The impacts of invasive alien species are determined by their abundance, a relationship that usually does not follow linear trends owing to the complexity of ecological interactions. Abundance-impact curves are essential for informing evidence-based management interventions because they can reveal how per-capita impact changes as the invader become more abundant. Across 12 invasion gradients occurring in two areas, we constructed abundance-impact curves for the invasive grass Urochloa decumbens in a tropical savanna (Cerrado). We used generalized additive models to assess how increases in the invader’s abundance influenced system properties from the microhabitat to ecosystem levels. At the microhabitat level, increasing invader abundance resulted in nonlinear effects on bare soil and illuminance but a linear reduction in temperature fluctuations. The specific leaf area of dominant plants linearly increased with invader abundance. We found higher per-capita effects of Urochloa decumbens on the native graminoid cover when the invader was at low levels of abundance. Conversely, the per-capita effects on native species richness were higher at moderate levels of invasion. These results indicate the immediate impacts of the invader on the abundance of functionally similar native grasses but greater impacts on species richness only at moderate levels of invasion. The total biomass increased through the invasion gradient. Despite these changes, the abundance of invasive species did not influence the ecosystem properties. Our findings support the functional redundancy between Urochloa decumbens and native dominant grasses; however, despite this similarity, Urochloa decumbens promotes negative impacts at the microhabitat, organism, and community levels.
