PART IX — TERRESTRIAL ECOSYSTEMS 



Figure IX-3 — MOSQUITO SUBMODEL 



WATER 



WATER 

 VOLUME 



5 



EVAP 



TEMP 



y 



HIBER- 

 NATING 

 FEMALES 



^ 



* 



EGGS 



M 



LARVAE 



-K 



Y, 



U 



PUPAE 



-DC 



^f 



ADULTS 



V, 



MORTALITY 



The figure is a submodel, or subsystem, of the larger desert ecosystem model. This 

 particular submodel is designed to elucidate how water level and temperature affect 

 the production of mosquitos in desert playas. Note that water volume is important 

 initially to the female and the laying of the eggs whereas temperature is important 

 throughout all stages in the mosquito's life; temperature is also important to the 

 effectiveness of the water volume. 



proach; but these data will them- 

 selves provide no validation of the 

 model. In general, however, the de- 

 velopment of separate submodels as 

 a first step is likely to lead to a more 

 sophisticated total model, with greater 

 variety in its content, than is likely 

 to be attained by using trial and error 

 to modify a complete model without 

 subdivisions. On the other hand, the 

 possibility that important elements 

 of interaction between submodels de- 

 veloped separately may affect their 

 behavior when brought together is 

 an ever present danger with this ap- 

 proach, and must be examined by 

 validation comparisons of model be- 

 havior with that of the ecosystem as a 

 whole. 



Data Base — Both approaches to 

 ecosystem modeling, and modifica- 

 tions of them, have been explored in 

 recent years, and various simple mod- 



els have been developed for aquatic 

 and terrestrial systems. Only excep- 

 tionally, however, have the data been 

 sufficient even for the construction 

 of a model, let alone for testing it. 

 The reason is that the data have 

 usually been collected earlier, perhaps 

 for a different purpose, and without 

 reference to the particular type of 

 model that was being built. Even 

 where data were collected with model- 

 ing requirements in mind, the de- 

 velopment of the model has often 

 indicated the need for data additional 

 to those already collected. 



Standard Models — The problem 

 of modeling does not need to be 

 tackled afresh and independently for 

 each new ecosystem that comes under 

 scrutiny. There is a great deal in 

 common in the general structure of 

 relationships within different terres- 

 trial ecosystems, even as diverse as 



tundra and tropical forest, though 

 none of the species are the same and 

 the balance of the various life forms 

 and processes is quite different. Even 

 between aquatic and terrestrial sys- 

 tems, there are numerous analogies. 

 Consequently, it may be expected 

 that experience in modeling one type 

 of ecosystem will greatly simplify the 

 problems when a new type of eco- 

 system is considered, though all pa- 

 rameters may have to be estimated 

 afresh. 



Moreover, the value of model de- 

 velopment is not limited by national 

 frontiers. Where the same landforms 

 and biota occur on both sides of a 

 frontier the same models can be ex- 

 pected to represent the ecosystems 

 there, so that a model for the Sonoran 

 desert in Arizona should also apply 

 to the Sonoran desert in Mexico. 

 Even where different biota are in- 

 volved in different countries, the gen- 

 eral ecosystem structure as repre- 

 sented in the model will often be the 

 same, and only the constants and data 

 used may need to be changed in order 

 that the same models should be 

 applicable. 



Terrestrial Systems — In general, 

 the more successful models have been 

 concerned with aquatic ecosystems; 

 they are simpler, with fewer com- 

 ponents, and limnologists are more 

 accustomed to recording a wide range 

 of data than are terrestrial ecolo- 

 gists. Few terrestrial models cover 

 more than a limited selection of eco- 

 system components. In the arid 

 lands, particularly, it is not possible 

 to point to any complete ecosystem 

 model based on well-authenticated 

 data. 



The paucity of models for com- 

 plete terrestrial systems does not 

 indicate a similar lack for subsystems. 

 Certain parts of terrestrial systems 

 have been the subject of considerable 

 modeling activity. Some aspects of 

 meteorology, for instance, are well 

 served in this way, as is hydrology. 

 There are models for soil nitrogen 

 cycling, for photosynthesis and plant 



282 



