Ultrasound treatments improve the microbiological quality of water reservoirs used for the irrigation of fresh produce

Food Research International 75 (2015) 140–147

Contents lists available at ScienceDirect

Food Research International journal homepage: www.elsevier.com/locate/foodres

Ultrasound treatments improve the microbiological quality of water reservoirs used for the irrigation of fresh produce Maria V. Villanueva a, Maria C. Luna b, Maria I. Gil b, Ana Allende b,⁎ a b

Contariego S.L., Polígono Industrial Oeste 39/4, Murcia E-30169, Spain Research Group on Quality, Safety and Bioactivity of Plant Foods, Department of Food Science and Technology, CEBAS-CSIC, Campus de Espinardo, 25, Murcia E-30100, Spain

a r t i c l e

i n f o

Article history: Received 30 January 2015 Received in revised form 18 May 2015 Accepted 27 May 2015 Available online 30 May 2015 Keywords: Food safety Water treatment Disinfection Primary production Microbial indicators

a b s t r a c t Irrigation water has been highlighted as a source of microbial contamination in produce. Water treatment has been recommended as an intervention strategy to reduce microbial risks associated to irrigation water. Commercial water treatments mostly depend on chemical agents; although growers search for greener alternatives to chemical biocides. Ultrasounds (US) have been proposed as an environmentally friendly technology for irrigation water. In the present study, the suitability of two US treatments (20 kHz: US20 and 40 kHz: US40 at a specific energy (Es) of 745 J/L) and one chlorine treatment (1–2 ppm free chlorine) was evaluated and compared to the untreated control. Five water reservoirs belonging to five commercial intensive farms were selected as representative of irrigation practices generally used in south of Europe. All tested water treatments were able to reduce microbial loads, including Escherichia coli (0.5–0.6 log units), to values that were accepted in most of the recommended guidelines of good agricultural practices (≤2 log units). The obtained reductions were lower than those previously reported for these water treatment technologies in lab-scale tests. High microbial reductions are commonly obtained in laboratory studies, yielding impressive results. However, when the same treatments are applied under real commercial conditions, microbial reductions are usually less impressive. All water treatments were able to reduce COD of irrigation water when compared to the untreated control. COD reductions obtained using chlorine (≥430 mg/L) and US 20 (~ 100 mg/L) were higher than those observed using US 40 (b 50 mg/L). The impact of the water quality on the efficacy of US treatments was evaluated in two types of water including surface water and treated wastewater. It was found that the quality of the irrigation water significantly influences the efficacy of the ultrasound treatment. Correlations between indicator parameters have been also evaluated. Obtained results showed that high algae counts were well correlated with high levels of coliforms and E. coli. It could be concluded that US can be proposed as an alternative water treatment to chemical treatments to preserve microbial quality of irrigation water stored in water reservoirs reducing the environmental impact. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction Several authors have reviewed the most important factors for microbial contamination of fruits and vegetables. They conclude that reducing microbial contamination of irrigation water and soil is the most effective target for the prevention and control of produce contamination (FAO/ WHO, 2008). Water from different sources can be used in crop production and it can be generally ranked by the microbial contamination hazard (Harris et al., 2012). In order of increasing risk these are potable or rain water, groundwater from deep wells, groundwater from shallow wells, surface water, and finally raw or inadequately treated wastewater (EFSA, 2014). Considerable seasonal and climatic variations in water quality are possible and particularly notable where supplies are drawn from surface sources and store in water reservoirs. If the irrigation ⁎ Corresponding author. E-mail address: [email protected] (A. Allende).

http://dx.doi.org/10.1016/j.foodres.2015.05.040 0963-9969/© 2015 Elsevier Ltd. All rights reserved.

water is obtained from a surface water reservoir, it is considered as a high risk for pathogen contamination with foodborne pathogens such as Salmonella spp. and pathogenic Escherichia coli, because they are open to many routes by which microorganism causing plant disease or human food-borne illness can enter (Jones, Worobo, & Smart, 2014). This is the case of most of the irrigation water used in the south-east production area of Spain (Castro-Ibáñez, Gil, Tudela, Ivanek, & Allende, 2015). Agricultural irrigation in this area is usually carried out by water from the Tajo-Segura river transfer. Water treatment is therefore, one of the most recommended mitigation options and prevention strategies of contamination for irrigation water, particularly for surface water in reservoirs (FAO, 2003). When there is water scarcity, treated wastewater is sometimes used to compensate it, although there is still averseness about its use. In most of the cases, the wastewater is municipal wastewater that passed through waste water treatment plants. This is currently carried out in crops where the irrigation water is not in direct contact with the edible

