Canadian Rose Annual 1995 pp 55-57
Dr. I Meneve,
Research Station for Ornamental Plant Growing Caritasstraat 21, B-9090-MELLE Belgium

[Editor's Note. This article was previously published in the magazine Horticulture Week and our gratitude is expressed to them for permission to republish it here.]

Rose breeding has a long history. Since the Dutch and French horticulturists started breeding roses in the 18th century, the number of rose cultivars has grown spectacularly.

Before about 1800 roses in cultivation were hardy shrub roses, mainly summer-flowering. With the introduction of perpetual flowering roses from the Far East at the end of the 18th century, the door to the so-called 'modern' roses was opened, resulting in Noisettes, Bourbons, Tea-Roses and Hybrid Perpetuals.

Modern roses are mostly tetraploid and very heterozygous. Crossing within the existing range results in a very varied progeny. Since the breeding work in the last few decades was done on a very large scale we can say that this group of roses is virtually outcrossed, and so it has become difficult to raise something new starting from this assortment.

The actual breeding is almost all based on line-breeding, i.e. auto-fertilization of roses, and family breeding where crosses are made between existing cultivars and unnamed seedlings. Selection in this seedling population over a period of several years results in new cultivars which are different from the existing ones, but not really "NEW" in the right sense of the word.

In the last few year some famous breeders (Illsink, Kordes, Lens, Meilland) have made a valuable contribution by introducing ground-covers and modern shrub roses, which are very popular for landscaping.

The history of garden roses gives us an excellent source of new ideas to introduce in the breeding programme of roses. The Masters Memorial Lecture of Ann P. Wylie, from the Department of Botany of the University of Manchester and published in 1954 under the title The History of Garden Roses, provided an excellent guide for this purpose. One of Mrs. Wylie's conclusions was: "Breeders will have to turn more and more to little or unused species in their quest for new characters".

The numerous sprayings of garden roses with fungicides against black-spot and powdery mildew, with more or less success, was the incentive to include the species R. hugonis, R. rugosa and R. fedtschenkoana in the breeding programme of the Station for ornamental Plant Growing at Malle (Belgium). Since the first two species are diploids, it was necessary to raise them to the tetraploid level in order to avoid the triploid stage (2n x 4n -> 3n), which is the cause of sterility in the F1 generation.

It should be mentioned here that Kordes already used R. rugosa in his breeding programme using R. kordesii Wulff. This rose originated by spontaneous chromosome doubling of 'Max Graf', which is probably a hybrid of R. rugosa and R. wichuraiana, both diploids. However this cannot be considered 100% rugosa and so I thought it was worth while to start over with the species R. rugosa itself.

Because R. fedtschenkoana is a tetraploid species, successive crossing with other roses gave no problem.

To obtain a tetraploid level, seedlings of R. hugonis and R. rugosa were treated at the cotyledon stage with colchicine at 0.2% concentration. A drop of this solution was placed between the cotyledons on the growing tip or apex. We only obtained about 50 seedlings of R. hugonis since seed setting is very poor. R. rugosa, on the other hand, gave abundant seeds so we had about 1200 seedlings at our disposal.

Due to the toxicity of colchicine a great percentage of the seedlings died or showed marked malformation after some weeks. The seedlings that survived the treatment were checked visually for abnormalities to foliage and plant growth. It is known that doubling of the chromosome numbers results in coarse leaves, bigger flowers and fruits and a sturdier growth. In the approximately 20 R. hugonis seedlings that survived, no aberrant seedlings were discovered. Microscopical checking of cells in the root tip still showed diploidy.

Out of the much greater population of R. rugosa seedlings, 12 plantlets were selected for their growth and foliage characteristics. After microscopical checking, 3 of them showed tetraploidy level. These plantlets were carefully planted in the glasshouse, before transplanting outside. The growth and the size of the foliage showed marked differences in comparison to the diploid species R. rugosa itself.

After two years the tetraploid plants flowered for the first time, the flowers were of a larger size but the colour was the same as the diploid species. The pollen of these flowers was carefully collected and flowers of existing varieties and unnamed seedlings were pollinated with it.

The first generation (F1) of this crossing were all shrubs which grew up to 1.50-2.00 m in the glasshouse. Of these approximately 300 seedlings some flowered during the first year, but it was too late in the season to use the pollen for breeding purposes. Many of the most vigorous seedlings were already discarded in the first year, due to lack of space in the glasshouse.

The second year most of the 120 conserved seedlings flowered in May-June and pollen was collected from the best flowers to obtain a F2 generation using rose bushes as mother plants. In this F2 generation, composed of about 500 seedlings, some seedlings showed the normal growth of bedding roses and flowered also the first year.

These seedlings were budded on R. canina in the open. The next year the budded plants were checked for disease resistance. Almost all of them showed a complete immunity to blackspot and powdery mildew. The best of them (according to the form of the flower and the growth habit) were planted in the glasshouse to continue the breeding work.

These F2 plants were notably fertile and were reciprocally crossed with existing bedding roses. The subsequent years selections of these seedlings were crossed over and over with the existing assortment.

As we would expect, a great percentage of the seedlings of the subsequent generations lost their immunity against the rose diseases because the R. rugosa blood was diluted by the repeated crossings with the existing assortment.

These seedlings were immediately discarded and we continued with the resistant ones. After 10 years some selections were ready to be sent to the rose trials for evaluation. In 1990 the varieties 'Ville du Roeulx' and 'Melrose' performed very well in the rose trials of Paris (Bagatelle), Geneva, Dublin, Genova (Italy) and Kortrijk and Le Roeulx in Belgium. These varieties are distinct from the existing ones because their foliage shows a markedly different form and the plants need no spraying for protection against fungal diseases. Since the colour of the two named varieties and the other unnamed seedling is mostly in the range of rose to deep-rose, we shall try to introduce other colours in the future without, of course, losing the resistance against diseases.

A similar breeding scheme with R. fedtschenkoana, a tall shrub up to 2.50 m with corymbs of white colours, resulted in the remarkable variety 'Florange', an F6 seedling. This variety shows a very good resistance to diseases and won the Golden Rose in Kortrijk (Belgium) and a silver medal in Genova (Italy). It has glossy leaves and it is rather late flowering.

In the course of this breeding work it was a pity that we had to discard so much interesting material due to a lack of place and staff. Without doubt some very valuable material has disappeared this way.

In any event, this long term breeding programme finally resulted in us obtaining some new valuable rose varieties. These varieties are not only just new looking cultivars but constitute an important contribution to a better environment since they need not be sprayed at all. It is my wish that the further developments along these lines will yield in the near future still more new attractive cultivars which are 'friendly to the environment'.