Comparison of Two Possible Routes of Pathogen Contamination of Spinach Leaves in a Hydroponic Cultivation System

Transcription

1 1536 Journal of Food Protection, Vol. 74, No. 9, 2011, Pages doi: / x.jfp Copyright G, International Association for Food Protection Research Note Comparison of Two Possible Routes of Pathogen Contamination of Spinach Leaves in a Hydroponic Cultivation System SHIGENOBU KOSEKI,* YASUKO MIZUNO, AND KAZUTAKA YAMAMOTO National Food Research Institute, , Kannondai, Tsukuba, Ibaraki , Japan MS : Received 23 January 2011/Accepted 12 April 2011 ABSTRACT The route of pathogen contamination (from roots versus from leaves) of spinach leaves was investigated with a hydroponic cultivation system. Three major bacterial pathogens, Escherichia coli O157:H7, Salmonella, and Listeria monocytogenes, were inoculated into the hydroponic solution, in which the spinach was grown to give concentrations of 10 6 and 10 3 CFU/ml. In parallel, the pathogens were inoculated onto the growing leaf surface by pipetting, to give concentrations of 10 6 and 10 3 CFU per leaf. Although contamination was observed at a high rate through the root system by the higher inoculum (10 6 CFU) for all the pathogens tested, the contamination was rare when the lower inoculum (10 3 CFU) was applied. In contrast, contamination through the leaf occurred at a very low rate, even when the inoculum level was high. For all the pathogens tested in the present study, the probability of contamination was promoted through the roots and with higher inoculum levels. The probability of contamination was analyzed with logistic regression. The logistic regression model showed that the odds ratio of contamination from the roots versus from the leaves was 6.93, which suggested that the risk of contamination from the roots was 6.93 times higher than the risk of contamination from the leaves. In addition, the risk of contamination by L. monocytogenes was about 0.3 times that of Salmonella enterica subsp. enterica serovars Typhimurium and Enteritidis and E. coli O157:H7. The results of the present study indicate that the principal route of pathogen contamination of growing spinach leaves in a hydroponic system is from the plant s roots, rather than from leaf contamination itself. Outbreaks of foodborne illness associated with the consumption of leafy vegetables have been a critical issue worldwide (18, 22, 28). These outbreaks are caused by contamination of the edible portions of plant leaves with foodborne pathogens. The potential cause of preharvest contamination of the edible portion of leaves has been considered direct contact with the pathogen through contaminated irrigation water and/or soil (5, 6, 10, 16, 23) and uptake from the root system through the plant vasculature (3, 4, 9, 19, 20, 24, 27, 29). However, few comparative studies of these two contamination routes have been conducted. It is unclear as to which of these two routes is the principal source of contamination of edible aerial leaves. If the main route of contamination of leafy green vegetables were clarified, measures more effective and critical for controlling microbial contamination during production would be possible. Thus, the elucidation of the contamination route will play a key role in ensuring the safety of leafy vegetables for consumption. In the present study, we aimed to determine the primary route of pathogen contamination of the edible portion of leafy greens. To clarify the translocation of the bacterial pathogens from the roots to the aerial leaves, we used a hydroponic growing system in a small plastic pot for * Author for correspondence. Tel: z ; Fax: z ; koseki@affrc.go.jp. assessing directly and simply the uptake of bacterial pathogens. The contamination from the leaf itself was also investigated by inoculating bacterial pathogens directly onto the leaves. In addition, because the contamination by bacterial pathogens through both of the possible contamination routes would be affected by the contamination level (inoculum size) (4, 9, 24), the effect of inoculum size was also investigated. Finally, the effects of the contamination route (from roots versus from leaves), inoculum level, and type of pathogens on the probability of contamination of the leave were analyzed with logistic regression. MATERIALS AND METHODS Bacterial strains. Six strains of Escherichia coli O157:H7 (ATCC [human feces], ATCC [feces of patient with hemolytic uremic syndrome], ATCC [raw hamburger meat implicated in hemorrhagic colitis outbreak], ATCC [clinical isolate], ATCC [human feces], and ATCC BAA-460 [human feces, 1996, Sakai City Institute of Public Health, Japan]); five strains of Salmonella enterica subsp. enterica serovars Enteritidis (ATCC BAA-708 [human clinical specimen] and ATCC 4931 [human gastroenteritis]) and Typhimurium (ATCC [unknown; designations: St. 403, cys B mutant], ATCC [unknown; designation: TA 1535], and ATCC [unknown; designation: TA 1537]); and six strains of Listeria monocytogenes (ATCC [poultry], ATCC [sheep], ATCC [chicken], ATCC [spinal fluid of child with meningitis], ATCC [rabbit] and ATCC [guinea pig])

