How To Wiki


Yield Trial Work Outline


Sample Promotion Page
Comments on sample promotion page

Design of Experiments

Generic Seeds Yield Trial Seed Preparation Work

In-house Experimental Design

Creating a Yield Trial Seedlist

Seedlist Format

In-house seed prep and counting seed

Seed Packets
Applying fungicides and Insecticides to seed

Preliminary Work for Commercial Yield Trials

Working With Pre-Designed Yield Trials

Field Design and Filling Planting Trays

Planting Plan
Laying out and filling the planting trays
Planting Tray Checksheet

Measuring and marking ranges

Amount Due Calculation Spreadsheet Sample


Plastic trays
Making a straight first row
Checking seed depth and location
Sound off
Comment on speed vs. growth


Days from Planting chart
Table of Contents page
Sample Table of Contents page
Planting Plans
New Designations and Abbreviations
Sample Planting Plans page
Sample New Designations and Abbreviations page
Format and Creation of Yield Trial Notebook page
Single Replicate Format
Single Replicate Format Example (as in computer
Single Replicate Format Example (as printed)
Three Replicates per Page Format Example (as printed)
In-House Format Example


Staking the Plots



Getting the harvest and observation data together

Yield Calculations


Harvest Indices

Yield Trial Report

Yield Trial Work Expanded Outline

Preliminary work
Receipt of Seed from Commercial Clients
Test Design
Seed Preparation
Planting Plan
Planter Checksheet (if needed)
Fill Planter Trays
Measuring the Field
Seedbed Preparation
Measuring and staking ranges
Marking Ranges
Planting Plan(s)
Cutting out ranges and thinning

Notebook Creation

Early notes - (Emergence if desired) , Vigour
Herbicide Application
Notetaking at Pollination
Days to 50% Pollen Shed (Tassel)
Days to 50% Siliking (Silk)
Notes taken prior to Harvest
Stalk Breakage below the ear
Root Lodging
Stalk Lodging
Population/Stand count

Harvest with Combine

Harvest tags, Harvest Data Sheets, Moisture testing

Fall field work Data Entry Yield Trial Reports

Commercial Clients
Cover Letter
Results and Averages
Statistical Analysis
Results and Averages
Statistical Analysis
Harvest indices and Sorts


Note: This material was written before it was known that there is company called "Generic Seeds" in the U.S.A. The new fictional corporate name for this material is "Big Yellow Seeds".

Each year Big Yellow Seeds sent out a fax to prospective clients stating the Yield Trial and Nursery services being offered and the prices per row or plot. This was sent some time from late February to the end of March. This was usually accompanied by shipping instructions for sending seed to Big Yellow Seeds from various areas.

A list of current clients and prospective clients should be prepared, kept and updated each year. Obtain prospective clients from the breeder and the various seed grower and other company directories available.

The sample promotional fax below gives a general outline. There was only slight variation in the basic outline. In 1995 for example, the fact that 30 acres of systematically tile drained land was available at one site was emphasized because tile is a rarity in the local region. The row width changed from 30 in. to 36 in. one year due to the loss of the use of a 30 in. plot planter and problems changing a used planter to 30 in. The plot length has changed at times too, as well as the prices. Ideally, you don’t want to be changing these specifications so much.

There used to be an INSPECTED PRIVATE YIELD TRIAL that was an officially recognized Yield Trial that was inspected by a member of the provincial Corn Committee who had a degree in Agronomy. The member would state that the Yield Trial was as specified by the contractor and passed a set of guidelines. This government group no longer exists as the government has gotten out of research in favour of promoting privately funded research.Such an inspection agency was useful as a neutral referee.

Big Yellow Seeds received material from CANADIAN COMPANIES via courier (Purolator,UPS etc.) at:

Big Yellow Seeds
201 Anywhere St
Anytown, Anyprovince

If the material was shipped by bus it came into a local gas station which phoned to notify Big Yellow Seeds that it had a package for the company to pick up.

U.S. COMPANY (via courier or parcel post)
Send to: Big Bert Seeds
c/o Convenient U.S. delivery place
1214 Anystreet
Closest border town, Closest state

Get the COMPANY NAME and the DATE the SHIPMENT was SENT.

The shipments to the U.S. location were picked up in person by a company representative.

Note that by 2001, these instructions for shipping and the method of promotion may have changed, so find out what the current procedures are.

Seed Solutions is a New Zealand company that offers Winter Nursery space to Northern Hemisphere researchers. It has a nice website promoting its services [1].

MAS Plant Breeding Services in Chile also has an excellent promotional website.[2]

Hawaii Seed Research Services, Inc. (Winter Nursery) [3]

As a good example of current practice, an application form for the LOUISIANA AGRICULTURAL EXPERIMENT STATION 2009 LOUISIANA CORN HYBRID PERFORMANCE TRIAL can be found here.[]



Big Bert Seeds will be contracting corn yield trials for the 1998 crop season. With this in mind, I wish to draw your attention to the following:

  • The excellent location with its [specific] daylength and [specific] CHU's provides an excellent selecting area.
  • I will have an official inspected private yield trial, a separate company preliminary trial and a strip trial.
  • I can offer several location sites with different soil types.
  • I will supply complete data for vigour, silking, lodging, stalk breakage, yield and moisture followed with a complete analysis at season end.
  • Plot Size: 2 rows, 76 cm wide by 6 m long***

45-50 planted seeds per row {C}rows thinned to 30 plants

  • Seed to be sent in packets of 45-50 seeds.


  • One Entry replicated 3 times ........................................$75.00 per Entry
  • One entry replicated 2 times .........................................$60.00 per Entry

A second location is available upon request.

  • Three testing sites are available. They are Location1, Location2, and Location3.
  • Isolation, Observation and Nursery space is available by negotiation.
Joe A. MaizeManager

Shipping Instructions

Canadian Company
Send to: Big Bert Seeds
Anytown, Anyprovince
U.S. COMPANY (via courier or parcel post)
Send to: Big Bert Seeds
201 Anywhere St
c/o Convenient U.S. delivery place
1214 Anystreet
Closest border town, Closest state

{C}Comments regarding the sample promotion page {C}The row width and plot length have varied somewhat due to equipment restrictions.

The row width and length shown is for a 1/1000 hectare plot Since then, the rows have been reduced to around 18 ft. A row width of 30 in. is preferred with 22 in. of serious interest for future use. A row width of 36 in. was used one year because the used planter that was purchased couldn't be modified to 30 in. rows. It was all that was available so it was used. The different plot widths and lengths make comparison of results over several years a bit difficult and should be avoided.

The promotional fax is very plain Jane and I was advised not to bother jazzing it up. I was told that Plant Breeders are not interested in flashy promotional material.

The line regarding prices for less than 3 replicates sometimes changed so that it was more expensive per plot to do a test with less than 3 replicates.

The contract work of conducting Yield Trials for other companies was a way to make money using the equipment and years of expertise that were available to the company. Companies with good funding might not bother offering such services to other companies. On the other hand, there might be a way to make such contract work a viable business.

You will note that Nursery and Crossing block space was offered but the only people interested in that service was the USDA Germplasm Enhancement Maize program (GEM) and that was on a volunteer basis. I should add however that seed production fields of a new type of corn developed by a local University/Government research station before it closed, were grown for the company that licensed it.

Design of Experiments

Yield Trials are experiments used to collect information about new hybrid varieties with an emphasis on the yield, usually as compared to the best commercially available variety or varieties grown in the local area. Some of the plots grown in a Yield Trial field might be referred to as Observation plots whose yield may not be a primary concern. There can also be Yield Trials in which comparison with a commercially available variety is not involved because one is searching for information on heterotic patterns and yields in crosses between families of inbreds or in specific inbred crosses. Generic Seeds didn’t conduct research beyond testing hybrids but in the future there will be a need for more sophisticated research and experiments. Some of the research will involve things like disease resistance, nutritional content of the seed and plant and the physiological performance of plants.

Considerable thought needs to go into the design of Yield Trials because there is a lot of expense involved in conducting them and the very large number of possible combinations of even just a dozen inbreds means it’s very easy to spin one’s wheels going nowhere on a trial and error basis searching for a hybrid with the desired performance characteristics. According to R.W. Jugenheimer, “Pioneer tested 9,800 new hybrids in 1974 plus 1,000 ‘retest’ hybrids.” As well, “20,000 performance plots were harvested” in 1974. That was 30 years ago, so today’s numbers might be even larger. The sooner the company gets someone on staff with training in experimental design and analysis, the better. This doesn’t have to be someone with a Ph.D or even an MSc. It might be difficult to justify hiring a full-time statistician until the company is testing thousands of varieties but such a specialist might be hired as a consultant on a short term basis. At the very least, one can consult the plant breeders that have been acting as mentors for the company. My personal preference is for a Bayesian viewpoint as discussed in the posthumously published PROBABILITY THEORY: THE LOGIC OF SCIENCE by Professor Edwin T. Jaynes.

Read this section first and use the following website information as reference material and an introduction to the subject. These website notes are:

  • 1) Professor Jennifer Kling's lectures on Designing Experiments
  • 2) EDGAR: Experimental Design Generator and Randomiser,
  • 3) Guide to Field Experimentation in Agriculture - experimental design,

My personal experience is that mere references to the websites won’t result in people reading them, so I prefer to have the text immediately to hand, either at the point of reference or included as an appendix. You can locate other websites devoted to the design of experiments or statistics on the web as you wish.

One of the things I noticed regarding experimental design that you will encounter online and in books at Universities is that the information is often directed at breeders of small grains such as wheat or barley rather than maize breeders. There are, as a result, a few differences in the illustrations of Big Yellow Seeds Yield Trials and the illustrations given for wheat yield trials. In Figure 4 (Randomized complete block design) of the Guide to Field Experimentation, the replicates are shown beside each other like columns in a spreadsheet while the Big Yellow Seeds design places the replicates behind one another like spreadsheet rows. The designs in textbooks on Experimental Design are also usually far more complicated than the simple design used by Big Yellow Seeds.

Try to keep the size of a Yield Trial Field to a day’s harvest work. This is about 350-400 plots. According to R.W. Jugenheimer, in the 1970s, a 3 man crew and a self -propelled picker sheller could do 350 2-row plots in 8 hrs. In the early 1990s somewhat more than this was done at Big Yellow Seeds by one person with a plot combine and automated data collection system. In the mid to late 1990’s, when the automated data collection system wasn’t working, 2 people in the plot combine did about 350 plots in about 8 hours with a third person doing a second moisture test on samples after each round.

For an idea of a maximum figure, the largest physical fields in the local area are 1 mile by 1 mile (640 ac.) and called Sections. Each Section would hold somewhere between 220 and 250 ranges (depending on the plot length and the path widths between ranges and the number of headland rows) and around 1000 to 1050 plots (also depending on number and size of pathways and guard rows in the other direction). That’s an extreme figure of about 248, 000 plots per Section. In practice, you would plant multiple fields of a standard size with borders of 2 to 4 guard rows separated by paths large enough to drive a combine and other vehicles on. The large paths would be seeded to a small grain like wheat and mowed with a field mower. Within each field the paths between ranges would be small (3 ft.) and bare or kept bare by hoeing out weeds.

You may see references to “exchange” yield trials. These are yield trials in which 2 companies in different locations exchange varieties to be tested and each company tests the other company’s material at no charge as if the material was one’s own. This can be a good way to get a variety grown in many different locations and possibly find a variety to sell outside the local area. However, with private funding, researchers are being told to focus on developing varieties exclusively for their local area, so an exchange trial would be considered a wasteful expense. This might be an exploitable weakness if one had franchise type operations in different areas or was part of some type of association of breeding companies in different areas and could transfer seed around rather than equipment. One of the inbreds that made it possible to grow maize in Southern Ontario, Canada was an inbred from the Canadian Government Research station in Morden, Manitoba.

Big Yellow Seeds used an experimental design known as Randomized Complete Block for its yield trials (See the Experimental Design websites ). You may hear or see references to another design known as a lattice. Randomized Complete Block (RCB) is a simple design in which each test consists of a number of ranges equal to the number of desired replicates. That is, the plots of each replicate are contained in one range. The width of each test in terms of plots is equal to the number of varieties being tested. Each variety gets one plot in a replicate. Each replicate contains plots of the same varieties but in different orders.

One of the first things to be decided is the number of varieties that there will be in a test. This also determines the width of the test since each variety has one plot in the first range. You want to keep the test compact to ensure somewhat that the soil type or topography of each plot isn’t different. Something like 50 2-row plots wide may be too wide. If you have more than 25 varieties to be tested, consider mutiple tests within a Yield Trial Field and multiple Yield Trial Fields of standard dimensions such as a serpentine pattern of 25 plots wide by 24 ranges long (if space permits). A pattern that was used in the early 1990s involved a 24 plot width perhaps to better facilitate using 4-row electric indexing trays and a four -row planter. This pattern made the Yield Trial Fields more modular and separated planting the Yield Trial Fields from planting guard rows between each field (There are some peculiarities of this pattern that I will discuss later).

For the seedlist example in these notes I’m going to use a single Yield Trial field pattern that is 25 plots wide and has 4 guard rows of a commercially available variety on each side planted straight through. You could probably get away with only 2 guard rows on each side.

Now, after saying that, I have to add that putting the tests received from other companies into this field may mess things up a bit at least for one Yield Trial Field. For instance, a company might send you a test for 30 varieties. Now your field should be 30 plots wide and any test to be added in that field with less than 30 varieties will need to have filler plots added to fill the extra width. If one company wants to test 60 varieties, another company wants to test 15 varieties and another wants to test 24 varieties, things get complicated.

