Abstract
Antimicrobial resistance is increasingly understood as an environmental as well as a clinical problem, yet the distribution of antibiotic resistance genes (ARGs) in Australian urban waterways remains poorly characterised. This thesis extract reports a cross-sectional survey of a hypothetical peri-urban catchment in temperate south-eastern Australia, in which water and sediment were sampled at twelve sites spanning forested headwaters, residential reaches and a wastewater treatment plant (WWTP) outfall. Quantitative PCR was used to measure four resistance genes (sul1, tetA, blaTEM and the class 1 integron-integrase gene intI1) alongside the 16S rRNA gene. Absolute and normalised ARG abundances increased by up to three orders of magnitude near the outfall and declined with downstream distance. Treated-effluent fraction, faecal indicator bacteria and nutrient concentrations were the strongest correlates of ARG abundance. The findings position treated wastewater as the dominant local driver and support catchment-scale surveillance within Australia’s One Health response to antimicrobial resistance.
Introduction
Antimicrobial resistance (AMR) is one of the most serious threats to human and animal health worldwide, and the World Health Organization has identified it as a priority requiring coordinated global action (World Health Organization [WHO], 2015). In Australia, the clinical burden of resistant infections is tracked through the Antimicrobial Use and Resistance in Australia (AURA) surveillance system (Australian Commission on Safety and Quality in Health Care [ACSQHC], 2021), and the national response is explicitly framed within a One Health paradigm that links human, animal and environmental health (Commonwealth of Australia, 2020). The environmental compartment, however, remains the least monitored, despite growing evidence that waterways act as reservoirs and dispersal routes for antibiotic resistance genes (ARGs) (Berendonk et al., 2015).
Urban catchments are of particular concern because they concentrate the principal sources of ARGs: human faecal waste, treated and untreated wastewater, stormwater runoff and, in some settings, agricultural inputs. Wastewater treatment plants (WWTPs) receive antibiotics, resistant bacteria and mobile genetic elements from entire urban populations, and even well-operated plants discharge measurable ARG loads to receiving waters (Rizzo et al., 2013). Yet most quantitative surveys have been conducted in Europe, North America and China, and the behaviour of ARGs in Australian urban waterways has received comparatively little attention.
This thesis extract addresses that gap through a catchment-scale survey. The aims were to: first, quantify the abundance of representative ARGs and the class 1 integron-integrase gene along an urban-to-estuarine gradient; second, test whether treated-effluent inputs, land use and nutrient status explain their spatial distribution; and third, interpret the results for catchment management within the Australian One Health and water-quality policy setting. The scope is observational, characterising spatial patterns rather than mechanisms or health outcomes.
Literature Review
Three themes dominate the environmental AMR literature and frame the present study. The first is the characterisation of WWTPs as resistance hotspots. Rizzo et al. (2013) synthesised evidence that the treatment process, while reducing overall bacterial numbers, can enrich the relative abundance of resistant organisms and mobile genetic elements, so that effluent remains an important point source. Pruden et al. (2006) were among the first to argue that ARGs should be treated as contaminants in their own right, demonstrating elevated tetracycline resistance genes downstream of urban and agricultural inputs.
The second theme concerns suitable molecular indicators. The class 1 integron-integrase gene, intI1, has been proposed as a proxy for anthropogenic pollution because it co-occurs with diverse resistance cassettes and correlates with human impact across many environments (Gillings et al., 2015). The sulfonamide resistance gene sul1, frequently embedded within class 1 integrons, is similarly abundant and persistent, making it a practical target for quantitative surveys.
The third theme is the relationship between resistance and faecal contamination. Karkman et al. (2019) argued that ARG abundances in many impacted waters can be explained largely by the extent of faecal pollution, implying that resistance patterns often track sewage inputs rather than in situ selection. At larger scales, Zhu et al. (2017) documented continental pollution of estuaries with ARGs, linking gene profiles to population density and wastewater discharge. Australian evidence remains comparatively sparse; national assessments emphasise the One Health importance of environmental surveillance (Commonwealth Scientific and Industrial Research Organisation [CSIRO], 2020) without yet providing dense, catchment-scale datasets. This study contributes such a dataset for a temperate Australian catchment.
