23 October 2014 - Tinna Austen Ng’ong’ola-Manani (IKBM)

Norwegian title:
Naturlig fermentering og fermentering med melkesyrebakterier av soyabønne- og soyabønne-mais blandinger: Effekt på ernæringsmessig kvalitet, mikrobiell diversitet, mattrygghet og forbrukeraksept

Prescribed subject of the trial lecture:
Microbial enhancement of the nutritional quality of fermented foods

Time and place for the trial lecture and the public defence:
Thursday 23.10.14, BTB Library at 12:15

Associate Professor Trude Wicklund, NMBU, main supervisor
Dr. Hilde Marit Østlie, NMBU
Dr. Agnes Mwangwela, Malawi University

Evaluation committee:
Professor Ulf Svanberg, Chalmers University of Technology, Sweden
Professor Lene Jespersen, University of Copenhagen, Denmark
Professor Judith Narvhus, IKBM 

The doctoral thesis is available for public review at the UMB library.
Thesis number 2014:84, ISSN 1894-6402, ISBN 978-82-575-1243-9


Traditional Malawian diets are predominantly maize-based and have been associated with widespread inadequate intakes of several nutrients. In addition to maize, legumes are an important source of protein and other nutrients in diets of many people in developing countries. Soybeans have the highest protein content among legumes and when consumed together with cereals, a high quality protein is provided because cereals and legumes arecomplementary in terms of limiting amino acids. However, soybean utilization in Malawi is minimal due to limited knowledge in processing.

In an effort to increase utilization and consumption of soybeans by all age groups in Malawi, solid-state fermented pastes of soybeans and soybean-maize blends were developed. The fermented pastes were to be used as relish and to serve as major sources of protein in maizebased diets. Spontaneous solid state fermentation of soybeans favors growth of Bacillus subtilis, a highly proteolytic organism that produces high amount of ammonia. High ammonia levels result in strong odor which some people find objectionable. On the other hand, lactic acid bacteria (LAB) are weakly proteolytic and do not lead to accumulation of high levels of organoleptically unpleasant metabolic products.

In this study, thobwa, a Malawian fermented cereal gruel prepared from maize flour and co fermented with malt flour from finger millet was used as a back – slopping material to facilitate lactic acid bacteria fermentations in LAB fermented pastes (LFP). Whereas pastes fermented without inoculum were referred to as naturally fermented pastes (NFP). Pastes composed of 100% soybeans, 90% soybeans and 10% maize, and 75% soybeans and 25% maize. Naturally fermented pastes were designated 100S, 90S and 75S, while LFP were designated 100SBS, 90SBS and 75SBS. Metabolite changes, microbial diversity, growth and survival of enteropathogens, sensory properties and consumer acceptance of pastes of soybeans and soybean-maize blends fermented naturally and by LAB were compared.

Both types of fermentation resulted in increases in soluble protein which were pronounced at 48 hrs in most samples and were highest in 100S (49%). High decreases in total amino acids were also observed at 48 hrs, with 6.8% and 7.4% reductions in 100S and in 100SBS, respectively. On a positive note, the imiting amino acids, cysteine (in 100S and 90S) and methionine (in 90S) increased throughout fermentation. Whereas in LFP, cysteine increased during 48 hrs of fermentation and this trend was also observed with methionine in 75SBS.


Both types of fermentation degraded anti-nutritional factors, phytic acid and trypsin inhibitors. However, natural fermentation was more effective in degrading phytic acid than LAB fermentation. In NFP, 33 to 54% reduction in phytic acid was achieved during 24 hrs fermentation and by 72 hrs, 85% reduction was noted and the phytate was not detected in some samples. Whereas, 18 to 32% reduction was achieved in LFP after 24 hrs, and by 72 hrs, 37 to 49% reduction had been achieved.

Lactic acid was the major end product of fermentation in both LFP and NFP. High lactic acid production in LFP was consistent with pH reduction. The pH was reduced from 6.44 - 6.48 to 4.20 – 4.64 representing a 28 – 35% reduction after 24 hrs fermentation. After 72 hrs, the pH was reduced to 3.91 – 4.26, representing 34 – 39% reduction. In NFP, pH was reduced from 6.88 – 6.95 to 6.15 – 6.74 during 24 hrs and to 5.36 – 5.81 during 72 hrs representing 3 – 12% and 17 – 23% reductions, respectively. Higher pH reduction in LFP could have been due to a higher LAB population which was 3 log10 cfu/g higher than in NFP at the beginning of the fermentations.

The fermenting LAB microflora in both NFP and LFP were heterofermentative rods and homofermentative cocci. The microbiota were phenotypically characterized as Lactobacillus brevis, Lactobacillus fermentum, Lactobacillus buchneri, Lactobacillus collonoides, Lactobacillus acidophilus, Lactobacillus plantarum, Lactobacillus delbrueckii subsp. delbrueckii, Lactobacillus pentosus, Leuconostoc mesenteroides, Weissella confusa, Lactococcus lactis subsp. lactis, Pediococcus pentosaceus and Pediococcus damnosus. The dominant microflora were Lb. fermentum, Lb. brevis, W. confusa and P. pentosaseus. These four species were confirmed as the dominant fermenting microflora by 16S rDNA genotyping.

In addition, Bacillus spp. and Enterococcus faecium/ Enterococcus durans were identified as part of the microflora. Denaturing gradient gel electrophoresis confirmed Lb. fermentum, W. confusa/Weissella cibaria, P. pentosaseus as dominant microflora. DGGE revealed microbial succession in NFP in which Bacillus spp. and Lactobacillus linderi were succeeded during later fermentation.

Microbial diversity was similar throughout fermentation in LFP. The following microorganisms were present in both NFP and LFP at the end of the fermentations: P. pentosaceus, Lb. fermentum, W. confusa/W. cibaria, and Weissella koreensis. 

In paper IV, natural fermentation, LAB fermentation through back-slopping and starter culture fermentation using Lb. fermentum were inoculated with Escherichia coli. All ix 5.0, the critical value for B. cereus growth. After 72 hrs, B. cereus cell counts ranged between 0 to 3 log10 cfu/g in back-slopped pastes. In Lb. fermentum fermentation, pH values ranged between 5.30 and 5.35 while cell counts were 3.7 to 5.3 log10 cfu/g after 72 hrs of fermentation. In natural fermentation, pH increased from 5.87 at 24 hrs to 7.2 during 72 hrs of fermentation in 90S. Consequently, B. cereus population increased from 2.2 log10 cfu/g to above 8.0 log10 cfu/g during 24 hrs of fermentation. Since the infectious dose for B. cereus is  ≥ 3.0 log10 cfu/g, it was concluded that back-slopping has a potential of producing pastes that are safe with regards to B. cereus poisoning. Nevertheless, a thermal treatment of the pastes prior to consumption was recommended to ensure safety.

Consumers unconsciously used type of fermentation to determine their preference patterns and preference was biased towards natural fermentation. Naturally fermented pastes were characterized by yellow color, higher pH, fried egg-like appearance and aroma, sweetness, softness, rancid odor, and raw soybean odor. These attributes were also considered as drivers of liking. Lactic acid bacteria fermented pastes were characterized by brown color, sourness, bitterness, saltiness, umami, burnt roasted soybeans and maize aromas. Optimization by enhancing the drivers of liking while suppressing drivers of dislike would increase utilization of soybean fermented pastes. fermentations could not reduce the pH to ≤ 4.4, the critical value for Escherichia 222 coli growth. Nevertheless, back-slopping inhibited E. coli growth more than the other fermentations. In back-slopped pastes, E. coli counts increased from 2.4 to 3.5 log10 cfu/g during 24 hrs and remained constant during further fermentation. While E. coli population increased from 2.0 – 2.3 log10 cfu/g to 6.8 – 7.6 log10 cfu/g in Lb. fermentum fermentation and from 2.3 log10 cfu/g to 8.8 – 9.2 log10 cfu/g in NFP during 24 hrs fermentation. The cell counts were above the infectious dose of 100 cells implying food safety concerns for some Shiga-toxin producing E. coli in the event of contamination during fermentation.

Published 31. oktober 2016 - 17:46 - Updated 23. mai 2017 - 19:26