Improving approaches to the analysis of functional and taxonomic biotic homogenization: beyond mean specialization
* Abadie et al., Journal of Ecology, 99, 2011, 1134 claim that ‘landscape disturbance causes small-scale functional homogenization but limited taxonomic homogenization’. This statement does not seem to accurately summarize their results. * Abadie et al. provide no strong arguments in favour of a cause and effect relationship between landscape disturbance and functional homogenization because their approach is correlational. * Abadie et al. associate an index of mean community specialization with functional biotic homogenization (BH), and they in turn associate functional BH with ecosystem functioning. However, the community specialization index is associated with a very specific kind of ‘function’ – the species response – which has no clear link with ecosystem functioning. This problem is frequent in the literature on biotic homogenization. * There is no clear sign in the data shown by Abadie et al. that metrics incorporating species attributes are ‘much more reliable’ than taxonomic diversity indices. * As frequently observed in the literature on biotic homogenization within communities, their results show no sign of loser or winner species or of ‘extirpation of specialist species’. Therefore, there seems to be no evidence in support of the biotic homogenization model they propose. * Synthesis. The Average Community Specialization – a mean specialization index– is in itself incapable of identifying loser and winner species and has no clear link with ecosystem functioning. Methods other than mean trait approaches should be used to study either functional homogenization or the extirpation of specialist species.