Daylilies forum: Pod or Pollen Parent

Views: 1289, Replies: 13 » Jump to the end
Name: Maurice
Grey County, Ontario (Zone 4b)
Image
admmad
Apr 8, 2015 8:23 PM CST
This thread continues from Post #824234.

The question of whether seedlings may inherit characteristics differently from the pod parent versus the pollen parent could be quite important for professional animal and plant breeders and geneticists. That is because they are often involved with breeding for increasing food production, etc. So it is known how to objectively test for any such differences in plants. The basis for finding whether there are differences is the reciprocal cross.

If we had chosen to cross 'Dragonfly Dawn' and 'Born to be Wild' then the reciprocal crosses would be (pod parent first):
('Dragonfly Dawn' X 'Born to be Wild') and ('Born to be Wild' X 'Dragonfly Dawn').

We can label the seedlings from the cross ('Dragonfly Dawn' X 'Born to be Wild') as DDBW, and the seedlings from the reciprocal cross, ('Born to be Wild' X 'Dragonfly Dawn') as BWDD.

[to be continued with edited additions]
Edit #1
Choosing scape height (h) as plant habit characteristic and flower size (s) as the flower characteristic and relying on the values in the registered information:
'Dragonfly Dawn' h = 36; s = 7.5
'Born to be Wild' h = 32; s = 6.5
Assuming quantitative inheritance, no dominance and no differential parental effects then the expected values for the reciprocal seedlings are:
DDBW average seedling h = 34; average seedling s = 7.0
BWDD average seedling h = 34; average seedling s = 7.0
No reciprocal difference.

Assuming quantitative inheritance, some dominance for say larger size and no differential parental effects then the expected values for the reciprocal seedlings are:
DDBW average seedling h = between 34 and 36, say 35; average seedling s = between 7.0 and 7.5 say 7.25
BWDD average seedling h = between 34 and 36, say 35; average seedling s = between 7.0 and 7.5 say 7.25
No reciprocal difference.

Assuming quantitative inheritance, no dominance but a differential parental effect then the expected values for the reciprocal seedlings are:
DDBW average seedling h = between 34 and 36, say 35; average seedling s = between 7.0 and 6.5 say 6.75
BWDD average seedling h = between 34 and 32, say 33; average seedling s = between 7.0 and 7.5 say 7.25
A reciprocal difference in scape height and a reciprocal difference in flower size. Scape heights would differ on average by two inches and flower sizes would differ on average by half an inch. In a normal garden setting these average size differences would not be noticeable due to the much larger environmental variability and the small number of seedlings grown. In a research setting, depending on how well environmental differences are controlled and how many hundreds of seedlings are grown and measured (sample size) they could be shown as statistically significantly different. Whether they would be of any practical difference is another question.

In a normal garden setting, with the number of seedlings many daylily hybridizers grow, many crosses will not be able to show whether there are 'real' differences in parental effects. Those crosses where the parents have large differences in the appropriate characteristics, in this example scape height and flower size, could be used to examine possible differences in parental effects.

Edit #2

Choosing a different cross using the cultivars Stella de Oro (SdO) and Joyce's Curls (JC) we have
Stella de Oro - scape height (h) = 11; flower size (s) = 2.75 inches
Joyce's Curls - scape height (h) = 45; flower size (s) = 6 inches.

Reciprocal crosses would be Stella de Oro x Joyce's Curls (SdOJC) and Joyce's Curls x Stella de Oro (JCSdO)

The expected scape heights with no dominance or differential parental effects would be (11+45)/2 = 28
The expected flower size with no dominance or differential parental effects would be (2.75 + 6)/2 = 4.4 (rounded)

If there are differential parental effects and no dominance then the average scape height of SdOJC seedlings should be smaller than 28 and that of JCSdO seedlings should be larger than 28. There is a range of 17 inches for a maternal effect to appear in scape height.
If there are differential parental effects and no dominance then the average flower size of SdOJC seedlings should be larger than 4.4 and that of JCSdO seedlings should be smaller than 4.4. There is a range of 1.6 inches for a paternal effect to appear in flower size.

It is not too difficult to test for differences in parental effects; several established daylily hybridizers did so for themselves, presumably early in their hybridizing careers, although they probably did not rigorously follow scientific objective methods. There are some 'rules' that need to be followed to guarantee that one's results are objective and not affected by unknown factors that bias the results. We have to treat the seeds from the reciprocal crosses in exactly the same manner. We cannot choose to plant only the largest or plumpest seeds or the seeds from the first pods to mature, etc. We have to treat the seeds equally and randomly. We cannot discard the last seeds to sprout or choose only to plant the largest seedlings, etc. We cannot simply plant the seeds or seedlings from the two reciprocal crosses in one row each side by side. We could plant the seeds or seedlings in two rows but at random in each row. An easy way to do this would be to take one die from a pair of dice and roll it before planting the first seed or seedling. If it was 1, 2, or 3 then one would plant a seedling of SdOJC in the first position in row 1 and plant a seed or seedling of JCSdO beside it in the first position of row 2. If it was 4, 5, or 6 then one would plant a seedling of JCSdO in the first position in row 1 and plant a seed or seedling of SdOJC beside it in the first position of row 2. The die would be rolled again to choose which reciprocal cross to plant in the second position of row 1 and so on until done.

