Using Species in MDB Breeding, Part 1: Iris lutescens

by Tom Waters

I’m a hobby hybridizer, working on a very small scale (I can only raise about 200 seedlings each year). So from the outset, I planned to focus on niche projects, rather than trying to compete with the large-scale hybridizers and their well-established breeding lines. I identified some projects that I thought might have value, but that few others were working on. One of these is to develop a line of true-breeding miniature dwarf bearded (MDB) irises.

As I’ve written previously, most MDBs produced today are small selections from standard dwarf bearded (SDB) breeding. Because the lines of the top SDB hybridizers are so advanced, this approach indeed produces many fine MDBs that display all the variety of color and refinement of form found in the modern SDBs. With a few exceptions, however, hybridizers do not have dedicated lines aimed specifically at producing MDBs. Rather, they select them from among their SDB seedlings that happen to fall below the height limit separating the two classes. Since these plants are genetically no different from SDBs, they are likely to produce SDB-sized seedlings, and some are prone to growing out of class, showing foliage and stalks out of proportion to their size, or having flowers that are too large and coarse.

In earlier times, most MDBs were produced by crossing SDBs with the species Iris pumila, which indeed produced smaller, daintier plants than those from pure SDB breeding. Unfortunately, these plants are unbalanced tetraploids with limited fertility, making them dead-ends.

If one is seeking a true-breeding line of MDBs, there are several options. One is to start with the SDBs and keep selecting for smaller size. Another option is that promoted by Ben Hager: cross tetraploid MTBs with I. pumila, then cross those seedlings with small SDBs or MDBs from SDB breeding. The idea here is that the MTB X pumila crosses will produce genetically smaller plants, and the genes for small size can be stabilized in the breeding line.

A third approach, which I have been exploring, is to expand the gene pool by using species (or combinations of species) that are small in size but belong to the same fertile family as the SDBs. This not only has the potential of creating a line of true-breeding MDBs, but also increases genetic diversity, which may give more variety of forms and colors. This is the first of a series of posts on this project, focused specifically on my hybridizing with I. lutescens and its close relations.

I. lutescens is a dwarf species found in the Mediterranean lands of southwestern Europe, particularly Spain, France, and Italy. There are many related but different populations in this range, which are sometimes given species status, but more often treated as synonyms of I. lutescens. I. subbiflora in Portugal and I. bicapitata in Italy, are almost always treated as separate species, but are part of the same continuum of types found in I. lutescens. Before the 1950s, most dwarf irises found in gardens in western Europe and North American were forms or hybrids of I. lutescens. After the first SDBs were produced, hybridizers quickly abandoned the older I. lutescens dwarfs in favor of the SDBs, which showed wider color possibilities and improved form.

Given this history, I did not at first think of using I. lutescens in my own hybridizing projects. On reflection, however, it seemed that those hybridizers of the 1950s, ‘60s, and ‘70s were excited about exploring the potential of the new SDBs, and not really concerned with creating true-breeding MDB lines. I. lutescens might indeed still have something to offer toward that goal. In nature, its forms span the MDB and SDB height ranges, and flowers are smaller than those of typical SDBs.

S004-01,
I. subbiflora ex Spain
(SIGNA seed)

 

S019-01,
I. lutescens (SRGC seed)

Early on, I grew a plant of I. subbiflora obtained from the SIGNA seed exchange, collected in Spain. Apparently this raises some doubts about its identification, as there is some question whether there are populations of I. subbiflora in Spain, or only I. lutescens. My plant is not as tall as most descriptions and photographs of I. subbiflora, normally being about nine inches in my garden.

Crossing the subbiflora with SDB ‘Kaching’ (Black, 2009) produced a lot of deep red seedlings, mostly of small SDB size, although height varied from year to year and with location in the garden. Although I enjoyed these seedlings, they were larger than what I was aiming for; so I made an effort to use MDBs, rather than SDBs, in future crosses, and to seek out small plants of I. lutescens to use instead of this I. subbiflora. A cross of I. subbiflora with MDB ‘Circa’ (Johnson, 2015) produced attractive plants near the MDB height limit, but still not as small and delicate as I would like.

