Using machine learning algorithms to analyse the scute structure and sex identification of sterlet Acipenser ruthenus (Acipenseridae).

Sturgeons are valued as specialty black caviar, which is very expensive. Only females are used in the technology of caviar aquaculture. Universal method of sex determination has not yet been developed. Most of known methods are not sufficiently accurate, or used at a relatively late age, or difficult to use. Perfect early determination of sex is considered to be impossible. Because of the dark colour of most sturgeons and important morphological differences, which fish of almost all ages have, were overlooked. We first found that the scute structure of sterlet sturgeon depends on the sex. The found dependencies with the help of machine learning algorithms open a possibility for creation of sex determination equipment using the artificial intelligence. Our results open a perspective for creation of sex determination methods for other 23 sturgeon species, which can increase the efficiency of caviar aquaculture and restoration of sturgeons in natural waters. [ABSTRACT FROM AUTHOR]