2 J. Food Prot., Vol. 74, No. 9 COMPARISON OF PATHOGEN CONTAMINATION ROUTE OF LEAFY GREENS 1537 FIGURE 1. Images of spinach plants used in the present study. (a) Spinach growing in a hydroponic solution, (b) inoculation of pathogens onto the growing leaves, and (c) inoculation of pathogens onto cut leaves. were used as inocula. (Detail information for all the strains is available at: tabid/112/default.aspx.) All strains were stored at 280uC in tryptic soy broth (TSB; Merck, Darmstadt, Germany) containing 10% glycerol. On the day before the experiment, a sterile, disposable plastic loop was used to transfer individually a small amount of the frozen bacterial cultures to 10 ml of TSB for E. coli O157:H7 and Salmonella in a glass tube. The frozen L. monocytogenes cultures were transferred to brain heat infusion broth (Merck). The cultures were incubated at 37uC for 24 h, without agitation. Cells of each bacterial strain were collected by centrifugation (2,000 g, 15 min, 20uC), and the resulting pellets were each resuspended in 5 ml of sterile distilled water. To make a single sample of each pathogen comprising every strain, equal volumes of cell suspensions of the five or six strains of each pathogen were combined to give approximately equal populations of each strain. Inoculation of spinach with the pathogens. Mature and intact spinach plants (Brassica rapa var. perviridis) that had been hydroponically grown for 10 weeks were used as test samples. The spinach plants were grown with the nutrient flow technique in a greenhouse maintained at 20uC. The hydroponic solution was prepared with commercial fertilizers (Otsuka House nos. 1 and 2, Otsuka AgriTechno Co., Ltd., Tokyo, Japan), according to the product instruction. The contents of the hydroponic solution used in this study were as follows: 224 ppm of NO32, 41 ppm of PO432, 312 ppm of Kz, 160 ppm of Ca2z, and 48 ppm of Mg2z. Whole plants (including leaves and roots) were removed from the nutrient flow technique system and then transplanted individually into plastic containers (118 mm in diameter by 151 mm tall, with a volume of 946 ml) with a styrene foam lid (25-mm thick) filled with the hydroponic solution (800 ml). The plants were used for experiments 2 days after transplanting, to give the plants a chance to stabilize (Fig. 1a). To examine the contamination from plant roots, 1 ml of the inoculum (108 and 105 CFU/ml) was put directly into the plastic container filled with hydroponic solution (800 ml) to give pathogen density levels of 106 and 103 CFU/ml. The plants growing in the plastic containers were held at 23uC and 50% relative humidity for 48 h in a safety cabinet with light and without airflow to avoid overdrying the leaf surfaces. Four spinach plants were used for each inoculum level of each pathogen, and the five to seven leaves of each plant were used for microbiological tests. Thus, 23 to 27 leaves were assessed in each inoculum level of each pathogen. Independently from the experiment mentioned above, to examine contamination of the plants from the leaf surface, the bacterial pathogen was separately inoculated onto each leaf of the growing spinach plants by pipetting 100 ml of inocula in 10 to 15 spots (Fig. 1b). Care was taken to avoid allowing the spotted inocula to run off. Inoculum levels of 103 and 106 CFU per leaf were achieved with serially diluted inocula. Although the numbers of inoculated leaves varied depending on the conditions of the plant, three to five per plant were inoculated with pathogen. The inoculated plants were held at 23uC and 50% relative humidity for 48 h in a safety cabinet with light and without airflow. Five spinach plants were used for each inoculum level of each pathogen, and the three to five leaves of each plant were used for microbiological tests. Therefore, 18 to 25 leaves were assessed in each inoculum level of each pathogen. Comparison of bacterial growth on leaves of growing plants versus on cut leaves. To examine the pathogen growth on leaves that were still attached to growing plants and on leaves cut from growing plants, the bacterial suspension was inoculated onto the leaves. For the leaves on growing plants, the bacterial pathogen was inoculated separately, as described in the previous section (103 CFU per leaf), onto three leaves of the seven to eight leaves. Three of the remaining leaves that were not inoculated with the pathogen were cut from the growing plant and then inoculated with the bacterial pathogens in the same manner to give pathogen density levels of 103 CFU per leaf (Fig. 1c). The inoculated plants and cut leaves were incubated at 23uC and 50% relative humidity for 24 h in a safety cabinet with light and without airflow to avoid overdrying the leaf surfaces. Three spinach plants were used for each pathogen. Thus, enumeration of bacterial number was conducted for a total of nine leaves in each leaf condition (growing and cutting) of each pathogen. Confirmation of the presence of bacterial population and its enumeration. The spinach leaves from the growing plant were removed individually with sterile scissors. Each spinach leaf (2 to 3 g) not including stem, as shown in Figure 1c, was combined with

