@Seedfork Larry, your question was very important because I found an error in the analysis in checking the conclusion. Thank you. However, when the problem was resolved the results were confusing. That lead to much more analysis. The figures below are the result of those analyses.



The percentages of the cultivars that rebloom in these diagrams is based on the number of each growth habit (Dor/Ev/Sev) in each bloom period. As an example, in diploids in the EE bloom season 45% of the Dor cultivars were registered as reblooming, 81% of the Ev cultivars were registered as reblooming and 71% of the Sev cultivars were registered as reblooming. All three percentages are independent of each other in this analysis.
Cultivars registered as having a VL bloom season appear to be unusual for both the diploids and the tetrpaloids. The percentage of those cultivars that rebloom is the same for each growth habit. That is unexpected. I am interested in possible explanations.
The main observation is that the percentage of cultivars that are registered as reblooming decreases the later in the bloom season the cultivars first bloom. Growth habit does not seem to have much effect on that decrease. Evergreen (evergrowing) cultivars have the highest proportion of rebloomers while dormant (deciduous) cultivars have the lowest - except for the VL bloom period.

Another question that arises from the observations is how do dormant/deciduous cultivars that first bloom very late manage to rebloom? Especially since there were 19 cultivars that were registered as Dor and reblooming out of a total of 36 tetraploid cultivars registered as very late and reblooming and 29 cultivars that were registered as Dor and reblooming out of a total of 46 cultivars that were registered as very late and reblooming in the diploids.

The number of cultivars registered as evergreen and reblooming by bloom time and the number of cultivars registered as dormant and reblooming by bloom time.

How likely is it that this is confounded by hybridizers in short season areas registering as rebloom based on having tested in long season areas?

That is something I had not even considered, I wonder how many others things I am not even aware of could affect the conclusions from such data? Still, I find it odd such registrations could sway the results. But, I guess there is no data to even began to evaluate that? I guess my surprise is that so many plants registered that bloom so late would ever have such a high percentage of rebloomers?
I was expecting to find that the earlier the plants bloomed the initial first bloom, the higher the percentage of rebloomers would actually be.

@Sooby

How likely is it that this is confounded by hybridizers in short season areas registering as rebloom based on having tested in long season areas?

Perhaps naively on my part, it should not be confounded by that possibility - aren't the details of each registered cultivar supposed to be as measured in the hybridizer's location and growing conditions?

I have to wonder how many ML are being registered as L and how many L are being registered as ML. I also wonder how many VL are correctly identified as Ev and Sev.

Isn't a problem with classifying by foliage habit that it can vary by growing conditions/location? Semi-ev particularly is open to interpretation also.

The reason I wondered about locations is that some hybridizers, maybe not many, hybridize in more than one area. Presumably they are only listed as being at one of those addresses but may have described the plant's behaviour in the other. Some do test in different climates also, and one would hope that the plant is described as it grows in the hybridizer's own garden but I wonder if that is always the case.

Reading the latest issue of the Daylily Journal, it appears that maybe a lot of hybridizers are not using the proper definitions for those terms. I see in the Robin Scott Elliott was pointing that out by posting the official way of determining those bloom time terms.

Sorry, cross posted.

From the AHS FAQ:

Extra Early (EE) – These daylilies are the first to bloom, and vary from March or April in the extreme South, to May or June in the North.
Early (E) – These daylilies bloom three to five weeks prior to the mass of bloom at midseason.
Early Midseason (EM) – These daylilies bloom one to three weeks before the height of bloom of most cultivars.
Midseason (M) – These daylilies bloom at the peak of the daylily bloom in your own garden. This ranges from May in the South to July in the North.
Late Midseason (LM) – These daylilies bloom one to three weeks after the height or peak of bloom in your garden.
Late (L) – These daylilies bloom when most others have finished blooming, usually four to six weeks after the peak of the season.
Very Late (VL) – These daylilies are the last to bloom, often late in the summer in the South, fall in the North.
Rebloomer (Re) – These daylilies bloom more than one time during a single season. Some of these bloom early (e.g., May or June) and then repeat in the fall. Others have a succession of bloom periods, one shortly after another for several months.

From the Registration guidelines:

EE - Extra Early.
A very few daylilies commence bloom earlier than 2-4 weeks before mid-season. These are Extra Early.
E - Early.
Daylilies that commence bloom 2-4 weeks before mid-season are considered Early.
EM - Early Midseason.
Daylilies that commence bloom 1-2 weeks before mid-season are Early-Mid.
M - Midseason.
Daylilies that commence bloom at "peak bloom" time are Mid.
MLa - Late Midseason.
Daylilies that commence bloom 1-2 weeks after mid-season bloom are Mid-Late.
La - Late.
Daylilies that commence bloom 2-4 weeks after mid-season are considered Late.
VLa - Very Late.
A very few daylilies commence bloom later than 2-4 weeks after mid-season. These are Very Late.

