AN ABSTRACT OF THE THESIS OF

Transcription

1

2 AN ABSTRACT OF THE THESIS OF Melissa M. Sales for the degree of Master of Science in Food Science and Technology presented on July 5, 03 Title: An Evaluation of Blackberry Harvest Sanitation and the Ability of Foodborne Pathogens to Survive in Blackberry Products Abstract approved: Mark A. Daeschel Blackberries, genus Rubus, are an important Oregon agricultural commodity. In their fresh and processed forms, they offer many health benefits and may be able to help Americans better achieve fruit consumption recommendations because of convenience and pleasant sensory qualities. However, the susceptibility of blackberry products to contamination with bacterial pathogens of human health concern is unknown. Blackberries have never directly been implicated in a food safety incident; however, raspberries, also in the Rubus genus, have been the vehicle for hepatitis A, norovirus, and Cyclospora cayetanensis outbreaks. Furthermore, strawberries were recently the source of an Escherichia coli O57:H7 outbreak in Oregon. To better understand the potential for microbial pathogen contamination and the ability of these microorganisms to survive or grow in blackberry products, several studies were conducted. Fresh berries from the Obsidian and Triple Crown cultivars were evaluated at different harvest periods for the aerobic plate count, coliforms, yeasts, and molds to establish a baseline microbial population. Environmental samples

3 were taken from a clean mechanical harvester and then from the same harvester that had been intentionally left soiled with berry harvest debris to determine the impact of harvester microbial quality. Samples from Marion and Black Diamond cultivars were hand harvested and evaluated for E. coli O57:H7 and Salmonella spp. by rapid detection methods via the NEOGEN Reveal.0 systems. Fresh, wild Himalaya blackberries and frozen blackberries from the Triple Crown cultivar were spot inoculated with Escherichia coli O57:H7, Salmonella Typhimurium, Listeria monocytogenes, and Staphylococcus aureus to determine the ability of these microorganisms to survive on the berry surface. Himalaya samples were stored for 3 days at ambient temperatures and Triple Crown for 6 months at -3.3 C. Lastly, juice and wine were made from Marion and Black Diamond purees. The juices and wines were used for pathogen survival studies using the aforementioned microorganisms to better understand what constituents of blackberries may contribute to bactericidal effects, as well as the survival patterns in these products. Aerobic plate counts for Obsidian and Triple Crown cultivars ranged from log CFU/g of berry with later harvests tending to have higher values. Triple Crown mid-late harvest samples were significantly higher than the early harvest samples (p = 0.005). Yeasts and molds ranged from log CFU/g of berry with later harvests having significantly higher values for both cultivars (p = Obsidian ; p <0.00 Triple Crown ). Coliforms were detected in Obsidian mid-harvest and Triple Crown earlyharvest samples at.0 and.40 log CFU/g of berry, respectively. The aerobic plate counts measured from the clean and intentionally soiled mechanical harvester were not

4 significantly different. Escherichia coli O57:H7 and Salmonella spp. were not detected using rapid detection methods in evaluated Marion and Black Diamond samples. Escherichia coli O57:H7 was not detectable in fresh or frozen inoculated samples. Salmonella Typhimurium was detected in frozen samples with.95 and 3. log reductions. Listeria monocytogenes was only detected in frozen samples and experienced log reductions.4. Staphylococcus aureus was detectable on every fresh and frozen berry inoculated with log reductions ranging from 0.67 to The greatest reductions occurred with fresh samples. Growth of microorganisms was not observed in any juice or wine samples. Maximum observed survival times in juices ranged from h for L. monocytogenes to 08 h for Salmonella Typhimurium. Maximum survival times in wines were 40 m for both E. coli O57:H7 and Salmonella Typhimurium, and 80 m for both L. monocytogenes and S. aureus. Adding ethanol to juice samples to equal that of their counterpart wines decreased survival time for all microorganisms evaluated by several hours. Increasing the ph of wines by approximately one unit increased the survival time from minutes to hours, and in some cases, days. The overall results suggest that blackberries are not an ideal environment for E. coli O57:H7, Salmonella Typhimurium, L. monocytogenes, and S. aureus to grow. However, these microorganisms may be able to survive depending on the type of blackberry product and its subsequent storage. Many constituents of blackberries may provide bactericidal activity, with organic acids appearing to have the greatest effect.

5 Copyright by Melissa M. Sales July 5, 03 All Rights Reserved

6 An Evaluation of Blackberry Harvest Sanitation and the Ability of Foodborne Pathogens to Survive in Blackberry Products by Melissa M. Sales A THESIS Submitted to Oregon State University in partial fulfillment of the requirements for the degree of Master of Science Presented July 5, 03 Commencement June 04

7 Master of Science thesis of Melissa M. Sales presented on July 5, 03. APPROVED: Major Professor, representing Food Science and Technology Head of the Department of Food Science and Technology Dean of the Graduate School I understand that my thesis will become part of the permanent collection of Oregon State University libraries. My signature authorizes release of thesis to any reader upon request. Melissa M. Sales, Author

8 ACKNOWLEDGEMENTS I would like to express my sincerest gratitude to Mark Daeschel, not only for being an incredibly patient and compassionate mentor, but for also giving me the freedom and opportunities to discover my strengths. For that, I am truly grateful. Thank you to David Bryla, Javier Fernandez-Salvador, Renee Harkins, Angela Tseng, Jooyeoun Jung, George Cavender and Mingyang Lui for all of their time, knowledge, and assistance with collecting samples. I would like to thank everyone that gave me support and advice during my time at OSU, especially Dan Smith, Linda Dunn, Lisbeth Goddik, Joy Waite-Cusic, Brian Yorgey, and Jeff Clawson. Whether you know it or not, you ve made many decisions that I ve had to make a little easier. Thank you to all of the wonderful and supportive friends that I have made at OSU that I have come to consider family. I would especially like to thank Jake Mattson for teaching me not to take myself too seriously and for helping me to become a larger geek than I ever could have become on my own. I would like to thank my significant other, Zak, for his incredible support and for never giving up on me during my academic journey even when I was completely insufferable. I would also like to thank my son, Alex, for being the greatest, most patient kid I could have ever asked for and for giving me a reason to establish higher goals. I only wish that my parents, Karen and Gilbert Sales, could have seen me come this far.

9 CONTRIBUTION OF AUTHORS Dr. Joy Waite-Cusic assisted with microbiological data interpretation. Jacob Mattson assisted with the statistical analysis of harvester and microflora data.

10 TABLE OF CONTENTS Page. Introduction.... Production.... Consumption and Health Benefits....3 Potential microbial Risks Potential Antimicrobial Properties An Evaluation of Blackberry Harvest Sanitation and the Ability of Escherichia coli O57:H7, Salmonella Typhimurium, Listeria monocytogenes, and Staphylococcus aureus to Survive on the Surface of Fresh and Frozen Blackberry Fruit Abstract Introduction Materials and Methods Fresh Field Samples Direct Pathogen Testing Mechanical Harvester Whole Berry Inoculation Data Analysis Results and Discussion Fresh Field Data Direct Pathogen Testing Mechanical Harvester Spot Inoculation Studies Conclusions An Evaluation of the Survival of Escherichia coli O57:H7, Salmonella Typhimurium, Listeria monocytogenes, and Staphylococcus aureus in Marion and Black Diamond Blackberry Juice and Wine... 37

11 TABLE OF CONTENTS (Continued) 3. Abstract Introduction Materials and Methods Juice and Wine Preparation Juice and Wine Properties Culture Preparation Survival Study Procedure Data Analysis Results and Discussion Properties of Purees, Juices, and Wines Survival Study Results Conclusions Overall Conclusions and Future Work Bibliography APPENDICES... 65

12 LIST OF FIGURES Figure Page Figure. Aerobic Count at Various Harvest Times... 7 Figure. Yeasts and Molds at Various Harvest Times... 7 Figure.3 Coliforms at Various Harvest Times... 8 Figure.4 Mechanical Harvester Aerobic Counts... 9 Figure 3. Survival of E. coli O57:H7 in Marion Products Figure 3. Survival of E. coli O57:H7 in Marion Wine Figure 3.3 Survival of E. coli O57:H7 in Black Diamond Products Figure 3.4 Survival of E. coli O57:H7 in Black Diamond Wine Figure 3.5 Survival of Salmonella Typhimurium in Marion Products Figure 3.6 Survival of Salmonella Typhimurium in Marion Wine... 5 Figure 3.7 Survival of Salmonella Typhimurium in Black Diamond Products... 5 Figure 3.8 Survival of Salmonella Typhimurium in Black Diamond Wine... 5 Figure 3.9 Survival of S. aureus in Marion Products Figure 3.0 Survival of S. aureus in Marion Wine Figure 3. Survival of S. aureus in Black Diamond Products Figure 3. Survival of S. aureus in Black Diamond Wine... 54

13 LIST OF TABLES Table Page Table. Cultivar Information... 7 Table. Detection of E. coli O57:H7 and Salmonella spp. in Marion and Black Diamond Cultivars... 8 Table.3 Himalaya inoculated with E. coli O57:H Table.4 Himalaya inoculated with Salmonella Typhimurium... 3 Table.5 Himalaya inoculated with L. monocytogenes... 3 Table.6 Himalaya inoculated with S. aureus... 3 Table.7 Frozen Triple Crown inoculated with E. coli O57:H Table.8 Frozen Triple Crown inoculated with Salmonella Typhimurium Table.9 Frozen Triple Crown inoculated with L. monocytogenes Table.0 Frozen Triple Crown inoculated with S. aureus Table 3. Cultivar Information Table 3. ph and Soluble Solid Content of Blackberry Puree Table 3.3 ph, Soluble Solid Content, and Titratable Acidity of Blackberry Juices Table 3.4 ph, Ethanol Content, and Titratable Acidity of Blackberry Wines Table 3.5 Maximum Observed Survival Times of L. monocytogenes in Marion and Black Diamond Juices and Wines: All Treatments... 5

14 LIST OF APPENDICES Appendix Page Appendix I. Fresh Field Samples Raw Data, Chapter Appendix II. Mechanical Harvester Raw Data and Photos, Chapter Appendix III. Himalaya Raw Data, Chapter... 7 Appendix IV. Triple Crown Raw Data, Chapter Appendix V. Juice Raw Data, Chapter Appendix VI. Juice Variables Raw Data, Chapter Appendix VII. Wine Raw Data, Chapter Appendix VIII. Wine Variable Raw Data, Chapter Appendix IX. Inocula Raw Data, Chapter

15 LIST OF APPENDIX FIGURES Figure Page Figure II. Harvester Location Figure II. Harvester Location Figure II-3 Harvester Location Figure II.4 Harvester Location Figure II.5 Harvester Location Figure II.6 Harvester Location Figure II.7 Harvester Location Figure II.8 Harvester Location

16 LIST OF APPENDIX TABLES Table Page Table I. Aerobic Count Raw Data, Fresh Samples Table I. Yeasts and Molds Raw Data, Fresh Samples Table I.3 Coliforms Raw Data, Fresh Samples Table II. Aerobic Count Raw Data Table III. E. coli O57:H7 on fresh 'Himalaya' Blackberries, Raw Data... 7 Table III. Salmonella Typhimurium on fresh 'Himalaya' Blackberries, Raw Data Table III.3 L. monocytogenes on fresh 'Himalaya' Blackberries, Raw Data Table III.4 S. aureus on fresh 'Himalaya' Blackberries, Raw Data Table IV. E. coli O57:H7 on Frozen 'Triple Crown' Blackberries, Raw Data Table IV. Salmonella Typhimurium on Frozen 'Triple Crown' Blackberries, Raw Data. 76 Table IV.3 L. monocytogenes on Frozen 'Triple Crown' Blackberries, Raw Data Table IV.4 S. aureus on Frozen 'Triple Crown' Blackberries, Raw Data Table V. E. coli O57:H7 in 'Marion' Juice, Raw Data Table V. E. coli O57:H7 in 'Black Diamond' Juice, Raw Data... 8 Table V.3 Salmonella Typhimurium in 'Marion' Juice, Raw Data... 8 Table V.4 Salmonella Typhimurium in 'Black Diamond' Juice, Raw Data Table V.5 L. monocytogenes in 'Marion' Juice, Raw Data Table V.6 L. monocytogenes in 'Black Diamond' Juice, Raw Data Table V.7 S. aureus in 'Marion' Juice, Raw Data... 87

