Interactions between small rodents and their food plants in tundra habitats

Abstract

Small rodents are key herbivores of arctic ecosystems, where their cyclic population dynamics have important implications for vegetation dynamics. The role of vegetation for shaping small rodent population dynamics remains, however, unclear. Evaluation of this interaction has been greatly hampered by the rather fragmentary knowledge of small rodents feeding ecology, which in turn is due to methodological challenges in studying small rodent diets. In this thesis, I investigated interactions between small rodents and vegetation, focusing on variability of both diets and plant defences induced by rodent herbivory. In order to overcome limitations set by methods, I have evaluated the use of two recently developed methods and one traditionally used method for studying small rodent diets. I found that novel DNA metabarcoding methods give the best taxonomic resolution whereas traditional microhistological methods may still be useful to elucidate which plant parts have been eaten. Moreover, use of stable isotope analysis in research of small rodent feeding habits can be useful when evaluating the temporal persistence of diets. In three observational studies, I described diets of small rodents in tundra habitats and assessed sources of variability in them. I found that diets of my study species were more diverse than previously believed, suggesting that diversity is an important but previously underrated trait of small rodent diets. Nutritional quality seems to determine which food items are preferred, as small rodents selected in general for most palatable food items. In addition to nutritional drivers, my results suggest that various ecological drivers, such as food availability, competition and predation risk avoidance, are important determinants of for small rodent diets in the wild. A better incorporation of ecological and nutritional drivers into a common framework is therefore needed to understand what shapes diets of small rodents, and herbivores in general, in natural habitats. I evaluated the impact of herbivory on grass silica defences in an experimental study. I found that levels of such defences vary between locations, probably due to heritable differences among grass populations in the response to herbivory and differences in local growth conditions. Based on such spatial variation and the large proportion of other food items included in the diets of the focal small rodent populations, it is unlikely that silica-based defences would have a strong direct role in shaping small rodent population dynamics in Finnmark. However, increased levels of silica in grasses reduce grass palatability for rodents, and hence probably impact rodent diet quality indirectly, through a shift to diets with less grasses. I therefore suggest that availability of alternative food items is likely to be an important factor shaping the interaction between small rodents and grasses. Variability in both small rodent diets and plant defences suggests that interaction between vegetation and a small rodent species may show pronounced spatial variation. This, together with my findings on the diversity of diets and its increase during high population densities, indicates that a single plant defence mechanism is unlikely to shape rodent-vegetation interactions. I suggest that deeper understanding of the role of food for small rodent population dynamics could rather be gained by focusing on the compensatory nutritional effects of different food items as well as diet diversity.