Principal component analysis and comprehensive evaluation of fruit quality in some advanced selections of Chinese bayberry

Abstract:【Objective】Chinese bayberry (Myrica rubar Sieb. et Zucc.) belonging to the genus Myrica in the family Myricaceae, has been cultivated mainly in southern China. Bayberry fruit is delicious with attractive color and flavor and high economic and medicinal values. Studies have shown that bayberry extracts contain antioxidants that exhibit bioactivities counteracting diabetes, cancer and other health problems. The main factor hindering the industry of bayberry is the shortage of excellent varieties. Therefore, comprehensive evaluation of the fruit quality and accurate evaluation of the advanced selections are the preconditions for breeding of excellent varieties.【Methods】In the years of 2016 and 2017, fruit sampling and quality analysis were carried out on 21 different accessions, including 7 advanced selections and 14 main cultivars. Twenty three fruit quality characters were collected including single fruit weight (FW) , lengthwise diameter (LD) , broadwise diameter (BD) , fruit shape index (FI) , stone weight (SW) , edible ratio (ER) , hardness (HN) , L* (L) , a* (a) , b* (b) , soluble solid (TSS) , total sugar (TS) , fructose (Fru) , glucose (Glu) , sucrose (Suc) , titratable acid (TA) , malic acid (MA) , oxalic acid (OA) , citric acid (CA) , acid-suger ratio (AS) , polyphenol (Pol) , flavone (Fla) and vitamin C (Vc) . We calculated the means and standard deviations of these data in the two years. By the dimensionless treatment of the data of the 23 traits, cluster analysis and principal component analysis were carried out and a comprehensiveevaluation model of fruit quality characters was established.【Results】The range of FW was 8.536-25.273 g; the FI was in the range of 0.919-1.018. LD was in the range of 23.880-35.551 mm, and BD in 25.299-36.043 mm. BD was slightly greater than LD. The range of ER was 91.001%-94.562%, and2010 B2 was the highest. ER had a close relationship with SW (0.635-1.713 g) . HN was distributed between 1.955-4.476 N. The range of TSS of different materials was 9.300%-13.370%, and the TSS content of WD was the highest. TS, Suc, Glu and Fru were in the ranges of 7.579-10.438, 4.010-6.135, 0.915-1.590 and 0.930-1.843 g·100 g^-1, respectively. The variation range of TA was 0.667%-1.586%, and the CA, MA and OA were in the ranges of 7.521-13.469, 0.554-1.287 g·kg^-1 and 16.085-34.304 mg·kg^-1, respectively. The range of AS was 7.469-18.555. The contents of Vc, Fla and Pol were 12.340-81.550 mg·100 g^-1, 0.697-2.943 and 0.830-3.602 mg · g-1, respectively. Following the cluster analysis, the 21 accessions were clustered into 4 groups. Group I consisted of‘Baishuituan' (BST) , 2010 B1, 2010 B3, 2011 B4 and 2010 B2; Group II included‘Biqizhong' (BQ) , 2008 X4 and‘Wandaoyangmei' (WD) ;Group III was formed by‘Anhaibianyingsi' (YS) , ‘Shenghongzhong' (SH) , ‘Dingaomei' (DA) , ‘Zaodamei' (ZDM) , ZX856, ‘Zaojia' (ZJ) , ‘Tongzimei' (TZM) , ‘Anhaizao' (AHZ) , ‘Chise' (CS) , ‘Zaose' (ZS) and‘Zaojimimei' (ZJMM) ; Group IV was composed of‘Dongkui' (WD) and 2008 LD.There were significant differences between accessions. Based on the result of principal component analysis, 6 factors were extracted from the converted data matrix with their cumulative contribution reached88.368%, and the eigenvalue of the 6 factors ranged from 1.360 to 7.052. Fru, Glu, Vc, Fla and Pol had higher values of rotated component than the other traits. The first factor was related to fruit nutrition with variance contribution of 30.663%. The second factor was related to fruit size (FW, LD, BD, SW) with variance contribution of 20.130%. The third factor was related to pericarp texture (AS) , with variance contribution of 16.469%. The fourth factor was related to sugar level (TSS, TS and Suc) , with variance contribution of 8.737%. The fifth factor was acid factor (TA and CA) , with variance contribution of 6.456%. A comprehensive evaluation model for the fruit quality of different materials was constructed: Fn=0.307 F1+ 0.201 F2+ 0.165 F3+ 0.087 F4+ 0.065 F5+ 0.059 F6. The order of the synthetical score of the 21 materials was WD> 2008 LD > 2008 X4 >DK > BQ > ZX856 > DA > ZJMM > ZDM > 2011 B4 >SH > 2010 B1 >YS > 2010 B2 > 2010 B3 > BST > ZJ> TZM > CS> ZS > AHZ. WD, DK, BQ and DA were regarded as the‘four great varieties' in the industry of Chinese bayberry. The fruit quality of the advanced selections, 2008 LD, 2008 X4 and ZX856, were comparable with the‘four great varieties' and could be further evaluated as the target excellent breeding materials.【Conclusion】Based on the principal component analysis and comprehensive evaluation model, three advanced selections (2008 LD, 2008 X4 and ZX856) were identified. The comprehensive evaluation model provides a reliable reference for selections of high-quality breeding materials.