A total of 9592 samples of half udder milk were collected monthly throughout lactation for bacteriological and somatic cell count (SCC) study from 1322 Churra ewe lactations from seven separate flocks enrolled in the recording scheme of the National Association of Spanish Churra Breeders in the Castile-Le6n region of Spain. Statistical analyses were carried out from a mixed model with random factor half udder or ewe for repeated measures. Test of significance of fixed effects of this mixed model showed significant effects of organisms, flock, parity, lactation stage, and birth type on SCC. Special reference must be made to novobiocin-sensitive coagulase-negative staphylococci, which represented more than 50% of the isolates and which elicited SCC geometric means of around 106/ml. In addition, the analysis of 4352 monthly test-day records for milk yield, SCC, and bacteriology showed that the ewes that were uninfected and infected by minor pathogens had the lowest SCC and the highest milk yields, whereas those infected by major pathogens had high SCC and milk yield losses between 8.8 and 10.1% according to the uni- or bilateral character of the infection. Finally, ewes infected by novobiocin-sensitive coagulase-negative staphylococci elicited SCC values similar to those of infections by major pathogens and milk yield losses ranging between those caused by minor and major pathogens. As a result, emphasis should be put on prevention of subclinical mastitis, particularly mastitis caused by novobiocin-sensitive coagulase-negative staphylococci in dairy sheep herds to improve microbiological and hygienic milk quality and to minimize losses in milk yield.

Genetic parameters for milk yield, contents of fat, total protein, casein and serum protein, individual laboratory cheese yield, and somatic cell counts (SCC) were estimated from 7492 monthly test-day records of 1119 Churra ewes. Estimates were from multivariate REML using analytical gradients (AG-REML) procedures. Except for fat content, estimates for the other routinely recorded traits (milk yield, protein content, and SCC) agreed with those previously obtained in this and other dairy sheep populations. Protein content and composition had the highest heritabilities and repeatabilities. Heritabilities for protein and casein contents were very similar (0.23 and 0.21, respectively), and genetic correlation between the traits was close to unity (0.99). Accordingly, casein content is not advisable as an alternative to protein content as a selection criterion in dairy ewes; it does not have any compelling advantages and costs more to measure. Individual laboratory cheese yield (ILCY) obtained with Churra ewes had a low heritability (0.08), suggesting that potential for selection for this parameter would be possible but is not recommended. All correlations with ILCY were high and positive except for milk yield. A high SCC was accompanied by an increase in serum protein content and involved a loss in milk yield.

A total of 3119 test day records for milk yield, somatic cell count (SCC), and contents of fat, protein, casein, serum protein, lactose, and total solids were obtained weekly from 155 lactations of Churra ewes during both the suckling and milking periods. All of the variables were significantly affected by stage of lactation, number of live lambs weaned, age at parity, and parity number within age. The lactation curves for SCC and for percentages of protein, casein, fat, and total solids were related inversely to the lactation curves for milk yield and lactose percentage. The SCC curve reached its lowest point at wk 5, which coincided with the maximum milk yield. At wk 3, lactose content reached its peak, and contents of fat, protein, casein, and total solids reached their lowest points. After weaning, monthly within-lactation correlations of the variables were simulated; values were 0.584 for yield, 0.468 for SCC, 0.371 for casein percentage, 0.360 for total solids percentage, 0.350 for protein percentage, 0.342 for fat percentage, 0.270 for lactose percentage, and 0.030 for serum protein percentage. Simplified procedures based on only one milking (a.m. or p.m.) could be effective in the testing programs for milk yield and SCC, thus, reducing the economic costs of milk recordings in mammary health and breeding programs.

Between January 2006 and December 2007, a total of 4,579 test-day observations for contents of milk fatty acids (FA) were obtained from 2,218 lactations of 1,109 ewes belonging to 14 Churra dairy flocks. The 36 analyzed FA were quantified as grams per 100 g of total FA and were grouped as 18 dependent variables: 10 FA, 6 groups of FA, and 2 FA indexes. Flock, day of testing within flock (TD), ewe age, stage of lactation, and season effects contributed significantly to variations in FA. The 2 most important variation factors were flock (3 to 30% of total variance) and TD (35 to 70% of total variance). The percentage of variance explained by the TD effect for conjugated linoleic acid (CLA, C18:2 cis-9, trans-11) and linolenic acid (C18:3 cis-9, cis-12, cis-15) was particularly high: 60.7 and 68.2%, respectively. The season effect was also a very important variation factor, closely linked to feeding. The most significant seasonal variations were observed in polyunsaturated FA, with the highest values occurring in spring and summer and the lowest in winter. More specifically, CLA and linolenic acid contents were 44 and 30% higher in spring-summer than in winter. As the age of the ewe increased, the monounsaturated and polyunsaturated FA decreased and the short- and medium-chain saturated FA increased. The CLA and the CLA/C18:1 trans-11 Delta(9)-desaturase index increased significantly throughout lactation. The correlation coefficient between rumenic acid (CLA) and vaccenic acid was high (0.47) because of the precursor-product relationship via the Delta(9)-desaturase enzyme. The correlation coefficients were high between C10:0 and C12:0 (0.79), C12:0 and C14:0 (0.73), and C14:0 and C16:0 (0.29), probably because of their similar metabolic origin. Positive correlations were also obtained among the C(18) FA family. All the studied factors of FA variation would be considered as fixed effects in the statistical models used for estimation of genetic and phenotypic parameters from test-day records of commercial flocks.

