In the current study, the majority of Enterococcus spp. isolates that were susceptible to macrolides were found in hospital cows, similar to previous work demonstrating that Enterococcus resistance to macrolides was found in isolates from clinical animals. Moreover, all MDR Enterococcus isolates were from hospital and fresh cows, indicating that macrolide resistance genes might originate from hospital cows that are being treated for a variety of medical conditions and then spread to fresh cows. Among Enterococcus isolates, a negative association was noted between the occurrence of tetracycline and macrolide genes, indicating that the presence of the tetracycline resistance genes was associated with a reduced risk of simultaneously finding the macrolide resistance gene in these fecal bacteria. This finding is interesting given the previous observation that resistance to tetracycline and macrolide–lincosamide–streptogramin group was observed through transposable elements. In dairy production, lincosamide is used to treat mastitis in conventional farms; however, lincosamide resistance genes were found in hutch calves based on the CARD database. This may indicate the transfer of resistance genes along the production line and calves can acquire resistance genes at this early age. Similarly, it was reported that calves at 1–2 weeks of age acquired tetracycline-resistant genes, likely due to colonization with resistant bacteria from their mothers and/or the dairy farm environment, greenhouse rolling benches given the ubiquity of manure. According to genes identified from the CARD database, resistance to ≥3 antimicrobial classes genes was commonly observed among E. coli and Enterococcus spp.
No significant links between resistance to tetracycline and fluoroquinolone were observed in this study, which may be due to the mechanism of resistance to fluoroquinolone being frequently related to chromosomal mutations, while the mechanism of resistance to tetracycline can occur due to genetic mobility.For this study, we identified genes by evaluating two publicly available databases in E. coli—namely, sulphonamide, trimethoprim, and beta-lactamase resistance genes from ResFinder, and tetracycline and aminoglycoside resistance genes from both ResFinder and CARD. These resistance genotypes were in concordance with resistance phenotypes we characterized previously. For Enterococcus spp., high levels of agreement between resistance genotypes and phenotypes were only found for tetracycline resistance genes from the ResFinder . Similarly, a previous study observed a lower concordance between phenotypes and genotypes of streptomycin in Salmonella isolates. In contrast, high correlations between the presence of resistance genotypes and observed phenotypes have been reported in nontyphoidal Salmonella from retail meat specimens and human cases. Another study reported 67.9–100% concordance between resistance phenotypes and genotypes and 98.0–99.6% concordance between susceptible phenotypes and genotypes in Campylobacter from retail poultry. Although relatively few studies have been performed on Gram-positive organisms using WGS to study AMR, a high correlation between resistance genotypes and phenotypes in Enterococcus isolates was reported. The lower correlations between resistance genotypes and phenotypes of Enterococcus in the current study could be due to the small numbers of bacterial isolates tested, availability of drugs for antimicrobial susceptibility test in the commercial kits, and a different method used to analyze correlations between genotypes and phenotypes. In addition, the lower correlations also could be due to discrepancies between genotype and phenotype resistance that vary with bacterial species and antimicrobials.
Therefore, a combination of genotypes for resistance prediction with phenotypes determined by antimicrobial susceptibility would provide a more accurate assessment of resistance of different bacterial species from different samples and against different antimicrobials. Results of genotypes in the current work and phenotypes in our previous work on the same bacterial strains allowed us to better understand the resistance of E. coli and Enterococcus spp. on dairy farms.Based on phylogenetic analysis of resistance genes in E. coli detected from the ResFinder database, a quarter of the isolates that were in cluster 2A were from hutch calves. Phylogenetic analysis of resistance genes of Enterococcus detected from ResFinder also indicated a unique cluster of MDR genes mainly from hutch calves . Similarly, phylogenetic analysis of genes detected from the CARD database found distinct clusters of genes in E. coli and Enterococcus from hutch calves. Therefore, these results indicate that bacteria from hutch calves had AMR characteristics that were distinct from isolates from cattle at other stages of dairy production. Most E. coli isolates from hutch calves were MDR to aminoglycoside, phenicol, sulphonamide, and tetracycline, which is consistent with other studies in that E. coli from calves were frequently resistant to multiple antimicrobials. For example, MDR bacteria were very common from integrated veal calves. A review article indicated that young dairy calves often carry high levels of AMR in their fecal E. coli and Salmonella enterica, which could provide a potential reservoir of AMR genes for the greater dairy farm environment depending on how calf manure is managed or mixed into the general manure stream on the dairy. Our results, in addition to these prior studies and reviews, suggest that monitoring of MDR bacteria in hutch calves may be important for reducing the spread of AMR bacterial genes to other production stages in dairy farm settings. On the other hand, heat maps and phylogenetic analyses indicated a wide distribution of multiple resistance genes among multiple adult cattle production stages for fecal E. coli based on the CARD database . Given that one adult dairy cow can produce in excess of 20 to 30 kg of feces a day, conventional dairy herd sizes in California often exceed 1000 adult cows, and the concentration of fecal E. coli in dairy manure typically exceeds 106 cfu/g , one can expect that MDR fecal bacteria are widely distributed throughout the greater dairy farm environment and likely in relatively high concentrations. A previous study reported that AMR gene profiles varied between farms and different types of samples but a greater proportion of genes were common to all types of samples, suggesting horizontal transfer of common resistance genes among production stages. Samples in this study were collected within one farm at one point in time, and the sample size from each production stage was small due to the cost of WGS and available funding; these constraints may limit the representativeness of our results. However, our study warrants further investigation of the relationship between AMR clusters in different cattle groups and different types of farm sample matrices to support the effort to better control the spread of AMR within modern conventional dairy farms.In our previous work, we characterized the antimicrobial resistance phenotypes of E. coli and Enterococcus spp. from cattle at different production stages on a commercial dairy farm in Central California, USA. Briefly, using convenient sampling, fecal samples were collected from the rectum of dairy cattle at twelve different production stages on a commercial farm in the San Joaquin Valley, the major dairy production region of California.The antimicrobial susceptibility of E. coli and Enterococcus strains was determined by minimum inhibition concentrations of tested antimicrobials using a microbroth dilution method. Antimicrobials tested for E. coli were cefoxitin, azithromycin, chloramphenicol, tetracycline, ceftriaxone, amoxicillin/clavulanic acid, ciprofloxacin, gentamycin, nalidixic acid, ceftiofur, sulfisoxazole, trimethoprim–sulfamethoxazole, ampicillin, and streptomycin. Antimicrobials tested for Enterococcus were tigecycline, tetracycline, chloramphenicol, daptomycin, streptomycin, tylosin tartrate, quinupristin/dalfopristin, linezolid, nitrofurantoin, penicillin, kanamycin, erythromycin, ciprofloxacin, vancomycin, lincomycin, and gentamycin. Resistance phenotypes of E. coli and Enterococcus from the previous work were used for the analysis of associations with genotypes in the current work. In the current study, greenhouse bench top based on the availability of strains from cattle at different production stages that determined resistance phenotypes in our previous work, 40 strains of E. coli and 49 strains of Enterococcus from our culture collections were selected for genotype characterization using whole-genome sequencing .
