Saturday, December 21, 2013

Breeding For Rust Resistance in Daylilies: Part 6

Breeding For Rust Resistance in Daylilies: Part 6

This post is arriving a little faster than the others in the series have because it is a continuation of the last post, so I wanted to get it up while the last post was still fresh in our minds and before we move into the holiday week. Let's just dive right in...

Practical Techniques For Resistance Breeding: Part 2

To refresh our memories before we proceed, the five systems of mating as outlined by Sewell Wright are listed below. For a more definition of each, please refer to the previous post in this series.

1)   Random Mating {RM}
2)   Genetic Assortative Mating {GAM}
3)   Phenotypic Assortative Mating {PAM}
4)   Genetic Disassortative Mating {GDM}
5)   Phenotypic Disassortative Mating {PDM

The first of these five systems, Random Mating (RM), can be of use in breeding for rust resistance if the entire population shows some level of resistance on the higher end of a bell curve (i.e., moderate to high resistance or immunity). In this system, allowing open pollination (OP) is one way to go, or one could simply mix any number of pollens from various donors (or all of the plants in the population) to apply to as many of the plants in the population as desired. Either way allows for random mating. Yet, with this system, while resistant individuals will likely be created and perhaps even plants desirable on other levels, there is much less control over the direction that the population moves in. With that said, even this system, if only highly resistant individuals are selected each generation, can generate increases in resistance and the creation of highly resistant plants.

Of these five systems of mating, the most effective for breeding for rust resistance is 2 and 3 – Genetic Assortative Mating (GAM) and Phenotypic Assortative Mating (PAM). In both of these methods, two plants showing rust resistance are being mated. In GAM, two related, rust resistant plants are mated, while in PAM two unrelated rust resistant plants are mated. The first then is inbreeding using ‘like x like’, while the second is outcrossing using ‘like x like’. Both can be very effective for concentrating rust resistance. As it is highly likely that we have multiple genes for rust resistance, #3 (PAM) has the added possibility of combining multiple genes for resistance, while #2 (GAM) will concentrate the genes already present in that family line. We can see then that both of these mating systems are very important for breeding rust resistance, and both then are foundational to that work. However, the 4th and 5th systems also have uses.

In regards to breeding for rust resistance, the 4th (GDM) and 5th (PDM) systems would generally involve crossing a highly resistant or immune individual to a moderate or low resistance individual though not exclusively. The 5th (PDM) could involve a wide cross for phenotypic reasons (flower, plant habit, foliage type, etc) that use two rust resistant plants or the 4th (GDM) could involve a cross of two resistant plants that have been shown to have very different genetic types of resistance. Both GDM and PDM can be used to blend multiple types of resistance to create a broader-based resistance genotype/phenotype. Where GDM focus more on different genetic types of rust resistance (genotype), PDM focuses more on different phenotypic expressions of rust resistance (i.e., a seemingly immune plant crossed to a plant showing consistent partial resistance or where each parent shows resistance to a different strain of rust). When either GDM or PDM are used to cross a resistant x non-resistant plant, this is generally what I refer to as a ‘salvage project’.

Now, let us look at these points in more practical terms. In resistance breeding, the most common and effective strategies involve breeding two resistant individuals. This is the fastest path to getting more resistant offspring. Of course, when dealing with more than one gene for resistance, we are not always assured of resistant offspring in the first generation (F1), especially if we mate two resistant plants carrying recessive resistance genes and those genes are not compatible (i.e., at different alleles - this however in practice seems to be very rare). However, when that F1 were intermated to create the second generation (F2), not only would each type of resistance reoccur in some of the offspring, but a small number should express both types of resistance in one plant, but I don’t want to get sidetracked because in most instances, two resistant plants throw some resistant offspring. 

As mentioned above, if two related resistant plants are crossed (GAM), then you are concentrating the genes for resistance within that family. You may get a few offspring with greater resistance than either parent in this way, many with the same levels of resistance as the two parents and some with less resistance than either parent. This is a very effective tool when a family line showing high resistance has been identified, as it will allow you to concentrate the best they have to offer, perhaps creating individuals with much greater resistance than the line began with. This is also why it is so important to identify specific resistant plants as well as family lines that show clearly heritable resistance – these plants need to be used both for outcrossing and then for subsequent inbreeding (GAM) to concentrate and explore their resistance factors.

