Theory 10/Food Science & Fermentation /Fermentation science/3 min
Flavour Is a Trajectory, Not a Recipe
Two batches end in the same place and taste different. The difference is the route.
The flavour of a fermented food is set by the sequence of microbial stages rather than by the final composition, so two fermentations ending with the same species and metabolites can taste different, and process control should target the trajectory.
01The evidence
- Flavour formation in a fermented food has been characterised through a three-stage model, with the mechanism of flavour development tied to the succession rather than to the endpoint.Food Chemistry
02The tension
Fermentation is controlled by endpoint specifications: final pH, final species composition, final metabolite concentrations. Those specifications are met by batches that taste different, and the difference is attributed to terroir or to chance rather than to the sequence.
03The theory
A fermentation is a succession. Each stage produces metabolites that the next stage may consume, convert or leave behind. The final profile therefore depends on which intermediates were produced and how long they persisted.
If the trajectory matters, then two batches with identical endpoints can differ because their intermediate histories differ, and the difference is in trace compounds that the endpoint specification does not include.
The prediction is that flavour should be predicted by the trajectory, measured as the sequence of dominant genera and metabolites, better than by the final composition. It also predicts that interventions targeting an early stage should change flavour even when the endpoint is unchanged.
The practical consequence is that process control should monitor the trajectory and that specifications should include intermediate markers rather than only endpoints.
04What this is not
Endpoint models judge a fermentation by where it finishes. This judges it by how it got there, which explains flavour differences between identical endpoints.
05Predictions
- P1Flavour should be predicted by trajectory better than by final composition.
- P2Early-stage interventions should change flavour at constant endpoint.
- P3Intermediate markers should improve specification.
06What would refute it
Flavour determined entirely by endpoint composition would leave the trajectory redundant.
07What it would change
It would change fermentation control and specification.
08Evidence base
- Mechanism of flavour formation in a three-stage fermentation model — Food Chemistryhttps://www.sciencedirect.com/science/article/abs/pii/S0308814625045686