Statistics and correlations for the 10 nut quantitative traits variables
The 10 nut quantitative traits of 112 C. illinoinensis germplasm were determined and the descriptive statistics of means, mean standard error, variance, range, maxima, minima, standard deviations and coefficients of variation (CV) are analyzed in this study (Table 2, Supplementary Table S2). The results showed that some nut traits had high CVs, representing extensive morphological variability. These included single nuts kernel mass (32.75%), single nut mass (28.80%), shell thinness (25.03%). The remaining traits showed comparatively low CV values (<20%). Oil yield and ratio of transverse longitudinal had the lowest CVs of 5.38% and 2.91%, respectively. The ANOVA showed that accessions effects in pecan were significant on all nut quantitative traits.
Table 2
Analysis of diversity for 10 nut quantitative traits in 112 pecan germplasm.
Trait | Mean | Mean SE | Sd. | Variance | CV% | Range | Max. | Min. |
Single nut mass (g) | 6.14 | 0.18 | 1.77 | 3.16 | 28.80 | 8.68 | 10.62 | 1.94 |
Nut transverse diameter (mm) | 21.67 | 0.23 | 2.43 | 5.81 | 11.19 | 12.35 | 26.37 | 14.02 |
Nut longitudinal diameter (mm) | 38.50 | 0.58 | 6.23 | 38.28 | 16.18 | 31.98 | 54.38 | 22.40 |
Nut lateral diameter (mm) | 20.59 | 0.22 | 2.38 | 5.64 | 11.54 | 12.12 | 25.86 | 13.74 |
Nut aspect ratio (Nut shape index) | 1.78 | 0.02 | 0.26 | 0.07 | 14.51 | 1.34 | 2.43 | 1.09 |
Ratio of transverse longitudinal | 1.05 | 0.03 | 0.03 | 0.00 | 2.91 | 0.13 | 1.13 | 1.00 |
Single nuts kernel mass (g) | 3.16 | 0.10 | 1.03 | 1.08 | 32.75 | 5.22 | 6.28 | 1.06 |
Kernel yield (%) | 51.22 | 0.54 | 5.70 | 32.60 | 11.13 | 31.72 | 62.85 | 31.13 |
Shell thinness (mm) | 0.78 | 0.02 | 0.20 | 0.04 | 25.03 | 1.36 | 1.77 | 0.41 |
Oil yield (%) | 65.92 | 0.32 | 3.54 | 11.41 | 5.38 | 15.19 | 73.82 | 55.36 |
EI mart had the highest single nut mass (10.62 g), followed by Choctaw (10.28 g), SD07 (9.82 g) and Mohawk (9.61g). The lightest single nut mass was SD09 (1.94g), followed by SD38 (2.25 g) and SD08 (2.43 g). The maximum nut transverse diameter was SD07 (26.37 mm), followed by Jackson (26.29 mm), Choctaw (25.56 mm) and Success (25.43 mm). Maximum nut longitudinal diameter was observed in Mahan (54.38 mm), followed by Nacono (50.38 mm), EI mart (50.11 mm) and Western (49.77 mm). SD09 had the minimum nut longitudinal diameter (22.40 mm). Maximum nut lateral diameter was observed in SD07 (25.86 mm), followed by Jackson (25.21 mm), Tejas (25.08 mm) and Choctaw (24.62 mm). SD38 had the minimum nut longitudinal diameter (13.74 mm). Nut aspect ratio (Nut shape index) indicated the shape of the nuts. The maximum of nut aspect ratio was Sauber (2.43), followed by Mahan (2.32) and Shawnne (2.30 mm). SD04 had the minimum nut aspect ratio of 1.09. Cape Fear had the thinnest shell thinness (0.41 mm), followed by SD02 (0.45 mm) and SD01 (0.47 mm). The thickest shell thinness was recorded in SD04 (1.77 mm), followed by SD13 (1.38 mm) and SD47(1.25 mm). The highest kernel yield was Sioux (62.85%), followed by SD02 (62.25%) and Mahan (61.36%), and the lowest kernel yield was SD04 (31.13%). The highest oil yield was SD05 (73.82%), followed by Wichita (72.44%) and Creek (72.03%), and the lowest oil yield was SD41 (55.36%).
