Fish feed conversion rate, growth rate and survival rate multi-character composite breeding method

Abstract

The invention relates to a fish feed conversion rate, growth rate and survival rate multi-character composite breeding method, and belongs to the field of aquatic animal genetic breeding. According to the method, composite characters are established by using phenotypic correlation, and then a genome selection method of multiple characters is established; and finally, establishing a comprehensive genome breeding value calculation method and a parent selection scheme by utilizing genetic correlation. According to the method, the problem that in a traditional method, the subjectivity of manually setting weighting parameters is too high is solved. Through comprehensive evaluation of a plurality of important economic characters, not only are the objectivity and accuracy of breeding selection improved, but also more effective resource allocation and higher genetic gain can be realized in multi-character breeding, and sustainable development of aquaculture is facilitated.

Description

Technical Field

[0001] The invention belongs to the field of aquatic animal genetic breeding, and specifically relates to a fish feed conversion rate, growth, survival rate and multiple traits composite breeding method. Background Art

[0002] Fish is one of the important sources of animal protein for humans. In recent years, my country's aquaculture output has continued to grow, especially in response to population growth and increased food demand. Feed costs account for 30% to 70% of the total production cost of fish farming and are the main expenditure of intensive farming. Feed conversion rate (FCR) is a key feature of aquaculture and directly affects the profitability and sustainability of the industry. Therefore, breeding new varieties with high FCR can reduce breeding costs and increase profits.

[0003] With the continuous development of the aquaculture industry, the breeding of a single trait can no longer meet the market's diverse demands for fish quality. Therefore, the breeding trend in recent years has begun to develop towards multi-trait breeding. However, there is still a lack of efficient composite breeding methods in multi-trait breeding. In order to effectively improve the comprehensive production performance of fish, it is urgent to develop an efficient multi-trait composite breeding technology that can take into account key traits such as feed conversion rate, growth rate and survival rate at the same time, so as to achieve better breeding results. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a highly efficient composite breeding method for fish feed conversion rate, growth, survival rate and multiple traits, so as to improve breeding efficiency and accelerate the breeding process.

[0005] The present invention is achieved through the following technical solutions: A method for composite breeding of fish feed conversion rate, growth and survival rate, wherein the method is as follows: (1) Establish more than 60 full-sib families of fish, mark the individuals in the family, and calculate the survival rate of the fish during the culture; (2) 20-30 fish were randomly selected from each family and raised individually in small cages. After a period of raising, the body weight and feed conversion rate of the individuals in the family were obtained; (3) After obtaining the two traits of body weight (BW) and feed conversion rate (FCR), genetic parameter evaluation was carried out to obtain the heritability of fish body weight and feed conversion rate traits and the phenotypic correlation between the traits. XY and genetic correlation A ; (4) The trait values ​​of body weight and feed conversion rate were Z-score standardized. The standardization formula was: Z = (X-μ) / σ, X is the original value, μ is the mean of the trait, and σ is the standard deviation of the trait; (5) Establish the composite trait CT, CT = BW Z值 + FCR Z值 (1+(1-r XY )) ; (6) DNA was extracted from all individuals and sequenced using a sequencing platform to obtain the SNP genotype of each individual; (7) For the three traits of BW, FCR and composite trait CT, three genomic selection prediction methods, GBLUP, ssGBLUP and rrBLUP, were used to estimate genomic estimated breeding values ​​(GEBVs). The prediction accuracy of the three genomic selection prediction methods was compared using the five-fold cross-validation method. (8) Select the genomic selection prediction method with the highest accuracy for the three traits of BW, FCR and composite trait CT, and calculate the individual genomic breeding value GBLUP; (9) Establish the composite genomic breeding value CGBLUP and its calculation method; CGBLUP = BW GBLUP值 + FCR GBLUP值 (1+ (1-r A )) + CT GBLUP值 ; All individuals are ranked according to the composite genome breeding value, and the families in the top 20% of the survival rate are screened. At the same time, the individuals with high composite genome breeding value are marked and used as reserve broodstock for subsequent breeding.

[0006] Furthermore, in step (1), during the family culture process, when the individuals grow to the size of commercial seedlings, more than 2,000 of them are randomly selected for special pond culture, and at the same time, the mortality is recorded and the family survival rate is calculated.

