Experimental development of nutritive media for micro-organisms

Abstract

When a nutritive medium is made up of many factors, the combined influence of the levels of the factors on the yield of the culture may generally be expressed in two rules: (1) the yield may be limited by the concentration of one of the factors, provided that the rest are in the indifference region; (2) when all the factors are in the indifference region the yield is limited by the least favourable combination of interacting factors. The second rule results from interactions between the factors which are in turn due to correlated processes involved in the utilization of the different factors by the organism. If there were no such interaction it would be possible to select the optimum concentration of each factor independently of the levels of the other factors, so that only the first rule above (originally formulated by Liebig) would be called into play. That this is an oversimplification is clear when it is realized that factors which by themselves do not influence yield are capable of affecting it through interaction with other factors. In designing an optimum nutritive medium two problems must be solved: (1) among the totality of possible factors, those must be selected which are capable of affecting the yield; (2) optimum levels of the selected factors and of factors interacting with them must be established. We have studied this problem in deriving a nutritive medium for the green sulphur bacterium, Chlorobium thiosulfatophilum. For the first step we have used the method of random balance (Budne, 1959). This not only allows us to select the important factors, but also provides an approximate estimate of the gradient of the response surface, i.e. the direction in which the set of optimum levels lies.