Wednesday, October 22, 2014

Some Thoughts on My Program...

Some Thoughts on My Program and Objectives

Breeding With an Eye to Many Traits

As this year is drawing to a close, I have been reviewing my program and my objectives. I find I am very pleased with where my program is heading. I see a lot of potential even though a great many of my seedlings are just bridge plants to be used in further breeding work. However, amongst those seedlings, I see many of my goals beginning to come together and I am encouraged that the images I see in my mind's eye can become a reality.

It is becoming apparent to me that because I often write about breeding for disease resistance with a focus on daylily rust, there are a good number of people out there who seem to associate my program strictly with rust resistance breeding, but that is far from the case. While resistance to rust is important to me, I simply view that resistance as one trait in a larger set of plant traits I consider important and am selecting for. However, just like every other daylily breeder on earth, my most exciting end-goals revolve around the flowers. Let's face it, we aren't growing daylilies as strictly foliage plants (even though we are stuck with just their foliage for most of the year - a thing to consider).

With that said though, there are many plant traits that I find equally important to the flower, and a beautiful flower on a terrible plant is a very disappoint thing for me. So my goals are beautiful flowers on attractive, hardy, well-performing plants. I see that as a complete package. Now I will stress that I am realistic. I don't believe there is such a thing as a completely perfect plant that has every positive trait you could ever want and will be that perfect in every environment and garden it is grown in. Sure, the attempt to produce such plants is a worthy goal and the Lenington award recognizes the best of such plants, but I doubt there is any one cultivar that is equally perfect in every situation. If nothing else, varying tastes in flower types will potentially discredit some of the best for some people.

I have a broad taste in flower types. I love spiders on the one hand, and ruffled full-forms on the other, unusual forms on the one hand and sculptural types on the other. Teeth are amazing and patterns are stunning. Eyes and edges entice and yet a good, bright yellow or near white self is a delight in the garden with carrying power and marvelous grouping effect in the landscape. I see a use and need for most of the phenotypes that exist in the daylily world and I am working with a wide range of phenotypes. I am not in any particular camp as to the 'right' flower type. I can't choose, I love most of them and I see different uses for them all, and for that matter, I like both diploids and tetraploids.

However, from growing daylilies for nearly forty years, I do have a long list of things I don't care for. Plants that don't increase, poor bud counts, low scape to fan ratios, weak scapes that lean or fall completely, ugly foliage, flowers that don't open, flowers on scapes so short that they are down in the foliage, muddy 'not-really-any-color' colors, canoed petals, spotty petals, high disease susceptibility, and infertility, are examples of traits I greatly dislike. I find all of these things disappointing, so in my own breeding program, I make every effort to cull out these undesirable traits when I find them in my seedling beds or the hybridizing garden.

Yet, there may be instances where any given seedling or cultivar is so good in one or two traits,  that I can overlook its other lesser traits, and I think we all have such instances. Such plants are regularly used in breeding programs, but it is nice if we can make the attempt to combine the good traits we so value with other good traits that such a plant may lack in an effort to make superior plants with many desirable traits. That is called breeding. So the goal of breeding is generally to produce superior plants with more and more advanced flowers, but ideals cannot always be realized in reality. That doesn't mean we abandon the attempt, though.

What it does mean is that for anyone who is being realistic and hopes to ever introduce a cultivar in their lifetime, there are always compromises, as the notion of the perfect plant may be more of an ideal than a reality. While it may be an ideal to always be reaching for, compromise may mean that you introduce something that is marginally less than perfect, or even if it is perfect to your eye or in your garden, you recognize that it may not be so to other people's eye or in their gardens. These things have to be accepted when we are dealing with an organism that is as environmentally influenced as the dayliliy.

Very few of us have the resources to breed plants as the professional plant breeders do or to utilize all of their techniques. We may incorporate what we can from their techniques and make our best effort, but we also have to recognize as hobbyist breeders that we all have some limitations. It is important to assess our abilities honestly, so as not to build unrealistic expectations that may well lead to disappointment - either for ourselves or others. We are probably all a bit unrealistic when we are getting started, though time and experience usually help to adjust that earlier idealism.

As an example, the beginner may say they only want to produce plants with six or more branches, or thirty or more buds, but then you get an amazing flower with four branches and fifteen buds, and you just can't cull it, so there is a compromise. We might make excuses, but the best approach is just to be honest - this flower was too pretty to cull. Now if you want to stick to your guns somewhat you might make that plant a bridge plant, breeding it to find an offspring just as lovely but with better branching and bud count. However, some may choose that plant as an introduction while others might use it is a bridge plant only and neither are really wrong. It is there choice and the marketplace will determine whether the breeder was right or not.

