Until recently, when simulating ecological aquatic systems we have combined hydrodynamic modelling with classical (Euler type) water quality modelling. The latter represents biological and chemical variables as mean concentrations within individual calculation nodes (~ water volumes) of the hydraulic model. This of course has the constraint that behaviour and interactions of single individuals cannot be simulated explicitly. However, by introducing the ABM concept in DHIs generic ecological modelling module, ECO lab, we now provide an opportunity for combining ABM simulated as discrete agents or particles (x,y,z) in combination with both hydraulic and classical water quality modelling.

By expanding ECO lab with ABM we now have the option to define ABM -agents and their behaviour in terms of rules logics and algorithms determining the movement and multiple states, e.g. a fish agent defined by multiple states could be length, biomass, fitness, fat content etc. Each state can be modelled as a function of the other fish states, or as a function of any hydraulic or Euler type variables included in the model. To simulate movement a number of neighbourhood functions are being developed in order to evaluate the environment surrounding each agent. These neighbourhood functions return for instance the oxygen or water velocity gradients in a specified radius around each agent and these gradients can be applied directly in the rules or algorithms defining the movement behaviour of the agents.

The ABM option is not yet available in DHIs software release but is planned to be included in the release 2010. However, we already apply this knew modelling approach in DHIs consultancy, e.g. The Hounting project.