Combining High-throughput Phenotyping and Multivariate Analysis to Assess Fruit Quality Traits in Southern Highbush Blueberry (Vaccinium corymbosum Interspecific Hybrids) Germplasm Collection

Ebrahiem Babiker; Stephen J. Stringer; Hamidou F. Sakhanokho; Patricio Muńoz

doi:10.21273/HORTSCI17185-23

Breeding blueberry cultivars with enhanced fruit quality requires simple, accurate, and cost-effective assays to select individuals from segregating populations. In this study, berry diameter, berry weight, firmness, pH, total polyphenol, total acids, D-glucose, D-fructose, total glucose, and total sugar content were quantified in 188 southern highbush blueberry selections and cultivars over 2 years. Significant variation between years, genotype, and year × genotype interaction was detected for all traits. Glucose and fructose were the predominant sugars, and they were in a range of 32.14–64.72 and 28.61–69.63 mg/mL, respectively. Total sugars content ranged from 62.22 to 131.15 mg/mL. Correlation analysis showed a strong positive correlation between total sugar content measured with the discrete analyzer and total soluble solids assessed as Brix (r² = 0.96). In addition, glucose, fructose, and total glucose showed high and positive correlation between them and with the total sugar content. The titratable acidity was positively correlated with total acids (r² = 0.60) and strong positive correlation between berry diameter and berry weight (r² = 0.94) was detected. Principal component analysis (PCA) showed that PC1 explained 44.9% of the variation and the major contributing traits for diversity were D-fructose, D-glucose, total glucose, and total sugars. PC2 accounted for 21.2% of the variation and was mainly attributed to berry weight and diameter. Cluster analysis showed that the blueberry genotypes fell into two major groups. Cluster-I comprised genotypes with the highest amounts of total acids, pH, polyphenol, D-glucose, D-fructose, total glucose, and total sugar, whereas Cluster-II has genotypes with distinctly lower amounts of tested compounds and larger berries. Information obtained from this study is critical to identify superior genotypes for future crosses and advance evaluation. In addition, the firmness tester and discrete analyzer used in this study were invaluable in improving the efficiency and precision of phenotyping.

The cultivated tetraploid southern highbush blueberry (SHB) is a result of interspecific hybridizations between northern highbush blueberry (NHB) and low-chill Vaccinium species native to the southern United States. Most of the old SHB blueberry cultivars have the NHB cultivar Bluecrop and the Vaccinium darrowii clone FL4B in their pedigree. However, the successful development of elite and region-specific cultivars necessitates the use of additional Vaccinium species in breeding. For example, the diploid species Vaccinium elliottii (2n = 2x = 24) is being used to introduce disease resistant, early flowering, and adaptation to dry upland soils (Dweikat and Lyrene 1991) and as result, three SHB cultivars Carteret, Snowchaser, and Kestrel were developed (Brevis et al. 2008). The diploid species Vaccinium tenellum has been successfully used in SHB breeding to introduce drought tolerance traits, and two cultivars Bladen and Reveille have resulted from these efforts (Ballington 2001; Ballington et al. 1993). These activities have resulted in SHB cultivars that possess a mixture of alleles from four or more Vaccinium species (Nishiyama et al. 2021). Despite significant improvements in adaptation traits obtained through traditional breeding methods, little is known about the level of variation in sugars, organic acids, and phenolic compounds existing within the SHB gene pool.

Blueberry consumption has increased markedly in the past few decades and, as a result, demand for high-quality berries has also increased (Gallardo et al. 2018; Gilbert et al. 2014). Developing SHB with improved berry qualities necessitates evaluation of a large and diverse germplasm collections in replicated field trials across years for different traits. Most of the fruit quality traits are under polygenic control displaying a continuous phenotypic expression, moderate heritability values, and are subject to significant genotype × environment interaction (G × E), making breeding for fruit quality traits a difficult task (Cellon et al. 2018; Edger et al. 2022). Berry texture, flavor, sugar, and acid content are among important traits determining blueberry quality. Blueberry flavor is influenced by the sugar-acid ratio and by aroma which is determined by quantity and composition of volatile organic compounds. Titratable acidity and pH are the determinant factors for fruit acidity. The two most abundant acids in blueberries are malic acid and citric acid, and glucose and fructose are the predominant sugars (Mikulic-Petkovsek et al. 2015; Wang et al. 2012, 2019).

