Comparison of types and antimicrobial susceptibility of Staphylococcus from conventional and organic dairies in west-central Minnesota, USA


Timna J.O. Wyckoff, Division of Science and Mathematics, University of Minnesota, Morris, 600 East 4th Street, Morris, MN 56267, USA. E-mail:


Aims:  To assess whether conventional and organic dairy management practices are associated with differences in the susceptibility of Staphylococcus to antimicrobial agents.

Methods and Results: Staphylococcus was isolated from milk samples collected from conventional and organic dairies in west-central Minnesota. Isolates were categorized as (1) coagulase-positive, (2) novobiocin-sensitive coagulase-negative or (3) novobiocin-resistant coagulase-negative. Novobiocin-resistant coagulase-negative Staphylococcus (CNS) was more common on conventional farms and novobiocin-sensitive CNS predominated the isolates from organic farms. Overall, a larger proportion of isolates from organic rather than conventional farms were susceptible to erythromycin, pirlimycin and tetracycline. However, for pirlimycin and tetracycline, different patterns of susceptibility were observed among Staphylococcus categories.

Conclusion:  In this study, organic dairy management was associated with more overall antimicrobial susceptibility among Staphylococcus than was conventional management. However, different patterns of susceptibility among Staphylococcus categories suggest that multiple management practices, including some unrelated to antimicrobial use, may contribute to the observed differences in susceptibility.

Significance and Impact of the Study:  This study adds to our understanding of the implications of dairy management choices.


Bacterial resistance to antimicrobial agents is a leading healthcare concern for the 21st century. As resistance is an expected result of antimicrobial usage, we must scrutinize and evaluate every application of antimicrobial agents to keep these drugs effective (Levy 2002). A large fraction of antibiotics are used in livestock production; therefore, agricultural practices make potentially important contributions to the overall development of antimicrobial resistance (FAO/OIE/WHO 2003; Wegener 2003). Small US dairies typically use few growth-promoting antimicrobial agents (USDA 2005), and thus offer an opportunity for focused study of the effect of routine therapeutic and prophylactic antimicrobial usage on the development of resistance.

The use of antimicrobial agents in US livestock production is widespread but not universal. US national organic livestock production standards forbid the use of antimicrobial agents; when these drugs must be used to save an animal, the treated animal is subsequently culled from the organic herd (, accessed 15 January 2007). Organic livestock operations therefore offer a useful setting for comparisons of bacterial antimicrobial resistance in environments with differing antimicrobial usage.

A major use of antimicrobial agents on dairies is for the treatment and prevention of mastitis (Zwald et al. 2004). As a result, contagious mastitis pathogens, such as Streptococcus spp. and coagulase-positive Staphylococcus (CPS) species (e.g. S. aureus), might bear the brunt of the selective pressure towards resistance on conventionally managed dairies. Coagulase-negative Staphylococcus (CNS) can also be important mastitis pathogens and are more prevalent udder colonizers than Saureus (Matthews et al. 1992). Therefore, CNS could potentially provide a reservoir of resistance genes for the more contagious pathogens.

Pol and Ruegg (2007) recently reported higher minimal inhibitory concentrations (MIC) for some antibiotics among CNS from dairies where those antibiotics are used. Studies comparing antimicrobial susceptibility of S. aureus and other contagious mastitis pathogens from conventional and organic dairies have reported mixed results – from significant differences between farm type for several antimicrobial agents in north-eastern US dairy herds (Tikofsky et al. 2003) to no significant differences between farm type for herds in Denmark and Switzerland (Bennedsgaard et al. 2006; Roesch et al. 2006). Interestingly, several recent studies have reported less resistance to antimicrobial drugs among faecal pathogens isolated from organic dairies when compared with their conventional counterparts (Sato et al. 2005; Halbert et al. 2006; Ray et al. 2006).

In this study, we isolated Staphylococcus from milk samples collected from healthy cows on conventional and organic dairies in west-central Minnesota, USA. We categorized these isolates as (1) coagulase-positive Staphylococcus, (2) novobiocin-sensitive CNS (NSCNS) or (3) novobiocin-resistant CNS (NRCNS). We then compared the antimicrobial susceptibility profiles of each category of Staphylococcus from conventional and organic farms to evaluate whether organic and conventional management practices are associated with differences in antimicrobial susceptibility of various Staphylococcus categories.