The second thing to be decided is the number of replicates to grow for each test. Statisticians recommend a minimum of 3 replicates but 2 replicates are considered adequate for their purposes by many plant breeders. I’ve seen plans for tests with 6 replicates. Both PERFORMANCE OF CORN HYBRIDS IN LOUISIANA, 1999 and AGRONOMIC AND TEST INFORMATION: GRANGER (see appended online pages) used 4 replicates.

The third thing to consider is the number of locations the test is to be grown in.

By 1999, some of the commercial companies were not interested in paying for more than 2 replicates and were often content with only one or two locations. In-house yield trials became merely observation plots with randomization often abandoned, only two replicates grown of each variety and ANOVA done only for the paying commercial trials.

Big Yellow Seeds Yield Trial Selection and Seed

Preparation Work

Parcels of seed packets will be received from other companies while you are preparing seed for your own tests. For information on how to handle these parcels and their contents, you can skip to PRELIMINARY WORK FOR COMMERCIAL YIELD TRIALS but there might be information in this next section that you need to know to understand the instructions in that later material.

I’m going to discuss in-house test design, yield trial seedlist creation and seed preparation work before discussion of what to do with parcels sent by other companies for their private yield trials because you need to know how to design and prepare your own tests from scratch. Most of the companies that send tests do all of the test design and seed preparation work for you so all you have to do is get the seed planted. However, they might just send bulk bags of seed of the varieties to be tested and a few notes on the design of the test such as the number of locations and number of replicates.

Experimental Design

Random Complete Block

{C}In a random complete block design, the test is as many plots wide as there are entries and as many ranges long as there are replicates. It is designed as a block with 3 foot pathways between replicates. The entries have one plot in each range/replicate but the position of the entry's plot in each range/replicate is determined randomly. A test with 3 or more replicates is best for accurate information. RCB design is used primarily for commercial yield trials and any material that looks especially promising.This design, however, is often abandoned for a simpler unrandomized design for Generic Seeds's material. {C}The Seedlists for each test are distinguished by the number of replicates and locations. For in-house material, Big Yellow Seeds tests involving multiple replicates and locations should be listed first in the Big Yellow Seeds Yield Trial Seedlist. Single plots are usually for observation purposes and listed last. Note that the Yield Trials involve two - row plots. All information on the seedlist is given in terms of plots not rows.

The Yield Trial Seedlists are printed in portrait size (8.5 x 11) but the format for Yield Trial Seedlists is completely different from the other seedlists.

The basic difference is that the plot numbers for an entry are listed beside the entry and its entry number.

Once you know the number of varieties there are to be in each test and the number of replicates for the test, you can create the randomization of the plot numbers for the test.This is an assignment of one plot number per replicate to each entry number.

Yield Trial Seed Selection

The breeder will need Inventories of hybrid seed produced the previous year in Crossing Blocks as well as Inventories of hybrids to be retested and other hybrid seed from other years. A list of currently available commercial hybrids, their traits and the opinions of growers regarding these hybrids may be needed to select possible standards to use as checks to compare against.

The breeder will have to decide what the entries to be tested will be, the design of the tests, in which tests the entries will be placed, how many replicates there will be in tests and the number of locations for each test.


First, create a list of all of the Big Yellow Seeds entries that will be worked with as per the breeder's selections and don't assign plot#'s until after the list of entries has been finalized. This list will contain as much information on an entry as the company has such as a Source Plot number or Source Row# which might also be referred to as a hybrid number, a pedigree, and possibly a family designation. Some varieties may have been given new designations so you need to have the new designation and pedigree information together in the list or in a separate reference list so people can find the proper seed based on the Seedlist information.

Group the entries according to the number of replicates to be planted for each entry. Start with entries in multiple replicate tests and end with entries in single replicate tests. Note that it may be necessary to move some material from a 3 replicate test to a two replicate test or from a two replicate test to one plot if there is not enough seed.

Number the entries consecutively in an Entry # column. You can use ‘#’ as the column heading rather than the wider ‘ENTRY #’ heading.

All of the plot #'s for the different replicates in each location are placed to one side of the entry# and pedigree of the entry. The plot#'s can be either on the extreme left or right but extreme left is preferred. It's best not to use plot #'s reserved for other maize fields such as 1-2999 (Inbreeding Nursery) or 7001-8999 (Sib Nursery). Plot#'s should also distinguish location. E.g. 19001-19499 for Town1, 30001 - 30500 for Town2 etc.

For each entry, there will be a plot# for each replicate listed beside the entry# and entry pedigree. Plot#'s for replicates of all locations of tests using the seed of the entry will be included so all the needed packets of seed of a variety are filled at one sitting.

E.g. A 25 entry, 2 replicate test for two locations, in Town1 and Town2. Assigned Town1 plot#'s are from 20,000-25,000. Assigned Town2 plot#'s are from 30,000 to 40,000.

1993 Yield Trial
Town1 Town2
2001 2026 30001 30026 1 MX1
2010 2049 30010 30049 2 P3979

Note that the same randomization scheme was used for the two locations. This can be used to make the creation of the seedlists simpler by adding digits to the front of plot#'s. E.g. All plot #'s plus 20000 or add 200 to the left of the digits.

Note also that a single name pedigree was used in this example rather than listing the pedigrees of the parents. This is simpler to look at and of shorter width but requires some form of index regarding the new names so the correct seed can be found by the person preparing the seed. It is best to list the female source row number and pedigree and the pedigree of the male parent in that order so the seed can be found.

A format that appears in some of the older notebooks lists a female pedigree in the first column followed by a CROSS column for the male and a THREE WAY column for a third parent if the cross was a three way cross.

When it is possible, all of the plots of a replicate should be in the same range. Check varieties are usually the last 2 or 3 entries of a test. Seed of some check varieties might not be available until shortly before planting.

A preferred number of plots per replicate is 25. This would be a 25 entry test. This fits very nicely on a landscape size page for the notebook but has not been done often because of differences in the number of entries for other tests in the same field and the large number of Big Yellow Seeds entries. Plot #'s for the first replicate are not randomized but plots for additional replicates usually will be randomized.

Often the trials for Big Yellow Seeds hybrids have not been randomized nor replicates kept within single ranges. The plot #'s were assigned to the entries consecutively but the entries with multiple replicates were listed first so that 2 replicate material was followed by observation plots. Listing of the pedigrees has also caused some space problems using portrait size paper for the Seedlist and even for landscape size pages for the notebook. Also, finding seed that has been given a new designation can cause problems. An index can be created to show the information needed but the breeder would like to see the pedigrees rather than the new designation. If used, a copy of the index should be included as one of the first few pages in the notebook for easy referencing by the breeder.


Each range in a test should have the plots of a separate replicate. If there are 3 replicates there should be 3 ranges. The number of plots in a range equals the number of entries. If there are 25 entries, then there will be 25 plots in each range for each replicate.

The first replicate of a trial is not randomized by Big Yellow Seeds. This is so that if there is a mistake in the planting, inspection of the seed in the first range can be helpful. Also, the sequence 1, 2, 3, 4 is just as random a sequence as 2, 4, 3, 1 or any of the other permutations of those 4 numbers.

Other replicates are randomized by rearranging the order of entry #'s for the replicate. Remember, however, that the Seedlist sort is by Entry # so the plot #'s shown in the column for the other replicates will be scrambled rather than the entry numbers.

MSTAT was used for creating randomizations.The randomization schemes were checked. The breeder did not like an entry to appear behind itself in another replicate. If this happened, the simplest way to fix it was to exchange the offending entry number in the other replicate with another entry number in another plot in the same range (or replicate).

A simple method to produce randomizations is to put the numbers of the entries or of the plots on pieces of paper and draw the numbers from a hat. If entry numbers are drawn, the first entry number drawn is placed in the first plot of that replicate. The next entry number drawn is placed in the second plot of that replicate.

If plot numbers are drawn, then the first plot number drawn corresponds to the first entry number in the list and the second plot# drawn is assigned to the second entry number on the list.

A user friendly program should be found or created to replace the MSTAT Experiment Plot Randomization program. Big Yellow Seeds may now have enough money to purchase software such as the AGRONOMIX AGROBASE II plant breeding software.

Your goal is to create a Yield Trial Seedlist similar in format to the sample below. With proper software, the creation of this seedlist requires only the number of varieties to be tested (including check varieties) and the number of replicates. It isn’t finished, however, until the ENTRY NAME column on the extreme right is filled so people know where to get the seed from. The creation of such a seedlist was a complicated cut and paste operation using MSTAT and VP-Planner.

Yield Trial Seedlist
5001 5047 5067 1 unassigned
5002 5042 5051 2 unassigned
5003 5037 5054 3 unassigned
5004 5041 5057 4 unassigned
5005 5032 5060 5 unassigned
5006 5026 5065 6 unassigned
5007 5049 5072 7 unassigned
5008 5048 5062 8 unassigned
5009 5043 5073 9 unassigned
5010 5046 5058 10 unassigned
5011 5027 5070 11 unassigned
5012 5050 5059 12 unassigned
5013 5044 5071 13 unassigned
5014 5034 5053 14 unassigned
5015 5035 5064 15 unassigned
5016 5040 5068 16 unassigned
5017 5045 5052 17 unassigned
5018 5039 5056 18 unassigned
5019 5038 5074 19 unassigned
5020 5033 5061 20 unassigned
5021 5031 5055 21 unassigned
5022 5030 5069 22 unassigned
5023 5029 5066 23 unassigned
5024 5028 5063 24 unassigned
5025 5036 5075 25 unassigned

Yield Trial (Test) Seedlist Format

Every YIELD TRIAL SEEDLIST should have a title stating that it is a Yield Trial Seedlist and possibly also other identifiers such as a Field # and Test # if such a system is in use or a company name if the trial is being done for another company. The second line shows the location or locations where the test is being planted.This is usually the name of the town or village closest to the test. I’ve shown only one town here as CLOSESTTOWN but there can be more.

The next line contains column headings. On the extreme right is the ENTRY NAME column. I’ve put unassigned here because you can make up the randomization scheme using only the number of varieties to be tested and the number of replicates. The ENTRY NAME is identification information that can be used to obtain the proper seed to put into the proper planting envelopes. The ENTRY NAME might be a name like HELIX or P3979, or it might be a Hybrid number, pedigree or a company name and # such as CIBA#1. This information is usually not printed in the notebook so it is not available to the people gathering information on the varieties during the test.

The column to the left of the ENTRY NAME column is the ENTRY NUMBER column. Each entry number is assigned to a specific variety in the test.

The 3 Replicate columns (REP1, REP2, REP3) contain plot numbers. The seedlist is not a map of the field but an assignment of ENTRY NUMBERS to a specific plot in each range. The first replicate contains plots 5001 - 5025, the second repicate contains plots 5026 to 5050 and the third replicate contains plots 5051-5075. For a two-row plot, there will be two envelopes to be filled with seed for each plot #. In the sample 3 replicate seedlist, a total of six envelopes is needed for each variety in the test. For Entry #1, there will be two envelopes marked as plot 5001, two envelopes marked as 5047 and two envelopes marked as 5067.

It’s a company policy not to randomize the assignment of entry numbers to plot numbers in the first replicate. The plot numbers of the second replicate of this sample seedlist were randomly assigned to an entry number using a basic random shuffling algorithm. The same was done for the third replicate. Remember that the seedlist is not a map of the field. The plot numbers in the field are sequential and arranged in a serpentine pattern. You can also randomly assign an entry # to a plot number in the field and then sort by entry # to get the same result. How you get such a list of plot numbers is up to you. I used MSTAT to get the randomizations but it involved a lot of cut and paste to create the seedlist as compared to the little BASIC program I made recently to create it all from just the number of varieties and the number of replicates. You could use Python or VBA too. You could also put the plot numbers for a replicate in a bag and draw them out one at a time assigning each one to the next entry number.

In my sample seedlist the plot numbers run from 5001 to 5075. I added a feature to the little BASIC program I made to set the first plot number to begin from. You could use 1 to 75 or 1001 to 1075. For the company, the first plot should always end with a 1. You might want to use 1 to 75 because you can then use the same test pattern for all the tests if they are the same size and just add prefixes like 500 to the plot numbers to get 501 to 575. In a spreadsheet this could be done with an addition formula. The plot numbers in each test and location should be unique.

If there are more locations, the name of additional locations is placed on the line with the first location name between the last Replicate of the first location and the Entry number column. The plot numbers for the next location are placed below in the same way as for the first location. In fact you could copy the plot numbers for the first location, add say, 6000 to each plot number so they were unique and paste them below the second location name. The idea is to get all of the envelopes to be filled with seed of a variety on one line and on one piece of paper so it can all be done in one sitting.

Seed Preparation

Seed preparation for each test is done according to the Seedlist for that specific test. The person preparing seed should have or create a set of envelopes with duplicates of each plot# required for the test according to the number of rows in a plot. This means that for two-row plots there are 2 envelopes marked with the same plot number for all the required plot numbers. The envelopes are #2 coin envelopes that fit into the small white trays. Old envelopes with already stamped numbers can be used but make very sure there are no seeds in the envelopes from the previous year. The plot numbers can be stamped on new blank envelopes if there are no old envelope sets with the required plot numbers so a stamper that can be set to change numbers after two printings, and ink may be needed.

The storage tray and envelope of the listed hybrid source row and pedigree will be obtained from the proper crossing block. Ensure that the source row and pedigree match the seedlist.

Once the envelopes are numbered, one will need a long flat ruler (foot, yard or meter) a large tray of 10 empty small white trays, and the first large box from the storage room that contains the required seed . Don't remove just the envelope of seed needed but take the entire large tray of storage envelopes to the work bench. A small balance will also be needed to help count seeds. If a balance is not available the counting can be done by hand using a pointer such as a pencil and corners of cardboard shoe box lids (See the section MANUAL (HAND) COUNTING below).