S027-07,
'Kaching' X S004-01

 

S044-04,
'Circa' X S004-01

I raised a small purple I. lutescens from seed from the Scottish Rock Garden Club (SRGC) seed exchange. It’s about half the size of the I. subbiflora plant. Crossing this with MDB ‘Miniseries’ (Keppel, 2011) gave me a couple promising seedlings: one with small reddish flowers have rather tall stems; the other has mid-sized purplish flowers on shorter stems. I’d be happier if I’d gotten the small reddish flowers on the small stems! But these are interesting, and I will continue working with them.

S039-02,
S019-01 X 'Miniseries'

 

S039-03,
S019-01 X 'Miniseries'

I also crossed the SRGC I. lutescens with MDB ‘Pearly Whites’ (Black, 2014) and MDB ‘Beetlejuice’ (Black, 2013); and a number of these seedlings bloomed this spring. They were all interesting to look at, mostly nicer in form than expected. It’s too early to be sure what height they will settle out at, but this year most of them looked more like small SDBs than like MDBs.

S055-05,
'Beetlejuice' X S019-01

  

S055-11,
'Beetlejuice' X S019-01

S056-01,
'Pearly Whites' X S019-01

I acquired some other small lutescens plants to work with: one from the Berkeley Botanical Garden, and one from Sean Zera, raised from SIGNA seed. I have not yet bloomed seedlings from these. I keep acquiring more I. lutescens seeds, hoping to expand my collection, particularly in colors other than purple. I. lutescens grows here, but doesn’t seem really happy. My high desert zone 5/6 garden in New Mexico is a long way from the south of France.

I am enjoying working with species in MDB breeding, but it must be emphasized that this is a long-term project. Any use of species in breeding must be followed up with several generations of crossing back to modern hybrids if one wishes to meet current expectations of width, ruffling, and substance. It appears that some persistence will be needed to combine the genes for short stems, small flowers, and narrow foliage in the same line. Still, I’m finding these explorations to be a very satisfying use of my small space and limited resources.

Bearded Iris Species for the Garden

 by Tom Waters

Most of us are quite familiar with the gorgeous hybrid bearded irises grown in so many gardens around the world. Whether dwarf, median, or tall bearded, these irises are the great show-stoppers of the springtime iris parade. Every year, hybridizers introduce new varieties, so there is a never-ending stream of new colors, patterns, and flower forms.

But before hybridizers took an interest in irises - and that was scarcely more than 100 years ago - bearded irises in gardens were much like irises in nature. Yes, bearded irises are wildflowers in Europe and western Asia, where numerous different species grow on hillsides or in valleys without gardeners to care for them.

These wild bearded irises are of interest to some iris growers, for a variety of reasons. Some of us are just attracted to wild flowers, and like to see the plants as nature made them. Others are students of iris history, and it is these bearded species that are ancestral to our modern hybrids; growing them is like tracing the roots of our family tree. Others are hybridizers, curious to explore the untapped potential of these species.

Here is just a smattering of bearded iris species that may be of interest, with an emphasis on those that have been mentioned frequently in books and articles or that have featured prominently in the development of our modern bearded iris hybrids.

Iris pallida,
a form with variegated foliage

Iris pallida is a European tall bearded (TB) species, with light blue-violet flowers and straight stems with branching clustered near the top. To my eye, it is the most stately of all the bearded iris species.

I. variegata, another European species, is smaller, similar to  miniature tall bearded (MTB) irises in size, with zig-zag stems and distinctively colored flowers with yellow standards and falls marked in wine-red or brick-red.

These two species are probably the sole ancestors of the diploid tall bearded irises grown in gardens in the 1800s and early 1900s. They hybridize in nature, and these natural hybrids (amoenas, bitones, and blends) were collected by flower lovers and grown in European gardens. Any modern TB iris that has yellow (or pink or orange) coloration owes its color to I. variegata.

