Reproductive Traits of Hermaphroditic Muscadine Cultivars

in HortScience
View More View Less
  • 1 Departamento de Fitotecnia e Fitossanitarismo, Universidade Federal do Paraná, Rua dos Funcionários, 1540, Curitiba, PR, Brasil 80035-050
  • | 2 Department of Horticulture, University of Georgia–Tifton Campus, 4604 Research Way, Tifton, GA 31793

Native muscadines (Vitis rotundifolia Michx.) are dioecious, but muscadine vineyards are usually planted with a mixture of female plants and hermaphroditic pollenizers. Hermaphroditic cultivars are derived from either of two separate sources original hermaphroditic plants, H1 or H2. Nine hermaphroditic and two female cultivars were studied to determine their potential as pollenizers. Inflorescences of each cultivar were collected in the field to evaluate the number of anthers per flower, the number of pollen grains per anther, and pollen viability and germination in the main and secondary flowering periods. The number of flower clusters per shoot ranged from one to four with most producing two or three flower clusters per shoot. The number of anthers per flower varied by cultivar and cluster position, but in general was between six and eight anthers per flower. ‘Noble’ showed the highest pollen grains production per anther and per flower, reaching 5777 and 39,860, respectively, in the first cluster and ‘Carlos’ produced the least amount of pollen. All cultivars that had secondary flowering showed lower pollen production per anther and per flower as compared with the main flowering period. Optimal muscadine pollen germination media contained 50 mg·L−1 boric acid, 145 mg·L−1 calcium nitrate, 188 g·L−1 sucrose, 10 g·L−1 agar, and 10 mm 2-(N-morpholino)ethanesulfonic acid (MES) pH 6.0. The pollen grain viability of hermaphroditic and female cultivars was high, but pollen grain germination was low in hermaphroditic cultivars and absent in female cultivars. H1-derived cultivars produced more flower buds per cluster and higher germination rates than H2-derived cultivars, indicating they may be better pollenizers. Chemical names used: 2-(N-morpholino)ethanesulfonic acid (MES).

Abstract

Native muscadines (Vitis rotundifolia Michx.) are dioecious, but muscadine vineyards are usually planted with a mixture of female plants and hermaphroditic pollenizers. Hermaphroditic cultivars are derived from either of two separate sources original hermaphroditic plants, H1 or H2. Nine hermaphroditic and two female cultivars were studied to determine their potential as pollenizers. Inflorescences of each cultivar were collected in the field to evaluate the number of anthers per flower, the number of pollen grains per anther, and pollen viability and germination in the main and secondary flowering periods. The number of flower clusters per shoot ranged from one to four with most producing two or three flower clusters per shoot. The number of anthers per flower varied by cultivar and cluster position, but in general was between six and eight anthers per flower. ‘Noble’ showed the highest pollen grains production per anther and per flower, reaching 5777 and 39,860, respectively, in the first cluster and ‘Carlos’ produced the least amount of pollen. All cultivars that had secondary flowering showed lower pollen production per anther and per flower as compared with the main flowering period. Optimal muscadine pollen germination media contained 50 mg·L−1 boric acid, 145 mg·L−1 calcium nitrate, 188 g·L−1 sucrose, 10 g·L−1 agar, and 10 mm 2-(N-morpholino)ethanesulfonic acid (MES) pH 6.0. The pollen grain viability of hermaphroditic and female cultivars was high, but pollen grain germination was low in hermaphroditic cultivars and absent in female cultivars. H1-derived cultivars produced more flower buds per cluster and higher germination rates than H2-derived cultivars, indicating they may be better pollenizers. Chemical names used: 2-(N-morpholino)ethanesulfonic acid (MES).

Muscadine is an important native grape species planted mainly in the southeastern United States. The natural range of V. rotundifolia extends from Delaware to central Florida and along the Gulf of Mexico to eastern Texas. The species extends north along the Mississippi River to Missouri and near the Appalachian Mountains from the east and west (Olien, 1990). Because muscadines are native to the southeastern United States, where the climate is warm, rainy, and humid during most of the year, they have strong resistance to many pests and diseases, including Pierce’s disease (Xylella fastidiosa Wells et al.), which limits the commercialization of bunch grapes in this region (Hopkins et al., 1974; Olmo, 1986). Muscadine fruits have great diversity in size, shape, color, skin, and pulp texture, and possess a unique flavor that distinguishes them from bunch grapes (Lane, 1997). Muscadines are planted for various commercial purposes including wine, juice, jellies, and fresh fruit consumption, and have high levels of anthocyanins and others phenolic compounds with high antioxidant capacity and great health benefits for consumers (God et al., 2007; Greenspan et al., 2005; Hudson et al., 2007).

Native muscadines are typically dioecious, with staminate vines being more common than female vines (Reimer and Detjen, 1910), although a few naturally occurring hermaphroditic vines have been described (Detjen, 1917). The perfect hermaphroditic muscadine flower consists of a normal pistil surrounded by five or more tall erect stamens producing functional pollen. The filaments equal or exceed the combined length of the ovary, style, and stigma (Detjen, 1917). In contrast, female flowers are imperfect hermaphrodites in which the ovary, style, and stigma are thickened and enlarged, and stamens are recurved, shorter, and the pollen produced is sterile (Detjen, 1917). The staminate vine flower has merely a whorl of erect stamens containing pollen (Dearing, 1917). It has been proposed that hermaphroditic vines are staminate vines which have regained the ability to produce a functional pistil (Detjen, 1917). Evidence for this comes from the existence of a vine, which bears blossoms of all stages of gradation from true staminate to true hermaphrodite (Dearing, 1917). Additionally, hermaphroditic and staminate flower clusters are much larger than female clusters (Detjen, 1917).

Before the 1940s, all fruiting muscadines were female and vineyards would typically contain female vines with interspersed male pollenizers. Early on breeders discovered hermpaphroditic vines were occasionally produced from crosses between male and female vines, and the resulting hermaphrodites were used to develop hermaphroditic cultivars (Dearing, 1917). The lineage of all present-day hermaphroditic cultivars can be traced back to two original hermaphroditic seedlings known as H1 and H2 (Dearing, 1948). H1 resulted from a cross between ‘Eden’ and a Vitis munsoniana variety, ‘Mission Male’, whereas H2 resulted from a cross between ‘Scuppernong’ and another V. rotundifolia variety, ‘New Smyrna’. Both hermaphrodites were spontaneously formed and identified from a group of male and female seedlings. These two hermaphroditic varieties were developed in 1911 and 1912, respectively, and all modern self-fertile muscadine cultivars are descended from them. As hermaphroditic cultivars with acceptable fruit quality were developed, the male pollenizer vines were replaced with hermaphroditic pollenizers so that more fruit could be produced (Dearing, 1948).

