projects. Quantitative research on entire ecosystems must be based on 

 the observed response of large landscape units exposed to 

 manipulation whether the manipulation is planned or not. The 

 experimental unit may consist of an entire watershed. By measuring 

 the ecological responses of the watershed to inputs from disturbances 

 caused by man, empirical data can be collected and mathematical 

 models can be developed and tested for their predictive abiUty and 

 management value. 



Availability of landscape units of suitable size for experimentation is 

 central to holistic ecosystem experiments. Some of these experimental 

 ecosystems should be in locations where they will be only sHghtly 

 altered by inadvertent release of pollutants or by uncontrolled human 

 activity. The experimental area should be of sufficient size so materials 

 such as nutrients, radionuclides, or heavy metals which cycle can do so 

 primarily within the ecosystem. In addition, these ecosystems should 

 be representative of much larger units so information gained may be 

 transferred with confidence throughout the region. Research 

 techniques need to be developed to indicate the reliability of 

 extrapolated data. 



Recent findings have shown that holistic models of large systems 

 based on scientific truths are difficult to assemble and will take years to 

 construct. Models which show the relationships of various parts of the 

 ecosystem and are not truly mathematical in flow relationships have 

 more immediate value. It is, therefore, necessary to proceed at once 

 with subsystem models that can be constructed with reasonable 

 accuracy. Since so many of man's activities alter ecological systems, 

 research is needed which emphasizes recovery rates or shows how 

 changed conditions can be modified to support more useful ecosystems. 

 One of our great needs is to learn how to rehabilitate damaged 

 ecosystems, such as strip mined lands or aquatic areas, or to design 

 ecosystem uses so the damage can be mitigated. This type of research 

 would be valuable in planning physical or chemical changes in 

 ecosystems whether surface waters, landscapes, or urban and 

 industrial sites. 



34 



