DEVELOPMENT OF A SPICE FORMULATION FOR PURE CULTURED 'FERMENTED CUCUMBER HCKLES Thesis. for the Degree of M. S? MlCHlGAN STATE UNWERSITY CECELIA KUHNLEY MARSHALL 1971 yaflkb‘. LIBRARY Michigan State University swims BY ‘3' HUAB 3: SUNS. UUUK‘BMDERY INC. LIBRARY BINDERS gmnaropf, mam; A; . My" . ‘ ABSTRACT DEVELOPMENT OF A SPICE FORMULATION FOR PURE CULTURE FERMENTED CUCUMBER PICKLES By Cecelia Kuhnley Marshall The objective of this study was to develop a pure culture fermented pickle with good flavor that did not have the spice formu- lation rendered unacceptable by the fermentation process. Nine different spice formulations were evaluated for this purpose. Various chemical analyses were used to monitor the course of fermentation of cucumbers by pure cultures of three strains of Lactobacillus plantarum, two strains of Pediococcus cerevisiae and mixtures of both. Total plate counts were used to follow the growth of the organisms and to determine the effects of various spice formulations on the cultures. The different spice formulations were evaluated for flavor acceptability, and the best formulation was submitted to taste panels. Attempts were made to determine the effects of the various microorganisms on a spice formulation Cecelia Kuhnley Marshall by gas-liquid chromatographic analysis of the headspace in sample jars. A spice formulation containing oil of dill weed, oleoresin- oil of garlic, oleoresin capsicum, caraway oil, pimenta leaf oil, and polysorbate 80, when added at a level of 0. 94% (v/v) to a brine containing 6. 7% (w/w) salt, was found to produce a good product. Differences in the flavor of the final product were noted, depending on which microorganism was responsible for the fermentation. Thus, the flavor of products fermented by various strains of Lactobacillus plantarum and Pediococcus cerevisiae-were evaluated subjectively and objectively. Subjective taste panel evaluations indicated that the pedio- cocci produced a bland pickle, lactobacilli produced sour pickles (the degree of sourness depended on the level of acid produced), and the mixed cultures of 3. cerevisiae and £1. plantarum tended to counterbalance each other by accentuating certain flavor character- istics and attenuating others. Hedonic ratings by taste panels indicated that preference in pickles varied with individual tastes and that, overall, no one sample was preferred above another. Gas-liquid chromatography, using flame ionization and electron capture detectors, failed to show any differences in the headspace profiles of the various samples. However, addition of Cecelia Kuhnley Marshall cucumbers to the brine, whether or not they were fermented, resulted in an unexplained disappearance of two peaks which appeared in the chromatograms of the brine. Although the differ- ences in flavor produced by the fermenting microorganisms were evident by organoleptic evaluation, they were not detected by this chromatographic system. The variation in flavor was the result of some change in trace volatile constituents or in the non -volatile flavor constituents. DEVELOPMENT OF A SPICE FORMULATION FOR PURE CULTURE FERMENTED CUCUMBER PICKLES By Cecelia Kuhnley Marshall A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1971 ACKNOWLEDGMENTS The author wishes to express her gratitude to Dr. R. V. Lechowich, for his guidance and assistance throughout the study and in the preparation of this manuscript. The author also wishes to thank Dr. C. L. Bedford of the Department of Food Science and Human Nutrition and Dr. R. N. Costilow of the Department of Microbiology and Public Health for their suggestions. Special thanks are given to Mrs. Marguerite Dynnik for her excellent technical assistance, and to Dr. Ken Fox and Mr. Merlin Breen for their help on the project. The author is indebted to Dr. L. G. Harmon for his interest in and support of this project. Finally the author is deeply grateful to her husband, Jim, for his continuing strength and support and understanding. ii L1 L1; RE TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION............. REVIEW OF LITERATURE The Natural Fermentation of Cucumbers Pure Culture Fermentation of Cucumbers Analysis of Flavor Use of Gas- Liquid Chromatography for Flavor Research . Spices: Oils and Oleoresins METHODS AND MATERIALS Preparation of Pickles by Pure Culture Fermentation . . . . . Cultures Brines Cucumbers Chemical Analyses of Pickles pH . . Titratable Acidity Reducing Sugar Total Count Salt . . . Organoleptic Analyses of Pickles Chromatographic Analyses of Pickles Column . . Gas Chromatograph . Preparation of Samples and Standards Interpretation of Results iii Page . viii 10 12 13 19 21 21 21 22 24 25 26 26 26 28 29 29 30 30 30 32 33 RESULTS AND DISCUSSION Chemical Analyses . Changes in Reducing Sugar, Titratable Acidity, and pH . Plate Counts Organoleptic Evaluations Gas Chromatographic Analyses SUMMARY AND CONCLUSIONS LITERATURE CITED General References APPENDIX iv Page 34 34 34 50 51 53 62 65 73 76 TABLE LIST OF TABLES Page pH, per cent titratable acidity (calculated as lactic acid), and per cent reducing sugar in heated, acidified, uninoculated control samples in various spice formulations . . . . . . . . . . 35 Changes in pH, per cent titratable acidity (calcu- lated as lactic acid), and per cent reducing sugar during pure culture fermentation of cucumbers in various spice formulations by Pediococcus cerevisiae FBB -39 . . . . . . . . 37 Changes in pH, per cent titratable acidity (calcu- latedas lactic acid), and per cent reducing sugar during pure culture fermentation of cucumbers in various spice formulations by the combination L. plantarum FBB -12 and E. cerevisiae FBB-39 . . . . . . . . . . . . 38 Changes in pH, per cent titratable acidity (calcu- lated as lactic acid), and per cent reducing sugar during pure culture fermentation of cucumbers in various spice formulations by Lactobacillus plantarum FBB-12 . . . . . . . . 39 Changes in pH, per cent titratable acidity (calcu- lated as lactic acid), and per cent reducing sugar during pure culture fermentation of cucumbers in brine flavored with the double strengthdill chipspice formulation by various strains of Lactobacillus plantarum and Pediococcus cerevisiae . . . . . . . . . . 44 TABLE 10. 11. 12. Changes in pH, per cent titratable acidity (calcu- lated as lactic acid), and per cent reducing sugar during pure culture fermentation of cucumbers in a brine flavored with selected whole spices plus garlic and dill by various strains of Lactobacillus plantarum and Pediococcus cerevisiae Per cent conversion of sugar to lactic acid during pure culture fermentation of cucumbers by various strains of Lactobacillus plantarum and Pediococcus cerevisiae Changes in total plate counts (organisms/ ml) of brines flavored by various spice formulations during pure culture fermentation by Pedio- coccus cerevisiae FBB-39 . . . Changes in total plate counts (organisms/ ml) of brines flavored by various spice formulations during pure culture fermentation by Lacto- bacillus plantarum FBB-12 Changes in total plate counts (organisms/ ml) of brines .flavored by various spice formulations during pure culture fermentation by Pedio- coccus cerevisiae FBB -39 and Lactobacillus plantarum FBB-12 Changes in total plate counts (organisms/ ml) of brines flavored by the double strengthdill chip formulation during pure culture fer- mentation by various strains of Pediococcus cerevisiae and Lactobacillus plantarum Changes in total plate counts (organisms/ ml) of brines flavored by selected whole spices plus garlic and dill during pure culture fer- mentation by various strains of Pediococcus cerevisiae and Lactobacillus plantarum vi Page 45 48 76 77 78 79 80 TABLE 13. 14. 15. 16. Analysis of Variance showing significance of differences between a commercial product and pure culture samples fermented by Lactobacillus plantarum 4, Pediococcus cerevisiae 3, and the combined cultures, based on data from the first hedonic flavor evaluation Analysis of Variance showing significance of differences between a commercial product and pure culture samples fermented by Lactobacillus plantarum 4, Pediococcus cerevisiae 3, and the combined cultures, based on data from the second hedonic flavor evaluation Analysis of Variance of data from the first hedonic evaluation of the flavor of a com- mercial product (Y) and three pure culture fermented samples . Analysis of Variance of data from the second hedonic evaluation of the flavor of a com- mercial product (Y) and three pure culture fermented samples . . . , vii Page 54 54 81 82 FIGURE LIST OF FIGURE S 1. Changes in pH, per cent reducing sugar, and per cent titratable acidity (calculated as lactic acid) during pure culture fermenta- tion of cucumbers by Pediococcus cere- visiae FBB -39 . U 2. Changes in pH, per cent reducing sugar, and per cent titratable acidity (calculated as lactic acid) during pure culture fermenta- tion of cucumbers by Lactobacillus plantarum FBB-12 and Pediococcus cere- visiae FBB-39 . 3. Changes in pH, per cent reducing sugar and per cent titratable acidity (calculated as lactic acid) during pure culture fermenta- tion of cucumbers by Lactobacillus plantarum FBB—12 4. Final per cent reducing sugar and titratable acidity 5. Gas (calculated as lactic acid) after 60 days in pickles fermented by various strains of Lactobacillus plantarum and Pediococcus cerevisiae chromatograms of the headspace vapors of a brine flavored with a double strength dill chip formulation, an unfermented spiced control, and pure culture fermented spiced samples analyzed by flame ionization (FID) and electron capture (EC) . viii Page 41 42 43 46 59 INTRODUCTION The pickling industry incurs large economic losses annually due to defective salt stock produced during the natural fermentation of cucumbers. Bloaters or hollow pickles may be produced as a result of a gaseous fermentation by coliforms, yeasts, and/or heterofermentative lactobacilli. Softening of pickles is the result of the action of pectin -splitting enzymes produced by molds growing on the cucumber blossoms. Mechanical harvesting increases the number of blossoms that remain on the cucumbers, allowing enzymatic action to occur during fermentation. Several treatments have been proposed to control the natural fermentation, but a consistent quality salt stock has not been achieved. Etchells, Bell, and Costilow (1968a) patented a process for a completely controlled pure culture fermentation. The advantages of such a process include reduced pickling time, consistent high quality products, reduction of losses due to spoilage and defects, suitability of consumer-sized containers for the "in jar" fermenta- tion, and lower labor costs since the produce must be processed only once. However, there have been flavor problems with pure culture pickles (Dr. R. N. Costilow, personal communication). The flavor produced by pure culture fermentation in the absence of spices is not acceptable. When spices have been added to the covering brines, the fermented products have had different flavors, depending on which microorganism was responsible for the fermen- tation. The objective of this study was to develop a spice formula- tion suitable for commercial production of a pure culture fermented pickle. An attempt was also made to objectively evaluate, by GLC headspace analysis, the differences in flavor of the final products as a result of the fermentation by strains of Lactobacillus plantarum and Pediococcus cerevisiae. REVIEW OF LITERATURE The Natural Fermentation of Cucumbers Extensive research on factors involved in the natural fermentation of cucumbers and other vegetables has been done since the early nineteen -thirties, much of it by the U. S. Food Fermenta- tion Laboratory and the North Carolina Agricultural Experiment Station, North Carolina State University, Raleigh, North Carolina. The process of brining cucumbers for natural fermentation has been reviewed elsewhere (Etchells and Jones, 1946; Etchells, Jones, and Bell, 1951c; Cruess, 1958; Binsted, Devey, and Dakin, 1962). There is an initial loss of water and water -soluble nutrients from cucumbers after brining, (Jones and Etchells, 1943; Jones fl. , 1940). The cucumber tissue becomes more permeable, allowing sugar and other organic substances to leave the cells and salt and acid to enter (Jones and Etchells, 1943; Pederson, and Albury, 1962). Salt brining exerts a selective action on the microorganisms indigenous to the cucumber, thus, the resulting fermentation will lead to the production of acid. The combined salinity and acidity of the product are responsible for its preservation. Etchells and Jones (1943) have reported that three groups of microorganisms may be involved in the natural fermentation. Species of Aerobacter are salt —tolerant and may begin the fermen- tation-with the production of equal amounts of hydrogen and carbon dioxide. The lactic acid bacteria are responsible for producing the desired level of acid. These organisms may be homofermentative, in which case lactic acid accounts for 90 to 95% of the end products of fermentation, small amounts of acetic and formic acids and carbon dioxide also being produced (Etchells and Jones, 1946; Christensen, Albury, and Pederson, 1958). The heterofermentative bacteria, e. g. Lactobacillus brevis, produce copious amounts of gas and acetic, lactic, and formic acids in decreasing amounts (Christensen SILL , 1958). Two groups of yeasts, fermentative and oxidative, may be involved in the natural fermentation of brined cucumbers (Etchells and Jones, 1951a). Those which produce an alcohol and gaseous fermentation in the brine include, primarily, Torulopsis holmii and Brettanomyces versatilis, as well as other species of these genera and of Hansenula, Torulaspora, and Zygosaccharomyces. Debaryomyces membranaefaciens was found to be the most preva - lent film yeast; others include species of Endomycopsis and Zygo- saccharomyces and Candida krusei (Etchells and Jones, 1951a; Costilow and Fabian, 1953). The film yeasts grow luxuriously on the tops of fermentation vats and can utilize the lactic acid as a source of carbohydrate, thus raising the pH of the brine and allow- ing the growth of other undesirable spoilage microorganisms. The course of fermentation is directed, at least in part, by the initial inoculum provided by the cucumber (Etchells and Jones, 1943). The production of lactic acid by the homofermentative bacteria, i. e. , Leuconostoc mesenteroides, Streptococcus faecalis, Pediococcus cerevisiae, and Lactobacillus plantarum, in the order of increasing prevalence (Pederson and Ward, 1949; Pederson and Albury, 1950; Costilow and Fabian, 1953; Costilow et_ail_. , 1956), is most desirable, resulting in preservation of the pickles. The gaseous fermentation by coliforms, heterofermentative lactic acid bacteria, and yeasts is considered to be a competitive and undesir- able fermentation (Etchells and Jones, 1943, 1951a; Etchells 3131' , 19510; Etchells, Borg, and Bell, 1968b), because "bloaters" or hollow pickles result. Bloaters reduce the value of the cured stock since they must be used for cut pickle or relish items, and repre- sent a tremendous loss of profit to the pickle industry annually (Etchells 2331. , 1951c). Many attempts have been made to control the natural fer- mentation to reduce the incidence of gas production and bloater formation. The salt-tolerant organisms, including species of Aerobacter, which can initiate the fermentation with the production of gas, are very sensitive to acid (Etchells and Jones, 1951a). Thus, it is important that there be rapid and immediate growth of the lactic acid bacteria after brining the cucumbers. Jones et al. (1940) studied the effects of initial salt concentration and the addition of acid or sugar to the brines. They concluded, as have others (Etchells and Jones, 1946; Pederson and Ward, 1949; Costilow and Fabian, 1953) that brines of lower salt concentrations (10% or less) result in higher acid production. The acid forming bacteria seem to be inhibited by higher salt concentrations (Jones 3E1. , 1940 ; Etchells and Jones, 1943; Pederson and Ward, 1949; Etchells and Jones, 1951a; Etchells _et_a1_. , 1951c), and the retardation of their growth allows the more salt -tolerant Aerobacter species to prolif- erate. The addition of sugar to the brine has been found to increase the populations of lactic acid bacteria, but it also results in the increased growth of gas and alcohol -producing yeasts, with a sub- sequent increase in the numbers of bloaters (Jones eta}; , 1940; Etchells _e_t_a_l. , 1951a). Addition of acetic acid (vinegar) to cucumber brines delayed acid production and encouraged yeast fermentation; when lactic acid was added, the acid -forming bacteria were markedly inhibited, and an active and prolonged yeast fermentation resulted in the formation of bloaters (Jones et a1. , 1940). The temperature of the brine, which is usually 80 to 86 F, does not appear to affect the fermentation to any great extent (Etchells and Jones, 1946), since the optimum temperature for the lactic acid bacteria, both gas - producers and non -gas -producers, is between 75 and 86 F (Peder- son and Albury, 1950). Lower temperatures seem to favor growth of Leuconostoc species (Etchells and Jones, 1946; Pederson and Albury, 1950). The control of film yeasts is easily accomplished by ultra - violet light, as in direct sunlight (Rahn, 1931; Binsted e_t_al_. , 1962), but this is not a suitable means of control since the fermentation 1 vats are located outdoors where gross contamination by dust, insects, rodents, and birds and dilution of the brine by rain can occur. Mustard oil, neutral mineral oil, or liquid paraffin could be floated on the surface of the vats in a thin layer (Blum and Fabian, 1943; Cruess, 1958; Binsted 3131. , 1962), but it is difficult to remove the last traces of these from the cucumbers after fermenta- tion. The volatility of mustard oil also precludes its use. Phillips and Mundt (1950) suggested using sorbic acid as a means of preventing the growth of scum yeasts, because it is non- toxic to humans and has no flavor or odor. Sorbic acid (2, 4- hexadienoic acid) appears to selectively inhibit all catalase positive bacteria, yeasts, and molds by suppression of fumarate oxidation (York and Vaughn, 1955). The inhibition by sorbic acid depends on the concentration of undissociated acid, which is greatest at pH 5. 0 or less (Costilow, Ferguson, and Ray, 1955; Etchells, Bell, and Borg, 1955). Levels of 0. 1% sorbic acid were found to effectively depress the growth of both fermentative and oxidative yeasts (Phillips and Mundt, 1950; Borg, Etchells, and Bell, 1955; Costilow 3131; , 1955); but some species of Candida, particularly 2. krusei, were still present in the sorbic acid treated brines (Borg fl} , 1955; Costilow et al. , 1955; Etchells, Borg, and Bell, 1961). Phillips and Mundt (1950) and Borg et_al_. (1955) reported reduced populations of acid —forming bacteria and a subsequently slower acid development in brines to which 0. 1% sorbic acid had been added. Experiments by Costilow fl. (1957b) showed that the adverse effect of sorbic acid on the acid fermentation increased with salt concentration, and that a level of 0. 03% sorbic acid would still inhibit yeasts without having an effect on the acid fermentation. However, Borg et_al. (1955) and Costilow (1957a) reported that all levels of sorbic acid slowed the rate and quality of the cure, and yielded pickles with inferior-color. Costilow _e_t_al. (1956) found that L. plantarum was slightly less active than L. brevis in sorbic acid treated brines. Etchells, Borg, and Bell (1968b) have con- firmed that bloaters still occur in sorbic acid treated brines as a result of the gaseous fermentation by L. brevis. Softening of pickle stock has also been a serious problem to the pickle industry. This problem has been thoroughly reviewed by Bell, Etchells, and Jones (1950); Etchells and Jones (1951b); Cruess .(1958); and Binsted et_al_. (1962). Pasteurization of unfermented, partially fermented, and fermented genuine dill pickles has been advocated for their preserva- tion (Etchells and Jones, 1942, 1944, 1951b). Heating the product to an internal temperature of 165 F for 15 min, followed by rapid cooling, was found to destroy vegetative microorganisms and to inactivate enzyme systems without adversely affecting the texture and crispness of the pickles. Pasteurization actually improved the tex- ture of genuine dill pickles, as these will soften during storage if not pasteurized (Jones eial. , 1941). More recent studies on the influ- ence of various acidities and over- and under-pasteurization of fresh pack pickles have been reported (Monroe e_t_a_l_. , 1969; Nicholas and Pflug, 1961). It was hoped that pasteurization would provide a standardized procedure for preserving and controlling the quality of the finished products. Pederson and Albury (1961) found that the course of the natural fermentation could be controlled to some extent by the addition of certain species of lactic acid bacteria to cucumber brines. However, because the fermentation depends to such a great 10 extent on the natural flora of the cucumbers, no one brining treatment discussed will consistently give the same quality salt stock from year to year. Pure Culture Fermentation of Cucumbers Etchells et al. (1964) studied the feasibility of a pure culture fermentation in the absence of competitive microorganisms. There was some precedence for such experiments when Engelland (1962) patented a process for the pure culture fermentation of shredded cabbage for sauerkraut. The use of a heat shock (66 to 82 C water) was found to be effective in destroying vegetative microorganisms and insured a rapid rate of acid production by the added pure culture. Acidification of the covering brines to pH 5. 0 or less with lactic acid (resulting in about 0. 10% acidity after equalization) was found to effectively inhibit the growth of sporeforming bacteria, especially anaerobic types, which survived the heat shock. Strains of Pedio- coccus cerevisiae, Lactobacillus plantarum, and Lactobacillus brevis grew to the highest populations, producing the highest amount of acid and lowest pH values. In mixed cultures, 3. cerevisiae grew first, followed by L. plantarum; L. brevis grew at a slower rate and reachedlower maximum populations. L. plantarum was found to consistently produce the highest levels of acid; L. brevis was the 11 lowest acid -producer of the three. Only strains of L. plantarum and E. cerevisiae were found to consistently produce enough acid to insure a brine pH of less than 4.0. L. brevis was also undesir- able due to the production of gas, resulting in hollow pickles, or bloaters. Based on the results of these studies reported in 1964, Etchells, Bell, and Costilow patented the pureculture fermentation process for pickled cucumbers (1968, U. S. Pat. No. 3, 403, 032). The advantages of such a process as described in the patent are 1) complete control over the fermentation, decreasing losses in brine stock due to bloaters, softening, bleaching and putrefaction; 2) a final product of consistent and predictable high quality free of undesirable changes in flavor and texture; 3) the self -limiting nature of the process (i. e. , development of a certain level of acid causes the organisms to die off), which allows pickled cucumbers and other vegetables to be manufactured in retail and wholesale con- tainers; 4) a consistently firm texture in the finished product which is about the same as that of the starting material; and 5) a reduction in fermentation time, as the process is essentially complete within a week. A major problem with pure culture fermented pickles has been flavor, associated with the spicing and organoleptic acceptance 12 of these products (Dr. R. N. Costilow, Department of Microbiology and Public Health, Michigan State University, East Lansing, Michi- gan, personal communication). Consultation withone of the pickle manufacturers in Michigan has confirmed the above analysis of the problem. Analysis of Flavor The course of fermentation of cucumbers either by pure culture or by the natural flora is followed by increases in brine acidity. But the ultimate evaluation of the fermentation is based on - the appearance and taste of the final product. A fermented pickle can easily be distinguished from a non -fermented, or fresh pack, pickle, even if the brines of both contain the same amount of salt and acid. Margalith and Schwartz (1970) discussed the effects of microorganisms on the flavor of fermented products. In order to establish the contribution of a particular species involved in a fer- mentation process to the flavor of the final product, the approximate chemical composition of the flavor components of that product must be known. Recognizing the need for work on the flavor of pure culture fermented pickles, Aurand e_t_a_l. (1965) attempted to identify the flavor constituents introduced by the organisms responsible for the 13 fermentation. Forss et_a_l_. (1962) used GLC to isolate and identify the flavor constituents of fresh cucumbers. Nona -2 -trans, 6 -cis- dienal and hex -2 -enal were found to be responsible for‘the pleasant "green" or "plant-like" flavor of the fresh