Euphytica 36 (1987) 535-539
Institute for Horticultural Plant Breeding (IVT), P.O. Box 16,6700 AA Wageningen, the Netherlands Received 27 October 1986


The inheritance of the dwarf character was studied in Fl-seedling populations arisen from crosses between diploid Polyantha cultivars and the diploid dwarf species R. chinensis minima (SIMMS) VOSS. Dwarfness is controlled by a single dominant gene D. R. chinensis minima and its dwarf descendants are heterozygous for D, while polyanthas are homozygous for d. The origin of R. chinensis minima and its potential for breeding new pot roses are discussed.


The most important class of dwarf roses is that of the Miniatures, the cultivars of which are all descended from the diploid (2n = 2x = 14), recurrent flowering dwarf species R. chinensis minima (SIMMS) VOSS. Its oldest cultivar, Pompon de Paris, was a popular pot plant in France between 1820 and 1850 (SHEPHERD, 1954), but fell in disuse in later years. The cultivar was rediscovered in Switzerland in 1920 and then, as a new species, denominated R. rouletii by Correvon (KRÜSSMANN, 1951), under which name it is well known among rosarians (Dubois, 1983). Not untill 1930, R. rouletii was used by Jan de Vink (Netherlands) and Pedro Dot (Spain) to create the first Miniatures (VAN DE LAAR, 1968; FITCH, 1977), which are, like the species, still the smallest roses in cultivation. The mode of inheritance of the dwarf character has not been investigated, but was assumed to be recessive.

In the scope of a breeding programme for pot roses aiming at the rare combination of dwarf and clustering habits, polyantha-like species and cultivars were pollinated with R. chinensis minima and two of its direct dwarf descendants. The present paper reports on the inheritance of the dwarf character in the progenies obtained.


Plant material consisted of diploid and recurrent flowering species, selections, and named cultivars. R. chinensis minima (SIMMS) VOSS was obtained as 'Pompon de paris' (MEIKLE, 1980) from the Royal British Rose Society. Selections 8026 and 8027 are pronounced dwarfs from a Fl-population of R. chinensis minima and the polyantha 'Gloria Mundi'. R. multiflora nana (MEIKLE, 1980), selections 8029, 8401 and the cultivars are classed as polyanthas (SHEPHERD, 1954). The above material was selfed and pollinated with R. chinensis minima, 8026 and 8027. Only successful crosses are presented in Table 1. Seedlings were grown in pots on benches in a heated greenhouse in the spring of 1984 and 1985. At anthesis seedlings were classed as 'dwarf' or 'normal'. In a few populations the length of the main shoot was measured at that time.


In all populations with R. chinensis minima, 8026 and 8027 as a progenitor, a number of dwarf seedlings occurred which, as to plant height, flower- and leaf size and internode length could be easily distinguished from the 'normal' ones (Fig. 1). No dwarfs occurred in the selfpopulations of the polyanthas (Table 1). On account of the ratios normal: dwarf obtained, a 1:1 or 3:1 segregation was supposed. Chi-square analysis yielded probabilities that showed a good to very good fit for this hypothesis. These results indicate that the dwarf character in crosses with R. chinensis minima is determined by a single dominant gene, which is proposed to be denominated D. Consequently, in R. chinensis minima, no's 8026 and 8027 the dwarf character is controlled by Dd, whereas in the polyanthas involved it is controlled by dd. Although the segregation into dwarfs and normal plants was quite conspicuous, within each category plants showed some variation for shoot length at anthesis, which is illustrated for the population 84381 'The Fairy' x R. chinensis minima in Fig. 2. It can be seen that variation for length was smaller for dwarf than for normal seedlings.

Particularly in the populations 84370, 84372, 84373 and 84448, a number of seedlings could be selected that combined the dwarfing and clustering habits.

Fig. 1. Segregation for dwarf (left) and giant (right) seedlings in the population 84330 'Meinadentel' x R. chinensis minima.

Table 1. Segregation for dwarf and normal seedlings, in diploid populations of
polyantha x R. chinensis minima (SIMS) VOSS, its direct descendants, and in selfings.

