Xerie Mediterranean-type shrubland associations of Alta and Baja California and the community/continuum debate

A survey of the xeric shrublands of Pacific coastal North America from San Francisco to El Rosario (Mexico), including the inner Channel Islands, was conducted using 99 sample sites of 0.063 ha size. TWINSPAN classification and DECORANA ordination confirmed the existence of two plant formations, distinguishable physiognomically: coastal sage scrub and coastal succulent scrub. Within coastal sage scrub, four floristic associations were recognized: Diablan, Venturan, Riversidian and Diegan. Within coastal succulent scrub, two floristic associations were defined: Martirian and Vizcainan. These associations occur in distinct geographical regions following the coastline, with the Riversidian association occurring in the basin inland from Venturan and Diegan regions. Their locations are strongly correlated with differences in evapotranspirative stress regimes. The Channel Island sites show affinities to several of the mainland associations. The Venturan association can be further subdivided floristically into two subassociations, dominated by Salvia mellifera and S. leucophylla respectively. These subassociations which are coextensive geographically at a regional scale, typically do not intermingle at a local scale but often meet along sharp boundaries in the landscape. The dominant species segregate by moisture preference, S. mellifera preferring coarser-texture soils and more southerly aspects than S. leucophylla. Richness and equitability of these sites are depressed relative to other xeric shrubland sites, reflecting the fact that the two subassociations partition the Venturan flora into substantially non-overlapping subsets of species. This segregation of associates between the two Salvia dominance types strongly suggests biotic influence of the dominants on subordinate species, perhaps mediated by allelopathy. This biotic interaction, leading to relatively strong floristic subassociations segregating independently in the landscape, would provide an example of the holistic community structure referred to by Clements and his followers, embedded within a larger pattern of continuity in species distributions.Nomenclature follows Munz & Keck (1959), Munz (1974) and Wiggins (1980).I am grateful to the following for research assistance: S. Coon, E. Hobbs, S. Lavinger, J. F. O'Leary, K. R. Preston, B. Rich and A. Troeger. I also thank the numerous public and private land owners who permitted access to the study sites. This research is based upon work supported by the National Science Foundation under grant DEB 76-81712.