cucumber, whereas, non -2 —enal apparently contributed to the more unpleasant, astringent flavor. Aurand 3:31. (1965), using GLC to separate and identify the volatile constituents produced during the pure culture fermentation of cucumbers in an unspiced brine, found that formaldehyde, acetalde- hyde, acetone, and propionaldehyde are formed during the fermenta- tion by strains of L. plantarum or E. cerevisiae. Butyraldehyde was also produced during pure culture fermentation by L. plantarum, but not by 3. cerevisiae. A peak representing ethyl alcohol appeared on the chromatograms, but there was some question as to whether this was a product of fermentation, since the jars and caps were rinsed-in 70% ethanol prior to packing. Fleming, Etchells, and Bell (1969) have since analyzed the volatiles produced during the pure culture fermentation of olives in a similar manner. Use of Gas -Liquid Chromatography for Flavor Research The classical approach to flavor analysis was to use distilla - tion of the food or other complicated and time -consuming methods (Burr, 1964). With the advent of GLC, Dimick and Corse (1956) 14 suggested its use as a simple and rapid means of analyzing the volatile constituents of flavor. The principal contributions to flavor are made by aroma and taste. Hornstein and Teranishi (1967) have compiled an excellent review of what is known about the physiology of taste and odor reception. The characteristic properties of food odors are outlined by Weurman and Lunteren (1967). If the sense of smell is deadened in some way, almost all of the flavor of a food disappears. Since the scientific approach to any problem is to identify it by some objective measurement, and since odor is such a vital part of flavor, attempts have been made to identify the volatile compounds which are responsible for the aroma of a food, in order to correlate odor (an objective measurement) with the subjective evaluation of total flavor of the food (Dimick and Corse, 1956; J en- nings, 1957; Teranishi $31.51!." 1963; Burr, 1964; Guadagni 3131., 1966; Rogers, 1966; Hornstein and Teranishi, 1967). Gas -liquid chromatography is especially suitable for separating highly volatile compounds. Flame ionization detectors show a high sensitivity to organic compounds, but are insensitive to inorganic gases and water; are capable of high resolutions; and can be used to detect a tremendous range of samples and sample quan- tities (Mackay, Lang, and Berdick, 1961; Nawar and Fagerson, 1962; Weurman, 1963; Burr, 1964; Nawar, 1966; Ryder, 1966; and 15 Hornstein and Teranishi, 1967). Rapid, accurate, reproducible, low temperature analyses of small volumes of samples can now be made without destroying the sample by extraction and concentration. With- drawing a portion of the headspace vapor above a food provides the simplest, fastest, most precise sampling, with little possibility for artifact formation; and, therefore, most closely resembles organo- leptic testing (Mackay fl. , 1961; Nawar and Fagerson, 1962; Teranishi, Buttery, and Lundin, 1962; Teranishi fl. , 1963; Kepner 3:31. , 1964; and Teranishi and Mon, 1968). Some of the early'work using headspace sampling (Jennings, 1957; Buttery and Teranishi, . 1961; Bassette, Ozeris, and Whitnah, 1962; and Teranishi, Buttery, and Mon, 1964) led to the suggestion that GLC profiles might be used in quality control to follow aroma quality by the appearance or dis- appearance of certain peaks. The volatile sulfides in garlic (Oaks, Hartmann, and Dimick, 1964), potatoes (Gumbmann and Burr, 1964), beer (Brenner, Owades, and Fazio, 1955), cabbage (Bailey 3331., 1961), and onions (Carson and Wong, 1961) have been analyzed by GLC. Several other foods have also been characterized by their aroma profiles: milk (Bassette fl. , 1963), raspberries (Weurman, 1961), orange juices (Wolford £331. , 1963), cocoa beans (Bailey e_’_c_a_l. , 1962), beer (Kloot, Tenny, and Bavisotto, 1958), and sauerkraut (Vorbeck £131. , . 1961). 16 There are, however, certain disadvantages to the use of GLC for flavor-analyses. Even the most sensitive flame ionization detectors, which can respond to as low a concentration as 10 pg/kg (ppb) (Kendall and Neilson, 1964), are not as sensitive as the human nose (Teranishi _e_t_a_l_., 1963, 1964; Bayer, 1966; Weurman and Lunteren, 1967; Hornstein and Teranishi, 1967). The gas chromato- graph can adequately detect only compounds with a fairly high odor threshold, the theoretical limit of detection being about that for n-amyl acetate--0. 05 1.1g per liter in air (Teranishi, 1970). Other highly odorous compounds, such as dimethyl sulfide, n-decanal, methyl mercaptan, and B-ionone, can be detected in much-lower quantities by human sensors (Guadagni, 1963). These compounds, present below the threshold. of the GLC detector, may comprise a mixture which is very characteristic of the flavor of the food product (Elberg, 1967). There is often no correlation between the magnitude of the peak and the contribution of that component to the total flavor or aroma; i. e. , the fraction with the most intense odor, or contain- ing a component which is present in the largest quantity, may give a very small peak (Burr, 1964; Sethi, Nigam, and Rao, 1965; Bayer, 1966; and Guadagni _e_t_al. , 1966). Trace compounds may not appear as a visible peak or could be hidden by a larger peak of a major com- ponent with the same retention time (Merritt and Walsh, 1962; Bayer, 17 1966; Elberg, 1967). The flavor of a food depends not only on which volatiles are present, but also on the concentration and relative ratios of these compounds in the food (Nawar, 1966 ; Weurman and Lunteren, 1967). For these reasons, some scientists (Weurman, 1963; Nawar, 1966; Ryder, 1966) feel that direct vapor or headspace analysis gives an oversimplified picture of the total flavor. The ratio of volatiles in the headspace vapor may or may not be the same as that in which they appear in the food; low threshold compounds may not be detected; and only compounds with low boiling points will be characterized. The detection of low -boiling compounds is important; but these do not duplicate the total flavor of most products (Ryder, 1966). Gas chromatography is an excellent method for separating components of a mixture, but cannot be used to determine the contri- bution of the various components to the flavor or aroma, nor to identify compounds (Burr, 1964; Bayer, 1966). Other methods, such as mass spectroscopy, infrared spectroscopy, or nuclear mag - netic resonance, must be used to identify the flavor volatiles after separation (Hornstein and Teranishi, 1967 ; Elberg, 1967; Scott, 1969). The relative contributions of the different fractions to the total aroma are determined by human judges who must sniff the gas fractions as they leave the chromatograph, and/ or taste known and 18 unknown mixtures of trapped volatiles to determine whether any alterations may have occurred (Burr, 1964; Bayer, 1966; Ryder, 1966; Elberg, 1966; and Teranishi and Mon, 1968). Thus, only experts with years of experience are qualified to make such evalua - tions, and even then, there are apt to be errors due to fatigue when complex mixtures are tested. The American Society for Testing and Materials published an excellent review in 1967 of the correlations between subjective and objective methods of flavor analysis. Boggs gt_a_l_. (1964) were able to correlate an increase in hexanal (as an indicationof rancidity) over potato granules during storage with the judges' evaluation of flavor deterioration. McCarthy 3331. (1963) found correlation of judges' flavor profiles of bananas and their gas chromatographic patterns. Guadagni 31:31. (1966) showed similar correlation between the judges' evaluations of Delicious apple essence and its corre- sponding chromatographic pattern, but found that the smallest chromatographic peak represented the volatile fraction with the most intense odor. Rhodes (1958) and Kendall and Neilson (1964) were not as successful in finding correlation between GLC and subjective odor analyses. Kendall and Neilson demonstrated that GLC may not detect certain odorants in low concentrations where they may pro- duceimportant changes in the odor characteristics, and that it may not give an indication of odor blends or how new odors are produced. 19 Spices: Oils and Oleoresins Spice oils and Oleoresins are used instead of whole spices to flavor most fresh -pack and processed pickles, because they allow more intense and uniform flavor and permit microbiological control. Since spice oils and Oleoresins are added to cucumber brines prior to pure culture fermentation (Etchells 31a}; , 1968a), the constituents of these should be known in order to evaluate changes in the flavor components which might occur during microbial fermentation. Most spice oils and Oleoresins have been fractionated by gas chromatography. Only dill, garlic, capsicum, mustard, caraway, and pimenta will be discussed, since these are the major spices used to flavor dill pickle products. Carvone and phellandrene are the major constituents of oil of dill (Anethum graveolens L. ). Guenther (1950) reported that phellandrene is responsible for the typical odor and flavor of the fresh herb and dominates the total flavor when there is less than 35% carvone present (carvone being found in higher con- centrations in the seed). Commercial oil of dill may be "cut" or standardized with limonene, a terpene of citrus origin, which may equal or exceed on a weight basis the weight of the dill oil (Sethi, Nigam, and Rao, 1965). The odorous components of garlic (Allium sativum) are released by enzymatic action from the parent compound alliin (Jacobs, 20 . 1951a). Enzymes have also been found to be essential to the release of volatiles in other substrates (Weurman, 1961; Fleming 1.311. , 1968). Alliin is converted to allicin with the release of pyruvate and ammonia by the enzyme alliinase (Jacobs, 1951a; Guenther, 1952). Allicin spontaneously decomposes to give methyl- and allyl -sulfides, disulfides, and trisulfides (Oaks 111: , 1964). In contrast to garlic and dill, in which the volatile oils are responsible for the typical odor and flavor, the non -volatiles of capsicum (Capsicum frutescens) are the source of its flavor and "heat" (Rogers, 1966). Capsicum is, therefore, sold as an oleo- resin, containing fixed oil, capsaicin, pigments, sugars, and resins, but no essential oil (Mathew e_t_a_l_. , 1971). Capsaicin, a substituted benzylamine derivative, is the pungent principle of capsicum (Rogers, 1966; Mathew e_t_al., 1971). The extremely volatile isothiocyanates of mustard oils (Brassica_r1i_g_r;a_ or L. juncea) are responsible for their odor and flavor (Guenther, 1949, 1952). Oil of caraway (Carum carvi) contains 50 to 60% carvone, and is usually diluted with limonene, as is dill (Guenther, 1950). Pimenta leaf oil (Pimenta officinalis) has the odor and flavor of allspice, eugenol being present in a concentration greater than 81% for high grade oil (Guenther, 1950). METHODS AND MATERIALS Preparation of Pickles by Pure Culture Fermentation Cultures Three strains of Lactobacillus plantarum, two of Pediococcus cerevisiae, and mixtures of both'were used in these studies. L. plantarum FBB -‘12 and L. cerevisiae FBB -39 were obtained through the courtesy of Dr. R. N. Costilow of the Department of Microbiology and Public Health, Michigan State University, East Lansing, Michi- gan. L. plantarum Nos. 1 and 4 and E. cerevisiae No. 3 were purchased from Chris Hansen Laboratories, Milwaukee, Wisconsin. Cultures were maintained in APT agar stabs with added calcium carbonate, and were transferred once daily in APT broth at least three days prior to their inoculation into cucumber samples. Incu- bation for the active broth cultures was 30 C; maintenance subcul- tures were incubated 24 hr at 30 C, stored at 4 C, and subcultured at 30 -day intervals. 21 22 Brines Covering brines contained water, salt, spices, and food coloring. Food grade lactic acid was added to acidify the brines to approximately pH 5. 0. Each brine contained a small amount of yellow no. 5 (0. 008% w/w) and 6. 7% (w/w) salt, to result in an approximate concentration of 2. 5 to 3. 0% (w/w) salt after equilibra- tion‘with the cucumbers. Two commercial spice formulations, containing only spice oils and oleoresins, and three formulations combining whole spices . with spice oils were used. Various spice mixtures were added to standard salt brines in the following amounts: 1XKD -- single strength kosher dillformula, containing oil of dill weed, oil of mustard, oleoresin -oil of garlic, oleoresin capsicum, and polysorbate 80-- 0. 