Methodology
Study area and sampling design
The study catchment is a hypothetical peri-urban catchment of approximately 78 square kilometres in temperate south-eastern Australia, draining from forested headwaters through residential suburbs to a tidal estuary. A secondary-treatment WWTP discharges to the lower freshwater reach. Twelve sampling sites (S1 to S12) were positioned to capture the land-use gradient, from a forested reference reach (S1), through residential and stormwater-influenced reaches, to the outfall mixing zone (S8) and the downstream estuarine sites (S9 to S12), as detailed in Table 1. Site selection followed the spatial logic recommended for water-quality assessment under the Australian and New Zealand guidelines for fresh and marine water quality (ANZG, 2018).
Sample collection and processing
At each site, triplicate surface-water grab samples (1 L) and surficial sediment samples were collected on three occasions during the 2024 autumn baseflow period, giving 36 water samples in total. Samples were transported on ice and processed within six hours. Water was filtered through 0.22 micrometre membranes to capture the microbial fraction, and community DNA was extracted from filters and sediment using a commercial soil and water DNA kit following the manufacturer’s protocol. Physico-chemical variables, namely total nitrogen, total phosphorus and Escherichia coli as a faecal indicator, were measured concurrently to characterise each site.
Gene quantification by qPCR
Absolute abundances of four resistance genes, the sulfonamide gene sul1, the tetracycline gene tetA, the beta-lactamase gene blaTEM and the class 1 integron-integrase gene intI1, were quantified by quantitative PCR (qPCR), alongside the bacterial 16S rRNA gene. Standard curves were prepared from plasmid dilutions spanning seven orders of magnitude, and abundances were expressed as gene copies per millilitre of water. Relative abundance was calculated by normalising each ARG to the 16S rRNA gene, which controls for differences in total bacterial biomass between sites. Figure 1 summarises the workflow from site selection to statistical analysis.
Quality assurance and quality control
Quality assurance and quality control (QA/QC) followed standard practice for environmental molecular work. Every qPCR run included no-template controls and duplicate reactions; assays were accepted only where amplification efficiencies fell between 90 and 110 per cent and correlation coefficients exceeded 0.99. The limit of detection was defined as the lowest standard reliably amplified, and values below this threshold were reported as not detected. Field and extraction blanks were processed identically to samples to identify any contamination.
Statistical analysis
Gene abundances were log-transformed to approximate normality. Spearman rank correlations were used to test associations between ARG abundances and environmental variables, because the relationships were monotonic but not necessarily linear. Differences among reaches were considered significant at p < 0.05. Analyses were conducted in R.
Results
ARG abundances varied by up to three orders of magnitude across the catchment and displayed a clear spatial gradient centred on the WWTP outfall (Table 1). At the forested reference site (S1), sul1 was near the lower end of quantification at 2.1 × 102 copies/mL and blaTEM was not detected. Abundances rose progressively through the residential and stormwater-influenced reaches and peaked sharply at the outfall mixing zone (S8), where sul1 reached 4.8 × 105 copies/mL and intI1 reached 6.2 × 105 copies/mL. Downstream of the outfall, abundances declined with distance but remained well above reference levels as far as the estuary mouth (S12).
Table 1: Absolute abundances of the 16S rRNA gene and three representative resistance genes in water at twelve sites, ordered along the catchment gradient from forested headwaters to the estuary mouth.
