Prevalence and transmission patterns of Mycoplasma bovis in comingled Holstein dairy heifers from two different parent farms

Abstract

Background

Mycoplasma bovis probably enters dairy herds when carrier animals are introduced. Comingled calves that become subclinical M. bovis carriers could promote cross-dairy transmission. A prospective cohort study in Holstein heifers from two unrelated herds (Farms A and B previously M. bovis positive and negative, respectively) comingled at a facility raising only their calves assessed: 1). prevalence of asymptomatic M. bovis infection; 2). associations between four anatomic sites (nares, eyes, ear canal, vagina; M. bovis culture with PCR confirmation). Fifteen calves per farm were enrolled every 4 months. Swabbing solutions were first collected at parent farms, thereafter monthly for 6 months, then quarterly to 21 months.

Results

Three heifers from each dairy were lost after enrollment leaving 144 heifers (72 per farm) in the analysis. On day 1, a Farm A calf vaginal sample was the single M. bovis positive. While comingled, positives increased dramatically. Days of age to first positive were not different between farms (Farm A median = 109, range 42–561 days; Farm B median = 110, range 33–404 days; P = 0.96). Overall, 125/144 heifers yielded 634 positive samples, intranasal (46.7%), ocular (25.7%), vaginal (17.8%), ear canal (9.8%). The most common combinations were eye/nose (55/634, 43.3%) and nose/vagina (21/634, 16.5%). Intranasal positives increased exponentially at 2–3 months of age, plateauing over 4–10 months, and were more frequent than non-nose (Incidence Rate Ratio 1.44, 95% CI 1.41–1.47; P < 0.001). Positive combinations involving the nose temporally lagged the intranasal alone positives, but eventually reached similar frequencies. After returning to the parent farms, frequency of intranasal positives declined more rapidly than did non-nose.

Conclusions

M. bovis was cultured from all sites, but the nose appears most critical for transmission. Once intranasal carriage escalated, virtually all calves were subsequently positive at least once at one or more sites, indicating how readily asymptomatic M. bovis can disseminate in a population of animals and potentially manifest as clinical disease later in life.

Background

Mycoplasma spp. infections occur in many areas of the world [1,2,3,4,5,6]. There are over 100 known species of mycoplasma [1], but Mycoplasma bovis (M. bovis), first described in the context of mycoplasma mastitis [7,8,9,10], is the primary mycoplasma of concern in the United States cattle industry in both bovine respiratory disease [11,12,13,14] and mastitis [2, 15,16,17,18]. The tendency of M. bovis-associated diseases to be chronic in nature can lead to economic costs/losses per case greater than many other pathogens [12, 19, 20]. Chronicity of conditions combined with poor treatment responses also has important animal welfare implications.

Mycoplasmas are very small bacteria that do not possess a cell wall and firmly attach to mucosal cell surfaces, consequently they have predilections for establishing respiratory, mammary, or genital infections [21]. Cattle can therefore present with numerous mycoplasma-associated clinical syndromes, including mastitis, conjunctivitis, otitis, arthritis, and pneumonia. Because of their simple genome and fastidious growth requirements, laboratory diagnosis of mycoplasma can be difficult [16], hence many infections might be under-diagnosed. Mycoplasmas are difficult to culture and positive identification can take days to weeks. Moreover, identification can be confused by the fact that Mycoplasma spp. other than M. bovis can cause the same clinical manifestation of disease. For example, although M. bovis is the most common cause of mycoplasma mastitis, mastitis outbreaks caused by other Mycoplasma spp. have been reported [22, 23]. Significant problems with false negative results due to sampling and handling techniques, or intermittent shedding are also recognized. Nevertheless, identification of positive animals is most commonly made using microbiological procedures. Serology and ELISA detection methods have also been developed but have some disadvantages when specifically studying the epidemiology of M. bovis infections in youngstock; the former is only indicative of exposure and multiple Mycoplasma spp. confound the latter. PCR techniques are also available with the advantages of speed [16] and mycoplasma speciation without the need for secondary tests. However, as with all PCRs, it does not confirm the presence of viable organisms.

