinguish in particular community ecology versus conservation applications.

Results/Conclusions

This study resulted in a decision tree that should help researchers chose when to use which type of index. While some criteria can be derived from the literature (e.g. some indices can represent the three facets of diversity while others can only represent PD or FD; some can include abundance while others only consider occurrence), simulations provided hints on other considerations that are equally important. Generally the types of questions related to community ecology will require contrasting empirical patterns to random expectation, while conservation applications will be concerned with correlation and spatial congruence between indices (i.e. whether or not adding FD and PD would make us protect different species or sites). These changes in emphasis mean that diversity indices in the first case do not need to increase monotonically (i.e. in some situations their values can decrease when a new species is added without disturbing the validity of the analysis), while monotonicity is required in the second case (we would not like to use an index that decreases the value of the overall protected network when a new site is added). There is currently a gap in terms of the indices available to apply the multifaceted perspective in conservation planning.