h~ternatiomdJoumalofFoodMicrobioko.% 15 (1992) 77-85
77
'g') 1992 Elsevier ~cience Publishers B.V. All rights reserved 0168-161)5/92/$05.00 FOOD (111457
Evaluation of substrates and storage conditions for preparing and maintaining starter cultures for tempeh fermentation Muambi Shambuyi i, Larry R. Beuchat 2, Yen-Con Hung 2 and Tommy Nakayama 2 I Department of Biology, UnirersiO"of Kinshasa, Kinshasa, Zaire and 2 Food Safety attd Quality Enhancement Laboratoo'. Department of Food Science and Technolo~,~,, Unitersity ¢( Georgia, Agricultural E.~periment Station, Griffin. Georgia, USA (Received 19 June 1991; accepted 14 October 1991)
Heat-pasteurized cassava root, cowpeas, partially defatted peanuts, rice and soybeans were evaluated for their suitability to support growth and sporulation of the tempeh mold, Rhizopus micro~portls var. oligosporus, and the oncom mold, Neurospora intermedia, at 25, 3(i a n d . 3 7oC. The molds grew best and sporulated most luxuriantly on cassava and rice incubated at 37°C. Viability of molds remained high for up to 3(1 weeks when dried (a,. 0.48), powdered substrates on which the molds had been cultured, were stored at 5, 25 and 37°C. Survival was best when powders were stored at 5°C, although storage at 25°C did not cause a marked decline in C F U / g . R. microxporus vat. oligosporus starter cultures produced on rice and then stored for 20 weeks at these temperatures compared most favorably with commercial starter cultures for preparing high-quality soybean tempeh. The simple technology required to prepare these starter cultures enhances the potential for their application in developing countries where the introduction of nontraditional fermented legume foods to low-protein diets could help to ameliorate malnutrition. Key words: T e m p e h ; Rhizopus microsporus; Rtdzopus oligosporu~': Fermentation; Soybean; Peanut; Rice; Cowpea: Cassava
rqeurospora intermedia:
Introduction Tempeh (tempe) is a popular Indonesian food made by fermenting soybeans with Rhizopus microsporus var. oligosponts (R. oligosponts). Other types of beans, cereals and coconut can also serve as fermentation substrates (Winaro and Reddy, 1986). Recent developments in tempeh research have been summarized by Nout and Rombouts (1990). Correspondence address: L R . Beuchat, Food Safety and Quality Enhancement Laboratory, Department of Food Science and Technology, University of Georgia, Agricultural Experiment Station, Griffin, G A 30223-1797, USA.
Traditional inoculum consists of dried, pulverized tempeh from a previous batch, hibiscus leaves on which Rhizopus has been cultured or steamed rice or cassava tlour inoculated with leaf-grown Rhizopus mycelium. Pure starter cultures using sterile soybeans, wheat and rice (Ko and Hesseltine, 1979) have been successfully developed, The storagc life of such starters may be up to one year at 4°C (Medwid and Grant, 1984). However, information on the shell-life and pcrflwmancc of tempeh starlcr ctdtures prepared using pasteurized cassava root, cow~cas, peanuts, rice and soybeans as substratcs is meager or nonexistent. This study was designed to evaluate the suitability of these materials as substratcs tk~r growth and sporulation of R. microsporus var. oligosporus. Survival of spores'in dried, powdered substrate and the quality of tempeh produced using stored starter cultures was investigated. In addition, production and storage stability of Neurospora httermedia starter cultures grown on these substrates was investigated. This mold is used to ferment peanut press cake in the preparation of oncom, and its application to fermented ccrcal- and legume-based foods in dcvcloping countries also holds grcat potential.
Materials and Methods
Source and maintenance of room cultures Rhizopus mictvsporus var. oligosponts (R. oligo~ponts) NRRL 2710 and Neurospora intetvwdia NRRL 2884, obtained from the USDA-ARS Northern Regional Research Laboratory, Peoria, IL, were used in all experiments involving starter culture preparation. Two commercial products, both obtained from The Tempeh Lab (156 Drakes Lane, Summertown, TN, U.S.A), were also evaluated. These were a 'tempeh starter' consisting of R. oligosporus grown on a grain, then dried, ground and combined with wheat flour and 'spore powder' which consisted of a dried, ground R. oligospoms culture grown on a grain substrate, without ad,!ed wheat flour. The NRRL strains were cultured on acidified potato dextrose agar (APDA) (pH 3.5) at 25°C tk~r 3-5 days to induce sporulation and then stored at 4°C until used to prepare inocula. Commercial tempeh starter and spore powder were likewise stored in sealed containers at 4°C until used in laboratory experiments.
Preparation of inocula for making starter cultures R. microsporus vat. oligo~7~orus and N. imermedia were inoculated onto the surface of ADPA in 90-ram Petri dishes and incubated at 25°C for 3 days. Spores ~;f ~. microsporus and conidia of N. intermedia were harvested by depositing 10 ml of sterile water on the surface of each culture and gently rubbing with a sterile bent glass rod. The spore and conidhd suspensions were removed from the cultures and the flooding procedure was repeated twice. Pooled suspensions were diluted with sterile water to ~ive viah!e populations of 104-1(} 5 C F U / m l . Viable populations were determined by surface-plating duplicate 0.1-ml a[iquots on APDA,
79 incubating at 25°C and enumerating colonies ~ffter 3-5 days. Suspensions were used as inocula fi~r substrates used to prepare starter cultures.
