Do you paint? One of my favorite painters is Monet. I love the way he uses light and shadow especially on reflective surfaces like water or the leaves of trees. Each painter has his or her own style and you can see it in the artwork that they produce. If you do paint, do you use paint brushes or your fingers? It has been years since I have finger painted with my kids.  It was lots of fun, but they always seemed to do a better job of it than I did.

Each spring for the last several years, my rose breeding has seemed very much like painting to me and because of the often unpredictable results, has become more of an art than a science. I mix this color rose with that color rose (via pollination), to see what happens. Sometimes when I mix red with white, I get pink, just like what you would expect. But other times, I may mix yellow with lavender and still get pink. Though many rose breeders use brushes when applying pollen to accomplish their pollinations (and to “mix” their colors), I just use my fingers. So, I guess that I still am finger painting!

When planning crosses, most people will try to cross one rose having a particular set of good traits with a second rose that has other good traits that would compliment the first rose.  An initial step then in setting up crosses might be to list potential parents along with their good traits.  I always start with the good seed parents since they are so much harder to find than the good pollen parents.  Then, I try to select the pollen parents that will make up for any deficits that the seed parents have.  Besides being familiar with the parents that you select, it is helpful to know something about the ancestry of your selected parents too.  Of course the parents that you select will depend to a large egree upon the characteristics that you value most in roses.

One sort of rose that I think would be very desirable to breed for would be a medium height bushy plant, with fewer thorns, having clean glossy foliage, and miniflora to floribunda sized blooms with exhibition form and at least some fragrance, carried in small clusters, and in warm colors.  With that ideal in mind, I would have to narrow down my list of potential rose parents in the garden that I might choose to use in breeding toward that goal.

To start with in the example below (see table), I have proposed using two seed parents that might work in moving toward that goal.  One is an orange miniflora seedling of mine, code named J29-4.  It has long lasting medium fragrant blooms coming one to a stem on a vigorous sturdy plant clothed in clean glossy foliage.  The other, code named J3-4, is a golden floribunda with a darker coppery reverse on a clean medium sized bush with clean semi-glossy foliage, with informal blooms.  Both of these roses set hips excellently, which is why I chose them to be seed parents.

Next, I have selected two possible pollen parents to use with the above seed parents to try to arrive at my goal.  The pollen parent code named I72-1 is an upright growing miniflora with dark yellow exhibition blooms growing on straight stems.  Although it does get some powdery mildew, it is rarely enough to be a problem.  The second proposed pollen parent is a floribunda with clean glossy foliage, carrying rich orange well-formed blooms coming most often one per stem.  The plant habit is a bit spreading, but growth produces a medium sized plant.

Table of selected rose traits, with goals and proposed parents

As I mentioned earlier, it is good to know the ancestry of your proposed parents in order to make some predictions about what you might see in the resulting seedlings.  The notation to describe parentage follows two simple rules.  The first rule is that the seed parent is listed first, followed by the pollen parent.  An “X” is used between them to indicate the two roses that were crossed together.  The second rule is that in complex parentages, brackets are used to indicate each level of parentage.  Similar to how they are used in algebra, they indicate the order of how each step is accomplished.  The most inner brackets happened first (they are the more remote parents), while the outer brackets indicate roses added to the ancestry more recently.  Different shaped brackets <{[()]}> are used below to help indicate the level of parentage.

With each of the four roses in mind, it is important to realize that there are more possible crosses that can be performed between them than what first meets the eye. First, each of the roses could be crossed with itself, as in a self-cross. Then each rose could be crossed with each of the other roses acting first as a seed parent and then in reciprocal crosses as a pollen parent. In all, there are 16 possible crosses that could be set up.

Since the preferred seed parents have already been determined and I do not want to do any self crosses since that has already been done to test germination rates, and I do not want to pair the seed parents together since neither has good exhibition form, there are only four possible crosses that I want to consider:

1. J29-4 X I72-1
2. J29-4 X J62-1
3. J3-4 X I72-1
4. J3-4 X J62-1

Looking at the table on the previous page, I see that all of the desirable traits are covered in the first and the fourth crosses listed above.  I could expect a tendency toward miniflora sized blooms when J29-4 is crossed with I72-1, while J3-4 X J62-1 would probably give more floribunda sized seedlings. These two combinations would probably also yield a greater color range than the above second and third crosses.  However, if the goal was to produce more yellow colored seedlings, I might want to try the third cross since both the seed and pollen parents have strong yellow tones.

When looking at the ancestry, if it would be desirable to cross 'Kardinal' traits with 'Fairhope' traits, to perhaps increase the likelihood of exhibition form, it can be seen that this could be accomplished with the first and fourth crosses.

In the end, I will probably do all four of the proposed crosses – mixing orange with dark yellow, orange with deep orange, gold with dark yellow, and gold with deep orange. And what do you think I will get? Probably mostly pink!