The desire for local, fresh produce year round is driving the growth of hydroponic growing systems in the United States. Many food crops, such as leafy greens and culinary herbs, grown within hydroponics systems have their root systems submerged in recirculating nutrient-dense fertilizer solutions from planting through harvest. If a foodborne pathogen were introduced into this water system, the risk of contamination to the entire crop would be high. Hence, this study was designed to determine whether Escherichia coli O157:H7, non-O157 Shiga toxin–producing E. coli, and Salmonella were able to survive and reproduce in two common hydroponic fertilizer solutions and in water or whether the bacteria would be killed or suppressed by the fertilizer solutions. All the pathogens grew by 1 to 6 log CFU/ml over a 24-h period, depending on the solution. E. coli O157:H7 reached higher levels in the fertilizer solution with plants (3.12 log CFU/ml), whereas non-O157 Shiga toxin–producing E. coli and Salmonella reached higher levels in the fertilizer solution without plants (1.36 to 3.77 log CFU/ml). The foodborne pathogens evaluated here survived for 24 h in the fertilizer solution, and populations grew more rapidly in these solutions than in plain water. Therefore, human pathogens entering the fertilizer solution tanks in hydroponic systems would be expected to rapidly propagate and spread throughout the system and potentially contaminate the entire crop.
The production of food crops in greenhouses and other controlled environment agriculture facilities has been experiencing rapid growth in the United States. Rabobank (15) reported that the value of U.S. greenhouse-grown food crops exceeded $3 billion in 2013 and is expected to exceed $4 billion by 2020. Between 2009 to 2012, the number of farms growing greenhouse vegetables (excluding tomato) and herbs increased from 4,056 to 8,712 and those growing fruits and berries increased from 244 to 655 (23). Growers of fresh produce in greenhouses may utilize various production systems such as a nutrient film technique, deep flow technique, Dutch buckets, and various types of gutter systems. These systems use a liquid fertilizer solution that is continually or periodically supplied to the crop from storage tanks. In most cases, the fertilizer solution is collected and reused (recirculating closed system), but in some situations the solution is allowed to drain after application and is discarded (drain-to-waste system).
Irrigation water can be a source of both Salmonella and Escherichia coli (1, 3, 9, 22, 26). When contaminated water is used to make fertilizer solutions, pathogens can be introduced into hydroponic systems used for the production of fresh produce and can in turn contaminate the fresh produce (5, 6, 10, 11). The bacteria can then persist on or in the contaminated plants. E. coli can enter the plants through the roots (2) or the leaves (5). Salmonella has been observed to enter the plant roots through natural openings or wounds (7), or pathogens can form biofilms on the plants' roots (3). As a result of the potential for water to introduce human pathogens into the production environment, the U.S. Food and Drug Administration, Food Safety Modernization Act final rule for produce (25) requires that agricultural water that comes into direct contact with produce must be free of E. coli, Salmonella, and Listeria monocytogenes.
Controlled environment fruit and vegetable production has been perceived as having fewer potential food safety concerns than field-grown produce because the produce does not come in contact with natural field soils and is grown in a controlled environment more isolated from animals. However, human pathogens can still be introduced into greenhouse production systems from various sources, including water, substrates, and human activity (12, 19–21). Because hydroponic systems used in greenhouses most often use recirculated fertilizer solutions, human pathogens introduced can easily and rapidly spread throughout the crop. In June 2014, a multistate foodborne outbreak of Salmonella Saintpaul infection was linked to greenhouse-grown cucumbers from Culiacan, Mexico that caused 84 infections and 17 hospitalizations in the United States. Like the majority of foodborne outbreaks, poor sanitation and failure to follow good agricultural and manufacturing practices precipitated this outbreak (24).
Although human pathogens can survive in water, it is unknown whether they survive in hydroponic fertilizer solutions. The objective of this research was to determine the survival and growth rate over 24 h of E. coli O157:H7, non-O157 Shiga toxin–producing E. coli (STEC), and Salmonella in water, a typical fertilizer solution without plants, and a typical fertilizer solution with plants.