09% (V/V) 2XKD--double strength kosher dill formula-- 0. 18% (v/v) 3XKD --triple strength kosher dill formula-- 0. 27% (v/v) 2XGD-KD -—double strength garlic and dill, single strength mustard and capsicum-- 0. 18% (v/v) 23 1XDC -- single strength dill chip formula, containing oil of dill weed, oleoresin -oil of garlic, oleoresin capsicum, caraway oil, pimenta leaf oil, and polysorbate 80 - - 0. 47% (v/v) 2XDC --double strength dill chip formula -- 0. 94% (v/v) WSG--—3 g per quart of a mixed pickling spice containing pepper, capsicum, ginger, bay leaves, mustard, cassia, coriander, cloves, dill, celery, and fennel, plus 0. 06% (v/v) oleoresin-oil of garlic WSG2 -- 3 g per quart of selected whole spices (ginger, cassia, cloves, and half of the bay leaves were removed from the whole mixed pickling spice), plus 0. 06% (v/v) oleoresin oil of garlic WSGD2 -- 3 g per quart of selected whole spices, plus 0. 06% (v/v) oleoresin -oil of garlic and 0.07% (v/v) oil of dill weed. The spice oils were added to the covering brines during preparation. The whole spices were soaked in 70% ethanol overnight, dried in a vacuum oven, and added to the jars as they were inoculated, taking care that each jar received only one capsicum berry. 24 The cover brines were heated to 180 F to destroy asporogenous, vegetative mocroorganisms, and were then kept at 4 C until used the following day. Cucumbers Whole no. 2 cucumbers were used in all experiments. The cucumbers, jars, and caps were obtained through the courtesy of a Michigan pickle processor. The cucumbers were stored at 4 C on the day of receipt, and on the next day were washed and packed according to the patent by Etchells 31131. (1968). Cucumbers were ' placed in a wire basket with a wire cover and were immersed in 180-185 F water for five minutes, bringing the surface temperature of the cucumbers to 140-160 F. The blanched cucumbers were drained and packed by hand into sterile quart jars, 560 to 570 g per jar. Plastic gloves dipped periodically into a solution containing 100 ppm chlorine were used in handling the cucumbers. The jars were sterilized separately with aluminum foil covers; the lids were heated at the time of use in 180 F water which had been acidified with food grade lactic acid to pH 3. 0 or less. The jars containing the blanched cucumbers were immediately filled with 375 to 385 ml cold brine, capped, and labeled as to spice formulation and inoculum. After the cold brine and hot cucumbers equilibrated to a temperature 25 of approximately 86 F, an inoculum. of 109 cells of the desired pure culture or mixture of pure cultures was aseptically added to each jar. The jars were tightly recapped and incubated at 86 F (30 C) for four days. After this primary fermentation, the product was stored at room temperature. Six to ten jars per spice formulation were inoculated with each inoculum variable. One jar was used .for each sampling period and extra jars were retained for organoleptic evaluation and chromato- graphic analyses. The same number of control jars were packed for each spice formulation. One control, designated C1, contained - blanched cucumbers, but no inoculum. A second control, C2, also contained blanched cucumbers and was not inoculated, but was acidified with food grade (85%) lactic acid to a level of approximately 0.8% lactic acid. A third control was included in the earliest experi- ments: washed cucumbers, receiving no heat shock, uninoculated and unacidified. These were not included in later experiments because all produced a gaseous natural fermentation, but the pH was greater than 5. 0. Chemical Analyses of Pickles The following chemical analyses were used to follow the fermentation: pH, titratable acidity, reducing sugar, total plate 26 count, and salt. The control jars and all samples were analyzed after 4, 8, and 30 days. One lot of pickleswas also analyzed after 60 days storage. El A Beckman model 1019 research pH meter was used to determine the pH of a portion of each brine sample. Titratable Acidity Ten milliliters of brine were titrated to a phenolphthalein endpoint with 0. 1000 N sodium hydroxide. The results were calcu- lated as per cent lactic acid. Reducing Sugar The sugar was extracted from each sample in the following manner. Equal weights of cucumber and brine were homogenized in a Waring blendor. To 20. 0 g of blended sample were added 2. 0 g of calcium carbonate and 200 -300 ml distilled water. This mixture was boiled for‘thirty minutes, cooled, and then transferred to a 500-ml volumetric flask. Two to three milliliters saturated neutral lead acetate were added slowly, until the supernatant fluid was clear. Distilled water'was added to bring the mixture to a volume of 500 ml; the'solution'was thoroughly mixed and filtered through 27 Whatman No. 2 paper. One gram anhydrous potassium oxalate was added to precipitate the excess lead; the product was mixed and refiltered. One milliliter of the resulting filtrate was diluted with one ml distilled water and analyzed by the Folin -Wu micro method for determining reducing sugars (A. O.A. C. , 1965). Concentrations of 0. 1, 0. 2, and 0. 5 mg standard dextrose solutions were used as a basis for calculating the per cent reducing sugar in the samples. The per cent reducing sugar in the sample was calculated by the following equation: per cent reducing sugar = *1 X d X 105:: EgIIg X 21:: where: Cs = concentration of dextrose standard 3 = absorbance of dextrose standard at 420 nm = absorbance of sample at 420 nm d = dilution factor. The per cent conversion of reducing sugar to lactic acid was calculated as (follows, using the data from cucumbers fermented byL. plantarum FBB-12 and _P_’. cerevisiae FBB-39 (Table 3) as an example: 28 Example: initial reducing sugar: (% sugar in unfermented control) (total wt. cucumbers + brine in jar) 100 = g initial reducing sugar in jar (1' 2133955—9 = 12. 1 g reducing sugar initially final or residual sugar: (0-4913355fl = 4, 7 g residual reducing sugar reducing sugar utilized: (12.1 g) - (4. 7 g) = 7. 4 g reducing sugar utilized theoretical yield of lactic acid, assuming 100% conversion: 7. 4 g lactic acid from 7. 4 g sugar actual yield of lactic acid: 0. 72% acid in brine after fermentation - 0. 10% acid in original brine 0. 62% acid produced by fermentation (0.62) (95m 100 = 5. 9 g lactic acid produced -—'—— X 100 = 80% conversion of sugar to acid Total Count Dilutions of sample brines were plated on All Purpose plus Tween Agar (APT; Difco Laboratories, Detroit), which contains the 29 following constituents per liter of distilled water: 7. 5 g Bacto -yeast extract; 12. 5 g Bacto -tryptone; 10. 0 g Bacto —dextrose; 5.0 g sodium citrate; 0.001 g thiamine hydrochloride; 5.0 g sodium chloride; 5.0 g dipotassium phosphate; 0. 14 g manganese chloride; 0. 8 g magnesium sulfate; 0. 04 g ferrous sulfate; 0.2 g sorbitan monooleate complex; 15 g Bacto -agar; and with final pH 6. 7. Plates were incubated at 30 C and counted after 48 hrs. Colonies were observed for the typical morphology of L plantarum and/ or _1: cerevisiae. Salt One milliliter of sample brine was diluted with four ml of distilled water and the per cent salt determined using Quantab Chloride Titrator No. 1176 (Ames Division of Miles Laboratories, Inc., Elkhart, Ind.). Organoleptic Analyses of Pickles All controls and samples, after complete fermentation, were subjected to an expert panel of a Michigan pickle processing firm who rated the samples on a five -point hedonic scale. Three of the most acceptable samples, flavored with the double strength (2X) dill chip‘spice formulation, and a commercial pickle sample (Kosher- style dill pickles) were submitted to two separate taste panels for flavor evaluation based on a nine -point hedonic scale. 30 The results were analyzed for preference by the analysis of variance procedure (Kramer and Twigg, 1966). Chromatographic Analyses of Pickles Column The column used for the chromatographic analyses was pre - packed by Analabs, North Haven, Conn. The ten-foot stainless steel column was packed with 60 to 80 mesh, acid washed, silanized Chromosorb W which had been coated with 10% Carbowax 20M. The column was conditioned for 72 hours at 210 C prior to injection of the samples. Gas Chroma togr'aph A Hewlett -Packard model 5750 B research chromatograph, equipped with a Moseley 7128 A dual pen strip chart recorder (F&M Scientific Co. , Avondale, Pa.) was used for the analyses. Both flame ionization and electron capture (EC) detectors were used. Energy in the form of beta -radiation was supplied to the electron capture detector by a tritium foil. A stream splitterwas attached to the column exit to send one -half of the effluent to each of the detectors. The operating parameters of the chromatograph were: Column: Support C oating: Sample Size: Carrier Gas: Purge Gas: Hydrogen: Air: Temperatures: Chart Speed: 31 10' X %" stainless steel, prepacked (Analabs) 60 to 80 mesh Chromosorb W, acid washed, silanized 10% Carbowax 20M 0. 5 ml for all samples and standards Helium at pressure of 40 psi 50 ml per min flow rate through column 25 ml per min flow rate through each detector 90% Argon-10% Methane, High Purity (Matheson Co. , East Rutherford, New Jersey), at pressure of 30 psi 63 ml per min through EC detector only 12 psi 33 psi Detectors-- 130 C Injection Port-- 1 10 C Column-- 80 C 1.0 in per min for first 5 min, then 0.25 in per min Sensitivity and Attentuation: 10 X 1 for flame ionization 102 X 4 for electron capture Pulse Interval (EC): 50 rsec 32 Preparation of Samples and Standards Half of the sample jar lids for the pickles made with a 2XDC spice formulation were fitted with a rubber-serum bottle septum. The septa were inserted into a hole drilled in each of the lids and were sealed to the lids with epoxy cement. Spice formulations for chromatographic analyses were prepared in the same fashion as the brines that were added to the cucumbers, but the cucumbers were omitted. Each component of the spice formulation in its proper concentration was added singly to an acidified salt brine of the same composition as that added to the pickle samples prior to fermentation. The brine -spice mixture (385 ml) was added to sterile jars, mixed, heated to 180 F, and cooled. The headspace of these single spices and of a brine con- taining the entire spice formulation was sampled through rubber septa in the jar lids, and the chromatographic profiles obtained were used as standards for comparison with the fermented samples. Headspace gas samples were obtained with a Pressure -Lok series B gas syringe, fitted with a 23 gauge X 2 inch-luer -lok needle (Precision Sampling Corporation, Baton Rouge, La. ). The syringe was washed between samples by withdrawing hexane from five dif - ferent baths, rinsing the barrel assembly several times at each bath, and drying in air between washings. 33 Interpretation of Results No attempt was made to quantitate or to identify the chromatographic peaks. Interpretation of results was based on. a visual examination for the presence or absence of peaks, the relative ratios of peak heights, and retention times. RESULTS AND DISCUSSION Chemical Analyses Changgs in Reducing Sugar, Titratable Acidity, and pH All heated unacidified controls receiving no pure culture inoculum (C1) supported the growth of sporeforming bacteria (long rods, presumably clostridia with some bacilli near the headspace of the jars) which survived the heat shock treatment. The cucumbers became soft, and a pH greater than 5. 0 was produced. These con- trols were not included in subsequent chemical or organoleptic analyses. Thus, the heated, acidified, uninoculated controls (C2) were used as a reference for the amount of reducing sugar originally present in the cucumbers. It was assumed that the reducing sugar content of the control samples did not change, since these were not fermented. Slight variations in sugar were noted (Table 1), but these , could be accounted for in part by the variable amount of reducing sugar-in different cucumbers. The Folin -Wu procedure for deter- mining reducing sugar was used because the principal sugars in 34 35 TABLE 1. -- pH, per cent titratable acidity (calculated as lactic acid), and per cent reducing sugar in heated, acidified, uninoculated control samples in various spice formulations. Formulation ‘ ( %) ( %) 1XKD 3.21 0. 66 1.03 2XKD 3. 12 0.83 1.39 2XGD-KD 3. 13 0.81 1. 28 1XDC 3. 10 0.82 1.41 2XDC 3. l4 0. 82 1. 44 2XDC 3. 11 0. 83 1.14 2XDC 3.14 0.82 1.32 ' WSG 3.13 0.79 1.37 WSG2 3.--1’2 0. 79 1. l8 WSGD2 3. 18 0. 79 1.18 Average 3. 14 0. 