Skip to main content
NCBI home page
Applied and Environmental Microbiology logoLink to Applied and Environmental Microbiology
. 2024 Jun 28;90(7):e00672-24. doi: 10.1128/aem.00672-24

Sanitization of hydroponic farming facilities in Singapore: what, why, and how

Cliff An Ting Tham 1,2, Ye Htut Zwe 1,3, Michelle Mei Zhen Ten 1, Geraldine Shang Ya Ng 1, Jillinda Yi Ling Toh 1, Bee Ling Poh 2, Weibiao Zhou 1, Dan Li 1,✉
Editor: Christopher A Elkins4
PMCID: PMC11267894  PMID: 38940566

ABSTRACT

This study performed microbial analysis of nutrient film technique (NFT) hydroponic systems on three indoor farms in Singapore (the “what”). To justify the necessity of sanitizing hydroponic systems, strong biofilm-forming bacteria were isolated from the facility and investigated for their influence on Salmonella colonization on polyvinyl chloride (PVC) coupons in hydroponic nutrient solutions (the “why”). Finally, sanitization solutions were evaluated with both laboratory-scale and field-scale tests (the “how”). As a result, the microbiome composition in NFT systems was found to be highly farm specific. The strong biofilm formers Corynebacterium tuberculostearicum C2 and Pseudoxanthomonas mexicana C3 were found to facilitate the attachment and colonization of Salmonella on PVC coupons. When forming dual-species biofilms, the presence of C2 and C3 also significantly promoted the growth of Salmonella (P < 0.05). Compared with hydrogen peroxide (H2O2) and sodium percarbonate (SPC), sodium hypochlorite (NaOCl) exhibited superior efficacy in biofilm removal. At 50 ppm, NaOCl reduced the Salmonella Typhimurium, C2, and C3 counts to <1 log CFU/cm2 within 12 h, whereas neither 3% H2O2 nor 1% SPC achieved this effect. In operational hydroponic systems, the concentration of NaOCl needed to achieve biofilm elimination increased to 500 ppm, likely due to the presence of organic matter accumulated during crop cultivation and the greater persistence of naturally formed multispecies biofilms. Sanitization using 500 ppm NaOCl for 12 h did not impede subsequent plant growth, but chlorination byproduct chlorate was detected at high levels in the hydroponic solution and in plants in the sanitized systems without rinsing.

IMPORTANCE

This study’s significance lies first in its elucidation of the necessity of sanitizing hydroponic farming systems. The microbiome in hydroponic systems, although mostly nonpathogenic, might serve as a hotbed for pathogen colonization and thus pose a risk for food safety. We thus explored sanitization solutions with both laboratory-scale and field-scale tests. Of the three tested sanitizers, NaOCl was the most effective and economical option, whereas one must note the vital importance of rinsing the hydroponic systems after sanitization with NaOCl.

KEYWORDS: biofilms, sanitization, hydroponics, Salmonella, sodium hypochlorite

INTRODUCTION

Urbanization poses significant challenges to food security worldwide, particularly in densely populated areas where arable land is scarce (1). Singapore, emblematic of such urban centers, grapples with the conundrum of sustaining its food supply amidst limited land availability for traditional agriculture. With only 1% of its land designated for farming due to competing land uses, Singapore relies heavily on food imports to meet more than 90% of its nutritional needs (2). This dependency exposes the nation to vulnerabilities stemming from disruptions in global food supply chains. To address this pressing issue, Singapore launched the “30 by 30” initiative in 2019, aiming to meet 30% of its nutritional requirements locally by 2030 (3). Central to this initiative is the exploration of innovative agricultural practices capable of maximizing food production within the constraints of urban spaces. Among these practices, hydroponic farming has emerged as a promising solution, offering the potential to cultivate vegetables sustainably in indoor environments with minimal land requirements (4).

However, indoor growth with precise control of environmental variables, such as temperature and carbon dioxide levels, does not automatically ensure that produce is exempt from food safety issues. Even with stringent controls, undesired microorganisms could still be introduced into indoor growing systems from multiple sources, including water, fertilizer, growth substrate, seeds/seedlings, and pests, as well as contamination from human manipulation. Unlike traditional soil-based farming, hydroponic systems provide a conducive environment for microbial biofilm formation on various surfaces, including those in nutrient film technique (NFT) systems commonly employed on indoor farms (5, 6). Our previous study demonstrated that once the pathogenic bacterium Salmonella enterica serovar Typhimurium was introduced, it could rapidly spread throughout the hydroponic system, likely through nutrient solution circulation (7). To date, there are no guidelines on the sanitization of indoor farming facilities in Singapore.

Salmonella is a leading cause of bacterial foodborne illnesses and outbreaks worldwide and is strongly correlated with the consumption of fresh produce, as supported by numerous pieces of evidence (8–11). In 2021, a Salmonella outbreak occurred in multiple states in the USA and was linked to eight varieties of hydroponic greenhouse-grown leafy green vegetables from the BrightFarms brand. Eleven people became ill, two were hospitalized, and there were likely more unreported cases of illness, which was estimated to be 30 times greater than the reported cases during this outbreak (12). During a screening study conducted by our group (13), multiple Salmonella strains belonging to seven different serovars were isolated from both the nutrient solution and the lettuce crops from local farms in Singapore, suggesting the need for local farms to improve sanitation and hygiene practices.

Addressing microbial contamination in hydroponic systems necessitates the development and implementation of robust cleaning and sanitization protocols, which must consider not only their efficacy but also their practicality, cost-effectiveness, and potential impact on plant health and system sustainability. Balancing these factors is crucial for ensuring the viability and resilience of hydroponic farming practices in urban settings such as Singapore. This study aimed to provide the first baseline knowledge of the microbial loads and composition in operational indoor hydroponic farms in Singapore. By investigating the biofilm-forming capabilities of bacterial isolates from hydroponic farms and the interactions of strong biofilm formers with Salmonella, one of the most representative pathogens, we aimed to elucidate the necessity of sanitizing hydroponic farming systems. Accordingly, our work tested and proposed effective sanitation measures to safeguard the food safety of indoor hydroponic farming systems.

RESULTS AND DISCUSSION

Microbial analysis of multitier NFT hydroponic systems

Microbial analysis was conducted by collecting samples from the NFT hydroponic systems in operation at three commercial indoor farms in Singapore. Ten surface swabbing samples were collected from each of the seven components of the NFT systems, as illustrated in Fig. 1b. Interestingly, at all three farms, the nutrient tank interior surfaces and main discharge points leading to the nutrient tanks always exhibited significantly greater microbial counts at 7.3–7.5 log CFU/cm2 than did the other tested parts (P < 0.05; Fig. 1a).

Fig 1.

Fig 1

Open in a new tab

The key microbial analysis results of NFT hydroponic systems at three commercial indoor farms in Singapore. (a) The microbial loads in different components of NFT systems. Values denoted with different letters are significantly different (P < 0.05). For each data point, 10 samples were collected from each farm (n = 30). (b) The illustration of the NFT systems. (c) The bacteria phyla identified by 16S rDNA sequencing (>1%) from selected samples. (d) Microbiome compositional difference (beta diversity) based on the bacterial composition analysis. Unweighted UniFrac (left, presence of taxa), weighted UniFrac (middle, presence and relative abundance of taxa), and Bray-Curtis (right, presence and relative abundance of taxa) distances are visualized by principal coordinate analysis plots.

