Effect of three commonly used aquaculture chemicals against enteropathogenic Escherichia coli and Salmonella spp.

Enteric bacteria such as Escherichia coli and Salmonella spp. are significant fish pathogens and related to thousands of cases of food-borne diseases every year in human. Since aquatic environments are reservoirs of these pathogens, they may contaminate the food fish and result in outbreaks. Therefore, it is crucial to reduce or eliminate these pathogens from aquaculture facilities. We tested effectiveness of three commonly used aquaculture chemicals, viz., lime, hydrogen peroxide and zeolite on bacterial load, Escherichia coli and Salmonella spp. under laboratory and earthen pond conditions where they were applied at a dose recommended for freshwater aquaculture. Results of the bacteriological study showed that lime had a significant role in reducing bacterial abundance from an initial value of 1.5 × 103 to 1.9 × 103 cfu/ml and 3.9 × 103 to 6.3 × 103 to a final value of 1.2 × 103 to 1.5 × 103 cfu/ml and 1.9 × 103 to 6.3 × 103 cfu/ml within 24 h under aquaria and pond condition, respectively. A complete inactivation of E. coli was observed in lime treated aquaria and ponds 24 h post-treatment, whereas Salmonella spp. remained unaffected in pond water but inactivated under aquarium condition after same period. However, neither hydrogen peroxide nor zeolite treatment reduced total bacterial count as well as E. coli and Salmonella even after 1-week post-treatment. It is, therefore, recommended that the water treatment chemicals should be applied in aquaculture ponds at appropriate dose, and farmers need to adopt biosecurity measures to ensure food safety.


Introduction
Aquaculture has become the fastest-growing food-producing sector in the world and currently provides approximately half of all the fish for human consumption (Ghasemi et al. 2018;Kyule-Muendo et al. 2022). This has been possible primarily due to the fact that the sector has diversified, intensified and advanced technologically. Inland finfish aquaculture is the most common type of aquaculture operation in Asia particularly in China, India, Viet Nam, Bangladesh, Thailand, Indonesia and the Philippines, and accounts for approximately 90% of global aquaculture production by volume (FAO 2020; Rocha et al. 2022). Aquafarmers in these countries use various chemicals in their farms during different stages of the culture system from pond preparation up to harvesting. It was reported that calcium carbonate, calcium hydroxide, calcium oxide, calcium sulfate dihydrate, sodium chloride, zeolite were used as water and sediment treatment compounds; copper sulfate; fentin acetate, malachite green, methylene blue, endosulfan, lindane, dichlorvos, trichlorfon, rotenone as pesticides; benzalkonium chloride, calcium hypochlorite, chlorine, copper complex solution, hydrogen peroxide, formaldehyde, potassium permanganate as disinfectant; ammonium phosphate, ammonium sulfate, urea, triple super phosphate as fertilizer and erythromycin, amoxicillin, ampicillin, sulfamethoxazole, chlortetracycline, doxycycline, oxytetracycline, trimethoprim and oxolinic acid etc. were used as antibiotics (Plumb 1992;Rico et al. 2012;Islam et al. 2014). Many of these compounds improve water quality parameters while some influence planktonic biomass and bacteria in the pond (Rafiee and Saad 2006;Mainous et al. 2010). While the use of the aforesaid chemicals possibly plays roles in increased aquaculture production, they are thought to have exerted deleterious effects on the structure and functioning of the ecosystem.
Microorganisms residing in the aquaculture facilities may have positive or negative effects on the outcome of aquaculture operations (Al-Harbi 2003;Al-Harbi and Uddin 2008;Ouattara et al. 2011) and among them faecal indicator bacteria (FIB) are usually enumerated to evaluate the level of microbial water contamination. The abundance of these FIBs is supposed to be correlated with the density of pathogenic microorganisms from faecal origin and is thus an indication of the sanitary risk associated with the various water utilizations (Ouattara et al. 2011). Escherichia coli and other intestinal enterococci such as Salmonella spp. are the most frequently used indicators of faecal pollution, and it was demonstrated by epidemiological studies that they were better indicators of the human risk associated with waters than coliforms (Edberg et al. 2000;Fewtrell and Bartram 2001). With the increasing concern over human health and food safety, enumeration of enteropathogenic E. coli and Salmonella spp. in aquaculture system has become increasingly important. In developed countries, information on the effects of various chemicals and drugs on various agents of the food web are available that allow them to approve or ban particular aqua chemical. Ecotox (2010) reported that many aquaculture disinfectants are moderately to highly toxic to planktonic and macroinvertebrate species with acute LC50s ranging between 1 and 100,000 µg/L, while Wollenberger et al. (2000) reported that relatively low concentrations of oxolinic acid disrupts reproduction in Daphnids. Antibiotics have also been reported to influence blue-green algae and bacterial abundance in the aquaculture system (Robinson et al. 2005;Brain et al. 2008;Haque et al. 2014;Hossain et al. 2014;Mannan et al. 2020), which may indirectly affect the food web.
In aquaculture systems, disinfection is generally used to eliminate pathogenic microorganisms from inorganic substances (APIC 1996). We previously reported the influence of three commonly used water treatment chemicals viz., lime, hydrogen peroxide (commercially available as oxy-more) and zeolite (commercially available as zeo-prime) on the water quality parameters and plankton biomass in pond and aquaria systems (Ferdous et al. 2013). Among these 3 chemicals, hydrogen peroxide sometime function as weak microbiocide and have the capability to remove sea lice in salmon (Bruno and Raynard 1994). Since, no comparable data are available for these chemicals that are already in widespread commercial use in developing countries in Asia including Bangladesh, the present study was designed to determine the efficacy of lime, hydrogen peroxide (oxy-more) and zeolite (zeo-prime) against bacterial load, E. coli and Salmonella spp. under laboratory and earthen pond conditions. Such information will help ensure rational use of various aquaculture chemicals and drugs, and environmental sustainability of aquaculture sector.

