In true Jared fashion, what began as a quick email turned into an article. Before he departed from his 17-year chapter at Organic Seed Alliance, Jared Zystro shared his top tips on breeding for disease resistance on-farm. A caveat from Jared: much of this he learned years ago, so his memory may be imperfect and the science may have moved since.
*This content has been edited by OSA’s communications team and does not reflect Jared’s email verbatim.
If you’re breeding for disease resistance on your own farm, one of the first forks in the road is whether to chase single-gene resistance or build up multi-gene resistance. The two behave differently in the field, take different amounts of time, and fail in different ways. Here’s how Jared thinks about the choice.
What’s the difference?
Many single-gene resistances, also called vertical or qualitative resistance, rely on the plant’s hypersensitive response (HR). Essentially, the gene lets plant cells recognize a specific configuration on the outside of the pathogen, a lock-and-key mechanism. When the pathogen is recognized, those cells self-destruct, creating little necrotic lesions around where the pathogen landed and stopping the spread.
Because there’s a lot of selection pressure for the pathogens to avoid triggering that response, they evolve to change their outer shape so that the cell no longer recognizes them, breaking that resistance. Breeders then need to find a new gene that helps the plant identify a different part of the pathogen, and the “race race” continues.
Multi-gene resistance, also called horizontal or quantitative, works differently. It can be physical (think thicker wax on the outside of the plant) or it can be a stack of weaker biochemical defenses adding up. Either way, pathogens can’t evolve around it as easily, because they can’t just change their shape a little the way they can to dodge single-gene resistance.
Some single genes are durable
This is something Jared didn’t learn until he took a class on breeding for disease resistance: not all single-gene resistance is fragile. Durability largely comes down to whether the part of the pathogen the plant is recognizing is so essential to how the pathogen functions (some key piece of its metabolism, say) that it can’t really be changed.
The classic example is the mlo gene in barley for powdery mildew, which has held up in European spring barley since the 1970s. It works by disrupting a host gene the pathogen actually needs to establish infection, so there’s no easy shape for the pathogen to change. (Worth noting: mlo is recessive, a bit of an exception to the “single-gene resistance is usually dominant” point below.)
Some pathogens also simply don’t evolve as fast, often because they may not have a sexual phase in their life cycle. That said, sexual reproduction isn’t the only thing driving pathogen evolution rate. Population size, generation time, and migration matter a lot too.
And that can change, as we can now see with late blight. For roughly 150 years after the Irish famine, only one mating type existed outside of Mexico, so it could only reproduce clonally. Once the second mating type spread out in the 1980s, sexual reproduction became possible and a wave of new, more aggressive strains followed. Climate change and human-mediated movement of life between continents make it a lot easier for different types of these pathogens to “meet” each other and perhaps be able to reproduce.
What the seed catalog is telling you
Single-gene resistance is usually much more effective than multi-gene resistance, in the sense that a plant carrying it will show much less effect from being exposed. With multi-gene resistance, you may still see some damage.
That’s the distinction behind the codes in catalogs like Johnny’s: HR = High Resistance, IR = Intermediate Resistance. HR is usually single-gene; IR is usually multi-gene, though not always. Sometimes IR reflects a single gene that only gives partial resistance.
Two other practical notes:
- Markers exist for many single genes. Genetic markers have been identified for a lot of them, letting breeders do genetic testing pretty easily and cheaply to see whether their plants carry those resistance genes.
- Single-gene resistance is often dominant. A plant carrying one copy from either parent will show resistance. That’s helpful for people breeding hybrids, who only need to bring resistance in from one parent, but it makes selecting harder, because you don’t know whether the resistant plant you’re looking at carries two copies or one.
Why strong genes hide subtle ones
Here’s the catch that shapes everything below. Because single-gene resistance (if it hasn’t been broken by pathogen evolution) is so effective, its presence makes it nearly impossible to select for multi-gene resistance in a variety or breeding population. The powerful single genes basically mask your ability to see the more subtle differences between plants with varying degrees of multi-gene resistance. This was a big point of Raoul Robinson’s Return to Resistance.