80 1.27 36 cucumbers are glucose and fructose, sucrose being present in only trace amounts (Jacobs, 1951b). The average reducing sugar content of all control samples (1.27%) was used as the initial level of Sugar in the jars. The amount of reducing sugar present in the fresh cucum- bers was calculated by dividing the average total grams of sugar per control jar by 570 g, the weight of cucumbers in the jars (see page 28). This figure, 2. 13%, agrees with the value of 2. 17% reported by Jones and Etchells.(1943, Table 1) for cucumbers of the size used in these experiments. 1 Data in Tables 2, 3, and 4 indicate that the different spice formulations had no significant effect on the fermentation produced by the pure cultures. Differences in data obtained when-spice vari- ables were repeated (c. f. , 2XKD and 2XDC) were probably due to the use of different lots of cucumbers and to experimental errOrs in sampling and technique. Due to the bacteriostatic effects of certain spice oils (e. g. , mustard and garlic), the organisms might be expected to grow better and produce a more rapid fermentation in the whole spiced brines. The data indicate only a slight advantage to the fermentation by the FBB-12 strain of L. plantarum (Tables 3 and 4). Also, as the concentration of the spice oil mixtures increased, there was some inhibition of fermentations by both L. plantarum and _P_. cerevisiae (Tables 2 to 4). 37 :55me mica—pop $3 .H bowmafinoummm 62:59:00 mflflfifi mama :5 2030.2 mm @3253on b55505 53385 9E 5 5355853355 5 .5 co m5 255:5 :5 8 8555205 5.83 555.5 682 555 555 555 555 555 35 55.5 55.5 :5 558535 E5 555 E5 555 555 555 55.5 55.5 55.5 5053 555 E5 555 555 555 555 55.5 55.5 55.5 05.3 55 E5 555 555 555 555 55.5 55.5 55.5 033 555 555 555 555 555 555 55.5 55.5 55.5 088 3 .5 55 .5 s5 5 55 .5 55 .5 55 5 55 .5 55 .5 5s .5 092 554 555 555 555 555 555 85 E5 55 55-9955 555 $5 555 555 555 555 55.5 55.5 55.5 55555 35 55.5 555 $5 555 555 55.5 55.5 55.5 avg $5 555 -- 555 555 555 55.5 55.5. 55.5 96% 555 555 555 $5 $5 555 55.5 55.5 55.5 ovum 55 5 5 55 5 w 55 5 5 mama 9me whom noflflaauoh 00a 35v 35v v a . m Law mas: o A0303 mm m :m . c m 5:335 5335559 .mm- mmh @5330qu mnooooowpom he. 933538.53 moEm 36:9, 5 mumngoso mo :oflflnmfiuom c.5330 9:5 mfinspummsm wfiosvmn «coo own can .335 0303 mm powwgofimov 35305 3855533 .58 non .mg 5 momcmav: .m admin. 38 .5596 mfiospou 555m .5 55535555555550.5555? 5655535500 55555555555 9555.5 :5 2030555 m5 553555550553 35305 5335.55.55 55.5 .o 33555553555555 6 .m .50 mg 5.55555 5555 3 62.55305 who? mmfium 5302 555 55.5 55.5 555 55.5 55.5 5.5.5 55.5 55.5 555.835 55.5 55.5 55.5 555 55.5 555 55.5 55.5 55.5 505.3 55.5 55.5 55.5 55.5 55.5 55.5 55 .5 55.5 55.5 653 55.5 555 55.5 555. 555 55.5 55.5 55.5 55.5 0955 555 55.5 55.5 55.5 55.5 55.5 55.5 55.5 55.5 0855 55 .5 5.55 55.5 55.5. 55.5 555 55 .5 55.5 55.5 0855 55.5 55.5 555 55.5 5.55 555 5.55 55.5 55.5 35-55955 55.5 555 55.5 5.55 555 55.5 55.5 55.5 55.5 55555 55 .5 555 555 55.5 55.5 55.5 55.5 55.5 55.5 5555555 55.5 55.5 -- 55.5 55.5 55.5 55.5 55.5 55.5 55555 55.5 55.5 55.5 55.5 55.5 55.5 55.5 55.5 55.5 55555555 55 5 5 55 5 5 55 5 5 .0.de 905D .5..th Goflgsgoh 5E 55 8555 55m 955055 m .0303 m3 Ma 55 . 5 M 5555535 55532555. I: .mmu mmm 55.553.555.58 .Mu paw maummm 55555535553 .4 55055555355500.0555 53 msg#35858 005mm 555505.555, 5 2255555555 .5o 5.505553555555553 55.535550 55.55555 955.556. ummsm wfiospou «coo .525 @5555 .3505 03055 5.5 553555505553 55305 5335.553 #580 .5055 .mg 555 momqmsvuu .m magma. . 2555955 95505695 555 .5 Dmpmgxonaam 55255535500 552.555.5555 55.5.5.5. =5 2030.55 m5 5535.55.55on b55555 55555.: 5.5 .5 55358552555 5 .5 5o 555 5555555 :5 9. 555555555 5.83 5555.555 8552 39 55.5 55.5 - 55.5 55.5 55.5 55.5 55.5 55.5 55.5 555.5535 55.5 55 .5 55.5 55.5 555 55.5 55.5 55.5 55.5 5053 55.5 55.5 55.5 55.5 55.5 55.5 55.5 55 .5 55.5 053 55.5 55 .5 . 55.5 55.5 55.5 5.55 55.5 55.5 55.5 0955 55.5 . 555 555 555 55.5 555 55.5 55.5 55.5 0955 55.5 55.5 555 555 55.5 55.5 55.5 55.5 55.5 0855 555 55.5 555 55.5 555 55.5 55.5 55.5 55.5 555555.855 555 55.5 55.5 55.5 55.5 55.5 55.5 55.5 55.5 555555 55.5 55.5 55.5. 55.5 55.5 55.5 55.5 55.5 55555 55 .5 55.5 -- 55.5 55.5 55.5 55.5 55.5 555555. 55.5 55.5 55.5 55.5 55.5 55.5 55.5 55.5 5555555 55 ,5 5 55 5 55 5 5 .5.de momma 995d Gofiwuaauoh 85555 AOEV A°§v Q 55w: macs m 3303 .013 m 5 . 5 55 5555535 555355555. .muummm 55555535553 mgmownoaowd he. mach—3555.55.58 5.55am 5555.555, :5 5555555555055 .50 aoflflnofiuom 953556 5.555 955.556.553.55 350565.55 «550.com E55 .8505 555.503 5.55 63355353 @5305 035555.33 «5.80 nod .53 :5 momcmsv I- .5 mgmdfl. 40 The data of Tables 2, 3, and 4 are also illustrated in Figures 1, 2, and 3 that show the average decreases in pH and reducing sugar and increases in titratable acidity for all spice formulations during pure culture fermentations by P. cerevisiae FBB -39, £.p1antarum FBB-12, and the mixed cultures. The figures show that the fermentation is rapid during the first two days and is essentially complete after the first week. Lactobacilli are more acid -tolerant and generally continue to produce a small amount of acid, even after 30 days. Tables 5 and 6 include data that indicate the differences in fermentations produced by various strains of P. cerevisiae, k. plantarum, and mixtures of both in a double strength dill chip formulation and a whole spice formulation. Figure 4 summarizes the effects of each of the inoculum variables on the fermentation in the double strength dill chip formulation, showing final titratable acidity and reducing sugar after 60 days. Pediococci produced the least acid and utilizedless than half of the fermentable sugars. L. plantarum strain No. 1 produced the most acid, leaving only a trace of reducing sugar in the cucumbers. The combined pure cultures of L. plantarum strain No. 1 and E. cerevisiae strain No. 3 produced the same level of acid as the Lactobacillus alone. .11: plantarum strains 12 and 4 are intermediate acid -producers. 0.60 0.20 PER CENT TITRATABLE ACIDITY 8 REDUCING SUGAR FIGURE 1. 41 X pH _. 5.0 ‘ TITRATABLE ACIDITY O REDUCING SUGAR 4‘30 l l 2.0 IO 20 30 TIME (DAYS) Changes in pH, per cent reducing sugar, and per cent titratable acidity (calculated as lactic acid) during pure culture fermentation of cucumbers by Pediococcus cerevisiae FBB-39. Hd 42 l.20 i pH __ a: g TITRATABLE ACIDITY 3 5 ' REDUCING SUGAR (D LOO ‘ E T g I o l m a: \ \ as 0.80 .\ _ \ ). t 9. o 4 0.60 __ ‘ m ““ § _1 _______ j m < .— < a: t 0.40 5 5 p. 1,. z m o a: 0.20 m a 0 1 l 0 IO 20 30 TIME (DAYS) FIGURE 2. -— Changes in pH, per cent reducing sugar, and per cerevisiae FBB —39. 5.0 4.0 2.0 cent titratable acidity (calculated as lactic acid) during pure culture fermentation of cucumbers by Lactobacillus plantarum FBB-12 and Pediococcus Hd I.20 I X pH ._ 5.0 I O TITRATABLE ACIDITY I g ‘ . REDUCING SUGAR g LCD 4. ”’ I o 2 I :5 I D I S 0.80 _ 4.0 \ a: \ °° \ >- \ I'- a 0.60 _ \ a ’K < LU ~55. ~ ~ ______ -‘ -— m < z 0.40 _ 3.0 I: t .— I.— z m 0 0.20 0: LL! 0. o I l 2.0 0 I0 20 30 TIME (DAYS) FIGURE 3. -— Changes in pH, per cent reducing sugar, and per 43 cent titratable acidity (calculated as lactic acid) Hd during pure culture fermentation of cucumbers by Lactobacillus plantarum FBB-12. 44 055356.500 0550000055555 u on “55555535555555 05555505553054 u AA .5596 95505605 555 .5 55505555555550.5005 025535.500 55555555555 5.5.55.5. 55.5 20550.55 mm 5535555055500 0.55505 555552555 .55 .5 55355555058555 5 .5 5o 555 5555555 55.5. 55. 555555555 5.83 5555.555 "352 55.0 mm 0 wmIo 0.0.0 50.0 00.0 05.0 00.0 505.5" 05.5 mm.» 5.5.5. 55 55.5 50.0 50.0 00.0 50.0 55.5 05.5 50.0 50.0 50.5 50.5 55.5 00.5 5 5515 55.0 05.0 $0 55.0 00.0 00.0 55.0 5.0 05.5 00.5 55.0 05.5.. «5 5515 05. .0 mm 0 55.0 500 N5. .0 05. .0 m5. .0 mm .0 05.5 mm .m 5.5 .m 55.5 m 055 + 5. 55.5 m0 .0 «0 .0 N55 .0 55 .0 5.5 .5 5.5 .5 00 .0 05 .0 05 .m 00.5 mm .5 55 .m m 0m + 5 95 mm .0 5.0 .0 mm .0 5.5 .0 55. .0 05 .0 555 .0 :0 05.5 05.. .m 00 .5 mm .m an 05.5 + m5 55.5 55. 0 mm 0 05.0 «0.0 5.5.0 55.0 5.5.0 05.0 00.0 55.5 50.5 05.5 m 055 E. .0 g 0 5.5-0 0.5. .0 mm .0 mm .0 mm .0 55.0 05.5 05 .m 00 .m 05.5 mm 0m 00 05 0 55 00 on 0 5. 00 05 0 5 99505 mhwm 95.50 50555.5..qu . 95555505555055 5.5. 5.50 55 555.555 35855 550 555 5596 was 5635 0.55505. 555355555. 055555556500 5550000055035 0555.. 555553555555 55555550555 ”5% .50 555.550 m=o555>mn 550555555555qu 005550355550 55.50.530.550 03:00 255,553 00.503555055559 555 203555550550 .50 55033550555505 0.535550 0.55555 $55,556 nmgmwfiozcon 55500. .5055 055.5 .8505 0550555 55 0055558553 b50505. 035555.555» 55.500 .5055- .5555 5 0035.550 .2. .m mama. 45 555535.550 960003va u om 58:53:55.5 5:555omn3owa5 n 55.5 £5555 55365.5 .5 55 .5 55355535955 353:8 555555525 5.35. 555 £03055 55 5355505on 5:33 30.32% .55 .o Emfifixouaam .o .5 50 55 53:5 5.. 9. 5.2.5533 983 3&5 "3oz 55.0 05.0 55.0 55.0 05.0 05.0 55.0 55.0 55.5 55.5 55 5 55.5 5 95 50.0 50.0 55.0 55.0 55 .5 55.5 55.0 55.0 55.5 55.5 55.5 55.5 5.5.5 55 0 55.0 55 0 55.0 55.0 55.0 55.0 55.0 55.5 55.5 55.5 55.5 5 am +5 55.5 50.0 50.0 55 0 55.0 05.5 55.5 55 .0 55.0 50.5 50.5 55.5 55 .5 5 om + 5 9.5 55.0 55.0 55.0 55.0 55.0 55.0 55.0 05.0 05.5 55.5 55.5 55.5 5 am 05 05 5 5 05 05 5 5 05 05 5 5 5.5505 .5me mme 8535qu 555588.855 550 550 Mo 5255 355055 53 mm 555:5 5505635 .3885 23.835. 6.55530qu 55800005595 new 855325555 5555505330515 .50 555.55 5565.35 be. 5555 555 055.555 5559 5505555 550553 5.305505 55? 55.59555 255.5 m :5 5.5555538 50 5835:5853 0.53555 5.55:5 55.556 55555. @5505va 280 Log 23 .5305 055055 55 535550553 55305 53.85.33 5580 .35 .5555 :5 5595.950 1. .5 "5.555555. 46 555530.550 5300000505n5 0:5 830552550 555550503055 .50 5:55.55 5305055, .3 005008.555 555503 E 5.550 05 555.55 50505 055055 55 05555505500 550505 53555053 0:5 55535 5503000 5:50 000 555mm -- .5 mmDOHM 530.003 53539.00 5320000500.. 520.542.0200 5000002051 f0 f0 _ +5 5+. to \ \\\\ N 1 05.0 \\\\\\\\\ ., \ \\\\\\\\ \\\\\\\\ 0¢.0 1 00.0 .LNBO 83d \\\\\KE\§\\\\\\\ k\\\\\\\\\\\\\\\ \ 5 1 00.0 5.4035 02.030wm § >._._0_U< uni/2.5255; 1 05.. 47 The combination of these strains with strains of P. cerevisiae slightly depressed the amount of acid produced. The final pH or titratable acidity may be used to evaluate acid fermentations. Different strains of Lactobacillus will produce varying amounts of acid, but there is not necessarily a correlation between the amount of acid produced and the hydrogen ion concen- tration (Pederson and Bagg, 1944). The final pH depends on the buffering capacity of the medium and on the relative proportion of lactic and acetic acids. (Homofermentative lactobacilli may produce 1 - 10% acetic acid, in addition to lactic acid.) The cucumber -brine system is poorly buffered; thus the hydrogen ion concentration increases rapidly, inhibiting further growth of the. organisms. How- ever, Pederson and Bagg. (1944) indicated that neither the pH nor titratable acidity alonewere limiting for fermentation. These authors contended that the combined effects of hydrogen ion con- centration, undissociated lactic and acetic acids, and/ or other growth products result in inhibition of fermentation. Since _I_J_. plantarum strain No. 1. is capable of utilizing essentially all the fermentablesugars in cucumbers, the-level of carbohydrate could be limiting in this case. Data showing. the per cent conversion of sugar to lactic acid are presented in Table 7. _I_.. plantarum strain No. I appeared 48 055530.500 55000005095 u 0n5 “850550550 5555505930515 u 0.5 ”05075 55 5 .5 5 .5 5.5. 3.05.3 5 00 + 5 0.5 55 5 .2 5 .2 50. 50053 5 00 + 5 0.5 E 5 .5 5 .5 5.5. 3053 5 0.5 55 5 .2 5 .2 50. 305B 5 0.5 55 5 .5 5 .5 50. .3053 5 00 E 5 -5 5 .5 50. 0855 5 00 + 5 0.5 55 5.3 5 .2 50. 09% 5 00 + 5 0.5 55 5 .5 5 .5 50. 0955 5 0.5 S 5.2 5 .2. 5.5. 0955 5 0.5 55 5 .5 5 .5 50. 0955 5 0.0 55 5 .5 5 .5 50. 0855 55 00 + 55 0.5 55 5 .5 5 .5 50. 0855 55 0.5 5.. 5-5 5 .5 50. 083 55 00 55 5 .5 5 .5 50. :5 0o 53.. 55. 00 + 55 0.5 55 o .5 o .5 E. :5 0o 5:5 55 0.5 55 5 .5 5 .5 N0. :5. 0o 52... 55 00 855505500 0055002000 .550 \ 00050055 .550 \ 00 5555555 05555.5. 005555080005 .05. 0505. 0550515 5 05555 5 0000005055 0035 000.008.50.55 . . . .6 55.3.5 0555556500 5500003095 005 Eaflnfl 555550503055 .50 5553.5 5:05.550? .3 500853050 .50 0005555008005 003530 0025 05.30 0505 055055 9.05595 50 005500280 5000. 00.5 E. .5 "5.555555. 