PLANT BREEDING Iris ‘It’s Magic’ with good yellow colour Canadian breeder and striking patterns. Note the wispy ALAN MCMURTRIE standards inherited from Iris danfordiae is revolutionising early-flowering bulbous irises by expanding the range of yellows and introducing new patterns Breeding new Reticulata ırises ULTIVARS DERIVED from New Reticulata iris cultivars are flowered species, that I collected in Iris reticulata, conveniently mostly sports from existing cultivars Turkey 30 years ago. C known as Reticulata irises, selected by companies growing them My aim is to produce cultivars have flowers that are mostly shades at field-scale. They include sports of with new colours and patterns that of blue or purple. Exceptions include blue-flowered ‘Harmony’, such as are robust, clump-forming and that near-white ‘Natascha’ from the ‘Alida’ (pale blue) and ‘Pixie’ (violet), come back year after year, while 1970s and pale blue and yellow and an as-yet unreleased purple and a giving a modest number of blooms. ‘Katharine Hodgkin’. The last, along near white. ‘Harmony’ is a reliable with the similar ‘Frank Elder’ and cultivar for forcing, so forcers favour New breeding stock ‘Sheila Ann Germaney’, are sterile others in its extended family. Purple- Iris danfordiae has beautiful lemon- hybrids between I. histrioides and flowered ‘George’ is well known for yellow flowers with a pleasant scent, I. winogradowii. sporting blue, and what was once but the plant does not reappear in Other recent developments known informally as ‘Blue George’ the garden. Dig it up after its first include sports from ‘Katharine is now ‘Palm Springs’. year or two of flowering and all you Hodgkin’ and ‘Harmony’, wild- However, I am improving the tend to find is numerous bulblets. collected bicolours, and a couple range of Reticulata irises by using Growers speak of the original bulbs of sterile hybrids between ‘Cantab’ material collected from the wild, as having ‘shattered’ – they planted and I. winogradowii. including I. danfordiae, a yellow- nice large bulbs, but all they find in 20 March 2016 PlantsmanThe Parent species used in the breeding programme: Iris danfordiae collected from near Darbogazˇ in Turkey (left); an unidentified purple-flowered species collected from near Çat in Turkey (centre); and Iris sophenensis (right) obtained from a cultivated source their place is numerous rice-grain- near Darbogazˇ I found what turned Iris histrioides and, to a degree, its size bulbs. When bought from a bulb out to be a fertile, diploid (2n=18) hybrids, also produce bulblets, but supplier they usually have sufficient form of the species. It also produces they are less numerous and tend to energy to regenerate two bloom-size numerous bulblets but has been be slightly larger. bulbs for the following year. But useful for breeding. Hybrids from Two other taxa have been conditions in most gardens are not this diploid I. danfordiae can be important in my breeding work. good enough to produce further vigorous enough to rebloom One is an unidentified purple species bloom-size bulbs. More every year. I collected near Çat in Turkey. The gardenworthy, yellow-flowered Diploid I. danfordiae and its other is I. sophenensis (syn. I. histrioides Reticulata irises are needed. hybrids tend to bloom in the first var. sophenensis), a blue-flowered This poor performance is a half to two-thirds of the flowering species given to me by Frank Kalich characteristic of the sterile, triploid season. Partly as a result of this, the in 1987. (3n=27), commercial form of flowers are typically up well before I. danfordiae. It was presumably the leaves, which shows them off Breeding basics initially selected because of its better. Iris reticulata hybrids bloom When you cross two species the larger flowers than diploid forms. in the final two-thirds of the season first generation of hybrids (F1) tends However, in 1985 I was fortunate to and their leaves tend to be at the to be fairly uniform in appearance. collect I. danfordiae in Turkey, and same level or higher than the flowers. Breeders will then cross these hybrids with each other, or back Iris ‘Eye Catcher’ is recently released onto their parents, and it is in and becoming available commercially the second generation (F2) and subsequent ones that recessive characteristics can express themselves. When I. danfordiae is crossed with I. sophenensis the F1 hybrids are blue, because the blue of the latter parent is dominant over the yellow of the former parent. Whites with blue accents start to appear in the F2 progeny. This is the result of the blue and yellow genes being turned All photographs by Alan McMurtrie by All photographs off, yielding