Knowledge of levels and patterns of phenotypic variation in a germplasm collection is essential for improving berry quality in breeding new blueberry cultivars. Manual phenotyping tools represent a bottleneck inhibiting massive phenotyping of fruit quality traits. Numerous analytical tools and techniques have been used by breeders to measure berry quality traits. Digital refractometers are the most widely used tools for measuring total sugars in fruits, and chromatographic methods are used to determine their composition and concentration (Jia et al. 2020). Titratable acidity is determined by neutralizing the acid using standard base (Chiabrando et al. 2009) and several chromatographic methods have been used to determine the composition and quantify the concentration of organic acids (Wang et al. 2019). FirmTech®, FruitFirm®, and Texture analyzer are among instruments widely used to determine the firmness of blueberries (Ehlenfeldt and Marttin 2002; Luby et al. 2023). These instruments measure the compression force (g) needed by a probe to press the surface of the fruit 1 mm. However, using these instruments is time-consuming and labor intensive and some require a large amount of juice to be extracted. Enzymatic analysis estimates sugar concentration based on enzyme metabolism and change in sample absorbance can measure small amounts of specific types of sugars in a short time (Blunden and Wilson 1985). The discrete analyzer system uses two measurement techniques for photometric and electrochemical analyses, allowing parallel determination of a variety of analytes present in a sample simultaneously. Thus, the system enables determination of berry pH, total polyphenol, total acids, D-glucose, D-fructose, total glucose (sucrose), and total sugar content, while overcoming limitations of standard berry quality analysis techniques.

In general, plant breeders measure different traits, but some of which may not be of sufficient discriminatory power to describe, and group sets of genotypes. One way to overcome this problem is to reduce the dimensionality of the data. Multivariate statistical methods such as principal component and cluster analyses are extensively used to find factors describing the greatest variation existing in a data set and estimating relationships between a set of genotypes while considering the relationships between different traits. The objectives of this research were to i) estimate berry firmness and berry diameter of 188 SHB blueberry genotypes using the firmness tester and determine the correlation between berry weight and berry diameter; ii) test the feasibility and accuracy of discrete analyzer to estimate total acids, pH, polyphenol, D-glucose, D-fructose, total glucose, and total sugar in 188 SHB blueberry genotypes and determine the correlation between total sugar content measured using the discrete analyzer and total soluble solids assessed using a handheld digital refractometer; and iii) use PCA and cluster analysis to evaluate the magnitude of genetic diversity among genotypes and to group our germplasm collection based on combination of fruit quality traits.

Materials and Methods

Plant materials.

A diverse panel containing 58 released cultivars from blueberry breeding programs in Mississippi, Florida, Georgia, and North Carolina and 130 breeding selections from the US Department of Agriculture Agricultural Research Service blueberry breeding program were established in Poplarville, MS, USA (30.8402°N, 89.5342°W) in a sandy loam soil (pH 5.2) amended with pine bark. The experimental design was a completely randomized design with four replications of each genotype. Standard cultural and production practices were used. These include drip irrigation, summer pruning, and application of 6.5 g/plant of 13–13–13 fertilizer after harvest. Fertilizer was broadcast on the soil surface under the plant canopy. Pruning consisted mainly of removing lower twiggy growth, dead or damaged shoots, and weaker growth. Two freeze events occurred in Feb (−8 °C) and Mar 2021 (−5 °C), whereas −3 °C was recorded in Feb and Mar 2022. However, extensive damage was observed in the 2022 season due to a mild winter in Dec 2021 (16 °C) compared with average high temperatures of (1 °C) in Dec 2020.

Approximately 50 berries from each of the 188 SHB genotypes were harvested when more than 50% of the berries were ripe in the 2021 and 2022 seasons. Berry weight of 32 randomly selected berries was estimated using a standard laboratory scale. Berry diameter and firmness of 32 berries were assessed on the same day of harvest using Agrosta®winterwood Firmness Tester (formerly Durofel DFT 100, Serqueux, France). Berry firmness was estimated by measuring the force (g/mm) required to compress the skin.

Chemical analyses.

Approximately 20 randomly selected berries were homogenized, filtered through a 0.22-μm cellulose membrane, and 2-mL of blueberry juice was added to a sample cup and used in an automated photometric system (Gallery discrete analyzer; Thermo Fisher Scientific, Waltham, MA, USA) to estimate pH, total polyphenol (PPhenol), total acids (TAcid), D-glucose (DGluc), D-fructose (DFructo), total glucose (TGlucos), and total sugar content (SuFrGl). A built-in barcode reader was used to enter the sample information and detect the kits. pH was estimated using electrochemical pH sensor and four calibration standards. The total acids test was estimated using a potentiometric titration with bromothymol blue as indicator and a self-made citric acid solution (15 g/L) for calibration. Change in color was measured at a wavelength of 620 nm and a side wavelength of 700 nm. Results were calculated automatically by the analyzer using calibration curve and expressed as citric acid. The total polyphenol test is based on colorimetric Folin Ciocalteu method using a 700-nm filter and gallic acid for calibration. Total phenolic contents were expressed as gallic acid equivalents.

