Environmental Engineering Reference
In-Depth Information
ecology propagate the coupling of different model types in which individual-based
models are thought to play an integrating role (Cury et al. 2008) because of their
flexible structure, which allows to combine knowledge and data from different sub-
disciplines, that can be used to analyze findings on heterogeneous ecosystem levels
and to understand the corresponding complex interaction and network structure.
In particular, since radio-tracking became possible, the modelling of the beha-
vioural repertoire within a spatial context, and especially models that explicitly
investigate animal movement and dispersal, have become an important domain of
IBM applications (Gustafson and Gardner 1996; Broekhuizen et al. 2003; Jopp
2003; Pe'er and Kramer-Schadt 2008). IBMs have greatly contributed to the study
of population dynamics in complex landscapes (Lima and Zollner 1996; Nathan
et al. 2008; Revilla and Wiegand 2008). Population development may be simulated
in dependence of complex behavioural modes or context dependent changes in
reactions (Shin and Cury 2001) also including the field of population viability
analysis (PVA, e.g. LePage and Cury 1997; Mazaris et al. 2005). IBM of invasion
processes allow combining dispersal processes with specifies properties, individual
behaviour and the properties of the invaded community (Higgins et al. 2001;
Nehrbass et al. 2007).
Since their beginning, IBMs have undergone a rapid development and have been
applied to almost all ecological topics and a large number of research questions
(DeAngelis and Gross
1992
; DeAngelis and Mooiji 2005; Grimm and Railsback
2005
). In the following, we illustrate the basic structure of IBMs and demonstrate
their functionality on the basis of simple model applications.
12.2 The Structure of Individual-Based Models
In order to give an overview on basic formal elements of individual-based models
we begin with a short introduction into the programming background and an outline
of the concept of object-oriented programming (OOP, e.g. Rumbaugh et al. 1991;
Silvert 1993; Hill
1996
). Then we look at the application of the OOP to conve-
niently describe structure and interaction of individual actors.
In OOP, the source code is organized in blocks which are delimited from each
other. There are different types of blocks with the so-called CLASS as the most
prominent one. A CLASS is a specific programme unit which consists of storage
reservations for variables and may additionally contain code (statements) how to
change the values of its variables. In such a CLASS it is possible to specify further
sub-classes which allows to implement a hierarchic programme structure. During
programme execution, a CLASS can be copied multiple times and these class
instances may be kept available in the computer storage. This is the decisive feature
in object-oriented programming. The copies of a class are called OBJECTS, thus
being eponymous for the whole approach. Each of these objects, which are avail-
able during runtime of the programme, consists of the same code as the class from
which it is derived, but may contain specific values stored in its variables. This
