Variety Trials
A seed catalog tells you what a variety is. A trial tells you what it did, once, in one place, under a stated protocol. AMSeed keeps those two kinds of evidence apart on purpose: nothing on this page is folded back into a variety's description, because a yield is a property of a season, not of a seed.
| Trial | Year | Season | Site | Entries | Design |
|---|---|---|---|---|---|
| Yield Characteristics of Indigo Purple Tomato Varieties in Kentucky | 2015 | full season | Frankfort, Kentucky | 10 (2 held here) | randomized complete block |
| Evaluation of Late Blight Resistant Tomato Varieties, 2013 | 2013 | full season | New York | 13 (11 held here) | not stated |
| Evaluation of Late Blight Resistant Tomato Varieties, 2012 | 2012 | full season | New York | 13 (10 held here) | not stated |
| Fresh-Market Tomato Variety Performance in 2009 | 2009 | spring | Lexington, Kentucky | 24 (7 held here) | randomized complete block |
| Yield and Income of Fall Staked Tomato Cultivars in Eastern Kentucky | 2009 | fall | Quicksand, Kentucky | 20 (6 held here) | not stated |
What is and is not recorded
These 5 trials evaluated 80 entries between them. AMSeed publishes results for 36 of those entries, covering 21 varieties, and lists the rest by name without their numbers. An entry goes unpublished for one of two honest reasons: the variety is not in this database yet, or the report's label maps onto more than one variety we hold and attaching a number would mean guessing which seed was in the ground.
Where a trial reports a figure we cannot attribute confidently, it is left out and the omission is stated on the trial's own page rather than quietly dropped. Each result carries the mean-separation letters its authors published, so the report's own statement about which differences are real survives into the database instead of being flattened into a ranking.
There is no "best variety" list here and there will not be one. Every trial in this record says in its own words that a single site in a single season does not settle how a variety performs, and a leaderboard built across them would contradict all three sources at once.
How this fits the rest of the data: methodology · variety database · research library