Arilpums Are Here!

by Tom Waters

When I last wrote about my progress with hybridizing arilpums in September, I was eagerly anticipating this spring's bloom, with many new seedlings expected to bloom on established plants. I was not disappointed!

Aladdin's Gem

First, a little recap. Arilpums are hybrids between pure arils and the dwarf bearded species Iris pumila. When a tetraploid aril is used as the aril parent, theory predicts that the arilpum seedlings will be fully fertile amphidiploids. Historically, most arilbred dwarfs and medians have been produced by using arilbreds, diploid arils, and SDBs in various combinations, and the result has been infertile offspring, making it impossible to linebreed the smaller arilbreds to refine them and extend their potential. Arilpums are the way out of that difficulty. However, only a few have been registered and introduced: 'Barbarella' (Mathes, 1991), 'Aladdin's Gem' (Thoolen, 2002), and 'Topaz Talisman' (Jensen, 2015). Although producing fertile pollen, arilpums so far have shown little or no pod fertility, a disappointing setback.

Last spring, I had some luck with the two arilpum seedlings I bloomed, Q025-01, 'Merlin's Magic' X Armenian pumila seedling, and Q053-01, 'Tadzhiki Bandit' X 'Sudden Butterflies'. Both produced seeds for me, as did a seedling of George Hildenbrandt's, GH-11-4-2, 'Dunshanbe' X 'Hidden Dragon'. So it seemed there was hope after all for creating a line of fertile arilpums!

This spring reinforced my optimisim, and also gave a preview of the great variety of color, pattern, and form that is possible with arilpums.

Q068-14

Q068-04, with "monster beards"

Earliest to bloom were seedlings from cross Q068, 'Dunshanbe' X 'Wild Whispers'. I had over a dozen of these, most dark violet selfs with some aril look but without much distinctiveness. Two stood out, however. Q068-04 has massive caterpillar beards. Not exactly beautiful, but certainly memorable. Q068-09 is the best of the lot, with a clean oncocyclus shape and elegant look. 

Q068-09, the standout

The second cross to bloom was really exciting: Q066, 'Tadzhiki Bandit' X Caucasus pumila seedling. These were much smaller and more varied than the 'Dunshanbe' seedlings, showing many exciting colors and patterns. I've kept almost everything from this cross for further evaluation.

Q066-06

Q066-02

There was another from cross Q053, looking almost the same as its sibling that bloomed last year. I also had a few seedlings from cross Q070, 'Vera' X 'Royal Wonder', one of which was deemed worthy of a closer look.

Q070-01

I made a number of crosses with this year's arilpums, but was not very systematic or thorough about it. Seeds have not been harvested yet, so it is premature to say much about their pod fertility. But there are pods, even some bee pods, so I am optimistic. 

It's been about ten years work to get to this point, but at last I have a small cadre of arilpum seedlings to work with, and the future looks very promising.

Progress with Arilpums

by Tom Waters

When I wrote my previous blog post on arilpums early in 2019, I had to focus mostly on history and theory, as I did not yet have much to show from my own seedlings. That has now changed, and the current state of this project is even better than I had dared hope.

As explained in the previous post, arilpums are arilbreds dwarfs that come from crossing pure arils with the dwarf bearded species Iris pumila. If the aril parent is a tetraploid, theory says the resulting arilpums should be fertile amphidiploids, a sort of arilbred analog of the SDBs. They thus have the potential of ushering in a whole new "fertile family" of miniature arilbreds.

Waters Q025-01

This spring, I bloomed two arilpum seedlings, both prolific two-year-old clumps covered in bloom. The first, Q025-01, from 'Merlin's Magic' (a selection of Iris stolonifera) x P002-05, a pumila seedling of Armenian origin, has a dainty Regelia look and some nice veining. The second seedling, Q053-01, from 'Tadzhiki Bandit' x 'Sudden Butterflies' is larger and huskier, with flowers reminiscent of tall arilbreds with Regelia ancestry, such as 'Stars Over Chicago' or 'Saffron Charm'.

