Salmonella and Listeria monocytogenes survival on Field Packed Cantaloupe Contact Surfaces

Field-packing of cantaloupes involves numerous food contact surfaces that can contamination melons with foodborne pathogens; the soil on these surfaces increases throughout the harvest day. Data is lacking on the cross-contamination risk from contaminated food contact surfaces under the dry conditions typical of cantaloupe field-packing operations. This study sought to evaluate the survival of Salmonella and Listeria monocytogenes on cantaloupe field-pack food contact surfaces using both a wet and dry inoculum to provide insights into managing foodborne pathogen contamination risks. Five clean or fouled materials (cotton gloves, nitrile gloves, rubber gloves, cotton rags, and stainless steel) were inoculated with a cocktail of either Salmonella or L. monocytogenes. A wet inoculum was spot inoculated (100 µL) onto coupons. A dry inoculum was prepared by mixing wet inoculum with 100 g of sterile sand, and shaking the coupons with the inoculated sand for 2min. Coupons were held at 35°C (35% RH) and enumerated at 0, 2, 4, 6 and 8 h. Significant differences in pathogen concentrations over time were calculated and the GInaFiT add-in tool for Excel was used to build Log-linear, Weibull, and Biphasic die-off models. Depending on the material type, coupon condition, and inoculum type, Salmonella and L. monocytogenes reductions over 8 h ranged from 0.3-3.3 and -0.4-4.2 log10 CFU/coupon, respectively. For all material types, Salmonella reductions were highest on wet-inoculated clean coupons; L. monocytogenes varied by material type. Weibull and biphasic models were a better fit of respective pathogen die-off curves than linear models. Overall, faster die-off rates were seen for wet inoculated and clean materials. Since pathogen populations remained viable over the study duration and both inoculum type and coupon condition impacted survival, frequent sanitation or replacement of food contact surfaces during the operational day is needed to reduce the risk of cross-contamination.

Southern Region Produce Safety Alliance Grower Training: Using Pre- and Post-Training Knowledge Assessments to Understand Training Effectiveness

The Produce Safety Alliance (PSA) grower training was introduced in 2016 as the standardized curriculum to meet the training requirements of the Food and Drug Administration's (FDA) Food Safety Modernization Act's (FSMA) Produce Safety Rule (PSR). The PSR states that at least one supervisor or responsible party from each farm must have successfully completed this food safety training or one equivalent to the standardized curriculum, as recognized by the FDA. This study evaluated the effectiveness of PSA trainings conducted between 2017 and 2019 in the Southern United States by the Southern Regional Center for Food Safety Training, Outreach, and Technical Assistance by analyzing pre- and posttest assessments. Effectiveness was based on a 25-question knowledge assessment administered to participants before (n = 2494) and after (n = 2460) each training. The knowledge assessment indicated the overall effectiveness of the training, with average scores increasing significantly from pretest (15.9/25, 63.4%) to posttest (20.3/25, 81.3%) (P < 0.001). The greatest knowledge gains were seen in the Postharvest Handling and Sanitation, How to Develop a Farm Food Safety Plan, and Agricultural Water modules. Notably, these modules had lower posttest scores compared to the other modules, indicating that the amount of knowledge gained did not necessarily correspond with a sufficient understanding of the material. To ensure that participants understand all aspects of the PSR and best practices to minimize food safety risks, additional or advanced trainings may be needed. Additionally, the current testing instrument (pre-/posttest) used for PSA grower training, while validated, may not be optimal, thus alternative methods to assess the training effectiveness are likely needed.

Factors Associated with the Prevalence of Salmonella, Generic Escherichia coli, and Coliforms in Florida's Agricultural Soils

Limited data exist on the environmental factors that impact pathogen prevalence in the soil. The prevalence of foodborne pathogens, Salmonella and Listeria monocytogenes, and the prevalence and concentration of generic E. coli in Florida's agricultural soils were evaluated to understand the potential risk of microbial contamination at the preharvest level. For all organisms but L. monocytogenes, a longitudinal field study was performed in three geographically distributed agricultural areas across Florida. At each location, 20 unique 5 by 5 m field sampling sites were selected, and soil was collected and evaluated for Salmonella presence (25 g) and E. coli and coliform concentrations (5 g). Complementary data collected from October 2021 to April 2022 included: weather; adjacent land use; soil properties, including macro- and micro-nutrients; and field management practices. The overall Salmonella and generic E. coli prevalence was 0.418% (1/239) and 11.3% (27/239), respectively; with mean E. coli concentrations in positive samples of 1.56 log CFU/g. Farm A had the highest prevalence of generic E. coli, 22.8% (18/79); followed by Farm B, 10% (8/80); and Farm C 1.25% (1/80). A significant relationship (p < 0.05) was observed between generic E. coli and coliforms, and farm and sampling trip. Variation in the prevalence of generic E. coli and changes in coliform concentrations between farms suggest environmental factors (e.g. soil properties) at the three farms were different. While Salmonella was only detected once, generic E. coli was detected in Florida soils throughout the duration of the growing season meaning activities that limit contact between soil and horticultural crops should continue to be emphasized. Samples collected during an independent sampling trip were evaluated for L. monocytogenes, which was not detected. The influence of local environmental factors on the prevalence of indicator organisms in the soil presents a unique challenge when evaluating the applicability of more global models to predict pathogen prevalence in preharvest produce environments.

Methodological differences between studies confound one-size-fits-all approaches to managing surface waterways for food and water safety

Even though differences in methodology (e.g., sample volume and detection method) have been shown to affect observed microbial water quality, multiple sampling and laboratory protocols continue to be used for water quality monitoring. Research is needed to determine how these differences impact the comparability of findings to generate best management practices and the ability to perform meta-analyses. This study addresses this knowledge gap by compiling and analyzing a data set representing 2,429,990 unique data points on at least one microbial water quality target (e.g., Salmonella presence and Escherichia coli concentration). Variance partitioning analysis was used to quantify the variance in likelihood of detecting each pathogenic target that was uniquely and jointly attributable to non-methodological versus methodological factors. The strength of the association between microbial water quality and select methodological and non-methodological factors was quantified using conditional forest and regression analysis. Fecal indicator bacteria concentrations were more strongly associated with non-methodological factors than methodological factors based on conditional forest analysis. Variance partitioning analysis could not disentangle non-methodological and methodological signals for pathogenic Escherichia coli, Salmonella, and Listeria. This suggests our current perceptions of foodborne pathogen ecology in water systems are confounded by methodological differences between studies. For example, 31% of total variance in likelihood of Salmonella detection was explained by methodological and/or non-methodological factors, 18% was jointly attributable to both methodological and non-methodological factors. Only 13% of total variance was uniquely attributable to non-methodological factors for Salmonella, highlighting the need for standardization of methods for microbiological water quality testing for comparison across studies.IMPORTANCEThe microbial ecology of water is already complex, without the added complications of methodological differences between studies. This study highlights the difficulty in comparing water quality data from projects that used different sampling or laboratory methods. These findings have direct implications for end users as there is no clear way to generalize findings in order to characterize broad-scale ecological phenomenon and develop science-based guidance. To best support development of risk assessments and guidance for monitoring and managing waters, data collection and methods need to be standardized across studies. A minimum set of data attributes that all studies should collect and report in a standardized way is needed. Given the diversity of methods used within applied and environmental microbiology, similar studies are needed for other microbiology subfields to ensure that guidance and policy are based on a robust interpretation of the literature.

Ranking Food Safety Priorities of the Fresh Produce Industry in the United States

A broad understanding of community member food safety priorities in the fresh produce supply chain does not currently exist. This information is essential to improve food safety knowledge and practices effectively and efficiently throughout the fresh produce industry; therefore, the goal of this study was to identify and rank community produce safety priorities in the United States. Survey questions were designed and approved by food safety experts for participants to rank 24 fresh produce safety priorities. The anonymous survey was distributed online via Qualtrics™ to fresh produce community members from November 2020 to May 2021. A score was calculated for each priority by summing weighted ranking scores across responses. Descriptive statistics and logistic regression were used to determine frequencies and distribution of response and identify factors (e.g., role in produce safety, size/location of organization/operation) that influenced rankings. A total of 281 respondents represented fourteen different roles in the fresh produce industry, with most identified as growers (39.5%). Produce operations were distributed across the U.S. and annual produce sales ranged from below $25,000 to over $5,000,000. Health and hygiene, training, postharvest sanitation, traceability, and harvest sanitation were ranked as the top five food safety priorities. These findings provide insight into community member priorities in fresh produce safety and can be used to inform intervention efforts, ranging from specialized training for produce growers and packers, industry-driven research projects, and gaps in risk communication strategies.

