SECT. 2] 



PKODUCTIVITY, DEFINITION AND MEASUREMENT 



147 



In experiments of short duration, dark fixation is usually of minor im- 

 portance only. In experiments of long duration, however, a very pronounced 

 growth of bacteria takes place. It is a well known fact, as already mentioned 

 on page 140, that a huge growth of bacteria starts as soon as sea- water is 

 enclosed in bottles. Jones et al. (1958) showed that after 24 h the number of 

 bacteria had increased by a factor of nearly 500. In oligotrophic water with few 

 algae, the dark fixation by the bacteria may, therefore, easily exceed the 

 fixation due to photosynthesis if the duration of the experiment is long. It 

 cannot be repeated often enough that the enclosure of a water sample in a 

 bottle is a severe interference. The duration of experiments must be kept short ! 



0) o 



o> '— 

 o o 



d JO 



X ■*- 



2 4 6 



mgC/m 3 /h at light saturation 



Fig. 8. Rate of dark fixation of 14 C as a function of the photosynthetic rate at light 

 saturation. Summer plankton from the North Atlantic ; filled circles from the surface, 

 open circles from the lower boundary of the photic zone. (After Steemann Nielsen.) 



In agreement with Anderson and Libby (1950), Steemann Nielsen (1955) 

 showed that 14 C02 is assimilated about 5% slower than is 12 C02- This estimate 

 was based partly on experiments made by van Norman and Brown (1952) and 

 partly on the author's own experiments. Not all workers have taken the in- 

 fluence of isotope discrimination into consideration when employing 14 C for 

 measuring primary production. Although the correction is small it should be 

 taken into consideration. 



It is likely that small amounts of assimilates are lost in most cases either by 

 excretion during the experiments or owing to damage during the final filtration 

 of the water sample. The experiments presented by Fogg (1958) seem to 

 indicate that such losses are usually of no great importance. 



However, in some cases the loss may be considerable (cf. McAllister et al., 

 1961). As shown by Lasker and Holmes (1957) it is due to the fragility of the 

 labelled cells. Many naked flagellates are destroyed beyond recognition on the 



