Effect of media composition on yield values of bacteria growing on binary and ternary substrate mixtures in continuous culture

Pseudomonas aeruginosa 142 and a presumed variant were grown axenically in chemostats on salicylate/benzoate or salicylate/glucose binary feeds. Each substrate was supplied at 2, 10, 50, 90, 98, or 100% of the total energy fl ux. Two experiments were also run with ternary mixtures using the same subs trates. Aliquots were transferred to fed-batch reactors receiving the same substrates at the same specific rates as the chemostat, but with one substr ate radiolabeled with C-14. Radiolabel incorporated into biomass, (CO2)-C-1 4, and soluble microbial products over a period of 8 minutes was used to es tablish the biomass yield, CO2 yield, and product yield, respectively, asso ciated with a given substrate. The effect of the percent substrate in the f eed on the yields depended on the pair of substrates supplied. When benzoat e comprised 50% or more of the applied substrate in salicylate/benzoate fee ds, the fraction of benzoate in the feed had a small effect on the yield va lues associated with benzoate. However, when benzoate constituted 2% or 10% of the feed, CO2 yields were lower, biomass yields were slightly lower, an d product yields were higher. In contrast, the percent of salicylate in the feed had little effect on any of the salicylate yields for cells growing o n the salicylate/benzoate feeds. When salicylate was mixed with glucose, th e yields associated with salicylate behaved quite differently. Biomass and CO2 yields were lower and product yields higher when salicylate was 2% or 1 0% of the feed than when it was higher. In the same substrate mixtures, glu cose-based biomass yields were higher and CO2 yields were lower when glucos e constituted 2% or 10% of the feed but were constant for higher percentage s. The results suggest that the fate of a substrate is relatively independe nt of the feed composition as long as the substrate in question constitutes a significant percentage of the mixture. Thus, in those situations the ass umption of a constant biomass yield in multicomponent substrate modeling is justified. However, when a given substrate constitutes a small percentage of the feed, significant changes in yield may occur. (C) 2001 John Wiley & Sons, Inc.