| Site | Land-use context | 16S rRNA (copies/mL) | sul1 (copies/mL) | intI1 (copies/mL) | blaTEM (copies/mL) |
|---|---|---|---|---|---|
| S1 | Forested headwater (reference) | 2.0 × 106 | 2.1 × 102 | 3.5 × 102 | n.d. |
| S2 | Rural-residential | 3.0 × 106 | 5.0 × 102 | 8.0 × 102 | 6.0 × 101 |
| S3 | Suburban tributary | 5.0 × 106 | 1.8 × 103 | 2.6 × 103 | 2.2 × 102 |
| S4 | Suburban tributary | 7.1 × 106 | 3.2 × 103 | 4.5 × 103 | 4.0 × 102 |
| S5 | High-density residential | 1.0 × 107 | 9.5 × 103 | 1.3 × 104 | 1.5 × 103 |
| S6 | Stormwater confluence | 1.4 × 107 | 2.4 × 104 | 3.1 × 104 | 4.2 × 103 |
| S7 | Upstream of outfall | 1.3 × 107 | 2.0 × 104 | 2.8 × 104 | 3.8 × 103 |
| S8 | WWTP outfall mixing zone | 7.9 × 107 | 4.8 × 105 | 6.2 × 105 | 9.5 × 104 |
| S9 | 0.5 km downstream | 4.0 × 107 | 1.9 × 105 | 2.5 × 105 | 3.6 × 104 |
| S10 | 2 km downstream | 2.0 × 107 | 6.0 × 104 | 8.4 × 104 | 1.1 × 104 |
| S11 | Lower estuarine reach | 1.2 × 107 | 1.5 × 104 | 2.1 × 104 | 2.4 × 103 |
| S12 | Estuary mouth | 7.1 × 106 | 4.2 × 103 | 6.0 × 103 | 5.5 × 102 |
Normalising to the 16S rRNA gene confirmed that the outfall signal reflected genuine enrichment rather than higher bacterial biomass alone. The relative abundance of sul1 rose from 2.1 × 102 / 2.0 × 106 = 1.1 × 10-4 at S1 to 4.8 × 105 / 7.9 × 107 = 6.1 × 10-3 at S8, an approximately 55-fold increase in the proportion of the community carrying the gene. The integron marker intI1 tracked sul1 closely across all sites, consistent with their frequent physical linkage on class 1 integrons.
Spearman correlations identified the treated-effluent fraction as the strongest single correlate of resistance, followed by faecal indicator bacteria and nutrient concentrations (Table 2). All associations were positive and statistically significant. The treated-effluent fraction explained ARG abundance more strongly than impervious catchment cover, indicating that point-source wastewater, rather than diffuse urban runoff alone, dominated the local resistance signal.
Table 2: Spearman rank correlation coefficients between environmental variables and the abundance of three resistance genes across the twelve sites.
| Environmental variable | sul1 | intI1 | blaTEM |
|---|---|---|---|
| Impervious catchment cover (%) | 0.72* | 0.75** | 0.68* |
| Treated-effluent fraction (%) | 0.91*** | 0.88*** | 0.85*** |
| Total nitrogen (mg/L) | 0.84** | 0.82** | 0.77** |
| Total phosphorus (mg/L) | 0.79** | 0.76** | 0.70* |
| Escherichia coli (MPN/100 mL) | 0.87*** | 0.85*** | 0.81** |
Note. Spearman rank correlation coefficients; n = 12 sites. Significance: * p < 0.05; ** p < 0.01; *** p < 0.001.
Discussion
The results provide clear evidence that treated wastewater is the dominant local driver of ARG abundance in this catchment. The sharp peak at the outfall, the downstream decay, and the strong correlation with treated-effluent fraction together point to the WWTP as a point source, echoing the hotspot model advanced by Rizzo et al. (2013). The tight coupling between intI1 and sul1 supports the use of the class 1 integron-integrase gene as an anthropogenic marker in Australian waters (Gillings et al., 2015), and the strong association with Escherichia coli is consistent with the argument that faecal contamination largely explains ARG patterns in impacted systems (Karkman et al., 2019).
The magnitude and gradient observed are broadly comparable with international surveys of urban rivers and estuaries (Pruden et al., 2006; Zhu et al., 2017), suggesting that Australian catchments are not exceptional in kind, even if local hydrology and effluent management shape the details. The persistence of elevated abundances to the estuary mouth is notable given the dilution and tidal exchange expected in that reach, and it indicates that resistance genes discharged to freshwater can reach receiving marine environments that are managed under the ANZG (2018) water-quality framework and the relevant state Environment Protection Authority (EPA) water-quality objectives.
