Reseeding and resprouting woody shrubs are predicted to be restricted to habitats with long and short disturbance return intervals, respectively. Nonetheless, resprouters and reseeders co-occur along Gulf of Mexico coastal ecoclines where lightning-ignited fires occur frequently. A previous study of reseeder and resprouter habitat associations hypothesized that coexistence might result from occupancy of different niches along ecoclines. Hypericum microsepalum (resprouter) is associated with upland, flatwoods pine savannas where soil moisture is lower and fires tend to burn frequently and evenly across the landscape. In contrast, H. brachyphyllum (facultative reseeder/resprouter) and H. chapmanii (reseeder) are associated with wet lowlands; fires spreading from uplands typically burn increasingly patchy in progressively wetter areas. Reseeder and reseeder/resprouter seedlings are associated with drier habitats than established plants of the same species. We tested for obligatory habitat associations using reciprocal transplants. In 2006, replicated 2700cm³ plots containing at least ten established plants of each species were transplanted, depending on the species, into locations containing the same species (0 cm) and the other two species (10cm and 20cm higher or lower along the ecocline). Control and transplant plots were established along eight ecoclines and sampled every two months over three years for survival, growth, and reproduction.

Results/Conclusions

Survival of transplants was greatest at the moist, lowland end of ecoclines regardless of originating location along the ecocline. Accelerated failure-time analyzes indicated that H. brachyphyllum and H. chapmanii transplants had the highest survival in those lower locations along ecoclines where they naturally occur in the field. Hypericum microsepalum had the greatest survival in habitats lower in elevation along the ecocline than where they occur naturally. Growth of transplanted Hypericum reseeders and resprouters was not significantly different from control plots. In all years, there was no difference in fecundity between H. microsepalum transplants and controls. Hypericum chapmanii and H. brachyphyllum did not flower in the first year after transplanting, and in years two and three, they showed no difference in fecundity between controls and transplant plots. Our data thus indicate that edaphic characteristics alone do not control the distinct habitat associations of established plants of reseeders and resprouters along coastal ecoclines. Instead, differential mortality of seedlings or juveniles may restrict the occurrence of established plants to certain portions of ecoclines. Alternatively, other ecological processes, such as flooding-fire interactions that create heterogeneous fires over space and time, may influence reseeder and resprouter distributions along ecoclines.