Integrated Modeling of Brook Trout Growth, Movement and Population Dynamics

Developing mark-recapture models in a state-space Bayesian framework is increasingly common.  The state-space framework allows separate modeling of the state process (e.g. growth, movement or survival) and the observation process (i.e. incomplete detection).  It is also an extremely flexible approach allowing key processes to be integrated and inform each other. For example, a growth model can fill in missing observations of body size, and be used to provide complete covariate information for the effect of body size on survival or movement. We have developed an integrated model of population processes and use it to estimate relationships among processes. The result is an integrated model that takes full advantage of the data, can be used for sensitivity analysis, and for forecasting effects of changes in environmental drivers (stream flow and temperature).  Brook trout have been sampled seasonally from the West brook (Whately, MA) and three of its tributaries seasonally for almost 15 years.  Model results indicate that the effect of body size on survival varies dramatically across seasons and there is a strong survival cost to movement among tributaries with little growth advantage. This modeling approach can easily be extended to get better estimates of recruitment, abundance, and density dependence.