Iris variegata

The other wild irises important in the development of the modern TBs were a range of different tetraploid plants from various locations in the eastern Mediterranean. These include I. cypriana, I. trojana, I. mesopotamica, and some named clones such as 'Amas' and 'Ricardi'. All are purple bitones and large, well-branched plants. The botanical status of these is not as clear as might be, but every modern TB descends from them in some fashion. It was hybridizers of the early 20th century who crossed the colorful diploid pallida/variegata hybrids with these large tetraploids to create the modern tetraploid TBs that have become the most popular and widely grown of all irises.

Iris aphylla

Moving down in height, we come next to I. aphylla, a species from eastern Europe that is usually similar to standard dwarf bearded (SDB) or intermediate bearded (IB) irises in height, but copiously branched, the lowest branch usually emerging right at ground level! The flowers are dark violet, and beards usually bluish. It has been used in two different ways in breeding modern bearded irises: first in breeding "black" TBs, and then later in the work of Ben Hager and others creating the tetraploid MTBs. See my earlier blog post Our Debt to Iris aphylla.

In the realm of dwarf species, the earliest known to western European gardeners was I. lutescens (once called I. chamaeiris), which is native to the warm Mediterranean areas of Spain, France, and Italy. It was the basis for our garden dwarf bearded irises until the 1950s. Its height varies, spanning the modern MDB and SDB classes and sometimes growing even a bit taller. Colors are yellows, violets, whites, and blended and bitoned versions of those colors. See my post Iris lutescens: The Dwarfs that Time Forgot.

Iris pumila seedlings

The most important dwarf species in the development of modern dwarf and median irises is the species I. pumila, native to cooler, higher regions of eastern Europe. It is quite tiny, usually only about 4 inches tall, with blooms that appear almost stemless and  spring straight from the rhizomes. It has an extensive color range, from white to blue, purple, rosy, and yellow, usually with a darker spot of violet, red, or brown. It came to the attention of iris breeders in western Europe and the US rather late, in the mid-20th century. Once "discovered," however, it became a power player in iris breeding. I. pumila, when crossed with TBs, produced the modern SDB class, and through them the modern MDBs and IBs as well. See Iris pumila: A Tiny Treasure.

Iris reichenbachii grows wild on the Balkan peninsula, where it exists in both diploid and tetraploid forms. The flowers are rather long in form, and most are either a dull yellow or a grayish violet. This species is notable in the history of iris breeding for introducing the "dominant amoena" gene into the TB gene pool through the work of Paul Cook. Most modern bicolors have I. reichenbachii back somewhere in their family tree. See The Untapped Potential of Iris reichenbachii.

Iris reichenbachii

Also worthy of note is I. reichenachii's diploid relation, I. suaveolens (once called I. mellita). It is smaller than I. reichenbachii, but otherwise similar. It is noted for its sickle-shaped leaves, in some forms edged in a narrow line of red-violet pigment.

I hope this line-up of bearded species has piqued your interest. They are nice additions to the garden for anyone who is curious about where our garden iris "came from" or is attracted to the simple beauty of wildflower irises. One sometimes finds a few bearded species available from commercial iris growers, but more are consistently available through various seed exchanges from societies like the Species Iris Group of North America, the Scottish Rock Garden Club, the British Iris Society, or the Dwarf Iris Society.

Why not "go wild" and try some bearded species this year!

Understanding the Historic Dwarf Bearded (DB) Class

By Tom Waters

Before 1958, the now familiar American Iris Society (AIS) horticultural classes for bearded irises did not exist. Certainly there were bearded irises of different sizes, but the precisely defined “dwarf,” “median,” and “tall” bearded classes were still gradually taking shape as a way to describing the diversity of the bearded irises in nature and in the garden. Today, we use “median” to refer to all the bearded irises except miniature dwarfs and tall bearded. 

Consequently, care must be taken in assessing how the earlier bearded irises were categorized. This post will look at just one such category, the dwarf bearded irises, abbreviated DB prior to 1958.