Procedure ]br makhtg starter cultures Cassava roots (Manihot utilisshna), cowpeas (Vigna unguiculata), peanuts (Arachis hypogaea ), rice ( Oryza satii'a ) and soybeans ( Glycine max) were evaluated as substrates for preparing starter cultures. The outer 2 mm of cassava roots were removed and roots were cut into 8-ram cubes. Peanuts (Florunner vat.) were heated under forced air at 90°C to loosen skins. Upon cooling to 22°C, skins were removed from cotyledons and discarded. Kernels were partially defatted usitlg a hydraulic Carver press (Fred S. Carver Inc., Menomonee Falls, Wi, USA). Cassava root, cowpeas and soybeans were separately submerged in cold tap water, brought to a boil and cooked tini:~ ~'~l,t~y ~uft a~ judged subjectively by determining the amount of force required to irreversibly compress individual cubes and seeds between the thumb and forefinger. Rice was combined with cold tap water and brought to a boil, whereas partially defatted peanuts were submerged in boiling water for 1 rain but were not subjected to a cook treatment. Upon reaching desired stages of readiness, all substrate materials were thoroughly drained of surface water and cooled to 22°C. Each substrate (1500 g) was inoculated and thoroughly mixed with 10 ml of R. microsporus var. oligosporus spore suspension or N. intermedia conidia suspension. Inoculated substrates were spread in 0.5-cm layers in sterile enamel pans, covered with sterile perforated aluminum foil and incubated at 25, 30 and 37°C. Observations on development of mycelium and production of spores and conidia were made at 2-8 h intervals over a 5-day period.
Procedure jbr dryfltg and storing starter culture powders Cultures grown on substrates at 37°C for 3-4 days were selected for further studies designed to determine survival of spores and conidia during storage in a dry state. The entire mass of substrate, mycelia and spores or conidia was dried in a forced-air oven at 40°C for 16-18 h. The dried material was then ground in a coffee mill and allowed to equilibrate to an a,, of 0.48 (atmospheric relative humidity of 48%) at 30°C for one day. Powders from each mold/substrate combination were hermetically sealed in glass containers to which lids were applied and stored under darkness at 5, 25 and 37°C for up to 30 weeks before being analysed for C F U / g .
Determinatton of t,iability After 0, 1, 2, 4, 8, 12, 20 and 28 weeks, duplicate 1.0-g samples of powdered R. microsporus var. oligosl)orus startel cultures were serially diluted in 0.1 M potassium phosphate buffer (pH 7.0) and surface plated (0.1 ml) in duplicate on APDA. Plates were incubated at 25°C and colonies were counted after 30 h. P6wdered samples prel ared from substrates inoculated with N. intermedia were analysed for C F U / g after 0, 1, 2, 4, 8, 12, 20 and 30 weeks using the same procedure described
80
fi)r analysing samples containing R. microsporus vat. oligosporus, except that colonies were counted after 48 h.
Prepamthm and analysis of tempeh Powdered cassava- and rice-based starter cultures prcpared using R. micro.~porus wtr. oligosporus, as well as commercial (control) tempeh starter and spore powder products, were evaluated for their suitability to make tempeh. Experimental starter cultures had been stored at 5, 25 and 37°C for 20 weeks prior to evaluation for this purpose. Soybeans were soaked in tap water at 22°C for 18 h and then boiled until slightly s/0ft before thoroughly draining and cooling to 37°C. Experimental and commercial starter cultures were added to the soybeans to give initial inoculum populations of approximately 5 X 104 C F U / g . Soybeans and inoculum were thor,ughly mixed, deposited in layers (1.5 to 1.8 cm thick) in perforated polyethylene bags and incubated at 32°C. The external and internal appearance of products was monitored over a 2-day period. Three replicate trials were performed. The pH and protein, lipid and moisture contents of tempeh were determined according to standard procedures described in an earlier report (Beuehat, 1977). Textural quality of three 20-g samples (ca. 2.6 x 6.0 x 1.6 cm) of tempeh was determined after 24 h of fermentation using a Kramer shear-compression test cell attached to an instron Universal Testing Machine (Instron). The lnstron was operated at a crosshead speed of 50 m m / m i n and a chart speed of 100 m m / m i n . Peak force required to shear compress each sample was recorded. Energy required to break and completely shear compress samples was calculated by integrating the area under the force deformation curve up to the completion of the shear compression motion.
Statistical analysis Data presented in this rcport represent means of values obtained from three replicate trials. Statistical analysis of data was performcd using ANOVA and Duncan's; Mt~,[fiplc Range Test procedures of Statistical Analysi~ System (SAS, 1985).
Results and Discussion
Suitability of suhstrate ]or preparing starter culture Ah test substratcs supported growth of R. microsporus var. oiigosporus and N. intermedia at 25, 30 and 37°C. Growth with subsequent spore and conidia production was most rapid at 37°C. Conidia formation by N. intermedia on defatted peanuts was poor. Cassava and rice clearly served as better substrates than
cowpca, peanut om soybean for development of mycclium as well as for spore and conidia formation. Preliminary studies revealed that the ratio of substrate to water during the cooking proccss, as well as the cooking time, markedly affected the rate of growth
81 6'
6'
,,
4 S°C 2S°C37°CI
L) 0
3
2
~
c~J
o
I
Cassava
Cowpea
o~
Q. 0 O.
8
0
1
Rice 0 4 6 1216202428
4
8 12 16 20 24 28
0
4
6 12 16 20 24 28
Storage time (weeks) Fig. 1. Survival of R. microsporus var. oligo.~ponts in dried (I).48 a,.), powdered starter cultmes produced using various seed substrates and stored ;tt 5, 25 and 37°C for up to 28 weeks.
