Abstract
Considerable demand exists for blue and green eggs, especially in Southeast Asia. Although some chicken breeds lay eggs with shell colors ranging from light blue to dark green (olive), genetic lines producing uniform blue eggs or green eggs are preferred for commercial production. Our previous studies demonstrated the difficulties of selection on human scoring of the visual eggshell color (AveObs; 1=light blue, 2=blue, 3=green, 4=olive), and suggested to select on 3 colorimeter measurements (L*, a*, and b*), and on Shell Color Index (SCI, =L*–a*–b*). In the present study, these 4 parameters were the criteria of selection for blue-eggshell (Blue) group (high SCI, high L*, low b*, and low a*) and for green-eggshell (Green) group (SCI≈75, L*≈80, b*≈12, and low a*). The base population was generation G5 of an experimental line called BG, of black-bone chickens with AveObs ranging from 1 to 4. Generations G6, G7, and G8, progeny of 3 cycles of selection, were evaluated in this study. In the Blue group, the first selection cycle significantly changed the G6 means of the colorimeter parameters and of AveObs, and significantly reduced the colors phenotypic variance (Vp). The following two cycles of selection in the Blue group hardly changed the means but further reduced Vp, and the heritabilities. In the Green group, in line with high heritabilities, the three selection cycles led to significant changes from G5 to G8 in the means of the eggshell colorimeter parameters, and mean AveObs increased from 2.6 to 2.9. However, the Vp in the Green group remained almost as high as in G5, probably due to the polygenic control of the combinations of the two pigments, biliverdin and protoporphyrin, that jointly determine eggshell shades from greenish blue to olive. Selection targets derived from bivariate plots suggest that one additional cycle of selection may fix uniform Blue group (AveObs=2), whereas more cycles of selection will be needed to reach uniformity of the green color (AveObs=3) in the Green group.
Keywords: Blue or green eggs, Eggshell color, Selection response, L*a*b* colorimeter parameters, Visual color
Introduction
Eggshell color is an important breeding goal for egg production, due to its association with eggshell quality, and also its attractiveness to consumers and the consequent economic value (Lukanov et al., 2015). The common eggshell colors range between white (the basic color of eggshell matrix) and brown, which is caused by the deposition of protoporphyrin IX mostly in the outside layer of the eggshell (Samiullah et al., 2015; Bi et al., 2018). Protoporphyrin IX is synthesized from heme via the δ-aminolevulinic acid pathway (Ryter and Tyrrell, 2000; Sachar et al., 2016), and regulated at the molecular level by key genes such as ALAS1 and ABCG2 (Li et al., 2013; Yang et al., 2022). The Minolta colorimeter has been widely used to measure the L* (lightness), a* (redness), and b* (yellowness) values of the eggshell color. By selecting for higher values of the calculated Shell Color Index (SCI, = L*–a*–b*), commercial brown-egg lines have been bred successfully to lay darker eggs (Cavero et al., 2012), for consumers who prefer eggs with uniform dark brown shells. However, there is also considerable consumer demand for eggs with blue or green shells, especially in Southeast Asia (Guyonnet, 2022). The main pigments responsible for the blue eggshell color are biliverdin and biliverdin zinc chelate, whereas the shell of greenish eggs contains, in addition to biliverdin, protoporphyrin in the outside layer of the eggshell (Zhao et al., 2006). The wide phenotypic range of blue-greenish colors, from very light blue to very dark green (olive green) reflects the relative contents of these pigments in the eggshell (Wang et al., 2009; Zeng et al., 2022). However, because consumers prefer eggs with uniform eggshell color (Cavero et al., 2012; Mulder et al., 2016; Guo et al., 2020) and the associated practical marketing considerations, it is preferred to have genetic chicken lines that lay either blue-only or green-only eggs.
Due to the involvement of these two pigments (biliverdin and protoporphyrin), it appears to be more complicated to breed for blue eggs or for green eggs, compared to selection for dark-brown eggs. Several studies have attempted to reveal the genetic control of blue-greenish eggs, by using L*a*b* measurements along with molecular detection of the Oocyane (O) gene (Xu, 2018; Dierks et al., 2019; Chen and Wang, 2022; Beranova et al., 2024). Three O genotypes (O/O, O/o and o/o) could be determined through the identification of the EAV-HP insertion in the SLCO1B3 gene, which has been proved to be the basic genetic difference between chickens that lay blue/green eggs and those laying non-blue/green eggs (Wang et al., 2013; Wragg et al., 2013). However, it is a challenge to produce a pure lines where all hens lay either blue eggs (i.e., with biliverdin only) or green eggs (i.e., with specific combination of biliverdin and protoporphyrin), because even in a population of homozygotes O/O that do not lay white or brown eggs, there still exists an appreciable range of blue-greenish shades (Chen and Wang, 2022; Beranova et al., 2024). In our previous study (Wang et al., 2023a), L*, a*, b* were measured on eggs laid by hens from an experimental line (named BG) bred for black bone and blue/green eggs. The values of L*, a*, b*, and the calculated SCI, were highly phenotypically and genetically correlated with the average of visual eggshell color classification by 4 independent human observers (AveObs), where each observer classified the shell color of a representative egg from each hen as: 1 = light (blue), 2 = blue, 3 = green, or 4 = olive (dark/brownish green). Based on these associations, Wang et al. (2023a) suggested using combinations of L*, a*, b* and SCI as criteria to select for blue eggs or for green eggs; the present study evaluates the responses to three cycles of actual selection using these criteria.
Materials and methods
The BG line and its management
The experimental line called BG has been bred at Hangzhou Academy of Agricultural Sciences (Wang et al., 2021, Wang et al., 2022, Wang et al., 2023a). It originated from the segregating progeny of a cross between Dongxiang blue eggshell chicken and Jiangshan black-bone chicken, with the aim of combining black body (and meat) and blue-greenish eggshell. In 2015, the base population of the BG line (G0) was produced after its paternal and maternal parents were confirmed to be homozygous (O/O) for the Oocyan (O) gene, using PCR technology (Wragg et al., 2013). From generations G1 to G3, the BG line was expanded by annual random mating. In 2019 and 2020, generations G4 and G5 were reproduced by family mating, with sires mated to non-sib dams.