M.V. Villanueva et al. / Food Research International 75 (2015) 140–147

141

part of the plant, such as peppers (cultivated in greenhouses) and artichokes (grown in open-field). The use of wastewater irrigation provides important nutrients for crops (Steele & Odumeru, 2004), but improperly treated wastewater can contain foodborne pathogens (FAO/WHO, 2008). Physico-chemical and biological characteristics of irrigation water should be also controlled due to the risk of micro and macroalgae accumulation which harms irrigation systems and blocks filters and pumps, slowing down the irrigation process (Nakayama & Bucks, 1991). Algae, periphyton and bottom sediments will also affect the microbial quality of irrigation water as the microbial concentrations will depend on exchange with these microbial reservoirs in direct contact with water (Pachepsky, Shelton, McLain, Patel, & Robert, 2011). Good Agricultural Practices (GAP) guidelines recommend growers to periodically test the water they use for microbial and chemical contaminants as a preventive measure (FAO/WHO, 2008). In the case of an insufficient water quality, intervention strategies based on water treatment should be taken. Nowadays, chemical sanitizers (mostly chlorine based sanitizers) are still the most commonly used water treatments. However, greener alternatives to chemical biocides are being demanded, particularly for organic production. In fact, concerns have risen in the last years regarding both situations, the absence of a water treatment and the excessive use of potentially toxic chemicals to treat irrigation water. Norton-Brandão, Scherrenberg, and Jules (2013) presented a critical review on urban reclamation technologies for irrigation including ozone, ultraviolet, filtration, sodium hypochlorite, titanium dioxide and electrolysis. Despite different technologies being available for application of treated wastewater in irrigation, the use of effluent in agriculture is not being properly managed in the majority of cases. Commercially available solutions found for these concerns are chemical disinfection such as sodium hypochlorite and calcium hypochlorite, UV-C light, filtration, electrochemical disinfection and ultrasounds (US) among others (Jones et al., 2014; Norton-Brandão et al., 2013). Among these methods, the efficacy of chlorination has been widely reported as the main technology used for the removal of biohazards to disinfect irrigation water (Suslow, 2013). However, the limitations of chlorine in terms of byproducts and the potential negative effect on the environment have limited its used in agricultural water. Recently, US treatments have been introduced as a greener technology based on the cavitation and sonolysis phenomenons. Cavitation can be defined as the process of formation, growth and subsequent collapse of microbubbles occurring in an extremely short time interval (milliseconds) (Mason & Peters, 2002). During the collapse of the microbubbles or so-called cavities, large amounts of energy are released. Sonolysis refers to the formation of radicals induced by cavitation bubbles during ultrasound. Thus, the advantages of ultrasound include potential chemical-free and simultaneous oxidation, thermolysis, shear degradation, and enhanced mass transfer processes together (Sharma, Sanghi, & Mudhoo, 2012). Recently, more research has been directed toward ultrasound application in disinfection of process wash water at a laboratory scale (Gómez-López et al., 2014). Ultrasound can also be used to degrade chemical contaminants such as those present in municipal wastewater as the conventional microbiological processes used in treatment plants are not designed or capable to remove these contaminants (Sharma et al., 2012). The aim of this study was to evaluate the disinfection capacity of two commercially available water treatment technologies (chlorine tablets at 1–2 ppm and US at 745 J/L), each one applied to a different water reservoir used to irrigate commercial growing fields of green peppers and artichokes. Correlations between physicochemical and microbial indicators were determined.

5 km were selected as representative of irrigation practices generally used in south of Europe, particularly in the southeast of Spain, one of the main vegetable producer areas in Europe. The climate in this area is semi-arid Mediterranean, with a mean annual rainfall of 235 mm and a mean annual temperature of 20 °C. The water temperature ranged between 15.2 and 19.2 °C. Over a ten weeks period (from March to June, 2013) each one of the five reservoirs was constantly subjected to a different treatment: (i) Reservoir 1 (60 × 60 × 5 m3) was the control reservoir without any disinfection treatment (control); (ii) Reservoir 2 (60 × 20 × 5 m3) was chemically treated with commercial chlorine tablets (chlorine); (iii) Reservoir 3 (90 × 90 × 5 m3) was treated with US 20 kHz frequency treatment (US20); (iv) Reservoir 4 (45 × 45 × 5 m3) was treated with US 40 kHz (US40) and (v) Reservoir 5 (50 × 50 × 5 m3), that received wastewater from a wastewater treatment plant (50% of total volume), was treated with US 20 kHz (US20 WW). All of the rest, received water from the Tajo-Segura river transfer. All the water reservoirs undergo influx and outflux during 10 weeks, which slightly modified the characteristics of the water. Water reservoirs 1, 3, 4 and 5 included in this study were not subjected to any disinfection treatment before this test. Reservoir 2 was previously treated with copper sulfate (CuSO4) up to 1 month before this study started. The water reservoirs 1, 3 and 4 were used to irrigate organic green-peppers cultivated in greenhouses, while water reservoirs 2 and 5 irrigated artichokes grown in open fields. The US system consists of two main parts: an emitter, which is placed into the water reservoir, and the electronic equipment, which is placed outside the water reservoir and connected to power supply. The US equipment (TISU 300 W®, CONTARIEGO, Murcia, Spain) was able to emit frequencies between 20 and 60 kHz. In this study frequencies of 20 kHz (US20) (reservoirs 3 and 5) and 40 kHz (US40) (reservoir 4) were selected. The emitted signals were registered by a sonar surround hydrophone (TC 4013, Ambient Recording Company, Duivendrecht, Netherlands). The specific power or intensity (I) and the specific energy (Es) were calculated as following:

2. Materials and methods

Water samples (200 mL each) were taken once per week for a period of 10 weeks between March and June, 2013. Each sampling time, three specific sampling-points were established for each water reservoir to obtain representative samples. Samples were taken avoiding the top layer of the water at an approximately 25 cm depth by using the deep water sampler. Samples were transported to the lab within 30 min.

