Antimicrobial Resistance in Indicator Microorganisms Escherichia coli and Enterococcus spp. from Healthy Dairy Cattle in Latvia

TL;DR

This study examines antibiotic resistance in gut bacteria from healthy Latvian dairy cows, finding low resistance overall but noting higher risks on larger farms.

Contribution

The study provides new data on antimicrobial resistance in indicator bacteria from clinically healthy Latvian dairy cattle, emphasizing farm size as a potential risk factor.

Findings

Resistance levels in Escherichia coli and Enterococcus spp. from healthy Latvian dairy cows were generally low.

Multidrug resistance was more common in isolates from larger farms and milk samples.

Enterococcus spp. showed high resistance to quinupristin–dalfopristin, an antibiotic not used in Latvian dairy farming.

Abstract

Antimicrobial resistance (AMR) is a growing global concern because bacteria that no longer respond to antibiotics can spread between animals, the environment, and humans. Within the One Health framework, dairy production systems play an important role in this process, as bacteria from healthy cows can be disseminated through milk, manure, and the farm environment. In this study, bacteria isolated from clinically healthy dairy cows in Latvia were examined to determine their resistance to antibiotics that are critically important in human medicine. Rectal swab samples were collected from cows to isolate two common gut bacteria as indicators of cumulative antimicrobial pressure, and milk samples were also taken from bulk tanks to assess AMR relevant to food safety. Overall, resistance levels were low, and most bacteria remained susceptible to critically important antibiotics. However, some bacteria showed resistance to multiple drugs, particularly those isolated from larger farms, suggesting that farm conditions may increase the risk of more resistant bacteria developing and spreading. These findings highlight the importance of responsible and prudent antibiotic use on farms and the need for continued monitoring of antibiotic resistance. Understanding how resistance develops and spreads helps protect animal health, food safety, and public health.

Antimicrobial resistance (AMR) in food-producing animals is a growing One Health concern. However, data on AMR in indicator microorganisms from clinically healthy dairy cattle in Latvia remain limited. This study aimed to characterize the AMR profiles of Escherichia coli and Enterococcus spp. isolated from rectal swabs and bulk-tank milk collected from 18 dairy farms between February and May 2025. Bacterial identification was performed using conventional culturing and MALDI-TOF mass spectrometry, and antimicrobial susceptibility was determined using the disk diffusion (Kirby–Bauer) method, interpreted according to European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines. Resistance levels were further quantified using resistance scores (R-scores) and the Multiple Antibiotic Resistance Index (MARI). In total, 582 E. coli and 428 Enterococcus spp. isolates were recovered from rectal swabs, with E. coli showing the highest resistance to ampicillin (12.5%) and amoxicillin–clavulanic acid (6.7%), whereas resistance to tetracycline was rare (0.3%). Enterobacteriaceae from milk exhibited higher resistance to ampicillin (45.8%) and amoxicillin–clavulanic acid (20.8%). Among Enterococcus spp., resistance was highest to an antibiotic not used in dairy cows in Latvia quinupristin–dalfopristin (69.2%), while resistance to vancomycin and linezolid remained low (0.5% each). Milk-derived enterococci showed a comparable pattern, with additional resistance to streptomycin (25%). Overall, resistance levels and multidrug resistance were low. However, the presence of sporadic resistant isolates and elevated MARI values, particularly in large-scale farms and milk-derived bacteria, highlights the importance of continued AMR surveillance and prudent antimicrobial use in the Latvian dairy sector.

Figures

Full text

1. Introduction

Antimicrobial resistance (AMR) is an increasing global concern within the framework of One Health that threatens the efficacy of treatment in both human and veterinary medicine [1,2]. The use of antibiotics in food-producing animals, particularly in intensive dairy production systems with larger herd sizes and higher animal density, contributes to the emergence and dissemination of antimicrobial-resistant bacteria. In addition to pathogenic microorganisms, commensal and opportunistic bacteria, such as Escherichia coli and Enterococcus spp., are also important, as they can contribute to the spread of resistance genes [2,3,4]. These genes pose a risk to human health through multiple routes of transmission, including the food chain, direct contact with animals, and environmental exposure [4,5].

In dairy farming, antimicrobials are frequently used for both therapeutic and metaphylactic purposes, particularly for the treatment and prevention of mastitis [5,6]. However, inappropriate or excessive antibiotic use significantly accelerates the development of resistance [3]. Monitoring resistance patterns in indicator organisms isolated from dairy cattle, including intestinal flora and milk-associated microbiota, provides essential information for evaluating on-farm AMR risk, assessing antimicrobial stewardship practices, and supporting evidence-based interventions [7,8].

