Seed dispersal in Juniperus has been largely attributed to frugivores that consume the female cones, which are fruit-like in some species such as western juniper (WJ) but dry and woody in others such as Utah juniper (UJ). Scatter-hoarding rodents have generally been ignored or regarded as juniper seed predators. We have been investigating potential roles of scatter-hoarding rodents in dispersal of western and Utah juniper seeds, and hypothesize that they act either as secondary dispersers in species with fruit-like cones (WJ), by removing and caching seeds from frugivore feces, or as primary dispersers in species with woody cones (UJ). At northeastern California sites, we compared bird and rodent removal of whole WJ cones, hand-cleaned seeds, and seeds removed from bird feces, and quantified microsite-specific distributions of bird-passed WJ seeds and subsequent removal by rodents. Additionally, seedling emergence experiments were conducted with whole WJ cones, hand-cleaned and bird-passed seeds buried at 1 cm or on the soil surface. To understand whether juniper seed and cone traits influence potential dispersers, removal of whole WJ and UJ cones and hand-cleaned seeds was monitored at a WJ and a UJ site. Removal and caching of radioactively-labeled seeds at these sites was also followed.
Results/Conclusions
In our removal experiment with WJ cones, hand-cleaned and bird-passed seeds, birds removed only cones, while rodents preferred seeds at one site but showed no preference between seed types. WJ seeds defecated by birds were most abundant under juniper canopies and were heavily depleted by rodents in all microsites from winter to the following summer. In seedling emergence experiments, no WJ seedlings resulted from seeds or cones on the soil surface. When buried, however, seeds produced more seedlings than cones and bird-passed seeds had greater emergence than hand-cleaned seeds. Removal experiments with UJ and WJ seeds and cones at a WJ and a UJ site showed that cones from the local juniper species were preferred, but removal of local and non-local seeds was similar. Piñon mice removed the most seeds and cones at both sites. In caching experiments, we recovered many radio-labeled seeds in scattered caches (WJ: 82 caches, 415 seeds, 63.0% of seeds found; UJ: 127 caches, 458 seeds, 39.5% found) – most being attributable to piñon mice. Taken together, our research suggests that WJ seeds are dispersed by frugivorous birds and scatter-hoarding rodents that act as secondary or complimentary dispersers, while UJ is dispersed directly by scatter-hoarding rodents.