2.1. Water sources and treatments Five reservoirs belonging to five intensive farms all located in the southeast of Spain (Murcia) (37°44′N, 0°57 W) within a distance of

I¼

PotencyðWÞ ; VolumeðLÞ

  W Es ¼ I  timeð secÞ ¼ J=L: L

In this case, the Es applied to each water reservoir treated with US was 745 J/L. The signal strength marked out the number of US emitters needed for each reservoir. Thus, reservoirs 4 and 5 were treated with one US emitter while two emitters were placed in reservoir 2, due to the bigger dimension. Therefore, although the water reservoirs had different dimensions, the Es was the same in all of them. The US emitters were always located in the same position, in the extreme of the water reservoir. In case two US emitters were needed, they were placed on opposite sides of the water reservoir. The water influx and outflux of the water reservoirs helped to homogenize the water. Reservoir 2 was treated with commercial chlorine tablets (Chlorilong 250, Bayrol Ibérica, S.A.U., Barcelona, Spain) following the manufacture recommendations. The concentration of free chlorine in the water was established between 1 and 2 ppm. Higher chlorine concentrations were avoided as they might cause damage to the plants. To maintain the selected concentration, free chlorine was monitored in the water reservoir at least twice per week and maintained by manual addition. The water treatments were applied constantly for up to 10 weeks. 2.2. Sampling

142

M.V. Villanueva et al. / Food Research International 75 (2015) 140–147

Table 1 Initial physicochemical properties of irrigation water in the different water reservoirs including chemical oxygen demand (COD), alkalinity, turbidity and pH. Different letters within the same column are significantly different (p b 0.05). Water reservoir

COD (mg O2/L)

Alkalinity (meq HCO− 3 /L)

Turbidity (NTU)

pH

1 2 3 4 5

1720.0b 845.5c 1627.5b 1777.5b 2498.0a

9.0bc 11.2a 9.8ab 9.0bc 8.0c

12.1a 2.4c 13.1a 3.1c 7.3b

8.2c 8.8a 8.4abc 8.2bc 8.7ab

Samples for microbiological counts were processed the same day and the rest of the water was stored at −20 °C until physicochemical measures were carried out. Before freezing, thiosulfate (Panreac, Barcelona, Spain) (1:10) was added. The incorporation of thiosulfate as a neutralizer was carried out to control the treatment time.

hydrochloric acid 0.1 N (Panreac, Barcelona, Spain). Turbidity was tested using the Turbiquant 3000R turbidimeter (Merck, Madrid, Spain). 2.5. Statistical analysis Microbial loads were log-transformed and introduced in Excel spreadsheet (Microsoft Excel, 2010). For the mean calculations only samples with numbers above the detection limit were included. IBM SPSS statistics 19 was used for statistical analysis. Shapiro–Wilk test was performed to assess the normality of the data (P N 0.05). Mann Whitney U and Kruskal–Wallis tests were used to examine the difference between the indicators and to define differences between irrigation water treatments. Bivariate correlation analysis (Spearman's Rank) and linear regression analysis were also implemented. Confidence intervals for the correlation analysis and for the uncertainty range of the estimated regression parameters were established at 95%. 3. Results

2.3. Microbial parameters 3.1. Physicochemical characteristics Total coliforms, E. coli spp. and enterococci were used as hygiene indicator organisms. Microbial quality of irrigation water samples was analyzed by filtering 100 mL of water through sterile 0.45-μm-pore-size membrane filters (QA Life Sciences, Santiago, California, USA). The filters were placed on the plates with appropriate media. E. coli and total coliforms were analyzed according to ISO 9308–1:2000 and were isolated by using Chromocult agar (Oxoid, Basingstoke, Hampshire, UK) after incubation for 24 h at 37 °C. Enterococci were detected and enumerated using the membrane filtration method ISO 7899–2 (i.e., membrane filtration) and were isolated by using Slanetz–Bartley agar (Oxoid, Basingstoke, Hampshire, UK) after incubation for 44 h at 37 °C. The membrane was examined under a bright light and all red or brown colonies were counted as presumptive enterococci. Confirmation of potential positive colonies was performed by inoculating on a medium with azide Esculin Bili (AEB), after incubation for 2 h at 44 °C. Confirmed positive colonies were recognized as brown-black colonies. The level and type of green algae and cyanobacteria were also evaluated. The analysis was performed following the Sedgewick-Rafter counting method using the counting chamber S-R (Sedgewick-Rafter, Pyser-SGI Limited, UK). One milliliter of each water sample was placed in the countingchamber. Microalgae were counted under an optical microscope (Olympus CX31, OLYMPUS IBERIA, S.A.U. Barcelona, Spain). Five replicates were analyzed per sample. The quantitative analysis was calculated by the formula: N¼

C  At As  S

where N (total density: microalgae/mL), C (microalgae counted), At (total area: 1000 mm2), As (square area of the counting-chamber = 10 mm2) and S (number of squares of counting-chamber counted) (LeGresley & McDermott, 2010). The total count is expressed as 104 units/mL.