[0007] Furthermore, in step (2), weighing is required at the beginning of breeding, the breeding environment is kept consistent during the breeding process, the fish are fed until full, leftover bait is cleaned, and the daily food intake of each fish is recorded.

[0008] Furthermore, in step (2), all family individuals are marked in separate cages and randomly distributed to various culture ponds.

[0009] Furthermore, in step (3), the genetic parameters are evaluated using an animal model, and the heritability h is obtained by considering the common environmental effect. 2 , phenotypic genetic correlation XY and genetic correlation A .

[0010] Furthermore, in step (6), the tissue used for DNA extraction is fish fins to avoid damaging the fish body.

[0011] Furthermore, in step (6), the DNA can be sequenced using an Illumina sequencing platform or a BGI T7 sequencing platform.

[0012] Furthermore, in step (7), the evaluation indicator is predictive accuracy (PA), which is defined as the correlation between genomic estimated breeding values ​​(GEBVs) and phenotypic values ​​divided by the square root of heritability.

[0013] The beneficial effects of the present invention compared with the prior art: The present invention innovatively proposes a high-efficiency composite breeding technology for multiple traits of fish feed conversion rate, growth, and survival rate. By using phenotypic correlation to establish composite traits and using genetic correlation to calculate comprehensive genomic breeding values, the problem of excessive subjectivity in artificially setting weighted parameters in traditional methods is overcome. This method not only improves the objectivity and accuracy of breeding selection, but also achieves more efficient resource allocation and higher genetic gain in multi-trait breeding. In addition, the present invention enhances the comprehensive production performance of selected individuals and improves breeding efficiency through a comprehensive evaluation of multiple important economic traits, making breeding work more in line with actual breeding needs, and ultimately promoting the sustainable development of aquaculture and the improvement of economic benefits. DETAILED DESCRIPTION

[0014] The technical scheme of the present invention is further explained by the following examples, but the protection scope of the present invention is not limited in any form by the examples. In this example, turbot is used as a research object to establish a high-efficiency composite breeding technology for multiple traits such as feed conversion rate, growth and survival rate.

[0015] Example 1 1. Family establishment and seedling cultivation A one-female-one-male mating design was used to establish full-sib families. In 2023, 70 F1 full-sib families were successfully constructed at Yantai Tianyuan Aquatic Products Co., Ltd. Each family was raised in a different breeding pond and marked with a label with the family number. When the family reached 9 months of age, it was used for weight and feed conversion rate measurement experiments.

[0016] 2. Centralized temporary care before testing Before the test, 1,000 fry were randomly selected from each family and kept in a workshop for 2 weeks. After the temporary rearing, 30 fry with good vitality and no damage were randomly selected from each family and kept in a small cage culture system for 3 days without feeding (the small cage system refers to the patent obtained by the applicant, patent number 2021102057610, patent name: A device and method for large-scale determination of individual feed conversion rate of turbot).

[0017] 3. Determination of body weight and feed conversion rate After the temporary feeding period, the weight of each fish was accurately weighed and recorded (recorded as the initial weight). The juveniles were put back into the small cages and numbered separately. The next day, they were fed with commercial feed of uniform specifications. They were fed until they were full. If there was any leftover bait, it was recovered in time. The daily food intake of each fish was recorded to calculate the feed conversion rate of each fish. The experiment lasted for 60 days. During the experiment, the water temperature of each breeding pond was kept consistent, the salinity was 30, and the dissolved oxygen was > 6.0 mg / L. After the experiment, the fish were stopped from feeding for 24 hours, and the weight of each fish was accurately weighed and recorded (recorded as the final weight). The feed conversion rate of each individual was calculated using the formula: Feed conversion ratio (FCR, %) = (final body weight - initial body weight) / total weight of dry matter in the feed × 100%; After weighing, part of the tail fin of each fish was cut off and placed in anhydrous ethanol and then frozen in a -80°C refrigerator for later use.

[0018] 4. Evaluation of genetic parameters of body weight and feed conversion rate traits The feed conversion rate, body weight and other descriptive parameters of turbot from different families were sorted and preliminarily analyzed using Excel software, and variance analysis of growth-related traits from different families was performed using SPSS software, as shown in Table 1. The ASReml software package in R software was used to estimate variance components and heritability, and heritability was estimated using the individual animal model. Heritability was expressed as h 2 The phenotypic correlation and genetic correlation are represented by r XY and genetic correlation A As shown in Table 2.

