MODELING STRUCTURAL DYNAMICAL CHANGES IN A DANISH SHALLOW LAKE

Authors
Citation
Sn. Nielsen, MODELING STRUCTURAL DYNAMICAL CHANGES IN A DANISH SHALLOW LAKE, Ecological modelling, 73(1-2), 1994, pp. 13-30
Citations number
38
Categorie Soggetti
Ecology
Journal title
ISSN journal
03043800
Volume
73
Issue
1-2
Year of publication
1994
Pages
13 - 30
Database
ISI
SICI code
0304-3800(1994)73:1-2<13:MSDCIA>2.0.ZU;2-C
Abstract
Biomanipulation and diversion of nutrients can affect not only the eut rophication level, but also the trophic structure and species composit ion of the phytoplankton society in lakes. Existing models often fail to describe the reactions of the ecosystem when such qualitative chang es are involved. For this purpose structural dynamic models are needed . This paper presents the first attempt to model the qualitative chang es that have been observed to occur in a Danish shallow lake as a resu lt of biomanipulation. A structural dynamic model was developed that i ncluded nine species or types of algae normally observed in Danish sha llow lakes. The high flow ratio of the lake was assumed to justify tha t summer conditions would represent a steady state of the system. The parameters of the model were calibrated until the model gave adequate results. The available mathematical tools normally used for calibratio n were found to give unsatisfactory results. The calibration therefore had to be carried out by running the model repetitively, adjusting pa rameters until it was able to perform according to the changes actuall y observed in the lake chosen for this case study. The use of paramete rs from literature to characterize the different types of algae is sho wn to have a great impact on this phase of model development. The cali bration was found to be very time consuming, mainly due to uncertainti es connected to the input parameters. The importance and uncertainty o f the parameters chosen indicate that we may probably have to choose n ew strategies for similar modelling approaches in the future, such as the application of goal functions. The implementations would allow par ameters to vary during simulations in order to simulate ecosystem prop erties, such as adaptational processes and buffering capacities which are features possessed by natural ecosystems.