Abstract
Salvia is the largest genus in the Lamiaceae with more than 1000 species. The species S. coccinea used in this study has naturalized in the southeastern United States and is an important plant for pollinators. This project aimed to improve phenotypic characteristics of S. coccinea for use in the landscape by selecting for increased petal size and unique petal color. Two elite accessions were selected for hybridization using the pedigree method. One selection displayed compact habit with bicolored coral and white flowers, while the other was slightly larger with solid red flowers. Selections were made based on improved flower color and larger petal size. The breeding program achieved a 25% increase in petal width and a more vivid petal color for the coral bicolored selections. Additionally, a 60% increase in petal width was achieved for red flowers. These novel selections are attractive plants for the landscape, displaying improved ornamental value and supporting local pollinator populations.
Salvia is the largest genus in the Lamiaceae with more than 1000 accepted species. With an intercontinental native range, Salvia species can be found growing in Europe, Asia, Africa, Australia, and the Americas (Kew Science, 2019). Several members of the genus are desirable for their ornamental value and ability to attract pollinators. Salvia coccinea is a valuable plant for the landscape with red flowers that support both bees and hummingbirds. Its natural distribution is not well known but likely ranges from Mexico to Central America or Brazil (Wester and Claßen-Bockhoff, 2011). However, it has naturalized in the southeastern United States from South Carolina through Texas (Kew Science, 2019). The species flowers from spring through late summer and can be used as an annual or herbaceous perennial depending on its geographic location (Clebsch, 2003).
Salvia coccinea is often grown as an annual seed crop and is included in wildflower mixes for the southeastern United States (American Meadows, 2022; Clebsch, 2003). This species is visited by bees for its valuable nectar reserve but is primarily pollinated by hummingbirds. Nectar guides are usually not present but can exist as small white spots on the corolla with about 3 to 10 µL of nectar held at the base of the flower (Wester and Claßen-Bockhoff, 2011). The inflorescence appears in panicles with bilabiate corollas and continually blooms throughout the season. Although the flowers are typically red, petal color can vary from coral to white, including bicolored flowers. The species displays uniform branching with an average height of 1 m. The deltoid leaves have crenate margins and increase in size down the stem from 2.5 to 7.0 cm in length (Clebsch, 2003; Starr, 1985).
Salvia coccinea is a diploid species with 11 base chromosomes (2n = 2x = 22) (Alberto et al., 2003). A previous genetic study showed red flower pigmentation is dominant with pink-colored cultivars arising as a recessive trait (Kumar, 1943). Several selections have been made commercially available. The Summer Jewel series, developed by Takii & Co., Ltd. (Kyoto, Japan) includes red, pink, white, and lavender flowering cultivars with compact plant habits. These were winners of All-America Selections marketed as displaying large, 1.3-cm blooms (All-America Selections, 2021). Other cultivars have been selected for compact growth, such as Brenthurst with bicolored white and coral petals. However, limited breeding work has been done to improve the species further.
To describe changes in observable flower color, samples must be compared with a reliable standard (Griesbach and Austin, 2005). In a study conducted by Tucker et al. (1991), the RHS color chart was recommended as the best option, based on ease of use. Although RHS values can be used as a standard to report the color of biological samples, variation in hues can be more easily described numerically by using the CIELAB color space (International Commission on Illumination, 2004). In this model, L* describes the lightness of a color (black to white) on a scale of 0 to 100. The other two parameters, a* and b*, describe mutually exclusive zones of color. Red and green are described by a*, where red is a positive value and green is negative. Similarly, b* represents yellow or blue, where yellow is positive and blue is negative (Luo, 2015). These three scales of lightness, red to green and yellow to blue, are considered independent of one another and can be used in combination to describe color space (Krauskopf et al., 1982).
Although S. coccinea has not been extensively bred, interspecific hybrids have been achieved between other Salvia species. In a study conducted by Tychonievich and Warner (2011), nine species of Salvia were crossed to determine hybridization ability. Due to autogamy in many Salvia species, the seed parent was emasculated before anther dehiscence. Pollination was achieved by touching the pollen parent’s anther to the seed parent’s stigma. All self-pollinated species successfully developed seeds using this method. However, only a few interspecific crosses resulted in seed set (Tychonievich and Warner, 2011). Therefore, intraspecific hybridization should be a reliable method to develop improved S. coccinea.
Intraspecific hybridization is a breeding tool used to improve the value of many plant species. For example, inbreeding depression was observed in self-pollinated Andrographis paniculata, leading to its classification as an endangered species. However, through intraspecific hybridization, researchers were able to increase the genetic diversity of A. paniculata and subsequently support conservation efforts (Valdiani et al., 2012). This breeding technique has also been used to increase biomass production in Chamaecrista fasciculata (Erickson and Fenster, 2006), induce variation in flower size in Plumbago auriculata (Chen et al., 2021), and enhance performance and disease resistance in Poa pratensis (Pepin and Funk, 1971).