The dimensions of the water reservoirs varied from 6000 m3 up to 40,500 m3. The origin of the water in the water reservoirs was also variable. All these factors affected the initial physicochemical characteristics of the irrigation water stored in the different water reservoirs (Tables 1). The initial COD values were, in all the cases, high (N 800 mg O2/L), although significant differences were observed among them (Table 1). Water from the Reservoir 5 (supplied with water from the Tajo-Segura river transfer and treated wastewater) showed the highest COD value. On the other hand, Reservoir 2, which was treated before this study with CuSO4, showed the lowest COD value (Table 1). COD levels higher than 800 mg O2/L can be considered high based on the previously published recommended guidelines and legislation, where values above 500 mg/L have been defined between moderate to high (Mujeriego, 1990; USEPA, 2004). Previous reported COD values for irrigation water (ranged between 15 and 107 mg O2/L) were significantly lower than the COD values found in the present study (Gemmell & Schmidt, 2012; Ijabadeniyi, Debusho, Vanderlinde, & Buys, 2011). The average alkalinity value was 9.4 ± 1.2 meq HCO− 3 /L. Ayers and Westcot (1985) established that a moderate alkalinity ranges between 1.5 and − 8.5 meq HCO− 3 /L and above that limit (N 8.5 meq HCO3 /L) alkalinity can be defined as high. The use of irrigation water with high levels of alkalinity should be restricted mainly due to a potential risk of filter collapse (Ayers & Westcot, 1985). Significant differences were observed among the different water reservoirs for the alkalinity values. Reservoir 2 was the one showing the highest initial value (Table 1). Differences were also observed between the turbidity values of the different water reservoirs, with values close to or below to 10 Nephelometric Turbidity Units (NTU) (Table 1). This value (10 NTU) has been reported as the maximum allowed level for irrigation water that comes into direct contact with fresh produce (RD, 1620, 2007). The turbidity levels found in this study were lower than previously reported values of about 20 NTU for water use to

2.4. Physicochemical analysis Briefly, changes in levels of free and total chlorine (mg/L), pH, and COD (mg O2/L) were measured at different time intervals. The pH was measured using a pH & redox multimeter (Crison, Barcelona, Spain). Free chlorine was measured using a portable photometer HI 96701 (Hanna Instruments, CITY; COUNTRY) which is based on a colorimetric assay. Samples of free chlorine were taken at least at 5 different locations of the water reservoir avoiding to take the superficial water. Chemical oxygen demand (COD) was determined by the standard photometric method using the Spectroquant NOVA 60 photometer (APHA, 1998). Alkalinity (meq CaCO3/L) was determined by acid titration using

Table 2 Initial microbial characterization of irrigation water in the different water reservoirs including total coliforms, E. coli spp., enterococci and algae. Different letters within the same column are significantly different (p b 0.05). ns: no significant. Water Total coliforms E. coli spp. Enterococci Algae reservoir (log cfu/100 mL) (log cfu/100 mL) (log cfu/100 mL) (104 units/mL) 1 2 3 4 5

2.5 ± 0.3ns 2.2 ± 0.1ns 2.2 ± 0.1ns 2.7 ± 0.2ns 2.6 ± 0.2ns

0.6 ± 0.1ns 0.2 ± 0.2ns 0.3 ± 0.2ns 0.5 ± 0.3ns 0.6 ± 0.1ns

0.9 ± 0.1b 0.6 ± 0.1b 0.9 ± 0.1b 1.8 ± 0.0a 0.7 ± 0.4b

4.1 ± 0.5b 1.6 ± 0.2c 3.9 ± 0.5b 2.3 ± 0.1c 5.9 ± 0.1a

M.V. Villanueva et al. / Food Research International 75 (2015) 140–147

1500

143

A

B

20

a b 500

a

ns

10

0 0

Turbidity reduction (NTU)

COD reduction (mg / L)

1000

-500

Chlorine

US 20

US 40

Chlorine

Treatments

US 20

US 40

Treatments

Fig. 1. Boxplot representing reductions in chemical oxygen demand (COD, mg O2/L) (A) and turbidity (NTU) (B) between untreated and treated irrigation water samples. Water treatments include chlorine (Reservoir 2), US 20 (Reservoir 3) and US 40 (Reservoir 4). Bottom and top of the boxes represent the 25th and 75th percentile and the end of the whiskers express the minimum and maximum of all the data.

3.2. Microbiological characteristics

A

B 3

3

2 2 ns

ns 1

1 0 0

a

D

3

6 b

b

a

2

4

E. coli reduction (log cfu/100 ml)

C

4

1

ns

2

0

Enterococci reduction (log cfu/100 ml)

Total coliforms and E. coli loads of the tested water reservoirs did not show significant differences among them (Table 2). Average values of total coliforms and E. coli counts were 2.4 ± 0.2 and 0.4 ± 0.2 log cfu/

Algae reduction 10 unit/mL

Total coliforms reduction (log cfu/100 ml)

irrigate crops in South Africa (Ijabadeniyi et al., 2011). According to Ayers and Westcot (1985), the pH levels in irrigation water can range around 6.5–8.4. The average pH was 8.5 ± 0.3, although differences were observed among reservoirs, with values of 8.8 and 8.2 as maximum and minimum pH levels, respectively (Table 1).

0 Chlorine

US 20 Treatments

US 40

Chlorine

US 20

US 40

Treatments

Fig. 2. Boxplot representing reductions in total coliforms (log cfu/100 mL) (A), enterococci (log cfu/100 mL) (B), E. coli spp. (log cfu/100 mL) (C) and algae (104 unit/mL) (D) between untreated and treated irrigation water samples. Water treatments include chlorine (Reservoir 2), US 20 (Reservoir 3) and US 40 (Reservoir 4). Bottom and top of the boxes represent the 25th and 75th percentile and the end of the whiskers express the minimum and maximum of all the data.