17 LIST OF APPENDIX TABLES (Continued) Table Page Table V.8 S. aureus in 'Black Diamond' Juice, Raw Data Table VI. E. coli O57:H7 in 'Marion' Juice with added ethanol, Raw Data Table VI. E. coli O57:H7 in 'Black Diamond' Juice with added ethanol, Raw Data Table VI.3 Salmonella Typhimurium in 'Marion' Juice with added ethanol, Raw Data Table VI.4 Salmonella Typhimurium in 'Black Diamond' Juice with added ethanol, Raw Data... 9 Table VI.5 L. monocytogenes in 'Marion' Juice with added ethanol, Raw Data... 9 Table VI.6 L. monocytogenes in 'Black Diamond' Juice with added ethanol, Raw Data... 9 Table VI.7 S. aureus in 'Marion' Juice with added ethanol, Raw Data... 9 Table VI.8 S. aureus in 'Black Diamond' Juice with added ethanol, Raw Data Table VII. E. coli O57:H7 in 'Marion' Wine, Raw Data Table VII. E. coli O57:H7 in 'Black Diamond' Wine, Raw Data Table VII.3 Salmonella Typhimurium in 'Marion' Wine, Raw Data Table VII. 4 Salmonella Typhimurium in 'Black Diamond' Wine, Raw Data Table VII.5 L. monocytogenes in 'Marion' Wine, Raw Data Table VII.6 L. monocytogenes in 'Black Diamond' Wine, Raw Data Table VII.7 S. aureus in 'Marion' Wine, Raw Data Table VII.8 S. aureus in 'Black Diamond' Wine, Raw Data Table VIII. E. coli O57:H7 in ph adjusted 'Marion' Wine, Raw Data... 0 Table VIII. E. coli O57:H7 in ph adjusted 'Black Diamond' Wine, Raw Data... 03

18 LIST OF APPENDIX TABLES (Continued) Table Page Table VIII.3 Salmonella Typhimurium in ph adjusted 'Marion' Wine, Raw Data Table VIII.4 Salmonella Typhimurium in ph adjusted 'Black Diamond' Wine, Raw Data 04 Table VIII.5 L. monocytogenes in ph adjusted 'Marion' Wine, Raw Data Table VIII.6 L. monocytogenes in ph adjusted 'Black Diamond' Wine, Raw Data Table VIII.7 S. aureus in ph adjusted 'Marion' Wine, Raw Data Table VIII.8 S. aureus in ph adjusted 'Black Diamond' Wine, Raw Data Table IX. Inocula for Juice Survival Studies Table IX. Inocula for Wine Survival Studies... 08

19 An Evaluation of Blackberry Harvest Sanitation and the Ability of Foodborne Pathogens to Survive in Blackberry Products. Introduction. Production Blackberries belong to the genus Rubus, subgenus Rubus, and are an important agricultural commodity in Oregon (Hummer 00). The United States is the largest producer of blackberries in the world, with 65% of the United States production occurring in Oregon in 005 (Strik and Others 007). Oregon experienced a 5% increase in blackberry production between 995 and 005, and a 6% increase between 005 and 0, a trend that is expected to continue (Strik and Others 007; USDA 0). Oregon is unique in that the majority of cultivated blackberries are trailing types, which are typically machine harvested. Most states and growing regions predominantly cultivate erect and semierect cultivars. Trailing types typically and with exception, do not produce fruit that is firm enough to withstand extensive shipping and handling to be used for fresh market. Blackberries that are machine harvested are typically destined for further processing into products such as individually quick frozen (IQF) berries, purees, jams, jellies, and even wine (Strik and Others 007). More than 95% of Oregon blackberries are destined for processing (Strik and Others 007). In Oregon, the predominant trailing cultivar is Marion and accounted for 5% of harvested blackberries in 0 (USDA 0). Marion is known for its desirable sensory qualities

20 for processing (aroma, acid, and sugar profile) (Strik and Others 007; Du and Others 00). Another trailing cultivar that has gained popularity in Oregon is Black Diamond. Black Diamond has the added benefit of being thornless and is suitable for processed products as well as fresh market due to its firmness. This cultivar may be mechanically or hand-harvested (Strik and Others 007; Du and Others 00). The trailing cultivar Obsidian is mainly grown for hand-harvested fresh market (Finn and Others 005). Triple Crown is also popular in Oregon and is a semierect cultivar. It is hand-harvested, which is typical for erect and semierect types (Strik and Others 007). Trailing types are harvested between June and August, with erect and semierect types extending into October, depending on cultivar (Strik and Others 007; Finn and Strik 008).. Consumption and Health Benefits Consumption of Rubus fruit, stems, and leaves for food and pharmacological purposes is believed to have begun approximately 0,000 years ago (Hummer 00). In Oregon, evidence of Rubus consumption can be dated back to 8000 BCE as determined by carbon dating of materials found at Newberry Crater in Bend, OR (Hummer 00). Ancient Greeks and Romans used nonedible portions of Rubus plants for everything from hair dye to relieving stomach aches. An interesting record of Hippocrates writings suggest using blackberry leaves soaked in wine to apply to wounds for antiseptic purposes, a property later understood to be attributed to their tannin content (Hummer 00). Native populations of North America used Rubus for many ailments related to

21 3 female reproductive function, including childbirth and menstrual cramps (Hummer 00). In modern times, blackberries still have much to offer human health. Obesity is understood to be an epidemic in the United States. Fortunately, rates of obesity among American adults appear to have reached a plateau, with obese men comprising 35.5% of the adult population and obese women, 35.8% (Flegal and Others 0). Obesity is associated with metabolic syndrome which is comprised of a set of risk factors that increase an individual s likelihood of developing chronic diseases such as diabetes and cardiovascular disease (Gropper and Others 009). One strategy to help combat obesity in the United States is to increase overall consumption of fruits and vegetables. Healthy People is a program under the Department of Health and Human Services. A primary goal for the Healthy People 00 objective period was to increase the consumption of fruits and vegetables among Americans. Specifically, the goal was to increase the percentage of individuals over the age of years that consume at least servings of fruit a day to 75% and at least 3 servings a day to 50% (CDC 00). Figures reported in 009 showed that the goal was far from being reached with only 3.5% of adults consuming at least servings of fruit daily, a significant decline from 000, and 6.3% consuming at least 3 servings of vegetables daily (CDC 00). Figures for Oregon were slightly better with 33% consuming at least servings of fruit daily and 30.5% consuming at least 3 servings of vegetables daily (CDC 00). To try to improve consumption of fruits and vegetables in the United States, the Centers for Disease

22 4 Control and Prevention (CDC) has released a guidance document that outlines a 0 point strategy to increase consumption. The overall approach to the strategy is to increase availability and visibility of fruits and vegetables while decreasing cost barriers and increasing consumer education about where their food is coming from. The CDC feels that this can be achieved through such practices as increasing farmers markets, expanding community agriculture programs, and encouraging the availability of fresh produce in markets that tend to not provide as much choice (CDC 0). Blackberries could serve to help increase overall consumption of fruit, especially when considering the portability and convenience of many blackberry products: for example, fresh, frozen, and freeze dried fruit and blackberry juice. Blackberries also have pleasant sensory properties that make them attractive to consumers across age categories (Du and Others 00). Nutritionally, they are low in calories, a good source of fiber, and contain vitamins and minerals (USDA Nutrient Database). Many components of blackberries could assist with weight management. Blackberries contain pectin, a soluble fiber. Consumption of pectin has been associated with increased satiety, which may reduce energy overconsumption (Perrigue and Others 00). Soluble fiber has also been associated with a reduction in blood lipids, which reduces the risk for developing cardiovascular disease (Jenkins and Others 00; Gropper and Others 009). Aside from weight loss and management, soluble fiber can be considered a prebiotic, since it is fermentable in the gut, promoting healthy intestinal flora (Gropper and Others 009).

23 5 Blackberries are also known to be high in anthocyanin content and antioxidant capacity (Puupponen-Pimia and Others 005; Tsuda 0). Consumption of anthocyanins has been demonstrated to assist with obesity and diabetes prevention and or control (Tsuda 0). Cyanidin-3-glucoside (C3G) is an anthocyanin found in blackberries. Consumption of this chemical has been shown to decrease body fat accumulation in mice even when high fat meals were consumed. The mechanism proposed involved a decrease in lipid synthesis in the liver and white adipose tissues (Tsuda 0). Cyanidin-3-glucoside was also found to upregulate the expression of adiponectin, which is associated with increased insulin sensitivity (Tschriter 003; Tsuda 0). Metabolic syndrome is also associated with increased inflammation, and anthocyanins were associated with the reduction of the inflammatory markers, tumor necrosis factor-α and monocyte chemotactic protein- (Sasaki and Others 007; Tsuda 0). Gallic acid is a metabolite of anthocyanins and is a powerful antioxidant (Yao and Others 004). Antioxidants can scavenge free radicals and prevent damage to proteins, lipids, and DNA within the human body. Gallic acid is present in blackberries, but is also a byproduct of anthocyanin metabolism by microorganisms in the human gastrointestinal tract (Aura and Others 005; Kempler and Humpf 005). Blackberries can be a healthy addition to the diet considering the numerous ways that they can positively impact human health.

24 6.3 Potential microbial Risks To our knowledge, blackberries have never been implicated in a food safety incident; however microbial safety risks may exist. While chemical and physical hazards are always a possibility (e.g., improperly applied pesticides, thorns), by and large, the most important potential hazards are biological in nature. Sources of microbial contamination could include, but are not limited to: birds, insects, and other animals in the field, microorganisms in soil, environmental molds, contaminated irrigation water, and humans practicing poor hygiene. Machine harvested fruit has less human contact than hand-harvested fruit. However, the impact of proper sanitation of mechanical harvesters is unknown. Additionally, it is possible for the harvester to disturb soil resulting in aerosols, an issue if the soil were contaminated with pathogenic microorganisms. This would be of particular concern when using improperly composted manure in the production system, as it has been found to be able to harbor E. coli for extended periods of time (Beuchat 00). Fruit that is destined for the fresh market is hand harvested. The safety of this fruit depends on the hygiene practices of the workers. Hand-harvested fruit is packed directly into the plastic clamshell containers that are sold in fresh markets. The fruit does not undergo any washing or microbial decontamination step before reaching the consumer. The clamshells are handled in a way to prevent them from becoming

25 7 contaminated, using boxes on elevated carts to store them so that they do not touch the ground (Strik and Others 007). Additionally, fruit that is visibly moldy or in contact with the ground is not harvested (personal observations). Safety becomes an even larger concern when considering that during the offseason, blackberries for fresh market and frozen use are often imported to the U.S. from Mexico and Guatemala (Strik and Others 007). Guatemalan raspberries (genus, Rubus) were found to be the source of several Cyclosporiasis outbreaks, eventually leading to them having a Detention Without Physical Examination (DWPE) Import Alert placed on them (Herwaldt and Others 997; FDA 0). This has resulted in Guatemalan raspberries being prohibited from import during the months of March-August every year (FDA 0). The DWPE only applies to raspberries; blackberries from Guatemala are allowed to be imported. Frozen raspberries have also been implicated in hepatitis A and norovirus outbreaks (Falkenhorst and Others 005; Reid and Robinson 987). At the time of this writing, frozen berries are being implicated in a hepatitis A outbreak that is affecting Denmark, Norway, Sweden, and Finland. At this time, the type of berry and origin are unknown (Gillesberg Lassen 03). The same strain of hepatitis A has surfaced in an outbreak currently affecting the United States. Again, a frozen berry blend is believed to be the source with the investigation focusing on imported pomegranate seeds that were part of that blend (CDC 03; Terry 03). Although not in the Rubus genus, Strawberries (genus Fragaria) are another berry that has been associated with foodborne illness. They were found to be the vehicle for a 0