MATERIALS AND METHODS
Bacterial culture preparation. The following cultures were used for inoculations: E. coli O157:H7 (ATCC 35150, 43895, and 43890); STEC O26:H11, O45:H2, O103:H2, O111:H2, O121:H19, and O145NM (STEC Center Database, Michigan State University, East Lansing, MI); Salmonella Typhimurium (ATCC 14028; SA 3250, Salmonella Genetic Stock Center, Calgary, Alberta, Canada); and Salmonella Enteritidis (ATCC 13076). Each strain was grown separately in brain heart infusion (BHI) broth (HiMedia, Mumbai, India) for 24 h at 35°C in 10-ml test tubes with two additional transfers into fresh BHI broth for 24 h at 35°C. For the final transfer into BHI broth, individual cultures of each pathogen type were combined into separate bottles of E. coli O157:H7, non-O157:H7 STEC, or Salmonella. Each pathogen cocktail was centrifuged at 4,000 rpm for 10 min, and the supernatant was removed. Cells were resuspended in 0.85% saline solution, and the centrifugation and rinsing process was repeated two more times. The viable count of each bacterial suspension was determined by plating on plate count agar (HiMedia).
Solution preparation. Three treatment solutions were utilized in this study: distilled water (water control), new fertilizer solution without plants, and used fertilizer solution with plants. The water control solution was prepared by autoclaving distilled water and adjusting the pH to 6.0 with sulfuric acid. The new fertilizer without plants and the used fertilizer with plants were prepared by mixing the fertilizer formulation (Table 1) with autoclaved distilled water. Fertilizer nutrients were those commonly used within the hydroponics industry. The fertilizer with plants and the fertilizer without plants were not significantly different and were monitored at the start of the experiment (Table 1). Pathogen culture (1 ml) was added to 1 gal (3.8 liters) of the solution to reach a final level 103 CFU/ml. The electrical conductivity of these solutions was 1.6 dS/m−1, and the pH was 6.0. The fertilizer solution with plants was produced by using the solution in a deep flow technique system with growing basil (Ocimum basilicum) plants for 4 weeks. The fertilizer solution was adjusted daily with fertilizer concentrate and water to maintain the desired electrical conductivity. Sulfuric acid was used to maintain the pH at 6.0. After 4 weeks in the deep flow technique system, the fertilizer solution was removed from the system tank, and the pathogens were introduced. Each pathogen was added separately to each solution to a final level of 1,000 CFU/ml. For the fertilizer with and without plants, samples were taken to ensure that the water and fertilizer source were free of all pathogens. All water samples were free from foodborne pathogens. Fertilizer solutions were monitored daily to maintain the desired pH and electrical conductivity. All solutions were standardized to a temperature of 21°C. Separate experiments were conducted for each of the three pathogen types.
Plating and bacterial counts. Samples were taken at 0, 1, 2, 3, 6, 9, 12, 15, 18, 21, and 24 h after inoculation of the solutions. Each sample was homogenized, diluted in 0.1% Bacto peptone water (Difco, BD, Le Pont de Clair, France) and then directly plated on MacConkey agar with sorbitol (SMAC; Hardy Diagnostics, Santa Maria, CA) for E. coli O157:H7 samples Possé agar for the non-O157:H7 STEC samples (14), or bismuth sulfite (BS) agar (Hardy Diagnostics) for Salmonella samples. The plates were incubated at 35°C, and colonies were counted after 24 h. E. coli O157:H7 forms clear colonies with no fermentation of the sorbitol on SMAC. Non-O157:H7 STEC strains on the Possé agar appear as follows: O103, purple with midnight blue center; O145, teal; O111, between royal and midnight blue; O121, teal with light outer ring; O45, purple to midnight blue with metallic sheen; and O26, bluish purple. S. enterica colonies are black or greenish gray and may have a sheen with no zones or halos on BS agar.
Data collection and analysis. The data collected from the plate counts were entered into SAS 9.3 software (SAS Institute, Cary, NC) for analysis. This experiment included three complete replications with two subsamples per combination of pathogen plus water source. To evaluate specific responses to variables, mixed model methodologies were used to allow for a determination of the change in the populations of E. coli O157:H7, non-O157:H7 STEC, and Salmonella. Inoculations after 24 h in the three solutions (control water, fertilizer without plants, and fertilizer with plants) were analyzed for potential interactions. Significance was assessed at the 0.05 level.
