SECT. 4] COMMUNITIES OF ORGANISMS 431 



over if one looks only at large units (trophic levels). For example. Teal (1957), 

 in his study of a temperate cold spring containing a very limited community, 

 used the spring as a thermostat when measuring the respiration of the animals. 

 He thus avoided uncertainties in correcting respiration for temperature, but, 

 as the respiratory quotient was not known, used an average oxycalorific 

 coefficient to convert to calories and assumed a linear relation between weight 

 and respiration whereas most studies have suggested a non-linear relation 

 (Bertalanffy, 1957). He obtained an indirect measure of mortality in the field 

 by measuring growth under constant conditions in the laboratory and assuming 

 a reasonable, but unproven, relation between population weight, growth rate 

 and respiratory rate. Using this estimated mortality rate, he corrected for 

 losses occurring between sampling periods. In some cases the caloric content 

 of a species was calculated from an analysis of a more or less closely related 

 species taken from the literature. He took multiple random samples so that he 

 could calculate confidence limits for the mean number of each species but only 

 presented these fully for one species, the fourth in importance in terms of 

 energy flow. Because the animals were aggregated, 90% confidence limits for 

 the mean often covered a range of one-half to twice the mean value. Therefore, 

 energy flow based on mean values of assimilation per individual and respira- 

 tion per individual could easily be in error by these factors. This makes the 

 perfect agreement between credit and debit columns in his community energy- 

 balance chart a bit difficult to believe unless the figure for deposition, 28% of 

 the debit column, was added to achieve the balance. 



None of the preceding should be taken as a condemnation of Teal's work. It 

 is presented only to outline some of the problems involved and the numerous 

 assumptions and simplifications, made without rigorous justification, required 

 in striking an energy balance in even a relatively simple community. In terms 

 of replicate determinations and attention to sampling theory, population 

 changes, feeding habits, etc., his study of community productivity is one of the 

 most complete and careful yet done and should be looked at by everyone 

 contemplating such a study. By comparison, some marine community pro- 

 ductivity studies (Harvey, 1950 ; Odum and Odum, 1955), although admittedly 

 done on much more complex communities, leave a great deal to be desired. 



Teal's results emphasize the necessity for work continuing over at least a 

 year; despite the fact that the spring he studied was almost constant chemically 

 and thermally, the species which used most of the energy available during the 

 summer was unimportant during the winter and over the year was only third 

 in terms of energy flow. 



Even though phytoplankton productivity is discussed in other chapters in 

 this volume (Steemann Nielsen, Chapter 7, page 129; Ryther, Chapter 17, 

 page 347), because of its primary importance as a source of energy in marine 

 communities it seems appropriate to mention briefly some of the limitations of 

 the methods now used for its estimation (see the review by Steele, 1959, for 

 documentation). A combination of chlorophyll, light intensity at the surface 

 and the extinction coefficient has been used to estimate photosynthesis which 