Although H1 and H2 are both hermaphroditic, there are differences between them. Genetic studies (Loomis et al., 1954) showed that when self-pollinated, hermaphrodites fell into two groups: those producing hermaphroditic and female progeny, and those producing hermaphroditic, female, and male progeny. The first group, which had the H1 source of hermaphroditism, produced three hermaphroditic plants for every one female plant. The second group, which had the H2 source of hermaphroditism, produced a ratio of nine hermaphrodites to three female to four male plants. More work needs to be done to better understand the inheritance of flower type in muscadines, but it has been hampered by the difficulty of emasculating muscadine flowers. Additionally, cluster size in H1 is larger than that of H2 and H1 has lower self-fertility than H2 (Dearing, 1917). What effect these differences have on their ability to function as pollinizers has not been studied.

Modern vineyards are typically a mixture of female vines and hermaphroditic pollenizers. Female cultivars are still planted because they typically have a larger berry size than hermaphroditic cultivars (Gupton, 2000; Williams, 1957). Modern female cultivars can have a berry size as large as 16–18 g, whereas current hermaphroditic cultivars are no larger than 10–13 g (Conner, 2009). However, female cultivars also often have lower yields due to lack of pollination and smaller cluster size (Conner, 2009), and so new hermaphroditic cultivars with very large berry size are being developed (Conner, 2014). Original recommendations were to plant pollinizer vines every third vine in every third row (Reimer and Detjen, 1910). Since then little formal work on pollinizer placement has been conducted and pollinizer placement recommendations have not changed (Krewer et al., 2000).

Muscadine flowers have nectar glands, compact stigmas, waxy pollen grains, and a strong sweet odor, all features common to insect-pollinated flowers (Detjen, 1917). Caging studies indicate that 81% of fruit set in female cultivars is due to insect pollination, with the remainder attributed to wind pollination (Sampson et al., 2001). The main pollinators of muscadines are the small mining bee (Halictus stultus Cress.), green bee (Agapostemon splendens Lep.), gray bee (Magachile sp.), and small bumblebee (Bombus impatiens Cress.) (Hegwood and Himelrick, 2001). The role of self-pollination in female cultivars is controversial. Pollen of female flowers has been shown to be overwhelmingly sterile in germination tests (Reimer and Detjen, 1910) and in bagging experiments in breeding work (Detjen 1917; Husmann and Dearing, 1916; Reimer and Detjen, 1910). Conversely, other pollination studies have suggested low, but significant, self-pollination can occur in female cultivars (Dearing, 1938, 1948; Sampson et al., 2001). It has also been suggested that some female cultivars may produce small amounts of functional pollen, the evidence of which is seen by high fruit set in some female cultivars such as ‘Supreme’, and high fruit set in secondary flowering shoots when relatively little pollen is available from pollenizer vines.

This work was undertaken to estimate the variability among hermaphroditic cultivars in the amount and functionality of pollen produced. Fertilization is affected by a lot of factors in the field, such as weather conditions, types and amount of pollenizers, distance from the source of pollen, nutritional status of plants, coincidence of flowering, and compatibility (Hegwood and Himelrick, 2001; Lu et al., 2005; NeSmith, 1999; Olien, 1990). Therefore, high production of pollen grains per flower with good germination is desirable hermaphroditic pollenizers. Two female cultivars were included in the study to assess whether any functional pollen is produced. The presence of functional pollen in female vines, even at low levels, would have important implications to muscadine breeding as it would allow female cultivars to serve as male parents.

Materials and Methods

Plant material.

Vines were grown at the University of Georgia–Tifton Campus located in Tifton, GA (lat. 33°53′7.69′′N, long. 83°25′20.30′′ W). In 2015, two female cultivars, Fry and Supreme, and nine hermaphroditic flowered cultivars, Carlos, Cowart, Doreen, Granny Val, Hall, Lane, Nesbitt, Noble, and Polyanna, were selected for study. Two vines of each cultivar were used. Vines were trained to a single-wire trellis with two cordons per vine and spaced 3 m between plants within the row and 4.5 m between rows. Vines were irrigated through a single line with a 3.6 L·h−1 drip emitter located 30 cm from each side of the vine. Diseases and insects were controlled according to commercial guidelines (Poling et al., 2003).

Muscadine vines can be quite variable in bud fertility and shoots from the same vine produced from one to four flower clusters. To reduce variation due to the number of clusters per shoot, when flower clusters were visible on 15 May 2015, the number of clusters formed per shoot was evaluated in 100 shoots per cultivar. Then, for each cultivar, 10 shoots with an equal number of clusters were chosen for study with preference given to the largest number of shoots per cluster so that five shoots with that number could be identified on each of two vines. In these shoots, the number of flower buds per cluster was counted in all clusters of all cultivars, at phenological stage 57—inflorescences fully developed, flowers separating according to the BBCH code for grapevine (Lorenz et al., 1995).

Pollen collection and counting.

Ten inflorescences from each available cluster position (primary, secondary, tertiary, or quaternary from shoot base) of each cultivar were collected in the field early in the day (8:00–10:00 am), from clusters that had a few open flowers. After the open flowers were eliminated, 20 flower buds were randomly sampled, the calyptras were removed with a pair of tweezers, and the number of anthers per flower was counted. To collect pollen, the clusters were rubbed through a sieve (Standard Test Sieve, No. 10, pore size = 2 mm, Fisher Sci., Pittsburgh, PA) to remove the calyptras and anthers. Anthers were then collected and dried in a petri dish for 3 d. The anthers and pollen were placed into 15-mL polypropylene centrifuge tubes (15 mL Conical Centrifuge Tubes, Fisher Sci.) and stored at 4 °C for further analysis. Flower clusters from cultivars with enough secondary flowering for adequate replication during the summer were analyzed using the same methods.