In Latvia, systematic surveillance of AMR in dairy cattle is still in its early stages. Existing studies have focused on clinical or foodborne pathogens, whereas data on commensal organisms from healthy animals, primarily calves, are scarce [9,10], limiting the establishment of national baseline resistance levels in lactating cows. Additionally, baseline information on the geographical variation in AMR within the country is lacking. Given the risk of zoonotic transmission and the growing need to preserve the effectiveness of existing antimicrobials, particularly those critical to human medicine, establishing comprehensive nationwide surveillance systems is crucial.

The present study aimed to determine the AMR profiles of Escherichia coli and Enterococcus spp. isolated from rectal swabs and bulk-tank milk samples from clinically healthy, lactating dairy cows in Latvia. By applying standardized antimicrobial susceptibility testing and resistance scoring methods, this study provides the first comprehensive insight into the regional distribution and burden of AMR among indicator bacteria from healthy dairy cattle in Latvia, a livestock population for which such data have not previously been available.

2. Materials and Methods

2.1. Study Area

Sampling was conducted between February and May 2025 across four administrative regions of Latvia: Latgale, Vidzeme, Zemgale, and Kurzeme. To ensure representation of different production systems, farms were selected proportionally, including medium (200–500 animals) and large-sized (>500) farms, which together accounted for 10% of each type of farm in Latvia (Figure 1) and were distributed equally across administrative regions.

In each region, samples were collected from three medium-sized farms (n = 12). In addition, two large-sized farms were sampled in both Vidzeme and Zemgale, and one in Latgale and Kurzeme (n = 6), resulting in a total of 18 farms.

2.2. Sample Collection

Rectal swab samples were collected from 25 clinically healthy cows in their second or higher lactation on each farm. Each animal was restrained in its housing area, and a sterile swab was carefully inserted into the rectum and rotated against the rectal mucosa. The swab was then placed into a transport medium and appropriately labeled.

At each farm, bulk-tank milk samples were collected from both the bottom outlet valve and, when accessible, the top hatch. Before sampling from the outlet valve, the first portion of milk was discarded to reduce contamination, and subsequent milk was collected into a sterile 100 mL container. For top sampling, milk was aseptically collected from the upper layer of the tank using a clean 75 mL milk-sampling pipette and transferred into a sterile 100 mL container.

All samples were immediately placed in a cooling container at 4 °C and transported to the laboratory within 6 h for immediate processing.

2.3. Rectal Swab Analysis

Rectal swab samples were examined to isolate Escherichia coli and Enterococcus spp. and to determine their phenotypic AMR to selected antibiotics.

For the isolation of E. coli, all samples were directly streaked onto Levine EMB blue agar (Biolife, IT, Bannockburn, IL, USA) using the original swab taken from the transport medium. The plates were incubated at 36 ± 1 °C for 22 ± 2 h. After incubation, colonies with typical morphology (2–3 mm in diameter, slightly raised, violet cyclamen colonies with a darker center, and greenish metallic sheen) and atypical morphology (3–5 mm in diameter, flatter, rough surface, and sometimes irregular margins) were selected for further analysis.

For the isolation of Enterococcus spp., samples were placed into 5 mL of Azide Dextrose Broth (Biolife, Milan, Italy), vortexed (Biosan, Riga, Latvia) for 5–10 s, and incubated at 36 ± 1 °C for 22 ± 2 h for selective enrichment. Subsequently, 10 µL of enrichment broth was streaked onto Slanetz–Bartley agar (Biolife, Milan, Italy) and incubated at 36 ± 1 °C for 44 ± 4 h. One typical red–brown colony per sample was then subcultured onto Bile Aesculin Azide Agar (ISO form, Biolife, Monza, Italy) and incubated at 36 ± 1 °C for 22 ± 2 h. The development of a brown/black halo surrounding the colony was considered confirmatory for Enterococcus spp. Where required, colonies were identified using a VITEK MS MALDI-TOF mass spectrometer (bioMérieux SA, Lyon, France), where the Log Score values were ≥2.0.

