Strawberry (Fragaria × ananassa Duch) and cherry tomato (Lycopersicon esculentum var. cerasiforme) produced by the Shenboleyuan farm in Hangzhou (Zhejiang, China) and red bayberry (Myrica rubra) produced by the Shenxianju garden in Xianju (Zhejiang, China) were selected for evaluation. A total of 3 kg of each type of produce was collected when the fruit maturity was approximately 80%, as indicated by a red surface color. For the microbial safety survey, fresh strawberry, cherry tomato, and red bayberry free of external defects were randomly purchased from local retail grocers, supermarkets, and fruit shops in Hangzhou (Zhejiang, China) on their day of their arrival from the producers. A total of 50 samples (1 kg each) of each fruit were collected and prepared for the comparison of microbiota before and after sanitizing treatment. The collected fruits were evaluated for microbial content in terms of total aerobic bacteria, E. coli, yeast, mold, and S. Typhimurium. The collected samples were transported immediately to the laboratory under refrigerated conditions. The samples were stored at 4°C in the laboratory, and the experimental treatments were completed within 24 h.

The E. coli O157:H7 (ATCC 43895) and S. Typhimurium (ATCC 14028) bacterial strains used in this study were preserved cultures originally obtained from the American Type Culture Collection (ATCC). The E. coli O157:H7 (E8) and S. Typhimurium (S9) strains were isolated from chickens in our laboratory. The strains of E. coli O157:H7 and S. Typhimurium were kept at -80°C in Luria-Bertani broth (LB, Difco) containing 20% (v/v) glycerol for long-term preservation and maintained on LB slants at 4°C for temporary use. For experimental preparation, a loop inoculum was transferred to LB broth and incubated for 24 h at 37°C for each food-borne microbial strain. Consecutive loop transfers of 24 h LB cultures were made, and cocktails of E. coli O157:H7 (ATCC 43895 and E8) and S. Typhimurium (ATCC 14028 and S9) were prepared from equal volumes of each strain suspension. The cocktail of each pathogen was used as the inoculum for the evaluation of each sanitizing method. The inoculum concentrations of E. coli O157:H7 and S. Typhimurium were standardized at 10⁷ cfu/mL by diluting with sterile 0.1% peptone water. The inoculum suspension of E. coli O157:H7 was applied to the strawberry, cherry tomato, and red bayberry samples, and the inoculum suspension of S. Typhimurium was applied to the strawberry and cherry tomato samples. Mesh bags containing the fresh fruits were submerged in the inoculum suspensions and gently agitated for 30 min. After inoculation, the produce samples were air-dried in a biosafety chamber (AIRTECH, Suzhou, China) for 30 min to facilitate bacterial adhesion before exposure to various disinfection treatments (Wang et al., 2013). All work involving biohazardous agents was performed under Biosefety Level 2 conditions.

The chemicals used to prepare the sanitizers were obtained from Sigma–Aldrich (Shanghai, China), and fresh sanitizing solutions were prepared with deionized water. The optimal sanitizing times and concentrations of the previously described sanitizers were applied to the three types of fresh produce in this study (Allende et al., 2009; Abadiasa et al., 2011; Sagong et al., 2011; Neal et al., 2012; Wang et al., 2013; Chen et al., 2016). The selected sanitizing solutions of 0.02% SH (200 mg/L free chlorine, pH 6.5), 0.1% SC (1.0 g/L, pH 9.4), 2% CA (20 g/L, pH 2.1), 2% LA (20 mL/L, pH 2.1), 2% MA (20 g/L, pH 3.3), and 0.1% ASC (containing 1.0 g/L SC, pH 2.3) were prepared according to previously reported methods (Allende et al., 2009; Abadiasa et al., 2011; Sagong et al., 2011; Neal et al., 2012; Wang et al., 2013; Chen et al., 2016). The sanitizing solutions were prepared immediately before use, and TW (pH 6.8) was used as the control. After the inoculation step (i.e., E. coli O157:H7 and S. Typhimurium), the sample contained in a mesh bag was dipped into the sanitizing solution for the amount of time specified in the relevant previous study (Table 1). Subsequently, the sample was transferred to a new sterile bag and rinsed with 100 mL sterile water. The mesh bag was gently massaged by hand for 1 min to remove chemical residues that might interfere with subsequent microbial analysis. To evaluate the sanitizing treatments against natural microbiota (i.e., aerobic bacteria, mold, and yeast), fresh produce samples were randomly selected, and the sanitizing treatments were performed without the artificial inoculation step. The microbiota populations were determined before and after the sanitizing treatments, and the D-values were calculated.