@Sooby

Isn't a problem with classifying by foliage habit that it can vary by growing conditions/location? Semi-ev particularly is open to interpretation also.

Yes. The growing conditions/location affect many of the characteristics recorded by the AHS when it registers daylily cultivar names. Scape height, number of branches, number of buds, flower size, rebloom, etc. can be affected by growing conditions.

The reason I wondered about locations is that some hybridizers, maybe not many, hybridize in more than one area. Presumably they are only listed as being at one of those addresses but may have described the plant's behaviour in the other.

They should not be describing the plant based on other growing locations.

Some do test in different climates also, and one would hope that the plant is described as it grows in the hybridizer's own garden but I wonder if that is always the case.

It would be inappropriate for a hybridizer to describe their introduction by its characteristics elsewhere.

It seems that the AHS is inconsistent.

If we want to study this a little more at any point, I'd volunteer to track data. If we wanted to pick plants to study, I could collect them in 2022 for study starting 2023, as long as it wasn't crazy like 200, but I might go as high as 50 or maybe 100. Seems like it would take hundreds to cover all the variables, though.

Or volunteers could just track all their registered daylilies. I'd probably need some help determining instant rebloom from a delinquent fan that just needed an extra week or 10 days to get going. But maybe it doesn't matter. Maybe we count and measure the visible scapes at the time of the first bloom for a plant, say there are 6, and if I end the main bloom season with 8, there were two "instant rebloom or delinquent" scapes. We might be able to determine when the plants bloomed vs registration by zone (or temperature?), and whether they are Ev, SEv, or Dormant in each zone, and ??? When I tracked all my plants in the past, there was a clear peak for using that "over 2-4 weeks before peak is EE" logic, and of course, if they rebloomed give all the previous factors.

Tim, this challenge could be given to the ADS/AHS. For example, many birdwatching clubs keep track of specific birds that visit their gardens all year round. Those data are then tallied to a national scale to look at migratory patterns etc.
I do not see why the ADS cannot just ask all of their members to keep track of blooming and reblooming cultivars every year, and provide a centralized database form to fill out. Ideally, even create a phone-app that can be used. I would imagine that is kinda their job to provide such insights into daylilies. Otherwise, why do they exist?

Well, I don't want to underestimate the ADS, but I think the problem is resources. When Cornell collects and tracks bird activity, they have a bunch of grad students to do the work and the university has a whole system for collecting contributions from wealthy people and graduates to fund projects. I'm not sure the ADS is designed to throw bodies or resources to develop the processes and the systems that would be required to put all that together.

I was thinking it would be easier to have a controlled amount of volunteers to spread an adequate but minimized amount of work.

But how cool would it be if the ADS connected with a university or two to get interns and resources for studies like this? My assumption is I might not be around to participate if the study has to wait for all that to be set up.

From the Daylily Encyclopedia - McNamara's chapter on Bloom Sequence (my adaptation).

The flowering time (FFO) of the cultivar 'Alan' was used to identify EM and of 'Aten' to identify ML.
You can read the Daylily Encyclopedia book at archive.org by "borrowing" it for an hour at a time (when it is "available").
Chapter 5 in that book has the Farr (Pennsylvania) Bloom Sequence chart, Webber Garden's Adaptation of the Farr chart for Maryland and also has a Florida Bloom Sequence chart by Watkins.

I was confused by the chart above at first, but looking at it helped me to actually understand why the categories are based on Mid or peak season bloom. It has always seemed to me that the first flower to bloom in the garden should be the VE, and the rest of the categories should be based on the weeks after that the rest of the plants bloom.
But then I realized that just one plant would actually be the basis of all the categories, where as with mid season, it is the largest class of flowers that determine the other divisions.
I am not sure there is a need for the VE and VVL categories shown in the above chart, but I seem to prefer the registration guidelines for the blooming categories. I have often thought that just for consistency maybe EE and EL or VE and VL should have been used, but maybe someone can tell me why that was not done?

I love the graph. As a gardener with possibly slightly obsessive compulsive issues, I don't like that there are 4 numbered weeks right of M, but 5 to the left. And that three categories get two weeks, and that 4 get only one. I do think that's how peak works in most people's gardens, though. I'm trying to balance my garden so the peak is spread out to more to both sides, but I know I'm an oddball with that. However, if you look at it from my balance-my-daylily-garden perspective, by definition M can't be 6th week. M should be 5 and 6, or as Spinal Tap might suggest, this one should go to 11. And while I'm thinking it through, if E and EM are based on first bloom, why aren't L and VL determined by the last bloom, if trying to define peak/M?

I still have house guests from Christmas, but when I get some time, I'm really looking forward to putting my actual bloom data into this time line, and the two already mentioned above, to see how things look.

Wow... sometimes I forget what a goofy nerd-geek I am, but now I'm looking forward to graphing my bloom tracking data like it's the Sunday crossword puzzle.

I'm looking forward to your graph Tim (and glad I'm not the only one bothered by the asymmetry)