In the four generations presented in this study (G5, G6, G7, G8), all birds were reared in stack-style brooding and growing group cages, from 0 d to 56 d and from 57 d to 105 d, respectively. Subsequently, females and males were housed in individual cages, each providing 700 cm2 and 1200 cm2 per female and male, respectively. After 16 weeks of age, day length was increased artificially in 30-min increments every two weeks until reaching 16 h of light per day. From day-old chicks to laying hens, most management procedures followed the guidelines of Hy-Line International (https://www.hyline.com/), except for nutrition. During the brooding, growing, and laying periods, the diets provided 19.5% crude protein (CP) and 2,900 kcal/kg metabolizable energy (ME), 16.5% CP and 2,800 kcal/kg ME, and 17.5% CP and 2,850 kcal/kg ME, respectively. All rearing and breeding practices were approved by the Animal Ethics Committee of Hangzhou Academy of Agricultural Sciences.
Eggshell colorimeter measurement and shell color classification
Table 1 shows the numbers of eggs whose eggshell colors were measured, and the number of hens that laid these eggs, by generation and age. The average, standard deviation (SD), minimum and maximum numbers of eggs per hen are also presented. Most averages ranged between 3 and 6 eggs/hen/age, with SD values ranging between 0.89 and 2.23. The Minolta colorimeter CR400, a portable instrument commonly applied in many fields, was used to measure the color parameters L*, a*, and b*, indicating the color’s lightness, position on the red-green scale, and position on the blue-yellow scale, respectively (Konica Minolta, 2007). These measurements, and the calculated parameter SCI (= L*–a*–b*), were recorded for each measured egg. For each hen, the averages (of all measured eggs) of these four colorimeter parameters at each age (around 200 d and around 300 d) were defined as L*200 and L*300, a*200 and a*300, b*200 and b*300, SCI200 and SCI300, respectively. However, to better represent the hens' genetic potentials, and based on the highly significant correlation between each hen's means at 200d and 300d in this study (data not shown) and a previous one (Wang et al., 2023a), the results presented and discussed in this article are the means of these parameters over both ages, defined as L*200&300, a*200&300, b*200&300, and SCI200&300.
Table 1.
Total and per-hen number of eggs that their shells were measured by Minolta colorimeter in each generation and age; only hens which had eggs at both ages (200d and 300d) were studied.
| Generation | Age | Total number of hens with eggs | Total number of eggs | Eggs per hen |
|||
|---|---|---|---|---|---|---|---|
| Mean | SD1 | Min2 | Max3 | ||||
| G5 | 200 | 392 | 1970 | 5.03 | 1.57 | 1 | 9 |
| 300 | 392 | 1501 | 3.83 | 1.15 | 1 | 7 | |
| 200&300 | 3471 | ||||||
| G6 | 200 | 356 | 1674 | 4.70 | 1.46 | 1 | 8 |
| 300 | 356 | 2138 | 6.01 | 2.22 | 1 | 10 | |
| 200&300 | 3812 | ||||||
| G7 | 200 | 183 | 934 | 5.10 | 2.23 | 1 | 9 |
| 300 | 183 | 1834 | - | - | - | - | |
| 200&300 | 1117 | ||||||
| G8 | 200 | 165 | 576 | 3.49 | 1.15 | 1 | 7 |
| 300 | 165 | 359 | 2.18 | 0.89 | 1 | 4 | |
| 200&300 | 935 | ||||||
Standard Deviation of the number of eggs per hen.
Minimum number of eggs per hen.
Maximum number of eggs per hen.
At 300d in G7, a single representative egg per hen was measured.
As described previously (Wang et al., 2023a), a single representative egg was selected from each hen at each age, and its visual color was classified by the same 4 trained human observers, as either “Light” or “Blue” or “Green” or “Olive”. These assessments at different ages were all performed in a specialized viewing room under illuminant D65 using full-spectrum light emitting diode lamps. The numerical values of the 4 visual color classes (Light = 1, Blue = 2, Green = 3, Olive = 4) from the 4 observers were averaged for each egg and termed AveObs (with 13 possible values from 1 to 4 in increments of 0.25). For each hen, the AveObs at around 200 d, around 300 d, and the mean over the two ages were termed AveObs200, AveObs300, and AveObs200&300.
Selection and breeding of eggshell color groups
In all generations, selection for skin color and egg production was applied (Wang et al., 2021, 2022). In G5, some females and males were selected to serve as dams and sires for blue-eggshell (Blue) and green-eggshell (Green) groups, based on the four colorimeter parameters (SCI, L*, a*, and b*). For the Blue group, the selected hens were those with eggs averaging high SCI (∼85), high L* (∼84), low b* (∼6), and low a* (∼–7.5), whereas for the Green group, the selected hens were those with eggs averaging SCI around 75, L* around 80, b* around 12, and low a* (∼–7) (Wang et al., 2023a). The sires of the Blue and Green groups were full-sib males of the selected hens. The same selection criteria were applied, within each color group (Blue or Green), for selecting G6 females and males as dams and sires to produce G7, and G7 females and males as dams and sires to produce G8. Each color-selected sire was mated with 3 to 6 randomly assigned non-sib dams selected for the same eggshell color (blue or green), producing two color-specific (Blue and Green) groups of full-pedigree progeny chicks, in 2-3 consecutive hatches per generation.
Number of hens per generation and group
The data used in this study were taken only from hens that had at least one measured egg around 200 and around 300 days of age: 392 in G5, 141 in G6, 99 in G7, and 105 in G8. Table 2 shows also the family structure in these generations (the average number of hens per dam and per sire), as well as the number of selected dams and sires in generations G5, G6, and G7, in the Blue and Green eggshell color groups. There were between 8 and 16 selected sires and between 28 and 43 selected dams, per group/generation, and the average number of progeny hens per sire and per dam ranged from 4 to 5 and from 1.5 to 2.1, respectively (Table 2). The reduced variation within the selected Blue and Green groups after the first cycle of selection (Fig. 2) allowed the reduction in number of hens per generation while adding two more generations (G7 and G8). The overall statistical reliability was assured by measuring sufficient numbers of hens over the 3 generations of the selected groups.
Table 2.
Number of all hens, selected dams, and selected sires in generations G5, G6, G7 and G8, in the Blue and Green eggshell color groups.
| Color group | G5 |
G6 |
G7 |
G8 |
||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| All-hens | Selected-G5 dams | Selected- G5 sires | All-hens | Hens/G5 dam | Hens/G5 sire | Selected-G6 dams | Selected-G6 sires | All-hens | Hens/G6 dam | Hens/G6 sire | Selected-G7 dams | Selected-G7 sires | All-hens | Hens/G7 dam | Hens/G7 sire | |
| Blue | 392 | 28 | 7 | 59 | 2.1 | 8.4 | 33 | 12 | 49 | 1.5 | 4.1 | 20 | 8 | 32 | 1.6 | 4.0 |
| Green | 32 | 8 | 82 | 2.6 | 10.3 | 30 | 10 | 50 | 1.7 | 5.0 | 43 | 16 | 73 | 1.7 | 4.6 | |
Fig. 2.
