Paeonia officinalis L.subsp. banatica (Rochel) Soó

Növényföldrajz ed. 1 146 1945

type: [herbaceous peony] – [species]

accepted name





By Prof. R. Soó


(Received December 12, 1959)

In my opinion P. banatica Rochel PI. Banat. Rar. 48 tab. 11. (1828) is a homogenous taxon which belongs to the "Formenkreis" of P. officinalis even if some specimens occasionally are suggestive of P. mascula (P. corallina). The apical leaflets of P. banatica consist namely almost invariably of 2—3 lobes or 2—3 segments, and are very seldom entire, while the lateral leaves are as a rule entire. This is particularly true for the plants of the sandy steppe of Deliblat (rochel's locus classicus) and the Mecsek Mountain, whereas on the specimens of Bazias the medial leaves are often entire, rarely with 2—3 lobes; for the most part both shapes of leaf can be found. The shape and width of the leaves and leaflets are very variable. The picture in rochel's book represents a typical specimen in natural size, whose medial leaflets, with the exception of one, are tripartite, even one of the lateral leaflets of the basal leaves is bipartite. The leaves may in nature be narrower or wider than in the picture. Also the picture refers to the Formenkreis of P. officinalis. Some terms of the original description e. g. foliolis... subtus arachnoideo-incanis— are not always valid. Its loci classic! are "Bielo-Berdo" and "Ulma", on the sandy steppe of Deliblat.

Geographical distribution of P. banatica :

1. Hungary, Mecsek Mountain: "Zengo" and "Hármashegy", between the villages Pecs-varad and Hosszúhetény. Discoverer: kitaibel (Addit. p. 184). Published by: janka (Öst. Bot. Zeitschr. 1866. 398;, majeb (Gimn. Ért. 1859, Pécs: Fünfkirchen, 31), nendtvich (ap. kerner Zool. Bot. Ges. 1853. 567, 571), simonkai (Math. Term. tud. Közl. IX. 179), horvát (Term. tud. Közl. 1924. 66, Gimn. Ért. 1934, Pécs 10, Mecsek flórája 72 etc.); it was collected also by javorka, Soó, Z. karpati, papp, baksay, J. szujko. Number of the specimens examined: 8. The medial leaves are bi- or tripartite, only two leaves are entire. Dimensions of the leaves; 8—16 cm long, (1.5) 2.5—5 (7—8) cm wide, varying between narrowly lanceolate and ovate-lanceolate. (Dimensions: 1.5,1.8, 2.0, 2.2, 2.5 X 8 cm, 2.3, 2.5, 3.1, 3.5 X 9 cm, 3.5 X 10 cm, 3 X 11 cm, 3.0, 4.0 X 12 cm, 3.0, 3.5, 4.5 X 13 cm, 4.0, 6.0, 7.0 X 14 cm, 7.5, 8.0 X 16cm).

2. Hungary, comitat Somogy (kit. in Host Flora Austr. II. 63). Dubious.

3. Yugoslavia, Syrmia, Fruska Gora Mountain: Cerevic (Schneller Verb. Pressbg. Ver. Naturk. IV. 80). I have not seen any specimens.

4. Yugoslavia, Banat, sandy steppe of Deliblat: Deliblato. Discoverer: rochel. Published by Ulma (rochel PI. Banat. Rar. 48), Grebenac (wiebzbicki Flora 1840. 368, heuffel Enum. pi. Banat. 13), "Susara" (wiebzb. I.e.), "Korn" (PAN6lC Ost. Bot. Zeitschr. 1867. 82, bobbas Homokp. 91, wagnek Erdeszeti Kfserl. 1913. ,280), "Flamunda" (wagner I.e.). Collected by Ulma-Mramorak (simonkai ap. tatab Acta Bot. Hung. II. 83), Flamunda (hollós, degen, javorka and timko ap. tatar I.e.), "Fontina fetje" (degen ap. tatar I.e.), Deliblat (muller ap. tatar I.e.), further thaisz (Deliblat), tl'zson (Fehertelep). Number of the specimens examined: 20.

