We examine the roles of two biotic interactions that have not previously been considered in demographic models: indirect interactions and hybridization. Lupinus tidestromii is a federally endangered plant that is restricted to 15 extant populations on coastal dunes in California, where it co-occurs and hybridizes with its common congener, L. chamissonis. Recent invasions of the grass, Ammophila arenaria, have resulted in elevated densities of the native seed predator, Peromyscus maniculatus, and high levels of predation on these Lupinus species, which is disproportionately intense for the rare species. Furthermore, the two species were historically parapatric in a portion of the range, but recent horticultural introductions of L. chamissonis have brought these species in contact, where hybridization now occurs. We use demographic modeling to assess the degree to which seed predation, hybridization, and their interaction contribute to population dynamics and extinction risk of L. tidestromii. Between 2008 and 2009 we monitored survival and growth of 2440 seedling, non-reproductive, and reproductive individuals in 13 populations of L. tidestromii. We sowed 100 seeds into seed baskets at four populations to measure germination rates. We quantified pre-dispersal seed predation by following marked flowering stalks to observe fruit fates.

Results/Conclusions

Survival, growth, and regression of individuals in each stage class were highly variable among populations. Mean survival of seedlings was 0.39 (0.18-0.67), of non-reproductive individuals was 0.75 (0.45-0.93) and of reproductive individuals was 0.73 (0.38-0.94); mean germination rate was 0.09. Preliminary analysis of predation showed it was variable among populations and ranged from 0.35 to 0.65 racemes consumed. These data suggest that management actions that reduce seed predation, such as exclusion cages or removing invasive grass that houses seed predators, will enhance fecundity; actions that reduce disturbance by cattle ranching or human traffic will enhance plant survival. We will extend this model by incorporating direct estimates of hybridization rates into a frequency-dependent, stage-structured model to evaluate the separate and joint effects of indirect interactions and hybridization on the population dynamics and viability of L. tidestromii throughout its range. This will be the first study to empirically quantify the effect of hybridization on population viability of a species.