Consumption of fermented camel milk, named shubat, is very popular in Central Asia and especially in Kazakhstan where it is known for its medicinal and dietary properties. To identify lactic acid bacteria (LAB) camel milk and shubat were sampled from 4 regions of Kazakhstan with important camel's population. In total, 26 dairy samples from 13 selected farms representing the variability of the farming system in the country were collected. Isolated strains were identified by genotypic approach including PCR using three different pairs of primers (338f/518r; W001/23S1; Lac1/Lac2/Lac3) and 16S rDNA gene sequencing. Three genus were in majority: Lactococcus, Lactobacillus and Enterococcus. The following microorganisms were identified: Enterococcus durans ; Enterococcus faecalis; Enterococcus faecium; Lactobacillus casei; Lactobacillus casei subsp. casei; Lactobacillus curvatus; Lactobacillus kefiri; Lactobacillus paracasei; Lactobacillus sakei; Lactococcus lactis subsp. lactis; Leuconostoc mesenteroides. Identification of camel milk and shubat microflora provides a theoretical foundation for developing starter cultures by using local LAB strains for industrial production of traditional fermented milk products.

The organoleptic properties of traditional dairy beverages done with non-conventional dairy species (horse, camel), popular in Central Asia, were rarely described in the literature. To characterize the Volatile Organic Compounds (VOC) profile of fermented mare milk, 12 samples of cow milk, used as matrix were inoculated by different strains from two types of bacteria (bacilli and cocci) isolated from natural fermented mare milk. The analysis performed by Gas Chromatography coupled with Solid-Phase Micro-Extraction allowed identification of 160 different compounds from the 12 strains, and 90 from natural fermented mare milk. After cluster analysis, 3 types of profiles were observed. Those profiles were distinct by the amount of acid compounds (low, medium, high), negatively related to aromatic and aliphatic hydrocarbons. The analysis of the mean volatile compound profile of each type of bacteria (bacilli and cocci) by factorial discriminant analysis showed that 3 molecules (oxime-methoxy-phenyl, propanedioic acid propyl and 2-propanamine) allowed to well class 100% of the samples. Further researches on bacterial identification and experiments with different fermentation matrices from other dairy species will be conducted.

No abstract