AN OBJECT-ORIENTED SIMULATION FRAMEWORK FOR INDIVIDUAL-BASED SIMULATIONS (OSIRIS) - DAPHNIA POPULATION-DYNAMICS AS AN EXAMPLE

Citation
Wm. Mooij et M. Boersma, AN OBJECT-ORIENTED SIMULATION FRAMEWORK FOR INDIVIDUAL-BASED SIMULATIONS (OSIRIS) - DAPHNIA POPULATION-DYNAMICS AS AN EXAMPLE, Ecological modelling, 93(1-3), 1996, pp. 139-153
Citations number
45
Categorie Soggetti
Ecology
Journal title
ISSN journal
03043800
Volume
93
Issue
1-3
Year of publication
1996
Pages
139 - 153
Database
ISI
SICI code
0304-3800(1996)93:1-3<139:AOSFFI>2.0.ZU;2-V
Abstract
A general framework for the implementation of ecological models direct ed towards the falsification of knowledge, as opposed to models direct ed at making predictions, is proposed. The framework is constructed by defining a set of classes, with their interrelationships, in an objec t-oriented programming language. The classes represent the major level s of the so-called levels-of-integration hierarchy: individual, popula tion and system. The abiotic physical and chemical environment is impl emented by the classes condition and resource, respectively. Class hab itat is used to represent the spatial structure of an ecosystem. The s imulation is controlled by a class called analyser. The simulation mec hanism is implemented by deriving all these real-life objects from a m ore abstract class simobject. The engine of the simulation is formed b y a dynamic list of references to simobjects, sorted according to the time each simobject should be activated next, The data of each object are implemented in class dataobject, from which simobject is derived. The applicability of this framework, called OSIRIS (object-oriented si mulation framework for individual-based simulations), is shown for a p opulation dynamical study on daphnids, The effects of variation among individual daphnids on the growth rate and structure of a population o f daphnids are studied by comparing the results of the individual-base d model with those of a life table. Moreover, variation in population growth rate over time, which parameter cannot be derived from a life t able, is calculated. Finally, the sensitivity of the model for the num ber of modelled individuals and the sampling interval is analysed.