of R. oligo,~I~orus and N. &termedia. Substrates cooked too long were characterized by stickiness, clumping and a generally poor performance. Procedures described abcwe were considered most acceptable for preparing substrates suitable for mold growth. The lower surface :volume ratio of whole seeds, i.e., cowpea and soybean, compared to rice and pressed peanut cotyledons undoubtedly influenced the relatively lower amount of mycelium, spores and conidia produced by test molds. According to Jurus and Sundberg (1976), hyphae of Rhizopus penetrate to a depth of about 25% of the width of a soybean cotyledon. In an.v event, the formation of sporangia by R. rnicrospora var. oligosporus occurred most quickly on cassava and rice substrates as evidenced by the development of gray coloration within 20-26 h of incubation at 37°C.
Surt,ival of R. microsporus t,ar. oligosporus and N. intermedia during storage Survival of R. microsporus var. oligosporus after drying and storing substrates on which it was cultured is illustrated in Fig. 1. Initial C F U / g were !.6 × 10 7 and 3.4 x 10 (' in cassava and rice, respectively, l_x)wer initial populations were noted in powders made from peanut, cownea and soybean substrates. Survival of R, microsporus was best when powde'cs were stored at 5°C, although storage at 25°C did not cause a marked decline in C F U / g during the 28-week test period. The inactivation rate was highest at 37°C. The population, in soybean powder was less than 10 z C F U / g (lower limit of detection) in samples stored for 28 weeks at 37°C. The survival of R. oligosporus spores on glass beads as affected by temperature
82 8 7L 6 \
5
:3
4
U.
3
t.0
2 1
"5
8
~O
6
~
5
"\
,
rs"c 2~'~ 37'c-........... L_..c .................
Cowpea
Cassava
,\
3 2 1
Soybean
Rice
0
4
8
12 16 20 24 28 32
0
4
8
12
16 20
24 28 32
Storage time (weeks) Fig. 2. Survival of N. intermedia in dried (0.48 a , ) , powdered starter cultures produced using various seed substrates and stored at 5, 25 and 37°C for up to 3(1 weeks.
and relative humidity (RH) was studied by Rusmin and Ko (1974). Combinations of high temperature and/or high humidity (45°C/100% RH, 45°C/50% RH, 25°C/100% RH, 4°C/100% RH)caused a rapid decrease in germination percentages during the first few weeks of storage. Viability was maintained at highest levels when spores were stored at 4°C and 0 or 50% RH. Observations on survival of R. microsporus var. oligospoms spores in test powders reported here are in general agreement with those of Rusmin and Ko (1974), using glass beads. The high level of spore viability maintained in all of the test suostrates held at 25°C/48% RH is encouraging, since these conditions are not unlike those of tropical environments in Zaire and other developing countries into which tempeh might be introduced. Shown in Fig. 2 are survivor curves for N. intermedia in cassava, cowpea, rice and soybean powders stored at 5, 25 and 37°C for up to 30 weeks• Peanut samples were not subjected to analysis because they contained vet3' low initial populations of conidia. As observed in experiments witil R. microsporus var. oligosporus, cassava and rice served as the best substrates for growth. These substrates also promoted prolific conidia r.produ':.tion. Storage at 5°C protected N. interrnedia against inactivation, whereas storage at 37°C had a detrimental effect. Less than 102 C F U / g of r;~.c was detected al'Ccr 30 weeks of storage at 37°C. Observatior.~ on survival characteristics of N. intermedia are promising to those intereste.d,_;,i developing starter cultures for making oncom. Low numbers of con.ld-:a'produced on peanut were somewhat surprising, since peanuts are the " substrate of choice for making oncom (Beuchat, 1986). Evidently, mycelium devel-
1,13 opment proceeded at the expense of conidia production, The suitability of N. intermedia starter cultures Ibr making oncom was not evaluated.
Et,ahtation o]" R. microspoms rar oligo,~poms starter cultures Experimental cassava- and rice-based starter cultures containing R. micro,~porus var. oligosporus which had been stored at 5, 25 and 37°C for 20 weeks compared favorably with commercial tempeh starter and spore powder for preparing tempeh. Since the C F U / g of experimental and commercial products differed considerably, th,, amount of inoculum was adjusted to result in approximately 5 x 1{}4 C F U / g of ,,-bean substrate. Thus, any effect inoculum size might have on preparation of tempeh was essentially eliminated. No discernible differences in incubation time required to reach good-quality tempeh were detected in .~oybean substrates containing experimental and commercial inocula. The addition of vinegar (120 grain, I ml per 100 g) to the cooked soybeans before inoculation was helpful in promoting uniform growth. The pH of cooked soybeans decreased from 6.43 to a range of 6.07-6.19 during fermentation. Moisture, lipid and protein contents ranged from 37.4 to 38.0%, 25.5 to 27.0% and 46.8 to 48%, respectively (dry weight basis). The temperature of the fermenting soybeans increased to 34°C during the 2-day incubation period. Statistical analysis of texture measurements indicated that replication and sampling effects were not significant, i.e., samples from the same treatment were uniform. Results of measurements of textural qualities of tempeh made using control and experimental starter cultures are summarized in Table I. The amount of peak force (N) required to shear compress tempeh is an indication of firmness. There were no differences in peak force values for tempeh made using experimental rice-based and control starter cultures. Tempeh made using the experimental
TABLE 1 Objective m e a s u r e m e n t of texture of tempeh made using commercial (control) and experimental R.