Three questions before you choose
Especially if you’re doing this on-farm with limited resources:
- Are there even known sources of single-gene resistance for the diseases you care about? If not, you’re going the multi-gene route. If there are sources, how hard would they be to get into your variety? Are they only found in very unadapted wild species, where a bunch of breeding work will be needed to select out the unwanted traits you might bring along?
- How important is it that your plants are “highly resistant”? Would some amount of visible disease mean the plants are no longer marketable? Are there other easy ways to control the pathogen through cultural practices, such that the combination of intermediate resistance and those practices keeps disease effects below an economic threshold?
A real-world example: Jared worked with a big spinach grower in Salinas. Downy mildew in spinach is a big problem, and the “race race” has been running with it for a long time, to the point where there are many single-gene resistance genes being juggled and many races that have evolved to get around them. This grower wanted to grow his own spinach seed to supply his baby spinach business, and he knew he couldn’t effectively juggle all those resistance genes in an open-pollinated variety. But he had worked out cultural practices, including his own garlic extract spray, plus a rapid production, harvest, and supply line, so that if the multi-gene, intermediate resistance was somewhat effective, he could still produce his baby spinach and meet market standards even with a little infection occurring. - Are there known, reliable markers for the genes you care about, and affordable testing? If you’re doing this on-farm rather than as part of a university or a big company, you need a reasonably priced service you can ship samples to. Jared isn’t here to be a salesman, but he has used Ag-Biotech in Monterey for this a lot in the past; they list the markers they can screen for on their website. Some people have raised concerns about the reliability of their marker testing, though marker testing mistakes are always something to plan for. More on that below.
If you go the single-gene route
Make sure the resistance is coming in from one of your parents. If it’s coming from a wild or less-adapted source, know that you’ll be bringing in other traits you’ll likely need to select out, and that some of those traits may be genetically tied to the resistance genes, at least until enough generations of mixing and crossing over have occurred.
Go slow with marker testing. If you’re using a testing service, it’s very easy to get excited and screen your population very hard. Why not send in hundreds of samples if you can get them tested for under a dollar each, especially if you can test for multiple diseases at once?
Here’s why not. Say you have a tomato breeding project and you care about resistance to late blight, bacterial spot, and fusarium, each with several strains and several resistance genes found for each. You also care about marketable yield, shape, and flavor, plus other things you really need to see to be able to select on. You cross your favorite delicious, beautiful, high-yielding variety to one that isn’t as good in productivity and quality but carries all those resistance genes. If you go hog wild in the second generation after the cross (the F2) and send in leaf samples from hundreds of plants at the seedling stage, then transplant only the ones carrying all those resistances, you’ve maybe shot yourself in the foot, because when you look at things in the field, you don’t have enough plants that are productive and look and taste good. This is even more of an issue with traits you don’t see every year, like drought tolerance, where you need the right environment.
The moral of the story: go slow with marker testing, let your population cross with itself to shuffle genes, and make some selections on the harder stuff you need to be in the field to see before going through a big population bottlenecking process with marker-assisted selection.
Verify your marker results in the field. Marker-assisted selection is not 100% reliable. It’s easy to be fooled, because you get back a spreadsheet saying whether each plant carried each gene as a heterozygote (a single copy from one parent), a homozygote (a copy from both), or not at all. But labs make mistakes; the markers they’re looking for are often adjacent to the resistance gene rather than in the gene itself, so sometimes the marker gets “detached” from the gene; and generally, life is fuzzy and doesn’t perfectly match the simplified models we carry in our heads. Grow those plants out and see whether they actually show resistance.
Remember you need the disease to be present. With either kind of resistance, the pathogen needs to be in your field at levels that affect the plant in order to select for resistance. That might not happen every year, or in every field you grow in, so keep it in mind when planning. You may want to team up with a fellow farmer in another location that often gets the disease, to help cover your bases in the years when the pathogen doesn’t hit your farm hard enough to select.
If you go the multi-gene route
Start with the best parents and the most diversity you can get. Because you’re trying to stack lots and lots of smaller-effect genes together, the key principles of quantitative breeding apply: choose parents with as high a level of resistance as you can identify, while ensuring you have large amounts of genetic diversity to select from. If you don’t know how diverse your parents are because you don’t know their genetic backgrounds, you can base it on differences in visible traits and on what you do know, choosing parents that came from different plant breeders, different breeding companies, or different university breeding programs. You’ll probably also want to start with more than just two parents.
