Vol. 4:11-17. Evaluation of conditions for fermentation of fish offal T. Mikael Lassén Division ofAnimal Nutrition, Department ofAnimal Science and Animal Health, Royal Veterinary and Agricultural University, Bulowsvej 13, DK-1870 Frederiksberg C., Denmark Conditions for the lactic acid fermentation of fish offal were evaluated regarding the effect of sub- strate concentration (2, 5 and 10% dextrose), preacidification with lactic acid (initial pH of6.8, to 6.5 or 6.0), and inoculum size of Lactobacillus plantarum (10 7 , 10* and 109 colony forming units (cfu)/g). pH and lactic acid production were monitored during a two-week storage period. A small-scale silo for fermenting fish offal was also constructed, and measurement of redox potential was evaluated as a means to estimate bacterial growth conditions. The most favourable conditions for fermentation, manifested by a low and stable pH and high lactic acid production, were achieved with an inoculum size of 10* cfu/g and 5% dextrose. Preacidification did not affect final pH. Redox potential was shown to give a reliable estimate of growth conditions for bacteria under anaerobic conditions by rapidly falling to below -550mV in silage with a high lactic acid concentration. Key words: Lactobacillus plantarum , preservation, animal by-product, silage, lactic acid, redox po- tential, spoilage ntroduction Lactic acid fermentation may be a useful alter- native to deep freezing, chemical acidification and protein concentrate as storage method for animal by-products, provided a product of good hygienic quality can be obtained. To achieve proper fermentation, the material must contain lactic acid bacteria and nutrients necessary for optimal fermentation, and have a temperature within the range of the fermenta- tive bacteria. The nutritional demands of the lac- tic acid bacteria include fermentable carbohy- drates, amino acids, nucleotides and vitamins (Beck 1978). The preserving properties of lactic acid bac- teria (LAB) are mainly due to production of or- ganic acids (Hurst 1973, Hurst and Collins- Thomsson 1979, Speck 1981), the undissociat- ed molecules of which are responsible for the antimicrobial activity, with decreasing pH of the product, the proportion of undissociated mole- cules of organic acid increases (Ingram et al. 1959, Baird-Parker 1980). Further, substances with specific activity against gram-positive bac- © Agricultural Science in Finland Manuscript received February 1994 11 AGRICULTURAL SCIENCE IN FINLAND https://www.c-info.fi/en/info/?token=R6EoAvoVSm1GMOMU.e1mCteAZA72l8_qOnEHOHg.CqEirrYCN3bUABOLDDKgrdTCblkhIVARqfGdOopT1r5kRWazs3dCdO_PaP7bK_390QbvNfGqYzr31M8w_ZosHjYvW0eNKPeVKv2u6s8w-oTX3bMn7xtfzMqxZ_UO9qC172Z21f-3Fu_lnVGIkMnVyyMO1sjiqwCavm2bidcX_DuIOqYftO7HPX6T2y8PYWxfw5xf0PpGUYjl4XQnk3LxkIT302V0KUjyR6eMYfwNLwgo8qA8g-3LEvyOkkYg2bRppUBV_Zo1ZTQLDSm9 Lassén, T. M.: Evaluation ofconditionsforfermentation offish offal teriähave been observed in herring silage (Lind- gren and Clevström 1978). Lactic acid fermentation of animal by-prod- ucts can be achieved by adding LAB and fer- mentable carbohydrates to minced animal by- products. The fermented products are usually stored at temperatures ranging from 20°C to 30°C (Nilsson and Rydin 1965, Mackie et al. 1971, Stanton and Yeoh 1977, Raa 1980, Lind- gren and Pleje 1983). The present investigation was part of a project studying the biopreservation of animal by-products with lactic acid bacteria. This part of the study investigated the optimum level of carbohydrate addition, initial pH and inoculum size for the fermentation process. In addition, a method for measuring redox potential was de- veloped and evaluated as a means for estimat- ing growth conditions for bacteria. Material and methods Fish offal (viscera and heads) from herring (Clu- pea harengus) caught in the North Sea in autumn was obtained from Gilleleje Processing Plant, Gilleleje, Denmark. The offal was minced coarsely through a 10-mm plate, and deep-fro- zen within 6 h. The offal was kept at -20°C until 12 h before