micro~portts var. oligosoorus starter cultures Type of starter
Storage
Texture measurement "
culture
temp. (°C)
Peak Force (N)
Energy (J)
T e m p e h starter Spore powder Rice starter n,
5 5 5 25 37 5 25
308 " 303 " 3(13 ,h 312 a 311 ~' 293 b 276 ~
2.36 2.54 2.10 2.09 2.12 1.90 1.80
Cassava starter i,
b ~ c c d c
~' Values in the same column which are not followed by the same letter are significantly different ( P _< 0.05). h Experimental d,'v powde'red ~t~rter cnlture~ were prepared from rice and cassava substrates on which R. micro~porus var. oligosporus was grown at 37°C for 3 - 4 (lays.
84
cassava-based starter culture which had been stored at 25°C, was significantly less firm than all other tempeh products. Tempeh made using cassava-based starter stored at 5°C was less firm than tempeh produced from all other starter cultures except the rice-based starter culture stored ~tt 5°C. These observations may have been duc Io differences in rate of germination of spores which, in turn, would result in differences in the extent of myeelial growth and ~.ubstrate breakdown during the 24-h fermentation period. Voisey (1971) defined chewiness as the energy ( J ) required to masticate a solid food product to a steady state for swallowing. Tempeh prepared using experimental starter cultures was less chewy than tempeh made using control cultures (Table 1). In turn, the use of cassawl-based starter cultures resulted in tempeh exhibiting significantly less chewiness compared to tempeh prepared using rice-based cultures. The induction and activity of various enzymes of R. microsporus var. oligosporus as a result of the cassava and rice constituents may have contributed to changes in chewiness. The starter cultures developed and evaluated in this study are not microbiologically pure. While bacterial populations were not determined, it is certain that spores of BaciUus and ClostrMium species survived and, indeed, were activated by the cooking processes used to prepare the starter cultures. Others have reported that dry rice-based tempeh starter cultures contain 104-105 C F U / g bacteria (Ko, 1985; Tanuwidjaja, 1985). Bacteria may in fact play a significant role in the development of high-quality tempeh (Mulyowidarso, 1988). Nout and Rombouts (199(i) support the use of pure culture starters only under aseptic fermentation conditions, since tempeh substrates are not sterile. From a practical viewpoint, the use of impure starter cultures containing predominantly R. microsporus var. ol~gosporus, not unlike those used in the present study, offer great potential for developing and introducing tetr~peh-making processes in areas of the world where people could benefit greatly by consuming this product of enhanced nutritional quality.
Acknowledgments This work was done while Prof. Dr. Muambi Shambuyi was on a Senior Fulbright Research Fellowship appointment at the University of Georgia. The authors are grateful for the technical assistance of Ms. Brenda Nail a~:d Ms. Kimberly Hortz.
References Beuchat, L.R. (1977) Modification of cooki,:-b=l,i,g properties of peanut flour by enzymatic and chemical hydrolysis. Cereal Chem. 54, 405-414. Beuchat, L.R. (1986) Oncom (fermented peanut press cake). In: N.R. Reddy et al. (Eds.), Legume-Based Fermented Foods. CRC Press, Inc., Boca Raton, FL, pp. 135-144.
~5 Jurus, A.M. and Sundberg, W,J. (1976) Penelration of Rhizopus oligo,sporus into soybeans in tentpeh. Appl, Environ. Microbiol. 32, 284-287. Ko, S.D. (19851 Growth and toxin pro(hlctJon by Psetuhmtonas cocol'enenans, the :+o called 't~ongkrek bacteria,' Asian Food J. I. 78-84. Ko, S.D. and Itessellin¢, C.W. (19791 Tempe anti related foods, in: A.It, Rose (Ed.), Economic Botany° "~'+,~, +. Microbi+ll Biomass. Academic Press, London. pp. 115-14(I. Medwid, R.D. and Grant. D.W, (1984) Gernfinatkm of Rhizopus o/ih,oSlJortts sporangiospores. Appl. Environ. Microbiol. 48, 1067-1071. Mulyowidarso, R.K. (19881 The microbiology and biochemistry of soybean so+lking for tempe fermentation. PhD Dissertation, The University of New South Wales, Australia. 267 pp. Nout, M.J.R. and Rombouts, F.M. (199(I) Recent developments in temps research. J. Appl. Bacteriol, 69, 609-633. Rusmin, S. and Ko, S.D. (19741 Rice-grown RhizolmS oligo,Tmrus inoc.Jlum for tempeh fermentation. Appl. Environ, Microbiol. 28, 347-,350. SAS. (1985) SAS User's Gtlidc: Basics, 5th Edn. Cary, NC. SAS Institute, Inc. Tanuwidjaja, L. (1985) Large scale tempe inoctdum production. Proc. Asian Symp. Nonsalted Soybean Fermentation, Tsukuba, Japan. Tsukuba Science City, Nail. Food Res. inst. pp. 3(15-309. Voisey, P.W. (1971) Use of the Ottawa texture measuring system lk~r testing fish products. Internal Report 7022, Eng. Res. Service. Ottawa, Canltda. Winaro, F.G. and Reddy, N.R. (1986) Tempe. In: N.R. Reddy et al. (Eds.), Legume-Based Fermented Foods. CRC Press, Inc., Boca Raton, FL. pp. 95-117.