The range and distributions of individual hen values of the 5 eggshell color parameters in G5 (the base population, 
) and in the Blue (
) and Green (
) groups in G6 and G8. The phenotypic variances (Vp) of each distribution are also shown. The asterisks above each pair of Vp values indicate the significance (* <0.05, ⁎⁎ <0.01, ⁎⁎⁎ <0.001) of the difference between Vp of Blue hens versus Vp of Green hens in G6 and in G8. In all 5 parameters, the Vp among the Blue hens in G6 were significantly lower than the corresponding Vp of all hens in G5, whereas the Vp among the Green hens did not differ significantly than those of G5 hens (except of AveObs).
Statistical analysis
The inbreeding coefficient was determined in each generation of the BG line and was found to be very low (G6: 0.009; G7: 0.015; G8: 0.017), thus neither affecting performance nor biasing the estimates of genetic parameters. The JMP (v. 14.0) software (https://www.jmp.com/) was used to conduct all the statistical tests. For each trait, t-tests between the means of the Blue versus Green groups were conducted in G6 and in G8. The Tukey HSD test was used to test the differences between the mean of G5 (the base generation), and the means of G6 and G8 in each color group (Blue and Green). Two-sided F-tests were used to compare the phenotypic variances (Vp) of individual hens in G5 (base population) versus those of the Blue and Green groups in G6, and to compare the Vp of Blue hens with that of Green hens in generations G6 and G8. The correlation coefficients between AveObs and each of the 4 colorimeter parameters (L*, a*, b*, SCI) were weighted by the number of individuals with each value of AveObs, in Blue and Green groups.
Using the individual animal model software ASReml 4.1 (https://asreml.kb.vsni.co.uk/), estimates of heritability (h2) of each trait, and of the genetic correlation (rG) and phenotypic correlation (rP) between the traits, were calculated for the base population (G5) and for the Blue and Green groups from the data combined over generations G6, G7, and G8. The data for these calculations were means of all measured eggs per hen, of the five color parameters. The ASReml calculations used available pedigree information in each generation as random effects, and generation and hatch-within-generation as fixed effects. The significance levels (P > 0.05, P < 0.05, P < 0.01, or P < 0.001) of h2, rG and rP were determined by the likelihood ratio test of ASReml.
Results and discussion
Means of color parameters in the base population and the blue and green groups
The selection for either Blue or Green eggshell color started in G5. The effects of these selections on the two color groups' means are presented in Fig. 1 by plotting the means of all hens in G5 (the base population), and of the hens in the Blue and Green groups in G6 and G8, for each of the 5 color parameters. The mean of the base population (G5) was intermediate between those of the Blue and Green groups in the selected generations (G6 and G8) for each color parameter, except a*. The differences between the means of the two selected groups were significant already after the first selection cycle, in the four colorimeter parameters (L*, a*, b*, and SCI) that served as selection criteria, and also in the mean visual color (AveObs). Except for a*, the differences between the means of the Blue and Green groups further increased after the second and third selection cycles (from G6 to G8) due to additional selection responses in the Green group, whereas the means of the Blue group hardly changed from G6 to G8 (Fig. 1). Within each color group, the pattern of changes in the means from G5 to G6 to G8 was similar in L* and SCI, reflecting the high positive correlation between L* and SCI. In the Blue group, the pattern of changes in the group means of a* and b*, from G5 to G6 and G8, was opposite to that in L* and SCI, reflecting the negative correlations between these two parameters (a* and b*) and the two other parameters. Correlations among all 5 color parameters were reported by Wang et al. (2023a), and the correlations (and their significance) that were calculated from the data of G5 (the base generation) are reported in Table 4 of this study.
Fig. 1.
The means of five eggshell color parameters in the base population (G5, ) and in the selected Blue group () and Green group () in generations G6 and G8. The blue and green horizontal dotted lines indicate the selection targets of each color group, as shown in Fig. 3.
Table 4.
Estimates of heritability (diagonal) and correlations (phenotypic above diagonal, genetic below diagonal) among the traits averaged over two ages (200d and 300d), calculated from the G5 (the base population) data, and from data combined over the three selected generations (G6, G7, G8) in the Blue and Green groups.
| Traits | Data | L*200&300 | a*200&300 | b*200&300 | SCI200&300 | AveObs200&300 |
|---|---|---|---|---|---|---|
| L*200&300 | Base population (G5) | 0.766⁎⁎⁎ | –0.384⁎⁎⁎ | –0.838⁎⁎⁎ | 0.917⁎⁎⁎ | –0.661⁎⁎⁎ |
| Blue (G6,G7,G8) | 0.415 | 0.521⁎⁎⁎ | –0.664⁎⁎⁎ | 0.842⁎⁎⁎ | –0.646⁎⁎⁎ | |
| Green (G6,G7,G8) | 0.551⁎⁎ | –0.236⁎⁎ | –0.735⁎⁎⁎ | 0.865⁎⁎⁎ | –0.748⁎⁎⁎ | |
| a*200&300 | Base population (G5) | –0.559⁎⁎⁎ | 0.721⁎⁎⁎ | 0.681⁎⁎⁎ | –0.680⁎⁎⁎ | 0.415⁎⁎⁎ |
| Blue (G6,G7,G8) | 0.731 | 0.173 | –0.009 | 0.076 | –0.365⁎⁎⁎ | |
| Green (G6,G7,G8) | –0.278 | 0.671⁎⁎⁎ | 0.659⁎⁎⁎ | –0.646⁎⁎⁎ | 0.465⁎⁎⁎ | |
| b*200&300 | Base population (G5) | –0.880⁎⁎⁎ | 0.809⁎⁎⁎ | 0.972⁎⁎⁎ | –0.976⁎⁎⁎ | 0.664⁎⁎⁎ |
| Blue (G6,G7,G8) | –0.627 | –0.242 | 0.284 | –0.943⁎⁎⁎ | 0.564⁎⁎⁎ | |
| Green (G6,G7,G8) | –0.666⁎⁎ | 0.750⁎⁎⁎ | 0.576⁎⁎⁎ | –0.961⁎⁎⁎ | 0.824⁎⁎⁎ | |
| SCI200&300 | Base population (G5) | 0.936⁎⁎⁎ | –0.792⁎⁎⁎ | –0.985⁎⁎⁎ | 0.905⁎⁎⁎ | –0.682⁎⁎⁎ |
| Blue (G6,G7,G8) | 0.890 | 0.399 | –0.932 | 0.296 | –0.609⁎⁎⁎ | |
| Green (G6,G7,G8) | 0.833⁎⁎⁎ | –0.734⁎⁎ | –0.952⁎⁎⁎ | 0.581⁎⁎⁎ | –0.842⁎⁎⁎ | |
| AveObs200&300 | Base population (G5) | –0.940⁎⁎⁎ | 0.710⁎⁎⁎ | 0.896⁎⁎⁎ | –0.933⁎⁎⁎ | 0.646⁎⁎⁎ |
| Blue (G6,G7,G8) | –0.646 | –0.685 | 0.743 | –0.793 | 0.099 | |
| Green (G6,G7,G8) | –0.799⁎⁎ | 0.881⁎⁎ | 0.929⁎⁎ | –0.994⁎⁎⁎ | 0.469* |
*,⁎⁎,⁎⁎⁎ The correlation coefficient differ significantly from zero at P < 0.05, P < 0.01, and P < 0.001, respectively.
