Spatial Variation in the Selective Scenarios of Hormathophylla spinosa (Cruciferae)

The effects of multispecific systems containing both mutualistic and antagonistic interacting organisms on the evolution of plant traits have seldom been analyzed. We studied the selection exerted by several species of herbivores and pollinators in three populations of Hormathophylla spinosa (Cruciferae) in the Sierra Nevada (Spain) over 4 yr by using path analyses and structural equation modeling (SEM). The main selective pressures in our study sites were ungulates and pollinators. However, the importance of each kind of interacting organism differed among populations. Our results indicate a selection mosaic among populations of H. spinosa in the Sierra Nevada caused by the spatial variation in the relative importance of different interactions as selective pressures. We found two main selective scenarios, depending on the presence or absence of ungulates. In the populations with low ungulate pressure, there was positive phenotypic selection in flower number per plant and in flower density (mediated by nectarivorous pollinators). In the two populations with high ungulate pressure, there was a strong positive, ungulate-mediated selection in thorn density. Our results suggest that the application of SEM to several populations simultaneously monitored might help to isolate the major selection pressures on local populations and identify potential differences in selection among populations, becoming a useful exploratory approach to study the geographical variation of selection in complex systems.