49 to be the most efficient and 3. cerevisiae FEB -39 the least efficient in this respect. Differences in spice formulations had no significant effect on the per cent conversion. No explanationecan be given for the-low value obtained for the conversion of sugar to lactic acid during the fermentation of cucumbers in a 2XDC brine by 3.. cerevisiae FEB -39. Mixtures of pure cultures of various strains of _I: plantarum and _1: cerevisiae had little effect on the final con- version produced during these fermentations. The fact that yields lower than the theoretical 100% conversion were obtained indicates that some of the sugar utilized by the bacteria was converted to other intermediary metabolites, building materials, and energy. The Folin -Wu determination for reducing sugars was not too precise. The standard deviation for values obtained for'any given inoculum variable was much too high to be accounted for by differ- ences in cucumbers alone (see Tables 2 to 6). . This method was chosen to obtain rapid approximations of the reducing sugar-in cucumbers during various stages of fermentation by different pure cultures. Only a value for residual sugarwafter fermentation'would be of interest to the pickle processor. Other methods for deter- mining. reducing sugar should be investigated. Equilibration of salt at 2. 5 to 3. 0% (w/w) was complete in all jars by the fourth day. 50 Plate Counts Total aerobic plate counts for all the inoculum and spicing variables are presented in Tables 8 to 12 (Appendix). There‘was no apparent effect of spice formulation on growth of the micro- organisms. Initial (O-day) plate counts were made by adding 1 ml of the inoculum to 385 ml brine and plating subsequent serial dilu- tions. An initial count was not made for the 3XKD inoculum variables; the inoculum was assumed to be of the order of 106 organisms per ml, since these were of the same lot as the 1XKD and 2XKD samples. Similarly, the-inoculum for 2XGD-KD samples was assumed to be approximately 109 organisms per ml. Initial counts were not determined for lots 3 and 4, due to a shortage of brine. Again, the inoculum was approximately 109 organisms per m1. Plate counts were made for lot 1 spice formulations 1XKD, 2XKD, and 3XKD two days after inoculation. These counts showed increases to about 1 X 109 microorganisms/ml of brine for all the pure culture variables. All populations were declining by the fourth day. This corresponds to the rapid production of acid shownin Figures 1 to 3. The numbers of microorganisms continued to drop after the fourth day by as much as five log cycles by the end of 60 days (see Tables 10 and 11). 51 Organoleptic Evaluations Samples of pickles fermented by each inoculum variable for each spice formulation were evaluated for flavor and accept- ability by members of a Michigan pickle processing firm. The samples were rated on a five -point hedonic scale as excellent (5), good, fair, poor, or not acceptable (1). The unfermented, acidified controls for all the spice formulations were rated "poor" or "not acceptable. " The control samples correspond to a commercial fresh -pack (unfermented) pickle, but were acidified with lactic rather than acetic acid. All of the pure culture fermented pickles had good texture and were crisp. The IX and 2XKD and 2XGD—KD samples were generally rated "fair" with a slight bland flavor. The combina- tion off. cerevisiae FBB -39 and .1: plantarum FBB-12 in the 2XKD flavored product was rated "good" by one of the judges, but "not acceptable” by another. Such differences in preference were noted for all spice formulations. The 3XKD samples were all unacceptable due to the high level of capsicum. Products flavored by the whole mixed picklingtspice plus garlic-were too aromatic. Thus, only selected whole spices were usedJin subsequent whole spice formulations. Addition of oil of dill was necessary because the dill seeds added little flavor unless they were crushed. The WSG2 and WSGD2 formulations resulted in 52 products with very pronounced flavor and a notable "bite" of capsicum. These products were rated higher by some of the judges than others, depending on the intensity of flavor that was acceptable to them. The dill chip formulation appeared to give the best flavored product. The flavor was more intense in the 2XDC samples; the 1XDC samples were slightly bland. The acid fermentation had a decided effect on the flavor of the pickles, since all of the fermented samples could be distinguished from the acidified controls. The fermentations by different strains ofk. plantarum and 3. cerevisiae also had varying effects on the flavor of the final product. In a given spice formulation, strains of 3. cerevisiae produced a more bland flavor. This could have been due to alteration of one of the spices and/ or to the lower level of acid produced by these organisms. The fermentation by k. plantarum strain No. 1 singly and in mixed culture resulted in too high a level of acid to be acceptable. The consensus of the expert panel was that the double strength dill chip (2XDC) formulation in combination with the fermentations produced by either 3. cerevisiae strain No. 3 or 39 or E. plantarum strain No. 4, or by mixed cultures of these, yielded the most acceptable products. Samples of these pickles (2XDC-PC 3, 2XDC-Lp 4, and 2XDC-PC 3 + Lp 4)- were submitted to taste panels for flavor 53 evaluation based on a nine -point hedonic scale (9 = liked extremely; 1 = disliked extremely). The judges individually rated each of the pure culture fermented samples and a commercial kosher—style dill pickle. Statistical analysis of the results of two taste panels indicated no significant difference in preference between any of the pure culture fermented samples or between the pure culture fer- mented samples and the commercial pickle. The F values, calcu- lated for sample variation (from raw data in Tables 15 and 16-- Appendix), are presented in Tables 13 and 14; these were smaller . in both cases than the value required for significance. Results of the first taste panel, Table 13, showed that there was a significant difference (P = 0.05) between judges. These results confirm the differences in preference, depending on the individual, that were noted during evaluation of the samples by the panel of experts. Gas Chromatggraphic Analyses The fermentation of cucumbers by 3. cerevisiae resulted in notable changes in the flavor of the pickles, regardless of spice formulation, that were different from the flavors produced by _I:. plantarum and the _I;.. plantarum-1:. cerevisiae mixed cultures. These changes might be expected to be the result of interaction of 54 TABLE 13. -- Analysis of Variance showing significance of differences between a commercial product and pure culture samples fermented by Lactobacillusplantarum 4, Pediococcus cerevisiae 3, and the combined cultures, based on data from the first hedonic .flavor evaluation. Source of Degrees of Sum of Mean V:;:ir;ce Tabular F Var1ance Freedom Squares Var1ance (F) 5% 1% Total 75 2 15. 41 Samples 3 11.41 3.80 .1.62 2.78 4.17 Judges 18 77.16 4.29 1.83 1.80 2.29 Residual 54 , 126. 84 2. 35 Note: Raw data presented in Table 15 (Appendix). TABLE 14. -- Analysis of Variance showing significance of differences between a commercial product and pure culture samples fermented by Lactobacillus plantarum 4, Pediococcus cerevisiae 3, and the combined cultures, based on data from the second hedonic flavor evaluation. Source of Degrees of Sum of Mean Veg-iagce Tabular F Variance Freedom Squares Variance (F) 5% 1% Total 79 201. 99 Samples 3 2. 24 0.75 0.30 2. 77 4.15 Judges . 19 58. 24 3. 06 1. 23 1.78 2. 26 Residual 57 . 141. 51 2.48 Note: Raw data presented in Table 16 (Appendix). 55 the microorganisms themselves or their metabolic end -products with one or more of the spices. Because of the involvement of sulfur compounds in many biochemical reactions, the changes in flavor could be due to alteration of the volatile sulfur constituents of garlic. Brenner e_t__a_l. (1955) found that beer yeasts produce many undesir- able sulfur compounds in addition to alcohol. These authors showed the interrelationship of the sulfur compounds and how they could be altered by the yeasts' metabolism. Similarly, the metabolism of lactic acid bacteria might be expected to involve transformations of sulfur compounds. A Oaks fl. (1964) used dual channel gas chromatography to separate and identify the volatile sulfur constituents of garlic. Dual channel GC refers to the use of a single column and two dif- ferent detectors. These authors used a flame ionization detector, because it is sensitive to a wide range of organic compounds, but not to inorganic gases or water vapor; and an electron capture detector, which is sensitive to only certain groups of organic com- pounds. The sensitivity of the electron capture detector varies with the electron affinities of such groups as conjugated carbonyls, halogens, and nitro- and sulfur-containing compounds. However, the detector response by electron-capture is affected by oxygen and water vapor (Buttery and Teranishi, 1961). Oaks et al. (1964) found 56 electron capture to bemore sensitive than flame ionization to the volatile sulfides in garlic, particularly the di- and trisulfides. Based on the work of Oaks et_al. (1964), dual channel gas chromatography was used in this experiment to determine whether the differences in the flavor of pickles fermented by pediococci and lactobacilli could be detected objectively. Carbowax 20M (poly- ethylene glycol) was used as a stationary phase because of its inter- actions with polar compounds. The Chromosorb W support was acid -washed to remove trace elements and treated with-dimethyl- dichlorosilane (DMCS, Analabs) to make the surface more inert and thus minimize the adsorption of volatiles by the support surfaces. Certain restrictions were imposed upon the operating parameters by theuse of the dual detectors and the choice of stationary phase. Loads of greater than 5% Carbowax 20M cause substrate "bleed" which affects the sensitivity of the electron capture detector. Sub- strate bleed can be minimized by usinga column temperature less than 100 C. Maximum separation of components was obtained by using acolumn temperature of 80 C. The recommended carrier gas flow for the Hewlett-Packard 5750 B flame detector is 25 ml per min for %-inchtcolumns. A purge gas (90% argon- 10% methane) was requiredfor the EC detector, to prevent overloading the detector cell with-sample. The recommended ratio of purge gas to column 57 effluent was between 2. 5 and 3 to 1. Moisture traps were connected to the gas lines to remove water-vapor from the carrier and purge gases. Representative chromatograms obtained upon GLC head- space analysis of2XDC brine (no cucumbers), an acidified control, and samples fermented by strains of _P cerevisiae and _L_. plantarum singly and in mixed cultures are shown in Figure 5. No significant differences between the chromatograms for any of the pureculture variables or control samples were noted. No peaks were deleted and none were added as a result of the different fermentations. Slight differences in peak heights occurred. but followed no pattern. These differences were probably due to lack of complete control over sample size and/ or carry -over of volatiles adsorbed on the column from sample to sample. The column-was allowed to condi— tion for at least an hour between sample injections and overnight to minimize error dueto residual sample. Although there was no difference among chromatograms of the fermented samples or between the fermented samples and con- trols, it is interesting to note that the addition of cucumbers to the brine resulted in the disappearance of peaks _3; and}: in Figure 5 (2XDC brine). No attempt was made to identify these peaks, except to say that peak a had the same retention time as a small peak 58 .6553 5.53550 58.55.0555 new 5553 50055555505 95555.5 53 5555555555 3555555555 vooam 553588.53 55.535550 5.55555 paw 40.555500 vooam 5535555550355: 55 5005555555505 @5550 :35 £98.53. minnow 55 55B 65.503555 5555.555 55 .50 950555 mowmmnmms 93 .50 mawumopwfiouso .5550 I- .m mmDOHm 59 Um 0.“. n m on. _. + — Q.— 555 .5... 55552.25 52: 20.525555 2x NU Nu m MEDOHE < c. .._ . .5 5 .555 — N! In 505 C. ‘ :6? 