Niche differences, not fitness differences, explain coexistence across ecological groups

Abstract

Understanding how species interactions affect community composition is an important objective in ecology. Yet, the multitude of methods to study coexistence has hampered cross-community comparisons. Here, we standardized niche and fitness differences across 1018 species pairs to compare the processes driving composition and outcomes, among four community types (annual plant, perennial plant, phytoplankton, and bacteria/yeast). First, we show that niche differences are more important drivers of coexistence than fitness differences. Second, in all community types negative frequency dependence is the most frequent process. Finally, the outcome of species interactions differs among community types. Coexistence was the most frequent outcome for perennial plants and phytoplankton, while competitive exclusion was the most prevalent outcome in annual plants and bacteria/yeasts. Overall, our results show that niche and fitness differences can be used as a common currency that allow cross community comparisons to understand species coexistence.

Building modern coexistence theory from the ground up: the role of community assembly

“…Second, for a non-coexisting species j we apply the methods of modern coexistence theory to the invasion growth rate of species j into this coexisting sub-community. Importantly, our new approach does not change any analysis for a community where all n species coexist and all n – 1 communities exist, which might be the majority of the existing empirical data (Buche et al ., 2022), but it does extend the applicability of the methods of modern coexistence theory.…”

“…A better understanding of non-coexistence is direly needed, as often models fitted to empirical data predict competitive exclusion despite co-occurring species (Buche et al ., 2022; Germain et al ., 2016; Kraft et al ., 2015b; Godoy & Levine, 2014). This mismatch between empirical observation may stem from difficulties of assessing the actual underlying species interactions (Adler et al ., 2018a) or from uncertainties (Bowler et al ., 2022).…”

“…But what are the “right” invasion growth rates and in what sense do they being positive ensure coexistence? The answer is clear for two species competitive com-munities, the bedrock of MCT (Buche et al ., 2022; Letten et al ., 2017). Each species should have a positive invasion growth rate when the other species is at equilibrium, i.e the mutual invasability criterion.…”

“…This is especially unfortunate, as modern coexistence theory explores more and more species rich communities (Spaak et al ., 2021a; Shoemaker et al ., 2019; Godoy et al ., 2018; Saavedra et al ., 2017; Petry et al ., 2018; Chesson, 2018; Spaak & De Laender, 2021). Knowledge about these two issues are wide-spread in modern coexistence theory (Spaak et al ., 2021b; Spaak & De Laender, 2020; Barabas et al ., 2018; Saavedra et al ., 2017; Pande et al ., 2020), yet the invasion growth rates remain widely used as if these problems did not exist (Spaak et al ., 2021a; Grainger et al ., 2019a; Letten et al ., 2018; Shoemaker et al ., 2019; Schreiber et al ., 2018; Buche et al ., 2022).…”

Modeling phytoplankton-zooplankton interactions: opportunities for species richness and challenges for modern coexistence theory

“…Conversely, niche differences were only weakly correlated with phytoplankton or zooplankton richness. This differs from earlier findings in phytoplankton or plant communities, which mostly found that niche differences were better predictors of species richness (Narwani et al ., 2017; Buche et al ., 2021; Levine & HilleRisLambers, 2009; Adler et al ., 2010). This difference may stem from many different sources, such as model complexity, the two trophic levels in our model, or our inability to compute niche and fitness differences for all communities (see next section).…”