Plant material.

Three blotched (‘Carolyn Whorton’, ‘White Christmas’, and ‘Florida Blizzard’) and six nonblotched (‘Candidum’, ‘Fannie Munson’, ‘Frieda Hemple’, ‘Gingerland’, ‘Miss Muffet’, and ‘Rosebud’) commercial cultivars were used as parents in 20 crosses. ‘Carolyn Whorton’ and ‘White Christmas’ have pink and white blotches, respectively. ‘Florida Blizzard’ is a progeny of a cross between ‘Aaron’ (nonblotched) and ‘White Christmas’ (Harbaugh et al., 2002) and has inherited the blotching trait from ‘White Christmas’. ‘Candidum’, ‘Fannie Munson’, ‘Frieda Hemple’, ‘Gingerland’, ‘Miss Muffet’, and ‘Rosebud’ are all nonblotched, but their leaves are in different colors (white with green veins, pink, red, white with red spots, lemon green with burgundy spots, and pink, respectively). All of these cultivars have been asexually propagated for many years (except ‘Florida Blizzard’). Most of them are among the 10 most popular commercial cultivars (Bell et al., 1998; Deng et al., 2005) and have been used frequently as breeding parents for cultivar development. Their phenotype and inferred genotype for leaf blotching and vein color are shown in Table 1.

Table 1.

Phenotype and genotype (inferred) of nine commercial caladium (Caladium ×hortulanum) cultivars used as parents for crosses in this study.

View Table

Flower induction, pollination, and progeny growing.

Large, dry tubers (6.4 cm or greater in diameter) were treated with a GA₃ solution (ProGibb T&O; Valent BioSciences, Libertyville, IL; 600 mg·L⁻¹) overnight at ambient temperature to induce flowers (Harbaugh and Wilfret, 1979). Treated tubers were planted in plastic pots (20 cm in diameter, 3.5 L in volume) filled with VerGro container mix A (Verlite Company, Tampa, FL) in late Aug. 2004. Plants were maintained in a glasshouse with 20% to 30% light exclusion and temperatures between 15 and 32 °C. Pollination was performed between 25 Oct. and 14 Nov. 2004 following the procedure of Deng and Harbaugh (2004). Seed heads were harvested in Dec. 2004 and Jan. 2005. Seeds were extracted and sown in Dec. 2004 and Jan. 2005 onto VerGrow container mix A (Verlite) in 200-cell trays and germinated at ≈21 °C under continuous cool fluorescent light (30 μmol·m⁻²·s⁻¹). Between 14 Feb. and 1 Mar. 2005, seedlings were transferred to six-pack plastic cells filled with VerGrow container mix A (Verlite) amended with Nutricote 13-13-13 (13N–6.4P–10.8K; The Scotts Co., Marysville, OH). Seedlings were grown in a greenhouse with 20% to 30% light exclusion, a photosynthetic photon flux 600 to 800 μmol·m⁻²·s⁻¹, and temperatures between 15 and 30 °C. Progeny in the six-pack cells were planted on raised ground beds in June 2005. The soil was EauGallie fine sand with ≈1% organic matter and a pH of 6.2 to 6.8. The beds were fumigated with methyl bromide and chloropicrin, covered with white-on-black plastic mulch, and irrigated with the seepage irrigation system (Geraldson et al., 1965). Plants were spaced 25.4 cm apart and grown on the beds for three seasons until July 2007. In each growing season, 15 g of Osmocote controlled-release fertilizer (18N–2.6P–10K, 8- to 9-month; Scotts Co.) was applied to each plant.

Progeny of four crosses (or four populations) were also propagated asexually through tuber division and planted in four replicates in randomized complete blocks to assess potential variation in leaf characteristics after asexual propagation. Tuber pieces were planted in new fumigated, mulched ground beds between 3 May and 13 June 2007. These plants were irrigated and fertilized through a drip system.