Statistically significant correlations among all the 10 nut traits were analyzed (Table 3). The positive correlations were observed between each pair of single nut mass, nut transverse diameter, nut longitudinal diameter and nut lateral diameter. Nut shape index is significantly positively correlated with nut longitudinal diameter, but negatively correlated with nut transverse diameter and nut lateral diameter. Single nuts kernel mass is significantly positively correlated with single nut mass, nut transverse diameter, nut longitudinal diameter and nut lateral diameter. There is a very significant negative correlation between shell thinness and kernel yield (-0.793). Oil yield is significantly positively correlated with single nut mass, single nuts kernel mass and kernel yield.
Table 3
Correlation coefficients among 10 nut quantitative traits in 112 pecan germplasm.
| SNM | NTD | NLOD | NLAD | NSI | RTO | SNKM | KY | ST | OY |
SNM | 1 | | | | | | | | | |
NTD | 0.792** | 1 | | | | | | | | |
NLOD | 0.706** | 0.485** | 1 | | | | | | | |
NLAD | 0.807** | 0.962** | 0.460** | 1 | | | | | | |
NSI | 0.126 | -0.273** | 0.701** | -0.272** | 1 | | | | | |
RTO | -0.107 | 0.045 | 0.048 | -0.214* | -0.005 | 1 | | | | |
SNKM | 0.937** | 0.700** | 0.679** | 0.703** | 0.160 | -0.043 | 1 | | | |
KY | 0.083 | -0.038 | 0.061 | -0.072 | 0.070 | 0.175 | 0.414** | 1 | | |
ST | 0.190 | 0.146 | 0.065 | 0.214* | -0.031 | -0.257* | -0.103 | -0.793** | 1 | |
OY | 0.386** | 0.135 | 0.181 | 0.187 | 0.084 | -0.176 | 0.418** | 0.262** | -0.032 | 1 |
Abbreviation: SNM, Single nut mass; NTD, Nut transverse diameter; NLOD, Nut longitudinal diameter; NLAD, Nut lateral diameter; NSI, Nut shape index; RTO, Ratio of transverse longitudinal; SNKM, Single nuts kernel mass; KY, Kernel yield; ST, Shell thinness; OY, Oil yield. ** indicated that correlation is significant at the 0.01 level; *indicated that correlation is significant at the 0.05 level. |
Principal component analysis (PCA) of 10 nut quantitative traits variables
10 pecan nut quantitative traits were put into four components that explained 89.078% of the total variation through PCA (Table 4). The first component, which accounted for 41.031% of the total variation, included single nut mass, nut transverse diameter, nut longitudinal diameter, nut lateral diameter and single nuts kernel mass. The second component, accounting for 23.498% of the total variation, included kernel yield and shell thinness. The third component, which explained 13.51% of the total variation, included nut shape index only. The fourth component, explaining 11.039% of the total variation, included ratio of transverse longitudinal and oil yield.
Table 4
First 4 components from the PCA analysis of 10 nut quantitative traits in 112 pecan germplasm.
