Somatic cells in milk represent epithelial cells and leukocytes (Paschino et al., 2019). Somatic cell count (SCC) is considered from many aspects as an indicator of udder health and generally is used for detection of subclinical mastitis in ewes (Gonzáles-Rodríguez, Gonzalo and San Primitivo, 1995; Pengov, 2001; Olechnowicz and Jaskowski, 2005). However, there is still a big discussion among scientists about the physiological level of SCC in milk of ewes for detection of their udder health (Persson et al., 2017).

Berthelot et al. (2006) reported in their study SCC <500 × 10³ cells.mL^-1 for healthy ewes and for infected ewes SCC >1000 × 10³ cells.mL^-1, if SCC was in flock ˃650 × 10³cells.mL^-1 it showed 15% incidence of udder disease to have subclinical mastitis. The results of Kern et al. (2013) indicated threshold of SCC 400 × 10³ cells.mL^-1 in meat breeds of sheep, 300 × 10³ cells.mL^-1 in dairy breeds and 100 × 10³ cells.mL^-1 in extensive breeds as right value in detecting problems with udder health. Hussein, El- Khabaz and Malek (2015) determined value of SCC ≥400 × 10³ cells.mL^-1 in Ossimi sheep as limit for detection subclinical mastitis. The limit for the detection of subclinical sheep mastitis was determined by Swiderek et al. (2016) as 200 × 10³ cells.mL^-1. Similar threshold of SCC for diagnosis of mastitis in Sarda sheep was considered at 265 × 10³ cells.mL^-1 (Caboni et al., 2017). Sutera et al. (2018) in their study showed value SCC ˃500 × 10³ cells.mL^-1 as a possible limit in relation to milk quality.

In the study in our breeding practise Idriss et al. (2015) reported 78% of the samples of individual ewes <600 × 10³ cells.mL^-1. Vršková et al. (2015) found out that 76% of Tsigai had SCC <300 × 10³ cells.mL^-1. In recent study Tančin et al. (2017) found out that 82.03% individual milk samples were <400 × 10³ cells.mL^-1, 71.79% milk samples were <200 × 10³ cells.mL^-1 and only 8.89% milk samples were >1000 × 10³cells.mL^-1. Oravcová, Mačuhová and Tančin (2018) found out 60% samples with SCC ≤200 × 10³cells.mL^-1.

The aim of this study was to describe the frequency of distribution of ewes in SCC groups on the basis somatic cell score (SCS) per whole lactation and estimate changes of SCS from1^st lactation to SCS in 2^nd lactation. The effect of farms was evaluated too.

The experiment was carried out during two periods in dairy practice. Seven ewes’ dairy farms were involved in the study during 1^st observed period in 2016 and 2017 and at eight farms during 2^nd observed period in 2017 and 2018. On the farms they were kept Tsigai breed, Lacaune and on one farm Slovak dairy sheep. Tsigai (TS) breed were kept on farm 1^st, 2^nd, 3^rd, 4^th and farm 5^th. Lacaune (LC) breed were kept on farm 6^th, 7^th, 8^th and farm 9a^th. On farm 9b^th they were kept Slovak dairy sheep (SD) in 2^nd observed period only. Within each of the period the same animals were sampled on their 1^st and following 2^nd lactation in next year of study. In 2 farms (1^st, 3^rd) hand milking was performed and remaining 7 flock were milked by machine milking. Milk sampling were taken once a month as a part of milk recording service. Milk samples were taken from April to August in 1^st and 2^nd observed periods. Analysis of milk samples has been performed in the certificated Central laboratory of Breeding services of the Slovak Republic (Plemenárske služby š.p. SR Bratislava).

For evaluation only ewes with minimum 3 and more sampling during each lactation within both 1^st and 2^nd periods were included into study. Thus minimum six observations were available per animal. A total of 1199 milk samples from 159 ewes (140 TS, 19 LC) were collected during 1^st observed period. From 219 ewes (130 TS, 63 LC, 26 SD) were collected 1653 milk samples during 2^nd observed period.

Impact of parity on SCC was not statistically significant in 1^st observed period (p <0.0868) but was significant in 2^nd observed period (p <0.0001). Similar results were reported by Romero et al. (2017). They found out that multiparous ewes had significant higher SCC compared with primiparous ewes (205 × 10³ cells.mL^-1 and 102 × 10³ cells.mL^-1, resp.). Also Takano et al. (2018) showed in their study that multiparous Lacaune ewes had a higher incidence of intramammary infections during early lactation than primiparous ewes. SCC were higher in multiparous than in primiparous goats (Diaz et al., 2011). The youngest ewes had the lowest SCC, while the oldest ewes showed in general the highest SCC (Arias et al., 2012). Subclinical mastitis occurred less frequently in primiparous ewes than those with two or more lactations significantly (p <0.05) and ewes on 3^rd lactation had the most cases of subclinical mastitis (Sani, Mahdavi and Moezifar, 2015).

