602 Journal o f Food Protection, Vol. 78, No. 3, 2015, Pages 602-608 doi: 10.4315/0362-028X.JFP-14-095 C o pyright © , Internatio nal A sso cia tio n fo r Food P rotection
Research Note
Cross-Sectional Survey of Indicator and Pathogenic Bacteria on Vegetables Sold from Asian Vendors at Farmers’ Markets in Northern California FENGGUANG PAN,1 XUNDE LI,23* JEN N IFER CARABEZ,3 GUY RAGOSTA,3 KRISTINE L. FERNANDEZ,3 ELAINE WANG,3 ANYARAT THIPTARA,3 ELIZABETH ANTAKI,3 a n d EDWARD R. ATW ILL123* 'Department o f Food Science and Engineering, Jilin University, 5333 Xi’an Road, Changchun 130062, People’s Republic o f China; 2Western Institute for Food Safety and Security, 1477 Drew Avenue, Suite 101, Davis, California 95616, USA; and 3Department o f Population Health and Reproduction, 1089 Veterinary Medicine Drive, Davis, California 95616, USA MS 14-095: Received 25 February 2014/Accepted 14 November 2014
ABSTRACT A cross-sectional survey was conducted during summer 2013 to determine the occurrence of Escherichia coli, fecal coliforms (FCs), E. coli 0157:H7, and Salmonella on raw vegetable commodities common to Asian cuisine from 21 vendors or farmers at six farmers’ markets in northern California. Based on 242 samples from six commodities (basil, yardlong beans, bitter squash, okra, squash stems and leaves, cilantro), 100% of samples had detectable FCs and 20% had detectable E. coli. The mean concentrations were 0.67 log CFU/g and 1.26 log CFU per bundle for E. coli and 4.00 log CFU/g and 6.26 log CFU per bundle for FCs. Vegetables irrigated with ground versus surface water contained lower concentrations of FCs, but this difference was not observed for E. coli. Yardlong beans, bitter squash, and okra had lower levels of FCs compared with basil, cilantro, and squash stems and leaves. Sixteen (6.6%) samples had detectable levels of Salmonella serovars (Newport, Enteritidis, Agona, and Worthington), with the majority of positives found in cilantro and squash stems and leaves. There was a twofold higher probability of Salmonella contamination in samples from growers or vendors who stated that they used organic farming practices compared with samples from those using conventional farming practices. Lastly, the concentrations of FC and E. coli bacteria were significantly associated with Salmonella contamination: for each additional 100 CFU/g or bundle, the probability of Salmonella contamination increased by —15 and —30%, respectively. None of the samples had detectable E. coli 0157:H7.
In the United States, the number of fanners’ markets selling produce increased 79% from 1994 to 2002 (40), especially after the passage of the Farmer-to-Consumer Direct Marketing Act of 1976 (6). Consumers often perceive produce from farmers’ markets as fresher, locally grown (11), and potentially more traceable to the farm of origin than produce from larger retailers, such as brand-name supermar kets (41). Although many consumers self-report that they prefer the higher quality and value of farmers’ market produce (41) and health experts recommend regular con sumption of produce to reduce the risk of various chronic diseases (10,17,20), eating raw or minimally treated produce from farmers’ markets and other direct marketing operations remains a risk factor for foodbome pathogens (16). The Centers for Disease Control and Prevention Food bome Disease Outbreaks Surveillance reported that 13% of foodbome outbreaks in the United States between 1998 and 2008 were associated with leafy vegetables (14). The same survey also reported that norovirus, Salmonella, and Shiga toxin-producing Escherichia coli were common pathogens that caused produce-related outbreaks of disease in the United
States between 1998 and 2008 (14). Elsewhere, researchers have traced outbreaks of Salmonella to contaminated lettuce (12, 26) and other vegetables from Mexico (30), Turkey (3), Canada (2), and Brazil (33, 37). Although minimal baseline information exists for the prevalence of foodbome pathogens for specialty crops sold at farmers’ markets in some countries, including Brazil (21), Canada (5, 28), Nigeria (25,31), Togo (1), Turkey (36), and the United States (18, 39), many countries (including the United States) have not conducted detailed cross-sectional surveys of specialty crops grown on small to medium-sized farms and sold at local retailers, by direct marketing, and at urban and peri-urban farmers’ markets (15). Given that bacterial pathogens represent a common cause of produce-associated foodbome illness (4, 14, 24) and that little information exists on the microbiological quality of ethnic-specific commodities such as yardlong beans and bitter squash from direct retail operations, we conducted a crosssectional survey of bacterial pathogens and indicators for a variety of specialty commodities at farmers’ markets in northern California during summer 2013.