While the plants were growing one would water and fertilize both rows in exactly the same way and with the same amounts. When the seedlings bloomed one would measure the heights of all the scapes and of all the flowers. If there were too many then one would have to choose at random which specific ones to measure. Or one could choose a rule, before any seedling bloomed, such as to measure the height of the first scape to bloom on each seedling and to measure the flower size of the first flower to bloom on each scape, etc. Finally the average scape heights and average flower sizes would be compared for the two reciprocal crosses.

Edit #3

There has been scientific objective research with daylilies that provides evidence for whether there are genetic differential parental effects. It is "Heritability and genetic correlation in daylily selection" and it can be found at http://link.springer.com/article/10.1007%2Fs10681-011-0478-y Only the abstract is free to read but I have read and analyzed the research. They used 20 cultivars in ten sets of reciprocal crosses and looked at several characteristics including scape heights and bud counts. There were no reciprocal differences. The pod and pollen parent did not affect the seedling characteristics differently for any of the characteristics measured by the researchers.

Although my analyses (reported in the article in the Daylily Journal mentioned by Sooby/Sue) do not involve reciprocal crosses they can provide a hint as to whether there might be genetic differential parental effects in daylilies for characteristics such as scape height and flower size. Those analyses used the information reported in the AHS daylily database. There was no evidence that scape height or flower size was affected differently by the two parents. Those analyses examined the characteristics reported for seedlings and both their parents from 856 crosses for scape height and from 1,889 crosses for flower size.

I have to conclude that it is very unlikely that there are general genetic effects that cause pod and pollen daylily parents to affect their seedling offspring characteristics differently. I cannot rule out that there may be a few cultivars which when crossed with specific other cultivars produce reciprocal differences in their offspring due to genetic effects. I cannot rule out that reciprocal differences may be present in some gardens sometimes (when measured objectively and not subjectively) caused by non-genetic effects.

There are many factors that may cause reciprocal differences. Although there is no overall objective evidence for reciprocal differences in daylilies it is possible that some of those factors may operate in some daylily crosses or in some locations or under some conditions some times. One of these is maternal effects. One can think of these as effects caused by the pod parent when it is making the seed. Some pod parents may make many seeds or they make large seeds or they may provide the seeds with poor nutrition or good nutrition and so on. Characteristics of the seed can have visible effects on the seedling characteristics. Those effects can last for a long time but they usually fade away with the passage of time.

Most hybridizers do not make reciprocal crosses. That means that once they have decided to cross cultivar A with cultivar B they must decide in which direction they make the cross. Do they cross (pod parent first) cultivar A x cultivar B or cultivar B x cultivar A? Those decisions may be made completely subconsciously or they may be made at least partly consciously. There may be unknown patterns in how those decisions are made that may have effects on the seedling characteristics, at least for some time. A hybridizer might decide to use the cultivar with the higher bud count as the pod parent, or the cultivar with the most seeds per pod as the pod parent, or the cultivar with the most fans (and therefore scapes) as the pod parent. Such conscious or subconscious decisions could have unknown effects on seedling characteristics, at least for some period of time. Typically such effects would fade with time, but they might need many years to fade completely away and therefore influence conclusions.

There are also possible subjective effects if one does not measure characteristics objectively. For example, in a cross between a plant with 45 inch scapes (pod) and one with 15 inch scapes (pollen) the expected seedling average scape height would be 30 inches. If one looked at a seedling with a 30 inch tall scape, without measuring it, would one classify it as tall or short? Would one classify it as being like the pod parent or like the pollen parent? If the cross produced five seedlings with scape heights of 40, 35, 30, 25, and 20 inches and after blooming one culled and discarded the seedlings with scape heights of 25 and 20 inches and then a year or two later looked at the remaining seedlings would they be classified as more like the pod parent or the pollen parent? Hint, the seedling average is 30 inches.

Maurice
[Last edited by admmad - Apr 9, 2015 3:20 PM (+)]
Give a thumbs up | Quote | Post #825433 (1)
Name: Muata Kamdibe
Diamond Bar, California (Zone 10a)
Killing plants since 1992
Region: California Daylilies Vegetable Grower Plumerias
Image
Muata
Apr 8, 2015 11:41 PM CST
Maurice, thank you for starting this thread, as I am new to hybridizing DLs too. So, I will be anxiously waiting for your next installment :)

¨You have to get up and plant the seed and see if it grows, but you can't just wait around, you have to water it and take care of it.¨ - Bootsy Collins

Name: Donald
Eastland county, Texas (Zone 8a)
Region: Texas Enjoys or suffers hot summers Raises cows Plant Identifier
Image
needrain
Apr 9, 2015 5:48 AM CST
I'll be following this thread also. It's never made sense to me that one parent would have more influence than the other since during meiosis genetic material is contributed equally - unless mitochondrial inheritance has more influence than is usually attributed to it. I've just never seen an explanation that seemed logical as to why it would be true, but there is a huge amount of anecdotal evidence from people that it happens. Maybe it will be clear in this thread.
Donald
Name: Char
Vermont (Zone 4b)
Daylilies Forum moderator Garden Ideas: Master Level Celebrating Gardening: 2015 Hosted a Not-A-Raffle-Raffle Region: Vermont
Dog Lover Hybridizer Enjoys or suffers cold winters Garden Photography Keeper of Poultry Organic Gardener
Image
Char
Apr 9, 2015 5:51 AM CST

Moderator

The discussion on Pod or Pollen Parent started in the April Seedling Bloom thread found here...
The thread "April Seedling Blooms" in Daylilies forum

These are the beginning posts for this new thread topic .....