In general, the most economically significant infections in dairy herds are those of the mammary gland [18]. Among these, M. bovis infections are contagious in nature and their control includes measures such as strict hygienic milking practices, identification of positive animals, and culling of those found to be infected [9, 10, 15, 24]. Presence of the disease on dairies is most often determined using bulk tank cultures [25], from which it is estimated that 1—8% of all herds have at least one cow with a mycoplasma infection [26,27,28]. However, the source of infection for dairy herds remains unclear. Attempts to define the epidemiology of the disease have not been entirely successful [2,3,4,5,6]. However, the disease occurs most often in large dairy herds and two factors were recognized as being most important in mycoplasma mastitis epidemiology; the introduction of diseased animals and poor milking procedures [10]. A third possible risk factor has also been identified. Mycoplasma spp. can enter dairy herds with the introduction of asymptomatic carrier animals purchased for replacements or expansion [16, 29]. Transmission from newly introduced animals with clinical disease has been documented, but it has also been shown that resident asymptomatic-carrier animals can transfer the agent [30]. It is clear from the literature that asymptomatic carriers are difficult to diagnose but continue to be a significant source of the organism on dairy herds. Moreover, the situation is further complicated by the finding that some healthy animals are dually colonized with apparently non-pathogenic Mycoplasma spp., and M. bovis. For example, M. dispar can be found in the upper respiratory tract of healthy cattle, while M. bovis is more often present in the lower respiratory tract of animals exhibiting bovine respiratory disease (BRD) than in healthy cattle [31]. However, M. bovis itself can often be present in the upper respiratory tract of cattle without clinical disease and can even be isolated from lungs of some cattle that have neither lesions nor clinical disease [32]. Because it is so well adapted to colonization of mucosal surfaces [21], regardless of the exposure route, during early infection M. bovis can be isolated from multiple body sites, particularly the upper respiratory tract, mammary gland, conjunctiva and urogenital tract [24, 33] where it can persist without causing clinical disease. Thus, while M. bovis alone can cause natural or experimentally induced clinical disease, its presence does not always lead to disease and similarly, the presence of clinical disease does not appear necessary for the maintenance and dissemination of M. bovis in the cattle population [34]. It has been recognized that asymptomatic carriers and intermittent shedding play critical roles in dissemination of mycoplasma infection between herds [35].

It is unlikely dairy herds will stop expanding, thus, to prevent establishment of mycoplasma infections, understanding which animals might be harboring infection and accurate identification of asymptomatic carriers are of vital importance to the dairy industry. If calves are infected by way of comingling at a heifer-rearing operation or from resident herd mates, they have the potential to become asymptomatic carriers, making them significant role players in the introduction and/or perpetuation of mycoplasma infections. We hypothesized that the comingling of calves at a heifer-rearing facility would be a significant source of transfer between calves with potential for introduction/reintroduction to the parent herds.

A prospective cohort study was conducted in Holstein heifers originating from two unrelated herds while comingled at a heifer-rearing facility raising only their calves; Farm A which had previously been identified as M. bovis positive and Farm B which had never been identified as positive. The objectives were to assess: 1). within and between herd prevalence of asymptomatic M. bovis infection in calves and 2). associations between positive cultures obtained from four different anatomic sites (eye, ear, nose, and vagina).

Methods

Parent herd characteristics

This study was conducted using two large dairy herds located in southeastern Pennsylvania USA milking approximately 1,500 (Farm A) and 750 (Farm B) Holstein dairy cows, respectively. The study herds were selected based on being closed herds of relatively large size, with excellent computerized records (PCDART, Dairy Records Management Systems, Raleigh, NC, USA, Farm A and DairyComp 305, Dairy One Cooperative Inc., Ithaca, NY, USA, Farm B), well-trained personnel, and willingness of the owners to comply with study requirements. Both farms were high producing herds (Farm A—rolling herd average 12,700 kg per cow annually; 3.75% fat and 3.2% protein and Farm B—rolling herd average 11,340 kg per cow annually; 3.75% fat and 3.2% protein). Farm A had previously diagnosed cases of mycoplasma infection including postmortem culture of Mycoplasma spp.; confirmed as M. bovis by PCR on postmortem lung tissue. In addition, the operation had clinical disease (otitis media, chronic pneumonia non-responsive to antimicrobials, severe conjunctivitis) and bulk tank positives – all PCR confirmed. Overall, compelling evidence of Mycoplasma bovis infection. Farm B sampled the bulk tank quarterly with occasional periods of monthly sampling for surveillance purposes and never had a mycoplasma positive result in more than 10 years of testing prior to the start of the current study. Neither were any clinical cases or sub-clinical carriers of mycoplasma identified over the same period. Moreover, all samples from clinical mastitis cases underwent culture for pathogen identification including M. bovis and never yielded a positive result for mycoplasma.

On both Farms A and B, cows were moved to the maternity areas, open straw bedded packs, within one week of their due date. To form study groups comprising calves as close in age as possible, records and careful monitoring were used to identify those cows likely to calve within two weeks of each other. All calves were fed 4 L of high-quality colostrum within 12 h of birth, received an oral vaccination against E. coli and Coronavirus (First Defense® Bolus; ImmuCell Corp., Portland, ME, USA) intranasal vaccination (TSV-2®; Zoetis Inc., Kalamazoo, MI, USA) against infectious bovine rhinotracheitis (IBR) virus and parainfluenza₃ (PI₃) virus, and navels were dipped using an iodine solution. After preventative measures, calves were housed in individual hutches until collected for transport to the heifer facility. The rearing operation picked up newborn calves from each herd twice a week on separate days. At pick-up, general attitude and navel scores were recorded and the first set of swabbing solutions for mycoplasma (see below) were collected.