[0019] Table 1 shows the variance components and heritability of body weight and feed conversion rate traits ; Table 2 shows the genetic and phenotypic correlations between body weight and feed conversion rate traits .

[0020] 5. Establishing composite trait CT Calculation method of composite traits: All individuals are Z-score standardized according to the trait values ​​of body weight and feed conversion rate. The standardization method can convert traits of different ranges into standard scores (Z-score) with the same dimension. The standardization formula is: Z = (X-μ) / σ, X is the original value, μ is the mean of the trait, and σ is the standard deviation of the trait. According to the phenotypic correlation obtained by genetic parameter evaluation, the final composite trait CT is the body weight Z value + feed conversion rate Z value (1+(1-phenotypic correlation)), that is, CT = BW Z值 + FCR Z值 (1+(1-rXY )).

[0021] 6. Genotype acquisition The library was sequenced using the Illumina sequencing platform, and the sequencing volume was required to be >6G / sample. Raw sequencing data (Raw Reads) were obtained after the machine was removed. To ensure the accuracy of subsequent analysis, the raw data needs to be quality controlled before mutation detection. The default parameters of the fastp software were used for sequencing read quality control, and the data CleanReads were obtained after quality control cleaning. Clean Reads were compared to the reference genome using BWA software. The comparison results were converted to a format using SAMtools software, and the sam file was converted into a bam file and sorted. Picard software was then used to remove redundancy, and Qualimap software was used to analyze the results. GATK software was used to detect SNPs, and sites with a missing rate of more than 20%, more than two completely different genotypes, and a minor allele frequency of less than 0.05 were filtered out. (Filtering parameters: QD < 2.0 || MQ <40.0 || FS > 60.0 || SOR > 3.0 || MQRankSum < -12.5 || ReadPosRankSum < -8.0). The software Annovar was used to annotate the SNP detection results. DNA was quantified and genotyped by whole genome resequencing, with an average sequencing depth of more than 10×. Sites with a deletion of more than 0.02 were filtered out, and self-filling was performed using the software beagle. Minor allele frequencies less than 0.05 (--MAF 0.05) and deviations from the Hartmann-Wenstein equilibrium (--hwe1e-5) were filtered out, and the final SNP sites were used for subsequent analysis.

[0022] 7. Genomic selection prediction model and cross-validation For body weight, feed conversion rate and composite traits, three methods, GBLUP (Genomic Best Linear Unbiased Prediction), ssGBLUP and rrBLUP, were used to estimate GEBVs. GBLUP and ssGBLUP belong to direct methods, that is, the variance components are estimated by iteration, and then the mixed linear model is solved to obtain the individual estimated breeding values. GBLUP is a molecular kinship matrix constructed by SNPs instead of the relationship matrix built based on pedigree. ssGBLUP integrates the relationship matrix built based on pedigree and the analytical kinship matrix built by SNPs to establish a new H relationship matrix. rrBLUP belongs to the indirect method, which first estimates the marker effect, and then accumulates the marker effect in combination with the genotype information to obtain the individual estimated breeding value.

[0023] In order to evaluate the accuracy of the three methods, GBLUP, ssGBLUP and rrBLUP, a five-fold cross-validation was performed. The reference population was randomly divided into five equal subsets, one of which had the phenotype set as missing, and the remaining four subsets were used for model training, and so on, each subset was calculated as missing once. In order to reduce the random error caused by grouping, the cross-validation results were repeated 50 times and the average was taken. The evaluation index was predictive accuracy (PA), which was defined as the correlation between GEBV and phenotypic value (r XY ) divided by the square root of heritability (h) 2 .

[0024] 8. Prediction accuracy of different GS models The prediction accuracy and bias of the three genomic selection prediction models of GBLUP, ssGBLUP and rrBLUP were compared by cross-validation method. As shown in Table 3, the prediction accuracy of the three models is high. For weight traits, the accuracy of ssGBLUP is the highest, 0.5411, followed by GBLUP at 0.5325, and rrBLUP at the lowest, 0.5203, with no significant difference among the three. For FCR traits, the accuracy of ssGBLUP is the highest, 0.5339, followed by GBLUP at 0.5278, and rrBLUP at the lowest, 0.5223, with no significant difference among the three. For composite traits, the accuracy of ssGBLUP is the highest, 0.5405, followed by GBLUP at 0.5316, and rrBLUP at the lowest, 0.5187, with no significant difference among the three.

