Journal of Heredity 62: 203-204. (1971)
Inheritance of recurrent blooming in Rosa wichuraiana
Pete Semeniuk

The polyploidy characteristic of modern roses and their interspecific hybrid nature has made segregation of genetic characters so complex that clear-cut examples of Mendelian inheritance in the genus have not been described, and very little is known about the genetic basis of any important character. Before 1810 nearly all the cultivated European roses were summer flowering. The introduction of perpetual flowering derivatives of R. chinensis Jacquin, and R. odorata gigantea var. (Crepin) Rehd. & Wils. From the Far East played an important part in the development of our modern recurrent blooming roses. In 1810, first generation nonrecurrent blooming hybrids began to appear between the recurrent blooming China derivatives and nonrecurrent flowering groups of roses. These hybrids, when used in crosses with other roses having an R. chinensis background, passed the recurrent blooming habit to some of the progeny. Hurst1 suggested that the habit of continuous flowering (recurrence) is due to the action of a single Mendelian recessive gene introduced into our modern roses by China roses. Other workers2,3 have made similar observations that indicate simple Mendelian inheritance, but final proof has been lacking.

A variant of R. wichuraiana Crep., upon selfing, segregated into two distinct classes, recurrent vs. nonrecurrent blooming. This phenotypic variation within a self fertile group of roses provided an opportunity to determine the genetic basis of this character.

Methods and Materials

Beginning in summer, 1960, and ending in 1968, all selfing and crossing was done in a screened greenhouse. Selfs and crosses made by hand-pollination each summer produced fruit that matured in late fall. After each harvest, seeds were after-ripened and seedlings grown in the greenhouse. Seeds of rose species are dormant when mature and require a period of after-ripening at low temperatures before germination. The cold requirements of selfs and crosses of R. wichuraiana, F1, F2, F3, testcrossed and backcrossed seeds were satisfied by continuous exposure to 40° F. for 120 days. The time from pollination to flowering of the next generation with nonrecurrent blooming roses was two years. After each fruit harvest in the fall, nonrecurrent blooming plants were transferred to cold frames. After adequate winter cold, they were brought back to the greenhouse for flowering and pollination.

Table 1. Inheritance of flowering types in progeny from selfing and crossing in R. wichuraiana

  Seedling flowering types
Pollination No. seed No. germ. nonrecurrent recurrent
Nonrecurrent R. wichuraiana selfed 294 226 226  
Recurrent R. wichuraiana selfed 494 373   373
F1 227 186 186  
F2 216 158 121 37
F1 x recurrent (testcross) 1144 960 517 443
F1 x nonrecurrent (backcross) 776 502 502  

Results and Discussion

Progeny from successful selfing of nonrecurrent R. wichuraiana in 1960 included both recurrent and nonrecurrent bloomers. Seedlings of both types were selected to produce another self-generation. The first self-generation recurrent blooming seedlings produced 494 seeds. All 373 seed that germinated were recurrent bloomers. These plants bred true in all cases, confirming that they were homozygous for recurrent blooming.

Selfing nonrecurrent blooming plants produced two kinds of progenies. Approximately one-third of the progenies included only nonrecurrent bloomers. Two-thirds of the plants produced progenies that included both recurrent and nonrecurrent bloomers. Two-hundred ninety-four seeds were produced in the first self-generation of the first kind of nonrecurrent blooming plants. All 226 seed that germinated grew into nonrecurrent plants.

When pure breeding nonrecurrent and recurrent plants were crossed to produce the F1, the progeny were all (186 plants) nonrecurrent. The nonrecurrent factor (R) is completely dominant over non recurrent (r).

Upon self-pollination, the F1, nonrecurrent (R/r) individuals produced an F2 that contained nonrecurrent and recurrent progeny in a 3:1 ration (121R/-:37r/r) (Table 1).

A testcross gives precise information about a segregating population. It also enables the experimenter virtually to look at the gametes of the F1 individual, since the recessive gametes can in no way mask the appearance of the dominant type, R. Crossing of heterozygous F1 by the recessive parent r/r produced 517 R/r plants and 443 r/r plants, very close to the expected 1:1 ratio (Table 1). Testcrosses between F1 and recurrent bloomers indicated that recurrent bloomers differ from nonrecurrent bloomers by a single recessive factor.

From a backcross between F1 (R/r) and homozygous dominant parent (R/R) 502 nonrecurrent bloomers were produced. The results of breeding tests indicate that nonrecurrent blooming is inherited as a simple Mendelian dominant.


In R. wichuraiana the dominant gene R for nonrecurrent flowering and its recessive allele r are inherited in simple Mendelian ratios. The F2 population indicated that nonrecurrent bloomers were dominant to recurrent, and due to a single gene difference. Emergence of recurrent bloomers in the F2 demonstrated that the recessive factor was not modified in the F1 hybrid. The recurrent factor had probably arisen by a single gene mutation from the normal, nonrecurrent wild-type condition.

Literature Cited

  1. Hurst, C. C. Notes on the origin and evolution of our garden roses. J. Roy. Hort. Soc. 66: 73-82. 1941.
  2. Shepherd, R. E. History of the Rose. Macmillan, New York. 1954.
  3. Wylie, A. E. Masters Memorial Lecture, The History of the Garden Roses. J. Roy. Hort. Soc. 79: 555-571. 1954.