In this study, intraspecific hybridization was used to improve S. coccinea for use in the landscape. Because interspecific hybridization has limited success and not much breeding work has been done in the species, two morphologically distinct accessions of S. coccinea were selected to introduce variability into the gene pool (Tychonievich and Warner, 2011). The objective of the current study was to make selections for larger flowers, unique petal color, and a compact growth habit.
Materials and Methods
To develop an improved hybrid line, an accession of S. coccinea with a compact habit and coral bicolored petals was collected from a wild population growing on Amelia Island in Nassau County, FL (30.631388, –81.468611). Two red accessions were trialed with the coral selection for use in the breeding program. The common species, also known as scarlet sage, was obtained from Trade Winds Fruit (Santa Rosa, CA). The other, Yucatan, was a larger cultivar from Salvias.com.ar (Buenos Aires, Argentina). These three selections were grown from seed in the Trial Gardens at the University of Georgia in Athens, GA (33.944507, –83.375774) to observe overall growth habits in Summer 2019. The common scarlet sage was selected to hybridize with the coral accession based on its compact, well-branched structure.
Vegetative cuttings of both plants were taken to maintain these selections during the winter in a greenhouse at the University of Georgia Horticulture Research Farm (33.886045–83.420179). Cuttings were made by selecting 15- to 20-cm segments of terminal growth from the parent plants. All leaves were stripped off the bottom of the cutting leaving one set of leaves at the top. The cuttings were dipped in a 1000 ppm KIBA solution then left to air dry for 10 min. The treated cuttings were stuck in a 3:1 perlite: PRO-MIX high porosity substrate with biofungicide and mycorrhizae (Premier Tech Horticulture, Quakertown, PA) in 280-mL Square Deep Vacuum pots (HC Companies, Twinsburg, OH). The prepared trays were placed on a bench with 70% shade, bottom heat, and misted for 5 s every 5 min from 7 am to 7 pm. The day/night greenhouse conditions were set to 25/20 °C and 40%/30% humidity.
Once cuttings had developed a strong root system, they were removed from mist but kept under shade for 2 weeks before moving into full sun. The pots were fertilized with 2.5 g of 8- to 9-month Osmocote Plus 15–9–12 (15N–4.4P–10.0K) (ICL Specialty Fertilizers, Summerville, SC). The liners were then moved into C300 2.8-L pots (Nursery Supplies Inc., Kissimmee, FL) with substrate consisting of 20% peatmoss, 28% 9.5 mm aged pine bark, 42% 15.9 mm aged pine bark, and 10% sand (Old Castle, Shady Dale, GA) and fertilized with 10.5 g of Osmocote Plus 15–9–12. Plants were also liquid fed once per week with 200 ppm Jack’s 20–10–20 Peat Lite (20N–4.9P–16.6K) (JR Peters Inc., Allentown, PA). A foliar spray of Venom (dinotefuran) (Valent U.S.A LLC, San Ramon, CA) at a rate of 0.3 g/L of water was used in conjunction with Pyrethrum TR (BASF, Ludwigshafen, Germany) at a rate of 1 can/279 m2 to reduce populations of southern pink moth Pyrausta inornatalis (Crambidae).
Since red is the dominant flower color, the coral flowered accession was selected as the female parent to easily confirm hybridization in the F1 population (Kumar, 1943). In January of 2020, controlled crosses were made between the two accessions, using the breeding technique described by Tychonievich and Warner (2011). The female parent was emasculated 1 d before anthesis to prevent autogamy. At this stage in floral development, the petals were still tightly encasing the reproductive structures and the anthers had not dehisced. Once the style had fully extended from the female flower, mature anthers were collected from the male parent for pollination. Using forceps, pollen was transferred from the male anther to the stigma of the female parent. Primary pollinators were considered excluded from plants in the greenhouse, so the inflorescences were not bagged after pollination.
Salvia flowers contain four ovules resulting in a maximum of four seeds produced per flower (Visco and Capon, 1970). In this study, the seed set was high with most flowers setting four seeds per flower. Seeds were harvested when the calyx had dried, and the seedcoat had become visibly brown. Thirty-eight hybrid seeds were individually sown on the surface of the PRO-MIX high porosity substrate in a 200-cell (22-mL) plug tray to form the F1 population. Trays were irrigated from above and kept consistently moist on a 70% shade bench maintained under the same greenhouse conditions previously mentioned.