Heifer selection and sampling schedule

Every 4 months, 30 Holstein calves (15 from each farm) were enrolled until a total of 150 (75 from each farm) was obtained. To keep ages as close as possible, the first 15 heifer calves born in each month of enrollment were assigned to the study. The first set of swabbing solutions from each calf (nares, sub-palpebral, ear canal, and vagina see below for details) were collected at the parent farm, then monthly for 6 months (whether still comingled at the heifer facility or returned to the home farm), followed by quarterly sampling up to 21 months.

Heifer facility procedures

The heifer rearing facility was in northeast Maryland, USA; approximately 48 and 32 km (30 and 20 miles) away from farms A and B, respectively. The operation had previously reared heifers from Farm A but had only just started receiving heifers from Farm B at the start of the study. There were no heifers from any other farms present during the study period. Once at the rearing facility, calves were further processed (blood collection for total protein assessment, ear punch for bovine viral diarrhea virus [BVDV] diagnosis, weight, and height measurements). The rearing process occurred in three phases:

Phase 1

All heifers were initially housed singly in one of four open-faced barns. Each barn had a single sloping roof and comprised 24 individual straw-bedded units which were filled as calves arrived at the facility so each barn could contain a mix of calves from Farms A and B. Units were separated by wooden panels with a metal gate at the front. Gates were constructed with openings to allow calves access to buckets in exterior brackets that contained water and calf starter grain.

Until completion of weaning over weeks 6 to 13, milk was delivered via bottles hung from the gate. Other notable procedures that occurred during this phase included disbudding by cauterization at 4-weeks of age, and intranasal vaccination (Inforce™ 3®; Zoetis Inc., Kalamazoo, MI, USA) against bovine respiratory viruses (IBR, PI₃, bovine respiratory syncytial virus [BRSV]) at 10 days prior to completion of weaning. Although not permissive to contact, during phase 1 calves could potentially make nose-to-nose contact with their neighbor at the front and possibly even over the sides of individual units.

Phase 2

Once weaned, calves were placed on straw-bedded packs in groups of 5–6 in one of three barns. Movement from phase 1 to phase 2 occurred as calves were weaned, thus groups could contain a mix of animals from each farm. Each phase 2 barn was partitioned to form five separate pens with a straw-bedded pack at the back and a concrete standing area at the front with access to the feed lane through fixed, non-locking head gates. The packs were separated by wooden rails and the standing areas by a metal swinging gate that could be used to lock heifers onto the packs while the standing area was scraped. Compared to phase 1, construction of phase 2 barns was more permissive to direct contact between neighboring pens. During phase 2, all Farm B heifers left the facility by 5 months of age to return to the home farm.

Phase 3

Farm A animals remained at the heifer facility into phase 3 which began at 6 months of age when heifers were moved to one of two barns in groups of 12–15. The most notable procedures that occurred during this phase were intramuscular vaccination (Bovi-shield Gold® 5; Zoetis Inc., Kalamazoo, MI, USA) against bovine respiratory viruses (IBR, PI₃, BRSV, BVDV Types 1 and 2), and subcutaneous vaccination against clostridial bacteria (Clostridium chauvoei, septicum, novyi, haemolyticum, sordellii) and C. perfringens types C and D bacterin toxoid (Ultrabac® 8; Zoetis Inc., Kalamazoo, MI, USA). Farm B heifers received intramuscular Bovi-shield Gold FP® 5L5 (Zoetis Inc., Kalamazoo, MI, USA) against bovine respiratory viruses (IBR, PI₃, BRSV, BVDV Types 1 and 2, plus leptospiral bacteria serovars (Leptospira interrogans Canicola, Grippotyphosa,, Icterohaemorrhagiae, Pomona and Leptospira borgpetersenii Hardjo-bovis) as well as subcutaneous Ultrabac® 8 back at the parent farm. The construction of the phase 3 barns was the same as phase 2 except, to accommodate the growing heifers, there were only three pens per barn. All Farm A heifers had returned to the home farm by 12 months of age.

Samples

Swabbing solutions were obtained using a sterile polyester-tipped applicator pre-moistened in a commercially prepared PPLO (pleuropneumonia-like organisms) broth recommended for the cultivation and maintenance of Mycoplasma spp. and Ureaplasma spp. (Hardy Diagnostics, Santa Maria, CA, USA). After collection, swabs were placed back into the media containing tubes.

• 1. Nares – the applicator tip (one for each nostril) was inserted beyond each naris approximately 10 cm into the nasal passage, rotated, then removed.

• 2. Eyes – the applicator tip (one for each eye) was inserted under the upper eyelid, moved medially to the medial canthus of eye, passed under the third eyelid, then removed.