The amount of pollen produced in each flower was determined by placing five flower buds into a 1.5-mL microcentrifuge tube and allowing them to dry overnight. Five tubes were used per cluster position. The next day, 0.5 mL of lactic acid (85%) was added to the tube and the anthers were crushed with the end of a No. 2 camel hair brush. Then 0.2 mL of the liquid suspension was placed on a hemocytometer with double chambers. The number of pollen grains contained in each of five big squares in each chamber was determined, which amounted to 10 counts per tube. The number of pollen grains per anther was calculated by:
UNDE1
where Npa is the number of pollen grains per anther, A is the average of the number of pollen grain counted in the 10 squares, a is the number of the anthers crushed, Vi is the volume of the initial solution of lactic acid in mm3 (500 mm3), and Vs is the volume of the one big square of the chamber (0.1 mm3). The number of pollen grains per flowers was calculated by:
UNDE2
where Npf is the number of pollen grains per flower, Npa = number of pollen grains per anther, and Na is the number of the anthers per flower.

Pollen viability and germination.

Pollen viability was tested with dry pollen that was placed onto a glass slide with a 1% acetocarmine solution. After 15 min, the number of stained pollen grains was counted from five fields per sample, each field containing 100 pollen grains.

To improve the germination media for muscadine pollen grains, preliminary germination trials were performed with pollen from Magnolia, which was the first cultivar with open flowers. Inflorescences of this cultivar were collected when the first flowers were open and taken to the laboratory. The open flowers were removed and eliminated. To force flower buds to open, the calyptras were removed with a pair of tweezers and the anthers were collected in a petri dish and dried for 3 d. The anthers and pollen were placed in plastic tubes at 4 °C for 10 d until the start of the trials. Tissue culture plates made of polystyrene with 24 wells (Corning Costar 3524, well volume = 3.4 mL; Corning Inc., Corning, NY) were used in all trials, and 3 mL of culture media was placed in each well. Only in the trial of agar concentrations, to evaluate the treatment without agar, cellophane booklets were used inside petri dishes (Alexander and Ganeshan, 1989). These booklets consisted of seven layers of filter paper (Fisher P5; Fisher Sci.) covered by a layer of cellophane (Research Products International Corp., Mount Prospect, IL), which were moistened with liquid germination media. The pollen was dispersed over the medium with a camel hair brush, which was shaken 2 cm above the plates. This process was observed in the light microscope to confirm a homogeneous distribution of pollen grains over the medium. The plates and petri dishes were put inside closed plastic bags and maintained at 25 °C for 24 h in the dark for all trials. In the germination tests, each well of the tissue culture plates was considered one replicate and 100 pollen grains were counted under a light microscope at ×40 magnification, with four replications per test. Pollen grains that had tubes at least the same diameter as the pollen grain were counted as germinated.

Optimization of germination media tests were carried out sequentially starting with an initial culture medium of 150 g·L−1 sucrose, 10 g·L−1 agar, and 50 mg·L−1 boric acid (Sharafi and Bahmani, 2011). The pH of this media was adjusted to 5.0, 5.5, 6.0, 6.5, and 7.0 through the addition of 0.1 n HCl. Calcium nitrate concentration (0, 100, 200, and 300 mg·L−1) was tested in a factorial experiment with 150 g·L−1 sucrose, 10 g·L−1 agar, 50 mg·L−1 boric acid media, which had a pH of 6.0 adjusted with either the addition or 0.1 nN HCl or using a 10 mm 2-(N-morpholino)ethanesulfonic acid (MES) buffer. Boric acid concentration was tested in a high concentration (50, 100, 200, 400, and 800 mg·L−1) and then a low concentration (0, 10, 20, 30, 40, and 50 mg·L−1) in a germination media consisting of 150 g·L−1 sucrose, 10 g·L−1 agar, 145 mg·L−1 calcium nitrate, and 10 mm MES pH 6.0. Sucrose concentration (0, 50, 100, 150, 200, and 250 g·L−1) was tested in a germination media consisting of 10 g·L−1 agar, 50 mg·L−1 boric acid, 145 mg·L−1 calcium nitrate, and 10 mm MES pH 6.0. Agar concentration (0, 5, 6, 7, 8, 9, and 10 g·L−1) was tested in a buffer consisting of 188 g·L−1 sucrose, 50 mg·L−1 boric acid, 145 mg·L−1 calcium nitrate, and 10 mm MES pH 6.0.

Pollen germinability of cultivars was tested in an optimized media consisting of 188 g·L−1 sucrose, 10 g·L−1 agar, 50 mg·L−1 boric acid, 145 mg·L−1 calcium nitrate, and 10 mm MES pH 6.0. Germination was evaluated in 24-well tissue culture plates maintained at 25 °C for 24 h in the dark with four replications per cultivar. Stored dry pollen, as described above, from the 11 cultivars was tested from the main and secondary flowering shoots.

Data analysis.

In all trials, the experimental design was completely randomized with 10 replicates to evaluate the number of flower buds per cluster, five replicates to evaluate the number of pollen grains per anther and per flower, and four replicates for the other trials. Comparison among cultivars for number of flower buds per cluster, flower bud weight, number of anthers and pollen grains per flower, and pollen grain viability and germination was tested through one-way analysis of variance with mean separation by Scott Knott test (Scott and Knott, 1974). The effect of cultivar and cluster position was evaluated by PROC GLM with cultivar × cluster position interactions not being evaluated due to missing cells resulting from an unequal number of clusters per shoot tested in different cultivars. The effect of pH, boric acid, calcium nitrate, and sucrose concentration was evaluated using polynomial regression analysis and models were evaluated by R2 values. Agar concentration was evaluated by Scott Knott test. Differences between H1 derived (‘Carlos’, ‘Granny Val’, and ‘Noble’) and H2 derived (‘Cowart’, ‘Doreen’, ‘Nesbitt’, and ‘Polyanna’) hermaphroditic cultivars in flower and pollen attributes were tested by considering only the results from the basal two clusters in the main flowering period which all cultivars produced. Cluster position was not significant, so data from both clusters were combined and differences between H1- and H2-derived cultivars were tested by a t test.

Results and Discussion

Vine characteristics.