Fate and Growth Kinetics of Salmonella and Listeria monocytogenes on Mangoes During Storage

Imported mangoes have been linked to outbreaks of salmonellosis in the USA. The purpose of this research was to evaluate the persistence and growth kinetics of Salmonella and Listeria monocytogenes on the intact surface of whole 'Ataulfo', 'Kent', and 'Tommy Atkins' mangoes stored at three different temperatures. L. monocytogenes was also evaluated on fresh-cut 'Tommy Atkins' mangoes stored at 4, 12, 20 ± 2°C. Whole mangoes were spot inoculated with rifampicin-resistant pathogen cocktails (6 log CFU/mango) onto the midsection of whole fruit (n = 6). Fruit was stored at 12, 20, or 30 ± 2°C and sampled for up to 28 days. The specific growth rates derived from DMFit models as a function of time were used to develop secondary models. On 'Kent' mangoes, Salmonella had a population increase from 0.3 to 1.1 log CFU/mango with a linear growth rate of ∼0.004, 0.01, and 0.06 log CFU/mango/h at 12, 20, and 30°C, respectively. At 20 and 30°C, Salmonella growth rates were significantly higher than 12°C (P < 0.05). No clear Salmonella growth trend was observed; populations decreased up to 1.6 log CFU/mango on 'Tommy Atkins' and 'Ataulfo' at 12°C. Populations of L. monocytogenes on whole and fresh-cut mangoes declined regardless of temperature and storage period. Food safety during storage should be the top priority for fresh-cut tropical fruit processors.

The prevalence and concentration of Salmonella enterica in poultry litter in the southern United States

Poultry litter is applied to crop production land in the southern United States as a waste management strategy as it is a nitrogen-rich fertilizer and plentiful throughout the region. While litter is a known reservoir for human enteric pathogens including Salmonella enterica, little is known regarding pathogen prevalence, concentration, and common serotypes within the material. Litter from thirteen farms across four southern states was examined for Salmonella. Samples (n = 490) from six of the thirteen (46.2%) farms tested positive. Thirty-three samples out of 490 (6.7%) were Salmonella positive. Salmonella was ca. 95% less likely to be collected from stacked litter piles than from the poultry house floor or pasture, and every day increase in litter age reduced the likelihood of recovering Salmonella by 5.1%. When present, concentrations of Salmonella in contaminated poultry litter were variable, ranging from <0.45 to >280,000 MPN/g. The most prevalent serotypes found were Kentucky (45.5%), Kiambu (18.2%), and Michigan (12.1%). Salmonella Kentucky also had the greatest distribution and was found on 4 of the 6 (66.7%) positive farms. Results from this survey demonstrated that Salmonella prevalence and concentration in poultry litter is highly variable, and good agricultural practices are critical to safely use poultry litter as a soil amendment on fresh produce fields.

Application of Chitosan Microparticles against Human Norovirus

ABSTRACT

Human norovirus (HuNoV) is the leading cause of foodborne illness outbreaks and the second most common cause of waterborne infections in the United States. The goal of this research was to investigate the antiviral activity of chitosan microparticles (CMs) against HuNoV GII.4 Sydney and its cultivable surrogate Tulane virus (TuV) in suspensions mimicking fecally contaminated water. CMs were prepared by cross-linking chitosan molecules with sodium sulfate, and the antiviral activity of CMs was assessed with an infectivity assay on TuV and by quantitative reverse transcription PCR on TuV and HuNoV. A 3% CM suspension in phosphate-buffered saline (pH 7.2) bound to TuV particles but had a negligible impact on virus infectivity (P > 0.05). A 10-min contact time resulted in a 1.5-log reduction in genomic copies per mL of TuV and HuNoV in fecal suspensions (P < 0.05). Despite the negligible impact on viral infectivity, CMs can moderately bind to infectious virus particles and help purify environmental water by removing these particles. In this study, TuV was a suitable surrogate for HuNoV with similar log reductions in fecal suspension. These findings highlight the potential application of CM as a novel treatment to minimize the spread of waterborne viral pathogens.

HIGHLIGHTS

Big Data Impacting Dynamic Food Safety Risk Management in the Food Chain

Foodborne pathogens are a major contributor to foodborne illness worldwide. The adaptation of a more quantitative risk-based approach, with metrics such as Food safety Objectives (FSO) and Performance Objectives (PO) necessitates quantitative inputs from all stages of the food value chain. The potential exists for utilization of big data, generated through digital transformational technologies, as inputs to a dynamic risk management concept for food safety microbiology. The industrial revolution in Internet of Things (IoT) will leverage data inputs from precision agriculture, connected factories/logistics, precision healthcare, and precision food safety, to improve the dynamism of microbial risk management. Furthermore, interconnectivity of public health databases, social media, and e-commerce tools as well as technologies such as blockchain will enhance traceability for retrospective and real-time management of foodborne cases. Despite the enormous potential of data volume and velocity, some challenges remain, including data ownership, interoperability, and accessibility. This paper gives insight to the prospective use of big data for dynamic risk management from a microbiological safety perspective in the context of the International Commission on Microbiological Specifications for Foods (ICMSF) conceptual equation, and describes examples of how a dynamic risk management system (DRMS) could be used in real-time to identify hazards and control Shiga toxin-producing Escherichia coli risks related to leafy greens.

Assessment of the Effect of Thermotherapy on 'Candidatus Liberibacter asiaticus' Viability in Woody Tissue of Citrus via Graft-Based Assays and RNA Assays

In 2019, citrus production in Florida declined by more than 70%, mostly because of Huanglongbing (HLB), which is caused by the bacterium 'Candidatus Liberibacter asiaticus' (CLas). Thermotherapy for HLB-affected trees was proposed as a short-term management solution to maintain field productivity. It was hypothesized that thermotherapy could eliminate HLB from affected branches; therefore, the study objectives were to show which time-temperature combinations eliminated CLas from woody tissues. Hardening, rounded Valencia twigs collected from HLB-affected field trees were treated in a steam chamber at different time-temperature combinations (50°C for 60 s; 55°C for 0, 30, 60, 90, and 120 s; 60°C for 30 s; and an untreated control). Three independent repetitions of 13 branches per treatment were grafted onto healthy rootstocks and tested to detect CLas after 6, 9, and 12 months. For the RNA-based CLas viability assay, three branches per treatment were treated and bark samples were peeled for RNA extraction and subsequent gene expression analyses. During the grafting study, at 12 months after grafting, a very low frequency of trees grafted with twigs treated at 55°C for 90 s and 55°C for 120 s had detectable CLas DNA. In the few individuals with CLas, titers were significantly lower (P ≤ 0.0001) and could have been remnants of degrading DNA. Additionally, there was a significant decrease (P ≤ 0.0001) in CLas 16S rRNA expression at 55°C for 90 s, 55°C for 120 s, and 60°C for 30 s (3.4-fold change, 3.4-fold change, and 2.3-fold change, respectively) in samples 5 days after treatment. Heat injury, not total CLas kill, could explain the limited changes in transcriptional activity; however, failed recovery and eventual death of CLas resulted in no CLas detection in most of the grafted trees treated with the highest temperatures or longest durations.

Planar Interdigitated Aptasensor for Flow-Through Detection of Listeria spp. in Hydroponic Lettuce Growth Media

Irrigation water is a primary source of fresh produce contamination by bacteria during the preharvest, particularly in hydroponic systems where the control of pests and pathogens is a major challenge. In this work, we demonstrate the development of a Listeria biosensor using platinum interdigitated microelectrodes (Pt-IME). The sensor is incorporated into a particle/sediment trap for the real-time analysis of irrigation water in a hydroponic lettuce system. We demonstrate the application of this system using a smartphone-based potentiostat for rapid on-site analysis of water quality. A detailed characterization of the electrochemical behavior was conducted in the presence/absence of DNA and Listeria spp., which was followed by calibration in various solutions with and without flow. In flow conditions (100 mL samples), the aptasensor had a sensitivity of 3.37 ± 0.21 k log-CFU-1 mL, and the LOD was 48 ± 12 CFU mL-1 with a linear range of 102 to 104 CFU mL-1. In stagnant solution with no flow, the aptasensor performance was significantly improved in buffer, vegetable broth, and hydroponic media. Sensor hysteresis ranged from 2 to 16% after rinsing in a strong basic solution (direct reuse) and was insignificant after removing the aptamer via washing in Piranha solution (reuse after adsorption with fresh aptamer). This is the first demonstration of an aptasensor used to monitor microbial water quality for hydroponic lettuce in real time using a smartphone-based acquisition system for volumes that conform with the regulatory standards. The aptasensor demonstrated a recovery of 90% and may be reused a limited number of times with minor washing steps.

