Tree-shrub associations in grazed woodlands: first rodents,then cattle?

Spatial associations of tree saplings with spiny or toxic plants in grazed woodlands are generally explained by associational resistance, i.e., protection against grazing via a well-defended neighbor. In this study, we tested whether directed seed dispersal and post-dispersal seed removal by wood mice are additional explanations for the observed spatial association between thorny shrubs and trees, hence before associational resistance. We performed three studies in grazed woodlands in the Netherlands to test this idea. Our first seed dispersal experiment with tagged acorns showed indeed that wood mice disperse acorns directed towards shrubs. The majority of these dispersed acorns were, however, consumed. Our second experiment revealed that post-dispersal removal of cached acorns was higher under shrubs than in grassland and under trees, but also indicated the importance of within-shrub position: shrub centers were faster depleted than shrub edges. Also the number of freshly emerged seedlings and older saplings differed between micro-habitats, being higher under trees and at shrub inner-edges than at shrub outer-edges, shrub centers, and in grassland. The spatial associations with shrubs got stronger with the age of saplings, which probably reflect accumulated grazing effects over time, being higher in the open than in shrubs. We conclude that directed acorn dispersal and post-dispersal acorn removal by wood mice are two additional explaining mechanisms for the spatial associations between thorny shrubs and trees in grazed woodlands, before associational resistance. Our findings change the view that large herbivores are the sole main drivers behind tree recruitment patterns in grazed woodlands.