• Apodemus sylvaticus;
  • density dependence;
  • frequency dependence;
  • functional response;
  • herbivory

1. The response of post-dispersal seed predators to changes in the absolute and relative abundance of seeds of two tree species: ash (Fraxinus excelsior) and wych elm (Ulmus glabra) was studied in seminatural woodland in County Durham, UK.

2. Analysis examined two components of seed predation: seed encounter (the probability of at least one seed being removed) and seed exploitation (the proportion of seeds removed once encountered). Exclosure studies identified small mammals, particularly the woodmouse (Apodemus sylvaticus) as the principal post-dispersal seed predators and revealed a marked preference (almost four-fold difference in encounter) for seeds of Ulmus over Fraxinus.

3. For both Fraxinus and Ulmus, a linear regression model described the relationship between seed density and the number of seeds removed by rodents more successfully than non-linear models. These relationships were maintained whether the species were presented singly or together with neither the slopes of Fraxinus or Ulmus changing significantly. These results indicate that, over the range of seed densities used, rodents consumed a constant proportion of seeds irrespective of seed density or frequency. The high rates of exploitation for Ulmus seeds suggests the slope of the relationship primarily reflects rates of encounter, whereas for Fraxinus it is a result of both low encounter and exploitation.

4. Linear regression identified seed removal by rodents to be frequency-independent both over the woodland as a whole and within each of six micro-habitats. This appeared to be a result of the marked preference for Ulmus over Fraxinus which was not reversed even when Ulmus was rare. Frequency-dependent seed predation by rodents is predicted to be most likely when both prey have similar low palatabilities.

5. When encountered by rodents, patches of Ulmus seeds were exploited almost completely, irrespective of seed density or frequency which suggests rodents have the potential to cause local extinction of Ulmus seed populations. In contrast, the lower rates of encounter and exploitation of Fraxinus seeds implies ample opportunities for prey escape.