16 



J. E. Hobbie 



Even in the pond ecosystem, the number of interactions is too great 

 and most cannot be included in any model. We decided what to include on 

 the basis of the carbon flow results (Figure 1-5). The modeling helped by 

 forcing all the workers to sample just one pond and by forcing us to work 

 on some processes which were virtually unknown (e.g., the detritus food 

 web). It quickly became obvious that we simply did not understand many 

 of the controls of the pond ecosystem. Despite this, the deterministic 

 model which resulted did an excellent job of mimicking or simulating the 

 annual cycles of carbon flow through the plankton and benthic systems. 

 For example, the submodel of benthic algal photosynthesis (Figure 1-11) 

 was developed with physiological data from 1973 but the simulation fits 

 the 1971 and 1972 field data very well. Unfortunately, there were several 

 assumptions that had to be included, in spite of our attempts to measure 

 every parameter, and the results of the calculations were very sensitive to 



J L 



I I I I I I I I I L 



FIGURE 1-11. Model simulation (solid 

 line) and measured estimates (circles) of the 

 photosynthesis of epipelic algae (mg C m'^ 

 hr'^). (Redrawn from Stanley 1974.) 