If two unrelated resistant plants are crossed (PAM), you may be dealing with the same type of gene(s) or you may be combining different types of genes. If one of those genes is dominant from either parent, then you can expect (in theory) as many as 50% (somewhat) resistant offspring from such a mating if the parent with the dominant gene is heterozygous for the gene and as many as (in theory) 100% showing some resistance (not immunity!!) if the parent with the dominant gene is homozygous for the gene, though in practice we rarely see such high numbers, or at least do not recognize them as the resistance levels of heterozygotes can be variable. If both plants are carrying multiple genes, then the outcomes can be very variable (this might be termed ‘quantitative expression’), but as long as some of the recessives are compatible genes, you should see some resistant offspring. It is likely you will see a few offspring as or more resistant than the parents, but not always (as mentioned above, you could even get none, though this would be unusual). However, in general practice, this method seems to work well to produce resistant offspring and often to increase the resistance of a few of them as well. Further, this method can be used to combine multiple genes for rust resistance in an effort to both increase resistance and create lines with a broader base of resistance that may offer resistance to more than one strain of rust. This will be an important challenge in time.

When we want to do a ‘salvage project’ (a specific example of GDM or PDM) we will generally be looking to cross a more resistant plant to a less resistant plant, and often a very susceptible plant, at that.  In doing this type of work, the best results are had when a highly resistant (or immune) plant is one of the parents, especially if the second parent is very susceptible. While even then we might not expect the offspring to be highly resistant, we may see some increase in resistance in the offspring, but even if we don’t we still have added some recessive genes for resistance to the offspring, so we should select through them, noting their resistance levels and where possible, choosing those with the highest levels and then mating those offspring to another resistant plant (usually the best route to increase the resistance in the next generation by brining out some homozygosity for the hidden recessives) or mating them together and watching for the most resistant offspring (which again, would indicate individuals with a concentration of the desired genes).

A salvage project may not be an overnight success. They take patience and one may have to go through two or more generations to start seeing the flower phenotype we want to save from the ‘salvage plant’ recombined with the resistance factors from the resistant plant. If we are careful and patient, taking our time to research resistant plants, and choose something that is resistant that is a good match for the ‘salvage phenotype’, then we can more quickly get back to that phenotype. Say, for instance, that we have a patterned seedlings or cultivar that shows exceptional pattern, but is very rusty. We want to utilize the genes for great pattern, but increase rust resistance. So we research know resistant cultivars to find something that is both highly resistant and shows pattern and we use that in our cross to the salvage plant. In this way we are both introducing rust resistance and making every effort to retain and increase the pattern genetics. Can you see then how this is really not that different from what you already do? Let’s change the word ‘rust’ to ‘bud count’ and see if the example is still so scary.

You have an excellent patterned cultivar with poor bud count. You want to increase the bud count while retaining the fine pattern. You research cultivars with high bud count and locate one that shows high bud count and is also patterned. You cross the two and work with the progeny to retain the fine patterning of the low bud count individual while recombining that trait with the high bud count of the outcross plant. This is the exact same thing I am describing, and you are probably all already doing projects just like that, so don’t get freaked out by this being about rust. Can you see how rust resistance can become just another trait you select for, just like any other? You already do salvage projects, but you do them for things like bud count, branching, flower opening, spotting, etc. Any flaw that keeps you from introducing a plant with otherwise good traits, but the plant has other traits that are so good you keep it as a breeder and breed from it, striving to increase its good traits and ‘iron out’ the flaw, is the exact equivalent to a wonderful plant with rust susceptibility.

The most important thing to remember about any salvage project, whether it is for rust susceptibility or low bud count, is that you may need to raise a large number of seedlings in order to get away from the flaw and replace it with its desirable opposite, and you may not be able to do it in one generation. So any salvage project may represent an investment of more time and space than other projects. In regards to rust resistance, you may get few resistant offspring and take longer to get resistant offspring than if you don’t pursue a salvage project at all. Always weigh the challenges of a salvage project to see if it is really worth pursuing. It may or may not be and that will entirely depend on you.

I now want to touch on a few final points in breeding schemes. There are three very basic mating strategies that can be very helpful. This first is to simply cross two individuals (called a hybrid single cross). We then call these plant ‘A’ (the pod parent) and plant ‘B’ (the pollen parent). Their offspring are ‘AB”. The ‘AB’ offspring can then be selfed, mated to other ‘AB’ siblings, backcrossed to either parent or outcrossed to an alternate plant/line.

The second is when plant ‘AB’ is outcrossed to a third plant (plant ‘C’), we get a three-way cross (AB X C or C x AB). The registry is full of plants reflecting this type of breeding and it is very common in daylily breeding. It also may utilize PAM where plant ‘C’ is unrelated to “AB’ (outcross) or GAM where plant ‘C’ is related to plant ‘AB’ (inbreeding). Any of these techniques are excellent for enhancing and concentrating rust resistance and are also useful in salvage projects, as well.

