4.1. Prevalence Data
With 72 farms from eight different German federal states, comprising 2816 individual samples of pre-weaned dairy calves and their dams, our study represents the largest study on the fecal ESBL/AmpC-
E. coli prevalence in German dairy farms to date. The results provide evidence for the occurrence of ESBL/AmpC-
E. coli in 100% of the examined dairy farms. The herd prevalence was higher than that of all previous European prevalence studies [
11,
13,
14,
15,
16,
22,
23,
24]. However, in those studies the dairy farms investigated were all, on average, considerably smaller in size and lower in the milk production performance level compared to the farms included in the present study. In contrast to other studies, our study omitted the enrichment of the sample material prior to cultural investigations. Omitting the enrichment step in our study could lead to an underestimation of prevalence. Therefore, our result that in 100% of the herds at least one animal tested positive for ESBL/AmpC-
E. coli is concerning. Only one other study, a Swedish study, used the same approach and determined a herd prevalence of 18% [
15]. Until now, the study with the highest ESBL/AmpC-
E. coli herd prevalence in European dairy herds (93%) was also carried out in southern Germany [
13]. In Germany, there is a north–south divide in cow numbers on dairy farms. In the north of Germany, there are larger farms with more than 100 cows, whereas in the south the herds are smaller. This is the first study with individual samples in large dairy herds in Germany. Due to convenient sampling, there could be some bias regarding the estimation of prevalence in our study. However, in a quarter of the herds (
n = 18) all calves present were sampled (
Supplement 1).
In the present study, the number of ESBL/AmpC-
E. coli-positive calves per farm varied between a single animal and all sampled animals, with an average calf prevalence of 63.5%. This enormous range in variation has been observed previously in other studies using a similar methodology [
15,
25,
26]. Although much lower on average (18.0%), prevalence variations in dams also showed a wide range (0% and 88.9%). This finding was consistent with a Czech study from Manga et al. (2019) where a maternal prevalence of 15% was determined. However, that study only investigated a single dairy farm [
26]. Other studies differed in the methodology used, which allows no comparison with the work described here [
13,
16,
27,
28,
29]. Further studies are needed to determine whether the ESBL/AmpC-
E. coli colonization of the calf originates from the dam or from other sources.
4.2. Risk Factors for the Occurrence of ESBLAmpC-E. coli in Calves
The most important risk factor associated with high prevalence of ESBL/AmpC-
E. coli in calves was the feeding of WM, whereas the use of MR was associated with a low ESBL/AmpC-
E. coli prevalence. Although prohibited for human consumption, feeding of WM to calves is a common practice worldwide [
30,
31], as it is both legal and economically beneficial. In general, WM often contains antibiotic residues, albeit far below the therapeutically effective dose [
32,
33]. It is known that feeding milk containing even very low concentrations of antibiotic residues can provide a selection advantage for multidrug-resistant pathogens [
34]. In a study by Pereira and colleagues, resistant bacteria were found in a significantly greater proportion of calves which were fed with antibiotic-spiked milk compared to calves, which were fed raw milk without antimicrobials. Moreover, they detected a clear difference in the calves’ microbial colonization profiles [
35]. Additionally, in 2014, Brunton et al. detected a higher amount of ESBL-
E. coli in the feces of calves fed with WM, as well as a prolonged excretion of the multidrug-resistant pathogens after weaning, and considered a selective effect of WM, although it is unclear at what concentrations the antibiotic residues of WM reach the intestines of the suckling calves [
32]. Likewise, contamination of WM with ESBL-
E. coli has been shown to occur, which could also lead to oral colonization of suckling calves [
36]. However, if the intake of vital ESBL-
E. coli via WM was causal for an increased prevalence in calves, pasteurization should be a protective factor, which could not be proven by several studies [
37,
38].
Our finding that feeding MR was associated with a low ESBL/AmpC-
E. coli prevalence was consistent with a Swedish study by Duse and colleagues in 2015, in which calves fed with MR excreted significantly less ESBL-
E. coli than those fed with raw milk [
15].
Another parameter significantly associated with a farm being classified as having a high ESBL/AmpC-
E. coli prevalence in calves was daily cleaning of the calf-feeding equipment, such as nursing buckets. This result is supported by a study by Heinemann and colleagues (2020), which showed that current hygiene practices in dairy farms are often inadequately implemented. In that study, the highest bacterial loads, including those of ESBL-
E. coli, were detected in the milk-feeding buckets and inner artificial teats, although sanitation measures were in place [
39]. It is conceivable that insufficient cleaning leads to a transmission of bacterial colonization between calves, especially when nursing buckets are not permanently allocated to the same calf. Another possible explanation is that the feeding equipment is contaminated through the use of a single batch of cleaning solution for all equipment, promoting a microbial transmission. Although this theory of the bacterial contamination of feeding equipment is as yet unproven, bacterially contaminated teat cleaning solutions have been identified as a cause of mastitis [
40].