Detailed numerical information regarding the selection in the two color groups and their consequences, and the required statistical tests, is presented in Table 3. The first column shows the means of All-hens in G5 (base population), followed by the selection differential (SelDif) between the means of All-hens in G5 and the means of the G5 hens selected to be dams of either the Blue or the Green groups in G6. The following column shows the means of all Blue hens and all Green hens in G6, after the first cycle of selection. The next column shows SelDif (G6+G7), the sum of the selection differentials in G6 and in G7, and it is followed by the means of all Blue hens and all Green hens in G8, after the third selection cycle. The selection criteria for blue eggshell were high SCI, high L*, low b*, and low a*. They yielded substantial SelDif(G5) values of the selected Blue dams in G5 for all 5 color parameters, and led to the significant changes from G5 (base population) to G6 in the Blue group (Table 3). Compared to the base population (All-hens in G5), the Blue group All-hens means of the 4 colorimeter parameters in G6 were significantly different (higher in L* and SCI, lower in a* and b*), and accordingly the mean of the eggshell visual color (AveObs) was significantly reduced, from 2.67 (more green than blue) to 2.38 (more blue than green). The selection criteria applied in the Green group were intermediate for SCI (around 75), for L* (around 80), and for b* (around 12), and accordingly they yielded low values of SelDif (G5) compared to those in the Blue group. Consequently, in the Green group, only two means of All-hens in G6 were significantly different from those in G5 (higher in b* and lower in SCI), and the mean visual color (AveObs) of the Green group only slightly increased, from 2.67 in G5 to 2.77 in G6 (Table 3, Fig. 1).
Table 3.
Means of the eggshell color parameters of all BG hens in generation G5 (the base population) and of all the hens in the Blue and Green groups in generations G6 and G8; selection differentials (SelDif) are presented for the first selection cycle (G5) and combined over the second and third cycles (G6+G7).
| Color parameter | Color group | All-hens (G5)1 | SelDif (G5)2 | All-hens (G6) | SelDif (G6+G7)3 | All-hens (G8) |
|---|---|---|---|---|---|---|
| L*200&300 | Blue | 81.92Ba | 2.63 | 83.27A | 1.08 | 82.84A |
| Green | –1.32 | 81.30a | –0.86 | 78.81b | ||
| a*200&300 | Blue | –6.22Aa | –0.84 | –7.17B | –0.87 | –7.27B |
| Green | –0.30 | –6.37a | –0.47 | –6.35a | ||
| b*200&300 | Blue | 9.42Ac | –3.09 | 8.03B | –2.69 | 7.67B |
| Green | 1.07 | 10.60b | 0.60 | 11.89a | ||
| SCI200&300 | Blue | 78.72Ba | 6.56 | 82.42A | 4.63 | 82.44A |
| Green | –2.09 | 77.07ab | –0.99 | 73.27b | ||
| AveObs200&300 | Blue | 2.67Ab | –0.47 | 2.38B | –0.42 | 2.30B |
| Green | 0.28 | 2.77ab | 0.40 | 2.91a |
All-hens (G5): all the hens in G5, the base generation of selection on eggshell colors) In G5, AveObs were determined at 250 d (instead of averaging 200 d and 300 d).
SelDif (G5): the differences between the means of all hens in G5 and the hens selected to be dams of the Blue and Green groups in the following generation (G6).
SelDif (G6+G7): the differences between the means of all hens and the hens selected to be dams of the following generation, calculated in G6 and G7, and combined.
,B In each color parameter, means of All-hens in G5 and the Blue group in G6 and G8 with no common superscript differ significantly at P < 0.05.
,b,c In each color parameter, means of All-hens in G5 and the Green group in G6 and G8 with no common superscript differ significantly at P < 0.05.
The differences between All-hens means in G8 versus G6 were significant in the Green group (lower L*, higher b* and lower SCI) and also their mean AveObs increased from 2.77 to 2.91, whereas the G8 means of the Blue group were similar to the corresponding G6 means in all 5 color parameters. Thus, in contrast to the first selection cycle, only in the Green group were there significant responses to the second and third selection cycles (Fig. 1, Table 3), with no corresponding responses in the Blue group. The high response to selection for Green, and the lack of response to selection for Blue, suggest different heritabilities in these two groups.
Heritability of color parameters in the base population and the blue and green groups
The heritability (h2) estimates of the 5 color parameters, and their significance, are presented in the diagonal of Table 4, whereas their standard errors (SE) are shown in the diagonal of Table 5. They were calculated from all the hens in G5 (the base population) and from the combined data (G6+G7+G8) of all hens in each of the two selected groups, Blue and Green. The data for each selected group were combined over the three generations due to insufficient data (limited number of sires and dams) in each single generation. The h2 estimates in G5 were very high, from 0.972 for b*, 0.905 for SCI, 0.766 for L* and 0.721 for a*, and relatively high (0.646) for AveObs. Due to the high h2 values (Table 4) and their low SE values (Table 5), all the h2 estimates in G5 were highly significant. The heritabilities in the base population, along with high SelDif values, led to the substantial responses to the first cycle of selection for blue eggshell (Fig. 1, Table 3). The responses to the first cycle of selection for green eggshells should reflect the same base-generation h2 estimates, but they were low and non-significant, because the Green SelDif (G5) values were lower than the corresponding Blue SelDif (G5) values (Table 3).
Table 5.