The person preparing the seed will first place the ruler on the seedlist beneath the first record which corresponds to the first entry. The ruler is to help with moving the eyes straight across a record horizontally. One could also use a blank piece of paper positioned sideways (landscape) below each line. Get the marked envelopes of each of the required plot#’s for each replicate in each location. Look in the Entry column at the source row or source plot number or hybrid number or name and find the storage envelopes with that source information in the storage trays. Look at the pedigree required and select the same pedigree from the storage envelopes of the same source row #. The pedigree on the storage envelope should match the pedigree or designation on the seedlist.

Fill each of the envelopes for the entry with 50 seeds per envelope using either a balance or manual counting.

Manual (Hand) Counting

You need 2 corners cut from the top of a cardboard shoe box. They should have two sides that meet at a corner. Pour an amount of seed from the appropriate storage envelope into one corner piece. Slide the other corner piece under the first so the back edge is twice as long as for one piece and the open side is closest to you. Use a pointer of some kind, like a sharp pencil, to separate 2 or three seeds from the others. Move these seeds using the pencil in a horizontal sweeping motion to the other corner,counting by 2's or threes. Continue transferring seed from the one side to the other until you have 50 seeds in the other corner. Use the sides of the corner piece to help pour the 50 seeds into one of the empty envelopes. Repeat for the second envelope and then put any remnant seed back in the storage envelope. Put the storage envelope away and then put the newly filled envelopes in the small white trays.

Using a Balance to Count

The balance is for weights of less than 1 kg and will have a unit counting feature. See the instructions for use of the balance regarding how to proceed. Use a small pan rather than a large one. Remember to always tare out the pan's weight so you can work just with the weight of seed or unit count.

First, prepare the balance to take an average seed weight. This can be based on 5,10 or 20 seeds depending on the accuracy desired. Use the 10 seed average. If using the 10 seed average, count out 10 seeds into the pan and begin according to the manufacturer’s documentation for use of the balance as a unit counter. The unit count should say 10 and adding one or two more should increase it by one or two units. Add more seed until you have 50 units, pour the 50 units into one of the empty envelopes. Pour 50 more units from the storage envelope into the pan and then pour these 50 units into the second empty envelope. Put the 2 newly filled envelopes into the small white tray and return the storage envelope to its proper place in the storage tray.

Since seed of different varieties will have different sizes and weights, the unit count procedure will have to be done for each new variety of seed. This means everytime the pedigree changes.

There are electronic seed counters that can be bought with multiple spouts for filling more than one envelope at a time.

Rough and Ready Count

It’s possible to measure out seed quantities by sight only, but this practice is frowned on because it isn’t very accurate. You should check with the breeder before using it. It might be o.k. to use this for Big Yellow Seeds Yield Trials since the plots are overplanted and thinned. However, if you don’t put enough seed in the envelope it becomes a major problem.

50 seeds can be placed in the palm to get an idea of the amount and then a similar amount is poured into the palm of one hand each time seed is needed. The seed is then poured from the palm into the envelope with the correct plot number. {C}Use any way of counting that works best for you. Be very careful not to drop seed. Seed that falls to the floor must stay there and be put into the garbage when the area is swept. This is to ensure that only correct seed is placed in envelopes for planting.

Sometimes it happens that there is not enough seed available for the number of plots planned according to the seedlist. If this happens, the person preparing the seed should notify the breeder. The breeder may simply abandon planting that seed and replace it with another variety, a check variety or a filler variety just so plot#'s don't have to be changed.

Any changes made to the seedlist should be indicated legibly in pen and updated to the computer after the Seedlist preparation is finished. Any changes in pen on the seedlist are senior to any changes recorded elsewhere. If the change is not recorded on the seedlist it was never made.

Place the newly filled envelopes with their flaps up, at the front of the first small white tray, with a bottle or some other object behind them to keep them erect. Put the storage envelope back in its proper place in the storage tray.

As much as possible, develop a routine habit of working that avoids confusion and keeps the work moving along smoothly. Always complete the task at hand and start fresh before beginning the next task. E.g. Put back the storage envelope used before getting a new storage envelope or put the storage envelope away and the newly filled envelopes in the white tray before starting with the next Entry. To make things easier you might want to put envelopes for different locations in a different white tray.

After all of the seed including the seed of check varieties has been put into envelopes, the envelopes should be either put into plot# order and checked that they are in plot# order.

The next step will be to plan the field design and place seed or seed packets into the planting trays.

Modern Seed Counting and Packet Filling

Seed Research Equipment Solutions,LLC ([4])

Seed packets

Big Yellow Seeds worked with seed packets that were #2 coin envelopes with the plot#’s stamped on one corner. Other companies send seed packets of various sizes and colours. The nicest seed packets to work with are about the same size as #2 coin envelopes and have a perforated top that is easy to rip off. The Information regarding the variety, entry#, and plot# is printed on the smaller top section of the envelope. Sometimes it seems the information should be put on the bottom section so that after the seed is poured one has the bottom sections in hand with the information and doesn’t have to find where one stashed the last two tops ripped off.


Standard preparation of seed by many companies includes the application of a fungicide/insecticide. Big Yellow Seeds does not do this. The reason is to help breed better material.(Actually, it was more like trying to avoid the hassle of getting the chemicals onto the small batches of seeds). [One year the bean nursery was practically wiped out by an infestation of cut worms]. One year, a kind of salt shaker was used to sprinkle powder on seeds in the plastic planting cells.

One method used to apply the fungicide was to create a kind of cement mixer with a 5 gallon pail attached at an angle to a rotating device. A liquid solution of the fungicide was prepared and put into a spray bottle which was then used to spray the seed of an entry as it rotated and mixed in the pail. A bulk of seed for an entry was treated, not just 100 seeds at a time.

Another possibility is to use a rock polisher. The batch of seed of a single cultivar/variety and the fungicide in powdered or liquid form are put in a short cylindrical rock polishing container with a sealable lid and mixed by rotating the container as one would to polish rocks. The container sits on two thin, cylindrical rods (one on each side of the container with the top of the container facing forward) with one rod driven by a motor to rotate and the other rod free to rotate in any direction.

Preliminary Work for Commercial Yield Trials

Receipt of Seed

1. Record in a small notebook the following for each Yield Trial:

  • Company name (E.g. Ioweah Seeds Co.)
  • Number of Entries (E.g. 60 entries)
  • Number of Replicates (E.g. 2)
  • Number of locations (E.g. 2 locations, Location1 and Location2)

Update this information right up until the day seed is received.Keep the information current to after the seed is planted.

Use this information to compute the amount due and to work out the size of the entire Yield Trial Field for all of the companies.

2. When the seed is received, check the envelopes in the boxes to confirm all of the above. Record all the data available on the seed envelopes such as Plot number, Pedigree, Randomization scheme etc. and check it with the info given by the company. Record to the computer any information regarding the trial that has been received on paper. Notify the company if there are any discrepancies. If you have a scanner, scan any and all paperwork received regarding the test.

3. If a test design is needed, create the randomization scheme and seedlist and get the proper seed in the proper marked envelopes and put them in plot number order.

If the test design sent doesn’t use a serpentine pattern of plot numbering or the first replicate of the test is not randomized see the section below entitled, WORKING WITH PREDESIGNED YIELD TRIALS:PROBLEMS

If you revise a test design for which seed packets for each row were sent, write the new plot number on each envelope with a Black Sharpie marker according to your revised test design and seedlist.

4. Put the received envelopes in proper planting order. Sometimes the seed packets are sent grouped by entry so you want to make sure you get the seed packets arranged by plot number.

There can be some problems with certain test designs so see the section below entitled, WORKING WITH PREDESIGNED YIELD TRIALS:PROBLEMS.

Working With Pre-Designed Yield Trials


  • Received Test design is not serpentine
  • Plot numbering conflict
  • First replicate is randomized
  • Minimizing Filler Plots

Received Test design is not serpentine

A problem that can occur is how to incorporate into the design of the YIELD TRIAL FIELD, an individual company’s pre-designed YIELD TRIAL which will be referred to here as a test. This pre-designed test is a specification of a randomization scheme with plot numbers and possibly also a planting plan, that the company has created for you to use as their test. Usually, the company will pre-package the seed in individual seed packets for each row as well and staple the packets for a 2-row plot together. One would think that having all the design work done by someone else would make things easier, but in practice, it makes things more difficult, particularly if there is little or no coordination between the company’s test designer and your YIELD TRIAL design policies. For example, BIG YELLOW SEEDS (our company) designs all YIELD TRIAL FIELDS with plots numbered in a serpentine pattern. This allows a person taking notes to move from one range to the next in an orderly and efficient manner. However, a company may design and send a test for their company plots that does not have a serpentine pattern. Instead, the first plot of each range, for example, is always on the left. As an example, KORN KING sends a test that assigns the plot#’s 1001-1020 to the first range with plot# 1001 on the bottom left side of the planting plan. The second range of 20 plots is assigned plot#’s 2001-2020. Plot# 2001 is directly behind plot# 1001 in the field. On a planting plan page, it’s above plot# 1001 on the left side of the page. There is a third range numbered from left to right from 3001 to 3020. Plot 3001 is directly behind plot 2001 in the field.

First of all, one should inform companies sending pre-designed trials that a serpentine pattern is used by Generic Seeds (our company) in all tests to facilitate the taking of notes. If the company designer chooses to ignore this information, you have no choice but to deal with the problem.

The design and analysis of a test can place importance on the position of a plot in the design and on the position of the immediate neighbours of that plot and their effects on that plot. For this reason, one needs to ask the designer about making any changes to the designed test and keep the original designer informed of any actions that affect the design. For example, if the field was designed without a serpentine pattern but is planted with a serpentine pattern, it creates two problems:

1) First, the company representatives will have a planting plan that is incorrect and when looking at what they think are the plants of plot# 2001, they will be looking at the plants of plot# 2020. The second range on their planting plan goes left to right, but in reality it has been planted with the plots in the second range going right to left.

2) Secondly, the designer may have sought to minimize or investigate the effects due to neighbouring plots, on yield or some other trait. Changing to a serpentine planting plan, puts all the plots of even numbered ranges in a different position with respect to the plots in front and behind. For shading, this is not critical because plots still have the same neighbours just on different sides but for some analyses of soil conditions etc., the planting plan used by the designer will be incorrect.

There are several ways of dealing with the problem of serpentine policy vs. unserpentine policy.

1) If it is acceptable to the original designer, redesign the planting plan with a serpentine pattern. Send a copy of this redesigned planting plan to the original designer and company for their records so that representatives sent to take a look at the plots have a correct planting plan. Make sure to properly stake and tag the plots with the serpentine pattern according to the revised planting plan. This is the simplest solution.

2) Arrange to plant the plots as specified in the original planting plan but alter the planting plan and plot#’s on the seed packets for the person filling the seed trays and alter the seedlist and notebook for the person taking notes.

If there were no plot numbering conflicts, you could use the original planting plan and give it to the person filling seed trays for planting but it’s not a good idea. The person that fills the seed trays for planting is used to serpentine planting plans and may accidentally overlook the difference in the planting plan. The different planting plan would also involve breaking the established serpentine routine of filling planting trays. The person filling the trays might forget not to use a serpentine pattern despite the planting plan and verbal or written instructions.

One way around this is to renumber the plots. However, you can’t just renumber the plots sequentially. Even numbered ranges in the design must be sorted from largest plot# to smallest plot#. This way, plot# 1020 is followed by plot# 2020 in the planting plan, seedlist, seed packet tray, planting trays and notebook.

As a matter of policy, for all pre-designed tests, retain in hard copy form all the original design information and any other information related to the tests. Be sure to get down in print any information printed on the received seed packets as well. Keep randomization, seedlist and planting plan information on the computer in the same form as for any of your own designs. It has happened that a company’s notebook did not match the information printed on their seed packets. If the seed had been planted using the plot numbers printed on the seed packets, the information in the company notebook would not match the expected phenotype of the plants in at least one plot and the company representatives would easily assume that there was a planting error and that Big Yellow Seeds Seeds planted the test incorrectly.

Break the data up into separate ranges on a spreadsheet and sort the even ranges in descending plot# order. Create a new plot# column to the left of the original plot#’s. Use Excel’s autofill feature to fill the new plot# column from the first new plot# to the last new plot# in sequential order. Print out this information as a revised seedlist. Get the seed packets and using the revised seedlist, rearrange the order of the seed packets in trays. Use a permanent marker to renumber the seed packets on the top left according to the revised seedlist.

Revise the planting plan to a serpentine pattern and print this out for use by the person filling planting trays. When printing the notebook, hide the old plot# column. When preparing reports, delete or hide the revised plot#’s.

It should not be a problem if a test must start from a different side of the field than the company planting plan if one was received. Merely inform the company designer that the directions of the ranges have been reversed so he can inform company representatives that wish to view the test. Be sure to properly stake and tag the test.

Plot numbering conflict

The YIELD TRIAL FIELD doesn’t need to be sequentially numbered from the start of the first plot of the first test to the last plot of the last test. Each separate test can use a different set of numbers. Just make sure there are no numbering conflicts between separate tests. It’s also best to avoid using numbers assigned to nurseries or crossing blocks but since the seed from tests is not kept it isn’t a problem.

If plot#’s need to be changed just add a separate column on the test spreadsheet or a new plot# field to the database file.