I also had good bloom this year on an arilpum seedling from

Waters Q053-01

George Hildenbrandt, GH-11-4-2, from 'Dunshanbe' x 'Hidden Dragon'. This is a charming dark arilpum with nice aril-like flower form.

Hildenbrandt GH-11-4-2

The remarkable thing is that all three of these seedling produced seed for me this year when crossed with other arilpums. In my previous post, I mentioned that earlier hybridizers working with arilpums had found them to be infertile as pod parents. That is clearly (and thankfully!) not a general rule! Here are my successful arilpum x arilpum crosses this year:

Q053-01 x Aladdin's Gem gave 38 seeds

Q025-01 x Q053-01 gave 9 seeds

GH-11-4-2 x Q053-01 gave 23 seeds

GH-11-4-2 x Aladdin's Gem gave 3 seeds

These are not huge seed yields, but they are certainly adequate for continued breeding. I am quite optimistic that a number of the second-generation seedlings from these crosses will be fully fertile, and it will be possible to line breed arilpums without having to constantly go back to the difficult aril x pumila initial cross. This will also open up arilpum breeding for hybridizers who live in areas where either arils or pumila are difficult to grow.

Next spring should offer even more excitement, as I have more than 30 arilpum seedlings from various crosses that will be blooming as two-year-old plants. I got a sneak preview of one this spring,  Q066-08, from 'Tadzhiki Bandit' x P018-02, a pumila seedling originating from the Caucasus mountains. It looks for all the world like a miniature rendition of its aril pod parent, giving a rather more "onco" impression than others mentioned in this post. It has not yet been tested for fertility.

Waters Q066-08

All this has been an exciting vindication of a breeding project motivated mostly by theoretic possibilities rather than concrete experience. I'm now confident that arilpums have a bright future ahead.

Arilpums: Fertile Arilbred Dwarfs!

by Tom Waters

'Aladdin's Gem' (Thoolen, 2002)
OGB/ABD

Most arilbred medians are produced by crossing standard dwarfs (SDBs) with fertile halfbreds (OGB). The results of such crosses tend to be pleasing garden subjects, occupying about the same niche as intermediates (IBs), but with the added interest of arilbred color patterns. Unfortunately, arilbred medians produced this way are usually infertile, or have very limited fertility at best. Also, being only ¼ aril (OGB-), they don’t always show enough pronounced aril characteristics to really stand out as arilbreds.

For those of us who enjoy the smaller arilbreds, the holy grail would be diminutive fertile arilbreds that are half aril in content. We might imagine miniaturized versions of the taller arilbreds that have come from breeding arils with tall bearded (TBs). One approach to realizing this dream is through the creation of “arilpums”: hybrids between arils and the tiny dwarf bearded species Iris pumila. This type of cross produces true arilbred dwarfs, around 10-12 inches (25-30 cm) in height, fertile, and half aril. Additionally, Iris pumila is expected to confer winter hardiness on its seedlings.

The potential of such breeding was first appreciated in the mid 20^th century, when iris breeders became aware of the cytogenetics of irises, and the possibility of creating “fertile families” of irises from wide crosses, if the parents are chosen to have the right chromosome configurations to produce fertile offspring. Both the SDBs (from TBs crossed with Iris pumila) and the fertile arilbreds (from TBs crossed with arils) are examples of such fertile families, and show how varied and successful they can be.

The most direct way to build a fertile family of arilpums is to cross Iris pumila with tetraploid arils. In nature, the only tetraploid arils are the Regelia species Iris hoogiana and Iris stolonifera. Indeed, the pioneering hybridizer Paul Cook (who also introduced the first SDBs from TB x pumila crosses) produced two hoogiana/pumila hybrids in the 1950s: ‘Hoogpum Blue’ (Cook, 1956) and ‘Hoogpum Purple’ (Cook, 1956). Both were fertile, but Iris hoogiana doesn’t have many characteristics that most of us think of as epitomizing the “aril look”, and these little offspring were not very distinguishable from MDBs, and in fact were registered as such.