The vines had two flowering periods. The main flowering period occurred in the spring shortly after budbreak. The secondary flowering period occurred in early summer when new shoots formed from lateral buds on the current season’s growth. Before the beginning of the main bloom period, a preliminary evaluation of the vines was conducted to determine the most common number of flower clusters produced on a shoot for each cultivar. The number of flower clusters per shoot ranged from one to four (Table 1). ‘Fry’, ‘Granny Val’, ‘Hall’, ‘Lane’, ‘Noble’, and ‘Polyanna’ had a higher number of shoots with two clusters, especially ‘Polyanna’ that showed 85% of shoots with two clusters. ‘Doreen’ showed similar proportion of shoots with one and two clusters, whereas ‘Carlos’, ‘Cowart’, and ‘Supreme’ had a similar proportion of shoots with two or three clusters. ‘Nesbitt’ presented a high percentage of shoots with four clusters, 23%.

Table 1.

Percentage of shoots with one, two, three, and four flower clusters.

Table 1.

The number the flowers buds per cluster varied strongly among cultivars and to a lesser extent by cluster position (Table 2). In the main flowering period, ‘Carlos’, ‘Granny Val’, ‘Hall’, ‘Lane’, ‘Noble’, and ‘Polyanna’ produced more than 100 flower buds per cluster, whereas ‘Doreen’, ‘Fry’, ‘Nesbitt’, and ‘Supreme’ were lower than other cultivars (Table 2).

Table 2.

Flower and pollen characteristics of 11 cultivars in the main and secondary flowering period.

Table 2.

Similar results were found for ‘Noble’ and ‘Doreen’ in North Carolina where the number of flowers per cluster was 128 and 84, respectively (Goldy, 1988). In the secondary flowering period, there was little variation between cultivars or among clusters for flower number per cluster. Only ‘Cowart’ had lower number of flower buds in the second cluster in the secondary flowering period (Table 2).

Current season’s flower clusters are differentiated just before and after flowering in the previous season (Vasconcelos et al., 2009), and physiological conditions of the vine during that time period are influential in determining the number of flower clusters produced on the shoot (Guilpart et al., 2014). External factors such as temperature, light intensity, water, and nutrient availability affect cytokinin biosynthesis, which is the most important hormone in the control of flowering in the grapevine (Srinivasan and Mullins, 1981). Warm weather favors further inflorescence differentiation, resulting in more clusters per shoot, whereas cool weather favors more flowers per clusters and fewer clusters per shoot in Vitisvinifera (Vasconcelos et al., 2009). Water and nitrogen stress during the period of floral differentiation determined 65% to 70% of ‘Shiraz’ and ‘Aranel’ grapevine yield of the next season, affected mainly bud fertility and berry number per bunch (Guilpart et al., 2014).

Flower characteristics.

The flower bud weight in the main flowering period ranged from 4.2 to 6.7 mg (Table 2). Most variation occurred between cultivars, but significant variation was also found between clusters. Flower bud weight did not appear to be clearly associated with flower sex, as one female cultivar, Fry, had relatively large buds, whereas the other female cultivar, Supreme, had smaller buds. No clear pattern was seen between cluster position as some cultivars had larger buds in the more distal clusters, whereas other cultivars had smaller buds in the distal position. In the secondary flowering period shoots, flower bud weight ranged from 3.3 to 6.3 mg, with generally lower bud weights as compared with the main flowering period (Table 2). Surprisingly, flower bud weight was not correlated with the number of anthers per flower (Table 3). However, the flower bud weight was positively correlated with the number of pollen grains per anther and the number of pollen grains per flower, in both main and secondary flowering shoots (Table 3), indicating that the heavier buds have larger anthers with a greater number of pollen grains.

Table 3.

Correlations among muscadine flower and pollen characteristics in the main and secondary flowering period.

Table 3.

The number of anthers produced in muscadine flowers is variable, while bunch grapes generally have five anthers per flower (Kelen and Demirtas, 2003). We observed anywhere from five to ten anthers per flower, but six, seven, or eight anthers per flower (Fig. 1A) was the most common configuration. The number of anthers produced per flower varied strongly by cultivar, but not by cluster position (Table 2). ‘Fry’, ‘Nesbitt’, ‘Polyanna’, and ‘Supreme’ produced the most anthers per flower in the main flowering period. Curiously, the female cultivars had the highest average, averaging 7.5 anthers per flower (Table 2; Fig. 1B), despite their pollen being nonfunctional. However, even nonfunctional pollen may play a role in the attraction of pollinators by serving as a food source. In the secondary flowering period, there was less difference among cultivars, and female cultivars had similar number of anthers as the hermaphroditic cultivars.

Fig. 1.
Fig. 1.

Flowers and pollen grains of muscadine cultivars. (A) Flowers of ‘Noble’ with six, seven, and eight anthers; (B) flower of ‘Fry’ with seven anthers; (C) germination of pollen grains of ‘Granny Val’; (D) pollen grains of ‘Supreme’ in the viability test, dark pollen grains are viable and arrows indicate inviable pollen; (E) pollen grains of ‘Carlos’ in the viability test, dark pollen grains are viable and arrows indicate inviable pollen; (F) acolporated pollen of ‘Fry’; (G) tricolporated pollen of ‘Granny Val’; and (H) tricolporated pollen of ‘Lane’. Bars: A–B, 2 mm; C, 50 µm; D–H, 10 µm.

Citation: HortScience horts 51, 3; 10.21273/HORTSCI.51.3.255

The number of pollen grains per anther varied by cultivar and cluster position in the main flowering period, but only by cultivar in the secondary flowering period (Table 2). ‘Noble’ stood out from the other cultivars for the most pollen grains produced per anther and per flower, reaching 5777 and 39,860, respectively, in the first cluster. It was also the cultivar with the highest flower bud weight, but not the highest number of anthers (Table 2), which reinforces the hypothesis that the pollen grains production is related to the size of anther and not the number of anthers. In a study with Zingiber officinale (Roscoe), the authors also found that bigger anthers have higher number of pollen grains (Subbarayadu et al., 2014). ‘Carlos’ had the lowest number of pollen grains produced per anther and per flower and the other cultivars were intermediate with some variation among the clusters. All cultivars that had secondary flowering produced fewer pollen grains per anther and per flower compared with the main flowering period. These estimates of muscadine pollen production are much higher than the results found with eight V. vinifera table grapes cultivars where the number of pollen grains per anther ranged from 581 to 1500 and the number of grains per flower ranged from 2906 to 9000 (Kelen and Demirtas, 2003). We could find no estimates of pollen production of muscadines in the literature.

Pollen viability and germination.