Chemicals and materials
Two types of water treatment chemicals viz., lime and zeolite and a disinfectant viz., hydrogen peroxide were used in the study. The characteristics of these chemicals are listed in Table 1.

Experimental setup
Nine glass aquaria (size 37 × 30 × 60 cm) were set at the Laboratory of Fish Harvesting, Department of Fisheries Technology, Bangladesh Agricultural University (BAU), Mymensingh, Bangladesh that were filled to a depth of 15 cm with water collected from earthen experimental ponds located at the vicinity of the Faculty of Fisheries, BAU during autumn (August-September) 2012 as a part of a series of experiments conducted from 2012 to 2018. They were held at room temperature range of 25.8-27.5 °C. For pond experiment, 9 experimental ponds with an average water area of 0.65 decimal with more or less similar basin conformation, bottom-soil type and contour were used.

Water treatment procedure and sampling
Water in aquaria was treated with lime, zeolite (zeo-prime) and hydrogen peroxide (oxy-more) at the rate of 0.27, 0.07 and 0.07 g/m 3 while in experimental ponds water was treated with lime at the rate of 1 kg/decimal, oxy-more at the rate of 3 g/decimal and zeolite at the rate of 200 g/decimal. The dose for lime was based on the requirement of the water and/ or soil for freshwater aquaculture in this region, and that for oxy-more and zeo-prime was based on manufacturer's recommendations. Water sample from aquaria were collected before treatment to determine physico-chemical and bacteriological parameters, data of which were regarded as control. Similarly, samples were collected every 24 h from each aquarium after treatment for consecutive 7 days and analyzed for water quality parameters and bacteriological study. As for pond treatment, 1 L water samples were taken from three points in each pond prior to the application of chemicals and regarded as control. Then samples were collected 24 h post-treatment, and daily thereafter for consecutive 7 days.

Water quality monitoring and bacteriological analysis
Measurements of temperature, dissolved oxygen (DO), total alkalinity, phosphate-phosphorous, nitrate-nitrogen, ammonia-nitrogen and pH were recorded every 24 h throughout the experimental period as described previously (Ferdous et al. 2013). Water transparency was recorded for pond only. As for bacteriological analyses, water samples were collected in sterile glass bottles (250 mL) 10-12 cm below the water surface from each aquarium and pond. Decimal dilutions were made in physiological saline (0.85% w/v NaCl in deionized water) and plated on to agar plates. For each dilution, 0.1 mL aliquots were plated on trypton soya agar (TSA), (Difco, Detroit, Michigan) in duplicate by using the spread plate technique (APHA 1995). Plates were incubated at 37 °C for 48 h. After incubation, the plates having 30-300 colonies were selected for counting with a Leica Quebec Darkfield Colony Counter (Leica, Inc., Buffalo, New York) and resulting colonies were enumerated by converting to cfu/mL by multiplication by dilution factors. Then each different colony types were streaked on selective media viz., EBM (Eosin methylene blue) agar plate and SS (Salmonella-Shigella) agar for detection of E. coli and Salmonella spp. respectively.