Post # 823076
spunky1 said:Becky the pod and pollen parent traits have been discussed since the first seedling bloomed, in my experience plant habit normally comes from the pod parent and the bloom will be influenced mostly by the pollen parent. This is not science and not always the case, but is only my opinion.


Post 823084
beckygardener said:Fred - Thanks for sharing your experience about genetic traits. I don't know where some of mine are getting their traits from? Must be the gene pool further back. I have been getting some odd blooms on some of my seedlings. Quite a few resemble the pod parent ... which is why I asked. And then there are those that look very similar to the pollen parent. But most look like something else! Rolling on the floor laughing Rolling on the floor laughing Rolling on the floor laughing It would be fun to do a gene tree on some just to see where those traits are coming from. I might make a collage some time of such a tree. That would be interesting and fun!


Post # 823130

sooby said:

Becky, Maurice wrote an article about this for the AHS Daylily Journal. It's available as a PDF here:

http://www.daylilynetwork.org/resource/resmgr/website_reprin...



Post # 823497

beckygardener said:Thanks, Sue! I will read that article!

Are the researchers still wanting strains of rust? I never mailed mine out. I had intended to, but apparently when we had company last year, I forgot to actually mail the rusty leaves to the university. I found them months later still in the envelope. (sigh)


Post #823823

admmad said:
Fred, I don't want to put you on the spot but I would like to please ask for some clarification.

I have seen the traditional belief that pod parents influence plant habit and pollen parents influence flower characteristics before. But no one ever describes what they specifically mean by 'plant habit' and 'flower'. There are a substantial number of possible characteristics that could describe plant habit and as many or more possible characteristics that could describe flowers.

What specific characteristics would you say were 'plant habit'?

Would these include height of the scape, number of buds, number of branches, how fast it produces new fans, width of leaves, length of leaves, dormant, evergreen, semi-evergreen, whether it has fibrous roots or tuberous roots, etc.? Which ones exactly do you include in the plant habit? Please name all the specific characteristics that seem to come from the pod parent in your crosses.

Also which plant characteristics would you include in those that were bloom and influenced mostly by the pollen parent in your crosses?

Would those include, when it flowers, if it reblooms, the flower width, the petal & sepal widths, the petal & sepal lengths, the flower colours, the eye patterns, the picotees, ruffling, whether it is fertile or sterile, the average number of seeds it produces per pod, the average number of pollinated flowers that produce pods with seeds, whether it is pod-fertile at all, whether it is pollen fertile or sterile, etc. Please name all the specific characteristics that seem to be influenced mostly by the pollen parent in your crosses.

And lastly, what proportion of the crosses you make are done reciprocally? That is, do you always cross Cultivar A x Cultivar B and Cultivar B x Cultivar A when you decide to cross two plants with each other? Do you sometimes make the crosses reciprocally? Or hardly ever? etc.


Post #824059

spiderjoe said:It's more scientific than saying that pod influences plant and pollen influence bloom. Go and study crosses and you will soon discover that theory is wrong. Some genes are stronger and rise to the top.


Post #824069

beckygardener said:Joe - I agree with what you said. I see the same thing in my garden with seedlings. I have many that I only know the pod parent. And many look a LOT like mama! Whistling Thumbs up

Post # 824234

spunky1 said:Very nice seedling Pat.

spunky1 said:
This is not science and not always the case, but is only my opinion.
POD PARENT
For myself Plant Habit= Scape Habit. Height, branching, bud count, strong enough to hold up the blooms, and must rebloom here. I always select seedlings on rebloom. I could care less about the width of leaves, length of leaves, dormant, evergreen, semi-evergreen, whether it has fibrous roots or tuberous roots as long as grows and preforms well everywhere. Of course I want a selected seedling to increase and I can tell after the second year if the increase is enough to keep it a third year.
POLLEN PARENT
When I look at a pollen parent I am looking at the bloom color, form and size. The color or pattern is the first thing that gets my attention. I could care less about the width of leaves, length of leaves, dormant, evergreen, semi-evergreen, whether it has fibrous roots or tuberous roots, I am only looking at the bloom because I have a pod parent that has all the other desired assets. I will make at least two of the same cross to ensure I get some seed from that cross, maby 5% I will make more than two cross's.
DESIRED RESULTS.
I really admire people that have all this scientific knowledge and hope it helps them in what there trying to achieve in the daylily world. I have been doing this for a long time and when you get down to the nitty gritty you never know what you will cross with what until you walk out the door each morning and see what's blooming. For myself I agree with Papa John "this is not rocket science its pizza and I love pizza." Again just my opinion.


Post # 824729

GDJCB said:I recall going to look at a bull with my grandfather when I was in my early teens, we walked up to the pen and grandpa immediately said "good looking young bull, he will do". The man selling the bull, who was young, I would guess mid-late 20's, obviously took my grandpa's statement "he will do" as almost an insult and begin to list all the "stats" (birth weight for grand sire, sire, finish weight, etc.). My grandfather waited and listened and when he was finished just said "yep, he will do". Bull threw light weight calves for the first time heifers that gained well and graded well at time of sale. They both were right and the end result was the same, one relied on experience and sight to see what was there, the other, pedigree and stats to tell them. I have my own opinion on which one has more value, but it is only one mans opinion and not meant to down grade those who choose to look rather than to see. More than one way to skin a cat, though to be clear, I have not tried any.

