Acid and rennet gelation properties of sheep, goat, and cow milks: Effects of processing and seasonal variation

Figure 1 presents the acid gelation parameters of all samples (n = 9) for the sheep and goat milks in comparison with the control cow milk. The RM of all species had comparable long acid gelation times (>130 min) and low G′ values (<50 Pa). The small difference in the G′ of the RM acid gels followed the order of the protein content of the milk (Table 1): sheep > cow > goat, as higher protein concentrations naturally contribute to stronger gels. The tan δ values of the RM acid gels were around 0.30 for all species. The gelation pH of the RM was highest for RM_cow (pH 4.85), followed by RM_sheep (pH 4.63), and was lowest for RM_goat (4.55). Previous studies also reported that RM_cow has a higher acid gelation pH of around 4.8 (; ) than those of RM_sheep and RM_goat (around pH 4.6; ), the mechanism of which remains to be investigated. Overall, the acid gelation capabilities of the RM of different species were all fairly poor and did not vary substantially.

The cow milk behaved as expected following homogenization and the intense heat treatment at 95°C for 5 min, with the gelation pH increasing from 4.85 to 5.24, a significant reduction in the gelation time, and the G′ increasing from below 30 Pa to over 500 Pa (Figure 1). The results are consistent with those widely demonstrated in previous studies indicating that heat-induced extensive whey protein denaturation greatly improves acid gelation properties of milk (, ; ). Homogenized and pasteurized cow milk showed intermediate gelation properties between RM_cow and HHM_cow.

Similar processing impacts were observed for the sheep and goat milks (i.e., HHM had the shortest gelation time, the highest gelation pH, and the highest G′; P < 0.001 in all cases). The large variation in the G′ of HHM_sheep acid gels (Figure 1C) arose from seasonal variations, which are presented and discussed later (Table 2). Consistent across the species, the tan δ values of the acid gels were mainly reduced by homogenization, although the impacts of the heat treatments were more significant for goat milk (Figure 1D). However, for goat milk, the reduction in the gelation time (34.2%) and the increase in the gelation pH (0.24) from RM_goat to HHM_goat were less pronounced than for the sheep and cow milks (>60% reduction in gelation time and approximately 0.40 increase in gelation pH; Figure 1). Still, HHM_goat had a slightly lower gelation pH (4.79) than RM_cow (4.85) and a very low G′ (<25 Pa), even after the improvements by processing, whereas the G′ values of the HHM_cow and HHM_sheep acid gels were a few hundred Pa (Figure 1C). Gelation time, gelation pH, and G′ were significantly affected by the interaction between processing treatments and animal species (sheep and goat; P < 0.01, 2-way ANOVA).

The results indicated that the goat milk had inferior acid gelation functionality to the cow and sheep milks, particularly after processing treatments (Figure 1). The extents of processing-induced enhancements of the acid gelation properties were lower for the goat milk than for the sheep and cow milks. Our findings agree with previous studies demonstrating that acid gels and yogurts produced from goat milk have weaker structures than those produced from cow milk and sheep milk (; ; ). A few characteristics of goat milk that are different from those of cow and sheep milks may contribute to its inferior acid gelation properties, particularly after processing. The low casein-to-whey protein ratio and the low concentration of α_S1-CN in goat milk have been considered to be the reasons for the weak texture of goat milk acid gels or yogurts in previous studies (; ; ). Consistent with these studies, the protein analysis of the goat and sheep milks used in this study () showed that the goat milk had a lower casein-to-whey protein ratio and a lower proportion of α_S1-CN than the sheep and cow milks (, ). These 2 factors would contribute to the poor gelation properties of goat milk, with or without heat treatment. Furthermore, we found that the goat milk also had a lower ratio of β-LG to α-LA than the sheep and cow milks (), which was consistent with results reported previously (; ). This difference would specifically affect the heat-induced improvements in the acid gelation properties of goat milk because whey proteins actively take part in the development of the acid gelation of milk only following their denaturation (). Previous studies have shown that α-LA is inferior to β-LG in improving acid gelation properties or yogurt texture (; ; ), probably because of its lower hydrophobicity, its lower isoelectric point, and the absence of free thiol groups (; ; ). suggested that the α-LA in milk may compete for the thiol groups of denatured β-LG and form protein complexes that are inferior in promoting acid milk gelation compared with β-LG/κ-CN complexes that contain less or no α-LA. The higher proportion of α-LA in the goat milk whey proteins may have contributed to the lower increase in the acid gelation pH following heat denaturation and the weak gel structure.