* Authors for correspondence. Tel: 530-754-2154; Fax: 530-752-7181; E-mail: [email protected] (X.L.), [email protected] (E.R.A.).
Sample collection. Vegetable commodities were selected to include those that are sometimes eaten raw and those that are
MATERIALS AND METHODS
603
BACTERIAL LEVELS ON FARMERS’ MARKET VEGETABLES
J. Food Prot., Vol. 78, No. 3
TABLE 1. Average EC and FC concentrations on vegetables grown using surface and groundwater irrigation sourcesa EC (log CFU/g)
EC (log CFU/bundle)
FC (log CFU/g)
FC (log CFU/bundle)
Groundwater Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil
0.23 0.94 0.38 0.38 1.13 0.75
(0.70) (1.23) (0.94) (0.90) (1.41) (1.16)
0.47 1.80 0.71 0.73 1.97 1.43
(1.45) (2.34) (1.73) (1.72) (2.46) (2.21)
3.73 4.02 3.99 3.91 4.18 3.98
(0.34) (0.32) (0.31) (0.32) (0.34) (0.37)
6.14 6.38 6.21 6.13 6.18 6.21
(0.31) (0.27) (0.27) (0.29) (0.35) (0.32)
Surface water Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil Organic Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil
0.61 (1.16) NA 0.0 (0.0) 0.41 (0.97) NA 0.91 (1.58)
1.15 (2.18) NA 0.0 (0.0) 0.79 (1.85) NA 1.63 (2.83)
3.86 (0.35) NA 4.28 (0.25) 4.01 (0.26) NA 4.28 (0.47)
6.23 (0.35) NA 6.46 (0.17) 6.31 (0.24) NA 6.51 (0.42)
0.48 1.15 0.19 0.40 0.76 0.93
(1.02) (1.23) (0.73) (0.93) (1.30) (1.29)
0.93 2.30 0.33 0.78 1.31 1.76
(1.96) (2.46) (1.27) (1.80) (2.25) (2.44)
3.83 3.98 3.98 3.89 4.08 4.08
(0.35) (0.19) (0.26) (0.37) (0.33) (0.25)
6.24 6.26 6.18 6.18 6.10 6.25
(0.34) (0.19) (0.26) (0.34) (0.33) (0.22)
0.22 0.86 0.49 0.35 2.41 0.69
(0.69) (1.25) (1.03) (0.87) (0.29) (1.16)
0.46 1.61 0.93 0.66 4.60 1.31
(1.43) (2.33) (1.94) (1.65) (0.00) (2.18)
3.77 4.03 4.08 3.96 4.45 3.98
(0.32) (0.35) (0.34) (0.28) (0.12) (0.44)
6.17 6.42 6.32 6.18 6.64 6.24
(0.30) (0.29) (0.30) (0.27) (0.17) (0.38)
0.31 0.94 0.41 0.36 1.13 0.86
(0.82) (1.23) (0.96) (0.88) (1.34) (1.22)
0.63 1.80 0.77 0.68 2.05 1.63
(L64) (2.34) (1.79) (1.67) (2.43) (2.29)
3.82 4.02 4.05 3.94 4.16 4.02
(0.33) (0.32) (0.31) (0.31) (0.33) (0.37)
6.22 6.38 6.28 6.18 6.22 6.25
(0.31) (0.27) (0.28) (0.29) (0.37) (0.32)
Conventional Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil Overall Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil
“ Values are means (standard deviations). EC, E. coli; FC, fecal coliform; NA, not applicable.