For sugar analyses, a sugar combination standard was used for calibration and five ready-to-use enzymatic kits [D-glucose # 984304, D-fructose 984302, D-glucose + D-fructose #984314, D-glucose + D-fructose + Sucrose # 984317, Sucrose (total glucose) # 984312] were used. The D-glucose, D-fructose, and D-glucose + D-fructose estimation was based on a method that uses hexokinase, phosphoglucose isomerase, and glucose-6-phosphate dehydrogenase, and the D-glucose + D-fructose + Sucrose and Sucrose estimation was based on a test that uses beta-fructosidase, hexokinase, phosphoglucose isomerase, and glucose-6-phosphate dehydrogenase. All analyses were performed at 37 °C using disposable optical cuvettes that are loaded with samples and reagents for each test, mixed, incubated, and measured at 340 nm and an additional 600-nm side wavelength was used for verification. The results were expressed as mg/mL. The total sugar content (SuFrGl) was calculated as the sum of D-glucose, D-fructose, and total glucose (sucrose). Total soluble solids (TSS) measured in Brix was determined for each sample using a handheld digital refractometer (Microsystems AR 200; Leica, Wetzlar, Germany). Titratable acidity was measured by titration with 0.1 N NaOH using an automatic compact titrator (862 titrosampler; Metrohm, Tokyo, Japan) and expressed as the percentage of citric acid content.

Statistical analysis.

Analysis of variance was performed using Proc General linear model in SAS (version 9.4; SAS institute, Cary, NC, USA). Means, standard deviation, coefficient of variation (CV), minimum and maximum values, genotypic (δ²_g), phenotypic (δ²_p), and environmental variances (δ²_e) were estimated using SAS. Broad-sense heritability (H²) was calculated using the following formula: $H^{2} = \frac{δ_{g}^{2}}{δ_{g}^{2} + \frac{δ_{g x e}^{2}}{y} + \frac{δ_{e}^{2}}{y r}},$ where yr represents the number of year replications (Holland et al. 2002).

To investigate the relationships among trait variables and the factors underlying genotype variation, Hierarchical clustering analysis based on Ward’s agglomeration methods were performed using LSmeans in JMP Genomics version 15. PCA was conducted in JMP and a biplot of PC1 and PC2 was constructed. Pearson’s coefficient of correlation was used to determine the significant relationships between different phenotypic variables using Proc Corr in SAS (version 9.4, SAS Institute).

Results

Analysis of variance revealed significant effects of genotype, year, and interaction of genotype by year (G × Y) for all measured berry quality traits (P < 0.001). Despite the highly significant G × Y effect, values for all measured traits except firmness were higher in the 2021 season than in 2022 (Table 1). Substantial variability for berry weight was observed among genotypes with largest berries from ‘Raven’ and smallest berries from US846. Berry weight means ranged between 0.5 g for US846 and US804 to 3.56 g for ‘Raven’. We observed a strong and significant correlation between berry diameter and berry weight (R = 0.94, Fig. 1). In this study, we also observed noteworthy variation in firmness where the consistently firmest genotypes were ‘Cooper’, ‘Pearl’, ‘Magnolia’, MS803, MS1670, MS2220, and MS2864, whereas MS2069 and ‘Springhigh’ were the softest.

Fig. 1. Pearson’s correlation coefficients between berry weight and berry diameter measured with the Firm Tester. Each point represents the LSmean of each genotype.
Citation: HortScience 58, 7; 10.21273/HORTSCI17185-23