Survival of Shiga Toxin-Producing Escherichia coli in Various Wild Animal Feces That May Contaminate Produce

ABSTRACT

Domestic and wild animal intrusions are identified as a food safety risk during fresh produce production. The purpose of this study was to evaluate the survival of Shiga toxin-producing Escherichia coli (STEC) in cattle, feral pig, waterfowl, deer, and raccoon feces from sources in California, Delaware, Florida, and Ohio. Fecal samples were inoculated with a cocktail of rifampin-resistant STEC serotypes (O103, O104, O111, O145, and O157) (104 to 106 CFU/g of feces). Inoculated feces were held at ambient temperature. Populations of surviving cells were monitored throughout 1 year (364 days), with viable populations being enumerated by spread plating and enrichment when the bacteria were no longer detected by plating. Representative colonies were collected at various time intervals based on availability from different locations to determine the persistence of surviving STEC serotypes. Over the 364-day storage period, similar survival trends were observed for each type of animal feces from all states except for cattle and deer feces from Ohio. STEC populations remained the highest in cattle and deer feces from all states between days 28 and 364, except for those from Ohio. Feral pig, waterfowl, and raccoon feces had populations of STEC of <1.0 log CFU/g starting from day 112 in feces from all states. E. coli O103 and O104 were the predominant serotypes throughout the entire storage period in feces from all animals and from all states. The survival of both O157 and non-O157 STEC strains in domesticated and wild animal feces indicates a potential risk of contamination from animal intrusion.

HIGHLIGHTS

Survival of Salmonella in Various Wild Animal Feces That May Contaminate Produce

ABSTRACT

Heightened concerns about wildlife on produce farms and possible introduction of pathogens to the food supply have resulted in required actions following intrusion events. The purpose of this study was to evaluate the survival of Salmonella in feces from cattle and various wild animals (feral pigs, waterfowl, deer, and raccoons) in California, Delaware, Florida, and Ohio. Feces were inoculated with rifampin-resistant Salmonella enterica cocktails that included six serotypes: Typhimurium, Montevideo, Anatum, Javiana, Braenderup, and Newport (104 to 106 CFU/g). Fecal samples were stored at ambient temperature. Populations were enumerated for up to 1 year (364 days) by spread plating onto tryptic soy agar supplemented with rifampin. When no colonies were detected, samples were enriched. Colonies were banked on various sampling days based on availability of serotyping in each state. During the 364-day storage period, Salmonella populations decreased to ≤2.0 log CFU/g by day 84 in pig, waterfowl, and raccoon feces from all states. Salmonella populations in cattle and deer feces were 3.3 to 6.1 log CFU/g on day 336 or 364; however, in Ohio Salmonella was not detected after 120 days. Salmonella serotypes Anatum, Braenderup, and Javiana were the predominant serotypes throughout the storage period in all animal feces and states. Determination of appropriate risk mitigation strategies following animal intrusions can improve our understanding of pathogen survival in animal feces.

HIGHLIGHTS

Quantification of Salmonella enterica transfer between tomatoes, soil, and plastic mulch

Tomatoes have been linked to Salmonella outbreaks in the United States (US). Plasticulture systems, that combine raised beds, plastic mulch, drip irrigation and fumigation, are common in commercial staked fresh tomato production in the US. The US FDA Produce Safety Rule prohibits the distribution of any produce covered by the rule (including fresh market tomatoes) that drops to the ground before harvest. This research was undertaken to better characterize the risks posed by tomatoes that touch plastic mulch or soil immediately before or during harvest. Research was conducted in three states (Florida, Maryland, and Ohio). Each state utilized tomatoes from their state at the point of harvest maturity most common in that state. Each state used indigenous soil and plastic mulch for transfer scenarios. New plastic mulch obtained directly from the application roll and used plastic mulch that had been present on beds for a growing season were evaluated. A five-strain cocktail of Salmonella enterica isolates obtained from tomato outbreaks was used. Mulch (new or used), soil, or tomatoes were spot inoculated with 100 μl of inoculum to obtain a final population of ~6 log CFU/surface. Items were either touched to each other immediately (1-2 s) after inoculation (wet contact) or allowed to dry at ambient temperature for 1 h or 24 h (dry contact). All surfaces remained in brief (1-5 s) or extended (24 h) contact at ambient temperature. Transfer of Salmonella between a tomato and plastic mulch or soil is dependent on contact time, dryness of the inoculum, type of soil, and contact surface. Transfer of Salmonella to and from the mulch and tomatoes for wet and 1 h dry inocula were similar with mean log % transfers varying from 0.7 ± 0.2 to 1.9 ± 0.1. The transfer of Salmonella between soil or plastic mulch to and from tomatoes was dependent on moisture with wet and 1 h dry inocula generally yielding significantly (p < 0.05) higher transfer than the 24 h dry inoculum. Results indicate that harvesting dry tomatoes significantly (p < 0.05) reduces the risk of contamination from soil or mulch contact. Transfer to tomatoes was generally significantly greater (p < 0.05) from new and used plastic mulch than from soil. If contamination and moisture levels are equivalent and contact times are equal to or <24 h before harvest, significantly (p < 0.05) more Salmonella transfers to tomatoes from mulch than from soil. Our findings support that harvesting tomatoes from soil has similar or lower risk than harvesting from plastic mulch.

Fate of Salmonella in Central Florida Surface Waters and Evaluation of EPA Worst Case Water as a Standard Medium

HIGHLIGHTS

Survival of Salmonella was studied in surface waters. Salmonella stayed alive in nonsterile microcosms for 168 days. Limited decline was seen in sterile surface, deionized, and EPA Worst Case water. EPA Worst Case water offers potential as a standardized medium.

Prevalence and concentration of stx+ E. coli and E. coli O157 in bovine manure from Florida farms

Fresh produce outbreaks due to Shiga toxin-producing Escherichia coli (STEC) continue to occur in the United States (US). Manure-amended soils can pose a public health risk when used for growing raw agricultural commodities. Knowing the prevalence and concentration of STEC in untreated biological soil amendments of animal origin (BSAAO) is important to help guide the most appropriate pre-harvest interval(s) following application to limit risks from these soil amendments. Bovine manure samples were collected from 12 farms in Florida, including samples from piles, lagoons, barns, and screened solids. Two methods were used to detect stx1/2 and rfbE genes in samples. A prevalence rate of 9% for stx1 and/or stx2 and 19% for rfbE was observed from the 518 bovine manure samples evaluated. A most probable number (MPN) assay was performed on stx+ samples when applicable. The geometric mean for stx+ samples (n = 20) was 3.37 MPN g^-1 (0.53 log MPN g^-1) with a maximum value of 6,800 MPN g^-1 (3.83 log MPN g^-1). This research was part of a larger nationwide geographical study on the prevalence and concentration of STEC in bovine manure to help guide regulations on feasible pre-harvest intervals for the application of untreated BSAAO.

Growth and Survival of Listeria monocytogenes and Salmonella on Whole and Sliced Cucumbers

Cucumbers were associated with four multistate outbreaks of Salmonella in the United States between 2013 and 2016. This study evaluated the fate of Listeria monocytogenes and Salmonella on whole and sliced cucumbers at various storage temperatures. Cucumbers were inoculated with five-strain cocktails of L. monocytogenes or Salmonella, air dried, and stored at 23 ± 2, 4 ± 2, and -18 ± 2°C. Whole and sliced cucumber samples were enumerated on nonselective and selective media at 0, 0.21, 1, 2, 3, and 4 days (23 ± 2°C); 0, 1, 2, 3, 7, 14, and 21 days (4 ± 2°C); and 0, 7, 28, 60, 90, and 120 days (-18 ± 2°C). For Salmonella, additional time points were added at 8 and 17 h (23 ± 2°C) and at 17 h (4 ± 2°C). Population levels were calculated for whole (CFU per cucumber) and sliced (CFU per gram) cucumbers. Both pathogens grew on whole and sliced cucumbers held at ambient temperatures. At 23 ± 2°C, L. monocytogenes and Salmonella populations significantly increased on whole (2.3 and 3.4 log CFU per cucumber, respectively) and sliced (1.7 and 3.2 log CFU/g, respectively) cucumbers within 1 day. Salmonella populations significantly increased on whole and sliced cucumbers after only 5 h (2.1 log CFU per cucumber and 1.5 log CFU/g, respectively), whereas L. monocytogenes populations were not significantly different on whole and sliced cucumbers at 5 h. L. monocytogenes and Salmonella populations survived up to 21 days on refrigerated whole and sliced cucumbers. At 4 ± 2°C, L. monocytogenes populations significantly increased on whole (2.8 log CFU per cucumber) and sliced (2.9 log CFU/g) cucumbers, whereas Salmonella populations significantly decreased on whole (0.6 log CFU per cucumber) and sliced (1.3 log CFU/g) cucumbers over 21 days. Both pathogens survived on frozen whole and sliced cucumbers for at least 120 days. The ability of L. monocytogenes and Salmonella to grow on whole and sliced cucumbers in short amounts of time at ambient temperatures, and to survive on whole and sliced cucumbers past the recommended shelf life at refrigeration temperatures, highlights the need to reduce the likelihood of contamination events throughout the cucumber supply chain.

