Prevalence of contagious mastitis pathogens in bulk tank milk in Québec

Abstract

The objective of this study was to estimate the prevalence of mycoplasma, Staphylococcus aureus, and Streptococcus agalactiae in bulk tank milk (BTM) in Québec dairy herds. BTM was sampled 3 times a month in 117 randomly selected dairy herds. Samples were submitted for S. aureus, S. agalactiae, and mycoplasma and for direct mycoplasma detection by polymerase chain reaction (PCR). Mycoplasma spp. was identified at least once in 3 herds (2.6%) by primary culture and/or PCR and in 4 herds (3.4%) by enrichment culture and/or PCR. Staphylococcus aureus was isolated at least once in 99 (84.6%) and 112 (95.7%) herds in primary culture and after enrichment, respectively. Streptococcus agalactiae was isolated at least once in 9 (7.7%) and 10 (8.6%) herds in primary culture and after enrichment, respectively. Herd prevalence of mycoplasma was similar to that previously reported in Canada. Staphylococcus aureus is still by far the most important contagious mastitis pathogen.

Introduction

Despite the effort involved in the control and prevention of mastitis by the implementation of udder health programs, mastitis is still the most expensive disease in dairy production. However, control programs clearly have a positive impact on bulk milk somatic cell count and prevalence of contagious mastitis pathogens (1–4). Detection of changes in the distribution of pathogens over time requires regular national or provincial monitoring of herd prevalence.

The main contagious mastitis pathogens for which the primary reservoir is the cow’s udder are Staphylococcus aureus, Streptococcus agalactiae, and Mycoplasma spp. (5). Prevalence of mycoplasma mastitis has been reported to have increased in various parts of the world (6–9). However, individual or bulk tank milk (BTM) prevalence data are frequently not available and may vary greatly between countries. In Canada, only 1 study is available on the prevalence of mycoplasma species in BTM. This study was performed in Prince Edward Island in 2004, and mycoplasma was cultured at least once in BTM samples from 5 of 258 herds (1.9%) (10).

Since Mycoplasma spp. are highly sensitive to freezing and storage (11–13), only fresh bulk milk samples can be used for mycoplasma culture. Based on this consideration, a national prevalence study may be difficult to implement and Olde Riekerink et al (14) recommended that herd-level prevalence studies of Mycoplasma spp. should be carried out regionally. The use of polymerase chain reaction (PCR), however, may be an excellent alternative to culture when storage for a long period may be expected. Pinnow et al (15) reported that the sensitivity of a nested PCR for detection of M. bovis in milk frozen for 2 y was 100% in comparison to 27% for culture.

The main objective of this study was to estimate the prevalence of mycoplasma species in BTM in Québec using conventional mycoplasmal culture and a conventional PCR. Secondary objectives were to estimate the prevalence S. aureus and S. agalactiae in BTM in Québec and to compare these results to those previously reported in Canada.

Materials and methods

Study population

Herds included in the study were randomly selected and stratified over the 17 administrative regions of the province of Québec. Herds were selected from those registered in the «Programme de contrôle de la qualité et de l’innocuité du lait cru prélevé à la ferme». All commercial dairy herds in Québec are registered in this program, which represented 6628 herds at the beginning of the study (February, 2009). The sample size was calculated using the sample size estimate of the Win Episcope 2.0 software (16). Based on an estimated mycoplasma bulk tank milk prevalence between 2% to 5%, an allowable error of 4%, and a confidence level of 95% a sample of 48 to 115 herds was required. Finally, a total of 125 herds were randomly selected and 117 agreed to participate in the study.

Sample collection

Three fresh bulk milk samples were collected monthly from the 117 herds in the study. Samples were taken by the herds’ veterinary practitioners as previously described (17). The milk in the bulk tank was agitated for 5 to 10 min, then, approximately 9 mL of milk was taken from the top of the bulk tank using a sterile syringe and pipette. The sample was placed in a 13 mL screw cap sterile plastic tube and transported at ~5°C to the provincial laboratory.

