Assembly processes in soil animal communities: Integrating phylogeny and trait-based approaches

Soils are among the most biodiverse systems on earth. The coexistence in soils of a multitude of animal species has long puzzled soil ecologists. How can so many species co-occur, and what are the processes driving and maintaining species coexistence in soil? Using a deductive approach, I propose that (1) there are assembly processes, (2) that work on, or are related to, certain objects, i.e., functional traits, to (3) produce particular patterns. I use a conceptual model combining patterns of evolution of species traits, trait similarity and phylogenetic relatedness between coexisting species, from which to infer assembly processes in soil Collembola (springtail) communities collected from habitats characterized by different disturbance regimes. In Chapter 2, I reconstruct a Collembola phylogeny and use phylogenetic comparative methods to explore phylogenetic signal, model of evolution and ancestral state for a variety of traits, including body shape, body length, pigmentation, number of ommatidia, vertical stratification and reproductive mode. The results demonstrate that body shape of Collembola evolved quickly early in their diversification but slowed down afterwards. In contrast, evolutionary transitions in pigmentation, number of ommatidia and reproductive mode depended on how deep in the soil that species live. Ancestral Collembola traits were likely slender body, hemiedaphic way of life, sexual reproduction, possession of many ommatidia and bright color, but these traits presumably changed several times during species diversification. The phylogenetic signal detected in these traits forms the basis of further community phylogenetic analyses. In Chapter 3, I propose the neutral lipid fatty acid composition of Collembola as a functional trait related to both food resources and physiological functions and test phylogenetic signal in fatty acid profiles. Long-chain polyunsaturated fatty acids related to physiological functions demonstrated phylogenetic signal. In contrast, most food resource biomarker fatty acids and the ratios between bacterial, fungal and plant biomarker fatty acids exhibited no phylogenetic signal. These results suggest that Collembola with close phylogenetic affinity experienced similar environments during divergence, while niche partitioning in food resources among closely related species favored species coexistence. In Chapter 4, I use both community phylogenetic and trait-based approaches to infer the assembly processes of Collembola communities inhabiting arable fields, grasslands and forests. The results indicate that Collembola communities in arable fields were mainly structured by environmental filtering, while niche partitioning dominated in forests. Epedaphic (surface-living) species showed phylogenetic clustering in grasslands and forests, while in forests they also possessed similar traits. Hemiedaphic (sub-surface-dwelling) species were phylogenetically clustered in arable fields and grasslands, but in forests they were phylogenetically overdispersed and had different traits. However, the assembly of euedaphic (soil-dwelling) communities did not differ from random patterns. Furthermore, different phylogenetic groups of Collembola showed different patterns in the three habitats. These results suggest that Collembola assemblages are driven by different mechanisms in different habitats, with the relative importance of these mechanisms different between soil strata and between phylogenetic lineages. Furthermore, applying community phylogenetic approaches to a manipulative soil block experiment (Auclerc et al. 2009; Soil Biology and Biochemistry 41, 1596–1604) in Chapter 5 shows that Collembola community composition during their succession in forest and meadow soil was determined by the interaction of dispersal and selection processes. Niche partitioning gradually strengthened at later successional stages, offsetting the effects of environmental filtering. As a consequence of dispersal, community composition changed gradually from that resembling the original habitats to that of the new habitats. In the final chapter I ascribe the above-mentioned patterns to the scenarios presented in the conceptual model and discuss the likely mechanisms, with reference to the four high-level processes, selection, dispersal, drift and speciation, proposed in The Theory of Ecological Communities (Vellend 2016). I provide a roadmap for integrating phylogenetic comparative methods, community phylogenetic analyses and trait-based approaches in studies on the assembly processes of soil communities. Overall, this thesis is the first application of new methods developed in community ecology and evolutionary biology to the study on assembly processes in the soil communities. Future studies using the conceptual model and roadmap proposed in this thesis will advance our understanding of the mechanisms driving and maintaining soil biodiversity from both ecological and evolutionary perspectives