Some traits may have more importance, but I wanted to start with a flower example, because that is where most of us leave our guidelines or rules behind. However, it is good if we can be less easily seduced by pretty faces. When a plant has a serious flaw (and the above example is not really a serious flaw) it is probably best if we can look past the pretty face to the flaw and either discard the plant or only use it as a bridge plant, carefully testing its offspring to find those with its good traits and without the major flaw. I suspect if this had been done more, there would be less disappointment regarding some cultivars. One thing I am finding though is that there are many pretty faces produced in every seedling crop, so personally, I don't find it impossible to cull a pretty flower if there are other serious flaws.

My philosophy is that I let the plants guide me. I set out a group of criteria as a guideline in my breeding and I eliminate those that don't fit the criteria, but I do try to make that set of criteria relatively realistic and I do not expect every plant to match all the criteria points. Some of the selection points are more important to me than others. However, beyond that static set of criteria, I observe what is actually occurring and let that be my main guide. Whatever is working is a gift and should be exploited, whether it fits our narrow expectations or not.

This may be a good point to discuss my actual goals with rust resistance. Classical disease resistance selection would focus solely on producing the most resistant plants, but I look at my program as being focused more on actively removing the most susceptible plants from the gene pool. I certainly note the recurrently most resistant cultivars and seedlings, and I note this every year I have rust, using those that show the most resistance each year as major breeders, but my goal is to produce reasonable garden resistance rather than the most extreme immunity that is 'forever and always'.

I consider my goals and methods to be more realistic because I can't use all aspects of a professional-level selection system as would be used in commercial operations and rust is constantly mutating so that a highly resistant/seemingly immune plant in my garden may not be equally resistant in another garden with another strain of rust, but for that matter a fast-increaser in my garden might be a slow-increaser in another garden or a plant with strong scapes in my garden may fall in another garden or on an immature division. I just find it to be important to be realistic as I pursue my goals.

All I can ever say is that a given cultivar I might introduce shows a given level of resistance in my garden for 'x' number of seasons. I can report the feedback I receive from others, and all of this is relevant information, but it doesn't mean that any given cultivar in my garden will perform the same in another garden, so it seems to me that the pursuit of rust resistance, and nothing else, is not a good use of my time and effort. I will say that I suspect there will be some cultivars that will show high levels of resistance in the face of many strains of rust and may exhibit that resistance for many years, but none are likely to be resistant in the face of every possible strain of rust or for all time. That is reality and to ignore it is to set myself and others up for disappointment. Again, I choose to be realistic.

Now, with that said, I still think it is important for me to select against rust susceptibility, as high disease susceptibility is a trait I find very undesirable, just as I find falling scapes, muddy colors, or flowers that won't open to be highly undesirable traits. So all these traits fall into the realm of 'things I cull against'. By removing such traits, I am working to move my gene pool in the opposite direction - toward disease resistance, strong scapes, clear colors, and flowers that consistently open well, etc. I see 'rust resistance' as no more or less important than any other trait. It is just another trait within the overall selection criteria I have set for myself and it is not, by a long shot, my only focus.

With the publication of the paper Identification of Pathotypes in the Daylily Rust Pathogen Puccinia hemerocallidis by Dr. Buck of the University of Georgia, we now have strong evidence that there are multiple strains of rust in the US. While this is no surprise to me, the publication of this paper gives us evidence of this obvious point - rust continues to mutate. The most encouraging point of this paper though is that there were a handful of cultivars tested in this study that showed good to high resistance to the isolates (potential genetical strains) of rust used in the test project. This is very encouraging and indicates that cultivars of broad-based genetic resistance do exist and can likely be bred. (I will likely offer more detailed look at this research paper in a later blog post.)

I believe that we in hobbyist breeding settings can make an impact through identifying cultivars with broad-based resistance and breeding value and any interested party should use whatever testing methodologies they feel comfortable with to work toward identifying such plants to breed from. As introductions are made from such plants I do feel it is important to report on the resistance that has been observed, without making that a major marketing strategy.

It may seem counterintuitive, but I don't really see rust as a 'premium' trait, in and of itself. While we may in time realize that some cultivars and the genes they carry offer broad-based, genetic resistance for a long period of time, there is probably no real way to know that or make such a claim as only time will tell. It is only in hindsight that we will know this to have been the case. Working with those that have been consistently resistant or seemingly-immune for a long period of time and in many locations does give us the best bet for long-term and broad-based genetic resistance to many strains of rust though. 

I see rust resistance as an issue of cosmetics. I feel this way because rust is generally not deadly (except for some instances in the deep south) and it only creates an unsightly mess. While this may be uncomfortable to look at and completely unacceptable for many gardeners, it is not a plague. So I tend to view rust susceptibility as merely another undesirable trait to select against, much like any other undesirable trait we would select against, such as spotting of the blooms or falling scapes (both cosmetic/aesthetic considerations).