First replicate is randomized

It is a Big Yellow Seeds company policy not to randomize the order of the entries in the first replicate. This minimizes confusion if there are problems in the planting. The plots of the first replicate are used as references to the identity of seed in subsequent ranges of the same test. Technically, 1,2,3 is just as random a sequence as 3,1,2. It is easier to work with plots and entries if the first plot is Entry 1, the second plot is Entry 2 and the third plot is Entry 3 etc.

The easiest way to handle a randomized first replicate is not to mention the actual entry numbers to the fieldworkers but to refer only to plot numbers. If the first plot is Entry #5, let the field worker assume it is Entry 1. Tell the field worker which plot#’s correspond to the same entry number for the first plot. This creates a bit of confusion for you if the fieldworker starts referring to the seed in the first plot as Entry #1, but other alternatives can be more complicated and create confusion for more people. A second possible way to handle a randomized first replicate is to assign new Entry #’s in a new spreadsheet column and hide the old Entry #’s until needed for reports. This method keeps everything the same but what the company may refer to as a cat you just call a feline.

A third method of handling a randomized replicate is to retain the original Entry #’s but to physically reorder the first replicate by Entry # and renumber the seed packets. This method changes the planting plan so the original designer would have to be contacted before doing so. It also creates more work for you.

Minimizing filler plots

When designing the YIELD TRIAL FIELD one wants to minimize the number of filler plots needed to create a rectangular field. At the same time, the plots of a replicate should all be in the same range. However, a replicate can be split into more than one range if needed. Just remember to inform the original designer if there is a change to the planting plan.

Determining the number of plots wide for the YIELD TRIAL FIELD can be done by finding the greatest common divisor of the number of entries in each test, or the closest number to the greatest common divisor.

If one receives a 2 rep, 60 entry test and a 3 rep, 20 entry test, the greatest common divisor of 60 and 20 is 20. Filler plots will not be needed at all if the 2 rep test with 60 entries is planted in 6 ranges with each replicate requiring 3 ranges and the 3 rep test with 20 entries planted in 3 ranges with each replicate requiring only one range.

If the first test was planted 60 plots wide, and the second test was planted 20 plots wide, there would be a block of 120 filler plots beside the 3 rep, 20 entry test. This is a large waste of inputs. {C}It’s possible, if absolutely necessary, to put 2 different tests in the same ranges. The only stipulation is that there should be at least 2 and preferably 4 guard rows between the tests in each range. Constructing the divided ranges can be a paperwork nightmare.

Each test should be a nice block with each range of the same test directly behind one another. This means plot numbers have to be assigned so that the planting trays can be filled in a serpentine manner. The first range of the first test will be followed by plots containing the filler/guard rows, followed by the first and second ranges of the second test, then filler/guard rows and the second range of the first test.

If one was using the plastic seed trays for the automatic indexer, one could keep the old plot#’s, fill the cells for the first test in a serpentine fashion, fill the cells for the guard/filler plots in the middle, and then fill the cells for the second test in a serpentine pattern.

Field Design and Filling Planting Trays

Once it has been decided where the Yield Trial field or fields will be grown, that part of the physical field will have to be measured and marked out. This can be done as soon as the weather and soil conditions permit. A rough measure can be taken and then a more accurate one just prior to planting (as a check and to set up field markers and guides for driving a straight line).

A planter pass is 10 ft.. (Find the correct width depending on the equipment used). Usually there are two headlands on each end of the field of two or three planter passes in width (8 - 12 rows, 20 –30 ft.). Usually the front headland has been 12 rows and the back headland has varied according to requirements.

You want to have as few empty plots in the last range as possible but make some allowance for possible late entries. Also note that 4 rows of guards will be planted on each side of the yield trial field.

You will need to set a range width (I’m using the word width here in its strict definition as the shortest measurement). This is the plot length + 3 ft. pathway. This should be set in the design phase. Usually this has been of about 21 feet including the path (18 feet of plot + 3 feet of path (or 2 x 1.5 ft of path at each end of the plot)).

Measure the field length available and subtract 60' for 12 rows of headland at each end. The amount of headland to grow may vary with circumstances. To find the number of ranges available,divide the available field length minus headlands by the range width and ignore anything after the decimal.

You will need a measurement for the available field width (left to right or right to left) so you should have some idea of the area in which you want the Yield Trial field or fields to be grown in and possibly an idea of the required area according to proposed plans. At some point you should have taken measurements of the entire available space and then later you can check how proposed plans fit with this. Maybe space needs to be reserved for a Nursery of some kind on one side of the field or there is only room for 50 plots wide and then there is a steep slope in the field or a low spot that would adversely affect the tests.

For In-house Yield Trial fields the company should try to use a standard test and field size when possible. This should be set in the design phase with only minor tweaking based on the measured available space. The number of tests in a field can change but as a rule of thumb the number of entries in each test and therefore, the field width should be kept the same if possible. One field design I have seen was for 3407 plots. In one location this was grown as a field 86 ranges long by 40 plots wide. In a second location, these 3407 plots were grown in 5 fields beside each other with 4 guard rows between each field. Four fields were 26 ranges by 26 plots and the fifth field was 26 ranges long by 28 plots wide.

Yield Trials conducted for other companies have usually been placed in a single Yield Trial field but the width of this field has never been the same each year since companies send a different number of entries to test each year. The test with the largest number of entries plays a dominant role in deciding how wide the field should be. Ideally, you want to grow the test with the most entries without splitting the replicates into more than one range, however, if the largest test is 60 entries and the next test is only 24 entries, there will be a large block of filler plots in the ranges containing the 24 entry test. You might prefer to use a field 30 plots wide and split the replicates of the first test into two ranges so you have fewer filler plots to deal with.

To find the field width required, one multiplies the number of plots by 2 times the row width, or multiply the number of rows (2 x number of plots) by the row width.

E.g. 21 plots x 2 rows per plot x 2.5 ft row width = 105 ft + guard rows

If 4 guard rows are planned for each side then 105 ft + (8 rows x 2.5) = 125 ft is needed. If there are multiple fields with guard rows between them and on the outside borders, you need to take all of the field rows and guard rows into account and make sure there is enough space available with maybe 5 ft. to 10 ft. extra. The extra space is needed because the row width btween planter passes can vary a few inches and the rows might not be planted exactly perpendicular to the physical field edge.

Once the available space has been measured and the number of ranges and number of plots per range has been set, one can create the Planting Plan showing the ranges of plots and the direction of plots in each range of a serpentine design.

Planting Plan

For each physical Yield Trial field that you create you will need a planting plan to guide the filling of planting trays with either seed (plastic trays) or seed packets. The planting plan was created with a spreadsheet program. It should have a title showing the name, location and possibly the year. The person filling the trays doesn’t need the year but you will print out the planting plan later for inclusion in the notebook.


2001 Private Company Yield Trials

A detailed Planting Plan

A basic planting plan shows 2 lists of plot numbers separated by arrows alternating in direction. The plot numbers on each side of the arrows are the first and last plot numbers on the sides of each of the ranges. The arrows show the direction in which the plot numbers increase. The planting plan should show range numbers on at least one side of the page, usually on the left. If the field is quite wide you might want a second list of the range numbers on the right side as well. The first range is the closest to the bottom of the page.There should be a title centred above the planting plan and on the left side you might want to add information regarding tests in certain ranges. You should also include information regarding guard rows on each side and if they are to be put into the trays or planted straight through after the trays are planted.

Make very sure that your arrows are alternating properly and going in the correct direction. For in-house designs the arrows should ALWAYS alternate to give a serpentine pattern. For tests from other companies there may be an exception. The test may have been designed to start on a different side than what it should if it was in a field created with a serpentine pattern throughout. It’s best to keep things going in a serpentine pattern if possible because if the arrow doesn’t follow a serpentine pattern, it may be missed or regarded as a mistake since that’s the usual pattern.

The plot numbers for a specific test should be sequential but they don’t have to be sequential for the entire field. You can have tests start at different plot numbers if necessary. Sometimes companies send tests with different starting plot numbers so it’s easier to use those numbers than change a lot of things.

There are several ways to get the plot numbers for the sides. Recently I found a way to calculate and list them with a simple program so I don’t have to calculate them at all. You might also consider using field designs of 25, 50 or 100 plots wide to make it easier to calculate the side numbers.

With the spreadsheet you can start on whichever side plot 1 is on and move to the other side and put the plot number for that side (call this number FW (FIeld Width for now) in the cell. Then move up one spreadsheet row and add 1 to FW and put that number in the cell. Then move to the other side and add one less than FW to the number you last entered and put the new number above the one. The pattern is move to the other side, add (FW-1), move to the other side, add 1 and repeat.

You can also number the sides going up or down the columns. The pattern going up is add 1, move up, add (2*FW-1) and repeat. One side starts with 1 and you add (2*FW-1) then add 1 and repeat while the other side starts at FW and you add 1 then add (2*FW-1) and repeat. You might also use formulas in the cells if you know how. Just be vary careful if you use formulas to eventually paste the values in. If a number gets altered when you leave the formulas in, it can mess up at least one more number and possibly a lot of other numbers. Use any method to get the numbers in there but be very careful not to make a mistake.

More than 18 ranges

If more ranges are needed than can be fit into the plastic trays (18) or the cardboard trays (25?), this is handled by using a second set of trays above the first set and labeling the first set as A and the second set as B. Tray 1A might hold the seed for ranges 1-18 of the first 4 rows/columns and Tray 1B would hold the seed for ranges 19-36 of the same first 4 rows/columns. The operator of the planter will have to stop at range 19 (plastic trays) so the empty tray 1A can be removed and Tray 1B can be loaded.

Show the tray assignments a few lines below the first range arrow in a manner similar to showing the ranges but with a very short arrow that always goes the same way according to whether the field is being planted left to right or right to left.

Tray Assignments
C1 C2 C3
R1 TRAYS:1D --> 20D 21D 22D --> 34D
R2 1C --> 20C 21C 22C --> 34C
R3 1B --> 20B 21B 22B --> 34B
R4 1A --> 20A 21A 22A --> 34A

The figure above shows the trays needed for two fields with a 4-guard-tray between them. The field on the left is 20 trays wide (40 plots/80 rows) by 4 trays long (72 ranges). The arrows show that the left field is being planted from left to right. The first planter round would plant trays 1A, 1B, 1C, 1D, 2D, 2C, 2B, and 2A in that order. The field on the right is 13 trays wide.

Guard rows can be planted straight through or as plots and should be indicated vertically on the planting plan outside the plot numbers and arrows. When there is only one field, the guard rows are planted straight through on each side of the field and are shown separate from the test plots.

When working with multiple fields, a modular system can be used in which the extreme left and right borders consist of two guard rows that are planted straight through while separate trays of 4 rows of guard-row seed is planted between each field. Each modular field then, has 2 guard rows on each side.

It might not be very convenient but one could plant the 4 guard rows between multiple fields straight through (4 Guard-Solid) rather than as plots (4 Guard-Tray) if an empty planter pass of 10 ft. was left between each field.

The plots of a test should all be in the same field but the numbering of plots can be continued from one field to the next.

One of the peculiarities of pre-1995 planting plans was that the plots for STANDARD TRIALS (1 rep/location and 1 standard variety) were always numbered starting with a 0 for the units digit. E.g. 0 or 5000 etc. You are advised not to do this but to start all numbering with a units digit of 1.

Another peculiarity of pre-1995 planting plans was that a plot numbering scheme was used in which the extreme left plot numbers of a field were not shown. Instead, the numbering would skip two plot#’s on the left and go from say, 48 in range 2 to 51 in range 3 and from 98 in range 4 to 101 in range 5. This may have been because these 2 rows on the extreme left contained seed for 2 guard rows. You are advised not to use this numbering system but fields of 24 plots may make the planting with a 4-row planter easier.

It might be a good idea to print out the planting plans for each field separately to reduce confusion when filling trays with seed but it might also be a good idea to be able to see the entire set of multiple field plans on one page to make sure one includes all the required guard rows.

When you are satisfied that the planting plan has all the necessary information and has no errors, you can print it out and send it to the workroom. If you had a network connection to the workroom you could just send the file.

Once the Planting Plan is printed out, one can lay out the empty planting trays whether large brown cardboard trays with exactly ten white trays each, or the plastic 4 cell electric indexer trays. Laying out and filling the planting trays Explaining about laying out the planting trays would best be done with a video or at least with the use of pictures.Two different types of trays have been used:

1) Plastic Trays

These trays are used in the electric indexer on the Kinze Planter. They are clear plastic with 4 bubbles/pockets/cells wide and eighteen cells long. There are some pros and cons of the electric indexer to consider which will be discussed later in a planting section. The seed in the Seedlist seed envelopes is poured into the cells.

2) Cardboard Trays

These are large brown cardboard trays with 10 empty small white cardboard trays inside. The seed envelopes are placed in the empty white cardboard trays in the large brown cardboard trays.

A large table is needed to layout the trays on. Two tables can be put together or even plywood sheeting can be used if it's strong enough to hold the weight. The longer the surface available, the better but a method will be shown regarding how to work with wide fields and a short surface.

A Planting Plan is essential to laying out the trays. First determine the number of trays needed depending on the number of rows wide the trial is to be. For a 60 row wide trial, at least 15 plastic trays would be needed or 6 large brown cardboard boxes.

The plastic trays are placed 4 cells wide side by side. 15 such trays beside each other will give 60 rows wide. That is, 30 2-row plots wide. The plastic trays hold eighteen ranges. How to add more ranges will be covered later. The brown cardboard trays will hold more ranges but it depends on how fat the seed envelopes are. If the table is too short to hold 15 trays, layout the trays as far as possible on one side of the table and then continue around the table laying out trays going in the other direction on the other side. This can be thought of as folding one long line of trays. Whether the fold occurs on the left side or right depends on which side of the field Plot # 1 is on. Make every possible effort to prevent movement around the table on the end opposite the fold. A physical barrier is better than a mental one. {C}Example: The Planting Plan is placed before you in the middle of the table. It calls for 60 rows (30 plots) wide by 10 ranges long. Plot#1 is on the left (The first range moves from left to right). The table holds only 10 plastic trays wide. 15 trays are needed.

