Tuesday, November 18, 2014

A Review of Identification of Pathotypes in the Daylily Rust Pathogen with an Eye Toward Breeding for Resistance

A Review of Identification of Pathotypes in the Daylily Rust Pathogen with an Eye Toward Breeding for Disease Resistance

Brian Reeder


Overview of Paper

In the paper entitled Identification of Pathotypes in the Daylily Rust Pathogen Puccinia hemerocallidis - James W. Buck (Department of Plant Pathology, University of Georgia, Griffin, GA 30223, USA - Received: January 23, 2013; accepted: May 16, 2013.  - Journal of Phytopathology 161 (2013) 784–790 - 2013), the abstract at the beginning of the paper states... 

"Daylily rust, caused by Puccinia hemerocallidis, has been present in the United States since 2000. In 2003, inoculations with a single isolate of P. hemerocallidis identified daylily cultivars with high levels of resistance to the fungus. The present study was carried out to determine if pathotypes of P. hemerocallidis are present in the southeastern United States. Sixteen isolates of P. hemerocallidis were each inoculated onto leaf segments from 19 daylily cultivars and the resulting disease phenotype assessed. A significant effect of rust isolate on host reaction phenotype was observed for nine of the 19 daylily cultivars. Five of the nine cultivars displayed reaction phenotypes with different isolates of P. hemerocallidis that included at least one susceptible or moderately susceptible and also resistant phenotypes. These results indicate that different pathotypes of the fungus are present in the southeast United States. Daylily hybridizers interested in screening for host resistance to P. hemerocallidis will need to include multiple isolates of the fungus to allow for this host specialization."

We can see from this statement that we now have good evidence that there is more than one strain of rust in the US, but there also seems to be more than one gene for resistance to the potential strains of rust pathogen studied in the paper.

The paper gives an overview of the methods used. The research is conducted in greenhouses for quarantine of each isolate. Each isolate may or may not be a unique strain of rust. I have received no data showing any gene typing of each isolate. To my knowledge at this time, the isolates are determined by unique accession. That is, each isolate came from a different garden, so that there were 16 accessions from sixteen Georgia gardens in the study. To the best of my knowledge at this time, all evidence of strain differences in rust accessions is based on the reactions of the host materials. So at this time I would stress that the most we can know from the paper is that it seems highly that there are at least two different strains of rust amongst the 16 isolates, though there could be more than two.

The same can be said for any hypothesis of ‘resistance genes’ based on this paper - there seems to be the indication of variation in resistance amongst the nineteen cultivars used in the research project. Of those nineteen, nine show different responses to the 16 isolates and of those nine, six were observed to have a wider divergence in rust presentation with reactions that were both resistant and susceptible, depending upon the isolate. As the most common type of apparent resistance variation is expressed as one rating up or down (i.e., HR to R or R to MR or MR to MS), those cultivars that show a wider jump between rating variations (i.e., R to S) are notable and may indicate actual strain differences in the rust pathogen as well as variable resistance/susceptibility genes in the host (daylily plant). However, as breeders, we may be more interested in those cultivars that showed the highest resistance to every isolate, as these may represent individuals with more than one resistance gene or with a resistance gene offering protection against more than one rust strain. Both instances could be useful in breeding for resistance.

These quotes of interest caught my eye in the paper’s Discussion section. I will quote each here and offer a bit of commentary.

In conclusion, the present study clearly shows that pathotypes of P. hemerocallidis that differ in virulence to daylily cultivars are present in the south east United States.

This is accurate. We can now make a fairly good case for there being more than one strain of rust. Further work may reveal much more about this, if any of the research being done is published or made available to the hobby in some form.

The number of resistance genes to P. hemerocallidis present in Hemerocallis sp. is unknown. Diploid and tetraploid cultivars, with dormant to evergreen growth habits, were included in this study with no apparent differences in susceptibility observed. The chrysanthemum – Puccinia horiana pathosystem is thought to contain at least seven resistance genes (De Backer et al. 2011). Incorporation of disease resistance into new daylily cultivars would be a valuable tool for reducing the impact of daylily rust on the ornamental industry… Development of DNA markers that can detect resistance in diploid and tetraploid Hemerocallis germplasm could greatly facilitate current breeding efforts… Daylily hybridizers interested in producing new cultivars resistant to P. hemerocallidis will need to include multiple isolates of the rust fungus to allow for this host specialization.”

This well surmises where the hobbyist breeder with an interest in increasing rust resistance in their gene pool now finds themselves. We can be relatively certain that there is more than one strain of rust in the US. We know that there are resistance genes to rust pathogens in other species. We know that such genetic resistance can be vertical (single dominant major genes) or horizontal (multiple genes combined).  

