Abstract: In this paper, seven types of substrates were used as co-metabolic substrates, the functional bacterial flora DDMZ1 was applied as the initial bacterial flora, and an azo dye, reactive black 5, was chosen as the target pollutant. The effects of different cometabolic substrates (sucrose, glucose, fructose, glucose + fructose, beef extract, peptone, sodium carbonate) on the decolorization performance and community structure of bacterial flora DDMZ1 on decolorization of reactive black 5 were studied by the stationary culture method. Decolorization results showed that added fructose co-substrate promoted bacterial decolorization performance. Under 500 mg·L^-1 of reactive black 5, culturing for 72 h, the decolorization rate reached 88.96%. Highthroughput sequencing analysis results showed that the bacterial community structure significantly differed with different cometabolic substrates. On the classification of genus level, Acinetobacter spp. occupied the leading position in samples with peptone, beef extract and carbonate as co-substrates, whereas contents of Lactococcus spp. and Burkholderia spp. were higher in samples with sucrose, glucose, fructose, glucose + fructose as co-metabolic substrates. Among them, Lactococcus reached its maximum ratio in sample with fructose as co-metabolic substrate, which indicated that Lactococcus might function as an important dominant genus in decolorizing process of reactive black 5. To sum up, compared to the initial flora DDMZ1, organic nitrogen sources and inorganic carbon co-metabolic substrates had little effects on the community structure, while sugars as co-metabolic substrates significantly affected the community structure. Furthermore, fructose was able to enrich Lactococcus as functional decolorizing bacteria, and optimized the bacterial community structure, thus promoting bacterial decolorization effects.

Key words: additive system, seemingly unrelated regression., stand-level biomass, error structure