A successful breeding program depends on the accurate identification of genetically superior individuals that will parent the next generation. Factors that affect accuracy include the number of animals enrolled in the program, the heritability and genetic architecture of the trait, and the method used for prediction breeding values. In addition to that, selection intensity and generation interval are factors that will affect genetic gain.
Traditionally, shrimp breeding strategies rely in selection within family. In such approach, individuals are selected according to their pedigree, and random animals from the best families are kept for reproduction. This approach is limited by its low accuracy, especially for traits with low heritability, which is usually the case for disease resistance. In addition, such methodology cannot identify the difference between animals of the same family, which may lead to an inaccurate selection of candidates.
In the past decade, the inclusion of dense genetic marker chips in genetic evaluations became a commonplace in livestock breeding programs. This technology can be used in shrimp aquaculture resulting in higher accuracies and shorter generation intervals, which leads to higher genetic gains. The single-step Genomic BLUP methodology has the ability to include phenotypes, genotypes, and pedigree of multiple animals, which makes it the method of choice for aquaculture commercial applications. Such method uses the same models as used in pedigree/family selection programs, with little change of statistical models. Moreover, this method allows inclusion of multiple traits and animals without genotypes in the evaluation.
Another challenge in breeding programs for disease resistance is the accuracy of phenotypes. Phenotypes are usually collected after challenge studies, in which individuals are exposed to pathogens and their survival is observed over time. Common phenotypes are a binary survival status at the end of the challenge, or survival days. While progress can be achieved with this information, better phenotypes would optimize the time and resources dedicated to challenge studies. Alternative and novel phenotypes include PCR for quantifying pathogen load, ELISA for toxin levels, microbiome for bacterial community analysis, among others.
Finally, shrimp breeding programs struggle to optimize the number of animals without reducing profitability. There are many questions regarding the number of animals, genotyping strategies, and chip density for shrimp. Simulations show that the current chips with 50k markers are adequate for accurate breeding programs. Moreover, data collection and genotyping must be carried out for at least 3 generations. Finally, collecting data on more animals per family will yield higher accuracies then studying the same number of individuals across several families.
Genomic selection programs can be implemented for accelerating the pace of genetic gain for disease resistance in shrimp. Genetic gains are cumulative and there are no concerns about pathogen resistance or environmental impacts. More importantly, genetic gains are spread to the whole population, which facilitates the observation of the results in commercial implementations. Finally, information about the ideal number of animals and markers will only become available once programs are implemented and those variables tested.