Once seedlings had three sets of true leaves, they were transplanted into 280-mL square pots using the same substrate. They were fertilized with 2.5 g of 8- to 9-month Osmocote Plus 15–9–12 and moved into full sun. Before transplanting to the field, final selections from the F4 generation were moved into C300 2.8-L pots (Nursery Supplies Inc., Kissimmee, FL) using the Old Castle substrate previously described and fertilized with 10.5 g of Osmocote Plus 15–9–12.
The F1 population was morphologically uniform, so no selections were made from this generation. Instead, each plant was allowed to self-pollinate, and 676 seeds were collected to form the F2 generation. The pedigree selection method was used in which each plant from the F2 was kept as a separate breeding line, and selections were made based on phenotypic traits (Love, 1927). Selections were made in the F2 and subsequent populations based on petal size, petal color, and the number of nodes to the first flower. In the F2, 40 plants were maintained in the breeding program. The pedigree breeding method was continued in the F3 and F4 generations to select plants with larger flowers, improved petal color, and compact growth from each breeding line. The number of seedlings grown out and selected in each generation is summarized in a simplified schematic in Fig. 1.
Red and coral accessions of Salvia coccinea were hybridized to select for improved petal color, flower size, and compactness. The F1 was morphologically uniform, so no selections were made in this generation. The population size was bulked up for the F2 generation, then selections were made for large flowers with vivid petal color. This process was repeated through the F4 making selections for both red and coral accessions.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Red and coral accessions of Salvia coccinea were hybridized to select for improved petal color, flower size, and compactness. The F1 was morphologically uniform, so no selections were made in this generation. The population size was bulked up for the F2 generation, then selections were made for large flowers with vivid petal color. This process was repeated through the F4 making selections for both red and coral accessions.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Red and coral accessions of Salvia coccinea were hybridized to select for improved petal color, flower size, and compactness. The F1 was morphologically uniform, so no selections were made in this generation. The population size was bulked up for the F2 generation, then selections were made for large flowers with vivid petal color. This process was repeated through the F4 making selections for both red and coral accessions.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
One selection criterion, flower color, was used to isolate plants with improved petal pigmentation compared with the parents. Flower color was determined by comparing the upper and lower lobes to the RHS color chart (Royal Horticultural Society, 2001). The RHS color sample was scanned with a color sensor to numerically describe variation in color using the CIELAB color space (Nix Mini 2 Color Sensor, Nix Sensor Ltd., Hamilton, Ontario, Canada). The bicolored coral parent had the RHS color 56D (L* = 89, a* = 8, b* = 3) on its upper lobe and a lower lobe color of 43D (L* = 70, a* = 42, b* = 20). The red parent had the same upper and lower lobe color with an RHS value of 44B (L* = 47, a* = 64, b* = 50). Individuals from this breeding program were selected for more vibrant coral coloration.
Another selection criterion, petal size, was the measured distance from one edge of the flower’s lower lobe to the opposite edge at the widest point. The petal lobe width was 1.3 cm at the widest point for the red parent and 1.6 cm for the bicolored coral parent. Any progeny with a petal width ≤1.6 cm was discarded. Finally, selections were made for a more compact plant by estimating final height based on growth in the 280-mL pots. The number of nodes from the substrate line to the first inflorescence was used to distinguish plants with the same size and color flowers. When a decision was made based on the number of nodes to the first flower, a plant with eight or fewer nodes was selected to remain in the breeding program.
In-ground beds were prepared for trialing the elite selections at the same research farm (33.886811, –83.419844) by forming three rows 30 m in length and 0.9 m in width. The planting area comprised Cecil soil, a well-draining soil with a surface layer of sandy clay loam and deep red clay subsoil (Robertson, 1968). Preplant fertilizer was applied at a rate of 9 kg N/ha using All Season’s Lawn and Garden Fertilizer 10–10–10 (10N–4.4P–8.3K) (Tri County Fertilizer and Specialty Co., Inc., Honea Path, SC). A single line of drip tape with emitters every 30 cm (Rivulis Irrigation Ltd., San Diego, CA) was laid on each bed for controlled irrigation. TIF Total Blockade plastic mulch (Berry Global Inc., Evansville, IN), 1.5 m wide and 1.25 mm thick, was applied to each row with a Model 2600 Series II raised bed plastic mulch layer (Rian-Flo Irrigation, East Earl, PA) to minimize weed pressure.
By June 2021, the elite selections were well established in the 2.8-L pots and were planted in the prepared field rows. Plants were spaced 1.2 m within the row and 0.9 m between rows. The plants were automatically irrigated in the morning for 2 h three times per week using the drip irrigation system. After 3 months in the field, total height was measured to determine the plant’s mature size. Final selections were made based on overall plant form, flower size, and flower color.