77
'g') 1992 Elsevier ~cience Publishers B.V. All rights reserved 0168-161)5/92/$05.00 FOOD (111457
Evaluation of substrates and storage conditions for preparing and maintaining starter cultures for tempeh fermentation Muambi Shambuyi i, Larry R. Beuchat 2, Yen-Con Hung 2 and Tommy Nakayama 2 I Department of Biology, UnirersiO"of Kinshasa, Kinshasa, Zaire and 2 Food Safety attd Quality Enhancement Laboratoo'. Department of Food Science and Technolo~,~,, Unitersity ¢( Georgia, Agricultural E.~periment Station, Griffin. Georgia, USA (Received 19 June 1991; accepted 14 October 1991)
Heat-pasteurized cassava root, cowpeas, partially defatted peanuts, rice and soybeans were evaluated for their suitability to support growth and sporulation of the tempeh mold, Rhizopus micro~portls var. oligosporus, and the oncom mold, Neurospora intermedia, at 25, 3(i a n d . 3 7oC. The molds grew best and sporulated most luxuriantly on cassava and rice incubated at 37°C. Viability of molds remained high for up to 3(1 weeks when dried (a,. 0.48), powdered substrates on which the molds had been cultured, were stored at 5, 25 and 37°C. Survival was best when powders were stored at 5°C, although storage at 25°C did not cause a marked decline in C F U / g . R. microxporus vat. oligosporus starter cultures produced on rice and then stored for 20 weeks at these temperatures compared most favorably with commercial starter cultures for preparing high-quality soybean tempeh. The simple technology required to prepare these starter cultures enhances the potential for their application in developing countries where the introduction of nontraditional fermented legume foods to low-protein diets could help to ameliorate malnutrition. Key words: T e m p e h ; Rhizopus microsporus; Rtdzopus oligosporu~': Fermentation; Soybean; Peanut; Rice; Cowpea: Cassava
rqeurospora intermedia:
Introduction Tempeh (tempe) is a popular Indonesian food made by fermenting soybeans with Rhizopus microsporus var. oligosponts (R. oligosponts). Other types of beans, cereals and coconut can also serve as fermentation substrates (Winaro and Reddy, 1986). Recent developments in tempeh research have been summarized by Nout and Rombouts (1990). Correspondence address: L R . Beuchat, Food Safety and Quality Enhancement Laboratory, Department of Food Science and Technology, University of Georgia, Agricultural Experiment Station, Griffin, G A 30223-1797, USA.
Traditional inoculum consists of dried, pulverized tempeh from a previous batch, hibiscus leaves on which Rhizopus has been cultured or steamed rice or cassava tlour inoculated with leaf-grown Rhizopus mycelium. Pure starter cultures using sterile soybeans, wheat and rice (Ko and Hesseltine, 1979) have been successfully developed, The storagc life of such starters may be up to one year at 4°C (Medwid and Grant, 1984). However, information on the shell-life and pcrflwmancc of tempeh starlcr ctdtures prepared using pasteurized cassava root, cow~cas, peanuts, rice and soybeans as substratcs is meager or nonexistent. This study was designed to evaluate the suitability of these materials as substratcs tk~r growth and sporulation of R. microsporus var. oligosporus. Survival of spores'in dried, powdered substrate and the quality of tempeh produced using stored starter cultures was investigated. In addition, production and storage stability of Neurospora httermedia starter cultures grown on these substrates was investigated. This mold is used to ferment peanut press cake in the preparation of oncom, and its application to fermented ccrcal- and legume-based foods in dcvcloping countries also holds grcat potential.
Materials and Methods
Source and maintenance of room cultures Rhizopus mictvsporus var. oligosponts (R. oligo~ponts) NRRL 2710 and Neurospora intetvwdia NRRL 2884, obtained from the USDA-ARS Northern Regional Research Laboratory, Peoria, IL, were used in all experiments involving starter culture preparation. Two commercial products, both obtained from The Tempeh Lab (156 Drakes Lane, Summertown, TN, U.S.A), were also evaluated. These were a 'tempeh starter' consisting of R. oligosporus grown on a grain, then dried, ground and combined with wheat flour and 'spore powder' which consisted of a dried, ground R. oligospoms culture grown on a grain substrate, without ad,!ed wheat flour. The NRRL strains were cultured on acidified potato dextrose agar (APDA) (pH 3.5) at 25°C tk~r 3-5 days to induce sporulation and then stored at 4°C until used to prepare inocula. Commercial tempeh starter and spore powder were likewise stored in sealed containers at 4°C until used in laboratory experiments.
Preparation of inocula for making starter cultures R. microsporus vat. oligo~7~orus and N. imermedia were inoculated onto the surface of ADPA in 90-ram Petri dishes and incubated at 25°C for 3 days. Spores ~;f ~. microsporus and conidia of N. intermedia were harvested by depositing 10 ml of sterile water on the surface of each culture and gently rubbing with a sterile bent glass rod. The spore and conidhd suspensions were removed from the cultures and the flooding procedure was repeated twice. Pooled suspensions were diluted with sterile water to ~ive viah!e populations of 104-1(} 5 C F U / m l . Viable populations were determined by surface-plating duplicate 0.1-ml a[iquots on APDA,
79 incubating at 25°C and enumerating colonies ~ffter 3-5 days. Suspensions were used as inocula fi~r substrates used to prepare starter cultures.