Statistical analysis
The data obtained in the experiment were recorded and preserved in computer. The data obtained in the experiment were analyzed by using SPSS version 11.5 (SPSS Inc., Chicago, USA). Significant differences were determined among treatments at the 5% level (p < 0.05).

Results
Over the entire experimental period, average water temperature was 26.4 ± 0.6 °C in aquarium and 30.6 ± 1.7 °C in experimental ponds (Supplementary Tables S1 and S2). There was little or no change in aquarium water temperature throughout the experimental period, but significant changes (p < 0.5) were observed in water temperature for pond experiment. Water transparency in pond, on the other hand, varied from 23.0 ± 0.9 to 24.0 ± 2.3 cm for control before chemical treatment that changed significantly to 27.0 ± 2.0 to 28.0 ± 0.6 cm, 1 day after all three chemical treatments (Fig. 1). The mean value of transparency of the pond water after 7 days was 27.5 ± 0.7 cm with highest recorded for lime treatment. Differences between initial and final values among the treatments was found to be significant (p < 0.05).
In the present study, average initial values of pH were 7.9 ± 0.10 in aquarium and 7.3 ± 0.05 in experimental ponds (Fig. 1). Changes in pH were not significant in different treatments except for lime treatment. It was observed that after lime treatment pond pH changed from 7.3 ± 0.05 to 8.2 ± 0.10, which indicated that lime has a significant role on water pH which was similar to that observed for aquarium treatment. These values remained more or less stable in both aquarium and pond condition with little fluctuations. DO, on the other hand, hand a mean value of 4.5 ± 2.0 ppm in pond water for lime, 3.8 ± 1.2 to 9. 5 ± 0.8 ppm for oxy-more and 3.5 ± 1 to 9.0 ± 1.1 for zeo-prime treatments. More or less similar observation was found for aquarium treatment. These values showed significant change (p < 0.05) after day 1 in both laboratory and pond conditions. Total alkalinity varied from 100 ± 15.0 to 140 ± 13.2 mg/l in pond with liming and 110 ± 4.0 to 140 ± 6.2 for zeo-prime, respectively. The highest value of total alkalinity content (160 ± 3.6 mg/L) was recorded for liming and showed no significant variation for oxy-more treatment condition (Data not shown). Total hardness, on the other hand, was above 960 ppm for pond water. Also the study showed that significant change in hardness in pond condition occurred after lime treatment (1130 ± 12.4 to 1250 ± 23.0).
The result of the quantitative estimation of aerobic bacteria showed that during the period of study bacterial loads ranged from 2.6 × 10 3 , 3.2 × 10 3 , 2.9 × 10 3 cfu/ml to 1.2 × 10 3 , 3.0 × 10 3 , 2.7 × 10 3 cfu/ml before and after 3rd day of treatment in pond with lime, oxy-more, zeo-prime treatment, respectively (Fig. 1). The study also showed that after liming E. coli population was disappeared on day 1 for both aquarium and pond experiments, indicating efficacy of lime as a disinfectant. Salmonella, on the other hand, was present in pond throughout the experimental period while was found to have disappeared on day 1 under aquarium condition ( Table 2). As for oxy-more and zeo-prime treatment, E. coli and Salmonella species were unaffected. Transparency (cm) Bacterial load (log cfu/ml)

Aquarium water
Pond water Transparency (cm)

Discussion
A comparative study on the effect of three commonly used aquaculture chemicals on bacteriological parameters showed substantial reduction of bacterial counts and enteropathogenic E. coli and Salmonella spp. using lime treatment after 24 h, but no evidence was found for inactivation of these pathogens using oxy-more and zeo-prime.