Data-taking and analysis.

Progeny was scored for leaf blotching (blotched or nonblotched) and main vein color (red, white, or green) in June to Nov. 2006 and again in June 2007. Each time, multiple fully expanded, mature leaves were examined for each progeny. Under high levels of light in the field or in the greenhouse, some white-veined leaves tended to develop some level of coloration. In these cases, young leaves before unfurling were examined to confirm vein color. Progeny of the four populations that were asexually propagated and replanted in 2007 were examined in Oct. 2007. Consistent phenotypes were observed among the four replicates of each progeny.

χ² tests for goodness of fit were performed to compare observed ratios with expected ratios for the segregation of leaf blotching, whereas contingency χ² tests were conducted to examine the possibility of genetic independence or linkage between leaf blotching and vein color. Calculation of the χ² values was conducted using the program “Calculation for the Chi-Square Test” developed by Preacher (2008).

Inheritance of leaf blotching.

Progeny from selfing three blotched cultivars, White Christmas, Carolyn Whorton (Fig. 1A), and Florida Blizzard, segregated in a 3:1 ratio (blotched:nonblotched) (Table 2; χ² = 0.11 to 0.40, P = 0.53 to 0.74). Progeny from selfing three nonblotched cultivars (Candidum, Miss Muffet, and Gingerland) were all nonblotched (140 individuals; Table 2). These results indicate that leaf blotching seems to be controlled by one single nuclear locus with two alleles and that blotching allele was dominant over the nonblotching allele.

Table 2.

Segregation for leaf blotching in progeny of 20 caladium crosses made in 2004.^z

View Table

Leaves of caladium cultivars Carolyn Whorton (A; blotched, red-veined) and Frieda Hemple (B; nonblotched, red-veined) showing the presence or absence of leaf blotching.

Citation: HortScience horts 44, 1; 10.21273/HORTSCI.44.1.40

• Download Figure

View Full Size

Leaves of caladium cultivars Carolyn Whorton (A; blotched, red-veined) and Frieda Hemple (B; nonblotched, red-veined) showing the presence or absence of leaf blotching.

Citation: HortScience horts 44, 1; 10.21273/HORTSCI.44.1.40

• Download Figure

Leaves of caladium cultivars Carolyn Whorton (A; blotched, red-veined) and Frieda Hemple (B; nonblotched, red-veined) showing the presence or absence of leaf blotching.

Citation: HortScience horts 44, 1; 10.21273/HORTSCI.44.1.40

• Download Figure

Five crosses were made between five nonblotched cultivars, including reciprocal crosses between ‘Candidum’ and ‘Gingerland’, ‘Fannie Munson’ and ‘Gingerland’, and a cross between ‘Frieda Hemple’ (Fig. 1B) and ‘Rosebud’. None of the progeny (Table 2; 389 individuals) were blotched, indicating that these cultivars were homozygous-recessive for leaf blotching and might be good parents for test crosses.

Five test crosses were made to test the segregation of leaf blotching in ‘White Christmas’. It was used as a seed parent in two crosses and as a pollen parent in three crosses. Leaf blotching segregated in a 1:1 ratio (Table 2; χ² = 0.06 to 2.13, P = 0.14 to 0.81) in the test crosses with ‘Rosebud’, ‘Frieda Hemple’, or ‘Miss Muffet’. However, the segregation was skewed toward more nonblotched individuals in progeny of the reciprocal crosses between ‘White Christmas’ and ‘Gingerland’ (Table 2; χ² = 9.04 or 6.40, P ≤ 0.01).

When ‘Carolyn Whorton’ was crossed with ‘Gingerland’ and ‘Miss Muffet’, the segregation of leaf blotching was in a 1:1 ratio (Table 2; χ² = 0.10, P = 0.75 or 0.76) as expected for a trait controlled by a single nuclear locus with complete dominance. This segregation was also observed when ‘Florida Blizzard’ and ‘Carolyn Whorton’ were crossed (Table 2; χ² = 0.05, P = 0.82). However, skewed segregation (Table 2; χ² = 6.72, P = 0.01) was observed again when the cross involves ‘Gingerland’. In this case, more blotched progeny appeared than expected.

