Understanding a plant’s microhabitat requirements may provide context for spatial patterning on larger scales, as well as being an important tool for the conservation of rare species.  This study quantitatively investigated the microhabitats of eight herb species that co-occur on Florida rosemary scrub—a unique, relatively open vegetation type located within a more densely vegetated landscape.  We studied four species previously classified as “scrub generalists,” and four classified as “rosemary specialists.”  Fifty focal plants of each species were randomly chosen as the center of concentric circular quadrats (CCQs).  We compared microhabitats among the species-centered CCQs and with 50 random CCQs.  For each CCQ, cover estimates and species counts were conducted within three concentric circles with radii of 40, 80, and 120 cm.  Additional environmental measures were made within 30 centimeters of the focal plant or point.  Specialists were predicted to be more associated with other specialists, have less variation in microhabitat variables, and occur more often in “gaps” between shrubs than generalists or random points.  A “gap specialization index” was created using Principle Components Analysis (PCA) to score species along a gradient highly correlated with “openness.”

Results/Conclusions:

Specialist species were found more often in specialist-centered CCQs than in generalist or random CCQs.  No generalist species was associated with other generalists.  For all species except Cnidoscolus stimulosus, and Liatris ohlingerae—both scrub generalists—microhabitat scale mattered.  Litter, herb, and shrub cover values were more different from random point cover values for the 40 cm plot than the 120 cm plot.  Variation in microhabitats was highest among random plots, and was higher among all four generalist microhabitats than among specialist microhabitats.  Specialist quadrats had lower litter and shrub covers, lower fine root mass, and higher distance to nearest shrub than generalist or random quadrats.  These three variables were correlated with each other and were the main components of the PCA axis representing openness.  The habitat generalist Liatris ohlingerae had the lowest gap specialization index while the specialist Hypericum cumulicola had the highest.  The results indicate the importance of local microhabitats in determining local distributions and might account for species distributions at larger scales.