Standard errors of heritability (diagonal) and correlations (phenotypic above diagonal, genetic below diagonal) among the traits averaged over two ages (200d and 300d), calculated from the G5 (the base population) data, and from data combined over the three selected generations (G6, G7, G8) in the Blue and Green groups.
| Traits | Data | L*200&300 | a*200&300 | b*200&300 | SCI200&300 | AveObs200&300 |
|---|---|---|---|---|---|---|
| L*200&300 | Base population (G5) | 0.133 | 0.055 | 0.020 | 0.011 | 0.034 |
| Blue (G6,G7,G8) | 0.248 | 0.068 | 0.055 | 0.028 | 0.057 | |
| Green (G6,G7,G8) | 0.188 | 0.082 | 0.040 | 0.022 | 0.036 | |
| a*200&300 | Base population (G5) | 0.131 | 0.144 | 0.035 | 0.035 | 0.053 |
| Blue (G6,G7,G8) | 0.319 | 0.211 | 0.092 | 0.092 | 0.078 | |
| Green (G6,G7,G8) | 0.265 | 0.178 | 0.048 | 0.049 | 0.063 | |
| b*200&300 | Base population (G5) | 0.042 | 0.065 | 0.130 | 0.003 | 0.035 |
| Blue (G6,G7,G8) | 0.311 | 0.691 | 0.214 | 0.010 | 0.061 | |
| Green (G6,G7,G8) | 0.146 | 0.134 | 0.173 | 0.007 | 0.023 | |
| SCI200&300 | Base population (G5) | 0.023 | 0.071 | 0.005 | 0.132 | 0.033 |
| Blue (G6,G7,G8) | 0.121 | 0.603 | 0.074 | 0.217 | 0.057 | |
| Green (G6,G7,G8) | 0.081 | 0.143 | 0.024 | 0.176 | 0.022 | |
| AveObs200&300 | Base population (G5) | 0.047 | 0.111 | 0.053 | 0.043 | 0.144 |
| Blue (G6,G7,G8) | 0.903 | 0.664 | 0.858 | 0.956 | 0.202 | |
| Green (G6,G7,G8) | 0.121 | 0.167 | 0.203 | 0.036 | 0.184 |
In the Green group, the heritability estimates were almost as high (and their SE as low, Table 5) as those in the base population (G5), hence they were similarly highly significant (Table 4). The high h2 estimates explain the substantial responses to the second and third cycles of selection (i.e., from G6 to G8) in the Green group (Fig. 1, Table 3). In contrast, most estimates of heritability in the Blue group were much lower (a*=0.173, b*=0.284, SCI=0.296, AveObs=0.099) than the corresponding ones in G5 and in the Green group (Table 4). Due to these low heritabilities, there were no responses to the second and third cycles of selection in the Blue group (Fig. 1, Table 3).
The phenotypic and genetic correlations (rP and rG) among the 5 color parameters are presented above and below the diagonal of Table 4, respectively, with their SE shown in the corresponding cells of Table 5. Almost all rP and rG were high and statistically significant, except in the Blue group, where the high estimates of rG were not statistically significant due to high values of SE, reflecting the small sample size (limited number of hens and families) in the Blue group. The very high genetic correlations among the 4 colorimeter parameters (L*, a*, b*, and SCI) in the base population (G5) were aligned with the combined selection criteria for the Blue and Green groups (Table 4). Most importantly, consistent high correlations were observed in both color groups: positive between L* and SCI, and negative between L* and b* as well as between b* and SCI (Table 4). Together with the high heritabilities, these correlations contributed to the significant responses to the combined selection criteria and the overall genetic changes in eggshell color from G5 to G8.
Distribution of color parameters in the base population and the blue and green groups
The distribution plots of all base-population hens (G5) and of all the Blue and Green hens in G6 and in G8, and their phenotypic variances (Vp), are presented in Fig. 2. Compared to the base population, the range and Vp of the 5 color parameters in the Blue group were much reduced (Fig. 2), in line with the significant responses to the first cycle of selection for blue eggshell (Fig. 1, Table 3). In spite of the minimal changes in the means of the Blue group in G8 compared to G6 (Table 3), the range and Vp of the 5 color parameters in the Blue group were further reduced in G8 (Fig. 2). In contrast, the range and Vp of the 4 colorimeter parameters (L*, a*, b*, and SCI) in the Green group in G6 were only slightly reduced compared to the base population (G5) and were similar or only slightly reduced in G8. Accordingly, the difference between the Vp in the Blue versus Green groups was significant in the 4 colorimeter parameters in G6 and in G8 (Fig. 2). Only in the average visual color score (AveObs), the range and Vp in the Green group were reduced from G5 (0.54) to G6 (0.32), and further reduced to G8 (0.17), but AveObs in the Blue group was much more uniform, with Vp=0.23 in G6, and Vp=0.09 in G8 (Fig. 2). However, due to the lower accuracy of visual classification (compared to the Minolta measurements), the difference between the Vp in the Blue versus Green groups was not statistically significant in the AveObs in G6 and in G8.
Comparing the range of the phenotypic values of the 5 color parameters in G5 to the corresponding range in the Blue group in G6, clearly shows that the first cycle of selection eliminated low values of L*, high values of a*, high values of b*, and low values of SCI (Fig. 2). Consequently, whereas AveObs values in G5 (the base population) ranged from 1 (Light) to 4 (Olive), AveObs values in the Blue group were mostly between 1.5 (intermediate between Light and Blue) and 3.0 (Green) in G6. In G8, the means of the Blue groups hardly changed compared to G6 (Fig. 1, Table 3), but the second+third cycles of selection further reduced the Blue group variation in the colorimeter parameters (in line with their reduced heritabilities). Consequently, in G8 the range of AveObs values in the Blue group was further reduced with most individuals between 2.0 (Blue) and 2.5 (intermediate between Blue and Green).
In the Green group, selection from G5 to G6, and from G6 to G8, eliminated only the very extreme high values of L*, a*, and SCI, and extreme low values of b*, and consequently the Green groups were far less uniform than the Blue groups (Fig. 2). In line with the minimal reduction in range and Vp in the Green group, most AveObs values in G6 ranged between 2 (Blue) and 3.5 (Green/Olive), not much narrower than the full range from 1 (Light) to 4 (Olive) in G5. In G8, the means of the Green groups changed significantly compared to G6 (Fig. 1, Table 3), and the second+third cycles of selection further reduced the Green group variation in the colorimeter parameters (in line with their significant heritabilities). The range of AveObs values in the Green group in G8 was further reduced, with most individuals ranging between 2.5 (intermediate between Blue and Green) and 3.5 (intermediate between Green and Olive).