3 1538 KOSEKI ET AL. J. Food Prot., Vol. 74, No ml of 0.1% peptone water (Merck) in a stomacher bag and pummeled for 2 min in the stomacher. As stated in the previous section, since the experiments were set up separately for each contamination route of leaf, sterilization of leaf surface to distinguish between the contamination of internal and surface was not conducted in the present study. Aliquots of the homogenates were serially diluted in 0.1% peptone water and plated in duplicate (0.1 ml each) on appropriate selective agar plates. The undiluted samples were directly plated in quadruplicate (0.25 ml) onto appropriate selective agar plates. E. coli O157:H7 was enumerated on sorbitol MacConkey agar (Merck) supplemented with cefixime-tellurite selective supplement (Merck) after incubation at 37uC for 24 h. The presence of E. coli O157:H7 was confirmed with a latex agglutination test (E.coli O157 Singlepath, Merck). Salmonella was enumerated on bismuth sulfate agar (Merck) after incubation at 37uC for 24 h. The presence of Salmonella was confirmed by testing reactions on triple sugar iron (Merck) slants. L. monocytogenes was enumerated on Oxford Listeria-selective agar supplemented with Oxford Listeria-selective supplement (Merck) after incubation for 24 h at 37uC. The presence of L. monocytogenes was confirmed with the Singlepath Listeria diagnosis kit. To confirm the pathogen presence that did not occur at detectable levels in the plate count experiments, 1.0-ml aliquots of the undiluted homogenates were added to 9 ml of TSB and incubated at 37uC for 24 h. The incubated mixture was streaked onto each type of selective agar plate and then incubated further at 37uC for 24 h. The presence or absence of colonies on each plate was scored as positive or negative. Estimation of the probability of contamination. Based on the assumption of a binomial distribution of the event positive or negative (1, 13), we assumed that the probability distribution of contamination of spinach leaves would also follow a binomial distribution. For each leaf (n ~ 268), the presence or absence of each pathogen was scored as 1 or 0, respectively. The data were fitted to a logistic regression model with R statistical software (version for Mac OS X; The model was of the form shown in equation 1: logit(p) ~ a 0 z a 1 site z a 2 pathogen z a 3 IC ð1þ where logit(p) is an abbreviation of ln [P/(1 2 P)], ln is the natural logarithm, P is the probability of contamination (in the range of 0 to 1), a i are coefficients to be estimated, site is the route of pathogen contamination (root or leaf), pathogen is the kind of bacteria (E. coli O157:H7, Salmonella, or L. monocytogenes), and IC is the inoculum level (log CFU per milliliter) of the pathogen applied to the roots or leaves. Here, the variables site and pathogen were treated as dummy variables. The receiver operating characteristic (ROC) curve, the Hosmer- Lemeshow goodness-of-fit statistic (13), and the le Cessie van Houwelingen goodness-of-fit statistic (12, 17) were used as measures of the goodness of fit of the model developed. All the measures were calculated with R software. The area under the ROC curve (AUC) is a measure of discrimination, obtained from a plot of sensitivity, i.e., the proportion of observed events that were correctly predicted to be events, versus specificity, i.e., the proportion of nonevents that were correctly predicted to be nonevents. The closer the value of AUC is to 1, the greater the discrimination. In epidemiological studies, an AUC value of 0.7 is considered acceptable discrimination, an AUC value of 0.8 is considered excellent discrimination, and an AUC value of 0.9 is considered outstanding discrimination (13). The Hosmer-Lemeshow goodnessof-fit statistic, which is determined by grouping objects into a contingency table and calculating a Pearson chi-square statistic, was proposed (13) as a means of estimating goodness of fit when there is no replication or insufficient replication in any of the subpopulations. Small values of the statistic (large P values) indicate a good fit of the model to the data. The le Cessie van Houwelingen test (17), another goodness-of-fit test for logistic regression, is a singledegree-of-freedom omnibus test of lack of fit for a binary logistic model that generally has more statistical power than the Hosmer- Lemeshow procedure (12). RESULTS Pathogen contamination of the growing spinach plants through the leaves or roots. Prior to the pathogen inoculation experiments, we confirmed that there were no pathogenic bacteria in 10 other plants harvested