THE DISTRIBUTION OF BACTERIA IN LAKES 



43 



undertaken, siiK-c tlicy will I'nniisli a guide 

 as to where and when to make collections. 

 It should be kept in mind, however, that 

 such purely ([uantitative studies furnish 

 only an imperfect and incomplete picture 

 of the lake bacteria. For instance, there 

 are more periphytic bacteria active in the 

 bottom meter of Lake Mendota in the 

 winter, the temperature 4° C, than at any 

 other level at any season. These must be 

 a particular group of cold-loving bacteria 

 that find optimum conditions at the bottom 

 of the lake Avhen it is frozen. AVliat species, 

 we do not know. 



The quantitative distribution of bacteria 

 in lakes has been studied by the use of four 

 distinct techniques. The one most com- 

 monly used is the method used in all 

 branches of bacteriology, that of plate 

 counting. A sample of water, properly 

 diluted, is mixed with melted nutrient agar, 

 poured into a Petri dish, incubated, and 

 the number of colonies counted. This 

 method has a serious disadvantage in that 

 no single medium will alloAV the growth of 

 all the cultivable bacteria, and some of the 

 water bacteria will not grow in any 

 medium. For the most part, only organic 

 media have been used, and only hetero- 

 trophic bacteria have been cultivated. 

 Earlier studies were conducted primarily 

 from the standpoint of sanitation, and the 

 medium used, standard beef infusion agar, 

 was more suitable for the cultivation of 

 parasitic bacteria. Such studies have not 

 yielded results that are significant from the 

 standpoint of hydrobiology. With the in- 

 troduction of media more suitable for work 

 on lake bacteria, particularly the dilute 

 sodium caseinate-glucose medium of Fred 

 and his associates, more significant results 

 were obtained. 



I have used a medium similar to that 

 mentioned, but modified to include a wider 

 variety of organic materials. It is com- 

 posed of sodium caseinate, peptone, gly- 

 cerol, starch, and dibasic potassium phos- 

 phate, 0.05 per cent of each in tap water, 

 with 1.5 per cent agar. This dilute medium 

 has given higher counts than the caseinate- 

 glucose medium. I began my work on lake 

 bacteria with a prejudice against the plate 



connt. As I have found the results of such 

 counts to correlate with the results of 

 microscopic counts and with other data, I 

 have been forced to revise my opinion. 

 The plate count method may be developed 

 into a very useful one with further study. 



Plate cultures must be prepared as soon 

 as possible after the sample has been col- 

 lected, since not only the total number but 

 also the relative proportions of different 

 kinds of bacteria will alter markedly if the 

 sample is stored. To be accurate, plate 

 counts must be made in replicates of at 

 least five for each dilution. The prepara- 

 tion of plates requires large amounts of 

 glassware and other equipment not easily 

 provided in a field laboratory. This has 

 been perhaps one of the most important 

 factors in limiting the study of lake bac- 

 teria. The author has found that plating 

 under field conditions can be greatly simpli- 

 fied by the use of flat-sided bottles instead 

 of the customary Petri dishes. Lotion 

 bottles of 120 cc capacity fitted with screw 

 caps are used. The agar medium is placed 

 in these bottles in 30 cc amounts in the home 

 laboratory, where they are sterilized. In 

 the field the agar can be melted by placing 

 the bottles in a pan of boiling water. 



The published literature on plate counts 

 of lake bacteria from a hydrobiological 

 standpoint has been extensively reviewed 

 by Baier (1935). 



The distribution of lake bacteria may 

 be further studied by the use of selec- 

 tive media. Thus one may prepare a solu- 

 tion containing no source of nitrogen. In 

 such a medium only those bacteria which 

 can use atmospheric nitrogen may grow. 

 Merely inoculating a tube of this medium 

 with some lake water Avill show whether 

 nitrogen-fixing bacteria are present or ab- 

 sent. By preparing a series of dilutions of 

 the water, and inoculating a number of 

 replicate tubes from each dilution, one may 

 estimate the numbers of bacteria belonging 

 to a particular physiological group in the 

 sample. By the use of a variety of selective 

 media, all sorts of physiological groups may 

 be determined. 



Selective media were used by Domogalla, 

 Fred, and Peterson (1926) to study the 