Descriptive statistics were used to examine the distribution of AMR and MDR genes in E. coli and Enterococcus detected from the CARD and ResFinder databases. Logistic regression analysis was used to identify the relationship between the presence of various resistance genes. The production stages and bacterial species were also included as independent variables for the regression models. Univariate regression analysis for all independent variables was screened for potential significance, and a p-value threshold of 0.05 was used as an inclusion criterion in the model. Kappa coefficient analysis was used to assess the level of agreement between a bacterial isolate having a specific resistance genotype and also having the corresponding resistant phenotype for the isolates. The resistance genotypes used in the Kappa analysis were the genes detected from the CARD and ResFinder databases, while the resistance phenotypes were determined from our work on the same set of samples published previously . For the Kappa analysis, each bacterial isolates’ phenotypic resistance was compared to its corresponding pattern of resistance genes by classes of antimicrobial drugs. The percentage generated by the Kappa analysis indicated the degree of agreement between a pair of resistance phenotype and genotype class. A Kappa value of 100% indicated perfect agreement, while a Kappa 0% means no agreement between the presence of an AMR genotype class and its associated phenotype and. All statistical analyses were performed using Stata version 14 . A p-value < 0.05 was considered as statistical significance.Absorption and scattering of shortwave solar irradiance in the Earth’s atmosphere is balanced by absorption, emission, and scattering of long wave radiation. This balance between the shortwave radiation and the long wave radiation determines the temperature structure of the atmosphere and local temperature values on Earth’s surface. The SW and LW balance is essential for understanding climate change, but also for the thermal design of radiant cooling systems, cooling towers, solar power plants, and of the built environment in general. Current interest in the optical designs of passive cooling devices that take advantage of atmospheric windows to reject heat to outer space requires detailed balance between the incoming thermal radiation from the atmosphere and the outer space and the outgoing emissive power from the coolers in order to calculate the equilibrium temperatures and cooling efficiencies. These figures of merit depend on the local atmospheric conditions, which include the convective environment around the device and the downwelling radiative flux from the atmosphere. The convective contribution can be minimized by design, but the thermal radiation from the atmosphere is geometrically constrained by the ability of passive cooling devices to radiate directly to outer space. Absorption bands of water vapor dominate the absorption and emission of infrared radiation in the atmosphere when conditions are wet . When the relative humidity is low,other contributors such as CO2 and aerosols contribute in a non-negligible way through specific bands of the spectrum to the overall thermal balance of the passive cooling devices. Therefore, a detailed spectral model for the long wave radiative transfer in the atmosphere is in need to calculate the thermal balances of such optically selective devices. Two distinct solar power technologies have emerged as most competitive in the renewable energy market of utility scale solar plants: direct photovoltaic conversion and concentrated solar power using heliostat fields to direct solar radiation to a central boiler. In addition to greenhouse gas emission offset, large scale solar farms also interact with the atmosphere through surface albedo replacement. While both PV and CSP technologies affect the local environment, the extent in which they do so has not been studied in detail. Thus a spectrally resolved shortwave radiative model is needed to quantitatively evaluate the effects of albedo replacement on the local shortwave radiative exchange between the ground and the atmosphere. Therefore, to better understand the thermal balances of the Earth-atmosphere system, passive cooling devices and solar power farms, the objective of this research is to develop detailed radiative model to simulate the shortwave solar and long wave atmospheric radiative transfer in the atmosphere, with and without the presence of clouds. The model is validated against ground measurements and other radiative models for varies meteorological conditions. By comparing the modeling results with ground telemetry, representative cloud characteristics for given surface conditions are proposed. Thus a complete spectral model is presented that allows for determination of long wave and shortwave irradiance and can be used for a wide range of meteorological conditions. With the developed radiative model, the albedo replacement effects of large scale PV and CSP farms can be qualified for varies conditions. In addition, the model also serves as a valuable tool to analysis the contribution of each atmospheric constituent to the thermal balance of the Earth-atmosphere system, for seven critical bands of the infrared spectrum. Furthermore, the determination of thermal equilibrium temperatures for radiative cooling devices, also requires knowledge of the spectral atmospheric solar and long wave radiation.