Bacteriological Properties Of Udder Surface And Milk Samples From Sheep Farms In Hungary

Background Ewe milk due to its beneficial composition and properties contributes to the growth of microorganisms. The primary prerequisite for making high-quality sheep milk product is the production of high-quality raw sheep milk by dairy farms. Thus, the aim of this study was to examine the bacteriological properties of the udder surface (US), individual ewe raw milk (IERM) and bulk tank milk (BTM) samples from sheep farms in Hungary. Methods Seventy-seven US, seventy-seven IERM, and ten BTM samples were examined from March 2018 to April 2019. Total plate count (TPC), Enterobacteriaceae count (EBC), Escherichia coli count (ECC), Staphylococcus aureus count (SAC), lactic acid bacteria count (LAB) and psychrotrophic bacteria count (PBC) of different ewe breeds were examined according to ISO standards in four Hungarian sheep farms. The differences in the microbiological status between the breed and farm were considered. Results High counts of TPC (3.2±1.1 lg cfu/cm 2 ) and EBC (2.4±0.7 lg cfu/cm 2 ) were found in Cigája breed of US samples in farm1. TPC of US sample of Lacaune (farm2), Lacaune (farm3) and British milk sheep was 2.7±0.8, 2.2±0.5 and 2.4±0.4 lg cfu/cm 2 , respectively. The mean value of TPC of IERM of Merino, Cigája and Dorper breeds in farm1 was 2.8±1.2, 3.5±1.2 and 3.3±0.7 lg cfu/ml, respectively. There was significant difference (p<0.05) between Merino and Cigája breeds for TPC. Although there was no significant difference (p<0.05) between breeds for EBC, the mean of EBC of IERM samples from Dorper breed was highest in farm1. Comparatively low TPC of IERM of British milk sheep breed was recorded. The mean of TPC of BTM was 7.4±0.6, 6.3±0.4

Both LAB and PBC were high in BTM of Lacaune breed. Except for BTM, SAC and ECC were not detected in most US and IERM samples. Conclusion The presence of microorganisms in BTM above the limit indicates high microbial contamination due to poor hygienic conditions during milking and milk handling. Practicing very good hygiene principles at the farms, in handling and transportation of milk, is a must.

Background
Dairy sheep industry is a promising branch of livestock production in Hungary. According to the report of Hungarian Central Statistical Office (KSH), the country has 798, 000 heads of ewe in 2018. In the country, most of ewe population milked belongs to Merino breeds, however, there are some milk breeds like Lacaune, Awassi, milking Tsigai and British milk sheep [1]. Data from Faostat [2] revealed that Hungary takes 15th place from EU member countries with the sheep milk production of 1149 tonnes in the year of 2017. The interest for ewe milk in Europe is increasing from time to time.
Similarly, in Hungary it was observed that there was yearly change in sheep milk production [2].
However, ewe milk is an excellent media for the growth of microorganisms due to the high content of fat, protein, total solids, essential vitamins and minerals [3].
Milk is virtually sterile when it is synthesized and secreted into a healthy alveoli of ewe mammary gland. However, the exterior of ewe's udder can contribute bacteria that are associated with the skin of animal and from the environment in which animals are housed and milked [3]. The microbiological quality of the sheep milk can be affected by contamination during and after milking, method of milking, breed, season and the hygiene of farms [4]. According to the Regulation (EC) 853/2004 of the European Parliament and the Council [5], the total count of microorganisms is a basic and mandatory indicator for evaluation of raw sheep milk quality. TPC of ewe milk that will undergo pasteurization before processing must not exceed 1.5 × 10 6 cfu/ml and shall not exceed TPC of 5 × 10 5 cfu/ml if the milk is intended for processing without heat treatment [5].
Unpasteurized ewe raw milk contains pathogenic and non-pathogenic bacterial population. Ewe milk is dominated by lactic acid bacteria (LAB) [6]. One of the main fermentation products of the metabolism of carbohydrates, lactic acid, was produced by LAB [7]. The consumption of raw ewe milk has a risk for the consumer, due to the possible presence of human pathogenic organisms in the raw ewe milk [8]. Oliver et al. [9] from USA reported that raw milk could be contaminated by pathogenic microorganisms as clearly shown by numerous epidemiological studies. Ewe milk is a source of undesirable bacteria like Enterobacteriaceae, Escherichia coli, psychotrophic bacteria and Staphylococcus aureus [10]. Apart from Hungary, there were literature which dealt with the microbiological quality of ewe milk, especially in BTM in different ewe breeds at different places and time. For instance, the study by Ombarak and Elbagory [11] from Egypt pointed out that the mean 4 value of total bacterial count was 2.04 × 10 6 cfu/ml, Enterobacteriaceae was 1.67 × 10 5 cfu/ml and S. aureus was 6.67 × 10 4 cfu/ml in raw ewe milk.
This study was based on the fact that bacteriological quality studies on sheep milk have so far been relatively neglected although the production (and the interest towards increased production) is showing an upward trend worldwide. This is also true for Hungary, where there is a paucity of information on bacteriological analysis of raw sheep milk. Therefore, this study was designed to study the bacteriological properties of udder surface, individual ewe raw milk and bulk tank milk samples of