A 25-g portion of fresh produce was homogenized in 225 mL sterile physiological saline (8.5 g/L of NaCl) for 1 min using a blender (Stomacher 400, Seward, London, United Kingdom). Serial, 10-fold dilutions of each homogenate were prepared with sterile 0.1% peptone water, and 0.1 mL volumes of undiluted or suitable dilutions of the homogenate were spread on triplicate plates of the corresponding medium. Specifically, aerobic bacteria were incubated on plate count agar (Hope-Bio, China) at 30°C for 48 h; yeast and mold were incubated on potato dextrose agar (Hope-Bio, China) at 28°C for 48–72 h (Murray et al., 2007). E. coli O157:H7 and S. Typhimurium were determined by plating dilutions onto LB agar (Difco) or xylose lysine deoxycholate agar (Difco) at 37°C for 24 h. The culture purity of E. coli O157:H7 and S. Typhimurium was confirmed by Gram staining and biochemical tests (Smibert and Krieg, 1994). After incubation, colony counting was carried out using an automatic bacteria counter (Shineso Science and Technology Co., Ltd., China).

A total of 50 samples of each fruit type were subjected to the optimal sanitizing methods determined from the evaluations of the methods, and the effects of the optimal sanitizing methods on microbial contamination were evaluated. The strawberry and cherry tomato samples were evaluated for total aerobic bacteria, E. coli, yeast, mold, and S. Typhimurium, and the red bayberry samples were evaluated for total aerobic bacteria, E. coli, yeast, and mold.

All the experiments of this study were performed in triplicate (i.e., three independent tests) with each independent test containing three replica platings. The average plate count from each treatment was converted to log cfu/g. GraphPad Prism software was used for data analysis. The reduction in microbiota or pathogens was the D-value obtained for no treatment and for sanitizing treatment. Data are presented as means ± SD. Analyses were assessed by Student’s t-test, and p-values < 0.05 were considered statistically significant.

Washing the surfaces of the fruits with TW is the method most frequently used by consumers to remove soil and some microorganisms; however, the efficacy of this method is limited (Neal et al., 2012). In this study, washing with TW for 10 min (TW-10) significantly reduced the amounts of aerobic bacteria and S. Typhimurium on strawberries compared to the control (washing with TW for 2 min; TW-2), but it did not significantly reduce E. coli O157:H7, mold, or yeast (Figure 2). In cherry tomato, TW-10 treatment significantly reduced aerobic bacteria, E. coli O157:H7, and mold on cherry tomato (Figure 3) but did not produce significant reductions in any microbiota or pathogen on red bayberry (Figure 4). The reductions in microbiota and pathogens resulting from TW-10 treatment were lower than 1.5 (0.15–1.44) log cfu/g, consistent with the results of Neal et al. (2012) (0.7 log cfu/g), Keskinen and Annous (2011) (0.77 log cfu/g), and others (Abadiasa et al., 2011; Wang et al., 2013). These results indicate that TW treatment is not sufficient for sanitizing strawberry, cherry tomato, and red bayberry. Generally, the microbial reductions achieved by washing with water were higher for strawberry and cherry tomato than for red bayberry; this may be attributed to the glossy surfaces of strawberries and cherry tomatoes (Velázquez et al., 2009).

Organic acids (LA, CA, and MA) that are recognized as GRAS agents (Chen et al., 2016) were evaluated for their ability to sanitize fresh produce. For strawberry, treatment with LA solution significantly reduced all microbiota and pathogens; treatment with CA solution significantly reduced aerobic bacteria, E. coli O157:H7, yeast, and S. Typhimurium but not mold; and treatment with MA solution significantly reduced aerobic bacteria, E. coli O157:H7, and S. Typhimurium but not mold and yeast (Figure 2). For cherry tomato, the three organic acids produced significant reductions in the tested microbiota and pathogens (Figure 3). For red bayberry, the three organic acids produced significant reductions in E. coli O157:H7; although treatment with LA and MA significantly reduced aerobic bacteria, the aerobic bacteria reduction achieved by CA was significantly lower than the reduction observed for the control treatment (TW-2). LA was the only organic acid effective at reducing yeast, and no tested organic acid effectively reduced mold on red bayberry (Figure 4).

