International Journal of Aquaculture, 2025, Vol.15, No.2, 76-87 http://www.aquapublisher.com/index.php/ija 80 (such as shell length, weight, survival, etc.) often require manual operation, which is inefficient and easily stressful to individuals. To meet the needs of large-scale family breeding, high-throughput, non-invasive phenotypic acquisition technology is gradually being applied in abalone breeding. Among them, underwater imaging and image analysis are important directions. By installing a high-definition camera or scanner in the breeding pond, images of abalone individuals are collected regularly, and then using image analysis algorithms to automatically measure growth indicators such as shell length, shell width and volume, the growth tracking of thousands of individuals can be achieved without taking measurements one by one (Dai et al., 2024). This method has been verified in some aquaculture, such as the prediction of fish length and behavioral monitoring. For attached growing abalone, combined with special lighting and background plates, an automatic shooting and size recognition system is expected to be established to significantly improve the efficiency of phenotypic data acquisition. In addition, technologies such as ultrasonic imaging have also introduced abalone breeding. Studies have tried to use ultrasound to non-invasively detect the gonad development status of abalone to distinguish gender and maturity. This method avoids the harm caused by traditional anatomy examination to pro-balm, making screening of breeding groups more efficient. Furthermore, environmental sensors and intelligent monitoring platforms are also under development, which can record breeding environmental parameters and phenotypic behavioral data such as abalone feeding and exercise in real time (Williams et al., 2018). Through the Internet of Things, this data is integrated into a breeding database, breeding experts can monitor the growth and health of each family in real time during the busy season and adjust their breeding strategies in a timely manner. 4.2 Automation and scale development of molecular marker detection technology The advancement of genotyping technology has greatly promoted the transition from experience to precise seed selection. In abalone breeding, molecular marker detection has evolved from the past low-throughput manual operation to a highly automated large-scale process. First of all, the high-density SNP classification technology based on chips has been applied. By customizing the SNP chip, genotypes of tens of thousands of loci can be detected on each abalone in one experiment. Secondly, the popularity of the next generation of sequencing (NGS) technology provides solutions for simplifying methods such as genome sequencing and target sequence capture. Using RAD-seq or GBS technology with restriction enzyme digestion, thousands of abalones can be sequenced in parallel, and thousands of SNP marker data can be obtained for genetic evaluation and selection. Compared with traditional methods, the data analysis process can be completed automatically, and data between different batches and laboratories have good consistency (Zhao et al., 2023). Furthermore, laboratory automation equipment greatly improves detection efficiency and accuracy. The adoption of fully automatic DNA extractor and PCR loading system enables 384 or even 1536 samples to be processed simultaneously per round of molecular detection, reducing artificial error and working intensity. In some large domestic aquatic breeding centers, the equipment of high-throughput sequencing platforms and computing servers makes it a reality to process tens of thousands of abalone samples each year. Finally, as the cost decreases, the application of whole-genome resequencing in breeding has also been explored. Some studies have conducted whole-genome sequencing of dozens to hundreds of abalones to identify functional mutations and provide a basis for marker-assisted selection. 4.3 The role of bioinformatics and cloud computing in data analysis As abalone breeding enters the genome era, the analytical and processing ability of large-scale data has become one of the keys to the success or failure of breeding. On the one hand, the amount of data generated by genome sequencing and large population typing is extremely large, requiring powerful bioinformatics tools for management and interpretation. A genome-wide association analysis (GWAS) may involve tens of millions of genotypes and trait data of thousands of individuals, and its computational complexity is very high. To this end, researchers have developed bioinformatics software specifically for aquaculture breeding, such as the "AquaGS" system based on a graphical user interface, which can support users to conveniently perform GS prediction model construction and cross-validation analysis. This type of software integrates a one-stop process from data import, quality control, model training to result visualization, greatly reducing the technical threshold for breeding experts to apply GS (Liu et al., 2023). At the same time, more and more research relies on high-performance computing (HPC) and cloud platforms to process breeding data. Uploading huge genomic data to the cloud can quickly complete tasks such as genome assembly, variation detection and breeding value prediction using the parallel
RkJQdWJsaXNoZXIy MjQ4ODYzNA==