The total polyphenol level in juice ranged from as low as 2.29 mg/mL in MS1509 to 4.63 mg/mL in MS2274, and the juice pH ranged from 3.85 in MS1014 to 5.6 in US846 (Table 1). Pearson correlation analysis showed a significant and positive correlation between total polyphenol content and juice pH (Table 2). The overall total acids ranged from 4.4 mg/mL in ‘Meadow lark’ and ‘Raven’ to 10.38 mg/mL in US846 (Table 1) and a significant and positive correlation between titratable acidity and total acids was detected (R = 0.62, Fig. 2). In all tested genotypes, glucose and fructose were the predominant sugars, and sucrose was the least abundant sugar type. Significant differences in D-fructose were observed among genotypes with the levels ranging from 32.1 mg/mL in MS1509 to 64.7 mg/mL in ‘Snowchaser’. D-glucose levels ranged from 28.6 mg/mL in MS1509 to 69.6 mg/mL in US846, whereas total glucose ranged from 31.1 mg/mL in MS1509 to 64.9 mg/mL in ‘Abundance’. The significant differences within and among sugars levels in all different genotypes resulted in a marked variability in total sugars, which ranged from 62.2 mg/mL in MS1509 to 131.1 mg/mL in ‘Abundance’ and MS1758. A high and positive correlation was detected among D-fructose, D-glucose, total glucose, and total sugars content (Table 2). Further, Pearson correlation analysis showed a strong and positive correlation between total sugars content measured with the discrete analyzer and TSS assessed as Brix (R = 0.96, Fig. 3). The significant differences in total sugars and total acids indicated a marked variability in sugar-to-total acids ratios, which ranged from 7.05 for MS1670 to 26.9 for MS2864. Eight genotypes, namely ‘Raven’, ‘KeeCrisp’, ‘Bobolink’, ‘MissAlice’, MS1125, MS1758, MS1847, and MS2864, had high sugar-to–total acids ratios in two years.

Fig. 2. Pearson’s correlation coefficients between total acids measured with the discrete analyzer and the titratable acidity. Each point represents the LSmean of each genotype.
Citation: HortScience 58, 7; 10.21273/HORTSCI17185-23

Fig. 3. Pearson’s correlation coefficients between total sugar content measured with the discrete analyzer and the total soluble solids assessed as Brix. Each point represents the LSmean of each genotype.
Citation: HortScience 58, 7; 10.21273/HORTSCI17185-23

Broad-sense heritability estimates showed that different proportions of the observed variation were controlled by genetic factors. D-glucose, D-fructose, total glucose, and total sugars had heritability estimates of 0.53 to 0.56, whereas berry weight and diameter displayed high estimate of broad-sense heritability (Table 1).

Phenotypic analysis by PCA.

To obtain information on similarity and differences among genotypes, PCA was performed for berry quality traits. The results of the PCA showed that of 10 PCs, five components accounted for 94.5% of the variation. Among all the PCs, the first two components described 66.1% of the total variability. PC1 explained 44.9% of the total phenotypic variation and the major contributing traits for diversity in PC1 were D-fructose, D-glucose, total glucose, and total sugars (Table 3). PC2 accounted for 21.2% of the variation, which was mainly attributed to berry weight and berry diameter. PC3 described 12.8% of the total variability and depicted mainly by total acids. PC4 accounted for 9.47% and ascribed by berry firmness, and PC5 explained 6.06% and illustrated by pH and total polyphenol (Table 3).

The PCA biplot showed the trait profiles of tested SHB genotypes and further explained the amount of diversity present (Fig. 4). For example, ‘Raven’ had the largest berries with higher berry weight but had low pH, low polyphenol, and low total acids, and conversely US845, US846, and MS978 had the smallest berries and high level of polyphenol and total acids. Traits such as berry weight and berry diameter displayed a high amount of variability, whereas berry firmness showed lowest variability. In addition, the PCA biplot confirmed the strong correlation between berry diameter and berry weight.

Factors associated with the 10 PCs were subject to cluster analysis using the Euclidean distance matrix on Ward’s hierarchical cluster method. The resulting constellation plot demonstrated that the blueberry genotypes fell into two major clusters (Fig. 5). Cluster-I split into six subclusters (C1 to C6), whereas Cluster-II consisted of seven subclusters (C7 to C13). Cluster-I comprised genotypes with highest amounts of total acids, pH, polyphenol, D-glucose, D-fructose, total glucose, and total sugar, whereas Cluster-II has genotypes with distinctly lower amounts of tested compounds and larger berries (Table 4).

Fig. 5. Constellation plot cluster analysis based on 188 southern highbush blueberry genotypes using phenotypic data of 10 fruit-related traits.
Citation: HortScience 58, 7; 10.21273/HORTSCI17185-23

Discussion

Breeding new SHB blueberry cultivars possessing enhanced fruit quality begins with characterizing the existing germplasm and subsequent production of segregating populations. Most of the fruit quality traits are quantitatively inherited, influenced by environmental factors, and require specialized equipment resulting in costly measurements. Historically, phenotypic recurrent selection strategies have been successful as efficient breeding techniques to improve blueberry germplasm. However, the selection methods have relied on subjective evaluation to assess most external fruit quality traits such as firmness. To overcome these challenges, it is advantageous to use fast and inexpensive methods to characterize breeding materials and generate information related to nature and degree of diversity in fruit quality traits.