Materials and methods


All microbiological media, additives and susceptibility test discs were from Becton, Dickinson and Company (Sparks, MD, USA) with the exception of the pirlimycin discs, which were from REMEL (Lenexa, KS, USA). Chemicals were from Sigma-Aldrich (St Louis, MO, USA) and Fisher Scientific ( Rabbit plasma was from Lampire Biological Laboratories (Pipersville, PA, USA).

Farm visits

Owners of 16 small dairies (20–100 cows) were identified and interviewed about their use of antimicrobial agents in compliance with protocols approved by the University of Minnesota Institutional Review Board. Eight of the dairies had been certified for at least 1 year under the standards of the USDA National Organic Program (, accessed 15 January 2007) and are termed ‘organic’ in this study. The other eight dairies are termed ‘conventional’ because they were neither certified nor seeking to follow organic management practices. Composite milk samples were collected from all healthy cows at each dairy according to National Mastitis Council guidelines (National Mastitis Council 1999) and in compliance with protocols approved by the University of Minnesota Institutional Animal Care and Use Committee. Sample collection took place between September 2004 and September 2005. Briefly, after teats were cleaned and stripped by the farm owners, laboratory personnel wiped each teat with a fresh alcohol wipe and collected several millilitres of milk from each lactating teat.

Staphylococcus isolation and identification

Milk samples were stored at 4°C and plated within 12 h of collection. To select for Staphylococcus, 10 μl of each milk sample was plated to mannitol salt agar plates and incubated at 35°C for 24 h. A representative colony from each plate showing presumptive Staphylococcus growth was restreaked and single colonies from the new plates were confirmed as Staphylococcus by Gram stain (Gram-positive cocci), catalase testing (positive) and thioglycollate growth pattern (facultative and nonmotile). Confirmed Staphylococcus isolates were identified as coagulase-positive or coagulase-negative by a tube coagulase test. The novobiocin susceptibility of coagulase-negative isolates was determined by disc diffusion (see below).

Antimicrobial susceptibility testing

Antimicrobial susceptibility testing was performed by agar disc diffusion and interpreted by breakpoints according to guidelines established by the Clinical and Laboratory Standards Institute (formerly NCCLS) for bacteria isolated from animals (NCCLS 2002). Briefly, Staphylococcus from agar culture was used to inoculate Mueller-Hinton broth to match a 0·5 McFarland turbidity standard. This broth was used to inoculate a Mueller-Hinton agar plate for disc diffusion. Susceptibility to cefoxitin [30 μg; for determination of methicillin-resistance (Velasco et al. 2005)], cephalothin (30 μg), erythromycin (15 μg), novobiocin (5 μg; for identification purposes only), penicillin (10 IU), pirlimycin (2 μg), tetracycline (30 μg) and vancomycin (30 μg) was determined.

Statistical analyses

Milk samples were identified as bacteriologically positive or negative for Staphylococcus. Staphylococcus isolates were identified as susceptible or not susceptible (intermediate or resistant) to each antimicrobial agent. Chi-squared analysis was used to compare populations, and the Rao–Scott second-order correction was used to account for clustering by farm (Rao and Scott 1992). The statistical analysis was performed using the ‘svydesign’ and ‘svychisq’ functions within the r software package ( We report P-values >0·05 as not significant (n.s.).


The profile of Staphylococcus species differs significantly by farm type

Milk samples were collected from 339 conventionally managed cows and 501 organically managed cows on sixteen total dairies (eight of each management type). Owners of organically managed cows confirmed that they had not used antimicrobial agents for at least 4 years (at least 1 year of certification and 3 years of transition). Owners of conventionally managed cows all reported usage of several antimicrobial drugs in the past year: cephalosporins (seven owners), penicillins (six owners), tetracyclines (five owners) and pirlimycin (five owners). Five owners of conventionally managed cows reported practicing dry cow therapy, the administration of antimicrobial drugs to all cows at the end of lactation. Staphylococcus species were isolated from approximately half of the milk samples from both types of farm [conventional = 166 (49·0%); organic = 239 (47·7%); n.s.].