DB 'Artoviolacea' (Todaro, 1856)
probably a natural hybrid
of I. lutescens and I. pumila
Photo: El Hutchison

In the early 1900s, botanists were aware of a number of dwarf bearded species, notably Iris pumila, I. lutescens (then mostly known by the name I. chamaeiris), I. aphylla, I. reichenbachii, and others. Gardeners in western Europe and the United States were also familiar with garden cultivars of dwarf bearded irises. These were almost exclusively forms of I. lutescens or hybrids between I. lutescens and other dwarf species. I. lutescens is native to the Mediterranean regions of France, Italy, and Spain. Other dwarf Iris species are native to Eastern Europe, and thus I. lutescens was more accessible to early commercial nurseries in France, England, Germany, and the Netherlands.

There was little concern about establishing a precise definition for the DB category. Dwarf cultivars being grown were easy to distinguish from tall beardeds (TBs), and that was really all that mattered. Crossing dwarfs with talls produced intermediates,  another widely used term. The intermediate beardeds (IBs) were a pretty obvious group of hybrids and easily distinguished from both dwarfs and talls. Keep in mind that this was long before separate awards were established for different irises. Terms like dwarf, intermediate, and tall were helpful descriptions for irises rather than a specific horticultural class that needed to be assigned unambiguously.

DB 'Bride' (Caparne, 1901)
probably pure I. lutescens,
6 to 8 inches in the author's garden
photo: Tom Waters

The AIS checklists of the 1920s defined dwarf bearded irises by listing some familiar dwarf species, and noting that the term included hybrids between these species. The 1939 checklist added some precision by stating that dwarf irises were up to 17 inches in height. It was not clear whether ancestry from dwarf species or height was to be decisive; it was just taken for granted that hybrids among the dwarf species and cultivars would fall into the expected height range.

What were these early dwarfs like? Like the species I. lutescens that dominated their ancestry, the early dwarfs were mostly between 6 and 12 inches in height, almost never branched, bearing one or two terminal buds. Colors ranged from yellow, cream, and white, to blue, violet, and purple. Most had self or bitone color patterns. The spot pattern from I. pumila, so familiar to us in modern dwarfs, was seen only occasionally. All blooms were tailored, and the plants bloomed before the intermediates and talls. Although I. lutescens in the wild often shows interesting markings, patterns, and blended colors, these were not common among garden varieties. Perhaps this reflects the preference for clear, smooth colors that persisted among iris fanciers until the later decades of the 20th century.

DB 'Path of Gold' (Hodson, 1941),
a child of 'Bride' and
probably pure I. lutescens, is
10-12 inches in the author's garden.
Photo: TomWaters

Hybridizing with dwarf irises took a back seat to TB hybridizing until the 1940s and 1950s. William J. Caparne of England produced many dwarfs around 1900, as did the German firm of Goos and Koenemann. A number of American hybridizers subsequently worked with dwarfs, most notably Hans and Jacob Sass of Nebraska.

As hybridizing progressed, it became clear that complicated hybrids might not always fall into one of the intuitive categories being employed. In 1948, the AIS adopted a new set of classes defined by L. F. Randolph. Randolph rejected height as a decisive factor, and offered instead a rather vague notion that advanced generation hybrids that had "most of the characteristics of typical dwarf bearded irises" (he had in mind things like season of bloom, short foliage and stalk, lack of branching, etc.) would be considered DBs.

This way of thinking about the class was soon pushed to the breaking point. Hybridizers had begun importing and working with different species, outside the familiar I. lutescens and its hybrids. Robert Schreiner had imported seeds of the tiny cold-hardy species I. pumila from eastern Europe, and this exciting new species gradually spread among dwarf hybridizers. Paul Cook introduced the first hybrids between I. pumila and TBs in 1951: 'Baria', 'Fairy Flax', and 'Green Spot'. Although the term had not yet been invented, these were the prototype of what would become the modern classification of standard dwarfs. Did these new irises have "most of the characteristics" of dwarfs? At a mere 10 inches, they were less than 18 inches tall, to be sure, but did having a TB parent make them intermediates? They were sometimes branched, not a characteristic of "typical" DBs of the time.