The bacterial composition of the swabbed surfaces was identified by 16S rDNA sequencing. At the phylum level, Proteobacteria was the most abundant group from all the tested samples, accounting for 36.0%–63.8%, except for the submerged pipes in the nutrient tank at Farm I. Next, Bacteroidota was also identified from all of the samples, with major relative abundances accounting for 6.8%–35.8% (Fig. 1c). The microbiome composition became very diverse at the genus level (Fig. S1), making it difficult to identify the commonly shared genera from all the samples. Instead, the genus composition tended to be more farm specific. At Farm I, unclassified Comamonadaceae, Nitrospira, Bacillus, unclassified Rhizobiaceae, and Aeromonas were identified with relative abundances greater than 1% in more than half of the samples. More genera were identified when the same criteria (higher than 1% relative abundances from more than half of the samples) were applied to the samples from Farm II, including unclassified Comamonadaceae, Sphingobium, Saccharimonadaceae TM7a, Pseudomonas, Hyphomicrobium, Terrimonas, unclassified Chitinophagaceae, Panacagrimonas, Microbacterium, Thermomonas, Saccharimonadales, Luteimonas, and Allorhizobium-Neorhizobium-Pararhizobium-Rhizobium. Diverse genera were also identified for samples from Farm III, which were mostly different from the samples from Farm I and Farm II. The genera with relative abundances greater than 1% in more than half of the samples at Farm III included Hydrogenophaga, Leptolyngbya PCC-6306, unclassified Saprospiraceae, unclassified Comamonadaceae, Cyanobacteria, Haliscomenobacter, Rhodobacter, Hyphomicrobium, Porphyrobacter, Hyphomonadaceae SWB02, Sediminibacterium, and Ferrovibrio. Indeed, beta diversity analysis of amplicon sequence variants (ASVs) also revealed similarities in the microbiomes of samples from the same farm. Clear clusters for samples from the three farms were identified from principal coordinate analysis (PCoA) plots based on weighted and unweighted UniFrac distances, as well as count-based Bray-Curtis distances (Fig. 1d). Since the three farms were all indoor hydroponic farms using NFT systems with polyvinyl chloride (PVC) materials, tap water was used to formulate the hydroponic solutions, and Brassica leafy greens were grown. The most likely reasons for this distinctive farm-specific microbiome composition could be personnel and other environmental factors.

In addition, we also performed 18S rDNA sequencing to identify fungi and algae from the surface samples. Microalgae belonging to Chlorophyta, Ochrophyta, and Diatomea and fungi belonging to Ascomycota, Cryptomycota, and LKM15 were identified in most of the samples from all three farms (Fig. S2). Specifically, the green algae Chlorophyta had high relative abundances in most of the samples, especially in the Farm III samples (>90%, Fig. S2). Very recently, Lipsman et al. (14) revealed that a shared algal-bacterial extracellular matrix could promote algal-bacterial aggregation in an environmentally relevant model system. In their study, algal exudates played a pivotal role in promoting bacterial colonization, stimulating bacterial exopolysaccharide production, and facilitating the joint formation of the algal-bacterial extracellular matrix. To date, the influence of green algae on the food safety of hydroponic systems, which is worthy of exploration in the future, remains largely unknown. In this study, the following parts were still focused on bacteria. The influence of the background microbiome on the microbial safety of hydroponic systems was investigated by evaluating the interactions of bacterial biofilm formers with Salmonella, one of the most representative foodborne pathogens. Sanitization efficacy was evaluated by plate counts on plate count agar, a nonselective agar mainly used to assess total viable bacteria.

Biofilm former identification

Ten bacterial colonies with various morphologies were isolated from the samples from the interior surfaces of the nutrient tanks. Their biofilm-forming abilities were observed by the measurement of biomass together with five Salmonella strains belonging to different serovars (Fig. 2a). With the use of the criteria established previously (15), four isolates, as well as five Salmonella strains, were classified as weak biofilm formers, three isolates were classified as moderate biofilm formers, and three isolates were classified as strong biofilm formers. We thus selected isolates C2 and C3, the strongest Gram-positive and Gram-negative biofilm formers, respectively, for the following studies. Whole-genome sequencing (WGS) revealed C2 as Corynebacterium tuberculostearicum (Fig. 2b) and C3 as Pseudoxanthomonas mexicana (Fig. 2c).

Fig 2.

Fig 2

Open in a new tab

Screening of strong biofilm formers from the interior surfaces of hydroponic systems. (a) Biofilm-forming capability and Gram-staining results of the 10 bacterial isolates. The results are based on the average of triplicates. (b) and (c) Single nucleotide polymorphis (SNP) trees based on the maximum likelihood method illustrating the phylogenetic relationships between C2 and C3 and closely related members of the same genus, respectively. The numbers represent bootstrap values. The red IDs represent human clinical isolates, the blue IDs represent environmental isolates, and the black IDs represent strains of unknown origin.

Effect of C. tuberculostearicum C2 and P. mexicana C3 biofilms on the colonization of Salmonella on PVC coupons

The Salmonella Brunei strain, which was self-isolated from a hydroponic farm in Singapore by our group in a previous study (13), showed weak biofilm-forming ability, as shown in Fig. 2a. When we dipped PVC coupons in a 5 log CFU/mL Salmonella suspension and rinsed the coupons with sterile deionized water, no Salmonella was enumerated from the PVC coupons (<1 log CFU/cm2) after up to 120 h of incubation in hydroponic nutrient solution (Fig. 3a). In contrast, on PVC coupons with preformed C. tuberculostearicum C2 and P. mexicana C3 biofilms, Salmonella was detected at 2.2 ± 0.3 and 1.9 ± 0.2 log CFU/cm2 after water rinsing, and the Salmonella counts continued to increase along with the incubation of the coupons in hydroponic nutrient solutions up to 3.3 ± 0.2 and 2.8 ± 0.1 log CFU/cm2, respectively, after 120 h (Fig. 3a). C. tuberculostearicum has been reported to be a ubiquitous bacterium that colonizes human skin (16). The occurrence of these bacteria in the hydroponic facility might be due to human manipulation. Although there have been occasional reports of human infection caused by this bacterial species (17), no evidence could be found of any food-related infections or risks. P. mexicana is an environmentally nonpathogenic bacterium (18). Our results suggest that background microbiomes such as C. tuberculostearicum and P. mexicana, although mostly nonpathogenic, could enhance the retention of foodborne pathogens such as Salmonella after being introduced to hydroponic circulating systems and might not be able to easily colonize a cleaner system.

Fig 3.

Fig 3

Open in a new tab

Experimental setups and results showing the effectiveness of C. tuberculostearicum C2 and P. mexicana C3 on the colonization of Salmonella on PVC coupons. (a) Adherence and colonization of S. Brunei on preexisting C. tuberculostearicum C2 and P. mexicana C3 biofilms. Each value is the average of triplicates from three independent experiments. The error bars indicate the standard deviations (SDs). The dotted line represents the limit of detection. Panels a–d show the significance of Salmonella enumeration at different time points on C2- or C3-preformed biofilms, respectively. (b) Salmonella [on xylose lysine deoxycholate (XLD) plates] and total bacteria counts [on tryptone soya agar (TSA) plates] from single- and dual-species biofilms. Each value is the average of triplicates from three independent experiments. The error bars indicate the SDs. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, respectively.

