Despite the many ecologically important, habitat-forming species that maintain both sexual and asexual reproductive modes, we have little understanding of how the relative allocation to each mode may impact the persistence, life history characteristics, or genetic structure of natural populations. However, several investigators have suggested situations when asexual reproduction is predicted to be important. For example, asexual reproduction is postulated to be more important for populations in stressful habitats than those occupying benign habitats. Here I compare the relative contribution of sexual and asexual reproduction to population persistence in two species of the seagrass genus Phyllospadix (the surfgrasses). I use the fact that these species span an intertidal stress gradient with P. scouleri occupying the relatively benign low intertidal and P. serrulatus inhabiting the stressful high intertidal to assess the impact of environmental stress on allocation between reproductive modes. I apply matrix projection models to four years of demographic data from two sites to ask how asexual and sexual growth contribute to population growth and use methods developed by Cochran and Ellner (1992) to understand how allocation to each reproductive mode affects plant longevity. Finally, I contrast matrix model results with genetic population structure estimates from AFLP markers.

Results/Conclusions
For both species, mean matrix estimates of population growth rate were approximately equal to 1 (λ= 0.99 ± 0.03 for P. serrulatus and λ = 0.97 ± 0.03 for P. scouleri). Elasticity analyses showed that asexual reproduction contributed more to population growth than sexual reproduction for both species, but the relative contribution of asexual reproduction was approximately 10 times greater in P. serrulatus than P. scouleri. Surfgrass plants were shown to have surprising longevity (expected age of largest size class ~ 70 and ~ 120 years for P. scouleri and P. serrulatus, respectively). Halving the success rate of asexual reproduction lowered the expected age in each class by 20-40%. AFLP results show high population differentiation over short distances (10s of km) for both species, but substantially higher Fst values for P. serrulatus. My results are in line with the prediction of increased reliance on asexual reproduction in stressful habitats and show how life history characteristics and genetic structure may be altered by the relative success of sexual versus asexual reproduction.