Gale



Post #825345

admmad said:This post is going to be a bit long and then I will start a new thread for pod and pollen parents. I am not attempting to change anyone's opinion. I use as a basic starting point that when someone makes observations in their growing conditions that their observations are correct. However, the biological causes or reasons behind those observations may not be the ones assumed.



In my opinion, people with long and deep experience can look at something and see what is there; those without such experience need to objectively measure what they see to have the same amount of information with a similar level of confidence. There are, however, some possible catch-22s in breeding - in effect, sometimes what one sees is not what one gets.

One possible example of the complications of hybridizing.

A hybridizer has a certain set of goals and has found five plants that fulfill the requirements. One of those requirements is that the plant has at least 25 buds per scape. All five plants have 25 buds per scape and all equally fulfill the other requirements. The hybridizer can only afford to purchase one of the plants. If the hybridizer relies only on the visible characteristics of the plants then the choice of which plant to use as a parent is effectively random. The hybridizer can only use the visible characteristics (the phenotypes) to estimate the breeding values (how well the plants are at passing on their characteristics to their seedlings) of the five plants. However, if the pedigrees of the plants are available and if information is available about the characteristics of the parents then the hybridizer may have more information to help make the choice.

Pod Parent A (32 buds) X Pollen Parent B (18 buds) -> Potential Breeding Plant 1 (25 buds)
Pod Parent C (25 buds) X Pollen Parent D (25 buds) -> Potential Breeding Plant 2 (25 buds)
Pod Parent E (10 buds) X Pollen Parent F (40 buds) -> Potential Breeding Plant 3 (25 buds)
Pod Parent G (35 buds) X Pollen Parent F (30 buds) -> Potential Breeding Plant 4 (25 buds)
Pod Parent H (15 buds) X Pollen Parent J (20 buds) -> Potential Breeding Plant 5 (25 buds)

With the added information about the characteristics of the parents, when the pedigrees are available and their stats are available there is more information that can be used to help make a decision.
Breeding Plant 1 has the expected number of buds that we would predict from a quantitative characteristic with no dominance, that is (32 +18)/2 = 25 buds.

Breeding Plant 2 has the expected number of buds that we would predict from a quantitative characteristic with no dominance, that is (25 +25)/2 = 25 buds. We would expect that the seedlings produced by this plant would be less varied in bud count than those produced by plant 1.

Breeding Plant 3 has the expected number of buds that we would predict from a quantitative characteristic with no dominance, that is (10 +40)/2 = 25 buds. We would expect that the seedlings produced by this plant would be more varied in bud count than those produced by plant 1.

Breeding Plant 4 has less than the expected number of buds that we would predict from a quantitative characteristic with no dominance, that is (35 +30)/2 = 32.5 versus 25 buds.

Breeding Plant 5 has more than the expected number of buds that we would predict from a quantitative characteristic with no dominance, that is (15 +20)/2 =17.5 versus 25 buds.

When we use a plant's visible characteristics (its phenotype) as the basis for choosing whether it will be used as a parent we are relying on the strength of the relationship between a plant's genotype and its visible characteristics. But those relationships (called heritability) can vary from zero to one. When it is zero the visible characteristics cannot tell us anything about its genotype. When it is one then the physical characteristics are a perfect measure of its genotype.

Returning to the example above. A plant with 25 buds may have a genotype for 25 buds, for more than 25 buds or for less than 25 buds. We cannot know what its genotype is for certain until we have made crosses and seen what bud counts its seedlings have (on average). That is called a progeny test. Nowadays, because of the possibility of using artificial insemination, progeny testing is done extensively in the beef and dairy cattle industry.

Many daylily hybridizers use progeny testing, even though they may not identify it as a method they use. When they choose a plant to use as a parent they often cross it with many other daylilies. If it does not produce at least a few seedlings of sufficient quality it is may not be used again as a parent - that is effectively progeny testing to choose 'good' parents.


Name: Muata Kamdibe
Diamond Bar, California (Zone 10a)
Killing plants since 1992
Region: California Daylilies Vegetable Grower Plumerias
Image
Muata
Apr 9, 2015 10:30 AM CST
Maurice, thank you for starting this thread, as I am new to hybridizing DLs too. So, I will be anxiously waiting for your next installment :)

¨You have to get up and plant the seed and see if it grows, but you can't just wait around, you have to water it and take care of it.¨ - Bootsy Collins

Name: Maurice
Grey County, Ontario (Zone 4b)
Image
admmad
Apr 9, 2015 3:15 PM CST
needrain said:unless mitochondrial inheritance has more influence than is usually attributed to it.


In many plant species both mitochondria and chloroplasts may be inherited from the pod parent predominantly and not from the pollen parent. They each have their own genes as well as relying on many other genes on the normal chromosomes in the nucleus for proper functioning. There are typically only 50 or so genes in mitochondria and a little over 100 in chloroplasts. There are typically 25,000 pairs of genes on the chromosomes which are usually inherited equally (one of each pair) from the pod and pollen parents.