With the exceptions of mold on red bayberry and E. coli O157:H7 on strawberry, LA produced the largest reductions in microbiota and pathogens among the three organic acids. LA and CA resulted in the greatest reductions in S. Typhimurium (5.71 log cfu/g; S. Typhimurium was completely removed) on cherry tomato among the organic acid treatments. Moreover, LA produced the largest reductions in E. coli O157:H7 on red bayberry (4.31 log cfu/g) and S. Typhimurium on strawberry (6.95 log cfu/g) among all the sanitizing treatments in this study (Table 2). Previous studies reported that treatment with 2% LA reduced E. coli O157:H7 by 1.74 log cfu/g and S. Typhimurium by 1.73 log cfu/g on fresh organic lettuce (Sagong et al., 2011). Treatment with 2% LA at 55°C was reported to reduce E. coli O157:H7 and Salmonella spp. on fresh spinach leaves by 2.7 log cfu/g and 2.3 log cfu/g, respectively (Neal et al., 2012). Wang et al. (2013) found that treatment with 2% LA reduced E. coli O157:H7, S. Typhimurium, and L. monocytogenes by approximately 3.0 log cfu/g on minimally processed lotus sprouts, and the sanitizing efficiency of different treatments followed the approximate order of 2% LA > 1% LA > 0.5% LA > SH. In this study, the inoculated E. coli O157:H7 and S. Typhimurium were efficiently removed (2.39–6.95 log cfu/g) on the three fresh fruits by sanitizing with 2% LA, and the results were more efficient than those of on other previous investigations (Sagong et al., 2011; Neal et al., 2012; Wang et al., 2013).

Compared to TW-2, sanitizing with SH resulted in significant reductions in aerobic bacteria, E. coli O157:H7, mold, and S. Typhimurium on strawberry (Figure 2). SH produced significant reductions in all microbiota and pathogens on cherry tomato (Figure 3) and significantly reduced aerobic bacteria and E. coli O157:H7 on red bayberry (Figure 4). However, yeast on strawberry along with yeast and mold on red bayberry were not significantly removed by sanitizing with SH (p > 0.05).

Sanitizing with SC resulted in significant reductions in aerobic bacteria, E. coli O157:H7, yeast, and S. Typhimurium on strawberry (Figure 2); significant reductions in all microbiota and pathogens on cherry tomato (Figure 3); and significant reductions in E. coli O157:H7 on red bayberry (Figure 4). However, the ability of SC to reduce aerobic bacteria was lower than that of the control treatment (TW-2) for red bayberry (Figure 4). Mold on strawberry along with yeast and mold on red bayberry were not significantly reduced by sanitizing with SC (p > 0.05). Comparatively, all microorganisms (i.e., microbiota and pathogens) on strawberry, cherry tomato, and red bayberry were significantly removed by treatment with 0.1% ASC. Thus, in this study, the efficiency of the different treatments decreased in the following approximate order: 0.1% ASC > 0.02% SH > 0.1% SC. The antimicrobial activity of ASC was much higher than those of SC and SH for the surface sterilization of strawberry and cherry tomato (Figures 2, 3).

