Background: In the environment, microbial communities are constantly exposed to invasion by antimicrobial resistant bacteria (ARB) and their associated antimicrobial resistance genes (ARGs) that were enriched in the anthroposphere. A successful invader has to overcome the biotic resilience of the habitat, which is more difficult with increasing biodiversity. The capacity to exploit resources in a given habitat is enhanced when communities exhibit greater diversity, reducing opportunities for invaders, leading to a lower persistence. In the context of antimicrobial resistance (AMR) dissemination, exogenous ARB reaching a natural community may persist longer if the biodiversity of the autochthonous community is low, increasing the chance of ARGs to transfer to community members. Reciprocally, high microbial diversity could serve as a natural long-term barrier towards invasion by ARB and ARGs. Results: To test this hypothesis, a sampling campaign across seven European countries was carried out to obtain 172 environmental samples from sites with low anthropogenic impact. Samples were collected from contrasting environments: stationary structured forest soils, or dynamic river biofilms and sediments. Microbial diversity and relative abundance of 27 ARGs and 5 mobile genetic element marker genes were determined. In soils, higher diversity, evenness and richness were all significantly negatively correlated with the relative abundance of the majority (>85%) of ARGs. Furthermore, the number of detected ARGs per sample was inversely correlated with diversity. However, no such effects were found for the more dynamic, regularly mixed rivers. Conclusions: In conclusion, we demonstrate that diversity can serve as barrier towards AMR dissemination in the environment. This effect is mainly observed in stationary, structured environments, where long-term, diversity-based resilience against invasion can evolve. Such barrier effects can in the future be exploited to limit the environmental proliferation of AMR.

Plant-derived foods are potential vehicles for microbial antibiotic resistance genes (ARGs), which can be transferred to the human microbiome if consumed raw or minimally processed. The aim of this study was to determine the prevalence and the amount of clinically relevant ARGs and mobile genetic elements (MGEs) in differently processed smoothies (freshly prepared, cold-pressed, pasteurized and high-pressure processed) and fresh produce samples (organically and conventionally cultivated) to assess potential health hazards associated with their consumption. The MGE ISPps and the class 1 integron-integrase gene intI1 were detected by probe-based qPCR in concentrations up to 10⁴ copies/mL in all smoothies, lettuce, carrots and a single tomato sample. The highest total (2.2 × 10⁵ copies/mL) and the most diverse ARG and MGE loads (16/26 targets) were observed in freshly prepared and the lowest prevalences (5/26) and concentrations (4.1 × 10³ copies/mL) in high-pressure-processed (HPP) smoothies. Bla_CTX-M-1-15 (1.2 × 10⁵ c/mL) and strB (6.3 × 10⁴ c/mL) were the most abundant, and qacEΔ1 (95%), bla_TEM1 (85%), ermB and sul1 (75%, each) were the most prevalent ARGs. QnrS, vanA, sat-4, bla_KPC, bla_NDM-1 and bla_OXA-10 were never detected. HPP treatment reduced the microbial loads by ca. 5 logs, also destroying extracellular DNA potentially encoding ARGs that could otherwise be transferred by bacterial transformation. The bacterial microbiome, potential pathogens, bacterial ARG carriers and competent bacteria able to take up ARGs were identified by Illumina 16S rRNA gene sequencing. To reduce the risk of AMR spread from smoothies, our data endorse the application of DNA-disintegrating processing techniques such as HPP.

Rivers receive continuous inputs of antibiotic-resistant bacteria (ARB) from wastewater, agriculture, and other anthropogenic sources, yet it remains unclear whether the recipient ecological component and its microbial communities determine whether introduced ARB establish or disappear. Ecological invasion theory predicts that invasion success depends on biodiversity, community stability, and occupation of ecological niche space, but these mechanisms have rarely been evaluated together in natural microbial communities. Here, we challenged river biofilms collected from 20 sites in 12 European rivers across six countries with a model antibiotic-resistant Escherichia coli carrying a conjugative IncP-1α plasmid The invasion assays were carried out under standardized laboratory flume conditions. River biofilms differed markedly in their permissiveness to invasion despite identical invasion conditions. Higher bacterial diversity consistently accelerated invader loss rates, whereas communities containing more abundant and diverse close phylogenetic neighbours of the invader exhibited stronger exclusion during early biofilm establishment. Diversity loss during transition into the experimental system emerged as the strongest explanatory variable of invasion resistance prior to biofilm maturation, whereas Shannon diversity became the dominant predictor in mature communities. Integrating these complementary ecological dimensions substantially improved explanatory prediction of ARB persistence compared with individual predictors alone. Particularly invasion-resistant biofilms also exhibited distinct ecological community composition consistent with mature, structurally complex microbial assemblages. Together, our findings demonstrate that the establishment of ARB in the environment is not stochastic but can be predicted from measurable ecological properties of recipient microbiomes, highlighting microbial biodiversity and community organization as natural barriers to antimicrobial resistance dissemination.