Although the effect of parity on SCS in between 1^st and 2^nd lactation wasn´t detected in 1^st observed period, we found out the effect of parity on SCC at the level of individual farms. During 1^st observed period we detected the effect of parity on SCS in farm 4^th and farm 9a^th (Table 1). Significant effects of parity on SCS during 2^nd observed period, and at farm level in farm 1^st, 3^rd, 5^th, 8^th and farm 9b^th (Table 2). Breeds didn´t have impact on change of SCS in monitored farms (Table 3). Significant differences between farms with the same breed could indicate the effect of management level on farms.

Distribution of ewes in SCC groups during 1^st observed period (2016 and 2017) was as followed: G1 (38.99%, 33.96% resp.), G2 (32.02%, 23.90% resp.), G3 cells.mL^-1 (6.92%, 10.69% resp.), G4 (6.29%, 6.29% resp.) and G5 (15.72%, 25.16% resp.). During 2^nd observed period (2017 and 2018) there were following distribution of ewes in SCC groups: G1 (57.99%, 35.16% resp.), G2 (21%, 20.09% resp.), G3 (6.39%, 9.13% resp.), G4 (3.2%, 8.68% resp.) and G5 (11.42%, 26.94% resp.). If compare changes from 1^st to 2^nd lactation in both observed periods the following changes occurred: In 1^stmonitored period there were 8.81% ewes in SCC group with ˂200 × 10³cells.mL^-1 during 1^st lactation which moved into SCC groups ≥600 × 10³ cells.mL^-1 during 2^nd lactation. Even 6.92% from these mentioned ewes moved into SCC group ≥ 1000 × 10³ cells.mL^-1. In 2^nd observed period 15.53% of ewes were in SCC group with ˂200 × 10³ cells.mL^-1 during 1^st lactation, which moved into SCC groups ≥600 × 10³ cells.mL^-1 in the following lactation. Even from these ewes 10.96% moved into SCC group ≥ 1000 × 10³ cells.mL^-1.

These changes from 1^st to 2^nd lactation among SCC groups and clear increase of percentage of samples in SCC group ≥1000 × 10³ cells.mL^-1 in 2^nd lactation indicate higher prevalence of subclinical mastitis. Persson et al. (2017) detected significant association between intramammary infection and high SCC in ewes. In contaminated samples were significantly higher SCC as compared with uncontaminated milk samples (Ozenc et al., 2011). From preliminary results of Tančin et al. (2018) there was shown that high SCC in milk samples were associated with presence of pathogens. Romero et al. (2017) observed significant higher SCC in milk of primiparous and multiparous ewes with mastitis. Early diagnosis and treatment of subclinical mastitis can significantly eliminate clinical forms of mastitis (Zigo et al., 2017).

Data shown in Figure 1 and Figure 2 represent examples of frequency of distribution of ewes from one farm with machine milking and another farm with hand milking during their 1^st and 2^nd lactation. On both figures there are presenting changes of udder heath from 1^st to 2^nd lactation by clear demonstration of difference between count of ewes in SCC group <200 × 10³ cells.mL^-1 and in SCC group ≥1000 × 10³ cells.mL^-1. In both farms during the 2^nd lactation there was a decrease in the distribution of ewes in the SCC group <200 × 10³ cells.mL^-1 regardless on the milking technique. Increase of percentage of ewes in SCC groups ≥1000 × 10³ cells.mL^-1 could be due to the increase prevalence of subclinical mastitis in these farms. In other study Marogna et al. (2010) found out that hand milking was associated with 62% higher risk of bacterial positive samples compared to machine milking which we did not confirmed in our study. Marogna et al. (2010) also observed that machine milking with portable devices was associated with 40% higher risk of bacterial positive samples compared to machine milking with fixed plants. Queiroga (2017) detected significantly higher prevalence of subclinical mastitis in herds with machine milking than those with hand milking (p <0.0001). Vasileiou et al. (2018) reported increased prevalence of mastitis in farms with hand milking.

In conclusion, high percentage of ewes had SCC <200 × 10³ cells.mL^-1 during 1^st lactation only. During 2^nd but not during 1^st observed period the ewes on 2^nd lactation had higher SCC compared with primiparous ewes, however, clear individual farm effect was recorded in both observed periods. Also significant effect of farm management on SCC was demonstrated without connection to hand or machine milking. Thus the level of management in dairy farm has to be considered.