The Editor’s Choice paper for volume 106 issue 3 is a paper on niche occupancy and community assembly by Yuanzhi Li and colleagues. Journal of Ecology Associate Editor Peter Vesk has given his thoughts on the paper below.

In the latest issue of Journal of Ecology, Li et al. present an exciting new investigation of niche packing and community assembly. This work picks up a classical Hutchinsonian approach to modelling the niche as an n-dimensional hypervolume, but using traits rather than resources, extending previous work by Blonder et al. (2014). It then seeks to assess niche packing under increased species richness and, in particular, expectations of habitat filtering vs limiting similarity in plant community assembly.

The authors develop a clear, explicit formulation of the niche hypervolume and relations between overlap and total hypervolume. Importantly, this work utilises intraspecific variation, without which the functional hypervolume of individual species cannot be calculated. Paraphrasing their presentation, there is a mathematical identity whereby the average species hypervolume multiplied by the number of species in the community is equal to the total community hypervolume plus the amount of overlap. The insight being that to make sense of species richness effects, all three elements – total community hypervolume, average species’ hypervolume, and the overlap – are needed.

Using this framework, Li et al. examine simple expectations under basic processes of habitat filtering and limiting similarity. These expectations are tested with a compilation of 313 communities across 21 community studies worldwide. Importantly, these range across growth forms from grasses and herbs, through shrubs to trees.

In a neutral assembly process, where average hypervolume will be unrelated to SR, as more species are randomly added to a community, the overlap should increase and the total hypervolume should also increase. Under limiting similarity, the amount of overlap should increase more slowly with species richness than expected under neutrality. And the total hypervolume should increase more rapidly than under neutrality. Conversely, if habitat filtering dominates, then as species are added, the total hypervolume should increase, but more slowly than under neutrality and with more overlap as they are being squeezed into some constrained functional space.

Read the full paper online: https://doi.org/10.1111/1365-2745.12802

What they demonstrate is that habitat filtering is pervasive in shaping communities. Now, this is in itself might not be particularly novel, but the global coverage of the datasets analysed is useful in demonstrating the prevalence of habitat filtering. Expectations under limiting similarity were not supported.

This work opens up avenues for theoretical and more standardized empirical study. Spatial scale matters to processes of competition and habitat filtering, with competition (and individual packing) expected more in neighbourhoods (de Bello et al. 2013). Extending from 2 to 3 trait hypervolumes would be useful, though rendering estimation of the hypervolume more difficult. And of course, the question of trait choice, some traits likely relate to competition more than to habitat filtering.

Peter Vesk, University of Melbourne, Australia, and Journal of Ecology Associate Editor

Read the full paper online: Habitat filtering determines the functional niche occupancy of plant communities worldwide