Pasteurization effects on yield and physicochemical parameters of cheese in cow and goat milk

Cheese production is one of the most common forms of valorization of dairy production, adding value and preserving milk. Various types of cheese produced from raw and pasteurized milk are known worldwide. On the present work, we are interested in studying the effect of the type of milk (raw and pasteurized) of two species (cow and goat) on the yield and physicochemical characteristics of the fresh cheese. On the northeastern Algeria, on 5 cow farms and 3 goat farms; 5 raw and 5 pasteurized milk cheese manufacturing trials were conducted. The analysis of the results of the 80 samples of milk and cheese of both cow and goat species showed that the latter contained significantly more fat and protein than cow’s milk and that pasteurized milk contained more protein than raw milk. As a result, actual cheese yield of goat cheese was higher than that of cows in pasteurized and raw milk. For higher yield, our result supported the use of pasteurized milk as a raw material in the manufacture of farmhouse cheese. Practical Application: For cheese-making under small scale using milk from cow or goatI: pasteurization can increase actual cheese yield.


Introduction
On Algeria, as in all regions of the world, consumption of dairy products like cheeses is an old tradition linked to livestock farming, since dairy products are made using ancient artisanal processes, employing milk or mixtures of milk from different species (Boudalia et al., 2016;Leksir et al., 2019;Leksir & Chemmam, 2015;Shori, 2017).
Cheese is one of the most common milk products, which is produced by raw or pasteurized cow's milk, or other species such as goats and sheep milk (Mahieddine et al., 2017). Whatever the species, the consumption of cheeses produced from raw milk has advantages and disadvantages, it tends to display greater variability in comparison with their cheese counterparts made of pasteurised milk, and characterised by strong and unique organoleptic, which is highly appreciated by the consumers in the case of raw cows' milk cheese Montel et al., 2014), or not always gaining the consumers' sensory acceptance in the case of raw goats' milk cheese due essentially to the short-chain fatty acids presence (Mituniewicz-Małek et al., 2019). Several studies show that raw milk cheese contains a wide variety of microflora including beneficial bacteria, especially lactic acid bacteria, which also contribute to more intense and stronger flavor production than that of pasteurized milk cheeses (Casalta et al., 2009;Grappin & Beuvier, 1997). These results have been attributed to several indigenous microbiota, such as Lactococcus spp., Lactobacillus spp., Leuconostoc spp., and Enterococcus spp. Moreover, indigenous microflora, especially lactic acid bacteria can control the proliferation of many contaminating bacterial pathogens and thus protect the cheeses from microbiological risk via growth inhibition of food-borne pathogens within raw milk cheese using bacteriocin-mediated mechanism (Yoon et al., 2016). On the same way, from literature, populations with a similar genetic background shown that children growing up on a farm have a lower risk of developing asthma and allergies due to the consumption of raw unpasteurized milk (Waser et al., 2007); especially, those who consume goat milk, which has been identified as more favorable allergenic characteristics (Verruck et al., 2019;Ranadheera et al., 2019).
However, the use of raw milk unpasteurized in the production of cheeses carries a potential health risk; linked to the possible occurrence of antibiotics and other drugs residues, mycotoxins, estrogenic hormones, pesticides, industrial pollutants and zoonotic pathogens (Alegbeleye et al., 2018). Ot also can be linked to the presence of Listeria monocytogenes, verocytotoxin-producing Escherichia coli, Staphylococcus aureus, Salmonella and Campylobacter and other pathogenic bacteria (Yoon et al., 2016;Kousta et al., 2010;Verraes et al., 2015).
To avoid risk for human health from raw milk consumption, pasteurization is considered the most widely used method in the dairy industry. Ot consists of a treatment process that heats raw milk with specified time and temperature combinations, designed to eliminate all known milk-borne pathogens and undesirable bacteria resulting in a not harmful product of constant quality. Hence, cheese manufactured from unpasteurized milk may seem to be superior to cheese made from pasteurized milk in its microbiological safety (Yoon et al., 2016). Also, this treatment can slightly affect milk composition, physicochemical characteristics, nutritive profile and sensory properties (Alegbeleye et al., 2018).
Pasteurization has been shown to have either a positive influence (Lau et al., 1990;Salwa & Galal, 2002;San Martín-González et al., 2007) or no effect on cheese yield (Drake et al., 1997). Such differences in results might be related to the breeds or the methodology of cheese making, thus assessing the effects of pasteurization of cheese yield on local breeds with a standard methodology is key and might shed light on the influence of pasteurization and give guidance to cheese producers. Besides, given that most local cheese is made from pasteurized milk, it is important to reveal potential differences in quality and yield of cheese using raw and pasteurized milk. Here, we investigate the effect of pasteurization on yield and physicochemical characteristics of cheese made from cow and goat milk collected from five different farms in Northeast Algeria. We have two hypotheses based on literatureI: (1) due to the interspecific similarity in milk content (Arora et al., 2013), we hypothesize that cows and goats have comparable cheese yield; (2) due to the dominance of the positive effect of pasteurization on cheese yield (Jørgensen et al., 2019), we expect the same outcome in both species. These aspects are of the primary social, economic and health importance for local dairy production.

