Paeonia arietina(1)

Perennials. Tap roots columnar, lateral roots always tuberous, sometimes tubers beaded. Stems 30-70 cm tall, green, usually entirely hirsute, rarely hirsute only above, with 4-7, rarely up to 15, yellowish green or purplish scales at the base. Lower leaves biternate; leaflets decurrent, usually with some segmented or shallowly dividid; leaflets/leaf segments most frequently 13-23, rarely down to 11 and up to 32 in number; petioles and petiolules densely hirsute; blades elliptic, oblong or ovatelanceolate, cuneate at the base, acute at the apex. 5-12 cm long, 3-6 cm wide, glabrous or villose along major veins above, but mostly densely, very rarely sparsely villose beneath. Flowers solitary and terminal; involucrate bracts 1-3 in number, leaf-like; sepals 3—5 in number, unequal in size, all rounded at the apex, villose outside, green but purple at the periphery; petals 6-9d in number, rose or red, obovate, entire or rarely deeply lobed; filaments purple, anthers yellow; disk waved, glabrous but sometimes hirsute; carpels 2 or 3, less frequently 4 or 5, occasionally 1 in number, yellow tomentose, hairs 2.5 mm long; styles very short or absent; stigmas red, c. 1 mm wide. Follicles 2-3 cm long.

Chromosome number: 2n=20 (tetraploid)

Growing usually in sparse oak or coniferous Woods, or in clearings of forests, also in pastures, on limestones but also granites, with a wide range of altitudes from 300 m in Emilia of Italy to 2,100 m in Kutahya Province of Turkey. Distributed in Turkey, Romania, Bosnia-Herzegovina, Albania, Croatia and Italy (Emilia) from the east to the west.

Although Sabine’s taxon has epithet “cretica”, his plant was not from Crete. Stern provides a clear explanation.

The taxonomic treatment of Paeonia arietina G. Anderson (= P. mascula subsp. arietina (G. Anderson) Cullen & Heywood) has been controversial. Paeonia arietina was described first by Anderson, who clearly indicated that his new species was characterised by tuberous roots, pilose stems and petioles, leaflets that are decurrent and glaucous-pilose beneath, and sepals that are pilose at the base. Two varieties were recognised by Anderson in this species: var. andersonii and var. oxoniensis. He described the first from a garden in England and presumed it to be a native of the Levant (near Istanbul, Turkey), and the second from the Oxford Botanic Garden, wrongly considering it a native of Crete following Mr Clusius. De Candolle recognised P. arietina and considered it to be from “Oriente”. Lynch completely followed Anderson, enumerating the two varieties and considering them originally from the Levant and Crete, respectively.
Stern not only recognised Paeonia arietina but also was the first author to indicate its distribution by quoting herbarium specimens. The specimens cited by Stern mostly match Anderson’s protologue very well, e.g. those from Turkey (Anatolia and Amienia), Bosnia and Italy (Emilia).

Cullen and Heywood reduced Paeonia arietina G. Anderson to P. mascula subsp. arietina (G. Anderson) Cullen & Heywood. They ignored Stern’s article, and thus continued to include Crete in the distribution range of this entity. Their treatment was followed by Davis and Cullen, Akeroyd, and Halda. Which treatment is natural? That is to say, is P. arietina an independent species or just a subspecies within P. mascula? In the molecular phylogenies of Paeonia sect. Paeonia produced by Sang and his co-workers, P. arietina, P. parnassica and P. officinalis (including P. humilis = P. officinalis subsp. microcarpa) could not be distinguished from each other on the matK and psbA-trnH trees. Nevertheless, these three species are differentiated clearly from P. mascula subsp. mascula and P. mascula subsp. hellenica on all of these trees. On the basis of ITS sequences for sect. Paeonia, P. arietina, P. parnassica and P. officinalis (including P. humilis) form a group, within which they are indistinguishable from each other, whereas P. mascula subsp. mascula and subsp. hellenica form another group with P. coriacea, P. broteri, P. clusii (including P. rhodia), P. mlokosewitschii (= P. daurica subsp. mlokosewitschii) etc.. The phylogeny of sect. Paeonia reconstructed from a synthesis of the ITS and matK phylogenies indicates that P. arietina, P. parnassica and P. officinalis (including P. humilis) form a group, which is divergent from another group comprising P. mascula (including two subspecies, subsp. mascula and subsp. hellenica, P. coriacea, P. clusii (including P. rhodia) and P. mlokosewitschii. Therefore, molecular phylogeny shows that the P. arietina—P. parnassica group is more closely related to P. officinalis than to P. mascula. Our Adhl and Adhz gene trees also demonstrate that the P. arietina—P. parnassica group form a clade that is separated from the P. mascula clade.

