| Summary: | Abstract In studies of community assembly, species are often assumed to have similar spatial distributions and responses to the environment regardless of age or size. Under this assumption, it is possible to use species and species‐level traits in community composition studies. Here, we test this assumption for two species of soil‐living arthropods (springtails: Collembola) with direct development but assumed differences in self‐organizing behavior. We expected that the species with more pronounced social interactions (Hypogastrura tullbergi) should be less influenced by environmental factors and species interactions across all age classes, than Folsomia quadrioculata that is not known to exhibit social behavior. We used variance partitioning to examine the relative contributions of soil variables, vegetation composition, and other Collembola, vs. spatial variables (as a proxy for intraspecific interactions, i.e., self‐organization), on the distribution of the two species and three of their age classes. We show that two coexisting species with clear aggregation patterns greatly differ in how much the environment contributes to affecting the species’ spatial structure. Local F. quadrioculata abundance was explained by different spatial and environmental variables depending on age class. In contrast, for H. tullbergi, spatial variables explained more of the abundance variation in all age classes. These differences have implications for the general predictability of changes in spatial structuring of species, as self‐organized species may be less likely to respond to changes in environmental factors. Our results show that because age classes may be differentially affected by environmental conditions, caution should be taken when assuming that species traits can be applied to all developmental stages in a species.
|
|---|
