Short-term impacts of a large cultural event on the microbial pollution status of a pre-alpine river

Many rivers are continuously impacted by microbial faecal pollution from point sources such as wastewater treatment plants or untreated sewage discharges (Servais et al. 2007; Kirschner et al. 2017). Therefore, the extent of faecal river pollution is generally dependent on the available wastewater treatment infrastructure. Rivers or river sections where state-of-the-art wastewater treatment is implemented should have low to moderate faecal pollution levels, while the absence of wastewater treatment infrastructure can result in critical to excessive faecal pollution levels (Kirschner et al. 2009, 2017). Nevertheless, even in countries with state-of-the-art wastewater treatment, short-term events can lead to extreme microbial contamination of rivers. It is well known that heavy rain events in particular lead to substantial increases in faecal indicator levels (Tornevi et al. 2014) which are caused by combined sewer overflows (CSOs) (Passerat et al. 2011; Mascher et al. 2017; McGinnis et al. 2018) or faecal run-off from manured fields, live-stock pastures or areas with wildlife droppings (Fernández-Alvarez et al. 1991; Aitken 2003; Gilfillan et al. 2018; Mushi et al. 2021). Such storm and flood events can also lead to increased faecal contamination due to resuspension of bacteria deposited in river sediments (Chu et al. 2011; Bradshaw et al. 2016; Reitz et al. 2021).

Less well-studied causes of short-term faecal pollution in rivers are recreational activities (e.g., bathing and water sports) and cultural activities, which can also have significant impacts on microbial water quality. Religious mass bathing events, after which outbreaks of faecal-borne diseases have been recorded, have been reported specifically for India (Vortmann et al. 2015; David & Roy 2016; Purohit et al. 2020). To the best of our knowledge, there are no studies examining the impact of mass cultural events on microbial water quality of rivers in other parts of the world.

Here, we assessed the impact of the large FM4 Frequency music festival on the microbial quality of the Traisen River, a pre-alpine tributary of the Danube River in Austria. An approximately 2-km long stretch of the river is part of the festival area; during the 4 days of the festival, people are excessively using the river for recreation and personal hygiene, bringing faecal bacteria and chemicals from personal care products (PCPs) such as sunscreen, soap, and shampoo into the water environment (Supplementary material, Figure S1) (Harjung et al. 2020). As pollution parameters, we determined the concentrations of two standard faecal indicators, enterococci and E. coli according to the Austrian Bathing Water Act (Austrian Bathing Water Act 2009) and the concentrations of organic and inorganic nutrients, in the water upstream and downstream of the festival site. In this way, we were able to demonstrate that large music festivals can have a significant short-term impact on microbial water quality in rivers when they are an integral part of the cultural event.

The Traisen River, a pre-alpine tributary of the Danube River, is located in the province of Lower Austria. The catchment area consists of agricultural, riparian, and urban areas. The river itself has been subjected to intensive river regulation measures (Haidvogl et al. 2018), such as the construction of shallow cross-section dams dividing the river into basins of 200–400 m length. The man-made river bed consists mainly of coarse gravel and cobbles. In this study, we investigated the short-term impact of the large cultural event FM4 Frequency music festival on a 2-km long and 40-m wide section of the river in the immediate vicinity of the city of St. Pölten (approximately. 60,000 inhabitants), which is most affected by the music festival with ∼200,000 visitors over the course of 4 days (up to 50,000 per day). From 18 km upstream to 32 km downstream of the festival area, no wastewater treatment plant effluents are discharged into the river. Additional samples taken over 1 year (October 2020–October 2021, n = 2 × 5) at one site 17 km upstream and one site 15 km downstream showed low to moderate faecal pollution levels in the river throughout the year (upstream site: 2.24 ± 0.55 log10 E. coli/100 ml, downstream site: 1.25 ± 0.15 log10 E. coli/100 ml).

