Forest Floor Bacterial Community Composition and Catabolic Profiles in Relation to Landscape Features in Québec’s Southern Boreal Forest

Abstract

Bacterial communities mediate many of the processes in boreal forest floors that determine the functioning of these ecosystems, yet it remains uncertain whether the composition of these communities is distributed nonrandomly across the landscape. In a study performed in the southern boreal mixed wood forest of Québec, Canada, we tested the hypothesis that stand type (spruce/fir, aspen, paper birch), stand age (57, 78-85, and 131 years old), and geologic parent material (clay and till) were correlated with forest floor bacterial community composition. Forest floors in 54 independent forest stands were sampled to comprise a full factorial array of the three predictor variables. Bacterial community structure was examined by terminal restriction fragment (T-RF) length polymorphism analysis of genes encoding for 16S rRNA. Distance-based redundancy analysis of T-RF assemblages revealed that each predictor variable, as well as their interaction terms, had a significant effect on bacterial community composition, geologic parent material being the most discriminating factor. A survey of the 15 T-RFs with the highest percentage fit on the first two ordination axes describing the main effects indicated that each landscape feature correlated to a distinct group of bacteria. A survey of the most discriminant T-RFs describing the effect of stand type within each combination of stand age and geologic parent material indicated a strong dependency of several T-RFs on geologic parent material. Given the possible link between bacterial community composition and forest floor functioning, we also assessed the effects of the same three landscape features on community-level catabolic profiles (CLCP) of the extractable forest floor microbiota. Geologic parent material and stand type had significant effects on CLCPs. On clay plots, the effects of landscape features on T-RF patterns were highly consistent with their effects on CLCPs. In light of our results, we suggest that future research examine whether bacterial community composition or CLCPs can be used to detect latent environmental changes across landscape units.