Abstract
Amasi is one of the most popular traditional fermented milk product in the Southern African region produced through a natural fermentation process. In South Africa unpasteurized milk is sold in local areas and can be used to make amasi at a household level. The most critical parameters in the processing of amasi are fermentation time and temperature, these are responsible for giving a standardized and good quality product. The aim of this study was to optimize the fermentation condition (time and temperature) of amasi and to investigate further its physicochemical, bacteria community, nutritional and health properties. The optimal conditions of amasi were carried out through a tool known as response surface methodology with assistance from k-means clustering, a machine learning technique. The optimal conditions obtained based on the physicochemical properties studied were (pH, titratable acidity (TTA), total soluble solids (TSS) and consistency) were 25 ℃ for 120 hr which is represented as optimal 25 (OP25) and 32 ℃ for 140 hr as optimal 32 (OP32). These were further used to produce starter culture-fermented amasi. The microbiological evaluation of raw milk and optimized amasi (OP25 and OP32) was performed using Amplicon sequencing. The nutritional properties were carried out by proximate composition following the Association of Official Analytical Chemists (AOAC) methods, mineral (Inductively-coupled plasma optical emission spectrometry (ICP-OES)) and amino acid composition. While for health properties, phytochemical was measured using total phenolic content (TPC) and total flavonoid content (TFC) as well as antioxidant activities using (2,2-Azinobis (3-Ethyl-Benzothiazone-6-Sulfonic acid)) (ABTS) and (2,2-diphenyl-1-picrylhydrazyl) (DPPH) measurements. The fatty acids present in the samples were analysed using base-catalysed methanolysis of glycerides based on the International Organization for Standardization-International Dairy Federation (ISO-IDF) method. The results showed significant inverse relationship between the pH and TTA among the samples. The pH reduced from 6.64 in raw milk to 3.44 in OP25 while TTA increased from 0.02% in raw milk to 0.23% in OP25. A reduction of the TSS and consistency was also observed with the increasing fermentation time and temperature. The study also revealed that the bacteria predominantly associated with amasi fermentation was the lactic acid bacteria from the family Lactobacillaceae while the strains most abundant in raw milk belong to the Prevotellaceae family. The results further showed that fermentation conditions significantly improved the moisture, ash, protein, carbohydrates and energy content of both fermented ii samples but the fat content fluctuated between the investigated samples (RM, OP25, OP32, SC25, SC32). The fermentation, however, increased most of the essential and non-essential amino acids with glutamic acid being the most abundant in all of the samples ranging from RM-SC32 with amounts ranging from 4.46 - 6.30 g/100 g. A similar trend was also observed in the mineral composition, in this case, potassium was the most abundant in all samples having values ranging from 110.18 - 134.46 μg/g. The phenols and flavonoids (TPC and TFC) in all samples as well as the antioxidant capacity (ABTS) were significantly (p≤0.05) improved by optimized fermentation. The free fatty acids were significantly decreased during the fermentation process. Thus, this study demonstrated the importance of the optimization of fermentation conditions in producing a standardized and good quality amasi. The overall results depicted the amasi with enhanced nutritional and health promoting properties than the raw milk which can be of utmost health benefits to consumers. Keywords: Amasi, fermentation, optimization, physicochemical, nutritional properties, health properties.
M.Sc. (Food Technology)