Small mammal tree seed selection in mixed forests of the Eastern United States

Small mammals play a critical role in forest ecosystems as both seed predators and dispersers; they have been shown to affect tree species composition within forests and may significantly reduce recruitment rates of certain tree species, many of which are commercially valuable. Thus, understanding small mammal seed preference is essential for both animal ecologists and foresters. Although extensive research on small mammal seed choice has been conducted both in North America and Europe, limited knowledge is available on: (1) the effects of silvicultural practices on seed choice; (2) seed selection - as most studies focus on seed use; and (3) certain understudied seed-small mammals interactions - e.g. few studies have concurrently examined both coniferous and deciduous seeds from North American mixed forests, and the seed selection of some small mammal species is not well-known (e.g. Napaeozapus). To contribute to filling these gaps, our study focused on the following objectives: (1) to quantify seed selection of seven forest seed species by small mammal species within the mixed forests of the eastern US; (2) to evaluate the influence of silvicultural practices on seed choice; (3) to explore relationships between seed choice and environmental factors such as weather and microhabitat. We conducted a series of cafeteria-style experiments in the field and in the laboratory; 2459 seed choice events, mostly by four small mammal species (Peromyscus maniculatus, Myodes gapperi, Napaeozapus insignis, and Tamiasciurus hudsonicus) were analyzed using multinomial mixed-effect models, allowing us to determine the probability of selection for each seed species. We identified a consistently high-preference seed (Pinus strobus) and one low-preference seed (Betula papyrifera). All other species (Acer rubrum, Picea rubens, Tsuga canadensis, Quercus rubra, and to some extent, Abies balsamea) had intermediate preference levels. Indeed, selection varied also by small mammal species (e.g. Acer rubrum was the top choice for Myodes gapperi in the field). Further, we found that the silvicultural practices examined here (even-aged management, two-stage shelterwood, and unmanaged) did not directly influence seed choice, whereas illumination (night- and day-time light levels), rain, and temperature did, and the magnitude of the effects varied by small mammal species. Our results show that in mixed forests, small mammals may impact seeds of economically valuable trees (e.g. Pinus strobus and Acer rubrum), whereas they may ignore seeds of less valuable trees (e.g. Betula papyrifera and Abies balsamea). We recommend that forest managers consider small mammal communities when developing forest regeneration plans.