The little apple can fall far from the tree: Seed dispersal of greenleaf manzanita (Arctostaphylos patula)

Background/Question/Methods

Three inclusive hypotheses have been suggested as to why seed dispersal benefits plants; the most understudied of these hypotheses is the directed dispersal hypothesis. This hypothesis posits that a benefit of animal-mediated seed dispersal is that seeds are dispersed to microsites that facilitate seed germination. We examined this hypothesis in a novel way by studying the interaction between plants, animals, and disturbance. We hypothesized that in fire-prone systems, seed-caching animals will provide an additional benefit to plants by caching seeds in microsites that are not only suitable for germination, but also safe from fire events. We chose greenleaf manzanita (Arctostaphylos patula, Ericaceae) to test this hypothesis because, like most manzanitas, its seeds are sensitive to heat and need charate scarification to germinate. Further, manzanitas (Arctostahpylos spp.) are appropriate to use as a model because they are the most diverse woody taxon in and ubiquitous throughout the California Floristic Province, an area that experiences frequent fires.

Results/Conclusions

We radiolabeled seeds over a period of two years to follow dispersal and found that scatter-caching rodents disperse seeds away from parent plants on the order of tens of meters and cache them in sites safe from fire events. After fire, we found a significant positive correlation between seed germination from rodent caches and fire intensity. By deploying full-shrub exclosures, we also found that treatments allowing rodents and no other potential dispersers (e.g., birds, large mammals) access had fruits depleted over one order of magnitude faster than other treatments that excluded rodents and allowed other potential dispersers access. Although we did find evidence of medium-to-large mammals dispersing seeds, the events were less frequent, presumably ineffective at delivering seeds to safe sites, but could potentially serve as long-distance dispersal function. These findings suggest that two-phased seed dispersal may be occurring in this system, with scatter-caching rodents not being effective at moving seeds horizontally but effective vertically in space and other dispersers having the opposite effects. We further speculate that in fire-prone systems, seeds that are scatter-cached have an advantage over those that are otherwise abiotically incorporated into the soil and that this syndrome is ubiquitous throughout these systems.