Lichen communities are formed by processes that act on the biology of at least two organisms: a fungus its photosynthesizing partner. Here, I develop a general framework that predicts how dispersal limitation, environmental filtering, competition, and disturbance should affect variation in functional strategy along environmental gradients for communities of obligate mutualists, using lichen epiphytes as a model system. I then ask whether a set of easily observed morphological traits can be used to identify these assembly processes by testing hypotheses predicted by the framework. In addition, I evaluate the utility of trait-based community metrics for studying variation in lichen epiphyte communities at different spatial scales. Lichen morphological traits and substrate environmental characteristics were assessed in 9x9 cm quadrats on 720 tree trunks in deciduous forest plots spanning a 400 km elevational gradient in the southeastern United States. Hypothesized trait-environment relationships were evaluated using generalized linear models of community-mean trait values and null models of trait dispersion. Linear mixed models were used to partition trait variance among spatial scales (tree, plot, site, ecoregion) and assess the extent to which spatial variation in community trait structure could be attributed to environmental covariates.
Results/Conclusions
Trait-environment relationships were consistent with weak effects of water-limitation, competitive constraints and disturbance due to substrate instability. Lichens on drier substrates with fewer bryophytes were more closely appressed to the bark, an effect which was stronger in lower elevation sites with drier climates, while lichens producing vegetative propagules occurred more frequently on loose bark. Null models found no evidence of reduced trait dispersion arising from hypothesized constraints imposed by assembly processes. A lack of evidence for reduced trait dispersion combined with overall small effects sizes and low levels of explained variance in linear models suggest either a strong role for stochasticity in the formation of lichen epiphyte communities in these forests or that simple morphological traits have limited utility for detecting processes structuring lichen communities in this habitat. Furthermore, morphological traits and functional composition exhibited much less variability at larger scales than diversity and abundance, which limits their ability to detect larger-scale processes. Future trait-based studies of lichen epiphyte community assembly and spatial structure may find greater success with quantitative traits.