In this regard, a diverse set of 188 SHB genotypes was evaluated for fruit quality traits and significant effects for genotypes, years, and genotypes × year were observed for all tested traits. This result confirms the previous finding of Howard et al. (2003), who reported seasonal variation in total sugar, phenolic content, and organic acids. Berry firmness is one of the important traits to develop cultivars amenable to machine harvest. Results from this study revealed that 27% to 39% of the tested genotypes displayed a firmness value of 160 g/mm or above what is considered as superior value (Ehlenfeldt and Marttin 2002). Interestingly, the firmest selections, MS803, MS1670, MS2220, and MS2864, have either ‘Pearl’, or ‘Magnolia’ as parent indicating that the firmness was transmitted from them. Further, two genotypes, MS2069 and ‘Springhigh’, exhibited the lowest firmness values. This confirmed a previous report by Olmstead and Finn (2014), who reported that the SHB cultivar Springhigh produces soft berries. In addition to firmness, berry diameter and berry weight are key traits that not only influence consumer acceptance of blueberry cultivars but also are important components contributing to yield. In this study, a marked degree of diversity for berry diameter and weight was observed. Consistent with previously reported results (Ferrao et al. 2018), our study demonstrated that a strong and positive correlation exists between berry weight and berry diameter. Previously, a study by Mengist et al. (2020) found a strong correlation between berry weight and berry volume and it was concluded that berry weight can be used as a proxy to select for berry volume. These findings indicated that berry diameter could be used to select for higher berry weight and thereby for high yield.

Fructose, pH, and several volatile compounds are key factors affecting consumer preference of blueberries (Gilbert et al. 2015). In this study, our finding that the major sugars in tested SHB genotypes are glucose and fructose, agreed with earlier finding by Jia et al. (2020). Significant variation was detected among tested genotypes regarding D-glucose, D-fructose, total glucose, and total sugars. Variation of soluble solids content has also been reported in previous studies (Mengist et al. 2020; Prior et al. 1998). Our results revealed a strong positive correlation between total sugar content measured with the discrete analyzer and the TSS assessed as Brix°. In addition, findings that total sugars content was highly correlated with D-glucose, D-fructose, and total glucose agreed with previous result by Jia et al. (2020). The strong correlation between D-glucose, D-fructose, total glucose, and total sugars enables indirect selection for high TSS based on individual sugar. Fructose levels in juice samples can be determined using only two kits, which cost ∼25 cents per sample. Significant variation was also observed among tested genotypes regarding the total acids. In addition, a significant and positive correlation between total acids and titratable acidity was detected. This finding confirmed results obtained by Jia et al. (2020), who found strong correlation between citric acid and titratable acidity. In blueberries, malic acid is the predominant organic acid in rabbiteye (Vaccinium virgatum) cultivars, whereas citric acid is the predominant in V. corymbosum, and Vaccinium myrtillus. Quinic and Shikimic acids are the major organic acids in the diploid V. darrowii and V. tenellum (Mikulic-Petkovsek et al. 2015; Wang et al. 2012). Because the SHB germplasm possesses a mixture of alleles from different Vaccinium species, low correlation between the titratable acidity and total acids could be attributed to the variability in organic acids composition. Aroma and sugar-to-acid ratio are crucial indexes that determine the taste balance of blueberry and subsequently the acceptance of a new cultivar. In this study, eight genotypes consistently displayed a high sugar-to-acids ratio over the two seasons. Mikulic-Petkovsek et al. (2015) attributed the high sugars to a low crop load. This could be true for certain genotypes, such as MS1758 and MS1847.

Despite success in breeding superior blueberry cultivars, making genetic progress is still difficult when traits have low heritability. Estimating variance components, heritability, and phenotypic correlations between traits is useful for predicting genetic progress. In this study, variation among blueberry genotypes was greater than that observed between the two growing seasons for all tested traits, contributing to results of intermediate and higher estimates of broad-sense heritability. Berry weight and berry diameter displayed high heritability estimates, and the heritability estimates of 0.8 for berry weight is similar to that of 74% reported by Mengist et al. (2020). High estimates of broad-sense heritability for a trait indicates that it may be controlled by a few major genes and elucidated the importance of nonadditive effects.