Procedure ]br makhtg starter cultures Cassava roots (Manihot utilisshna), cowpeas (Vigna unguiculata), peanuts (Arachis hypogaea ), rice ( Oryza satii'a ) and soybeans ( Glycine max) were evaluated as substrates for preparing starter cultures. The outer 2 mm of cassava roots were removed and roots were cut into 8-ram cubes. Peanuts (Florunner vat.) were heated under forced air at 90°C to loosen skins. Upon cooling to 22°C, skins were removed from cotyledons and discarded. Kernels were partially defatted usitlg a hydraulic Carver press (Fred S. Carver Inc., Menomonee Falls, Wi, USA). Cassava root, cowpeas and soybeans were separately submerged in cold tap water, brought to a boil and cooked tini:~ ~'~l,t~y ~uft a~ judged subjectively by determining the amount of force required to irreversibly compress individual cubes and seeds between the thumb and forefinger. Rice was combined with cold tap water and brought to a boil, whereas partially defatted peanuts were submerged in boiling water for 1 rain but were not subjected to a cook treatment. Upon reaching desired stages of readiness, all substrate materials were thoroughly drained of surface water and cooled to 22°C. Each substrate (1500 g) was inoculated and thoroughly mixed with 10 ml of R. microsporus var. oligosporus spore suspension or N. intermedia conidia suspension. Inoculated substrates were spread in 0.5-cm layers in sterile enamel pans, covered with sterile perforated aluminum foil and incubated at 25, 30 and 37°C. Observations on development of mycelium and production of spores and conidia were made at 2-8 h intervals over a 5-day period.
Procedure jbr dryfltg and storing starter culture powders Cultures grown on substrates at 37°C for 3-4 days were selected for further studies designed to determine survival of spores and conidia during storage in a dry state. The entire mass of substrate, mycelia and spores or conidia was dried in a forced-air oven at 40°C for 16-18 h. The dried material was then ground in a coffee mill and allowed to equilibrate to an a,, of 0.48 (atmospheric relative humidity of 48%) at 30°C for one day. Powders from each mold/substrate combination were hermetically sealed in glass containers to which lids were applied and stored under darkness at 5, 25 and 37°C for up to 30 weeks before being analysed for C F U / g .
Determinatton of t,iability After 0, 1, 2, 4, 8, 12, 20 and 28 weeks, duplicate 1.0-g samples of powdered R. microsporus var. oligosl)orus startel cultures were serially diluted in 0.1 M potassium phosphate buffer (pH 7.0) and surface plated (0.1 ml) in duplicate on APDA. Plates were incubated at 25°C and colonies were counted after 30 h. P6wdered samples prel ared from substrates inoculated with N. intermedia were analysed for C F U / g after 0, 1, 2, 4, 8, 12, 20 and 30 weeks using the same procedure described
80
fi)r analysing samples containing R. microsporus vat. oligosporus, except that colonies were counted after 48 h.
Prepamthm and analysis of tempeh Powdered cassava- and rice-based starter cultures prcpared using R. micro.~porus wtr. oligosporus, as well as commercial (control) tempeh starter and spore powder products, were evaluated for their suitability to make tempeh. Experimental starter cultures had been stored at 5, 25 and 37°C for 20 weeks prior to evaluation for this purpose. Soybeans were soaked in tap water at 22°C for 18 h and then boiled until slightly s/0ft before thoroughly draining and cooling to 37°C. Experimental and commercial starter cultures were added to the soybeans to give initial inoculum populations of approximately 5 X 104 C F U / g . Soybeans and inoculum were thor,ughly mixed, deposited in layers (1.5 to 1.8 cm thick) in perforated polyethylene bags and incubated at 32°C. The external and internal appearance of products was monitored over a 2-day period. Three replicate trials were performed. The pH and protein, lipid and moisture contents of tempeh were determined according to standard procedures described in an earlier report (Beuehat, 1977). Textural quality of three 20-g samples (ca. 2.6 x 6.0 x 1.6 cm) of tempeh was determined after 24 h of fermentation using a Kramer shear-compression test cell attached to an instron Universal Testing Machine (Instron). The lnstron was operated at a crosshead speed of 50 m m / m i n and a chart speed of 100 m m / m i n . Peak force required to shear compress each sample was recorded. Energy required to break and completely shear compress samples was calculated by integrating the area under the force deformation curve up to the completion of the shear compression motion.
Statistical analysis Data presented in this rcport represent means of values obtained from three replicate trials. Statistical analysis of data was performcd using ANOVA and Duncan's; Mt~,[fiplc Range Test procedures of Statistical Analysi~ System (SAS, 1985).
Results and Discussion
Suitability of suhstrate ]or preparing starter culture Ah test substratcs supported growth of R. microsporus var. oiigosporus and N. intermedia at 25, 30 and 37°C. Growth with subsequent spore and conidia production was most rapid at 37°C. Conidia formation by N. intermedia on defatted peanuts was poor. Cassava and rice clearly served as better substrates than
cowpca, peanut om soybean for development of mycclium as well as for spore and conidia formation. Preliminary studies revealed that the ratio of substrate to water during the cooking proccss, as well as the cooking time, markedly affected the rate of growth
81 6'
6'
,,
4 S°C 2S°C37°CI
L) 0
3
2
~
c~J
o
I
Cassava
Cowpea
o~
Q. 0 O.
8
0
1
Rice 0 4 6 1216202428
4
8 12 16 20 24 28
0
4
6 12 16 20 24 28
Storage time (weeks) Fig. 1. Survival of R. microsporus var. oligo.~ponts in dried (I).48 a,.), powdered starter cultmes produced using various seed substrates and stored ;tt 5, 25 and 37°C for up to 28 weeks.