Table 2 shows the seasonal variations in the acid gelation properties of sheep milk. Consistent acid gelation properties were found in RM_sheep over the seasons. In contrast, PM_sheep and HPM_sheep showed significant seasonal variations (i.e., early-season sheep milk had a slightly shorter gelation time and a higher gelation pH than its late-season counterpart, whereas the G′ of the PM_sheep acid gel was higher in the late season). The most prominent seasonal differences were found for the HHM_sheep acid gels. The HHM process increased the G′ of the acid gels in the early and mid seasons to around 500 Pa but resulted in a much lower increase in G′ for the late-season sheep milk (mean of 158.9 Pa), despite its high protein content (Table 1). Late-season HHM_sheep had a significantly longer gelation time (75 min) and a lower gelation pH (pH 4.93) than the milks from the other seasons. Season × processing interaction significantly affected the G′ (P < 0.01, 2-way ANOVA) and gelation pH (P < 0.05, 2-way ANOVA) of sheep milk acid gels. In addition, for the early- and mid-season HHM_sheep samples, the tan δ value reached a maximum after gel formation (tan δ_max) at pH 4.99 ± 0.02, as has been well demonstrated in heated cow milk (; ), which was surprisingly absent for the late-season HHM_sheep samples (Table 2, Figure 2). Correlation analysis showed that the desirable acid gelation properties of HHM_sheep, including the higher gelation pH and the higher G′, were significantly correlated with the smaller casein micelle size of the heated milk (P < 0.001). To lesser extents, the β-LG content, the ionic calcium concentration, and the viscosity of HHM_sheep were also negatively correlated with the gelation pH and the G′ (P < 0.05). As reported in , the greater heat-induced increase in the average casein micelle size and the viscosity in late-season sheep milk was likely to be contributed mainly by the heat-induced micelle–micelle aggregation, which was promoted by the higher concentrations of protein and ionic calcium in late-season sheep milk. We hypothesize that (1) the larger and partially aggregated casein micelles may have hindered the optimal packing of the micelles during gelation and (2) the elevated viscosity may limit the mobility of components in the system and delay gel formation and rearrangement.

Similar to the seasonal trend observed in sheep milk, acid gels made from similarly processed cow milk from a seasonal-calving herd had a significantly longer acid gelation time, a lower gelation pH, and a lower gel G′ in the late season (). However, the extent of the decrease in G′ in the late season compared with the early and mid seasons was much greater for sheep milk in the present study (about 70%, Table 2) than for cow milk (20% to 40%) in the study by . In addition, revisiting the data reported in , we found that all late-season heated cow milk samples had a tan δ_max following gel formation despite their lower gelation pH, which was absent for the late-season heated sheep milk in the present study. The presence or absence of a tan δ_max appears to be crucial for the development of acid gel structure, which is determined by the gelation pH (; ). We further discuss this observation in the next section. It is unclear whether the similar trends in the acid gelation properties of seasonal lambing/calving sheep milk and cow milk were (partially) contributed by any similar factors relating to the changes in the milk properties in late lactation. The heated cow milk (90°C for 6 min) in the study of did not vary in its casein micelle size over the seasons, and thus the impact of micelle aggregation just discussed for sheep milk would not apply to cow milk.

