Spatial-scale dependencies in the predation of seeds by rodents

Abstract

Summary 1. Previous studies have shown rates of seed predation in deciduous woodland to be high (on average 60%) and extremely variable in space and time. 2. Post-dispersal seed predation by forest rodents was examined simultaneously in five areas of deciduous woodland surrounding Durham City, Co. Durham. 3. This study investigated the influence of seed density, seed burial and seed species in determining the rates of seed encounter and exploitation by rodents of both native and exotic seed. The effect of within- and between-sites differences were also examined. 4. Live-trapping studies and the use of exclosure treatments revealed that Apodemus sylvaticus and Clethrionomys glareolus were the major seed predators. Together, they were responsible for removing between 55-80% of the experimental seed supply. 5. Seed burial accounted for a high proportion of the variation and significantly reduced the frequency of rodent encounter compared to surface seeds at both seed densities. Furthermore, it increased the variation in encounter due to density, species and site effects. It had no significant influence, however, on the extent to which groups of ten seeds were exploited once encountered. 6. Variations in the frequency of seed encounter and exploitation were both strongly affected by changes in seed density, with high densities increasing the chance of seed detection and removal. Significant species effects were also detected for seed encounter and exploitation. 7. Seed predation was spatially patchy, between and within experimental sites, possibly reflecting variation in the spatial distribution of seed predators. 8. The selective nature of the seed predators, plus the relative patchiness of predation intensity in space, suggest that post-dispersal seed predation may play an important role in determining the distribution and/or abundance of deciduous woodland plants.