At each sampling occasion, quadruplicate water samples were collected by hand from a depth of 30 cm below the water surface in autoclaved 250-ml glass bottles at both the upstream and downstream sites. Sampling took place between noon and 1 p.m. The bottles were kept in a cool bag during transport to the laboratory, where they were stored at 4 °C and processed within 24 h.

During each sampling event, water temperature, oxygen content, electrical conductivity, and pH were measured on-site with handheld devices (Xylem Analytics Germany Sales GmbH & Co. KG, Weilheim, Germany: pH WTW 33110, conductivity WTW 3310, dissolved oxygen and temperature WTW Multi 3510 IDS). To assess the potential input of dissolved organic matter, dissolved organic carbon (DOC) was analysed. DOC-free vials were filled by releasing river water from a DOC-free syringe with a sterilised glass microfiber GF/F filter on the tip. The vials were then placed in the cooling box and stored at 4 °C until analysis (within 1 week after sampling). DOC analysis was performed on a GE-Sievers 900 TOC analyzer (SUEZ Water Technologies & Solutions, Trevose, PA USA) equipped with a persulfate oxidation and an inorganic carbon removal unit (LOD 0.1 mg/l). Total phosphorus (TP) was measured after wet combustion of the raw water sample with a spectrophotometer at 890 nm (Hach-Lange DR 2800; LOD 0.5 μg/l) according to the Austrian standard OENORM EN ISO 6878. Discharge data were obtained from the office of the provincial government of Lower Austria.

Escherichia coli and intestinal enterococci were analysed according to international standards (ISO 1998A, 1998B). MUG/MUD microtiter plates (Biorad Austria GmbH, Vienna, Austria), four replicates each, were incubated at 44 ± 0.5 °C for 36–42 h. After incubation, the plates were analysed in a dark room using a handheld UV light-source. The fluorescent wells were counted and the most probable number (MPN) was determined according to the manufacturers MPN table. By applying two dilutions (original and 1:20 dilution), as recommended for bathing waters, the limit of detection of both microtiter plate methods were 15 MPN/100 ml (1.18 log10 MPN/100 ml).

For statistical analysis, E. coli and enterococci concentrations were log10 transformed. Differences between sites were tested by Students t-test for paired samples. For assessing temporal differences, the investigation period was divided into a pre-festival stage (before 13 August), the festival stage (15–19 August) and the post-festival stage. Differences between stages were assessed by one-way ANOVA and Tukey post hoc test was applied. Analyses were performed with IBM programme SPSSv24; a significance level of p < 0.05 was chosen for significant differences.

In the pre-festival stage, E. coli and enterococcci concentrations were at similar low levels around 1.5 log10 MPN/100 ml at both the upstream and the downstream site of the festival area (Figure 2). No significant differences between upstream and downstream sites were observed (t-test; p > 0.05). While E. coli concentrations increased only slightly during the festival (up to 2.3 log10 MPN/100 ml at the downstream site), enterococci exhibited a massive increase up to 4.7 log10 MPN/100 ml (=5 × 10⁴ MPN/100 ml) during the festival (Figure 2), far above the limit values for short-term pollution events in bathing waters according to Austrian legislation (400 MPN/100 ml (Austrian_Bathing_Water_Act 2009)). The increases started on the first day of the festival and were statistically significant only for the enterococci, both at the upstream and at the downstream site (ANOVA; Tukey post hoc, p < 0.001). Slightly higher values were observed downstream of the festival area (4.5–4.7 log10 MPN/100 ml) in comparison to the upstream site (4.1–4.5 log10 MPN/100 ml), but the difference was not statistically significant (t-test, p > 0.05). Three weeks after the festival, the enterococci concentrations dropped back down to levels that were similar to what was observed before the festival (ANOVA; Tukey post hoc, p < 0.001). In contrast, E. coli concentrations remained at the same level and exhibited the highest value at the end of the study period.

