Viola nana (De Candolle, 1824) Le Jolis, 1860

Magrini, Sara & Scoppola, Anna, 2015, Further studies in Viola Sect. Melanium (Violaceae). Identity and typification of Viola nana and V. henriquesii, two neglected European Atlantic taxa, Phytotaxa 230 (3), pp. 259-266 : 260-261

Viola nana

The earliest legitimate name of the taxon at the species-rank is Viola nana (DC.) Le Jolis in Plantes Vasculaires des Environs de Cherbourg (1860: 27–28) (Art. 11.4, McNeill et al. 2012). Le Jolis validly published the name at new rank by writing “ Viola nana DC. ”, which is regarded as an indirect reference to the basionym (Art. 41.3, McNeill et al. 2012) V. tricolor L. var. nana DC. published in De Candolle’s Prodromus (1824: 304). Furthermore, he reported in brackets “ V. parvula Brébiss. fl. norm. non Tin.; V. nemausensis Lloyd fl. Ouest, non Jord.”, thus linking his V. nana to the “ violette naine ” of the “ sables maritimes ” of Normandy, to which other names were wrongly given by Brébisson in Flore de la Normandie (1859: 40) and by Lloyd in Flore de l’Ouest de la France (1854: 60).

De Candolle’s protologue is brief and poor in regard to diagnostic characters: “ Glabriuscola, caule brevissimo, cotyledonibus per anthesim persistentibus”, but it reports a clear indication of the habitat and distribution of V. tricolor var. nana , “In arenosis maritimis Neustriae, Syriae, Asiae minoris”, thus making it possible to link this taxon to Le Jolis’s species. Le Jolis gave a more specific indication of the distribution of V. nana in the Cherbourg area (Manche, France), part of the so-called Neustria reported by De Candolle, particularly, “ dans les sables maritimes de la côte Ouest: Vauville, Biville, Le Rozel, etc. ”.

De Candolle cited the Asiatic species V. tenella Poiret (1808: 644) at the end of the protologue, probably as a synonym. On the other hand, Le Jolis (1860) reported that “... la description que Poiret a donnée du V. tenella ..., plante de Syrie, convient bien à notre espèce du littoral de l’Océan; mais il n’est pas probable que les deux plantes soient identiques. Notre plante est bien celle que De Candolle a appelée V. tricolor π nana , et cette dernière épithète me semble devoir lui être conservée. ”. In accordance with Le Jolis, the two names, V. nana and V. tenella , should be referred to different species and the epithet “ nana ” should refer only to the taxon of the Western Atlantic coast of France.

Following Le Jolis, many botanists reported V. nana , even if with different author citations, in their floras or cytological studies as an autonomous taxon occurring on the maritime sands of the Atlantic coast of France, but also on the Channel Islands and on the Isles of Scilly ( UK) ( Godron 1875, Corbière 1893, Gerbault 1923, Drabble 1928, 1930, Clausen 1931, Fothergill 1944, Lousley 1956, Pettet 1964), until the end of the 1960’s, when Valentine et al. (1968) in Flora Europaea included it in V. kitaibeliana . On the other hand, the Atlantic authors consider it as a distinct taxon even today, as highlighted also by its vernacular names, Dwarf Pansy in English and Violette naine instead of Pensée de Kitaibel in French. Recently, Tison and de Foucault (2014) have highlighted the cytological differences within Viola kitaibeliana s.l. in France, suggesting that the Atlantic hexaploid populations could be referred to a taxon of higher rank.

In accordance with most of the cited authors and following the suggestion of Merxmüller (in litt.) in Guinochet & de Vilmorin (1982: 1221): “Le taxon du Koelerion albescentis doit correspondre au V. nana (DC.) Corbière. ”, this taxon cannot be included in V. kitaibeliana s.s. nor can it be placed in synonymy with it. In fact, V. nana is a welldefined species that differs from V. kitaibeliana s.s., especially from a cytological, chorological, and ecological point of view. Viola nana has a chromosome number 2n = 48, as determined in plants from Jersey, the Channel Islands ( Clausen 1931, C!, Pettet 1964), from Tresco and Bryher, Isles of Scilly ( Pettet 1964), and, more recently, from Charente-Maritime, Oloron Island ( Verlaque & Espeut 2007), while V. kitaibeliana s.s. has a chromosome number 2n = 16 ( Schmidt 1964, Fernández Casado 1984, Erben 1985, Verlaque & Espeut 2007, Hand & Gregor 2011, Scoppola et al. 2014). Furthermore, V. nana is an Atlantic species with a distribution that is well separated from V. kitaibeliana . In fact, it is restricted to the Western Atlantic coast of France (Gironde, Charente-Inférieure, Vendée, Morbihan, Finistère, Manche, Calvados, and Pas-de-Calais), the Isles of Scilly (Tresco, Teän, and Bryher) and the Channel Islands ( Guernsey, Herm and Jersey), as reported by previous authors and verified by our studies in several European herbaria (see Appendix 1.). Finally, it is well characterized by a high phytoclimatic and ecological specificity. Viola nana grows in pioneer grasslands of the grey dunes, dominated by hemicryptophytes and rich in annuals, colonized by more or less abundant mosses and lichens (Koelerion arenariae Tüxen 1937 nom. mut. propos. Rivas-Martínez et al. 2002). It is also regarded as an indicator species of the habitat 2130 * Fixed coastal dunes with herbaceous vegetation (grey dunes), representing the Atlantic biogeographical area, particularly of the sub-types 2130-1 * Grey dunes of North Sea and Manche, distributed along the Northern coast of France (East Manche to the Cotentin Peninsula) with a Temperate oceanic bioclimate, and 2130-2 * Grey dunes of the Atlantic coasts, along the Western coast (West Atlantic Manche) with a Temperate hyperoceanic bioclimate that becomes submediterranean on the Morbihan coast, and Temperate oceanic submediterranean on the Atlantic Loire ( Rivas-Martínez et al. 2004, Géhu 2007).

For all these reasons, we propose here to reaffirm this taxon as a distinct cryptic species, accepting the name Viola nana (DC.) Le Jol.