Effect of C. tuberculostearicum C2 and P. mexicana C3 on the growth of Salmonella in dual-species biofilms

A different setup was designed, and it was demonstrated that C. tuberculostearicum C2 and P. mexicana C3 could not only assist the adherence of Salmonella on surfaces but also promote the growth of Salmonella to higher levels when mixed in dual-species biofilms. As shown in Fig. 3b, the S. Brunei counts were significantly greater in dual-species biofilms cocultured with C. tuberculostearicum C2 (6.7 ± 0.1 log CFU/cm2; P = 0.0002) and with P. mexicana C3 (6.3 ± 0.2 log CFU/cm2; P = 0.0018) than in biofilms cocultured with single-species biofilms (5.1 ± 0.2 log CFU/cm2). Due to a lack of selective culture media for C. tuberculostearicum and P. mexicana, it was impossible to make a direct comparison of the C2 and C3 counts in single- and dual-species biofilms. However, we did observe significantly greater total bacterial counts in the dual-species biofilms (7.9 ± 0.5 and 7.8 ± 0.2 log CFU/cm2) than in the C2 and C3 single-species biofilms (6.9 ± 0.2 and 6.6 ± 0.1 log CFU/cm2; P < 0.05; Fig. 3b), suggesting a symbiotic interaction between S. Brunei and C. tuberculostearicum and P. mexicana.

Among the five tested Salmonella strains, S. Typhimurium had the strongest biofilm formation (Fig. 2a). However, its biofilm-forming abilities, as measured by biomass, were also significantly lower than those of C. tuberculostearicum C2 and P. mexicana C3 isolated from the interior surfaces of nutrient tanks (P < 0.05; Fig. 2a). When tested in dual-species biofilms, a symbiotic interaction was also observed between S. Typhimurium and C. tuberculostearicum and between S. Typhimurium and P. mexicana (Fig. 3b).

Such symbiotic effects between different bacterial species in biofilms have been observed in previous studies, such as coculture biofilm assays involving Staphylococcus aureus with Pseudomonas aeruginosa (19) and Staphylococcus epidermidis with Candida albicans (20). However, not all bacterial interactions within mixed-species biofilms are synergistic. For instance, in one of our previous studies, we found that a strain of Chromobacterium hemolyticum isolated from a hydroponic nutrient solution exhibited highly antagonistic properties, eliminating Salmonella, Escherichia coli, Listeria monocytogenes, and S. aureus in coculture biofilms (21). Nevertheless, with respect to exceptionally antagonistic strains, our findings again suggest that the presence of resident biofilms in hydroponic systems may increase the risk of persistent foodborne pathogens.

Laboratory-scale test: efficacy of various sanitizers on S. Typhimurium, C. tuberculostearicum C2, and P. mexicana C3 biofilms

Using S. Typhimurium, C. tuberculostearicum C2, and P. mexicana C3 biofilms grown on PVC coupons as a representative and robust model, we evaluated the potential of a few commonly used sanitizers, including sodium hypochlorite (NaOCl), hydrogen peroxide (H2O2), and sodium percarbonate (SPC), to sanitize hydroponic systems. The tested concentrations of each sanitizer (50 and 500 ppm NaOCl, 1.5% and 3.0% H2O2, and 1.0% SPC) were selected based on commonly used concentrations under other application scenarios (22–24). The tested durations (1, 6, and 12 h) were selected based on practical considerations not to compromise farming productivity.

As a result, compared with H2O2 and SPC, NaOCl exhibited superior efficacy in removing biofilms. For C. tuberculostearicum C2 and P. mexicana C3 biofilms, NaOCl at 500 ppm reduced the bacterial count to <1 log CFU/cm2 from the biofilm grown on PVC coupons within 1 h. For S. Typhimurium biofilm, NaOCl at 500 ppm reduced the bacterial count to <1 log CFU/cm2 from the biofilm grown on PVC coupons within 6 h. At a lower concentration of 50 ppm, C. tuberculostearicum C2 counts were reduced to <1 log CFU/cm2 within 12 h. S. Typhimurium and P. mexicana C3 counts were reduced to <1 log CFU/cm2 within 6 h. In comparison, neither 3% H2O2 nor 1% SPC achieved such an effect within the tested 12 h (Fig. 4).

Fig 4.

Fig 4

Open in a new tab

Experimental setup and results of the sanitization of S. Typhimurium, C. tuberculostearicum C2, and P. mexicana C3 biofilms grown on PVC coupons. Each value (Log CFU/cm2 ± SD) is the average of triplicates from three independent experiments. N.d., not detected, <1 log CFU/cm2.

Field-scale test: sanitization of commercial NFT hydroponic systems

The hydroponic systems were sanitized after the crops were harvested. The concentrated sanitizers were directly added to the hydroponic solution in the NFT systems to reach the defined working concentrations. The dose of NaOCl and duration required for the sanitization of commercial NFT hydroponic systems in the field-scale test increased remarkably in comparison with those in the laboratory-scale test. As shown in Table 1, the microbial loads on the NFT system surfaces were only reduced to <1 log CFU/cm2 by 500 ppm NaOCl treatment for 12 h. On the one hand, mixed-species biofilms have been reported to have greater resistance to multiple inactivation treatments than single-species biofilms (25–27). On the other hand, this was also likely due to the presence of organic matter accumulated during crop cultivation. Organic matter, such as proteins and carbohydrates, is well known to significantly interfere with the sanitizing effect of NaOCl (28, 29). Although replacing the hydroponic solution with fresh sanitizers is expected to generate better efficacy by reducing organic matter loads, one must consider sustainability and economic factors at the same time when designing the protocol.

TABLE 1.

Efficacy of 500 ppm NaOCl and 3% H2O2 in sanitizing the multi-tier NFT hydroponic systems at three commercial indoor farms in Singaporea

Hydroponic system components Microbial loads (log CFU/mL in solution and log CFU/cm2 on surfaces ± SD)
Control 500 ppm NaOCl 6 h 500 ppm NaOCl 12 h Control 3% H2O2 6 h 3% H2O2 12 h
Nutrient solution 5.6 ± 0.4 N.d. N.d. 6.0 ± 0.3 4.3 ± 0.2 2.3 ± 0.4
Nutrient tank wall interior surface 7.5 ± 0.2 3.1 ± 0.2 7.3 ± 0.2 5.2 ± 0.2 3.6 ± 0.1
Submersible pumps in nutrient reservoir 7.3 ± 0.4 3.0 ± 0.2 7.5 ± 0.3 5.8 ± 0.2 3.2 ± 0.2
Submerged pipes in nutrient reservoir 6.7 ± 0.1 2.1 ± 0.2 6.9 ± 0.1 4.3 ± 0.2 2.8 ± 0.1
Main discharge point leading to nutrient reservoirs 7.8 ± 0.2 3.2 ± 0.1 7.7 ± 0.2 5.6 ± 0.2 3.9 ± 0.2
Nutrient inlet of hydroponics trough 6.9 ± 0.1 2.5 ± 0.1 6.9 ± 0.2 4.8 ± 0.3 2.8 ± 0.2
Surface of hydroponic trough 6.9 ± 0.7 2.4 ± 0.1 6.4 ± 0.2 4.7 ± 0.3 3.2 ± 0.2
Drainage outlet of hydroponics trough 6.2 ± 0.5 2.5 ± 0.2 6.5 ± 0.4 4.0 ± 0.1 2.3 ± 0.1
Open in a new tab
a

For each data point, 10 samples were collected from each farm (n = 30). N.d., not detected, <1 log CFU/cm2 or <1 log CFU/mL.

