Read the current Progressive Forage digital edition

How to use data from forage trials

Chad Hale Published on 30 January 2015

The purpose of this article is to discuss how producers should view forage trial data and what to look for when studying the numbers. Some trials are useful and some aren’t; the question is how to tell the difference.

Silage per ton acre

Table 1 shows a trial report that was given to me this fall. There are yield figures, of course, but more importantly, there is a line at the bottom called the LSD; this is a statistical term called “least significant difference.” If the difference between two varieties is smaller than this number, we can’t really be sure that one variety is better than the other.

In the example table, if you used Entry 5 as your corn hybrid and were satisfied but looking for an improvement, all the entries from Entry 3 down to Entry 7 are the same, mathematically speaking. Entry 5 had a yield of 24.41 tons of silage per acre. That figure, plus or minus the LSD of 3.35 tons, covers the yield differences from Entry 3 to Entry 7.

Every trial is unique, and there are often more to the results than meets the eye. Take the bottom two entries in this trial. They are lagging way behind in yield – why? You can’t know the answer from looking at the information provided.

Since I saw the plots first-hand, I know that those two entries just happened to be planted in low spots in two of the reps, which were under water early in the season and then had severe lodging to boot.

So they aren’t at the bottom necessarily because of poor genetics. When I plant them next year, they might just win, but I can’t recommend them based on this year’s data.

Much of the data producers see is imbedded in advertising. If you see an ad with data and don’t see an LSD number, you can’t really be sure if the different numbers you see are due to one variety really being better than the other. There are some other things to realize about data used in advertising:

1. Advertising space costs money, so oftentimes some data is omitted for space reasons.

2. Advertisers will pick the very best data to place in an advertisement.

If you have access to third-party research from a university or other research institute, there will almost always be an LSD associated with the results.

Another thing that can happen is that the LSD number is bigger than the spread in yields from the top of the trial to the bottom. In that case, you can’t say which variety in the trial has the highest yield potential because they were all too close together.

In my job, I see quite a bit of trial data, and I can tell you that no variety wins every trial every time. Variety development is a competitive business, and most of the varieties entered in a trial are good varieties.

It’s a bit like saying the person who took the bronze in the Olympics is not as good as the person who took the gold. In reality, they are both great athletes.

If you aren’t careful, you can be like a guy I know in the South who lives near a university trial site and plants the winning ryegrass entry each year. He has yet to plant the same variety two years in a row.

So rather than chasing the winning entry all the time, the thing to look for is long-term trends. If a variety is in the top third of a trial year for several years, that is pretty solid performance. If the variety is in the top third of trials in multiple locations for multiple years, that is a real winner.

Data in advertising is viewed with some skepticism, and that is justified for several reasons, a few of which are mentioned above. If you have ever gone to the fair or a trade show, each booth you stop at offers ways to increase your production by 5 or 10 percent.

If you bought everything from every vendor, your beef calves should gain 10 pounds a day, you should have 15-ton hay yields, and cows should milk 200 pounds a day.

Are the vendors lying about their products? No, they aren’t lying – but production can only go up so high, and some products are developed to address a specific problem. If you don’t have the problem, you won’t see the benefits of the product.

Variety trials are the same way. Just because the university plot in your county gets 10-ton alfalfa yields, that is not a promise you will get the same tonnage. A variety that performs best in a swamp probably won’t do well in the desert.

And more importantly, the variety that wins in Vermont probably won’t win in California. One thing that has changed with the Internet age is that you can access trials from all over the world from the comfort of your living room. I run into producers all the time who literally live in a place like Vermont but make variety decisions based on a trial they saw from California.

We have moved from struggling to find data to being overwhelmed with data that might not be appropriate for a specific situation. The task now is to sift through all the information available and pick out the information you should look at.

Talk to your seed dealer and extension agent to find out if there is a forage trial location near you that you can utilize in your decisions. Ask them what data they use when making planting decisions.

And to be clear, there is nothing wrong with looking at a different state for information, as long as the growing conditions would be similar to your location. In general, the testing locations nearest you will be your most meaningful sources of information. While you should be skeptical of data, ignoring data is a huge mistake that can cost a lot of money.

So how can you use the information from forage tests? In my opinion, trial data is a starting point, not an end point. Trial data can help you narrow the choices from hundreds of varieties available down to a handful of choices. To narrow from that handful down to the final winner, there are a couple more steps that need to be taken.

For example, let’s say an alfalfa variety has topped the local trials five years in a row. That is definitely a variety you should look at. But before you plant the whole farm, talk to neighbors to see if anyone else has tried it and what their results have been. It is always cheaper to learn from the neighbor’s mistakes than from your own.

If you planted this new alfalfa on all your acres this spring and it did really well, what have you learned? You learned that planting that variety on whatever open fields you had in the conditions of that specific year was a good decision.

On your operation, the reality is that every field is different and each growing season is different. If you had only heavy-clay fields open for planting this spring and had great success with the new variety, you may not have the same results next year when your sandy field is due to be planted.

I nearly always suggest doing some sort of testing on your own operation. Plant some test strips big enough to be harvested separately and weigh the bales from each strip, or incorporate one new variety per year and try it two or three times on a small scale before adopting it as your variety of choice.

Finding a trial site near you is a good way to get a variety adapted to your location, but that variety may or may not be adapted to your management style. The only way to know that for sure is to plant it.

I realize failed plantings are costly, but never trying to improve is costly, too. Small introductions of new varieties limit the risk of failure but allow you to constantly look for improvements in your variety arsenal.  FG

Chad Hale
  • Chad Hale
  • Research Manager
  • Byron Seeds