R. canescens? (Dawson Forest, GA)
(S and J Perkins)
R. colemanii (Soapcreek, GA)
(S and J Perkins)
R. austrinum 'Alba'
(S and J Perkins)
Pink sport of 'Millie Mac'
(S and J Perkins)
Cahas White R. calendulacaleum seedling
(S and J Perkins)
Interaction of a pink 2n with 4n R. calendulaceum?
(S and J Perkins)
Pink 4N
(Ron Miller)
Historically most deciduous azaleas of North America are assigned to the Pentanthera section of the genus Rhododendron where R. canadense and R. vaseyi are assigned to the sister section Rhodora (Kron). More recent work by Hall, Ranney, Jones, Miller, Towe suggests that Pentanthera divides into 2 clades along ploidy levels (diploids and tetraploids). This work ,also, suggests that 2n (diploid) R. vaseyi is distinct from the Pentanthera. R. canadense and R. molle are aligned with the 4n clade but unlike the other 4ns (tetraploids) are themselves most likely 2ns (diploids). All the members of the 2n-clade are 2ns (diploids).
2N-Clade
R. alabamense
R. arborescens
R. canescens
R. cumberlandense
R. eastmanii
R. flammeum
R. occidentale
R. periclymenoides
R. prinophyllum
R. prunifolium
R. viscosum
4N-Clade
4ns in 4N-Clade
R. atlanticum (resembles R. viscosum)
R. austrinum (resembles R. canescens?)
R. calendulaceum (resembles R. cumberlandense)
R. colemanii (resembles R. alabamense)
R. luteum (Europe)
2ns in 4N-Clade
R. canadense
R. molle (Asia)
Neither Clade
R. vaseyi
Notes:
The 2n R. cumberlandense and the 4n R. calendulaceum resemble one another very closely.
Ranney, Miller, Jones determined that R. colemanii is a 4n that is separate from the 2n R. alabamense which it resembles.
Miller, Towe, Ranney have located 4ns that are pink resembling the 2n R. canescens but are closely aligned with the 4n R. austrinum. Yellow, pink, and white flowered forms of these 4ns are sometimes colocated but the 4n pinks are also found in separate populations from the other color forms. Miller suggests there is mounting evidence these pink 4ns although closely aligned with yellow austrinum may well be a separate species; however, no final determination of clasification for this 4n pink population has been made to date.
Ranney, Perkins, Clyburn have evidence strongly suggesting that R. canadense is a 2n and not a 4n as historically suggested.
Perkins speculate based on limited hybridization attempts that the 4n members of the 4n-clade normally reject pollen from the 2n-clade. The 2n-clade accepts pollen from 4n members of the 4n-clade. Members of 2n-clade accept pollen in both directions. The 4n members of the 4n-clade accept pollen in both directions.
Hall suggests that the 4n-clade and 2n-clade separated nearly 40 million years. Hall suggests that 4ns and 2ns do interact to a limited degree.
The hybrid swarm on Gregory Bald appears to be composed of only 2ns. R. calendulaceum does not appear to be present on Gregory Bald as previously conjectured.
No hybridization between the 2n R. arborescens and the 4n R. calendulaceum appears to take place at Wayah Bald. Ranney, Towe, Skinner, James have strong evidence that natural hybridization occurs between other 2ns and 4ns especially involving the 4n R. calendulaceum and pink 2ns. Miller suggests that natural hybridization between 2ns and 4ns is rare.
Cahas Mountain in Virginia contains a natural population of fertile 4n R. calendulaceum in shades of orange, yellow, white with yellow blotch, and pastels (James).
Although the normal interaction of 4ns and 2ns is infertile 3ns, some have speculated that such interactions in deciduous azaleas may cause 4ns to result in proportions higher than the norm. Some have noted "fertile" 3ns. The evidence for such interaction is sparse but well documented especially with regard to interaction of the pink 2ns with the 4n R. calendulaceum. However, most controlled crosses to date between known 2ns and 4ns have resulted in infertile 3ns.
Ranney and Towe have confirmed 4n and 3n intermediates in northwestern South Carolina in an area that includes the 2n R. periclymenoides and the 4n R. calendulaceum. In this area there are pink 4ns as well as other more pastel colors resembling R. calendulaceum and the R. periclymenoides are often lavendar or dark pink with glands on the corolla indicating the interaction may be influencing both populations. However the evidence for influence is much stronger in the direction of the 2n influencing the 4n than in the other direction since glandular forms of the 2n R. periclymenoides often appear in normal populations.
Ranney and Towe have confirmed what appear to be 4n intermediates in northeastern Georgia in an area that includes the 2n R. canescens and the 4n R. calendulaceum.
James has reported intermediates of pink 2ns and 4n R. calendulaceum in the mountains of northwestern Georgia.
Individual members of both deciduous azaleas clades appear not to self although no formal tests have been conducted to verify this.
'Marydel', 'My Mary', 'Snowbird' are 4ns indicating they are most likely not the result of a cross between a 2n and a 4n as often documented. Although unreduced gametes can cause a 4n / 2n interaction to produce fertile 4ns, there is no indication that these three named cultivars are more willing to accept pollen from 2ns than other normal 4ns.
Miller suggests that complex glands are often found on flower buds on many members of the 4n-clade which are never found on the flower buds of the 2n-clade. The 4n R. atlanticum does not appear to have such complex glands or at least to a much lesser degree on the flower bud.
Geographical location for the 2n and 4n species seems to indicate that 2ns may be hardier than 4ns.
A 4n population of the normally 2n R. occidentale is known to exist. This 4n population appears to have derived directly from the 2n population.
The relationship of 2n R. molle and 2n R. canadense to the remainder of the 4n-clade is open to much debate. The ploidy level of R. molle has not been extensively studied. Individual garden plants of R. molle are thought to be 4ns; however, all wild collected R. molle to date tested for ploidy are 2ns. The ability of these 2 species to breed with either the 4n-clade or the 2n-clade is worthy of investigation.
All documented crosses involving the interaction of the 4n-clade and the 2n-clade are open to some question since individual members of the 2 clades are easily mistaken for one another. This is especially true for hybridization of deciduous azalea done in the 19th century.
Fertility between species appears to be a good indicator of "DNA closeness" in the case of North American deciduous azaleas. The 2n R. vaseyi which is not known to interbreed is the most distance from the rest. There is indication that the 4n-clade and 2n-clade do not interbreed as easily in nature or the garden as the species within the clades do. The 2n R. molle and the 2n R. canadense are known to interbreed but fertile offspring are the exception rather than the rule.
Considering the evidence suggests that the 4n-clade and the 2n-clade interact to produce fertile offspring sparingly, the question of why do the 4n members of the 4n-clade resemble 2n-clade members who are somewhat colocated is open to much debate. Is it just a manner of shared pollinators? Is limited interaction enough to cause such closeness in appearance? Is the closeness explained by having shared ancestors? Do periods of increased hybridization in the 2n clade result in increased hydridization between the 2n and 4n clades. Are we misreading the evidence?
John and Sally Perkins