Study sites
The samples are collected from five county (Tamlouka, Ain Makhlouf, Ain Larbi, Belkheir, and Bouchegouf) located in the wilaya of Guelma which is located in North-East Algeria. The region is characterized by a subhumid climate in the center and in the North and semi-arid in the South. The climate is mild and rainy in winter and hot and dry in summer.

Milk sampling
A total of 80 samples of milk from both cow (n = 50) and goat (n = 30) were studied. Raw milk samples were collected every other week during a period of two and a half months (from May to mid-July 2018) from five cow and three goat breeding farms. A quantity of 1.3 liters of a mixture milk was collected in sterile flasks, and then transported in a cooler with an ice-bath to the cheese factory in Ain Makhlouf (Guelma, Algeria). All bottles are previously autoclaved at a temperature of 121 °C, under a pressure of 1 bar for 15 minutes. The transport times vary depending on the remoteness of the sampling sites. On order to take account of the real field conditions, no conservative was added. For physicochemical analysis, flasks are transported in a cooler to the laboratory of Biochemistry of the University May 8, 1945, Guelma-Algeria. For the manufacture of the two types of cheese (raw and pasteurized), 0.5 liter of the milk was pasteurized at a temperature of 82 °C for 10 seconds and 0.5 liter was used raw.

Milk analysis
pH and acidity were measured during the manufacture of cheeses; pH was measured using a pH meter Adwa, AD1000 and acidity was determined according to the method described by Rhiat et al. (2013). At the dairy processing units, fat, protein content, temperature and density of milk were measured with Lactoscan milk analyser (series number 15-1030, made in Bulgaria).

Cheese making
At a temperature of 20 °C, both types of milk are inoculated with commercial mesophyll ferment of the brand Dupont-Danisco (Dangé-Saint-Romain, France) which contains lactic ferments Lactococcus lactis sub. lactis, Lactococcus lactis sub. cremori and Lactococcus lactis sub. lactis biovar diacetylactis at a dose of 2 g/100 L then is smoothed and matured for three hours, then is rennet with a rennet of animal origin dose 1 L/15000 and left curdled for 13 hours, this curd is molded for 24 hours. Finally it is demolded and drained for 24 hours. The manufacturing process lasts about three days. On total, 40 samples of raw milk cheese and 40 samples of pasteurized milk cheese were analyzed to determine the physico-chemical properties and yield.

Cheese analysis
The real cheese yield for all cheeses is calculated after demolding and theoretical cheese yield was determined using pradal model equation, which based on fat and protein content of milkI: Theoretical yield = 0.211 × FC + 0.093 PC + 4.0 (FC, PC designated fat and protein content respectively).
Cheese physico-chemical analyses were carried out on the fifth day of manufacture in the university laboratory. For fresh cheese, the following analyses were carried outI: pH measurement, determination of titratable acidity, total dry extract, ash content humidity and theoretical and real cheese yield. Hydrogen potential (pH) is measured using a pH-meter "Adwa, AD1000". Total dry extract (TDE) was determined using a method which consists of evaporating water from the test socket in an oven (Memmert) at a temperature of 103 °C and weighing the residue according to the method ADAC 926.08 (Association of Dfficial Analytical Chemists, 1995). The determination of the ash rate is made by incineration of the sample at 550 °C for 3 hours in an electric muffler furnace with thermostat. The water content (humidity) is determined according to Amariglio (1986).

Statistical analyses
The results were expressed in form of the mean ± SD (Standard Deviation). The experiments were carried out in duplicates. Two-way ANDVA was conducted to test if there was a significant difference in actual cheese yield between species (cow vs. goat) and cheese type (raw vs. pasteurized).
Linear mixed-effects models were carried out to calculate average yield across farms (using farm as a random intercept), and assess differences in fat and protein content and humidity among species and milk type, and reveal potential relationship between theoretical and actual yield. PCA was carried out to see which physicochemical variables best explain yield. Significance was considered at p <0.05 using R 3.5.1 (R Development Core Team, 2018).