DB 'Sapphire Night' (Nicholls, 1935),
perhaps a hybrid involving
 I. lutescens and I. aphylla,
12-14 inches in the author's garden.
Photo: Tom Waters

Cook registered his new hybrids as IBs, but to many iris fanciers that just seemed wrong. In 1954, AIS introduced a new height-based classification, where any bearded iris 15 inches tall or less was a DB, regardless of ancestry or other characteristics. This caused a schism with many dwarf enthusiasts.

Eventually the matter was settled by splitting the DB class in two: a miniature dwarf bearded (MDB) class with an upper height limit of 10 inches (reduced to 8 inches in 1976), and a standard dwarf bearded (SDB) class from 10 to 15 inches. The MDBs were the province of the Dwarf Iris Society, whereas the SDBs were regarded as medians and promoted by the Median Iris Society.

The effects of Cook's new hybrids were not limited to the classification system. Excited by the developments, dwarf hybridizers scrambled to use I. pumila and the new SDBs in their breeding. The 1950s was a transition decade: it began with nearly all DBs being cultivars or hybrids of I. lutescens, and ended with nearly all MDBs and SDBs being hybrids of I. pumila and TBs in varied proportions. The influx of TB genes brought plicatas, pinks, and ruffled blooms into the dwarf iris world; I. pumila brought dramatic and varied spot patterns.

What does this all mean for the collector of historic dwarf bearded irises? First, it is essential to understand that the DB class does not correspond to any single modern class. It spans the height range of both MDB and SDB classes, even extending into the IB class in some instances. In principle, every historic DB could be assigned to a modern class on the basis of height. However, height was not recorded in early registration data; so such determinations would require catalog descriptions or garden observations. The Median Iris Society attempted to make such reassignments for the IB, MTB, and BB classes, although I do not know of any similar undertaking having been made for the MDBs and SDBs.

Even if we were able to sort historic DBs into modern classes based on height, that would not capture the essential nature of the old category. It represents irises of different ancestry and genetic characteristics than the modern classes. The 8-inch height boundary is even less meaningful for the old DBs than it is for their modern successors. I. lutescens and its hybrids span this boundary, and there is seldom any meaningful difference between a 7-inch and a 10-inch DB.

I think the best advice for modern growers interested in collecting historic dwarfs is to approach them on their own terms: viewing the DB category in its own right with context that differs from modern hybrids. Starting from that perspective, one can then notice some similarities and comparisons that might be made: A DB that is a I. lutescens/pumila hybrid, for example, might resemble an MDB that is an SDB/I. pumila hybrid.

The historic DBs are harder to come by than modern cuiltivars, but it is worthwhile to acquire and grow a few. Not only will you be helping to preserve a window into iris history, you will also get to enjoy a type of iris that really has no modern counterpart.

Iris lutescens: The Dwarfs that Time Forgot

by Tom Waters

Dwarf bearded irises may be found growing wild throughout much of southwestern Europe, from Spain and Portugal, through southern France, and into northern and central Italy. Through the centuries, different botanists have encountered them in different localities and assigned different names to them: Iris chamaeiris, I. italica, I. olbiensis, I. lutescens, I. virescens, I. subbiflora, I. bicapitata.

Iris lutescens, raised from seed

By the twentieth century, it was clear that most, if not all, of these were really irises of the same species. Gardeners were most familiar with those from southern France, going by the name of I. chamaeiris, so began referring to the whole species as the “chamaeiris complex”. But the rules of botanical nomenclature require that synonyms be resolved by using the earliest published name for the species. In this case, that honor goes to I. lutescens, the name used by Lamarck in 1789. This is now the correct name for all these irises, with the exception of two irises at the extremities of the species’ range,  I. subbiflora in Portugal and I. bicapitata in the Gargano peninsula of eastern Italy, which are regarded by many (though not all) botanists as distinct species in their own right. Even if these are not regarded as belonging to I. lutescens, they are indisputably very close relatives.

I. lutescens is a delightfully varied species. The flowers are most often yellow, cream, or violet, but there are near-white forms, purples, blends, and bitones. In height, they range from about 6 to 12 inches (15 to 30 cm). The stem is unbranched, with one or two terminal flowers.

