Ecosystem dynamics viewed from an endoperspective

The connection between ecosystems, thermodynamics and information theory is analyzed. We intend to show that the laws on computation can be applied in a much simpler and direct way than thermodynamic laws. The focus is on terrestrial ecosystems, for which boundaries are defined according to the statistical properties of energy and matter fluxes. The internal model which organisms inside ecosystems have about their environment is investigated. The evolution of an ecosystem model is considered as an alternation of model testing and model revision phases. The problem of knowledge acquisition for an observer placed inside the system and intimately connected with it (endoperspective) is discussed in detail. The bounds for the maximal information gain depending on thermodynamic conditions are derived under different assumptions. The proposed framework has not been rigorously tested yet, but it would explain much of the current dilemma in ecological modelling. Its application leads to testable predictions.