Abstract
An 8-week feeding trial was conducted to evaluate the application of common carbohydrate sources, cornstarch (CS), wheat starch (WS), and wheat flour (WF), to different gibel carp genotypes, Dongting, CASIII, and CASV. The results of the growth and physical responses were analysed by data visualization and unsupervised machine learning. As revealed by a self-organizing map (SOM) and the cluster of growth and biochemical indicators, CASV had superior growth and feed utilization and better regulation of postprandial glucose, followed by CASIII, while Dongting showed a high level of plasma glucose with poor growth performance. CS, WS, and WF were differently utilized by the gibel carp, and WF was associated with greater zootechnical performance based on higher specific growth rate (SGR), feed efficiency (FE), protein retention efficiency (PRE), and lipid retention efficiency (LRE), induced hepatic lipogenesis, increased liver lipids, and enhanced muscle glycogen. Spearman's correlation analysis of the physiological responses indicated that plasma glucose had a significantly negative correlation with growth, feed utilization, glycogen storage, and plasma cholesterol level, and it was positively related to liver fat content in gibel carp. Transcriptional variabilities were observed: CASIII showed increased expression of pklr, which is involved in hepatic glycolysis, and pck and g6p, which are involved in gluconeogenesis. Interestingly, Dongting showed upregulation of genes involved in glycolysis and fatty acid oxidation in muscle. Furthermore, there were numerous interactions between carbohydrate sources and strains for growth, metabolites, and transcriptional control, confirming the existence of genetic polymorphisms in carbohydrate use in gibel carp. Globally, CASV showed relatively better growth and carbohydrate utilization, and wheat flour seemed to be more efficiently utilized by gibel carp.