| Macrobrachium rosenbergii is an important economic species for freshwater aquaculture in China with fast growth,easy cultivation,strong adaptability,and abundant nutrition.Traditional breeding projects only take weight traits as the main breeding objectives ignoring other important economic traits that may affect weight traits,such as survival rate,disease resistance,and different morphological threshold traits.Accurate estimation of genetic parameters is the prerequisite for the smooth development of breeding work.Traditional genetic evaluation methods are usually based on pedigree information to estimate the breeding value of a single trait.However,this method does not consider the covariance between traits.It may lead to inaccurate estimation of breeding values.The use of multi-trait genetic assessment can obtain higher accuracy.With the continuous development of molecular marker technology,the genomic selection method is used in aquaculture breeding,further improving selection accuracy.The primary purpose of this study was to construct p BLUP,p BLUP-GG,ss GBLUP,and ss GBLUP-MF models under single trait and two trait methods,and to provide a reference breeding analysis model and evaluation method for accurately evaluating the breeding population of M.rosenbergii.The genetic parameters of survival traits in different generations and their genetic correlation with harvest weight were estimated.The selection potential of the proportion of male and female M.rosenbergii in different morphology,and the effect of improved traits on the growth rate of M.rosenbergii were evaluated.The main contents and results of this study are as follows:1.Prediction ability analysis of genome genetic assessment model for growth traits of M.rosenbergiiIn this study,two commercial and two wild populations were collected as founder populations.The optimal contribution method was used to develop mating schemes,produce families,and genotype the parents of G2 and G3 generations.Four models,including p BLUP,p BLUP-GG,ss GBLUP,and ss GBLUP-MF with Metafounders,were used to compare the accuracy and deviation of estimating the breeding value of harvest weight using different breeding analysis models.The prediction accuracy and bias of the estimated breeding value of harvest weight using single trait and multiple trait breeding analysis models were compared for the harvest weight and harvest length traits.When the multi-trait model is used,the prediction accuracy obtained by the p BLUP-GG and ss GBLUP-MF models is 7.69%higher than that by the p BLUP model,and the deviations are reduced by 27.54%and 26.81%.The results showed that the prediction ability（0.28）of the ss GBLUP-MF model considering the effect of the unknown parent group was higher than that of the p BLUP model（0.25）under the multi-trait model,and a lower deviation（0.99 vs.1.38）was obtained,which could more accurately estimate the additive genetic variance.In this study,the two-character model did not show apparent advantages compared with the single-character model.It is speculated that the lack of phenotypic information of the two traits is less and the heritability of the two traits is similar,so the two-character model cannot be used.2.Genetic correlation analysis of growth and survival traits of M.rosenbergiiIn this study,we collected the records of individual survival status of 116,700individuals from four generations of the core breeding population of M.rosenbergii in China at high rearing densities（10.65～14.39 individuals/m~2）,and the genotyping information of 410 individuals.The genetic parameters of the four generations of survival traits of M.rosenbergii were estimated using p BLUP and ss GBLUP methods.The results showed that the variation coefficients of survival traits in different generations of M.rosenbergii were 10.9%～27.3%,belonging to the medium and high variation levels,indicating that genetic improvement of survival traits had great potential.The estimated heritability of the four generations from G0 to G3 was low,from 0.05±0.01 to 0.20±0.02.The genetic correlations between harvest weight and survival traits were low and negative（-0.02～-0.40）in G0,G2 and G3 generations,and was low and positive in G1 generation.The prediction accuracy of heritability estimates obtained by ss GBLUP model has not been significantly improved,which may be due to the low heritability of survival threshold traits in this study and the small proportion of genotyping individuals in the total.3.Genetic correlation analysis of growth and morphological threshold traits of M.rosenbergiiIn this study,the morphological and body weight information of 40,935individuals in the two generations（G3～G4）of the core breeding population of M.rosenbergii was recorded.The genetic parameters of threshold traits of different morphological dichotomies（0 and 1）between generations and within generations were estimated using the parental breeding analysis model and the threshold statistical model,and the genetic correlation between threshold traits and body weight traits was estimated using the two-character model.The estimation results of variance components showed that the heritability estimates of male and female shrimps with different morphological thresholds were 0.08～0.12 and 0.04～0.22,respectively.There was a low positive correlation between morphological OC,BC threshold traits and harvest weight（0.01～0.43）.There were moderate correlations between IF,BF,SF threshold traits and harvest weight traits of female shrimp（-0.61～0.53）.There are available additive genetic variations in different morphological threshold traits of male and female M.rosenbergii,and there is a large room for improvement.Selecting OC male shrimp and increasing its proportion in breeding population will not reduce its growth rate;selecting IF female shrimp and increasing its proportion in breeding population may reduce its growth rate. |
PDF Full Text Request