common to Asian cuisine and are sold at local farmers’ markets in San Francisco, Stockton, and Sacramento. After visits to several northern California fanners’ markets, the final vegetable selection included basil (Ocimum basilicum), okra (Abelmoschus esculentus), yardlong beans (Vigna unguiculata subsp. sesquipedalis), cilantro (Coriandrum sativum), squash stems and leaves (Cucurbita sp.), and bitter squash (Momordica charantia). We visited each farmers’ market once between 1 July and 30 September 2013 and purchased 242 vegetable samples (100 to 200 g per bundle) of the six commodities from 21 different vendors at six different farmers’ markets in northern California. The vendor placed each sample directly into a plastic bag, which we maintained at 4°C during transport and storage prior to testing. At the time of purchase, we asked the fanner or the vendor to indicate the source of water (surface or groundwater) used to irrigate their produce fields and whether they classified their growing operation as organic or conventional. Market location was the only identifying information we collected, to maintain anonymity of the farmers and/or vendors whose produce was used for this survey. Because we did not record the specific identity of each vendor or fanner, we
could not conduct further investigations into survey results to determine whether the vendors or farmers grew the produce using certified organic methods. Immunomagnetic separation of Salmonella and E. coli 0157:H7. Samples were weighed (100 to 200 g), placed into a Whirl-Pak bag (Nasco, Fort Atkinson, WI) containing 200 ml of tryptic soy broth (TSB; BD, Sparks, MD), and gently shaken by hand for 1 min. Samples were enriched and bacteria were isolated using previously described methods for E. coli 0157:H7 (8) and Salmonella (13). Following enrichment, Salmonella and E. coli 0157:H7 were isolated via immunomagnetic separation, using a Bead Retriever according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA). These analytical methods were able to detect 8 to 9 CFU of E. coli 0157 and Salmonella Enteritidis, respectively, when these concentrations were spiked with labora tory-grown strains onto 200 g of the selected commodities. Enumeration of indicator bacteria on vegetables. E. coli and fecal coliforms (FCs) were enumerated using conventional
604
PAN ET AL.
J. Food Prot., Vol. 78, No. 3
TABLE 2. The association between type o f vegetable, source o f
water fo r irrigation, and the concentration o f fecal coliforms for vegetables sold at northern California farmers’ markets during summer 2013 Factor
r
95% Cl
CR*
P
TABLE 4. The association between type o f vegetable and the concentration o f indicator E. coli fo r vegetables sold at northern California farmers’ markets during summer 2013 Factor
P"
95% Cl
CR*
P
value*
value' Vegetable
Vegetable Yardlong beans'* Basil Bitter squash Okra Squash stem/ leaves Cilantro
0.0 0.65 0.31 0.59
(0.10, 1.20) (0.05, 0.57) (0.04, 1.14)
1.0 1.92 1.36 1.80
0.59 1.01
(0.19, 0.98) (0.26, 1.77)
1.80 2.75
—
—
0.02 0.02 0.04
Research Note
Cross-Sectional Survey of Indicator and Pathogenic Bacteria on Vegetables Sold from Asian Vendors at Farmers’ Markets in Northern California FENGGUANG PAN,1 XUNDE LI,23* JEN N IFER CARABEZ,3 GUY RAGOSTA,3 KRISTINE L. FERNANDEZ,3 ELAINE WANG,3 ANYARAT THIPTARA,3 ELIZABETH ANTAKI,3 a n d EDWARD R. ATW ILL123* 'Department o f Food Science and Engineering, Jilin University, 5333 Xi’an Road, Changchun 130062, People’s Republic o f China; 2Western Institute for Food Safety and Security, 1477 Drew Avenue, Suite 101, Davis, California 95616, USA; and 3Department o f Population Health and Reproduction, 1089 Veterinary Medicine Drive, Davis, California 95616, USA MS 14-095: Received 25 February 2014/Accepted 14 November 2014
ABSTRACT A cross-sectional survey was conducted during summer 2013 to determine the occurrence of Escherichia coli, fecal coliforms (FCs), E. coli 0157:H7, and Salmonella on raw vegetable commodities common to Asian cuisine from 21 vendors or farmers at six farmers’ markets in northern California. Based on 242 samples from six commodities (basil, yardlong beans, bitter squash, okra, squash stems and leaves, cilantro), 100% of samples had detectable FCs and 20% had detectable E. coli. The mean concentrations were 0.67 log CFU/g and 1.26 log CFU per bundle for E. coli and 4.00 log CFU/g and 6.26 log CFU per bundle for FCs. Vegetables irrigated with ground versus surface water contained lower concentrations of FCs, but this difference was not observed for E. coli. Yardlong beans, bitter squash, and okra had lower levels of FCs compared with basil, cilantro, and squash stems and leaves. Sixteen (6.6%) samples had detectable levels of Salmonella serovars (Newport, Enteritidis, Agona, and Worthington), with the majority of positives found in cilantro and squash stems and leaves. There was a twofold higher probability of Salmonella contamination in samples from growers or vendors who stated that they used organic farming practices compared with samples from those using conventional farming practices. Lastly, the concentrations of FC and E. coli bacteria were significantly associated with Salmonella contamination: for each additional 100 CFU/g or bundle, the probability of Salmonella contamination increased by —15 and —30%, respectively. None of the samples had detectable E. coli 0157:H7.
