| Summary: | Trophic relationships, amongst others, define the structure of an ecosystem. They are mostly simplified and described as plant-herbivore and predator-prey interactions. Modelling trophic interactions are one way to improve our understanding of the functioning, impact and management of ecosystems. In this study, I explore how the cyclic vole and lemming populations affect the dynamics of the boreal forest in Fennoscandia. Specifically, I ask what mechanism controls the food web in years with peak and low densities of small rodents, the impact of small rodents on primary producers and how predator densities influence small rodents. To strengthen the conclusions, I test how robust the models are to ± 20% changes in parameter values. To answer these questions, I applied Ecopath, a mass-balance modelling approach, to explain trophic relationships in a system. The main output of the model is Ecotrophic Efficiency (EE), a measure to capture the consumed production of each trophic level. I modelled the vertebrate food web primarily connected to the cyclic voles and lemmings in the boreal forests, and built models according to their cycle phases. This is the first time this boreal forest community is modelled using Ecopath. The models showed a top down control on the bottom layer (mosses, lichens and fungi) in peak rodent years. The densities of small rodents would need to increase 16 fold from observed densities to negatively affect the field layer (shrubs, herbs, grasses and grass-like species). Predator density would need to increase 4 times to be able to control their prey. In addition the model were robust to parameter changes up to 20%. The system shows a strong herbivore-plant interaction in peak rodent years, but in low rodent years no control mechanism was apparent, indicating surplus resources for all components of the food web. Small rodents, specifically lemmings, deplete the bottom layer (mosses) in peak density years. Predators seem to only have a minor influence on the cycle dynamic. With this model ...
|
|---|