Traits | F1 | F2 | F3 | F4 |
Single nut mass | 0.967 | -0.095 | 0.058 | -0.012 |
Nut transverse diameter | 0.847 | -0.347 | -0.223 | 0.290 |
Nut longitudinal diameter | 0.772 | 0.334 | 0.506 | 0.126 |
Nut lateral diameter | 0.847 | -0.436 | -0.181 | 0.100 |
Nut shape index | 0.206 | 0.611 | 0.756 | -0.086 |
Ratio of transverse longitudinal | -0.088 | 0.396 | -0.170 | 0.692 |
Single nuts kernel mass | 0.946 | 0.199 | -0.089 | -0.051 |
Kernel yield | 0.225 | 0.805 | -0.454 | -0.123 |
Shell thinness | 0.005 | -0.830 | 0.367 | -0.077 |
Oil yield | 0.375 | 0.108 | -0.245 | -0.696 |
Eigenvalue of correlation matrix | 4.103 | 2.350 | 1.351 | 1.104 |
Explained proportion of total variance % | 41.031 | 23.498 | 13.510 | 11.039 |
Cumulative proportion of total variance % | 41.031 | 64.529 | 78.039 | 89.078 |
Because the first two PCs explained 64.5% of the total variation among the cultivars (Table 4), the approximation of the real multivariate diversity of the cultivars on the 2-PC axis is quite acceptable for the most important discriminating (contributing) traits. The scatter plot of the first two principal components (Fig. 1) shows geometrical distances among the cultivars in the plot that reflect a genetic similarity for the 10 measured nut traits and the relations between the four groups delivered by the cluster analysis. According to phenotypic resemblance and morphological characteristics, the samples were plot grouped. For example, accessions SD08, SD09, SD38, SD50 and Kanza had the smallest nut transverse diameter and lateral diameter, minimum single nut mass and nuts kernel mass. SD58, SD59 and SD60 had the smaller single nut mass, nuts kernel mass and lateral diameter. SD04, SD13, SD14 and Zhongshan 25 had the thickest shell thinness (> 1.0mm) and minimum kernel yield. SD07, Choctaw, Jackson, Kiowa, Success and Waco had the biggest nut lateral diameter and nut transverse diameter, bigger nut fruit weight, higher single nuts kernel mass and medium shell thinness. Mahan, Wichta, Melrose, Nacono, Pownee, Western shelly, Desirable, Mohawk and EI mart had largest nut longitudinal diameter and single nuts kernel mass, biggest nut weight, thinner shell thinness. These results demonstrate that single nut mass, nut transverse diameter, nut longitudinal diameter, nut lateral diameter, single nuts kernel mass, kernel yield and shell thinness are highly positively correlated and as a result, these morphological traits led to the highest loading factors in this PCA analysis. The distribution of the cultivars in Fig. 1 showed that their rich genetic diversity (profiles) for the studied traits.
Fatty acid components and content analysis, statistics and correlations for the 15 fatty acids variables
36 fatty acids (Supplementary Table S1) were measured in the oil of 112 pecan germplasm, and 15 fatty acids were detected (Table 5, Supplementary Table S3). The highest fatty acid is octadecenoic acid (252.485 mg/g), followed by octadecadienoic acid (133.952 mg/g), hexadecanoic acid (26.393mg/g), and octadecatrienoic acid (7.233 mg/g), decanoic acid (0.014 mg/g) and octadecanoic acid (0.014 mg/g) were the lowest fatty acid. Ratio of UFA was up to 93.55%. Unsaturated fat is broken down into two subgroups known as mono- and polyunsaturated. These are the healthy forms attributed to cholesterol reduction and heart health. Ratio of mono- and polyunsaturated fatty acid were 64.54% and 35.46% respectively. The maximum hexadecanoic acid was detected in Pawnee (50.579 mg/g), followed by SD49 (42.780 mg/g), SD39 (38.701 mg/g) and SD64 (38.651 mg/g). The lightest hexadecanoic acid was detected in SD45 (0.408 mg/g). SD49 had the highest octadecenoic acid (401.802 mg/g), followed by Pawnee (379.83 mg/g) and Oconee (323.365 mg/g). SD49 had the highest octadecadienoic acid (244.335 mg/g) and octadecatrienoic acid (17.742 mg/g), followed by SD42 (231.960 and 12.685 mg/g, respectively) and SD64 (227.208 and 12.509 mg/g, respectively). SD49 had the highest Eicosenoic acid (3.477 mg/g) and tricosanoic acid (0.416 mg/g). SD49 had the highest total of fatty acids (713.911 mg/g) and total of unsaturated fatty acids (668.853 mg/g), followed by Pawnee (620.707 mg/g and 568.167 mg/g) and SD64 (587.681 mg/g and 547.813 mg/g). Ratio of unsaturated fatty acids in all Carya illinoinensis germplasm were more than 91%, and the highest was SD45 (99.64%). The maximum of mono- (406.776 mg/g) and polyunsaturated fatty acids (262.077 mg/g) were in SD49, followed by Pawnee (384.170 mg/g) and Oconee (326.585 mg/g) for monounsaturated fatty acids, and by SD42 (244.645 mg/g) and SD64 (239.717 mg/g) for polyunsaturated fatty acids. The maximum of monounsaturated fatty acid ratio was Forkert (80.88%), followed by Sauber (78.24%) and SD04 (77.74%).