from the same nutrient flow technique system and in the hydroponic solution in the cultivation system. Although a high rate of contamination by all the pathogens tested was observed through the roots when a higher inoculum (10 6 CFU) was used, contamination by any of the pathogens was rare when a lower inoculum (10 3 CFU) was applied (Table 1). For those plants that were contaminated through the roots, the number of pathogens detected on the leaves attained a maximum of 4 log CFU per leaf. In contrast, there were few cases of contamination through the leaf surface, even when the inoculum level was high. In cases where contamination was confirmed, the number of pathogens was at most ca. 3.0 log CFU per leaf. Probability of contamination. The estimated parameters of the logistic regression model (equation 1) are shown in Table 2. The model developed showed good performance, as indicated by the high values of all the goodness-of-fit statistics shown in Table 2. The estimates indicated that the probability of contamination was promoted through the roots and at higher inoculum levels for all the pathogens tested in the present study. These results are illustrated as changes in the probability of contamination as a function of inoculum levels (Fig. 2). The odds ratio of contamination from the roots to the leaves was 6.93, which suggests that the risk of contamination from the roots was 6.93 times higher than the risk of contamination from the leaves. The probability of contamination by L. monocytogenes was lower than the probability of contamination by the other two bacterial pathogens. The odds ratio of kind of pathogen represented that the risk of contamination of L. monocytogenes was about 0.30 times that of Salmonella and E. coli O157:H7. Comparison of pathogen growth on the leaves. The pathogen growth on cut leaves versus leaves still attached to the growing plant was compared (Table 3). Although the levels of pathogens inoculated onto growing leaves decreased by a nondetectable level after 24 h of storage, the levels of pathogens inoculated onto cut leaves increased by 2 log CFU per leaf. This result was found for all of the pathogens tested in the present study. DISCUSSION Transmission of the pathogens to the growing leaves through the roots was confirmed in the hydroponic system

4 J. Food Prot., Vol. 74, No. 9 COMPARISON OF PATHOGEN CONTAMINATION ROUTE OF LEAFY GREENS 1539 TABLE 1. Summary of the detection of pathogens after inoculation of pathogens to spinach plants Organism(s) Site Inoculum level Range of detected viable count (log CFU/ml or leaf) a Recovery b (log CFU/leaf) c E. coli O157:H7 Root 6 18/27, /25,1.7 Leaf 6 4/20, /18 ND d Salmonella Root 6 20/25, /24,1.7 Leaf 6 3/20, /25,1.7 L. monocytogenes Root 6 12/24, /23,1.7 Leaf 6 1/19, /18 ND a The pathogens were inoculated into the hydroponic solution surrounding the root at 6 or 3 log CFU/ml, and inoculation onto growing leaves by spotting the inoculum to 6 or 3 log CFU per leaf. b Recovery was confirmed by enrichment in TSB. Results are shown as the number of positive leaves/total number of leaves examined. c Detection limit was 1.7 log CFU per leaf. d ND, not detected. used in this study. Although the higher pathogen concentration (.10 6 CFU/ml) contaminated growing leaves at a high rate, the lower concentration (,10 3 CFU/ml) did not transmit pathogens from roots to the growing leaves. This result was consistent with those of other reports (4, 9, 24). Although there have been some reports regarding the translocation of pathogens (3, 4, 9, 19, 20, 24, 27, 29), to the best of our knowledge, no probabilistic comparative analysis of pathogen contamination has been conducted. In the present study, we elucidated that the probability of the contamination of the growing leaves from the roots indicated a risk of contamination 6.93 times higher from the roots than from the leaf surfaces. Furthermore, the statistical analysis indicated that the contamination was dependent on the inoculum level, which represented the contamination level. Although pathogen concentrations.10 6 CFU/ml would not be present in a hydroponic solution in a real situation, the results of the present study showed low probability of contamination from the root system, even when the concentration of the pathogen was low. Thus, sanitary control of the solution directly in contact TABLE 2. Estimated parameters of logistic regression a Parameter Estimate SE Intercept Site Root Leaf Organism(s) E. coli L. monocytogenes Salmonella Inoculum concn a AUC, 0.875; Hosmer-Lemeshow test, 3.65 (df ~ 8) with P ~ 0.89; le Cessie van Houwelingen test, P ~ with the roots must play a key role in ensuring the safety of hydroponically grown leafy greens. In most of the previous studies, the translocation of pathogens through the root system was investigated in soilgrown plants. As Warriner et al. (29) mentioned, there might be a different response of pathogenic bacteria to the growing roots in the