2x. 2x— , loo wZEn r 8 x5 .28 5 erO— SSNOdSQH 31VDS TIM 1:43;)an 60 present in chromatograms of the headspace vapors of both the garlic-brine and dill-brine standards. A larger peak with the same retention time as peak _b_was present in the chromatogram of the dill-brine standard. The fact that both garlic and dill standards would contain components with the same retention time, but in different amounts, illustrates the possibilities for masking of small peaks by larger ones which may or may not be identical constituents (Elberg, 1967). With the maximum sensitivity of this instrument and the resolving power of the column used, it was not possible to characterize the differences between the fermentations by the various microorganisms. Perhaps the flavor differences were due to changes in the non -volatile components, or to higher boiling volatile constituents, or to trace levels of volatiles which were not detected. Further work could be done using different stationary phases, such as Apiezon C, Apiezon L, or SF 96-50 (methyl silicone oil), which have been used for GLC analyses of other foods and flavorings. Pederson fl' (1964) showed that changes occurred in the lipid fraction of dill pickles during natural fermentation and suggested that these changes might affect the flavor of the product. Chromato- graphic analyses of the lipid fractions of pure culture fermented pickles might help to characterize the flavor differences. Other 61 suggestions for further investigation include using a larger sample size or concentrating the sample by increasing the vapor pressure (by raising the temperature) or by the addition of sodium sulfate (Bassette 3131. , 1962; Kepner, 1964). Concentrating the headspace vaporwould alter the ratio of the various volatile constituents so that it would not give an accurate picture of their relative importance to the flavor, but might facilitate detection of differences between the fermented samples (Nawar and Fagerson, 1962). SUMMARY AND CONC LUSIONS The purpose of this study was to develop an adequate‘spice formulation for pure culture fermented pickles. The spice formu- lationwas evaluated subjectively by expert and non -expert taste panels and objectively by GLC headspace analysis. Several spicing and inoculum variables were studied. The different spice formulations had little effect on the fermentation by various pure cultures, but the organisms produced .flavor differences in the finished pickles when the same spice formulation was used. Strains of L: plantarum produced the highest levels of lactic acid, and this acid affected the flavor of the final product. The .1: plantarum strain No. 1 produced a pickle which was much too sour. Strains of _P_ cerevisiae produced a bland product, regardless of which spice formulation was used. Samples of headspace vapor in jars of pickles fermented by strains of_1_3_. cerevisiae and _D. plantarum, either singly. or’in mixed cultures, were analyzed by dual channel gas chromatography in an attempt to characterize the flavor differences. Under the'con- ditions of the experiment, no differences between the headspace 62 63 vapors of any of the samples were evident. Therefore, the differences in flavor must have been due to changes in highly flavored volatile constituents present in. trace amounts or to changes in the non -volatile flavor constituents. A spice formulation containing oil of dill weed, oleoresin- oil of garlic, oleoresin capsicum, caraway oil', pimenta leaf oil, and polysorbate 80 was found to produce adequately flavored pure culture pickles when added in a concentration of 0. 94% (v/v) to the covering brines. Samples of pickles processed in brine con- taining this spice formulation and fermented by _13. cerevisiae strain No. 3, L. plantarum strain No. 4, and the mixture of both cultures were evaluated for flavor and acceptability by taste panels. The results indicated no overall significant difference in preference between these samples and a commercial kosher -style dill pickle. The product fermented by _P_. cerevisiae had a milder flavor, whereas the L. plantarum fermented product was more sour. The product fermented by the mixed culture was preferred by repre - sentatives of a Michigan pickle processing firm. since the two cul- tures seemed .to complement the effects of each other on the flavor. _I_.._. plantarum strain No. 4 accentuated certain flavor characteristics, while _P_’. cerevisiae strain No. 3 attenuated otherflavor character- istics, producing a desirable blend of flavors. However, preference in pickles varies solwidely among individuals that these preliminary 64 evaluations indicate any of the three pure cultures (_P_. cerevisiae strain No. 3, _L: plantarum strain No. 4, or the mixture of both) could produce a pickle of commercial potential. The pickle industry should test these products on a more widespread scale to determine whether-the potential market justifies the cost of educating con- sumers to a new product. LITERATURE CITED LITERATURE CITED American Society for Testing and Materials.’ 1967. "Correlation of Subjective -Objective Methods in the Study of Odors and Taste. " STP 440. ASTM, Philadelphia. A.O.A. C. 1965. Official Methods of Analysis. 10th ed. Asso- ciation of Official Agricultural Chemists, Washington, D. C. p. 499. Aurand, L. W., Singleton, J. A., Bell, T. A., and Etchells, J. L. 1965. Identification of volatile constituents from pure- _ culture fermentations of brined cucumbers. J. Food Sci. 30:288. Bailey, S. D., Bazinet, M. L., Driscoll, J. L., and McCarthy, A. I. 1961. Volatile sulfurcomponents of cabbage. J. Food Sci. £363. Bailey, S. D., Mitchell, D. G., Bazinet, M. L., and Weurman, C. . 1962. Studies on the volatile components of different varieties of cocoa beans. J. Food Sci. 21:165. Bassette, R., Ozeris, S., and Whitnah, C. H. 1962. Gas Chrom- atographic analysis of headspace gas of dilute aqueous solutions. Anal. Chem. 34:1540. Bassette, R., Ozeris, S., and Whitnah, C. H. 1963. Direct chromatographic analysis of milk. J. Food Sci. 23:84. Bayer, E. 1966. Quality and flavor by gas chromatography. J. Gas Chromatogr. 4:67. Bell, T. A., Etchells, J. L., and Jones, I. D. 1950. Softening of commercial cucumber-salt -stock in relation to poly- galacturonase activity. Food Tech. 4:157. 65 66 Binsted, R., Devey, J. D., and Dakin, J. C. .1962. Pickle and Sauc_e Making. Food Trade Press, LTD. , London. pp. 78-85, 204-213. Blum, H. B. , and Fabian, F. W. 1943. Spice oils and their com- ponents for controlling microbial surface growth. Fruit Prod. J. and Amer. Vinegar Ind. _2__2_.:326. Boggs, M. M., Buttery, R. G., Venstrom, D. W. , and Belote, M. L. . 1964. Relation of hexanal in vapor above-stored potato granules to subjective flavor estimates. J. Food Sci. _2__9_:487. Borg, A. F., Etchells, J. L., and Bell, T. A. 1955. The influ- ence of sorbic acid on microbial activity in commercial cucumber fermentations. Bacterial. Proc., Abstract. p. 19. Brenner, M. W., Owades, J. L., and Fazio, T. 1955. Determina- tion of volatile sulfur compounds. IV. Further notes on mercaptans. Amer. Soc. Brewing Chemists, Proc. 1955: 125. Burr, H. K. . 1964. Gas chromatography in flavor research: its uses and limitations. Food Tech. fiz1884. Buttery, R. G. , and Teranishi, R. . 1961. Gas -liquid chromatog- raphy of aroma of vegetables and fruit: direct injection of aqueous vapors. Anal. Chem. $331439. Carson, J. F., and Wong, F. F. 1961. The-volatile flavor com- ponents of onions. J. Agric. Food Chem. _9_:140. Christensen, M. D., Albury, M. N., and Pederson, C. S. 1958. Variation in the acetic acid-lactic acid ratio among the glacticacid bacteria. Appl. Microbiol. _6_:316. Costilow, R. N. 1957a. Sorbic acid as a selective agent for cucumber fermentations. . VIII. Evaluation of salt stock from sorbic acid treated cucumber fermentations. Food Tech. 1_1_:591. Costilow, R. N., Coughlin, F. M., Robach, D. L., and Ragheb,. H. S. 1956. A study of the acid -forming bacteria from cucumber fermentations invMichigan. Food Res. 21:27. 67 Costilow, R. N. , Coughlin, F. M., Robbins, E. K., and Hsu, Wen- Tah. 1957b. Sorbic acid as a selective agent in cucumber fermentations. II. Effect of sorbic acid on the yeast and lactic acid fermentations in brined cucumbers. Appl. Micro- biol. 5:373. Costilow, R. N., and Fabian, F. W. 1953. Microbiological studies of cucumber fermentations. Appl. Microbiol. 1:314. Costilow, R. N., Ferguson, W. E., and Ray, S. 1955. Sorbic acid as a selective agent in cucumber fermentations. I. Effect of sorbic acid. on microorganisms associated with cucumber fermentations. Appl. Microbiol. 3:341. Cruess, W. V. 1958. Commercial Fruit and Vegetable Products. 884 pp. McGraw-Hill Book Co. , Inc. , New York. Dimick, K. P., and Corse, J. 1956. Gas chromatography: a new method for the separation and identification of volatile materials in foods. Food Tech. £0360. Elberg, P. L. 1967. Flavour mixtures. Essential trace com- pounds and less important main components. Ind. Chim. Belge. 32:772. Engelland, G. C. 1962. Food processing. U.S. Patent No. 3, 024, 116. Etchells, J. L.,-Bell, T. A., and Borg, A. F. 1955. The influ- ence of sorbic acid on the growth of certain species of yeasts, molds, and bacteria. Bacteriol. Proc. , Abstract. p. 19. Etchells, J. L., Bell, T. A., and Costilow, R. N. 1968a. Pure culture fermentation process for pickled cucumbers. U. S. Patent No. 3, 403, 032. Etchells, J. L., Borg, A. F., and Bell, T. A'. 1961. Influence of sorbic acid on populations and species of yeasts occurring in cucumber fermentations. Appl. Microbiol. 9:139. Etchells, J. L., Borg, A. F., and Bell, T. A. 1968b. Bloater formation by gas -forming lactic acid bacteria in cucumber fermentations. Appl. Microbiol. _1_§_:1029. 68 Etchells, J. L., Borg, A. F., Kittel, I. D., Bell, T. A., and Fleming, H. P. 1966. Pure culture fermentation of green olives. Appl. Microbiol. 1_4_:1027. Etchells, J. L., Costilow, R. N., Anderson, T- E., and Bell, T. A. 1964. Pure culture fermentation of brined cucumbers. Appl. Microbiol. 2523. Etchells, J. L., and Jones, I. D. 1942. Pasteurization of pickle products. Fruit Prod. J. 2_1_:330. Etchells, J. L., and Jones, I. D. 1943. Bacteriological changes in cucumber fermentation. Food Ind. 15:54. Etchells, J. L., and Jones, I. D. 1944. Procedure for pasteurizing pickle products. Glass Packer. 2_3_:519. Etchells, J. L., and Jones, I. D. 1946. Characteristics of lactic acid bacteria from commercial cucumber fermentations. J. Bacteriol. _5_2_:593. Etchells, J. L., and Jones, I. D. 1951a. Progress in pickle research. Part I. Glass Packer. 30:264. Etchells, J. L., and Jones, I. D. 1951b. Progress in pickle research. Part II. Glass Packer. 32:358. Etchells, J. L., Jones, I. D., and Bell, T. A. 1951c. Brigands in brine. Research and Farming. 12:3. Agr. Exp. Sta., Raleigh, North Carolina. Fleming, H. P., Cobb, W. Y., Etchells, J. L., and Bell, T. A. 1968. The formation of carbonyl compounds in cucumbers. J. Food Sci. 33:572. Fleming, H. P., Etchells, J. L., and Bell, T. A. 1969. Vapor analysis of fermented Spanish -type green olives by gas chromatography. J. Food Sci. 3_4:419. Forss, D. A., Dunstone, E. A., Ramshaw, E. H., and Stark, W. 1962. The flavor of cucumbers. J. Food Sci. 2_7:90. Guadagni, D. G., Buttery, R. G. , and Okano, S. 1963. Odour thresholds for some organic compounds associated with flavor. J. Sci. Food Agr. 14:761. 69 Guadagni, D. G., Okano, S., Buttery, R. G., and Burr, H. K. 1966. Correlation of sensory gas-liquid chromatographic measurements of apple volatiles. Food Tech. 22:518. Guenther, E. 1949. The Essential Oils, Vol. II. D. van Nostrand, Co., Inc., New York. pp. 164, 319, 734. Guenther, E. 1950. The Essential Oils, Vol. IV. D. van Nostrand, Co., Inc., New York. pp. 370-377, 573-579, 619-632. Guenther, E. 1952. The Essential Oils, Vol. VI. D. van Nostrand, Co., Inc., New York. pp. 55-69. Gumbmann, M. R. , and Burr, H. K. 1964. Volatile sulfur com- pounds in potatoes. J. Agr. Food Chem. 3:404. Hornstein, I., and Teranishi, R. 1967. The chemistry offlavor. Chem. and Eng. News. 1967:93. ' Jacobs, M. B. 1951a. A study of garlic flavor. Amer. Perfumer. 57:301. Jacobs, M. B. 1951b. The Chemistry and Technolcfiy of Food and Food Products, Vol. II. Interscience Publishers, Inc. , New York. pp. 1215-1216. Jennings, W. G. 1957. Application of gas-liquid partition chroma- tography to the study of volatile flavor compounds. J. Dairy Sci. 42:271. Jones, I. D., and Etchells, J. L. 1943. Physical and chemical changes in cucumber fermentation. Food Ind. 15:62. Jones, I. D., Etchells, J. L., Veldhuis, M. K., and Veerhoff, O. 1941. Pasteurization of genuine dill pickles. Fruit Prod. J. @3304. Jones, I. D., Velhuis, M. K., Etchells, J. L., and Veerhoff, O. 1940. Chemical and bacteriological changes in dill -pickle brines during fermentation. Food Res. _5_:533. Kendall, D. A. , and Nielson, A. J. 1964. Correlation of subjective and objective odor-responses. Ann. N. Y. Acad. Sci. 116:567. 70 Kepner, R. E., and Maarse, H., and Strating, J. 1964. Gas chromatographic headspace techniques for the quantitative determination of volatile components in multicomponent aqueous solution. Anal. Chem. §_6_:77. Kloot, A. P. van der, Tenny, R. I., and Bavisotto, V. 1958. An approach to flavor definition with gas chromatography. Amer. Soc. of Brewing Chemists, Proc. 1958:96. Kramer, A., and Twigg, B. A. 1966. Fundamentals of Quality Control for the Food Industry. 