According to our observations, there are distinct differences between the Paeonia mascula group on one side and the P. arietina—P. parnassica group together with the P. officinalis group on the other. In the P. mascula group, the roots are always carrot-shaped, whereas they are tuberous, sometimes even tandem-tuberous, in all populations observed in the P. arietina—P. parnassica and P. officinalis groups. Root shape, whether tuberous or carrot-shaped, has been ignored by nearly all previous authors working on the genus Paeonia, but is a very stable character and thus of great significance in taxonomy. All the individuals observed in the P. arietina—P. parnassica group have hirsute stems, petioles and sepals, whereas these parts are always glabrous in the P. mascula group. These two groups also differ in the indumentum on the lower surface of leaves. Leaves are mostly glabrous, less frequently sparsely hispid beneath in the P. mascula group, but mostly rather densely, very rarely sparsely, villose beneath in the P. arietina—P. parnassica and P. officinalis groups. There are further differences in the shape and size of leaflets/leaf segments. Those in the P. mascula group are obovate, broad-ovate or broad-elliptic, 6-15 cm long, 4-9 cm wide, whereas those in the other two groups are elliptic, ovate-lanceolate or oblong, 5-12 cm long, 1.5—6 cm wide. Therefore, the closest relative of P. arietina—P. parnassica is the P. officinalis group, not P. mascula as in the treatments of Cullen and Heywood, Akeroyd, Davis and Cullen, and Halda.

Paeonia arietina, P. parnassica, P. officinalis and P. banatica are similar to each other in morphology. They share a number of characters: roots tuberous, leaflets mostly segmented, stems, petioles, lower leaf surfaces and sepals nearly always hairy, and chromosome number 2n = 20 (tetraploid). According to our observations, P. arietina and P. parnassica are more similar to each other than to the P. officinalis — P. banatica group. Generally in the P. arietina—P. parnassica group, stems and petioles are always rather densely hirsute, and leaflets/leaf segments are oblong, ovate-lanceolate, or rarely elliptic. By contrast, in the P. officinalis—P. banatica group, stems and petioles are sparsely hirsute or glabrous, and leaflets/leaf segments are ovate-lanceolate, elliptic or linear-elliptic. The most distinct difference between these two groups is perhaps in the indumentum on the abaxial side of the sepals, which is densely villose in the P. arietina—P. parnassica group but densely to sparsely hispidulous or even glabrous in the other group. Within the P. arietina—P. parnassica group, P. arietina has a greater number of leaflets or segments, rose to red petals, and yellow anthers, whereas P. parnassica has dark purple petals and purple anthers.

An addition to the overview by Hong: Paeonia arietina Cullen & Heywood ssp. arasicola G.Kaynak, Ö.Yılmaz & R.Daşkın
This subspecies is barely documented, and mostly in Turkish literature. By the shape of its root system and leaflets it is probably a subspecies of P. arietina. The leaflets below are tomentose as is the stem. Leaflets acute, elliptic, narrowly obovate, oblanceolate, 7,5-9,5 x 4-5 cm. Petals white to pale pink. Growing at around 1,500-2,000 m altitude in rocky places in forests of Fagus and Quercus. Narrowly restricted to the Bursa region, Turkey, more precisely the Aras River Valley from where it gets its name. Population localities: Afyonkarahisar; Sultandagi mountain, 1400-1500 m, and Bursa, Uludag Mountain. 2n=20.
Original introduction: Daşkın, R., Yılmaz Ö., Kaynak G., Açık L., Çelebi A., 2005. “A New Taxon of Paeonia (Paeoniaceae) from Northwest Anatolia. Turkey.” (Probably published online only, to have been removed later). The species is usually referred to as: Paeonia mascula subsp. arasicola (Kaynak, Daykin & Yilmaz) subsp. nov.  A formal english introduction in any written publication cannot yet be found, although it is sometimes referred to. Notwithstanding the authors denomation of it as P. mascula, it is clearly P. arietina because it has tuberous roots with small beads (personal observation), has indumentum on stems and on the leaflets below, whilst those leaflets are also narrower.
More info about species peonies in Turkey can be found here (Turkish language only, except for the abstract):
Kaya, Erdal. “Türkiye Şakayıklarının (Paeonia spp.) Kültüre Alınması ve Islahı.” In:  IV. SÜS BİTKİLERİ KONGRESİ BİLDİRİLER, 2010, pp. 226-238.

Paeonia arietina ssp arasicola in Turkey, Bursa, Uludag Mountain.
Images and video by Adem Çevikbaş

Paeonia arietina ssp arasicola in Turkey, Bursa.
Images by Yıldız Konca.

  1. Hong, De-Yuan. “Peonies of the World. Taxonomy and phytogeography.” Kew: Royal Botanic Gardens, 2010, p. 221-225.[]

2021, The Peony Society -

Log in with your credentials


Forgot your details?

Create Account