Seasonality did not greatly affect the acid gelation properties of goat milk (Table 3). The only significant differences were found in the summer, when the RM_goat had the shortest gelation time and the HHM_goat gels had the lowest G′ (P < 0.05, Table 3). For goat milk, the seasonal variation in its milk solids content appeared to be the only factor that had an impact on its acid gelation properties. The protein and fat contents, which were lowest in summer goat milk, were correlated negatively with the gelation time of RM_goat and positively with the G′ of HHM_goat gels (P < 0.05). Summer goat milk with fewer milk solids had a lower buffering capacity and thus the gelation pH of the RM_goat (pH 4.55 in all seasons, Table 3) was reached earlier during acidification. It is also expected that the lower solids content of summer goat milk contributed to the lower G′ of the gels. Other than these impacts of varying solids content, the acid gelation properties of goat milk in the present study were stable over the year. This contrasted with the sheep milk in the present study and the New Zealand cow milk in previous studies (, ), for which the late-season milk had worse acid gelation properties beyond the apparent effect of the protein or fat concentrations. Unlike the sheep and cow milks, which were produced from typical seasonal calving/lambing systems, the goat milk in the present study was produced from a mix of spring-kidding and autumn-kidding goats. The results indicate that this practice effectively stabilized the acid gelation properties of goat milk over the seasons. Similarly, reported that mixing spring- and autumn-calving cows in a herd eliminated the seasonal variations in most of the compositional parameters and processing properties of the milk.

This section is dedicated to discussing the acid gelation results and the potentially crucial role of a critical gelation pH in deciding the G′ of the final acid milk gels across different ruminant species and processing treatments. Two intriguing results from the acid gelation analysis were found in the present study: (1) acid goat milk gels had very low G′ values despite the intense heat treatment (95°C for 5 min) that induced extensive whey protein denaturation and an increase in the gelation pH (Figure 1); (2) late-season sheep milk had a markedly lower G′, a slightly lower gelation pH of about 4.9, and the absence of a tan δ_max in contrast to its earlier-season counterparts (Table 2, Figure 2). It appears that the acid gel rigidity G′ will be fairly low when the gelation pH is lower than a critical value, regardless of the species or the processing history of the milk.

To further illustrate the potential impact of a critical acid gelation pH on the final G′ of acid milk gels, we plotted the G′ data of the acid milk gels against the gelation pH from the present study (goat, sheep, and cow milks) and previous studies (, ; ; ) in Figure 3. All of the included studies acidified the milk with GDL, defined the point of gelation as the point at which the G′ was greater than 1 Pa, and reported the final G′ by the end of the gelation process (from 6 to 13 h). Consistent across the studies, the final G′ of the acid gels formed at pH in the range 4.5 to 4.9 did not change greatly with the gelation pH (<100 Pa in all cases), whereas, above a critical gelation pH of around 4.9 to 5.0, the final G′ of the acid gels increased markedly (in a linear manner) up to around 1,000 Pa with increasing gelation pH (up to approximately 5.5). Intriguingly, the patterns were consistent in different studies despite the fact that the different gelation pH were induced by various factors, including the ruminant species (present study), seasonality or stage of lactation (present study; ), heating intensity (present study; ), pH of the heat treatment (90°C for 30 min, ), and the use of a low level of rennet in addition to GDL ().

It is commonly believed that the denaturation of the whey proteins in milk increases the acid gel stiffness (indicated by G′) by increasing the amount of gelling materials and connecting bonds (especially the stronger disulfide bonds) and elevating the gelation pH to around 5.2 because of the higher isoelectric points of the whey proteins (approximately 4.8 to 5.3) than those of the caseins (, ). In addition, it has also been demonstrated that an increase in the gelation pH to around 5.2 results in a structural rearrangement at around pH 5.0, as indicated by the presence of a tan δ_max, which leads to the formation of straighter strands in the gel network that contribute to a higher gel stiffness (; ; ). We hypothesize that the latter effect of structural rearrangement (via increasing the gelation pH over a critical level) plays the dominant role in enhancing the G′ of the final acid milk gels. Below the critical gelation pH, there is no structural rearrangement and the G′ remains low. Above the critical gelation pH, colloidal calcium phosphate takes part in the formation of the initial gel structures (; ), and its involvement in the initial gel matrix increases with increasing gelation pH. The structures formed at a higher pH will undergo a longer and greater extent of rearrangement at around pH 5.0, which will contribute to forming a final gel network that has straighter strands and a higher G′. The hypothesis is supported by the following observations (Figure 3): (1) below the critical gelation pH, the G′ of the differently processed goat milk gels (gelation pH at or below 4.87) was below 25 Pa despite the differences in the level of whey protein denaturation (from 0% to around 95%, ); (2) for the HHM_sheep samples, which had greater than 90% of the whey proteins denatured (), the low gelation pH of 4.93 for the late-season samples led to G′ values that were 70% lower than those of the early- and mid-season samples, whose gelation pH values were only approximately 0.2 units higher but were sufficient to induce a tan δ_max; (3) in unheated cow milk, in which the contribution of whey protein is negligible, the addition of rennet increased the acid gelation pH to 5.40, induced a tan δ_max, and enhanced the final G′ to about 800 Pa, which was greater than for the heated milk without the addition of rennet (about 400 Pa, ); (4) in heated cow milk, the G′ of the acid gels increased with increasing gelation pH despite similar extents of whey protein denaturation (; ; , ). Based on these observations, it appears that the effect of heat-denatured whey proteins in enhancing the final G′ of acid milk gels arises primarily via the increase in the gelation pH from 4.8 to approximately 5.2, which induces the structural rearrangement at pH 4.9 to 5.0, in addition to increasing gelling components and connecting bonds. The latter effect can also be seen in the results presented in Figure 3, between samples with similar gelation pH and various extents of whey protein denaturation. For instance, an HPM_sheep sample (<20% whey protein denaturation) that gelled at pH 4.90 had a final G′ of 80 Pa (dashed-line circle, Figure 3), whereas the 3 late-season HHM_sheep samples (>95% whey protein denaturation) with marginally higher gelation pH (4.91 to 4.94) had a mean final G′ of 159 Pa (solid-line circle, Figure 3). However, the magnitude of this difference in G′ was far smaller than that between samples with similar levels of whey protein denaturation but a larger difference in gelation pH above 5.0 (Figure 3).

