Keywords
data limitations, ecological response models, statistical modelling
Start Date
1-7-2010 12:00 AM
Abstract
Many techniques used to model ecosystems cannot be meaningfully applied tolarge-scale ecological problems due to data constraints. Disparate collection methods, datatypes and incomplete data sets, or limited theoretical understanding mean that a wide rangeof modelling techniques used to model physical processes or for problems specific tospecies or populations cannot be used at an ecosystem scale. In developing an ecologicalresponse model for the Coorong, a South Australian hypersaline estuary, we combinedseveral flexible modelling approaches in a statistical framework to develop an approach wecall ‘ecosystem states’. This model uses simulated hydrodynamic conditions as input topredict one of a suite of states per space and time, allowing prediction of likely ecologicalconditions under a variety of scenarios. Each ecosystem state has defined sets of biota andphysico-chemical parameters. The existing model is limited in that its predictions have yetto be tested and, as yet, no spatial or temporal connectivity has been incorporated intosimulated time series of ecosystem states. This approach can be used in a wide range ofecosystems, where enough data are available to model ecosystem states. We are in theprocess of applying the technique to a nearby lake system. This has been more difficult thanfor the Coorong as there is little overlap in the spatial and temporal coverage of biologicaldata sets for that region. The approach is robust to low-quality biological data and missingenvironmental data, so should suit situations where community or management monitoringprograms have occurred through time.
Overcoming data constraints to create meaningful ecological models
Many techniques used to model ecosystems cannot be meaningfully applied tolarge-scale ecological problems due to data constraints. Disparate collection methods, datatypes and incomplete data sets, or limited theoretical understanding mean that a wide rangeof modelling techniques used to model physical processes or for problems specific tospecies or populations cannot be used at an ecosystem scale. In developing an ecologicalresponse model for the Coorong, a South Australian hypersaline estuary, we combinedseveral flexible modelling approaches in a statistical framework to develop an approach wecall ‘ecosystem states’. This model uses simulated hydrodynamic conditions as input topredict one of a suite of states per space and time, allowing prediction of likely ecologicalconditions under a variety of scenarios. Each ecosystem state has defined sets of biota andphysico-chemical parameters. The existing model is limited in that its predictions have yetto be tested and, as yet, no spatial or temporal connectivity has been incorporated intosimulated time series of ecosystem states. This approach can be used in a wide range ofecosystems, where enough data are available to model ecosystem states. We are in theprocess of applying the technique to a nearby lake system. This has been more difficult thanfor the Coorong as there is little overlap in the spatial and temporal coverage of biologicaldata sets for that region. The approach is robust to low-quality biological data and missingenvironmental data, so should suit situations where community or management monitoringprograms have occurred through time.