The 3:1 ratio of red to coral flowers in the F2 was analyzed with a chi-square test. The mean and standard deviation were determined for the distribution of lower lobe width and number of nodes to first flower. Finally, relationship between measured parameters was assessed for each of the field-trialed plants using the linear regression function in RStudio. These analyses were completed using the R programming language (RStudio, Boston, MA).
Results and Discussion
Hybridization of the tall, red S. coccinea accessions with the compact, bicolored coral accession successfully generated improved intraspecific hybrids. The red-petaled progeny had larger flowers than the original red parent. For the coral-petaled progeny, a more vivid color was generated in addition to an increase in flower size. Compact growth habit was achieved for both red and coral selections.
Uniform flower color was observed in the F1 population, with all 38 plants having red petals. Because the female parent had coral-colored flowers, the occurrence of all red progeny confirmed a successful cross. In the subsequent generation, the observance of 517 red-flowered plants: 155 coral-flowered plants (Fig. 2) indicated a strong 3:1 ratio of red to coral flowers. This ratio was confirmed with a chi-square test indicating that flower color is inherited in a Mendelian fashion, with red expressing complete dominance χ2 (1, N = 672) = 1.34, P > 0.05. These findings are consistent with observations made by Kumar (1943) on intraspecific hybrids of S. coccinea.
One red and one coral accession of Salvia coccinea were hybridized to form an improved selection. A 3:1 ratio of red to coral flowers was observed in the F2 population, χ2 (1, N = 672) = 1.34, P > 0.05.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
One red and one coral accession of Salvia coccinea were hybridized to form an improved selection. A 3:1 ratio of red to coral flowers was observed in the F2 population, χ2 (1, N = 672) = 1.34, P > 0.05.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
One red and one coral accession of Salvia coccinea were hybridized to form an improved selection. A 3:1 ratio of red to coral flowers was observed in the F2 population, χ2 (1, N = 672) = 1.34, P > 0.05.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Although the F1 population was morphologically uniform, one plant from this generation did not set seed in the greenhouse. This selection did not appear to produce any pollen and was unable to be bred through later generations. However, this plant was maintained in the breeding program to determine fertility in the field trial. Despite all other S. coccinea selections setting seed in the field, this selection did not set seed even after pollinator visitation. Additionally, it could not be induced to set seed by hand-crossing. Therefore, this selection is suspected to be both male and female sterile, which has not previously been reported in S. coccinea. This selection was maintained through vegetative cuttings for further testing and potential release as a sterile cultivar.
No variation in petal color was observed in the red-petaled progeny. However, in the F2 and subsequent generations, wide variation was observed in the bicolored progeny from pale pink to deep coral. Therefore, progeny with a deeper orange color than the parent were considered elite selections to be maintained in the breeding program. All a* and b* are positive, indicating that flower color is represented by a combination of red and yellow alone according to the CIELAB color system. The measured RHS color values and associated CIELAB values for the elite coral-colored progeny are summarized in Table 1. The difference in upper and lower lobe color between parents and progeny is illustrated in Fig. 3.
Hybridization of a red, male accession of Salvia coccinea with a bicolored coral, female accession resulted in a distribution of observed petal colors. (A) The color of the upper petal of the male (bottom left sphere) and female (top right sphere) relative to selected coral progeny. (B) The color of the lower petal of the male (bottom left sphere) and female (top right sphere) relative to selected coral progeny. Values are given in the CIELAB color space.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Hybridization of a red, male accession of Salvia coccinea with a bicolored coral, female accession resulted in a distribution of observed petal colors. (A) The color of the upper petal of the male (bottom left sphere) and female (top right sphere) relative to selected coral progeny. (B) The color of the lower petal of the male (bottom left sphere) and female (top right sphere) relative to selected coral progeny. Values are given in the CIELAB color space.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Hybridization of a red, male accession of Salvia coccinea with a bicolored coral, female accession resulted in a distribution of observed petal colors. (A) The color of the upper petal of the male (bottom left sphere) and female (top right sphere) relative to selected coral progeny. (B) The color of the lower petal of the male (bottom left sphere) and female (top right sphere) relative to selected coral progeny. Values are given in the CIELAB color space.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
One red and one coral accession of Salvia coccinea were hybridized to select for larger flowers with more vivid coral petal color. The lower lobe width of elite selections from the F2, F3, and F4 populations are given with their RHS Color Chart and CIELAB values.
For the upper lobe of the bicolored parent, the CIELAB values were L* = 89, a* = 8, b* = 3. The first parameter, L* = 89 indicates the upper lobe of the parent had a significant amount of white in the pigment. Comparatively, the upper lobe of the red parent had CIELAB values of L* = 47, a* = 64, b* = 50. All selected progeny listed in Table 1 have L* values between the two parents, indicating their pigmentation was darker than the bicolored parent, but lighter than the red parent. With a* and b* values of 8 and 3, respectively, the bicolored parent was found to have minimal red or yellow coloration. However, the red parent had comparatively high values with a* = 64 and b* = 50. All selected coral progeny had high a* values indicating an increase in red compared with the bicolored parent, with variations in yellow indicated by the b* value.
