≡ Erpetion DC. ex Sweet, Brit. Fl. Gard. 2: nr. 170. 1826
≡ Viola subg. Erpetion (DC. ex Sweet) Y. S. Chen in Raven & Hong, Fl. China 13: 111. 2007
Type: Viola hederacea Labill.
Description.—Perennial herbs. Axes seemingly morphologically differentiated in a perennial stem with lateral sympodial stolons. Perennating stem densely or occasionally remotely noded. Sympodial stolons with a pair of bracts between each cluster of leaves. Stipules small, lanceolate. Lamina ovate-rhomboid to broadly reniform, margin crenate, long-petiolate. Corolla white to dark violet, often with a darker throat; corolla sometimes highly reduced. Spur reduced to a gibba and a green blotch on the inside of the bottom petal. Lateral petals with a broad dense pad of papillae. Style filiform, beardless. Cleistogamous flowers not produced. Allo-octoploid. Secondary base chromosome number x’ = 25.
Diagnostic characters.—Sympodial stolons present. Spur reduced to a gibba. Lateral petals with a broad dense pad of papillae.
Ploidy and accepted chromosome counts.—8x, 16x, 24x; 2n = 50 (V. banksii).
Age.—Crown node age 3.7 (3.2–3.9) Ma, stem node age 7.4 (6.5–7.7) Ma [1].
Included species.—11.
Viola banksii K. R. Thiele & Prober,
Viola cleistogamoides (L. G. Adams) Seppelt,
Viola curtisiae (L. G. Adams) K. R. Thiele,
Viola eminens K. R. Thiele & Prober,
Viola fuscoviolacea (L. G. Adams) T. A. James,
Viola hederacea Labill.,
Viola improcera L. G. Adams,
Viola perreniformis (L. G. Adams) R. J. Little & Leiper,
Viola serpentinicola de Salas,
Viola silicestris K. R. Thiele & Prober
Distribution.—Southern and eastern Australia; Tasmania
Discussion.—Phylogenetically, sect. Erpetion is an allo-octoploid lineage with two CHAM genomes and another two genomes in common with sect. Chilenium, indicating that sect. Erpetion experienced a second genome duplication after the two sections diverged. There is no indication that this ancestral tetraploid Erpetion still exists. Section Erpetion is characterised karyologically by the secondary base chromosome number x’ = 25 [2]. The estimate of 10x for sect. Erpetion by Marcussen et al. [1] was based on unconfirmed (and probably erroneous) counts of 2n = 60 and 2n = 120 on “representatives of the Viola hederacea complex in the Kosciusko area” by Moore in [3].
Members of sect. Erpetion can be recognised immediately by two unique synapomorphies, i.e., the presence of sympodial stolons, which differ from true stolons by their clustered leaves and bibracteolate stem segments, and the pad of papillae on the lateral petals in place of the beard of trichomes some members of other lineages exhibit. Anatomically, the sympodial stolon consists of a potentially infinite chain of bibracteolate stem segments each ending in a leaf rosette, which in turn produces a new segment from the axil of its lowermost leaf. Adventitious roots are produced at the base of each rosette only. In Fragaria (Rosaceae), both sympodial and monopodial stolons can be found among closely related species (e.g., F. viridis vs. F. vesca, respectively), suggesting that the underlying genetics can be quite simple.
We follow the original delimitation of Becker [1] for the section. At the time only one variable species was recognised, Viola hederacea, but c. 11 species are now recognised [4,5]. Genome size data (2C DNA) indicate that sect. Erpetion forms a polyploid series based on 8x, i.e., with V. banksii at the 8x level (two accessions with 1.26 and 1.27 pg), V. fuscoviolacea at the 16x level (2.57 pg), and V. hederacea at the 24x level (3.45 pg; T.M., unpublished data, and [2]). Indeed, the occurrence of autogamous taxa with very small corollas, i.e., V. cleistogamoides and V. fuscoviolacea, agree with the observation of high ploidy in this section. A cultivar attributed to V. banksii is frequently grown as an ornamental, but appears to be a hybrid, based on having low pollen fertility [2].
The sister lineage of sect. Erpetion is the South American sect. Chilenium [1], from which it may have diverged c. 7.4 Ma ago [1]. This relationship is surprisingly from a morphological perspective, as the two taxa are rather dissimilar and lack obvious synapomorphies.
Fig 1 Global distribution of Viola sect. Erpetion.
1. Marcussen, T.; Heier, L.; Brysting, A.K.; Oxelman, B.; Jakobsen, K.S. From gene trees to a dated allopolyploid network: Insights from the angiosperm genus Viola (Violaceae). Syst. Biol. 2015, 64, 84–101. [Google Scholar] [CrossRef]
2. Kwiatkowska, M.; Żabicka, J.; Migdałek, G.; Żabicki, P.; Cubała, M.; Bohdanowicz, J.; Słomka, A.; Jedrzejczyk-Korycinska, M.; Sliwinska, E.; Sychta, K.; et al. Comprehensive characteristics and genetic diversity of the endemic Australian Viola banksii (section Erpetion, Violaceae). Aust. J. Bot. 2019, 67, 81–98. [Google Scholar] [CrossRef]
3. Smith-White, S. Cytological evolution in the Australian flora. Cold Spring Harb. Symp. Quant. Biol. 1959, 24, 273–289. [Google Scholar] [CrossRef]
4. De Salas, M.F. Viola serpentinicola (Violaceae), a new Tasmanian species endemic to serpentinised ultramafic soils. Muelleria 2018, 36, 112–117. [Google Scholar] [CrossRef]
5. Thiele, K.; De Salas, M.F.; Walsh, N.G.; Messina, A.; Little, R.J.; Prober, S.M. Viola curtisiae, a new rank for a poorly understood species, with notes on V. hederacea subsp. seppeltiana. Muelleria 2018, 36, 107–111. [Google Scholar] [CrossRef]