Laboratory analysis

Milk samples submitted to the provincial laboratory were divided into 2 aliquots. The first one was used for culture of S. aureus, S. agalactiae, and mycoplasma at the provincial laboratory. The second one was refrigerated and submitted to the molecular diagnostic laboratory of the Faculté de médecine vétérinaire de l’Université de Montréal for direct mycoplasma detection by PCR.

Mycoplasma culture

An inoculum of 10 μL of milk was spread on a modified Hayflick’s agar plate (primary culture) and after this inoculation, 0.2 mL of milk was added to 2 mL of modified Hayflick’s broth for Mycoplasma spp. enrichment. The broth with pH indicator was incubated at 35°C aerobically for 48 h before it was spread on modified Hayflick’s agar plates. All agar plates were incubated at 35°C in a candle jar for up to 7 d. The plates were observed daily for the typical fried egg appearance of mycoplasma colonies. Suspected colonies on agar plates were identified by immunofluorescence (18,19) at the Laboratoire d’expertise en pathologie animale du Québec du Ministère de l’Agriculture, des Pêcheries et de l’Alimentation du Québec.

Mycoplasma detection by polymerase chain reaction (PCR)

Extraction of total DNA in bulk tank milk was performed with a commercial kit (QIAamp DNA mini kit, QIAGEN, Montreal) at the molecular diagnostic laboratory of the Faculté de médecine vétérinaire de l’Université de Montréal. Two PCR amplification protocols were used in this study, one for the detection of Mycoplasma spp. and one for the specific detection of Mycoplasma bovis, as previously described (20,21).

Staphylococcus aureus and Streptococcus agalactiae cultures

An inoculum of 10 μL of milk was spread on a Columbia agar plate supplemented with 5% sheep blood (primary culture). Additionally, milk samples were frozen at −20°C for 24 h before inoculation on 5% sheep blood agar plates. The frozen samples were also incubated aerobically for 24 h at 35°C before inoculation as described above. After inoculation, all plates were atmosphere and examined 24 h incubated at 35°C in 5% CO₂ and 48 h after incubation. Staphylococcus aureus was identified by Gram stain, positive catalase test, α- and β-hemolysis, and positive tube coagulase test. Streptococcus agalactiae was identified by Gram stain, negative catalase test, negative esculin hydrolysis, positive CAMP test, and positive reaction to Lancefield group B antiserum.

Statistical analysis

Herd prevalence was calculated for Mycoplasma spp., S. aureus, and S. agalactiae. A herd was considered positive when at least 1 BTM was positive. Apparent prevalence and 95% exact confidence intervals were calculated using the results of primary culture; a second calculation was made using the results of culture after enrichment. A Chi-squared test was used to determine herd prevalence heterogeneity between administrative regions. Statistical analyses were performed using SAS v9.1 (SAS, Cary, North Carolina, USA).

Results

In 2009, 6487 dairy herds with an average herd size of 56.2 cows and 364 000 cows were listed in Québec (22,23). The average herd size of the 117 participating herds was 56.0 cows (SD: 34.5) with a total of 6544 cows. In total 351 samples originating from these 117 herds were analyzed. Culture was performed within 24 h for 298 samples (84.9%), between 24 and 48 h for 52 samples (14.8%) and at 96 h for 1 sample, which was negative by culture and PCR. Delay between sampling and PCR analysis ranged from < 72 h for 331 samples (94.3%), between 96 and 120 h for 16 samples (4.6%) and between 6 to 7 d for 4 samples.

Mycoplasma spp. was identified at least once by primary culture and/or PCR in 3 herds [2.6%; 95% confidence interval (CI): 0.5% to 7.3%] and in 4 herds (3.4%; 95% CI: 0.9% to 8.5%) by enrichment culture and/or PCR (Table 1). Two of the 4 positive herds were positive for M. bovis. Mycoplasma culture was positive once in each of 2 herds. Mycoplasma spp. PCR was positive once in each of 3 herds and specific M. bovis PCR was positive once in each of 2 herds. In all herds considered positive by culture or PCR, only 1 of the 3 consecutive samples was positive. The positive herds originated from 2 administrative areas.