Relatively low pollen germination was seen in initial trials with a pollen germination media composed of 150 g·L−1 sucrose, 10 g·L−1 agar, and 50 mg·L−1 boric acid (Sharafi and Bahmani, 2011). Pollen germination media must be similar to key components of the stigmatic fluid to produce optimal germination, so a series of experiments were performed to optimize the germination media for muscadine pollen. Media pH had a significant effect on pollen germination, with maximal germination of pollen grains obtained at pH 6.0 (Fig. 2A). Results of adjusting pH with HCl or MES buffer were similar, so MES buffer was used for all further experiments as it was more convenient to prepare and use.

Fig. 2.
Fig. 2.

Effect of germination media on germination ‘Magnolia’ muscadine pollen. (A) Test of pH; (B) Test of calcium citrate; (C) Test of high concentrations of boric acid; (D) Test of low concentrations of boric acid; (E) Test of sucrose concentration; (F) Test of agar concentration. Same letters over the bars indicate that the means are not significantly different by Scott Knott test (P > 0.05).

Citation: HortScience horts 51, 3; 10.21273/HORTSCI.51.3.255

The effect of the calcium nitrate on germination was best modeled with a quadratic regression, producing a maximum germination of pollen grains at 145 mg·L−1 (Fig. 2B). Boric acid concentration was tested in a range of concentrations both higher (Fig. 2C) and lower (Fig. 2D) than the 50 mg·L−1 used in the original media. The higher range of concentrations produced a decreasing linear regression with increasing concentrations of boric acid, whereas the low concentrations produced an increasing linear regression with the increase in boric acid concentration. The original rate of 50 mg·L−1 was found to be optimal producing rate of almost 14% germination, while a complete lack of boric acid reduced germination to about 4%.

Having the correct osmotic potential in the germination media is vital for maximal pollen germination, and this role is usually fulfilled through the addition of sucrose. Sucrose concentration greatly affected pollen germination and there was no germination in the culture medium without sucrose, and with only 50 g·L−1 sucrose germination was near zero (Fig. 2E). With increasing sucrose concentration, there was increased the germination with a maximum value of 188 g·L−1 estimated by the quadratic regression equation. Sucrose is also needed for the germination V. vinifera table grape pollen (Kelen and Demirtas, 2003).

In the last test with agar concentrations, germination was lower in liquid media without agar than in the solid culture media with agar, but germination percentage was unaffected by agar concentration (Fig. 2F). We chose to use a concentration of 10 g·L−1 agar since it is the most commonly used proportion in other germination tests with grape pollen (Kelen and Demirtas, 2003; Sharafi and Bahmani, 2011).

The final optimized culture medium containing 50 mg·L−1 boric acid, 145 mg·L−1 calcium nitrate, 188 g·L−1 sucrose, 10 g·L−1 agar, and 10 mm MES pH 6.0 was used to evaluate pollen germination in all muscadine cultivars. Even with the optimized germination media, the highest germination values obtained for ‘Magnolia’ were relatively low at ≈14%. Germination values lower than 15% were also obtained with the seedless V. vinifera table grapes ‘Mystery’, ‘Crimson Seedless’, ‘Autumn Royal’, and ‘Sugraone’, despite several attempts to increase the germination rate. For these cultivars, the best culture medium was 100 mg·L−1 boric acid, 300 mg·L−1 calcium nitrate, and 200 g·L−1 sucrose (Carreño et al., 2009).

Pollen germination rate varied strongly by cultivar and less so by cluster position (Table 2). Germination rates ranged from 2.5% to 18.2% in the hermaphroditic cultivars, and no pollen germination was seen in the female cultivars. Different germination rates are commonly observed in grapevine cultivars (Amjad et al., 1969; Kelen and Demirtas, 2003; Sharafi and Bahmani, 2011). Germination rates were similar to that found with other muscadine cultivars, which range from 10% to 14% (Lu et al., 2005). The highest germination rate was observed with the second cluster of ‘Granny Val’, 18.2% (Fig. 1C). This cultivar was the only one in the main flowering period that had higher germination in the second cluster. In all other cultivars, germination decreased or remained the same from the first cluster. This behavior was also observed in the secondary flowering period. In some cultivars, reduced germination was quite pronounced, as in ‘Cowart’, ‘Noble’, and ‘Polyanna’. This reduction in germination lowers the amount of functional pollen available. Comparing the first and second cluster of ‘Noble’, which did not differ regarding the amount of pollen produced per flower, but which did differ in pollen germination, we can estimate the production of 5,038,304 pollen grains in the first cluster and 5,312,916 pollen grains in the second cluster. But as the first cluster germination rate is 7.7% and the second cluster just 2.7%, the number of function pollen grains is 387,949 in the first cluster and just 143,449 in the second cluster. This amount is less than what can be produced by the second cluster of Carlos (195,828), which was the cultivar with the lowest production of pollen grains per flower.

Pollen grain viability as assessed by acetocarmine staining was high in all cultivars in both flowering periods, and pollen grains of female (Fig. 1D) and hermaphroditic (Fig. 1E) cultivars showed intense red color. Only ‘Carlos’ in the main flowering period had a viability below 90% (Table 2). High levels of muscadine pollen viability were also found by Sampson et al. (2001). However, despite the high viability, pollen grain germination was low in all cultivars, and no germination was seen in the female cultivars in either flowering period. To confirm the female cultivars pollen inability to germinate, the whole plate was evaluated and not a single pollen germinated pollen grain was observed. Our results confirm that even in female cultivars such as Supreme, which commonly set a large crop, the flowers need to be cross-pollinated with functional pollen for fruit production. Observations by the author (P. Conner) of ‘Supreme’ vines over the years indicate that this cultivar is lower in vigor with shorter and less numerous shoots. This may allow pollinators better access to flower clusters, which are obscured by foliage in more vigorously growing cultivars. Better pollinator access would also explain the high set on clusters of the secondary flowering shoots, which are usually produced on top of the main canopy.

