Background and Objective: Antimicrobial resistance (AMR) is an important environmental and public health concern, and aquatic sediments can serve as reservoirs of antibiotic resistance genes (ARGs). Increasing evidence indicates that non-antibiotic contaminants, including heavy metals and pesticides, may also be associated with the environmental resistome. Zinc (Zn) is an essential trace element, and azinphos-methyl is an organophosphorus insecticide that may enter aquatic ecosystems through agricultural runoff. The southern Caspian Sea receives inputs from agricultural, industrial, and urban sources, and its coastal sediments provide a suitable environment for the accumulation of chemical contaminants and diverse microbial communities. However, limited information is available concerning the simultaneous spatial association of Zn and azinphos-methyl with ARGs and bacterial community structure in this region.



Aims: This study investigated the associations of Zn and azinphos-methyl with ARG abundance and bacterial community structure in coastal sediments of the southern Caspian Sea using shotgun metagenomic analysis.



Materials and Methods: Eight coastal sampling stations were selected (n = 8). Genomic DNA was extracted from 300 mg of sediment, and shotgun metagenomic libraries were sequenced on the Illumina NovaSeq 6000 platform. ARGs were identified using CARD and ResFinder, and ARG abundance was normalized as reads per million (RPM). Bacterial taxonomic composition was characterized at the phylum level. Zn concentrations were determined by ICP-MS, and azinphos-methyl was quantified using GC-MS. Alpha diversity was assessed using the Shannon index, and beta diversity was assessed using principal coordinates analysis (PCoA). Associations were evaluated using Spearman rank correlation analysis, with P < 0.05 considered statistically significant.



Results: Proteobacteria was the dominant bacterial phylum across most stations, reaching up to 94.4% relative abundance. The Shannon diversity analysis indicated that L1 and L2 had the highest diversity, whereas L8 showed the lowest diversity, with PCoA demonstrating separation of L8 from other stations. Zn was measurable at only two stations (342.5 mg/kg at L1 and 122.76 mg/kg at L3). Azinphos-methyl was detected at three stations, with the highest concentration at L4 (0.463 ng/g). Spearman correlation analysis identified eight ARGs significantly and positively associated with Zn, with qacA showing the strongest association (ρ = 0.990, P < 0.001). Azinphos-methyl showed significant positive correlations with 25 ARGs, with bacA showing the strongest association (ρ = 0.892, P = 0.0029). At the bacterial phylum level, Zn showed four significant positive associations (SAR324 clade, Margulisbacteria, LCP-89, and Spirochaetota), whereas azinphos-methyl showed a significant negative association with Bacteroidota (ρ = −0.810, P = 0.015).



Discussion: The observed correlations indicate that spatial variation in Zn and azinphos-methyl concentrations was accompanied by differences in the abundance of several ARGs. The strongest Zn–ARG association was observed for qacA, a gene associated with multidrug efflux systems. Azinphos-methyl showed significant positive associations with a larger number of ARGs than Zn, with the strongest association observed for bacA. The bacterial community analysis provided an additional ecological context for the observed ARG patterns. However, such mechanisms cannot be directly inferred from the observational metagenomic data generated in the present study. The interpretation of the findings should particularly consider the limited sample size and the distribution of the investigated contaminants.



Conclusion: This study investigated the spatial associations of Zn and azinphos-methyl with ARGs and bacterial community structure in coastal sediments of the southern Caspian Sea. Zn showed significant positive correlations with eight ARGs, with qacA showing the strongest association (ρ = 0.990, P < 0.001). Azinphos-methyl showed significant positive correlations with 25 ARGs, with bacA showing the strongest association (ρ = 0.892, P = 0.0029). At the bacterial phylum level, Zn was positively associated with SAR324 clade, Margulisbacteria, LCP-89, and Spirochaetota, whereas azinphos-methyl showed a significant negative association with Bacteroidota. Overall, the findings demonstrate statistical co-variation among non-antibiotic contaminants, ARG abundance, and bacterial community structure. However, the observational design, the small number of independent samples (n = 8), and the highly uneven distribution of Zn and azinphos-methyl require cautious interpretation. The identified relationships should be considered exploratory and hypothesis-generating candidate associations rather than evidence of causal effects.