Differential effects of contrasting phenotypes of a foundation legume shrub drive plant–plant interactions in a M editerranean mountain

Questions Are variable effects of different phenotypes of foundation plant species on subordinates across contrasting habitat conditions (with varying stress and disturbance levels) due to either varying environmental conditions or heritable differences in traits between phenotypes? To evaluate the contribution of environmental effects we quantified the effects of contrasting phenotypes of a foundation legume shrub on their subordinate species across exposure (drought stress) and grazing conditions. Location A sub‐alpine xerophytic community of western Mount Lebanon, Lebanon. Methods For two phenotypes of the spiny cushion shrub O nobrychis cornuta , a facilitative phenotype that occurs in concave topographies (mesic soil) and a competitive phenotype that occurs in convex topographies (xeric soil), we quantified cushion traits, environmental conditions and subordinate plant species abundances (within and outside the cushions) for the two phenotypes in northern (low stress) and southern (high stress) exposures, and with and without grazing. Relative interaction index ( RII ) for subordinate species richness and abundance was calculated in the eight treatment combinations and a correspondence analysis ( CA ) was conducted on species composition. Results Drought stress exacerbated phenotypic effects in southern exposure, with loose phenotypes being more facilitative and tight phenotypes more competitive than in northern exposure. This was related to both changes in cushion traits of the two phenotypes and to an increase in the pool of subordinate species sensitive to cushion effects. In contrast, grazing increased cushion positive effects of both phenotypes through the occurrence of indirect facilitation, with loose phenotypes becoming more facilitative and tight phenotypes less competitive. This was due both to changes in cushion traits of the two phenotypes and to their shared spiny phenotype limiting grazing effects. Conclusions Because increasing stress from northern to southern exposure did not increase competitive effects of either phenotype, and because the cessation of grazing did not cancel out differences in facilitative effects between phenotypes, neither drought stress nor grazing disturbance appeared to be the main drivers of the observed phenotypic effects on subordinate species across habitats. We conclude that differences in phenotypic effects of this legume shrub are very likely due to heritable differences in traits between phenotypes.