COS 48-2 - Disparity in phylogenetic and trait similarity patterns suggests differences in assembly mechanisms for dominant and non-dominant plant species across biomes

Background/Question/Methods

Community assembly mechanisms are often assumed to affect all community members in a similar way. If a plant community is strongly affected by environmental filters, all coexisting species should share more similar traits or phylogenetic history than expected by chance (under-dispersion), while strong interspecific interactions (e.g., competition) should lead to greater differentiation among species (over-dispersion). Yet, dominant and non-dominant plants differ profoundly in a given community, not only in their biomass, but also in functional traits and phylogenetic relatedness. Further, recent studies in grassland communities found phylogenetic under-dispersion among dominant plants, but not among non-dominants, suggesting that community assembly mechanisms may differ between dominant and non-dominant plants.

Using sPlot, a global database of community plots with plant composition, cover, trait, and phylogenetic information, we explored the generality of the disparity in assembly mechanisms between the dominant and non-dominant plants across biomes. For each of the 238,248 selected plots, we split the community in three dominance groups with the same number of species based on their cover, and estimated the phylogenetic relatedness and trait distance among species of each group separately. We also tested if some clades were more likely to be either dominant or non-dominants across biomes.

Results/Conclusions

Of the 21 tested combinations of biome and biogeographic realm, in 13 the coexisting dominant plants were more under-dispersed than the non-dominants. The other 8 combinations included 2631 plots, with most of them having a similar, but non-significant trend. Coexisting dominants tended to be under-dispersed in terms of their phylogenetic relatedness (19 out 21), while non-dominants had more heterogeneous patterns of relatedness (3 over-dispersed, 8 under-dispersed, 10 non-significant). The trait analysis suggests that most traits had heterogeneous responses, but dominant species tended to be more under-dispersed than non-dominants (similarly to the trend observed for their phylogenetic relatedness) in terms of the phosphorus concentration in their leaves, their specific leaf area, and the number of seeds they produced, while plant height, seed length and leaf dry matter content returned an opposite result (dominants more over-dispersed than non-dominants). Our work reveals a disparity in phylogenetic and trait patterns between coexisting dominant plants and the non-dominant ones on a global scale, suggesting that assembly mechanisms among dominants are different from those among non-dominants. We present potential environmental and phylogenetic drivers of this disparity, as well as the potential existence of trait syndromes associated with dominance and non-dominance in plant communities.