Abstract:

We established a field experiment to explore effects of different releasing methods and coverage treatments on predation and removal of Quercus wutaishanica, Prunus salicina and Pinus armandii seeds by rodents. We used three release (separately releasing, assembly releasing, and blend releasing) and cover (control, litter coverage, and soil coverage) treatments in young Q. wutaishanica stands in the Liupan Mountains of Ningxia in spring 2010. We detected significant effects of seeds of different species, release methods, and coverage treatments on the in situ seed predation rate. Seed predation and hoarding rates after removal by rodents were influenced by seed size, thickness and hardness of seed testa (endocarp), release methods, and coverage treatments. Although the average distance of seed movement during predation and hoarding events was within 5 m, the hoarding distance was longer than the predation distance for all three species. Specifically, we found that the predation distance of P. armandii seeds was significantly longer than those of Q. wutaishanica and P. salicina seeds, whereas the hoarding distance of the former was significantly shorter than the other two species. Sites that rodents used as predation and cache sites tended to contain only a single seed at a majority of the detected cache and predation sites. Soil burial was the primary strategy that rodents used to cache seeds. In summary, seed physical characteristics, distributional patterns, and ground coverage in forest communities may affect seed predation and hoarding behaviors of rodents by extending seed handling time and increasing predation risk of estimation by rodents and as a result, determined seed fates.

Key words: rodent, seed fate, scatter hoarding, seed dispersal, Liupan Mountains