In the United States, the number of fanners’ markets selling produce increased 79% from 1994 to 2002 (40), especially after the passage of the Farmer-to-Consumer Direct Marketing Act of 1976 (6). Consumers often perceive produce from farmers’ markets as fresher, locally grown (11), and potentially more traceable to the farm of origin than produce from larger retailers, such as brand-name supermar kets (41). Although many consumers self-report that they prefer the higher quality and value of farmers’ market produce (41) and health experts recommend regular con sumption of produce to reduce the risk of various chronic diseases (10,17,20), eating raw or minimally treated produce from farmers’ markets and other direct marketing operations remains a risk factor for foodbome pathogens (16). The Centers for Disease Control and Prevention Food bome Disease Outbreaks Surveillance reported that 13% of foodbome outbreaks in the United States between 1998 and 2008 were associated with leafy vegetables (14). The same survey also reported that norovirus, Salmonella, and Shiga toxin-producing Escherichia coli were common pathogens that caused produce-related outbreaks of disease in the United
States between 1998 and 2008 (14). Elsewhere, researchers have traced outbreaks of Salmonella to contaminated lettuce (12, 26) and other vegetables from Mexico (30), Turkey (3), Canada (2), and Brazil (33, 37). Although minimal baseline information exists for the prevalence of foodbome pathogens for specialty crops sold at farmers’ markets in some countries, including Brazil (21), Canada (5, 28), Nigeria (25,31), Togo (1), Turkey (36), and the United States (18, 39), many countries (including the United States) have not conducted detailed cross-sectional surveys of specialty crops grown on small to medium-sized farms and sold at local retailers, by direct marketing, and at urban and peri-urban farmers’ markets (15). Given that bacterial pathogens represent a common cause of produce-associated foodbome illness (4, 14, 24) and that little information exists on the microbiological quality of ethnic-specific commodities such as yardlong beans and bitter squash from direct retail operations, we conducted a crosssectional survey of bacterial pathogens and indicators for a variety of specialty commodities at farmers’ markets in northern California during summer 2013.
* Authors for correspondence. Tel: 530-754-2154; Fax: 530-752-7181; E-mail: [email protected] (X.L.), [email protected] (E.R.A.).
Sample collection. Vegetable commodities were selected to include those that are sometimes eaten raw and those that are
MATERIALS AND METHODS
603
BACTERIAL LEVELS ON FARMERS’ MARKET VEGETABLES
J. Food Prot., Vol. 78, No. 3
TABLE 1. Average EC and FC concentrations on vegetables grown using surface and groundwater irrigation sourcesa EC (log CFU/g)
EC (log CFU/bundle)
FC (log CFU/g)
FC (log CFU/bundle)
Groundwater Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil
0.23 0.94 0.38 0.38 1.13 0.75
(0.70) (1.23) (0.94) (0.90) (1.41) (1.16)
0.47 1.80 0.71 0.73 1.97 1.43
(1.45) (2.34) (1.73) (1.72) (2.46) (2.21)
3.73 4.02 3.99 3.91 4.18 3.98
(0.34) (0.32) (0.31) (0.32) (0.34) (0.37)
6.14 6.38 6.21 6.13 6.18 6.21
(0.31) (0.27) (0.27) (0.29) (0.35) (0.32)
Surface water Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil Organic Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil
0.61 (1.16) NA 0.0 (0.0) 0.41 (0.97) NA 0.91 (1.58)
1.15 (2.18) NA 0.0 (0.0) 0.79 (1.85) NA 1.63 (2.83)
3.86 (0.35) NA 4.28 (0.25) 4.01 (0.26) NA 4.28 (0.47)
6.23 (0.35) NA 6.46 (0.17) 6.31 (0.24) NA 6.51 (0.42)
0.48 1.15 0.19 0.40 0.76 0.93
(1.02) (1.23) (0.73) (0.93) (1.30) (1.29)
0.93 2.30 0.33 0.78 1.31 1.76
(1.96) (2.46) (1.27) (1.80) (2.25) (2.44)
3.83 3.98 3.98 3.89 4.08 4.08
(0.35) (0.19) (0.26) (0.37) (0.33) (0.25)
6.24 6.26 6.18 6.18 6.10 6.25
(0.34) (0.19) (0.26) (0.34) (0.33) (0.22)
