Let's Talk Hybridizing: Incest Among Plants
Pat Halligan
Freeland, Washington

During this October's meeting of the Seattle Chapter Hybridizer's Group, we were discussing the merits of selfing. We've all heard the old saw: cross two plants to bring together some desirable traits, and then go on to the F2 generation by selfing the best progeny and, if you're lucky or raise enough babies, you'll get a plant with all the desired traits combined in one plant. Or maybe you've heard this one: here's a great plant, but I can make it even better if I self it and raise enough seedlings

Is this true? What happens in the real world?

I'll pose the question this way: how many selfs have you seen in the plant registries? We all know that lots of selfs are made, but why do so few of them ever get registered? Here's the answer. Ever heard of the Habsburg Jaw? Or hemophilia in royal families? Inbreeding causes two things, both by the same process: the expression of undesirable recessive traits, and the loss of genetic diversity. The process can be predicted mathematically and can be prevented if you follow some simple rules.

Maybe I should first talk about how our genes are arranged. The genes are arranged on two sets of chromosomes. Each gene can have several different variations called alleles, but each individual can only have two alleles, one for each set of chromosomes. For instance, eyes can be brown, blue, or hazel or, in rare cases, other uncommon eye colors. People with brown eyes have two "brown" alleles. Those with blue eyes have two "blue" alleles, and those with hazel eyes have one brown and one blue allele. Defective alleles could produce other variations in eye color.

So what does this mean in plants? To breeders, plants are black boxes. We have absolutely no idea what's inside a plant's genome, but we know that there are a lot of genes, and a lot of alleles. We also know that many important alleles produce effects that we cannot see. This means that we must breed not only for what we can see but also for all that hidden stuff that's so important for our plants. We have to devise a strategy that preserves the essential genes that we cannot see, as well as the genetic diversity that allows us to continue forward generation after generation.

Selfing is the worst form of incest. Each time you self a plant, you've just destroyed half of the genetic diversity within that plant. This is usually the kiss of death to a breeding program, but in some specialized breeding the loss of diversity is a desirable goal. Here I'm talking about the production of hybrid corn. In this specialized process, two vary inbred lines of corn are produced which are very uniform. These lines have very low vigor and produce very poor plants. But here's the trick: if you cross the two inbred lines of corn, you get "hybrid corn," which is not only uniform but also has superior vigor. How did that happen? By combining the two inbred lines, you restore the genetic diversity within the individual plants.

So what's the lesson? You'll almost never produce a superior plant by inbreeding, but inbred plants can be good parents, if bred to unrelated plants.

In my own breeding project, I started out with a limited number of species, and I have continued breeding now for five generations. Now that's a prescription for genetic disaster! With each generation I'll have less to work with as genes are lost from my population. This loss of genes from a population is called "genetic drift" and it occurs with a predictable mathematical precision. Soon, with this loss of genes I will have bred myself into a genetic box. My plants will become more and more uniform with each generation.

So how can I defeat genetic drift? First, you start with as wide a base as possible. Let's say you start with three species: R. racemosum , R. edgeworthii , and R. campylogynum . Not much to work with here, right? Wrong! Buy five different really nice forms of each species. Wait a minute! Don't I want to use only the very best forms? That's what the experts always say. Forget the experts; you need all those genes you never see in those other plants. So now you've increased your population from three plants to fifteen, and the amount of genetic diversity accordingly.

With each succeeding generation, instead of just selecting only the very best plants to breed, you select all the acceptable plants, including many that aren't really what you're looking for, but are healthy, vigorous plants. By doing this, you've accomplished two things. First, you've increased your breeding population several fold, and as a result reduced your genetic drift accordingly; and second, you've selected not only for the floral traits you were looking for, but also for those hidden alleles which confer vigor. Doing this will keep you out of that genetic box.

Now, I know that very few people pursue the type of breeding program that requires this type of care. Most people pick two commercial hybrids and cross them. The population they work with is outside their garden. So how does this all relate to the average breeder?

For an example, let's look at a "yak" cross. We believe that the FCC form ('Koichiro Wada') of Rhododendron yakushimanum is the best one, so let's cross it with 'Nancy Evans'. Gee, one of the plants is really nice with some yellow and some indumentum as well. So let's self it, and if we raise enough seedlings we'll get an even better plant. Right? Wrong!

The correct way to do this cross is to get two similar hybrids with totally different parentage. For example, one parent could be 'Koichiro Wada' X 'Nancy Evans' and the other parent could be Teddy Bear' X 'Lila Pedigo'. I'm using this cross only to illustrate how you can widen your base, and thus have more to work with. You can just about bet that you won't get what you want in the first generation anyway, so let's plan ahead and stay out of that genetic box.

If the whole population for your breeding program will be growing in your garden, then you better prepare to combat genetic drift. First, you go get several different forms of yak. But you don't stop there. If you want really good indumentum, you'll also need some other plants: maybe one of those new Taliensia species, or maybe even something really wild like R. mallotum . If you use hybrids, find out what parents were used, and get hybrids that used different forms. Do the same thing with the yellow. We have many yellow hybrids so use them.

They don't have to be perfect, but they should have different parents. Now start breeding by "mixing and matching." Assume you won't get what you want right away, so keep your population large enough to go several generations.

For most breeders, the population is outside your garden. You pluck plants from all that stuff that's available out there and breed them. So why are you using the same parents as everyone else? Don't you want something different? Go out and find little used, new or forgotten species. Look for plants that everyone else has overlooked, not because they're inferior but because they haven't been released, or are not well-known.

In my own program, I'm so far off the deep end that I can't find any plants outside my own project that can contribute to my goals. I knew from the start that I would have to start from scratch and breed in splendid isolation for many generations before reaching my goals. This year I planted seeds from over 120 crosses. I used about sixty parents, not just the best plants, but a wide variety of acceptable plants. I'll end up planting around 2000 seedlings, of which maybe half will make it to flowering. Am I breeding individuals? No! I'm breeding a population.

Can I offer any proof that breeding a population really works? All I can say is that I have a great deal of variability within my plants, and that they seem to be more variable now after five generations than when I started. With each generation, the population inches towards my goals, so that the very best plants a few years ago would now be unacceptable as parents. In a few years I'll consider today's plants inferior. I suspect that after five more generations, I will still have plenty to work with, and new stuff will continually pop up unexpectedly that will surprise me.

Pat Halligan is a member of the Whidbey Island Chapter.