Consistent with the laboratory-scale test, H2O2 was less effective in eliminating the microbial loads in the NFT systems. Although a time-dependent effect was observed for all tested components, resulting in significant reductions (P < 0.05), microbial counts of 2.3–3.9 log CFU/cm2 were still obtained after treatment with 3% H2O2 for 12 h (Table 1). H2O2 has been gaining popularity among hydroponic growers due to concerns about chemical residues, leading to many studies evaluating its efficacy (30–34) and commercial companies providing H2O2-based hydroponic decontamination solutions such as Waterco from Australia. Furthermore, the literature has also extensively documented the potential role of H2O2 as a signaling molecule in the abiotic stress response and photosynthesis, contributing to growth performance in plants (35–37) and thus further enhancing the appeal of H2O2 for hydroponic growers. However, the study conducted by Kriem et al. (33) was the only study measuring the microbial load reductions in bioponic nutrient solutions toward H2O2. Similar to our data, the greatest microbial reductions reported in this study were 97.8% for bacteria and 99.1% for fungi after a 200 mg/L H2O2 production time of 60 min (33). In addition to efficacy, economic and practical factors also favor NaOCl. For instance, in Singapore, the cost per liter of 3% H2O2 exceeds that of 5% NaOCl (household bleach) by approximately 3.5 times. Since our results advocate diluting commercially available 5% NaOCl by a factor of 100 before application, this further reduces the cost of sanitizing the systems with NaOCl.

Influence of NaOCl sanitization on the yield and chemical safety of the indoor-grown agricultural products

Although NaOCl is effective and economical as a commonly used sanitizer, it is associated with significant chemical safety concerns due to the generation of harmful byproducts (38–41). For instance, the use of chlorine in fresh-cut produce washing is prohibited in some European Union countries, including Germany, Switzerland, the Netherlands, Denmark, and Belgium (42, 43).

In this study, following a 12-h treatment with 500 ppm NaOCl, the hydroponic system was drained and replenished with a fresh nutrient solution without additional rinsing. Komatsuna (Brassica rapa var. perviridis) seeds were germinated in this system. As shown in Table 2, after 7 days, 98% of the komatsuna seeds germinated in the sanitized system. At the end of the growth cycle (day 35), the harvested crops had fresh weights of 109.0 ± 9.6 g/plant and 12.5 ± 2.9 leaves/plant. In comparison, seeds germinated in systems without sanitization achieved a germination rate of 99%. At the end of the growth cycle (day 35), the crops had fresh weights of 109.0 ± 9.2 g/plant and 12.8 ± 2.7 leaves/plant. No significant differences (P > 0.05) in fresh weight or number of leaves were detected.

TABLE 2.

Germination rate (n = 100, day 7) and yield (fresh weight and number of leaves, n = 30, day 35) of komatsuna (Brassica rapa var. perviridis) grown in nutrient solution after sanitization with 500 ppm NaOCl for 12 h

Germination rate (%) Fresh weight (g) Number of leaves
In sanitized system 98 109.6 ± 9.6 13.0 ± 3.0
In unsanitized system 99 109.7 ± 9.2 13.0 ± 3.0
Open in a new tab

At the end of the growth cycle (day 35), we collected nutrient solution samples from the hydroponic system and analyzed a series of chlorination byproducts. As a result, chlorate was detected at an alarmingly high level (2,383 mg/L), while the remaining concentrations were below the detection limits (Table 3). We thus also analyzed the level of chlorate in the harvested komatsuna and detected a level of 0.054 mg/kg chlorate in the vegetables (Table 3).

TABLE 3.

Analysis of chlorination byproducts present in the hydroponics nutrient solution in the NFT systems and in ready-to-harvest komatsuna (Brassica rapa var. perviridis) grown in the NFT systems for 35 days

Byproducts Analysis method Control (tap water) In systems without rinsing after sanitation In systems rinsed twice after sanitation
In hydroponics nutrient solution Chlorate as ClO3 (mg/L) APHA 4110 D <0.5 2,383 <0.5
Chloroacetic acid (μg/L) EPA 552.3 <1 <1 <1
Monobromoacetic acid (μg/L) EPA 552.3 <1 <1 <1
Dichloroacetic acid (μg/L) EPA 552.3 <1 <1 <1
Trichloroacetic acid (μg/L) EPA 552.3 <1 <1 <1
Dibromoacetic acid (μg/L) EPA 552.3 <1 <1 <1
Total haloacetic acids (μg/L) EPA 552.3 <1 <1 <1
Chloroform (μg/L) EPA 8260 D <5 <5 <5
Bromodichloromethane (μg/L) EPA 8260 D <5 <5 <5
Dibromochloromethane (μg/L) EPA 8260 D <5 <5 <5
Bromoform (μg/L) EPA 8260 D <5 <5 <5
Total trihalomethane (μg/L) EPA 8260 D <5 <5 <5
In ready-to-harvest komatsuna Chlorate as ClO3 (mg/kg) QuPPe-PO (EURL-SRM version 12.2.2023) N. A. 0.054 <0.01
Open in a new tab

The European Food Safety Authority (EFSA) has estimated the risks of chlorate to human health (44). The recommended safe intake level for daily intake of chlorate is 36 µg/kg of body weight per day, as a high intake of chlorate can limit the blood’s ability to absorb oxygen, leading to kidney failure. Over time, chronic exposure to chlorate can inhibit iodine uptake. EFSA also has a tolerable daily intake (TDI) of 3 µg/kg of body weight per day for long-term exposure to chlorate in food. Based on this recommendation, chlorate intake due to the consumption of vegetables grown in the sanitized hydroponic systems in this study would barely reach the TDI. However, one may not rule out the possibility of chlorate intake from other sources, such as drinking water. Therefore, rinsing steps after sanitization became essential to avoid introducing new hazards into hydroponic-grown agricultural products. In our next trial, the systems were rinsed with tap water twice after sanitization before growth. The chlorate levels of both the hydroponic solution and the vegetables were reduced to levels below the detection limits (<0.5 mg/L in the hydroponic solution and <0.01 mg/kg in the vegetables; Table 3).

Taken together, our results suggest that omitting the rinsing steps would not diminish the crop yield, but the harmful byproducts could remain at high levels in the hydroponic solutions and be internalized and accumulated in the edible parts of the crops grown in the systems, which are consequences that could be easily ignored by growers and consumers and could become a new food safety issue in the long run. It is therefore vitally important to set regulatory guidelines for the technical details of sanitizing hydroponic systems with NaOCl, especially to prevent overdose and to ensure the appropriate rinsing steps.

Conclusions

Utilizing hydroponics to address food security concerns in a highly urbanized country such as Singapore comes with its own set of unique challenges. Without proper sanitization measures, the closed-loop process heightens the risk of microbial contamination and persistence within the systems. In this study, we demonstrated high microbial loads as well as diverse and farm-specific microbiome compositions in hydroponic systems at three commercial farms in Singapore. Two strong biofilm-forming bacteria isolated from the interior surfaces of the hydroponic systems were able to facilitate the colonization of Salmonella on PVC coupons in hydroponic solutions. Finally, we compared the efficacy of NaOCl, H2O2, and SPC in sanitizing hydroponic systems. NaOCl turned out to be the most effective and economical option, whereas a rinsing step was found to be of vital importance to avoid harmful chlorination byproducts in the final products. This study contributes valuable insights into the management of microbial contamination in hydroponic farming, facilitating the advancement of sustainable agricultural practices in urban environments such as Singapore. Although both laboratory-scale and field-scale tests were conducted in this study, since the microbiome compositions in the hydroponic systems were found to be diverse and farm specific, the conclusions still need to be validated in other regions with different environmental conditions in the future.