For the lower lobe of the bicolored parent, the CIELAB values were L* = 70, a* = 42, b* = 20. Comparatively, these values were L* = 47, a* = 64, b* = 50 for the red parent. Thus, when looking at the progeny, it was evident that they had lightness values between the two parents. Additionally, the elevation of both a* and b* values compared with the bicolored parent resulted in a more orange color in the progeny. Overall, the lower lobe of selected progeny was coral in color with intermediate darkness between the two parents.
The second measured trait, petal width, was also uniform in the F1, with the entire population having a lower lobe width of 1.6 cm. In the F2, however, petal width varied, as shown in Fig. 4. The median lower lobe width was 1.6 with a mean value of 1.6 ± 0.1 cm. The red and coral parents had a lower lobe width of 1.4 and 1.6 cm, respectively. Therefore, a mean value of 1.6 cm in the F2 indicates the average population had a lobe width equivalent to the female parent. However, the lobe width of some progeny exceeded the width of either parent. The red-flowered progeny ranged in size from 1.2 to 2.0 cm, and selections were made to maintain flowers 1.6 cm or larger. The coral-flowered progeny had lower lobe widths of 1.2 to 1.8 cm in size. Because the coral-colored progeny also showed variance in petal color, selections were made for improved pigmentation and size.
Two accessions of Salvia coccinea were hybridized to select for larger flowers. A symmetrical distribution of petal lobe width was observed in the F2 population. The average lobe width was 1.6 cm with a standard deviation of 0.1 cm.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for larger flowers. A symmetrical distribution of petal lobe width was observed in the F2 population. The average lobe width was 1.6 cm with a standard deviation of 0.1 cm.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for larger flowers. A symmetrical distribution of petal lobe width was observed in the F2 population. The average lobe width was 1.6 cm with a standard deviation of 0.1 cm.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
By the F4 generation, size improvements had been made in both the coral and red flowers. A side-by-side comparison of the original bicolored parent (Fig. 5A) and an elite coral-colored selection (Fig. 5B) from the F4 is shown in Fig. 5. With a 2.0-cm bottom lobe width, elite coral selections had up to a 25% increase in size compared with the 1.6-cm parent. Because selections were not limited by flower color, an even more significant improvement in lobe width was achieved for the red flowers. Figure 6A is the original red parent with a bottom lobe width of 1.3 cm. A 60% increase in petal size (Fig. 6B) was achieved through the breeding program by selecting red flowers up to 2.1 cm in size.
One red and one coral accession of Salvia coccinea were hybridized to select for larger flowers with more vivid coral petal color. (A) Original bicolored, female parent with a bottom lobe width of 1.6 cm. The female parent had CIELAB values of L* = 89, a* = 8, b* = 3 for the upper lobe and L* = 70, a* = 42, b* = 20 for the lower lobe. (B) The improved hybrid (AD9-2-5) from the F4 population with a 2.0-cm bottom lobe width. The CIELAB values were L* = 69, a* = 40, b* = 2 for the upper lobe and L* = 57, a* = 56, b* = 37 for the lower lobe.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
One red and one coral accession of Salvia coccinea were hybridized to select for larger flowers with more vivid coral petal color. (A) Original bicolored, female parent with a bottom lobe width of 1.6 cm. The female parent had CIELAB values of L* = 89, a* = 8, b* = 3 for the upper lobe and L* = 70, a* = 42, b* = 20 for the lower lobe. (B) The improved hybrid (AD9-2-5) from the F4 population with a 2.0-cm bottom lobe width. The CIELAB values were L* = 69, a* = 40, b* = 2 for the upper lobe and L* = 57, a* = 56, b* = 37 for the lower lobe.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
One red and one coral accession of Salvia coccinea were hybridized to select for larger flowers with more vivid coral petal color. (A) Original bicolored, female parent with a bottom lobe width of 1.6 cm. The female parent had CIELAB values of L* = 89, a* = 8, b* = 3 for the upper lobe and L* = 70, a* = 42, b* = 20 for the lower lobe. (B) The improved hybrid (AD9-2-5) from the F4 population with a 2.0-cm bottom lobe width. The CIELAB values were L* = 69, a* = 40, b* = 2 for the upper lobe and L* = 57, a* = 56, b* = 37 for the lower lobe.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for larger flowers. (A) The original red, male parent with a bottom lobe width of 1.3 cm. (B) The improved selection (AK10-4-3) from the F4 population with a 2.1-cm bottom lobe width.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for larger flowers. (A) The original red, male parent with a bottom lobe width of 1.3 cm. (B) The improved selection (AK10-4-3) from the F4 population with a 2.1-cm bottom lobe width.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for larger flowers. (A) The original red, male parent with a bottom lobe width of 1.3 cm. (B) The improved selection (AK10-4-3) from the F4 population with a 2.1-cm bottom lobe width.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Selections were made primarily on petal size and color. However, the number of nodes to the first flower was used as the final selection criterion. This measurement was used to project the mature height of a plant and make selections for compactness. In the F2 population, the number of nodes to the first flower varied widely, as shown in Fig. 7. The average number of nodes to first flower was nine nodes in the F2 with a standard deviation of one node. When a decision had to be made between plants with indistinguishable flowers, plants with eight or fewer nodes were selected to remain in the breeding program. However, final height measurements taken after the 3-month growth in the field indicated no relationship between height and number of nodes to the first flower (R2 = 0.09; Fig. 8). Additionally, no relationship was identified between final height and lower petal lobe width (R2 = 0.04; Fig. 9). Final selections were made for plants with large flowers, improved petal color, and compact habit in the field (Table 2).