The mitochondria and chloroplast are very important to the plant. Mutations can occur anywhere. Certain pollen sterility is inherited through mutant mitochondria. Certain leaf variegation is inherited through mutations in the chloroplasts. But other pollen sterility and leaf variegation is inherited from both parents through the nuclear chromosomes.

However, there are many other factors that can affect crosses and cause the seedlings from reciprocal crosses to be different. One cannot know whether there are differential parental effects by knowing about all the possible factors that can make seedlings from reciprocal crosses different. Typically those effects do not show patterns across characteristics - they would affect or not affect each individual characteristic independently. The chloroplasts make the food the plant needs and the mitochondria enable the plant to use that food for whatever purposes it needs. Characteristics such as scape height, number of buds, rebloom or not, etc. use that food but so do characteristics such as flower size and flower colour, pigment patterns, etc. Basically, it is likely that every characteristic of a plant relies on that food, along with other material from the soil, etc. Every characteristic of a plant relies on both the chloroplast and the mitochondria. In that respect, scape height and flower size are not different. Characteristics thought of as plant habit are not different from those thought of as affecting the flower.
Maurice
Name: Becky
Sebastian, Florida (Zone 10a)
Celebrating Gardening: 2015 Daylilies Hummingbirder Butterflies Seed Starter Container Gardener
Charter ATP Member I was one of the first 300 contributors to the plant database! Garden Ideas: Master Level Lover of wildlife (Black bear badge) Birds Ponds
Image
beckygardener
Apr 9, 2015 7:53 PM CST
A great discussion, Maurice! Always an education when you post!

I would really like to know more about bloom colors, ruffle/teeth, patterning, eye size, etc. and how those genes can be passed to it's children. I haven't done enough crossing to figure all that out. Is there an article already written in layman terms that I could read? Or could you do us the honor of delving into that aspect of the genetic pool?

I often wonder how growers can cull plants the very first year, when those very plants might improve with successive years? How do the growers KNOW what to cull and what not to cull?
What lies behind us and what lies before us are tiny matters, compared to what lies within us.
Garden Rooms and Becky's Budget Garden
Name: Donald
Eastland county, Texas (Zone 8a)
Region: Texas Enjoys or suffers hot summers Raises cows Plant Identifier
Image
needrain
Apr 9, 2015 9:38 PM CST
admmad said:

...... There are typically only 50 or so genes in mitochondria and a little over 100 in chloroplasts. There are typically 25,000 pairs of genes on the chromosomes which are usually inherited equally (one of each pair) from the pod and pollen parents.

However, there are many other factors that can affect crosses and cause the seedlings from reciprocal crosses to be different.

.........Every characteristic of a plant relies on both the chloroplast and the mitochondria. In that respect, scape height and flower size are not different. Characteristics thought of as plant habit are not different from those thought of as affecting the flower.


Less than 200 genes outside meiosis from the pod parent and 25,000 pairs equally shared by both parents involved in producing offspring. If it is 'equal', then what are the factors that effect a cross and tilt characteristics to one or the other of the parents? If it's truly equal, then it's similar to a coin toss. Completely arbitrary on whether offspring favors one parent or the other. If there are factors aside from chloroplasts and mitochondria, then those are the ones I'd like to hear information about. Fertility and variegated foliage are pretty well known as things influenced maternally. It's the other factors affecting those 25,000 pairs contributed by the parents. Where does the favoritism come into play?
Donald
Name: Maurice
Grey County, Ontario (Zone 4b)
Image
admmad
Apr 10, 2015 8:12 AM CST
If it is 'equal', then what are the factors that effect a cross and tilt characteristics to one or the other of the parents? If it's truly equal, then it's similar to a coin toss. Completely arbitrary on whether offspring favors one parent or the other. If there are factors aside from chloroplasts and mitochondria, then those are the ones I'd like to hear information about.


Those other effects that may cause a genetic (not environmental) reciprocal difference to appear in a characteristic:
gene imprinting or parent-of-origin expression differences
segregation distortion
transmission distortion of entire chromosomes in tetraploids
gametic selection or competition
linkage with the self-incompatibility genes
sex differences in recombination

These are effects that have been found to occur in one or more other plant species and that can result in differences in reciprocal crosses. Without specific research in daylilies we cannot know if any of these effects occur. If they do occur there is no way to predict which characteristic or characteristics might be affected nor whether the seedlings would resemble the pod parent or the pollen parent more for any specific characteristic. When they do occur in a plant species they are likely to occur only in specific crosses of certain cultivars and not in others. Without specific research we cannot predict how frequently any such effect might occur. It is unlikely that any of these effects would be generally present in a majority of daylily crosses but it is not impossible. These effects are not typically large nor extremely common in the species in which they occur.

Professional plant breeders and geneticists have been working with a knowledge of the basic (Mendelian) genetics for a little over a hundred years. Most of their work is done without making reciprocal crosses. Sometimes they make reciprocal crosses. Most of the time the reciprocal crosses do not show any differences. Sometimes they do show an effect. To place any confidence that those effects are real and not just caused by normal variability and random chance the same experiments have to be repeated by other researchers and the effects shown to be consistently present and caused by the same factor(s) consistently. Research is infrequently repeated and negative findings are often never published. Recently, with the rise of 'molecular' genetics many of the methods used by Mendelian geneticists to study inheritance have declined in use. That often makes it very difficult to interpret some modern observations.