Table 5
Analysis of diversity for 15 fatty acids contents in 112 pecan germplasm.
Symbol | Fatty acids | Mean | Mean SE | Sd. | Variance | CV | Range | Max. | Min. |
C10:0 | Decanoic acid (mg/g) | 0.014 | 0.000 | 0.004 | 0.000 | 31.80 | 0.03 | 0.037 | 0.006 |
C12:0 | Dodecanoic acid(mg/g) | 0.159 | 0.000 | 0.005 | 0.000 | 2.93 | 0.04 | 0.188 | 0.151 |
C14:0 | Tetradecanoic acid (mg/g) | 0.077 | 0.004 | 0.044 | 0.002 | 57.50 | 0.26 | 0.260 | 0 |
C15:0 | Pentadecanoic acid (mg/g) | 0.096 | 0.002 | 0.024 | 0.001 | 24.58 | 0.11 | 0.167 | 0.055 |
C16:0 | Hexadecanoic acid (mg/g) | 26.393 | 0.537 | 5.660 | 32.268 | 21.45 | 50.17 | 50.579 | 0.408 |
C16:1 | Hexadecenoic acid (mg/g) | 0.450 | 0.014 | 0.151 | 0.023 | 33.64 | 0.66 | 0.852 | 0.188 |
C17:0 | Heptadecanoic acid (mg/g) | 0.135 | 0.005 | 0.051 | 0.003 | 37.47 | 0.34 | 0.344 | 0 |
C17:1 | Heptadecenoic acid (mg/g) | 0.361 | 0.007 | 0.078 | 0.006 | 21.57 | 0.57 | 0.758 | 0.184 |
C18:0 | Octadecanoic acid (mg/g) | 0.014 | 0.004 | 0.004 | 0.000 | 27.34 | 0.03 | 0.032 | 0.006 |
C18:1 | Octadecenoic acid (mg/g) | 252.485 | 3.723 | 39.118 | 1557.053 | 15.49 | 236.74 | 401.801 | 165.063 |
C18:2 | Octadecadienoic acid (mg/g) | 133.952 | 3.151 | 33.336 | 1112.203 | 24.89 | 185.70 | 244.335 | 58.635 |
C18:3 | Octadecatrienoic acid (mg/g) | 7.233 | 0.215 | 2.290 | 5.173 | 31.65 | 15.61 | 17.742 | 2.130 |
C20:0 | Eicosanoic acid (mg/g) | 0.351 | 0.013 | 0.132 | 0.018 | 37.67 | 1.00 | 0.999 | 0 |
C20:1 | Eicosenoic acid (mg/g) | 1.623 | 0.041 | 0.434 | 0.192 | 26.75 | 2.80 | 3.477 | 0.675 |
C23:0 | Tricosanoic acid (mg/g) | 0.171 | 0.035 | 0.037 | 0.001 | 21.63 | 0.31 | 0.416 | 0.102 |
Total of fatty acids (mg/g) | 423.514 | 5.964 | 62.747 | 3984.974 | 14.82 | 462.66 | 713.911 | 251.251 |
Total of NFA (mg/g) | 396.105 | 5.525 | 58.104 | 3419.001 | 14.67 | 432.97 | 668.853 | 235.884 |
Ratio of NFA (%) | 93.55 | 0.072 | 0.76 | 0.576 | 0.81 | 8.10 | 99.64 | 91.54 |
Total of MNFA (mg/g) | 254.919 | 3.779 | 39.646 | 1599.351 | 15.55 | 240.54 | 406.776 | 166.234 |
Ratio of MNFA (%) | 64.54 | 0.574 | 6.09 | 36.913 | 9.43 | 33.22 | 80.88 | 47.67 |
Total of PNFA (mg/g) | 141.185 | 3.337 | 35.318 | 1247.150 | 25.02 | 201.31 | 262.077 | 60.765 |
Ratio of PNFA (%) | 35.46 | 0.574 | 6.09 | 36.913 | 17.17 | 33.22 | 52.33 | 19.12 |
The ANOVA showed that accessions effects in pecan were significant on all detected 15 fatty acids components. The descriptive statistics of means, mean standard error, variance, range, maxima, minima, standard deviations and CV of 15 fatty acids are also analyzed (Table 5). The results showed that some fatty acids had high CVs, representing extensive fatty acid component variability. These included eicosanoic acid (37.67%), heptadecanoic acid (37.47%), hexadecenoic acid (33.64%), decanoic acid (31.8%) and octadecatrienoic acid (31.65%). The remaining fatty acids showed comparatively low CV values (<30%). Ratio of unsaturated fatty acids had the lowest CV of 0.81%.
