Abstract: This study was designed to explore the effect of controlled freezing-point storage (CFPS) on the quality of
Sinonovacula constricta. The structure and dynamics of the microbial community in Sinonovacula constricta during CFPS
was investigated by Illumina high-throughput sequencing technology. The growth kinetics of specific spoilage bacteria
(Pseudomonas spp.) was modeled and validated at different temperatures. The results showed that the initial bacterial flora
in fresh Sinonovacula constricta was composed of Arcobacter spp. (16.2%), Bacteroidales (8.4%) and Psychrilyobacter
spp. (7.2%). At the end of the survival period, Pseudomonas spp. became the specific spoilage bacteria, which accounted for
26.2% of the total bacterial population, followed by Vibrionaceae, Psychromonas spp., Arcobacter spp. and Psychrilyobacter
spp., with proportions of 17.8%, 14.4%, 13.0% and 11.8%, respectively. The growth dynamics of Pseudomonas spp. at
low temperatures (−1.5?15.0 ℃) could be described by the modified Gompertz model. The determination coefficient (R2)
of the model exceeded 0.98 and both the bias factor and accuracy factor were within a statistically acceptable range. The
square-root model for temperature dependence of the specific growth rate and lag-phase duration showed good linearity with
determination coefficients (R2) of 0.963 and 0.970, respectively. The Gompertz model established could predict the growth
of Pseudomonas spp. in the range of −1.5?15.0 ℃.

Key words: Sinonovacula constricta, controlled freezing-point storage, high-throughput sequencing, microfloral analysis, dynamic model