Chapter 48 



Regional Variability of Bacteria in 

 North Atlantic Sediments 



E. H. Anthony 

 INTRODUCTION 



T 



he accuracy with which bacteria in lake sediments may be 

 counted was discussed in a previous paper (4). The errors ex- 

 hibited by agar plate and membrane filter counts of bacteria in 

 309 sediment samples from the depths of sixteen lakes were 

 analysed and the following conclusion drawn. If four trips are 

 made to a lake, preferably over several seasons, and if upon each 

 trip four cores of sediment are taken, and if from each of these 

 cores quadruplicate membrane filter counts are made, then the 

 mean number of bacteria in the sediment may be estimated with 

 a standard percentage error of about ± 36 per cent. The ob- 

 jectives of the present study were to assess the precision with 

 which bacteria in marine sediments can be counted by the filter 

 method, and to determine whether sediments from different 

 regions can be distinguished from one another on the basis of 

 such counts. In addition, filter counts have been compared with 

 direct counts by fluorescence microscopy, and the possibilit>^ of 

 freezing marine sediments without changing their bacterial num- 

 bers has been examined. 



PROCEDURE 



Coastal sediments were obtained with a Naumann sampler; 

 Labrador and Scotian shelf sediments with a Phleger corer. These 

 devices were lined with plastic tubing which also served to trans- 

 port the coastal sediments to the laboratory. Supernatant water 

 was removed by aspiration as the core was pushed to the top of 

 the tube by a plunger. A bacterial sample was obtained by thrust- 



522 