• 3. Ear canal – the applicator tip (one for each ear) was inserted into the ear canal, advanced approximately 3–5 cm, rotated back and forth, then removed.

• 4. Vagina – the applicator tip was inserted past the vulvar lips, beyond the vestibular junction, into the vagina, rotated back and forth, then removed.

Mycoplasma culture

Tubes were stored at refrigeration temperatures for no longer than 4 days until being shipped overnight on ice to the microbiology laboratory at Washington State University College of Veterinary Medicine, USA. Although the laboratory had each calf’s unique ID number, they were unaware of which farm any specific calf belonged to. Tubes of enrichment medium containing swab specimens (swabbing solutions) were incubated at 37 °C in a 10% CO₂ atmosphere for 4 days [36, 37]. On the fourth day, 100 μL of swabbing solution was streaked onto aa agar plate for selective isolation of Mycoplasmataceae (PPLO) organisms (Hardy Diagnostics, Santa Maria, CA, USA). Plates were incubated at 37 °C in a 10% CO₂ atmosphere for 10 days, then examined with a 15X dissecting microscope for colonies with the distinctive “fried egg” appearance. Samples were considered positive if any mycoplasma colonies were seen, and negative if there was no evidence of mycoplasma growth.

Mycoplasma bovis PCR

Confirmation that Mycoplasma-positive samples were M. bovis was by PCR. The PCR reaction was performed as described in Boonyayatra et al. [38] Briefly, genomic DNA was extracted using PureLink Genomic DNA Kits (Invitrogen, Carlsbad, CA), a silica-based column purification method, performed on multiple colonies collected from a single plate, which were visually identified as Mycoplasma spp., following the manufacturer's instructions. The PCR reaction was initiated by combining 5 μL of the extracted DNA with 12.5 μL of 2 × Master Mix: HotStart Taq DNA polymerase, dNTP mix, ROX passive reference dye, and QuantiTect Probe real-time-PCR buffer [containing Tris–Cl, KCl, (NH4)2SO4, 8 mM MgCl2, pH 8.7], 10 pmol of primer (MbF 5′-TAATGCACGCAAACTCTCGTAGT-3′, fusA gene of Mycoplasma bovis), and 5 pmol of probe according to the QuantiTect Probe PCR kit (Qiagen Sciences, Germantown, MD) instructions. All real-time PCR assays were performed on a Step One Plus (Applied Biosystems, Foster City, CA) real-time PCR instrument. The thermal cycling protocol included 15 min of predenaturation at 95 °C, 45 cycles of 15 s of denaturation at 94 °C, and 60 s of annealing at 60 °C. Based on a cycle threshold (CT) of 37, each reaction was classified as positive or negative for M. bovis [36, 38,39,40].

Statistics

Data were analyzed using STATA 17 (Stata Corp., College Station, TX, USA). Continuous data were assessed for normality using the Shapiro Wilk test. For descriptive statistics, a t-test (normal) or Wilcoxon rank sum (non-parametric), which makes no assumptions regarding distributions, were used to determine whether there were significant differences between calves that did or did not become positive. Results are presented as mean and standard deviation (SD) or medians with first and third (25th-75th) percentiles (interquartile range, IQR), and range. Depending on the number of groups being compared, non-parametric Wilcoxon rank sum or Kruskal Wallis tests were also used for preliminary examination of count data (number of positive sites), but final analyses of count data were by means of Poisson regression. Nominal categorical data (binomial and multinomial) were analyzed with a x² or Fisher’s exact test, where appropriate. Where applicable, two-sample tests of proportionality (where the null hypothesis is that there is no difference) were used to evaluate the groups. After checking that the data fit a Poisson distribution, Poisson regression was used to compare positivity rates at all sites, or combinations of sites when more than one site was positive. For the Poisson models, predictors examined included Farm, season of birth, season of sample collection, sampling day i.e., session number (also a surrogate for age), and age in days. As the animals were handled in a close to identical manner, with the exception that one cohort returned to the home farm earlier than the other, there were few potential confounders. However, the duration of comingling and seasons were examined as possible confounders. Robust variance estimation at the calf level was used to control for clustering in all regressions. Incidence rate ratios (IRR) with 95% confidence intervals (CI) were used to quantify associations. Statistical significance was inferred when P < 0.05.

Results

Group characteristics and comingling timeline

The age of calves when transported from home to the heifer facility was not statistically significantly different between farms (Farm A, median 4 days, IQR [Interquartile range, 25th and 75th percentile] 3–4, range 1–9; Farm B, median 3 days, IQR 3–4, range 1–9; P = 0.42). Even so, because Farm A had more calves born per week, the time taken to complete assembly of the groups was significantly shorter for Farm A (median 8, IQR 7–9, range 4–14 days) compared to Farm B (median 20, IQR 17–21, range 13–32 days, P = 0.02). In terms of season of birth, there was no difference between farms (P = 0.97) in the proportions of calves born in Winter (December-February; Farm A 15/72, 20.8%, Farm B 15/72, 20.8%), Spring (March–May; Farm A 30/72, 41.7%, Farm B 28/72, 38.9%), Summer (June–August; Farm A 12/72, 16.7%, Farm B 14/72, 19.4%) or Autumn (September–November Farm A 15/72, 20.8%, Farm B 15/72, 20.8%). Three heifers from each farm died of causes unrelated to mycoplasma soon after enrolment and were excluded, leaving 144 heifers (72 from each farm) in the analysis.