2.4. Milk Analysis

In parallel, after thorough mixing, 10 µL and 100 µL aliquots were inoculated onto Columbia Agar, Columbia CNA Agar, Slanetz–Bartley Agar, Baird–Parker Agar (Biolife, Milan, Italy), and MacConkey II Agar (BD Sparks Glencoe, Sparks Glencoe, MD, USA). The 10 µL aliquots were streaked using bacteriological loops, whereas the 100 µL volumes were pipetted and spread using the spread-plate method. Plates were incubated at 36 °C for 24–48 h. From each sample, the three most dominant colony types were subcultured onto Columbia Agar (Biolife, Milan, Italy) for pure culture isolation and subsequently identified using VITEK MS MALDI-TOF mass spectrometry (bioMérieux SA, Chemin de I’Orme, France).

2.5. Antibiotic Susceptibility Testing

The antimicrobial testing panel included 12 antibiotics for E. coli isolates and 9 for Enterococcus isolates, primarily those considered critically important and highly important in human medicine according to the World Health Organization (WHO) [11], as well as agents belonging to categories A, B, C, and D of the European Medicines Agency (EMA) classification of antibiotic classes in veterinary medicine [12] (Table 1).

For E. coli isolates, the following antimicrobial agents were tested: amoxicillin–clavulanic acid (30 µg, Oxoid, Basingstoke, UK), ceftriaxone (30 µg, (Oxoid, Basingstoke, UK), gentamicin (10 µg, (Oxoid, Basingstoke, UK)), cephalexin (30 µg, Oxoid, Basingstoke, UK), cefepime (30 µg, Oxoid, Basingstoke, UK), eravacycline (20 µg, Liofilchem, IT, Roseto degli Abruzzi, Italy), trimethoprim (5 µg, Oxoid, Basingstoke, UK), sulfamethoxazole–trimethoprim (25 µg, Oxoid, Basingstoke, UK), ciprofloxacin (5 µg, Oxoid, Basingstoke, UK), ampicillin (10 µg, Oxoid, Basingstoke, UK), azithromycin (15 µg, Oxoid, Basingstoke, UK), and imipenem (10 µg, Oxoid, Basingstoke, UK).

For Enterococcus spp. isolates, the antimicrobial panel included ampicillin (2 µg, Liofilchem, IT), gentamicin (30 µg, Oxoid, Basingstoke, UK), streptomycin (300 µg, Oxoid, UK), vancomycin (30 µg, Oxoid, UK), linezolid (30 µg, Oxoid, Basingstoke, UK), eravacycline (20 µg, Liofilchem, Roseto degli Abruzzi, IT), quinupristin–dalfopristin (15 µg, Liofilchem, Roseto degli Abruzzi, IT), ciprofloxacin (5 µg, Oxoid, Basingstoke, UK), and imipenem (10 µg, Oxoid, Basingstoke, UK).

Inhibition zone diameters were interpreted according to European Committee on Antimicrobial Susceptibility Testing EUCAST 2025 breakpoints [13], and isolates were classified as susceptible, intermediate, or resistant. Enterococcus faecalis ATCC 29212 and E. coli ATCC 25922 were used as quality control strains.

2.6. Calculation of Multiple Antibiotic Resistance Index (MARI) and Resistance Score (R-Score)

To assess multidrug resistance among Escherichia coli and Enterococcus spp. isolates, the MARI was calculated as described by Jalil et al. [14]. MARI was determined using the following equation:

$$[eqn]$$

where a represents the number of antibiotics to which the isolates were resistant, and b represents the total number of antibiotics tested against the isolates.

Additionally, the resistance score (R-Score) was calculated for each isolate to quantify the overall burden of AMR. For each antimicrobial agent, isolates classified as “intermediate” were assigned a value of 0.5, while those classified as “resistant” were assigned a value of 1.0. The R-score was calculated as the sum of these values across all tested antimicrobials, providing a cumulative measure of resistance for each isolate.

Multidrug resistance (MDR) was defined as resistance to three or more antimicrobial drugs tested in this study [14].

3. Results

3.1. Escherichia coli Isolates in Feces and Their Resistance Patterns

A total of 582 typical and atypical Escherichia coli isolates were recovered from 450 bovine rectal swab samples. For clarity, antimicrobial agents were grouped according to the antibiotic classification in veterinary medicine, as outlined in EMA guidelines: categories A, B, C, and D (Figure 2). Of the 12 antimicrobials tested, all isolates were susceptible to azithromycin and imipenem, both classified as Category A and regarded as critically important in human medicine according to the WHO.

