Decoupling of plant and microbial communities following long‐term removal of an invasive grass

Ecosystem function can be shaped by the relationship between plant and soil microbial communities. When invasive plants disrupt these plant–soil feedbacks, they can leave soil legacy effects that persist even after removal. However, such legacy effects vary among species, and links between aboveground and belowground recovery after invasive plant removal remain unclear. We examined the potential legacy effects of the invasive yellow bluestem (Bothriochloa ischaemum var. songarica) in a grassland in central Texas, USA. In a 9‐year removal experiment, we measured the diversity and composition of plant and soil microbial communities in invaded plots (dominated by the invasive grass), removal plots (where the invasive was continually removed), and native‐dominated plots. With only passive recovery (i.e. no active seeding or planting), plant richness, native grass cover, and native forb cover in removal plots became similar to native‐dominated plots in 1–3 years, although species composition was still different 8 years after removal. In contrast, soil microbial diversity and composition did not differ among the treatments, including no differences between invaded and native‐dominated plots. Plant species composition may be influenced by dispersal limitations because removal plots are still surrounded by the invasive grass. The observed decoupling between aboveground and belowground responses to invasion and removal efforts could be due to the soil microbial community's resistance to influence by the invasive grass. These results suggest that microbial soil legacy effects will likely not limit grassland restoration and that active seeding or planting may be required for a more complete restoration of native plant communities after yellow bluestem removal.