Bacteriological quality of udder surface samples
The result revealed that the mean value of TPC was 2.5±1.0, 3.2±1.1 and 2.5±0.8 lg cfu/cm 2 in US samples of Merino, Tsigai and Dorper breed from F1, respectively (  (Table 2).

Bacteriological quality of ewe milk
The mean values of TPC were 2.8±1.2, 3.5±1.2 and 3.3±0.7 lg cfu/ml in IERM samples of Merino, Tsigai and Dorper breeds from F1, respectively (Table 3). There was significant difference (p<0.05) between Merino and Tsigai breeds for TPC. The mean of TPC of IERM samples from Tsigai breed was significantly (p<0.05) higher than TPC of IERM samples from Merino breeds. In the case of F2 and F3, the mean TPC of Lacaune breed was 3.3±1.0 and 3.5±0.9 lg cfu/ml, respectively (Table 3). Even though, there was no significant (p>0.05) difference between two farms, the mean of TPC of IERM samples from F3 was higher than F2. The mean of TPC of IERM samples from British milk sheep breed in F4 was 1.8±0.4 lg cfu/ml.
In the case of F2 and F3, the mean EBC from Lacaune breed was 2.2±1.1 and 2.2±0.0 lg cfu/ml, respectively ( Table 3). The mean of EBC of IERM samples of British milk sheep breed in F4 was 1.4±0.0 lg cfu/ml. SAC was 2.6±0.7 and 2.8±0.3 lg cfu/ml in in IERM F3 and F4, respectively (Table   3). ECC was not detected in IERM from F1 and F4 (Table 3). Regardless of the breed, LAB was the same in IERM of F3 and F4 ( Table 3).
The bacteriological count results by years of sample collection are shown in table 5. The mean of TPC of US samples examined in all breeds has got significant (p<0.05) difference between years (Table 5).
A significantly higher value was recorded in 2018 than 2019 for TPC of US samples. Also, there was significant (p<0.05) difference between years for EBC of US samples taken from Tsigai and Dorper 6 breeds. In the case of IERM, TPC was significantly (p<0.05) different between years in Merino and Tsigai (Table 5). A significantly higher value was recorded in 2018 than in 2019.

Correlation between microbiological parameters
Linear correlation coefficient of Pearson was calculated to evaluate the correlation between US and IERM for TBC and EBC (Table 6) Table 6).