The survey conducted among the farmers suggests an association between high ESBL/AmpC-
E. coli colonization of the calves and the general application of preventive treatment to dairy cows on the farms. However, due to the wide range of possible prophylactic actions taken between the participants (e.g., milk fever/ketosis/acidosis prophylaxis), interpretation of this parameter is difficult. Nevertheless, regardless of the exact treatment regimen, disease prevention measures appear to improve maternal fitness and thereby lead to a decrease in ESBL/AmpC
E. coli prevalence. One possible explanation for this association is that farms which practice the general application of preventative treatments also have a generally higher quality of farm management and therefore cows with overall greater fitness. It is assumed that cows with a greater fitness are less colonized with ESBL/AmpC-
E. coli, which could result in a lower infection risk for the newborn calves at the time of birth. Moreover, cows from well managed farms are potentially in a better constitution for parturition and are monitored better during birth. This might result in easier births and subsequently fitter and healthier calves, given that difficult births are known to weaken calves [
41]. It is also conceivable that cows with a high fitness respond immunologically well to their environment and secrete high-quality colostrum. This in turn has a positive effect on the newborn calves and makes them more resistant to colonization with ESBL/AmpC-
E. coli present in their environment [
42].
4.3. Risk Factors for the Occurrence of ESBLAmpC-E. coli in Cows
Another parameter aiming to achieve a better fitness is the result of the linear relationship showing a significantly negative correlation between the actual lifetime production of the dairy herd and the maternal ESBL/AmpC-
E. coli prevalence. On the one hand, this observation may point to better hygiene, housing, feeding and health management within the dairy herds, leading to greater general health, well-being and fitness of the cows, resulting in a lower susceptibility to ESBL/AmpC-
E. coli [
43]. On the other hand, a possible explanation for the lower prevalence of ESBL/AmpC-
E. coli in dams on farms with higher actual lifetime production could be the lower use of antibiotic therapeutics within these farms due to a generally better physical constitution of these animals. The positive relationship between the culling rates and ESBL/AmpC-
E. coli prevalence of the dams supports the association between better health and the lower occurrence of multi-resistant bacteria.
Farms that practiced dry teat cleaning before milking showed significantly lower ESBL/AmpC-
E. coli prevalence in the examined dams. A prerequisite for dry teat cleaning is good barn hygiene. Therefore, a possible explanation for the lower maternal prevalence in herds with dry teat cleaning could be a better hygiene standard in general. Inadequate farm cleaning has repeatedly been identified as a risk factor for increased ESBL-
E. coli prevalence [
16,
44]. Moreover, the practice of wet teat cleaning has the potential for promoting bacterial transmission due to contaminated cleaning solutions, which could result in higher ESBL/AmpC-
E. coli pressure within the dairy herd [
40].
Surprisingly, disinfection of the calving area increases the occurrence of ESBL/AmpC-
E. coli in fresh cows, although the use of biocides is known to be a reliable method for the reduction of pathogenic bacteria [
45]. Unfortunately, the questionnaire did not record the frequency of cleaning and disinfection in the calving area. One possible explanation for the increased prevalence in the dams could be the use of the disinfection measure to compensate for inadequate and reduced cleaning frequency of the calving pen. It is also conceivable that frequent disinfection leads to ESBL/AmpC-
E. coli developing decreased sensitivity to biocidal products, as this has been shown to occur for other bacteria [
46]. Biocide resistance can cause selection advantages, which result in an increased growth of ESBL/AmpC-
E. coli, leading to increased colonization rates of the dams. Considering that both the daily cleaning of the calf’s feeding equipment and the disinfection of the calving area was associated with high ESBL/AmpC-
E. coli prevalence in calves and in dams, respectively, suggest that inaccurate cleaning and disinfection procedures carry the risk of worsening the prevalence of ESBL/AmpC-
E. coli.
Finally, self-production of basic feeds increased the ESBL/AmpC-
E. coli prevalence in the dams. In each examined farm, the feed ration of the dairy herd was based on corn and grass silage. More than half of the participants self-produced the silages on their own farmland using manure from the calves and dairy cows as organic fertilizers. This may lead to the steady circulation of ESBL/AmpC-
E. coli between livestock and farmland, as microbial contamination of both grassland and cattle feed has already been proven in the past [
47,
48]. The constant re-introduction of resistant bacteria via the self-produced feed rations could thus cause increased maternal colonization rates.