Understanding relationships between different berry quality traits is crucial to breeders. PCA summarizes the data and allows for extraction of valuable information from measurements of correlated traits. The biplot revealed associations among traits and trait profiles of tested genotypes and showed positive relationships between the traits in the same group. Interestingly, the clustering reflected genetic relationships among blueberry genotypes. For example, several siblings, such as MS1150 and MS1385 and MS1375 and MS1414, were grouped together. Further, genotypes possessing a common ancestral parent consistently clustered together. The tight relationship observed between D-glucose, D-fructose, total glucose, and total sugars is useful in indirect selection. The maximum variation was observed in PC1 followed by PC2 in comparison with other PCs, therefore the traits contributing to PC1 and PC2 should indeed be included in future genetic diversity studies. Further, the PC scores obtained from the study is useful to develop precise selection indices to select genotypes for advanced testing. Grouping of germplasm in a PCA plot indicates specific genotypic properties, information useful in selecting parents for use in crossing and in selecting individuals for further testing. Furthermore, the PCA results point out that the genetic background has a major effect on fruit quality traits. Hierarchical cluster analysis detects the relationship between genotypes and allow us to identify diverse parental combinations to create segregating populations that could result in superior hybrids.

In the present work, we hypothesized that an accurate and efficient phenotyping platform will greatly enhance the blueberry breeding efficiency and help breeders in selecting superior genotypes for crosses and further testing. For this purpose, we have used numerous analytical techniques to assess different fruit quality traits. Berry diameter and firmness were assessed simultaneously using a firmness tester, and the discrete analyzer was used to characterize hundreds of genotypes and obtain precise measurements of desirable fruit quality traits. The two high-throughput phenotyping techniques have shown enormous potential for improving both efficiency and precision in blueberry selection. In addition, use of the barcode system in the two techniques not only reduced time and labor, but it also eliminated the risk of human error. Characterizing the SHB germplasm collection revealed that significant variation exists among genotypes, demonstrating the potential of the collection in breeding high-quality cultivars. Pearson correlation analyses highlighted the strong correlation between different fruit quality traits. Further, constellation plot cluster and genotype by trait biplot based on PCA proved to be an efficient tool to group genotypes for future breeding activities. Information obtained from this study will be applied to select superior genotypes for future crosses and further advance testing.

[1] Ballington JR
. 2001. Collection, utilization, and preservation of genetic resources in Vaccinium. HortScience. 38:213–220. https://doi.org/10.21273/HORTSCI.36.2.206.
OpenURL
PubMed
Google Scholar
Crossref

[2] OpenURL

[3] PubMed

[4] Google Scholar

[5] Crossref

[6] Ballington JR
,
Rooks SD
,
Milholland RD
,
Cline WO
,
Meyers JR
. 1993. Breeding blueberries for pest resistance in North Carolina. Acta Hortic.
346
:87–94. https://doi.org/10.17660/ActaHortic.1993.346.11.
OpenURL
PubMed
Google Scholar
Crossref

[7] OpenURL

[8] PubMed

[9] Google Scholar

[10] Crossref

[11] Blunden CA
,
Wilson MF
. 1985. A specific method for the determination of soluble sugars in plant extracts using enzymatic analysis and its application to the sugar content of developing pear fruit buds. Anal Biochem. 151:403–408. https://doi.org/10.1016/0003-2697(85)90195-2.
OpenURL
PubMed
Google Scholar
Crossref

[12] OpenURL

[13] PubMed

[14] Google Scholar

[15] Crossref

[16] Brevis P
,
Bassil NV
,
Ballington JR
,
Hancock JF
. 2008. Impact of wide hybridization on highbush blueberry breeding. J Am Soc Hortic Sci. 133:427–437. https://doi.org/10.21273/JASHS.133.3.427.
OpenURL
PubMed
Google Scholar
Crossref

[17] OpenURL

[18] PubMed

[19] Google Scholar

[20] Crossref

[21] Cellon C
,
Amadeu RR
,
Olmstead JW
,
Mattia MR
,
Ferrao LF
,
Munoz PR
. 2018. Estimation of genetic parameters and prediction of breeding values in an autotetraploid blueberry breeding population with extensive pedigree data. Euphytica. 214:87. https://doi.org/10.1007/s10681-018-2165-8.
OpenURL
PubMed
Google Scholar
Crossref

[22] OpenURL

[23] PubMed

[24] Google Scholar

[25] Crossref

[26] Chiabrando V
,
Giacalone G
,
Rolle L
. 2009. Mechanical behaviour and quality traits of highbush blueberry during postharvest storage. J Sci Food Agr. 89:989–992. https://doi.org/10.1002/jsfa.3544.
OpenURL
PubMed
Google Scholar
Crossref

