Seed dispersal of two juniper species: Do scatterhoarding rodents play a role?
Knowledge of seed dispersal processes is vital for understanding the ongoing expansion of the range and increase in density of juniper woodlands. Juniper seeds develop within fruit-like female cones often called “berries.” Juniper seed dispersal has largely been attributed to frugivores and endozoochory, while little attention has been paid to scatterhoarding rodents. We conducted field experiments to explore the potential role of small mammals in seed dispersal of two closely related junipers. First, we examined removal rates of intact juniper berries versus cleaned juniper seeds over 17 days at a Juniperus occidentalis site in northeastern California (Shinn Peak, SP) and a J. osteosperma site in western Nevada (Pinenut Mountains, PM) by focusing trail cameras on dishes of 100 seeds and 100 berries placed under juniper canopies. Survival analysis was used to compare removal rate curves between sites and between seeds versus berries. J. occidentalis has smaller seeds and berries with greater water content than J. osteosperma. Second, we used trail cameras to identify rodent species that removed and scatterhoarded seeds from groups of 100 radiolabeled juniper seeds placed under juniper canopies. Using a portable Geiger counter we located seeds in scatterhoards and larderhoards, recording microsite, cache data and dispersal distance.
Removal of both seeds and berries was significantly higher for J. osteosperma at PM (96.4% of berries and 87.2% of seeds removed) than for J. occidentalis at SP (60.0% berries, 50.3% seeds), and berries had significantly higher removal rates at each site than seeds. Caching trials at the two sites yielded a total of 175 scatterhoards. At both sites Peromyscus truei was the main disperser making 169 of the caches found. Dipodomys panamintinus made 5 caches of J. osteosperma at PM, and a congener at SP, D. californicus, removed all 100 seeds in two trials, but no caches were found within 50 m of the seed source. The average dispersal distance for P. truei was 4.39 m at both sites and 8.97 m for D. panamintinus at PM. Average cache size was 3.3 seeds per cache for J. osteosperma and 5.0 for J. occidentalis. Shrub canopies were the most used microsites with 44.3% of caches occurring under shrubs at PM and 37.7% at SP. Seed burial is essential to successful seedling establishment in juniper species, and scatterhoarding is an effective means of burial. Our research thus suggests that scatterhoarding rodents may be important seed dispersers of these two juniper species.