The leaves of rochels originals (5 herb. papers!) are comparatively narrow, 1.8—3.5 cm wide, 6.5—11 cm long, the medial leaf, very exceptionally one or even both extreme leaflets bi- or tripartite. Further specimens from the collections of wierzbicki, heuffel, müller, tuzson, wagner, Hollós, javobka and Timkó. Medial leaves bi- or tripartite, very seldom entire, (1.5) 2—4 (—5) cm wide, 6—11 cm long, narrowly lanceolate to broad lanceolate, basal leaves broad ovate. Dimensions: 2.5 X 6.5 cm, 3x8 cm, 2.0, 2.5, 3.0, 3.5, 4.0 X 9 cm, 2.0 X 3.0 X 9.5 cm, 3.0, 3.5, 4.0, 4.5 X 10 cm, 3.0 X 11.5 cm, 2.5, 3.5, 5.5 X 11 cm, 4.0 X 12 cm. In 3 out of 20 specimens there were entire medial leaflets.

5. Rumania, Banat, Bazias (near the Monastery and in the forest). Discoverer: rochel. Published by heuffel (I.e. 13), borbas (I.e. 91), bernatsky Bot. Kozl. 1902. 31, javorka and TlMK.6 Fl. Exs. Hung. No. 739, collected also by wierzbicki, bohatsch, simoxkai, seymann, wagner, lindauer, degen, ap. tatar I.e.). Number of specimens examined: 20. Medial leaves entire or bipartite (very often on the leaves of the same specimen), in 5 out of the 20 specimens examined only entire, in 3 only divided, in the others mixed. It is a population with rather broad and large leaves. Leaves (1.8) 2.5—5 (6) cm wide, 6.5—13 cm long, narrowly lanceolate to ovate lanceolate and oblong respectively. Dimensions: 2.0 X 6.5 cm,

1.8, 2.0, 2.5, 2.8 x 7 cm, 2.0, 3.0 X 7.5 cm, 2.0, 2.0, 2.2, 2.5, 3.0 X 8 cm, 3.5 X 8.5 cm, 2.0,

2.5, 3.0, 3.5, 4.5, 5.0 X 9 cm, 3.0. 3.5, 4.0, 5.0 X 10 cm, 3.5, 4.0, 4.5, 5.0 X 12 cm, 3.5, 4.5, 6.0 X 13 cm.

6. Rumania, Lugos (Borb. Öst. Bot. Zeitschr. 1892. 286, Huth Engler's Jahrb. XIV. 270), incorrect, instead of Deliblat.

7. Rumania, Orsova (Haynald's Herb.), surely erroneous, instead of Bazias. Basal leaves everywhere broader, ovate or ovate-oblong, 4—5 X 8 X 10 cm.

It is hardly possible to state the number of the leaflets as the lower leaves are in most cases missing. The hairs of the leaves are variable, below with very little, short hairs or becoming glabrous.

P. banatica differs both in its habit and the partition of its leaves from the typical P. officinalis, whose lateral leaves, too, are generally divided into 2—4 segments, on the specimens from South-Tyrol examined by myself (cf. Herb. Norm. 3501, FEAH. 90, from the collections of Andreanszky, Behrendsen, Foletto, Porta, Rigo, chiefly from Monte Baldo and Val di Ledro), leaves 2,0, 2,5, 3 cm wide and 6—11 cm long. It differs likewise from P. officinalis ssp. peregrina, with the specimens of which originating from Spain I compared it (cf. Herb. Norm. 3005, Baenitz PI. Eur. 5903, from the collections of Reverchon, Derben, Pons), their leaves are divided to a larger extent with narrow leaflets. It can be still less identified with P. mascula (P. corallina) whose leaflets are wide, elliptic, ovate or orbicular and all entire.

Thus we cannot share in Nyárády's opinion who takes the plant of Bazias partly to P. corallina Retz (P. mascula Mill.) and describes a form of it with broad ovate or nearly roundish leaves by the name var. triternatiformis (which as an extreme variety may remain in the Formenkreis of P. banatica) — p. 403 —, partly to P. officinalis p. 407 —. He mentions here as new habitats Lugos, Bazias (surely instead of Bazias) and Dumbravita (Kisdombró) from the district of Belényes; these habitats are even in his own opinion dubious. According to Nyárády the leaves of the plant from Deliblat are 3—4 times, while those from Bazias 1,4—2(3) times as long as wide. The great number of the data quoted above shows that this ratio is varying in both habitats.