26 8 outbreak of E. coli O57:H7 in Oregon that left one person dead and several more ill: an outbreak later found to be due to deer feces being present in the strawberry fields (FDA 0; Stone 0; Terry 0). Raspberries and blackberries have a similar anatomy. They are comprised of small structures called drupelets that are held together by small hairs that intertwine (Bowling 000). Both have what is called a receptacle, where the fruit is attached to the plant. Blackberries detach above this and are harvested with the receptacle still in place, which is seen as the white center when looking at the top of a blackberry. Raspberries leave the receptacle behind when harvested, which leaves a hollow center in the fruit (Bowling 000). The anatomy of the blackberry could potentially give microorganisms of health concern the ability to contaminate and survive on fruit, even if later washing was implemented. The berry skin may not be an ideal location for microorganisms to adhere due its hydrophobic nature; however, the many crevices that exist where drupelets meet, and the receptacle, may provide niches for microorganisms to survive (Bowling 000). This would be a concern largely for fresh market blackberries, but also exists for the minimally processed IQF blackberries, since many microorganisms can survive freezing. Other blackberry products may not be as susceptible due to the processing method. Puree is often pasteurized, and then frozen, and juice is required to be pasteurized before reaching the consumer. Jams and jellies are heated to an extent that would kill any vegetative bacterial cells. The hostile environment of fermentation

27 9 during the berry wine making process would be unfavorable to bacterial pathogens, particularly when coupled with the dominance of yeast in the system. The concern with these further processed products is the potential for pathogenic bacteria to be able to survive if post processing contamination occurred. There is evidence that frozen purees and juice concentrates from various fruits can support the survival of E. coli O57:H7, Salmonella spp., and Listeria monocytogenes for at least weeks (Oyarzabal and others 003). Additionally, E. coli O57:H7, Salmonella spp, and Listeria innocua were able to survive on the surface of frozen strawberries and in strawberry juice over various periods of storage time (Duan and Zhao 009; Knudsen and Others 00). Aside from bacterial contamination, blackberries, like other agricultural products, have an opportunistic microflora that will spoil harvested fruit thus, limiting shelf-life. The naturally occurring microflora largely consists of yeasts and molds and will cause spoilage over time. Many fungal species can survive well in the low ph environment of blackberries and consume the organic acids present, which can lead to an increase in the ph of their environment (Tournas and Katsoudas 005). Depending on the rate at which that occurred, it could be possible for the ph to rise enough to allow some of the more acid tolerant bacterial pathogens to be able to perhaps survive (e.g., E. coli and Salmonella spp.). In retail samples taken from the Washington, D.C. metro area, 00% of blackberries had some sort of fungal population with 78% containing Botrytis cinerea, 33% containing Cladosporium, % Fusarium, % Penicillin, and % Rhizopus (Tournas and Katsoudas 005).

28 0.4 Potential Antimicrobial Properties Blackberries and blackberry products have many components that may have various degrees of antimicrobial activity. Blackberries naturally have a low ph that ranges from 3.0 to 4. (Beuchat 987). The organic acids present in blackberries are predominantly citric and malic acids. These organic acids can exist in a variety of ratios, depending on cultivar (Fan-Chiang 999). Organic acids exist in their undissociated form at low ph, which allows them to permeate the plasma membrane of microorganisms (Brul and Coote 999). Once inside the cell, the organic acids can then dissociate and cause a buildup of protons, eventually becoming toxic to the microorganism (Brul and Coote 999). Essential oils have been found to be effective against some microorganisms. For example, geraniol, found in blackberries (Du and Others 00), has been shown to have antimicrobial activity against Salmonella enterica, E. coli O57:H7, and L. innocua (Friedman and Others 004; Raybaudi-Massilia and Others 006). Blackberries are a rich source of phenolic compounds including phenolic acids, anthocyanins, and ellagitannins (Puupponen-Pimia and Others 00; Ruyi and Others 00). Phenolic compounds may also contribute to antimicrobial activity. Phenolic acids have been demonstrated to be effective against Gram-negative bacteria (Puupponen-Pimia and Others 00). Some of the phenolic acids and flavonoid compounds found in blackberries include: gallic acid, caffeic acid, coumaric acid, ferulic

29 acid, ellagic acid, catechin, quercetin, and myricetin (Sellappan and Others 00; Bilyk and Sapers 986). While tannins may not directly act against microorganisms, they may act indirectly by binding substances needed for their survival such as nutrients, and interfering with microbial extracellular enzyme functions (Puupponen-Pimia and Others 005). Nohynek and Others (006) found that extracts from berries in the genus Rubus, cloudberry and raspberry, were able to permeabilize and in some cases disintegrate the outer membrane of Gram-negative bacteria. They attributed this to a synergistic effect of ph and phenolic compounds, noting that gallic acid appeared to have the greatest effect which they believed was due to its ability to chelate divalent cations from the membrane, destabilizing it (Nohynek and Others 006). This trait is also shared by citric acid (Brul and Coote 999). In addition, the extracts were found to have an inhibitory effect against several Gram-positive microorganisms, including Staphylococcus aureus. Various wines, and even their unfermented juices, have been demonstrated to have bactericidal properties. Trials with chardonnay juice resulted in bacterial populations of E. coli O57:H7 and Salmonella Typhimurium being reduced to undetectable levels between 3- days, and the corresponding wine between 0-57 minutes; pinot noir juice between 0-6 days, and 0-60 minutes in the wine (Just and Daeschel 003). Another study with unknown grape varieties found that white and red wines resulted in the reduction of E. coli O57:H7 and Salmonella Enteritidis to undetectable levels within 30 minutes (Sugita-Konishi and Others 00). Interestingly, they did not associate the ethanol content of the wine with the lethality observed due

30 to a solution of 4% ethanol in phosphate-buffered saline not resulting in any reduction in the bacterial populations over a 60 minute period. That conclusion may have neglected to account for the synergistic effect of ethanol, organic acids, and phenolic compounds acting within the same system. Ethanol is known to have a solubilizing effect on the membranes of bacteria, which may allow for easier permealization of other constituents that are toxic to the bacterial cell (Willey and Others 008). Blackberries are harvested with a soluble solid content far below what is standard for wine grapes (~0- Brix versus 4 Brix) (Du and Others 00; Zoecklein and Others 990). Blackberry wines would naturally result in a lower ethanol content than grape wines due to the reduced sugar content unless supplemented. The effect on lethality that this would have in a blackberry wine fermented without sugar supplementation has not, to our knowledge at this time, been previously investigated. While the ethanol content of blackberry wines would be lower, the ph would not be dissimilar, yet the phenolic content would vary widely, depending on fruit type and cultivar.

31 . An Evaluation of Blackberry Harvest Sanitation and the Ability of Escherichia coli O57:H7, Salmonella Typhimurium, Listeria monocytogenes, and Staphylococcus aureus to Survive on the Surface of Fresh and Frozen Blackberry Fruit 3 Melissa M. Sales and Mark A. Daeschel Oregon State University Department of Food Science and Technology, Corvallis OR 9733 To be submitted to: Journal of Food Science Institute of Food Technologists 55 W. Van Buren Ste 000 Chicago Ill 60607

32 . Abstract Blackberries, genus Rubus, are an important Oregon agricultural commodity. In their fresh and processed forms, they offer many health benefits and may be able to help Americans better achieve fruit consumption recommendations because of convenience and pleasant sensory qualities. However, the susceptibility of blackberry products to contamination with bacterial pathogens of human health concern is unknown. Blackberries have never directly been implicated in a food safety incident; however, raspberries, also in the Rubus genus, have been the vehicle for hepatitis A, norovirus, and Cyclospora cayetanensis outbreaks. Furthermore, strawberries were recently the source of an Escherichia coli O57:H7 outbreak in Oregon. 4 To better understand the potential for microbial pathogen contamination and the ability of these organisms to survive or grow in blackberry products, several studies were conducted. Fresh berries from the Obsidian and Triple Crown cultivars were evaluated at different harvest periods for the aerobic plate count, coliforms, yeasts, and molds to establish a baseline microbial population. Environmental samples were taken from a clean mechanical harvester and then from the same harvester that had been intentionally left soiled with berry harvest debris to determine the impact of harvester sanitation. Samples from Marion and Black Diamond cultivars were hand harvested and evaluated for E. coli O57:H7 and Salmonella spp. by rapid detection methods via the NEOGEN Reveal.0 systems. Fresh, wild Himalaya blackberries and frozen blackberries from the Triple Crown cultivar were spot inoculated with E. coli O57:H7,

33 5 Salmonella Typhimurium, Listeria monocytogenes, and Staphylococcus aureus to determine the ability of these microorganisms to survive on the berry surface. Himalaya samples were stored for 3 days at ambient temperatures and Triple Crown for 6 months at -3.3 C. Aerobic plate counts (APC) for Obsidian and Triple Crown cultivars ranged from log CFU/g of berry with the late harvest Triple Crown samples having a significantly higher APC than early harvest samples (p = 0.005). Yeasts and molds ranged from log CFU/g berry with both cultivars having significantly higher counts at later harvest times (p = Obsidian ; p < 0.00 Triple Crown ). Coliforms were detected in Obsidian mid-harvest and Triple Crown early-harvest samples at.0 and.40 log CFU/g of berry, respectively. The overall aerobic plate counts measured from the mechanical harvester were not affected by machine cleanliness. Escherichia coli O57:H7 and Salmonella spp. were not detected using rapid detection methods in evaluated Marion and Black Diamond samples. Escherichia coli O57:H7 was not detectable in fresh or frozen inoculated samples. Salmonella Typhimurium was detected in frozen samples with.95 and 3. log reductions. Listeria monocytogenes was only detected in frozen samples and experienced log reductions.4. Staphylococcus aureus was detectable on every fresh and frozen berry inoculated with log reductions ranging from 0.67 to 3.48, the greatest reductions occurring in fresh samples.

34 6 The overall results suggest that blackberries are not an ideal environment for E. coli O57:H7, Salmonella Typhimurium, L. monocytogenes, and S. aureus to grow. However, these microorganisms may be able to survive depending on the type of blackberry product and its subsequent storage.. Introduction Blackberries, genus Rubus, are an important agricultural commodity in Oregon. In 0, 4,,760 kg of cultivated blackberries were produced in Oregon and valued at over 43 million USD (USDA 0). There are three types of blackberries plants: trailing, erect, and semierect. Oregon is unique in that the majority of cultivated blackberries in this state are trailing cultivars, most notably Marion, whereas many other growing regions cultivate erect and semierect types (Strik and Others 007; USDA 0). Trailing cultivars are typically machine harvested for processed markets, while erect and semierect cultivars are hand-harvested for fresh market. There are exceptions, and the popular Black Diamond cultivar is an example. Black Diamond is a trailing cultivar, but has the benefit of being thornless and produces fruit firm enough to be able to withstand shipping and handling required for fresh market. It can be mechanically or hand-harvested depending on its intended use (Finn and Others 005; Strik and Others 007). Machine harvested fruit is most often destined for further processing into individually quick frozen (IQF) berries, jams, purees, juices, and even wines (Strik and Others 007). The trailing cultivar Obsidian is hand-harvested for fresh market (Finn and Others 005). Triple Crown is a popular semierect cultivar in

35 7 Oregon and is hand harvested for fresh market (Finn and Strik 008; Strik and Others 007). Himalaya is considered an invasive species, but is still popular for noncommercial harvesting in Oregon (Finn and Strik 008). Table. Cultivar Information Cultivar Type Farming Method Marion Trailing Certified Organic Black Diamond Trailing Certified Organic Obsidian Trailing Certified Organic Triple Crown Semierect Certified Organic Himalaya N/A (similar to semierect) Wild Blackberries are harvested from June through October, depending on the cultivar and type, with each cultivar having a fruiting season of 3-6 weeks (Finn and Strik 008; Strik and Others 007). When considering the value of the Oregon blackberry harvest combined with the short fruiting season, the economic impact of a food safety recall could devastate the industry. The season is of such short duration that economic recovery following a recall would be very unlikely. Blackberries have never been implicated in a food safety incident; however, other berry fruits have. Frozen raspberries, also in the Rubus genus, have been the source of Cyclospora cayetanensis, norovirus, and hepatitis A outbreaks (Ho and Others 00; Reid and Robinson 987; Sarvikivi and Others 0). The anatomy of raspberries is similar enough to blackberries to warrant concern of blackberries being susceptible to similar contamination. Currently, frozen berries are believed to be the source of a hepatitis A outbreak in Denmark, Norway, Sweden, and Finland (Gillesberg and Lassen

36 8 03). The same strain of hepatitis A has surfaced in an outbreak affecting the United States, again a frozen berry blend believed to be the source with the investigation focusing on imported pomegranate seeds that were included in the blend (CDC 03; Terry 03). Although not in the Rubus genus, fresh strawberries (genus Fragaria) were the source of an outbreak of Escherichia coli O57:H7 in Oregon that left one dead and several others ill, demonstrating that other berry fruits are vulnerable to contamination as well (FDA 0; Terry 0). The outbreaks associated with raspberries and other berry blends have in common that humans are the reservoir for hepatitis A, Cyclospora cayetanensis, and norovirus. This suggests that berries may be susceptible to contamination with other causal agents of foodborne illness that humans are known to be carriers of. These could include Staphylococcus aureus, pathogenic strains of Escherichia coli, and Salmonella spp. Thus, poor hygiene of workers involved with harvesting and/or handling fruit could result in contamination of blackberries. Other sources of contamination are possible. Birds are known reservoirs of Salmonella. Flies have been demonstrated to be able to transmit E. coli O57:H7 to damaged apples (Janisiewicz and Others 999). Deer were determined to be the source of the E. coli O57:H7 outbreak associated with strawberries (Stone 0; Terry 0). Additionally, contaminated soil or irrigation water, and perhaps even the harvesting equipment could be sources of contamination.