white and revealing an underlying pattern of blue ribbing on the style arms and blue dotted or veined markings on the fall blade. Sometimes there is a yellow infusion around the fall ridge. Occasionally ➤ 21 Now that I have new seedlings blooming every year, one priority is replanting the promising ones, so bulbs of the most interesting can be sent to the Netherlands for trial as soon as possible. At the same time I need to maintain enough bloom-size bulbs in Canada for use in hybridizing. An even slower process is the building-up of stock by the grower to It took ten years for the breeding the point where sales can begin. It programme to yield good whites takes at least 12 years to get tens of such as Iris ‘Starlight’ thousands of bulbs, and after that it might need to increase to hundreds the accents are green, which is a Breeding limitations of thousands as, hopefully, demand combination of blue anthocyanins Working with Reticulata iris is a picks up. For this to work optimally and yellow carotenes. very slow process as each generation you need good bulblet production in Iris danfordiae has a short bristle is five years. For crosses that should the early years. With typical hybrids as a standard, in contrast to other work, hybridization success rates the overall increase each year is species where it is 7–10mm in width. vary year by year from 25 to 65%. about 2.1 to 2.4 times. In the case In F1 hybrids with I. sophenensis this Germination rates are around 30 to of diploid 2n=18 hybrids, which results in standards that are 1mm or 33%, with 5 to 8% losses during that produce many small bulblets, that less in width but of normal length. period, giving a net of about 25%. I rate is more than 3 times. In subsequent generations they are am now applying gibberellic acid to For common cultivars propagated somewhere in between, in both the seeds at various stages to try to in large quantities, growers get 3–4 width and length. improve this. Euro cents per bulb and their costs are perhaps 2.5 cents per bulb. They Iris ‘White Caucasus’ was an early face a significant risk if bulbs are selection from crosses using triploid unsold, as they have already spent parents available commercially money planting the bulbs out prior to orders coming in. Wholesalers and exporters wait as long as possible before placing orders with, so they are not left with unsold bulbs. For 200,000 bulbs a grower’s costs are around ¤3,000, so they need to sell 100,000 bulbs just to break even. Flower size Bulb growers usually say that wholesalers and exporters want cultivars with large flowers – 70mm from fall tip to fall tip. Although I. sophenensis has flowers of 70mm, the other parents in the 2n=18 group are smaller. For instance, the species from Çat is 38mm, and diploid I. danfordiae is 33mm compared to 45mm for the triploid form. I may be happy with the cultivars I have raised, but my primary goal is for March 2016 Iris ‘Spot On’ (above left) and Iris ‘Scent Sational’ (above right) were introduced in 2014. A hybrid derived from ‘White Caucasus’, 03-AN-3 (below), produces up to five flowers per bulb ,so could be particularly useful in further breeding. others to grow and enjoy them. This My more successful breeding is is one of the reasons why I am having with 2n=18 parents. It took 10 years a lab create tetraploid versions of to get my first white-flowered F2 some of my hybrids. cultivar, ‘Starlight’, and several more One of my cultivars, ‘Eye Catcher’, years to see further F2 hybrids. is a good size (55mm tip to tip) and When the F1 first bloomed from should do well commercially. diploid I. danfordiae crossed with ‘Orange Glow’ is attractive but at I. sophenensis the next question was 40mm it is seen as a bit small. A what to cross them with. The pollen tetraploid variant, which bloomed looked good under a microscope, so for the first time in 2015, is 50mm, crossing them with each other was so should sell better. I therefore most likely to work, which is what have to take a gamble on this as it I did. This is in contrast to existing takes 10 years to build stock to the cultivars, such as ‘Harmony’, which point where sales can begin, and seemingly produce good fluffy pollen, then two more years to gauge public but under a microscope you can see reaction.With these long timelines the pollen grains are misshapen. you need multiple strategies. Three of the F1 plants showed a bit of yellow influence, which made Early cultivars their blue look a touch dark. I Crossing existing commercial thought the F2 plants might turn out triploid cultivars essentially gives muddy and would need cleaning up more of the same. In the early days in subsequent generations.