Ten independent isolates from each evolution experiment were
used for comparing fermentation performance with the original
strain. For these experiments, synthetic must sugar concentration
was set to 28%, in order to be able to show technological advantages
of the evolved strains under conditions that are expected to result in
a stuck fermentation. Indeed, all the strains lef t considerable
amounts of residual sugars, but the trend for all the evolved strains
was for faster fermentation kinetics, higher weight loss (CO2
pro-duction), and lower residual sugar content, than the original in-dustrial strains (Fig. 2). The differences among the strains indicate
that the fi nal population of the evolution experiments was not ho-mogeneous, but it was rather a mixed culture of a number of strains,
a result in agreement with the observation by Brown and Oliver
(1982) . Four strains among those leaving the lowest amounts of
residual sugars, L76M2, L76M4, L76M8 and L78M4, were chosen for
further characterization in pilot scale assays in Bodegas Torres S.A .