532 CYCLES OF ORGANIC AND INORGANIC SUBSTANCES 



the same dimension. For the very simple calculations made here, 

 this is close enough and on the right side of the minimum, and 

 suggests that an approach of this type gives an indication of the 

 expected dimensions of patches. 



It implies that in inshore areas, where the production rate is 

 often high, the marked degree of patchiness may be due partly 

 to the above reasons as well as to the "patchiness" of the hydro- 

 graphic features. To be even more hypothetical, it may also be 

 that zooplankton patchiness of such organisms as Calanus can 

 provide a means for grazing on higher than average plant concen- 

 trations and thus permitting rather lower numbers of animals to 

 have a higher feeding rate than they could ha\'e if grazing were 

 evenly distributed. The importance of this, through the depend- 

 ence of egg laying on food supply, has been shown by Marshall and 

 Orr (1955). On the other hand, for areas with low productive rates, 

 the patches could have only very small gradients and so be situated 

 at large distances apart. As these dimensions increased, it is possi- 

 ble that the advantages of patchiness in animals would decrease 

 on account of the large horizontal distances involved so that in 

 very poor waters patchiness would disappear altogether, as seems 

 to be the case in unproducti\'e tropical areas. 



In this way, I have tried to show roughly how the time and 

 length dimensions of plankton changes may depend on production 

 rate per unit of populations. For temperate waters in particular it 

 suggests that the grids of stations often used, and repeated at 

 monthly or even weekly intervals, can be expected to show con- 

 siderable apparently random features. On this basis, samples on 

 successive days, or at distances of 10 km, may be required to 

 show up the detailed patterns. 



A consequence of this description of patchiness is its relation 

 to pelagic fish, such as herring, which feed on the zooplankton. 

 Since the small-scale relations between fish and zooplankton may 

 be similar to those between the animals and the plants [with the 

 same processes of aggregation, grazing, and disengagement as 

 Gushing (1955) and Alanteufel (1941) have shown for herring], 

 conventional sampling again can be expected to produce the same 

 apparent lack of any real relation. It seems likely that the time 



