Paeonia daurica(1)

Perennials. Roots carrot—shaped, attenuate downwards, up to 4.6 cm in diameter. Lower leaves biternate with 9 leaflets, occasionally 1-2 leaflets segmented, and thus leaflets/leaf segments usually 9, rarely 10 or 11, in number, entire, broad—obovate, oblong, rarely wide—elliptic, sometimes undulate, truncate or rounded, rarely acute or even short—acuminate at the apex, 8—17 cm long, 4.8—11.5 cm wide, glabrous above, glabrous or sparsely to densely villose, or sparsely to rather densely puberulous beneath. Flowers solitary and terminal; involucrate bracts 0—2 in number, leaf-like; sepals mostly 3, less frequently 2 in number, green but with purple periphery or entirely purple, orbicular or ovateorbicular, up to 3.5 cm long, all rounded at the apex; petals 5—8 in number, usually red or rose, less frequently yellow, pale yellow, white, or yellow but with a red spot at the base or with red periphery; filaments purple; disk c. 1 mm high, waved, tomentose above; carpels 1-5 but mostly 2 or 3 in number, mostly tomentose, less frequently glabrous, hairs 2.5-3 mm long; stigmas nearly sessile, red, 1.5—2 mm wide.

Chromosome number: 2n = 10 and 20 (see each subspecies for detail).

Three lowland subspecies, subsp. daurica, subsp. coriifolia and subsp. mlokosewitschii are diploid, whereas three alpine or subalpine subspecies, subsp. macrophylla, subsp. tomentosa and subsp. wittmanniana are tetraploids. No chromosome information is yet available for subsp. velebitensis.

Growing in deciduous broad—leaved or mixed forests, or at the edges of forests, on various media and at altitudes from 350 to 2,740 m. Widely distributed from Croatia in the west to N Iran in the east via Turkey and the Caucasus. Within the range of this species, the populations in the Caucasus (including the Transcaucasus), the Talish Mountains and the Elburz Mountains are diverse and polytypical, both in morphology and chromosome number. They have been treated as five allopatric subspecies with subsp. coriifolia (diploid) in the lowlands of Georgia, subsp. mlokosewitschii (diploid) in E Georgia, NW Azerbaijan and Dagestan in Russia, subsp. wittmanniana (tetraploid) in the highlands of Abchasia (Georgia) and adjacent regions, subsp. macrophylla (tetraploid) in the western Transcaucasus (Armenia, Georgia and NE Turkey), and subsp. tomentosa (tetraploid) in the Talish Mountains (Azerbaijan and Iran) and the Elburz Mountains (Iran) (Hong & Zhou, 2003)

In Paeonia daurica, the number of leaflets/ leaf segments of lower leaves is mostly 9, rarely 10, very occasionally 11, whereas in P. mascula this ranges from 11 to 22, rarely 10, very occasionally 9. This character is rather stable within populations. ln the population D. Y. Hong et al. H02215 (Mt Amanos, Hatay, Turkey), for example, we observed 37 individuals that all had 9 leaflets/leaf segments. On this mountain, there were hundreds of P. daurica individuals at altitudes from 1,300 to 1,550 m, but very few of them had more than 9 leaflets/leaf segments. The length of the terminal leaflets varied from 5.5 to 15.2 cm (9.2 ± 1.9 cm) in P. daurica, compared with 7.7—16.7 cm (11.8 ± 2.5 cm) in P. mascula. The ratio of the length to width ranged from 1.01 to 1.82 (1.45 ± 0.23) in P. daurica, and from 1.33 to 2.18 (1.79 ± 0.22) in P. mascula. The apex of terminal leaflets was mostly truncate, broad- rounded or rounded in P. daurica, but mostly acute, cuspidate or rounded—cuspidate in P. mascula. The widest point of the terminal leaflets also differed between the two species. Although the widest point was above the middle of the terminal leaflets in both entities, it was much above the middle, about halfway between the top and the middle in P. daurica, but not much above the middle in P. mascula. Thus, the terminal leaflets were broad-obovate or nearly orbicular in P. daurica, but obovate, oblong or ovate in P. mascula. As indicated by Hong et al. (2006), P. daurica was clearly, though not distantly, differentiated from P. mascula in morphology. They were not intermingled with each other, even in S Turkey where the two entities were sympatric. Morphologically, they differed in number of leaflets/ leaf segments of lower leaves, and also in the shape of the terminal leaflets. The great majority of individuals of these two species could be distinguished.

Stearn and Davis (1984) stated that P. daurica (= P. mascula subsp. triternata) and P. mascula geographically overlapped in the eastern Aegean islands (Lesvos and Samos), and that they showed some morphological overlap in Anatolia. We critically examined all the specimens of this group from the regions mentioned above at the herbaria ATH, E, G, LD and UPA, but found that the specimens from Lesvos and Samos, including E. Stamatiadou 2666 (ATH) determined by P. H. Davis as P. mascula subsp. triternata, all belonged to P. mascula subsp. mascula. Thus, no subsp. triternata was present on these two islands. In Anatolia, subsp. triternata and subsp. mascula did coexist in some areas, e.g. in Hatay Province, Turkey, but they were not found to overlap morphologically. Instead, they were rather distinct. Their morphological differentiation can be easily understood when their difference in chromosome number is considered. No hybrids between them have been discovered.