These findings carry policy relevance. Australia’s National Antimicrobial Resistance Strategy (Commonwealth of Australia, 2020) and CSIRO’s One Health assessment (CSIRO, 2020) both call for environmental surveillance, yet routine monitoring of receiving waters does not currently include ARGs. Guidance on managing health risks from recycled water (National Health and Medical Research Council [NHMRC], 2018) shows the regulatory attention given to wastewater reuse, though ARGs are not yet covered. The present data suggest that a small set of markers, sul1 and intI1 normalised to 16S rRNA, could provide a cost-effective screening tool. Several limitations apply. The survey was cross-sectional and confined to autumn baseflow, when effluent dominates streamflow, so wet-weather flows may redistribute the signal. Moreover, qPCR quantifies gene presence but not viability or expression, so the public-health implications cannot be inferred directly.
Conclusion and Management Implications
This catchment-scale survey demonstrates a pronounced, wastewater-associated gradient in antibiotic resistance genes across a temperate Australian urban waterway, with abundances peaking at the treatment-plant outfall and remaining elevated to the estuary. Treated-effluent fraction, faecal indicators and nutrients were the strongest correlates, identifying point-source wastewater as the principal local driver. For catchment managers and water utilities, the practical implication is that targeted upgrades to effluent treatment, such as tertiary disinfection at high-load plants, are likely to yield the greatest reduction in downstream ARG loads. Incorporating a minimal ARG panel into existing water-quality monitoring, aligned with the ANZG framework and the National Antimicrobial Resistance Strategy, would allow trends to be tracked and interventions evaluated. Future work should extend sampling across seasons and flow conditions and pair gene quantification with metagenomic and viability methods, linking environmental surveillance to the human and animal components of Australia’s One Health response.
References
ANZG. (2018). Australian and New Zealand guidelines for fresh and marine water quality. Australian and New Zealand Governments and Australian state and territory governments.
Australian Commission on Safety and Quality in Health Care. (2021). AURA 2021: Fourth Australian report on antimicrobial use and resistance in human health. ACSQHC.
Berendonk, T. U., Manaia, C. M., Merlin, C., Fatta-Kassinos, D., Cytryn, E., Walsh, F., & Martinez, J. L. (2015). Tackling antibiotic resistance: The environmental framework. Nature Reviews Microbiology, 13(5), 310-317.
Commonwealth of Australia. (2020). Australia’s national antimicrobial resistance strategy: 2020 and beyond. Department of Health.
Commonwealth Scientific and Industrial Research Organisation. (2020). Antimicrobial resistance in Australia: A One Health perspective. CSIRO.
Gillings, M. R., Gaze, W. H., Pruden, A., Smalla, K., Tiedje, J. M., & Zhu, Y. G. (2015). Using the class 1 integron-integrase gene as a proxy for anthropogenic pollution. The ISME Journal, 9(6), 1269-1279.
Karkman, A., Parnanen, K., & Larsson, D. G. J. (2019). Fecal pollution can explain antibiotic resistance gene abundances in anthropogenically impacted environments. Nature Communications, 10, Article 80.
National Health and Medical Research Council. (2018). Australian guidelines for water recycling: Managing health and environmental risks. NHMRC.
Pruden, A., Pei, R., Storteboom, H., & Carlson, K. H. (2006). Antibiotic resistance genes as emerging contaminants: Studies in northern Colorado. Environmental Science & Technology, 40(23), 7445-7450.
Rizzo, L., Manaia, C., Merlin, C., Schwartz, T., Dagot, C., Ploy, M. C., & Fatta-Kassinos, D. (2013). Urban wastewater treatment plants as hotspots for antibiotic resistant bacteria and genes spread into the environment: A review. Science of the Total Environment, 447, 345-360.
World Health Organization. (2015). Global action plan on antimicrobial resistance. WHO.
Zhu, Y. G., Zhao, Y., Li, B., Huang, C. L., Zhang, S. Y., Yu, S., & Penuelas, J. (2017). Continental-scale pollution of estuaries with antibiotic resistance genes. Nature Microbiology, 2, Article 16270.