To further categorize our Staphylococcus isolates, all were tested for their ability to coagulate rabbit plasma and their susceptibility to novobiocin, a characteristic that can be used to subcategorize CNS of bovine origin (White et al. 1989; National Mastitis Council. 1999). The proportion of total Staphylococcus isolates falling into each of these three categories – CPS, NRCNS and NSCNS – differed strikingly by farm type (Table 1). NRCNS predominated the isolates from conventionally managed cows while NSCNS predominated those from organic farms. The number of CPS isolates was low on both types of farm.

Table 1.   Proportion of total isolates from conventional and organic dairies from each category of Staphylococcus
Category of StaphylococcusNo. (%) of total StaphylococcusP-value*
  1. CPS, coagulase-positive Staphylococcus; NSCNS, novobiocin-sensitive coagulase-negative Staphylococcus; NRCNS, novobiocin-resistant coagulase-negative Staphylococcus; n.s., not significant.

  2. *Based on Rao–Scott second-order correction to Pearson’s chi-squared statistic.

CPS9 (5·4)27 (11·3)n.s.
NSCNS55 (33·1)155 (64·9)0·004
NRCNS102 (61·5)57 (23·8)<0·001
Total Staphylococcus166 (100)239 (100) 

Isolates from conventional and organic farms differ in their susceptibility to some antimicrobial agents

All Staphylococcus isolates were tested for susceptibility to several antimicrobial agents. Results obtained with reference strain Staphylococcus aureus ATCC 25923 fell within established quality control ranges for all agents tested. All isolates were susceptible to cephalothin, cefoxitin and vancomycin. The proportions of total Staphylococcus isolates susceptible to erythromycin, penicillin, pirlimycin and tetracycline are shown in Table 2. A larger proportion of isolates from organic farms were susceptible to pirlimycin and tetracycline compared with those from conventional farms. Susceptibility to erythromycin and penicillin did not differ significantly by farm type.

Table 2.   Comparison of antimicrobial susceptibility of Staphylococcus isolates from conventional and organic dairies
AgentNo. (%) of isolates susceptibleP-value*
  1. n.s., not significant.

  2. *Based on Rao–Scott second-order correction to Pearson’s chi-squared statistic.

Erythromycin129 (77·7)213 (89·1)n.s.
Penicillin135 (81·3)202 (84·5)n.s.
Pirlimycin143 (86·1)235 (98·3)<0·001
Tetracycline136 (81·9)220 (92·1)0·044

The susceptibility profiles of various categories of Staphylococcus differ by antimicrobial agent

When susceptibility is broken down by category of Staphylococcus, some interesting patterns emerge. Isolates within both CNS categories from organic farms were more likely to be susceptible to pirlimycin than CNS from conventional dairies (Table 3, Panel A). However, NSCNS isolates were more likely to be susceptible to tetracycline than NRCNS isolates for both farm types; no difference in tetracycline susceptibility was seen between farm types within either CNS category (Table 3, Panel B). Significant patterns do not emerge for erythromycin or penicillin susceptibility when broken down in this way (data not shown).

Table 3.   Antimicrobial susceptibility of each category of coagulase-negative Staphylococcus from conventional and organic dairies*
Category of StaphylococcusNo. (%) of isolates susceptibleP-value† (conventional vs organic)
  1. NSCNS, novobiocin-sensitive coagulase-negative Staphylococcus; NRCNS, novobiocin-resistant coagulase-negative Staphylococcus; n.s., not significant.

  2. *Coagulase-positive Staphylococcus is not included in this comparison. All were susceptible to pirlimycin and only a single isolate from an organic farm was not susceptible to tetracycline.

  3. †Based on Rao–Scott second-order correction to Pearson’s chi-squared statistic.

Panel A. Pirlimycin
 NSCNS50 (90·9)152 (98·0)0·025
 NRCNS84 (82·4)56 (98·2)0·012
 P-value†  (NSCNS vs NRCNS)n.s.n.s. 
Panel B. Tetracycline
 NSCNS55 (100)149 (96·1)n.s.
 NRCNS72 (70·5)45 (78·9)n.s.
 P-value†  (NSCNS vs NRCNS)0·0020·005 


Our data show that organic and conventional management practices are associated with significantly different profiles of Staphylococcus in dairy cows (Table 1). Conventional cows were colonized predominately with NRCNS, while nearly two-thirds of the Staphylococcus isolates from organic cows were NSCNS. A previous study of conventionally raised dairy cows found more NRCNS in the udders of animals kept in confinement and more NSCNS in those kept on pasture (White et al. 1989). This may be because, for much of the day, cows in confinement may be exposed to bedding and hay often containing environmental Staphylococcus, which are predominantly NRCNS (Matos et al. 1991). Organic management requires that livestock be kept on pasture for parts of each day (weather permitting), so this management difference may contribute to the different profiles of CNS in our study.

