Jennifer M. Jones, Katy D. Heath, Astrid Ferrer, James W. Dalling

Decomposition of woody material is a critical ecosystem function, releasing nutrients and carbon into the environment where they can be reused by other organisms. Woody material, however, is made up of two distinct tissues: wood, which has low nutrient concentrations, and bark, which functions in protection and sugar transport and has much higher nutrient concentrations. The extra nitrogen present in bark may play an important role in decomposition processes, particularly for the production of lignocellulolytic enzymes by wood decay fungi that break down woody tissue. We therefore hypothesized that the presence of bark would speed the decomposition of underlying wood, and that this effect would be greater in streams, where nutrients are less available to decomposers than on land. We placed branches of three tree species, with and without bark, in streams and on land in a lowland tropical forest in Panama.

After 3 and 11 months of decomposition, we measured mass loss and nitrogen (N) concentration and sequenced the fungal and bacterial communities of both wood and bark tissues. We found that bark decomposed faster than the underlying wood and that bark had higher nitrogen concentration. However, contrary to our expectation, when bark was present, wood decomposed more slowly than when bark was removed. We also found that bark mass loss, but not wood mass loss, was faster in streams than on land, suggesting fragmentation, or the physical breaking apart of woody tissue, is more important for bark mass loss in streams. Finally, differences in fungal and bacterial community composition between bark and wood tissues were significant but small. Our results show that the presence of bark slows wood decomposition while the primary drivers of decomposition rate differs between these tissues. Differences in the factors driving decomposition rate between bark and wood suggest that the contribution of bark to the decomposition of woody biomass may depend on habitat.

Read the paper in full here.