Validation of Enterococcus faecium as a surrogate for Salmonella under different processing conditions for peanuts and pecans

Food Safety and Modernization Act (FSMA) Preventive Control rules require nut processors validate thermal processes to ensure a desirable log reduction of Salmonella is achieved. Due to the complex nature of nut and nut products, processes and equipment, it is difficult to use one validation study for all and may requires individual equipment be validated at the plant level. In plant validation studies, pathogens such as Salmonella cannot be used due to the risk of contamination, thus the suitability of a non-pathogenic organism, Enterococcus faecium as a surrogate for Salmonella was evaluated for peanut and pecan thermal processing. Stagnant and forced dry air heating conditions, (120 °C (20, 30, 40 min), 130 °C (10, 20, 30 min), 140 °C (10, 20, 30 min)) were evaluated for unblanched peanut kernels. Oil heating conditions (116 °C, 121 °C, and 127 °C for 0.5, 1.0, 1.5, 2.0, 2.5 min) were evaluated for pecan kernels. Inshell pecans are conditioned in hot or cold water to facilitate the shelling process. Water heating conditions (75 °C (20, 40, 80, 120 s), 80 °C (20, 40, 80, 120 s), 85 °C (20, 40, 80, 120 s), 90 °C (20, 40, 60, 80 s), and 95 °C (20, 40, 60, 80 s)) were evaluated for inshell pecans. Under conditions, except forced air treatment, E. faecium reductions (Log N/N₀) were either not significantly different (P > 0.05) or significantly lower than Salmonella (P < 0.05), making it a suitable surrogate for the processes evaluated.

Identification of Biological Hazards in Produce Consumed in Industrialized Countries: A Review

Microbial contamination of fresh produce (fresh fruits and vegetables) poses serious public health concerns worldwide. This study was conducted as a comprehensive analysis of biological hazards in the global fresh produce chain. Data about produce-related outbreaks and illness were collected from the annual reports and databases of foodborne outbreak surveillance systems in different regions and countries from 2010 to 2015. The global patterns of and regional differences in documented outbreaks and cases were analyzed, and produce commodities and pathogens of greatest concern were identified. Data on sporadic illnesses were also collected through a comprehensive literature review of case-control studies. We found 988 produce-related outbreaks (with known agents) and 45,723 cases in all regions and countries. The numbers of produce-related outbreaks per million person-years were approximately 0.76, 0.26, 0.25, 0.13, 0.12, and 0.05 in New Zealand, Australia, the United States, the European Union, Canada, and Japan, respectively. The top three food categories and pathogens contributing to produce-related outbreaks were vegetables and nonfruits (i.e., food other than fruits; 27.0%), unspecified vegetables (12.2%), and vegetable row crops (11.7%) and norovirus (42.4%), Salmonella enterica (19.9%), and Staphylococcus aureus (7.9%), respectively. Produce consumption was identified as a protective factor, a risk factor, and either a protective or risk factor for sporadic illnesses in 11, 5, and 5 studies, respectively, among 21 case-control studies. Risks associated with produce consumption in the United States and the European Union have been linked to various factors such as irrigation water, cross-contamination, storage time and temperature abuse, infected food handlers, and unprocessed contaminated ingredients. The results of the current study indicate the complexity of produce products consumed across the globe and the difficulty in tracing illnesses back to specific food ingredients.

Characterization of Four Novel Bacteriophages Isolated from British Columbia for Control of Non-typhoidal Salmonella in Vitro and on Sprouting Alfalfa Seeds

Alfalfa sprouts have been linked to numerous North American outbreaks of Salmonella in recent years. Conventionally, treatments involving chlorine, heat, and irradiation are used for alfalfa seed sanitation. However, such treatments may be highly variable in their efficacy for pathogen control and/or detrimental to sprout quality, therefore negatively perceived by consumers advocating for natural alternatives. The usage of bacteriophages for pathogen control in sprouts has been previously explored, although with conflicting and inconsistent results. Lytic phages, viral predators of bacteria, represent an attractive approach as they provide several advantages compared to conventional treatments, such as their high specificity for bacterial targets and their ubiquity in nature. In this study, four Salmonella phages were isolated from British Columbia, Canada and characterized with respect to host range, burst size, latent period, and environmental stability to assess their potential to control Salmonella. Phage isolate SI1 showed the greatest host range, highest burst size and shortest latent period, greatest stability across all pH and temperatures and was the most effective in control of S. Enteritidis in vitro. Therefore, SI1 was chosen for treatment of sprouting alfalfa seeds artificially contaminated with S. Enteritidis with a multiplicity of infection (MOI) of ∼110 PFU/CFU. A significant (p < 0.05) reduction of 38.3 ± 3.0% of viable Salmonella cells was observed following two h of phage treatment. On days two to six of the sprouting process, reductions of Salmonella were also observed, but were not significant compared to the control (p > 0.05). It was further demonstrated that the sprout yield was not significantly (p > 0.05) affected by phage treatment. These results highlight the potential of phages recovered from the British Columbia environment for use as biocontrol agents against Salmonella, although differing efficacies in vitro was observed. Moreover, the effectiveness of SI1 to significantly (p < 0.05) control Salmonella on sprouting alfalfa seeds on day 1 of treatment was demonstrated. Although promising, future work should aim to optimize this treatment to achieve more effective, and longer lasting, biocontrol of Salmonella in sprouting alfalfa seeds.

Evaluating the U.S. Food Safety Modernization Act Produce Safety Rule Standard for Microbial Quality of Agricultural Water for Growing Produce

The U.S. Food and Drug Administration (FDA) has defined standards for the microbial quality of agricultural surface water used for irrigation. According to the FDA produce safety rule (PSR), a microbial water quality profile requires analysis of a minimum of 20 samples for Escherichia coli over 2 to 4 years. The geometric mean (GM) level of E. coli should not exceed 126 CFU/100 mL, and the statistical threshold value (STV) should not exceed 410 CFU/100 mL. The water quality profile should be updated by analysis of a minimum of five samples per year. We used an extensive set of data on levels of E. coli and other fecal indicator organisms, the presence or absence of Salmonella, and physicochemical parameters in six agricultural irrigation ponds in West Central Florida to evaluate the empirical and theoretical basis of this PSR. We found highly variable log-transformed E. coli levels, with standard deviations exceeding those assumed in the PSR by up to threefold. Lognormal distributions provided an acceptable fit to the data in most cases but may underestimate extreme levels. Replacing censored data with the detection limit of the microbial tests underestimated the true variability, leading to biased estimates of GM and STV. Maximum likelihood estimation using truncated lognormal distributions is recommended. Twenty samples are not sufficient to characterize the bacteriological quality of irrigation ponds, and a rolling data set of five samples per year used to update GM and STV values results in highly uncertain results and delays in detecting a shift in water quality. In these ponds, E. coli was an adequate predictor of the presence of Salmonella in 150-mL samples, and turbidity was a second significant variable. The variability in levels of E. coli in agricultural water was higher than that anticipated when the PSR was finalized, and more detailed information based on mechanistic modeling is necessary to develop targeted risk management strategies.

A Syst-OMICS Approach to Ensuring Food Safety and Reducing the Economic Burden of Salmonellosis

The Salmonella Syst-OMICS consortium is sequencing 4,500 Salmonella genomes and building an analysis pipeline for the study of Salmonella genome evolution, antibiotic resistance and virulence genes. Metadata, including phenotypic as well as genomic data, for isolates of the collection are provided through the Salmonella Foodborne Syst-OMICS database (SalFoS), at https://salfos.ibis.ulaval.ca/. Here, we present our strategy and the analysis of the first 3,377 genomes. Our data will be used to draw potential links between strains found in fresh produce, humans, animals and the environment. The ultimate goals are to understand how Salmonella evolves over time, improve the accuracy of diagnostic methods, develop control methods in the field, and identify prognostic markers for evidence-based decisions in epidemiology and surveillance.