MATERIALS AND METHODS

Microbial analysis of multitier NFT hydroponic systems

Samples were collected from three indoor farms using multitier NFT hydroponic systems to grow Brassica leafy vegetables in Singapore between August and September 2023. All three farms had maintained continuous operation for a period exceeding 3 months before sampling. After harvest, the nutrient solution was drained, and seven distinct components of the hydroponic system, as shown in Fig. 1a and illustrated in Fig. 1b, were swabbed in a 1 × 1 cm area with sterile cotton swabs as one sample. Ten samples were collected for each component at each farm.

The swabs were submerged in 10 mL of phosphate-buffered saline (PBS; Vivantis) in individual 50 mL Falcon tubes, transported to our laboratory on ice, and analyzed on the same day. After vortexing for 10 s, 1 mL of each 10 mL aliquot of the suspension in the Falcon tubes was serially diluted in 0.1% peptone water (Oxoid) and enumerated on plate count agar (PCA; Oxoid) after 3 days of incubation at 30°C.

DNA extraction and 16S and 18S rDNA gene sequencing analysis

DNA was extracted from the remaining 9 mL suspensions from the swabs after 1 mL was collected for microbial enumeration. Each 9 mL suspension was centrifuged at 9,000 × g for 5 min (Centrifuge 5810 R, Eppendorf, Hamburg, Germany). DNA extraction from the pellet was performed using the GeneJET Genomic DNA Purification Kit (Thermo Fisher Scientific, Vilnius, Lithuania) according to the manufacturer’s instructions. The quantity and quality of the DNA samples were assessed with a BioDrop DUO (BioDrop, Cambridge, UK), after which the samples were stored at −80°C for downstream sequencing and analysis.

16S and 18S rDNA gene amplification and sequencing were performed at Azenta US, Inc. (NJ, USA). Briefly, the next-generation sequencing library was constructed using the MetaVX Library Preparation Kit (Azenta US, Inc.). For the 16S rDNA gene, the amplicons generated covered the V3 and V4 hypervariable regions with the forward primer (5′ CCTACGGRRBGCASCAGKVRVGAAT 3′) and the reverse primer (5′ GGACTACNVGGGTWTCTAATCC 3′). The 18S rDNA gene was amplified at the V4 hypervariable region with the forward primer (5′ GGCAAGTCTGGTGCCAG 3′) and reverse primer (5′ ACTACGACGGTATCTRATCRTCTTCG 3′). Next-generation sequencing was carried out on an Illumina NovaSeq platform (Illumina, San Diego, USA). Automated cluster generation and 250 bp paired-end sequencing with dual reads were performed according to the manufacturer’s instructions.

The data were processed with open-source Quantitative Insights into Microbial Ecology 2 (version 2023.9) (45). The raw sequences were processed with Cutadapt (46) to trim primer and adapter sequences and remove low-quality sequences (quality threshold < 30 Phred score). The high-quality clean reads were then merged using VSEARCH (47). Sequences that failed to merge or did not contain primers were discarded. Subsequently, the reads were processed with deblur (48), which denoised the sequences and removed chimeras. The resulting ASVs were assigned taxonomies using the SILVA 138 reference database (49). To determine the differences in community structure between the farm samples from different farms, PCoA was performed using beta diversity metrics (weighted UniFrac, unweighted UniFrac and Bray–Curtis dissimilarity) after rarefaction to 4,000 sequences per sample. The microbiome composition was analyzed by constructing taxonomic bar plots at the phylum level and a heatmap at the genus level using the ggplot2 (version 3.4.4) (50) and complex heatmap (version 2.15.4) (51) packages, respectively, in RStudio (version 2023.12.1).

Biofilm former isolation and biofilm formation evaluation

Ten colonies with diverse morphologies were randomly picked from the PCA plates of the swab samples from the interior surfaces of the water tanks and subjected to Gram staining. The S. Brunei strain was previously isolated from the irrigation water of a hydroponic farm (13). Salmonella Enteritidis ATCC 13076, Salmonella Heidelberg ATCC 8326, Salmonella Newport ATCC 6962, and S. Typhimurium ATCC 14028 were purchased from the American Type Culture Collection.

The 10 environmental bacterial isolates and 5 Salmonella strains were cultured in tryptone soya broth (Oxoid) for 24 h at 30°C (for the 10 environmental isolates) and 37°C (for Salmonella), respectively. The bacterial suspensions were subsequently washed twice in PBS, diluted with a hydroponic nutrient solution to 5 log CFU/mL, and loaded into 6-well plates (2 mL/well). The hydroponic solution was formulated by mixing solution A and solution B (Watercircle Hydroponics Pte Ltd.) at 1:1 and diluting it with tap water to an electrical conductivity of 2 mS/cm. The solution was then filtered through a 0.22 µm filter (Millipore) for sterilization purposes. After a 5-day incubation at 25°C, each well was washed three times with sterile deionized water, stained with 2 mL of 0.1% (wt/vol) crystal violet (Sigma-Aldrich) solution for 20 min at room temperature, rewashed three times with deionized water, and incubated with 2 mL of 95% ethanol at 4°C for 45 min. The absorbance was read at 595 nm using a microplate photometer (IVD model, MultiskanTM FC, Thermo Fisher Scientific Inc., MA, USA).

The biofilm-forming capacity (measured by the total biofilm biomass) was determined according to the criteria established by Stepanović et al. (15). The optical density cutoff value (ODc) was the sum of the average OD of the negative control and three times the standard deviation of the negative control. The biofilm-forming ability was classified as “weak” if the OD < 2 ODc, “moderate” if the 2ODc < OD < 4 ODc, and “strong” if the OD > 4 ODc.

WGS analysis of the strong biofilm formers C2 and C3

DNA extraction was performed using the GeneJET Genomic DNA Purification Kit (Thermo Fisher Scientific, Vilnius, Lithuania) according to the manufacturer’s instructions. WGS was conducted at NovogeneAIT Genomics Singapore Pte Ltd. using an Illumina HiSeq 4000 with 350-bp insert DNA library preparation. These raw reads were then assembled into contigs using SPAdes (version 3.15.3) (52). The assembled contigs were used to identify isolated species via the KmerFinder tool (version 3.2) (53) available at the Center for Genome Epidemiology (CGE) of the Technical University of Denmark (https://cge.food.dtu.dk/services/KmerFinder/).

A total of 14 P. mexicana and 28 C. tuberculostearicum genomes were downloaded from the NCBI database, and a phylogenetic analysis was performed with the assembled contigs of the isolates. The analysis was performed based on the single nucleotide polymorphism tree using CSI Phylogeny (version 1.4) (54) at CGE. The phylogenetic tree was visualized using FigTree (version 1.4.4).

Effect of preformed C2 and C3 biofilms on Salmonella adhesion on PVC coupons

As illustrated in Fig. 3a, fully grown C. tuberculostearicum C2 and P. mexicana C3 suspensions were washed with PBS and diluted with a sterile hydroponic nutrient solution prepared as described above to 5 log CFU/mL. Each PVC coupon (2 × 4 cm, food-grade materials purchased from a local market in Singapore, presterilized by autoclave) was submerged in 20 mL of C. tuberculostearicum C2 or P. mexicana C3 suspensions and incubated at 25°C for 5 days.

