The non-contribution of the EPS from the LB2 starter to viscosity
and serum retention, in comparison with the EPS from the LB1 and
ST1 starters, may have been due to the high level of branching and
the flexibility of its EPS backbone, causing a smaller radius of volume.
van den Berg et al. (1995) showed that neutral EPS dissolve
easily in the serum because they interact less with the positively
charged caseins than anionic EPS do, and thus neutral EPS cause
less syneresis. This effect was observed in the present study for the
neutral EPS from the LB1 starter, which retained more serum than
the anionic EPS from the ST1 starter.