CERVUS vs. COLONY for successful parentage and sibship determinations in freshwater prawn Macrobrachium rosenbergii de Man

This study assesses the properties of 10 microsatellite loci and two popular computer programs for likelihood-based parentage identification in commercial strains of the giant freshwater prawn Macrobrachium rosenbergii. It was our interest to determine the number of markers needed to optimize between genotyping cost and power of the test. Nine loci were informative, with average expected heterozygosity of 0.80 and PIC score of 0.76. Low and non-significant frequencies of null alleles (< 0.05) were detected at the nine loci. Simulations implemented in the CERVUS program were performed to determine the number of informative loci needed to maximize assignment success. The accuracy in assignment was determined in 21 full-sib and two paternal half-sib families using two contrasting methods, a pair-wise likelihood comparison approach in CERVUS and a full-pedigree likelihood method in the COLONY program. When genotyping error was set at 0.05, use of four highly informative loci was sufficient for COLONY to resolve the genetic structure of this population, while seven loci would be required to obtain 94–99% correct assignment with CERVUS. As typing rate of error increased to 0.2, COLONY required only four loci to assign parentage to 85–90% of the offspring, while CERVUS required five to seven loci to obtain the similar proportions of assignment. Results suggest that this set of microsatellites, used in conjunction with COLONY would be an effective tool for parentage and sibship identification, testing performance of families and estimating genetic parameters in selective breeding programs of the giant freshwater prawn.

► We compare performance of 2 computer programs for parentage analysis in giant freshwater prawn.
► Nine microsatellite loci with expected heterozygosity of 0.80 and PIC score of 0.76 were used.
► COLONY required four loci to obtain 90% assignment rate with 0.01–0.2 error rates.
► CERVUS required five to seven loci to obtain similar assignment accuracy.