The plastic planting trays need to be labeled numerically to specify the end that corresponds to the first range and the order in which they are to be planted. Masking tape and a black marker can be used. Flip the trays over and put masking tape on the bottom of a corner cell of each tray. For now, put it on the right corner cell of each tray. This means, with the cells upside down,there will be 3 empty cells, then a taped cell with a number 1 on it, then 3 empty cells and a taped cell with the number 2 on it etc. until the last cell on the right on the other side is labeled with the number 15.It makes no difference if the label is on the left of a tray or the right. Just be consistent. Do the same to each tray.The end of the tray that the label is on is inserted first into the indexer if one is about to plant the first range and last if one is on the other side of the field about to plant the last ranges to the first range.

The trays are flipped back over so the cells can be filled with seed and laid out from the left side of the table to the right. Then move around the right end of the table so that from this other side the table now extends to the right. (The fold is on the right end of the table when standing in front of the first 10 trays on the one side of the table). Continue laying trays on this other side from left to right. 5 trays are so placed. {C}Return to tray 1 and place a physical barrier that prevents movement around the table on the left from tray 1 to tray 15. It is easy to remember not to go around the table on this end when one side has only a few trays. Sometimes though there can be an equal number of trays on each side of the table and it’s easy to move in a complete circle. {C}Place the Planting Plan in front of you. Pick up a white cardboard tray of seed envelopes from the large trays labeled for the trial. Always hold the tray of seed envelopes sideways and close to you and never over the planting trays. If seed should be spilled, let it fall on the floor rather than into planting tray cells/envelopes empty or filled.

Consult the Planting Plan in front of you on the table. The first cell to be filled is on the extreme bottom on the left end if you are standing in front of the first 10 trays. The white cardboard tray holding the seed envelopes should contain 2 envelopes with the first plot number on them. The envelopes should be taken out one by one in the order they are placed in the tray. (This is to keep one from picking envelopes out two at a time and changing their order for fields with single row numbering). {C}Take out the first envelope, ensure it has the correct plot number and pour its content into the first cell. DO NOT OVERFILL THE CELL. A metal cover has to be placed over the cells of a plastic planting tray after all the planting trays are filled. Seeds that are higher than the rim of the cell will be pushed into other cells when putting the cover on. This is not at all acceptable.

Pour the second envelope into the second cell (to the right of the first). The seed of these two cells are identical because they are the same variety in two rows. Sometimes, companies use different colour fungicides for each variety or for their material. The colour can help keep one on track. {C}Continue filling cells along the first range to tray #10 then imagine the fold and continue filling the first range of cells on the other side of the table from left to right to Tray #15. DO NOT go around the right end of the table when standing in front of the last five trays. Turn the planting plan around so you can read it and check the Planting Plan regarding the Plot# on the extreme right of the first range. Make sure it matches the last two cells filled. The next cell filled is directly above the last cell filled on the extreme right when standing in front of the last five trays. Continue filling trays moving in the opposite direction from before. So move from right to left in the second range of cells until reaching tray #1 and filling the extreme left cell in range two.

Using the plastic trays, knowing where one left off when a white cardboard seed envelope tray is emptied is usually simple since the next cell to be filled is an empty cell beside a filled cell. It can be a problem,however, if you finish a range in the last tray and go for lunch. If you don’t mark where you left off or consult the Planting Plan you could start a new range in the first tray going the wrong way and not know there was a problem until you consulted the Planting Plan at the last tray.

For the large brown cardboard planting trays with 10 white cardboard trays in each box, it is difficult to know where one left off. One way to mark where the last filled place was is to put the empty seed envelope tray on top of the last filled position with the rest of the seed envelope tray over already filled cells. {C}Range three is filled starting at the cell in the third range on the extreme left when standing in front of the first ten trays. Range three is filled moving left to right the same as range 1. NOTE: DON'T go around the left end of the table from Tray 1 to Tray 15. The process of filling the planter trays in this serpentine pattern is repeated until all the ranges have been filled as per the Planting Plan. Consult it at the end of each side of a range. WIth the cardboard trays and envelopes one can miss a place easily.

Once the plastic trays have been filled according to the planting plan the covers can be put on. Be very careful not to push seed from one cell into another cell.

For the cardboard planting trays a Planter Checksheet is needed to verify that the seed envelopes have been placed properly. This is just an expansion of the Planting Plan to show all plot#'s. The columns of plot numbers are compared to the contents of the white trays checking each set of two trays from range one to the last range. The two trays should match because each seed envelope is one row of a two row plot. If one of the two trays has 18 envelopes and the other has 19, there is a problem. Or if the 10th envelope of tray 1 is for Plot# 200 but the 10th envelope in tray #2 is for Plot# 306 there is also a problem. When placing envelopes, try to make sure an envelope is placed behind all previously placed envelopes and not slid in front of a previously placed envelope.

The check of the white trays is perpendicular to the direction in which the trays were filled with seed envelopes so this is a cross check of sorts.

The Planter Checksheet can be formatted in pairs of plots to match the four rows of the planter. There can be several pairs of columns on a page.

When the envelope placement in the cardboard trays has been verified to be correct, a post-it note or masking tape and marker can be used to assign a Yield Trial name or location and tray numbers in numerical order to each tray. The trays can be stacked with higher numbers below lower numbers so the trays are planted from 1 to n.

Doubling up with seed packets

If the cardboard planting trays are being used, one has the option of stapling the two envelopes for the two rows of a plot together and putting them into one tray rather than two. Most of the seed received from other companies is packaged in this way. It is nice not to have to separate the pairs. The problem, however, is that this doubling up leads to the need for a tray on the planter that must be longer than normal for a field of 15 ranges or more. The alternative is to use twice as many white trays and to split the pairs in the office and put them into the white trays just as one would pour individual envelopes into cells for the plastic trays.

Planting Tray Checksheet

Each spring that the cardboard planting trays were used, the Planting Plan for the field was created and given to whomever was filling the planting trays with seed packets (usually me). Then, after all the seed packets were placed in the planting trays, I had to create a full table of all the plot numbers in each Yield Trial field to check that the planting trays had been filled correctly.The plot#’s of envelopes in the planting trays were verified using the checksheet by checking each pair of column of numbers in each planting tray from the first to last range. Each column of seed packets was one row and the plots were 2-row plots so the pairs of plot#’s in adjacent columns in the planting tray had to match for each range.

E.g. A field that is 5 ranges by 5 plots wide.

Planting Plan
5 21 ---> 25
4 20 <--- 16
3 11---> 15
2 10 <--- 6
1 1 ---> 5
21 21 22 22 23 23 24 24 25 25
20 20 19 19 18 18 17 17 16 16
11 11 12 12 13 13 14 14 15 15
10 10 9 9 8 8 7 7 6 6
1 1 2 2 3 3 4 4 5 5

The Planting Plan is used to place the seed packets in the planting tray in a serpentine order from Plot number 1 to plot number 25. Since each plot has two rows there are two envelopes beside each other with the same plot number for each plot number on the planting plan. The Planting Plan is expanded to create the checksheet. This was done in a spreadsheet. I found using the data fill command in EXCEL to be the simplest way. I knew from the planting plan what the left and right end numbers were and just entered the left side numbers along one side and maybe the next column of numbers to help data fill know if the sequence in a row increased or decreased and used datafill to fill in each row of the table. Newer versions of Excel don’t call this data fill but completing a sequence. You highlight the first two cells of the row and drag the cursor across the row using the handle. The numbers were filled in automatically. My fields were never more than 30 plots wide or 20 ranges deep so this didn’t take too long.

What you are trying to catch with the checksheet is an envelope out of place in the Planting tray. Sometimes a spot can be missed going across the range or an envelope can get slid in in front of another envelope. Usually the error of missing a spot is caught at the end of the range using the Planting plan but envelopes in the wrong range are not. According to the Checksheet, columns one and two of the planting tray should have the plot numbers 1,10,11, 20 and 21 from range 1 to range 5. You can check this with each hand tilting envelopes at the same time to see the numbers in each column. The numbers should match the Checksheet order and each other. If you had an envelope with plot number 10 in the first column and the envelope in the second column was for plot #9, the envelopes are out of place because they shouold both be plot number 10. Perhaps somewhere a spot was missed going across the second range. So you would now correct the problem by finding where the plot numbers start going wrong and putting the proper envelopes in their proper place.

Continue checking each column until you are done all of them.

You might also want to format your checksheet print-out into pairs of columns (4 row planter) to help determine plot numbers using planter passes in the field if there is a problem.

Field Preparation (Field technician’s hat)

Material relating to the Field Technician's hat should be put in a separate document. It should include very practical information regarding the operation of the tractor, implements and field work.

Some points might be:

  • Don't work the tractor hard and then just shut it off. Let the engine cool itself first by idling the tractor for about 20 minutes. (This might not be applicable to all tractor engines but the idea is that if the engine block experiences abrupt temperature changes it can crack and be very costly to repair.)
  • BEFORE striking out inwardly across a field from a corner, clean off all plant debris from the shovels of the field cultivator or other implement to prevent creating a trail of transplanted weeds across the field.
  • Use depth gauges on field implements to prevent working the soil too deep.
  • When passing through lighter soil, raise the field implement slightly so the soil isn't worked deeper than in other areas.
  • It is better to have seed in the ground one or even two weeks before the soil is the optimum temperature for germination rather than to wait until the soil germination temperature is reached and be rained out of the field.

John Deere put out a series of volumes in the mid 1970s entitled Fundamentals of Machinery Operation that could be consulted.

The research director of a private plant breeding company claimed in the late 1990s that despite all the Ph.D.'s,M.Sc.'s,B.Sc.'s and college technicians he had on staff, the field technicians were not setting the planters properly and were inexperienced.

This isn't really surprising since if one is born in a city, one isn't going to get much opportunity to drive a tractor or plow a field. As a University field technician, one only operates a tractor for a few weeks each year.

Perhaps there is an opportunity here to create an Agricultural College that intensively trains tractor operators. A related cooperative might contract work with agricultural producers and could cycle graduate students through training on different tractors and implements.

A field technician that was plowing fields on a Fordson Major tractor at the age of 12, and who was an agricultural producer for decades, still claimed in his late 50's that there were other agricultural producers he would love to shadow regarding field work (if it wasn't a safety issue, he would ride in the tractor cab with them and just watch what they do). These other producers knew how to read the weather and soil so that if a rain came, their crop would sprout through a crack over the row while other people's maize would be sealed beneath the surface.

Interestingly, several of these star-rated operators were trained in Europe, so there might be something in the training there that is more practical and wise.

Seedbed creation (Field technician’s hat)

Application of Nitrogen (Field technician’s hat)

Measuring and Marking Ranges

From the side where the planting will begin and with the heel of a boot, make a mark parallel to the field edge and about 1 foot in. This represents the first row of the headland. Pull out a tape to a width that matches the planter width. Make a heel mark and continue measuring the headland for the planned number of planter passes for the headland. From the last mark, add and mark 1.5 feet and place a small wooden stake at the mark. The stakes are about 1/4" thick,1" wide and about 12" long. This stake is in the center of the 3 foot pathway in front of the first range. It may be necessary to angle the stake both vertically and horizontally to catch the sunlight. If possible use stakes painted white. {C}From this first stake, the ranges are going to be marked off on this side of the field. One person will hold the tape at the plot length with the thumbnail on the plot length number. The other person has a bundle of stakes and holds the front end of the tape.

A small stake is held next to the front of the tape in such a way that the stake can be placed in the ground in one movement down to the ground. The stake is placed so the width parallels the tape and its center is on the end of the tape. {C}The measuring pair of people then move forward until the second person is at the first stake and the measuring tape is taut. One or both yell out the range number "ONE!" at the same time that the first person places the second stake. Then the pair move forward until the second person reaches the second stake and both say "TWO" as the next stake is placed. This continues until the required number of ranges are staked out. The measurers need to move perpendicular to the field edge as much as possible perhaps by sighting out a distant object. The Stakes should be placed several feet outside the desired side of the field. This will ensure the marks can be seen when planting the first and last passes.

After marking the first side of the field, the other side of the field is marked. The ranges can be marked out starting from the far end of the field or the near end. For a large field or a set of fields all being planted at the same time, a third set of stakes may be needed between the 2 outside stakes to help keep the tractor driver going straight when marking the ranges. The third set is usually added by lining up the center stakes between the outside stakes.

The ranges are marked using a two disk marker with the disks set about 3 feet apart.

The driver lines up with the 2 stakes marking each side of the field in the center of his path and possibly also using center of range stakes and drives to the far stake on the other side as the disks mark a three foot pathway. At the other side, the tractor operator lines up the second set of stakes. It may be necessary for someone on each side to stand behind stakes to help the driver see the stakes and not also miss a stake and drive diagonally across the field. Marks should be made a little past the stakes so one can see them off to the side as the first and last passes are made with the planter. {C}Once a range has been marked, the stakes for that range can be removed. The stakes can be removed as one moves up the ranges on one side.