Expected further responses to selection in the blue and green groups
The Blue and Green groups differ substantially in their selection criteria, in the response to the 3 cycles of selection, and consequently in the expected further response to additional selection. The selection criteria for blue eggshell color were unidirectional: selecting for high values of L* and SCI, and for low values of a* and b*. In contrast, 3 criteria of selection for green eggshell color were intermediate values: L* around 80, b* around 12, and SCI around 75. Accordingly, the selection differentials of the G5 dams (SelDif (G5)) in the Blue group were larger than the corresponding SelDif (G5) of the Green dams, and consequently, only in the Blue group did the means of the 5 color parameters in G6 indicate significant responses (Table 3, Fig. 1).
The first cycle of selection for blue eggshell also significantly reduced the phenotypic variance (Vp) in all five color parameters, compared to the phenotypic variances in the base population (Fig. 2). These significant reductions in Vp for all color parameters, combined with the lower heritability estimates (compared to the base population) in the selected generations of the Blue group (Table 4), indicate a substantial reduction in the genetic variance (Vg) in this group. Thus, in line with the reduced Vp and lower SelDif (G6+G7) values (compared to SelDif (G5) values) in the Blue group (Table 3), and the reduced heritability, the two additional selection cycles (from G6 to G8) did not change the means of the color parameters (Table 3, Fig. 1) but further reduced Vp (Fig. 2). However, the four colorimeter parameters (L*, a*, b*, and SCI) were merely the selection criteria, whereas the breeding objective was to establish a genetically uniform line where all hens would lay blue eggs (i.e. scored AveObs=2). The AveObs scores of individual Blue hens in G8 ranged from 1.75 to 2.75 (Fig. 2) indicating that light-egg layers were eliminated, but some Blue-group hens laid eggs between blue (2.00) and green (3.00). Fig. 3 shows the bivariate association between AveObs values of the Blue hens in G8, plotted versus the values of the four colorimeter parameters of the same eggs, with the correlations (r values) and significances (superscripted asterisks) shown at the upper right corner. Based on the bubble plots of the Blue group in Fig. 3, it appears that further selection against the greenish shades can be achieved by culling hens with low L*, high b* and low SCI (Fig. 3). The horizontal blue dotted lines in Fig. 3 indicate the mean value of each colorimeter parameter that matches blue eggshell (AveObs=2). These values, higher than the G8 means of L* and SCI, lower than the G8 mean of b*, and equal to the G8 mean of a*, were drawn also in Fig. 1, to indicate the expected breeding target value of each parameter in the Blue group.
Fig. 3.
Bubble plots of the association between AveObs and the four shell colorimeter parameters in eggs of G8 hens from the Blue group (
) and the Green group (
). The size of each bubble represents the number of individual hens with each AveObs value. The r values are the correlation coefficients, with asterisks indicating their significance (* <0.05, ⁎⁎⁎ <0.001). The blue and green vertical dotted lines indicate the desired AveObs of Blue (2) and Green (3), and the horizontal dotted lines indicate the values of the colorimeter parameters that match the AveObs target of each color group.
The target value of L* is slightly above 84, about one unit higher than the G8 mean (82.84, Table 3), and it can be achieved by eliminating the hens with greenish shade (AveObs 2.5 and 2.75, Fig. 3). The target value of SCI is close to 86, higher than the G8 mean (82.44, Table 3), and — as for L*, it can be achieved by eliminating the hens with greenish shade (AveObs 2.5 and 2.75, Fig. 3). The target value of b* is slightly below 6, quite lower than the G8 mean (7.67, Table 3), and it can be achieved by eliminating the few hens with light blue (AveObs 1.5 and 1.75), and the hens with greenish shade (AveObs 2.5 and 2.75, Fig. 3). In these three colorimeter parameters, the selection criteria suggested by Wang et al. (2023a) appear to be valid also in G8 — highest L* and SCI, and lowest b*. The target value of a* is about –7.4, similar to the G8 mean (–7.27, Table 3), and therefore a* should not be considered for further selection in the Blue group. The moderate heritabilities of L*, b* and SCI in the selected Blue group, ranging from 0.284 to 0.296 (b* and SCI) to 0.415 (L*) (Table 4) suggest the feasibility of stabilizing a uniform line of blue-egg layers. This outcome could be expected, because the blue eggshell color is theoretically determined by the level of a single pigment, biliverdin, which appeared to be controlled by a major gene (Oocyane, O) (Wang et al., 2013; Wragg et al., 2013). Although previous studies reported simultaneous presence of high biliverdin and low protoporphyrin in blue eggshells (Wang et al., 2009; Zeng et al., 2022, 2023b), they tested blue-greenish eggs similar to those in the base population of this study, not pure blue ones.
In the Green group, the responses to the selection reflected the selection criteria: L* around 80, b* around 12, SCI around 75 and low a*. Accordingly, after 3 cycles of selection, the mean L* decreased from 81.92 (in G5) to 78.81 (in G8), the mean b* increased from 9.42 to 11.89 (almost 12!), the mean SCI decreased from 78.72 to 73.27, and the mean a* was only slightly reduced from –6.22 to –6.35 (Table 3, Fig. 1). Fig. 3 shows the bivariate association and the correlations (r values) between AveObs values of the Green hens in G8, and the values of the four colorimeter parameters of the same eggs. Based on the bubble plots showing highly significant correlations (Fig. 3) for the Green group, it appears that further selection towards the same or similar selection criteria may eliminate the Green-group hens that lay eggs with blue shell (AveObs<3) or olive shell (AveObs>3), and consequently improve the uniformity of the green eggshell color in this group. The horizontal green dotted lines in Fig. 3 indicate the mean value of each colorimeter parameter that matches green eggshell (AveObs=3). These values were also drawn in Fig. 1, to indicate the suggested breeding target value of each parameter in the Green group. These targets are very similar to the selection criteria that were used in the selection of the Green group (L* about 80, b* about 12, SCI about 75 and low a*) and are quite similar to the Green group means in G8. Therefore, further selection in the Green group should aim only at improving its uniformity. Based on the highly significant correlations between AveObs and the colorimeter parameters in the Green group in G8 (Fig. 3), and the high heritability of these parameters (Table 4), it is expected that by adhering to the selection targets for green shells (L*≈78, a*≈–7, b*≈12 and SCI≈73; Fig. 3), the uniformity of eggshell color in the Green group may further improve.