After 5 days of incubation, the coupons were rinsed with sterile deionized water. Each coupon was then dipped in a 5 log CFU/mL S. Brunei suspension for 5 s and rinsed thrice with sterile deionized water. Thereafter, each coupon was placed into 20 mL of fresh hydroponic nutrient solution and incubated at 25°C for another 5 days. At 0, 24, 72, and 120 h, the bacteria were collected from the individual coupons by cell scrapers and enumerated on xylose lysine deoxycholate (XLD, Oxoid) agar, a selective culture medium for Salmonella.

Single- and dual-species biofilm formation by C2, C3, and Salmonella

Fully grown C. tuberculostearicum C2, P. mexicana C3, S. Brunei, and S. Typhimurium cultures were washed with PBS and diluted with a sterile hydroponic nutrient solution prepared as described above to 5 log CFU/mL. Sterile PVC coupons were submerged in 20 mL of bacterial suspension (either single or 1:1 mixed suspension, as illustrated in Fig. 3b) and incubated at 25°C for 5 days.

After 5 days of incubation, the coupons were rinsed with sterile deionized water. The bacteria were collected from the individual coupons by cell scrapers and plated on XLD (Oxoid) agar plates for Salmonella enumeration and on tryptone soya agar (TSA; Oxoid) plates for total bacteria enumeration.

Efficacy of sanitizers on S. Typhimurium, C2, and C3 biofilms on PVC coupons

Fully grown S. Typhimurium, C. tuberculostearicum C2, and P. mexicana C3 cultures were washed with PBS and diluted with a sterile hydroponic nutrient solution prepared as described above to 5 log CFU/mL. Sterile PVC coupons were submerged in 20 mL of bacterial suspension and incubated at 25°C for 5 days.

After 5 days of incubation, the coupons were rinsed with sterile deionized water and submerged in different sanitizer solutions, including 50 and 500 ppm NaOCl [Clorox Bleach, free and total chlorine concentrations determined by the sodium thiosulfate titration method (55)], 1.5% and 3.0% H2O2, and 1.0% SPC (20 mL sanitizer for each coupon in each tube). After 1, 6, and 12 h, the coupons were removed from the sanitizer solutions and rinsed with sterile deionized water. Bacteria were collected from the individual coupons by cell scrapers and enumerated on XLD (Oxoid) agar plates for Salmonella enumeration and on TSA (Oxoid) plates for C2 and C3 enumerations.

Efficacy of NaOCl and H2O2 in sanitizing NFT hydroponic systems

The sanitization trials were conducted at the same farms as the microbial analysis study between October 2023 and January 2024. All three farms had maintained continuous operation for a period exceeding 3 months before sampling. After harvest, NaOCl and H2O2 were directly added to the hydroponic solutions (~200 L hydroponic solution in each system), reaching final concentrations of 500 ppm NaOCl and 3.0% H2O2. The solutions continued to circulate in the systems at a flow rate of 0.06 m/s (same as during the growing period). After 6 and 12 h, the hydroponic solutions with sanitizers were drained, and swabbing sampling and bacterial enumeration were performed at the seven locations as described above.

Germination rate and yield of komatsuna grown in sanitized hydroponic systems

After one hydroponic system was sanitized with 500 ppm NaOCl for 12 h, the system was drained and subsequently refilled with a fresh hydroponic solution. In parallel, a control system was established without any sanitization treatment for comparative purposes. To assess the impact of chemical residues on seed germination, 10 mL of the hydroponic solution from the sanitized system and 10 mL from the control system were used to germinate 100 komatsuna seeds each. The germination rate was recorded on day 7 (n = 100). Thirty seedlings from each were then randomly selected and transferred back to the two hydroponic systems (sanitized and unsanitized). Twenty-eight days after the end of the growth cycle (day 35), the fresh weight and number of leaves of each crop were assessed (n = 30).

The chlorination byproduct analysis

At the end of the growth of komatsuna (day 35), the hydroponic solution from the sanitized systems (without rinsing or rinsing twice with tap water, each time for 1 h with a flow rate of 0.06 m/s) was collected for the analysis of a panel of chlorination byproducts (Table 3). The analysis was conducted at Eurofins Mechem Pte Ltd., according to standardized methods, as listed inTable 3. The chlorate levels in the harvested komatsuna were analyzed at SGS Testing & Control Services Singapore Pte Ltd., according to the standardized method QuPPe-PO (EURL-SRM version 12.2.2023), as listed in Table 3.

Statistical analysis

Statistical analyses were performed using Microsoft 365 and Excel 2021. A paired t test was used for data with two groups, and one-way analysis of variance was used for data with more than two groups. Differences were considered significant when P was <0.05.

ACKNOWLEDGMENTS

This study was supported by a Singapore Food Story (SFS) R&D Programme Theme 3 Competitive Grant on Food Safety and Consumer Science (W22W3D0001) funded by the Singapore Food Agency.

C.A.T.H. and D.L. conceived and designed the research. C.A.T.H. and G.S.Y.N. conducted the experiments. Y.H.Z., M.M.Z.T., and J.Y.L.T. contributed analytical tools. C.A.T.H., Y.H.Z., and M.M.Z.T. analyzed the data. B.L.P., W.Z., and D.L. secured grants and supervised the projects. C.A.T.H. and D.L. wrote the manuscript. All the authors have read and approved the manuscript.

Contributor Information

Dan Li, Email: dan.li@nus.edu.sg.

Christopher A. Elkins, Centers for Disease Control and Prevention, Atlanta, Georgia, USA

DATA AVAILABILITY

The raw read files were uploaded to the NCBI Sequence Read Archive (SRA) with accession no. PRJNA1093099, SAMN40647016, and SAMN40647017.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/aem.00672-24.

Supplemental figures. aem.00672-24-s0001.docx.

Figures S1 and S2.

aem.00672-24-s0001.docx (894.7KB, docx)
DOI: 10.1128/aem.00672-24.SuF1

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.