This hypothesis can explain the differences in the G′ and the gelation pH observed in the present study as induced by processing, species, and seasonality. Sheep and goat milks have lower natural acid gelation pH values. The gelation pH of the processed milks remained lower than (goat milk) or just greater than (sheep milk) the critical gelation pH of around 4.9 to 5.0. As a result, the G′ of the acid goat milk gels remained low regardless of the processing improvements (Figure 1), whereas the late-season HHM_sheep without a tan δ_max had markedly lower G′ values than those from other seasons (Table 2). In contrast, RM_cow has a higher natural gelation pH (pH around 4.8) than RM_sheep and RM_goat, leading to the observations that a greater heating intensity (and thus a higher degree of whey protein denaturation) increased the gelation pH and the G′ of acid gels consistently throughout the typical gelation pH range of heated cow milk (5.0 to 5.4). As a result, the potential effect of a critical gelation pH (and structural rearrangement) per se on the G′ of the acid gels made from heated cow milk is indistinguishable from the effect of the simultaneously increasing level of whey protein denaturation.

The hypothesis can be tested by studying the gelation pH/G′ relationship around and over the critical pH range (e.g., by attempting to further increase the gelation pH of goat milk into the sensitive range of 4.9 to 5.0). It should be noted that the G′ of acid gels that is enhanced by the structural rearrangement process, compared with that contributed by additional gelling proteins, probably translates differently into other textural characteristics of acid gels or yogurt-type products. Intense heating of milk increases the stiffness G′ but also lowers the fracture strain of the acid gel and makes it more brittle (), which is typical for a coarse protein gel with straight strands (). demonstrated that the G′ of cow milk yogurt (approximately 800 Pa) was about 7 times higher than that of goat milk yogurt (approximately 100 Pa), whereas the difference in their firmnesses, as determined by textural analysis, was less than 40% (0.32 N for cow milk yogurt and 0.23 N for goat milk yogurt). Further investigation is needed to elucidate how different factors contribute to the rigidity G′ of acid milk gels and the simultaneous effects on other structural characteristics, such as large deformation properties.

The rennet gelation properties of the sheep milk, the goat milk, and the control cow milk are presented in Figure 4. For the RM samples, the rennet gelation times of RM_sheep and RM_goat were slightly shorter than that of RM_cow, in agreement with previous reports (; ). Similar to the acid gels, the difference in the G′ of the RM rennet gels followed the same order as the solids content of the milk: sheep > cow > goat (Figure 4B). The tan δ of the RM rennet gels followed the order goat > sheep > cow (Figure 4C). The RM rennet gels had higher tan δ and larger between-species variation than the RM acid gels, which was probably associated with the larger pore size and the higher structural flexibility and tendency for syneresis of rennet gels compared with acid gels (; ).

