Southern California shrublands are threatened by nonnative European annual grasses, which contribute to substantial fragmentation and loss of these ecologically important communities. Both natural and anthropogenic disturbances provide opportunities for competitive exotic herbaceous species to colonize in gaps in shrublands and overtake native grasses and shrubs. Remote sensing using moderate resolution imagery can provide a spatially comprehensive, cost-effective, and repeatable approach to monitor the establishment and expansion of nonnative herbaceous cover into sensitive habitats, and provide the potential for detecting native herbaceous cover where it occurs in high fractions. The underlying goal of this study is to better understand the temporal, spectral, and spatial characteristics of herbaceous vegetation in California shrublands in the context of effectively mapping and monitoring herbaceous vegetation cover using remote sensing methods.

Chapter 2 provides the details of the phenological response and spectral properties of native and nonnative herbaceous cover, based on plot-based fieldwork conducted in CSS habitat in the San Diego National Wildlife Refuge (SDNWR). High temporal resolution field spectroscopic data were shown to be useful for determining when native and nonnative herbaceous vegetation in CSS are most spectrally separable in the course of a growing season. In Chapter 3, intraannual remote sensing methods are explored to map herbaceous cover fraction in shrublands. Optimal discernment of herbaceous cover was found using single-date imagery timed in the early summer months (early June in 2010); fractional cover of growth forms was mapped with errors of 0.08 or less. Chapter 4 investigates the stability and reliability of cover fractions modeled using a multitemporal remote sensing approach.

MESMA applied to three dates (2008, 2010, and 2011) of SPOT imagery was used successfully to model growth form fractions over a large extent of shrubland vegetation with low individual and overall error. Findings from this study indicate that MESMA applied to moderate resolution imagery timed to maximize spectral differences between growth forms is an effective approach for quantifying changes in fractional cover that exceeds ten percent, providing a way to delineate and quantify herbaceous invasions and expansions following disturbance or succession.