Synergistic impacts of co‐occurring invasive grasses cause persistent effects in the soil‐plant system after selective removal

Human influence on the environment is so extensive that virtually all ecosystems on the planet are now affected by biological invasions. And, often, ecosystems are invaded by multiple co‐occurring non‐native species. Hence, it is important to understand the impacts these invasions are producing on biodiversity and ecosystem processes. Here, we present results of a 2‐year long field experiment where we tested the effects of co‐occurring invasive C₄ African grasses in a Cerrado area in central Brazil. We compared plant and arthropod communities, plant biomass, and soil nitrogen dynamics and soil chemical characteristics across five experimental treatments: Urochloa decumbens removal; Melinis minutiflora removal; both U. decumbens and M. minutiflora removal; U. decumbens and M. minutiflora invaded plots; and uninvaded Cerrado. We hypothesized that selective removal of invasive grasses would have distinct effects on the native ecosystem structure and functioning. We expected that each invasive grass would produce a different type of impact on the native ecosystem and that their impacts would be synergistic when co‐occurring. Removal of M. minutiflora doubled native plant diversity and biomass when compared to invaded plots, whereas removal of U. decumbens did not alter these parameters. Cerrado plots had four times more plant species than plots cleared of invasives. Removal of invasive grasses did not affect the species richness or community composition of soil epigeal fauna. Cerrado soils had lower fertility, organic matter content and pH than invaded soils. The effects were generally higher when both invasive grasses were removed, suggesting impacts were synergistic, but M. minutiflora had greater effects on plants and soils than U. decumbens. Both invasive species produced negative impacts, but a single species was the main driver. We also detected persistent effects of the invasive grass species on the ecosystem after 2 years of removal. We conclude that invasive species of the same functional group have similar types of effects in native ecosystems, but the magnitude of impact was largely dependent on invasive species biomass and cover. Where multiple invasive species are present, research and management of invaded ecosystems should tackle the interacting effects of co‐occurring invaders.