In summary, results suggest that leaf blotching in caladium is controlled by a single nuclear locus. We propose to name this locus B with allele B for blotching and having complete dominance over the nonblotching allele b. The three major blotched cultivars, White Christmas, Carolyn Whorton, and Florida Blizzard, are all heterozygous (Bb) at this locus. The skewed segregation in crosses involving ‘Gingerland’ as seed or the pollen parent might indicate the existence of other factors and/or mode of gene interaction that influenced the expression of leaf blotching.

Genetic relationship between leaf blotching and vein color.

Previous results show that the color of the main vein in caladium is controlled by three alleles at the V locus (red vein allele V^r > white vein allele V^w > green vein allele V^g) (Deng and Harbaugh, 2006; Wilfret, 1983, 1986). In the cross ‘White Christmas’ (green-veined, V^gV^g) and ‘Rosebud’ (red-veined, V^rV^w), four types of progeny (red-veined blotched or nonblotched and white-veined blotched or nonblotched) segregated in a 1:1:1:1 ratio (Table 3; χ² = 1.96, P = 0.58). The cross ‘Miss Muffet’ (white-veined, V^wV^g) and ‘White Christmas’ had another four types of progeny (white-veined blotched or nonblotched and green-veined blotched or nonblotched) in a 1:1:1:1 ratio (Table 3; χ² = 1.75, P = 0.63). In these two crosses, both independent segregation and complete linkage between the two concerning traits could result in the observed segregation ratio (Table 3).

Table 3.

Joint segregation of leaf blotching (blotched and nonblotched) and main vein color (red, white, and green) in progeny of seven caladium crosses.

View Table

Three types of progeny were observed in the ‘Carolyn Whorton’ (red-veined, V^rV^g) × self progeny with the double-recessive (green-veined, nonblotched; V^gV^gbb) type missing (Table 3). χ² analysis indicates a greater probability for linkage between leaf blotching and vein color (χ² = 0.79, P = 0.85) than for independent segregation (χ² = 6.89, P = 0.08). A similar segregation pattern (three types of progeny and with double-recessive progeny missing) appeared when ‘White Christmas’ and ‘Carolyn Whorton’ were crossed, which rejects an independent segregation hypothesis (Table 3; χ² = 43.67, P = 0) but fit the ratio expected for complete linkage between the two loci in the repulsion phase (Table 3; χ² = 3.54, P = 0.32). The segregation of three types of progeny in the cross ‘Carolyn Whorton’ and ‘Miss Muffet’ also supports linkage between leaf blotching and vein color (Table 3; χ² = 4.41, P = 0.49). If the two traits were inherited independently, six types of progeny (three vein colors and blotching or nonblotching) should be in a ratio of 2 (red-veined, blotched):2 (red-veined, nonblotched):1 (white-veined, blotched):1 (white-veined, nonblotched):1 (green-veined, blotched):1 (green-veined, nonblotched) in this cross. Segregation ratios in the ‘Florida Blizzard’ × self or ‘Florida Blizzard’ × ‘Carolyn Whorton’ also supported genetic linkage between leaf blotching and vein color (P = 0.86 and 0.83) and rejected independent segregation between the two traits (Table 3).

Thus, segregation data from all the five crosses that could differentiate genetic linkage from independent inheritance support complete (or close) linkage between leaf blotching and vein color with the red (V^r) or white (V^w) vein alleles in coupling with the nonblotching allele b and the green vein allele V^g in coupling with the blotching allele B. Thus, the genotype for ‘Carolyn Whorton’, ‘White Christmas’, and ‘Florida Blizzard’ in vein color and leaf blotching would be V^rb//V^gB, V^gb//V^gB, and V^wb//V^gB, respectively. Collectively, 357 progeny were examined in the five crosses, and no double-recessive recombinants were observed. This suggests very tight linkage between leaf blotching and vein color.