M.V. Villanueva et al. / Food Research International 75 (2015) 140–147

3.3. Suitability of chlorine and US to disinfect irrigation water Each reservoir was treated with a different disinfection technology, including chemical disinfection with chlorine and US treatments (20 and 40 kHz) for 10 weeks. To evaluate the efficacy of the disinfection technologies, differences between untreated irrigation water from water reservoir 1 and irrigation water treated with each technology were compared (Figs. 1 and 2). All water treatments were able to reduce COD of irrigation water when compared to the control. This means that the water treatments were able to dissolve the organic matter present in the water reservoirs. COD reductions obtained using chlorine

(≥430 mg/L) and US 20 (~100 mg/L) were higher than those observed using US 40 (b50 mg/L) (Fig. 1A). These findings agree with previous data which reported a low COD reduction (≤5%) after 60 min US treatment in landfill leachate containing a high initial COD (4470 mg O2/L) (Wang, Wu, Wang, Li, & Tao, 2008). However, Gómez-López et al. (2014) reported the ineffectiveness of US to reduce COD levels, which could be due to the type of organic matter present in the water. Regarding turbidity, water treatments reduced the levels when compared to the control. However, no significant differences were observed in the water turbidity among treatments (Fig. 1B). Turbidity is caused by suspended substances or dissolved substances such as clay, silt, finely divided inorganic and organic matter, soluble colored organic compounds, plankton and other microscopic organisms (USEPA, 1999). Some studies already reported the capacity of US to reduce water turbidity, showing that among several treatments, the use of 28 kHz was the most effective treatment with 75% efficiency (Doosti, Kargar, & Sayadi, 2012). Regarding the microbial quality of irrigation water it was observed that all the tested water treatments were able to reduce the levels of microbial indicators. Total coliforms and enterococci counts were reduced by more than 1 log unit when compared to untreated water (Reservoir 1), without significant differences between them (Fig. 2A and B). Similarly, E. coli levels were significantly reduced by the disinfection treatments when compared to untreated water (Fig. 2C). No significant differences were found among the water treatments. Chlorine is the most common chemical sanitizer used to reduce microbial loads in process wash water and its efficacy and mode of action has been widely reported (Gil, Selma, López-Gálvez, & Allende, 2009). In the present study, the free chlorine concentrations (1–2 ppm) maintained during the present study confirmed the microbial inactivation already reported in

Total coliforms (log cfu/100 ml)

A 3

2

a

b

a

a a

a

a

a

3

a

2

b b

b b b

1

a b

a a

a 1

b

0

b b

D a

3

0

b b

C

US 20 US 20 WW

E. coli (log cfu/100 ml)

B

US 20 US 20 WW

a

6 a

2

4

a

b b

a a

1

b

a

b

b

a b b

0

b 2

4 6 8 Time (weeks)

10

Enterococci (log cfu/100 ml)

100 mL, respectively, without differences among reservoirs. In the present study, the average values for enterococci were 1.0 ± 0.5 log cfu/ 100 mL. Similar values were already reported for irrigation water (Holvoet, Sampers, Seynnaeve, & Uyttendaele, 2014). However, Reservoir 4 showed the highest enterococci level (around 2 log cfu/100 mL) which has been associated with an increase in the probability for the presence of enteric pathogens (Holvoet et al., 2014). Significant differences in the microalgae counts were also observed among water reservoirs. Microalgae levels are very relevant in irrigation systems because they can collapse different parts of them. According to Nakayama and Bucks (1991), there is a simple relationship between the level of microalgae and irrigation plugging, being low, medium and high (b 10,000, 10,000–50,000 and N 50,000 units/mL, respectively) the risk of filter collapse. All of them showed the microalgae levels considered to be within the medium range (Table 2). Only in the case of Reservoir 5, which was supplied with treated wastewater, the microalgae level was higher than 50,000 which can be classified as a high risk of filter collapse.

2

4 6 8 Time (weeks)

a a b

b

2

Algae (104 unit/mL)

144

0

10

Fig. 3. Mean values of total coliforms (log cfu/100 mL) (A), enterococci (log cfu/100 mL) (B), E. coli counts (log cfu/100 mL) (C) and algae (104 units/mL) (D) of irrigation water treated with a US treatment of 20 kHz from two water reservoirs: Reservoir 3 fed with surface water and Reservoir 5 fed with a 50% mix of surface water and treated wastewater. Significant differences between procedures were determined by Kruskal–Wallis and Mann–Whitney test and (P b 0.005) and represented with different letters.

M.V. Villanueva et al. / Food Research International 75 (2015) 140–147

Total coliforms (Log cfu/100ml)

A 3.0

2.0

1.0

0.0

B

E. coli (Log cfu/100ml)

1.5

1.0

0.5

0.0 0

1

2 3 4 5 Algae (104 units/mL)

6

Fig. 4. Scatter plot showing the relationship between total coliforms (log cfu/100 mL) and algae (104 units/mL) (A), and E. coli spp. (log cfu/100 mL) and algae (104 units/mL) (B). Central regression lines are shown.