The first approach to arilpums with oncocyclus ancestry came from Peter Werckmeister, the great expositor of cytogenetic theory to English-speaking irisarians. In the 1970s, he produced “Miltonia” (not registered) from a lucky cross, in which the aril parent (an oncogelia from ‘Teucros’ x Iris susiana), although a diploid, acted as a tetraploid by producing an unreduced gamete. Werckmeister reported eventually creating a whole array of fertile arilpums from “Miltonia”. Alas, they were not widely distributed and are now lost to us.

Thoolen arilpum seedling AP9007-1

Interest in arilpums underwent something of a revival in the 1990s, after Samuel Norris and John Holden created a line of tetraploid oncogelias from using colchicine treatment. Crossing these with Iris pumila produced arilpums with much more “onco look” than had been possible previously. The first of these to be introduced was ‘Barbarella’ (Mathes, 1991), followed by ‘Aladdin’s Gem’ (Thoolen, 2002). Both Harald Mathes and Francesca Thoolen raised a number of different arilpum seedlings, but these were the only two registered and introduced into commerce. Both ‘Barbarella’ and ‘Aladdin’s Gem’ produce fertile pollen, but are unwilling pod parents. Mathes and Thoolen noted similar difficulties with their other arilpums, and their breeding programs were eventually abandoned.

The lack of pod fertility was an unfortunate setback, because the simplest way to build up this family would be to cross the existing arilpums among themselves, rather than repeating the initial aril x pumila crosses.

'Topaz Talisman' (Jensen, 2015)
RB/ABD

The most recent arilpum is the lovely stolonifera/pumila hybrid ‘Topaz Talisman’ (Jensen, 2015). It has not yet been tested extensively for pod fertility.

Why has this family not made more progress? It has been more half a century since its potential was first understood, theoretically. I do not believe the pod fertility issue is insurmountable, although it is a setback and perhaps kept the family from taking off in the 1990s, when interest was greatest. Since then, the number of hybridizers working with arilbreds has declined, and not all are interested in such an experimental project. The parent plants, tetraploid arils and Iris pumila, are not easy to grow in all locations: the arils need a very dry summer, and pumila needs a cold winter.

Another obstacle is that the initial crosses are difficult, not many viable seeds are produced, germination is poor, and so a lot of persistence is needed to bring arilpum seedlings to bloom. Since I began working in this area six years ago, I have made 365 pollinations, resulting in 525 seeds, and at long last four seedlings which I hope to see bloom this year or next.

I would love to see others get involved; the prospects for eventual success obviously depend on a number of people working with determination over the course of a number of years. Here is the recipe:

1. Acquire tetraploid arils, such as Iris stolonifera, ‘Tadzhiki Bandit’, ‘Tadzhiki Eclipse’, ‘Werckmeister’s Beauty’, ‘Balalaika Music’, and ‘Dunshanbe’. Keep in mind that you must use tetraploid arils – crossing Iris pumila with random arils and arilbreds offers no special hope of fertility in the seedlings.
2. Acquire Iris pumila, either from seed exchanges, or named cultivars such as ‘Royal Wonder’, ‘Wild Whispers’, ‘Little Drummer Boy’, ‘Hobbit’, and others.
3. Make crosses between these two groups, as often as possible, and mixing and matching parents as much as possible.
4. Also get any available arilpums (pretty much just ‘Aladdin’s Gem’ and ‘Topaz Talisman’ these days), to cross amongst themselves or with arilpum seedlings you produced from the aril/pumila crosses.

Hildebrandt arilpum seedling 11-4-2
(photo: George Hildebrandt)

I think this a project that offers enormous potential in the long term, and is ideal for a small, backyard hybridizer who does not have the resources to grow thousands of seedlings and does not need to worry about immediate commercial success. There is a whole new world waiting to be opened up!