RESULTS AND DISCUSSION
Hydroponic systems use recirculating fertilizer solutions that provide all of the essential macro- and micronutrients required for crop growth, including nitrogen, phosphorus, potassium, calcium, magnesium, sulfur, copper, boron, zinc, iron, and manganese (17). Similar to food crops, bacteria require a blend of nutrients, such as carbon, oxygen, nitrogen, hydrogen, phosphorus, sulfur, potassium, magnesium, calcium, and iron. These nutrients are typically taken from the environment. In the presence of a source with nutrients, such as a hydroponic fertilizer solution, bacteria such as E. coli and Salmonella can survive and grow to levels that could cause serious illness in humans.
Within our study design, the new (unused) fertilizer and used fertilizer with plants had many of these nutrients available, but the distilled water had limited nutrients. Although there was a general trend of increase of approximately 3 log CFU/ml for bacteria in all three solutions over the 24-h period, bacterial survival and solution differences were not consistent. The E. coli O157:H7 results did not show an interaction between treatment and sampling time, but the treatments were significantly different. The mean populations in the water control and the fertilizer without plants were not significantly different (4.38 and 5.51 log CFU/ml, respectively), but the fertilizer with plants had a significantly higher mean population (6.1 log CFU/ml; P = 0.0296) (Fig. 1).
For the non-O157 STEC (Table 2), no significant differences among solutions were found before 18 h, but at 18 h the fertilizer without plants had significantly higher mean populations (1.23 to 1.69 log CFU/ml; P = 0.0001) than did the fertilizer with plants, and at 21 h the fertilizer without plants had significantly higher mean populations (1.52 to 1.97 log CFU/ml; P < 0.05) than did the water and the fertilizer with plants. Non-O157 STEC growth was 3.38 log CFU/ml over 24 h in the control but was 3.77 log CFU/ml in the fertilizer with plants and 1.6 log CFU/ml in the fertilizer without plants.
Salmonella levels (Table 3) were not significantly different among the solutions before 18 h, but at 18 and 21 h the fertilizer without plants had significantly higher mean populations (1 to2 log CFU/ml; P < 0.05) than did the water and the fertilizer with plants. Salmonella growth was 2.69 log CFU/ml over 24 h in the control but was 3.84 log CFU/ml in the fertilizer with plants and 1.34 log CFU/ml in the fertilizer without plants.
Our study results are consistent with those of previous studies in suggesting that bacteria are able to adapt to their environment. E. coli O157:H7 and Salmonella can quickly adapt in natural environmental conditions and in the presence of background bacteria (4, 8, 9, 13, 18), which was the case in the hydroponic fertilizer solutions. In the present study, the used fertilizer with plants probably had many bacteria that the pathogens had to outcompete to survive. Although E. coli O157:H7 was able to outcompete the background flora, non-O157 STEC and Salmonella had higher growth rates in the water and in the fertilizer without plants than in the fertilizer with plants. The higher growth for both the non-O157 STEC and Salmonella may be due to the lack of competition within those systems, allowing those organisms to have access to all of the available nutrients within the system. When nutrients are not as readily available, growth will occur at a much lower rate (16). The new fertilizer solution without plants also had more inorganic nutrients needed for bacterial growth (13, 18), which may be why there was more growth in those systems compared with the other treatments. Another potential difference between the fertilizer with and without plants is that plant exudates could be impacting pathogen survival or death and/or promoting the growth of other microbes that competed with the pathogens. Over the 4 weeks, the fertilizer with plants may have had different ratios of mineral elements as a result of the uptake and release of ions from the plant roots, which can affect the survival and growth of pathogens.
Results from this study provide evidence that even over a short time period, E. coli O157:H7, non-O157 STEC, and Salmonella can survive and multiply. In the two fertilizer solution treatments, these bacteria survived and grew rapidly. Thus, if introduced into a hydroponic system, the potential for pathogen survival, rapidly growth, and spread is very high. In the fertilizer solution that had been used for a while and presumably had exudates from roots and other microbes introduced, the pathogens still grew better than in the plain water. In hydroponic systems with recirculating nutrient solutions, this potential for pathogen growth poses a major food safety concern. The water, once contaminated, would be in contact with the food product being grown over an extended period of time, which would increase the risk of a foodborne disease outbreak. During production and harvesting, growers should seek to minimize and completely eliminate contact between the hydroponic nutrient solution and the stems and/or leaves (i.e., saleable parts) of the plant. Further studies are needed to provide a more thorough understanding of pathogen growth and interactions within hydroponic production units over complete crop cycles and to identify food safety interventions that can be utilized in the presence of different crops without depleting quality and nutritional attributes.