soil environment versus hydroponic systems. Because the roots of spinach in a hydroponic system would be more accessible to pathogenic bacteria than would be those roots in a soil environment, translocation of pathogens from the roots to the leaves might occur relatively easily in a hydroponic system. Thus, the results of the present study might be difficult to extrapolate to a soil environment. However, because the fundamental mechanism underpinning the internalization and translocation of pathogenic bacteria through the roots of leafy vegetables would be common regardless of the environment, a similar trend would be observed in hydroponic systems and soil environments. The results presented here are of importance to the expanding hydroponics industry, and suggest that root uptake of bacteria still must be considered for soil grown plants. Contamination from the leaf surfaces was not observed in most cases, regardless of the concentration of the inoculum (Table 1). This result was consistent with previous studies (2, 9). Although a different trend in the contamination of growing leaves was confirmed in some reports (5, 10, 16, 23), the experimental conditions in the present study differed from those conditions in the previous reports. Pathogen could survive on leaves after inoculation onto the leaves under high relative humidity in a closed experimental chamber (5, 10, 16, 23). In contrast, pathogens could not survive on the leaves under relatively low humidity (,55%), as in this study and in field studies (2, 9). This information suggests that the relative humidity in the environment of the growing plant plays an important role in the survival of the inoculated bacteria on the leaf

5 1540 KOSEKI ET AL. J. Food Prot., Vol. 74, No. 9 surface. However, maintaining a high relative humidity is not likely in a real field cultivation environment. If the plants were cultivated in a field, humans could not control weather conditions, and high relative humidity would not be expected to last throughout the cultivation period. Thus, in terms of relative humidity conditions, the results of the present study might be relevant to field cultivation conditions (2, 7, 9, 25). On the contrary, because the relative humidity in real hydroponic green house cultivation would be higher than the relative humidity of a field, such an environment would assist the survival of pathogens on leaves. However, pathogen contamination from leaf surface would rarely occur in a hydroponic closed environment. Although pathogens inoculated onto cut leaves survived and multiplied within 24 h at room temperature, pathogens inoculated onto the leaves of growing plants decreased to an undetectable level (Table 3). One possible reason for the death of pathogens on the growing leaves would be low relative humidity (50%) in the experimental environment, as mentioned above. However, the low relative humidity cannot explain the increase in the number of pathogens on the cut leaves that were held at the same humidity condition. Another possible explanation is that a plant defense system was responsible for the death of pathogens on the growing leaves of entire plants. Cutting the leaves might disable this defense system, so that the pathogens on cut leaves could survive and grow. Crosscontamination of leafy vegetables with pathogenic bacteria in the postharvest process could thus occur (26). In addition, the growth of pathogenic bacteria on cut produce has been reported (14, 15), and the growth in the present study was consistent with the findings of previous studies (14, 15). Thus, the results suggest that prevention of pathogen crosscontamination in the postharvest process, as well as control of contamination during the growing period, are important for pathogen control in leafy vegetables. A significantly lower contamination rate was observed for L. monocytogenes than for Salmonella and E. coli O157:H7. Although the reason for this result was unclear, it might be due to differences between gram-positive and gramnegative bacteria. The lower rates of contamination by L. monocytogenes are consistent with the fact that only two freshly cut produce related listeriosis outbreaks have been documented in the United States (11, 21). The low rate of contamination by L. monocytogenes might merit future study. Leaf age (5) and damage to the leaf (2, 8, 16) have been reported to affect pathogen contamination of leafy greens. These factors affect pathogen uptake through the roots and/ or pathogen persistence on the leaves. However, we used mature and intact leaves to focus on the comparison of contamination routes in the present study. The effects of leaf FIGURE 2. Changes in the probability of pathogen contamination of the edible portion of spinach leaves through roots (observed r [#], estimated [solid line]) or leaves (observed [N], estimated [dashed line]) in a hydroponic system. (a) Escherichia coli O157:H7, (b) Salmonella, and (c) Listeria monocytogenes.