We have seen fourteen years of data, both scientific research and anecdotal observations, which allows us to surmise that there are genes for resistance in the daylily gene pool and we can make use of any such genes in our own field selection programs for rust resistance. Identification of resistance/susceptibility genes with laboratory methods is not currently available to the hobbyist breeder. There is no way to know if such testing will ever be made available or if it will be kept as a proprietary intellectual property. In the mean time, the hobbyist simply needs to do what we have always done, and breed best to best using what data we can find, what research is made available and our own observations.
Prairie Blue Eyes

Discussion of Practical Applications

So how do we hobbyist breeders apply the information gained through Dr. Buck’s paper? What is actually applicable to us? For me, the most important take-away is something I had previously taken for granted anyhow - there is not just one ‘rust’ pathogen. We will generally note variations in response in any given cultivar based potentially upon both environmental factors and/or strain interactions. Of special interest are those cultivars that show high resistance in many locations and over many years. It is possible that, if multiple genes for resistance occur in the daylily (which seems likely), then there are individuals that, quite by chance, already have more than one gene for resistance combined into their genotype. Such individuals would be highly useful in any effort to breed for resistance. Even if a new, more virulent strain of pathogen arises that defeats the resistance of any of these long term resistant individuals, they are still useful breeders for the resistance they convey to other strains of the pathogen. In short, such cultivars may be useful for a gene-pyramiding project.
Joan Senior

Discussion of Most Resistant Cultivars

For brevity, in considering the cultivars in the study, I will focus on those six cultivars that showed the highest and most consistent resistance with no susceptibility ratings in Dr. Buck’s paper - Hush Little Baby, Joan Senior, Prairie Blue Eyes, Going Bananas, Stella De Oro and Mardi Gras Parade.

It is by chance that I grow and have grown all six of these cultivars for a long period of time, some as long as thirty + years. It is also by chance that I had identified each of these cultivars as showing strong resistance to rust in my own field observations and so have been using them in breeding for several years already. I am already into the second generation with some of these cultivar’s descendants and I am seeing, what suggests to me, heritability of that high resistance. I believe any of these cultivars can offer a base upon which to add more resistance genes through outcrossing to other cultivars that also show high resistance. It is a place to start.

Now I know the arguments. These are all plain, old, cheap cultivars, so why would anyone want to use them? Stella De Oro, after all, is universally disparaged and seen as a near blight by refined daylily personages. However, my own breeding tests show that ideas of “old” or “plain” are more in the mind of the person thinking the thought. By combining modern faces and advanced phenotypes onto these older cultivars, new advances can be seen that were not possible back in the day when the cultivars were being bred from. By dipping the best of now back into the best of before, advances are possible and this also allows us to keep identifying potentially resistant new daylilies to make intensified gene combinations by breeding them back onto older, more proven cultivars. Both advanced faces and plant traits can be selected for in one program, after all.
Stella De Oro

Uses in Breeding Programs and Methods

We would typically suggest that individuals showing broad resistance, much as the six I detailed above, might have ‘vertical’ resistance, while a plant showing less complete resistance may show ‘horizontal’ resistance, but things may not always be so cut and dry. There can be instances where vertical resistance may give resistance to only one specific strain, or there may be instances where a single gene infers resistance to multiple strains. Conversely, horizontal resistance, being quantitative, can fall anywhere on a spectrum, depending on the concentration of genes in any given individual. In instances where many resistance genes are combined into one individual, we may see resistance to multiple strains. In some instances, this can be perceived as ‘vertical’ resistance, especially where the resistance is very complete. 

Gene Pyramiding is a concept in genomics that involves stacking multiple genes for resistance into one genome. While pyramiding is in many ways similar to the concept of horizontal resistance, in that there are multiple genes concentrating greater resistance, the difference is that in pyramiding we are making a conscious effort to add as many resistance factors as we can, possibly also tolerance and slow progression of sporulation traits, into one individual to increase field performance against the rust pathogen(s).

Some might suggest that a path would be to inbreed each of these six cultivars from the research paper and to thus concentrate their own resistance genes into homozygous lines, and this would certainly be interesting from a research standpoint, but from a hobby standpoint, I think this is one of our least productive routes for many reasons.

Another potential route is to cross the six cultivars from the paper to create lines of concentrated rust resistance. While this is feasible, and certainly could be interesting, I think it has two major drawbacks. The first is that none of the six cultivars offer anything terribly modern from the point of view of the flower. The second is that we do not know how many resistance genes may or may not be contained within all six of those cultivars. So strictly relying on those six cultivars, and only those six cultivars, is much too limiting for most modern breeding programs.