Two accessions of Salvia coccinea were hybridized to select for plants with compact habit. To estimate final plant height, the number of nodes from the substrate line to the first flower were counted. A symmetrical distribution of the number of nodes to the first flower was observed in the F2 population. The average number of nodes to first flower was nine nodes with a standard deviation of one node.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for plants with compact habit. To estimate final plant height, the number of nodes from the substrate line to the first flower were counted. A symmetrical distribution of the number of nodes to the first flower was observed in the F2 population. The average number of nodes to first flower was nine nodes with a standard deviation of one node.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for plants with compact habit. To estimate final plant height, the number of nodes from the substrate line to the first flower were counted. A symmetrical distribution of the number of nodes to the first flower was observed in the F2 population. The average number of nodes to first flower was nine nodes with a standard deviation of one node.
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for plants with compact habit. The number of notes from the substrate level to first flower was used to predict final plant height. However, comparison of the number of nodes to the first flower with mature height of Salvia coccinea hybrids indicated no relationship exists between these two parameters (R2 = 0.09).
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for plants with compact habit. The number of notes from the substrate level to first flower was used to predict final plant height. However, comparison of the number of nodes to the first flower with mature height of Salvia coccinea hybrids indicated no relationship exists between these two parameters (R2 = 0.09).
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for plants with compact habit. The number of notes from the substrate level to first flower was used to predict final plant height. However, comparison of the number of nodes to the first flower with mature height of Salvia coccinea hybrids indicated no relationship exists between these two parameters (R2 = 0.09).
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for plants with compact habit. Flower size was observed to vary among the intraspecific hybrids. However, comparison of the lower petal lobe width with mature height of hybrids indicated no relationship exists between these two parameters (R2 = 0.04).
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for plants with compact habit. Flower size was observed to vary among the intraspecific hybrids. However, comparison of the lower petal lobe width with mature height of hybrids indicated no relationship exists between these two parameters (R2 = 0.04).
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
Two accessions of Salvia coccinea were hybridized to select for plants with compact habit. Flower size was observed to vary among the intraspecific hybrids. However, comparison of the lower petal lobe width with mature height of hybrids indicated no relationship exists between these two parameters (R2 = 0.04).
Citation: HortScience 57, 9; 10.21273/HORTSCI16629-22
One red and one coral accession of Salvia coccinea were hybridized to select for plants with improved petal color, larger flowers, and compact habit. Elite selections are summarized with their RHS Color Chart values, width of the lower petal lobe, and height and width of mature plants (Royal Horticultural Society, 2001).
The number of nodes to the first flower did not correlate with plant height in S. coccinea; however, this is not the case for other species. In previous research, total plant height and number of nodes were found to increase together in Ocimum basilicum and Tecoma stans (Torres and Lopez, 2011; Walters and Currey, 2019). In both studies, the number of nodes was a positive predictor of plant height, across different environmental conditions. However, the number of nodes to the first flower did not correlate with plant height in S. coccinea, and therefore, it is not a reliable way to select for compact growth.
Future research may improve the selections in at least two ways. The first objective would be to develop a selection with reduced fertility or self-incompatibility to inhibit self-seeding in the landscape. Dense populations of S. coccinea have spread from gardens to uncultivated areas of South Africa, leading researchers to believe it has become invasive (Moshobane et al., 2020). Intraspecific hybridization is believed to contribute to the invasive spread of species, including Phalaris arundinacea (Lavergne and Molofsky, 2007), Pyrus calleryana (Culley and Hardiman, 2009), and Schinus terebinthifolius (Williams et al., 2005). For example, Pyrus calleryana is self-incompatible but readily sets seeds and forms thickets when crossed with cultivars of the same species (Culley and Hardiman, 2009). Although S. coccinea is beneficial to pollinators and is not currently on a list of invasive species for the United States, its ability to reseed may be undesirable.