Changes in bacteriological parameters
There were two types of water samples used in this study and their bacteriological parameters were measured at room temperature (26.4 ± 0.6 °C). Untreated water from pond was used as control for both aquaria and pond treatments.
The changes in parameters for treatment aquaria and ponds were statistically analyzed over a period of 7 days to identify their effect. The results of the viable bacterial cell count for various treatments from 0 to 7 days for control and treatment showed that at the start of the experiment (0 h), the average microbial load in the aquaria and pond water was 1.7 × 10 3 and 5.1 × 10 3 cfu/ml respectively with 1.5 × 10 3 to 1.9 × 10 3 cfu/ml for aquaria and 3.9 × 10 3 to 6.3 × 10 3 cfu/ml for pond water. Although the source of water was from same three earthen ponds, the variation obtained from triplicate source could have arisen due to mixing of water in the pond and error during sampling. The 24 h post-treatment values were recorded to be 1.2 × 10 3 to 1.5 × 10 3 cfu/ml for aquaria and 1.9 × 10 3 to 6.3 × 10 3 cfu/ml for pond water, where a significant decline from 1.9 × 10 3 to 1.2 × 10 3 cfu/ml was only observed for lime treatment (p < 0.05) (Fig. 1A). With the application of lime, pH rose from an initial value of 7.3 ± 0.05 to 7.9 ± 0.05 at temperature 27.6-32.5 °C, indicating effect of alkaline pH on bacterial load. It was reported that microbiological quality of water and sludge was related to the pH after liming treatment (Ganguly et al. 2000;Bina et al. 2004;Jin and Kirk 2018) and water composition had little effect on bacterial survival. In our study, pH drop was most significant 24 h post-treatment, which was more or less steady at about 8.0 throughout the experimental period. Other water quality parameters including DO, alkalinity, transparency and hardness also showed more or less similar trend as pH from day 0 to 7. We previously reported changes in these physico-chemical parameters (Ferdous et al. 2013).
In the cases of hydrogen peroxide treatment, bacterial load significantly decreased from 3.9 × 10 3 to 3.1 × 10 3 cfu/ml at 24 h post-treatment for aquaria only, while there was little or no change for pond water (Fig. 1B). Generally, hydrogen peroxide works as a weak mirobiocide and used as a preoxidant in municipal water treatment (Linley et al. 2012).
In the present study, hydrogen peroxide was no able to exert its microbiocidal effect under pond condition while its influence was evident under aquarium condition. This is probably related to fact that the function of hydrogen peroxide is largely dependent on physico-chemical condition of water e.g., temperature, pH, and presence of other catalyzers in the pond water (Rudra et al. 2005). In the cases of zeolite treatment, however, bacterial load showed little or no change during the study period under both aquarium and pond condition (Fig. 1C). The presence of E. coli and Salmonella was also observed for three chemicals under both aquaria and pond conditions. It was observed that after liming, E. coli population disappeared under both aquarium and pond conditions, but inactivation of Salmonella was only possible for aquarium treatment ( Table 2). The effect of lime on pathogenic bacteria is well documented (Grabow et al. 1978;Strauch 1999;Bina et al. 2004). In the present experiment, its effectiveness in pond condition, however, was not achieved. This is possibly related to comparatively lower increment in water pH. It was reported that the efficacy of inactivation by Ca(OH) 2  of Salmonella spp. depended on the pH, and was inactivated at pH 10.0-10.7 after 24 h (Dixon et al. 2011), while pH 11.0-11.5 with a retention time of 4 h was necessary to inactivate pathogenic E. coli and Salmonella typhae (Riehl et al. 1952). In our study, we conducted sampling of water every 24 h, and we predict that longer retention time probably resulted in inactivation of E. coli population in both aquarium and pond conditions, but inactivation of Salmonella was only accomplished for aquarium treatment alone. We therefore, predict that application of higher lime dose is necessary to achieve the desired microbiocidal effect in the water. In the cases of hydrogen peroxide and zeolite treatments, the pathogenic bacteria were not inactivated under both aquarium and pond condition ( Table 2).

