What actually happens to functional diversity during a large extinction?

Background/Question/Methods

Because no two species are equal, there has been an increasing drive to go beyond simple species richness when studying community dynamics and use taxon free approaches like continuous indices of functional diversity.  Of much practical importance to conservationists is understanding how species loss alters functional diversity and although there are many simulations that examine this relationship, there are few studies looking at real extinctions.  By examining the 99 largest mammals in North America over the last 50,000 years, we can see how functional diversity changed throughout immigrations; introductions; and a massive, continental scale extinction.

I coded each species for 10 quantitative and pseudo quantitative functional traits covering mass; percentage diet; running, climbing, digging, and swimming abilities; and sociality.  Presence/absence bins of 1,000 years were constructed based on species’ first and last appearance dates.  Using new methods I developed that allow for pseudo quantitative traits, species were placed in a multidimensional functional space and on a functional dendrogram both constructed from trait weighted interspecies Euclidean distances.  For each time bin, I calculated species richness, dendrogram based functional diversity (FD), Functional Richness (FRic), Functional Dispersion (FDis), and Rao’s Quadratic Entropy (RaoQ).

Results/Conclusions

Though the Pleistocene extinction is clearly size driven, examining additional traits besides mass provides a much more nuanced picture of functional diversity change over the last 50,000 years.  Both FD and FRic dropped precipitously during the megafaunal extinction.  However, FRic, which is more susceptible to outliers like large mammoths, dropped far below values predicted by null models of species loss.  FDis and RaoQ also decreased meaning that current species are more closely packed into a smaller functional space than pre glacial faunas.  The first species to go extinct were actually not those with distinctive trait values at the edge of functional space but instead those with average values and close functional equivalents that made them redundant.  Surprisingly, many “weirdo” species with unusual trait combinations survive into the present, making extinct species no more functionally distinct on average than extant taxa.  Large mass, and poor climbing and digging ability were the best predictors of extinction.

This study represents the highest temporal, taxonomic, and functional resolution analysis of the terminal Pleistocene to date while providing conservation and theoretical ecologists with a much needed example of functional change during a large extinction caused by humans and climate change.