The five islands comprising Channel Islands National Park (CHIS) experience natural gradients in temperature and moisture driven by ocean currents. Additionally, the islands were used as ranchlands and military land before becoming a national park, resulting in widespread erosion and vegetation change. As a result, CHIS spans gradients in climate as well as ranching duration and time since animal removal. Vegetation monitoring was initiated in 1984 on three islands (Anacapa, Santa Barbara, San Miguel), in 1990 on Santa Rosa Island, and in 1998 on Santa Cruz Island, with the goal of documenting the long-term response of island vegetation to ranch animal removal and climate fluctuations. Since that time, monitoring has documented the range of natural fluctuation in island environments over decades and provided insights into vegetation change in ecosystems unencumbered by ongoing development. Long-term vegetation monitoring at CHIS is therefore a rare example of an ecosystem experiment that demonstrates the results of management actions and serves as a baseline for land managers and scientists worldwide. 

Terrestrial vegetation data collected between 1984 and 2018 were modeled to estimate trends over time and to characterize relationships with covariates related to site characteristics, nonnative mammal removal programs, and water balance metrics. Data were analyzed for trends in vegetation cover, woody plant density, and plant community diversity grouped by life form and nativity across all islands and within individual islands, as well as for several individual species that dominate plant communities or present challenges to native plant recovery. In all, a total of 162 trend and covariate models were tested in this study, the details of which are provided in this report. Briefly, results reflect a decline in nonnative annual disturbance-thriving species with the reduction in animal grazing and trampling. Increasing trends were observed in native shrub density and native shrub recruitment density, as well as native shrub cover across all islands averaged together and on Santa Cruz Island. However, opposite trends were seen on the smaller islands of Santa Barbara and Anacapa, where increasing seabird activity may be damaging vegetation. Further results indicate the importance of soil moisture, relative humidity, fog, precipitation, site exposure, and solar radiation for vegetation patterns and trends. In many instances, there are apparent interacting effects of environmental variables with trends related to nonnative mammal removal and site location. Vegetation patterns in space and time emerge in the dataset as nuanced responses to interacting drivers.