The influence of cache site and rodent pilferage on whitebark pine seed germination in the Northern and Central Rocky Mountains

Plant regeneration is a multistage process that includes: 1) production of viable seeds, 2) dispersal to microsite types suitable for germination, 3) persistence of seeds, 4) germination, 5) seedling survival, and 6) plant establishment. Any of these stages may act as a bottleneck to mature tree recruitment. Cache site selection by seed dispersers influences regeneration by determining both germination potential and risk of seed removal by rodent seed predators (i.e., pilferage). Whitebark pine (Pinus albicaulis Engelm.)---a subalpine and treeline conifer distributed throughout the western United States and Canada---depends on the Clark's nutcracker (Nucifraga columbiana Wilson) for seed dispersal. Nutcrackers store seeds in many caches comprised of 1-15 seeds, which are buried 2-3 cm under substrate. In two study areas, I determined 1) germination rates of whitebark pine, 2) cache pilferage rates, 3) how germination and pilferage rates vary with elevation, microsite type, and cache size, 4) which rodent species are present and might be responsible for pilfering nutcracker caches, 5) whether a relationship exists between granivorous rodent density and pilferage rate.;In 2012, I created 735 simulated caches in six microsite types, reported as commonly used by nutcrackers on White Calf Mountain, Glacier National Park, Montana, and Tibbs Butte, Shoshone National Forest, Wyoming. Cache sizes were randomly selected from a Poisson distribution that ranged in size from 1 to 7 seeds created with cache size data from the literature. In 2013, I checked each cache for seed germination and pilferage. I determined whether the odds of germination and pilferage differed based on cache site characteristics by determining whether the 95% confidence intervals of the odds ratios included 1.0. Odds of germination in the subalpine zone of White Calf Mountain were lower than at treeline (OR = 0.585, 95% CI = 0.360 - 0.952); differences are likely caused by the late seral successional status of the forest, which is not conducive to whitebark pine regeneration. On Tibbs Butte, odds of germination were higher near rocks and near no object relative to trees (OR = 2.26, 95% CI = 1.31 - 3.91; OR = 2.51, 95% CI = 1.46 - 4.33, respectively). Odds of pilferage were higher at treeline than in the subalpine (OR = 0.524, 95% CI = 0.389 - 0.706). These results suggest that the influences of cache site characteristics on whitebark pine germination vary greatly geographically and with community type (both among and within Tibbs Butte and White Calf Mountain).