Total coliforms (log cfu/100 ml)

4

3

2

1

0 0.0

0.2

0.4 0.6 0.8 E. coli (log cfu/100 ml)

1.0

1.2

Fig. 5. Scatter plot showing the relationship between total coliforms (log cfu/100 mL) and E. coli spp. (log cfu/100 mL). Central regression line is shown.

145

process wash water (Van Haute, Sampers, Holvoet, & Uyttendaele, 2013). Thus, the present study confirms the efficacy of low doses of free chlorine to reduce microbial loads while reducing the environmental impact of high chlorine doses. On the other hand, US at low frequencies (from 20 to 100 kHz) has been widely reported as an efficient water treatment to reduce microbial levels, mainly in wastewater and processed water (Mason, Riera, Vercet, & Lopez-Buesa, 2005; Piyasena, Mohareb, & Mckellar, 2003). Previously reported reductions in water using similar US treatments were much higher than those obtained in the present study (São José et al., 2014; Gómez-López et al., 2014). It should be taken into account that water reservoirs monitored in this study were used to irrigate commercial fields of fresh produce. These water reservoirs were subjected to regular agricultural practices, which include the addition of surface water to the water reservoirs whenever growers needed. This makes the water treatment in these reservoirs a dynamic process where water with new organic matter is constantly added to the water reservoirs. Therefore, the observed microbial reductions not only depended on the disinfection capacity of the water treatment but also on the addition of untreated surface water to the water reservoir. Although previous studies reported higher disinfection capacity, under our conditions, both chemical and chemicalfree treatments were able to control microbial loads in the monitored water reservoirs. However, it should be taken into account that the water sources were highly variable that made identifying the effect of ultrasonic problematic. Microalgae levels were affected by the disinfection treatments as both chlorine and US treatments reduced their concentration when compared to untreated water (Fig. 2D), therefore reducing the risk of filter collapse from medium (Reservoir 1) to low (treated Reservoirs) (Nakayama & Bucks, 1991). Previous studies already highlighted the potential of ultrasound to disrupt algae cells, although the efficacy of the applied treatments (20 kHz, power 500 W) depended on the type of microalgae (Ranjan, Patil, & Moholkar, 2010). Recently, Joyce, King, and Mason (2014) reported that the efficacy of a 20 kHz US treatment to disrupt microalgae cells was specie dependent. Cell disruption of Dunaliella salina and Chlamydomonas concordia was observed after 4 and 16 min, respectively, while under the same conditions, there was no visible disruption for Nannochloropsis oculata (Joyce et al., 2014). Based on these results, the types of algae present in the tested irrigation water seem to be relatively susceptible to the US treatments tested in this study. 3.4. Influence of the water quality in the efficacy of US treatment The impact of the water quality, particularly the organic matter content, on the disinfection capacity of 20 kHz US was tested in two different water reservoirs, which have been fed with two very different types of water. Reservoir 3 was fed with surface water while Reservoir 5 was fed with a 50% mix of surface water and treated wastewater. Water quality of Reservoirs 3 and 5 is shown in Tables 1 and 2. Although the initial differences on the physicochemical properties between irrigation water in Reservoirs 3 and 5, the initial microbial counts were similar except for the algae levels (Table 2). However, the efficiency of the applied ultrasound treatment (20 kHz for 10 weeks) was different in the two water reservoirs. A 1 log reduction in total coliforms was observed in Reservoir 3, fed with surface water, while the same treatment did not have any effect in Reservoir 5, fed with treated wastewater (Fig. 3A), highlighting the influence of the quality of the water on the inactivation efficacy of the US treatment. Similar tendency was observed in the rest of the tested microorganisms including enterococci, E. coli and microalgae. Therefore, in the present study, the quality of the irrigation water significantly influences the efficacy of the ultrasound treatment (Fig. 3B to D). However, previous studies reported that the quality of the water did not influence the microbial inactivation by US (Gómez-López et al., 2014; Madge & Jensen, 2002). The reason why different results were obtained could be due to the different nature of the water treated with the US

146

M.V. Villanueva et al. / Food Research International 75 (2015) 140–147

and also the fact that water reservoirs were filling with water during the duration of the study.

3.5. Correlations between indicator parameters Pearson correlation between coliforms and microalgae counts for all the samples (n = 200) was 0.709. Therefore, high algae counts were well correlated with high levels of coliforms (Fig. 4A). On the other hand, when the levels of E. coli and microalgae were correlated, a Pearson correlation of 0.768 was found (Fig. 4B). Both correlations were significant at P b 0.01 level, highlighting the strong relationship between the two variables. Additionally, correlation between microalgae levels and enterococci loads was also significant but with a lower Pearson correlation of 0.494 (data not shown). Based on the obtained results, it could be concluded that the probability of high levels of coliforms and E. coli counts was higher at higher levels of microalgae while Enterococcus were not very well correlated with the microalgae counts. Previous studies already found high concentration of both enterococci and E. coli in mats of macroalgae Cladophora glomerata along Lake Michigan shores (Whitman, Shively, Pawlik, Nevers, & Byappanahalli, 2003). These studies suggest that algae may be an important environmental source of bacteria and further studies should be carried out to establish the suitability of algae levels as a good indicator for the presence of foodborne pathogens in irrigation water stored in water reservoirs. No correlation was found between the physicochemical (COD, turbidity and alkalinity) and the microbial indicators (total coliforms, E. coli, enterococci and algae) (data not shown). When the relationship between coliforms and E. coli counts was evaluated a Pearson correlation of 0.750 was found (Fig. 5). Similar or even higher correlation between these indicators has been previously reported indicating that it is not necessary to enumerate all hygiene indicators as they are strongly correlated to each other (Holvoet et al., 2014; Wilkes et al., 2009).

