Reciprocal bark exchange helps to disentangle tree species dependent bark and wood trait effects on invertebrate diversity

1. Previous studies showed that bark cover at early-decay stage had profound control on the invertebrate assemblages of bark and wood, with possible consequence for the decomposition process. However, previous experimental designs could not disentangle how bark versus wood traits affect the invertebrate assemblage process in bark and/or wood separately because wood traits of different tree species may vary independently from bark traits. Furthermore, we do not know whether such tree species specific bark trait effects are still influential at mid-decay stage. 2. To unravel whether and how bark and wood traits influence invertebrate communities in tree logs at mid-decay stage, we introduce reciprocal bark transplantation within pairs of different tree species as a new method. We applied this method to two pairs of phylogenetically contrasting species of gymnosperms (pair I: Araucaria araucana and Cryptomeria japonica, pair II: Picea abies and Thuja plicata) and another gymnosperm (Chamaecyparis lawsoniana) set as disturbance control to test for potential bark manipulation artefacts on invertebrate community composition. 3. Our bark exchange experiment revealed that both bark and wood host abundant and divergent subsets of invertebrates on mid-decay logs of different tree species. We further documented that the invertebrate community composition was predominantly shaped by the traits of host tissue per se, while also being significantly but less strongly affected by the traits of the other tissue, i.e. the adjacent bark or wood. Our results indicated that bark trait effects faded with time and how long bark trait effects persist greatly depends on bark thickness. 4. Synthesis. Our study suggests that maintaining deadwood heterogeneity related to variation between tree species, and to bark versus wood, is important for nursing a large biodiversity of invertebrates. Combined with bark removal methodology, our bark exchange method can be further extended to more decay stages and more forest biomes to track bark trait effects and bark induced priority effects on deadwood decomposition, and its associated invertebrate and microbial communities.