1) Know your germplasm (a lifelong pursuit!)
2) Identify best germplasm (through observation plots) for your needs & situation.
Trials can be used as selection nurseries as well.
3) Enrich and widen your material.
4) Put material in the best form.

A good trial is a test.
- Make sure it is not a beauty contest.
- Make sure it is a challenge.
- Less than optimum conditions gives you a better chance to find the workhorses and eliminate the prima donnas.

Why trial?
1) True comparision is the only way you'll know if your variety is worth the time.
2) You must be ruthless.
3) Its not what you keep, its what you throw away.
4) Often items need more than one look!

What do you compare your variety to?
1) Popular commercial varieties
2) Older standards
3) Learn the varieties your growers are using. Go see it in their field and learn their needs
4) Keep an open mind:
- your views on value will change
- the old will become new

Essentials for meaningful trials
1) Multiple replication
2) Lay out in a block design to reduce field effects
3) Multiple locations (and soil types) if possible
4) Multiple plantings per location for different environments, experiencing different growing years
5) Score important traits (no more than 4-5 traits at a time).
6) Use a 1-9 scale, not a 1-5 scale. 1 is poorest and 9 is best. Find the poorest and best plants first and rate them 1 and 9 respectively. Average the score for each accession over 3 reps.

To really execute, plant and evaluate a true trial properly takes time, energy and ultimately, money. Not everyone in this room will do real trials any time soon. Some of you have started and more will in the future. We already have two replicated trials scheduled for this year at Abundant Life and we also have a bunch of observation plots because we realize there's no way we can do trials on everything. Our friends here at Cornell do real replicated trials. Are they painful, George?

A: It depends on whether you're doing it for hands-on observation or whether you're doing it for statisfical publication.
J: True. But even if you're not doing it for statistical publication you have to get serious about taking numbers. Don't feel bad if all y ou can do is an observation plot. But observation plots don't give you as much real good true information as trials.

Know your germplasm. Get to know the crop intimately. You've got to know what you put in the trials. A lot of people who put time and money into trials waste it by putting junk in the trial, though sometimes the only way you find out its junk is by putting it in the trial. Sometimes you can find out it is junk just by doing observation plots or looking at it in other people's fields. A natural outgrowth of knowing the crop intimately is to be able to know the best germplasm for your needs or situation. Really you can only do that with a trial­evaluate and ultimately select. Trials can also be used as selection nurseries where you pull plants out. However, sometimes you do a trial and unfortunately just eat it, sell it in salads, or plow it up. A trial is really a test.

A lot of people misuse the word trial. I like to use the word evaluate. When people say they trialed that variety, did they really evaluate it or did they just run it out in an observation plot and peek at it? A field note like "looks good here" doesn't tell you anything. A trial is an evaluation for that thing on that day in that place under those conditions and nothing more. The same variety can look different in different conditions. I like to make it a little more challenging (but not too challenging) by putting it out in less than optimal conditions. Less than optimal conditions give you a better chance to find the workhorses and eliminate the prima donnas. You want to make sure its not a beauty contest.

Why trial? Why evaluate? True comparision is the only way you'll know if your variety is worth your time. If you see a pepper variety in a farmer's field that looks really good but is the only pepper they're growing, you don't have a frame of reference to put that observation into. You don't know how good another one would do in comparison.

You must be ruthless and dispassionate to be able to say, "This is not good enough for me." Its not what you keep but what you throw away that's ultimately important. Watch out for being kinder to your stuff than to the other person's stuff. I've done this myself. That's why it is better to be blindfolded or use a chart where you don't know what is in what position.

Do you want to base those choices on only one look-see? I have had many times and heard many stories of where I put something in the trial 1 year, looked at it hard, decided it wasn't good, then 3-4 years later started hearing how everybody grows it and likes it and it's the latest thing. I realized I had goofed up and hadn't given it a good enough look. How did I miss that? Probably because I only looked at it one time in one place under one enviromental condition.

Q: Tom: Could you eliminate 20 of 100 varieties after year 1 and exclude them from your improvement program without having to worry about losing anything? In year 3 would you include all who had scored well in each of the first two years under different conditions?

