The aim of the present study was to identify the nature of bonds established between protein particles after stirring that are responsible for the texture improvement of stirred yoghurts, called rebodying. Using a constant model yoghurt at pH 4·4, the effects of changes in the physicochemical conditions at stirring were studied on the subsequent rebodying. Short term rebodying was measured as the changes in viscoelastic properties at 4°C during 20 h after stirring, while long-term rebodying was measured as the viscosity changes during 28 d storage at 4°C. Moreover, stirred gels obtained from either set gels that were allowed time or not for ionic equilibration were compared. Increasing or decreasing ionic strength did not change the properties of stirred gels. Calcium chloride addition significantly decreased G′0 h, G′20 h and tan20 h but did not induce changes in the gel microstructure as observed by confocal scanning microscopy. Yoghurt rebodying could not be explained by fulfilling ionic equilibrium. Moreover, N-ethyl maleimide addition had no effect on the stirred yoghurt. Attractive electrostatic and disulphide interactions were not involved in the gel rebodying and increasing calcium concentration in the set gel limited rebodying.

The enhancement of the strength of set acid gels by heating milk was related to rheological parameters (water retention capacity, storage modulus) of corresponding stirred gels. To obtain accurate rheological data from stirred gel it was necessary to maintain a constant granulometry of gel particles and to recognize time after stirring as a contributing factor. Two hours after stirring, the gel exhibited a higher storage modulus when milk was heated above 80 °C. A measurement of viscosity of just-stirred yoghurt was sufficient to predict correctly the quality of a stirred gel analysed by viscoelastic measurements. Increased resstance to syneresis of just-stirred gels was related to higher viscosity. The quantity of β-lactoglobulin (β-lg) bound to casein micelles explains the improvement of these gel qualities. We have considered that the structure of the initial firm gel (mesostructure level) was conserved in fragments within the stirred gel. Consequently, the explanation given by various authors for the effect of heating milk on the properties of set gels can also be applied to stirred gels. The same mechanism, described in literature for structure formation of set gels from acidified milk is purposed to explain the role of heating milk on the recovery of gel structure after stirring. The β-lg association with casein micelles during heating favoured micelle connections during the acidification. It also favoured the association of gel fragments after stirring during the recovery in gel structure.