‘Eigenvector estimation of phylogenetic and functional diversity’: from patterns to processes


Correspondence author. E-mail: ksafi@orn.mpg.de

The spatial distribution of diversity on Earth was the first global pattern to be considered by ecologists (Hutchinson 1959; Stehli, Douglas & Newell 1969; Kerr & Packer 1997; Gaston 2000; Hawkins 2001; Jetz & Rahbek 2002), and describing the patterns and understanding the processes governing global diversity has been referred to as the ‘Holy Grail of Ecology’ (Huston 1994). Most often, diversity is understood as the number of species per unit of geographic space. A main shortcoming of this definition is that it poorly captures the fundamental differences in the ecological roles that species play in communities (Faith 1992; Losos & Glor 2003; Magurran 2004; Hooper et al. 2005; Wiens & Graham 2005). Assessing the different roles played by species in communities, however, is key to providing more ecologically and mechanistically realistic descriptions of biodiversity, and ultimately determines our understanding of the complex processes driving species composition and ecosystem functioning (Diaz & Cabido 2001; Hooper et al. 2002, 2005; Flynn et al. 2009).

Recent quantitative solutions to assess such different roles are to calculate phylogenetic (Faith 1992) or functional diversity (Diaz & Cabido 2001). Phylogenetic diversity incorporates the common ancestry of species into the measure of diversity. The rationale is that closely related species are more likely to occupy similar ecological niches, and thus their contribution to the local diversity should be weighed less (Faith 1992; Rodrigues & Gaston 2002). Likewise, functional diversity weighs the contribution of species according to their ecological similarity by taking into account the morphological, physiological and behavioural differences among species (Diaz & Cabido 2001). Importantly, assessing functional or phylogenetic diversity requires knowledge of life-history traits and the phylogenetic relationships (respectively) for entire communities (Diaz & Cabido 2001).

With both phylogenies and life history information becoming increasingly available (Jones et al. 2009), it is now possible not only to quantify these measures but also to empirically investigate how they relate to each other (Devictor et al. 2010; Safi et al. 2011). If we think of functional diversity as the amount of trait similarity in an ecological community, and of phylogenetic diversity as the amount of shared evolutionary time, then comparing how functional and phylogenetic diversity relate to each other should allow us to (i) directly quantify the rate of trait evolution and (ii) understand more about potential processes that generate biodiversity.

Diniz Filho et al. (2011) present a novel and integrative way to measure functional and phylogenetic diversity with more facets than previous methods were providing. By dissecting functional diversity into an acquired (specific) and an inherited (phylogenetic) component they highlight the possibility to directly quantify differences in selective pressure in a spatially explicit way. The specific component (FDs) in particular marks the proportion of the entire functional diversity contained in a community (FD) that cannot be attributed to the common ancestry of the species (FDp) and that was therefore acquired as an adaptive response to the (past and present) local environmental conditions. The decomposition of functional diversity into specific and phylogenetic components allows us to discern assemblages along an additional axis, a continuum where communities are defined from being fully determined by recent adaptive processes (FD = FDs) to communities in which phylogenetic constraints in concert with habitat filtering determined species composition (FD = FDp). This distinction is a first step if we want to understand where and under which circumstances species radiated and adapted to local environmental conditions versus under which circumstances environmental conditions determined species composition by filtering communities according to inherited characteristics that enable species to invade and persist in the local communities. This distinction is also important in view of the rapid anthropogenic environmental changes that many communities face these days. These changes might affect species assemblages defined by habitat filtering processes differently than assemblages in which recent adaptive processes have shaped species composition. In fact, energy availability (measured by evapotranspiration) was found to be a good correlate of the high functional redundancy characterising tropical communities (Safi et al. 2011), suggesting that increasing fluctuations in environmental conditions would have a much larger impact on tropical ecosystems (Dillon, Wang & Huey 2010).

Another major leap forward provided by this method is the ability to distinguish between the contributions of deep versus recent nodes in phylogenetic diversity in different communities. The concept of phylogenetic diversity is partially based on the idea that, rather than species richness, attention should be paid to protect processes that generate biodiversity (Crozier 1997; Mace, Gittleman & Purvis 2003). One suggestion for a global prioritisation and preservation of phylogenetic diversity is to prioritise species according to their contribution to evolutionary history; where more evolutionarily distinct species receive more attention (Isaac et al. 2007). The rationale behind conservation of phylogenetic distinct species is that these species have deep evolutionary roots and thus a long representation in communities, highlighting their ecological importance. However, it is unclear as to what role evolutionary distinct species really play in ecosystems and their functioning. With this new method it will be possible to quantify the evolutionary contribution of species to the phylogenetic diversity at the community level by quantifying the importance of deep nodes in the phylogenetic tree. By being spatially explicit, it will be possible to investigate when and where evolutionary distinct species actually define communities and ecosystem functioning.

The current pattern of species diversity is, as Davies et al. (2008) pointed out, the result of ‘[…] a complex history of speciation, extinction, anagenesis, and dispersal […]’. The method proposed by Diniz Filho et al. will draw a finer grained picture of diversity, helping us to further identify the mechanisms and ultimately the evolutionary forces acting on communities and unveil the signature of evolution in species communities.


I would like to thank Chuck Fox for inviting me to write this commentary. Nathalie Pettorelli has kindly commented on earlier versions.