Seven subspecies are recognised and keyed out below.

1a. Sepals often villose on the abaxial side; leaves rather densely villose on the lower side: subsp. velebitensis D. Y. Hong

1b. Sepals glabrous; leaves sparsely villose or puberulous, less frequently densely villose on the lower side or glabrous.

     2a. Carpels glabrous or nearly glabrous; petals yellow.

          3a. Leaflets/leaf segments densely villose and thus greyish beneath: subsp. macrophylla (Albov) D. Y. Hong

          3b. Leaflets/leaf segments usually sparsely villose beneath: subsp. wittmanniana (Hartwiss ex Lindl.) D. Y. Hong

     2b. Carpels tomentose; petals red, rose, white or yellow.

          4a. Leaflets/leaf segments puberulous or glabrous beneath, obovate, apex rounded or obtuse, often with a short mucro: subsp. mlokosewitschii (Lomakin) D. Y. Hong

          4b. Leaflets/leaf segments villose or glabrous beneath, obovate, oblong or wide-elliptic, apex rounded to short—acuminate.

               5a. Petals red or rose; leaflets/leaf segments glabrous or sparsely villose beneath.

                    6a. Leaflets/leaf segments broad-obovate, truncate to rounded at apex: subsp. daurica

                    6b. Leaflets/leaf segments obovate to oblong, rounded to acute at apex: subsp. coriifolia (Rupr.) D. Y. Hong

               5b. Petals yellow or yellowish white, but sometimes red at periphery or with a red spot at base; leaflets or segments villose beneath.

                    7a. Leaflets/leaf segments mostly densely villose and thus greyish beneath: subsp. tomentosa (Lomakin) D. Y. Hong

                    7b. Leaflets/leaf segments usually sparsely villose beneath: subsp. wittmanniana (Hartwiss ex Lindl.) D. Y. Hong

Note from the editor:

The placing of P. mlokosewitschii as a subspecies to P. daurica is controversial for many people, as previously it was always considered a separate species. Adding to the controversy, another (sub)species, wendelboi, has been named since Hong’s book was published. It is unclear whether this should be considered separate or as a subspecies of mlokosewitschii or not.

Assadi considers the main morphological difference to be the foliage, but he states “[In P. mlokosewitschii] leaflets are narrow ovate to ovate (not broadly elliptic to broadly ovate [as in P. wendelboi])”(2). This would clearly separate them both, but Hong states that “its leaflets [of mlokosewitschii] are usually obovate, rounded and mucronate at the apex.”(3) Both statements are at odds with one another unfortunately. From the measurements given, it is clear that mlokosewitschii has different foliage however. Assadi gives the following for wendelboi: “4-8 × 3.5-6 cm, orbicular to broadly elliptic” whilst Hong gives these for daurica (alas no specifics for mlokosewitschii): “8—17 cm long, 4.8—11.5 cm wide.” Thus wendelboi has both smaller leaflets that are nearly as wide as they are long (thus ‘elliptical’ indeed), whilst mlokosewitschii has larger leaflets that are less wide in comparison.

It is quite difficult to separate the species, as Hong also states: “…the P. daurica complex in the Caucasus has notable polymorphism, and no distinct differences have been found between the populations. It is thus considered that the group has differentiated only at the stage of geographical races…”.(4) The natural population of wendelboi is located some 500 km southwards of mlokosewitschii and whereas mlokosewitschii grows on slopes in or at the edge of deciduous forests, wendelboi grows on rocky slopes with very sparse low vegetation in full sun, facing south west. Wendelboi also has a much shorter stature (this remains so when grown at other localities) and the petal color is always ‘sulphur yellow’ whereas in mlokosewitschii petal color varies widely.(5)

Note 2: A new species or subspecies has recently been added to the puzzle: Paeonia archibaldii. For now it has been placed as a subspecies to daurica on this site.

  1. Hong, De-Yuan. “Peonies of the World. Taxonomy and phytogeography.” Kew: Royal Botanic Gardens, 2010, pp. 168-171.[]
  2. Assadi, M. “A taxonomic revision of the genus Paeonia (Paeoniaceae) in Iran.” In: Iran. J. Bot., 2016, vol 22, no 2, pp. 75-78.[]
  3. Hong, De-Yuan. “Peonies of the World. Taxonomy and phytogeography.” Kew: Royal Botanic Gardens, 2010, pp. 168-176.[]
  4. Hong, De-Yuan & Shi-Liang Zhou. “Paeonia (Paeoniacea) in the Caucasus.” In: Botanical Journal of the Linnean Society, 2003, vol 143, pp 135-150.[]
  5. Ruksans, Janis & Henrik Zetterlund. “An Iranian peony to honour Per Wendelbo.” In: The Alpine Gardener, 2014, vol 82, no 2, pp. 230-237.[]

2020, The Peony Society -

Privacy Preference Center

Log in with your credentials


Forgot your details?

Create Account