The sterility of female muscadine pollen grains was associated with their morphology. All pollen grains of female cultivars were of the acolporated type, which are spheroidal and without furrows (Fig. 1F), while the perfect flowered cultivars were generally tricolporated, with three furrows (Fig. 1G and H). In V. vinifera and Vitis lubruscana pollen, both acolporated and tricolporated pollen stain as viable, but only tricolporated pollen is able to germinate (Abreu et al., 2006; Ahmedullah, 1983). As in muscadine, cultivars only producing acolporated pollen have reflexed stamens (Fig. 1B) and flowers that are functionally female. These two pollen types also have differences related to starch accumulation. Tricolporated pollen have plastids filled with numerous starch granules, whereas the plastids in acolporated pollen lack starch grains. Additionally, acolporated pollen plasma membranes presented an irregular shape and in some regions detached from the pollen wall (Abreu et al., 2006). Pollen grains may also differ in size, shape, and exine characteristics that can be used as parameters in cultivar identification of grapes. Pollen of 43 cultivars of V. labruscana and V. vinifera was observed and the majority of cultivars had tricolporated pollen, with three longitudinal furrows, that were classified as wide furrow, narrow furrow, and narrow furrow with open ends (Ahmedullah, 1983). The muscadine cultivars evaluated in this research appeared to have only dimorphism with acolporated and tricolporated pollen. However, future studies should be performed to quantify the two types of pollen grains to determine if the cause of poor germination is partially due to the existence of acolporated pollen grains in the hermaphroditic cultivars.

Modern hermaphroditic cultivars are derived from either of two original hermaphroditic parents known as H1 and H2. The pedigrees of three cultivars (Carlos, Granny Val, and Noble) can be traced back to H1; and four cultivars (Cowart, Doreen, Nesbitt, and Polyanna) can be traced back to H2. The pedigrees of ‘Hall’ and ‘Lane’ indicate they originated from H1; however, molecular marker data casts doubt on this pedigree (P. Conner, unpublished data) and so these two cultivars were not included in the analysis. When looking at the first two clusters of the H1- and H2-derived cultivars, significant differences were found for the number of flower buds per cluster and pollen grain germination (Table 4). H1-derived cultivars averaged more than 120 flower buds per cluster, whereas H2-derived cultivars averaged less than 100. This is in agreement with observations that the original H2 vine produced smaller clusters than the H1 vine (Dearing, 1917) and with observations that H1 derived ‘Noble’ produced more flowers per cluster than the H2 derived ‘Sterling’ and ‘Doreen’ (Goldy, 1988). H1 was produced from a cross of a female V. rotundifolia to a male V. munsoniana, whereas both the male and female parents of H2 were V. rotundifolia, thus the increase in flower buds may be a feature introduced from V. munsoniana. ‘Polyanna’ was an outlier as it is H2 derived, but produced more than 130 buds per cluster (Table 2). Germination rates were higher in H1-derived cultivars. The combination of higher germination rates and more flowers per cluster, with no differences in the amount of pollen produced per flower means that H1-derived cultivars produce more functional pollen than H2-derived cultivars. However, on a practical level the differences do not appear to be so great so as to discourage the use of H2-derived cultivars as pollenizers, although it may be beneficial to increase the number of pollenizers when they are H2-derived cultivars.

Table 4.

Flower and pollen characteristics of H1 and H2 cultivars.

Table 4.

Literature Cited

  • Abreu, I., Costa, I., Oliveira, M., Cunha, M. & de Castro, R. 2006 Ultrastructure and germination of Vitis vinifera cv. Loureiro pollen Protoplasma 228 131 135

    • Search Google Scholar
    • Export Citation
  • Ahmedullah, M. 1983 Pollen morphology of selected Vitis cultivars J. Amer. Soc. Hort. Sci. 108 155 160

  • Alexander, M. & Ganeshan, S. 1989 An improved cellophane method for in vitrogermination of recalcitrant pollen Stain Technol. 64 225 227

  • Amjad, S., Satyanarayana, G. & Raj, B. 1969 Studies in the pollen morphology and physiology of sixty grape varieties J. Palynol. 5 30 36

  • Carreño, J., Oncina, R. & Carreño, I. 2009 In vitro studies of pollen germination capability and preservation of different cultivars of Vitis vinifera L Acta Hort. 827 493 496

    • Search Google Scholar
    • Export Citation
  • Conner, P. 2009 Performance of muscadine grape cultivars in southern Georgia J. Amer. Pom. Soc. 63 101 107

  • Conner, P. 2014 Characteristics of promising muscadine grape (Vitis rotundifolia Michx.) selections from the University of Georgia (USA) breeding program Acta Hort. 1046 303 307

    • Search Google Scholar
    • Export Citation
  • Dearing, C. 1917 The production of self-fertile muscadine grapes Proc. Amer. Soc. Hort. Sci. 14 30 34

  • Dearing, C. 1938 Muscadine grapes. U.S. Dept. Agr. Farmer’s Bul. 1785

  • Dearing, C. 1948 New muscadine grapes. U.S. Dept. Agr. Circ. 769

  • Detjen, L. 1917 Inheritance of sex in Vitis rotundifolia. North Carolina Agr. Expt. Sta. Tech. Bul. 17

  • God, J.M., Tate, P. & Larcom, L.L. 2007 Anticancer effects of four varieties of muscadine grape J. Med. Food 10 54 59

  • Goldy, R.G. 1988 Variation in some yield determining components in muscadine grapes and their correlation to yield Euphytica 39 39 42

  • Greenspan, P., Bauer, J.D., Pollock, S.H., Gangemi, J.D., Mayer, E.P., Ghaffar, A., Hargrove, J.L. & Hartle, D.K. 2005 Anti-inflammatory properties of the muscadine grape (Vitis rotundifolia) J. Agr. Food Chem. 53 8481 8484

    • Search Google Scholar
    • Export Citation
  • Guilpart, N., Metay, A. & Gary, C. 2014 Grapevine bud fertility and number of berries per bunch are determined by water and nitrogen stress around flowering in the previous year Eur. J. Agron. 54 9 20

    • Search Google Scholar
    • Export Citation
  • Gupton, C.L. 2000 Muscadine traits potentially useful in breeding J. Amer. Pomol. Soc. 54 114 117

  • Hegwood, C.P. Jr & Himelrick, D.G. 2001 Growth and development, p. 117–131. In: F.M. Basiouny and D.G. Himelrick (eds.). Muscadine grapes. ASHS Press, Alexandria

  • Hopkins, D.L., Mollenhauer, H.H. & Mortensen, J.A. 1974 Tolerance to Pierce’s disease and the associated Rickettsia-like bacterium in muscadine grape J. Amer. Soc. Hort. Sci. 99 436 439

    • Search Google Scholar
    • Export Citation
  • Hudson, T.S., Hartle, D.K., Hursting, S.D., Nunez, N.P., Wang, T.T.Y., Young, H.A., Arany, P. & Green, J.E. 2007 Inhibition of prostate cancer growth by muscadine grape skin extract and resveratrol through distinct mechanisms Cancer Res. 67 8396 8405