6 J. Food Prot., Vol. 74, No. 9 COMPARISON OF PATHOGEN CONTAMINATION ROUTE OF LEAFY GREENS 1541 TABLE 3. Comparison of pathogen growth at 23uC for 24 h on growing and cut spinach leaves Viable cell count on leaf (log CFU/leaf) a Salmonella E. coli O157:H7 L. monocytogenes Growing leaf Cut leaf Growing leaf Cut leaf Growing leaf Cut leaf Initial h after ND b ND ND a Values are means standard deviations (n ~ 9). b ND, not detected; the detection limit was 1.7 log CFU per leaf. age and damage on the contamination of leaves with pathogens should be investigated in a future work. In summary, the results of the present study indicate that the principal route of pathogen contamination of growing leaves in a hydroponic system is from the plant roots, rather than from direct contamination of the leaves. In addition, the contamination of the leaves from direct contact with the pathogens would occur mainly because of crosscontamination during the postharvest process. Thus, appropriate sanitary control of the hydroponic solution during cultivation and careful handling during the postharvest process play the key roles in the prevention of pathogen contamination of edible leaves. ACKNOWLEDGMENT This work was supported by a grant (research project for ensuring food safety from farm to table, D1-7107) from the Ministry of Agriculture, Forestry, and Fisheries of Japan. REFERENCES 1. Agresti, A An introduction to categorical data analysis, 2nd edition. John Wiley and Sons, Inc., Hoboken, NJ. 2. Aruscavage, D., S. A. Miller, M. L. L. Ivey, K. Lee, and J. T. LeJeune Survival and dissemination of Escherichia coli O157:H7 on physically and biologically damaged lettuce plants. J. Food Prot. 71: Bernstein, N Evidence for internalization of Escherichia coli into the aerial parts of maize via the root system. J. Food Prot. 70: Bernstein, N., S. Sela, and S. Neder-Lavon Assessment of contamination potential of lettuce by Salmonella enterica serovar Newport added to the plant growing medium. J. Food Prot. 70: Brandl, M. T., and R. Amundson Leaf age as a risk factor in contamination of lettuce with Escherichia coli O157:H7 and Salmonella enterica. Appl. Environ. Microbiol. 74: Delaquis, P., S. Bach, and L. D. Dinu Behavior of Escherichia coli O157:H7 in leafy vegetables. J. Food Prot. 70: Dreux, N., C. Albagnac, F. Carlin, C. E. Morris, and C. Nguyen-The Fate of Listeria spp. on parsley leaves grown in laboratory and field cultures. J. Appl. Microbiol. 103: Erickson, M. C Preharvest internalization of Escherichia coli O157:H7 into lettuce leaves, as affected by insect and physical damage. J. Food Prot. 73: Erickson, M. C., C. C. Webb, J. C. Diaz-Perez, S. C. Phatak, J. J. Silvoy, L. Davey, A. S. Payton, J. Liao, L. Ma, and M. P. Doyle Surface and internalized Escherichia coli O157:H7 on fieldgrown spinach and lettuce treated with spray-contaminated irrigation water. J. Food Prot. 73: Harapas, D., R. Premier, B. Tomkins, P. Franz, and S. Ajlouni Persistence of Escherichia coli on injured vegetable plants. Int. J. Food Microbiol. 