I am going to describe a test of the pod versus pollen parent belief based on the information from the AHS registration database. This is not a scientific example but it looks for evidence that the two parents contribute differently to their seedlings flower size or scape height.
Using the recorded information from the AHS registrations this is what I did.
Since ploidy may have an effect on flower size and scape height only diploid registrations were analyzed.
First I needed to check that there was not a relationship between scape height and when a cultivar was registered. I ran regressions of plant height over time for various time periods looking for the end of the period of selection of shorter plant heights. It turned out that hybridizers selected for shorter scapes until about the early 1990s. Before that time there was a bias that caused scape heights of registered cultivars to be shorter each decade.
I then chose a time period after that point during which there was no significant regression (1994-2000). I chose only those registrations with both a single known pod and a single known pollen parent from that time period. There was no other selection of records. I sorted on pod parent name and I then looked up the parents' characteristics (flower size, height separately) until an arbitrary number had been accumulated. I then resorted on pollen parent name and I then looked up the parents' characteristics until an arbitrary number had been accumulated. I then resorted on both fields being numerical rather than strings and extracted all those that matched.
Scape Height
When I analyzed the relationship between the pod parent's scape height and its seedling's scape height I found that seedlings and their pod parent have a correlation coefficient of 0.47 and a standard error of ± 0.03 for scape height for a sample size (N) of 856 crosses. When I analyzed the relationship between the pollen parent's scape height and its seedling's scape height I found that the correlation between the seedlings and their pollen parent is 0.52 ± 0.03. The correlations are not statistically different. Correlation measures how similar the seedlings are to their parents in the characteristic being measured. If the pod parent determined scape height then the correlation between the pod parent and its seedling would be high and the correlation between the pollen parent and its seedling would be low, near zero and not significant. What we find is that both parents make equal contributions to their seedlings scape height and that is exactly what we expect from quantitative genetic theory.
Daylily seedlings resemble their pod and pollen parents equally in scape height. Scape height is inherited equally from the pod and the pollen parent.
Flower Size
When I looked at the relationship between the pod parent's flower size and its seedling's flower size I found that the correlation for flower size between seedlings and their pod parent is 0.62 ± 0.02, with a sample size of N = 1889 crosses. When I looked at the relationship between the pollen parent's flower size and its seedling's flower size I found that the flower size of seedlings is correlated 0.59 ± 0.02 with that of their pollen parent. The correlations are not statistically different.
Again, if the pod parent determined the seedling's flower size then the correlation between the pod parent's flower size and its seedling's flower size would be large and that between the pollen parent's flower size and its seedling's flower size would be small, near zero and not significant. That is not what I found. Both parents contribute equally to their seedling's flower size and that is what we expect from quantitative genetic theory.
Daylily seedlings resemble their pod and pollen parents equally in flower size. Flower size is inherited equally from the pod and the pollen parent.
There is another way to look at the registration information. I have also analyzed specific types of reciprocal crosses from the registration database to double check the more general results. For example, when cultivars with four inch flowers were crossed with cultivars with six inch flowers (4” X 6”) the seedlings had an average flower size of 5.11”, N=20 crosses (I did not choose which specific crosses to analyze; they were simply the first 20 in the listing). If the pod parent determines flower size the seedlings should have had flower sizes closer to four inches. The seedlings from the reciprocal cross (6” X 4”) had an average flower size of 5.09”, N=26 crosses. If the pod parent determines flower size the seedlings should have had an average flower size closer to six inches. The reciprocal difference is not significant. In any case it is in the pollen parent's direction not the pod parent's. The average flower size of the seedlings is the same no matter whether the pod parent has a flower size of 4 inches or of six inches and that is what quantitative genetic theory tells us should be the situation.
Again the seedling does not resemble its pod parent any more than it does its pollen parent.
What about flower colour? We could look at the inheritance of red flower colour, but Arisumi, a geneticist already did so - he did not find that the pod parent determined flower colour. Both parents determined flower colour.
Most of the material in this post was from my article Daylily Genetics Part 4 Pod or Pollen Parent: Do they determine different seedling characteristics? in the AHS Daylily Journal Winter issue 2011.