Median ages in weeks from birth to weaning at the heifer facility were not different between farms (Farm A 7.6 weeks, IQR 7.0–8.8, range 6.1–10.9; Farm B 7.6 weeks, IQR 7.0–8.3, range 5.6–13.1, P = 0.79). The period of “weaning” in Phase 1 (see methods) covers the time of milk withdrawal until movement to group housing and was very occasionally prolonged because the latter was full, so space was not always available in a timely fashion. Farms A and B had slightly different contracts with the heifer facility such that Farm B heifers returned to their home farm earlier (median 4.0 months, IQR 3.6–4.1, range 2.9–4.9) than did Farm A heifers (median 8.0 months, IQR 7.3–9.6, range 6.4–11.6, P < 0.001). A few Farm B calves remained at the heifer facility after 4 months of age, but all had returned to the home farm by 5 months of age.

Culture results

There was one mycoplasma positive swabbing solution on the first test day (collected while still at parent farm), a vaginal sample in a calf from the previously positive Farm A. On the first sampling day (≤ 9 days of age), one calf from Farm A was missing an ear canal swabbing solution. On sampling day 5 (4 months), one Farm B calf was missed completely (0 samples) and another from Farm A had a vaginal sample only. All other samples were collected from all calves through sample day 4 (3 months), and 96–97% of calves were fully tested on sampling days 5–7 (i.e., 4–6 months of age, Table 1). From sample 8 (9 months of age) to 12 (21 months of age) the number of animals sampled declined, such that sampling of the majority of heifers (63.9%) still occurred on sample day 9 (12 months) but on subsequent test days declined to 34.7% (sample day 10, 15 months) and 11.1% (sample day 11, 18 months), respectively.

While comingled at the heifer facility, the number of positive calves increased over time, and continued to increase even after calves were separated when Farm B heifers returned home. Ultimately, only 19/144 (13.2%) heifers were never positive (13/72, 18.1% and 6/72, 8.3% from Farms A and B, respectively; P = 0.09). Excluding the single positive at sampling 1, days of age to first positive were not different between farms (A median = 109, range 42–561 days; B median = 110, range 33–404 days; P = 0.96). Because Farm B heifers left the heifer facility before those from farm A, the culture results obtained from each sampling were analyzed separately to determine whether there were any differences in the rates or site distribution of mycoplasma positives during or after comingling. Apart from the first sample taken at the parent farm i.e., a single positive obtained from Farm A, the only significant difference in incidence rate ratios (IRR) for any parameter, occurred when Farm B heifers returned home while those from Farm A remained at the heifer facility (Table 1). During this transition there were significant decreases in the number of both Farm B positive heifers and their total positive samples (sampling days 5 and 6, 4–5 months; Table 1). However, on sampling day 7 (6 months) there was a significant rebound increase in Farm B positive heifers which appeared to be driven by nose positives in combination with another positive site(s) (IRR 2.02, 95% confidence interval [CI] 1.23–3.29; P = 0.005) and non-nose positive samples (IRR 3.75, CI 1.52–9.27; P = 0.004) rather than nose only positives which were not different (IRR 1.31, CI 0.70–2.46; P = 0.40). Nevertheless, because the differences between farms were limited in time frame, we felt it was reasonable to summarize results from both farms combined over 15 months (sample days 1–10) when 35–100% of heifers were sampled (Fig. 1 and Table 2). Considering positive sites from all sampling days together, the only significant difference between farms was that Farm A had more calves with ear canal positives (either alone or in combination with other sites; 31/72, 43.1%; IRR 1.37, CI 1.22–1.55; P < 0.001) than did Farm B (6/72, 8.3%). There were no significant differences between farms in positives from any other site or combinations thereof.