Also in the manuscript elaboration of the Flora Europaea the plant of Bazias (though this is no locus classicus!) approaches partly to the P. officinalis, partly to the P. mascula, not excluding the possibility that at Bazias hybridization has taken place between the spontaneous and some introduced species, cultivated in the Monastery Garden. On the other hand, in my unchanged opinion, P. banatica is varying but remains within the above indicated limits of variation a uniform population which, although it is partly suggestive with its entire and broader leaves to a certain extent of P. mascula, it belongs, however, to the Formenkreis of P. officinalis with the nearly always divided leaflets of the upper leaves and its whole build, and constitutes the endemic, vicariant subspecies of that species in the Bánság district, in the Mecsek and Fruska-Gora Mountains; thus its correct name is:

P. officinalis L. (s. 1.) ssp. banatica Soó Növényföldrajz 1945 p. 146. (P. peregrina var. ban. Kittel 1844, P.feminea var. ban. Gürke 1903, P. off. var. ban. Graebn. 1923, Hayek 1924). The data from Macedonia and the Lesbos island are undoubtedly concerning another plant.


Cullen & Heywood

P. officinalis L., Sp. PI. 530 (1753)

(a) subsp. officinalis

Distribution: Albania, Austria, France, Switzerland, Hungary, Italy, Jugoslavia.

(b) subsp. villosa (Huth) Cullen & Heywocd, stat.nov.

Syn. P. peregrina var. villosa Huth, Engl. Bot. Jahrh. 14: 270 (1891) P.humilis var. villosa (Huth) F.C.Stern, Jour. Roy. Hort. Soc. 68:129 (1943); A study of the genus Paeonia 106 (1946). ? P. villosa Desf., Tabl. Ecole Bot. 126 (1804), nom. nud.

Distribution: France, Italy.

The combination P. officinalis subsp. villosa (Desf.) was proposed by Ciferri & Giacomini, Nomenclator Florae Italicae Pars altera: Dycotiledones, Fasc. I: 308 (1954) but without reference to the place of publication of the basionym and thereby invalid under the Code of Nomenclature.

P. villosa Desf. which is sometimes cited as the basionym of this taxon is a nomen nudum. huth (op. cit.) does not mention P. villosa Desf. except in the index when he gives "villosa Desf. Sweet = peregrina Mill".

(c) subsp. humilis (Retz.) Cullen & Heywood, comb. et stat. nov.

Syn. P. humilis Retz., Obs. Bot. 3: 35 (1783), Rothmaler, Feddes Repert. 49: 176 (1940); F. C. Stern, A study of the genus Paeonia 104 (1946).

Distribution: Portugal, Spain, France.

(d) subsp. banatica (Rochel) Soó, Növényfoldrajz 146 (1945).

Syn. P. banatica Rochel, PI. Banat. Rar. 48, t. 11 (1828); F. C. Stern , op. cit. p. 72 (1946).

Distribution: Jugoslavia, Hungary, Romania.

The identity of P. banatica Rochel has been the subject of considerable discussion in the literature. Stern (op. cit. 1946) regarded it as endemic to a small district in the Banat and cited only specimens from the vicinity of Bazias in the Romanian Banat, The plant from this region has no bearing on the typification of P. banatica which was described by rcchel from Deliblat (Jugoslavia) and figured by him in Plantae Banatus rariores icombus et descriptionibus illustratae, t. 11, fig.25. According to borza (Consp. Fl. Romaniae (1947) and in litt.) the figure illustrate a plant which is at most only varietally distinct from P. officinalis subsp. officinalis. nyárády (Fl. R.P.R. 2: 401 (1953) and in litt.) also considers P. banatica as a variety of P. officinalis. Soó (1960) has made a thorough investigation of the taxon and concludes that P. banatica is best treated as a vicariant subspecies of P. officinalis, distributed in Hungary, Jugoslavia and Romania.

The plants from Bazias (the basis of stei.n's concept of this taxon) form a variable population and are regarded by nyabady (loc. cit.) as corresponding to P. corallina (P. mascula (L.) mill.) who considers forms with leaves broadly ovate, 1.4—2 times as long as broad, as a separate variant, var. triternatiformis, so called because of its resemblance in leaf shape to P. triternata pallas. Soó, on the other hand, disagrees with this, but it does appear that some of the Bazias plants identified by many authors as P. banatica are in fact variants of P. mascula.

Viewed as a whole these four subspecies form a geographical replacement complex, extending from Portugal, to Jugoslavia and Romania. Intermediates frequently occur in the areas of overlap between these taxa and there seems to be no justification lot-following Stern in regarding them as different in status


Cullen & Heywood in Tutin: Flora Europaea

Eric Schmitt (1998)

subsp. banatica (Rochel) Soó, 1945

Basionyme : P. banatica Rochel, 1828

Synonymes : P. peregrina var. banatica (Rochel) Kittel, 1844

P. foemina var. banatica (Rochel) Gürcke, 1903

P. officinalis var. banatica (Rochel) Graebner, 1923

P. officinalis subvar. banatica (Rochel) Hayek, 1924

Tiges de 35 à 50 cm, glabres. Feuilles biternées, composées de 12 à 15 folioles glabres ou partiellement pubescentes, elliptiques, lancéolées à oblancéolées, de 8 à 16 cm de long et (1,5) 2,5 à 5 (7-8) cm de large, vert foncé et glabres au-dessus, glauques et glabres à légèrement pubescentes au-dessous, décurrentes à la base et aiguës à presque obtuses à l'extrémité. Foliole terminale entière ou généralement divisée jusqu'à la base en 2 ou 3 lobes, folioles latérales toujours entières. Fleurs rose-rouge de 8 à 12 cm de diamètre, filet des étamines rouge, 2 carpelles fortement tomenteux. Floraison avril-mai (juin).

2 n = 20 : Lovka & al. (1971, Yougoslavie, région de Deliblatska)

Présente de façon limitée en Europe centrale. Nord-est de la Yougoslavie (monts Fruska Gora, plaines de Deliblatska, près de la frontière roumaine), ouest de la Roumanie (région de Bàziâs, dans les forêts et près d'un monastère ; Oradea, Lugoj, Buzias, Orsova), sud-ouest de la Hongrie (monts Mecsek).

Les localités de Oradea et Lugoj sont citées par Nyàràdy qui les considère douteuses.

Stern (1946) supposait P. banatica d'origine hybride échappée d'un jardin, car elle se rencontre près d'un monastère. De par l'aspect de ses folioles, parfois largement elliptiques à ovoïdes, .glabres ou légèrement pubescentes, il la plaçait dans le groupe Mascula mais lui reconnaissait également des caractères de P. arietina (folioles parfois étroitement elliptiques, plus ou moins pubescentes). Se basant sur ces caractères intermédiaires entre les deux espèces, il maintenait P. banatica à un rang spécifique.

Nyàràdy (1953) signale dans la région de Bàziâs P. officinalis var. banatica, P. corallina var. typica (P. mascula subsp. mascula), P. corallina var. triternatiformis, et identifie comme P. corallina un échantillon d'herbier (FEH n° 739) récolté par Javorka dans le Banat.

Cullen & Heywood (1964) signalent également que des plantes de Bàziâs à folioles typiques de P. mascula ont été identifiées comme P. banatica par de nombreux auteurs.

Soô (1960) conteste le point de vue de Nyàràdy et considère que l'échantillon FEH n° 739 appartient bien à P. officinalis subsp. banatica, dans laquelle il inclut P. corallina var. triternatiformis, correspondant aux formes à feuilles largement elliptiques à ovoïdes.

Il donne également les caractéristiques des feuilles de spécimens récoltés dans diverses localités (Yougoslavie, Hongrie, Roumanie) dont les dimensions peuvent varier de 1,5 à 8 cm de large et de 6,5 à 16 cm de long. En conclusion, Soô considère que P. banatica est une pivoine variable de par la morphologie de ses feuilles, mais qui constitue une population homogène à l'intérieur de son aire de répartition, où elle représente une forme endémique de P. officinalis.

Jalas & Suominen (1991) suivent l'opinion de Soô et incluent les populations roumaines de P. mascula dans P. officinalis subsp. banatica.

Toutefois, P. officinalis subsp. banatica diffère nettement de P. mascula par ses feuilles avec la base des folioles longuement décurrente et par ses carpelles fortement tomenteux couverts d'un duvet blanc et très peu arqués.

Gürcke (1903) et Rechinger (1943) situaient cette pivoine dans la mer Egée (îles de Lesbos, Thasos); Hayek (1924) la mentionne en Macédoine. Ces indications correspondent probablement à P. mascula.

Passalaqua & Bernardo (2004)

in The Genus Paeonia in Italy (2004)


Paeonia officinalis ssp. banatica (Marlene Ahlburg)

my plant 1) Tag der Pflanzung/date of planting: 14.11.97 ,geliefert durch / delivered by: Frei, Wildensbusch, Schweiz,

Upper surface/Oberseite

Lower surface/Unterseite



Carsten Burkhardt's Web Project Paeonia - The Peony Database


Free counters!




© 2006 Dr. Carsten Burkhardt