Resistance was detected primarily among antimicrobials belonging to categories B, C, and D. In Category B, complete resistance was observed in 22 isolates (3.8%) to cefalexin, three isolates (0.5%) to trimethoprim–sulfamethoxazole, and one isolate (0.17%) to ciprofloxacin. In addition, intermediate susceptibility was recorded for ciprofloxacin in two isolates (0.3%) and for trimethoprim–sulfamethoxazole in two isolates (0.3%).

Within Category C, resistance to amoxicillin–clavulanic acid was observed in 39 isolates (6.7%), while resistance to gentamicin was detected in two (0.3%). In Category D, ampicillin resistance was found in 73 isolates (12.5%), and resistance to trimethoprim alone was identified in seven isolates (1.2%).

Resistance to critically important antimicrobials azithromycin and imipenem in Category A was not detected, while resistance to other high-priority agents was rare: eight isolates (1.4%) were resistant to cefepime, five (0.9%) to ceftriaxone, and two (0.3%) to eravacycline.

Resistance profiling using the R-score showed that 475 isolates (81.6%) had an R-score of 0, indicating complete susceptibility to all tested antimicrobials. In comparison, 107 isolates (18.4%) were resistant to one or more antimicrobial agents (R-score ≥ 1) (Figure 3A). Among all isolates, 79 (13.6%) were resistant to a single antimicrobial, and 28 (4.8%) were classified as multidrug-resistant (MDR), with the majority showing an R-score of 2 (n = 14). The highest R-score observed was 6, detected in two isolates.

Analysis of the Multiple Antibiotic Resistance Index revealed that 15 isolates (2.6%) had values exceeding 0.2, indicating high resistance pressure (Figure 3B). When MARI values were compared across farm sizes, isolates with MARI > 0.2 were predominantly detected on large-scale farms with ≥501 animals, rather than on medium-scale farms with 200–500 animals (Figure 3C,D). It should be noted that the observed variability reflects biological differences among isolates from different farms within the same size category, rather than technical variability.

On one large farm, MARI values ranged from 0.16 to 0.58, with 7 out of 32 isolates (21.9%) exhibiting a median MARI of >0.2. In contrast, another large farm showed no detectable resistance. Among medium-scale farms, the highest MARI was 0.33, observed in two farms (0.3%), each represented by a single isolate.

3.2. Enterococcus Isolates in Feces and Their Resistance Patterns

Of the 450 rectal swab samples collected from dairy cattle, 22 (4.9%) demonstrated no detectable growth of Enterococcus spp. on any culture medium. The remaining 428 isolates were tested against nine antimicrobials. Antimicrobials were classified into categories A, B, C, and D according to EMA guidelines (Figure 4).

Among Category A antimicrobials, resistance was detected in two isolates (0.5%) to linezolid and two isolates (0.5%) to vancomycin. In contrast, high levels of resistance were observed to quinupristin/dalfopristin, eravacycline, and imipenem, with 296 (69.2%), 41 (9.6%), and 28 isolates (6.5%) fully resistant, respectively. Imipenem was the only antimicrobial for which intermediate susceptibility was detected, affecting 398 isolates (93.0%).

Eight isolates (1.9%) were resistant to ciprofloxacin (Category B) and ampicillin (Category D). All isolates were fully susceptible to gentamicin (Category C), whereas resistance to streptomycin (Category C) was observed in ten isolates (2.3%).

For Enterococcus spp. isolates, R-score analysis showed that 103 isolates (24.1%) had an R-score < 1, indicating complete susceptibility or intermediate resistance to only one antimicrobial agent (Figure 5A). An R-score of 1 was observed in 262 isolates (61.2%), indicating resistance to a single antimicrobial. In contrast, 63 isolates (14.7%) had R-scores > 1, indicating resistance to two or three antimicrobials.

High resistance, defined by a MARI > 0.2, was observed for 63 (14.7%) isolates (Figure 5B). Of these, seven isolates (1.6%) were recovered from large-scale farms, whereas 56 (13.1%) originated from medium-scale farms. The remaining isolates (143 from large-scale and 244 from medium-scale farms) had MARI values ≤ 0.2. The highest MARI was observed among Enterococcus spp. isolates and was 0.39, detected in three isolates, one from a large-scale farm and two from medium-scale farms (Figure 5C,D).

3.3. Isolates in Milk and Their Resistance Patterns

Thirty-one bulk-tank milk samples were collected for microbiological analysis, including 18 samples from the bottom outlet valve and 13 from the top hatch. In total, 35 Gram-negative and 56 Gram-positive bacterial isolates were recovered and tested for antimicrobial susceptibility.