Select lightly at first. Let your population do some initial crossing and genetic mixing so you don’t lose too much genetics. Early on, the genes are still “stuck” together, traveling with the neighboring genes they started with in the parent. Each generation, chromosomes cross over during meiosis, and those linkages get broken, which gives you access to more combinations later.
Watch for single genes in your population. If they’re present, they can overpower and mask the multi-gene resistances you’re trying to see.
Field design: making small effects visible
Multi-gene resistance is typically more subtle in its effects, and there will be a gradient of resistance depending on how many small-effect genes are in a given plant. That means working hard on experimental designs that let you see and distinguish those small effects.
Deliberately increase disease pressure. Without going overboard and growing your plants in ways so different that it no longer looks like your normal growing conditions, think about your site and your cultural practices in reverse. Is there a low field with no good airflow where you’d normally avoid planting a susceptible crop? Plant your breeding trial there. Do you normally prune and tie your tomatoes in the afternoons when the dew is off them? Do the opposite. Go in there and mess with them in the mornings and after rains. If there are sprays or other ways you manage these diseases, skip them in the breeding field. There is a limit here: if you go too far, all your plants die, and you can’t select. You’ll have to feel that out.
Consider inoculation. A lot of university and private company breeding programs intentionally inoculate their plants with the disease of interest, and it can yield nice, strong, uniform disease pressure. Jared hasn’t done that in his own breeding work, because it often involves some care and feeding of disease organisms to make the inoculum, as well as a lot of manual infecting of each plant. Sometimes there are also phytopathology rules that limit which diseases you can do that with and when in the season you’re allowed to do it. But there may be a time and place where it’s the right tool for the job.
Design around uneven pressure. Pathogen pressure is generally not uniform across a field, and that makes it tricky to tell whether you have plants that are actually more resistant or whether they just escaped the pathogen because of their location. The solutions follow the usual trial design principles: select a field that’s as uniform as possible based on what you know (same soil across it, equal access to water, equal light and shade), and keep an equal approach and timing with any of your cultural practices. Many wind-borne pathogens follow patterns where they come to your field first from one side, hitting the edges before they get to the middle. This is where borders of non-breeding population plants can help, as can breeding and trial schemes that let you replicate your breeding families in multiple places in the field. Some breeders will also “seed” the disease throughout the field by including strips or blocks of a known, very susceptible variety.
Plan for the long haul. It will take many generations to stack up all these small-effect resistance genes, especially if you don’t get the disease every year. Think about a plan you can reasonably sustain for all those years.
Combining both approaches
These two approaches don’t have to happen in isolation. You may have some diseases you’re breeding for using single-gene resistance and others you’re taking the multi-gene route on. You can also take both approaches to the same disease in the same population. You just need some tricks.
- Do the multi-gene work first. For that, you want parents that don’t carry any known single-gene resistances. If you’re not sure, you can use marker screening to do the opposite of what you usually would: look for the varieties or plants that don’t carry any of the known genes.
- Go through the steps and years of selection for multi-gene resistance as described above.
- Bring in the single gene once you’ve plateaued. That’s the point where you’ve reached what feels like the limit of what you can accomplish resistance-wise with multi-gene breeding.
- Cross carefully to preserve diversity. Bring in the single-gene resistance by crossing your population with a variety with known resistance. Because your population is likely diverse and your source of multi-gene resistance is probably pretty complex, make many, many crosses between your population and the single-gene resistant “donor” parent. Advance all those crosses to the F2 or F3 and use marker screening to find the plants that are homozygous for the resistance genes.
You’ll now have the strong single-gene resistance, backstopped by the more durable multi-gene resistance. Of course, more breeding will still be needed to select back all the good traits your population had and that your disease-resistant donor may have been lacking.
One long-term caution: over time, if the single-gene resistance doesn’t get out-evolved by the pathogen and break down, you should expect your population is probably losing some of that background of multi-gene resistance, just because it’s being masked by the stronger single-gene resistance and not being selected on anymore. If that’s a worry for you, you can keep maintaining your earlier population that only had the multi-gene resistance, with an eye toward re-implementing the single-gene crossing plan again if you need to.