Illustrative video showing the marking of range aisles

NOTE: A video of real people doing the measuring and marking would be best but the overhead crane shots might be expensive. I would like to create a 3D version in Second Life or with Blender 3d and Platinum Arts Sandbox or a machinma with SimTractor 3d but no longer own a desktop computer.The video is part of creating ranges for an Inbreeding Nursery that is staked every five rows.It is incomplete and rough as I am neither an artist nor a programmer. I still need to make 2D RPG sprite characters of my own and add measuring animations to the video. The perspectives of the tractor and stakes in the video are inconsistent and need to be either top-down stakes or 3/4 overhead tractor. Can you draw? I need lots of artwork to illustrate these notes.

Amount Due Calculation Spreadsheet Sample

A simple spreadsheet was created to calculate the amounts to be paid by clients. Some companies preferred to pay half prior to planting and half after harvest so you might want to add columns for each payment.

  • Number of plots = (# of entries) * (# of replicates) * (# of locations)
  • Amt Due = Price per plot * number of plots
  • Note that a lower price is being charged for 3 replicate tests in this example
Sample Amount Due Calculation
Company Ents Reps Locs Plots Price/plot Amt Due
Adam Seeds 60 2 2 240 $30.00 $7,200.00
Bill Corn 36 3 1 108 $25.00 $2,700.00
Corey's Hybrids 25 3 2 150 $25.00 $3,750.00

Planting the green mile

Loading the trays(Envelopes)

Have a small notebook handy for recording things like planter settings (primarily the numbers assigned to the two sprockets that control the plot length and their position as drive sprocket or follow sprocket) or location of misplants etc. Also bring the planting plan and the planter checksheet to help if there is a problem planting. Know the direction of the first range from the Planting Plan and imagine it as a number line at the front of the field. The envelopes at the front of the small white trays should be in the same proper sequential plot number order in the large brown trays.

The people filling the cones on the planter are in seats between 2 cones and a few feet behind the cones, facing forward with two small white cardboard trays in a tilted metal or cardboard tray in front of them. This is the only arrangement I will be using.

( I noticed some other companies have the people sitting on the planter facing backward (their backs are to the tractor). The seed packets are in front of them. I find this arrangement quite disorienting but it may have its merits. )

Planting Yield Trials is easier than inbred rows because each person is planting two rows of the same seed and plot number. If the envelope in the left hand is poured into the cone on the right, it is not a problem. As long as the envelope in the right hand was poured into the cone on the left. However, it is best to enforce the rule against crossing sides with seed packets so there is no trouble when planting the Inbred Nursery.

The large brown trays should be packed for shipping so that the first ten rows are on top and can be used first followed by the next ten rows etc. {C}First, place bottles for the first eight rows. For the first four rows, the bottles are put in front of the first envelope. For the next four rows, the bottles are placed behind the last envelope. Each time you complete a round, you will do this to the 8 small white trays you pick up for the next round. {C}Don't even think of ever trying to take all 8 small white trays to the planter in one trip! Take 2 or 4 trays at a time,whatever you can handle without spilling, and maybe arrange to have the other person or the driver help you. {C}Here is the tricky part. Each small white tray of the first four rows must be individually spun 180 degrees so that the bottle is away from you and the back of the envelopes are towards you. It's best to do this in the large brown tray so that now you have 8 rows with bottles away from you but the first four rows have envelope backs to you. Now the trays are in the proper order as viewed from the back of the planter.

Next, I pull the first four trays out as a block. I put my hands on each side of the ouside trays and press gently towards the center as I lift up and get my fingers under two trays on each side. Don't press the trays too hard together or they will buckle upwards and spill. You can also take them out two at a time. {C}Give the 2 trays on the right to the person on the right side of the planter and then put the 2 trays for the left side in their proper boxes on the planter. Return and get the next four trays for the round. Don't change the sequential order of the next four trays or spin them around. Place them in the fertilizer box with the bottles to the back of the planter and make sure they are positioned and supported so they won't spill.

Check that the 2 plot numbers match for each side. Each person can check their two top envelopes behind the bottle and the one person can sound off regarding the plot# and that the match is ok. The plot # should be one higher or lower depending on the direction of the first range.

The general rule for planting envelopes is TAKE IT FROM THE TOP. Always draw the next envelope from the top of the tray. This is the envelope farthest from you. Don't take an envelope from the bottom of the tray. The reason envelopes are taken from the top is because when envelopes are taken from the bottom, there is pressure on them from the bottle and the other envelopes that makes them rip on staples or pull other envelopes out with them when they are taken out. {C}When you are confident the envelopes are in their proper rows on the planter, draw the top envelopes from your two rows. Draw the left envelope with the left hand and the right envelope with the right hand. Look at the plot#'s. They should be the same. You can sit them on your lap to help get a better grip on them or to be ready. Pour each envelope into its respective cone. This is the signal to the other person that they may do the same. When they are finished, and you are ready to go, check with the other person that they are ready to go and if so, give the signal to the driver to pro seed. ( pun intended!). You can also tell the driver to ‘make it sew!”.

When the planter is in the proper position with respect to the first range marker line in the soil, press the button to drop the seed. Push the button confidently and firmly. If you press it quickly as if you were afraid it would shock you, the press might not register at all and the seed won't drop. If you press slowly, it might register as two presses and seed just put in for the next range may drop in with seed of the previous range. This is a problem someone can work on.

Get the next two envelopes and pour them into the head ready to be dropped onto the cones. You should be about in the middle of the range or even three quarters through if you had some problems pulling the envelopes out. Get into a habit of press and fill,press and fill. If you don't, you will be doing something else and miss doing one of the two. Don't do anything else like sounding off to check the plot#'s or checking the cones until the head has been filled.

Locate the marker line for the end of the first range and the marker line for the start of the second range. The two lines should be separated by 3 feet while the other lines are separated by about 18 feet. For short plots like in the bean nursery, it can be confusing to know what is path and what is plot. Prepare to press the button when the planter is in the proper position with the second of the two range marking lines ahead. This is the line marking the start of the next range. Don't use the line that marks the end of the first range or you will be planting into the pathway. Sometimes you can note whether seed is still being planted in the cone from the first range. All the seed should have gone down the hole before you press the button.

When the planter is in the proper position for the second range, press the button,get the next two envelopes and fill the head.

Things to check for:

  • Do the plot #'s match for your pair of envelopes?
  • Is the plot# being planted by the other person in the same range as yours?
  • Is the cone working properly?
  • Is it going around at a steady pace or is it stopping and starting as if it's catching on something?
  • Is all the seed out of the cone before the button is pressed?
  • Is all of the seed dropping onto the cone?
  • Is the seed evenly spread around the cone? Are the cups working properly to drop the seed?
  • Is the cup opening and closing properly?
  • Is the cup opening and closing when the button is pressed?
  • Is the cup opening and closing more than once when the button is pressed?
  • Is the seed being spread evenly around the cup before being dropped?

Make prior arrangements with the driver regarding what signal will be used to have him stop at about the 5th or sixth range to check seed placement and depth. {C}Stop the driving after a range has completed planting and before starting to plant the next range.

The main correction to be made will be regarding when to press the button. {C}After checking the seed placement the driver will have to back up a bit to get the cones moving before continuing the planting.

In the last few ranges you may encounter empty envelopes. It may seem stupid but just go through the motions of pouring them into the head as if they contained seed. The reason for this is that it maintains the routine of press and fill. Otherwise, with nothing to do, one will be looking at something else and forget to press the button. It also helps ensure that everything is in its proper place especially when starting from the empty range.

Try to see where the last seed of the last range drops to get an idea of the range length and where the range is starting in relation to the marked lines. {C}The driver will turn the tractor and get into position to start the next pass and then stop and lower the planter.

Here is another tricky part.Take the next four envelopes from the fertilizer bin. The bottles should already be in the right places but the two trays on the left will be put on the right side of the planter and the two trays on the right will be put on the left side of the planter. Have in your mind the first range number line and how it would look from this end of the field. The plot#'s must go in the same direction as the number line. If the number line is moving from west to east, the plot numbers on the planter should also increase from west to east when looking from behind the planter at the plot#'s on the envelopes nearest the people. The plot#'s nearest you are planted last into the first range at the other end of the field.

Plant the second pass of the first round the same as the first pass but with corrections for seed placement. Check part way down the field how the seed placement is. Check for a few more passes and when it looks good, stop checking.

Electric Indexer (plastic trays)

The trays are shipped to the field in the order they are to be used. Each tray has one end marked with the tray number. The planter has storage space for a dozen trays or so.

Ensure that the tubes from the indexer holes to the heads are going to the correct rows. The tray is inserted with the metal covering on the bottom. This means that for a left to right first range, the seed for row 1 falls down the hole farthest from you and the seed for row 4 falls down the hole closest to you. For a right to left range, the seed for row 1 falls down the hole closest to you and the seed for row 4 falls down the hole farthest from you. {C}The cones on the planter are marked from behind from left to right as 1 on the left and 4 on the right. If the first range is from left to right, the seed for row 1 falls into cone #1. If the first range is from right to left the seed for row 1 falls into cone #4.

The first tray is inserted into the indexer with the numbered end first and slid from left to right. The metal covering will hit a metal ledge but the tray will keep sliding. Position the tray so that the first gap between cells is directly below the indexing bar that advances the tray one cell to the right each time the button is pressed. I don't remember if this is after the first set of four cells empties down the holes or if you must advance the tray to the holes. Either way advance the tray so that the first set of four cells empties into the heads below.

The indexer and the button that is pressed to activate it have been troublesome at times. If the tray is not positioned properly, the bar of the indexer will crush the cell and the indexer will have to be taken off with wrenches and repositioned.

It is very important to press the button firmly and confidently. If the button is pressed very quickly, the press may not register and the indexer won't advance the tray one cell to the right. If the button is pressed too slowly, the press may register as two presses and seed from 8 cells will be mixed together in the heads. One group of researchers solved this button problem by replacing the electrically operated button with a hand crank.For a field with lots of plots this might be too hard on a person's arm.

There is another button for the horn to notify the driver to stop if there are any problems. This is usually if the button doesn't work properly and the tray doesn't move or moves too far. Sometimes, the cones may not be working properly, but a disadvantage of using the automatic indexer is that the operator is too far up and too busy to see how things are going.

When the operator of the indexer is ready, the signal is given to the driver to pro seed or make it sew! Usually, the position of the marker arm on the planter in relation to the range marker line in the soil is used to determine when to press the button that advances the tray one set of cells to the right. Note that this position will have to change with the speed of the tractor.

At the other end of the field, the next tray is inserted with the number of the tray on the left. So the plots of the first range are planted last.

Often a field is not 18 ranges long so there are empty cells in a tray. Sometimes there are only a few trays with empty cells for the last few rounds of the planting.

If there are empty cells, the tray must be advanced to the proper place to begin planting. If the last range was completely filled, the tray can be advanced to the first cells with seed.

If the last range has empty plots, the first cells, possibly empty, are the ones of the last range. E.g. if the trays hold 18 ranges. The last range is 12 but there are empty plots on one side of the range. Position the tray so that 6 empty sets of cells have passed through. The next cell will be empty and correspond to range 12. Depending on the circumstances, you can advance the tray past the empty cells for range 12 and move the planter to range 11 before starting to plant. Note, however, that no fertilizer is being placed for the empty plots. Otherwise, start the planting for range 12 as for any other start from a field end,press the button at the proper time to move the tray one set of cells to the right, and prepare to plant the seed of the next set of cells at the proper time and place in the next range. This method, however, places fertilizer into the empty range and marks 4 rows.

When the entire Yield Trial Field has been planted place flags to mark the corners of the field.

Making a straight first row

Usually the first pass of four rows is made with a common hybrid in the planter buckets and then the buckets are taken off and replaced with the cones. These four rows are guard rows.

The basic method of planting a straight first row is to mark at both ends of the field and between, where the first row is going to be, and put up large white stakes where the center of the planter will be to guide the driver.

After marking where the first row should go, measure across to where the center of the planter should run and place a large white stake. Do this for a point further down the field and for a point at the other end of the field. Take a sight from behind the first stake to the last stake and correct the position of the middle stake if it is not in line.

The driver should make sure to put the planter marking disc down.

Checking seed depth and location

The seed depth and plot length should be checked on the headlands or some other place (where stray plants pose no problem) prior to planting the Yield Trial plots.

For the first few rounds of planting the Yield Trials the driver and the people on the planter should periodically check the placement of the seed. The driver must work together with the planters. Plant about 4 ranges and stop before planting the fifth range. First, check the seed depth in all rows and make adjustments to the planter if needed. Then check the location of the seed in several rows to find where the first seed dropped to where the last seed was planted for the range. The different colours of fungicides and insecticides on the seeds can help.The wooden stakes work well for exposing the soil to find the seed. Once one learns where to look on the planter track it is pretty straightforward. Remove the first inch of soil using the stake sideways and then use the point to find the narrow channel that the seed is planted in and slowly flick soil away until a seed is exposed. Be sure to replant and cover the seed. {C}Determining if the poor placement of seed is because the plot length is too long or too short or if it's because the seed is being dropped too early or too late is a bit tricky. If the seed was dropped after the range marking line and the last seed is several feet before the next range marking line, the plot length is too short. If the seed is dropped close to the range marking line and the last seed is well into the pathway between the next two range marking lines, the plot length is too long. If the seed is dropped a foot too soon and ends a foot short of the range marking line for that range, the plot length is perfect but the cue to drop is too soon. If the first seed found for the range is a foot past the range marking line and the last seed is a foot past the range marking line, the plot length is perfect but the cue to drop the seed is too late. {C}If the seed is dropping too soon or too late, adjust the point used for determining when to let seed drop.

If pressing the button at the proper time seems to be a big problem, there are systems one can get with wires that are laid on the ground and something on the wire trips the seed cups at the proper place.

