Local Temperature and Humidity are Associated with Proportion of Antimicrobial-Resistant Escherichia coli isolates in Farm Environments: Considerations for On-Farm Surveillance

Evidence suggests that increased local temperatures are associated with higher prevalence of antimicrobial resistance (AMR) in environmental bacteria. This study investigates the association between local climate and the proportion of antimicrobial-resistant Escherichia coli isolated from 2,766 farm environment samples from 53 English dairy farms. To do this, a non-linear Bayesian model that specifically accounts for decreased test sensitivity at low E. coli abundance was developed and used to estimate the proportion of isolates resistant to four antimicrobials (amoxicillin, cephalexin, streptomycin and tetracycline) from colony count data. Mean 7-day temperature and relative humidity at the farm location was modelled using a generalised additive model formulation. A higher proportion of E. coli isolates were resistant to cephalexin and streptomycin in samples collected from adult cow collecting yards, than heifer housing sheds. In contrast, a greater proportion of E. coli isolates from heifer housing sheds were resistant to amoxicillin and tetracycline. Evidence that local temperature is associated with an increase in the proportion of E. coli isolates resistant to streptomycin (20{degrees}C increase associated with a 5.0-fold increase; 95% CI: 1.03-33.0) and tetracycline (2.6-fold increase; 90% CI: 1.1-5.2) was observed. Additionally, relative humidity was associated with an increase in the proportion of isolates resistant to amoxicillin streptomycin and tetracycline. The influence of weather on the proportion of antimicrobial-resistant E. coli varied between samples collected from adult animals in collecting yards and heifers in housing sheds. These findings highlight the importance of considering weather conditions, sample characterises and seasonality when designing on-farm AMR surveillance systems. ImportanceUnderstanding how environmental conditions are associated with variability in AMR prevalence is critical for developing robust livestock AMR surveillance and anticipating the potential effects of climate change. The non-linear Bayesian modelling approach developed here adjusts for E. coli abundance associated variability in test sensitivity, enabling the influence of risk factors associated with the proportion of antimicrobial-resistant E. coli within samples to be more accurately estimated. Applying this approach to 2,766 faecal samples from 53 dairy farms in Southwest England indicated that the proportion of antimicrobial-resistant E. coli generally increased under warmer and wetter conditions. These findings suggest that environmental conditions can influence the prevalence of AMR E. coli in dairy farm environments and demonstrate the importance of accounting for weather related variability in livestock AMR surveillance. Adjusting for these associations in livestock AMR surveillance could improve the accuracy of modelling AMR trends and strengthen the assessment of climate-associated AMR risks.