Quantification of Transfer of Salmonella from Citrus Fruits to Peel, Edible Portion, and Gloved Hands during Hand Peeling

Although studies have quantified bacterial transfer between hands and various materials, cross-contamination between the surface of fresh citrus fruit and the edible portions during hand peeling has not been reported. This study quantifies transfer of Salmonella to the edible portion of citrus fruit from a contaminated peel during hand peeling. Citrus fruits used for this study were Citrus sinensis (sweet orange) cultivars 'Valencia' and 'Navel', Citrus unshiu (Satsuma mandarins), Citrus reticulata × Citrus paradisi ('Minneola' tangelo or 'Honeybell'), and C. paradisi (grapefruit) cultivar 'Marsh'. An avirulent Salmonella Typhimurium LT2 (ATCC 700720) resistant to rifampin was used for all experiments. The inoculum containing approximately 9 log CFU/mL (50 μL) was spot inoculated onto the equator, stem, or styler of each fruit and allowed to dry for 24 h. Six volunteers put on single-use latex gloves and peeled inoculated fruit. Peel, edible fruit portion, and gloves were collected and enumerated separately. Three replicates of the study were performed in which each volunteer peeled two inoculated fruit of each variety (n = 36 fruit per variety). Cross-contamination from contaminated surface of citrus fruits to edible portion or gloved hands during peeling was affected by inoculation sites. Average Salmonella transfer to the edible portion ranged from 0.16% (Valencia inoculated at the equator) to 5.41% (navel inoculated at the stem). Average Salmonella transfer to gloved hands ranged from 0.41% (grapefruit inoculated at the stem) to 8.97% (navel inoculated at the stem). Most Salmonella remained on the peel of citrus fruits. The average level of Salmonella remaining on the peel ranged from 5.37% (Minneola inoculated at the equator) to 66.3% (Satsuma inoculated at the styler). When grapefruit was inoculated, the Salmonella that remained on the peel showed a bimodal pattern in which some individuals left almost all Salmonella on the peel, while others left substantially less.

Microbial quality of agricultural water in Central Florida

The microbial quality of water that comes into the edible portion of produce is believed to directly relate to the safety of produce, and metrics describing indicator organisms are commonly used to ensure safety. The US FDA Produce Safety Rule (PSR) sets very specific microbiological water quality metrics for agricultural water that contacts the harvestable portion of produce. Validation of these metrics for agricultural water is essential for produce safety. Water samples (500 mL) from six agricultural ponds were collected during the 2012/2013 and 2013/2014 growing seasons (46 and 44 samples respectively, 540 from all ponds). Microbial indicator populations (total coliforms, generic Escherichia coli, and enterococci) were enumerated, environmental variables (temperature, pH, conductivity, redox potential, and turbidity) measured, and pathogen presence evaluated by PCR. Salmonella isolates were serotyped and analyzed by pulsed-field gel electrophoresis. Following rain events, coliforms increased up to 4.2 log MPN/100 mL. Populations of coliforms and enterococci ranged from 2 to 8 and 1 to 5 log MPN/100 mL, respectively. Microbial indicators did not correlate with environmental variables, except pH (P<0.0001). The invA gene (Salmonella) was detected in 26/540 (4.8%) samples, in all ponds and growing seasons, and 14 serotypes detected. Six STEC genes were detected in samples: hly (83.3%), fliC (51.8%), eaeA (17.4%), rfbE (17.4%), stx-I (32.6%), stx-II (9.4%). While all ponds met the PSR requirements, at least one virulence gene from Salmonella (invA-4.8%) or STEC (stx-I-32.6%, stx-II-9.4%) was detected in each pond. Water quality for tested agricultural ponds, below recommended standards, did not guarantee the absence of pathogens. Investigating the relationships among physicochemical attributes, environmental factors, indicator microorganisms, and pathogen presence allows researchers to have a greater understanding of contamination risks from agricultural surface waters in the field.

Quantifying Bacterial Cross-Contamination Rates between Fresh-Cut Produce and Hands

This study quantifies the cross-contamination rates between fresh-cut produce and hands using a nalidixic acid-resistant nonpathogenic Enterobacter aerogenes and cocktails of rifampin-resistant Salmonella or Escherichia coli O157:H7 strains. Volunteers performed the E. aerogenes experiments (n = 20), and one of the authors performed the Salmonella and E. coli O157:H7 experiments multiple times (n =15 and n =10, respectively). Each participant handled 25 g of fresh-cut carrots, celery, or cantaloupe in two different scenarios. In the first scenario, gloved hands were inoculated with 6 log CFU per hand of the bacteria, and in the second scenario, five 25-g pieces of fresh produce were inoculated to a concentration of 6 log CFU/25 g. The glove juice method was used to quantify the bacterial concentration on the gloved hands. About 30% of E. aerogenes on gloved hands was transferred to the carrots and celery, and 18% of E. aerogenes on gloved hands was transferred to the cantaloupe. When carrots or cantaloupe was inoculated with E. aerogenes , 1% was transferred to gloved hands; from inoculated celery, about 0.3% of E. aerogenes was transferred to gloved hands. There was not a significant difference between E. aerogenes and Salmonella cross-contamination rates (P > 0.05). When gloved hands were contaminated with E. coli O157:H7, about 30% was transferred to carrots, about 10% to celery, and about 3% to cantaloupe. When carrots and celery were inoculated with E. coli O157:H7, about 1% was transferred to gloved hands, but from inoculated cantaloupe only about 0.3% was transferred. Direction of transfer (to versus from produce), difference in type of produce, and differences among the bacterial species all had significant effects on the transfer rate. Understanding transfer rates to and from fresh-cut produce will allow for better risk assessment and management of microbial food safety risk related to fresh-cut produce.

Influence of Temperature Differential between Tomatoes and Postharvest Water on Salmonella Internalization

Salmonella bacteria may internalize into tomato pulp when warm tomatoes from the field are submerged into colder water. Several washing steps may follow the initial washing and packing of tomatoes at the packinghouses; the potential for internalization into tomatoes in subsequent washing steps when tomatoes have a cooler pulp temperature is unknown. Our objective was to evaluate Salmonella internalization into mature green and red tomatoes with ambient (21°C) and refrigeration (4°C) pulp temperatures when they were submerged into water at various temperature differentials, simulating repacking and fresh-cut operations. Red (4°C and 21°C) and mature green (21°C) tomatoes were submerged (6 cm) into a six-strain Salmonella cocktail (6 log CFU/ml) and maintained at ±5 and 0°C temperature differentials for varying time intervals, ranging from 30 s to 5 min. Following submersion, tomatoes were surface sterilized using 70% ethanol, the stem abscission zone and blossom end epidermis were removed, and cores were recovered, separated into three segments, and analyzed. Salmonella populations in the segments were enumerated by most probable number (MPN). The effects of temperature differential and maturity on Salmonella populations were analyzed; results were considered significant at a P value of ≥0.5. Internalized populations were not significantly different (P ≥0.5) across temperature differentials. Salmonella internalization was seen in tomatoes under all treatment conditions and was highest in the segment immediately below the stem abscission zone. However, populations were low (typically >1 log MPN per segment) and varied greatly across temperature differentials. This suggests that the temperature differential between tomatoes and water beyond the initial packinghouse may be less important than submersion time in Salmonella internalization.

Prevalence, Level, and Types of Salmonella Isolated from North American In-Shell Pecans over Four Harvest Years

In-shell pecan samples (500 g) were collected over four harvest seasons (2010 to 2014) from seven pecan shelling facilities located in five U.S. states. Four varieties of pecans were analyzed: Mexican Improved, Native Seedlings, Southern Improved, and Western Improved. Pecan samples (100 g) were sent to a third party laboratory for initial Salmonella screening. When a sample was positive for Salmonella, the pathogen level was determined by the most-probable-number (MPN) method (25, 2.5, and 0.25 g). Two sample preparation strategies were used for the MPN analysis, and both strategies were combined for the reported MPN values. Forty-four (0.95%) of 4,641 in-shell pecan samples were positive for Salmonella during initial screening; prevalence by year was 0.47 to 1.4%. Prevalence was not significantly different between varieties: Mexican Improved, 1.2%; Native/Seedling, 0.99%; Southern Improved, 0.97%; and Western Improved, 0.75%. Salmonella was not isolated from 31 of 44 samples upon retesting during MPN analysis (<0.47 MPN/100 g). When Salmonella was detected, the levels were 0.47 to 39 MPN/100 g, with a mean of 2.4 MPN/100 g. Thirty-one Salmonella serotypes were obtained from 42 Salmonella-positive pecan samples; Enteritidis was the most common (12% of samples) followed by Javiana (9%) and Braenderup (7%). All Salmonella Enteritidis isolates were phage type 8. Pulsed-field gel electrophoresis analysis (XbaI) revealed within-serotype diversity, indicating introduction of contamination from a variety of sources. Most (64%) of the isolates were resistant to streptomycin or tetracycline, and 13% were resistant to three or more antibiotics. Salmonella prevalence and level on in-shell pecans is comparable to that on other nuts.