Among the Gram-negative isolates, 24 (69%) belonged to the family Enterobacteriaceae (Figure 6A), consisting mainly of Escherichia coli (n = 11), Klebsiella spp. (n = 8), Citrobacter freundii (n = 3) and Enterobacter cloacae (n = 2), isolated from 21 (77%) bulk-tank milk samples.

The highest resistance among Enterobacteriaceae isolates was to ampicillin (EMA Category D), with 11 isolates (45.8%) showing no or minimal inhibition zones. Resistance to trimethoprim (Category D) was observed in two isolates (8.3%). Resistance to amoxicillin–clavulanic acid (Category C) was detected in five isolates (20.8%). Resistance was also observed to Category B antimicrobials: three isolates (12.5%) were resistant to cefalexin, one (4.2%) to ciprofloxacin, and one (4.2%) to trimethoprim–sulfamethoxazole. In Category A, four isolates (16.7%) were resistant to eravacycline, and one isolate (4.2%) was resistant to imipenem. All isolates were fully susceptible to gentamicin (Category C), azithromycin, cefepime, and ceftriaxone (Category A), and no intermediate susceptibility was observed.

Other Gram-negative bacteria (Figure 6B) included Butiauxella agrestis, Hafnia alvei, Lelliottia amnigena, and Stenotrophomonas maltophilia (one isolate of each), Raoultella spp. (n = 4), and Pseudomonas aeruginosa (n = 3).

Complete susceptibility was observed only to gentamicin (Category C), while resistance to amoxicillin and clavulanic acid was detected in five isolates (45.5%). In Category D, resistance to ampicillin and trimethoprim was detected in five (45.5%) and four (36.4%) isolates, respectively. Among Category B antimicrobials, resistance to cefalexin was observed in five isolates (45.5%), resistance to trimethoprim–sulfamethoxazole in three (27.3%), and intermediate resistance to ciprofloxacin in three (27.3%). Intermediate susceptibility to Category A antimicrobials was common: seven isolates (63.6%) were intermediate to imipenem, three isolates (27.3%) to cefepime, and one isolate (9.1%) to both imipenem and cefepime. Ceftriaxone resistance was observed in four isolates (36.4%), while three isolates (27.3%) were fully resistant to azithromycin and eravacycline. Within the Enterobacteriaceae group, eight isolates (33.3%) had an R-score > 1, including three isolates (12.5%) resistant to four antimicrobials simultaneously. Four isolates (16.7%) were resistant to one antimicrobial agent only, while the remaining 12 (50%) showed no resistance (R-score = 0).

Furthermore, non-Enterobacteriacea Gram-negative bacteria exhibited higher resistance levels, with three isolates (27.3%) displaying the maximum R-score of 8 and one isolate (9.1%) having an R-score of 6. One isolate (9.1%) showed monodrug resistance, while five (20.8%) showed complete susceptibility (Figure 7A).

MARI analysis showed that five Enterobacteriaceae isolates (20.8%) exceeded the high-risk threshold (MARI > 0.2), with the highest value being 0.33 in three isolates (13.1%). Among other Gram-negative bacteria, the highest MARI score was 0.93, found in one isolate (9.1%), while three isolates (27.3%) showed MARI values of 0.79 (Figure 7B).

Gram-positive isolates were grouped as Enterococcus spp. (n = 32, 57.2%), Staphylococcus aureus (n = 13, 23.2%), and coagulase-negative staphylococci (n = 11, 19.6%). Resistance patterns for Staphylococcus spp. were not further analyzed due to a lack of EUCAST clinical breakpoints for some antimicrobials, and because the focus of this study is on indicator organisms.

The Enterococcus group included Enterococcus faecalis (n = 28, 87.5%), E. faecium (n = 2, 6.3%), and E. durans (n = 2, 6.3%). All isolates were fully susceptible to linezolid and vancomycin (Category A), as well as gentamicin (Category C) and ampicillin (Category D) (Figure 8).

The lowest resistance was observed in Category B, where only one isolate (3.1%) was resistant to ciprofloxacin; however, the highest level was observed in Category A, where 27 isolates (84.4%) were resistant to quinupristin–dalfopristin. Resistance to eravacycline was observed in three isolates (9.4%), while all 32 isolates (100%) exhibited intermediate susceptibility to imipenem. In Category C, eight isolates (25%) were resistant to streptomycin.