Statistically significant correlations among all the 15 fatty acids contents were analyzed (Table 6). The positive correlations were observed between each pair of C12, C14, C15 and C16. The positive correlations were observed between each pair of C18:1, C18:2, C18:3, C20, C20:1 and C23, except for the pair of C18:2 and C23. C17 was positive correlations with other fatty acids components except for the C14, C16:1 and C18. C17:1, C18:1 and C20:1 were positive correlations with other fatty acids components except for the C16:1 and C18. C18:2 was positive correlations with other fatty acids components except for the C10, C16:1, C18 and C23. C18:3 was positive correlations with other fatty acids components except for the C10, C16:1 and C18. However, C16:1 and C18 were not correlations with other fatty acids components.
Table 6
Correlation coefficients among 15 fatty acids contents in 112 pecan germplasm.
| C10 | C12 | C14 | C15 | C16 | C16:1 | C17 | C17:1 | C18 | C18:1 | C18:2 | C18:3 | C20 | C20:1 | C23 |
C10 | 1 | | | | | | | | | | | | | | |
C12 | 0.533** | 1 | | | | | | | | | | | | | |
C14 | -0.114 | 0.510** | 1 | | | | | | | | | | | | |
C15 | -0.122 | 0.235* | 0.691** | 1 | | | | | | | | | | | |
C16 | 0.045 | 0.344** | 0.710** | 0.657** | 1 | | | | | | | | | | |
C16:1 | 0.124 | 0.084 | -0.007 | -0.058 | 0.032 | 1 | | | | | | | | | |
C17 | 0.381** | 0.270** | 0.144 | 0.271** | 0.445** | -0.005 | 1 | | | | | | | | |
C17:1 | 0.412** | 0.653** | 0.516** | 0.456** | 0.604** | 0.114 | 0.453** | 1 | | | | | | | |
C18 | 0.102 | 0.01 | -0.047 | -0.005 | 0.01 | -0.102 | -0.025 | -0.064 | 1 | | | | | | |
C18:1 | 0.391** | 0.578** | 0.426** | 0.288** | 0.547** | 0.137 | 0.499** | 0.787** | -0.082 | 1 | | | | | |
C18:2 | -0.105 | 0.289** | 0.611** | 0.769** | 0.659** | -0.012 | 0.225* | 0.508** | -0.12 | 0.190* | 1 | | | | |
C18:3 | 0.001 | 0.299** | 0.450** | 0.682** | 0.577** | -0.049 | 0.420** | 0.473** | -0.137 | 0.229* | 0.857** | 1 | | | |
C20 | 0.465** | 0.338** | 0.065 | 0.239* | 0.303** | -0.068 | 0.707** | 0.345** | 0.014 | 0.452** | 0.215* | 0.482** | 1 | | |
C20:1 | 0.312** | 0.563** | 0.507** | 0.413** | 0.567** | 0.085 | 0.525** | 0.746** | -0.098 | 0.894** | 0.350** | 0.366** | 0.462** | 1 | |
C23 | 0.569** | 0.385** | 0.083 | 0.162 | 0.327** | 0.078 | 0.353** | 0.565** | 0.153 | 0.587** | 0.161 | 0.227* | 0.456** | 0.524** | 1 |
Principal component analysis (PCA) for the fatty acids variable
PCA put the 15 fatty acids, total of fatty acids, total and ratio of UFA and total and ratio of MUFA and PUFA into 5 components which explained 86.201% of the total variation (Supplementary Table S5). The first component included C12, C14, C15, C16, C16:1, C17:1, C18:1, C18:2, C18:3, C20:1, total of fatty acids, total of UFA, total of MUFA, total of PUFA, accounting for 44.595% of the total variation. The second component included C10, ratio of MUFA and PUFA, accounted for 22.123% of the total variation. The third, fourth and fifth component included C20, ratio of UFA and C18, respectively, accounting for 7.757%, 6.888% and 5.024% of the total variation, respectively.