The third is when two the offspring of two hybrid single crosses are mated and is called a double cross. In this instance, you might notate it as Plant ‘AB’ and plant “CD”. Either plant can be pod or pollen parent, provided they are fertile that direction. This type of mating combines a greater diversity of genes, especially if none of the main plants (plant A, B, C or D) are closely related. If the phenotypes that we are seeking to combine are not genetically compatible (the same genes), we may not see much expression of the target trait in the F1 from crossing ‘AB’ and ‘CD’, but then that F1 can be selfed or intermated to bring out the target traits. The double mating scheme is a great way to combine multiple genes for rust into a later generation, bringing up to four different packages, eventually, into one group of offspring, with some few perhaps expressing homozygosity for all the genes involved. Where all four initial plants have the same genes for resistance, or are equally expressing those genes (i.e., all highly resistant), this method (as with the other two) can be a way to combine flower traits while also combining rust resistance. So you see, there are ways to select for rust resistance and the all-important ‘face’ at the same time!

Another point I want to discuss in regards to breeding strategies is recurrent selection, which is where one identifies the most resistant individuals and uses them to produce the next generation. This is the essence of resistance breeding. This strategy is used to increase and concentrate the targeted trait(s) within a population. The ‘keepers’ are selected repeatedly based on the highest levels of resistance and bred together to increase the amount of resistance. You do the same thing when selecting for any other trait, such as teeth or patterns or edges, thinness or roundness of petals, etc, when you keep the biggest teeth, edges or most strongly patterned, the thinnest or roundest petaled and keep breeding those together every generation. We know from those examples what we can achieve. We can increase resistance to diseases, rust and others, in just exactly the same way we have with other phenotype traits. It is most strongly through recurrent selection that we can do this.

Finally, one last technique I want to mention is the use of a recurrent parent. The recurrent parent is a more focused or narrow form of recurrent selection and is basically a backcross method of using recurrent selection. I have referred to this in the past as making good use of an exceptional individual and have made great strides with this technique in resistance breeding in poultry. The exceptional individual is very rare. They encompass many good traits and score highly for all those traits. In addition, they also have high breeding value, which means they can pass their good traits on in reasonable to high numbers to their progeny. Not all exceptional seeming individuals will have high breeding value, so when you find that combined with many good traits, you have found a real jewel worth its weight in gold!

In using the exceptional individual (recurrent parent) with high breeding value in recurrent selection, we cross it to another plant (called the donor parent) to make F1 offspring and then we backcross those offspring to the exceptional individual, to make a BC1. If this inbreeding does not reveal hidden deleterious traits and we see no indications of inbreeding depression, we may then continue to take the most resistant offspring from each BC over the recurrent parent for several generations. In doing this, we must watch carefully for signs of deleterious traits appearing or for inbreeding depression, but should none appear, backcrossing to the exceptional individual can concentrate and strengthen the exceptional traits and breeding value of the line, creating a line that is highly homozygous for good traits and that can then be used in many other directions with confidence of imparting a highly concentrated package of genes to those offspring. For instance, highly resistant plants from such a backcrossing scheme, as well as the recurrent parent, would make excellent subjects to use in salvage projects for crossing to the ‘salvage plant’.

In closing this section, I would mention that these are just a few of the possible breeding schemes that can be pursued. For a greater discussion of schemes, see the bibliography I will present at the end of this series. I do want to stress that there is not just one strain of rust, so in time, some of the more complex schemes and techniques I have described here will be necessary to begin combining lines of resistance to create plants with multiple forms of genetic resistance, offering a broad based resistance to multiple strains of rust. However, until we know more about the various strains of rust and can gather data on which cultivars show resistance to which strain of rust, we should approach this breeding in simpler ways, dealing with the rust that we encounter in our own gardens and working from there, allowing that a cultivar that was resistant one or more years, but suddenly shows significantly lower resistant in a subsequent year may have encountered a new strain of rust for the first time and to which is has lesser resistance. This will happen and we shouldn’t freak out when it does. It is highly likely that the plant still has resistance to one or more strains of rust and didn’t ‘fail’. The answer would then be to outcross that cultivar to another cultivar that did show resistance to the strain of rust encountered in that year, and in that way begin to blend the two packages of resistance genes.

It is important to realize at the start that there will never be a time when rust resistance breeding it ‘done’. That is highly unrealistic. What we should be striving for instead is to remain calm and remember that our goal is to stay one step ahead, not to somehow defeat the evolution of this pathogen forever, which is something we likely can’t do. Our goal is to improve the situation and reduce the number of highly rusty plants that enter into registration and the gardening world.

In the next installment, we will look at evaluating and rating rust resistance levels as well as various rating systems that are commonly in use.