Contents
xv
6.3 Results ................................................................................................... 97
6.3.1
Classification Trees ....................................................................... 97
6.3.1.1 Model Development and Validation .............................................. 97
6.3.1.2 Application of Predictive Classification Trees for River
6.3.2
6.3.2.1
6.3.2.2
Management .................................................................................. 98
Artificial Neural Networks .......................................................... 100
Model Development and Validation ............................................ 100
Application of Predictive Artificial Neural Networks for
River Management ....................................................................... 102
6.3.2.2.1
Prediction of Environmental Standards ................................ 102
6.3.2.2.2
Feasibility Analysis ofRiver Restoration Options ............... 103
6.4 Discussion ........................................................................................... 104
Acknowledgements ............................................................................. 105
References ........................................................................................... 105
7. Modelling Ecologlcal Interrelations in Running Water
Ecosystems with Artiflclal Neural Networks ........................ 109
7.1 Introduction ......................................................................................... 109
7.2 Materials and Methods ........................................................................ 110
7.2.1
DataBase ..................................................................................... 110
7.2.2
Data Pre-Processing ..................................................................... 110
7.2.3
Artificial Neural Network Types ................................................. 111
7.2.4
Dimension Reduction .................................................................. 111
7.2.5
Quality Measures ......................................................................... 111
7.3 Data Exploration with Unsupervised Leaming Systems ..................... 112
7.4 Correlations and Predictions with Supervised Leaming Systems ....... 115
7.4.1
Correlations and Predictions of Environmental Variables ........... 117
7.4.2
Dependencies of Colonisation Patterns of Macro-Invertebrates
on Water Quality and Habitat Characteristics .............................. 117
7.4.2.1 Aquatic Insects in a Natural Stream, the Breitenbach ................. 117
7.4.2.2 Anthropogenically Altered Streams ............................................. 120
7.4.3
Bioindication ............................................................................... 121
7.5 Assessment of Model Quality and Visualisation Possibilities:
Hybrid Networks ................................................................................. 122
7.6 Conclusions ......................................................................................... 123
Acknowledgements ............................................................................. 125
References ........................................................................................... 125
8. Non-linear
Approach
to
Grouping,
Dynamics
and
Organizational Informatics of
Benthic Macroinvertebrate
Communities in Streams by Artificial Neural Networks ....... 127
8.1 Introduction ......................................................................................... 127
8.2 Grouping Through Self-Organization ................................................. 130
8.2.1 Static Grouping ................................................................................. 130
8.2.2 Grouping Community Changes ......................................................... 143
8.3 Prediction ofCommunity Changes ..................................................... 147
8.3.1 Multilayer Perceptron with Time Delay ............................................ 147