37 9 Blackberries are a portable, convenient fruit with a pleasant aroma and flavor that could encourage individuals in the United States to consume the recommended daily servings of fruits (Du and Others 00). Nationally, only 3.5% of Americans consume at least two servings of fruit daily (CDC 00). Increasing fruit and vegetable consumption as a strategy to combat the obesity epidemic is a goal of the Centers for Disease Control (CDC 0). These goals increase the need to understand how foodborne pathogens behave in various fruit and vegetable environments to avoid increased consumption having unintended consequences. Understanding how certain microorganisms behave in the blackberry environment will help the blackberry industry better employ prevention strategies, understand the risk involved with their product, and to be prepared to quickly respond in the event of any contamination. Blackberries are known to naturally have a diverse microflora of yeasts and molds on their surface (Tournas and Katsoudas 005); however, not much information that quantifies the microflora is available. The relationship that these yeasts and molds may have with bacterial pathogens, if present, is not well understood. It has been suggested that they may inhibit pathogenic species by dominating the environment or may even encourage growth by making nutrients available and altering the surrounding ph (Beuchat 00; Tournas and Katsoudas 005). Blackberries have chemical constituents that could affect the ability of foodborne pathogens to survive. Organic acids are abundant in blackberries, resulting in the fruit having a low ph; conditions known to prevent growth and cause bacterial

38 0 death in many species (Beuchat 987; Brul and Coote 999; Jay and Others 005). Blackberries are also a rich source of phenolic compounds, many of which have been found to be bactericidal. Phenolic acids have been shown to be effective against Gramnegative bacteria and tannins are known to bind substances the bacterial cells need to survive (Puupponen-Pimia and Others 005, 007). Furthermore, extracts from other berries in the Rubus genus have been found to be antagonistic toward Gram-negative and Gram-positive bacteria (Nohynek and Others 006). To better understand the microflora associated with blackberries and how pathogenic bacteria may behave in whole berries, three primary goals for this study were established: ) establish quantifiable information about the microflora populations on blackberries; ) evaluate the sanitation of mechanical harvesters; and 3) evaluate the ability of foodborne pathogens to survive on fresh and frozen blackberries..3 Materials and Methods.3. Fresh Field Samples Samples of Obsidian and Triple Crown blackberries at their early, mid, mid/late, or late harvest periods were obtained from Riverbend Organic Farms in Jefferson, OR and evaluated for the aerobic plate count (APC), yeast and molds, and coliforms. Samples arrived in plastic clamshell containers and were analyzed the same day as harvest. Two clamshells were obtained for each cultivar at each harvest point evaluated. A 50 g sample taken from each clamshell was homogenized at 00 RPM for minutes (Stomacher 400 Circulator; Seward Laboratory). Homogenized samples were

39 immediately serially diluted in sterile Butterfield s phosphate buffer (referred to as phosphate buffer for simplicity) and plated on Count Agar (PCA;Difco), Dichloran Rose Bengal Chloramphenicol Agar (DRBC; EMD), and Violet Red Bile Agar (VRB; Cole- Parmer Instrument Company). Count Agar plates were incubated for 48h at 30 C, DRBC plates for 48-7h depending on growth at 30 C, and VRB plates for 4h at 35 C..3. Direct Pathogen Testing Early, mid, and late harvest samples of Marion and Black Diamond blackberries were collected by hand-harvesting at the North Willamette Research and Extension Center (NWREC) in Aurora, OR. Berries were separated by cultivar and placed in plastic bags. They were transported under refrigeration to Oregon State University and evaluated the same day. Samples were evaluated for Salmonella spp. and E. coli O57:H7 using modified protocols for the NEOGEN Reveal.0 Salmonella and NEOGEN Reveal.0 E. coli O57:H7 Complete Systems for rapid detection. From each cultivar, 5 g samples were placed in a sterile 500 ml media bottle and covered with 00 ml of phosphate buffer. This was conducted in duplicate for each cultivar. The berries were gently agitated in the solution at room temperature. For Salmonella testing, one bottle of REVIVE recovery medium was added to a sterile 500 ml media bottle and mixed with 00 ml of sterile deionized (DI) water at 4-43 C. From the berry sample, 5 ml of phosphate buffer was removed, added to the REVIVE mixture and then incubated for 4 h at 36 C. The contents of one bottle of provided xrv selective media was added to 00 ml of sterile DI water at 37 C and then

40 held at 4 C until ready to use. After the solution containing the sample completed its 4 h incubation, the xrv solution was added to it and then incubated at 4 C for an additional 4 h. Using the provided sterile dropper, 00 µl were transferred to a sample cup. The test strip was inserted and allowed to stand for 5 m at room temperature before reading. For E. coli testing, one bottle of Reveal.0 E. coli O57:H7 media was placed in a sterile 500 ml media bottle and mixed with 35 ml of sterile DI water at 4 C. From the berry sample, 65 ml of the phosphate buffer was removed, added to the media, and incubated at 4 C for 0 h. After incubation, 00 µl was transferred to the test cup provided. A drop of provided promoter agent was added and then incubated for an additional 5 m. The test strip was placed in the sample cup while it was still in the incubator and was read after 5 m. All media and reagents used were included in the test kits and experiments were conducted in duplicate. All incubation periods occurred with media bottles loosely lidded to allow for air exchange. Efficacy of the NEOGEN Reveal systems was verified using pure culture spiked berry samples and negative controls..3.3 Mechanical Harvester Sterile swabs were used to take 8 environmental samples from a clean mechanical harvester (over-the-row Littau Harvester; Stayton,OR) at the NWREC just prior to harvesting Marion blackberries. Locations sampled were photographed for future reference and can be viewed in Appendix II. Swabs were moistened in phosphate

41 3 buffer prior to location sampling. Once sampled, the swabs were agitated in culture tubes containing 4 ml of phosphate buffer. Samples from the tubes were serially diluted and plated on PCA. s were incubated for 48 h at 30 C. The mechanical harvester remained intentionally unwashed after the Marion harvest and post-harvest environmental samples were taken 48 h after the initial samples. The same 8 locations were sampled. Sample handling and plating occurred in the same manner as the initial sampling..3.4 Whole Berry Inoculation.3.4. Culture Preparation Cultures of Escherichia coli O57:H7 (ATCC 43894), Salmonella Typhimurium (ATCC 408), Listeria monocytogenes (Scott A), and Staphylococcus aureus (general food isolate, 648 in OSU Culture Collection) were used for spot inoculation studies. All cultures used were from the culture collection at Oregon State University, Department of Food Science and Technology. Cultures maintained in Trypticase Soy Broth (TSB) were transferred to fresh TSB and incubated for 4h at 35 C (BBL Trypticase Soy Broth, BD). One hundred microliters from each were again transferred to fresh TSB and incubated for 8 h at the same temperature. These cultures were then serially diluted in phosphate buffer and used for inoculation. Enumeration was determined by serial dilution on Trypticase Soy Agar (TSA) (Bacto Agar, BD; BBL Trypticase Soy Broth, BD).

42 .3.4. Fresh Berry Samples Himalaya blackberries were obtained from Bald Hill Vineyard in Corvallis, OR and kept at 8 C. Inoculation occurred the day following harvest. Individual berries were placed on a sterile cap, which were then placed in a Nalgene tray. Berries were then inoculated with 0 µl aliquots of a single bacterial strain in 5 locations for a total of 50 µl per berry ( log CFU/g of berry). Inoculated berries were allowed to dry overnight under a biosafety hood. After drying, the tray containing the berries was moved to an incubator held at room temperature and rotated several times per day for 3 d. 4 Each berry was then placed in a beaker, had 0 ml of phosphate buffer added to it and then gently agitated. This rinse solution was then serially diluted and plated on TSA. All plates were incubated at 35 C for a period of 6 h to allow injured cells an opportunity to recover (Mahmoud and Others 00). Using the weight of the berry sample, enough phosphate buffer was added to the berry/phosphate sample to equal a :0 dilution and then homogenized at 30 RPM for 30 s (Stomacher 400 Circulator, Seward). The resulting mixture was serially diluted and plated on TSA, allowing for the same recovery period as described above. After the recovery period, a layer of appropriate selective media was placed over the TSA: Sorbitol MacConkey Agar(SMAC; Difco) for E. coli O57:H7; Xylose Lysine Deoxycholate Agar (XLD; Difco) for Salmonella Typhimurium; Oxford Listeria Agar Base with Oxford Listeria Selective Supplement (EMB

43 5 and Sigma Aldrich) for L. monocytogenes; and Mannitol Salt Agar (MSA; BBL) for S. aureus Frozen Berry Samples Triple Crown blackberries were hand harvested from Riverbend Organic Farms and kept at 8 C. Inoculation occurred the day following harvest. Individual berries were placed in sterile glass jars and inoculated with a single bacterial strain ( log CFU/g of berry). Berries were inoculated with 5 µl aliquots in 0 locations for a total of 50 µl per berry. Berries were allowed to remain under a biosafety hood until visibly dry, covered with sterilized foil, and frozen at -3.3 C for 6 months. For evaluation, each berry was removed from the freezer just prior to analysis. Evaluation occurred in the same manner as described above for fresh berry evaluation with the following modification: initial rinse with 0 ml phosphate buffer occurred in the glass jar containing the berry, after which the berry was gently removed with forceps and placed into a filtered stomacher bag..3.5 Data Analysis All data were analyzed using Microsoft Excel with the exception of analyses of the aerobic plate count, yeasts and molds, and mechanical harvester aerobic plate counts, which were evaluated by hypothesis tests with Statgraphics Centurion..4 Results and Discussion.4. Fresh Field Data The aerobic plate counts (APC) for Obsidian and Triple Crown cultivars at their early, mid, mid/late, or late harvest periods ranged from 3.5 to 4.6 log CFU/g of berry

44 6 (Figure.). Counts were higher for both cultivars at their later harvest time, with Triple Crown mid/late harvest having significantly higher APC (p = 0.005). Figure. shows that yeasts and molds ranged from 3.0 to 4.73 log CFU/g of berry, with Obsidian late harvest having the highest values. Later harvests for both cultivars were observed to have significantly higher yeast and mold counts than earlier harvests (p = Obsidian ; p < 0.00 Triple Crown ). Caution should be taken when interpreting these findings considering that the summer of 0, when these samples were taken, experienced an exceptional amount of rain (Oregon Climate Service 0). This could have resulted in lower values than normal if microorganisms were being washed off fruit by the rain, or higher values if the increased moisture was providing better growth/survival conditions. Published values for the aerobic plate count, yeasts, and molds for blackberries were not found in a search of the literature; however, our data were consistent with values obtained from fresh strawberry rinse water (Jensen and others 0). Coliforms were detected in Obsidian mid harvest and Triple Crown early harvest samples at.0 and.40 log CFU/g of berry, respectively (Figure.3). The detection limit for the experimental design was 0.70 log CFU/g. Follow up tests were not conducted to determine if these coliforms were fecal in origin.