4. Conclusions In the search of greener alternatives to reduce microbial risk of irrigation water, US treatments were evaluated as potential water treatments to reduce microbial loads including algae. In the present study it was observed that US treatments (20 kHz: US20 and 40 kHz: US40 at an Es of 745 J/L) were able to reduce microbial loads in irrigation water when compared to untreated water. No significant differences were observed between total coliforms and E. coli loads in irrigation water treated with US and the most commonly used chemical water treatment, chlorine. However, chlorine treatment achieved the máximum COD reduction in the irrigation water. The impact of the water quality on the disinfection capacity of 20 kHz US was tested in two different water reservoirs. Based on the obtained results, more studies should be done to clarify the impact of the organic matter on the microbial inactivation capacity of US. Correlations between physicochemical and microbial indicators were also determined. Microalgae levels were well correlated with total coliforms and E. coli counts. Although the levels of E. coli spp. present in the untreated irrigation water were within the local E. coli criteria (RD, 1620, 2007) or some standards included in GAP guidelines (≤100 cfu/100 mL), the obtained results suggest the use of US as a potential alternative to reduce microbial risks of irrigation water.

Acknowledgments This work was supported by MINECO (project AGL2013-48529-R) and from the EUFP7 Veg-i-Trade Grant agreement no 244994. Support provided by the COST ACTION FA1202 BacFoodNet is also appreciated.

References American Public Health Association (APHA) (1998). Standard methods for the examination of water and waste water (20th edition ). The American Water Works Association (AWWA) and the Water Environment Federation (WEF). Ayers, R.S., & Westcot, D.W. (1985). Water quality for agriculture. Irrigation and Drainage Paper. 29 Rev 1. (pp. 174). FAO (http://www.fao.org/DOCREP/003/T0234E/T0234E00. HTM. Last access Nov 24th 2014). Castro-Ibáñez, I., Gil, M.I., Tudela, J.A., Ivanek, R., & Allende, A. (2015). Assessment of microbial risk factors and impact of meteorological conditions during production of baby spinach in the Southeast of Spain. Food Microbiology, 49, 173–181. São José, J.F.B.D, Andrade, J.N., Ramos, A.M., Vanetti, D.M.C., Stringheta, P.C., & Chaves, P.J.B. (2014). Decontamination by ultrasound application in fresh fruits and vegetables. Food Control, 45, 36–50. Doosti, M.R., Kargar, R., & Sayadi, M.H. (2012). Water treatment using ultrasonic assistance: A review. Proceedings of the International Academy of Ecology and Environmental Sciences, 2, 96–110. EFSA Panel on Biological Hazards (BIOHAZ) (2014). Scientific Opinion on the risk posed by pathogens in food of non-animal origin. Part 2 (Salmonella and Norovirus in berries). EFSA Journal, 12(3706), 1–118 (http://www.efsa.europa.eu/en/efsajournal/ pub/3600.htm. Last access Nov 24th 2014). Food and Agriculture Organization of the United Nations (FAO), & Committee of Agriculture (2003). Development of a framework for good agricultural practices. Seventh session, COAG/2003/6 http://www.fao.org/docrep/meeting/006/y8704e.htm. Last access Nov 24th 2014. Food and Agriculture Organization of the United Nations/ World Health Organization (FAO/WHO) (2008). Microbiological hazards in fresh leafy vegetables and herbs. Meeting report. Microbial risk assessment series, 14. (pp. 1–138) (ftp://ftp.fao.org/ docrep/fao/011/i0452e/i0452e00.pdf. Last access Nov 24th 2014). Gemmell, M.E., & Schmidt, S. (2012). Microbiological assessment of river water used for the irrigation of fresh produce in a sub-urban community in Sobantu, South Africa. Food Research International, 47, 300–305. Gil, M.I., Selma, M.V., López-Gálvez, F., & Allende, A. (2009). Fresh-cut product sanitation and wash water disinfection: Problems and solutions. International Journal of Food Microbiology, 134, 37–45. Gómez-López, V.M., Gil, M.I., Allende, A., Blancke, L., Schouteten, L., & Selma, M.V. (2014). Disinfection capacity of high-power ultrasounds against E. coli O157:H7 in process water of fresh-cut industry. Food and Bioprocess Technology, 7, 3390–3397. Harris, L.J., Bender, J., Bihn, E.A., Blessington, T., Danyluk, M.D., Delaquis, P., et al. (2012). A framework for developing research protocols for evaluation of microbial hazards and controls during production that pertain to the quality of agricultural water contacting fresh produce that may be consumed raw. Journal of Food Protection, 75, 2251–2273. Holvoet, K., Sampers, I., Seynnaeve, M., & Uyttendaele, M. (2014). Relationships among hygiene indicators and enteric pathogens in irrigation water, soil and lettuce and the impact of climatic conditions on contamination in the lettuce primary production. International Journal of Food Microbiology, 171, 21–31. Ijabadeniyi, A.O., Debusho, L.K., Vanderlinde, M., & Buys, E.M. (2011). Irrigation water as a potential preharvest source of bacterial contamination of vegetables. Journal of Food Safety, 31, 452–461. Jones, L.A., Worobo, R.W., & Smart, C.D. (2014). UV light inactivation of human and plant pathogens in unfiltered surface irrigation water. Applied and Environmental Microbiology, 80, 849–854. Joyce, E.M., King, P.M., & Mason, T.J. (2014). The effect of ultrasound on the growth and viability of microalgae cells. Journal of Applied Phycology, 26, 1741–1748. LeGresley, M., & McDermott, G. (2010). Counting chamber methods for quantitative phytoplankton analysis — Haemocytometer, Palmer–Maloney cell and Sedgewick-Rafter cell. In B. Karlson, C. Cusack, & E. Bresnan (Eds.), Microscopic and molecular methods for quantitative phytoplankton analysis. IOC Manuals and Guides, no. 55, . Paris: UNESCO (IOC/2010/MG/55). Madge, B.A., & Jensen, J.N. (2002). Disinfection of wastewater using a 20 kHz ultrasound unit. Water Environment Research, 74, 159–169. Mason, T.J., & Peters, D. (2002). Practical sonochemistry: Power ultrasound uses and applications (Second ed.). Horwood Publishing, 12–21. Mason, J.M., Riera, E., Vercet, A., & Lopez-Buesa, P. (2005). Application of ultrasound. In D.W. Sun (Ed.), Emerging technologies for food processing (pp. 323–351). London: Elsevier. Mujeriego, R. (1990). Riego con agua residual municipal regenerada. Generalitat de Catalunya. Nakayama, F.S., & Bucks, D.A. (1991). Water quality in drip/trickle irrigation: A review. Irrigation Science, 12, 187–192. Norton-Brandão, D., Scherrenberg, S.M., & Jules, B. (2013). Reclamation of used urban waters for irrigation purposes — A review of treatment technologies. Journal of Environmental Management, 122, 85–98. Pachepsky, Y., Shelton, D.R., McLain, J.E.T., Patel, J., & Robert, E. (2011). Irrigation waters as a source of pathogenic microorganisms in produce: A review. Advances in Agronomy, 113, 73–138. Piyasena, P., Mohareb, E., & Mckellar, R.C. (2003). Inactivation of microbes using ultrasound: A review. International Journal of Food Microbiology, 87, 207–216. Ranjan, A., Patil, C., & Moholkar, V.S. (2010). Mechanistic assessment of microalgal lipid extraction. Industrial & Engineering Chemistry Research, 49, 2979–2985. RD 1620 (2007). Real Decreto por el que se establece el régimen jurídico de la reutilización de las aguas depuradas. http://www.boe.es/boe/dias/2007/12/08/pdfs/ A50639-50661.pdf (Last access Nov 24th 2014) Sharma, S.K., Sanghi, R., & Mudhoo, A. (2012). Green practices to save our precious “water resource”. In S.K. Sharma, & R. Sanghi (Eds.), Advances in Water Treatment and Pollution Prevention (pp. 1–36). London: Springer.