For example, parent-of-origin expression differences:
A reciprocal cross is made, A x B producing AB seedlings and B x A producing BA seedlings. Then the researchers look at the amount of material produced by thousands of different genes in the A, B, AB and BA plants and compare those amounts statistically. Basically every measurement made on different plants will differ. So the researchers have to set a level which they will consider as indicating that a difference is 'real' and not just due to normal random variability. They will extract from their observations those genes that show differences in expression above their cut-off level and then examine whether the allele from the pod parent or the one from the pollen parent is the one that is expressed more. However, there is no information about how those differences in expression affect the visible phenotype/characteristics that we see in the plant and that would be important to breeding goals. And then the same research would need to be repeated by other researchers. Because of the complexity of the research that may or may not happen. Most such research is relatively recent and I have not followed any of it closely enough to know whether any of it has been repeated by other researchers and confirmed often enough to be accepted.

One of the catch-22s of much modern genetic research is that the characteristics being examined, such as the level of expression of different alleles, are potentially affected strongly by the environment. Environment in these cases means not only, how much water, fertilizer, sunlight, etc., the plants receive but also where they were grown, in what year they were grown, how closely they were planted together, what insects or other pests affected them, what diseases affected them, etc.

When we look at a plant we have a basic foundation for genetic effects for any characteristic that we can classify or measure. That is:
Phenotype (appearance) = Genotype + Environment + Genotype X Environment Interaction.
For some genes and some characteristics the effects of the genotype are so large and the effects of the environment and its interaction so small that they can safely be ignored. Some of the characteristics Mendel worked with are like that. But for other genes and characteristics the effects of the environment and its interaction are much more important. In those cases researchers make the environment as constant as possible so that they can ignore its effects (as well as those of its interaction). That may make the results applicable only to the conditions used in that research. Other times the researchers themselves repeat their research in different years and in different locations but usually holding all other known important environmental factors constant. This provides some information about the effects of the genotype X environment interaction.

Lets look at gametic selection or competition:
The pod parent makes the seed and its various parts while the pollen parent makes the pollen and its various parts. To do so each parent uses some of those 25,000 genes. But seed and pollen are different. So different genes (of the 25000) are used by the pod parent to make the seed and its parts than are used by the pollen parent to make the pollen. Some of the genes used to make the seed will be used by the seedling for other purposes. Some of the genes used to make the pollen will be used by the seedling for other purposes. This example is the sort of effect that can happen.
Cultivar A has a the genotype 11 for a gene that is used to make something for pollen. The gene is also used in making scapes. Cultivar B has the genotype 12 for the same gene. When cultivar A makes pollen all the pollen grains have the genotype 1. When cultivar B makes pollen half the pollen grains have the genotype 1 and half the pollen grains have the genotype 2. The gene is not used to make any part of the seed. Pollen grains with genotype 2 always beat pollen grains of genotype 1.
The cross B x A produces equal numbers of 11 and 12 seedlings because it produces only genotype 1 pollen grains.
The reciprocal cross A x B produces only 12 seedlings because genotype 2 pollen grains beat genotype 1 pollen grains.
The average scape heights of the seedlings from the reciprocal crosses will be different.

I can go through each effect in the list above and explain how it can cause differences in reciprocal crosses. But we have no information whether those effects are present in daylilies, we have no way of predicting which characteristics might be affected or in which direction the effects would change the seedlings.

The best we currently can do are the sorts of analyses I did for the characteristics and crosses recorded in the AHS cultivar registration database or in the future do our own specific well designed research to investigate characteristics we consider important in our own breeding populations. In the analysis I did for scape height and flower size I compared the relationship between the pod parent's scape height and its seedling scape height and the relationship between the pollen parent's height and its seedlings scape height. If any of those factors in the list were present (unless more than one was present and they balanced themselves out) then the seedlings should resemble one parent more than the other for scape height or flower size. That was not the case. So we can assume that none of those factors act strongly enough and generally enough in daylilies to affect scape height or flower size. But to conclude that none of those factors affect any specific cross under specific conditions in specific locations, etc. we would have to design and conduct the necessary research on the specific crosses in daylilies.

Edit #1
Maternal effects
Let's think about breeding animals for a few moments. I'll make the cross between two breeds of dogs, St. Bernards (S) and chihuahas, (C) and I will look at two characteristics, the average birth weight of puppies and the average number puppies in a litter. The data is from websites and it does not matter if it is not exactly correct. Both characteristics are different between the two breeds, with the average birth weight of St Bernards apparently one or two pounds and that of Chihuahuas is about four ounces. St Bernard litter sizes range from six to eight while those of Chihuahuas range from two to three. I make a reciprocal cross B X C and C X B and I examine their birth weights and litter sizes. There will be reciprocal differences. That is because both characteristics are mainly the mother's and not the offspring's and the father has no effect (or very little) on them. Differences in birth weight may cause differences in other characteristics in the puppies that last for years or their entire lives but would not be genetic.

Daylilies have equivalent characteristics, average seed size and the average number of seeds in each pod. Both characteristics are mainly the pod parent's and not the offspring's and the pollen parent has no effect (or very little) on them. Differences in seed weight and the nutrients stored in the seed may cause differences in other characteristics of the seedlings. Those differences may last for years but are less considerably less likely to last forever in the perennial daylily. That is because animals more or less stop growing and developing when they reach adult size. Daylilies do not stop developing, they more or less renew themselves every time they produce a new fan. Maternal effects that influenced the characteristics of the seedling's first fan or first few fans will typically decrease as each new fan is created.