Population Cross   Number of seedlings Expected
  dwarf normal
83042 R. multi. nana x R. chin. minima 16 13 1 : 1 0.58
83400 8029 x 60 54 1 : 1 0.57
84317 8401 x 19 17 1 : 1 0.74
84330 Meinadcntcl x 24 25 1 : 1 0.89
84342 New Penny x 4 3 1 : 1 0.71
84370 Katharina Zeimet x 9 11 1 : 1 0.66
84381 The Fairy x 38 36 1 : 1 0.82
84442 Marie Pavic x 5 5 1 : 1 1.00
84448 Yvonne Rabier x 3 4 1 : 1 0.71
84466 Kathleen x 17 14 1 : 1 0.59
84346 New Penny x 8026 4 5 1 : 1 0.74
84372 Katharina Zeimet x 55 50 1 : 1 0.63
84387 The Fairy x 69 64 1 : 1 0.67
84331 Meinadentel x 8027 25 24 1 : 1 0.89
84373 Katharina Zeimet x 90 81 1 : 1 0.49
84388 The Fairy x 6 7 1 : 1 0.78
83402 8027 x R. chin. minima 10 4 3 : 1 0.76
83404 8026 x 17 4 3 : 1 0.53
84366 8026 x 32 11 3 : 1 0.93
84325 8401 x self 0 64 - -
84334 Meinadentel x self 0 28 - -
84340 New Penny x self 0 13 - -
84380 Katharina Zeimet x self 0 52 - -
84392 The Fairy x self 0 81 - -
84447 Marie Pavic x self 0 11 - -
84454 Yvonne Rabier x self 0 23 - -
84465 Kathleen x self 0 38 - -

Fig. 2. Distribution of dwarf (Dd) and normal (dd) seedlings of 'The Fairy' x R. chinensis minima (SIMS) VOSS, over classes of shoot length at anthesis (n = 74).


In Rosa single gene control of characters appears to be rare; so far it has only been shown for recurrent flowering (SEMENIUK, 1971a; 1971b; DE VRIES & DUBOIS, 1978) and for moss (DE VRIES & DUBOIS, 1984).

Dominant control of dwarfness in roses is of importance to breeding. Firstly, it explains why the pioneers De Vink and Dot could select Miniatures in Fl-populations of R. chinensis minima and vigorous cultivars, and how the dwarf character could be transmitted to current dwarf roses at the triploid and tetraploid levels. For that matter, it is notable that triploid and particularly tetraploid 'Miniatures' are more vigorous than their diploid ancestors. This may be either due to heterozygosity for D, or to a generally greater vigour of polyploid roses.

Dominance of dwarfness allows the early selection of dwarf seedlings in populations arisen from crosses between Miniatures and desirable, but non-recurrent flowering rose species (DUBOIS & DE VRIES, 1984). As dwarfs can be recognized already a few months after seed germination, which is about one year before actual flowering occurs, dominance also greatly attributes to the efficient use of greenhouse area and hence to energy.

The origin of R. chinensis minima is uncertain. As the species has not been found in natural conditions in East Asia (SHEPHERD, 1954; HARKNESS, 1978), its import from that region was considered to be improbable. However, in the beginning of 1800, the botanists R. Sweet and A. P. De Candolle (SHEPHERD, 1954) both claim to have received it directly from the island of Mauritius, which at the time was an important intermediate station in the route from the Far East to the Cape.

*Ralph Moore (1967) raised minis from self-seeds of 'Old Blush'.

According to HURST (1941), R. chinensis minima would be a Europe-born open pollinated seedling of the Stud China 'Parsons Pink China',* whereas HARKNESS (1978) claims it to be an open-pollinated seedling of 'Slaters Crimson China'. Because of corresponding leaf shapes the latter supposition would be the most likely one, but in view of the dominance of the dwarf character segregation of either two normally vigorous cultivars into dwarfs seems improbable. It is not unlikely therefore that the species was imported from Mauritius after all, where, since roses are not indigenous there, it may have temporarily been nursed on its way from East Asia to West Europe.