Table 1 Summary of M. bovis culture results Farms A and B over the 21-month sampling period: nose vs. non-nose positive samples and incident rate ratios for total positives, Farm B compared to Farm A

Fig. 1

Time-course of M. bovis positive samples from four anatomic sites (nares, eyes, ear canal, vagina) or combinations thereof over 10 sampling days (birth [0* ≤ 9 days of age] to 15 months of age) in all calves from Farm A (N = 72) and Farm B (N = 72) that were comingled at a heifer rearing facility. The graph shows the overall dominance of nasal positives and their apparent importance in transmission

Table 2 Summary of M. bovis culture results combined from both farms A and B over the 21-month sampling period: comparison of nose vs. non-nose positive samples

Cumulatively, 634/1,276 (49.7%) positive samples were obtained from 125/144 heifers (86.8%; Table 2). Among positives at each of the four test sites, some of which were part of combinations when more than one site was positive on a given test day, most were intranasal (296/634, 46.7%) followed by ocular (163/634, 25.7%), vagina (113/634, 17.8%) and ear canal (62/634, 9.8%). Among all positives, 127/634 (20.0%) were combinations; eye/nose (55/634, 8.7% of all positives, or 43.3% of all combinations) and nose/vagina (21/634, 3.3% of all positives or 16.5% of all combinations) were most frequent. Other combinations were observed between 1–15 times, 0.2–2.4% of all positives (0.8–11.8% of combinations). As the study took place over all four seasons results were analyzed to determine whether there were any seasonal effects on the number of M. bovis positive heifers. Overall, there was no difference between seasons in the number of animals being classified as either positive (regardless of the number of sites) or negative and there were no differences between farms where Farm A was the referent in IRR for Winter (December-February; IRR 0.97, CI 0.72–1.29; P = 0.82), Spring (March–May; IRR 1.07, CI 0.84–1.37; P = 0.57), Summer (June–August; IRR 1.05, CI 0.61–1.81; P = 0.87), or Autumn (September–November; IRR 0.91, CI 0.69–1.20; P = 0.49).

Transmission timeline

Only one calf was positive on the first sample, at the second sampling (1 month of age) 12 calves were positive (1 ear, 2 eye, 4 nose, 4 vagina and 1 ear/vagina). However, over samplings 3 and 4 (i.e., 2–3 months of age) there was an exponential increase in calves with positive intranasal swabbing solutions; 16 and 36 (peak frequency), respectively. Combinations of positives involving the nose began to appear after a temporal lag (samplings 4–8; 3–9 months of age) but eventually reached similar frequencies to intranasal alone positives. There was also an increase in ocular positives not involving the nose, but they never achieved the same levels as those involving the nose alone or nose/eye combinations (Fig. 1).

Over the course of the study there were 91/144 (63.2%) calves that had intranasal positives and were subsequently identified as positive at additional sites either separate from or in combination with the nose (Fig. 1 and Table 2). Among those 91 calves, 63 (69.2%) were first positive in the nose alone, while 28 (30.8%) were first positive simultaneously (i.e., on the same sampling day) in the nose and at one or more other sites. Among combinations, nose and eye was by far the most common (16/91, 17.6%), followed by nose and vagina (5/91, 5.5%), nose and ear (4/91, 4.4%). There were only 3/91 calves first positive at three sites (two of nose, eye, ear 2.2% and one nose, eye, vagina 1.1%). In most calves, subsequent positives involved the nose one or more times (72/91, 79.1%), whereas subsequent positives that never involved the nose were less frequent (19/91, 20.9%). Regardless of site, 119 heifers were positive on two or more sampling days (119/125 = 95.2% of positive heifers, range = positive 2–7 times on a sampling day; Table 3).

Table 3 Number of heifers never positive (0 sampling days), or positive on one or more sampling days (1 – 7 times) at one or more of four sites tested on each sampling day

The number of heifers with intranasal positives per sampling event plateaued at 28 to 22 over sampling 5–8 (4–9 months of age). When combinations involving the nose are included, the number of positive calves at each sampling was in the 40–48 range over samplings 4–8 (3–9 months) except for sampling day 7 (6 months) which peaked at 68. During the study, intranasal positives alone or in combination with other sites occurred more frequently than sites or combinations not involving the nose (IRR 1.44, 95% CI 1.41–1.47: P < 0.001). After cessation of comingling when heifers returned home to Farm B, the frequency of intranasal positives in both Farm A and Farm B heifers declined rapidly after sampling day 7 (6 months of age), while other sites and positive combinations initially increased then declined more gradually (Fig. 1 and Table 1).

On the last scheduled sampling day, i.e. sample day 12 (21 months), no heifers from Farm A and only two from Farm B were sampled (Table 1) because most animals had already calved and entered the milking herd. On sampling days 10 and 11 (15 and 18 months), numbers were markedly reduced (Table 1), either because animals had calved early or were moved to locations without suitable handling for safe sampling. Additionally, a few animals left the herds due to voluntary or involuntary culling and were unavailable for testing.