I believe the most productive route is for each breeder who has the interest to use some of these six cultivars and to combine them with their own favorite cultivars, seedlings, and interests in form, color or pattern, especially those alternate cultivars that have also been shown to have or are known to have some resistance themselves. In other words, these six cultivars would possibly have applications both in pyramiding with other resistant lines as well as being useful in salvage projects to bring desired flower traits into more resistant genetic backgrounds.

First, let us look at using any of these cultivars in salvage projects. I believe that it is this type of breeding where the most interest is likely to be found in the daylily world, as a number of remarkable and popular cultivars show various levels of susceptibility. The desire to salvage the good traits into a more desirable background setting is standard fair in ornamental plant breeding and daylily breeders have been doing this type of thing as long as daylilies have been bred.

In short, a salvage project is a cross of highly resistant x less resistant/more susceptible. This cross is done to bring the expression of resistance together with desired flower traits in the F1 and/or later generations. Generally, a field test of F1 offspring will reveal some with higher resistance. Select toward resistance where possible. The F1 can then be interbred, backcrossed to the more resistant parent or outcrossed to another resistant cultivar or seedling. Selection is toward resistance in at least the second generation (F2, Bc1, etc.), if not the first (F1). From this technique, plants of usable resistance can be secured for further breeding work and in some instances, resistance may be increased even in the F1. Salvage projects may become a very important aspect of daylily breeding for spreading resistance genes further through the gene pool, but it is not the most important method for increasing resistance within individuals.

The most important form of breeding for increasing resistance genes within given individuals is to stack multiple genes for resistance together through the pyramiding of genes. We do this by breeding resistant individuals carrying different genes (or suspected different genes, in the hobby setting) together. We are seeking to combine the different genes from each parent into the offspring for broader-based, multigenic resistance to the given pathogen in future generations. We may expect to see some progress in the F1, selecting those that field trial with the most resistance or we may see little to no increase in the F1 and seek more expression in the F2 and later generations. Much will depend on the nature of the genes involved.

Regardless, we have many options once we have our F1. In my own situation, I prefer to field trial such seedlings for resistance levels. I then select for acceptable expressions of field resistance. I may repeat this for one or more years before I ever begin breeding. Once I begin breeding such F1 plants, I might go in any of many directions. I can backcross to either parent to concentrate that particular form of resistance, making a BC1 generation. This can often be a wise thing to do in both directions. Another thing I can do is to interbreed the F1 to make an F2. This is the best way to find the new recombination of both forms of rust resistance in one plant. Whether we need to breed to full homozygosity for any/all of the resistance genes will depend on their penetrance and their nature (dominant/recessive). Finally, the F1 can be outcrossed to other lines, cultivars or seedlings, in an effort to combine the resistance of both parents with another type of resistance. This later example tends to only work well when dealing with highly penetrant dominant genes, typically. 

As you can see, there are many directions and possibilities for our F1 seedlings. Our main aim is to combine the genes for resistance from both parents into the offspring. Once we have a plant or plants (whether as an F1 expression of dominant genes or in the F2) that show the targeted combination of traits we are looking for, we can then move forward. We may choose to work within the line to concentrate traits or we may seek to further build our pyramid by going out to a different lineage, also showing resistance, in the hopes of adding yet another genetically distinct set of genes for resistance.


Pyramiding of genes can go on indefinitely, so long as there are new genes to add. When an older, proven cultivar seems to ‘fail’, it may well be that it has encountered a new strain of rust that it does not have resistance for, so such cultivars should not be discarded, but used in pyramiding with new resistance genes to the new strain of rust to keep building a broad base of genetic resistance. The cultivar reacting to a new strain of rust that has been resistant for many years, through many tests and in many locations in the past has not suddenly ‘lost’ its resistance. It still has resistance to the rust pathogen strains it was previously tested against and those genes can and should still be put to use. 

Over time, we can assume that any and all genetic resistance will fail as new strains of the pathogen emerge. The speed at which this will occur is anyone’s guess, but with time, that failure is inevitable. Rather than see this as a problem, through pyramiding of resistance genes, we can continue to build on past successes and resistant bloodlines. In this way, we incorporate a broader range of resistance so that plants can safely overcome the challenges of many strains of rust. 

An excellent place to start will be to make some use of those cultivars that have been tested in research projects (such as that detailed in the Buck paper) and field trials, as well as those with a number of anecdotal reports and personal experiences. Some of the six cultivars with the highest scores in the Buck paper should find a place in any and all diploid breeding programs. As well, tetraploid conversions exist of several of these six cultivars. 

We will look more in-depth at some of these six cultivars in a coming blog post.
Hush Little Baby