Agricultural Soils Amended With Thermally-Dried Anaerobically-Digested Sewage Sludge Showed Increased Risk of Antibiotic Resistance Dissemination

Abstract

The application of sewage sludge (SS) to agricultural soil can help meet crop nutrient requirements and enhance soil properties, while reusing an organic by-product. However, SS can be a source of antibiotic resistance genes (ARGs) and mobile genetic elements (MGEs), resulting in an increased risk of antibiotic resistance dissemination. We studied the effect of the application of thermally-dried anaerobically-digested SS on (i) soil physicochemical and microbial properties, and (ii) the relative abundance of 85 ARGs and 10 MGE-genes in soil. Soil samples were taken from a variety of SS-amended agricultural fields differing in three factors: dose of application, dosage of application, and elapsed time after the last application. The relative abundance of both ARGs and MGE-genes was higher in SS-amended soils, compared to non-amended soils, particularly in those with a more recent SS application. Some physicochemical parameters (i.e., cation exchange capacity, copper concentration, phosphorus content) were positively correlated with the relative abundance of ARGs and MGE-genes. Sewage sludge application was the key factor to explain the distribution pattern of ARGs and MGE-genes. The 30 most abundant families within the soil prokaryotic community accounted for 66% of the total variation of ARG and MGE-gene relative abundances. Soil prokaryotic α-diversity was negatively correlated with the relative abundance of ARGs and MGE-genes. We concluded that agricultural soils amended with thermally-dried anaerobically-digested sewage sludge showed increased risk of antibiotic resistance dissemination.

Keywords: antibiotic resistance genes; emerging contaminants; mobile genetic elements; organic fertilization; soil microbial diversity; soil quality.

FIGURE 1

Relative abundances of ARGs and MGE-genes in amended soils, unamended soils and SS.

FIGURE 2

Radial diagram from RDA data. Solid lines represent a significant (p < 0.05) effect of the corresponding variable. Dotted lines represent the lack of significant effect. The variable “SS management” includes the dosage of application, the elapsed time after the last application, and the presence/absence of SS. ARGs and MGE-genes: relative abundance (relative to the 16S rRNA gene) of ARGs and MGE-genes. α-diversity: richness, Shannon’s diversity and Simpson’s diversity. Physicochemical properties: OM, pH, CEC, EC, NO3^–, NH4⁺, total N, Olsen P, K⁺, texture, and pseudo-total metal concentrations (Cd, Cr, Cu, Ni, Pb, and Zn).

FIGURE 3

Biplot of the RDA performed with SS management (i.e., dosage of application, elapsed time after the last application, and presence/absence of SS) as explanatory variables, ARGs and MGE-genes relative abundances as response variables, and field and current crop as covariates. Only statistically significant explanatory variables and response variables with the best fit are shown. The explanatory variables explained 23.1% of the variation in ARG and MGE-gene relative abundances.

FIGURE 4

Biplot of the RDA performed with soil physicochemical properties (explaining 43.3% of the variation in ARG and MGE-gene relative abundances) as explanatory variables, the relative abundance of ARGs and MGE-genes as response variables, and field and current crop as covariates. Only statistically significant explanatory variables and response variables with the best fit are shown.

FIGURE 5

Biplot of the RDA performed with the 30 most abundant families (explaining 65.8% of the variation in ARG and MGE-gene abundances) as explanatory variables, the relative abundance of ARGs and MGE-genes as response variables, and field and current crop as covariates. Only statistically significant explanatory variables and response variables with the best fit are shown.

FIGURE 6

Biplot of the RDA performed with α-diversity indices (explaining 18.9% of the variation in the abundance of ARGs and MGE-genes) as explanatory variables, the relative abundance of ARGs and MGE-genes as response variables, and field and current crop as covariates. Only statistically significant explanatory variables and response variables with the best fit are shown.

FIGURE 7

Network analysis of ARGs, MGE-genes (relative abundances) and 12 multi-resistant bacterial families based on Kendall’s correlations. Node size is proportional to the number of connections (degree). An edge represents a significant correlation, where the edge thickness is proportional to Kendall’s correlation coefficient (weight).

FIGURE 8

Structural equation models showing direct and indirect effects of prokaryotic diversity indices, dosage of SS application, elapsed time after the last application, HM pollution index, PC1 and PC2 on ARGs and MGE-genes. Arrows indicate positive and negative relationships by solid and dashed lines, respectively. Numbers next to arrows represent standardized estimated regression weights (p < 0.05). The r²-values indicate the proportion of variance explained for each variable. χ² = 31.5, AIC = 133, P = 0.416, df = 34.