Returning to genetic effects.
Basically under normal circumstances (disregarding mutations that may occur during growth) every cell in a daylily is genetically identical. Yet when one looks at a daylily there are enormous differences in appearances and functions. We see the differences - roots, crown, leaves, scape, petals, sepals, pistils, stamens, seeds, pollen, etc. yet they are all genetically identical. That is because although all cells are genetically identical not all genes have to be used by all cells.

Lets just consider seeds and pollen. Some of the genes used to produce seeds will be used only for that. Some of the genes used to produce seeds will also be used in other parts of the daylily. Some of the genes used to produce seeds will also be used to produce pollen. Some of the genes used to produce seeds will also be used to produce pollen and will also be used to produce other parts of the daylily. Some of the genes used to produce pollen will only be used for that, and so on.

Some of the genes used to produce seeds will not be used to produce pollen but will be used to produce, and in other parts of the daylily.
Some of the genes used to produce pollen will not be used to produce seeds but will be used to produce, and in other parts of the daylily.
If there is genetic variability in these two categories of genes that has visible effects in characteristics then there can be reciprocal differences.
Maurice
[Last edited by admmad - Apr 11, 2015 8:17 AM (+)]
Give a thumbs up | Quote | Post #826265 (9)
Name: Maurice
Grey County, Ontario (Zone 4b)
Image
admmad
Apr 10, 2015 9:06 AM CST
beckygardener said:A great discussion, Maurice! Always an education when you post!

Thank you.
I would really like to know more about bloom colors, ruffle/teeth, patterning, eye size, etc. and how those genes can be passed to it's children. I haven't done enough crossing to figure all that out. Is there an article already written in layman terms that I could read? Or could you do us the honor of delving into that aspect of the genetic pool?

The simple answer is that we do not know.
Mendel worked with a very specific set of conditions and characteristics, as was necessary to discover the basic underlying concepts of genetics. But even diploid daylilies do not fit those simple conditions and characteristics. Tetraploid genetics are much more complicated than diploid genetics.

You may find that some hybridizers indicate that eye patterns are 'recessive'. Others may indicate that picotee edges are 'recessive', etc. I suspect that none of the hybridizers have done the definitive research to identify the inheritance patterns of any daylily characteristic. I have results from crosses that suggest that picotee edges are not recessive but may follow the expectations of quantitative genetics. When characteristics are found to be actually recessive they may only be recessive to certain other characteristics. So one must always indicate what a characteristic is recessive to. As a theoretical example, pink flower colour could be dominant to white flower colour but recessive to purple flower colour.

The bottom line is we do not know how any characteristic is inherited in daylilies. Arisumi tried to investigate the inheritance of red flower colour but because of the self-incompatibility and strong inbreeding depression present in daylilies was not able to take the analysis far. Daylilies are not a very good material to understand genetics. No one else has even attempted an in depth research study of the inheritance of any daylily characteristic.

My advice is to consider most characteristics to be quantitative in daylilies. That would probably apply to depth of flower colour, width of eyezones, width of picotees, teeth size and number, etc. With quantitative characteristics when one does not know exactly how they are inherited one starts by assuming the offspring will be mid-way between the two parental values on average. That is not by any means what one must find when one actually does a cross and examines the seedlings but it is the starting guesstimate. Daylilies have the potential to show hybrid vigour or heterosis. In crosses were that is present then the seedlings may average better than the better parent. If heterosis for scape height was present in a particular cross then if tall was considered better and the parents had scape heights of 36 and 24 inches the beginning expectation would be that the seedlings would average 30 inches but if heterosis was present in this cross then the seedlings would average more than 36 inches. It is important to know that even if heterosis is not present in the cross there may be seedlings that are taller than 36 inches and seedlings that are shorter than 24 inches. And then again there may not be. The amount of variability in the seedling heights can also vary even when the parents have the same values. In one cross of plants with heights of 36" and 24" (36" x 24") the seedlings might range in height from 34" to 22"; in another cross using different parents with the same scape heights (36" x 24") the seedlings might range from 38" to 22", etc.

Edit #1

I would suggest doing progeny tests and not being surprised if what appear to be very similar crosses do not produce similar seedlings.

Mendel had very special circumstances in that he investigated characteristics that were determined strongly by single genes and his material was very inbred and did not have much if any genetic variability for those characteristics. We now know, from research on other plants, that flower colour is determined by at least 35 genes and probably 60 or more. Mendel was either working with plants that had variability in only one of those genes (or that had very minor variability in a few other of those genes) or he ignored any minor variability in flower colour. Daylily hybridizers do not tend to ignore flower colour differences, even minor ones. Since most daylilies are not produced by self-pollinations and since they are not inbred lines there is probably genetic variability in nearly all, if not all, the genes that affect flower colour. It is relatively simple to understand and predict the results of crosses when only one (segregating) gene is involved but by the time one has four or five it becomes easier to treat the characteristic as quantitative.
Maurice
[Last edited by admmad - Apr 11, 2015 7:28 AM (+)]
Give a thumbs up | Quote | Post #826287 (10)
Name: Becky
Sebastian, Florida (Zone 10a)
Celebrating Gardening: 2015 Daylilies Hummingbirder Butterflies Seed Starter Container Gardener
Charter ATP Member I was one of the first 300 contributors to the plant database! Garden Ideas: Master Level Lover of wildlife (Black bear badge) Birds Ponds
Image
beckygardener
Apr 10, 2015 3:54 PM CST
Maurice - "The simple answer is that we do not know." I suspected this might be your answer. Thank you !!!!