Quantifying the effect of hand wash duration, soap use, ground beef debris, and drying methods on the removal of Enterobacter aerogenes on hands

Hand washing is recognized as a crucial step in preventing foodborne disease transmission by mitigating crosscontamination among hands, surfaces, and foods. This research was undertaken to establish the importance of several keys factors (soap, soil, time, and drying method) in reducing microorganisms during hand washing. A nonpathogenic nalidixic acid-resistant Enterobacter aerogenes surrogate for Salmonella was used to assess the efficacy of using soap or no soap for 5 or 20 s on hands with or without ground beef debris and drying with paper towel or air. Each experiment consisted of 20 replicates, each from a different individual with ∼ 6 log CFU/ml E. aerogenes on their hands. A reduction of 1.0 ± 0.4 and 1.7 ± 0.8 log CFU of E. aerogenes was observed for a 5-s wash with no soap and a 20-s wash with soap, respectively. When there was no debris on the hands, there was no significant difference between washing with and without soap for 20 s (P > 0.05). Likewise, there was no significant difference in the reductions achieved when washing without soap, whether or not debris was on the hands (P > 0.05). A significantly greater reduction (P < 0.05) in E. aerogenes (0.5 log CFU greater reduction) was observed with soap when there was ground beef debris on the hands. The greatest difference (1.1 log CFU greater average reduction) in effectiveness occurred when ground beef debris was on the hands and a 20-s wash with water was compared with a 20-s wash with soap. Significantly greater (P < 0.05) reductions were observed with paper towel drying compared with air (0.5 log CFU greater reductions). Used paper towels may contain high bacterial levels (>4.0 log CFU per towel) when hands are highly contaminated. Our results support future quantitative microbial risk assessments needed to effectively manage risks of foodborne illness in which food workers' hands are a primary cause.

Cross contamination of Escherichia coli O157:H7 between lettuce and wash water during home-scale washing

Lettuce and leafy greens have been implicated in multiple foodborne disease outbreaks. This study quantifies cross contamination between lettuce pieces in a small-scale home environment. A five-strain cocktail of relevant Escherichia coli O157:H7 strains was used. Bacterial transfer between single inoculated lettuce leaf pieces to 10 non-inoculated lettuce leaf pieces that were washed in a stainless steel bowl of water for 30 s, 1 min, 2 min, and 5 min was quantified. Regardless of washing time, the wash water became contaminated with 90-99% of bacteria originally present on the inoculated lettuce leaf piece. The E. coli O157:H7 concentration on initially inoculated leaf pieces was reduced ∼ 2 log CFU. Each initially uncontaminated lettuce leaf piece had ∼ 1% of the E. coli O157:H7 from the inoculated lettuce piece transferred to it after washing, with more transfer occurring during the shortest (30 s) and longest (5 min) wash times. In all cases the log percent transfer rates were essentially normally distributed. In all scenarios, most of the E. coli O157:H7 (90-99%) transferred from the inoculated lettuce pieces to the wash water. Washing with plain tap water reduces levels of E. coli O157:H7 on the inoculated lettuce leaf pieces, but also spreads contamination to previously uncontaminated leaf pieces.

Survival of Salmonella, Escherichia coli O157:H7, and Listeria monocytogenes on raw peanut and pecan kernels stored at -24, 4, and 22°C

Cocktails of lawn-collected cells were used to determine the survival of Salmonella, Escherichia coli O157:H7, and Listeria monocytogenes on the surface of raw peanut and pecan kernels. Kernels were inoculated with mixtures of four to five strains at 3 or 6 log CFU/g, dried at room temperature, and then stored at -24 ± 1, 4 ± 2, and 22 ± 1°C for 28 or 365 days. In most cases, rates of decline of the pathogens did not differ significantly between the two inoculum concentrations in the 28-day study. At 6 log CFU/g, populations of all pathogens were reduced by 0.5 to 1.6 log CFU/g during an initial 3-day drying period on both peanuts and pecans. The moisture content of peanuts and pecans remained stable at -24 ± 1 and 22 ± 1°C; at 4 ± 2°C, the moisture content increased from 3.8 to 5.6% on peanuts and from 2.6 to 3% on pecans over 365 days. Pathogen populations were stable on pecans stored under frozen and refrigerated conditions, except for L. monocytogenes, which declined at a rate of 0.03 log CFU/g/30 days at 4 ± 2°C. Salmonella populations were stable on peanuts stored at -24 ± 1 and 4 ± 2°C, but E. coli O157:H7 and L. monocytogenes declined at rates of 0.03 to 0.12 log CFU/g/30 days. At 22 ± 1°C, Salmonella, E. coli O157:H7, and L. monocytogenes declined at a rate of 0.22, 0.37, and 0.59 log CFU/g/30 days, respectively, on peanuts, and at 0.15, 0.34, and 1.17 log CFU/g/30 days, respectively, on pecans. Salmonella counts were above the limit of detection (0.30 log CFU/g) throughout the study. In most cases during storage, counts obtained from pecans were higher than from peanuts.

Survival or growth of inoculated Escherichia coli O157:H7 and Salmonella on yellow onions (Allium cepa) under conditions simulating food service and consumer handling and storage

Whole and diced yellow onions (Allium cepa) were inoculated with five-strain cocktails of rifampin-resistant Escherichia coli O157:H7 or Salmonella and stored under conditions to simulate food service or consumer handling. The inoculum was grown in broth (for both whole and diced onion experiments) or on agar plates (for whole onion experiments). Marked circles (3.3 cm in diameter) on the outer papery skin of whole onions were spot inoculated (10 μl in 10 drops) at 7 log CFU per circle, and onions were stored at 4°C, 30 to 50 % relative humidity, or at ambient conditions (23°C, 30 to 50 % relative humidity). Diced onions were inoculated at 3 log CFU/g and then stored in open or closed containers at 4°C or ambient conditions. Previously inoculated and ambient-stored diced onions were also mixed 1:9 (wt/wt) with refrigerated uninoculated freshly diced onions and stored in closed containers at ambient conditions. Inoculated pathogens were recovered in 0.1 % peptone and plated onto selective and nonselective media supplemented with 50 μg/ml rifampin. Both E. coli O157:H7 and Salmonella populations declined more rapidly on onion skins when the inoculum was prepared in broth rather than on agar. Agar-prepared E. coli O157:H7 and Salmonella declined by 0.4 and 0.3 log CFU per sample per day, respectively, at ambient conditions; at 4°C the rates of reduction were 0.08 and 0.06 log CFU per sample per day for E. coli O157:H7 and Salmonella, respectively. Populations of E. coli O157:H7 and Salmonella did not change over 6 days of storage at 4°C in diced onions. Lag times of 6 to 9 h were observed with freshly inoculated onion at ambient conditions; no lag was observed when previously inoculated and uninoculated onions were mixed. Growth rates at ambient conditions were 0.2 to 0.3 log CFU/g/h for E. coli O157:H7 and Salmonella in freshly inoculated onion and 0.2 log CFU/g/h in mixed product. Diced onions support pathogen growth and should be kept refrigerated.

Fate of Escherichia coli O157:H7 and Salmonella on whole strawberries and blueberries of two maturities under different storage conditions

Strawberries and blueberries harvested at or near full-ripe maturity tend to be less firm and more susceptible to bruising during harvest and transport. The objective of this research was to determine the fate of Escherichia coli O157:H7 and Salmonella on bruised and intact surfaces of whole strawberries and blueberries at shipping (2°C) and retail display (15.5°C) temperatures. Strawberries and blueberries were either purchased from a supermarket or were harvested immediately prior to use; they were bruised using established protocols, were spot inoculated, and were incubated at 2 and 15.5°C. Strawberries, subjected to modified atmospheres, were further transferred to bags and were sealed in with an initial atmosphere of ca. 10% CO2 and 5% O2. Strawberries were sampled at 0, 2, 5, and 24 h and on days 3 and 7; blueberries were sampled on days 0, 1, 3, and 7. After stomaching, samples were enumerated on nonselective and selective media, and populations were recorded as log CFU per berry. At both storage temperatures, population declines for both E. coli O157:H7 and Salmonella were seen under all conditions for strawberries. At 2 ± 2°C, E. coli O157:H7 and Salmonella populations on blueberries declined over 7 days under all conditions. At 15.5 ± 2°C, E. coli O157:H7 populations declined; however, Salmonella populations initially declined but increased to populations near or above initial populations over 7 days on blueberries. No overall significant differences were observed between bruised and intact treatments or between the two maturity levels for strawberries and blueberries. Modified atmospheric conditions did not affect the behavior of E. coli O157:H7 and Salmonella on strawberries at both temperatures. This research indicates that E. coli O157:H7 and Salmonella do not grow on strawberries at shipping or retail display temperatures, even when they are harvested at a maturity prone to bruising; however, Salmonella growth may occur on bruised full ripe blueberries under retail display temperatures.