Discussion
Natural reservoirs of bacteria like the body of the animal, especially, the udder surface of sheep contributes the contamination bacteria in raw milk [3]. The udder surface of ewe certainly become contaminated with manure and mud while they are lying. This was one of reasons US samples were examined. Also, the diameter of US was easily measured during sampling. The higher value of TPC of US samples in Tsigai breed than Merino and Dorper breeds is might be due to the difference in the degree of udder surface dirtiness with manure and dust particles. In this study, the mean of TPC of US samples from F2 was significantly (p<0.05) higher than F3. This could be related to factors such as the location of the farm, hygienic condition of farm and handling of animals. The mean of EBC of US in Tsigai breed was significantly (p<0.05) higher than Merino breeds. This was maybe due to a hairy udder surface which stick with dust and feaces in the case of Tsigai breeds. In general, the low EBC of US samples indicates a good hygienic condition of ewe housing. Comparable data on a bacteriological load of US samples of sheep breeds are rare in literature, and this study is the first dealing with this specific parameter in the studied area.
In the present study, the significant difference (p<0.05) between Merino and Tsigai breed for TPC in IERM was observed. This finding was in contrast with the finding of Alexopoulos et al. [4] reported that the difference between ewe breed was not significant (p>0.05). Even though, the difference between 7 F2 and F3 was not significant (p>0.05), the mean of TPC of IERM samples from F3 higher than F2. This was might be because of hand milking and absence of pre-milking disinfection in F3. This finding differed from the report of Bytyqi et al. [12] from Kosovo who reported that the farm had a significant effect on TPC. Along with the poor hygienic conditions of hand milking, the hand can act as a vector for transmission of environmental and contagious pathogens, increasing their counts in milk [13].
Comparatively low TPC of IERM of British milk sheep breed could be because of extremely hardy and strong characteristics of British milk sheep. The hygienic practices during milking, raw milk and milk products quality can be indicated by examining total plate count of bacteria [14].
The higher value of EBC (1.4±0.0 lg cfu/ml) of IERM was found in Dorper breeds in F1 (Table 3). Our result was lower than the finding of Ombarak and Elbagory [11] in Egypt, which was 5.2 lg cfu/ml in raw ewe milk. SAC was <1 lg cfu/ml in F1 and not detected in F2 (Table 3). However, Alexopoulos et al. [4] reported that S. aureus was detected in raw ewe milk at an average level of 3.94 lg cfu/ml.
S.aureus can easily grow in ewe milk due to an excellent substrates of it. The mean value of LAB (3.3 lg cuf/ml) was the same in IERM of Lacaune and British milk sheep (Table 3). This result was lower than the finding of Kalhotka et al. [15] from Czech Republic which was 6.4 lg cfu/ml in raw milk of Lacaune breed.
The higher values of EBC in BTM is an indication of the possibility of bacterial contamination via the udder, by milking equipment or faeces [6]. Increased numbers of Enterobacteriaceae in BTM can also occur when Enterobacteriaceae grow on residual milk left in poorly sanitized milking equipment. The high count of LAB in BTM in F3, could be one of the reasons for the better quality of ewe's milk compared to the cow's milk. The PBC mean value in this study was less than the value reported by de Garnica et al. [13] from Spain, which was 5.7 lg/cfu in BTM of sheep. Storing milk more than 24 hr had an impact on PBC results leading to higher microbial counts.
In this study, the microbiological quality of ewe milk from BTM based on TPC values, except TPC of F4, were at unacceptable levels according to European Parliament and of the Council 853/2004 standard; TPC of ewe milk that will undergo pasteurization before processing must not exceed 6.2 lg cfu/ml. Thus the milk originated from the farms, where this study was undertaken, should be pasteurized before fermentation since the highest TPC count reached 7.4±06. Our result was higher than the finding of Zweifel et al. [16] from Switzerland which was 4.79 lg cfu/ml of TPC in BTM. Besides, mean values of TPC were higher than the report of Alexopoulos et al. [4] that the average TPC of 5.48 lg cfu/ml in Greece. Therefore, our finding indicated that there is contamination of milk during milking and milk handling. At present, there is no legislative limit for the EBC and coagulase-negative staphylococci in ewe milk, they are considered to be an indicators of hygienic circumstances in ewe milk production [17].
The mean of TPC of US from all breeds had significant (p<0.05) difference between years ( Table 5).
The explanation may be the dependence of bacterial load on the season and lactation dynamics. Also, there was significant (p<0.05) difference between years for EBC of US from Tsigai and Dorper breeds.
The higher value of EBC during 2018 might be supported by higher contamination of US of ewe. It could be due to the rising awareness of the sheep farmers about the urgent need for increasing the hygiene standards during the milking process. In the case of IERM, TPC was significantly (p<0.05) different between years in Merino and Tsigai (Table 5). Similarly, the finding by Gonzalo et al. [18] from Spain confirmed that year was significant source of variation for TBC in BTM. In fact, from the results, it is evident that the hygienic milk quality has improved during 2019 compared to 2018.
Moderate and positive correlations were observed between EBC IERM and EBC US (r = 0.56). This was might be the evidence that the environmental sources, such as dirty bed and contaminated feed are predisposing factors for the presence of this bacteria in raw milk In general, poor cleaning practices in the flocks tend to results in higher bacterial load in the milk.

