Multiple spatial scales affect direct and indirect interactions between a non-native and a native species

Plant–plant interactions influence community assembly and species responses to environmental change. However, species interactions are complex phenomena influenced by context and scale. We conducted a one-year replacement experiment between two grasses in subtropical grasslands (native Axonopus fissifolius and non-native Paspalum notatum) in central Florida, USA. We evaluated interactions between these species at three ecological scales, ‘pairwise interactions’, ‘patch’ (33 cm x 33 cm) and ‘plot’ (1m x 3m), and along a gradient (15 levels) of increasing non-native and decreasing native plot groundcover within enclosures in semi-native pastures. We transplanted 18 individuals of each species per plot (18 × 30 plots = 1080 plants in total) in a 2×2 design intersecting direct pairwise interactions (additional transplanted neighbor: absent/present) with recipient patch type (Axonopus fissifolius/Paspalum notatum). Leaf length, leaf number and plant biomass were measured at the beginning and the end of the experiment along with soil nutrients and pH at patch level. Over 92% of the transplants survived. We observed an interactive effect between patch type (non-native vs. native dominated) and plant abundance at plot level on plant performance, suggesting that indirect effects at larger spatial scales can influence effects at smaller scales. Surprisingly, both species exhibited enhanced performance with increasing abundance of the non-native species at plot level. We discuss several mechanisms explaining these indirect effects including Paspalum notatum induced changes in soil pH, soil feedbacks from the soil microbial community, preconditioning effects of the native species on the recipient soil, and positive density dependence effects after transplantation.