The idea of plowing seed tape into the soil in a manner similar to the way subsurface drainage tubing is installed was briefly considered. The main objection that was raised at the time was that it was absolutely necessary that no seed ever come loose from the seed tape and get planted in the wrong area but it was commonly observed that seed glued to tapes often became dislodged. This assumes a circular tape roll and glues on paper tapes. Maybe there are other packing and gluing or attachment possibilities. Another concern raised was about the amount of tension on the seed tape that would cause the tape to break while planting. Note: There is a video of plowing seed tape into soil.Tape Seeder for Wheat

I always thought it would be neat to set up a laser and reflectors or prisms on each side of each range to trip the seed cups when the sensor on the planter broke a beam.

Maybe a B52 bombsight of some kind would help(?).Possibly use differential GPS and lasers???

Sound off

Sounding off is just announcing to the other planter what plot#(s) was just dropped into the cone. It is a double check on the plot #'s within a range. It is most useful in the Inbreeding nursery but it doesn't hurt to use once in a while when planting Yield trials. Do a sound off before starting at each end of the field to make sure the trays were properly placed on the planter and are being picked from properly. As long as the other plot# for a commercial yield trial is one higher or lower, the planters are planting the same range. If one of the plot#'s is not one higher or lower, an envelope may have been missed and not planted or planted in the wrong range. {C}For commercial yield trials, the colour of the envelopes can help more to check that both planters are planting the same range at the same time. One company may use blue envelopes and the other yellow. Both planters should start planting blue envelopes at the same time. At all times the planters should be planting the same colour envelopes except if there is a range holding plots for more than one company.

Comments on speed vs. growth

A speed of planting that is the same as an agricultural producer would use is ideal. However, proper placement of the seed is paramount. The driver of the planting should not go faster than the person(s) doing the planting on the planter. A plot that is 5 feet out of its proper range is no good to anyone. Nor is a plot with large 10" gaps because the planter went that far before the next seed fell. Going slow does seem to give a growing advantage to plots so the driver should try to drive at a consistent speed. It is better if all the plots are planted well than if all the plots are planted poorly. The person(s) on the planter should work as quickly as possible without making mistakes as they can. A speed of planting that is too slow can easily lead to more mistakes because someone has shifted their attention to something else while waiting to pour seed into the planter or to press the button to put the seed in the cones. Find a comfortable speed to work at. Also, the rotation of the cones on some planters is geared to one of the wheels of the planter. When travelling through light soil types or possibly muddy areas, the planter wheels might not rotate at the same rate as elsewhere and the plot might lengthen as a result.


If there is a malfunction of the cones, stop the planting. Seed from cones can be obtained by catching the seed below the planting tube as the cone is spun manually.

If the seed has different colour insecticides, it can be sorted out and planted again with the planter, or planted manually. If there is no way to distinguish what seed belongs where, the plot will have to be left empty and reseeded later with a check variety or possibly new seed obtained from the sponsoring company.

The location of a misplant should be noted so that people can be vigilant later regarding the possible existence of carry over plants in previous ranges.

Record any planting errors and their corrections as well as replantings (R), carry over (CO), transplants to (TT) and transplants from (TF), double or partially double plantings (D). Get the correction information typed up and put into the notebook as well as make the necessary corrections to the affected notebook pages.

Yield Trial Notebook



pasteboard landscape (12 in. x 9 in.) covers for 2 prong Acco bindings Notebook Colour (BLACK) Title Label on Front Cover Company name, address and contact particulars (Inside Front Cover) Days from Planting chart (Inside Back Cover)


Company name, address and contact particulars List of Yield Trials being conducted with locations, plot #’s, # of replicates etc.


Planting Plans New Designations and abbreviations Note Pages

Front Cover

Black Notebook covers and Front Cover Label The notebooks have been created using 2 prong ACCO bindings in landscape size. For ease of identification on the shelf, BLACK has been preserved for the YIELD TRIAL NOTEBOOK. If the colour is maintained from year to year, identification becomes simple. An easy to read label for the Notebook should be on the fron t cover. The label does not need to be fancy but background pictures etc. can be included. The main priority is that the letters are big and very visible such as black on white. The title should state at least the year and the general contents. E.g. Yield Trials 2001 or 2001 YIELD TRIALS. It could also be three lines for three Yield Trials e.g. 2001,Commercial Yield Trials, Our Company Yield Trials.

Inside Covers

Company name, address and contact particulars on a label applied to the Inside front cover.This information will be helpful if the notebook is lost. Days from Planting (inside back cover) Post a DAYS FROM PLANTING CHART on the inside of the back cover. The chart can be taped on with clear scotch tape. A plastic pocket might be a good idea to help waterproof it as well as being convenient. The chart should also be printed as a page and placed just before the actual yield trial records.

The DAYS FROM PLANTING CHART should have at least two or three columns. Column one could be the day of the month,column two could be the month, and column three could be the number of days from planting. It is easy to create the chart using a spreadsheet. The range needed is about 60 days to 80. Most hybrids have been in the 70's. It is nice to also include a statement of the traits being rated and the high/low range. (E.g. Vigour 1-5, 1 is best, 5 is worst)

Planted: May 15, 1975
Days from Planting
Date Month Days from Planting
15 May 0
16 May 1
17 May 2
20 July 67
21 July 68
22 July 69
23 July 70
24 July 71
25 July 72
26 July 73
27 July 74
28 July 75
29 July 76
30 July 77
31 July 78
1 Aug. 79
2 Aug. 80
3 Aug. 81

It may prove helpful to include the day of the week as well.

Table of Contents Page

This page should have the Company name, location and contact information as a header. It should list the Yield Trials being conducted. It might list them as locations and Company names with the row or plot numbers for each company test listed. It might also list tests by type and location with relevant plot#'s. A sample Table of Contents page is shown on page 42. The example is on a portrait page but the notebook pages are usually printed as landscape pages.

If a standardized field size is used for all Yield Trials for Generic Seeds, a listing could be made of the Field numbers, # of tests in each field and the type of material being tested. The word field is being used here to a portion of a physical field planted to a set number of rows with a set number of ranges.

Sample Table of Contents Page


123 Anystreet
Anytown, Anyprovince
T6J 0H5
Ph: (403) 555- 5555
STANDARD TRIALS: 1 rep/location
Location : TOWN1 TOWN2 TOWN3
Plot #’s : 1-3407 5001-8407 10001-13407
Standard: P3984

ADVANCED LINE TRIALS: 2 reps/location
Location: TOWN1 TOWN2 TOWN3
Plot #’s : 20001-22100 25001-27100 3001-32100
Standard 1: P3979
Standard 2: P3967
Standard 3: KO30

FLUNKS TRIAL: 3 reps/location
Location: TOWN4 TOWN2
Plot #’s : 40001-40060 50001-50060

Location: TOWN4 TOWN2
Plot #’s : 41001-41060 51001-51060

HIGH LANDS TRIAL: 2 reps/location
Location: TOWN4 TOWN2
Plot #’s : 101-140 101-140

201-240 201-240

There should be a page summarizing the characteristics to be observed and rated as well as the rating scale.
Vigour: 1 to 5. (1is best rating. 5 is worst rating)
BRK (Breakage): 0 to 60. The number of stalks in the plot that are broken below the ear.

Planting Plans

This page should show the width of the field(s) in plot#'s, the number of ranges, direction of increasing plot#'s in a range, the presence and number of guard rows on each side of the field(s), names of fields by Company, location or type and information regarding directions if needed such as North, South, East or West.

If the Notebook does not contain a page showing the planting plan, this makes it impossible for the person working on the reports to match Harvest numbers to Plot numbers. The existence and location of filler plots should also be stated on the planting plan.

Note on the planting plan or harvest data sheets if there is any change in the number of ranges or plots/range in any part of the field prior to harvesting the plots. There is a possible problem if part of the last range contains filler plots that are harvested as a block and data for those plots is not recorded. This changes the number of harvest plots for which data will be obtained and changes the number of ranges harvested on one side of the field. E.g. Last range on left side was 19 but first range on right side was not 19 but 18. This means the 20th harvest# was not for the plot in range 19 on the right side but for the plot in range 18 which normally would have been harvest # 21. The person matching plot#'s to harvest #'s needs to know if the plots in range 19 were skipped. If the person doesn’t have this information it puts the matching out of order and the number of plots in the field won’t match the number of plots harvested.

New Designations and Abbreviations

This page is needed for In-house Generic Seeds Yield Trials. It is for the Breeder's reference purposes. The breeeder may not be familiar with a new abbreviation or name for a new variety or may desire to know more about the pedigree derivation of a variety. I don’t know much about how this was done. New crosses were given shortened names that contained some reference to the parental names. E.g. M17 = CM7 x MO17 or SK380 = SK02 x SK16 x CM380

At S5, Inbreds were considered pure and assigned short names of about 5 or 6 letters and numbers. The inbreds were grouped by pedigree and the new names for a family of inbreds or a group of inbred sibs from a cross would contain a sequential number given to each of the related lines.

When giving names you should be aware of the abbreviations used by the Federal Government Research Stations and U.S. Universities and Government Researchers. Canadian government designations start with C and the first letter of the station’s location followed by a number. CB = Canada Brandon, CM = Canada Morden, CL=Canada Lethbridge, CO = Canada Ottawa, CG=Canada Guelph etc.

PLANTING PLANS EXAMPLE PAGE (entire physical field)


My predecessor didn’t give me any information regarding the system that was used for creating and recording new designations or even hybrid numbers. I’ve only seen a few of the old lists from the early 1990s in several different arrangements. One thing I don’t know is why some of the new designations don’t include a full sequential list as one or two numbers in numerical sequences are missing. E.g. CSK1, CSK2, CSK4, CSK5, CSK6, CSK7, CSK8, and CSK10 to CSK25 is missing CSK3 and CSK9.

Most designations were not more than 5 alphanumeric characters.

When creating a new designation you want to keep some of the information from the pedigree in the new desigation if possible.

Basically you just make a list with the new name on the left and all the parental pedigree or long pedigree information on the left. If it’s a cross put the female pedigree first followed by the male pedigree. {C}INSERT SPREADSHEET 9

The crosses I made new designations for were often impure and quite messy. The pedigrees were listed completely and many inbreds were past S5 when they should have been given a shortened pure designation. Often the pedigrees involved multiple crosses on both the male and female sides. Sometimes it’s easier to show the pedigree vertically with the male pedigree on the top and the female pedigree on the bottom.


In 1999, a cross was made involving a pure Leaming family female inbred that was a three-way cross selfed to S7 and a Lancaster family male (seed from 1998 and 1997 and a sib row) that was a cross.Both should have been renamed earlier.

New Name: LEAKEY?? A.K.A. Hybrid 996574
Pedigree: ((LEAM*KE372*KE403)-(1-7)-3-2-1-1-2-1) * (CB18*P651)

male seed from 98-8676,98-8746, and 97-441-(1-8) {C}Certainly a mongrel cross!

Derivation? (LEAM x KE372 x KE403) x (CB18 x P651)
NOTE: I have used the asterisk for an x in the pedigrees to avoid confusion with the letter x.Unfortunately, this use of the asterisk conflicts with Purdy notation in which the asterisk is used to denote a backcross.


New Designation, Old Designation, A.K.A.

SO you need some kind of system for keeping track of the new names and you post a listing near the front of the Yield Trial Notebook.

{C}Format and Creation of Yield Trial Notebook page {C}The NOTEBOOK is printed as landscape pages. That is an 11 in. x 8.5 in. page. It is longer across the page than down.

The first line of the page header should contain the year of the trial, the name of the trial, the location of the trial, and the date of planting. The date of harvest can be shown on the summary report in the fall.

The second line contains the column headings for the records. These may change somewhat for different trials.

Ideally, each replicate would have 25 entries and all of these would be held in one range. Each range would therefore have 25 plots to be evaluated. A notebook page would have 25 records so the page would not have to be turned until notes on a range of plots is finished.

Often, however, the Commercial trials are of various sizes and a range may be 36 or some other number of plots wide and more than one page. {C}Sometimes it is possible to put up to three replicates on a single page. It depends on the space needed for writing numbers and the size of the pedigrees in the ENTRY column.

A replicate may also be split into two ranges with a page in the notebook for each range. E.g. 60 Entries, 30 entries per range. Note that if the next trial is 36 plots wide, the 6 filler plots in each range of the first trial are not listed and they form a block on one side rather than being interspersed with the other plots.

The columns for data entry should be a sufficient size in width and height for entering two digit numbers with a pencil by most people, not just small letter writers. This must be a priority since the purpose of the Notebook is to collect those numbers as they are observed.

The column headings would be PLOT#, ENTRY, ENTRY# (Usually I just put # to keep the column width small), ENTRY, VIG, TAS, SLK, BRK, POP, and one or two blank columns for possible extra data columns.

The ENTRY column is shown twice because the notebook on the computer should hide the second ENTRY column data with the variety names or pedigrees and print only the first ENTRY column data. The first entry column data contains only the pedigrees of check varieties. The second ENTRY column contains the names or pedigrees of all the varieties in the test. Note that since the ENTRY column with all the pedigrees is hidden and not printed, there is more available space for columns on the printed page.

The idea behind hiding the second ENTRYcolumn is to blind the observer as to the identity of the material in the trial. The first entry column showing only the check varieties really shouldn’t be necessary and would not be used but the breeder wanted to know which plots the check varieties were in.. {C}Place the observation columns in the order that data will be enterered into them. This helps in getting the data in the correct columns.

VIG is short for VIGOR, TAS- Days to 50% Tassel shed, SLK - Days to 50% of the plants silking, BRK - Number of plants with breakage of the stalk below the ear, POP - Plant Population of the plot. {C}Some other columns used have been an EMERGENCE column before the VIGOR column. It is for the number of plants that emerged after planting. Stalk Lodging, Goose neck, COLUMN (Stalk Breakage above and below the ear), Root Lodging, etc.

