The complexity of an ecosystem is only seen in parts and its interrelationships are never well understood. Ecological indicators are able to represent, integrate and characterize information embodied in comprehensive data sets when direct measurements cannot be realized. Ecological indicator system (EIS) is normally made up of a number of ecological indicators with certain structure to represent the most important features of the environmental state, the complex ensembles of ecosystem elements and the multiple webs of actions, reactions and interactions.
As an effective tool or approach, the ecological indicator system (EIS) is widely used in ecosystem monitoring, assessment and management. While there are many different forms of EIS based on the different purposes for EIS uses, they are all concerned with the following two key steps: structure of the integrated EIS frame and the assignment of rational weights to the ecological indicators. The frame of the EIS should integrally reflect not only the ecological hierarchy, but also the relationships among the complex ensembles of the investigated ecosystem. The status or the importance of an indicator in the EIS will be represented by its weight, which, to some extent, will be determined by the frame of an EIS.
This paper reviewed the branch shape ecological indicator system (BEIS) which is in common use at present, and point out its deficiency: (1) the overlapping or intersecting connections of the indicators are often overlooked or simplified. (2) Incorrect weights are assigned to the indicators because of disintegrated representation of the ecosystem complexity. Secondly, a net shape ecological indicator system (NEIS) was presented, including regular NEIS and irregular NEIS. It is developed by adding the overlapping and intersecting relation of the ecological indicators to the BEIS hierarchy frame and thus will better represent the integration and hierarchy of ecosystem. Then a twice assigning weight method (TAW) was demonstrated in both regular NEIS and irregular NEIS. This was used to resolve the problem of assigning the rational weights to NEIS indicators. For further study and comparison of BEIS and NEIS, two indicator systems for ecosystem health assessment were derived from hierarchical ecosystem indicators respectively, using the frames of BEIS and NEIS. The weights of the indicators in BEIS were determined by the AHP and those in NEIS were determined by TAW. It was easy to see the changes of the weights of the indicators in between the BEIS and NEIS. The indicators′ weights in NEIS have a significant trend to the average compared with those in BEIS. Two problems about the NEIS design and application were discussed.