Nineteen Enterococcus spp. isolates (59.4%) had an R-score of 1, while ten (31.3%) had an R-score of 2, indicating resistance to one or two antimicrobials, respectively (Figure 9A). MARI analysis showed that ten isolates (31.3%) exceeded the threshold of 0.2, corresponding to a MARI value of 0.28 (Figure 9B).

4. Discussion

This study provides comprehensive insight into antimicrobial resistance among the indicator microorganisms, Escherichia coli and Enterococcus spp., in clinically healthy dairy cattle in Latvia, based on an analysis of fecal samples collected across farms of varying sizes. Additionally, the bulk-tank milk isolates represent pooled samples that reflect not only the herd’s status but also the milking environment. Using both rectal swab samples and pooled milk samples allowed to assess AMR both in intestinal commensals and the dairy food chain. These findings contribute to establishing national baseline AMR data for adult dairy cattle in Latvia, a population for which systematic surveillance has been limited.

Given that farm involvement was voluntary, the study population may not be fully representative of all dairy herds in Latvia, and some degree of selection bias is possible. Farms with better herd health management, higher biosecurity standards, and more prudent antimicrobial use are more likely to participate in national research studies. However, including an equal number of farms from all administrative regions and across different herd sizes still provided valuable insights into current AMR patterns and the situation in the Latvian dairy cattle sector.

Although some of the antibiotics that were used to determine resistance are not widely used in food-producing animals, they remain critically important in human medicine, aligning within surveillance according to the One Health framework. The observed resistance to these agents may be due to instinct species-specific resistance, cross-reactions with other antimicrobial drug classes, or acquired resistance mechanisms, rather than a direct selective pressure from antibiotics used for treatment [15]. Therefore, it is essential to evaluate the current distribution of resistance.

4.1. Rectal Swab Isolate Resistance Patterns

Not only can antimicrobial drug residues be transmitted through manure, but AMR genes can also be disseminated, thereby serving as a significant reservoir, particularly for Gram-negative bacteria such as E. coli and Gram-positive bacteria such as Enterococcus spp. As manure is widely used as a soil fertilizer, this practice poses a substantial risk of disseminating AMR in the environment and potentially reaching humans [5,16]. Therefore, bacterial isolates from clinically healthy animals, as normal gut commensals, are crucial targets for AMR surveillance [7,17].

4.1.1. Escherichia coli Resistance Patterns

In this study, the highest prevalence of resistance among E. coli isolates was observed against ampicillin (category D) and amoxicillin/clavulanic acid (category C), 12.5% and 6.7%, respectively, while 3.8% of isolates were resistant to cephalexin (category B). In comparison with reports from other European countries, notable differences are observed. In Austria (years 2020 and 2022), the highest resistance was reported against tetracyclines (23% and 31%), sulfonamides (20% and 24%), and penicillins (19% and 22%) [18]. According to Denmark’s national surveillance report (2023), 15% of isolates were resistant to tetracyclines and sulfonamides, while 12% were resistant to ampicillin [19].

Interestingly, in Latvia’s neighboring country, Estonia, analysis of samples collected between 2010 and 2015 revealed the highest resistance to aminoglycosides and tetracycline, at 7.0–8.8% and 7.7%, respectively [20]. In contrast, in this study, resistance to aminoglycosides and tetracycline-class antimicrobials was among the lowest (0.3%), whereas resistance to penicillin-class antimicrobials remained among the top three categories, similar to the patterns observed in Austria and Denmark.

Penicillin-group antimicrobial drugs remain among the most widely used agents in food-producing animals, not only for therapy, particularly in mastitis treatment, but also for metaphylaxis, especially during dry cow therapy [21,22,23]. For example, Werner et al. [18] found that the most frequently used antibiotics on Austrian cattle farms were cephalosporins and penicillins, which is consistent with the resistance patterns observed in their study.

In this study, however, fewer antimicrobial agents from categories C and D were tested, namely amoxicillin/clavulanic acid and ampicillin. Nevertheless, higher resistance levels were observed in these categories, suggesting more frequent use of these antibiotics. This is supported by the European Medicines Agency’s annual surveillance report for 2023 [24], which indicated that penicillin, tetracyclines, and macrolides had the highest sales volumes among food-producing animals in Latvia. However, despite the reported high sales of tetracyclines, resistance to this class remained low, suggesting that resistance levels do not necessarily directly reflect on overall sales data and farm-level usage level should be assessed.

