1. Whether wild herbivores confer biotic resistance to invasion by exotic plants remains a key question in ecology. There is evidence that wild herbivores can impede invasion by exotic plants, but it is unclear whether and how this generalises across ecosystems with varying wild herbivore diversity and functional groups of plants, particularly over long-term (decadal) time frames.

2. Using data from three long-term (13- to 26-year) exclosure experiments in central Kenya, we tested the effects of wild herbivores on the density of exotic invasive cacti, Opuntia stricta and O. ficus-indica (collectively, Opuntia), which are among the worst invasive species globally. We also examined relationships between wild herbivore richness and elephant occurrence probability with the probability of O. stricta presence at the landscape level (6,150 km²).

3. Opuntia densities were 74% to 99% lower in almost all plots accessible to wild herbivores and increased more rapidly in plots excluding wild herbivores. These effects were largely driven by megaherbivores (≥1,000 kg), particularly elephants.

4. At the landscape level, modelled Opuntia occurrence probability was negatively correlated with estimated species richness of wild herbivores and elephant occurrence probability. On average, O. stricta occurrence probability fell from ~0.56 to ~0.45 as wild herbivore richness increased from 6 to 10 species and fell from ~0.57 to ~0.40 as elephant occurrence probability increased from ~0.41 to ~0.84. These multi-scale results suggest that any facilitative effects of Opuntia by wild herbivores (e.g., seed/vegetative dispersal) are overridden by suppression (e.g., consumption, uprooting, trampling).

5. Synthesis. Our experimental and observational findings that wild herbivores confer resistance to invasion by exotic cacti add to evidence that conserving and restoring native herbivore assemblages (particularly megaherbivores) can increase community resistance to plant invasions.