M.V. Villanueva et al. / Food Research International 75 (2015) 140–147 Steele, M., & Odumeru, J. (2004). Irrigation water as source of foodborne pathogens on fruit and vegetables. Journal of Food Protection, 67, 2839–2849. Suslow, T. (2013). It's time to “farmer-up”. GrowingProduce.com AVG section, 10j. (pp. 43–44), 43–44 (http://ucanr.edu/datastoreFiles/234-2605.pdf. Last access Nov 24th 2014). U.S. Environmental protection Agency (1999). Guidance manual for compliance with the interim enhanced surface water treatment rule: Turbidity provisions. Office of water. Document # EPA 815-R-99-010 USEPA (http://yosemite.epa.gov/water/ owrccatalog.nsf/1ffc8769fdecb48085256ad3006f39fa/ 484543f6ef92155a85256b0600723b47!OpenDocument. Last access Nov 24th 2014). U.S. Environmental protection Agency (2004). Guidelines for water reuse. EPA 645-R-04108Washington, D.C.: U.S. Environmental Protection Agency (Available on-line at www.epa.gov/ORD/NRMRL/pubs/625r04108/625r04108.pdf. Last access Nov 24th 2014). 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

147

for fresh-cut lettuce washing. Applied and Environmental Microbiology, 79, 2850–2861. Wang, S., Wu, X., Wang, Y., Li, Q., & Tao, M. (2008). Removal of organic matter and ammonia nitrogen from landfill leachate by ultrasound. Ultrasonics Sonochemistry, 15, 933–937. Whitman, R.L., Shively, D.A., Pawlik, H., Nevers, M.B., & Byappanahalli, M.N. (2003). Ocurrence of Escherichia coli and enterococci in Cladophora (Chlorophyta) in near shore water and beach sand of lake Michigan. Applied and Environmental Microbiology, 69, 4714–4719. Wilkes, G., Edge, T., Gannon, V., Jokinen, C., Lyautey, E., Medeiros, D., et al. (2009). Seasonal relationships among indicator bacteria, pathogenic bacteria, Cryptosporidium oocysts, Giardia cysts, and hydrological indices for surface waters within an agricultural landscape. Water Research, 43, 2209–2223.