    • Search Google Scholar
    • Export Citation
  • Husmann, G. & Dearing, C. 1916 Muscadine grapes. USDA Farmers’ Bul. 709

  • Kelen, M. & Demirtas, I. 2003 Pollen viability, germination capability and pollen production level of some grape varieties (Vitis vinifera L.) Acta Physiol. Plant. 25 229 233

    • Search Google Scholar
    • Export Citation
  • Krewer, G., Hall, M., NeSmith, D., Horton, D., Sherm, H., Sumner, P., Tyson, T. & Westberry, G. 2000 Georgia muscadine production guide. Georgia Coop. Ext. Serv. Bull. 739, Athens, GA

  • Lane, R. 1997 Breeding muscadine and southern bunch grapes Fruit Var. J. 51 144 148

  • Loomis, N., Williams, C. & Murphy, M. 1954 Inheritance of flower types in muscadine grapes Proc. Amer. Soc. Hort. Sci. 64 279 283

  • Lorenz, D.H., Eichhorn, K.W., Bleiholder, H., Klose, R., Meier, U. & Weber, E. 1995 Growth stages of the grapevine: Phenological growth stages of the grapevine (Vitis vinifera L. ssp. vinifera)—codes and descriptions according to the extended BBCH scale Austral. J. Grape Wine Res. 1 100 110

    • Search Google Scholar
    • Export Citation
  • Lu, J., Schell, L. & Ramming, D.W. 2005 Interspecific hybridization between Vitis rotundifolia and Vitis vinifera and evaluation of the hybrids Acta Hort. 528 479 486

    • Search Google Scholar
    • Export Citation
  • NeSmith, D.S. 1999 Fruit set and berry size of ‘Fry’ muscadine grape in response to six pollen sources HortScience 34 355

  • Olien, W.C. 1990 The muscadine grape: Botany, viticulture, history, and current industry HortScience 25 732 739

  • Olmo, H.P. 1986 The potential role of (vinifera x rotundifolia) hybrids in grape variety improvement Experientia 42 921 926

  • Poling, B., Mainland, C., Bland, W., Cline, B. & Sorenson, K. 2003 Muscadine grape production guide. NC State Ext. Ser. Bul. AG-94. Raleigh, NC

  • Reimer, F. & Detjen, L. 1910 Self-sterility of the Scuppernong and other muscadine grapes. NC Agr. Exp. Sta. Bul. 209

  • Sampson, B., Noffsinger, S., Gupton, C. & Magee, J. 2001 Pollination biology of the muscadine grape HortScience 36 120 124

  • Scott, R.J. & Knott, M. 1974 A cluster analysis method for grouping means in the analysis of variance Biometrics 30 507 512

  • Sharafi, Y. & Bahmani, A. 2011 Pollen germination, tube growth and longevity in some cultivars of Vitis vinifera L Afr. J. Microbiol. Res. 5 1102 1107

    • Search Google Scholar
    • Export Citation
  • Srinivasan, C. & Mullins, M.G. 1981 Physiology of flowering in the grapevine—a review Amer. J. Enol. Viticult. 32 47 63

  • Subbarayadu, S., Naik, B.S., Devi, H.S., Bhau, B.S. & Khan, P.S.S.V 2014 Microsporogenesis and pollen formation in Zingiber officinale Roscoe Plant Syst. Evol. 300 619 632

    • Search Google Scholar
    • Export Citation
  • Vasconcelos, M.C., Greven, M., Winefield, C.S., Trought, M.C.T. & Raw, V. 2009 The flowering process of Vitis vinifera: A review Amer. J. Enol. Viticult. 60 411 434

    • Search Google Scholar
    • Export Citation
  • Williams, C.F. 1957 Relation of berry size to flower type of seedlings in muscadine grape crosses Proc. Amer. Soc. Hort. Sci. 69 254 260

Contributor Notes

Corresponding author. E-mail: pconner@uga.edu.

  • View in gallery

    Flowers and pollen grains of muscadine cultivars. (A) Flowers of ‘Noble’ with six, seven, and eight anthers; (B) flower of ‘Fry’ with seven anthers; (C) germination of pollen grains of ‘Granny Val’; (D) pollen grains of ‘Supreme’ in the viability test, dark pollen grains are viable and arrows indicate inviable pollen; (E) pollen grains of ‘Carlos’ in the viability test, dark pollen grains are viable and arrows indicate inviable pollen; (F) acolporated pollen of ‘Fry’; (G) tricolporated pollen of ‘Granny Val’; and (H) tricolporated pollen of ‘Lane’. Bars: A–B, 2 mm; C, 50 µm; D–H, 10 µm.

  • View in gallery

    Effect of germination media on germination ‘Magnolia’ muscadine pollen. (A) Test of pH; (B) Test of calcium citrate; (C) Test of high concentrations of boric acid; (D) Test of low concentrations of boric acid; (E) Test of sucrose concentration; (F) Test of agar concentration. Same letters over the bars indicate that the means are not significantly different by Scott Knott test (P > 0.05).

  • Abreu, I., Costa, I., Oliveira, M., Cunha, M. & de Castro, R. 2006 Ultrastructure and germination of Vitis vinifera cv. Loureiro pollen Protoplasma 228 131 135

    • Search Google Scholar
    • Export Citation
  • Ahmedullah, M. 1983 Pollen morphology of selected Vitis cultivars J. Amer. Soc. Hort. Sci. 108 155 160

  • Alexander, M. & Ganeshan, S. 1989 An improved cellophane method for in vitrogermination of recalcitrant pollen Stain Technol. 64 225 227

  • Amjad, S., Satyanarayana, G. & Raj, B. 1969 Studies in the pollen morphology and physiology of sixty grape varieties J. Palynol. 5 30 36

  • Carreño, J., Oncina, R. & Carreño, I. 2009 In vitro studies of pollen germination capability and preservation of different cultivars of Vitis vinifera L Acta Hort. 827 493 496

    • Search Google Scholar
    • Export Citation
  • Conner, P. 2009 Performance of muscadine grape cultivars in southern Georgia J. Amer. Pom. Soc. 63 101 107

  • Conner, P. 2014 Characteristics of promising muscadine grape (Vitis rotundifolia Michx.) selections from the University of Georgia (USA) breeding program Acta Hort. 1046 303 307