it was to be used for fermentation. In experiment 1, batches of herring offal, each weighing 150 g, were prepared for fermen- tation according to a 3x3x3 factorial design re- garding level of fermentable carbohydrate (2, 5 and 10% )(v/w) dextrose), inoculum size (107, 108 , and 109 colony forming units (cfu)/g,Lacto- bacillus plantarum, (L.p), Chr. Hansen’sLab. A/ S Hprsholm, Denmark), and initial pH (original pH of 6.8 in the offal lowered to 6.5 and 6.0, respectively, with 90% (v/v) L-lactic acid, Merck Nr, 366). All combinations of substrate, in- oculum size and initial pH were prepared in trip- licate and treated in a waterbath at 25°C. pH and lactic acid were measured after one, two, three, seven and fourteen days. Lactic acid production in the fermented ma- terial was measuredby titrating 5g of silage sus- pended in 25 ml of distilled water to pH 7 with 0.1 M NaOH. pH was measured using a Radi- ometer PH2B pH meter equipped with a GK24018 probe whereafter the sample was ti- trated with a TTT II titrator (Radiometer A/S, Rpdovre, Denmark). The lactic acid content (%) was determined by multiplying the volume (L) of the NaOH by the molarity (M) of the NaOH and the molecular weight of lactic acid (90.06) and dividing it by the sample weight (Formula 1)The % lactic acid was determinedby subtract- ing the % LA (w/w) before fermentation from the % LA of the fermented sample. (L NaOH x M NaOHx 90.06) x 100 (1)% LA = 5 g (sample weight) In experiment 2 herring offal was fermented with 107 efu/g L.p, 0, 2 and 5% dextrose at 25°C in order to produce a poor quality silage for use in the biochemical studies (Lassén 1995b). Meas- urement of redox potential was tested and eval- uated as a tool for estimating bacterial growth conditions in the product. All silos were prepared in triplicate and subjected to all measurements. Redox potential (E) was measured directly on the material with a platinum electrode (PI 101) and a saturated calomel reference electrode (K401) using a pH meter (PH2B). E at pH 7 was calculated from formula 2 E(pHTO) =Eo + 58.1 + 244. 1X (pH- 1) (2) where E 0 is the actual reading, 244.1 is the potential of the calomel electrode at the temper- ature at which the reading takes place, 58.1 is a constant, and pH is the measuredpH of the sam- ple (Huss and Larsen 1979). Statistical analyses for experiment 1 were performed according to the GLM-procedure in SAS (SAS 1985). Changes in pH development and lactic acid production were evaluated ac- cording to the following split plot design mod- el: 12 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 11-17. Y uum= m+a,+tx + ck + d(ijk) + e,+ (ab)tf + (ac),,+ (ae)., + (bc)y,+ (be) ;, + (ce)„+ (abc),., + (abe),., + (bee),,, + (ace),,, + (abce).,, + e. „m where, Y = the ijklm th observation, (i = general mean, a,= fixed effect of dextrose level, b = fixed effect of inoculum size,J c,= fixed effect of initial pH, d(ijk) = random silo within dextrose level, inoc- ulum size and initial pH, e, = fixed effect of storage time, (ab),y = interaction between dextrose level and inoculum size, (ac)„= interaction between dextrose level and initial pH, (ae)„ = interaction between dextrose level and storage time, (bc) ;, = interaction between inoculum size and initial pH, (be),, = interaction between inoculum size and storage time. (ce)„ = interaction between initial pH and stor- age time, (abc),,, = interaction between dextrose level, in- oculum size and initial pH, (abe),,, = interaction between dextrose level, in- oculum size and storage time, (bce),„ = interaction between inoculum size, in- itial pH and storage time, (ace),„ = interaction between dextrose level, in- itial pH and storage time, (abce) ( ,„ = interaction between dextrose level, inoculum size, initial pH, and storage time, e ... = random error.ijklm Treatment effects were tested against sam- ple effect, using sample within dextrose level, inoculum size and initial pH as an error term. Results The effect of substrate concentration, inoculum size and preacidification on pH