45 Yeasts and Molds Log CFU/g APC Log CFU/g Figure. Aerobic Count at Various Harvest Times for 'Obsidian' and 'Triple Crown' Cultivars. Bars indicate the range (high and low values); weeks are for Obsidian Mid Harvest: Week 9 Obsidian Late Triple Crown Early Harvest: Week 30 Harvest: Week 34 Triple Crown Mid/Late Harvest: Week 35 Figure. Yeasts and Molds at Various Harvest Times for 'Obsidian' and 'Triple Crown' Cultivars. Bars indicate the range (high and low values); weeks are for Obsidian Mid Harvest: Week 9 Obsidian Late Harvest: Week 30 Triple Crown Early Harvest: Week 34 Triple Crown Mid/Late Harvest: Week 35

46 Coliforms Log CFU/g Figure.3 Coliforms at Various Harvest Times for 'Obsidian' and 'Triple Crown' Cultivars. Dotted line indicates the lower limit for the detection of coliforms; Bars indicate the range (high and low values); weeks are for <0.70 Log CFU/g <0.70 Log CFU/g Obsidian Mid Harvest: Week 9 Obsidian Late Harvest: Week 30 Triple Crown Early Harvest: Week 34 Triple Crown Mid/Late Harvest: Week Direct Pathogen Testing The results for the direct testing of E. coli O57:H7 and Salmonella spp. are shown in Table.. Neither pathogen was detected in any of the samples evaluated. Table. Detection of E. coli O57:H7 and Salmonella spp. in Marion and Black Diamond Cultivars using rapid detection methods Early 7/9/ Mid 7// Late 7/9/ Marion Black Diamond Marion Black Diamond Marion Black Diamond E. coli Rep Negative Negative Negative Negative Negative Negative Rep Negative Negative Negative Negative Negative Negative Salmonella Rep Negative Negative Negative Negative Negative Negative Rep Negative Negative Negative Negative Negative Negative

47 Log CFU/Swab Mechanical Harvester Microbial populations for individual sample locations on the mechanical harvester varied substantially (Figure.4). Little change occurred pre- and post-harvest at locations, 5, 6, and 7. Photographs of sampled locations can be viewed in Appendix II. A.03 log reduction was observed at location 3 post-harvest. Moderate increases were observed at locations, 4, and 8 with location 8 experiencing the largest increase (.0 log). Figure.4 Mechanical Harvester Aerobic Counts: Clean and Unwashed Clean 48 h After Intentionally Left Unwahsed Location on Mechanical Harvester When an overall aerobic plate count for the mechanical harvester was determined using all locations sampled, there was not a significant difference in clean

48 30 and intentionally dirtied values (p = 0.45). These results indicate that the microbial quality of the mechanical harvester does not change 48h post-harvest, even with washing not occurring between sampling periods. This is likely due to the organic acids released from the berries during harvest combined with UV radiation inhibiting growth (Brul and Coote 999; Rico and Others 007)..4.4 Spot Inoculation Studies.4.4. Fresh Berry Results Escherichia coli O57:H7, Salmonella Typhimurium, and L. monocytogenes were not detected in the Himalaya samples 3 d after inoculation (Tables.3-.5). Staphylococcus aureus was recovered from both inoculated berries, but only in the evaluation of the homogenized samples (Table.6). Staphylococcus aureus experienced log reductions of 3.03 and 3.48 in these samples. Table.3 Himalaya inoculated with E. coli O57:H7, evaluated after 3 d held at ambient temperatures Sample Berry Berry Inoculum per gram of 5.74 Log CFU/g 5.69 Log CFU/g berry Rinse Recovery <0.70 Log CFU/g < 0.70 Log CFU/g Homogenized Recovery <.70 Log CFU/g <.70 Log CFU/g

49 3 Table.4 Himalaya inoculated with Salmonella Typhimurium, evaluated after 3 d held at ambient temperatures Sample Berry Berry Inoculum per gram of 5.80 Log CFU/g 5.87 Log CFU/g berry Rinse Recovery < 0.70 Log CFU/g <0.70 Log CFU/g Homogenized Recovery <.70 Log CFU/g <.70 Log CFU/g Table.5 Himalaya inoculated with L. monocytogenes, evaluated after 3 d held at ambient temperatures Sample Berry Berry Inoculum per gram of 6.5 Log CFU/g 6.47 Log CFU/g berry Rinse Recovery < 0.70 Log CFU/g < 0.70 Log CFU/g Homogenized Recovery <.70 Log CFU/g <.70 Log CFU/g Table.6 Himalaya inoculated with S. aureus, evaluated after 3 d held at ambient temperatures Sample Berry Berry Inoculum per gram of 5.33 Log CFU/g 5.8 Log CFU/g berry Rinse Recovery < 0.70 Log CFU/g < 0.70 Log CFU/g Homogenized Recovery.30 Log CFU/g.70 Log CFU/g.4.4. Frozen Berry Results Escherichia coli O57:H7 was not detected in any of the frozen Triple Crown samples (Table.7). Salmonella Typhimurium was detected in of the 7 inoculated berries, and only in the rinse water with.95 and 3. log reductions (Table.8). It

50 3 should be noted that one of the berries was evaluated at 3 months to establish to the procedure protocol. Listeria monocytogenes was detected in 3 of the 7 inoculated berries, all 3 in the homogenized samples with recovery also occurring in one of the rinse samples (Table.9). Log reductions of L. monocytogenes ranged from when detected. Staphylococcus aureus was detected in all 7 inoculated berries, 4 of the 7 in the rinse water and 5 of the 7 in homogenized samples (Table.0). Log reductions of S. aureus ranged from The detection limit for rinse water was 0.70 log CFU/g berry and.70 log CFU/g berry for homogenized samples. Table.7 Frozen Triple Crown inoculated with E. coli O57:H7, evaluated after 6 months stored at -3.3 C Sample Inoculum per Rinse Recovery Homogenized gram of berry Recovery Berry 4.40 Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry 4.50 Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g

51 33 Table.8 Frozen Triple Crown inoculated with Salmonella Typhimurium, evaluated after 6 months stored at -3.3 C (*Berry evaluated after 3 months) Sample Inoculum per Rinse Recovery Homogenized gram of berry Recovery Berry * 3.95 Log CFU/g.00 Log CFU/g <.70 Log CFU/g Berry 3.90 Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g 0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Table.9 Frozen Triple Crown inoculated with L. monocytogenes, evaluated after 6 months stored at -3.3 C Sample Inoculum per Rinse Recovery Homogenized gram of berry Recovery Berry 4.3 Log CFU/g <0.70 Log CFU/g.70 Log CFU/g Berry 4.4 Log CFU/g.00 Log CFU/g.00 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g <.70 Log CFU/g

52 Table.0 Frozen Triple Crown inoculated with S. aureus, evaluated after 6 months stored at C Sample Inoculum per Rinse Recovery Homogenized gram of berry Recovery Berry 3.59 Log CFU/g <0.70 Log CFU/g.8 Log CFU/g Berry 3.6 Log CFU/g 0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g 0.70 Log CFU/g <.70 Log CFU/g Berry Log CFU/g.00 Log CFU/g.70 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g.93 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g.00 Log CFU/g Berry Log CFU/g <0.70 Log CFU/g.70 Log CFU/g 34 These results indicate that the surface of blackberry fruit is not suitable for these bacteria to grow; however, they may survive. Escherichia coli O57:H7 appears to be the least able to survive either on fresh or frozen berry surfaces, whereas S. aureus was the most able to persist. Freezing of the inoculated fruit appears to improve survival, which was to be expected considering that freezing is a method frequently used to preserve microorganisms (Jay and Others 005). Furthermore, E. coli O57:H7 and Salmonella spp. have been found to be able to survive on the surface of frozen strawberries for at least 30 d (Knudsen and Others 00). The aqueous nature of the inocula may have also contributed to the minimal survival observed. Blackberry fruit has a waxy surface that protects it from a variety of stresses, including osmotic pressure (Bowling 000; Shepherd and Griffiths 006). It was

53 35 observed during the inoculation procedure that the inoculum would form beads on the hydrophobic berry surface. This may have resulted in some of the bacterial cells experiencing desiccation as the phosphate buffer evaporated, leaving few viable cells in contact with the actual berry. In harvesting/processing settings, fruit would become contaminated by feces, hands, soil, etc. These other sources of contamination may lead to better bacterial survival on the blackberry surface..5 Conclusions The blackberry surface is not an environment that will allow the growth of E. coli O57:H7, Salmonella Typhimurium, L. monocytogenes, and S. aureus. The microflora that was observed in fresh blackberry samples may offer some protection by outcompeting/antagonizing bacterial pathogens (Beuchat 00). Moreover, E. coli O57:H7 was not observed to survive in any fresh and frozen samples within the experimental detection limits. However, the observation that there was survival by some pathogens reinforces the need for sanitary harvesting conditions and handling to prevent the contamination of blackberries that are destined for fresh market or IQF processing. The mechanical harvester used to harvest trailing type blackberries is not likely to be a source of contamination, particularly if it is exposed to sunlight (Rico 007). Any blackberry residue may actually reduce microbial populations on the harvester by exposure to organic acids; however, these results should not be taken as a suggestion to not wash mechanical harvesters post-harvest. The larger concern lies with the hygiene

54 36 of workers handling the blackberry fruit. Humans are known reservoirs of S. aureus and can also harbor pathogenic strains of E. coli and Salmonella spp. Furthermore, they are also known reservoirs of Hepatitis A, norovirus, and Cyclospora cayetanensis, all of which have been the cause of foodborne illnesses associated with raspberries indicating that humans can contaminate berries in the Rubus genus (Ho and Others 00; Reid and Robinson 987; Sarvikivi and Others 0). Although E. coli O57:H7 and Salmonella spp. were not detected in directly harvested and tested samples, it is important to consider other non-human sources of contamination. Animals, contaminated soil, or irrigation water could result in contamination with a variety of microorganisms of human concern. These sources can be avoided by establishing and maintaining sanitary growing, harvesting, and handling practices.