A 2D PCA plot was made according to the first two components (Fig. 2). The samples were plot grouped according to fatty acids content. For example, accessions SD49, Pawnee, SD46, Oconee, SD67, SD37, SD65, SD 68, SD39 and SD64 had the most mount of C14, C15, C16, C16:1, C17:1, C18:1, C18:2, C20:1 and the total of fatty acids, total of UFA and total of MUFA. Accessions SD42, Jinhua, SD36, Wichita, SD38, SD69, and Odom had the most mount of C14, C15, C16, C16:1 and C18:2, the secondly of C17:1, and the total of fatty acids, total of UFA, the less of ratio of MUFA. Accessions SD25, SD13, SD16, SD18, SD04, Elliot and Forkert had minimum of C14, C15, C16, C16:1, C17:1, C18:2, C18:3, and the total of fatty acids, total of UFA and total of PUFA. Accessions SD54, SD40, SD08, Shaoxing and Caddo had the minimum of C10 and C18:1, lesser of C17:1, and the total of fatty acids and total of UFA, and the highest of ratio of PUFA. These results demonstrate that C10, C14, C15, C16, C16:1, C17:1, C18:1, C18:2, C20:1, total of fatty acids, total of UFA, total of MUFA, total of PUFA and ratio of PUFA were highly positively correlated, and as a result, these fatty acids compositions and content led to the highest loading factors in this PCA analysis.
Cluster analysis
Two dendrograms were obtained from the Ward’s method using the euclidean distance based on the 112 accessions nut quantitative traits (Fig. 3A) and fatty acids content (Fig. 3B). Based on the nut quantitative traits, the 112 accessions were grouped in four distinct groups (Fig. 3A). C1 contained 31 accessions which had larger relatively singer nut mass and higher singer nuts kernel mass. C2 contained 12 accessions which had maximum singer nut mass and singer nut kernel mass, the longer nut transverse and lateral diameter. C3 contained 28 accessions which had minor singer nut mass and singer nut kernel mass, shorter nut transverse and longitudinal diameter. C4 contained 41 accessions which had miner singer nut mass and singer nut kernel mass, shorter nut longitudinal diameter, miner nut shape index, and thicker shell thinness.
Based on the 15 fatty acids contents, the 112 accessions were grouped in three major clusters (Fig. 3B). Cluster D1 included 32 accessions which had higher contents of C16, C18:1, C20:1, total of fatty acids, total of UFA and total of MUFA. Cluster D1 included two accessions which had highest contents of C16, C18:1, C20:1, total of fatty acids, total of UFA and total of MUFA. However, cluster D3 included 78 accessions, which had lower contents of C16, C18:1, C20:1, total of fatty acids, total of UFA and total of MUFA.