Fifteen days of harvest in the field, eight in the high tunnel, from the same December 1 plant date. That is the tradeoff Kate Cassity-Duffey found running snapdragons both ways, and the first harvest date did not move at all.
Cassity-Duffey (University of Georgia Department of Horticulture) is a soil scientist who came to cut flowers through organic vegetable growers. She reports two trials. The first sets a high tunnel against the field with Madame Butterfly Bronze with White and Chantilly Bronze, about a hundred plants per plot, replicated four times. Tunnel stems ran about 10 centimeters longer in both cultivars, and the tunnel paid for it in density: flowers stretched down the stem instead of sitting compactly, which she ties to a winter that put February in the 80s and swung back to 20 degrees.
The second trial runs a January 7 plant date against March 11 across eight Johnny’s cultivars, fertilized at 100 pounds of nitrogen at planting plus 50 side-dressed, with potassium as potash. Sixty days between plantings did not buy 60 days between harvests. The March date gave three to four stems per plant against two.
She also walks the eight cultivars one at a time, with stem lengths: Antibes Apricot I and Avignon Apricot short, Avignon Lavender 2 around 50 centimeters, Cannes Light Bronze 60 to 65, Costa Summer 65 to 70 with very strong stems.
On organic fertility, her own specialty: nitrogen arriving as poultry manure, compost, or feather meal has to clear mineralization and nitrification before a plant can use it, while immobilization, leaching, and volatilization take a share. She thinks cut flower growers may be under-fertilizing.
Recorded July 23, 2026 at the second annual Cut Flower Symposium, hosted by the UGA Center for Urban Agriculture with the Georgia Green Industry Association. Segment 4 of 8. Runtime 20:18. CEU Categories: [PENDING].
Chapters
0:00 Introduction and Recorded Bio
0:36 Part One, and Why Snapdragons
1:32 Why Organic Growers Are Adding Cut Flowers
2:12 High Tunnel or Field, and Why the South Is Hard
3:20 Organic Fertility and the Maryland Gap
3:53 The Nitrogen Cycle, and Why Organic Is Harder
5:28 Trial One: High Tunnel Versus Field, and the Rating Scale
6:48 Results: Same First Harvest, Different Last
8:01 Harvest Duration: Fifteen Days Against Eight
8:40 Stem Length, Density, and a Strange Winter
9:41 High Tunnel Versus Field Takeaways
10:30 Trial Two: Two Plant Dates, and the Fertilizer Rates
11:47 The Eight Cultivars, One by One
13:48 Potomac Lavender and Costa Summer
14:34 Plant Date Results and Stems Per Plant
15:57 Takeaways, Year Two, and Contact
17:24 Q&A: Seeding Potomac Now for a Fall Bloom
18:11 Q&A: Same Cultivars in Year Two
18:37 Q&A: Spacing and Rows Per Bed
19:07 Q&A: What the Snapdragon Groups Mean
Questions Answered
Did the high tunnel move the first harvest date? No, and she says she was surprised. Both cultivars hit the same first harvest in the field and the tunnel, since flowering answers to light as well as temperature. Everything after that differed.
How were the snapdragons spaced? Four inch spacing, four rows per plot on three foot centers.
What do the snapdragon groups mean? They index when a cultivar performs best and how long it takes to flower. Group ones run cooler and faster, twos and threes longer. Some cultivars are groupless, Madame Butterfly among them.
Can you seed Potomac now for a fall bloom? She has the same question, knows growers doing it, and is starting one in August to find out. About 10 weeks from seed until transplant, with light the deciding factor. She says on air she thinks Potomac is a group one, then reads a chat correction giving it as typically a three to four and accepts it.
Resources
- University of Maryland Extension cut flower document, named at 3:30 as the only fertility guidance available and silent on rates. No title or address is spoken.
- Johnny’s, the seed source for the eight-cultivar trial.
- American Takii, thanked alongside Johnny’s Seeds.
- Her email is on the closing slide but is not spoken on the recording. It has to be added by hand before publication.
Platform Note (not part of the pasted description)
Length. Everything from the opening line through the Resources block measures 4,137 characters, inside YouTube’s roughly 5,000-character field, with about 860 characters of headroom. Segment 03 shipped with 26. The margin here is real but not infinite, so re-count after any edit.
All 20 chapter markers are kept, costing 902 characters. Each was verified by script against a block start time in the corrected transcript at Stage 2.
Eight of the twelve Q&A pairs were cut. Four carry over. What went, and why:
| Cut pair | Reason |
|---|---|
| Why snapdragons and not sunflowers | Interesting to an archivist, not to a grower scrubbing for a result. Kept on the website version, where the missing part two matters. |
| Why organic fertility is harder | The mineralization chain is in the lede at half the length. |
| How much the tunnel shortened harvest | The 15-against-8 figure opens the description. |
| Stem length and the density tradeoff | Both halves are in the lede paragraph verbatim. |
| What fertility went on, and why the rates look high | Rates are in the lede. The release-rate explanation is on the website version. |
| Short-stem versus long-stem cultivars | The stem lengths are in the lede as a list, which is the form a grower scans anyway. |
| Whether two months apart gave harvests two months apart | The finding is in the lede. |
| Stems per plant by plant date | Also in the lede. |
All twelve pairs are on the website version. The four kept here are the three chat questions a grower would search for plus the first-harvest result, which is the single most surprising finding in the talk.
No email address is printed, because none was spoken. She says her email is “at the bottom” of a slide and never reads it. Same shape as segment 03’s QR code: nothing an audio sweep could recover, and it must be filled by hand before publishing. This is the one item that blocks publication.
The Potomac group exchange is stated as an exchange. She gives group one, a participant corrects it in chat, she accepts. Writing either number alone would misrepresent the recording, and her own cultivar walk puts Potomac Orange and Potomac Lavender in group three or four.
Timestamp format is M:SS and MM:SS throughout, since the segment runs under an hour.
Scaling Note
Target. Stage 3 derives from the Stage 2 package and scales with it. Segment 03’s YouTube file ran 10.2 KB against an 18.2 KB archive summary, a ratio of 0.56. This archive summary is 15.0 KB, so the proportional target was about 8.4 KB.
Delivered. 8.2 KB, of which the pasted description is 4,137 characters. The description came in well under the ceiling for the first time in the symposium, and the reason is structural: three fewer chapters and a shorter talk to summarize. A 20-minute segment does not need segment 03’s condensing pressure.
What was not scaled. The chapter list and the Platform Note’s cut-pair table. Both are the parts a person editing this file later actually needs.
Processed for UGA Center for Urban Agriculture / Cut Flower Symposium Archives
Source: Stage 2 Archive Package, GTBOP_Archive_Summary_2026-07-23_CassityDuffeySnapdragons.md