Matos et al. (1991) also found more S. aureus on pastured dairy cows than those in confinement, perhaps explained by increased lesions from fly bites. While our sample size did not reveal a statistically significant difference in CPS proportions between farm types, the proportion of CPS from organic farms was twice that from conventional farms (Table 1). In addition to differences in pasture access, five of the eight conventional dairy owners self-reported practicing dry cow treatment (antimicrobial prophylaxis against mastitis); this may have contributed to the difference in CPS occurrence.

All of the Staphylococcus isolates in this study were susceptible to vancomycin and methicillin (as tested by cefoxitin susceptibility), antimicrobial agents important to human medicine, but not used on any of the dairies in this study. Interestingly, all isolates were also susceptible to cephalothin, although all but one of the owners of conventionally managed cows reported using cephalosporin antimicrobials in the past year. This result is similar to those of other recent studies that have found high levels of susceptibility to cephalosporins among Staphylococcus (Tikofsky et al. 2003; Rajala-Schultz et al. 2004), despite the fact that cephalosporin antimicrobials are commonly used on US dairies (Zwald et al. 2004).

Our data do show resistance to erythromycin, penicillin, pirlimycin and tetracycline among Staphylococcus isolates from both farm types (Table 2), although only some of the conventionally managed cows were exposed to penicillins, tetracyclines and pirlimycin in the past year. In addition, our data show that a significantly larger proportion of Staphylococcus from organic dairies were susceptible to pirlimycin and tetracycline than were those from conventional dairies (Table 2).

One may be tempted to conclude from this data that the cessation of antimicrobial usage and subsequent release of selective pressure towards resistance has allowed susceptible Staphylococcus to overtake the resistant strains in organic herds. Indeed, in the case of susceptibility to pirlimycin (Table 3, Panel A), an antimicrobial agent used only on dairy farms, this could be the case. On organic farms, where pirlimycin has not been used for several years, a larger proportion of isolates from both CNS categories were susceptible to pirlimycin when compared with isolates from conventional farms, where five of eight owners reported using pirlimycin. This is in agreement with the recent study by Pol and Ruegg (2007), in which they reported a correlation between pirlimycin use and increased MIC values for pirlimycin among several mastitis pathogens, including CNS.

In the case of susceptibility to tetracycline (Table 3, Panel B), an antimicrobial agent more widely used in livestock production and in human health care, however, a different pattern emerges. A larger proportion of NSCNS than NRCNS from both farm types were susceptible to tetracycline, but no difference was observed between farm types within a category of Staphylococcus. Thus, the larger proportion of tetracycline-susceptible Staphylococcus on organic farms (Table 2) correlates with the larger proportion of NSCNS on those farms (Table 1). This pattern may be due to the usage of tetracycline on five of the eight conventional farms, or may be the result of management practices unrelated to antimicrobial usage. Pol and Ruegg (2007) reported that farm type (organic, conventional-low antibiotic use or conventional-high antibiotic use in their study) was associated with MIC for tetracycline among CNS, but did not further break down that category.

In conclusion, organic and conventional dairy management practices in west-central Minnesota are associated with different Staphylococcus profiles and differences in antimicrobial susceptibility among Staphylococcus species. However, the mechanisms behind these differences remain unclear and may be the result of a combination of many management practices. These mechanisms, as well as the timeline of changes in resistance patterns, merit further study. Our laboratory plans to investigate conventional-grazing operations and transitioning dairies to address these additional questions.


This work was supported by a Grant-in-Aid of Research and Artistry from the University of Minnesota Graduate School to T.J.O.W. and University of Minnesota Undergraduate Research Opportunities Program grants to A.L.B., C.E.D., E.J.S. and L.R.S. We thank Dr Dennis Johnson for dairy advice; Justin Greiman and Ruth Patten for help with milk collection; Drs Engin Sungur and Peter Wyckoff for statistical help; and the sixteen anonymous dairy farmers for allowing us to disrupt their busy schedules.