Iris lutescens 'Bride' (Caparne, 1901)

Until the second half of the twentieth century, I. lutescens  and a handful of its accidental hybrids with other species were the only dwarf bearded irises known to gardeners in western Europe and North America. Named cultivars were produced by the firm of Goos and Koenemann in Germany, by W. J. Caparne in England, and later by Hans and Jacob Sass, among others, in the US. At this time, the modern dwarf and median classes did not exist, so there was no distinction between miniature dwarfs and standard dwarfs; they were all simply “dwarf bearded”, and spanned the whole natural height range of the species, which straddles both of the modern categories.

Iris lutescens 'Path of Gold' (Hodson, 1941)

Although many people grew a few dwarfs, appreciating their charm and early bloom, almost all the attention of iris enthusiasts in the first half of the twentieth century was focused on the tall bearded. The dwarfs were rather taken for granted, by both gardeners and hybridizers. That began to change with the formation of the Dwarf Iris Society under the leadership of Walter Welch in the 1940s. Welch and his friends were determined to learn all they could to advance dwarf hybridizing, and their interest went beyond the I. lutescens cultivars to investigate other dwarf species, such as I. pumila from eastern Europe.

I. pumila is a diminutive species, about half the height of I. lutescens, single-flowered and almost stemless. Robert Schreiner had imported some seeds in the 1930s, and the species gradually became available to the new dwarf hybridizing enthusiasts. The turning point came in 1951, when Paul Cook in Indiana, who had exchanged his pumila pollen for TB pollen from his friend Geddes Douglas in Tennessee, introduced the first pumila/TB hybrids: ‘Baria’, ‘Green Spot’, and ‘Fairy Flax’. Although technically “intermediates” (as the word was used then, it meant a hybrid between dwarf and tall bearded irises), these new irises were no larger than many I. lutescens dwarfs, even though they often had a branch and a total of three buds! This launched a vigorous debate about classification, which led ultimately to the formation of the Median Iris Society and the four median classes we have today. The SDB class was created to accommodate the new pumila/TB hybrids and the taller I. lutescens cultivars, with the MDB class left for the “true dwarfs”, with a maximum height limit of 10 inches, later adjusted to 8 inches.

From the 1960s on, the SDBs from pumila/TB breeding totally dominated the world of dwarf irises. These SDBs carry an extraordinary genetic legacy (dramatic spot patterns from I. pumila, pinks and plicatas from TBs, not to mention more modern form). There was no interest any more in producing more of the overly familiar yellow or violet I. lutescens cultivars. Even the MDB class was taken over by the new SDBs. Most MDBs from the 1960s onward were produced by crossing the new SDBs back to I. pumila, or (especially in recent decades), just selecting irises from SDB breeding that happen to be under the height limit.

I. lutescens, once the very archetype of the dwarf bearded irises in gardens, is now a curiosity known only to species enthusiasts.

Is there any hope for a lutescens renaissance? At first blush, it would seem unlikely. The modern SDBs have been so developed by decades of dedicated hybridizing that they would seem to have nothing to gain (and much to lose, in terms of present-day expectations of the class) by the injection of I. lutescens into hybridizing lines.

If I. lutescens is to be heard from again in dwarf hybridizing, the opportunity may be in the MDB class. Some MDB enthusiasts have been grumbling of late that the class has been taken over by short SDBs, and is losing something of its distinctive charm. There may be some niche here for MDBs with more of a “wildflower” look, breaking away from the stiffness, width, and ruffling that comes from pure SDB breeding. Just as the MTB class has given a home to the simpler, more modest look of the diploids, perhaps there is an opening for more “retro” MDBs. I. lutescens is fully fertile with SDBs and their MDB progeny, and might add a breath of fresh air to a class that is starting to feel overworked.

Everything old is new again?

Iris lutescens campbelli

The World of Irises is the official blog of The American Iris Society. Now in its 99^th year, The American Iris Society exists to promote all types of irises. If you wish to comment on a post, you can do so at the end of the page and the author or the editors will reply. If you wish to learn more about The American Iris Society, follow the link.