Between January and July 1992, 8403 monthly test day records were obtained from 3202 ewes from 22 Churra dairy flocks in the Castilla-León region of Spain. Variables measured were milk yield, SCC, and fat and protein percentages; means were 912 ml, 1501 x 10(3) cells/ml, 6.76%, and 6.28%. Analysis of variance showed significant effects of flock, stage of lactation, parity, and type of birth on milk yield, SCC, and composition. Variability in SCC among flocks was high and ranged from 420 to 2719 x 10(3) cells/ml. During lactation, SCC and fat and protein percentages showed increases of 31.2, 37.6, and 20.3%. Twin birth ewes had higher SCC (8.6%) and yield (4.4%) than single birth ewes. However, fat and protein fell 1.7 and .8% in twin birth ewes. The parity effect was highly significant for yield, SCC, and fat; first lactation was always lower. A statistical model similar to that for dairy cattle was used to predict milk yield on the basis of flock, ewes within the flock, lactation stage, parity number, type of birth, and SCC. The model predicted 73% of the variation in milk yield.

A total of 6620 monthly test day records of SCC, milk yield, and protein percentage from single lactations of 2374 Spanish Churra ewes from 10 flocks was used to estimate genetic and environmental parameters. A subset of 4278 records containing data from healthy udders (SCC < or = 250,000 cells/ml) was also analyzed. Genetic parameters were estimated by REML using an animal model. Herd test date, parity, and lactation stage contributed significantly to variation of most variables, and birth type significantly affected milk yield only. The SCC increased markedly as parity number and stage of lactation increased. Heritabilities (+/- SE) for test day milk yield, log SCC, protein percentage, and log SCC (< or = 250,000 cells/ml) were 0.18 (0.03), 0.09 (0.02), 0.16 (0.03), and 0.03 (0.02), respectively. The corresponding repeatabilities were 0.54, 0.38, 0.38, and 0.10. Genetic correlations of log SCC with milk yield and protein percentage were -0.23 and 0.18, respectively. Phenotypic correlations were -0.15 and 0.16. Genetic and environmental reduction of SCC for dairy ewes could be achieved using practices similar to those for dairy cows. The negative genetic correlation between milk yield and SCC suggested that selection for increased milk yield alone is expected to result in a decrease in SCC.

Using the Fossomatic method (FSCC) a total of 23,003 analytical SCC observations were carried out on 6400 aliquots taken from 80 individual ewe milk samples with the objective of studying the influence of 4 preservation procedures (without preservation, potassium dichromate, azidiol, and bronopol), 2 storage temperatures (ambient and refrigeration), 10 milk ages (3,6,12, and 24h, and 2,3,4,5,7, and 9d postcollection), and two analytical temperatures (40 and 60 degrees C). In addition, each sample was analyzed with direct microscopic method (DMSCC), using 3 different stainings for each sample: methylene blue (MB), May-Grünwald-Giemsa (MGG) and Pyronin Y-methyl green (PGM). This allowed DMSCC and FSCC (at 24 h of age) to be compared. The reference DMSCC from MB staining was a reliable method in ewe milk, though more specific stainings such as MGG and PMG slightly improve the residual standard deviation for repeated SCC. Between DMSCC and FSCC, the highest coefficients of correlation (0.972 to 0.996) corresponded to preserved and refrigerated milk, and the lowest (0.708 to 0.919) to unpreserved and ambient stored aliquots. Except for the unpreserved and ambient stored aliquots, SCC values were similar in all aliquots. Under FSCC, preservation, storage and analytical temperature, milk age, and most of the interactions showed a significant effect on SCC variation. In preserved samples, logSCC values ranged between 5.67 (bronopol) and 5.62 (azidiol). The higest values (5.72) were for unpreserved milk, which showed false overestimation of SCC due to bacterial proliferation. LogSCC was higher at 60 degrees C (5.68) than at 40 degrees C (5.65). The interaction between age, preservation and storage temperature showed no cell degeneration in properly handled samples over the 9 d of study.

Records from monthly test days for milk yield, natural log-transformed SCC, and protein percentage of 10,171 records of 3832 Spanish Churra ewes in 24 herds were analyzed to estimate genetic parameters. Heritabilities and genetic and phenotypic correlations for test day records were estimated using multivariate REML with an animal model, accounting for the effects of litter size, test day within herd group, age at lambing, and month in milk. Heritabilities for test day records of milk, log SCC, and protein percentage were .34, .04, and .13, respectively. Genetic correlations of test day measures of SCC with milk yield and protein percentage were -.37 and .37, respectively; phenotypic correlations were -.05 and .19. Considered as a random effect, test day within herd accounted for 28 and 17% of the total variance of test day milk yield and test day SCC, respectively. The correlation between milk yield and SCC because of the test day within herd, the covariation factor, was estimated to be -.16. Results indicate a moderately high heritability for milk yield, a low heritability for SCC, and moderate, negative genetic relationships between measures of SCC in milk and milk yield.