Physicochemical analysis of milk
The mean ± SD of fat, protein, temperature, acidity, density and pH of the raw and pasteurized milk of cow and goat are presented in Table 1. There was a significantly higher fat content in goat milk than in cow milk (LMEI: p = 0.0001), but no significant difference between raw and pasteurized milk (LMEI: p = 0.95). The average fat content was 52.7 ± 6.36 g/L (n = 30) in goats and 47.7 ± 6.85 g/L in cows (n = 50). Similarly, there was a higher protein content in goat milk than in cow milk (LMEI: p<0.0001), and in pasteurized milk than in raw milk (LMEI: p = 0.002). The average protein content was 37.5 ± 2.35 g/L (n = 30) in goats and 30.5 ± 1.62 g/L in cows (n = 50), whereas it was 33.8 g/L in pasteurized milk (n = 40) and 32.5 g/L in raw milk (n = 40).

Actual cheese yield
On cows, the analysis of the actual cheese yield measured on 50 individuals in five farms with two methods (raw and pasteurized milk) showed that there was a significant effect of the type of cheese on the actual cheese yield (ANDVAI: F 1,40 = 6.31, p = 0.01) (Figure 1). However, neither was farm (ANDVAI: F 4,40 = 1.93, p = 0.12) nor the interaction between farm and type of cheese was not significant (ANDVAI: F 4,40 = 0.94, p = 0.44). When we consider farm as a random effect, the mixed effect linear model showed a significant effect of the type of cheese (t-value = 2.51, p = 0.01), revealing that the actual cheese yield in pasteurized milk (24.0 ± 2.64 kg) was 8.1% higher than in raw milk (22.2 ± 2.64 kg).
By analyzing data on 30 goats from three farms, we found that the type of cheese (ANDVAI: F 1,24 = 12.68, p = 0.001) and the farm (ANDVAI: F 2,24 = 4.14, p = 0.02) had a significant effect on actual cheese yield (Figure 2). The interaction between farm and type of cheese was not significant (ANDVAI: F 2,24 = 2.40, p = 0.11). By estimating the difference in the actual cheese yield between raw and pasteurized milk using a mixed effect model, we found that cheese yield with pasteurized milk (27.1 ± 2.7 kg) was 13.9% higher than with raw milk (23.8 ± 3.18 kg) (t-value = 3.38, p = 0.002).
The comparison of the actual cheese yield between the two species (cow and goat) showed that goats had higher actual cheese yield than cows (LMEI: species effect, p = 0.04) in both pasteurized and raw milk (LMEI: milk type effect, p = 0.0001). Dn average, goat milk produced 10.3% more cheese yield than cow milk. Pasteurized goat milk yielded 12.9% (difference = 3.1 kg) more cheese than pasteurized cow milk, whereas raw goat milk yielded 6.7% (difference = 1.6 kg) more cheese than raw cow milk (Figure 3).   six factors, 80 samples of the two species showed that the two first components explained 92.2% of the variance (Figure 4). PC1 which explained 66.6% of the variance was positively correlated with actual cheese yield (r = 0.81, p < 0.0001) and cheese humidity (r = 0.72, p < 0.0001), but negatively correlated with pH (r = -0.33, p = 0.0002) and ash percent (r = -0.71, p < 0.0001). PC2 which explained 25.5% of the variance was highly positively correlated with acidity (r = 0.91, p < 0.0001) and weakly negatively correlated with actual cheese yield, pH, humidity and ash percent (r = -0.30 --0.23, p < 0.05). Thus, based on the PCA, our data suggest that humidity was a good explanatory variable for actual cheese yield.
Linear mixed-effects model regressing actual cheese yield against humidity, type of cheese and species (farm as a random effect) showedI: (1) no significant effect of species and its interactions with other variables (two and three-way interactions); and (2) no significant effect of type of cheese and its interactions, expect a significant effect in the interaction between humidity and type of cheese (p = 0.04) ( Table 3). The latter significant interaction indicates that only humidity of pasteurized cheese showed a positive relationship with actual cheese yield in both cows and goats.
However, there was no significant effect in the humidity of cheese between raw and pasteurized milk in both cows (LMEI: t-value = 0.35, p = 0.72) and goats (LMEI: t-value = 1.79, p = 0.08).
The average humidity of cow cheese was 61.7 ± 5.61% using raw milk and 62.3 ± 7.16% using pasteurized milk, whereas the average humidity of goat cheese was 62.9 ± 4.05% using raw milk and 65.2 ± 2.94% using pasteurized milk.

Comparison between actual and theoretical yield
The theoretical and actual yield were positively correlated in raw and pasteurized cheese of both cows and goats (LMI: slope = 1.02 ± 0.18 (±SE), R 2 = 0.48, p <0.0001; Figure 5). The comparison between theoretical and actual yield using a linear mixed-effects model with milk sample as a random effect showed that although theoretical yield was on average greater than the actual yield (Table 4; LME, YieldI: p <0.0001), this different occurred only in raw cheese and not in pasteurized cheese (Table 4; LME, Yield × CheeseI: p = 0.04) (Figure 6). The difference between theoretical and actual yield in raw cheese was 2.38 ± 2.28 kg in cows and 3.67 ± 2.95 kg in goats, which represent 10.7% and 15.4% of the actual yield in cows and goats, respectively. However, the difference between theoretical and actual yield in pasteurized cheese was 0.57 ± 2.07 kg in cows and 1.49 ± 2.24 kg in goats, which account for 2.41% and 5.48% of the actual yield in cows and goats, respectively.