It seems unlikely that a differential shedding of the two FIBs took place and more enterococci than E. coli entered the river water during the festival. In contrast, faeces from humans (Wright 1982; Farnleitner et al. 2010) and waterfowl (Kirschner et al. 2004) usually contains higher average concentrations of E. coli than enterococci. A plausible explanation for the observed differential increase could have been a different persistence of the two FIBs. Such an uncoupling of enterococci and E. coli values has been observed previously for saline environments (Kirschner et al. 2004), with E. coli showing lower persistence at higher salinities (Sagarduy et al. 2019).

Solar irradiation and water temperature are other key factors influencing the decay of FIB, which may affect E. coli vs. enterococci differently (Brooks & Field 2016; Sagarduy et al. 2019). We observed only a slight increase in river water salinity during the festival event at the downstream site, as measured by electrical conductivity (from 430 to 445 μS/cm, Figure 3), which precludes a significant effect on E. coli persistence. Water temperature (Supplementary material, Figure S3) and solar radiation were high throughout the festival and may have exerted a strong influence on the FIB. The water levels of the river were very low (see Supplementary material, Figure S2), and water stayed longer in the separate river basins formed by the cross-sectional shallow dams, which allowed the water to warm up quickly and sunlight to penetrate to the river bottom. Alternatively to the hypothesised differential persistence of the faecal indicators, a possible impact of the PCPs on the assay performance cannot be completely excluded. The two fluorogenic substrates 4-methylumbelliferyl-β-d-glucuronide (MUG) for E. coli and 4-methylumbelliferyl-β-d-glucoside (MUD) for enterococci could theoretically be differently impacted by certain compounds contained in PCPs, although no data are available in the literature or from the manufacturer (Biorad, Marnes-la-Coquette, France, personal communication).

In comparison to other short-term events, the observed elevated faecal indicator levels during the festival were surprisingly high. Combined sewer overflow events in rivers with state-of-the-art wastewater treatment have been reported to result in increased faecal indicator concentrations that are at least two orders of magnitude higher than during dry weather. In the Seine River (France), directly below a CSO outfall, maximum concentrations of 2.9 × 10⁵E. coli and 7.6 × 10⁴ intestinal enterococci per 100-ml were recorded (Passerat et al. 2011), and in the Mohawk River (New York; USA) an activated CSO resulted in E. coli and enterococci concentrations above their upper detection limit of >2.4 × 10⁴ colony forming units/100 ml (Lininger et al. 2022). The intestinal enterococci concentrations found in our study (5 × 10⁴ MPN/100 ml) were thus in a comparable order of magnitude to CSOs in rivers. To the best of our knowledge, scientific literature on the short-term impacts of cultural events on river water quality is limited to mass bathing events in India. Vortmann et al. (2015) reported that total coliform concentrations in the Ganga River increased from 1.5 × 10⁴ to 5 × 10⁴ per 100 ml during a mass bathing event at a downstream site. Much higher levels were reported for the Kshipra River in Central India, where an increase from 3.7 × 10⁵ to 5.1 × 10⁶E. coli per 100 ml was observed during such an event (Purohit et al. 2020). Since no information on enterococci was provided in these studies, and rivers in India generally display higher faecal pollution levels due to poor wastewater treatment, a direct comparison is not appropriate.

Next to the increased microbial faecal pollution, we measured a significant increase of DOC and P_tot concentrations at the downstream site during the festival (Figure 3). Similarly, electrical conductivity was also elevated during the festival, with significantly higher values at the downstream site. Immediately after the festival ended, DOC and P_tot dropped to values comparable to the pre-festival period. The other measured physico-chemical parameters temperature, dissolved oxygen, and pH values did not change significantly during the cultural event ( Supplementary material, Figure S3).

The observed input of organic and inorganic matter was likely caused by the presence of festival participants and potentially originating from urine, faecal matter but also from various kinds of PCP. We hypothesise that the massive presence of certain PCPs in the river water (Harjung et al. 2020) also contributed to the differential persistence of the FIB. The effects of PCPs at both upstream and downstream sites of the festival area was likely enhanced by the low water levels observed during the festival (see above), which increased the exposure time of the FIBs to the PCPs. Indeed, Harjung et al. (2020) reported a dramatic increase in chemically stable UV-B filter phenylbenzimidazole sulphonic acid (PBSA) during the same festival, indicating that organic compounds in sunscreens and other PCPs are the sources of elevated DOC.