and lactic acid production in the silage is shown in Table 1. Generally, the results of measurements on indi- vidual silos within treatments showed very little variation between silos, but treatment effects (dextrose level, inoculum size and initial pH) were all highly significant (P<0.001). Lactic acid production and final pH decreased with increas- ing dextrose concentration, and an increase in inoculum level from 107 to 109 cfu/g resulted in faster lactic acid production and a decrease in pH during the first 24 hours. Slow initial fer- mentation (2% dextrose and 107 cfu/g) led to low- er lactic acid concentrations and high final pH. Despite a significant treatment effect, preacidi- fication ofherring offal before fermentation had little effect on lactic acid production and final pH within the inoculum and dextrose level (Table 1). Results of small-scale silo fermentation (Ex- periment 2) showed that fermentation without added carbohydrate caused pH to decrease from an initial value of 6.8 to 6.4 after 24 h, and then to increase to 6.8 in two days. The final pH after seven days at 25°C was about 6.5, but very little lactic acid was produced in this silage. In silage produced with 107 cfu/g L.p and 5% dextrose, PH decreased from an initial 6.8 to 4.4 in one week; over 2% lactic acid was produced during the first week of fermentation. Lactic acid fer- mentation with 2% dextrose as carbohydrate source produced poor quality silage. Slow ini- tial lactic acid production and a slow initial pH decline resulted in a putrefied product after two weeks (Fig. la). Measurement of redox potential in the fer- mented silage yielded identical results for with- in time and treatment measurements, and showed that the redox potential only decreased slightly in batches without dextrose and then increased to its initial level. With 2% dextrose addition the redox potential decreased from 0 to -550 mV during the first 24 hours but then increased to its initial level within seven days. The redox potential remained below -550 mV throughout the experimental period in the silage produced by addition of 5% dextrose (Fig. lb). 13 AGRICULTURAL SCIENCE IN FINLAND Lassén, T. M.: Evaluation of conditionsforfermentation offish offal Table 1.Effect of level of dextrose addition, inoculum size and preacidification on changes in pH and % lactic acid produc- tion (LA) in fermented herring offal. (Means of 3 measurements per treatment for which standard deviation ranges from 0 to 0.12 for pH and 0 to 0.12 for LA), (cfu = colony forming units) Storage time, days Dextrose Cfu/g 0 1 2 7 14 % pH LA pH LA pH LA pH LA pH LA 6.8 0.0 5.2 0.1 4.8 1.3 5.3 0,9 5.6 0.2 107 6.5 0.0 5.1 0.4 4.7 1.6 5.6 0.3 5.8 0.1 6.0 0.0 5.1 0.4 4,7 1.6 5.1 0.2 5.2 0.2 6.8 0.0 5.1 0.4 4.6 1.9 4.9 1.3 5.0 1.1 2 10* 6.5 0.0 5.0 0.7 4,7 1.6 5.3 0.3 5.2 0.3 6.0 0.0 4.9 1.0 4.6 1.9 5.1 1.3 5.0 1.3 6.8 0.0 4.8 1.3 4.5 2.2 4.3 2.8 4.4 2.5 109 6.5 0.0 4.7 1.6 4.4 2.5 4.4 2.6 4.2 2.5 6.0 0.0 4.6 1.8 4.4 2.5 4.2 3.1 4.3 2.9 6.8 0.0 5.0 0.7 4.7 1.6 4.4 2.5 4.7 1.7 107 6.5 0.0 4.9 1.0 4.7 1.6 4.6 1.9 4.7 1.7 6.0 0.0 4.9 1.0 4,6 1.9 4.5 2.2 4.7 2.0 6.8 0.0 5.0 0.8 4.6 1.9 4.3 2.8 4.6 2.3 5 10" 6.5 0.0 4.9 1.1 4.5 2.2 4.3 2.8 4.5 2.4 6.0 0.0 4.8 1.3 4.5 2.1 4.3 2.7 4.8 1.8 6.8 0.0 4.5 2.2 4,4 2.5 4.1 3.4 4.2 3.0 109 6.5 0.0 4.6 1.8 4.3 2.7 4.1 3.3 4.2 3.0 6.0 0.0 4.5 2.2 4.3 2.8 4.0 3.7 4.2 3.3 6.8 0.0 5.1 0.4 4,7 1.6 4.4 2.5 4.2 3.1 107 6.5 0.0 5.0 0.7 4.7 1.6 4.4 2.5 4.2 3.1 6.0 0.0 4.9 1.0 4.6 1.9 4.4 2.5 4.1 3.4 6.8 0.0 5.1 0.4 4.7 1.6 4.5 2.2 4.4 2.5 10 108 6.5 0.0 4.9 1.0 4.6 1.9 4.4 2.5 4.3 2.8 6.0 0.0 4.8 1.3 4.5 2.2 4.3 2.8 4.2 3.1 6.8 0.0 4.9 1.0 4.5 2.2 4.4 2.5 4.2 3.0 109 6.5 0.0 4.6 1.9 4.3 2.8 4.1 3.4 4.1 3.4 6.0 0.0 4,6 2.0 4.3 2.8 4.1 3.4 4.1 3.6 Discussion The amount of carbohydrate necessary to fer- ment animal offal successfully with Lactobacil- lus plantarum was evaluated with dextrose be- cause it is readily fermentable, and the results would not be confounded by a more complex carbohydrate or by-product source. Dextrose lev- els of 2-10% resulted in decreasedpH after two days. However, 2% dextrose led to putrefied si- lage after one week, which was probably caused by competing bacteria in the silage, which could grow if the pH did not decrease to below 4.5 in two days. Another possibility is that only small amounts of lactic acid were produced during the initial pH decrease. 10% dextrose, on the other hand, led to slower initial fermentation but more 14 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 11-17. stable silage during storage. The slower initial fermentation might be caused by a high osmotic pressure, which also slows initial growth of lac- tic acid bacteria (Nester et al. 1978). Hence, the 5% dextrose addition level was considered the most suitable for fermentation under the condi- tions tested. The results obtained conform with those from fermentation of poultry offal (Url- ings 1992), where 4% dextrose was the substrate level giving the most successful fermentation with Lactobacillus plantarum. The combined effects of decreased pH and increased concentration of undissociated lactic acid molecules inhibit growth of spoilage bac- teria, which explains why more lactic acid was produced with a faster initial pH decrease. Slow initial lactic acid production followed by a slow pH decrease and the availability of substrate fa- cilitate growth of spoilage bacteria. Certain spoilage bacteria can also use lactic acid as an energy source when no fermentable sugars are available (Pelczar et al. 1977). This may result in decreased lactic acid content and putrefaction of the silage (Raa 1980), as was observed here after one week at low dextrose concentrations (2%). The lactic acid concentrations which in- hibit growth of spoilage bacteria and result in a pH of under 4.4 in herring offal have been re- ported by Lindgren and Pleje (1983) and Has- san and Heath (1986) to be above 4.0%. This finding differs from that of the present study, where the highest lactic acid concentration was 3.7%, which corresponded to a pH of 4.0 after one week of fermentation with 109 cfu/g and 5% dextrose. Lactic acid production decreased when 10% dextrose was used as a carbohydrate source, which again may be caused by initial suppres- sion of the growth of lactic acid bacteria owing to high osmotic pressure in the raw material (Hassan and Heath 1986).The decline in the lac- tic acid content of samples prepared with less than 10% dextrose, and also in that of samples with a pH below 4.2, in which growth of spoil- age bacteria could be considered to have ceased, may be explained by hydrolysis of fish bone, which neutralizes the silage. Consequently, more lactic acid has to be produced to retain a low pH. This is, however, not possible when no more substrate is available or if the lactic acid bacte- ria have been inhibited by the high lactic acid concentration (Jay 1986). Fermentation was more effective when the inoculum size was increased from 107 to 109 cfu/ g as shown by a lower final pH and increased lactic acid production. Owing to the great ex- pense of increasing the inoculum size, preheat- ing the offal might be a preferable alternative to improve fermentation with lower inoculum size. Studies by Hassan and Heath (1986), however, showed no advantage to initial fermentation when fish offal was preheated before fermenta- tion. Fig. 1.Fermentation of herring offal with 107 colony forming units/g of Lactobacillus plantarum at 25°C. a) changes in pH % LA. b)— redox potential. 