Woody plant richness does not influence invertebrate community reassembly trajectories in a tree diversity experiment.

Understanding the relationship between plant diversity and diversity at higher trophic levels is important from both conservation and restoration perspectives. Although there is strong evidence for bottom-up maintenance of biodiversity, this is based largely on studies of simplified grassland systems. Recently, studies in the TreeDivNet global network of tree diversity experiments have begun to test whether these findings are generalizable to more complex ecosystems, such as woodlands. We monitored invertebrate community reassembly over 5 yr of experimental woodland restoration at the TreeDivNet Ridgefield site in southwest Australia, testing the effects of woody plant species richness and herb-layer manipulation on invertebrate community structure and ant species composition. From 2010 to 2014, we sampled ground-dwelling invertebrates using pitfall traps in herbicide vs. no-herbicide subplots nested within each of 10 woody plant treatments varying in richness from zero (bare controls) to eight species, which produced a total of 211, 235 invertebrates, including 98, 979 ants belonging to 74 species. In mixed model analyses, the presence of woody plants was an important driver of faunal community reassembly (relative to bare control plots), but faunal responses to woody plant treatment combinations were idiosyncratic and unrelated to woody plant richness across treatments. We also found that a herbicide-induced reduction in herbaceous plant cover and richness had a positive effect on ant richness and caused more rapid convergence of invertebrate community composition toward the composition of a woodland reference site. These findings show that woody plant richness did not have direct positive effects on the diversity and community reassembly trajectories of higher trophic levels in our woodland system. From a management perspective, this suggests that even low-diversity restoration or carbon sequestration plantings can potentially lead to faunal reassembly outcomes that are comparable to more complex re-planting designs.