Tan δ was the only rennet gelation parameter that displayed consistent processing impacts across the ruminant species (Figure 4C). The tan δ of rennet gels was reduced by homogenization (P < 0.05) and was not affected by the intense heating of the homogenized milk (P > 0.1 between HPM and HHM). The differences in tan δ between species observed for RM (goat > sheep > cow) were much reduced following homogenization. Native fat globules do not actively participate in the structure of milk gels and disrupt the development of a continuous protein network, which may have contributed to a higher tan δ of the gels. Following homogenization, the reduction in the fat globule size and the adsorption of caseins on the newly formed fat globule surface probably promoted active interactions between small fat droplets and the protein network, making the resulting gel more solid-like with a lower tan δ and thus a lower propensity for syneresis (; ).

Processing impacts on the rennet gelation time and the G′ showed some species-dependent differences. For cow milk, the HPM_cow rennet gel had a shorter gelation time and a higher G′ than the RM_cow rennet gel. Previous studies have reported that homogenization reduces the rennet coagulation time of cow milk (; ; ). suggested that less κ-CN cleavage is required to induce the rennet gelation of homogenized milk because of the spreading of casein micelles at the surfaces of homogenized fat globules. The reported effects of homogenization on rennet gel rigidity or firmness are mixed (; ). Consistent with our findings, some previous studies have reported that homogenization increased the rigidity G′ or firmness of rennet cow milk gels (; ; ), whereas other studies have shown that the homogenization of milk lowered the rennet gel firmness () or enhanced the gel rigidity only within a limited period (). The intense heating of cow milk showed detrimental effects on its rennet gelation properties, as the HHM_cow rennet gels had the longest rennet gelation time (20.2 min) and the lowest G′ (14 Pa) among all samples analyzed in the present study (Figure 4). Heat-induced denaturation of the whey proteins is well known to prolong the rennet gelation time, weaken the curds, and cause the resulting curd to contain more moisture (; ; ).

Compared with cow milk, the processing treatments of the goat and sheep milks had less pronounced effects on the rennet gelation time and the G′ of the gels. The rennet gelation time of sheep milk was not significantly affected by the treatments, whereas HPM_goat had a slightly longer rennet gelation time (15.9 min) than RM_goat and HHM_goat (13.5 and 13.3 min, respectively). The G′ of the sheep milk rennet gels was highest for RM_sheep, whereas the HPM_goat rennet gels had the lowest G′ among the differently processed goat milks (P < 0.05). However, these processing effects were less pronounced than those on cow milk (Figure 4) and were not consistent throughout the seasons (Table 4, Table 5). Homogenization did not affect the rennet gelation properties of sheep milk and slightly impaired those of goat milk (prolonging the gelation time and lowering the G′, Figure 4). Homogenization appeared to have both promoting and retarding effects on rennet coagulation, which are sensitive to the physicochemical differences in the milks of different ruminant species. Cow milk has a larger native fat globule size (and thus less surface area and milk fat globule membrane material), a smaller casein micelle size, and a higher casein-to-whey protein ratio than goat milk (), all of which may contribute to a more complete and even coverage of the casein micelles on the newly formed surfaces of the homogenized fat globules that contribute to the rennet gel structure. In addition, a higher proportion of serum-phase β-CN was present in the RM_goat (16.0%) than in the RM_sheep (6.7%, ) or cow milk (approximately 4%, revisiting data from ). β-Caseins are known to be excellent emulsifiers that probably compete for adsorption at the fat globule surfaces with the casein micelles. Altogether, the surfaces of homogenized goat milk fat globules are likely to be covered by fewer casein micelles (which are larger in size and potentially less evenly coated), more remaining milk fat globule membrane material, and more soluble nonmicellar β-CN than cow milk, which may have shifted the balance between the promoting and retarding effects, leading to the slight negative effects of homogenization on the rennet gelation time and the G′ of rennet goat milk gels.

