INTRODUCTION

As the dairy industry moves towards increased production of products with extended shelf life, the bacterial quality of raw milk is increasingly important to final product quality (Boor et al., 1998). To consumers, quality means that the product tastes good and that it keeps well in their home refrigerator (Boor, 2001). Moreover she reported that from a fluid milk processor's or regulatory point of view, quality is measured more objectively by comparing product conformance to established standards on the last day of sale. High SCC levels are not known to pose a direct public health risk, yet they reflected mammary infection and over all quality management (Schalk et al., 2002). Moreover lower SCC levels have been shown to be related to higher milk yield and better dairy products quality and are, therefore, of economic value (Ma et al., 2000). Another elements under regulation is bacterial count in milk (Schalk et al., 2002). Moreover, Hayes et al. (2001) reported that of the various measures of raw milk quality, the Total Bacterial Count (TBC) is of particular interest to the dairy farmers and processor. Increasing awareness of public health and food safety issues in present years has lead to a greater interest in milk quality (Schalk et al., 2002). The TBC frequently factors into the price farmers received for their milk, as many raw milk purchasers establish price incentives for milk with a low TBC (Hayes et al., 2001). Moreover they also added that the TBC serves as a rough gauge of herd health farm sanitation efficacy and proper milk handling and storage temperature. The volume of the official hygiene regulation for food processing establishments has been growing continuously over the past 20 to 30 years. This led to a decrease in hygiene risk awareness in food processing establishments which was partly replaced by strong reliance on legislative measures in food hygiene (Untermann, 1996).

There has been interest in recent years in expanding the shelf life of the fluid milk because of potential advantages for both processor and consumer (Gruetzmacher and Bradley, 1999). The shelf life and flavour changes of pasteurized milk are affected by processing conditions, packaging materials and bacterial growth (Allen and Joseph, 1985). One of the principle factors associated with this concern about milk quality is its shelf life (Gruetzmacher and Bradley, 1999). They also added that the consumer determines the acceptance of the fluid milk by flavour and length of time before milk spoils in the refrigerator.

Boor (2001) mentioned that to maintain or increase market share, however, the processor's goal should be to meet, or perhaps, create the consumer's quality expectations. To meet these challenges, dairy processors must focus on improving the quality and extending the shelf lives of their products. Unfairly good flavour and acceptable keeping quality are essentials in maintaining fluid milk sales. Proper selection of a milk carton is essential to provide barrier properties against the transmission of light, organic flavour compounds and oxygen from the air into the package. A good barrier will retain the aroma and flavour of a product to achieve a reasonable shelf life (Simon and Hansen, 2001).

MATERIALS AND METHODS

Source and Analysis of Milk Samples
The present study involves 8 different milk companies who process and distribute pasteurized milk in the Western Cape of South Africa. The pasteurized milk samples were randomly collected from the 5 different food stores and retailers distributed in the 5 major different social economical groups. The milk samples were brought to the Laboratory of the Medical Microbiology, University of Western Cape in an ice container. Part of the milk samples were spilt aseptically into 40 mL sterile bottles for bacteriological analysis, coded and refrigerated over night. All microbiological evaluations were done at Provincial Veterinary Laboratory, Stellenbosch. The somatic cell counts (SCC) were estimated using coulter counter (Beckman, Z1 series, England) according to the manufacturers recommended procedures. South African Bureau of Standards (SABS) methods were applied to the total bacterial count (ISO 6222, 1999), enumeration of coliforms (SABS ISO 4831,1991), detection of Escherichia coli (SABS ISO 7251, 1993), detection of Salmonella (ISO 6579, 1993) and detection of Listeria monocytogenes (SABS ISO 11290/1and 2,1996 and 1998). Standard cultures for the isolation of Staphylococcus spp., Streptococcus spp., Enterococcus spp. and Bacillus cereus were also done according to Quinn et al. (1994) and Bergey’s manual (Holt et al., 1994). Similarly another 50 mL of the same milk samples were coded and brought to the ARC-Animal Nutrition and Product Institute, Elsenburg for the determination of the chemical composition of milk. Percentages of fat, protein, lactose and SNF were done, using infrared spectrophotometer (Milko Scan 133B analyzer, A/S N. Foss Electric, Hillerford, Denmark). Whereas total solids and the ash contents were obtained by subtraction. The freezing point, to detect the percentage of the added water was also done by the advanced Cryscope (Fiske, USA).

Statistical and Data Analysis
The rate of isolation of each organism in the pasteurized milk sample were calculated as a percentage of the total number of the isolates. Those isolates were further regrouped in to three categories (major pathogens: 1; minor pathogens: 2 and negative: 0), according to Berning and Shook (1992) to facilitate their statistical analysis. Descriptive statistical (mean, standard deviation, variance, maximum and minimum) and ANOVA test of the paired t-test analysis were also performed, using the Statistical Packages for Social Science (SPSS, 10). Correlations and their significant level among the measurements were estimated using Pearson correlation using the same program (SPSS, 10).