Relationship between limnological parameter and bacterial abundance
During application of water treatment chemicals, the limnological parameters play an important role on the effectiveness of the applied chemicals. Among them temperature, transparency, DO, pH, nitrate and hardness can be regarded as most important. In the present study, we determined the changes in those parameters. Water temperatures measured in different aquaria and ponds ranged between 26.4 and 30.6 °C, which was more or less similar to the suitable range of 26-28 °C for tropical and subtropical species (Boyd 1990;Payne et al. 2016). Water transparency is another factor which indirectly is an estimate for abundance of suspended materials in water. The value for pond water changed from 23.0 ± 0.9 to 24.0 ± 2.3 cm before treatment to 27.0 ± 2.0 to 28.0 ± 0.6 cm, 1 day after lime, zeolite and hydrogen peroxide treatments. Generally, agricultural lime increases buffering capacity of pond water and absorption of dissolved inorganic phosphorus by sediments, which ultimately increases availability of food organisms for fish (Boyd 1982). While the significant increase in transparency for lime treatment was previously reported by others (Sipaúba-Tavares et al. 2003), our observation is similar to those where we observed highest significant increment from 23 ± 0.9 to 28 ± 0.6 cm at 24 h post-treatment. This increment was negatively correlated to bacterial abundance for lime treatment (Table 3). Similarly, zeolite and hydrogen peroxide showed significant increase in transparency 24 h after treatment. Generally, zeolite functions as an ion exchanger and adsorbent to decontaminate water and wastewater in aquaculture (Ghasemi et al. 2018) that is achieved by either flocculation of suspended solids, ionic exchange and absorption of ammonium ions by zeolites (Briggs and Funge-Smith 1996), and our observation of increased transparency suggests that many of the suspended solids present in pond water was effectively precipitated by zeolite. Similar to lime, this increment was also negatively correlated with bacterial abundance (Table 3). In recent years, the use of hydrogen peroxide has increased in aquaculture primarily due to many of its advantages including their broad-spectrum activity, e.g., efficacy against bacterial endospores, their lack of environmental toxicity following their complete degradation into non-toxic by-products (water and oxygen), their surface corrosiveness and lack of off odor (Repine et al. 1981;Linley et al. 2012). Therefore, the significant increase in water transparency for hydrogen peroxide 24 h post-treatment is probably related to removal of suspended materials from the experimental pond water. This parameter was, however, did not show any consistent correlation with bacterial abundance. pH is an important factor in fish culture. During the study period, we observed that the pH ranged from 7.9 ± 0.10 to 7.3 ± 0.05 in aquarium and pond water. It was found that changes in pH was negatively correlated to bacterial abundance for lime treatment (Table 3). Similarly, the changes in pH were also negatively correlated to bacterial abudance for hydrogen peroxide treatment for both aquarim and pond condition but not consistnet for zeolite. DO of water body is also very important factor for fish culture. During the experimental period the DO content in aquarium water changed dynamically for lime, oxy-more and zeo-prime treatments. It was found that DO content did not change on day 1 after lime treatment, where there was a slight increase on day 2 (p > 0.05) and a sharp decline until day 7. Generally, lime treatment does not cause any change in DO content in water and generally related to aeration of water. The values recorded, however, were suitable for survival of fish and aquatic organisms including microbes. Zeolite and hydrogen peroxide treatment, however, resulted in significant rise in DO content during the experimental period. The rise in DO content during zeolite treatment is not well understood, but it was reported that they are formed by AlO 4 and SiO 4 tetrahedra and connected by a shared oxygen atom (Xue et al. 2018). Hydrogen peroxide is the chemical of choice that is used to enhance the amount of DO in aquaculture ponds and improve water quality. Its microbiocidal effect was evident in aquarium and pond condition where bacterial load was reduced significantly during 24 h post-treatment.

Implications for aquaculture
Application of optimal doses of aquaculture chemicals to treat pond water and sediment and inactivate enteropathogenic E. coli and Salmonella spp. will ensure safety and increased production in aquaculture. Fish farmers need to be trained to ensure an optimum alkaline environment in their ponds so that stressful environment is created for pathogens, and suitable environment is created for increased primary and secondary production.

Conclusion
In aquaculture, problems that are frequently encountered include soft acidic waters, low natural productivity, high clay turbidity, oxygen depletion and acid sulfate soils. Under such circumstances application of lime, zeolite and hydrogen peroxide can render water quality parameters suitable for fish culture provided that they are applied at appropriate dose. Farmers also need to adopt fish health management practices and biosecurity measures in their ponds. Since environmental parameters remain are more or less same throughout the year in Bangladesh, rational use of aqua drugs is recommended to increase productivity and ensure sustainability.