REFERENCES

  • 1. Bloem S, de Pee S. 2017. Developing approaches to achieve adequate nutrition among urban populations requires an understanding of urban development. Global Food Security 12:80–88. doi: 10.1016/j.gfs.2016.09.001 [DOI] [Google Scholar]
  • 2. SFA . 2023. Food farms in Singapore. Available from: https://www.sfa.gov.sg/food-farming/food-farms/farming-in-singapore. Retrieved 29 Mar 2024.
  • 3. SFA . 2022. Singapore food statistics 2021. Available from: https://www.sfa.gov.sg/docs/default-source/publication/sg-food-statistics/singapore-food-statistics-2021.pdf
  • 4. Godge M. 2022. Hydroponics: getting to the root of the myths. Singapore Food agency [Google Scholar]
  • 5. Resh HM. 2022. Hydroponic food production: a definitive guidebook for the advanced home gardener and the commercial hydroponic grower. CRC press, New York. [Google Scholar]
  • 6. Ge C, Rymut S, Lee C, Lee J. 2014. Salmonella internalization in mung bean sprouts and pre-and postharvest intervention methods in a hydroponic system. J Food Prot 77:752–757. doi: 10.4315/0362-028X.JFP-13-370 [DOI] [PubMed] [Google Scholar]
  • 7. Li Y, Zwe YH, Tham CAT, Zou Y, Li W, Li D. 2022. Fate and mitigation of Salmonella contaminated in lettuce (Lactuca sativa) seeds grown in a hydroponic system. J Appl Microbiol 132:1449–1456. doi: 10.1111/jam.15295 [DOI] [PubMed] [Google Scholar]
  • 8. Popa GL, Popa MI. 2021. Salmonella spp. infection-a continuous threat worldwide. Germs 11:88–96. doi: 10.18683/germs.2021.1244 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. WHO . 2015. WHO estimates of the global burden of foodborne diseases: foodborne disease burden epidemiology reference group 2007–2015. Geneva, Switzerland: WHO. https://apps.who.int/iris/handle/10665/199350. [Google Scholar]
  • 10. Aiyedun SO, Onarinde BA, Swainson M, Dixon RA. 2021. Foodborne outbreaks of microbial infection from fresh produce in Europe and North America: a systematic review of data from this millennium. Int J of Food Sci Tech 56:2215–2223. doi: 10.1111/ijfs.14884 [DOI] [Google Scholar]
  • 11. Grivokostopoulos NC, Makariti IP, Tsadaris S, Skandamis PN. 2022. Impact of population density and stress adaptation on the internalization of Salmonella in leafy greens. Food Microbiol 106:104053. doi: 10.1016/j.fm.2022.104053 [DOI] [PubMed] [Google Scholar]
  • 12. McClure M, Whitney B, Gardenhire I, Crosby A, Wellman A, Patel K, McCormic ZD, Gieraltowski L, Gollarza L, Low MSF, Adams J, Pightling A, Bell RL, Nolte K, Tijerina M, Frost JT, Beix JA, Boegler KA, Dow J, Altman S, Wise ME, Bazaco MC, Viazis S. 2023. An outbreak investigation of Salmonella Typhimurium illnesses in the United States linked to packaged leafy greens produced at a controlled environment agriculture indoor hydroponic operation–2021. J Food Prot 86:100079. doi: 10.1016/j.jfp.2023.100079 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Tham CAT, Zwe YH, Li D. 2021. Microbial study of lettuce and agriculture water used for lettuce production at Singapore urban farms. Food Control 126:108065. doi: 10.1016/j.foodcont.2021.108065 [DOI] [Google Scholar]
  • 14. Lipsman V, Shlakhter O, Rocha J, Segev E. 2024. Bacteria contribute exopolysaccharides to an algal-bacterial joint extracellular matrix. NPJ Biofilms Microbiomes 10:36. doi: 10.1038/s41522-024-00510-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Stepanović S, Vuković D, Hola V, Di Bonaventura G, Djukić S, Cirković I, Ruzicka F. 2007. Quantification of biofilm in microtiter plates: overview of testing conditions and practical recommendations for assessment of biofilm production by staphylococci. APMIS 115:891–899. doi: 10.1111/j.1600-0463.2007.apm_630.x [DOI] [PubMed] [Google Scholar]
  • 16. Altonsy MO, Kurwa HA, Lauzon GJ, Amrein M, Gerber AN, Almishri W, Mydlarski PR. 2020. Corynebacterium tuberculostearicum, a human skin colonizer, induces the canonical nuclear factor‐kB inflammatory signaling pathway in human skin cells. Immun Inflamm Dis 8:62–79. doi: 10.1002/iid3.284 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Mule P, Patil N, Gaikwad S. 2018. Corynebacterium Tuberculostearicum a potential pathogen in breast abscess—a case report. Int J Med Microbiol Trop Dis 4:42–44. doi: 10.18231/2455-6807.2018.0009 [DOI] [Google Scholar]
  • 18. Selvaraj GK, Wang H, Zhang Y, Tian Z, Chai W, Lu H. 2022. Class 1 In-Tn5393c array contributed to antibiotic resistance of non-pathogenic Pseudoxanthomonas mexicana isolated from a wastewater bioreactor treating streptomycin. Sci Total Environ 821:153537. doi: 10.1016/j.scitotenv.2022.153537 [DOI] [PubMed] [Google Scholar]
  • 19. Pouget C, Dunyach-Remy C, Magnan C, Pantel A, Sotto A, Lavigne J-P. 2022. Polymicrobial biofilm organization of Staphylococcus aureus and Pseudomonas aeruginosa in a chronic wound environment. Int J Mol Sci 23:10761. doi: 10.3390/ijms231810761 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. El-Azizi MA, Starks SE, Khardori N. 2004. Interactions of Candida albicans with other Candida spp. and bacteria in the biofilms. J Appl Microbiol 96:1067–1073. doi: 10.1111/j.1365-2672.2004.02213.x [DOI] [PubMed] [Google Scholar]
  • 21. Zwe YH, Yadav M, Ten MMZ, Srinivasan M, Jobichen C, Sivaraman J, Li D. 2022. Bacterial antagonism of Chromobacterium haemolyticum and characterization of its putative type VI secretion system. Res Microbiol 173:103918. doi: 10.1016/j.resmic.2021.103918 [DOI] [PubMed] [Google Scholar]
  • 22. McGlynn W. 2004. Guidelines for the use of chlorine bleach as a sanitizer in food processing operations. Oklahoma cooperative extension service. [Google Scholar]
  • 23. Abdelshafy AM, Hu Q, Luo Z, Ban Z, Li L. 2024. Hydrogen peroxide from traditional sanitizer to promising disinfection agent in food industry. Food Rev Int 40:658–690. doi: 10.1080/87559129.2023.2191690 [DOI] [Google Scholar]
  • 24. Mileto D, Mancon A, Staurenghi F, Rizzo A, Econdi S, Gismondo MR, Guidotti M. 2021. Inactivation of SARS-CoV-2 in the liquid phase: are aqueous hydrogen peroxide and sodium percarbonate efficient decontamination agents? ACS Chem Health Saf 28:260–267. doi: 10.1021/acs.chas.0c00095 [DOI] [PubMed] [Google Scholar]
  • 25. Wang R, Kalchayanand N, Schmidt JW, Harhay DM. 2013. Mixed biofilm formation by shiga toxin–producing Escherichia coli and Salmonella enterica serovar Typhimurium enhanced bacterial resistance to sanitization due to extracellular polymeric substances. J Food Prot 76:1513–1522. doi: 10.4315/0362-028X.JFP-13-077 [DOI] [PubMed] [Google Scholar]
  • 26. Lin Z, Wang G, Li S, Zhou L, Yang H. 2022. Dual-species biofilms formed by Escherichia coli and Salmonella enhance chlorine tolerance. Appl Environ Microbiol 88:e0148222. doi: 10.1128/aem.01482-22 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Li D, Wong CH, Seet MF, Kuan N. 2019. Isolation, characterization, and inactivation of Stenotrophomonas maltophilia from leafy green vegetables and urban agriculture systems. Front Microbiol 10:2718. doi: 10.3389/fmicb.2019.02718 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Teng Z, Luo Y, Alborzi S, Zhou B, Chen L, Zhang J, Zhang B, Millner P, Wang Q. 2018. Investigation on chlorine-based sanitization under stabilized conditions in the presence of organic load. Int J Food Microbiol 266:150–157. doi: 10.1016/j.ijfoodmicro.2017.11.027 [DOI] [PubMed] [Google Scholar]