138: Ho, J. L., K. N. Shands, G. Friedland, P. Eckind, and D. W. Fraser An outbreak of type 4b Listeria monocytogenes infection involving patients from eight Boston hospitals. Arch. Intern. Med. 146: Hosmer, D. W., T. Hosmer, S. Le Cessie, and S. Lemeshow A comparison of goodness-of-fit tests for the logistic regression model. Stat. Med. 16: Hosmer, D. W., and S. Lemeshow Applied logistic regression, 2nd ed. Wiley and Sons, New York. 14. Koseki, S., and S. Isobe Growth of Listeria monocytogenes on iceberg lettuce and solid media. Int. J. Food Microbiol. 101: Koseki, S., and S. Isobe Prediction of pathogen growth on iceberg lettuce under real temperature history during distribution from farm to table. Int. J. Food Microbiol. 104: Lapidotand, A Transfer of Salmonella enterica serovar Typhimurium from contaminated irrigation water to parsley is dependent on curli and cellulose, the biofilm matrix components. J. Food Prot. 72: le Cessie, S., and J. C. van Houwelingen A goodness-of-fit test for binary regression models, based on smoothing methods. Biometrics 47: Lynch, M. F., R. V. Tauxe, and C. W. Hedberg The growing burden of foodborne outbreaks due to contaminated fresh produce: risks and opportunities. Epidemiol. Infect. 137: Mahbub, I Persistence of enterohemorrhagic Escherichia coli O157:H7 in soil and on leaf lettuce and parsley grown in fields treated with contaminated manure composts or irrigation water. J. Food Prot. 67: Mitra, R., E. Cuesta-Alonso, A. Wayadande, J. Talley, S. Gilliland, and J. Fletcher Effect of route of introduction and host cultivar on the colonization, internalization, and movement of the human pathogen Escherichia coli O157:H7 in spinach. J. Food Prot. 72: Schlech, W. F., P. M. Lavigne, R. A. Bortolussi, A. C. Allen, E. V. Haldane, A. J. Wort, A. W. Hightower, S. E. Johnson, S. H. King, E. S. Nicholls, and C. V. Broome Epidemic listeriosis evidence for transmission by food. N. Engl. J. Med. 308: Sivapalasingam, S., C. R. Friedman, L. Cohen, and R. V. Tauxe Fresh produce: a growing cause of outbreaks of foodborne illness in the United States, 1973 through J. Food Prot. 67: Solomon, E. B., H. J. Pang, and K. R. Matthews Persistence of Escherichia coli O157:H7 on lettuce plants following spray irrigation with contaminated water. J. Food Prot. 66: Solomon, E. B., S. Yaron, and K. R. Matthews Transmission of Escherichia coli O157:H7 from contaminated manure and irrigation water to lettuce plant tissue and its subsequent internalization. Appl. Environ. Microbiol. 68: Stine, S. W., I. Song, C. Y. Choi, and C. P. Gerba Effect of relative humidity on preharvest survival of bacterial and viral pathogens on the surface of cantaloupe, lettuce, and bell peppers. J. Food Prot. 68:

7 1542 KOSEKI ET AL. J. Food Prot., Vol. 74, No Wachtel, M. R., and A. O. Charkowski Cross-contamination of lettuce with Escherichia coli O157:H7. J. Food Prot. 65: Wachtel, M. R., L. C. Whitehand, and R. E. Mandrell Association of Escherichia coli O157:H7 with preharvest leaf lettuce upon exposure to contaminated irrigation water. J. Food Prot. 65: Warriner, K., A. Huber, A. Namvar, W. Fan, and K. Dunfield Recent advances in the microbial safety of fresh fruits and vegetables. Adv. Food Nutr. Res. 57: Warriner, K., F. Ibrahim, M. Dickinson, C. Wright, and W. M. Waites Interaction of Escherichia coli with growing salad spinach plants. J. Food Prot. 66: