A revision of taxonomic relation between Oenothera perangusta and O. ersteinensis (Onagraceae) based on morphometric research and statistical analyses Author Woźniak-Chodacka, Monika text Phytotaxa 2018 2018-12-13 383 1 55 74 http://dx.doi.org/10.11646/phytotaxa.383.1.3 journal article 10.11646/phytotaxa.383.1.3 1179-3163 13725223 Oenothera ersteinensis Linder & Jean, Bull. Soc. Bot. France 116: 523. 1969 . Type :— FRANCE . Bas-Rhin : Erstein , at junction of national road 83 with 426 [cultivated from seeds], July 1968 , Jean s.n. ( holotype : LILLE ; paratype : STR-40811!). Figure 3 . Genetic features: —Number of chromosomes: 2n=14; Renner’s complexes combination: cruens virens ; catenation figure: ring of 14 chromosomes ( Linder & Jean 1969 ); permanent translocation heterozygote ( PTH ); main types of genomes combination: AB ( Dietrich et al 1997 ). Distribution: —Western and Central Europe: France , Germany , Belgium and Switzerland ( Rostański et al. 2010 , Rostański & Verloove 2015 ). FIGURE 1. The holotype of Oenothera perangusta Gates (GH-00073029). FIGURE 2. The holotype of Oenothera perangusta Gates var. rubricalyx Gates (GH-00073030). FIGURE 3. The paratype of Oenothera ersteinensis Linder & Jean (STR-40811). Representatives of the American school of taxonomy treat the names O. ersteinensis and O. perangusta as synonyms assigned to separate and distinct species within subsect. Oenothera , O. biennis s.l. and O. oakesiana s.l. , respectively. In the opinion of those researchers, both taxa are clearly unlike, not only morphologically but also genetically, due to the different haploid genome types combinations ( Dietrich et al. 1997 ). The analysis of the interpretation of examined taxa by European researchers is much more complicated due to the inconsistent usage of particular names. The only one Oenothera taxonomist who conducted research on the whole European continent was K. Rostański and that is why mainly his papers were analyzed. In his earlier publications, Rostański used the name O. perangusta (without giving any synonyms) in reference to specimens found in Russia ( Rostański 1975 ) and Wales ( Rostański & Ellis 1979 ). In 1985 he proposed a new systematics of subsect. Oenothera , in which O. perangusta and O. ersteinensis were treated as one species (within series Oenothera ) under the name O. strigosa ( Rydberg 1900: 278 ) Mackenzie & Bush (1902: 139) . After the year 2001, Rostański started to consider O. ersteinensis and O. perangusta as separate species: O. perangusta was supposed to reach Scandinavia and Eastern Europe ( Rostański & Ramst 2001 , Rostański et al. 2004 , Rostański 2006, Rostański & Karlsson 2010 ) whereas the range of O. ersteienensis was thought to be limited to Western Europe mostly ( Rostański & Meierott 2006 , Rostański 2007 ). However, roughly at the same time, Rostański came back to the conclusion that both names refer to the same taxon ( Rostański 1995 , Rostański 2007 ). Finally, in his latest papers, the status of O. ersteinensis and O. perangusta is quite unclear; they are considered as separate, but “possible identical” species ( Rostański et al. 2010 , Rostański & Verloove 2015 ). Such unstable treatment leads to taxonomic and nomenclatural chaos. In this paper I present the results of morphometric research and statistical analyses performed on specimens labelled as O. perangusta , O. perangusta var. rubricalyx (including the holotypes ) and O. ersteinensis (including the paratype ). European representatives of two distant species— O. oakesiana and O. biennis , have served as a background to show the relations between the studied taxa.