As mentioned earlier, because of the potential ornamental values, developing new blotched cultivars has been an important objective in caladium breeding. The information gained from this study may help plan parental crosses and breeding populations for this breeding objective. These results indicate that new blotched progeny or breeding lines can be obtained if one of the two breeding parents in a given cross is blotched. All three blotched cultivars tested in this study (‘Carolyn Whorton’, ‘White Christmas’, and ‘Florida Blizzard’) are heterozygous for leaf blotching. Several other commercially available blotched cultivars, including Florida Calypso, Florida Elise, and Florida Roselight, are progeny of ‘Carolyn Whorton’ and nonblotched cultivars (Wilfret, 1992, 1995). These cultivars are expected to be heterozygous for leaf blotching as well. When one of these cultivars is crossed with nonblotched cultivars, approximately half of the progeny is likely to express leaf blotching. When these blotched cultivars are crossed, three-fourths of their progeny are likely to be blotched. Therefore, a significant percentage of blotched progeny will appear in a breeding population derived from crossing blotched cultivars or crossing blotched cultivars with nonblotched cultivars. This high percentage of blotched progeny should facilitate developing new blotched cultivars with diverse leaf colors, shapes, and growth habits and improved stress (sunburn, cold, drought, and so on) tolerance and disease resistance {Fusarium [Fusarium solani (Mart.) Saa.] tuber rot, for example}.

The available information on the mode of inheritance of leaf blotching may also provide a basis for further investigation into the origin of this trait. Leaf blotching is intriguing in cultivated caladium. The trait occurs commonly in Caladium humboldtii Schott, but it has been observed only rarely in the species accessions of C. bicolor, C. marmoratum, C. picturatum, or C. schomburgkii that are believed to have contributed to the development of cultivated caladium (C. ×hortulanum) (Birdsey, 1951; Hayward, 1950; Wilfret, 1993). It was suspected that C. humboldtii might be part of the parentage of ‘White Christmas’ and contributed to the leaf blotching trait. However, molecular marker and genetic relationship analysis indicate that ‘White Christmas’ and C. humboldtii were not closely related (Deng et al., 2007). Demonstration of leaf blotching as a trait under the control of a single dominant nuclear allele may facilitate future efforts to tag the trait with molecular markers and potentially trace the species origin of leaf blotching.

A previous study showed that leaf spotting is linked to vein color at an average recombination frequency of 4.4% (Deng et al., 2008). Considering such, one may ask what are the genetic or linkage relationships between these three traits (vein color, leaf spotting, and leaf blotching)? Efforts are being made to identify plant materials with appropriate phenotypes and genotypes for these traits and to develop appropriate families to investigate the question.

Ornamental aroids are the most prominent plants used in the foliage plant industry (Henny, 1988; Henny and Chen, 2003). Leaf characteristics are one of the most important groups of traits for the commercial values of ornamental aroids. A thorough understanding of the inheritance of foliar traits is critical for improving breeding efficiency for producing desirable cultivars in these aroids (Henny, 1988). Studies to date on the inheritance of foliar traits in caladium (leaf shape, vein color, spotting, blotching, and so on) have shown that the traits are generally under simple genetic control (Deng and Harbaugh, 2006; Deng et al., 2007, 2008; Gager, 1991; Wilfret, 1983, 1986). Diverse variation is available in a number of other leaf characteristics in caladium (Deng, personal observations). It will be interesting to determine how these other foliar traits are genetically controlled. If these traits are also under simple genetic control, caladium may serve as a good model for investigating the genetic relationships among foliar traits and studying the genetic interactions among the underlying loci in ornamental aroids.