concentration is multiplied by an application factor (a factor which estimates sensi- 

 tivities not measured in the critical concentration). 



There are a number of drawbacks to this assumption. It is by no means certain that 

 the most sensitive of the relatively small array of species exposed in the "real world" 

 will be represented among the tested in the laboratory (see Mount and Gillett, this 

 Monograph). In fact, the most sensitive species quite likely cannot be cultured in the 

 laboratory. Furthermore single species tests cannot, by definition, include such 

 factors as predation. competition, and other relationships which are so important in 

 the functioning of communities and ecosystems. Neither can these tests determine 

 effects on nutrient and energy transfer and other relationships equally vital to the 

 functioning of communities and ecosystems. Finally, single species laboratory tests 

 cannot accurately mimic the environmental partitioning, the transformations, and 

 other processes and changes that characterize a chemical's movement fate in a 

 natural ecosystem. 



Until these important factors are considered, an accurate estimate of hazard from 

 the use of a particular chemical is not likely.-'' 



INAPPROPRIATE INTERVENTION 



Many ecosystems go through periodic ecological upheavals as a consequence of 

 fires, floods, and other natural events. Often the fairly regular frequency of these 

 events causes ecosystems to depend upon them. Vogl-*'--' and others have termed 

 these "perturbation-dependent ecosystems." The differences between perturbation- 

 independent systems with regard to disturbance is illustrated in Figure 4 from Vogl.-' 

 Sometimes when the perturbation is thought to be caused by disease or insect pests, 

 ecologists are tempted to intervene and "restore" the system to its "original" 

 condition. 



A situation in which such a temptation was successfully avoided is described by 

 Mueller-Dombois.'** It involved the 'ohi'a dieback phenomenon in the Hawaiian 

 rainforest. This was first reported by Mueller-Dombois and Krajina-'' about a 

 decade ago on the island of Hawaii. The decline was described as a "severe epi- 

 demic. "and a prediction was made that the native rainforest would be eliminated in 

 15-25 years if the rate of damage had continued. This prediction implied that the 

 native rainforest had been stricken by a newly introduced disease. But intensive 

 disease research was begun in 1972 by the U.S. Forest Service and simultaneously by 

 Mueller-Dombois with the assumption that the dieback might be a recurring natural 

 phenomenon in primary succession. Mueller-Dombois-** concluded that the dieback 

 was not due to a newly introduced insect/ pest or disease-causing organism but rather 

 to climatic instability. Fortunately in this case massive intervention did not occur. 



However, this particular case history and a number of others involving floods and 

 fires indicate that intervention was inappropriate and would probably actually 

 endanger the survival of certain species in perturbation-dependent ecosystems. EPA 

 and other regulatory agencies would be well advised to take note of such case 

 histories and be certain that intervention to restore an ecosystem to its original 

 condition is appropriate and will arrest degradation resulting from man-induced 

 deterioration rather than natural cyclic phenomena. Such long-term studies as those 

 found in the Woods Hole Conference,'" Loucks," Botkin,'-Toyryla,-'' Bormann and 

 Likens,''^ and Edmondson-'-'* will provide the type of information needed to determine 

 when intervention is required. 



ECOLOGICAL REFUGES 



The prospects for natural recolonization of damaged ecosystems and for rapid 

 return to original conditions (or to a condition which included many of the original 

 qualities) will be enhanced if invading organisms can reach damaged areas from 

 nearby ecological refuges. These refuges not only protect the basic genetic endow- 



228 