  • 29. Hilgren J, Swanson KMJ, Diez-Gonzalez F, Cords B. 2007. Inactivation of Bacillus anthracis spores by liquid biocides in the presence of food residue. Appl Environ Microbiol 73:6370–6377. doi: 10.1128/AEM.00974-07 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Rodrigues PHV, Trientini MF, Fisher PR. 2022. Biofilm management in irrigation lines and hydroponic lettuce solutions using sanitizing chemicals. Acta Hortic 1335:703–710. doi: 10.17660/ActaHortic.2022.1335.89 [DOI] [Google Scholar]
  • 31. Thakulla D. 2021. Timing and rates of two hydrogen peroxide products to control algae and evaluation of water temperature on lettuce quality for hydroponic production. Oklahoma State University. [Google Scholar]
  • 32. Lau V, Mattson N. 2021. Effects of hydrogen peroxide on organically fertilized hydroponic lettuce (Lactuca sativa L.). Horticulturae 7:106. doi: 10.3390/horticulturae7050106 [DOI] [Google Scholar]
  • 33. Kriem LS, Pietzka C, Beckett M, Gärtling L, Wriedt B. 2023. Electrochemical in situ hydrogen peroxide production can reduce microbial load in bioponic nutrient solutions derived from organic waste. Agriculture 13:2122. doi: 10.3390/agriculture13112122 [DOI] [Google Scholar]
  • 34. Thakulla D, Dunn BL, Goad C, Hu B. 2022. Timing and rates of two products using hydrogen peroxide (H2O2) to control algae in ebb and flow hydroponic systems. HortScience 57:32–39. doi: 10.21273/HORTSCI16193-21 [DOI] [Google Scholar]
  • 35. Khan MIR, Khan NA, Masood A, Per TS, Asgher M. 2016. Hydrogen peroxide alleviates nickel-inhibited photosynthetic responses through increase in use-efficiency of nitrogen and sulfur, and glutathione production in mustard. Front Plant Sci 7:44. doi: 10.3389/fpls.2016.00044 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Zhang XL, Jia XF, Yu B, Gao Y, Bai JG. 2011. Exogenous hydrogen peroxide influences antioxidant enzyme activity and lipid peroxidation in cucumber leaves at low light. Sci Horticul 129:656–662. doi: 10.1016/j.scienta.2011.05.009 [DOI] [Google Scholar]
  • 37. Zhou J, Wang J, Shi K, Xia XJ, Zhou YH, Yu JQ. 2012. Hydrogen peroxide is involved in the cold acclimation-induced chilling tolerance of tomato plants. Plant Physiol Biochem 60:141–149. doi: 10.1016/j.plaphy.2012.07.010 [DOI] [PubMed] [Google Scholar]
  • 38. Goslan EH, Krasner SW, Bower M, Rocks SA, Holmes P, Levy LS, Parsons SA. 2009. A comparison of disinfection by-products found in chlorinated and chloraminated drinking waters in Scotland. Water Res 43:4698–4706. doi: 10.1016/j.watres.2009.07.029 [DOI] [PubMed] [Google Scholar]
  • 39. Legay C, Rodriguez MJ, Sérodes JB, Levallois P. 2010. Estimation of chlorination by-products presence in drinking water in epidemiological studies on adverse reproductive outcomes: a review. Sci Total Environ 408:456–472. doi: 10.1016/j.scitotenv.2009.10.047 [DOI] [PubMed] [Google Scholar]
  • 40. Nieuwenhuijsen MJ, Toledano MB, Elliott P. 2000. Uptake of chlorination disinfection by-products; a review and a discussion of its implications for exposure assessment in epidemiological studies. J Expo Anal Environ Epidemiol 10:586–599. doi: 10.1038/sj.jea.7500139 [DOI] [PubMed] [Google Scholar]
  • 41. Van Haute S, Sampers I, Holvoet K, Uyttendaele M. 2013. Physicochemical quality and chemical safety of chlorine as a reconditioning agent and wash water disinfectant for fresh-cut lettuce washing. Appl Environ Microbiol 79:2850–2861. doi: 10.1128/AEM.03283-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Rico D, Martín-Diana AB, Barat JM, Barry-Ryan C. 2007. Extending and measuring the quality of fresh-cut fruit and vegetables: a review. Trends Food Sci Technol 18:373–386. doi: 10.1016/j.tifs.2007.03.011 [DOI] [Google Scholar]
  • 43. Artés F, Gómez P, Aguayo E, Escalona V, Artés-Hernández F. 2009. Sustainable sanitation techniques for keeping quality and safety of freshcut plant commodities. Postharvest Biol Technol 51:287–296. doi: 10.1016/j.postharvbio.2008.10.003 [DOI] [Google Scholar]
  • 44. EFSA Panel on Contaminants in the Food Chain (CONTAM) . 2015. Risks for public health related to the presence of chlorate in food. EFSA J 13:4135. doi: 10.2903/j.efsa.2015.4135 [DOI] [Google Scholar]
  • 45. Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, Alexander H, Alm EJ, Arumugam M, Asnicar F, et al. 2019. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol 37:852–857. doi: 10.1038/s41587-019-0209-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Martin M. 2011. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J 17:10. doi: 10.14806/ej.17.1.200 [DOI] [Google Scholar]
  • 47. Rognes T, Flouri T, Nichols B, Quince C, Mahé F. 2016. VSEARCH: a versatile open source tool for metagenomics. PeerJ 4:e2584. doi: 10.7717/peerj.2584 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48. Amir A, McDonald D, Navas-Molina JA, Kopylova E, Morton JT, Zech Xu Z, Kightley EP, Thompson LR, Hyde ER, Gonzalez A, Knight R. 2017. Deblur rapidly resolves single-nucleotide community sequence patterns. mSystems 2:00191–16. doi: 10.1128/mSystems.00191-16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, Peplies J, Glöckner FO. 2013. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res 41:D590–6. doi: 10.1093/nar/gks1219 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50. Valero-Mora PM. 2010. ggplot2: elegant graphics for data analysis. Journal of statistical software, book reviews 35:1–3. [Google Scholar]
  • 51. Gu Z. 2022. Complex heatmap visualization. iMeta 1:e43. doi: 10.1002/imt2.43 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52. Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, Lesin VM, Nikolenko SI, Pham S, Prjibelski AD, Pyshkin AV, Sirotkin AV, Vyahhi N, Tesler G, Alekseyev MA, Pevzner PA. 2012. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol 19:455–477. doi: 10.1089/cmb.2012.0021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53. Clausen PTLC, Aarestrup FM, Lund O. 2018. Rapid and precise alignment of raw reads against redundant databases with KMA. BMC Bioinformatics 19:307. doi: 10.1186/s12859-018-2336-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54. Kaas RS, Leekitcharoenphon P, Aarestrup FM, Lund O. 2014. Solving the problem of comparing whole bacterial genomes across different sequencing platforms. PLoS One 9:e104984. doi: 10.1371/journal.pone.0104984 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55. Saboe D, Hristovski KD, Burge SR, Burge RG, Taylor E, Hoffman DA. 2021. Measurement of free chlorine levels in water using potentiometric responses of biofilms and applications for monitoring and managing the quality of potable water. Sci Total Environ 766:144424. doi: 10.1016/j.scitotenv.2020.144424 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental figures. aem.00672-24-s0001.docx.

Figures S1 and S2.

aem.00672-24-s0001.docx (894.7KB, docx)
DOI: 10.1128/aem.00672-24.SuF1

Data Availability Statement

The raw read files were uploaded to the NCBI Sequence Read Archive (SRA) with accession no. PRJNA1093099, SAMN40647016, and SAMN40647017.

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES