Fermentation is a process by which primary food products are modified biochemically by the action of microorganisms and/or their enzymes. Several societies have, over the years, intentionally carried it out to enhance the taste, aroma, shelf-life, texture, nutritional value and other properties of food. It is used in many parts (lithe world. However, there are regional differences in use and these depend on the availability of raw materials, consumption habits. and other socio-cultural factors. This study was aimed at (comparatively) assessing, local commercial and homemade amahewu with respect to nutritional value, hygiene and other health benefits to the commirn ity. Methods employed were Thin Layer Chromatography (TLC) (mycotoxins), High Perliffmance Liquid Chromatography (HPLC) (mycotoxins, sugars and amino acids), Dumas (proteins), SOxhlet (lipids) and intubation technique (metabolisable energy) to analyse maize meal and amahewu samples from various regions. The regions sampled included mal3heleni (South Coast) and kwaNgcolosi (North Coast) villages. Commercial amahewu was analysed with kind permission from Clover SA. Species from the following genera were isolated and identified from amahewu samples: Lactobacillus, Saccharonivccs, Lcuconostoc, Lactococcus, Panioca, Entcrobacter and kleb•iella. Saccharotnyces was detected in commercial samples only. Gram-negative strains were identified in most of manheleni village samples. No traceable amounts of aflatoxin BI (AFB1), fumonisin B 1 (FBI) and zearalenone (ZEA) were found in Clover SA samples. AFB I was detected in 40% of both maize meal and amahewu samples from maBheleni (range 0.55 — 0.84ng/g and 8.3x10 5 — 9.1x10-5ng/g respectively). From the same village, 100% of the maize meal and 80% of the amahewu samples were contaminated with FBI (range 4.1 47.2ng/g and 1.4 ---- 6.9ng/g respectively). ZEA was detected in all maize meal samples (range 0.9 — 4.3ng/g). None of the amahewu samples contained detectable levels of ZEA. All maize meal and amahewu samples from kwaNgcolosi were contaminated with AF13 1 (range 8.3 — 30.I ng/g and 0.04 - 0.102ng/g respectively). FB I was detected in 75% of both maize meal and amahewu samples from the same village (range 0.5 — 4.1ng/g and 0.04 0.56ng/g respectively). ZEA was also found in all maize meal samples and 75% of amahewu samples (range 3.7 — 16.4ng/g and 0.03 -- 0.06ng/g respectively). MaBheleni, Clover SA and kwaNgcolosi maize meal and amahewu samples contained vitamins B1, 13 2 and B6 with a range of 0.31+0.21 - 4.48±0.81 B 1 ; 0.15±0.14 - 1.67±0.33 B2 and 0.05±0.07 - 0.77±1.45 lig/g B6. Fat levels ranged from 0.28±0.40 to 4.54±0.05 percentage by weight. The levels of proteins varied from 4.02±0.02 to 8.40±0.04 percentage by weight. Starch concentrations ranged from 31.51.5.28 to 75.911.92g/100g. Maize meal samples contained glucose and maltose, while glucose, fructose, sucrose, maltose, M-triose, DP 4 and 5 and DP >15 were detected in amahewu. Apparent and true metabolisable energy for homemade and commercial Freeze-dried amahewu was 13.194 and 13.696MJ/kg (AME N ); and 13.605 and 14.106M.Ekv ( 1 MEN ), respectively. This study has shown that lactic acid maize fermentation reduce' the levels of AF13 1 , FB I and ZEA toxins in maize meal, inhibits the growth of most Gram-negative bacteria, and in some instances, fermentation did improve the nutritional value. Metabolisable energy analysis represents an important tool to assess whether or not compounds ingested are converted to sources of energy in the body and utilised. Amahewu fermentation yielded beneficial products (probiotics: reduced mycotoxins levels and reduced starch). In conclusion, natural lactic acid maize fermentation to produce amahewu will do more good than harm to the consumer, therefore, people need to be advised on how to safely store their maize and also to be encouraged to consume their stored maize in fermented form. / Thesis (M.Med.Sc.)-University of Natal, Durban, 2003.

submitted in fulfilment of the academic requirements for the degree of Master of Applied Science in Food Science and Technology, Durban University of Technology, 2015. / Vitamin A deficiency (VAD) is a major health problem in sub-Saharan Africa where maize is a staple food. Amahewu, a fermented non-alcoholic,maize-based beverage is a popular drink in southern Africa.The aim of this study is to produce a provitamin A enriched and acceptable amahewu, using provitamin A biofortified maize which can be used to alleviate VAD. The optimal processing parameters for the production of amahewu using provitamin A-biofortified maize were determined. Amahewu samples were prepared with reference to a traditional method by boiling a mixture of maize meal and water (rato:1:7) at 90ᴼC, with occasional stirring, for 15 minutes. The resulting porridge was left to cool to approximately 40ᴼC, before inoculation and fermentation at 37oC. Processing parameters investigated were inoculum types (wheat bran (WB), maize malt (MM) and Lactobacillus mixed starter culture) and inoculum concentration (0.5,1 and 2% (w/w)) and varieties of provitamin A maize (PVAH 62 and PVAH 19). Wheat flour (at 2%) was used as reference inoculum to conform to the traditional practice. White maize amahewu samples processed in the same way as those of provitamin A-biofortified maize were used as references. Provitamin A amahewu samples were produced using the optimized processing parameters and then analysed for nutrient composition, including carotenoids, protein, ash, amino acids, mineral profile and invitro protein digestibility. The consumer acceptability of amahewu samples was evaluated using regular consumers of amahewu (n= 54), who rated the acceptability of the samples on a 9-point hedonic scale (1:disliked extremely, 9:liked extremely). The storage stability of the provitamin A biofortified amahewu samples was assessed by subjecting the samples to different storage conditions: 4ᴼC, 25ᴼC and 37ᴼC. The microbiological quality of the stored samples was monitored by taking samples every day for a period of five days to analyse for the presence of aerobic and anaerobic bacterial spore formers, E.coli and moulds. The provitamin A maize variety did not influence pH and Total titratable acidity (TTA) of amahewu samples during fermentation. As expected, there was a substantial drop in pH with fermentation time. After 24 hours, all the samples of amahewu, including those made with white maize, prepared using malted maize and wheat bran inocula reached a pH of 3.3-3.8 and TTA of 0.3-0.6, which were within acceptable range for amahewu. The addition of a starter culture substantially reduced fermentation time, from 24 to six hours. The inoculum of WB and MM, respectively, at a concentration of 0.5%, with or without starter culture (5%), were found to be suitable for the production of amahewu using provitamin A biofortified maize. The total provitamin A content of amahewu samples, produced using optimised parameters (i.e one variety of provitamin A biofortified maize, 0.5% MM, WB with or without starter culture), ranged from 3.3-3.8 μg/g (DW). The percentage retention of total provitamin A ranged from 79%- 90% (DW). The lowest percentage retention was observed in products fermented with the addition of starter culture. The gross energy of the amahewu samples was approx. 20 MJ/kg. There was a slight increase in the lysine content of amahewu after fermentation. The protein digestibility (approx. 91%) of amahewu samples was slightly higher than that of raw provitamin A maize (86%). Amahewu processed using starter cultures had a slightly higher iron content than those processed without a starter culture. Consumer acceptability data showed that amahewu samples made with provitamin A biofortified maize were slightly more acceptable (average rating for overall acceptability was 7.0 ± 1.2), compared to those made with white maize (average rating for overall acceptability was 6.4 ± 0.8). Principal component analysis (PCA) of Amahewu sensory data showed that 71% of variation was due to maize types and 18% of variation may be due to the inoculum used during fermentation. The use of a starter culture improves the taste and aroma acceptability of amahewu. Segmentation of consumers based on overall linking for amahewu revealed three clusters, named A, B and C. Cluster A consisted of most consumers (43%), who liked amahewu moderately. About 60% of these consumers were females. Cluster B consisted of most of the consumers (31%) who were undecided about their liking for the product. Approximately 52% of the consumers in this cluster were female. Cluster C consisted of consumers (26%) who liked amahewu very much. Sixty-four percent (64%) of these consumers were female. It appeared that gender may have some influence on overall liking for amahewu, as cluster B, consisting of undecided consumers, had more male consumers compared to clusters A and C. Age did not seem to be significantly associated with the liking of amahewu. Provitamin A biofortified amahewu samples stored under refrigerated conditions (4ᴼC) had better microbiological quality compared to those stored at 25ᴼC and 37ᴼC. Refrigeration effectively maintains the microbiological quality of amahewu for about three of days. Provitamin A biofortified maize can be used to produce β-carotene enriched amahewu that is acceptable to consumers following the processing method that is traditionally employed for white amahewu at both domestic and commercial level. Provitamin A biofortified amahewu has the potential to make a significant contribution towards alleviating VAD among rural communities, who are the most vulnerable to VAD.

A importância da Listeria monocytogenes como patógeno veiculado por alimentos, causador de grave quadro de infecção é bem conhecida. Devido a sua grande capacidade de sobrevivência às condições adversas do ambiente, o presente trabalho teve como objetivos verificar a freqüência de L. monocytogenes em mortadelas comercializadas no município de São Paulo, SP; avaliar o comportamento de 2 níveis de inóculos de L. monocytogenes frente ao processamento térmico de mortadelas de 2 formulações em diferentes condições de estocagem e avaliar o comportamento de L. monocytogenes naturalmente presente em mortadelas fatiadas, embaladas a vácuo, comercializadas e estocadas sob refrigeração, durante sua vida de prateleira. As amostras foram analisadas pelos métodos de Presença/Ausência ou NMP/g. Verificou-se que 26,7% das amostras de mortadelas analisadas foram positivas para L. monocytogenes. O processamento térmico convencional aplicado na cocção das mortadelas (74ºC no ponto mais frio) foi suficiente para redução dos inóculos em até 3 ciclos log independente da formulação do produto. A temperatura de estocagem, refrigeração ou temperatura ambiente, não interferiu na recuperação do microrganismo. Nas mortadelas fatiadas e embaladas a vácuo foi observado aumento médio de 2,5 ciclos log do NMP de L. monocytogeneslg durante a vida de prateleira do produto, com níveis de L. monocytogenes próximos a 2,0 log NMP/g no final da vida de prateleira do produto. De acordo com os resultados obtidos concluiu-se que L. monocytogenes é um patógeno freqüente em mortadelas; o tratamento térmico de mortadelas é suficiente para redução de até 3 ciclos logarítmicos de L. monocytogenes, não tendo sido observada presença do microrganismo durante estocagem por 30 dias e que mortadelas fatiadas e embaladas a vácuo mantidas sob refrigeração são susceptíveis a multiplicação de L. monocytogenes durante sua estocagem refrigerada, representando um risco à população susceptível. / L. monocytogenes is an important foodborne pathogen that can cause life threating infections. The importance of L. monocytogenes as a foodborne pathogen is well known. Its capacity to resist to adverse environmental conditions makes this microorganism a cause of concern for the food industry. The objectives of this study were to evaluate: the incidence of L. monocytogenes in mortadella samples commercialised in São Paulo, SP, BR; the behaviour of 2 leveis of L. monocytogenes during the processing of 2 formulations of the product stored at difterent conditions and the behaviour of indigenous L. monocytogenes present in vaccum-packed sliced mortadella during the shelf life of the product. Samples were analised by the Presence/Absence technique or By MPN. 26.7% of the mortadella samples were positive for L. monocytogenes. The conventional cooking process (74°C in the coldest point) was sufficient to reduce the population of the microorganism in up to 3 log and the formulation of the product did not interfere with this reduction. L. monocytogenes was not detected during storage of the product for up to 30 days under cold storage (5-8°C) or room temperature (25°C). The vaccum-packed sliced product showed a mean increase of 2.5 log of MPN/g during its shelf life with the bacterium reaching populations around 2 log MPN/g by the end of the período It could be concluded that L. monocytogenes is very frequent in mortadellas; for this product the cooking process is sufticient to reduce up to 3 log of L. monocytogenes. This microorganism could not be recovered during storage for up to 30 days, indicating no sub-Iethal damage. The vaccum-packed sliced product stored under refrigeration is susceptible to L. monocytogenes growth and can be considered risky to susceptible population.

A importância da Listeria monocytogenes como patógeno veiculado por alimentos, causador de grave quadro de infecção é bem conhecida. Devido a sua grande capacidade de sobrevivência às condições adversas do ambiente, o presente trabalho teve como objetivos verificar a freqüência de L. monocytogenes em mortadelas comercializadas no município de São Paulo, SP; avaliar o comportamento de 2 níveis de inóculos de L. monocytogenes frente ao processamento térmico de mortadelas de 2 formulações em diferentes condições de estocagem e avaliar o comportamento de L. monocytogenes naturalmente presente em mortadelas fatiadas, embaladas a vácuo, comercializadas e estocadas sob refrigeração, durante sua vida de prateleira. As amostras foram analisadas pelos métodos de Presença/Ausência ou NMP/g. Verificou-se que 26,7% das amostras de mortadelas analisadas foram positivas para L. monocytogenes. O processamento térmico convencional aplicado na cocção das mortadelas (74ºC no ponto mais frio) foi suficiente para redução dos inóculos em até 3 ciclos log independente da formulação do produto. A temperatura de estocagem, refrigeração ou temperatura ambiente, não interferiu na recuperação do microrganismo. Nas mortadelas fatiadas e embaladas a vácuo foi observado aumento médio de 2,5 ciclos log do NMP de L. monocytogeneslg durante a vida de prateleira do produto, com níveis de L. monocytogenes próximos a 2,0 log NMP/g no final da vida de prateleira do produto. De acordo com os resultados obtidos concluiu-se que L. monocytogenes é um patógeno freqüente em mortadelas; o tratamento térmico de mortadelas é suficiente para redução de até 3 ciclos logarítmicos de L. monocytogenes, não tendo sido observada presença do microrganismo durante estocagem por 30 dias e que mortadelas fatiadas e embaladas a vácuo mantidas sob refrigeração são susceptíveis a multiplicação de L. monocytogenes durante sua estocagem refrigerada, representando um risco à população susceptível. / L. monocytogenes is an important foodborne pathogen that can cause life threating infections. The importance of L. monocytogenes as a foodborne pathogen is well known. Its capacity to resist to adverse environmental conditions makes this microorganism a cause of concern for the food industry. The objectives of this study were to evaluate: the incidence of L. monocytogenes in mortadella samples commercialised in São Paulo, SP, BR; the behaviour of 2 leveis of L. monocytogenes during the processing of 2 formulations of the product stored at difterent conditions and the behaviour of indigenous L. monocytogenes present in vaccum-packed sliced mortadella during the shelf life of the product. Samples were analised by the Presence/Absence technique or By MPN. 26.7% of the mortadella samples were positive for L. monocytogenes. The conventional cooking process (74°C in the coldest point) was sufficient to reduce the population of the microorganism in up to 3 log and the formulation of the product did not interfere with this reduction. L. monocytogenes was not detected during storage of the product for up to 30 days under cold storage (5-8°C) or room temperature (25°C). The vaccum-packed sliced product showed a mean increase of 2.5 log of MPN/g during its shelf life with the bacterium reaching populations around 2 log MPN/g by the end of the período It could be concluded that L. monocytogenes is very frequent in mortadellas; for this product the cooking process is sufticient to reduce up to 3 log of L. monocytogenes. This microorganism could not be recovered during storage for up to 30 days, indicating no sub-Iethal damage. The vaccum-packed sliced product stored under refrigeration is susceptible to L. monocytogenes growth and can be considered risky to susceptible population.

Text in English / Fungal species and their mycotoxins are the most predorminant contaminants of dried agricultural products in sub-Saharan Africa (SSA) and the main species of fungi that can synthesize mycotoxins are Aspergillus, Fusarium and Penicillium. In Africa, aflatoxin is labelled as a great threat to human and animal health due to its high contamination levels reported of aflatoxins in foods. The aim of this study was to survey fungi and aflatoxin contamination of small-scale processed foodstuffs sold at informal retail outlets in the Johannesburg metropolis, South Africa. A total of 270 food samples (10 starch and legume based foods, 11 meat and fish based foods, 22 spices and local condiments, 14 dried fruits and vegetables) were collected from retailers; and analysed four (4) times in different seasons of spring, summer, autumn and winter. Out of the 270 samples analysed, only 27.8% were contaminated with fungal. Of all the six categories of foods analysed, roots and tubers (60.0%), nuts and seeds (40.0%), dried vegetables (37.1%), and the Meat and Insect foods (33.3%) respectively, had the most contaminated samples with fungal respectively. The least contaminated food groups were the fish foods (10.0%) and spices and local condiments (16.7%) respectively. Twenty percent of the 270 dried food analysed were contaminated by Aspergillus species out of which 61.1% of the contaminated samples had fungal counts above 103 cfu/g. Aspergillus niger was the most predominant Aspergillus species identified in all the categories of food samples analysed. Fruits and vegetables (24.4%) and the nuts and seeds (20.0%) food groups had the highest number of samples contaminated with aflatoxin. Peanut flour and Cardamom had the most incidence of aflatoxin. AFB1, AFB2 & AFG1 were the most prominent aflatoxin types recovered from the food samples. Almost all the food samples in which aflatoxin were identified had aflatoxin values above 10μg/ml. / Life and Consumer Sciences / M.Sc. (Life Science)

Bifidobacterium spp. são microrganismos probióticos que podem ser incorporados em produtos alimentícios. Entretanto, para que seus efeitos benéficos à saúde humana ocorram, é necessário que o número de células viáveis na hora do consumo seja, no mínimo, 106UFC/g. As bifidobactérias são sensíveis à elevada acidez e, por isso, torna-se necessária a busca por métodos que possam proteger a integridade da célula, sendo um deles a microencapsulação. Em uma primeira etapa do trabalho, Bifidobacterium lactis foi encapsulado em micropartículas de alginato e alginato modificado (alginatoquitosana, alginato-quitosana-sureteric e alginato-quitosana-acryl-eze) e sua sobrevivência e liberação das micropartículas em fluidos simulados do trato gastrintestinal foram mensuradas utilizando-se soluções tampão com pH 1,5, 5,6 e 7,5, na presença e na ausência de pepsina (3g/L), pancreatina (1g/L) e bile (10g/L). A liberação de células das micropartículas teve uma relação direta com o pH do tampão. A microencapsulação aumentou a taxa de sobrevivência de B. lactis, em comparação com células não encapsuladas, em soluções tampão com pH 1,5 sem a presença de enzimas. Em suco gástrico simulado com enzimas digestivas, por outro lado, foi observado que a pepsina proporcionou um efeito protetor sobre as células de B. lactis, e nesse caso, as taxas de sobrevivência do microrganismo estavam diretamente relacionadas com o grau de injúria das células. Em uma segunda etapa do trabalho, leites fermentados com Streptococcus salivarius ssp. thermophilus e Lactobacillus delbrueckii ssp. bulgaricus foram enriquecidos com culturas de Bifidobacterium lactis submetidas a quatro tratamentos diferentes: desidratação em temperatura ambiente, liofilização/congelamento, encapsulação em alginatoquitosana e encapsulação em alginato-quitosana-acryl-eze. A população sobrevivente de B. lactis foi determinada semanalmente no leite fermentado e também após tratamento simulando condições do trato gastrintestinal. Os resultados indicaram que na ausência de pepsina, as populações de B. lactis foram reduzidas drasticamente após o contato com tampão pH 1,5, não sendo possível a detecção de células viáveis livres ou encapsuladas após 120 minutos de teste. A presença de pepsina influenciou positivamente a recuperação de células viáveis de B. lactis em todas as condições testadas, mas as culturas na forma desidratada apresentaram melhores resultados que as culturas microencapsuladas ou liofilizadas. No caso do leite fermentado contendo as células desidratadas, a população de B. lactis, após o tratamento em suco gástrico com enzimas, foi superior à detectada no produto antes desse tratamento. Conclui-se que a microencapsulação não foi eficiente para proteger B. lactis em leite fermentado contra injúrias causadas pelo trato gastrintestinal simulado. / Bifidobacterium spp. are microorganisms that can be added to foods. However, the benefits for the human health occur when the numbers of viable cells in the moment of the consumption is at least 106CFU/g. Bifidobacteria are acid sensitive, and methods to protect cell integrity, such as microencapsulation, are needed. In the first part of the present study, Bifidobacterium lactis was encapsulated in microparticles of alginate and modified alginate (alginate-chitosan, alginate-chitosan-sureteric and alginate-chitosan-acryl-eze) and the survival and release from microparticles in simulated gastrointestinal conditions were measured, using buffers (pH 1.5, 5.6 and 7.5), in the absence and presence of pepsin (3g/L), pancreatin (1g/L) and bile. The release from microparticles presented a direct relationship with pH. When the pH was 1.5 and no enzyme was present, encapsulation improved the survival of B. lactis, when compared to free cells. However, pepsin had a protective effect on B. lactis, and the survival rate was directly related to the cells injury degree. In the second part of the study, fermented milk samples containing Streptococcus salivarius ssp. thermophilus and Lactobacillus delbrueckii ssp. Bulgaricus were supplemented with B. lactis submitted to four different treatments: dehydration at room temperature, freeze drying, encapsulation in alginate-chitosan and encapsulation in alginate-chitosaacryl-eze. The number of viable B. lactis cells in the fermented milk was determined weekly and also after treatment with simulated gastrointestinal conditions. Results indicated that in the absence of pepsin, the number of viable cells decreased significantly after contact with buffers (pH 1.5), and no viable cell was detected after 120 minutes. Pepsin improved the recovery of viable cells in the assayed gastric conditions, being the dehydrated cultures more resistant than other cultures. In fermented milk containing the dehydrated cells, the number of viable cells increased after treatment with simulated gastrointestinal fluids. Microencapsulation was not an effective procedure to protect B. lactis in fermented milk against injury caused by the simulated gastrointestinal tract.