15 AGRICULTURAL SCIENCE IN FINLAND Lassén, T. M.: Evaluation ofconditionsforfermentation offish offal The fermentation process might also be im- proved by pre-acidification with other organic acids as more favourable growth conditions wouldbe created for the lactic acid bacteria. Pro- pionic acid was successfully used by Skrede and Nes (1988) to lower the initial pH in slaughter offal before fermentation with lactic acid bacte- ria. However, in the present study adjusting the initial pH to 6.5 and 6.0 resulted in almost the same final pH values as in non pre-acidified sam- ples. This finding supports theresults ofUrlings (1992), suggesting that pre-acidification is un- necessary for a successful fermentation. Redox potential may be measured to estimate the growth conditions of anaerobic bacteria (Wolstrup 1972) and to determine the oxygen content of biological material (Harrison 1971). Aerobic bacteria such as Bacillus demand high (positive) redox values, whereas anaerobic bac- teria such as Clostridium demand low (negative) redox values (Jay 1986). Redox potential is al- ways positive in the muscle and negative in the viscera of freshly caught herring (Huss and Larsen 1979). The redox potential in warm- blooded animals reaches a negative value short- ly after the animals have been slaughtered and remains negative until aerobic deterioration (Barnes and Ingram 1955, Wirth and Leistner 1970). The results reported here clearly show that within treatment and sample measurements were very reproducible. Appropriate fermentation led to a low and stable redox potential (-550 mV) after two days' fermentation, whereas inappro- priate fermentation led to an increasing redox potentia increased pH and decreased lactic acid concentration after two days' fermentation. Measurements of redox potential thus provide an efficient, fast and inexpensive means of esti- mating bacterial spoilage. The feasibility of using lactic acid fermenta- tion to preserve raw fish by-products was dem- onstrated with herring offal as a model. It was concluded that fermentation was most effective with an inoculum size of 109 cfu/g L.p and 5% dextrose at 25°C, as shown by sufficient initial lactic acid production, a rapid pH decrease and an eventual low and stable final pH. For eco- nomic reasons, however, an inoculum size of 108 cfu/g offal was considered more realistic for bi- opreservation of animal by-products. Measure- ment of redox potential was found to be a relia- ble and simple way of expressing conditions for bacterial growth under anaerobic fermentation and storage. Acknowledgements. Financial support for this study was provided by the Academy of Finland. Sincere gratitude is due to Mrs Birthe lessen of Chr. Hansen’s Lab A/S, Den- mark for providing the starter cultures used in the study, to Associate Professor Niels Enggaard Hansen for valuable suggestions and commentson the manuscript, and to Asso- ciate Professor Anne-Helene Tauson for her constructive criticism, and for her skilful help with the statistical ana- lysis. References Baird-Parker, A. C. 1980. Organic Acids. In: Silliker, J. H. (ed.). Microbial Ecology of Foods. Vol. 1. Academic Press, London, p. 126-135. Barnes, E. M. & Ingram, M. 1955. Changes in the oxi- dation-reduction potential of the sterno-cephalicus mus- cle of the horse after death in relation to the develop- ment of bacteria. Journal of Science of Food and Agri- culture 6: 448-455. Beck, T. 1978. The microbiology of silage fermentation. In; McCullough, M. E. (ed.). Fermentation of Silage - A Review National Feed Ingredients Association, lowa. p. 61-115. Harrison, D. E. F. 1971. Abstracts from Fift International Symposium and study Group on continuous culture of Microorganisms. Oxford. 39 p. Hassan, T. L. & Heath, J. L. 1986. Biological fermenta- tion of fish waste for potential use in animal and poultry feeds. Agricultural Wastes 15: 1-15. Hurst, A. 1973. Microbial antagonists in foods. Can. Inst. Food Sei. Technol. J. 6: 80-90. - & Collins-Thomsson, D. L. 1979. Food as a bacterial habit. In: Alexander, M. (ed ). Advances in Microbial Ecol- ogy. New York and London, Plenum Press, p. 79-134. Huss, H. H. & Larsen, A. 1979. The post mortem chang- 16 AGRICULTURAL SCIENCE IN FINLAND Vol. 4: 11-17. es in the oxidation-reduction potential of fish muscle and internal organs. In: Sobolenska-Ceronik, K. et al. (eds.). Food as an Ecological Environment for Pathogenic and Index Organism. Proceedings of the 10th International Symposium of lAMS (International Association of Micro- bial Societies), Ars, Poland, p. 265-279. Ingram, M., Ottaway, F. J. H. & Coppoch, J. B. M. 1959. The preservative action of acid substances in food. Chem- istry Index. Part 2. p. 1154-1162. Jay, J. M. 1986. Modern Food Microbiology. Third edi- tion, Van Nostrand, C.0., New York. 622 p. Lassén, T. M. 1995a. Lactic acid fermentation offish of- fal and chicken by-product with different starter cultures. Agricultural Science in Finland 4: 19-26. -1995b. Biological quality of biofermented fish offal and chicken by-product. Agricultural Science in Finland 4: 27- 33. Lindgren, S. E. & Clevström, G. 1978. Antibacterial Activity of Lactic Acid Bacteria. 