Restoration of abandoned rangeland to native plant communities may be compromised if the effects of exotic range grasses persist beyond their removal. We have started examining persistent effects of two non-native pasture grasses, bahia grass (Paspalum notatum) and natal grass (Rhynchelytrum repens), on the restoration of a central Florida scrub ecosystem. We removed grasses using herbicide and followed potential biogeochemical and microbial legacies. Specifically, we measured soil nitrogen, biological soil crusts, and soil fungi in intact and removed bahia and natal grass pastures compared to native scrub sites in three different seasons. Not surprisingly, bahia and natal pastures were significantly different from native scrub sites in all seasons, with more available nitrogen, fewer active algae, and more extensive fungal networks. Grass removal in pasture sites led to an increase in available nitrogen, likely a temporary pulse due to plant litter left on-site post-herbicide. Grass removal also affected soil microbial communities, with reductions in active algae and soil fungi, probably due to the direct effect of herbicide on algae and the decrease in roots for mycorrhizal colonization. In terms of restoration, initial seed germination of three native herbs was 3-13 times higher in native scrub compared to both intact and removed pasture sites. We have now attempted to manipulate the microbial community through the addition of native biological soil crusts to grass removal plots and will continue with seasonal measurements. Ultimately, both biogeochemical and biological legacies are likely to be responsible for poor restoration success.