Materials and methods
The eight sampled crossbred sows were housed at the research farm of the Institute of Animal Breeding and Husbandry of the Christian-Albrechts-University Kiel, with 120 sows in total (average parity 2.5). The sows were managed with a 28-day lactation period. One week prior to expected parturition, they were washed and confined in the cleaned and disinfected farrowing house, with eight pens per compartment. The floor in the single pens (2.47 m × 1.75 m) was partially slatted and no straw was provided.
All samples were collected during the period between the beginning of parturition and the birth of the last piglet. All sampled sows showed neither fever nor clinical signs of mastitis during the sampling period, and none developed mastitis subsequently. Skin sampling from all available glands was performed in six sows. In two sows, skin sampling was restricted because of the animal’s behavior, and only single glands were swabbed. Swab samples were taken from defined skin areas of mammary glands (10 cm in diameter), with the teat as the central point. In addition, colostrum samples from all sows were collected before the piglets started to suckle. For 19 single glands, milk sampling according to the sample scheme was not possible due to aggressive behavior of the sow or blocked access to the teat. The first streams of secretion from each teat were discarded in order to “wash out” bacteria in the distal end of the teat duct; the following streams were collected in sterile test tubes. In all, 96 skin samples (2, 8, 15, 16, 14, 14, 14, and 13 samples per sow, respectively) and 77 colostrum samples (2, 3, 12, 13, 14, 6, 14, and 13 samples per sow, respectively) were examined. Bacteriological analysis was performed.
A broad but similar spectrum of bacterial genera was isolated from both skin and colostrum samples. All 96 skin samples were bacteriologically positive, with 496 isolates in total and 5.2 isolates per sample on average. Only three skin samples, all from a single sow, were negative for Staphylococcaceae. Between one and five different species of Staphylococcaceae were isolated from the other 93 skin samples (96.9% of all skin samples), with 2.7 Staphylococcus species per sample on average. Staphylococcaceae were isolated most frequently from the skin (50.4% of all skin isolates), with Staphylococcus simulans the predominant species. Streptococcaceae were isolated from 61 of 96 skin samples (63.5% of all skin samples, 14.1% of all skin isolates), with Aerococcus viridans the dominant species. Seven species of Enterobacteriaceae were isolated from 64 of the 96 skin samples (66.7% of all skin samples, 19.4% of all skin isolates), and Escherichia coli was the dominant species.
Among the colostrum samples, 16 of 77 (20.8%) were culturally negative, while 61 of 77 (79.2%) were bacteriologically positive, with 122 isolates in total and two isolates on average per positive sample. Staphylococcaceae were isolated most frequently (54.1%) from both the colostrum and skin samples, with S simulans the dominant species. Streptococcaceae, with A viridans as the dominant species, represented 30.3% of all isolates from colostrum samples. Enterobacteriaceae were isolated from only four colostrum samples (3.9% of all colostrum isolates), with E coli the dominant species. In one colostrum sample, two morphologically different E coli isolates were identified, but no E coli was found in the corresponding skin sample. Also, in two other sows, E coli was detected in only a single gland.
The difference between the total numbers of bacterial isolates in skin and in colostral milk samples was significant, with a higher number in the skin samples. Regarding the agreement between colostrum and skin samples, in 35 of the 61 positive milk samples (57.4%), at least one bacterial species was isolated, which was also isolated from the skin sample of the corresponding teat. On isolate level, 54 of the 122 species isolated from colostrum (44.3%) were found on the corresponding teat skin as well. This correlation was especially distinct in S stimulans (13 of 54 species; 24.1%) and Staphylococcus warneri (12 of 54 species; 22.4%). However, paired analysis for each sow and teat revealed no significant relationship (P > .05) between the bacterial families in the skin flora and in colostrum flora.
This study clearly showed that several bacterial species are present both on the skin and in the colostral milk of clinically healthy sows, with Staphylococcaceae as the predominant family. Furthermore, the results of culturing the milk samples indicated that the bacterial flora differ between individual teats. To clarify a possible relationship between skin and milk flora and their potential influence on the occurrence of clinical mastitis in sows, further studies, for instance, quantitative bacteriology in healthy and diseased animals, are recommended.
• The bacterial flora on the mammary gland skin of sows consists mainly of Staphylococcaceae.
• Streptococcaceae and Enterobacteriaceae, including E coli, may be present transiently on the mammary gland skin of sows.
• Under the conditions of this study, the colostrum of many healthy sows is not free of bacteria; therefore, isolation of bacterial species from the milk of sows with clinical coliform mastitis for diagnostic purposes should be regarded against this background.
Kemper N, Preissler R. Bacterial flora on the mammary gland skin of sows and in their colostrum. J Swine Health Prod. 2011;19(2):112–115.