Whatever method you choose to get final yield data, the fundamentals remain the same.

1. You should avoid harvesting from ‘damaged’ parts of plots. Damage is unrelated to the treatments applied and tends to vary randomly among plots, so it misrepresents true treatment performance.
2. Edge effects can have a major influence on trial results. Try to harvest along plot centres, although this isn’t always possible when planter and harvester widths are the same. The advent of wider machinery has reduced the impact of this problem somewhat.
3. When calculating yield, accurate assessment of the area harvested is just as important as accurate assessment of the mass of material harvested. Pay careful attention to both.
4. Record all primary data on sheets made for the purpose. This will help to identify missing data and will help to track and correct any mistakes that may occur when making calculations.

Hand harvesting is an alternative that is usually overlooked with on-farm trials. The “On-Farm Trial Guide” gives Tips for Crops.

MORE

Harvesting a trial is much like harvesting a commercial crop. You want to gather all of the crop product without loss and you want to know how much you got.

Addition considerations when harvesting trials are outlined below.

• Ensure that you have accurately located plot boundaries. You don’t want to be including areas that didn’t receive any treatment in your plot harvest. Nor do you want to be mixing treatments in the one harvest.
• Ensure that you do not include parts of the plot that have non-treatment effects in them. These might include areas that have been trampled by rampant stock, have not been established properly, have suffered from flooding or have copped some spray drift.
• Exercise careful judgement when excluding parts of a plot from harvest. You don’t want to exclude them just because they don’t look any good. You need to assess whether the ‘problem zone’ is related to the treatment. If the problem zone is related to the treatment in any way – either directly or indirectly – then it should probably remain in the harvest area. The reasoning here is that if the problem is related to the treatment it could easily recur if conditions allow it. You need to know this before adopting the practice in your commercial crops.

Your records of in-season observations and measurements will prove invaluable in helping you to sort out whether trends or events that you see in the plots at maturity are likely to be related systematically to the treatments.

Ensure that you have accurately determined the harvest area. Yield is always expressed on a mass per unit area basis – tonnes per hectare, for example. If your assessment of harvest area isn’t accurate then no amount of precision in measuring the mass of harvested material will give you an accurate yield estimate. Use a tape measure for this purpose. Nothing else will give as reliable a result.

Try to minimise edge effects. You’ll already have given thought to this when designing your trial. Basically, you want to minimise the harvesting of plot edges because they usually differ from the plot centres.

Ensure that measuring equipment has been calibrated – recently and reliably. You can never assume that a measuring device works as intended. No matter how expensive or high-tech, it needs to be calibrated and checked against standards.

Data Recording

Good record keeping has several advantages. It will enable you to more fully and reliably interpret the data. This can be important when new information comes to light at some future occasion. Good records enable you to share results and can provide the basis for discussion with people with specialist skills or experience. Most importantly, good records provide the basis for you to repeat your wins and avoid repeating your losses.

The primary requirement of a data sheet is that it enables you to record:

• primary data that your require (eg. mass, number)
• appropriate units of measurement (eg. kg, number)
• the area from which data were obtained (eg. 1 m2, or 3 plants from each of 2 rows)
• the unique plot number or name from which the data were obtained
• the date that data were obtained

It’s almost certain that you’ll make some mistakes when making or recording measurements. You can most easily fix these errors if you record as much primary information as possible.

For example, if you suspect that your plot area measurement doesn’t seem right you can spot the error most easily if you’ve entered both the length and width on the data sheet, rather than just the calculated area. By the same token, you’ll most easily spot a mistake in yield if you have each of the harvest mass, moisture content and corrected harvest mass (at a standard moisture %) entered separately.

A well-designed data sheet will enable to most readily check that you have filled all of the necessary data ‘holes’.

Some of your data will come from subsamples of harvested material. It is important that subsamples from each plot are kept separate. If they are pooled you will lose most of the benefits that you gained from implementing a good trial design.

End of section critical decision point

Have you harvested the trial from a part of the plot with known dimensions? Was the harvest area free of unrepresentative ‘problem zones’? Did you measure harvested masses with calibrated equipment? Did you record the results on a plot-by-plot basis in a form that will enable you to retrieve them for later analysis? If not, can you be confident that you’ve got accurate, reliable and retrievable data to go on with? If, on the other hand, it all looks OK, then move on to the next section – “After the trial”.