Evaluating Garden Irises

The cornerstone of any breeding project is selection. From hundreds or thousands of seedlings, breeders select a few for their potential as garden plants. In essence, the hybridizer becomes a garden judge, evaluating each seedling against standards and expectations for the class of iris to which it belongs.

I have been an AIS accredited garden judge at two different times in my life, and have pretty much internalized the evaluation criteria given in the AIS Handbook for Judges and Show Officials. It behooves every iris hybridizer to be familiar with this book, even if only to understand how his or her irises are likely to be viewed by other iris hobbyists. When I first studied garden judging as a teenager, I accepted most of its precepts as "givens". Later on, I took to questioning what I was reading. What interested me in particular was the fact that some evaluation criteria were much more objective than others.

One example that has been remarked on by several writers is that of haft markings in tall bearded irises. In the first half of the twentieth century, many breeders were intent on creating self-colored irises with as much color purity as possible: the even, unmarked blue, yellow, or white. Most irises tend to show veining in the throat of the flower, on either side of the beard, and at the base of the standards. These haft markings were anathema to breeders pursuing the goal of pure self-colored blooms. For many years, judges were advised that an iris with haft marks was inferior to one that had none.

But why should this be so? It makes no difference to the health of the plant. One might argue that clear colors make a better display in the garden, but then one has to come to terms with the popularity of patterned irises: plicatas, blends, arilbreds, medians with spot or ray patterns, etc. There is actually no objective reason to deem a clean-hafted flower superior to a haft-marked one. It's in the eye of the beholder. Today, long after the elimination of haft marking was achieved in all color classes, the issue has faded. Many new irises have markings in patterns undreamed of 50 years ago, and are judged on the overall effect of the coloration.

Many judging standards, I believe, are like this. They promote the preferences of individuals of influence, often breeders, who have a particular vision of perfection in each class of iris. In this article, I am going to step through the various qualities that are used in evaluating irises, with a particular focus on teasing out objective criteria from criteria that reflect individual preference.

Vigor, Health, and Floriferousness

For me, these are some of the most important and most objective criteria for evaluating an iris as a garden plant. It would be difficult indeed to argue that a plant which grows and flowers well is not superior to one that is weak or reluctant to bloom. Although one might argue that it is possible to get too much of a good thing - say, an iris that increases so rapidly that it requires annual division to maintain its health - in reality, this is hardly ever an issue.

Vigor and floriferousness can be quantified. A simple way is to look at a 2-year-old clump and count the number of blooming fans and the number of nonblooming increases. If you do this with each clump you grow, you can readily determine what is an average number of blooms and an average number of increases, then select seedlings that are above average in either or both of these qualities. I think this is a better approach that using an arbitrary number or increase/stalk ratio, because it is adaptable to different kinds of irises, and accommodates the fact that standards may change over time.

(By the way, I have settled on the two-year-old clump as the standard for making evaluations. One-year-old plants, whether seedlings or new acquisitions, are not fully established. Clumps three years old and older vary too much in whether they are suffering from crowding. A two-year plant, established but not yet crowded, provides a consistent point of comparison between different plants.)

Evaluating the overall health of a plant is a little trickier; subjective impressions often come into play. A plant that succumbs to a disease or pest may just be unlucky one year. Planted elsewhere under different conditions it may be fine. Still, once one has some confidence that a plant is particularly healthy or particularly susceptible to problems, this forms an important part of the overall evaluation.

A somewhat more subjective issue, but one that is nevertheless important for most gardeners, is the attractiveness of the foliage. We look at the blooms for a week or two, but we have to look at the foliage year round. As a gardener, I like foliage that is green and healthy-looking through summer and fall. The natural growth cycle of irises can work against this, as leaf growth slows during summer dormancy and old leaves dry off faster than new ones replace them. This is especially true of arilbreds. I'm of somewhat divided mind about this. On the one hand, I like a crisp, green garden. On the other hand, I'm reluctant to fault a plant for doing what it does naturally, with no detriment to its health or vitality.