Acidified sodium chlorite produced the largest reductions in microbiota and pathogens among the chloric sanitizing methods. ASC and SC generated the largest reductions (6.95 log cfu/g) in S. Typhimurium on strawberry among all tested sanitization methods. Moreover, ASC also produced the greatest reductions in aerobic bacteria (4.01 log cfu/g) on strawberry and yeast (3.04 log cfu/g) and mold (1.52 log cfu/g) on cherry tomato (Table 2). The inoculated E. coli O157:H7 and S. Typhimurium on strawberry and cherry tomato were completely inactivated by 0.1% ASC. Previous studies showed similar results as those obtained for red bayberry in this study. For example, 0.02% SH reduced microbiota and pathogens by 1.0–1.3 log cfu/g on fresh-cut cilantro, reduced pathogens by 1.0–1.5 log cfu/g on fresh spinach leaves, and reduced pathogens by approximately 0.5–1.0 log cfu/g on fresh-cut apple (Allende et al., 2009; Abadiasa et al., 2011; Neal et al., 2012). However, the reductions obtained on red bayberry in this study were clearly lower than those obtained for strawberry and cherry tomato in this study and the reductions reported in the past studies mentioned above. This discrepancy suggests that the nature of the fruit surface can strongly influence the sanitizing effect (Velázquez et al., 2009). For fresh-cut cilantro, treatment with 0.025 and 0.05% ASC reduced microbial populations by approximately 2 log cfu/g, while the reduction achieved by 0.1% ASC was more than 3 log cfu/g (Allende et al., 2009). Although 0.1% SC was as effective as 0.1% ASC in reducing aerobic bacteria and E. coli O157:H7, SC was not as effective as ASC in reducing yeast and mold populations on cut cilantro (Allende et al., 2009). In this study, the ability of ASC to reduce aerobic bacteria, E. coli O157:H7 and yeast on red bayberry was similar to that of LA; however, only ASC significantly reduced the mold population. In general, 0.1% ASC and 2% LA were the most effective treatments for the three types of produce in this study, and treatment with 0.1% ASC was optimal because of the shorter sanitizing time and lower SC content.

A total of 50 samples of each fruit type from different retail grocers were subjected to treatment with 0.1% ASC, and the natural microbiota were evaluated before and after treatment. With the exception of E. coli on red bayberry, significant reductions were obtained for all microbiota and pathogens. The average microbial reductions on strawberry were 2.14 log cfu/g for aerobic bacteria, 0.40 log cfu/g for E. coli, 1.33 log cfu/g for mold, and 1.27 log cfu/g for yeast (Figure 5). The average microbial reductions on cherry tomato were 2.63 log cfu/g for aerobic bacteria, 1.53 log cfu/g for E. coli, 1.55 log cfu/g for mold, and 2.04 log cfu/g for yeast (Figure 6). The average microbial reductions on red bayberry were 0.68 log cfu/g for aerobic bacteria, 0.13 log cfu/g for E. coli, 0.39 log cfu/g for mold, and 0.30 log cfu/g for yeast (Figure 7). The reductions in natural microbiota were much higher for cherry tomato and strawberry than for red bayberry. The ASC application results on the three produce types from retail grocers were consistent with the sanitizing methods comparisons on different fresh fruits in this study. The sanitization of red bayberry with ASC solution is made difficult by the rough surface of red bayberry fruit. Although the reductions in E. coli obtained on the studied fruits were less than 1.55 log cfu/g, this does not indicate low sanitizing efficiency. E. coli contamination on the fresh produce samples was efficiently inactivated by ASC treatment, similar to the results obtained in the screening of sanitizing methods. The low reductions in E. coli on the retail grocers samples can be attributed to low contamination levels. S. Typhimurium contamination was not observed on any of the samples.

As indicated by López-Velasco et al. (2012) and other studies, fresh produce (including tomato) is usually washed by chlorinated water [e.g., 50–200 mg L^-1 of active chlorine or active ClO₂ (>3 mg/L)] to inactivate Salmonella spp. and other microbial contaminants. However, chlorine residue is limited in foods because of its toxic by-products (Richardson et al., 1998; Gil et al., 2016). In this study, chlorinated antimicrobial agents (i.e., NaClO and SC) were selected as the conventional sanitizing agents according to the EFSA Opinions on Salmonella and Norovirus [EFSA Panel on Biological Hazards (BIOHAZ), 2014]. GRAS organic acids (i.e., CA, LA, and MA) and ASC were selected as new methods for comparison. In addition to Salmonella spp., the effects of the traditional and new treatment methods were evaluated for pathogenic E. coli O157:H7, yeast, mold, and aerobic bacteria on strawberry, cherry tomato, and red bayberry. The results indicate that treatment with 0.1% ASC had the best sterilizing effect on these fruits; ASC treatment also required the shortest sanitizing time and low chlorite content. This method was further confirmed by examining 50 samples of each produce type.