At our 1575 ha Audubon preserve we use experimental and observational (monitoring) data to guide land management decisions. Rare habitat restoration commences the second year of non-chemical control of the exotic herbaceous perennial, Cynara cardunculus, which has invaded 283 ha of degraded and native grassland habitat. Native grasslands that dominate 181 ha are monitored then maintained or enhanced. Long-term monitoring data indicated a decline in dominant native bunchgrass density and cover after a record-breaking drought season, which instigated initiation of experiments for re-invigoration. An experiment on timing and frequency of mowing indicated that a second mid-season treatment could have negative effects, so we commenced an experiment to test effects of a single early mow. In anticipation of mowing an entire grassland stand, we collect long term data on bunchgrass density and cover in six stands, with treatment pending positive effects in our experiments. Because analyses of aerial photoseries taken over 48 years revealed gradual coastal sage scrub colonization of some grassland stands, 102 ha of C. cardunculus-infested sites in which shrub species have begun colonization will be restored to coastal sage scrub. Results from a factorial experiment established an optimum shrub seeding rate and soil tamping technique. Restoration practices were further refined with a second experiment that showed that a combination of direct seeding and plug planting spaced over the growing season would be beneficial in our semiarid region. Monitoring data from restoration sites led us to switch to a combination of active and passive approaches.