55 37 3. An Evaluation of the Survival of Escherichia coli O57:H7, Salmonella Typhimurium, Listeria monocytogenes, and Staphylococcus aureus in Marion and Black Diamond Blackberry Juice and Wine Melissa M. Sales and Mark A. Daeschel Oregon State University Department of Food Science and Technology, Corvallis OR 9733 To be submitted to: Journal of Food Science Institute of Food Technologists 55 W. Van Buren Ste 000 Chicago Ill 60607

56 3. Abstract Blackberries, genus Rubus, are an important Oregon agricultural commodity that is frequently processed into various products. These products include individually quick frozen (IQF) berries, jams, juices, purees, and even wines. However, the susceptibility of blackberry products to contamination with bacterial pathogens of human health concern is unknown. Previous studies and food safety incidents have demonstrated that many pathogenic microorganisms are able to survive in purees, juices, and frozen concentrates made from various fruits. 38 Survival studies were conducted in juices and wines made from Marion and Black Diamond purees to understand the potential for pathogenic bacteria to survive if post-processing contamination were to occur. Furthermore, the studies were designed to yield information about what chemical constituents of the juices and wines may contribute to antibacterial activity. Escherichia coli O57:H7, Salmonella Typhimurium, Listeria monocytogenes, and Staphylococcus aureus were evaluated for their ability to survive in these products. Growth of microorganisms was not observed in any juice or wine samples. Maximum observed survival times in juices ranged from h for L. monocytogenes to 08 h for Salmonella Typhimurium. Maximum survival times in wines were 40 m for both E. coli O57:H7 and Salmonella Typhimurium, and 80 m for both L. monocytogenes and S. aureus. Adding ethanol to juice samples to equal that of their counterpart wines decreased survival time for all microorganisms evaluated by several hours. Increasing

57 the ph of wines by approximately one unit increased the survival time from minutes to hours, and in some cases, days. 39 These results demonstrate that blackberry juice and wine do not support the growth of E. coli O57:H7, Salmonella Typhimurium, L. monocytogenes, and S. aureus. However, these microorganisms may be able to survive for various periods of time depending on the type of blackberry product. Many constituents of blackberries may offer bactericidal activity, with organic acids appearing to have the greatest effect. 3. Introduction Blackberries are an important agricultural commodity in Oregon with 65% of the U.S. production occurring in this state (Strik and Others 007). There are three types of blackberry plants: trailing, erect, and semierect. Oregon is unique in that cultivated blackberries are predominantly trailing types, with the primary cultivar being Marion, whereas other growing regions typically grow erect and semierect types (Strik and Others 007; USDA 00). Black Diamond is another popular trailing type in Oregon, but has the added benefit of being thornless and produces firm enough fruit to withstand shipping and handling, resulting in this cultivar being suitable for both mechanical and hand-harvesting (Finn and Strik 008; Strik and Others 007). Table 3. Cultivar Information Cultivar Type Farming Method Marion Trailing Conventional Black Diamond Trailing Certified Organic

58 40 Machine harvested blackberries are most often destined for further processing, whereas hand-harvested fruit is used for fresh market. Machine harvested fruit may be processed into individually quick frozen (IQF) berries, purees, jams, juices, and even wines (Strik and Others 007). Individually quick frozen berries are minimally processed and could be at risk for contamination. Jams, juices, and some purees undergo a thermal processing procedure that would effectively kill any vegetative pathogenic bacterial cells present. These cells would also not be likely to survive fermentation during the blackberry wine making process. For these fermented and thermally processed products, the concern is whether or not bacterial pathogens of human concern could survive in the event of post-processing contamination. The scope of this study will focus on the ability of foodborne pathogens to survive in processed blackberry juices and wines. There is evidence that frozen purees and juice concentrates from various fruits can support the survival of Escherichia coli O57:H7, Salmonella spp., and Listeria monocytogenes for at least weeks (Oyarzabal and Others 003). Previous studies using chardonnay juice resulted in bacterial populations of E. coli O57:H7 and Salmonella Typhimurium being reduced to undetectable levels between 3- days, and the corresponding wine between 0-57 minutes; pinot noir juice between 0-6 days, and 0-60 minutes in the wine (Just and Daeschel 003). Another study with unknown grape varieties found that white and red wines resulted in the reduction of E. coli

59 4 O57:H7 and Salmonella Enteritidis to undetectable levels within 30 minutes (Sugita- Konishi and Others 00). There are several chemical constituents of blackberry juices and wines that may contribute to bactericidal activity. Organic acids are abundant in blackberries, typically as citric and/or malic acids in varying ratios, depending on cultivar (Fan-Chiang 999). This results in the fruit having a low ph, conditions known to prevent growth and cause cellular death in many bacterial species (Beuchat 987; Brul and Coote 999; Jay and Others 005). Blackberries are also a rich source of phenolic compounds, such as phenolic acids, anthocyanins, and ellagitannins (Puupponen-Pimia and Others 00; Wu and Others 00). Phenolic acids have been shown to be effective against Gramnegative bacteria and the tannins bind substances the bacterial cells need to survive (Puupponen-Pimia and Others 00, 005). Furthermore, extracts from other berries in the Rubus genus have been found to be effective against Gram-negative and Grampositive bacteria (Nohynek and Others 006). Ethanol, present in wines, is known to have a solubilizing effect on the membranes of bacteria, which may allow for easier permealization of other constituents that are toxic to the bacterial cell (Willey and Others 008). The objectives for this study were to evaluate the ability of E. coli O57:H7, Salmonella Typhimurium, L. monocytogenes, and Staphylococcus aureus to survive in blackberry juices and wines, and to understand which constituents of the juices and wines contribute to antibacterial activity.

60 4 3.3 Materials and Methods 3.3. Juice and Wine Preparation Approximately 34 kg each of Marion and Black Diamond blackberries were collected for the purpose of making juice and wine. Marion fruit was harvested by machine in a single harvest from the North Willamette Research and Extension Center (NWREC) in Aurora, OR. After collection, the fruit was pureed and frozen at -3.8 C for further use. Black Diamond blackberries were hand-harvested from Riverbend Organic Farms in Jefferson, OR. Fruit were harvested over the Black Diamond fruiting season (July, 0) and frozen at -3.8 C until 34 kg was accumulated, after which, the berries were thawed and pureed before being frozen again for future use. Purees were thawed and equally divided. Half of each puree was hand pressed through a mesh bag in order to release juice. The juices were then bottled in 87 ml glass bottles, capped with crown caps, and pasteurized in a water bath at 7 C for 5 m. The other half of each puree was placed into 38 L plastic buckets and inoculated with Saccharomyces cerevisiae for the purpose of making wine. After 5 d of fermentation at ambient temperatures, both the liquid and pulp components of the wine were strained through a mesh bag and the filtered liquids were transferred to L carboys for continued fermentation for 5 d. The wine was then racked and bottled into 87 ml glass bottles, crown capped, and pasteurized as described above. Bottles of juice and wine were boxed and kept out of light at room temperature for further use.

61 3.3. Juice and Wine Properties Sterility of pasteurized juice and wine was determined by ATP bioluminescence (Firefly Luminometer, Arrow Scientific). The ph and titratable acidity of juice and wine samples were also measured. Titratable acidity was determined by titration with 0.N NaOH to an end point of ph 8. and calculated as grams of citric acid/l (ph Meter, Cole- Parmer). Soluble solid content of purees and juices were determined by refractometry (RFM 8 Multi Scale Automatic Refractometer, Bellingham Stanley, Inc.). Ethanol content of wines was determined by ebulliometry (Zoecklein 990; Dujardin-Salleron Laboratoires) Culture Preparation Cultures of Escherichia coli O57:H7 (ATCC 43894), Salmonella Typhimurium (ATCC 408), Listeria monocytogenes (Scott A), and Staphylococcus aureus (general food isolate, 648 From OSU Culture Collection) were used for juice and wine survival studies. All cultures used were from the culture collection at Oregon State University, Department of Food Science and Technology. Cultures were maintained in Trypticase Soy Broth (TSB) then were transferred to fresh TSB and incubated for 4 h at 35 C (BBL Trypticase Soy Broth, BD). One hundred µl from each were then again transferred to fresh TSB and incubated for 8 h at the same temperature. These cultures were then serially diluted in sterile Butterfield s phosphate buffer (referred to as phosphate buffer for simplicity) and used for inoculation. Enumeration was determined by serial dilution on Trypticase Soy Agar (TSA) (Bacto Agar, BD; BBL Trypticase Soy Broth, BD).

62 Survival Study Procedure Controls Due to the presence of pectin, juices were centrifuged at 63 x g for 5 m prior to use (IEC Clinical Centrifuge). Of each juice supernatant and wine, 9.9 ml was placed in culture tubes, to which 00 µl of a single bacterial strain was added. Samples were serially diluted in phosphate buffer and plated on TSA at appropriate time intervals to determine the maximum survival time of each bacterium in the control samples Variables Juice supernatant samples were adjusted to have an ethanol content that matched their wine counterpart. For Marion juice, ml of juice was mixed with ml of 00 % ethanol for a total sample volume of 9.9 ml with an ethanol content of 5%. Of Black Diamond juice, ml was mixed with 0.46 ml of 00% ethanol for a total volume of 9.9 ml with an ethanol content of 4.%. Wine samples were adjusted with 6M NaOH to increase their ph values by approximately unit in order to evaluate the impact of ph on bacterial survival. One hundred µl of culture was added to each of the variable samples and plated in the method described above for control samples Data Analysis All data analyses were conducted using Microsoft Excel. 3.4 Results and Discussion 3.4. Properties of Purees, Juices, and Wines The ph of Marion and Black Diamond purees, juices, and wines were fairly similar, ranging from 3. to 3.30 for all products (Tables 3.-4). The soluble solids

63 45 content of Marion puree and juice was slightly higher than Black Diamond puree and juice (Tables 3.-3). As would be expected based on soluble solids content, Marion wine had slightly higher ethanol content than Black Diamond wine, at 5.0% and 4.%, respectively (Table 3.4). Titratable acidity was higher in the Black Diamond juice and wine than in the Marion juice and wine (Tables 3.3-4). These values are consistent with other reported ph, titratable acidity, and soluble solid content values for blackberry juice (Vasquez-Araujo and Others 00). Table 3. ph and Soluble Solid Content of Blackberry Puree Marion Juice Black Diamond Juice ph Brix Table 3.3 ph, Soluble Solid Content, and Titratable Acidity of Blackberry Juices Marion Juice Black Diamond Juice ph Brix TA 3.95 g citric acid/l 4.76 g citric acid/l Table 3.4 ph, Ethanol Content, and Titratable Acidity of Blackberry Wines Marion Wine Black Diamond Wine ph % Ethanol 5% 4.% TA 5.85 g citric acid/l 5.99 g citric acid/l 3.4. Survival Study Results Growth was not observed for any of the microorganisms in any of the treatments. Escherichia coli O57:H7 was observed to no longer be detectable at 84 h in both Marion and Black Diamond juices (Figures 3. and 3.3). The detection limit for

64 46 the plating procedure used was 0.70 Log. Salmonella Typhimurium was no longer detectable at a maximum of 60 h in Marion juice and 08 h in Black Diamond juice (Figures 3.5 and 3.7). It should be noted that in Black Diamond trials, Salmonella Typhimurium was observed to no longer be detectable at 60 h in one trial, then remained at or near the detection limit from h in the second trial (Figure 3.7). This may have been the result of a few particularly acid tolerant cells being present. Staphylococcus aureus survived for 48 and 84 h in Marion and Black Diamond juices, respectively (Figures 3.9 and 3.). Results for L. monocytogenes tended to be inconsistent among trials, making it difficult to establish means and ranges. For this reason, results for L. monocytogenes are reported as maximum survival times in Table 3.5 for all treatments. Listeria monocytogenes was observed to have the shortest survival time, at h in both juices (Table 3.5). Adding ethanol to juice samples had the effect of reducing the survival time of all microorganisms, often by more than half. The survival time of E. coli O57:H7 was reduced to 36 h and Salmonella Typhimurium was reduced to 6 h in both juices with added ethanol (Figures 3., 3.3, 3.5, and 3.7). The survival time of S. aureus was reduced to 4 h and 36 h, L. monocytogenes to h and h in Marion and Black Diamond juices with added ethanol, respectively (Figures 3.9 and 3., table 3.5). This reduction in survival time was to be expected due to the ability of ethanol to solubilize

65 47 the bacterial membrane, making it easier for undissociated organic acids to be able to penetrate the bacterial cell (Willey and Others 006). Escherichia coli O57:H7 was observed to survive for a maximum of 60 m in Marion wine and 40 m in Black Diamond wine (Figures 3. and 3.4). Salmonella Typhimurium and S. aureus were observed to survive for a maximum of 40 m and 80 m in both wine samples, respectively (Figures 3.6, 3.8, 3.0, and 3.). Listeria monocytogenes survived for a maximum of 80 m in Marion wine and 40 m in Black Diamond wine (Table 3.5). There was a dramatic difference in survival times when comparing the wine to the juice with added ethanol results, even though both sets of samples had the same ethanol contents and similar ph values. This may have been due to the juice samples still having plenty of nutrients, such as carbohydrates and amino acids, available to the bacterial cells as they attempted to survive the hostile environment of organic acids and ethanol. The wines would have had many of these nutrients depleted by the yeast during the fermentation process. Other differences may include the wine containing additional yeast metabolites that may exhibit antimicrobial activity, such as acetate (Davison and Stephanopoulos 986). Increasing the ph of the wines had the effect of increasing the survival times of all microorganisms from 80 m to 6-48 h. In the case of L. monocytogenes, total kill was not observed; however, a 4.6 log reduction occurred in ph adjusted Marion wine and