0.22 0.86 0.49 0.35 2.41 0.69
(0.69) (1.25) (1.03) (0.87) (0.29) (1.16)
0.46 1.61 0.93 0.66 4.60 1.31
(1.43) (2.33) (1.94) (1.65) (0.00) (2.18)
3.77 4.03 4.08 3.96 4.45 3.98
(0.32) (0.35) (0.34) (0.28) (0.12) (0.44)
6.17 6.42 6.32 6.18 6.64 6.24
(0.30) (0.29) (0.30) (0.27) (0.17) (0.38)
0.31 0.94 0.41 0.36 1.13 0.86
(0.82) (1.23) (0.96) (0.88) (1.34) (1.22)
0.63 1.80 0.77 0.68 2.05 1.63
(L64) (2.34) (1.79) (1.67) (2.43) (2.29)
3.82 4.02 4.05 3.94 4.16 4.02
(0.33) (0.32) (0.31) (0.31) (0.33) (0.37)
6.22 6.38 6.28 6.18 6.22 6.25
(0.31) (0.27) (0.28) (0.29) (0.37) (0.32)
Conventional Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil Overall Yardlong beans Squash stem/leaves Okra Bitter squash Cilantro Basil
“ Values are means (standard deviations). EC, E. coli; FC, fecal coliform; NA, not applicable.
common to Asian cuisine and are sold at local farmers’ markets in San Francisco, Stockton, and Sacramento. After visits to several northern California fanners’ markets, the final vegetable selection included basil (Ocimum basilicum), okra (Abelmoschus esculentus), yardlong beans (Vigna unguiculata subsp. sesquipedalis), cilantro (Coriandrum sativum), squash stems and leaves (Cucurbita sp.), and bitter squash (Momordica charantia). We visited each farmers’ market once between 1 July and 30 September 2013 and purchased 242 vegetable samples (100 to 200 g per bundle) of the six commodities from 21 different vendors at six different farmers’ markets in northern California. The vendor placed each sample directly into a plastic bag, which we maintained at 4°C during transport and storage prior to testing. At the time of purchase, we asked the fanner or the vendor to indicate the source of water (surface or groundwater) used to irrigate their produce fields and whether they classified their growing operation as organic or conventional. Market location was the only identifying information we collected, to maintain anonymity of the farmers and/or vendors whose produce was used for this survey. Because we did not record the specific identity of each vendor or fanner, we
could not conduct further investigations into survey results to determine whether the vendors or farmers grew the produce using certified organic methods. Immunomagnetic separation of Salmonella and E. coli 0157:H7. Samples were weighed (100 to 200 g), placed into a Whirl-Pak bag (Nasco, Fort Atkinson, WI) containing 200 ml of tryptic soy broth (TSB; BD, Sparks, MD), and gently shaken by hand for 1 min. Samples were enriched and bacteria were isolated using previously described methods for E. coli 0157:H7 (8) and Salmonella (13). Following enrichment, Salmonella and E. coli 0157:H7 were isolated via immunomagnetic separation, using a Bead Retriever according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA). These analytical methods were able to detect 8 to 9 CFU of E. coli 0157 and Salmonella Enteritidis, respectively, when these concentrations were spiked with labora tory-grown strains onto 200 g of the selected commodities. Enumeration of indicator bacteria on vegetables. E. coli and fecal coliforms (FCs) were enumerated using conventional
604
PAN ET AL.
J. Food Prot., Vol. 78, No. 3
TABLE 2. The association between type o f vegetable, source o f
water fo r irrigation, and the concentration o f fecal coliforms for vegetables sold at northern California farmers’ markets during summer 2013 Factor
r
95% Cl
CR*
P
TABLE 4. The association between type o f vegetable and the concentration o f indicator E. coli fo r vegetables sold at northern California farmers’ markets during summer 2013 Factor
P"
95% Cl
CR*
P
value*
value' Vegetable
Vegetable Yardlong beans'* Basil Bitter squash Okra Squash stem/ leaves Cilantro
0.0 0.65 0.31 0.59
(0.10, 1.20) (0.05, 0.57) (0.04, 1.14)
1.0 1.92 1.36 1.80
0.59 1.01
(0.19, 0.98) (0.26, 1.77)
1.80 2.75
—
—
0.02 0.02 0.04
