Lactic acid bacteria grow faster when you feed them more sugar — up to a point. Beyond a critical substrate concentration, growth slows, then stalls. This is substrate inhibition, and it has been modeled for decades with single-pathway kinetics: one equation relating growth rate to substrate concentration, with an inhibitory term that kicks in at high concentrations. The models fit batch data reasonably well. They also fail to capture what actually happens inside the cell.
A mechanistic two-pathway model separates the bacterial response into distinct metabolic routes that activate at different substrate concentrations. At low sugar levels, the bacterium channels substrate through a high-efficiency pathway — maximizing yield per molecule consumed. As concentration rises past the threshold, a second pathway activates, one that processes substrate faster but less efficiently, producing more lactic acid per unit time at the cost of more waste heat and lower biomass yield. The switch is not a breakdown — it is an adaptive reallocation of metabolic resources.
The two-pathway model achieves higher predictive accuracy than single-pathway alternatives, particularly in the transition zone where inhibition begins. The improvement is not just statistical. It is mechanistic: the model identifies the concentration at which the cell shifts strategy, and that transition point is where fermentation engineers have the most leverage. Controlling substrate feed rate to hold the culture just below the switching threshold maximizes both growth and product formation — a practical insight that emerges only when the model captures the dual-pathway architecture.
The through-claim reaches beyond fermentation. Any system that responds to increasing load by switching between operating modes — efficient-but-slow versus fast-but-wasteful — will show apparent inhibition at the transition point. The inhibition is not failure; it is the cost of changing strategy. Understanding the transition as a switch rather than a collapse changes what you optimize for.