Conclusions
In conclusion, there was no significant difference between ewe breeds for Enterobacteriaceae count of individual ewe raw milk. However, total plate count of individual ewe raw milk samples from farm1 was significantly different between breeds. Therefore, the breed affected the bacteriological status of individual ewe raw milk samples. Total plate count of individual ewe raw milk samples from Lacaune breeds has got no significance difference between farms. Hence, the farm had no significant effect on the bacteriological status of raw milk samples. There was a significant difference between the year of examination in the case of total plate count of individual ewe raw milk. Therefore, the year had an

Bacteriological examination
Upon arrival at the laboratory, milk samples were kept in the refrigerator until examinations. Nine ml peptone water was prepared for decimal dilution as described by Petróczki et al. [19]. Total plate count (TPC) was performed according to MSZ EN ISO 4833-1 [20] as described by Tonamo et al. [21].
Enterobacteriaceae count (EBC) was performed on Violet Red Bile Glucose (VRBG) agar (Biolab Ltd., Hungary) following MSZ EN ISO 21528-2 [22]. Escherichia coli count (ECC) was performed according to ISO 16649-2:2001 on Tryptone Bile Glucoronide (TBX) agar (Biolab Ltd., Hungary) medium and was incubated at 37 o C for 24 hrs. All plates were incubated at 37 o C for 24 hrs. Culturing and Latex agglutination test of Staphylococcus aureus were performed according to MSZ EN ISO 6888-1 [23] standard as described by Petróczki et al. [19]. Lactic acid bacteria count (LAB) was examined according to MSZ EN ISO 15214 [24] standard. Using the pour plate method, 1 ml of the test sample was inoculated into the Petri dish before pouring over the molten de Man, Rogosa and Sharpe agar (MRS). The incubation was at 30 o C for 72 hrs. Psychrotrophic bacteria count (PBC) was examined on plate count agar medium (Biolab Ltd, Hungary) according to MSZ ISO 17410-2 [25]. Plates were incubated at 7 o C for 10 days. All samples were plated in duplicate. Accordingly, bacteriological properties of US samples were carried out and the results were divided by 20.

Statistical analysis
Data on bacterial count were entered into MS Excel sheet and log-transformed before analysis. SPSS

Availability of data and materials
All data generated and analysed during this study are included in this article. Further information on the data can be obtained from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This study was supported by the EFOP-3.6.3-VEKOP-16-2017-00008 project. The project is co-financed by the European Union and the European Social Fund.

Authors' contributions
AT, IK and FP collected all the required data; AT performed the laboratory works; IK and FP participated in coordination and supervision; IK, FP and AT designed the study and drafted the manuscript; AT analysed and interpreted the data; IK, CL and FP critically and substantially revised the manuscript. All authors read and approved the final manuscript.