While models of species coexistence largely focus on how competition defines biological communities, over recent decades, a number of studies show positive plant-plant species interactions (facilitation) can also promote stable coexistence. The long-lived, co-dominant shrubs California buckwheat (E. fasciculatum) and California sagebrush (A. californica) share a well-documented positive association at the habitat level in their native California coastal sage scrub ecosystem, but mechanisms underlying their interactions remain unclear at finer spatial scales. Here, a hypothesis that E. fasciculatum acidifies CSS’s alkaline soils and facilitates A. californica through amelioration of alkalinity stress is tested in a greenhouse experiment and association tests in the field. Greenhouse results demonstrate facilitation at early growth stages. In late growth stages, water competition is known to determine the shrubs’ interactions with each other, but here, field observations of the shrubs in late growth stages show positive associations between A. californica and E. fasciculatum that have a positive linear relationship to increasing soil pH. These results highlight the importance of understanding lifecycle-long interactions among species in evaluating facilitation’s impacts on community structure.

The presence of E. fasciculatum and A. californica in late growth forms (canopies > 1 m diameter) as well as all other plant species grouped together were recorded in 560 1 m² quadrats. Sampling sites were located in five CSS preserves and nature parks (see Table S1) along a N-S coastal gradient from San Onofre State Beach (33.376068, -117.567275) to Zuma Canyon State Park in Malibu (34.044757, -118.809499). At each site, 10 m² quadrats were chosen to filter microsites that supported growth of both target species. Within each 10 m² quadrat, three parallel 1 x 10 m belt transects were randomly chosen, except in Zuma Canyon and San Onofre Beach. In these two latter sites, 1x10 m belt transects were sampled in the only areas in which E. fasciculatum and A. californica were observed growing: along washes with thin soils in Zuma Canyon and along the edges of vernal pools atop seaside bluffs in San Onofre Beach. The presence of E. fasciculatum and A. californica as well as any other plants was recorded in the 10 1 m² quadrats within the 1 x 10 m belt transects. Plant species other than the target spcies were analyzed as one group. Associations were analyzed at 1 m², because it was assumed E. fasciculatum, A. californica and other species would interact at a species-to-species level for limited water, nutrients, and light. (See Table1 for site-specific count numbers.)

The 560 1 m² quadrats were aggregated with species data in five 50 m point-intercept transects from five additional CSS sites that were independently surveyed the same year by UC Irvine’s Center for Environmental Biology (CEB) along a coast-to-inland gradient (see Table S1 for coordinates of all locations). The presence of E. fasciculatum, A. californica and all other plant species grouped together in CEB transects was counted in adjacent 1 m blocks to conform surveying methods as much as possible with belt transects. For example, if E. fasciculatum was present at the 4 m point-intercept and A. californica was present at the 5 m point-intercept, then they were counted as co-occurring at a scale of 1 m. Since both shrub species typically have canopies > 1 m in late growth stages, these counting methods mean some shrubs were likely counted twice, but these methods better reflect the relatively large area of vegetative cover and effects on the distribution of other species that E. fasciculatum and A. californica in adult growth forms exhibit.

For analysis of spatial associations using chi-squared contingency tables, percent covers were converted to presence/absence of A. californica, E. fasciculatum, and all other plant species.