Regarding MDR patterns, the results indicate a lower MDR burden compared with reports from other European countries. For example, in 2020, Werner et al. [18] found that 17.7% of E. coli isolates had MDR (defined as resistance to more than three antimicrobial agents). In contrast, only 2.4% isolates in present study were resistant to three or more antibiotics, corresponding to an R-score ≥ 3.

Evaluating MARI between large and medium-sized farms showed that higher MARI (>0.2) was observed in some large-scale farms. This finding is consistent with previous research. Krogh et al. [25] demonstrated that increasing herd size is associated with higher antimicrobial use. In their Danish study, antimicrobial treatment was expressed as Animal Daily Dose (ADD) per 100 animals per day, and it was shown that antimicrobial use increased by 0.07 ADD/100 animals/day for every 100 cows, indicating that larger herd sizes are associated with greater antimicrobial use. Similar trends were reported in Germany, where Hommerich et al. [26] found a positive association between herd size and antimicrobial usage.

The underlying reasons for such an association are multifactorial. However, management practices such as the use of blanket dry cow therapy (BDCT) versus selective dry cow therapy (SDCT) and the initiation of mastitis treatment without prior antibiogram and susceptibility testing may play a significant role in the development and spread of AMR [5].

4.1.2. Enterococcus spp. Resistance Patterns

Enterococcus spp. isolates can persist in soil, water, and farm environments, contributing to a widespread dissemination of AMR over extended periods. This makes this commensal microorganism a relevant indicator organism and a concern for resistance transmission to humans as E. coli [17]. Enterococcus spp. are among the most significant causes of hospital-acquired infections in human medicine and are characterized by high levels of AMR, particularly to critically important antibiotics such as vancomycin [27].

To assess resistance levels in Enterococcus spp., a different set of antimicrobial agents were used, focusing primarily on those classified as critically important in human medicine, such as carbapenems, macrolides, and glycopeptides. In this study, resistance to vancomycin and linezolid was detected in 0.5% of isolates, whereas more than half (69.2%) exhibited resistance to quinupristin/dalfopristin, representing one of the highest resistance rates reported in European studies.

In a study by Morandi et al. [28] in Italy using a comparable AMR screening panel, resistance to quinupristin/dalfopristin was detected in 36.1% of isolates, while resistance to linezolid was observed in nearly half (45.9%) of the isolates tested. However, this study was limited to 39 cows housed in an experimental facility and may therefore not reflect field conditions, where herd management practices, antimicrobial usage patterns, and personnel-related factors shape resistance development over time. Consequently, dataset from the present study may better represent long-term real-world resistance patterns under commercial farming conditions.

The high prevalence of resistance to quinupristin/dalfopristin observed most likely reflects intrinsic, rather than acquired resistance mechanisms, as some of the Enterococcus species are known to exhibit intrinsic reduced susceptibility to streptogramins. Alternatively, there may be direct selective pressure from streptogramin use in dairy cattle [29], but it is forbidden to use them in food-producing animals in the EU [12]. This resistance is plausibly connected to co-selection caused by the use of other antimicrobial classes or due to the long-term persistence of resistant enterococci in the environment of the farm as well as due to mediation by transferable resistance genes [17], but it should be noted that resistance to this antibiotic in certain Enterococcus species, such as E. faecalis has intrinsic resistance, meaning that this resistance may not be acquired or due to selective pressure from antibiotic use [30].

The origin and underlying mechanisms of quinupristin/dalfopristin resistance in Enterococcus spp. isolates in Latvia remain unclear and warrant further investigation from a One Health perspective that also targets the environment, farm equipment and milking systems and resistome of farm workers.

4.2. Raw Milk as a Reservoir of Antimicrobial Resistance

In addition to fecal isolates, raw milk samples were taken and examined to assess the AMR patterns directly relevant to food safety and potential risk for humans.

Raw milk can be a significant source of AMR, disseminated through multiple pathways, including environmental contamination of soil and water and direct exposure in humans if consumed unpasteurized or used for further processing [5,31]. In our study, members of the Enterobacteriaceae family isolated from bulk-tank milk demonstrated the highest resistance to ampicillin (45.8%), followed by amoxicillin/clavulanic acid (20.8%), which aligns with the resistance patterns observed in rectal swab isolates, at 12.5% and 6.7% respectively. These findings may reflect selective pressure connected to intramammary antibiotic use in therapy for mastitis or in the dry period [21,22]. In comparison, fecal isolates can more accurately reflect systemic antimicrobial usage [5,31]. It should be noted that bacteria from the Enterobacteriaceae family were isolated in lower numbers and not detected in all of the bulk-tank milk samples; therefore, potentially amplifying the specific resistance patterns within the mammary glands rather than reflecting herd-level resistance.

