|Premise of the study: Californiaís vascular fl ora is the most diverse and threatened in temperate North America. Previous studies of spatial patterns of Californian plant diversity have been limited by traditional metrics, non-uniform geographic units, and distributional data derived from fl oristic descrip-tions for only a subset of species.
Methods: We revisited patterns of sampling intensity, species richness, and relative endemism in California based on equal-area spatial units, the full vascular fl ora, and specimen-based distributional data. We estimated richness, weighted endemism (inverse range-weighting of species), and corrected weighted endemism (weighted endemism corrected for richness), and performed a randomization test for signifi cantly high endemism.
Key results: Possible biases in herbarium data do not obscure patterns of high richness and endemism at the spatial resolution studied. High species richness was sometimes associated with signifi cantly high endemism (e.g., Klamath Ranges) but often not. In Stebbins and Majorís (1965) main endemism hotspot, Southwestern California, species richness is high across much of the Peninsular and Transverse ranges but signifi cantly high endemism is mostly localized to the Santa Rosa and San Bernardino mountains. In contrast, species richness is low in the Channel Islands, where endemism is signifi cantly high, as also found for much of the Death Valley region.
Conclusions: Measures of taxonomic richness, even with greater weighting of range-restricted taxa, are insuffi cient for identifying areas of signifi cantly high endemism that warrant conservation attention. Diff erences between our fi ndings and those in previous studies appear to mostly refl ect the source and scale of distributional data, and recent analytical refi nements.|