Stalk, Bud Count, and Proportion

Judging standards emphasize a particular ideal for the bloomstalk: blooms held above the foliage, branching that does not crowd the flowers but produces a pleasing composition, large number of buds. There is objective rationale for most of these criteria. We presume that an iris with more buds will bloom over a longer period of time, enhancing its garden vale. We admire the shape of the iris bloom, so we want to see blooms held in such a way that the shape is not compromised. Beyond that, though, we can quickly get into the area of personal preference. Iris pallida carries most of its blooms near the top of the stalk, Iris aphylla branches from the bottom. Neither matches the image of the idea "show stalk" which TB judges have so internalized. And yet, are these two species inherently less attractive because of their branching? For myself, I would say not. I regard this as just part of the interesting variation in the genus. If the "TB show stalk" look were inherently superior, then we would want to see it in all kinds of irises: Siberians, spurias, Louisianas, oncocyclus, reticulatas... This, of course, would be ridiculous. Another example arises in the MTB class, where a particular proportion of bloom and stem size is held as an ideal. If this proportion were objectively superior, then wouldn't we want every IB and BB to show it also?

In the end, I tend to think of "proportion" as a subjective judgment which takes into account the various proportions of the different species. An oncocyclus has different proportions than an MTB, and that is fine. Even within a single class, I tend to step back and simply ask whether the stalk is visually appealing, without insisting that every stalk be appealing in the same way.

Although additional buds and the extended season of bloom they provide are important, one must also consider the quality of those final blooms on the stalk. If they are so small and tatty that you're almost tempted to remove the stalk to make the garden look cleaner, then I don't think much is gained from that second crop of buds.


This is an area where personal preferences have become ingrained and often masquerade as objective criteria. The trend for over a century has been toward bearded irises with wider petals (particularly the falls). Some justify this by saying wider petals present a larger surface of color. True, but then why would anyone grow spider-formed daylilies? I'm personally fascinated by some of the irises with long, narrow petals: the Regelias and the amazing oncocyclus Iris acutiloba lineolata. The favoring of closed standards, like the matter of haft markings, comes from early hybridizing frustrations. Some older bearded irises were weak substanced, resulting in standards that would flop or droop erratically, spoiling the symmetry of the bloom. I see nothing wrong with open standards as a variation if they are symmetrical and firmly held. They can be especially nice if the interior of the flower is interesting, with contrasting style arms or bearded standards as in the Regelias.

So my approach to form is parallel to my thinking about stalk proportion. I rely on my personal artistic sense, and welcome variety.

Although substance is a generally desirable quality (we all prefer flowers that don't turn to mush in a little wind or rain, and that keep their shape), I don't think one needs to increase the thickness and stiffness of petals without limit. There is something to be said for an iris that retains the ability to flutter a bit and that does not appear to be carved out of wax or plastic.


I regard color preference as an almost entirely subjective matter. Truly, some of us like bright clean colors, others are attracted by subtle blends. Depending on your point of view, the bright colors might be labeled "garish" and the blended ones "murky". Some people love broken-color irises, others dislike them strongly. To each his own. Even the notion that an iris should hold its color until it fades is open to debate. I've grown a type of viola whose selling point was that the flowers passed from deep violet to lavender to near white as they faded, offering three colors on one plant!

With color, as with form, I think my role as an evaluator is first to simply observe accurately, and second to indulge my personal preferences. I see no reason in adopting a pretense of knowing what shape or color will make a flower attractive to others.

Putting It All Together

My evaluation method uses three quantitative objective measures: vigor, floriferousness, and fertility (the latter being important because I'm trying to expand the fertile families and want my seedlings to be useful in that context). In addition, I rate two more-or-less subjective qualities: garden value and distinctiveness. The first is an overall judgment of how much pleasure I derive from having this particular plant in my garden, the latter is an assessment of whether the iris is something new, or just another version of varieties already in commerce. Rating each of these five qualities on a 5-point scale is simple and pragmatic, and more comfortable for me than the point scales used by the AIS. The other advantage of this approach is that it adapts readily to evaluating wide crosses between species, which may not conform to the pre-established ideals for any of the currently recognized classes.

I'm grateful for my judges' training, as it allows me to approach "garden value" with an educated eye and a thoroughness I might not otherwise have. But in the final analysis, I think "garden value" is more that the sum of its parts, and that a given plant can give great pleasure - or little pleasure - for reasons that have little to do with a priori standards.




 Tom Waters

February 2011

More Articles

Return to Telperion Oasis

Unless otherwise noted, all text and illustrations copyright Tom Waters and all photographs copyright Tom or Karen Waters. Please do not reproduce without permission.