A limited number of studies have evaluated AMR, particularly in bulk-tank milk samples, especially across Europe. They mainly focus on bacterial presence, overall counts, and the presence of extended-spectrum β-lactamase-producing Enterobacteriaceae [31,32]. Some studies have evaluated the resistance levels in individual cows with subclinical or clinical mastitis [33,34]. For example, in Romania in 2022, E. coli isolated from cow milk samples suspected of mastitis showed ampicillin resistance of 28.57%, indicating moderate resistance, and higher (up to 90.8%) resistance to three additional antibiotics. Still, amoxicillin remained completely susceptible in this study [33]. In Sweden, from 2013 to 2018, the highest resistance in mastitis milk samples was also to ampicillin (8.6%), which is vastly lower than that found in this research [34]. Although milk samples from individual animals were not examined, resistance patterns were shown that were similar to those in samples obtained from animals with mastitis; however, this should be investigated further to determine if the resistant isolates in bulk-tank milk are derived from the mammary microflora or are solely from environmental contamination.

Meanwhile, resistance patterns for Enterococcus spp. slightly differ from those found in rectal swab samples. The highest resistance in both sample types was against quinupristin–dalfopristin (84.4% in milk), with the second-highest resistance in milk samples against streptomycin (25%), while in rectal swab samples, it was against eravacycline. The slight differences in resistance patterns between the milk and rectal samples likely reflect the diverse origins of milk bacteria, such as environmental factors, teat skin, bedding, and contamination from milking equipment, rather than solely fecal contamination or bacterial spread to mammary glands. These contamination sources can be exposed to different selective pressures, such as disinfectants, detergents, and water contamination, which can alter resistance patterns compared to those of intestinal microflora [34,35,36]. The same is considered for AMR connected to Enterobacteriaceae isolates.

These findings are consistent with some European studies. In Italy, Enterococcus spp. resistance in milk samples was 29.0% against quinupristin–dalfopristin and 84.4% against streptomycin; these resistance levels were different from in fecal isolates, where the highest resistance was seen against linezolid, quinupristin–dalfopristin, and vancomycin, accordingly [28]. In a study by Hanzelová et al. [36] in Slovakia, resistance to these antimicrobial drugs was not tested; however, the highest resistance was observed against ampicillin (56.8%), whereas no such resistance was found in this study.

The findings from the fecal and milk isolates highlight the current situation of AMR in the Latvian dairy production sector and the gaps in surveillance that can limit routine interpretation on the national level.

Although the European Union has established AMR surveillance requirements in veterinary clinical isolates, including E. coli as an indicator organism, under Directive 2003/99/EC [37] and subsequent decision 2013/652/EU [38] and 2020/1729/EU [39], only a limited number of European countries (e.g., Denmark, France, Germany, the Netherlands, Portugal, Sweden, and the United Kingdom) have fully implemented these measures as part of their AMR surveillance programs [5,15]. In Latvia, an AMR action plan under the “One Health” approach for 2023–2027 has been adopted; however, implementation remains partial and still under development, particularly regarding systematic surveillance in food-producing animals and raw milk.

5. Conclusions

The overall resistance levels among indicator microorganisms isolated from healthy dairy cows and milk bulk tanks in Latvia were low, with most isolates remaining susceptible to antimicrobials considered critically important in human medicine. It cannot be ruled out that farm managers and veterinarians who adopt a more prudent approach to antimicrobial use were more likely to participate in the study, potentially leading to an underestimation of resistance prevalence.

Nevertheless, the detection of sporadic multidrug-resistant isolates with MARI values above 0.2 indicates localized resistance hotspots, particularly on large farms and among environmental microorganisms isolated from raw milk.

Further studies focusing on antimicrobial use at the farm level are required to better understand the link between AMR patterns in indicator microorganisms and the selection and application of antimicrobial agents, as well as including small farms to ensure a more comprehensive national overview of AMR.

In addition, molecular detection of resistance genes would substantially improve the interpretation of observed phenotypic resistance patterns in dairy cattle, allowing to identify species specific resistance mechanisms and possibility for resistance gene horizontal transfer, and should be addressed in further studies.

These findings emphasize the necessity of integrated surveillance within a One Health framework to mitigate AMR at the animal–human–environment interface.