of R. oligo,~I~orus and N. &termedia. Substrates cooked too long were characterized by stickiness, clumping and a generally poor performance. Procedures described abcwe were considered most acceptable for preparing substrates suitable for mold growth. The lower surface :volume ratio of whole seeds, i.e., cowpea and soybean, compared to rice and pressed peanut cotyledons undoubtedly influenced the relatively lower amount of mycelium, spores and conidia produced by test molds. According to Jurus and Sundberg (1976), hyphae of Rhizopus penetrate to a depth of about 25% of the width of a soybean cotyledon. In an.v event, the formation of sporangia by R. rnicrospora var. oligosporus occurred most quickly on cassava and rice substrates as evidenced by the development of gray coloration within 20-26 h of incubation at 37°C.
Surt,ival of R. microsporus t,ar. oligosporus and N. intermedia during storage Survival of R. microsporus var. oligosporus after drying and storing substrates on which it was cultured is illustrated in Fig. 1. Initial C F U / g were !.6 × 10 7 and 3.4 x 10 (' in cassava and rice, respectively, l_x)wer initial populations were noted in powders made from peanut, cownea and soybean substrates. Survival of R, microsporus was best when powde'cs were stored at 5°C, although storage at 25°C did not cause a marked decline in C F U / g during the 28-week test period. The inactivation rate was highest at 37°C. The population, in soybean powder was less than 10 z C F U / g (lower limit of detection) in samples stored for 28 weeks at 37°C. The survival of R. oligosporus spores on glass beads as affected by temperature
82 8 7L 6 \
5
:3
4
U.
3
t.0
2 1
"5
8
~O
6
~
5
"\
,
rs"c 2~'~ 37'c-........... L_..c .................
Cowpea
Cassava
,\
3 2 1
Soybean
Rice
0
4
8
12 16 20 24 28 32
0
4
8
12
16 20
24 28 32
Storage time (weeks) Fig. 2. Survival of N. intermedia in dried (0.48 a , ) , powdered starter cultures produced using various seed substrates and stored at 5, 25 and 37°C for up to 3(1 weeks.
and relative humidity (RH) was studied by Rusmin and Ko (1974). Combinations of high temperature and/or high humidity (45°C/100% RH, 45°C/50% RH, 25°C/100% RH, 4°C/100% RH)caused a rapid decrease in germination percentages during the first few weeks of storage. Viability was maintained at highest levels when spores were stored at 4°C and 0 or 50% RH. Observations on survival of R. microsporus var. oligospoms spores in test powders reported here are in general agreement with those of Rusmin and Ko (1974), using glass beads. The high level of spore viability maintained in all of the test suostrates held at 25°C/48% RH is encouraging, since these conditions are not unlike those of tropical environments in Zaire and other developing countries into which tempeh might be introduced. Shown in Fig. 2 are survivor curves for N. intermedia in cassava, cowpea, rice and soybean powders stored at 5, 25 and 37°C for up to 30 weeks• Peanut samples were not subjected to analysis because they contained vet3' low initial populations of conidia. As observed in experiments witil R. microsporus var. oligosporus, cassava and rice served as the best substrates for growth. These substrates also promoted prolific conidia r.produ':.tion. Storage at 5°C protected N. interrnedia against inactivation, whereas storage at 37°C had a detrimental effect. Less than 102 C F U / g of r;~.c was detected al'Ccr 30 weeks of storage at 37°C. Observatior.~ on survival characteristics of N. intermedia are promising to those intereste.d,_;,i developing starter cultures for making oncom. Low numbers of con.ld-:a'produced on peanut were somewhat surprising, since peanuts are the " substrate of choice for making oncom (Beuchat, 1986). Evidently, mycelium devel-
1,13 opment proceeded at the expense of conidia production, The suitability of N. intermedia starter cultures Ibr making oncom was not evaluated.
Et,ahtation o]" R. microspoms rar oligo,~poms starter cultures Experimental cassava- and rice-based starter cultures containing R. micro,~porus var. oligosporus which had been stored at 5, 25 and 37°C for 20 weeks compared favorably with commercial tempeh starter and spore powder for preparing tempeh. Since the C F U / g of experimental and commercial products differed considerably, th,, amount of inoculum was adjusted to result in approximately 5 x 1{}4 C F U / g of ,,-bean substrate. Thus, any effect inoculum size might have on preparation of tempeh was essentially eliminated. No discernible differences in incubation time required to reach good-quality tempeh were detected in .~oybean substrates containing experimental and commercial inocula. The addition of vinegar (120 grain, I ml per 100 g) to the cooked soybeans before inoculation was helpful in promoting uniform growth. The pH of cooked soybeans decreased from 6.43 to a range of 6.07-6.19 during fermentation. Moisture, lipid and protein contents ranged from 37.4 to 38.0%, 25.5 to 27.0% and 46.8 to 48%, respectively (dry weight basis). The temperature of the fermenting soybeans increased to 34°C during the 2-day incubation period. Statistical analysis of texture measurements indicated that replication and sampling effects were not significant, i.e., samples from the same treatment were uniform. Results of measurements of textural qualities of tempeh made using control and experimental starter cultures are summarized in Table I. The amount of peak force (N) required to shear compress tempeh is an indication of firmness. There were no differences in peak force values for tempeh made using experimental rice-based and control starter cultures. Tempeh made using the experimental
TABLE 1 Objective m e a s u r e m e n t of texture of tempeh made using commercial (control) and experimental R.
