Echinochloa pyramidalis (Lam.) Hitchc. & Chase is an African grass with C4 photosynthesis, high biomass production, and high vegetative propagation that is tolerant to grazing and able to grow in flooded and dry conditions. Thus, it is highly invasive in tropical freshwater marshes where it is intentionally planted by ranchers to increase cattle production. This invasion is reducing plant biodiversity by increasing the invader's aerial coverage, changing wetland hydrology and causing soil physicochemical changes such as vertical accretion. Reducing the dominance of this species and increasing the density of native wetland species is a difficult, expensive, and time-consuming process. We applied a series of disturbance treatments aimed at eliminating E. pyramidalis and recovering the native vegetation of a partially invaded freshwater marsh. Treatments included physical (cutting, soil disking, transplanting individuals of the key native species Sagittaria lancifolia subsp. media (Micheli) Bogin, and/or reducing light with shade mesh) and/or chemical (spraying Round-Up™ herbicide) disturbances. At the end of the experiment, four of the five treatments used were effective in increasing the cover and biomass of native species and reducing that of E. pyramidalis. The combination of these treatments should be used to generate a proposal for the restoration of tropical wetlands invaded by non-native grasses. A promising treatment is using soil disked to soften the soil and destroy belowground structures such as roots and rhizomes. This treatment would be more promising if combined with the use of shade cloth. If it is desirable not to impact the soil or if there is not enough budget to make an effort to include active restoration disking soil, the use of shade cloth will suffice, although the recovery of native vegetation will be slower.

No abstract