The cutting edge: How decay trajectories and biodiversity vary between bark and wood within and across tree species

Deadwood is an important carbon stock and nutrient pool both in terrestrial and aquatic ecosystems, while it is also vital to help nurse the world’s biodiversity. Disentangling the factors dominating the deadwood decay process and how they interact with each other at different decay stages can help us to better predict how the ecosystem carbon stock and biodiversity response to the global climate change and to make preparations for the potential challenges. Deadwood decomposition is a long and complex process affected by diverse biotic and abiotic drivers; the dominant factors may change with, for instance, habitat, tree species, decay stages, and climate zones. In my thesis, I mainly focused on the decay dynamics of different tissue types within logs both within and across tree species, and how these variables interact with environmental factors in controlling microbial and invertebrate community compositions at different decay stages. To disentangle the interactive contributions of these factors on decay rates and biodiversity associated with tree logs, we have conducted two sets of long term in situ decomposition studies in an alpine forest and two environmentally contrasting temperate forests. In the alpine forest, the deadwood of Abies faxoniana, the dominant tree species in this region, was selected to answer (1) How do the decay trajectories and decay rates of bark, sapwood, and heartwood vary with forest spatial heterogeneity (Q1)? (2) How do biotic and abiotic factors interact to affect microbial community composition in decaying bark, sapwood and heartwood (Q3)? I mainly answered Q1 in Chapter 2. The results showed forest canopy cover significantly stimulated decay rates and altered the best-fit decay models of bark and sapwood. In contrast, heartwood decayed faster at the forest edge than in the other two positions within forest and followed a single exponential as the only best-fit decay model across forest locations. The results also showed that bark generally decayed faster than sapwood and heartwood at least at early and mid decay stages. In Chapter 4, we studied the response of microbial biomass and community composition to the presence versus absence of epixylic vegetation on decaying deadwood. The results showed the influence of experimental epixylic vegetation removal on microbial community composition significantly interacted with structural components of the logs and sampling seasons, while it was less pronouncedly affected by decay classes and the relative degree of deadwood contact with the soil (Q3). In the temperate forests, 9 gymnosperms and 16 angiosperms were incubated in the field aiming to answer the following questions: (3) Are bark and wood mass loss coordinated across tree species (Q2)? and (4) Do bark traits still filter invertebrate communities at mid decay stage and to which extent do bark and wood quality each contribute to their assemblage (Q4)? I found a generally doubled decay rate of bark compared with coupled underlying wood across tree species, and faster decaying bark came with faster decaying wood, albeit with much scatter around the mean pattern in the study of Chapter 3. I answered Q4 in Chapter 5 by applying a novel bark reciprocal exchange method. By comparing the invertebrates’ community composition between five gymnosperms, I found bark traits, mainly bark pH, thickness and tannin content, and wood traits, such as density, pH and water holding capacity, significantly shaping the tree species specific invertebrate assemblages in decaying bark and wood, respectively. I also found a clear difference of invertebrates in the underlying wood between different tree species, the variation of which was larger than that between decaying bark of different tree species.