Bifidobacterium spp. são microrganismos probióticos que podem ser incorporados em produtos alimentícios. Entretanto, para que seus efeitos benéficos à saúde humana ocorram, é necessário que o número de células viáveis na hora do consumo seja, no mínimo, 106UFC/g. As bifidobactérias são sensíveis à elevada acidez e, por isso, torna-se necessária a busca por métodos que possam proteger a integridade da célula, sendo um deles a microencapsulação. Em uma primeira etapa do trabalho, Bifidobacterium lactis foi encapsulado em micropartículas de alginato e alginato modificado (alginatoquitosana, alginato-quitosana-sureteric e alginato-quitosana-acryl-eze) e sua sobrevivência e liberação das micropartículas em fluidos simulados do trato gastrintestinal foram mensuradas utilizando-se soluções tampão com pH 1,5, 5,6 e 7,5, na presença e na ausência de pepsina (3g/L), pancreatina (1g/L) e bile (10g/L). A liberação de células das micropartículas teve uma relação direta com o pH do tampão. A microencapsulação aumentou a taxa de sobrevivência de B. lactis, em comparação com células não encapsuladas, em soluções tampão com pH 1,5 sem a presença de enzimas. Em suco gástrico simulado com enzimas digestivas, por outro lado, foi observado que a pepsina proporcionou um efeito protetor sobre as células de B. lactis, e nesse caso, as taxas de sobrevivência do microrganismo estavam diretamente relacionadas com o grau de injúria das células. Em uma segunda etapa do trabalho, leites fermentados com Streptococcus salivarius ssp. thermophilus e Lactobacillus delbrueckii ssp. bulgaricus foram enriquecidos com culturas de Bifidobacterium lactis submetidas a quatro tratamentos diferentes: desidratação em temperatura ambiente, liofilização/congelamento, encapsulação em alginatoquitosana e encapsulação em alginato-quitosana-acryl-eze. A população sobrevivente de B. lactis foi determinada semanalmente no leite fermentado e também após tratamento simulando condições do trato gastrintestinal. Os resultados indicaram que na ausência de pepsina, as populações de B. lactis foram reduzidas drasticamente após o contato com tampão pH 1,5, não sendo possível a detecção de células viáveis livres ou encapsuladas após 120 minutos de teste. A presença de pepsina influenciou positivamente a recuperação de células viáveis de B. lactis em todas as condições testadas, mas as culturas na forma desidratada apresentaram melhores resultados que as culturas microencapsuladas ou liofilizadas. No caso do leite fermentado contendo as células desidratadas, a população de B. lactis, após o tratamento em suco gástrico com enzimas, foi superior à detectada no produto antes desse tratamento. Conclui-se que a microencapsulação não foi eficiente para proteger B. lactis em leite fermentado contra injúrias causadas pelo trato gastrintestinal simulado. / Bifidobacterium spp. are microorganisms that can be added to foods. However, the benefits for the human health occur when the numbers of viable cells in the moment of the consumption is at least 106CFU/g. Bifidobacteria are acid sensitive, and methods to protect cell integrity, such as microencapsulation, are needed. In the first part of the present study, Bifidobacterium lactis was encapsulated in microparticles of alginate and modified alginate (alginate-chitosan, alginate-chitosan-sureteric and alginate-chitosan-acryl-eze) and the survival and release from microparticles in simulated gastrointestinal conditions were measured, using buffers (pH 1.5, 5.6 and 7.5), in the absence and presence of pepsin (3g/L), pancreatin (1g/L) and bile. The release from microparticles presented a direct relationship with pH. When the pH was 1.5 and no enzyme was present, encapsulation improved the survival of B. lactis, when compared to free cells. However, pepsin had a protective effect on B. lactis, and the survival rate was directly related to the cells injury degree. In the second part of the study, fermented milk samples containing Streptococcus salivarius ssp. thermophilus and Lactobacillus delbrueckii ssp. Bulgaricus were supplemented with B. lactis submitted to four different treatments: dehydration at room temperature, freeze drying, encapsulation in alginate-chitosan and encapsulation in alginate-chitosaacryl-eze. The number of viable B. lactis cells in the fermented milk was determined weekly and also after treatment with simulated gastrointestinal conditions. Results indicated that in the absence of pepsin, the number of viable cells decreased significantly after contact with buffers (pH 1.5), and no viable cell was detected after 120 minutes. Pepsin improved the recovery of viable cells in the assayed gastric conditions, being the dehydrated cultures more resistant than other cultures. In fermented milk containing the dehydrated cells, the number of viable cells increased after treatment with simulated gastrointestinal fluids. Microencapsulation was not an effective procedure to protect B. lactis in fermented milk against injury caused by the simulated gastrointestinal tract.