Beaver digging during highwater with an agent based model : Case studies in Dutch river sections

In this report an improved model of the behaviour of beavers during high-water is presented with an Agent-Based Model (ABM). The inputs of this model are based on an extensive overview of parameters that influence beaver behaviour, which is based on literature and consultation with beaver experts. When comparing the model result with field observations, the model appears to be able to predict the locations of the beaver burrows with acceptable accuracy, but it is not yet very sensitive to hydro-meteorological conditions. The validation data is not very comprehensive, hence a challenge remains in differentiating between hydro-meteorological conditions at the burrow locations and their time of detection. A more complete set of data would greatly improve calibration effectiveness and advance the model even more. It is always necessary to run the model various times on the same settings to account for randomness of hydro-meteorological sensitivity, lodge location (beaver’s home), and burrow location selection. The more information about the territory is available, such as lodge locations and number of beavers in the river section, the more accurate the prediction will be. Nevertheless, it is important to note that there is still much uncertainty related to beaver behaviour and the extent to which it can be modelled.