Distributions of Salmonella subtypes differ between two U.S. produce-growing regions

Salmonella accounts for approximately 50% of produce-associated outbreaks in the United States, several of which have been traced back to contamination in the produce production environment. To quantify Salmonella diversity and aid in identification of Salmonella contamination sources, we characterized Salmonella isolates from two geographically diverse produce-growing regions in the United States. Initially, we characterized the Salmonella serotype and subtype diversity associated with 1,677 samples collected from 33 produce farms in New York State (NYS). Among these 1,677 samples, 74 were Salmonella positive, yielding 80 unique isolates (from 147 total isolates), which represented 14 serovars and 23 different pulsed-field gel electrophoresis (PFGE) types. To explore regional Salmonella diversity associated with production environments, we collected a smaller set of samples (n = 65) from South Florida (SFL) production environments and compared the Salmonella diversity associated with these samples with the diversity found among NYS production environments. Among these 65 samples, 23 were Salmonella positive, yielding 32 unique isolates (from 81 total isolates), which represented 11 serovars and 17 different PFGE types. The most common serovars isolated in NYS were Salmonella enterica serovars Newport, Cerro, and Thompson, while common serovars isolated in SFL were Salmonella serovars Saphra and Newport and S. enterica subsp. diarizonae serovar 50:r:z. High PFGE type diversity (Simpson's diversity index, 0.90 ± 0.02) was observed among Salmonella isolates across both regions; only three PFGE types were shared between the two regions. The probability of three or fewer shared PFGE types was <0.000001; therefore, Salmonella isolates were considerably different between the two sampled regions. These findings suggest the potential for PFGE-based source tracking of Salmonella in production environments.

Salmonella transfer potential onto tomatoes during laboratory-simulated in-field debris removal

Florida Tomato Good Agricultural Practices (T-GAPs) mandate the removal of dirt and debris from tomatoes during harvest but do not provide any specific regulations or guidance; thus, the current practice of using cloths needs to be evaluated. This study examined Salmonella transfer from inoculated green tomatoes to uninoculated cloths and from inoculated cloths to uninoculated tomatoes, upon single and multiple touches. Tomatoes were spot inoculated with a rifampin-resistant Salmonella cocktail (10(7) CFU per tomato) and were touched with cloth (clean, dirty-dry, dirty-wet) at 0, 1, or 24 h postinoculation. Salmonella was enumerated on tryptic soy agar, followed by enrichments when necessary. The transfer direction was then reversed by touching freshly inoculated cloths with uninoculated tomatoes. Transfer coefficients (TCs) were then calculated. Salmonella TCs from inoculated tomato and cloth were highest when the inoculum was wet (0.44 ± 0.13 to 0.32 ± 0.12), regardless of the condition of the cloth. Although Salmonella TCs from inoculated tomato to uninoculated cloth decreased significantly when the inoculum was dried (0.17 ± 0.23 to 0.01 ± 0.00), low levels of Salmonella were detected on cloth even after 24 h of drying. Inoculated dirty cloth did not transfer more Salmonella compared with inoculated clean cloth, and Salmonella survival was not higher on dirty cloth. When inoculated clean cloth (wet) was touched with 25 tomatoes, significantly higher levels of Salmonella were transferred to the first, second, and fourth tomatoes (0.03 ± 0.10 to 0.09 ± 0.02). However, inoculated dirty-wet (below limit of detection) and dirty-dry (0.00 to 0.04 ± 0.01) cloths transferred similar levels of Salmonella to all 25 tomatoes. Results indicate a low risk of potential Salmonella contamination when the same cloth is used multiple times for debris removal, especially under high moisture levels. Results also show that the use of dirty cloths did not increase the risk of Salmonella cross-contamination.

Colorimetric paper-based detection of Escherichia coli, Salmonella spp., and Listeria monocytogenes from large volumes of agricultural water

This protocol describes rapid colorimetric detection of Escherichia coli, Salmonella spp., and Listeria monocytogenes from large volumes (10 L) of agricultural waters. Here, water is filtered through sterile Modified Moore Swabs (MMS), which consist of a simple gauze filter enclosed in a plastic cartridge, to concentrate bacteria. Following filtration, non-selective or selective enrichments for the target bacteria are performed in the MMS. For colorimetric detection of the target bacteria, the enrichments are then assayed using paper-based analytical devices (µPADs) embedded with bacteria-indicative substrates. Each substrate reacts with target-indicative bacterial enzymes, generating colored products that can be detected visually (qualitative detection) on the µPAD. Alternatively, digital images of the reacted µPADs can be generated with common scanning or photographic devices and analyzed using ImageJ software, allowing for more objective and standardized interpretation of results. Although the biochemical screening procedures are designed to identify the aforementioned bacterial pathogens, in some cases enzymes produced by background microbiota or the degradation of the colorimetric substrates may produce a false positive. Therefore, confirmation using a more discriminatory diagnostic is needed. Nonetheless, this bacterial concentration and detection platform is inexpensive, sensitive (0.1 CFU/ml detection limit), easy to perform, and rapid (concentration, enrichment, and detection are performed within approximately 24 hr), justifying its use as an initial screening method for the microbiological quality of agricultural water.

Quantifying transfer rates of Salmonella and Escherichia coli O157:H7 between fresh-cut produce and common kitchen surfaces

Cross-contamination between foods and surfaces in food processing environments and home kitchens may play a significant role in foodborne disease transmission. This study quantifies the cross-contamination rates between a variety of fresh-cut produce and common kitchen surfaces (ceramic, stainless steel, glass, and plastic) using scenarios that differ by cross-contamination direction, surface type, produce type, and drying time/moisture level. A five-strain cocktail of rifampin-resistant Salmonella was used in transfer scenarios involving celery, carrot, and watermelon, and a five-strain cocktail of rifampin-resistant Escherichia coli O157:H7 was used in transfer scenarios involving lettuce. Produce or surface coupons were placed in buffer-filled filter bags and homogenized or massaged, respectively, to recover cells. The resulting solutions were serially diluted in 0.1% peptone and surface plated onto tryptic soy agar with 80 μg/ml rifampin and bismuth sulfite agar with 80 μg/ml rifampin for Salmonella or sorbitol MacConkey agar with 80 μg/ml rifampin for E. coli O157:H7. When the food contact surface was freshly inoculated, a high amount (>90%) of the inoculum was almost always transferred to the cut produce item. If the inoculated food contact surfaces were allowed to dry for 1 h, median transfer was generally >90% for carrots and watermelon but ranged from <1 to ∼70% for celery and lettuce. Freshly inoculated celery or lettuce transferred more bacteria (<2 to ∼25% of the inoculum) compared with freshly inoculated carrots or watermelon (approximately <1 to 8%). After 1 h of drying, the rate of transfer from inoculated celery, carrot, and lettuce was <0.01 to ∼5% and <1 to ∼5% for watermelon. Surface moisture and direction of transfer have the greatest influence on microbial transfer rates.

Modeling the growth of Listeria monocytogenes on cut cantaloupe, honeydew and watermelon

A recent outbreak linked to whole cantaloupes underscores the importance of understanding growth kinetics of Listeria monocytogenes in cut melons at different temperatures. Whole cantaloupe, watermelon, and honeydew purchased from a local supermarket were cut into 10 ± 1 g cubes. A four-strain cocktail of L. monocytogenes from food related outbreaks was used to inoculate fruit, resulting in ~10(3) CFU/10 g. Samples were stored at 4, 10, 15, 20, or 25 °C and L. monocytogenes were enumerated at appropriate time intervals. The square root model was used to describe L. monocytogenes growth rate as a function of temperature. The model was compared to prior models for Salmonella and Escherichia coli O157:H7 growth on cut melon, as well as models for L. monocytogenes on cantaloupe and L. monocytogenes ComBase models. The current model predicts faster growth of L. monocytogenes vs. Salmonella and E. coli O157:H7 at temperatures below 20 °C, and agrees with estimates from ComBase Predictor, and a corrected published model for L. monocytogenes on cut cantaloupe. The model predicts ~4 log CFU increase following 15 days at 5 °C, and ∼1 log CFU increase following 6 days at 4 °C. The model can also be used in subsequent quantitative microbial risk assessments.

Salmonella transfer potential during hand harvesting of tomatoes under laboratory conditions

Tomato good agricultural practices, mandatory guidelines in Florida, do not have specific regulations for glove use during tomato harvesting. The objective of the research reported here was to evaluate Salmonella transfer from contaminated gloves to tomatoes and vice versa upon single and subsequent touches. Experiments were performed using mature, green, round tomatoes with two types of gloves (reusable and single use) and two hygienic conditions of reusable glove (clean and dirty [fouled with tomato leaves]). The transfer scenarios used during experiments were glove to tomato, tomato to glove, and glove to up to 25 subsequently touched tomatoes. The inoculated surface (6 log CFU per surface), after drying for 24 h, touched the uninoculated surface for 5 s. Salmonella populations from gloves and tomatoes were enumerated on nonselective and selective agar supplemented with 80 μg/ml rifampin. Enrichments were performed when counts fell below the detection limit. The rates of Salmonella transfer to tomatoes during a single touch were similar for single-use and reusable gloves; transfer from tomatoes to gloves was higher to single-use gloves than to reusable gloves under wet (0 h) inoculation conditions. Dirty reusable gloves did not transfer more Salmonella than clean reusable gloves during single contact under any conditions. When a single glove was sequentially touched to multiple tomatoes, clean reusable gloves transferred higher levels of Salmonella to the first few tomatoes touched than did single-use gloves and dirty reusable gloves. As workers' gloves became dirty over time during harvest, the risk of Salmonella transfer to tomatoes did not increase.