66 Log 48 a.96 log reduction occurred in ph adjusted Black Diamond wine at 48 h (Table 3.5). In ph adjusted Marion wine, E. coli O57:H7 survived for 36 h, Salmonella Typhimurium for 6 h, and S. aureus for 36 h (Figures 3., 3.5, and 3.9). In ph adjusted Black Diamond wine, E. coli O57:H7 survived for 48 h, Salmonella Typhimurium for 36 h, and S. aureus for 48 h (Figures 3.3, 3.7, and 3.). Increasing the ph of the wines had a dramatic effect on the survival time of all microorganisms. This was likely due to the increased ph causing more of the organic acids to exist in their dissociated state, making entry into the bacterial cell more difficult. The dramatic increase in survival observed with L. monocytogenes is consistent with another study that found some strains were able to persist over ph 3.5 and 4.0 for several hours (Phan-Thanh and Others 000). Figure 3. Survival of E. coli O57:H7 in Marion Products. Bars indicate the range (high and low values) Marion Juice Marion Juice Trial Marion Juice 5% ethanol Marion Wine Marion Wine ph 4.3 Detection Limit Time in Hours

67 Log Log 49 Figure 3. Survival of E. coli O57:H7 in Marion Wine. Bars indicate the range (high and low values) Marion Wine Marion WineTrial Detection Limit Time in Minutes Figure 3.3 Survival of E. coli O57:H7 in Black Diamond Products. Bars indicate the range (high and low values) Time in Hours Black Diamond Juice Black Diamond Juice 4.% ethanol Black Diamond Wine Black Diamond Wine ph 4.4 Detection Limit

68 Log Log Figure 3.4 Survival of E. coli O57:H7 in Black Diamond Wine. Bars indicate the range (high and low values) Black Diamond Wine Black Diamond Wine Trial Detection Limit Time in Minutes Figure 3.5 Survival of Salmonella Typhimurium in Marion Products. Bars indicate the range (high and low values) Marion Juice Marion Juice Trial Marion Juice 5% ethanol Marion Wine Marion Wine ph 4.3 Detection Limit Time in Hours

69 Log Log 5 Figure 3.6 Survival of Salmonella Typhimurium in Marion Wine. Bars indicate the range (high and low values) Marion Wine Detection Limit Time in Minutes Figure 3.7 Survival of Salmonella Typhimurium in Black Diamond Products. Bars indicate the range (high and low values) Time in Hours Black Diamond Juice Black Diamond Juice Trial Juice 4.% Ethanol Black Diamond Wine Wine ph 4.4 Detection Limit

70 Log Figure 3.8 Survival of Salmonella Typhimurium in Black Diamond Wine. Bars indicate the range (high and low values) Black Diamond Wine Detection Limit Time in Minutes Table 3.5 Maximum Observed Survival Times of L. monocytogenes in Marion and Black Diamond Juices and Wines: All Treatments Treatment Marion Juice Marion Juice 5.0% Ethanol Marion Wine Marion Wine ph 4.3 Black Diamond Juice Black Diamond Juice 4.% Ethanol Black Diamond Wine Black Diamond Wine ph 4.4 (Total kill was not observed in ph adjusted wine samples) Survival Time h h 80 m 4.6 Log Reduction in 48 h h h 40 m.96 Log Reduction in 48 h

71 Log Log 53 Figure 3.9 Survival of S. aureus in Marion Products. Bars indicate the range (high and low values) Marion Juice Marion Juice Trial Marion Juice 5.0% Ethanol Marion Wine Marion Wine ph Time in Hours Detection Limit Figure 3.0 Survival of S. aureus in Marion Wine. Bars indicate the range (high and low values) Marion Wine Marion Wine Trial Detection Limit Time in Minutes

72 Log Log Figure 3. Survival of S. aureus in Black Diamond Products. Bars indicate the range (high and low values) Time in Hours Black Diamond Juice Black Diamond Juice Trial Black Diamond Juice 4.% Ethanol Black Diamond Wine Black Diamond Wine ph 4.3 Detection Limit Figure 3. Survival of S. aureus in Black Diamond Wine. Bars indicate the range (high and low values) Black Diamond Wine Black Diamond Trial Detection Limit Time in Minutes

73 3.5 Conclusions Marion and Black Diamond juices and wines did not support the growth of E. coli O57:H7, Salmonella Typhimurium, L. monocytogenes, and S. aureus. However, survival will vary depending on acid and ethanol content. The maximum time that any of the tested microorganisms was observed to survive was 80 m in the wines and 08 h in the juices. These findings are consistent with other survival studies conducted using juice and wine made from wine grapes (Just and Daeschel 003; Sugita-Konishi and Others 00). These results indicate that if blackberry juice or wine became contaminated post-processing, these particular microorganisms would not be likely to survive by the time the product reached the consumer. Caution should be taken when interpreting these findings for juices because these experiments were conducted at ambient temperatures. If juices were contaminated and held under refrigeration or freezing conditions, survival may be extended. This could be of major concern for products such as unpasteurized frozen puree. Further research would need to be conducted to determine survival times under those conditions. 55 The constituents of the juices and wines that likely contribute to antibacterial activity include organic acids, phenolic compounds, and ethanol (wine). These constituents being present in the same system make it very likely that they act synergistically. The dramatic increase in survival time that was observed when the ph of wines was increased by approximately ph unit indicates that the organic acids are a major antibacterial contributor. Understanding the contribution of each constituent

74 56 would require conducting survival studies with them individually. Further research would be necessary to fully understand how the phenolic compounds, in the quantities that they occur in blackberry juice and wine, affect the ability of the studied bacteria to survive.

75 57 4. Overall Conclusions and Future Work The bacterial pathogens used in these studies were not observed to grow on fresh and frozen blackberries. Survival was observed for Listeria monocytogenes and Staphylococcus aureus on fresh, wild Himalaya after 3 d and on frozen Triple Crown after 6 months. Salmonella Typhimurium was detected on frozen Triple Crown at 3 and 6 months. These results emphasize the need for sanitary growing, harvesting, and handling procedures to prevent contamination. Maintaining such procedures will help ensure that blackberries do not become the center of a food safety incident as raspberries have in the past (Gillesberg and Lassen 03; Ho and Others 00; Reid and Robinson 987; Sarvikivi and Others 0). Many of the causative agents involved with raspberry recalls are harbored by humans, for example, hepatitis A, norovirus, and Cyclospora cayetanensis. This indicates that blackberries may be susceptible to contamination with microorganisms that humans are known reservoirs for including S. aureus, pathogenic strains of Escherichia coli, and Salmonella spp. in addition to those previously mentioned. Preventing contamination with these microorganisms requires emphasis and attention paid to worker hygiene. Furthermore, considering current recalls, processors are advised to have procedures in place to verify the safety of imported supplies especially if adding them to berry blends. This could be in the form of a certificate of acceptance/analysis or letter of guarantee from a foreign supplier of berry product.

76 58 Aside from humans, other sources of contamination are possible in the field. Animals, insects, and contaminated soil or irrigation water could all lead to bacterial deposits on blackberry fruit (Beuchat 00; Janisiewicz and Others 999; Terry 0). The mechanical harvester evaluated did not appear to a potential source of contamination, even when left intentionally unwashed. A diverse microflora was observed on the surface of fresh blackberries. How this microflora affects the ability of pathogenic bacteria to adhere to the berry surface and survive is not well understood and may be worth investigating further. The inability of E. coli O57:H7, Salmonella Typhimurium, L. monocytogenes, and S. aureus to survive beyond 80 m in wines and 08 h in juices demonstrate that blackberries inherently have antibacterial properties. There are several constituents of blackberry juices and wines that are likely to contribute to bacterial death, including organic acids, phenolic compounds, and ethanol in the case of wines. These constituents likely behave synergistically to cause bacterial cells to die. The dramatic increase in survival time observed when the ph of the wines was increased suggests that the organic acids are the primary antibacterial constituent. These studies were conducted at ambient temperatures. Future work should look at what effect the sample temperature has on survival and should be extended to frozen blackberry puree. Furthermore, conducting survival studies with isolated phenolic compounds from blackberries may yield valuable information.

77 59 Bibliography Aura A, Martin-Lopez P, O Leary KA & Others. (005). In Vitro Metabolism of Anthocyanins by Human Gut Flora. European Journal of Nutrition 44: Beuchat LR. (987). Food and Beverage Mycology. nd ed. New York: Van Nostrand Reinhold. 0 p. Beuchat LR. (00). Ecological Factors Influencing Survival and Growth of Human Pathogens on Raw Fruits and Vegetables. Microbes and Infection 4: Brul S & Coote P. (999). Preservative Agents in Foods: Mode of Action and Microbial Resistance Mechanisms. International Journal of Food Microbiology 50:-7. Bilyk A & Sapers GM. (986). Varietal Differences in the Quercetin, Kaempferol, and Myricetin Contents of Highbush Blueberry, Cranberry, and Thornless Blackberry Fruits. Journal of Agricultural and Food Chemistry 34(4): Bowling B. (000). The Berry Grower s Companion. Portland: Timbers Press p. [CDC] Centers for Disease Control and Prevention. (00). State-Specific Trends in Fruit and Vegetable Consumption Among Adults-United States, Morbidity and Mortality Weekly Report 59(35):5-30. [CDC] Centers for Disease Control and Prevention. (0). Strategies to Prevent Obesity and Other Chronic Diseases: The CDC Guide to Strategies to Increase the Consumption of Fruits and Vegetables. Atlanta: U.S. Department of Health and Human Services [CDC] Centers for Disease Control and Prevention. (03). Multistate Outbreak of Hepatitis A infections Linked to Pomegranate Seeds from Turkey. [Accessed June 8, 03] Davison BH & Stephanopoulos G. (986). Coexistence of S. cerevisiae and E. coli in Hemostat Under Substrate Competition and Product Inhibition. Biotechnology & Bioengineering 8():74-5. Du XF, Kurnianta A, McDaniel M, & Others. (00). Flavour Profiling of Marion and Thornless Blackberries by Instrumental and Sensory Analysis. Food Chemistry :080-8.

78 Duan J and Zhao Y. (009). Antimicrobial Efficiency of Essential Oil and Freeze-Thaw Treatments against Escherichia coli O57:H7 and Salmonella enterica Ser. Enteritidis in Strawberry Juice. Journal of Food Science 74(3):M3-7. Falkenhorst G, Krusell L, Lisby M & Others. (005). Imported Frozen Raspberries Cause a Series of Norovirus Outbreaks in Denmark, 005. Eurosurveillance 0(38): pii=795. [Accessed June 7, 03] Available online: Fan-Chian, H. December 4, 999. Anthocyanin Pigment, Nonvolatile Acid and Sugar Composition of Blackberries, MS Thesis. Corvallis, OR: Oregon State University. 90p. [Accessed June 9, 03] Available from Oregon State University Scholars Archive: pdf?sequence= Finn CE, Yorgey BM, Strik BC & Others. (005). Metilius Trailing Blackberry. HortScience 40(7):89-9. Finn CE and Strik BC. (008). Blackberry Cultivars for Oregon. Oregon State University Extension Service. EC 67-E. Flegal KM, Carroll MD, Kit BK, & Others. (0). Prevalence of Obesity and Trends in the Distribution of Body Mass Index Among US Adults, Journal of the American Medical Association 307(5):49-7. [FDA] Food and Drug Administration. (0). Fresh Strawberries from Washington County Farm Implicated in E. coli O57 Outbreak in NW Oregon. [Accessed June 8, 03] Friedman M, Henika PR, Levin CE & Others. (004). Antibacterial Activities of Plant Essential Oils and Their Components against Escherichia coli O57:H7 and Salmonella enterica in Apple Juice. Journal of Agricultural and Food Chemistry 5: Gillesberg Lassen S, Soborg B, Midgley SE & Others. (03). Ongoing Multi-strain Foodborne Hepatitis A Outbreak with Frozen Berries as Suspected Vehicle: Four Nordic Countries Affected, October 0 to April 03. Eurosurveillance 8(7):pii= Gropper SS, Smith JL, and Groff JL. (008) Advanced Nutrition and Human Metabolism. 5 th ed. Belmont: Wadsworth. 55, 3, 64 p. 60