Table 1.

Detection of Mycoplasma in bulk tank milk for 4 herds that were positive by culture or PCR in 3 consecutive months

		Month 1			Month 2			Month 3
	Administrative area	Culture	Mspp PCR	Mb PCR	Culture	Mspp PCR	Mb PCR	Culture	Mspp PCR	Mb PCR
Herd 58	Chaudière-Appalaches	—	−	−	−	+	+	−	−	−
Herd 63	Chaudière-Appalaches	M. bovis	+	+	−	−	−	−	−	−
Herd 71	Chaudière-Appalaches	A. laidlawiia	−	−	−	−	−	−	−	−
Herd 91	Montérégie	—	−	−	−	−	−	−	+	−

^aAcholeplasma laidlawii, culture after enrichment.

Mssp — Mycoplasma ssp.

Mb — Mycoplasma bovis.

Staphylococcus aureus was isolated at least once in 99 and 112 herds in primary culture or after enrichment, respectively (Table 2). Thirty-four and 95 herds were positive on 3 consecutive samples in primary culture and after enrichment, respectively (Table 3). Each administrative area had at least 50% positive herds except in Abitibi-Témiscamingue for which no positive herds were found in primary culture (0 of 3 herds) (Table 4). Estimated herd level prevalence differed by administrative area only for primary culture results (P = 0.006).

Table 2.

Québec bulk milk samples (117 herds) that were culture-positive for S. aureus and S. agalactiae in 3 consecutive months

	Month 1 Herds (%)	Month 2 Herds (%)	Month 3 Herds (%)	Cumulative prevalence Herds (%; CI*)
Staphylococcus aureus
Primary culture	74 (63.2%)	64 (54.7%)	71 (60.7%)	99 (84.6%; 76.8–90.6)
Enrichment	107 (91.4%)	103 (88.0%)	106 (90.6%)	112 (95.7%; 90.3–98.6)
Streptococcus agalactiae
Primary culture	5 (4.3%)	4 (3.4%)	7 (6.0%)	9 (7.7%; 3.6–14.1)
Enrichment	5 (4.3%)	5 (4.3%)	8 (6.8%)	10 (8.6%; 4.2–15.2)

^*95% exact confidence interval (CI).

Table 3.

Frequency of S. aureus and S. agalactiae isolation in 3 successive milk samples

	0 out of 3 times Herds (%)	1 out of 3 times Herds (%)	2 out of 3 times Herds (%)	3 out of 3 times Herds (%)
Staphylococcus aureus
Primary culture	18 (15.4%)	23 (19.7%)	42 (35.9%)	34 (29.0%)
Enrichment	5 (4.3%)	3 (2.5%)	14 (12.0%)	95 (81.2%)
Streptococcus agalactiae
Primary culture	108 (92.3%)	5 (4.3%)	1 (0.8%)	3 (2.5%)
Enrichment	107 (91.4%)	5 (4.3%)	2 (1.7%)	3 (2.5%)

Table 4.

Proportion of herds positive for S. aureus and S. agalactiae by administrative area

	% of herds positive
	S. aureus		S. agalactiae
	Primary culture*	Enrichment	Primary culture	Enrichment
Bas-Saint-Laurent	90.9 (10/11)	100.0 (11/11)	0.0 (0/11)	0.0 (0/11)
Saguenay-Lac-Saint-Jean	80.0 (4/5)	100.0 (5/5)	20.0 (1/5)	40.0 (2/5)
Capitale-Nationale	50.0 (2/4)	75.0 (3/4)	0.0 (0/4)	0.0 (0/4)
Mauricie	83.3 (5/6)	100.0 (6/6)	0.0 (0/6)	0.0 (0/6)
Estrie	80.0 (8/10)	100.0 (10/10)	0.0 (0/10)	0.0 (0/10)
Outaouais	100.0 (2/2)	100.0 (2/2)	0.0 (0/2)	0.0 (0/2)
Abitibi-Témiscamingue	0.0 (0/3)	66.7 (2/3)	0.0 (0/3)	0.0 (0/3)
Chaudière-Appalaches	96.0 (24/25)	100.0 (25/25)	8.0 (2/25)	8.0 (2/25)
Lanaudière	100.0 (5/5)	100.0 (5/5)	0.0 (0/5)	0.0 (0/5)
Laurentides	100.0 (4/4)	100.0 (4/4)	0.0 (0/4)	0.0 (0/4)
Montérégie	79.2 (19/24)	91.7 (22/24)	16.7 (4/24)	16.7 (4/24)
Centre-du-Québec	88.9 (16/18)	94.4 (17/18)	11.1 (2/18)	11.1 (2/18)
Total	84.6 (99/117)	95.7 (112/117)	7.7 (9/117)	8.5 (10/117)

