PRIMARY PRODUCTION 521 



limitation, and, especially, the effect of respired carbon. For the 

 last of these, Steemann Nielsen (1955) and Ryther (1956) in 

 rather similar experiments but with different organisms reached 

 different conclusions. Ryther's results imply that somewhere within 

 the respiratory cycle all the respired carbon is reassimilated by 

 phytosynthesis. while Steemann-Nielsen found an interaction 

 amounting to only 50 to 70% of the respiratory rate. Because of 

 such possible disagreement it may be that the correct interpreta- 

 tion of C^"* values will not be expressible in one simple lormula 

 as we had hoped. 



Another example of the type of complexities which have arisen 

 comes from the data on nutrients. In addition to phosphate, ni- 

 trate and silicate are now measured regularly (Mullin and Riley, 

 1955; Armstrong, 1951), and variations in the relative proportions 

 of these nutrients are found to be more widespread than had 

 perhaps been thought (Ketchum et al., 1958). This makes it more 

 difficult to think in terms of a single index for the effects of nutri- 

 ent deficiency on a plant population, yet this index has been one 

 of the main features of the early hypothetical systems to explain 

 plant changes (Steele, 1959). 



However, the most difificult sampling problem is probably that 

 raised by the zooplankton. With the plants, filtered water samples 

 catch effectively all the organisms, and chlorophyll extraction 

 measures a common factor which is directly connected with their 

 growth. For animals the displacement volume, or the dry weight, 

 of net hauls has been used to provide single values which can be 

 converted to carbon content for comparison. But no single mesh 

 size or speed of hauling of a net can catch all the animals, nor do 

 the values obtained differentiate between herbivores and carni- 

 vores. Thus the figures are rather poor indices of the possible ef- 

 fects of grazing on the phytoplankton and this is one of the main 

 factors we need to know. This inadequacy is perhaps the main 

 difficulty in sampling for quantitative information about produc- 

 tion cycles. 



Apart from the C^^ technique there are three other main methods 

 that have been used to estimate production. 



One of the first methods (Gaarder and Gran, 1927) was to find 