[27] OpenURL

[28] PubMed

[29] Google Scholar

[30] Crossref

[31] Dweikat IM
,
Lyrene PM
. 1991. Induced tetraploidy in a Vaccinium elliottii facilitates crossing with cultivated highbush blueberry. J Am Soc Hortic Sci. 116:1063–1066. https://doi.org/10.21273/JASHS.116.6.1063.
OpenURL
PubMed
Google Scholar
Crossref

[32] OpenURL

[33] PubMed

[34] Google Scholar

[35] Crossref

[36] Edger PP
,
Iorizzo M
,
Bassil NV
,
Benevenuto J
,
Ferrúo LF
,
Giongo L
,
Hummer K
,
Lawas LM
,
Leisner C
,
Li C
,
Munoz PR
,
Ashrafi H
,
Atucha A
,
Babiker EM
,
Canales E
,
Chagn D
,
DeVetter L
,
Ehlenfeldt M
,
Espley RV
,
Gallardo K
,
Gnther CS
,
Hardigan M
,
Hulse-Kemp AM
,
Jacobs M
,
Lila MA
,
Luby C
,
Main D
,
Mengist MF
,
Owens GL
,
Perkins-Veazie P
,
Polashock J
,
Pottorff M
,
Rowland LJ
,
Sims CA
,
Song GJ
,
Spencer J
,
Vorsa N
,
Yocca AE
,
Zalapa J
. 2022. There and back again; Historical perspective and future directions for Vaccinium breeding and research studies. Hortic. Res. 9:uhac083. https://doi.org/10.1093/hr/uhac083.
OpenURL
PubMed
Google Scholar
Crossref

[37] OpenURL

[38] PubMed

[39] Google Scholar

[40] Crossref

[41] Ehlenfeldt MK
,
Marttin RB Jr
. 2002. A survey of fruit firmness in highbush blueberry and species introgressed blueberry cultivars. HortScience. 37:386–389. https://doi.org/10.21273/HORTSCI.37.2.386.
OpenURL
PubMed
Google Scholar
Crossref

[42] OpenURL

[43] PubMed

[44] Google Scholar

[45] Crossref

[46] Ferrao LF
,
Benevenuto J
,
Oliveira B
,
Cellon C
,
Olmstead J
,
Kirst M
,
Resende MFR
,
Munoz P
. 2018. Insights into the genetic basis of blueberry fruit-related traits using diploid and polyploid models in a GWAS context. Front Ecol Evol. 6. https://doi.org/10.3389/fevo.2018.00107.
OpenURL
PubMed
Google Scholar
Crossref

[47] OpenURL

[48] PubMed

[49] Google Scholar

[50] Crossref

[51] Gallardo RK
,
Stafne ET
,
DeVetter LW
,
Zhang Q
,
Li C
,
Takeda F
,
Williamson J
,
Yang WQ
,
Cline WO
,
Beaudry R
,
Allen R
. 2018. Blueberry producers’ attitudes toward harvest mechanization for fresh market. HortTechnology. 28:10–16. https://doi.org/1021273/HORTTECH03872-17.
OpenURL
PubMed
Google Scholar
Crossref

[52] OpenURL

[53] PubMed

[54] Google Scholar

[55] Crossref

[56] Gilbert JL
,
Guthart MJ
,
Gezan SA
,
De Carvalo MP
,
Schwieterman ML
,
Colquhoum TA
,
Bartoshuk LM
,
Sims CA
,
Clark DG
,
Olmstead JW
. 2015. Identifying breeding priorities for blueberry flavor using biochemical, sensory, and genotype by environment analyses. PLoS One. https://doi.org/10.1371/journal.pone.0138494.
OpenURL
PubMed
Google Scholar
Crossref

[57] OpenURL

[58] PubMed

[59] Google Scholar

[60] Crossref

[61] Gilbert JL
,
Olmstead JW
,
Colquhoun TA
,
Levin LA
,
Clark DG
,
Moskowitz HR
. 2014. Consumer-assisted selection of blueberry fruit quality traits. HortScience. 49:864–873. https://doi.org/1021273/HORTSCI497864.
OpenURL
PubMed
Google Scholar
Crossref