1: Activity of fish silage, a cerealstarter and isolated organisms. Swedish Journal of agricultural Research 8: 61-66. - & Pleje, M. 1983. Silage fermentation of fish and fish waste products with lactic acid bacteria. Journal of the Science of Food and Agriculture 34: 1057-1067. Mackie, J. M„ Hardey, R. & Hobbs, G. (eds.) 1971. Fermented fish products. FAO Fisheries Reports, No. 100. 54 p. Nester, E. W., Roberts, E. C., Pearsall, N. N. & McCa- rthy, B. J. 1978. Microbiology. 2nd edition. Holt-Saun- ders International Editions, N Y. USA. 747 p. Nilsson, R. & Rydin, C. 1965. A new method of ensiling foodstuffs and feedstutfs ofvegetable and animal origin. Enzymologia 29: 126-142. Pelczar, M. J., Reid, R. D. & Chan, E. C. S. 1977. Micro- biology. Mcgrave-Hill Inc. 952 p. Raa, J. 1980. Biochemistry of microbial fish spoilage and preservation by lactic acid bacteria and added acid. In: Emejuavie, S. O. et al. (eds.). Global Impacts of Applied Microbiology, Sixth International Conference, Congress Proceedings Lagos, Nigeria, p. 3-17. SAS 1985 SAS/STAT® User's Guide: Statistics, Version 5 Edition. SAS Institute Inc. Cary, NC. 956 p. Skrede, A. & Nes, I. F. 1988. Slaughterhouse by-prod- ucts preserved by Lactobacillus plantarum fermentation as feed for mink and foxes. Animal Feed Science and Technology 20: 287-298. Speck, M .L. 1981 . Use of microbial cultures. Dairy Prod- ucts. Food Technology 35: 71-73. Stanton, W. R. & Yeoh, Q. L. 1977. Low salt fermenta- tion method for conserving trash fish waste under SE Asian conditions. Proceedingsof the Conference on Han- dling, Processing and Marketing of Tropical Fish. TPI, London UK. p. 268-277. Urlings, H. A. P. 1992. Fermentation of animal by-prod- ucts. Microbiological aspects of processing, epidemiolo- gy and animal nutrition. Diss. Utrecht University, Faculty of Veterinary Medicine, Department of the Science of Food of Animal Origin, Utrecht, The Netherlands. 135 p. Wirth, F. & Leistner, L. 1970. Redoxpotentiale in Fleis- chkonserven. Fleischwirtschaft 50: 491-492. Wolstrup, J. 1972. Mikrobiologiske undersogelser af vomindholdet fra koer fodret med sukkerroer, graesensi- lage, ho og grass. Thesis. Mikrobiologisk Laboratorium, Royal Veterinary and Agricultural University, Denmark, 47 p. SELOSTUS Kalajätteen fermentaatio-olosuhteiden arvionti T. Mikael Lassén Royal Veterinary and Agricultural University, Tanska Kalajätteen maitohappofermentaation olosuhteita tutkittiin erilaisilla kasvualustoilla (2,5 ja 10 % dekstroo- sia), maitohapolla ennalta hapottamisella (pH:n las- keminen 6,B:sta 6,s:een tai 6,o:aan) ja erikokoisilla Lactobacillus plantarum siirrostuksilla (10 7 , 108 ja 109 cfu/g). pH:ta ja maitohappopitoisuutta seurattiin kahden viikon varastointijaksolla. Pienoismalli kala- jätteen fermentaatiosiilosta rakennettiin ja selvitettiin redox-potentiaalin mittaamisen soveltuvuutta baktee- rien kasvuolosuhteiden arvioimiseen. Fermentaatiolle suotuisimmat olosuhteet (alhai- nen ja vakaa pH sekä korkea maitohappopitoisuus) saavutettiin 10* cfu/g siirrostuksella ja 5 % dekstroo- sipitoisuudella kasvualustassa. Ennalta hapottaminen maitohapolla ei vaikuttanut lopulliseen pH:een. Re- dox-potentiaalin osoitettiin antavan luotettavan arvi- on kasvuolosuhteista bakteereille, jotka elävät hapet- tomissa olosuhteissa. Tällöin redox-potentiaalisuus laski nopeasti alle -550 mV kalasäilörehussa, jonka maitohappopitoisuus oli korkea. 17 AGRICULTURAL SCIENCE IN FINLAND