Joint effect of nitrogen and phosphorous on glucose oxidase production by Aspergillus niger: Discussion of an experimental design with a risk of co-linearity

The combined effects of nitrogen and phosphorous on the production of glucose oxidase and gluconic acid by Aspergillus niger cannot be adequately described with Monod-type model, neither do they fit well to linear equations with interactions N × P, nor quadratic with N2 and P2 terms. On the other hand, the interactions of type N2P and NP2, although common in real cases such as enzymatic kinetics in the presence of inhibitors, should be verified – if included in empiric models – by means of designs that can lead to artefactual results derived from the co-linearity. To avoid this risk we propose a procedure, based on the ‘bootstrap’ algorithm, which provided consistent results in the mentioned bioproductions. Applied together with methods of response surface and gradient, said procedure allowed to optimize the enzyme production as a function of the concentrations of N and P, to quintuple the initially obtained levels, and to explain other culture behaviours related with the sources of these nutrients.