E. coli as a Gram-negative bacterium might be more sensitive to components of PCPs than Gram-positive bacteria such as enterococci, due to the lipopolysaccharide characteristics of their outer cell membrane. For example, EDTA (Voss 1967) and dodecyldiethanolamine (Lambert & Smith 1976), substances added to cosmetic products, have been reported to have effective bactericidal activity against E. coli and other Gram-negative bacteria. More recently, nanoparticles in sunscreen products were shown to have inhibitory and toxic effects on E. coli (Baek et al. 2017). Also antibacterial agents such as triclosan or triclocarban that are often present in PCPs, have been found in rivers such as in Italy (Palmiotto et al. 2018) or Spain (Carmona et al. 2014) and could have a differential effect on the persistence of the two faecal indicators investigated.

However, due to the large number and diversity of ingredients in PCPs, an enhanced effect of combinations of different chemicals on Gram negatives – in contrast to Gram positives – can only be speculated. To date, the presence of PCPs in rivers has generally been associated with their emission from sewage effluents (van Wijnen et al. 2018; Homem et al. 2022; Rapp-Wright et al. 2023). Concentrations in the magnitude of several hundred ng L⁻¹ have been recorded for various substances, significantly exceeding predicted no effect concentrations (PNEC) (Homem et al. 2022). One study indirectly linked the elevated occurrence of PCPs (particularly ingredients of sunscreen products) in alpine rivers to the number of tourists and residential population in the region, showing that the highest concentrations up to 6 μg L⁻¹ occurred in summer (Mandaric et al. 2017). To the best of our knowledge, the study by Harjung et al. (2020) is the only one that directly links the input of such PCPs into river water to bathing guests/festival visitors. PBSA concentrations of 45 μg L⁻¹ measured during the festival by far exceeded those found in other rivers and lakes (Harjung et al. 2020), corroborating our hypothesis of a potential impact on E. coli and enterococci detection.

We demonstrated that a large cultural festival event triggered a significant short-term faecal pollution event in a medium-sized pre-alpine tributary of the Danube River. The specific situation at the festival area and the intensive use of the river for recreational and hygienic purposes by festival participants led to a remarkable rise in enterococci concentrations in the river water, far above the limits for short-term pollution events in bathing waters according to Austrian legislation. This poses a significant potential threat to the health of the festival participants and people recreating in the festival area and downstream who could become infected with pathogens concomitantly shed into the water. In contrast to the enterococci, there was only a minimal increase in E. coli concentrations. This uncoupling could have been caused by the different persistence of the two FIB investigated in relation to high water temperatures, solar radiation and the presence of potentially large amounts of PCPs. Alternatively, an impact of the PCPs on the assay performance cannot be excluded. More detailed investigations on the persistence of FIB in response to different PCP ingredients and on the potential impact of these products on the performance of the two assays are highly needed. Such scenarios could also be relevant for other intensively used marine and freshwater bathing sites. We further conclude that more effective management is necessary for upcoming festivals at this river or other festivals taking place under similar settings, to prevent deterioration in water quality in terms of faecal pollution and to minimise the health risk to visitors and recreational users downstream. Management measures could include improved sanitation infrastructure, better waste management and campaigns to raise awareness of the problem among festival visitors.

This study was partly funded by the GFF project LS18-007 and by the State Government of Lower Austria (K3-F-799/001-2018). Hubert Kraill performed ion chromatography and P_tot analyses, Astrid Harjung provided DOC analyses. Astrid Harjung and Victor Aigner kindly assisted in the field work and supported the study with fruitful discussions. The authors also thank Katharina Besemer and Domenico Savio for their inspiring discussions and specific insights concerning the impact of the Frequency festival on the Traisen River.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.