Discussion

Mastitis is a major problem in the dairy industry and of great concern economically as well as for the health, welfare, and productivity of affected animals. In the United States, the most recent USDA-NAHMS report on U.S. dairies indicated that 99.7% of dairy operations had at least one case of mastitis and overall, 24.8% of all cows developed clinical disease [18]. Among the 57% of dairies that performed any milk cultures, mycoplasma was identified on 8.7% of all operations; primarily those categorized as medium (100–499) to large (500 + cows) [18] (USDA 2016). Since the earliest reports in the US [7,8,9,10], mycoplasma mastitis, most notably that associated with M. bovis, has become an increasingly important cause of contagious mastitis [17]. Its contagious nature can result in multiple animals being infected, making control efforts more difficult. Perhaps the most problematic characteristic of mycoplasma is that asymptomatic-carrier animals are common and become a “silent” source of infection for animals that comingle with shedding carriers. Moreover, whether carrier animals will ever become clinical or what may trigger them to become shedders of the organism is unclear.

The trend towards expansion of dairy herds in the US and concomitant growth of heifer-rearing facilities that comingle heifer calves from different dairy operations is unlikely to change. Consequently, there is a need to better understand the genesis of M. bovis asymptomatic carriers and more fully characterize their role in M. bovis epidemiology. To this end, the current study specifically focused on M. bovis prevalence and transmission in dairy heifers from two separate herds while comingled at a heifer-rearing facility.

Results of the current study show how rapidly and completely a population of calves from an apparently naïve herd (Farm B) can become infected with M. bovis when they contact calves from a herd with a history of M. bovis infection (Farm A). Despite the fact that 91.7% (66/72) of animals from Farm B (a disease-free herd that had never previously been identified as M. bovis positive) were M. bovis positive at least once at a least one of four anatomic sites and were more commonly positive multiple times at multiple sites, during the course of the study none of these calves exhibited any clinical disease that could be ascribed to M. bovis infection. The majority of calves from Farm A (81.9%, 59/72), which had a history of M. bovis associated clinical disease, also tested positive for M. bovis during the study and like those from Farm B never developed clinical disease.

There is evidence that M. bovis can be transmitted via aerosols, direct contact and indirectly via feed, water, housing, or other fomites [2,3,4, 15, 34, 35, 41, 42]. However, in terms of transmission dynamics, our findings indicate that colonization of the nasal passages might be predominantly responsible for spread of M. bovis among comingled calves. Early studies found that nasal prevalence of M. bovis in California dairy heifers up to 8 months of age was 34% and 6% in diseased and non-diseased herds, respectively [42]. More recently, longitudinal studies showed that almost all calves in diseased herds become infected [43]. In the current study, the number of heifers with a positive nasal sample increased exponentially at and immediately post-weaning (2–3 months of age, samplings 3 and 4), subsequently, there was a concomitant rise in the total number of calves found to be positive and those positive at multiple sites (Fig. 1 and Table 1). Several factors might be associated with the timing of the increase in positive samples: stress of weaning, decline of maternal antibodies (theoretically, this might only apply to Farm A), and transition to groups of 5–6 calves, some of which were comprised of calves from both Farms A and B, in housing more permissive to contact between groups.

Time-dependent increases in the prevalence of M. bovis positive nasal swabs have been observed in other production systems with both calves and older animals. For example, in veal calves raised in temperature-controlled housing in which they can interact with neighboring calves, attack rates for M. bovis colonization of the upper respiratory tract, characterized by positive nasal swabs, increased over time peaking at 3–4 months of age [44]. By contrast, 300–350 kg bulls, mostly M. bovis free when introduced into a fattening system, exhibited a dramatic increase in positive nasal swabs 15 days later [45]. Very similar results were reported for calves (6-month-old mixed-breed heifers, approx. 250 kg) entering a Canadian feedlot where prevalence of M. bovis in bronchiolar lavage fluid was initially 1.7% but had increased to 72.2% by 15 days and was only slightly greater (85.7%) 55 days after arrival [46]. Comingled feedlot animals exhibit the same increase in M. bovis prevalence of positive samples from the respiratory tract that likely originate in the nose as those seen in veal calves and the dairy heifers in our study. It is, however, noteworthy that in feedlot animals the dramatic increase in prevalence was far more rapid, about two weeks, compared to the veal and dairy calves where the same increase in nasal prevalence occurred at 3–4 months. Although there could be an age-related component, it is far more likely that the longer time to increased nasal positives is a function of significant separation between animals in the heifer facility and veal barns, whereas animals at feed lots are housed in large groups, thereby promoting spread.

Results from our study indicate that once the increase in nasal positives is apparent, widespread infection appears inevitable even after comingling ceases. Compared to Farm A there were more Farm B heifers with M. bovis nasal positives on sampling days 3 and 4 (Table 1). However, during the transition period as Farm B animals returned to the parent farm (sample days 5 and 6, 4–5 months of age) the number of nasal positives in Farm B heifers declined to levels below that of Farm A. Nevertheless, once the entire cohort had returned home, there was a rebound such that the number of Farm B positive heifers was significantly greater than Farm A (Table 1). It is very likely that those Farm B calves remaining at the heifer facility between 4 and 5 months of age were housed with Farm A calves during that time. It is, however, unclear whether extended comingling had any role in the increase in Farm B positive heifers seen after the full cohort had returned to the home farm. Unlike Farm A, Farm B had no history of mycoplasma infection, naïvety of the herd might also have contributed to the slight increase in total positive heifers overall and the suggestion that some animals remained positive for longer than those from Farm A, but more research is required to provide direct evidence of a role for enhanced immunity in the exposed herd.