Ok! So it is basically a prayer and a hope that you get what you are hoping for in a bloom through a particular cross? That would explain why so many seedlings are culled by daylily breeders. Breeders are looking for something in their seedlings that is very specific or on a narrow scope. If they don't get it in a seedling, the seedling is culled?
What lies behind us and what lies before us are tiny matters, compared to what lies within us.
Garden Rooms and Becky's Budget Garden
Name: Maurice
Grey County, Ontario (Zone 4b)
Image
admmad
Apr 10, 2015 5:07 PM CST
beckygardener said:So it is basically a prayer and a hope that you get what you are hoping for in a bloom through a particular cross?

To some extent.
Genetics is a numbers game. The more seedlings one grows from a cross the more likely it is that something "more extreme" or "better" will appear.

Some hybridizers will make crosses with the hope that they will find seedlings that are better than their parents. They might be looking for longer teeth, or for a wider picotee edge or for longer petals, or for taller scapes, etc. They are selecting for the characteristic and they expect to make relatively slow but steady progress in the direction of their goal. They might cross their best cultivar with a half inch gold picotee with their next best flower with a half inch gold picotee and if they found an equally good seedling with a gold picotee that was wider than a half inch they would keep it and discard the rest. They might do this for purples, pinks, lavenders, etc.

Other hybridizers may have a different sort of goal. For example, they might feel that there were too few spidery cultivars with patterned eyes. Or they might feel that there were too few spidery cultivars with gold picotees, etc. They might cross spidery cultivars without picotees with non-spiders with the widest gold picotees they could obtain in the hopes that some of the seedlings would not be too round and would have a bit of a gold picotee. Even if they did not find any seedlings with the characteristics they were looking for they might still keep some of the seedlings to use in other crosses hoping that the next generation might have some seedlings with both characteristics to some extent.

That would explain why so many seedlings are culled by daylily breeders. Breeders are looking for something in their seedlings that is very specific or on a narrow scope. If they don't get it in a seedling, the seedling is culled?

They are usually looking for something that is noticeably better than what is readily available. Lets say I measured the flower size of 1000 cultivars and found that only 1% had flowers ten inches or larger. I take all my daylilies that have eight inch or larger flowers and cross them with each other. I am going to discard all the seedlings that do not have at least a nine inch flower. I might only be able to keep three out every hundred seedlings. I cross all the seedlings with nine inch flowers together and in this next generation I am only going to keep seedlings that had at least ten inch flowers. I might only be able to keep one or two seedlings out of every hundred. The more severely one selects the quicker the characteristic changes in our breeding population. But unfortunately, usually also the more inbred the population becomes and the less generally robust. What proportion of seedlings one keeps (and what proportion one discards) depends on how strongly one is selecting.


Maurice
Name: Becky
Sebastian, Florida (Zone 10a)
Celebrating Gardening: 2015 Daylilies Hummingbirder Butterflies Seed Starter Container Gardener
Charter ATP Member I was one of the first 300 contributors to the plant database! Garden Ideas: Master Level Lover of wildlife (Black bear badge) Birds Ponds
Image
beckygardener
Apr 10, 2015 5:36 PM CST
Maurice - You are so VERY helpful! I want you to know that I (for one) really appreciate you taking the time and making the effort to answer my hybridizing questions.

I am just a home hybridizer. Not a major breeder. So I can not grow 100-1000's of seedling every year. My pitiful efforts are basically just a fluke ... should I happen to get something worthwhile. And I am not looking to hybridize to sell. My main goal is to create some "nice and a little different" daylilies to name after loved ones (to honor or remember them) and then share those plants around with friends and relatives. I also do not spray for rust. I am hoping to get lucky and grow some nice crosses that show some rust resistance. Those are the 2 goals in my meager hybridizing efforts. I have over 300 daylilies in my garden beds. All but 3 plants are seedlings. I will eventually have to cull some as I have many rust bucket daylily seedlings. I find it quite fascinating and fun to try my hand at hybridizing. Thank you for answering so many of my questions! I now understand about the fan and scape genetics and how to cross to get a taller or shorter or more branching scape. You helped me make sense of the part of hybridizing that perhaps we can create. The bloom itself is hit and miss .... many more misses than hits .... by far! Thumbs up Thank You!
What lies behind us and what lies before us are tiny matters, compared to what lies within us.
Garden Rooms and Becky's Budget Garden
Name: Muata Kamdibe
Diamond Bar, California (Zone 10a)
Killing plants since 1992
Region: California Daylilies Vegetable Grower Plumerias
Image
Muata
Apr 11, 2015 4:31 AM CST
I agree
¨You have to get up and plant the seed and see if it grows, but you can't just wait around, you have to water it and take care of it.¨ - Bootsy Collins

« Garden.org Homepage
« Back to the top
« Forums List
« Daylilies forum
You must first create a username and login before you can reply to this thread.

Today's site banner is by dirtdorphins and is called "Dianthus 'Nyewood Cream'"