Assessing the sentinel capacity of molluscs, seawater and sediments for monitoring antimicrobial resistance in the marine environment

Gabriel Arriagada^1*Libertad Canales²Fabián Flores²Nicolás Reyes³Ismael Maldonado¹Lisette Lapierre⁴Nicolás Galarce⁴

• ¹Institute of Agri-food, Animal and Environmental Sciences, Universidad de O'Higgins, San Fernando, Chile
• ²School of Agri-food, Animal and Environmental Sciences, Universidad de O'Higgins, San Fernando, Chile
• ³Master in Biological Sciences, Faculty of Sciences, Universidad del Bio-bio, Chillan, Chile
• ⁴Department of Preventive Animal Medicine, Faculty of Veterinary and Livestock Sciences,, Universidad de Chile, Santiago, Chile

Introduction: Although it is widely recognized that aquaculture activity is particularly relevant to the development and spread of AMR in the aquatic environment, national action plans against AMR typically do not include the marine environment among the compartments targeted for AMR monitoring and surveillance. This study aimed to compare the sentinel capacity of molluscs, seawater, and sediments for AMR surveillance in the marine environment of the Los Lagos region, Chile, using Escherichia coli as indicator bacteria and florfenicol, oxytetracycline, oxolinic acid, and flumequine as target antimicrobials. Methods: 101 mollusc, 76 seawater and 76 sediment samples were collected simultaneously from 76 sites in the coastal area of the Los Lagos region, Chile, between 2023 and 2024. All samples were subjected to conventional laboratory procedures for E. coli isolation. Minimum inhibitory concentrations (MICs) for florfenicol, oxytetracycline, oxolinic acid, and flumequine were estimated for each E. coli isolate, which was then classified as either "wild-type" (WT) (i.e., susceptible) or "non-wild-type" (NWT) (i.e. tolerant) based on local epidemiological cutoff values (COwt) calculated from the MIC results. The frequency of NWT E. coli isolates was calculated for each of the three compartments; significant differences in the probabilities of isolating E. coli and detecting NWT E. coli were assessed using logistic regression models. Results: E. coli was isolated in 82.2% of the mollusc, 93.4% of the seawater, and 38.7% of the sediment samples. The COwt values were estimated in 32 μg/mL for florfenicol, 64 μg/mL for oxytetracycline, 1 μg/mL for oxolinic acid, and 2 μg/mL for flumequine. The proportion of NWT E. coli among the four antimicrobials was consistently higher in seawater (25.0% on average), followed by sediments (10.8%) and then molluscs (5.4%). Logistic models indicated that the probabilities of isolating E. coli and detecting NWT E. coli for the four antimicrobials studied significantly depend on the environmental compartment, with seawater having the highest probability. These results should be considered by authorities developing plans to monitor AMR in the marine environment.