RESULTS

Table 1 showed that E. coli was isolated from 3.9% of the pasteurized milk samples. Their counts were found to be more than 190 cfu mL^-1 in one sample for each of 3 companies. Similarly S. epidermidis and E. faecalis were found in milk collected from each of the 3 companies supplying the pasteurized milk (Table 1). Enterococcus faecium (2.6%) was found in 2 samples of the milk from one factory, while S. intermedius was isolated from one sample (1.3%) of one company. All the examined pasteurized milk samples during the present study revealed no growth for Listeria spp. and Salmonells spp.

The somatic cell count (SCCs) was found to range from 1.2x10⁴-3.81x10⁵ cell mL^-1 with a mean of 6.426x10⁴±5.429x10⁴ cell mL^-1. The TBC was found to range from 60-1.0x10⁷ cfu mL^-1 with a mean count of 9.94x10⁵±2.9x10⁶ cfu mL^-1 (Table 2). The mean count of coliform bacteria was 2.844x10⁴±1.2x10⁵ cfu mL^-1, the minimum was zero and the maximum was 1.0x10⁶ cfu mL^-1 (Table 2). However E. coli revealed counts of 2.8x105 and 0 cfu mL¯1 for the mean, maximum and minimum values, respectively (Table 2).

The mean, minimum and the maximum values of fat and solids not fat of the pasteurized milk were 2.187±0.828, 6.093±1.423%; 0.54, 3.63, 3.94, 9.22%, respectively. The protein and lactose content of the pasteurized milk revealed 2.168±0.592, 3.195±0.835%; 1.19, 1.72, 3.47 and 5.04% for mean, minimum and maximum values, respectively.

Table 1:	Frequency of isolation of some bacteria from pasteurized milk in the western cape

Table 2:	Frequency analysis of the quality of pasteurized milk contents in western cape of South Africa

Table 3a:	Comparison of variations of bacteriological quality of pasteurized milk consumed in the Western Cape
In this and the following tables: A-I indicate various selected companies of milk processing

Table 3b:	Comparison of various chemical quality of pasteurized milk consumed in the Western Cape

Table 4:	Mean square and the level of significant of the quality of pasteurized milk in Western Cape of South Africa
*p<0.05, **p<0.01, ***p<0.001, NS = Non significant

The total solids and the ash contents of the pasteurized milk were estimated as 8.279±2.155, 0.72%±0.0005, 4.17, 0.71, 13.16 and 0.73% for mean, minimum and maximum values, respectively. The added water revealed 24.153±14.833, 0 and 53.90%, respectively.

Descriptive analysis of the different measurements of the pasteurized milk of the individual company showed higher means and standard deviations for somatic cell count, total bacterial count and coliform count (Table 3a). However, E. coli count and standard cultures revealed noticeable variations among the milk of the different companies. Similarly, the average compositional quality (fat, protein, lactose, SNF and TS) of the pasteurized milk were found to show quite differences between the companies manufacturing the milk. Moreover, all estimated values were found to be lower than the standard. Similarly, the percentages of the added water were also high for all collected pasteurized milk samples regardless of the variations between the individual companies (Table 3b).

The data in Table 4 also showed that there were highly significant differences (p<0.001) for SCCs, TBC, coliforms count and standard cultures of the pasteurized milk samples produced by the different companies. The different food stores from which the pasteurized milk samples were purchased, revealed significant variations for TBC (p<0.01), coliforms count and standard culture at p<0.001 and fat % at p<0.05 (Table 4).

Table 5:	Correlation coefficient of pasteurized milk contents in Western Cape of South Africa
*p<0.05, **p<0.01, ***p<0.001, ND= Not done

Similarly comparison of the different socioeconomic areas of the Western Cape also revealed significant variations for TBC and coliforms count (p<0.05) and the standard culture (p<0.01). Moreover, highly significant variations were also reported for protein, fat, lactose, SNF and TS (p<0.001). The percentages of the added water also revealed significant differences (p<0.05) when comparing the areas from which the milk was purchased. The lower or higher fat content of the milk was found to affect significantly the estimated fat% (p<0.001) and the total solids% (p<0.001). However, the different expired dates of the pasteurized milk were found to show significant variations only with the different SCCs. Neither the volume nor the types of the containers for packing the pasteurized milk were found to show significant variations on the quality measurements undertaken during the present study.

Correlations Between the Compositional and the Hygienic Quality Measurements for the Pasteurized Milk Samples
Significant (p<0.001) positive correlations were observed when comparing fat% of the pasteurized milk and each of standard culture (r = 0.848), protein (r = 0.969), lactose (r = 0.991) added water (r = 0.980) and ash (r = 0.328) as shown in Table 5. Significant (p<0.01) positive correlations were reported when comparing protein content of the pasteurized milk with lactose (r = 0.990), added water (r = 00.980) and ash (r = 0.338). Comparison of lactose with added water and ash showed significant (p<0.01) positive correlation (r = 0.976, r = 339) and significant negative correlation when compared with SCCs (r = -0.239, p<0.05). However comparison of SNF and each of fat and the added water revealed significant (p<00.315 and 353, respectively). The ash content of the pasteurized milk revealed significant (p<0.05) correlations when compared with SCCs, TBCs, coliform bacteria count and standard culture (r = -214, -0.297, 0.294 and 0.50, respectively). The standard culture showed significant positive correlation’s when compared with coliform bacteria count and E. coli count (r = 0.509 and 0.298, p<0.01). Similarly, the coliform count revealed significant (p<0.01) positive correlations’s when compared with TBC (r = 0.512). Moreover, the TBC revealed significant (p<0.05) positive correlations’s when compared to standard culture (r = 0.282). Other non significant correlations were also represented between other measurements as presented in Table 5.