    • Search Google Scholar
    • Export Citation
  • Dearing, C. 1917 The production of self-fertile muscadine grapes Proc. Amer. Soc. Hort. Sci. 14 30 34

  • Dearing, C. 1938 Muscadine grapes. U.S. Dept. Agr. Farmer’s Bul. 1785

  • Dearing, C. 1948 New muscadine grapes. U.S. Dept. Agr. Circ. 769

  • Detjen, L. 1917 Inheritance of sex in Vitis rotundifolia. North Carolina Agr. Expt. Sta. Tech. Bul. 17

  • God, J.M., Tate, P. & Larcom, L.L. 2007 Anticancer effects of four varieties of muscadine grape J. Med. Food 10 54 59

  • Goldy, R.G. 1988 Variation in some yield determining components in muscadine grapes and their correlation to yield Euphytica 39 39 42

  • Greenspan, P., Bauer, J.D., Pollock, S.H., Gangemi, J.D., Mayer, E.P., Ghaffar, A., Hargrove, J.L. & Hartle, D.K. 2005 Anti-inflammatory properties of the muscadine grape (Vitis rotundifolia) J. Agr. Food Chem. 53 8481 8484

    • Search Google Scholar
    • Export Citation
  • Guilpart, N., Metay, A. & Gary, C. 2014 Grapevine bud fertility and number of berries per bunch are determined by water and nitrogen stress around flowering in the previous year Eur. J. Agron. 54 9 20

    • Search Google Scholar
    • Export Citation
  • Gupton, C.L. 2000 Muscadine traits potentially useful in breeding J. Amer. Pomol. Soc. 54 114 117

  • Hegwood, C.P. Jr & Himelrick, D.G. 2001 Growth and development, p. 117–131. In: F.M. Basiouny and D.G. Himelrick (eds.). Muscadine grapes. ASHS Press, Alexandria

  • Hopkins, D.L., Mollenhauer, H.H. & Mortensen, J.A. 1974 Tolerance to Pierce’s disease and the associated Rickettsia-like bacterium in muscadine grape J. Amer. Soc. Hort. Sci. 99 436 439

    • Search Google Scholar
    • Export Citation
  • Hudson, T.S., Hartle, D.K., Hursting, S.D., Nunez, N.P., Wang, T.T.Y., Young, H.A., Arany, P. & Green, J.E. 2007 Inhibition of prostate cancer growth by muscadine grape skin extract and resveratrol through distinct mechanisms Cancer Res. 67 8396 8405

    • Search Google Scholar
    • Export Citation
  • Husmann, G. & Dearing, C. 1916 Muscadine grapes. USDA Farmers’ Bul. 709

  • Kelen, M. & Demirtas, I. 2003 Pollen viability, germination capability and pollen production level of some grape varieties (Vitis vinifera L.) Acta Physiol. Plant. 25 229 233

    • Search Google Scholar
    • Export Citation
  • Krewer, G., Hall, M., NeSmith, D., Horton, D., Sherm, H., Sumner, P., Tyson, T. & Westberry, G. 2000 Georgia muscadine production guide. Georgia Coop. Ext. Serv. Bull. 739, Athens, GA

  • Lane, R. 1997 Breeding muscadine and southern bunch grapes Fruit Var. J. 51 144 148

  • Loomis, N., Williams, C. & Murphy, M. 1954 Inheritance of flower types in muscadine grapes Proc. Amer. Soc. Hort. Sci. 64 279 283

  • Lorenz, D.H., Eichhorn, K.W., Bleiholder, H., Klose, R., Meier, U. & Weber, E. 1995 Growth stages of the grapevine: Phenological growth stages of the grapevine (Vitis vinifera L. ssp. vinifera)—codes and descriptions according to the extended BBCH scale Austral. J. Grape Wine Res. 1 100 110

    • Search Google Scholar
    • Export Citation
  • Lu, J., Schell, L. & Ramming, D.W. 2005 Interspecific hybridization between Vitis rotundifolia and Vitis vinifera and evaluation of the hybrids Acta Hort. 528 479 486

    • Search Google Scholar
    • Export Citation
  • NeSmith, D.S. 1999 Fruit set and berry size of ‘Fry’ muscadine grape in response to six pollen sources HortScience 34 355

  • Olien, W.C. 1990 The muscadine grape: Botany, viticulture, history, and current industry HortScience 25 732 739

  • Olmo, H.P. 1986 The potential role of (vinifera x rotundifolia) hybrids in grape variety improvement Experientia 42 921 926

  • Poling, B., Mainland, C., Bland, W., Cline, B. & Sorenson, K. 2003 Muscadine grape production guide. NC State Ext. Ser. Bul. AG-94. Raleigh, NC

  • Reimer, F. & Detjen, L. 1910 Self-sterility of the Scuppernong and other muscadine grapes. NC Agr. Exp. Sta. Bul. 209

  • Sampson, B., Noffsinger, S., Gupton, C. & Magee, J. 2001 Pollination biology of the muscadine grape HortScience 36 120 124

  • Scott, R.J. & Knott, M. 1974 A cluster analysis method for grouping means in the analysis of variance Biometrics 30 507 512

  • Sharafi, Y. & Bahmani, A. 2011 Pollen germination, tube growth and longevity in some cultivars of Vitis vinifera L Afr. J. Microbiol. Res. 5 1102 1107

    • Search Google Scholar
    • Export Citation
  • Srinivasan, C. & Mullins, M.G. 1981 Physiology of flowering in the grapevine—a review Amer. J. Enol. Viticult. 32 47 63

  • Subbarayadu, S., Naik, B.S., Devi, H.S., Bhau, B.S. & Khan, P.S.S.V 2014 Microsporogenesis and pollen formation in Zingiber officinale Roscoe Plant Syst. Evol. 300 619 632

    • Search Google Scholar
    • Export Citation
  • Vasconcelos, M.C., Greven, M., Winefield, C.S., Trought, M.C.T. & Raw, V. 2009 The flowering process of Vitis vinifera: A review Amer. J. Enol. Viticult. 60 411 434

    • Search Google Scholar
    • Export Citation
  • Williams, C.F. 1957 Relation of berry size to flower type of seedlings in muscadine grape crosses Proc. Amer. Soc. Hort. Sci. 69 254 260

All Time Past Year Past 30 Days
Abstract Views 14 0 0
Full Text Views 351 87 3
PDF Downloads 208 53 16