However, the blue-green eggshell colors between blue (AveObs=2) and olive (AveObs=4) are determined by combinations of two pigments, biliverdin and protoporphyrin (Wang et al., 2009; Liu et al., 2010; Dalirsefat, 2015; Zeng et al., 2022; Chen et al., 2023; Wang et al., 2023b). Reports on the continuous response to the selection for dark brown eggshell indicate the presence of substantial polygenic variation in the content of protoporphyrin, which determines the level of darkness of brown eggshell, in addition to the major genes like ALAS1 and ABCG2 (Li et al., 2013; Samiullah et al., 2015; Lu et al., 2021; Yang et al., 2022; Zhang et al., 2024). Similarly, there are indications that although the presence of biliverdin is controlled by a single major gene (Oocyane), the exact content level of biliverdin, which determines the intensity of the blue eggshell color from light (AveObs=1) to blue (AveObs=2), is controlled by polygenic variation (Darwish et al., 2019; Wang et al., 2023b). Therefore, it is assumed that the range of shades from blue-greenish eggshell (AveObs>2) to olive (AveObs=4) are determined by numerous combinations of the polygenic variants of the two pigments (Darwish et al., 2019; Chen et al., 2023). Moreover, specific different polygenic combinations may lead to the same color, i.e. parents with colorimeter parameters identical to the selection targets and AveObs=3, may differ in their polygenic genotypes, and therefore their progeny segregate to many polygenic genotypes that code for different eggshell shades with AveObs>2. The genetic variation in all color parameters in the BG line is due to segregation of modifier genes (polygenes) that affect, in each individual, the levels of the two pigments — biliverdin, which is produced by all individuals in the BG line but at varying levels, and protoporphyrin, which is produced only in individuals with greenish eggshell shades, also at varying levels. Therefore, the genetic variance (and consequently the h²) of the color traits in the BG line is due to the segregation of these two polygenic systems — those affecting biliverdin and those affecting protoporphyrin. Following the selection against greenish shades, the majority of chickens in the Blue group produce biliverdin and zero or negligible levels of protoporphyrin. Therefore, the polygenes responsible for variation in the level of protoporphyrin are not expressed in the Blue group, and that is why the genetic variation and heritability are reduced in this group, compared to the Green group.
Summary and conclusions
Three selection cycles with colorimeter parameters (mainly L*, b* and SCI) as criteria, yielded two distinctive groups, Blue where most hens laying blue eggs, and Green where most hens lay green eggs. The Blue group was quite uniform in all eggshell color parameters, reflecting reduction in the phenotypic variance (Vp) and genetic variance (lower heritabilities). Although h2 estimates of the selection criteria are low in the Blue group, but rP and rG among them are high and therefore it is expected that just one additional selection cycle against hens whose all color parameters differ from the breeding targets for blue eggs, will yield a genetic line in which all hens lay blue eggs (AveObs=2). The means of the Green group changed in the desired directions, but the group was less uniform, with Vp and heritabilities only slightly lower than in the base population. The lack of uniformity is possibly due to genetic architecture underlying the green eggshell shades, consisted of combinations of polygenic genotypes controlling the two pigments (i.e., biliverdin and protoporphyrin) that determine the wide range of colors, from blue-greenish to olive. However, with high h2 of the four colorimeter parameters and the high rP and rG between them, it can be expected that a few additional selection cycles will lead to a "true breeding" genetic line where all hens lay green eggs (AveObs=3).
CRediT authorship contribution statement
Huanhuan Wang: Writing – original draft, Project administration, Funding acquisition. Avigdor Cahaner: Writing – review & editing. Lei Zhang: Data curation. Ying Ge: Formal analysis. Hang Liu: Visualization. Yinghui Wei: Investigation. Qinghai Li: Formal analysis. Xuedong Zhang: Writing – original draft, Validation, Supervision, Resources.
Disclosures
None. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in the present study.
Acknowledgments
This work was supported by the project of Agricultural New Breeds from Zhejiang Science and Technology Agency [grant number: 2021C02068-9] and the ‘115′ project of Foreign Intelligence Import from Hangzhou Science and Technology Bureau [grant number: 20230450, 2023]
Footnotes
Scientific section for the paper: Genetics and Breeding
References
- Beranova K., Jozova E., Anderle V., Rost M., Zita L., Beran P., Landova H.T., Curn V., Tyller M. Molecular tool for efficient breeding of DOMINANT Greenshell laying hens and significant refinement of phenotypic selection focused on eggshell color. Poult. Sci. 2024;103 doi: 10.1016/j.psj.2024.104425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bi H.J., Liu Z.., Sun C.J., Li G.Q., Wu G.Q., Shi F.Y., Liu A.Q., Yang N. Brown eggshell fading with layer ageing: dynamic change in the content of protoporphyrin IX. Poult. Sci. 2018;97:1948–1953. doi: 10.3382/ps/pey044. [DOI] [PubMed] [Google Scholar]
- Cavero D., Schmutz M., Icken W., Preisinger R. Attractive eggshell color as a breeding goal. Lohmann Inform. 2012;47:15–21. https://d-nb.info/1029309485/34 [Google Scholar]
- Chen Q., Wang Z.P. A new molecular mechanism supports that blue-greenish egg color evolved independently across chicken breeds. Poult. Sci. 2022;101 doi: 10.1016/j.psj.2022.102223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen Q., Huang J.J., Wang Z.P. Establishment of quantization method and genetic basis analysis of brown eggshell color in the Lueyang black-boned chicken. Scientia Agricultura Sinica. 2023;56(17):3452–3460. doi: 10.3864/j.issn.0578-1752.2023.17.017. Chinese with English abstract. [DOI] [Google Scholar]
- Dalirsefat S.B. China Agricultural University; Beijing: 2015. Analysis of Eggshell Pigmentation in the Chinese Indigenous Blue-shelled Chickens. Docter thesisChinese with English abstract. [Google Scholar]
- Darwish H.Y.A., Dalirsefat S.B., Dong X.G., Hua G.Y., Chen J.F., Zhang Y.Y., Li J.X., Xu J.S., Li J.Y., Deng X.M., Wu C.X. Genome-wide association study and a post replication analysis revealed a promising genomic region and candidate genes for chicken eggshell blueness. PLoS One. 2019;14(1) doi: 10.1371/journal.pone.0209181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dierks, C., N. T. Ha, H. Simianer, D. Cavero, M. Schmutz, B. Andersson, R. Preisinger, and S. Weigend. 2019. Introgression of blue eggshell color from a gene bank collection into a White Leghorn breeding line. 70th EAAP Annual Meeting, Ghent, Belgium. https://docs.eaap.org/2019/theater/33/1200_Dierks.pdf.