Development and validation of a mathematical model for growth of pathogens in cut melons

Many outbreaks of foodborne illness associated with the consumption of fresh-cut melons have been reported. The objective of our research was to develop a mathematical model that predicts the growth rate of Salmonella on fresh-cut cantaloupe over a range of storage temperatures and to validate that model by using Salmonella and Escherichia coli O157:H7 on cantaloupe, honeydew, and watermelon, using both new data and data from the published studies. The growth of Salmonella on honeydew and watermelon and E. coli O157:H7 on cantaloupe, honeydew, and watermelon was monitored at temperatures of 4 to 25°C. The Ratkowsky (or square-root model) was used to describe Salmonella growth on cantaloupe as a function of storage temperature. Our results show that the levels of Salmonella on fresh-cut cantaloupe with an initial load of 3 log CFU/g can reach over 7 log CFU/g at 25°C within 24 h. No growth was observed at 4°C. A linear correlation was observed between the square root of Salmonella growth rate and temperature, such that √growth rate = 0.026 × (T - 5.613), R(2) = 0.9779. The model was generally suitable for predicting the growth of both Salmonella and E. coli O157:H7 on cantaloupe, honeydew, and watermelon, for both new data and data from the published literature. When compared with existing models for growth of Salmonella, the new model predicts a theoretic minimum growth temperature similar to the ComBase Predictive Models and Pathogen Modeling Program models but lower than other food-specific models. The ComBase Prediction Models results are very similar to the model developed in this study. Our research confirms that Salmonella can grow quickly and reach high concentrations when cut cantaloupe is stored at ambient temperatures, without visual signs of spoilage. Our model provides a fast and cost-effective method to estimate the effects of storage temperature on fresh-cut melon safety and could also be used in subsequent quantitative microbial risk assessments.

Issues to consider when setting intervention targets with limited data for low-moisture food commodities: a peanut case study

Peanuts and peanut-containing products have been linked to at least seven salmonellosis outbreaks worldwide in the past two decades. In response, the Technical Committee on Food Microbiology of the North American Branch of the International Life Sciences Institute collaborated with the American Peanut Council to convene a workshop to develop a framework for managing risk in low-moisture food commodities where large data sets are unavailable (using peanuts as the example). Workshop attendees were charged with answering questions regarding the appropriate statistical and scientific methods for setting log reduction targets with limited pathogen prevalence and concentration data, suitable quantities of data needed for determining appropriate log reduction targets, whether the requirement of a 5-log reduction in the absence of data to establish a target log reduction is appropriate, and what targeted log reduction would protect public health. This report concludes that the judgment about sufficient data is not solely scientific, but is instead a science-informed policy decision that must weigh additional societal issues. The participants noted that modeling efforts should proceed with sampling efforts, allowing one to compare various assumptions about prevalence and concentration and how they are combined. The discussions made clear that data and risk models developed for other low-moisture foods like almonds and pistachios may be applicable to peanuts. Workshop participants were comfortable with the use of a 5-log reduction for controlling risk in products like peanuts when the level of contamination of the raw ingredients is low (<1 CFU/g) and the process well controlled, even when limited data are available. The relevant stakeholders from the food safety community may eventually conclude that as additional data, assumptions, and models are developed, alternatives to a 5-log reduction might also result in the desired level of protection for peanuts and peanut products.

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

Agricultural water may contact fresh produce during irrigation and/or when crop protection sprays (e.g., cooling to prevent sunburn, frost protection, and agrochemical mixtures) are applied. This document provides a framework for designing research studies that would add to our understanding of preharvest microbial food safety hazards and control measures pertaining to agricultural water. Researchers will be able to use this document to design studies, to anticipate the scope and detail of data required, and to evaluate previously published work. This document should also be useful for evaluating the strength of existing data and thus should aid in identifying future research needs. Use of this document by the research community may lead to greater consistency or comparability than currently exists among research studies, which may ultimately facilitate direct comparison of hazards and efficacy of controls among different commodities, conditions, and practices.

Adhesion and splash dispersal of Salmonella enterica Typhimurium on tomato leaflets: effects of rdar morphotype and trichome density

Salmonella enterica strains with rdar (red dry and rough) and saw (smooth and white) morphotypes have previously been associated with tomato outbreaks but the dispersal mechanisms of these morphotypes are still poorly understood. In this study, Salmonella adhesion was distinguished from attachment by comparing different contact periods. Initial adhesion of rdar and saw morphotypes of Salmonella was compared in relation to tomato plants with different leaf trichome densities. Trichome densities were increased or reduced by treatment with jasmonic or salicylic acid, respectively. The overall effect of Salmonella morphotype and trichome density on splash dispersal was assessed in a rain simulator and correlated to cell hydrophobicity and initial adhesion. The presence of the rdar morphotype increased initial adhesion at high trichome densities but not at low trichome densities. Attachment of the rdar strain occurred after 30s contact time regardless of trichome density. Splash dispersal was slightly further for the saw morphotype than the rdar morphotype of S. enterica at all trichome densities. Salmonella cells of both morphotypes survived significantly better on the surface of high trichome density leaflets.

Growth of Clostridium perfringens during cooling of refried beans

Outbreaks of Clostridium perfringens have been associated with dishes containing refried beans from food service establishments. However, growth of C. perfringens in refried beans has not been investigated, and predictive models have not been validated in this food matrix. We investigated the growth of C. perfringens during the cooling of refried beans. Refried beans (pinto and black, with and without salt added) were inoculated with 3 log CFU/g C. perfringens spores and incubated isothermally at 12, 23, 30, 35, 40, 45, and 50°C. The levels of C. perfringens were monitored 3, 5, 8, and 10 h after inoculation, and then fitted to the Baranyi primary model and the Rosso secondary model prior to solving the Baranyi differential equation. The final model was validated by dynamic cooling experiments carried out in stockpots, thus mimicking the worst possible food service conditions. All refried beans samples supported the growth of C. perfringens, and all models fit the data with pseudo-R(2) values of 0.95 or greater and mean square errors of 0.3 or lower. The estimated maximum specific growth rates were generally higher in pinto beans, with or without salt added (2.64 and 1.95 h(-1), respectively), when compared with black beans, with or without salt added (1.78 and 1.61 h(-1), respectively). After 10 h of incubation, maximum populations of C. perfringens were significantly higher in samples with no salt added (7.9 log CFU/g for both pinto and black beans) than in samples with salt added (7.3 and 7.2 log CFU/g for pinto and black beans, respectively). The dynamic model predicted the growth of C. perfringens during cooling, with an average root mean squared error of 0.44. The use of large stockpots to cool refried beans led to an observed 1.2-log increase (1.5-log increase predicted by model) in levels of C. perfringens during cooling. The use of shallower pans for cooling is recommended, because they cool faster, therefore limiting the growth of C. perfringens.

Survival of Salmonella, Escherichia coli O157:H7, and Listeria monocytogenes on inoculated almonds and pistachios stored at -19, 4, and 24° C

The survival of Salmonella, Escherichia coli O157:H7, and Listeria monocytogenes was determined on almonds and pistachios held at typical storage temperatures. Almond kernels and inshell pistachios were inoculated with four- to six-strain cocktails of nalidixic acid-resistant Salmonella, E. coli O157:H7, or L. monocytogenes at 6 log CFU/g and then dried for 72 h. After drying, inoculated nuts were stored at -19, 4, or 24°C for up to 12 months. During the initial drying period after inoculation, levels of all pathogens declined by 1 to -log CFU/g on both almonds and pistachios. During storage, moisture content (4.8%) and water activity (0.4) of the almonds and pistachios were consistent at -19°C; increased slowly to 6% and 0.6, respectively, at 4°C; and fluctuated from 4 to 5% and 0.3 to 0.5 at 24°C, respectively. Every 1 or 2 months, levels of each pathogen were enumerated by plating; samples were enriched when levels fell below the limit of detection. No reduction in population level was observed at -19 or 4°C for either pathogen, with the exception of E. coli O157:H7-inoculated almonds stored at 4°C (decline of 0.09 log CFU/g/month). At 24°C, initial rates of decline were 0.20, 0.60, and 0.71 log CFU/g/month on almonds and 0.15, 0.35, and 0.86 log CFU/g/month on pistachios for Salmonella, E. coli O157:H7, and L. monocytogenes, respectively, but distinct tailing of the survival curves was noted for both E. coli O157:H7 and L. monocytogenes.