Physicochemical analysis of cheese
The mean ± SD of ash, moisture, temperature, acidity, dry matter and pH of the raw and pasteurized milk of cow and goat are presented in Table 2. A principal component analysis using

Discussion
Here, we assess the difference in the cheese yield using raw and pasteurized milk of cows and goats and it reveals that (1) in both cows and goats, pasteurized cheese had higher yield than raw cheese; (2) actual cheese yield was best explained by moisture content; and (3) although actual and theoretical yield were positively correlated, they presented a significant difference using raw milk but no difference using pasteurized milk. Cheese yield is influenced by different factors divided into two groupsI: i) milk quality, which depends essentially on the animal breed, species, genetic variants of proteins, lactation stage, feeding system, animal management, and environmental    conditions; ii) cheese-making processes as cold storage of milk, heat treatment, standardization, coagulation, curd cutting and cooking, draining, pressing, and salting or ripening (Lucey & Kelly, 1994).
Dur results showed that pasteurized cheese produced more actual yield than the raw cheese in both cows and goats. Similar findings were found in other studies on different kinds of cheeses, namely in Cheddar cheese with 10.32% vs. 10.56% (San Martín-González et al., 2007) and 10.13 vs. 10.21% under 63 °C for 30 min) (Lau et al., 1990) and in Mashanza cheese in Africa with 29.30 vs. 30.95% (under 72 °C for 15s (Salwa & Galal, 2002)). Dther studies that used pasteurization at 72 °C showed no difference in actual yield between raw and pasteurized cheese (10.6 vs. 10.5%) (Drake et al., 1997).
The increased yield in pasteurized cheese is probably, due to the higher cheese moisture retention which is known to increase in pasteurized milk leads to higher recovery of whey proteins and soluble solids (Goff, 2008;Lucey & Kelly, 1994). The other main finding of this study is that there was an interspecific difference in cheese yield between cows and goats.
While some studies show that cheese yield is higher in cows than goat (Rasheed et al., 2016), our study and others show the opposite (Hamidi et al., 2018;Mallatou & Pappa, 2005). Different factors might produce this difference, including those related to the milk composition and quality such as genetic variants of casein, fat and protein (Banks et al., 1981;Fenelon & Guinee, 1999;Verdier-Metz et al., 2001) and cheese-processing methodology (Lawrence, 1993). Here, we found higher fat and protein content in goat milk, which probably contributed to the increase in cheese yield (Lucey & Kelly, 1994). Hamidi et al. (2018) found lower fat content in goat than cow's milk in semiarid regions of Algeria, in these regions plants abundance and richness are very low. These differences in cheese yield and milk content are of interest to the local and regional cheesemakers (North Africa) where such information is still not well known and goat cheese-making is still a minority. Moisture content is one of the key factors that determine the cheese yield (Banks et al., 1981;Emmons et al., 2001).
Dur results showed that pasteurized cheese had more water content that raw cheese in both cows and goats, which could to some, explain the higher yield in pasteurized cheese.
Ot has been shown that cheese moisture retention increases with increasing temperature of milk pasteurization. Higher moisture reflects higher recovery of whey solids and salt in the cheese. On addition, a 1% increase in moisture result in a 1.8% increase in Cheddar cheese yield; in other words, 90 kg cheese/1000 kg milk, a moisture adjustment to 36% would result in 91.6 kg cheese/1000 kg milk (Hill, 2016).
On summary, our study assessing the effect of pasteurization on yield of cow's and goat's cheese reveals that using pasteurized milk increases yield through an increased moisture content of the cheese. Moreover, goat milk produces higher cheese yield than cow milk, showing higher protein and fat content. These results are of primary important for local and region milk and cheese industry, particularly for goat whose cheese production has received more local attention in recent years. These results show the importance of the techniques used in the dairy industry, in particular, the heat treatment. Raw milk pasteurization has not only a positive effect on cheese yield but also protects consumer following the elimination of pathogenic flora. Ot can also increase more profits for the farmers following milk transformation and cheese yield increasing.
Although yield was higher in pasteurized cheese, there are other components such as vitamins and enzymes that might be ruptured during the heating process; an aspect that warrant investigation. Finally, further studies should assess the microbiological quality of the two types of milks in both species and determine the best conditions under which each type of cheese should be produced and conserved.