A second objective for future research would be to create resistance to the southern pink moth, Pyrausta inornatalis (Crambidae). The larvae of the southern pink moth bore into and feed on unopened flowers exclusively in the genus Salvia (Oregon Department of Agriculture, 2021). Moths of the family Crambidae have successfully been repelled through treatment with the essential oil of Salvia officinalis (Göttig et al., 2017). Essential oil content is low in S. coccinea. However, increasing essential oil production in the species may deter southern pink moth with an environmental benefit and reduced production cost.
Essential oil production is related to the ploidy level. For example, chromosomes in S. leriifolia were induced to double from n = 11 to n = 22 using colchicine. As a result, while the diploid species contained 14 distinct essential oils, the tetraploid developed with eight additional compounds (Estaji et al., 2017). An increase in aromatic compounds was also reported in S. muticaulis (Tavan et al., 2022) and other polyploid-induced plants (Dhawan and Lavania, 1996). Therefore, the quantity and diversity of essential oils may be increased by inducing polyploidy in S. coccinea.
By crossing two accessions of Salvia coccinea, improved selections were isolated with characteristics superior to either parent. A 25% increase in petal size was achieved for coral petaled flowers, and a 60% increase was achieved for red-petaled flowers. Additionally, deeper orange, bicolored petals were isolated from this breeding program. From the field trial, plants were found to have a uniform shape and compact habit. Overall, the selections from this breeding program bring improved growth habit, larger flowers, and a more vivid color pallet to the available S. coccinea. These selections can be used as attractive flowering plants in the landscape to support local pollinator populations and are available for commercial release by vegetative propagation.
Literature Cited
Alberto, C.M., Sanso, A.M. & Xifreda, C.C. 2003 Chromosomal studies in species of Salvia (Lamiaceae) from Argentina J. Linn. Soc. Bot. 141 483 490 https://doi.org/10.1046/j.1095-8339.2003.t01-1-00178.x
All-America Selections 2021 Complete AAS winner list 29 July 2021. https://all-americaselections.org/complete-aas-winner-list/
American Meadows 2022 Southeast wildflower seed mix 25 Apr. 2022. https://www.americanmeadows.com/wildflower-seeds/wildflower-mix/southeast-wildflower-seed-mix
Chen, X., Gao, S., Shen, P., Liu, Y., Lei, T., Shi, L., Li, W., Li, Y., Yu, X., Yang, L. & Li, J. 2021 Genetic diversity analysis of intraspecific hybridization between Plumbago auriculata and Plumbago auriculata f. alba based on horticultural traits and molecular markers Acta Physiol. Plant. 43 31 https://doi.org/10.1007/s11738-020-03188-9
Clebsch, B 2003 The new book of Salvias: Sages for every garden 2nd ed. Timber Press Portland, OR
Culley, T.M. & Hardiman, N.A. 2009 The role of intraspecific hybridization in the evolution of invasiveness: A case study of the ornamental pear tree Pyrus calleryana Biol. Invasions 11 1107 1119 https://doi.org/10.1007/s10530-008-9386-z
Dhawan, O.P. & Lavania, U.C. 1996 Enhancing the productivity of secondary metabolites via induced polyploidy: A review Euphytica 87 81 89 https://doi.org/10.1007/BF00021879
Erickson, D.L. & Fenster, C.B. 2006 Intraspecific hybridization and the recovery of fitness in the native legume Chamaecrista fasciculata Evolution 60 225 233 https://doi.org/10.1111/j.0014-3820.2006.tb01101.x
Estaji, A., Hosseini, A.B., Ravandi, E.G., Dehghan, E. & Sefidkon, F. 2017 The effects of colchicine-induced autotetraploidy on selected characteristics of Nuruozak (Salvia leriifolia) Cytol. Genet. 51 74 81 https://doi.org/10.3103/S0095452717010042
Göttig, S., Korn, S. & Herz, A. 2017 Repellent and toxic properties of plant oils and extracts on Cydalima perspectalis Walker (Lepidoptera: Crambidae) Arch. Phytopathol. Pflanzenschutz 50 658 673 https://doi.org/10.1080/03235408.2017.1358506
