BIOGEOGRAPHICAL BOUXDARIES 221 



tion of pelagic fish and lobsters in New England. Crisp (1958) has 

 shown that the spread of the European distribution of the barnacle 

 Elminius modestiis may be partly explained by transport on the 

 hulls of ships; and that Balanus amphitrite has become established 

 near the warm effluents of electricity power stations in Britain 

 (Crisp and Southward, 1959). 



I have, therefore, done no more than provide examples of dif- 

 ferent aspects of the shape of distribution and abundance in time 

 and in space without any reference to environmental factors which 

 may be related to these distributions. The point I want to make 

 is this: that the ecologist who wishes to extend his work beyond 

 the limits of local exploration and who aspires to study fundamental 

 principles of general importance will need quantitative estimates 

 of abundance, distribution, and species composition of the kind 

 which I have shown. The fact that my examples were taken from 

 the plankton of a limited area is irrelevant. The same require- 

 ments will apply in studies of the seashore or the bottom of the 

 ocean. 



We are all aware of the limitations of isolated cruises but most 

 marine biogeography is based on the findings of surveys or expe- 

 ditions which were isolated in time, space, and technique. At the 

 symposium on "Perspectives in Marine Biology" held at Scripps 

 in 1956, there was an impressive number of pleas for repetition 

 of sampling of coral reefs, on the shore and in the open sea. 



I suggested earlier that repeated surveys using standardized 

 techniques are necessary. I mean "standardized" within one pro- 

 gram and throughout its duration. I would not advocate uni\-ersal 

 standardization of sampling methods, although this may be possi- 

 ble in some fields, and it has already been achieved in the physical 

 and chemical recording of such factors as salinity, temperature, 

 and oxygen. 



Sampling methods will obviously depend on the nature, size, 

 and abundance of the organisms that are to be captured. Most 

 studies of the dynamics of distribution will be served better by 

 many, frequent, small samples than by a smaller number of large 

 ones. Some research will be needed in each project to discover the 

 size of the smallest sample which will yield useful information. 