J: The first step is to go a couple of years, especially if they are really different years. Plant breeders deal with that by trialing in different places at the same time. I did the same beet trial in Oregon, Washington and Wisconsin the same year. The companies wanted wide adaptation. If you're doing it regionally you might double up sites in New England so you don't have to look at it for several years. I'd be real nervous to throw something out after just 1 year looking at it.

CR: In observations, you need to toss out 80 of 100 the first year.

Tom: That's commercial varieties you're comparing to each other and evaluating.

John: It's different for plant improvement trials. But I'm even talking about observation plots even though I understand that under the economic constraints of so much stuff coming at you you've got to have a way to be quick and dirty. There's no right answer to this­whether to look at it twice­but occasionally you're gonna miss one if you don't.

CR: That's right, you swallow your pride, admit it and put it in.
J: I'll show you ways to set up a trial that minimize the chances of having it be in that one spot where it was stressed. Cucumbers get bitter when stressed.

Q: What about physically trying to control the environment so that it is more normal in an unusual year? Or creating a different stress. Nice to have one dry year and one wet y ear. Or putting half under hoop house, half not and irrigating half of each of those plots.

J: That becomes an experment in itself.

George: A trial doesn't have to be large. You can have 3 replication plots under different conditions, and get an idea how it is going to perform. You'd have to define what you are looking for, whether it be disease resistance, drought tolerance, or both. You have a mindset when you go in: I'm going to evaluate these for whatever. You have to take the same notes in the same depth for all the reps.

J: If you're doing to have a trial for disease resistance you need to ensure you're gonna get hammered by the disease. If you have a disease nursery you have to manage the disease as well as the plant. But a lot of times the trial is more like: "I live in Unity, ME and I want to find out how cukes do in Unity, and here it is, one of many environments and I'm going to give them a normal Unity, ME experience rather than doing any heavy tweaking. Oh, I might withhold normal irrigation. I did a trial in Wisconsin where we tried to keep cucumbers from getting rained on at all. But its hard when you start managing weather. If that's your intent, you've gotta figure out a way to do it. Don't try to manipulate too much. Start with the environment we're dealing with, then make an observation, maybe with 3 reps in different soil types. There are a lot of ways to be creative and no one set way to do it.

Q: What about sharing the data?

A: State university agricultural schools often publish trial results for the people of the state. You can call them up and ask for them. Some post results on the internet. Garden City Seeds published some small charts of trial results on carrots and let people see how our carrots matched up with Peto's carrots when I worked there. For any given crop there are three or four experts in the field whom you can get to know. We need more avenues to get this information out. I encourage all of us 1) to be more in contact with people like these folks from Cornell and 2) to publish it in our catalogs, even give some line-by-line trial results. It is good for companies to publish trial results with basic statistics to show the differences and compare with industry standards.

You have to know what the popular commercial varieties are to compare. If everyone in New England is growing Copra Onions, that's the commercial standard.

Tom: More and more at markets the varieties are being sold by name, not just labeled as "onions." The best quality market growers in Vermont have named varieties, outstanding signage and stories on their displays. If you go to any market and see varieties not labeled, ask them, "what is that?" Then you'll usually get more information than you wanted. Knowledgeable customers now want specific heirloom tomatoes.

Amy: If there are 3 or 4 places where you can buy the same named heirloom tomatoes, when in the trials process do you get down to the beauty contest because that is an issue in the tomato business?

J: You can't say, "I don't do hybrids." If you're working with onions you've got to put Copra in because that's the standard by which everything is compared in this part of the country, like it or not. Otherwise, why will people want to buy your OP if Copra works? Copra is a workhorse variety. A friend of mine at Alf Christensen always said "that makes." It means, it always makes a crop no matter what the conditions. If you do a beet trial without Red Ace, you're sticking your head in the sand. Red Ace "makes."

Don't forget the older standards. I used to overlook the old standby Red Core Chantenay, but as I get more into what true quality is, Red Core Chantenay is my favorite carrot. If they've been old standards for a long time there is usually a reason.

Learn the varieties your growers are using. Successful seed company heads in California drive around in their Bronocs from ranch to ranch talking to the farmers and find out what they're growing and why, and know exactly what's doing on.

Keep an open mind. Your views and values will change. What is old will be new and what is new will be old.