Griesbach, R.J. & Austin, S. 2005 Comparison of the Munsell and Royal Horticultural Society’s color charts in describing flower color Taxon 54 771 773 https://doi.org/10.2307/25065433
International Commission on Illumination 2004 CIE 15: Technical report: Colorimetry 3rd ed. 21 Nov. 2021. https://www.cdvplus.cz/file/3-publikace-cie15-2004/
Kew Science 2019 Plants of the world online Salvia L. Royal Botanic Gardens, Kew 14 Jan. 2021. https://powo.science.kew.org/taxon/urn:lsid:ipni.org:names:30000096-2
Krauskopf, J., Williams, D.R. & Heeley, D.W. 1982 Cardinal directions of color space Vision Res. 22 1123 1131 https://doi.org/10.1016/0042-6989(82)90077-3
Kumar, L.S.S 1943 Inheritance of flower color in Salvia coccinea Juss ex Murr. Proc. India Acad. Sci. Sect. B. 17 37 40
Lavergne, S. & Molofsky, J. 2007 Increased genetic variation and evolutionary potential drive the success of an invasive grass Proc. Natl. Acad. Sci. USA 104 3883 3888 https://doi.org/10.1073/pnas.0607324104
Love, H.H 1927 A program for selecting and testing small grains in successive generations following hybridization J. Am. Soc. Agron. 19 705 712 https://doi.org/10.2134/agronj1927.00021962001900080002x
Luo, M.R 2015 Encyclopedia of color science and technology Springer Berlin, Germany https://doi.org/10.1007/978-3-642-27851-8_11-1
Moshobane, M.C., Nelufule, T., Shivambu, T.C. & Shivambu, N. 2020 Occurrence record of and possible invasion by Scarlet Sage (Salvia coccinea Buc’hoz ex Etl.) in South Africa Russ. J. Biol. Invasions 11 393 398 https://doi.org/10.1134/s2075111720040098
Oregon Department of Agriculture 2021 Pest alert: Southern pink moth Pyrausta inornatalis 28 July 2021. https://www.oregon.gov/
Pepin, G.W. & Funk, C.R. 1971 Intraspecific hybridization as a method of breeding Kentucky bluegrass (Poa pratensis L.) for turf Crop Sci. 11 445 448 https://doi.org/10.2135/cropsci1971.0011183X001100030042x
Robertson, S.M 1968 Soil survey of Clarke and Oconee counties, Georgia United States Department of Agriculture Athens, GA
Royal Horticultural Society 2001 RHS color chart 4th ed. Royal Horticultural Society London United Kingdom
Starr, G 1985 New World Salvias for cultivation in southern Arizona Desert Plants 7 167 207
Tavan, M., Azizi, A., Sarikhani, H., Mirjalili, M.H. & Rigano, M.M. 2022 Induced polyploidy and broad variation in phytochemical traits and altered gene expression in Salvia multicaulis Scientia Hort. 291 https://doi.org/10.1016/j.scienta.2021.110592
Torres, A.P. & Lopez, R.G. 2011 Photosynthetic daily light integral during propagation of Tecoma stans influences seedling rooting and growth HortScience 46 282 286 https://doi.org/10.21273/HORTSCI.46.2.282
Tucker, A.O., Maciarello, M.J. & Tucker, S.S. 1991 A survey of color charts for biological descriptions Taxon 40 201 214 https://doi.org/10.2307/1222974
Tychonievich, J. & Warner, R.M. 2011 Interspecific crossability of selected Salvia species and potential use for crop improvement J. Amer. Soc. Hort. Sci. 136 41 47 https://doi.org/10.21273/JASHS.136.1.41
Valdiani, A., Kadir, M.A., Saad, M.S., Talei, D. & Tan, S. 2012 Intra-specific hybridization: Generation of genetic diversification and heterosis in Andrographis paniculata Nees. A bridge from extinction to survival Gene 505 23 36 https://doi.org/10.1016/j.gene.2012.05.056
Visco, F.J. & Capon, B. 1970 Pollination mechanisms in three species of Salvia native to southern California Aliso J. Systematic Evolution. Bot. 7 231 242 doi: 10.5642/aliso.19700702.11
Walters, K.J. & Currey, C.J. 2019 Growth and development of basil species in response to temperature HortScience 54 1915 1920 https://doi.org/10.21273/HORTSCI12976-18
Wester, P. & Claßen-Bockhoff, R. 2011 Pollination syndromes of new world Salvia species with special reference to bird pollination Ann. Mo. Bot. Gard. 98 101 155 https://doi.org/10.3417/2007035
Williams, D.A., Overholt, W.A., Cuda, J.P. & Hughes, C.R. 2005 Chloroplast and microsatellite DNA diversities reveal the introduction history of Brazilian peppertree (Schinus terebinthifolius) in Florida Mol. Ecol. 14 3643 3656 https://doi.org/10.1111/j.1365-294x.2005.02666.x