With respect to the effect of heating, in agreement with the present study, previous studies reported that intense heat treatments did not greatly affect the rennet gelation properties of sheep and goat milks, as compared with cow milk (; ; ). reported that heating (95°C for 1 to 10 min) reduced the rennet coagulation time of goat milk, in contrast to that of cow milk, which agreed with our result for HHM_goat (in comparison to HPM_goat, Figure 4). found that varying the heating conditions did not affect the rennet coagulation time of sheep milk. The rennet coagulation time of goat milk was reduced by the same heat treatments but did not consistently correlate with the heating intensity or the level of whey protein denaturation (). There are a few possible reasons for the lower impacts of heating on the rennet gelation of sheep and goat milks compared with cow milk. First, the casein micelles have lower colloidal stability in sheep and goat milks than in cow milk (; ,). As a result, less hydrolysis of κ-CN is needed to induce gelation in goat and sheep milks and the heat-induced association between whey proteins and κ-CN would have a less hindering effect on goat and sheep milks than on cow milk (). Second, the ionic calcium concentration is higher in goat milk and lower in cow milk, with sheep milk being intermediate (; ). The ionic calcium concentration in cow milk may be close to a critical level for rennet gelation, and the heat-induced precipitation of calcium plays a key part in retarding the rennet coagulation of cow milk but not of goat and sheep milks (). reported that calcium addition markedly reduced the coagulation time of intensely heated cow milk but had limited impact on the rennet coagulation of heated goat milk or raw cow milk. Third, it was suggested that the more uniform distribution of the denatured whey proteins on the micelle surface in heated goat milk hinders the coagulation and fusion of casein micelles less compared with heated cow milk, in which larger whey protein aggregates are present ().

Seasonality did not greatly affect the rennet gelation properties of the sheep and goat milks (Table 4, Table 5). Most of the rennet gelation properties of sheep milk were not affected by the season, except for the lower G′ of HHM_sheep and the higher tan δ of RM_sheep and PM_sheep in the late season (P < 0.05, Table 4). Overall, the effect of seasonality on the rennet gelation properties of sheep milk was small compared with that on its acid gelation properties (Table 2). Late-season HHM_sheep had the lowest rennet gel rigidity G′, which was negatively correlated with the casein micelle size of HHM_sheep (r = −0.738, P < 0.05). This observation appears to agree with previous studies, in which milk with smaller casein micelles was associated with higher rennet gel strength, but not shorter gelation time (). Smaller casein micelles may interact with each other more readily during the gelation process because of their larger total surface area (; ). However, these studies found correlations between the native micelle size of RM_cow and its rennet gelation properties. The increase in micelle size with heating observed in the heated sheep milk (95°C for 5 min), because of the association of denatured whey proteins and micelle–micelle aggregations (; ), possibly contributes to the rennet gelation of milk somewhat differently compared with the naturally varying casein micelle size; this remains to be elucidated. Several previous studies investigated the rennet coagulation properties of RM_sheep and reported mixed effects of the lactation stages. A few studies found that the rennet curd firmness increased with lactation (; ), whereas other studies reported the opposite trend (; ). The rennet coagulation time has been reported to increase with the stage of lactation (; ), to decrease with the stage of lactation (), or to be the lowest in mid lactation (; ; ). Similar to the present study, some studies did not find a consistent effect of lactation on the rennet coagulation properties of sheep milk (; ). In summary, it is difficult to draw a conclusion on the effect of the season or the stage of lactation on the rennet coagulation properties of sheep milk, which can be confounded by factors such as parity (; ), breed (), lambing season (), and diet quality ().

For goat milk, the G′ of the RM_goat rennet gel was lower in summer than in winter. The tan δ of the rennet gels was higher in spring than in winter only for RM_goat but not the processed goat milks (Table 5). Similar to sheep milk, the rennet gelation time of goat milk was not affected by seasonality. The seasonal variation in rennet gel rigidity G′ can be explained by the varying concentration of the milk components, as with the acid gelation properties discussed earlier. The G′ of RM_goat was positively correlated with the concentrations of fat, total calcium, and colloidal calcium (P < 0.05), all of which were lower in summer than in winter (). A greater rennet gel strength was reported to be associated with higher contents of fat () and colloidal calcium phosphate () in milk. For all processed goat milks, there were no significant seasonal variations in any of the rennet gelation properties (Table 5). The year-round goat milk production by mixing spring-kidding and autumn-kidding herds largely maintained both the acid gelation and the rennet gelation properties of goat milk over the seasons.