micro~portts var. oligosoorus starter cultures Type of starter
Storage
Texture measurement "
culture
temp. (°C)
Peak Force (N)
Energy (J)
T e m p e h starter Spore powder Rice starter n,
5 5 5 25 37 5 25
308 " 303 " 3(13 ,h 312 a 311 ~' 293 b 276 ~
2.36 2.54 2.10 2.09 2.12 1.90 1.80
Cassava starter i,
b ~ c c d c
~' Values in the same column which are not followed by the same letter are significantly different ( P _< 0.05). h Experimental d,'v powde'red ~t~rter cnlture~ were prepared from rice and cassava substrates on which R. micro~porus var. oligosporus was grown at 37°C for 3 - 4 (lays.
84
cassava-based starter culture which had been stored at 25°C, was significantly less firm than all other tempeh products. Tempeh made using cassava-based starter stored at 5°C was less firm than tempeh produced from all other starter cultures except the rice-based starter culture stored ~tt 5°C. These observations may have been duc Io differences in rate of germination of spores which, in turn, would result in differences in the extent of myeelial growth and ~.ubstrate breakdown during the 24-h fermentation period. Voisey (1971) defined chewiness as the energy ( J ) required to masticate a solid food product to a steady state for swallowing. Tempeh prepared using experimental starter cultures was less chewy than tempeh made using control cultures (Table 1). In turn, the use of cassawl-based starter cultures resulted in tempeh exhibiting significantly less chewiness compared to tempeh prepared using rice-based cultures. The induction and activity of various enzymes of R. microsporus var. oligosporus as a result of the cassava and rice constituents may have contributed to changes in chewiness. The starter cultures developed and evaluated in this study are not microbiologically pure. While bacterial populations were not determined, it is certain that spores of BaciUus and ClostrMium species survived and, indeed, were activated by the cooking processes used to prepare the starter cultures. Others have reported that dry rice-based tempeh starter cultures contain 104-105 C F U / g bacteria (Ko, 1985; Tanuwidjaja, 1985). Bacteria may in fact play a significant role in the development of high-quality tempeh (Mulyowidarso, 1988). Nout and Rombouts (199(i) support the use of pure culture starters only under aseptic fermentation conditions, since tempeh substrates are not sterile. From a practical viewpoint, the use of impure starter cultures containing predominantly R. microsporus var. ol~gosporus, not unlike those used in the present study, offer great potential for developing and introducing tetr~peh-making processes in areas of the world where people could benefit greatly by consuming this product of enhanced nutritional quality.
Acknowledgments This work was done while Prof. Dr. Muambi Shambuyi was on a Senior Fulbright Research Fellowship appointment at the University of Georgia. The authors are grateful for the technical assistance of Ms. Brenda Nail a~:d Ms. Kimberly Hortz.
References Beuchat, L.R. (1977) Modification of cooki,:-b=l,i,g properties of peanut flour by enzymatic and chemical hydrolysis. Cereal Chem. 54, 405-414. Beuchat, L.R. (1986) Oncom (fermented peanut press cake). In: N.R. Reddy et al. (Eds.), Legume-Based Fermented Foods. CRC Press, Inc., Boca Raton, FL, pp. 135-144.
~5 Jurus, A.M. and Sundberg, W,J. (1976) Penelration of Rhizopus oligo,sporus into soybeans in tentpeh. Appl, Environ. Microbiol. 32, 284-287. Ko, S.D. (19851 Growth and toxin pro(hlctJon by Psetuhmtonas cocol'enenans, the :+o called 't~ongkrek bacteria,' Asian Food J. I. 78-84. Ko, S.D. and Itessellin¢, C.W. (19791 Tempe anti related foods, in: A.It, Rose (Ed.), Economic Botany° "~'+,~, +. Microbi+ll Biomass. Academic Press, London. pp. 115-14(I. Medwid, R.D. and Grant. D.W, (1984) Gernfinatkm of Rhizopus o/ih,oSlJortts sporangiospores. Appl. Environ. Microbiol. 48, 1067-1071. Mulyowidarso, R.K. (19881 The microbiology and biochemistry of soybean so+lking for tempe fermentation. PhD Dissertation, The University of New South Wales, Australia. 267 pp. Nout, M.J.R. and Rombouts, F.M. (199(I) Recent developments in temps research. J. Appl. Bacteriol, 69, 609-633. Rusmin, S. and Ko, S.D. (19741 Rice-grown RhizolmS oligo,Tmrus inoc.Jlum for tempeh fermentation. Appl. Environ, Microbiol. 28, 347-,350. SAS. (1985) SAS User's Gtlidc: Basics, 5th Edn. Cary, NC. SAS Institute, Inc. Tanuwidjaja, L. (1985) Large scale tempe inoctdum production. Proc. Asian Symp. Nonsalted Soybean Fermentation, Tsukuba, Japan. Tsukuba Science City, Nail. Food Res. inst. pp. 3(15-309. Voisey, P.W. (1971) Use of the Ottawa texture measuring system lk~r testing fish products. Internal Report 7022, Eng. Res. Service. Ottawa, Canltda. Winaro, F.G. and Reddy, N.R. (1986) Tempe. In: N.R. Reddy et al. (Eds.), Legume-Based Fermented Foods. CRC Press, Inc., Boca Raton, FL. pp. 95-117.