[0025] Table 3 Prediction accuracy and bias of GBLUP, ssGBLUP and rrBLUP models .

[0026] 9. Establish the composite genomic breeding value CGBLUP and its calculation method CGBLUP calculation method for composite genomic breeding values; CGBLUP = BW GBLUP值 + FCR GBLUP值 (1+ (1-r A )) + CT GBLUP值； Taking the benefit obtained from weight gain minus the actual feed cost as the cross-validation phenotypic value, the prediction accuracy of ssGBLUP was 0.5526±0.0312 and the deviation was 0.9661±0.1253, and the deviation was close to 1, indicating that the model was an unbiased prediction.

[0027] 10. Selection of reserve broodstock All individuals are ranked according to the composite genome breeding value, and the families in the top 20% of the survival rate are screened. At the same time, individuals with high composite genome breeding values ​​are electronically marked and used as reserve broodstock for subsequent breeding.

[0028] 11. Evaluation of breeding effect After two consecutive generations of breeding, compared with ordinary seedlings, the growth rate of the selected lines increased by more than 15%, the survival rate increased by more than 20%, and the feed conversion rate increased by more than 10%.

Claims

1. A method for compound breeding of fish feed conversion rate, growth and survival rate, characterized in that: The method is specifically as follows: (1) Establish more than 60 full-sib families of fish, mark the individuals in the family, and calculate the survival rate of the fish during the culture; (2) 20-30 fish were randomly selected from each family and raised individually in small cages. After a period of raising, the body weight and feed conversion rate of the individuals in the family were obtained; (3) After obtaining the two traits of body weight (BW) and feed conversion rate (FCR), genetic parameter evaluation was carried out to obtain the heritability of fish body weight and feed conversion rate traits and the phenotypic correlation between the traits. XY and genetic correlation A ; (4) The trait values ​​of body weight and feed conversion rate were Z-score standardized. The standardization formula was: Z = (X-μ) / σ, X is the original value, μ is the mean of the trait, and σ is the standard deviation of the trait; (5) Establish the composite trait CT, CT = BW Z值 + FCR Z值 (1+(1-r XY )) ; (6) DNA was extracted from all individuals and sequenced using a sequencing platform to obtain the SNP genotype of each individual; (7) For the three traits of BW, FCR and composite trait CT, three genomic selection prediction methods, GBLUP, ssGBLUP and rrBLUP, were used to estimate genomic estimated breeding values ​​(GEBVs). The prediction accuracy of the three genomic selection prediction methods was compared using the five-fold cross-validation method. (8) Select the genomic selection prediction method with the highest accuracy for the three traits of BW, FCR and composite trait CT, and calculate the individual genomic breeding value GBLUP; (9) Establish the composite genomic breeding value CGBLUP and its calculation method; CGBLUP = BW GBLUP值 + FCR GBLUP值 (1+ (1-r A )) + CT GBLUP值 ; All individuals are ranked according to the composite genome breeding value, and the families in the top 20% of the survival rate are screened. At the same time, the individuals with high composite genome breeding value are marked and used as reserve broodstock for subsequent breeding.

2. The method according to claim 1, characterized in that In the step 2), weighing operation is required at the beginning of breeding, keeping the breeding environment consistent during the breeding process, feeding the fish until they are full, cleaning up the leftover bait, and recording the daily food intake of each fish.

3. The method according to claim 1, characterized in that In step 2), all family individuals are marked in separate cages and randomly allocated to each culture pond.

4. The method according to claim 1, characterized in that: In step 3), the genetic parameters are evaluated using animal models, and the heritability h is obtained by considering the common environmental effects. 2 , phenotypic genetic correlation XY and genetic correlation A .

5. The method according to claim 1, characterized in that In step 7), the evaluation index is prediction accuracy, and prediction accuracy is defined as the correlation between the genomic predicted breeding value and the phenotypic value divided by the square root of heritability.