Background/Question/Methods The soil seed bank plays a dynamic role in the regeneration of plant communities after both natural and anthropogenic disturbances. Understanding how different types and intensities of disturbance alter the composition and spatial-temporal structure of the soil seed bank and corresponding vegetation may lend insight into what steps must be taken to restore anthropogenically disturbed habitats. We are conducting a study investigating the spatial-temporal relationship between standing vegetation and the soil seed bank among three vegetation types: native rosemary scrub, disturbed scrub and semi-improved pasture. We hypothesize that the composition and spatial distribution of the seed bank and aboveground vegetation will reflect contrasting abilities of species to cope with human and natural disturbances. The spatial-temporal relationship of vegetation and the seed bank was examined in nine 16 x 16 m permanent grid plots during the winter and summer seasons. Within each grid plot soil samples were collected from 130 randomly selected 40 x 40 cm quadrats and percent vegetation cover was assessed. The seedling emergence method was used to determine species composition of the soil seed bank. Moran's I spatial autocorrelation and the Mantel test will be used to evaluate the spatial-temporal relationship between the seed bank and aboveground vegetation. Results/Conclusions We identified 53 species aboveground and 45 in the seed bank across all three habitats. Aboveground, the disturbed scrub was the most species-rich and the semi-improved pastures the least. Belowground, the pastures were the most species-rich and the native rosemary scrub the least. Aboveground vegetation was dominated by long-lived perennials while the seed bank was dominated by few annual or short-lived perennial species. The greatest disparity between above and belowground species composition was observed in the pastures where the majority of species found in the seed bank were either absent or rare aboveground. Species pools between the disturbed and rosemary scrub were similar; however, percent cover of shrubs and some scrub herbs was reduced in the disturbed scrub. Occurrence of scrub herbs belowground was reduced in the disturbed scrub and extremely rare or absent from the pastures. Preliminary spatial structure analysis revealed that some species and functional groups with contrasting aboveground densities among the sites exhibited different patterns of spatial autocorrelation. Restoration of scrub populations in the pastures will require reintroduction of either seeds or transplants of desired species. Conversely, restoration of disturbed scrub sites will require the reinstatement of native spatial structure, distribution, and abundance of scrub species.