Optimization and application of probiotics complex for grouper based on dynamic model

This study applied systematic microecological method to develop multi-strain probiotics for grouper aquaculture, focusing on rational strain selection and interaction optimization. Five strains from marine source were initially screened through functional and safety assessments. Based on single-strain growth kinetics and pairwise interaction data, the generalized Lotka-Volterra model and multiple linear regression were employed to identify an optimal consortium consisting of Bacillus subtilis, Bacillus megaterium, and Bacillus velezensis (1: 1: 1), which exhibited significantly antagonism against Vibrio harveyi (p < 0.05) and higher digestive enzyme activities among thirty predefined combinations. In vitro tests using intestinal communities from the purebred grouper Epinephelus coioides (Ec) and the hybrid grouper E. fuscoguttatus ♀ × E. lanceolatus ♂ (Elf) revealed host-specific microbiome characteristics. The consortium exerted stronger effects in Elf than single-strain treatments. Supplementation of probiotic complex enhanced network stability, enriched beneficial genera (e.g., Bacillus), and suppressed opportunistic pathogens (e.g., Vibrio). Finally, in a 42-day feeding trial in juvenile Elf grouper, both single bacteria (B. velezensis at 10⁷ CFU/g feed) and the multi-strain consortium (total 10⁷ CFU/g feed) improved growth performance and disease resistance compared to the control. The consortium further increased weight-specific growth rate by 23.41% (p < 0.05), feed intake by 25.92% (p < 0.05), and survival rate after pathogenic challenge by 40.0% (p < 0.05). Both treatments reshaped gut microbiome correlation networks, reducing the dominance of Vibrio and Photobacterium while increasing the centrality of beneficial Cetobacterium in the consortium group. This work validated the effectiveness of the gLV model for screening and designing probiotic consortia, providing a promising solution for precision aquaculture of grouper.