- Guo J., Wang K.H., Qu L., Dou T.C., Ma M., Shen M.M., Hu Y.P. Genetic evaluation of eggshell color based on additive and dominance models in laying hens. Asian-Australas J. Anim. Sci. 2020;33(8):1217–1223. doi: 10.5713/ajas.19.0345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Guyonnet V. Does eggshell color really matter? WATT Poult. Intl. 2022;61:22–23. https://www.wattagnet.com/egg/egg-production/article/15534985/does-eggshell-color-really-matter [Google Scholar]
- Li G.Q., Chen S..R., Duan Z.Y., Qu L.J., Xu G.Y., Yang N. Comparison of protoporphyrin IX content and related gene expression in the tissues of chickens laying brown-shelled eggs. Poult. Sci. 2013;92:3120–3124. doi: 10.3382/ps.2013-03484. [DOI] [PubMed] [Google Scholar]
- Liu H.C., Hsiao M..C., Hu Y.H., Lee S.R., Cheng W.T.K. Eggshell pigmentation study in blue-shelled and white-shelled ducks. Asian-Aust. J. Anim. Sci. 2010;23(2):162–168. doi: 10.5713/ajas.2010.90256. [DOI] [Google Scholar]
- Lu M.Y., Xu L.., Qi G.H., Zhang H.J., Qiu K., Wang J., Wu S.G. Mechanisms associated with the depigmentation of brown eggshells: a review. Poult. Sci. 2021;100 doi: 10.1016/j.psj.2021.101273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lukanov H., Genchev A., Pavlov A. Colour traits of chicken eggs with different eggshell pigmentataion. Trakia J. Sci. 2015;2:149–158. doi: 10.15547/tjs.2015.02.007. [DOI] [Google Scholar]
- Konica Minolta Precise color communication, part I, let’s look at some color spaces. 2007. https://www.konicaminolta.com/instruments/knowledge/color/part1/07.html
- Mulder H.A., Visscher J.., Fablet J. Estimating the purebred-crossbred genetic correlation for uniformity of eggshell color in laying hens. Genet. Sel. Evol. 2016;48(1):39. doi: 10.1186/s12711-016-0212-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ryter S.W., Tyrrell R.M. The heme synthesis and degradation pathways: role in oxidant sensitivity: heme oxygenase has both pro- and antioxidant properties. Free Radic. Biol. Med. 2000;28(2):289–309. doi: 10.1016/S0891-5849(99)00223-3. [DOI] [PubMed] [Google Scholar]
- Sachar M., Anderson K.E., Ma X.C. Protoporphyrin IX: the good, the bad, and the ugly. J. Pharmacol. Exp. Ther. 2016;356:267–275. doi: 10.1124/jpet.115.228130. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Samiullah S., Roberts J.R., Chousalkar K. Eggshell color in brown-egg laying hens - a review. Poult. Sci. 2015;94:2566–2575. doi: 10.3382/ps/pev202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang X.T., Zhao C..J., Li J.Y., Xu G.Y., Lian L.S., Wu C.X., Deng X.M. Comparison of the total amount of eggshell pigments in Dongxiang brown-shelled eggs and Dongxiang blue-shelled eggs. Poult. Sci. 2009;88:1735–1739. doi: 10.3382/ps.2008-00434. [DOI] [PubMed] [Google Scholar]
- Wang Z.P., Qu L..J., Yao J.F., Yang X.L., Li G.Q., Zhang Y.Y., Li J.Y., Wang X.T., Bai J.R., Xu G.Y., Deng X.M., Yang N., Wu C.X. An EAV-HP insertion in 5’ flanking region of SLCO1B3 causes blue eggshell in the chicken. PLoS Genet. 2013;9(1) doi: 10.1371/journal.pgen.1003183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang H.H., Cahaner A.., Lou L.F., Zhang L., Ge Y., Li Q.H., Zhang X.D. Genetics and breeding of a black-bone and blue eggshell chicken line. 1. Body weight, skin color, and their combined selection. Poult. Sci. 2021;100 doi: 10.1016/j.psj.2021.101035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang H.H., Cahaner A.., Lou L.F., Zhang L., Ge Y., Li Q.H., Zhang X.D. Genetics and breeding of a black-bone and blue eggshell chicken line. 2. Laying pattern and egg production in two consecutive generations. Poult. Sci. 2022;101 doi: 10.1016/j.psj.2021.101679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang H.H., Cahaner A.., Lou L.F., Zhang L., Ge Y., Li Q.H., Pan Y.C., Zhang X.D. Genetics and breeding of a black-bone and blue eggshell chicken line. 3. Visual eggshell color and colorimeter parameters in 3 consecutive generations. Poult. Sci. 2023;102 doi: 10.1016/j.psj.2023.103052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang H.H., Ge Y.., Zhang L., Wei Y.H., Li Q.H., Zhang X.D., Pan Y.C. The pigments in eggshell with different colour and the pigment regulatory gene expression in corresponding chicken’s shell gland. Animal. 2023;17 doi: 10.1016/j.animal.2023.100776. [DOI] [PubMed] [Google Scholar]
- Wragg D., Mwacharo J.M., Alcalde J.A., Wang C., Han J.L., Gongora J., Gourichon D., Tixier-Boichard M., Hanotte O. Endogenous retrovirus EAV-HP linked to blue egg phenotype in Mapuche fowl. PLoS One. 2013;8(8) doi: 10.1371/journal.pone.0071393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xu J.M. Foshan University; Foshan: 2018. Breeding Study of Changshun Blue-eggshell Chicken. Master thesisChinese with English abstract. [Google Scholar]
- Yang J., Mao Z.Q., Wang X.Q., Zhuang J.J., Gong S.J., Gao Z.Y., Xu G.Y., Yang N., Sun C.J. Identification of crucial genes and metabolites regulating the eggshell brownness in chicken. BMC Genom. 2022;23:761. doi: 10.1186/s12864-022-08987-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zeng L.S., Xu G..Y., Jiang C.Y., Li J.Y., Zheng J.X. Research note: l*a*b* color space for prediction of eggshell pigment content in differently colored eggs. Poult. Sci. 2022;101 doi: 10.1016/j.psj.2022.101942. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang D.Y., Zhou W..C., Li J.L., Wang Z.P. A non-synonymous variant of C21R in the ABCG2 gene is significantly associated with brown eggshell color of Lueyang black-boned chicken. Hereditas (Beijing) 2024;46(12):1066–1075. doi: 10.16288/j.yczz.24-225. Chinese with English abstract. [DOI] [PubMed] [Google Scholar]
- Zhao R., Xu G.Y., Liu Z.Z., Li J.Y., Yang N. A study on eggshell pigmentation: biliverdin in blue-shelled chickens. Poult. Sci. 2006;85:546–549. doi: 10.1093/ps/85.3.546. [DOI] [PubMed] [Google Scholar]