[62] OpenURL

[63] PubMed

[64] Google Scholar

[65] Crossref

[66] Holland JB
,
Nyquist WE
,
Cervantes Martinez CE
. 2002. Estimating and interpreting heritability for plant breeding: An update. Plant Breed Rev. 22:9–112. https://doi.org/101002/9780470650202ch2.
OpenURL
PubMed
Google Scholar
Crossref

[67] OpenURL

[68] PubMed

[69] Google Scholar

[70] Crossref

[71] Howard LR
,
Clark JR
,
Brownmiller C
. 2003. Antioxidant capacity and phenolic content in blueberries as affected by genotype and growing season. J Sci Food Agr. 83:1238–1247. https://doi.org/101002/jsfa1532.
OpenURL
PubMed
Google Scholar
Crossref

[72] OpenURL

[73] PubMed

[74] Google Scholar

[75] Crossref

[76] Jia Z
,
Ji-yun N
,
Jing L
,
Hui Z
,
Ye L
,
Farooq S
,
Bacha SA
,
Jie W
. 2020. Evaluation of sugar and organic acid composition and their levels in highbush blueberries from two regions of China. J Integr Agric. 19:2352–2361. https://doi.org/101016/S2095-3119(20)63236-1.
OpenURL
PubMed
Google Scholar
Crossref

[77] OpenURL

[78] PubMed

[79] Google Scholar

[80] Crossref

[81] Luby CH
,
Doane S
,
Mackey T
,
Yang WQ
. 2023. A comparison of two firmness-testing machines for measuring blueberry firmness and size. HortTechnology. 33:98–102. https://doi.org/1021273/HORTTECH05060-22.
OpenURL
PubMed
Google Scholar
Crossref

[82] OpenURL

[83] PubMed

[84] Google Scholar

[85] Crossref

[86] Mengist MF
,
Grace MH
,
Xiong J
,
Kay CD
,
Bassil NV
,
Hummer K
,
Ferruzzi MG
,
Lila MA
,
Iorizzo M
. 2020. Diversity in metabolites and fruit quality traits in blueberry enables ploidy and species differentiation and establishes a strategy for future genetic studies. Front. Plant Sci. 11. https://doi.org/103389/fpls202000370.
OpenURL
PubMed
Google Scholar
Crossref

[87] OpenURL

[88] PubMed

[89] Google Scholar

[90] Crossref

[91] Mikulic-Petkovsek M
,
Schmitzer V
,
Slatnar A
,
Stampar F
,
Veberic R
. 2015. A comparison of fruit quality parameters of wild bilberry (Vaccinium myrtillus L) growing at different locations. J Sci Food Agr. 95:776–785. https://doi.org/10.1002/jsfa6897.
OpenURL
PubMed
Google Scholar
Crossref

[92] OpenURL

[93] PubMed

[94] Google Scholar

[95] Crossref

[96] Nishiyama S
,
Fujikawa M
,
Yamane H
,
Shirasawa K
,
Babiker E
,
Tao R
. 2021. Genomic insight into the developmental history of southern highbush blueberry populations. Heredity. 126:194–205. https://doi.org/101038/s41437-020-00362-0.
OpenURL
PubMed
Google Scholar
Crossref

[97] OpenURL

[98] PubMed

[99] Google Scholar

[100] Crossref

[101] Olmstead JW
,
Finn CE
. 2014. Breeding highbush blueberry cultivars adapted to machine harvest for the fresh market. HortTechnology. 24:290–294. https://doi.org/1021273/HORTTECH243290.
OpenURL
PubMed
Google Scholar
Crossref

[102] OpenURL

[103] PubMed

[104] Google Scholar

[105] Crossref

[106] Prior RL
,
Cao G
,
Martin A
,
Sofic E
,
McEwen J
,
O’Brien C
,
Lischner N
,
Ehlenfeldt M
,
Kalt W
,
Krewer G
,
Mainland CM
. 1998. Antioxidant capacity as influenced by total phenolic and anthocyanin content, maturity, and variety of Vaccinium species. J Agr Food Chem. 46:2686–2693. https://doi.org/101021/jf980145d.
OpenURL
PubMed
Google Scholar
Crossref

[107] OpenURL

[108] PubMed

[109] Google Scholar

[110] Crossref

[111] Wang S
,
Chen H
,
Camp MJ
,
Ehlenfeldt MK
. 2012. Genotype and growing season influence blueberry antioxidant capacity and other quality attributes. Int J Food Sci Technol. 47:1540–1549. https://doi.org/101111/j1365-2621201203004x.
OpenURL
PubMed
Google Scholar
Crossref

[112] OpenURL

[113] PubMed

[114] Google Scholar

[115] Crossref