How foreign is the past?

Abstract

Since Humboldt and Darwin, ecologists have puzzled over what determines community assembly and structure and how community structure may change with time. Human activity is one potential driver. Impacts of modern human societies on the environment and its biota are massive, with many forms of pollution, loss and fragmentation of habitats, and extensive introductions of exotic species changing many ecological and biogeographical patterns. Prehistoric societies might be expected to have had a much lower impact on their environment. However, Lyons and colleagues1 propose that plant and mammal assemblages are so fragile that the limited settlements, agriculture, and associated activities 6000 years ago in North America were sufficient to fundamentally change community-assembly rules. They base this conclusion on temporal patterns in the proportion of aggregated taxon-pairs in species presence-absence data-sets from the past 300 million years. We demonstrate that their conclusions result from the use of inappropriate modern data-sets and from biases when detecting segregated taxon-pairs in different sized data-sets. Lyons et al. analysed co-occurrence patterns between taxon pairs in 71 fossil (age 307 Myr – 100 yr) and 9 recent (<100 yr) presence-absence assemblages using an empirical Bayes approach2 that classifies each taxon-pair as random, aggregated, or segregated. Lyons et al. supplemented their analyses by including 39 modern ‘mainland’ data-sets, and 59 ‘island’ data-sets in some analyses. 2