As mentioned previously, despite the vast majority of heifers from both Farms A and B being positive at least once for M. bovis at one or more anatomic sites, none of the affected heifers exhibited any clinical signs that could be attributed to M. bovis infection. Other studies have demonstrated marked clonality of disease-associated M. bovis such that even if it can recrudesce from carriers after stressful events such as transportation and comingling, increased prevalence of M. bovis pulmonary infection observed during respiratory disease outbreaks in feedlot cattle seems primarily due to the horizontal transmission of a single clone rather than any M. bovis being capable of initiating disease [47, 48]. Moreover, these strains can persist in animals after the clinical signs have resolved [47]. Similar results have been reported in dairy cattle where one strain of M. bovis was found to be associated with a herd outbreak of mastitis [33]. Additionally, following introduction of an M. bovis mastitic cow to the herd’s hospital pen, the same strain infected presumed negative cows which then developed clinical mastitis [49]. On the first sampling day of the current study, only one site (vagina) from a single Farm A calf was identified as positive for M. bovis. Although it is tempting to imagine that this could have been a single clone, not associated with the development of clinical disease, that subsequently spread through the population of comingled heifers, it is quite possible that other day 1 positives were initially missed on culture. Moreover, Farm A calves had been reared at the heifer facility for a prolonged period prior to introduction of calves from Farm B. It seems reasonable to assume that there could be environmental contamination with mycoplasma that cannot be ruled out as an additional source of infection.

To elicit disease, pathogens including M. bovis must first cause infection by colonizing the target tissues such as the respiratory tract [50]. It is increasingly recognized that resident microbial communities (i.e., microbiota) have an important role in inhibiting this first step and where colonization occurs, might be able to suppress pathogens sufficiently to prevent overgrowth, inflammation and local or possibly systemic spread [51]. A positive influence for local microbial populations is supported by the finding that those cattle that remained healthy in a feedlot had a more diverse bacterial community in their nasopharynx than did cattle that developed respiratory disease [52, 53]. Composition of the nasopharyngeal microbiota in dairy calves is strongly influenced by the maternal vaginal microbiota, likely transferred to the neonate in the vaginal canal during birth [54]. Over the first 35 days of life calves share more 70% of their nasopharyngeal microbiota with the vaginal microbiota of their dam. Among the most prominent organisms of the vaginal microbiota (Mannheimia, Moraxella, Bacteroides, Streptococcus, and Pseudomonas), higher levels of Mannheimia have been associated with better health of the respiratory tract and middle ear of their calves. Whereas progeny of dams with less abundant Mannheimia in the vaginal microbiota appeared at greater risk of developing pneumonia and/or otitis [54]. Although most research has been done on the respiratory tract, there is no reason to believe that the microbiota of other areas such as the urogenital tract, ocular and aural regions would not be equally important in protection against pathogenic microbes including mycoplasma. While the microbiota of calves in the present study might have been involved in protection against clinical disease, as with possible clonal spread, any association of the microbiota to lack of disease development cannot be assessed without further research.

While it is clear that M. bovis alone can cause natural and experimentally induced clinical disease, our study confirms previous reports that presence of M. bovis does not always result in disease and that clinical disease is not necessarily a pre-requisite for the maintenance and dissemination of M. bovis in cattle [3, 4, 13, 34]. Chronic asymptomatic infection with intermittent shedding appears critical to the epidemiology of M. bovis. Although many cattle shed M. bovis for a few months or less [33, 42, 55], some can shed for many months or even years [27, 33, 42, 55]. Moreover, evidence indicates that stressful events such as transportation, comingling, entry into a feedlot and cold stress are associated with possible recrudescence and increased rates of nasal shedding [30, 56]. Despite the absence of clinical disease in these comingled heifers over the study period, there might therefore be circumstances later in life that lead to increased shedding of M. bovis. However, the significance of being infected at this stage of life on subsequent health and productivity can only be evaluated by prospectively following the affected animals through entry until leaving the main herd. All heifers in the current study were monitored through their lives in the respective parent herds and the findings will be presented in a companion paper.

Conclusions

These data demonstrate that over time M. bovis can be cultured from all sites tested, but it appears that nasal colonization is most critical for establishing transmission between comingled heifers. Moreover, when the level of intranasal carriage escalates at around 2–3 months of age, virtually all heifers will subsequently be positive at least once at one or more sites even if comingling ceases. These results indicate how readily asymptomatic M. bovis can be disseminated in a population of animals and potentially manifest as clinical disease later in life with concomitant effects on health and productivity.