79 Herwaldt BL, Ackers ML and the Cyclospora Working Group. (997). An Outbreak in 996 of Cyclosporiasis Associated with Imported Raspberries. New England Journal of Medicine 336(): Ho AY, Lopez AS, Eberhart MG & Others. (00). Outbreak of Cyclosporiasis Associated with Imported Raspberries, Philadelphia, Pennsylvania, 000. Emerging Infectious Diseases 8(8): Hummer K. (00). Rubus Pharmacology: Antiquity to the Present. HortScience 45(): Janisiewicz W, Conway WS, Brown MW & Others. (999). Fate of Escherichia coli O57:H7 on Fresh-But Apple Tissue and Its Potential for Transmission by Fruit Flies. Applied Environmental Microbiology 65(): -5. Jay JM, Loessner MJ, & Golden DA. (005). Modern Food Microbiology. 7 th ed. New York: Springer Science , 40 p. Jenkins DJA, Kendall CWC, Vuksan V & Others. (00). Soluble Fiber Intake at a Dose Approved by the US Food and Drug Administration for a Claim of Health Benefits: Serum Lipid Risk Factors for Cardiovascular Disease Assessed in a Randomized Controlled Crossover Trial. The American Journal of Clinical Nutrition 75: Jensen B, Knudsen IMB, Andersen B & Others. (03). Characterization of Microbial Communities and Fungal Metabolites on Field Grown Strawberries from Organic and Conventional Production. International Journal of Food Microbiology 60:33-. Just JR & Daeschel MA. (003). Antimicrobial Effects of Wine on Escherichia coli O57:H7 and Salmonella Typhimurium in a Model Stomach System. Journal of Food Science 68(): Kempler K & Humpf HU. (005). Metabolism of Anthocyanins and their Phenolic Degradation Products by the Intestinal Microflora. Bioorganic and Medicinal Chemistry 3: Knudsen DM, Yamamoto SA, & Harris LJ. (00). Survival of Salmonella spp. And Escherichia coli O57:H7 on Fresh and Frozen Strawberries. Journal of Food Protection 64(0):

80 Mahmoud BSM, Bachman G, & Linton RH. (00). Inactivation of Escherichia coli O57:H7, Listeria monocytogenes, Salmonella enterica and Shigella flexneri on Spinach Leaves by X-ray. Food Microbiology 7:4-8. Nohynek LJ, Alakomi H, Kahkonen MP & Others. (006). Berry Phenolics: Antimicrobial Properties and Mechanisms of Action Against Severe Human Pathogens. Nutrition and Cancer 54():8-3. Oregon Climate Service. (July 0). College of Oceanic and Atmospheric Sciences: Oregon State University. [Accessed June, 03]. Available online: Oyarzabal OA, Nogueira MCL & Gombas DE. (003). Survival of Escherichia coli O57:H7, Listeria monocytogenes, and Salmonella in Juice Concentrates. Journal of Food Protection 66(9): Perrigue M, Carter B, Roberts SA, & Others. (00). A Low-Calorie Beverage Supplemented with Low-Viscosity Pectin Reduces Energy Intake at a Subsequent Meal. Journal of Food Science 75(9): H Phan-Thanh L, Mahouin F & Alige S. (000). Acid Responses of Listeria monocytogenes. International Journal of Food Microbiology 55:-6. Puupponen-Pimia R, Nohynek L, Meier C & Others. (00). Antimicrobial Properties of Phenolic Compounds from Berries. Journal of Applied Microbiology 90: Puupponen-Pimia R, Nohynek L, Alakomi H & Others. (005). Bioactive Berry Compounds-Novel Tools Against Human Pathogens. Applied Microbiology and Biotechnology 67: 8-8. Raybaudi-Massilia RM, Mosqueda-Melgar J, & Martin-Belloso O. (006). Antimicrobial Activity of Essential Oils on Salmonella enteritidis, Escherichia coli, and Listeria innocua in Fruit Juices. Journal of Food Protection 69(7): Reid TMS & Robinson HG. (987). Frozen Raspberries and Hepatitis A. Epidemiological Infection 98:09-. Rico D, Martin-Diana AB, Barat JM & Others. (007). Extending and Measuring the Quality of Fresh-cut Fruit and Vegetables: A Review. Trends in Food Science & Technology 8:

81 63 Ruyi W, Frei B, Kennedy JA & Others. (00). Effects of Refrigerated Storage and Processing Technologies on the Bioactive Compounds and Antioxidant Capacities of Marion and Evergreen Blackberries. Food Science and Technology 43: Sarvikivi E, Roivainen M, Maunula L & Others. (0). Multiple Norovirus Outbreaks Linked to Imported Frozen Raspberries. Epidemiology and Infection 40:60-7. Sasaki R, Nishimura N, Hoshino H, & Others. (007). Cyanidin-3-glucoside Ameliorates Hyperglycemia and Insulin Sensitivity Due to Downregulation of Retinol Binding Protein 4 Expression in Diabetic Mice. Biochemical Pharmacology 74: Sellappan S, Akoh CC & Krewer G. (00). Phenolic Compounds and Antioxidant Capacity of Georgia-Grown Blueberries and Blackberries. Journal of Agricultural and Food Chemistry 50: Shepherd T & Griffiths DW. (006).The Effects of Stress on Plant Cuticular Waxes. New Phytologist 7: Stone C. (0). Fresh Strawberries Implicated in E. coli O57 Outbreak in NW Oregon. New York Berry News 0(7):. Strik BC, Clark JR, Finn CE & Others. (007). Worldwide Blackberry Production. HortTechnology 7(): Sugita-Konishi Y, Hara-Kudo Y, Iwamoto T & Others. (00). Wine has Activity against Entero-pathogenic Bacteria in Vitro but not in Vivo. Bioscience, Biotechnology, and Biochemistry 65(4): Tournas VH & Katsoudas E. (005). Mould and Yeast Flora in Fresh Berries, Grapes and Citrus Fruits. International Journal of Food Microbiology 05:-7. Terry L. (03). Hepatitis A Outbreak: Gresham Company Recalls Pomegranate Seeds. The Oregonian, June 7, 03. [Accessed June 8, 03] c.html Terry L. (0). Tests Reveal E. coli in Deer Droppings Found in Oregon Strawberry Field. The Oregonian. August, 0. [Accessed April 5, 03] in_deer_dr.html

82 Tschriter O, Fritsche A, Thamer C & Others. (003). Plasma Adiponectin Concentrations Predict Insulin Sensitivity of Both Glucose and Lipid Metabolism. Diabetes 5: Tsuda, T. (0). Dietary Anthocyanin-rich Plants: Biochemical Basis and Recent Progress in Health Benefits Studies. Molecular Nutrition and Food Research 56: [USDA]. United States Department of Agriculture. National Nutrient Database for Standard Reference. Blackberries, Raw.[Accessed May 5, 03] [USDA] United States Department of Agriculture. (0). Noncitrus Fruits and Nuts 0 Preliminary Summary. Vasquez-Araujo, L, Chambers IV E, Adhikari K & Others. (00). Sensory and Physicochemical Characterization of Juices Made with Pomegranate and Blueberries, Blackberries, or Raspberries. Journal of Food Science 75(7):S Willey JM, Sherwood LM, & Woolverton CJ. (008). Prescott, Harley, & Klein s Microbiology. Customized Version for Oregon State University, 7 th ed. Indianapolis: McGraw Hill Learning Solutions p. Yao LH, Jiang YM, Shi J & Others. (004). Flavonoids in Foods and Their Health Benefits. Plant Foods for Human Nutrition 59: 3-. Zoecklein BW, Fugelsang KC, Gump BH & Others. (990). Production Wine Analysis. New York: Van Nostrand Reinhold International Company. 3-34, 43-5, 57-6 p. 64

83 APPENDICES 65

84 Appendix I. Fresh Field Samples Raw Data, Chapter. Bold indicates value used in calculation. Detection limit calculated as: < colony/plate at the lowest dilution with 0. ml on each of plates /0. = 5 CFU/g or 0.70 Log CFU/g. Table I. Aerobic Count Raw Data, Fresh Samples 'Obsidian' Mid Harvest 07/0/ Rep Rep CFU/g Log 66 CFU/g Log CFU/g CFU/g 0^ ^ ^ ^ Obsidian' Late-Harvest 07/9/ Rep Rep CFU/g Log CFU/g CFU/ g Log CFU/g 0^ ^ ^ ^-5 0 Contaminated 'Triple Crown' Early-Harvest 08/3/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^ ^ ^ ^ Triple Crown' Mid/Late Harvest 08/30/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^- TNTC TNTC ^ ^ ^

85 67 Table I. Yeasts and Molds Raw Data, Fresh Samples 'Obsidian' Mid-Harvest 07/0/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^- TNTC TNTC TNTC TNTC ^ ^ Obsidian' Late-Harvest 07/9/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^- TNTC TNTC TNTC TNTC ^ ^ 'Triple Crown' Early-Harvest 08/3/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^ ^ ^-3 0 Triple Crown' Mid/Late Harvest 08/30/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^ ^ ^

86 68 Table I.3 Coliforms Raw Data, Fresh Samples 'Obsidian' Mid-Harvest 07/0/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^- 0 0 <5 < ^ ^ 'Obsidian' Late-Harvest 07/9/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^- 0 0 <5 < <5 <0.70 0^ ^ 'Triple Crown' Early-Harvest 08/3/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^ <5 <0.70 0^ ^ 'Triple Crown' Mid/Late-Harvest 08/30/ Rep Rep CFU/g Log CFU/g CFU/g Log CFU/g 0^- 0 0 <5 < <5 <0.70 0^ ^

87 Appendix II. Mechanical Harvester Raw Data and Photos, Chapter. Bold indicates value used in calculation. Table II. Aerobic Count Raw Data, Harvester Location Pre-Harvest Post-Harvest CFU/swab Log CFU/swab CFU/swab Log CFU/swab 0^- TNTC TNTC TNTC TNTC ^ ^ Location Pre-Harvest Post-Harvest CFU/swab Log CFU/swab CFU/swab Log CFU/swab 0^ TNTC ^ ^ Location 3 Pre-Harvest Post-Harvest CFU/swab Log CFU/swab CFU/swab Log CFU/swab 0^- TNTC TNTC ^- TNTC 4 0^ Location 4 Pre-Harvest Post-Harvest CFU/swab Log CFU/swab CFU/swab Log CFU/swab 0^ ^- 4 0^ Location 5 Pre-Harvest Post-Harvest CFU/swab Log CFU/swab CFU/swab Log CFU/swab 0^- 350 TNTC ^ ^ Location 6 Pre-Harvest Post-Harvest CFU/swab Log CFU/swab CFU/swab 69 Log CFU/swab

88 0^ ^ ^ Location 7 Pre-Harvest Post-Harvest CFU/swab Log CFU/swab CFU/swab Log CFU/swab 0^ ^ ^ Location 8 Pre-Harvest Post-Harvest CFU/swab Log CFU/swab CFU/swab Log CFU/swab 0^ ^ ^ Figure II. Harvester Location Figure II. Harvester Location

89 7 Figure II-3 Harvester Location 3 Figure II.4 Harvester Location 4 (beater bar) Figure II.5 Harvester Location 5 Figure II.6 Harvester Location 6

90 7 Figure II.7 Harvester Location 7 Figure II.8 Harvester Location 8 Appendix III. Himalaya Raw Data, Chapter. Bold indicates value used in calculation. Detection limit calculated as: < colony/plate at the lowest dilution with 0. ml on each of plates /0. = 5 CFU/g or 0.70 Log CFU/g; /0.0 = 50 CFU/g or.70 Log CFU/g for homogenized samples. Table III. E. coli O57:H7 on fresh 'Himalaya' Blackberries, Raw Data Rep Rep Berry Weight.4 g.6 g Inoculum 5.74 Log CFU/g Rinse Solution 5.69 Log CFU/g CFU/g CFU/g 0^- 0 0 <5 < <5 <0.70 0^ ^ Homogenized :0