producers, consumers and storage 

 compartments . 



A conceptual model of this scope 

 is complex, even in this summary 

 form. Perturbation of the system by 

 a stress, such as low flow, induces 

 many concurrent changes in com- 

 partments, functions of organisms and 

 trophic flows. The human mind has 

 difficulty dealing with such simul- 

 taneous changes and their ramifica- 

 tions. Computers form a powerful 

 tool for dealing with such systems. 

 Therefore a simulation program 

 (designated Chesapeake Bay Ecosystem 

 Model -CBEM) was created to assist in 

 dealing with secondary effects. This 

 model supplements the conceptual 

 model and provides a tool to analyze 

 the sensitivity of particular spe- 

 cies or compartments to low flow ef- 

 fects . 



CBEM has been initially defined 

 for the lower mesohaline Venice 

 zones, using data from the Patuxent 

 River, one of the western shore 

 tributaries of the bay. Under 

 average flow conditions, the species 

 and compartments utilized are shown 

 in Figure 10. The system includes 

 phytoplankton as the major producers, 

 grazed by two competitive species 

 of copepod zooplankters . These are 

 grazed in turn by ctenophores during 

 particular time periods. Other feed- 

 ers include benthic grazers (oys- 

 ters), juvenile fish and menhaden. 



The model is based on sets of 

 quasi-linear differential equations 

 which are periodically corrected to 

 compensate for non-linear behavior in 

 the ecosystem. The model operates by 

 calculating daily productivities and 

 abundances based on changing physical 

 and chemical conditions which can be 

 either input as data or calculated as 

 the program evolves in time. Typical 

 simulated timeframes range from 1 to 

 5 years for a run. Thus the effects 



of a low flow period can be tracked 

 for long-term biological changes 

 which occur after the system has 

 returned to a physical or chemical 

 equilibrium. 



The model was calibrated with 

 data on known responses and growth 

 rates of organisms, predominantly in 

 the Chesapeake Bay; however, data 

 from other estuaries were used where 

 Chesapeake data did not exist. Tests 

 of the model's validity under average 

 flow conditions were then made, using 

 existing historical data from the 

 Patuxent River, which had not been 

 used to create or calibrate the 

 model. An example appears in Figure 

 11 for phytoplankton abundance. The 

 figure shows that, in general, simu- 

 lated phytoplankton abundance agrees 

 well with observed data, although 

 there is a fall bloom in the simu- 

 lation which is not typically mani- 

 fested in the data. Knowledge of 

 such discrepancies prompt investi- 

 gation of biological properties of 

 the ecosystem which might supress or 

 mask such a plankton bloom. 



The effects of predation, and 

 the usefulness of CBEM as a tool to 

 study predation effects are illus- 

 trated in Figure 12. The changes may 

 differentially affect several other 

 species in the food web, and often 

 are manifested quite differently at 

 times throughout the year as popu- 

 lation abundances, growth potential, 

 and feeding behavior are altered. 



The main usefulness of both 

 conceptual and mathematical models is 

 to study effects of low flows, parti- 

 cularly those associated with sa- 

 linity changes. Average annual 

 stream flow into Chesapeake Bay has 

 historically ranged from lows near 

 50,000 cfs to highs near 150,000 cfs, 

 a dramatic range of values, parti- 

 cularly as it impacts salinity and 

 water quality. Figure 13 shows two 



141 