^*Apparent herd prevalence differed between administrative areas (P = 0.006).

Streptococcus agalactiae was isolated at least once in 9 and 10 herds in primary culture and after enrichment, respectively (Table 2). Three herds were positive on 3 consecutive samples (Table 3). Positive herds were found in only 4 administrative regions (Table 4).

Discussion

To our knowledge, this is the first report of the prevalence of mycoplasma in BTM in Québec. The apparent herd-prevalence of Mycoplasma spp. infection observed in this study is in agreement with those previously reported in other provinces in Canada in which infection ranged from 1.9% in Prince Edward Island (10) to 3% in Ontario (12). Mycoplasma bovis was identified in 2 herds in this study (1.7%) and was also the main mycoplasma species isolated in the Ontario and Prince Edward Island studies with herd prevalences of 2.4% and 1.5%, respectively (10,12). Data from the USA seemed to show a higher prevalence of mycoplasma in BTM but with a wide variation between States. In the NAHMS Dairy 2002 study (24), 7.9% of 871 dairy herds were culture-positive for Mycoplasma spp. on a single BTM sample. The prevalence was higher in the western States (9.4%) than in the southeastern (6.6%), northeastern (2.8%) and midwestern (2.2%) States. Two other studies on consecutive BTM cultures reported prevalence rates of 20% in the Northwest (monthly BTM culture between 1998 and 2000) (25) and 7% in Utah (5 consecutive BTM cultures at 3–4 day intervals) (26).

Mycoplasma bovis and Acholeplasma laidlawii were the 2 bacteria of the class Mollicutes identified in this study. Twelve Mycoplasma and Acheloplasma species have been reported to be associated with natural or experimental cases of mastitis in cattle (7). Among them, M. bovis is considered the most pathogenic and the most frequently isolated in North America. The pathogenicity of A. laidlawii is controversial. Researchers have suggested the possibility of nonpathogenic and pathogenic strains (27–29). This mollicute is considered a contaminant that can be found in milk especially during rainy months (7,27). Additionally, Gonzalez and Wilson (7) do not recommend the use of pre-enrichment for BTM since that increases the rate of isolation of A. laidlawii. Consequently, it is difficult to interpret the significance of A. laidlawii isolation in the present study.

Two techniques (culture and PCR) were used for the determination of the prevalence of mycoplasma in BTM in this study. The apparent herd prevalences of mycoplasma are 1.7% (2/117) and 2.5% (3/117) when detected by culture and PCR, respectively. To our knowledge, only 2 other studies reported the simultaneous use of culture and PCR for the determination of the prevalence of M. bovis in BTM; one was in France (30) and the other was in New Zealand (31). In these studies, no mycoplasma was detected by either culture or PCR making comparison between these two techniques difficult. Cai et al (12) reported an almost perfect agreement between real-time PCR and culture performed simultaneously on mastitis milk samples. In the present study, 2 BTM samples were PCR positive and culture negative. This could be explained by the lack of sensitivity of the culture or because the mycoplasma were no longer viable in the sample. One sample, however, was positive by enrichment and negative on Mycoplasma spp. PCR possibly because the PCR used cannot detect Acholeplasma or because PCR inhibitors were present in the sample. Sensitivity may be an issue when conventional PCR is used directly on clinical samples; however, PCR utilizing a DNA extraction or real-time PCR may have detection limits as low as 10 cfu/mL compared with 100 cfu/mL for conventional culture (12,32,33). Moreover, PCR is less affected by freezing or storage since it can detect both viable and non-viable organisms (12,15). The PCR used in this study also has the advantage of being less expensive ($17.50 CAN for 1 PCR) than culture ($26.00 CAN for culture and $24.50 CAN for identification). However, the main disadvantage of PCR is that it can only detect and identify mycoplasma species for which it has been developed (M. bovis in this study) (12).