DISCUSSION

The data in Table 2, 3a and 4 showed the pasteurized milk samples produced by the factories in Western Cape of South Africa are compliant with strict of the most world criterion for somatic cell count. This might be due to the system of payment for milk which depends on the SCCs as a criterion for the quality. This supported Dekkers et al. (1996) who reported that an alternative to deriving the economic value for SCCs is to derive the economic important of improved milk quality directly from it's impacting efficiency of milk processing. Moreover, Ma et al. (2000) reported that the key milk quality element being regulated is SCCs. However they reported that high SCCs levels are not known to pose a direct public health risk, yet they reflect mammary infection and overall quality of management. However, the present study recorded the presence of some microorganisms in the pasteurized milk like E. coli, Enterococcus Facium and Enterococcus faecalis, Staphylococcus epidermidis and S. intermedius (Table 1) suggested the contamination of the milk as stated by IDF (1994). Moreover, the frequency distribution for the microbial quality of the pasteurized milk (Table 2 and 3a) gives a comprehensive overview of pasteurized milk quality. Since the number of colony forming unit/ml of the TBC, coliforms count and E. coli counts in some of the milk samples were found to be above the quality standards. This finding supported O' Ferrall-Berndt (2003) who reported some pathogens and high microbial counts of milk from shop of South Africa. The quality standard in Brazil for TBC are 8x10⁴ and 3x10⁵ cfu mL^-1 and for coliforms count are 4 and 10 MPN/mL (Silva et al., 2001). Moreover, the presence of other isolates (Table 1) might be the reason for the lower bacteriological quality measures for some of the samples. This was in accord with Manie et al. (1999) who reported that coliform bacteria commonly contaminate raw milk as they do not survive pasteurization and is frequently used as an indicator of inadequate processing or post processing contamination. Lombard (1976) reported that pasteurization of milk provides protection for the consumers against pathogens which may be present in the raw milk and improves its keeping quality. Hence attention should be given to the sources of the contamination of the pasteurized milk. They include the microbial quality of raw milk, time and temperature of pasteurization, presence and activity of post pasteurization contaminants, types and activity of pasteurization resistant microorganisms and the storage temperature of milk after pasteurization (Gruetzmacher and Bradley, 1999). The higher count of E. coli might be due to unrefrigerater transportation and poor microbiological quality of water (Adesiyun, 1997). Moreover, member of E. coli may be associated with definite signs of illness and sometimes death as discribed by Osman (1998).

Further those contaminants and the high count were found to affect the compositional quality of the pasteurized milk (Table 4 and 5), which agreed with Kitchen (1981), Munro et al. (1984) and Mohamed et al. (1997). Moreover, the percentages of fat, protein, lactose, total solids and solid not fat were found to be lower than the standards. However Boor (2001) reported that currently pasteurization is the most common methods of destroying pathogenic organisms and reducing or eliminating spoilage organisms in dairy products by the HTST methods. Gruetzmacher and Bradley (1999) reported that the consumers determine the acceptance of fluid milk by flavour and length of time before milk spoils in the refrigerator. They also added that uniformly good flavour and acceptable keeping quality are essential in maintaining fluid milk sales. However, during the present result the expiry dates of milk issued by producers were found to have non significant effect on the quality of milk. This might be due as reported by Gruetzmacher and Bradley (1999) that elimination of post pasteurization contamination and proper cleaning and sanitation increased the shelf life of milk to 20.4 days instead of 9 days. Similarly Boor (2001) reported that consumers expect fluid milk products to be nutritious, fresh testing and wholesome. These lower values might be due to the high percent of the added water (Table 2, 3b, 4 and 5). It might reflect either the adulteration by water and/or the extra skimming of milk fat. Added water reduces the value of the milk by diluting the protein and other milk components that will influence products yield. Also the high percent of the added water might be due to technical faults during the processing of the pasteurized milk. Moreover, their variations were pronounced between the companies producing the milk and the areas to which they were supplied. This might suggested that different quality milk is produced and distributed. Since it was noticed during the present study that one of the company produced it's pasteurized milk under two commercial names. One of the products was of good quality and it was found in specific food stores, while the other was of low quality and was picked up from the areas of less standard of live (poor and rural areas). Both they need regulations, as Boor (2001) reported that increasing public awareness and regulatory attention directed toward food safety issues highlight the need for the dairy industry to proactively address and eliminate emerging food safety issues that may negatively impact the image of dairy products sales.

In conclusion we supported reports which stated that education, training and incentives are probably the key components of a total milk quality assurance programs, for producers, processors and consumers.