177 



increase turbidity in coastal waters, which will most likely reduce 

 seagrass and phytoplankton production and may foul the filtering mechanisms 

 of filter feeders such as clams and oysters. 



11.3 RESEARCH NEEDS 



Prediction of the effects of sea level rise on the future extent of 

 the various coastal ecosystems would be greatly enhanced by the development 

 of a model that incorporates the determinants of the vertical and 

 horizontal growth components of submerged, intertidal, and supratidal 

 coastal ecosystems. Furthermore, to understand the human consequences of 

 changes in extent of each ecosystem, a quantitative understanding of the 

 value of these systems to humans is essential. Energy analysis and natural 

 resource economics techniques (e.g., Odum et al., 1987) may hold promise in 

 this regard if specific fish, wildlife, ecological diversity, and aesthetic 

 values can be included. 



Synthesis of existing information on growth determinants and 

 quantative ecosystem values is the first step. The responses of some of 

 the more prevalent coastal plants and animals to some environmental changes 

 are known. As predictions improve the effects of sea level rise on 

 ecologically important physical variables, better predictions of the nature 

 and timing of ecological changes can be made using existing information. 

 For the best predictions, however, new ecological and physiological 

 information may also be required. The most important information needs can 

 be identified with the aid of a literature synthesis and subsequent 

 simulation model. The model will simulate ecological responses to 

 predicted environmental changes (e.g., rising water level, encroaching 

 salinity, increased tidal range and wave energy). Estimates of functions 

 and parameter values will be based on the best available information, but 

 some guesswork is anticipated. Uncertainty in ecological generalizations 

 can be explored by a sensitivity analysis of the model. Needed information 

 can be ranked according to a combination of the uncertainty involved in an 

 estimate and the sensitivity of the model to changes in the estimate. 

 Thus, not only can the most important information needs be identified, but 

 also the consequences of a lack of this knowledge can be demonstrated with 

 the model (Montague et al . , 1982). 