The apparent S. aureus and S. agalactiae herd prevalences reported herein are higher than those reported in a recently published Canadian study (74%; 95% CI: 61% to 86% for S. aureus) (14) and in another Canadian province (74% and 1.6%, respectively) (10) as well as those reported in the USA (31 to 37% and 0 to 10%, respectively) (34,35) and in Mexico (30% and 0%, respectively) (36). The findings are similar to the S. aureus herd prevalence estimated when applying a statistical-prediction model (100%; 95% CI: 80 to 100%) in the other Canadian provincial study (10) and the S. agalatiae prevalence reported in Pennsylvania (10%) (34). However, comparison between results from different prevalence studies must be performed with caution since sample selection and bacterial cultivation techniques may differ.

The apparent S. aureus herd prevalence presented here is similar to that reported in Québec in 1992 (93.4%) (37), suggesting that no improvements have been achieved over the past 20 years, and that S. aureus is still the most frequently isolated contagious mastitis pathogen in Québec and important improvements can still be achieved. On the other hand, the S. agalactiae herd prevalence is considerably lower than that previously reported in Québec in 1992 (43%). These results support the statement of Olde Riekerink et al (14) that S. agalactiae mastitis is on its way to being eradicated in Canada. However, differences observed between provinces demonstrate that efforts should be maintained and additional work should be done to achieve this goal.

Herd prevalences reported here are apparent prevalences. True herd prevalence, defined as herds that have contagious pathogen-infected udders, can only be determined if the sensitivity and the specificity of testing BTM are known. The sensitivity of a single BTM culture is low, between 21% and 54% for S. aureus and S. agalactiae and between 33% and 59% for mycoplasma. However, if consecutive samples are taken sensitivity will increase. The specificity of the test for the 3 pathogens can be assumed to be close to 100%. True prevalence is therefore probably higher than the apparent prevalence reported here.

Various factors may influence the detection of a contagious mastitis pathogen in BTM if only 1 cow was positive in the herd. In BTM, milk from infected mammary glands or cows is diluted by milk from uninfected mammary glands; therefore, it can be expected that the bacterial concentration or CFU/mL may be too low to be detected by culture if more cows contribute to the BTM (i.e., increased herd size and daily milk production). The shedding pattern, and particularly the number of bacteria shed by an infected mammary gland, however, may attenuate or amplify the dilution effect. Streptococcus agalactiae or mycoplasma infected mammary glands are reported to shed a large number of bacteria compared with S. aureus (17,38,39). Gonzalez et al (39) reported that variation in the rate of bacterial shedding in affected mammary gland could explain the different sensitivity values for BTM cultures observed for the detection of S. agalactiae infection in a herd. As an example, Biddle et al (40) evaluated that in Washington State (average herd size 330 cows, average daily milk production per cow 33 L) BTM culture would not detect mycoplasma infection from a dairy herd with a single infected cow approximately 39% of the time. The inoculum volume also affects the number of bacteria present in the inoculum and consequently the ability to detect infection (38).

Results of this study demonstrate that the herd prevalence of Mycoplasma spp. infection is similar to prevalences previously reported in other Canadian provinces, but lower than those reported in other parts of North America. The use of 2 techniques (culture and PCR) has increased the detection of the number of positive herds. However, since few samples were positive by either culture or PCR, comparison between these 2 techniques was difficult. As in other countries, S. aureus is the most prevalent contagious mastitis pathogen in Québec and herd prevalence has not decreased during the past 20 years. On the other hand, S. agalactiae herd prevalence has declined considerably during the same period.