efficient way to extract resources and the quantification of economic losses 

 resulting from ecological impacts thus become very relevant for modelers to 

 consider. The socio-economic viewpoint permits a critical examination of the 

 economic value of specific components of the ecosystem. In many cases ecosystems 

 and ecosystem components are valued differently by different people for different 

 reasons, uses, or purposes. Often the perceived values are based on both 

 qualitative and quantitative criteria. Thus, the modeler must be willing to 

 incorporate qualitative factors into the conceptual model and quantify them in 

 the mathematical model. This need to recognize and quantify social and economic 

 factors within the ecosystem model requires that the modeling effort be structured 

 from a total system perspective of which the ecology/biology is only one portion 

 of the input and output data/information bases. 



Delineation and definition of the spatial and temporal characteristics of 

 an ecosystem model must be compatible with and relevant to the decisionmaking 

 needs of "political" and resource management units. This point cannot be 

 stressed enough. Spatial characteristics must be pertinent to considerations 

 of boundaries of jurisdiction and "area of impact" concerns. Similarly, temporal 

 characteristics must be pertinent to the timing with which relevant decisions 

 will be made. Ecosystem models structured on mesoscale magnitudes will not be 

 applicable to small scale management needs. 



Ecosystem models are often quite complex, and the addition of variables 

 inherent to other systems might make models more complex and perhaps unmanageable. 

 The Panel stressed and gave considerable emphasis to the notion that modelers 

 need to tailor model complexity to the specific problem at hand. This should 

 be accomplished through a systematic examination of all the variables and a 

 subsequent elimination of those that seem inappropriate. If this is not done, 

 inappropriate variables may be given undue attention at too great a cost, not 

 only in time, effort, and money, but also in terms of the decisionmaking value 

 and usefulness of the model's output data. Tailoring model complexity to the 

 problem at hand will require close consultation between modelers and managers 

 (model users) before the model is developed. 



The Panel expressed a concern for the amount of effort that should be 

 expended on ecosystem models - are they worthwhile exercises, and how much 

 should be expended on their development and implementation? It is difficult to 

 answer this question in abstract without a particular well-defined problem or 

 set of socio-economic questions. The Panel concurred that any modeling effort 

 will have some value to decisionmakers, given that: 



1) the needs of the decisionmakers have been anticipated; 



2) the model is relevant to the needs of the decisionmaker; 



3) the model has some predictive value; 



4) the model focuses on critical decision outcomes; 



5) the model gives results with an acceptable level of variability; and, 



6) the modeling effort in the context of the above criteria is cost 

 effective. 



266 