In breeding for any trait, I find it very important to test potential breeders, noting their combining ability and breeding value, so as to locate those breeders with the best chances of passing on desirable traits and breaks in phenotype, while not passing on undesirable traits. This would apply equally to all traits. I don't see one trait as paramount and of more importance than any other trait. I am seeking balance and a balanced approach. Imbalanced approaches tend to allow a lot of small problems to accumulate over time through being ignored, while all focus is going to one or two traits. In time those many little problems can add up to major problems. That which is ignored does not go away.

So in my own program I am not seeking to go to any particular extremes with one single trait, but to create balanced plants with beautiful flowers. So while I do watch rust resistance and go out of my way to select against rust susceptibility that is not, nor can it be, my only focus. As an example, I am just as likely to cull a highly resistant plant with horrible flowers or some other serious flaw as I am to cull a high susceptible plant with beautiful flowers. Selection is a sword that cuts both ways. Or, I could say that I don't suffer serious flaws gladly, whatever those flaws might be.

In my program, I produce thousands of seed every year. In the first years, I produced around 20-30,000. Last year I produced 70,000+/- and this year closer to 100,000+/-. Yes...that is ridiculous. However, there is method to the madness, as I am selecting for many traits and only keep a few seedlings from any batch for any length of time. So it actually makes sense for me to have a long list of criteria for culling. The more traits I am selecting for, the more plants I can potentially cull and that allows me to get down to manageable numbers more quickly and potential be producing really superior plants in the process. 

I actually don't like to breed from huge numbers of individual plants, preferring to focus my efforts on a few plants that are well tested, but to get there, many plants have to be tested and evaluated. In time, of course, I will produce fewer seeds from fewer but more well understood cultivars of high breeding value. Until then though, I actually need a lot of things to cull for, just to reduce the numbers of seedlings I have, and of course, this also allows there to be a few seedlings with a high concentration of many desirable traits and/or real breaks in the flower on superior plants. That is the greater goal anyhow. So it makes sense for me to use rust resistance/susceptibility as a culling and selection tool. I can always stand to cull out a few more seedlings!

Culling starts early, continues as long as the plant is here and is not just focused on the flowers, as I cull lots of seedlings for various reasons before I ever see the first flower. However, I do not turn over my plants every year or two as is common in many operations. I tend to get first flower in two years, with some taking to the third year and only a few in the first year. (Those which take more than three years to produce the first flower are simply culled as well.) However, I find the flowers from the first year of flowering to be a highly unreliable guide to cull by. I have seen too many examples already of a flower that is gorgeous in the first year if flower only to look horrid the next to believe that selection based solely on the flowers of the first year of flowering is a wise idea. What I look for is a plant that has lovely flowers year after year. To do that, I need to evaluate the plant for several years.

Five years is my average evaluation time, both for a new cultivar I am considering for a major position in the breeding program and for my seedlings. My oldest group of seedlings, mainly from seeds I purchased in 2010 will be five years old next fall. I have been culling on those seedlings now for four full years. They have been through rust culling and foliage culling and flower culling. Of the 7,000 seedlings I started with, there are now about 100 plants left. Of those, there are four or five that I am considering for registration in 2015 and introduction in 2016 and maybe ten or twenty more that are likely to be retained as long-term breeding stock. I can't say I am completely sure yet on any of them though.

The first round of seedlings I produced here were from 2011 breeding and are now three years old (they were planted the summer of the year they were produced and were small seedlings before winter). They have been through several rounds of culling, but flower culling only began on these in earnest this year and will predominantly be done next year. Once they reach their fourth and fifth years, I will have a much clearer idea of their destinies, and of course there will be far fewer of them by that time as well. I am not sure any of these will make the cut for introduction, as most were created specifically to be bridge plants, but time will tell. I have now bred from a few of them in the 2013 and 2014 seasons and am beginning to gather preliminary data on the breeding value of some of the more promising. This is part of the five year evaluation and is another point of consideration - breeding value. This will determine who continues on in my breeding program and who gets composted.

While breeders further south are able to produce an introducible number of mature fans very quickly, and norther breeders can often do the same with greenhouses, I don't feel anyone can really understand their plants without a few years of observation, especially those of us growing in more temperate climates and strictly growing outside. Yet even if I were in the position to introduce plants more quickly, I think I would want to take my time and be as sure as possible about the consistency of any given plant and its flowers before I made an introduction. I think this would help me to avoid unpleasant surprises on down the line, though no one can foresee all possibilities or make perfect decisions all the time. We will all make mistakes and fail at times. These are gifts that we can learn from and no daylily is going to be perfect for all time and in every location, no matter what efforts we make toward that end. This may be one of the hardest lessons of all to learn.

In closing, I hope my thoughts have offered you insights into my process and perhaps also given you food for thought that may help your own programs. In some ways, many of my posts are more missives to myself, sort of a way to refocus or sort out my thoughts, than they are directions for others. They are sort of 'me talking to me' while allowing you to see the process. I do hope it can be of some help and guidance to us all :-)