541 pp. Avi Publishing Co. , Inc. , Westport, Conn. McCarthy, A. L, Palmer, J. K., Shaw, C. P., and Anderson, E. E. 1963. Correlation of gas chromatographic data with flavor profiles of fresh banana fruit. J. Food Sci. 2_8_:379. Mackay, D. A. M., Land, D. A., and Berdick, M. 1961. Objective measurement of odor. Ionization detection of food volatiles. Anal. Chem. 391369. Margalith, P. , and Schwartz, Y. 1970. Flavor and microorganisms. _13. D. Perlman (ed. ), Advances in Applied Microbiology, Vol. 12. Academic Press, New York. pp. 35-71. Mathew, A. G., Lewis, Y. S., Jagadishan, R., Nambudiri, E. S., and Krishnamurthy, N. 1971. Oleoresin capsicum. Flavour Ind. _2_(1):23. Merritt, C., and Walsh, J. T. . 1962. Qualitative gas chromato- graphic analysis by means of retention volume constants. Anal. Chem. 34:903. Monroe, R. J., Etchells, J. L., Pacilio, J. C., Borg, A. F., Wallace, D. H., Rogers, M. P., Turney, L. J., and Schoene, E. S. 1969. Influence of various acidities and pasteurizing temperatures on the keeping quality of fresh packdill pickles. Food Tech. 23:71. Nawar, W. W. 1966. Some considerations in interpretation of direct headspace gas chromatographic analyses of food volatiles. Food Tech. 20:115. 71 Nawar, W. W., and Fagerson, I. S. 1962. Direct gas chromatographic analysis as an objective method of flavor measurement. Food Tech. 1_6_:107. Nicholas, R. C., and Pflug, I. J. 1961. Over- and underpasteuri- zation of fresh cucumber pickles. Food Tech. 16:104. Oaks, D. M., Hartmann, H., and Dimick, K. P. 1964. Analysis of sulfur compounds with electron capture/hydrogen flame dual channel gas chromatography. Anal. Chem. 3651560. Pederson, C. S., and Albury, M. N. 1950. The effect of tempera- ture on bacteriological and chemical changes in fermenting cucumbers. N. Y. State Agr. Exp. Sta. (Geneva) Bull. 744. Pederson, C. S., and Albury, M. N. 1961. The effect of pure- culture inoculation on fermentation of cucumbers. Food Tech. 15:351. Pederson, C. S., and Albury, M. N. 1962. The absorption of salt by cucumbers during fermentation. Food Tech. 16:126. Pederson, C. S., and Bagg, J. V. 1944. The cause of variation in the relationship between titratable acidity and hydrogen ion concentration among lactic acid bacteria. J. Bacteriol. 48:559. Pederson, C. S., Mattick, L. R., Lee, F. A., and Butts, R. M. 1964. Lipid alterations during the fermentation of dill pickles. Appl. Microbiol. 1_2_:513. Pederson, C. S. , and Ward, L. 1949. The effect of salt on bacteriological and chemical changes in fermenting cucumbers. N. Y. State Agr. Exp. Sta. (Geneva) Tech. Bull. 228. Phillips, G. F., and Mundt, J. O. 1950. Sorbic acid as inhibitor of scum yeast in cucumber fermentations. Food Tech. 4:291. Rahn, O. 1931. Prevention of scum on brine pickles. The Canner and Dried Fruit Packer. 31:44. 72 Rhodes, J. W. 1958. Sampling method for analysis of coffee volatiles by gas chromatography. Food Res. 2_3_:254. Rogers, J. A. 1966. Advances in spice flavor and oleoresin chemistry. _In. R. F. Gould (ed.), Flavor Chemistry. Advances in Chem. Series 56. Amer. Chem. Soc., Washington, D.C. pp. 203-224. Ryder, W. S. 1966. Progress and limitations in identification of flavor components. In. R. F. Gould (ed.), Flavor Chemistry. Advances in Chém. Series 56. Amer. Chem. Soc., Wash- ington, D.C. pp. 70-93. Scott, R. P. W. 1969. Recent techniques in flavour analysis. Chem. and Ind. 1969:797. Sethi, S., Nigam, M. C., and Rao, P. R. 1965. Studies on chemical and economic utilization of Anethum sowa. Indian Perfumer. 9:17. Teranishi, R. 1970. High resolution gas chromatography in aroma research. Flavour Ind. 1_(1)':35. Teranishi, R., Buttery, R. G., and Lundin, R. E. 1962. Gas chromatography. Direct vapor analyses of food products with programmed temperature control of dual columns with dual flame ionization detectors. Anal. Chem. 3_4:1033. Teranishi, R., Buttery, R. G., Lundin, R. E., McFadden, W. H., and Mon, T. R. 1963. Role of gas chromatography in aroma research. Amer. Soc. Brewing Chemists, Proc. 1963:52. Teranishi, R., Buttery, R. G., and Mon, T. R. 1964. Direct vapor analyses with gas chromatography. Ann. N. Y. Acad. Sci. . 116:583. Teranishi, R. , and Mon, T. R. 1968. The use of gas chromatog- raphy in aroma research. In. Gas Chromatography. Grune‘and Stratton, Inc. , New York. pp. 293-305. Vorbeck, M. L., Mattick, L. R., Lee, F. A., and Pederson, C. S. . 1961. Volatile flavor of sauerkraut. Gas chromato- graphic identification of a volatile acidic off -odor. J. Food Sci. 23:569. 73 Weurman, C. 1961. Gas -liquid chromatographic studies on the enzymatic formation of volatile compounds in raspberries. Food Tech. 1_5_:531. Weurman, C. 1963. Recent developments in food odour research methods. Recent Advances in Food Sci. 3:137. Weurman, C. , and Lunteren, G. van. 1967. Application of gas chromatography to the study of food odours. Aroma—und Geschmacksstoffe in Lebensmitteln. 1967:21. Wolford, R. W., Attaway, J. A., Alberding, G. E., and Atkins, C. D. 1963. Analysis of the flavor and aroma constituents of Florida orange juices by gas chromatography. J. Food Sci. 2_8_:320. York, G. K., II, and Vaughn, R. H. 1955. Site of microbial inhibi- tion by sorbic acid. Bacteriol. Proc., Abstract. p. 20. _ General References Bradley, R. L., and Stine, C. M. 1962. Simple‘device for obtain- ing samples of headspace gas directly from sealed con- tainers for analysis by gas chromatography. J. Dairy Sci. 45:1259. Buttery, R. G., Ling, L. C., and Guadagni, D. G. 1969. Food volatiles: volatilities of aldehydes, ketones and esters in dilute water solution. J. Agr. and Food Chem. 11:385. Davis, P. 1970. A simple method to prevent loss of volatiles during headspace analysis. J. Chrom. Sci. 8: 423. Etchells, J. L., Bell, T. A., and Costilow, R. N. 1966. Con- trolled fermentation improves pickling. Food Processing and Marketing. 1966:150. Guadagni, D. G., Buttery, R. G., Okano, S., and Burr, H. K. 1963. Additive effect of subthreshold concentrations of some organic compounds associated with food aromas. Nature. 200:1288. 74 Guenther, E., Gilbertson, G., and Koenig, R. T. 1969. Essential oils and related products. Anal. Chem. 41(5):4OR. Hoff, J. E., and Feit, E. D. 1964. New technique for functional group analysis in gas chromatography. Syringe reactions. Anal. Chem. 3651002. Lovelock, J. E. . 1963. Electron absorption detectors and technique for use in quantitative and qualitative analysis by gas chromatography. Anal. Chem. 3_5:474. Lunteren, G. van, Straten, S. van, and Weurman, C. 1967. Irregularities in gas chromatographic direct vapour analysis. Int. Colloq. Chem. Coffee, 3rd. (Publ. 1968) pp. 191-196. McNair, H. M., and Bonelli, E. J. 1968. Basic Gas Chromatog- raphy. 303 pp. Varian Aerograph, Walnut Creek, Cali-- fornia. Merritt, C., Bresnick, s. R., Bazinet, M. L., Walsh, J. T., and Angelini, P. 1959. Determination of volatile components of foodstuffs. J. Agr. Food Chem. 1:784. Miller, J. T. 1968. Courseinbasic instrumentation. X. Chroma- tography. Bio-Med. Eng. 3:66. Pederson, C. S. 1936. A study of the species Lactobacillus plantarum (Orla-Jensen) Bergey et al. J. Bacteriol. 31:217. Pederson, C. S. 1949. The genus Pediococcus. New York State Agr. Exp. Sta. (Geneva) Bull. 780. Pflug, I. J. , and Schmidt, E. D. 1965. pH as a function of acetic acid concentration in fresh cucumbers and salt stock pickles. Quarterly Bulletin, Michigan Agr. Exp. Sta., Mich. State Univ., E. Lansing. 48(2):247. Purnell, J. H. , ed. 1968. Advances in Analytical Chemistry and Instrumentation, Vol. 6. Progess in Gas Chromatography. Interscience Publishers, Inc. , New York. 75 Vorbeck, M. L., Albury, M. N., Mattick, L. R., Lee, F. A., and Pederson, C. S. 1963. Lipid alterations during the fer- mentation of vegetables by the lactic acid bacteria. J. Food Sci. 2_8_:495. Woollen, A., ed. 1970. Food Industries Manual. 20th ed. Chemi- cal Publ. Co., Inc., New York. pp. 422 -59. APPENDIX 76 TABLE 8. -- Changes in total plate counts (organisms/ ml) of brines flavored by various spice formulations during pure culture fermentation by Pediococcus cerevisiae FBB -39. ——.—— Spice Days Formulation O 4 8 30 1XKDa 5.5x 1o6 2.0x 108 1.5x 108 > 3.0x 105 2XKDa 3.7 x 106 2.2 x 107 5.7 x 107 > 3.0x 105 2XKDb 2.5x 109 1.1x 108 6.0X 107 4.0x 107 a 7 6 5 .3XKD NODATA 1.1x 10 5.6x 10 >3.o>< 10 b 7 7 . 6 2XGD-KD No DATA 8.8x 10 1.2x 10 2.4x 10 c 8 8 6 1XDC No DATA 3.9x 10 1.1x 1o 2.4x 10 c 8 7 7 2XDC No DATA 5.6x 10 7.2 x 10 1.2 x 10 d 8 8 7 2XDC NO DATA 2.0x 10 1.5x 10 . 1.1x 10 c 8 8 7 WSG No DATA 7.0x 1o 1.3x 10 3.4x 10 d 8 8 7 WSG2 NO DATA 2.2 x 10 3.1x 10 6.0x 10 aLot 1 bLot2 cLot3 d Lot 4 77 TABLE 9. -- Changes in total plate counts (organisms /ml) of brines flavored by various spice formulations during pure culture fermentation by Lactobacillus plantarum FBB-12. Spice Days Formulation 0 4 8 30 1XKDa 9. 5 x 105 .1 x 107 1. 9 x 107‘3 > 3.0 x 105 a 6 6 3 2XKD No DATA .5x 10 1.8x 1o 2.7x 10 2XKDb 7.1 x 108 .2 x 108 6.9x 107 1.6 x 106 3XKDa No DATA .5 x 105 8.0 x 104 2.1x 104 b 7 8 6 2XGD-KD No DATA .8 x 10 . 1.1 x 10 6.2 x 10 c 8 8 7 1XDC No DATA .4>< 10 1.7x 10 9.8x 10 c 8 8 6 2XDC No DATA .9 x 10 7. 2 x 10 5. 6 x 10 2XDCd NO DATA .8 x 107 5. 6 x 107 1. 1 x m6 c 8 8 7 WSG NO DATA .4 x 10 5. 2 x 10 8.1 x 10 d 8 8 7 WSG2 No DATA .0x 10 6.0x 10 7.4x 10 aLot 1 bLot 2 cLot3 dLot 4 eEstimat'ed 78 TABLE 10. -- Changes in total plate counts (organisms/ml) of brines flavored by various spice formulations during pure culture fermentation by Pediococcus cerevisiae FBB -39 and Lactobacillus plantarum FBB-12. Spice Days Formulation 0 4 8 30 1XKDa 5.9x 106 1.9x io8 5.8x 107 > 3.0x 105 2XKDa 2.4x io6 2.4x 107 1.1x 107 >3.ox io5 2XKDb 3.1x io9 1.6x 108 4.3x io7 1.4x 106 a 6 6 4 3XKD NO DATA 8.7 x 10 1.7 x 10 5.7 x 10 b 7 7 _ 6 2XGD-KD No DATA 7.1x 10 2.3x 10 1.9x 10 c 8 8 7 1XDC NO DATA 2.6x 10 1.3x 10 5.3x 10 c 8 8 7 2XDC No DATA 5.1x 10 1.2 x 10 6.9x 10 d 8 7 6 _2XDC No DATA 2.2 x 10 8.2 x 10 1.2 x 10 c 8 8 6 WSG NO DATA 7.0x 10 4.5x 10 3.0x 10 d 8 8 7 WSG2 NO DATA 5.2x 10 3.2 x 10 5.9x io aLot 1 bLot2 cLot 3 d Lot 4 79 m “wvo 0-. as» An oflmgouoo msooooogom u onA “Eaficwa msfiwownopomA n aAm . , . . . e . AM #3me moAXmo woAXom moAxAA moAme ow A . . . e . Q moAXNA poAme woAme woAX¢m moAXmm oA A #3 Xo.m v we XAA. >3 Xm.m r3 Xw.m 3 Xm.m we Xm.m A woAxpm on m . . . . 0 m3 Xv m >3 XA A m3 Xm A m3 Xo m 4295 OZ Dam m on on w w 0 68396.30 wcflcmfiuom mama mo 5.93m .guflcma mgfiomngomA paw mmAmfiwouoo msoooooficom mo mfiwfim 3.3.35 13 :OSSGoSnom 8.2530 8.25 mare—.6 23335898 330 Aflv,fimco$w 83:26.83» 13 veneered $5.5 mo AAE\mEmAamwuov 3560 33m Afloat 5 mowcmnu t- .3 mAmSwB 80 m «0A A ofimrwouoo maoooooApmnA u on “Eagawa maAAAomnopomA u aAm . . . o . . a voAXmm @3va m3wa moAXoA moAXmm av A . . . o 6 Q moAXom moAXm woAXAp woAXmm m3 Xmm DA A . e . 6 a . o + Q moAXmm bcAXm woAXom moAXmN. moAXmA pm m v A . e e o . o + Q vofixmm poflxp woAXmm moAXmm moAXmA pm m A A . . . . . o ¢oAXmm N.oAXm chXrA moAxAm $3er an m 8 on w e c «82530 wfiucofiuom mama mo 585m .Ensfinma mgfloanopomA cam oflmtfiuoo maoooooAconA mo mfiwfim 363.? he coflficmfiuom c.2530 9:5 anti. :2. van 025% mean mooam Eon? @3038 13 62952.“, 85.5 .3 SSFEEGmMro 3550 33a :59‘ 5 momawnu .2. .NA HAmSE. 81 TABLE 15. -- Analysis of Variance of data from the first hedonic evaluation of the .flavor of a commercial product (Y) and three pure culture fermented samples. Samples Judges Lp 4 Totals Y Pc 3 Lp 4 + Pc 3 1 7 6 8 6 27 2 4 7 5 4 20 3 9 8 9 8 34 4 7 6 5 6 24 5 6 4 4 3 17 6 8 4 5 7 24 7 7 6 5 6 24 ‘ 8 7 4 5 3 19 9 8 4 4 4 20 10 4 2 4 8 18 11 8 5 4 6 23 12 2 3 5 6 '16 13 6 7 6 6 25 14 8 5 5 6 24 15 3 8 8 5 24 16 8 5 8 5 26 17 4 6 6 5 21 18 5 5 7 6 23 19 8 4 4 4 20 Sums 119 99 107 104 429 Means 6.2 5.2 5. 6 5.4 Correction factor (C. F.) = 2421. 59 Total sum of squares (O'TZ) = 215. 41 Sum samples squares (O' 2) = 11. 41 F Sum judges squared (O’Jz) = 77. 16 Residual error (O'Rz) = 126. 84 82 TABLE 16. -- Analysis of Variance of data from the second hedonic evaluation of the flavor of a commercial product (Y) and three pure culture fermented samples. Samples Judges Lp 4 Totals + Pc 3 Y Lp 4 Pc 3 1 2 7 8 5 22 2 7 6 8 7 28 3 4 4 6 5 19 4 6 7 7 7 27 5 6 7 6 6 25 6 4 3 5 7 19 7 7 8 4 8 27 8 4 5 6 8 23 9 9 8 3 6 26 10 7 3 7 7 24 11 7 6 7 5 25 12 7 6 7 5 25 13 6 5 7 5 23 14 5 7 5 6 23 15 7 7 8 6 28 16 4 3 8 6 21 17 5 6 7 6 24 18 8 7 7 7 29 19 8 4 2 5 19 20 9 8 6 9 32 Sums 122 117 124 126 489 Means 6. 1 5. 8 6.2 6. 3 Correction factor (C. F.) = 2989.01 Total sum of squares (0' 2) = 201. 99 T Sum samples squared (O'Fz) = 2.24 Sum judges squared (O'Jz) = 58.24 2 Residual error (O'R ) = 141.51 MICHIGAN STATE UNIVERSITY LIBRARIES III lljlll IILIJIIIH 3 1293 3196 4 74 i