The Yield Trial Notebook page should have double lines or a thick line below every 5th record. This isn't really necessary for Yield Trials but can help in keeping one's place when entering data in the field and at the office. It is more important for the Inbreeding nursery.

Use dashed lines rather than solid lines. Dashed lines are easier on the eyes and on ink usage.

Colour formatting is nice if one has the time.

Page numbering would be useful except there are frequent changes to the notebook. It isn’t necessary.

Sectional page numbering might be an option since changes in other sections would have no effect. E.g. Section A Pages A100-A151,Section B Pages B100-B125.

You can create Excel Styles and Templates for the notebook pages.


For the notebook, each replicate had to be sorted by Plot# with the Entry# and Entry name. The data for each replicate had to be listed below the previous replicate rather than beside it. Later more columns will be added for entering observed data. For Commercial Yield Trials, Entry Names and especially Entry pedigrees as Entry Names would be entered and stored but hidden and not printed out. The column was hidden by reducing its width to 0. By hiding the Entry names one is not changing an observation to match a preconceived idea of what the data should be or giving a higher grade based on the pedigree.

1993 Yield Trial
Replicate 1
5101 1 CIBA#1
5102 2 CIBA#2
5103 3 CIBA#3
ETC. (Plots (5101 to 5125))
Next Notebook page
1993 Yield Trial
5126 1 CIBA#1
5127 24 P3984
5128 25 P3979
Etc. (Plots 5126 to 5150) for this test

P3984 and P3979 were the best hybrids for the area {C}at the time and were used as the checks

1993 Yield Trial
5151 10 CIBA#10
5152 21 CIBA#21
5153 6 CIBA#6
Etc. (Plots 5151 to 5175)

The use of PLOT#'s in sequential order rather than ROW#'s is preferred. Use of Plot#'s halves the number of records to be printed or stored. If Row#'s are used, they can be listed as the set of odd numbers (E.g. 1,3,5,7,9 ...). Don’t get confused between rows and plots and record data for each row of the plot. There may be confusion about what one means by the term ‘row’ so see the Terminology page that explains how I am using the term.

For the commercial yield trials, a source row column is not needed. This information is usually not released. For Generic Seeds material, the Source row number from a Top Cross field could be used as a unique hybrid # identifier.

If a Yield Trial is 20 plots (40 rows) wide. Print only 20 lines per page for data from one range of 20 plots. If ROW#'s are used this would be 1,3,5,7,...39 rather than PLOT#'s 1 to 20. Print one range per page. Either way is just as easy using Excel’s datafill command but I prefer the use of Plot #’s.

Single page example

Page 47 is a sample Notebook page as it would be printed. It is a landscape page (11 cm x 8.5 cm)

{C}Be sure to leave a proper amount of margin on the left so the printing is not covered up by the Notebook Cover center binding part.

The first 2 rows on the page have a yellow background and no lines. This isn’t really necessary but looks nice. Black and white is fine.

Row 3 containing the Column headings has a grey background and grid lines. Again this is just window dressing. A dark grey isn’t a good idea as it makes it hard to see the black letters.

Rows 1 to 3 are header rows except for the replicate designation. They can be designated as spreadsheet title rows for printing or created separately as a page header.

Perforated grid lines are easier on the eyes. It also helps to keep track of ones place if the grid lines beneath every fifth plot are doubled or thickened.

(E.g. Below Plots 12005, 12010, 12015, 12020, and 12025)


Computer version of theNotebook Page

SINGLE PAGE AS PRINTED (Oops! Forgot the thick line every fifth plot! {C}

A Printed Notebook Page


Three replicates on one page printed Notebook Version



I am told that thinning is not necessary if one uses a vacuum planter. Generic Seeds would like to acquire one but the cost is prohibitive so a cone planter is being used.


  • sharp triangular hoes for the thinning work
  • metal file
  • chopping hoes or triangular hoes for cutting out ranges
  • tape measure
  • hats
  • sunblock
  • water (no ice!)
  • Yield Trial Planting Plan at least and if possible, Yield Trial Notebook.

Avoid Heatstroke !

Everyone must have a hat. Hats that also protect the neck from sunburn are best. Wide brimmed straw hats are preferred by many workers.
Protection from the sun by wearing proper clothing and/or sunblock is necessary to avoid sunburns. There is a safety issue regarding footwear and sharp hoes that needs to be addressed but most people seem to avoid accidents while wearing sneakers, sandals and even bare feet. Arrangements might be made to work early in the morning and in the evening when it is cool to avoid problems with the afternoon heat.Water brought to the field should never be ice cold.

Emergence counts

Do any Emergence counts first, if required, before chopping out the ranges with the chopping hoes. Triangular hoes work just as well as chopping hoes for cutting out ranges. The chopping hoes have a blade that looks like a slice of North American bread loaf or an upside down small case letter m.

Once all the emergence data is recorded the thinning can begin.

Thinning Procedures

The thinning is done when the corn plants are about 8 inches high or so. They come up to about half way from the ground to one’s knees. Check this and correct this line.(I think they were shorter)

Triangular hoes are the best to use. Have a file on hand to sharpen all hoes and to keep them sharp.

Before entering the field, know exactly how many plants are to be in each plot for each company. The usual plant population used has been about 30,000 ppa (60 plants for a plot 1/500 of an acre in size. Pioneer hybrids are recommended for planting at about 24,500 ppa and Novartis liked that plant population as well. The KWS varieties were bred for around 44,000 ppa.

Assign two people to work together to cut out the ranges ahead of the others.

Cut a straight line along the range marker line at the front of the field and chop out plants on the first pathway. Go up to the front of the next range and cut out a 3 foot pathway that straddles the plants of the first and second ranges. Use the plants as a guide as to where the path should be and not the path marks in the soil. Find a compromise between chopping out a lot of plants on each side, and going around rows that stick out. If a planter pass ends or starts too early or late and is way out, consider going around the end of that 4 rows or only chopping a few plants out so the plots will have enough plants. Chop an equal distance from each side rather than chopping along the end of the first range only. It is better to make a 2 foot pathway first, rather than chopping out a 3 foot pathway along the path marker lines in the soil and not having enough plants in a plot. Use a tape measure once in a while to check the plot length.

Plant Count

Each row is counted and then thinned to the proper number of plants. The best way to count is by holding the hoe vertically at one's side along the side of an arm and placing the tip over the whorl of a plant as one counts it. (The whorl is the central inverted cone of the plant). Pay special attention to clusters of plants to get an accurate count. Look for plants on the other side of the counted plant.Use the hoe tip to move the plants to see better. If you lose count. Start over. Do not guess. Subtract the desired row count from the actual row count and remove the difference.

E.g. If the plant count is 42 and the desired row count is 30, then remove 12 plants.

The ideal scene is an evenly spaced row.

First, thin clusters of plants down to one plant. If there is a large space near a cluster of two plants, let the two plants compete with each other rather than let one plant get food meant for several plants. Next, thin out plants that are too close together. Aim for a more even spacing. Knowing which plant to remove can be a bit tricky at times but one improves with experience.

Plants should be removed by the roots rather than chopped off. Place the uprooted plant in the center of the row. Do not fling plants around the field or into the next plot. Removal of the plant with the roots prevents regrowth and the plant should die in the sun in the middle of the row.

Use the tip of the hoe to get behind and a bit under a plant, bite into the plant with the hoe or get the hoe just beneath it and pull it toward you. Hopefully, the plant will be uprooted. If the plant is cut off and the roots remain, get the remaining plant and roots removed and placed in the middle of the row.

If the plant is part of a pair of plants and you want to remove the plant farthest from you, get the hoe tip in front of that plant and push the plant out, making sure the roots come out too.

Weeds can be hoed out of the row a few inches on each side of the row and between plants at the same time as thinning. Be careful not to cut off plants by accident after thinning a row. When weeding, the hoe should be run shallow below the soil to cut the roots of the weeds. If the hoe is run deep, the hoer just gets tired faster. Weeds can also be killed by covering them with soil so soil can be pushed or pulled up along the row as well. Soil around a plant can also help support it if a bit of a hole was dug next to it getting roots of another plant out. Fill any gouges made in the soil.

It is best if the group works together to finish one range at a time. A person can start at each side of a range and work towards the middle. The supervisor/leadhand should set a good example of workmanship but ensuring that things are being done properly is his/her top priority. Working together with the group rather than standing around is best for morale. Checks can be made every so often as the work progresses.

The supervisor should discourage a lot of banter in the field as it can lead to mistakes in counting. Conversation not directly applicable to the work should be saved for breaks.

Staking the Plots

Stakes may already have been placed on the sides of the YIELD TRIAL field as per the planting plan and may need to be moved to a better position in front of the first plant of a row. The stake should be placed on an angle between 45 and 60 degrees and far enough ahead of the first plant that it is not going to interfere with the growth of that plant.

The stakes are about 8 in. x 1 in. x 1/4 in.. Stakes painted white are nice but not necessary. The plot numbers can be put on at the office using black Sharpie markers. Test any new permanent marker to be used on stakes in the field before using it to mark plot numbers. Sharpie markers almost seem to create a tattoo in the wood so the numbers stay on after a lot of weathering.

Lay the stakes for one side of the field out on a table with the pointed end on the right and the square end on the left and write the plot numbers on from left to right as per the planting plan and bundle these stakes together from first range to last range. Do all the plot numbers for the other side of the field and bundle them from last range to first range. Do plot numbers between the two sides. Decide if going to use stakes at every fifth plot# or every fifth file/paper column/field row. For every fifth plot, the stakes will be staggered in the field and if the range width is not a multiple of five then there might be stakes beside each other on the sides.. For every fifth file/paper column/field row, the stake in the next range is up the center of the two rows but the first row of the plot in the next range is on the other side of the next range plot because of the serpentine direction of the ranges.

Staking by files(paper columns/field rows is best to work with in the field, but makes numbering the stakes difficult if not worked out beforehand.

Put the stakes in for one side moving from first range to last and then put the other side stakes in moving from last to first range. Put in the stakes between the sides.

Vigour Notes

Notes on vigour can be taken once the plots are thinned and staked. The vigour rating is based on how strong and healthy the plants look. A plot of skinny yellow plants with lots of gaps where plants did not emerge would get the worst rating while one with dark green, strong, tall plants would get the best rating. The rating is 1 for the best and 5 for the worst.


There should be only one set of data for each 2-row plot. Don’t record data for each single row. If one row of a two-row plot is earlier than another, consult the breeder as to how to rate the plot. Record to this file what you are instructed to do. If one row silks on day 70 and the other on day 72, you might record this as silking on day 71.

50% Anthesis - I may be wrong about this, but I believe the 50% refers to how much of the tassel is shedding anthers rather than the percentage of plants shedding pollen. You will be looking at about 60 tassels very quickly for each plot so if the majority of them are shedding anthers and have anthers hanging from the branches of the tassel, that should be good enough.

50% Silking (mid silk) - I’m not sure about this one either. Basically, if you have a few inches of silk showing on the majority of the plants in the plot that should be good enough to mark down. Sometimes, plants on one side of the plot will silk first for some reason and I’m not sure how you should rate the entire plot but the other side will pop out withinna day so you can split the difference and use the average.

It’s best to enter data into the computer as soon as possible after it’s recorded in the notebook. If you can use a pocket pc in the field, it would not be necessary to print a notebook or enter the numbers manually into the office computer.

In EXCEL97 one can set the cursor to move down one cell after entering a number into a cell and pressing return. So one can enter the data from the notebook straight down the page.

In VP-Planner there were some quirks due to the lineless nature of the spreasheets, the filling of cells with dashes to make notebook lines and the lack of movement of the cursor in the down direction except by pressing the down arrow. The data was all entered at harvest time after sorting the spreadsheet rows by plot number to get all the notebook lines together and deleted so they were out of the way. Then one could enter the data without having to press the arrow key several times to move the cursor past the lines.

A VP-PLANNER macro was created based on an example in the VP-Planner book that allowed me to enter the data in one area on the spreadsheet and avoid using the cursor. This macro transferred the entered data to columns on the right hand side.


Tagging Plots

Yield trial plots were tagged only for showing the plots to company representatives. This was usually in mid to late September. The tagging was done about a few days before a visit. The tags were 1 in. x 11 in. strips of heavy construction paper (>24 lbs) on which the Plot #'s and Company given variety name or pedigree were written neatly with a black SHARPIE permanent ink marker. One year the information was printed on labels and the labels put on the tags. With the serpentine range pattern, the plot #'s and information will be printed on different ends of the tag depending on the direction of the range. The tag was wrapped around the stalk at around 5 or 6 feet high and stapled tightly to the stalk. (Consider a looser stapling to allow for growth if the tags are placed earlier and will be on for a long time). Place the tag below the first leaf rather than around the base of the tassel since the tassel can break off taking the tag with it. If the tag sticks out too much it will bend, droop, or crease etc. and the Plot #'s will not be visible. {C}A local Government Research station used waterproof plastic pockets into which a piece of paper was placed with the plot # and any other desired info printed on it. The tag was attached to the stalk loosely with the attached nylon plastic string and fastener.

Notes Taken Prior to Harvest

LODGING - There are several different types of lodging. If the plant has fallen over from the roots, it’s a case of root lodging. If the plant bent over and fell from some part of the stalk, it’s stalk lodging. You might be asked to record each type of lodging separately or together. For these notes record all lodging together. Lodging of any kind is a bad trait.

BREAKAGE - This can have more than one meaning depending on where the plant has broken. Some plants can break above the ear and this isn’t considered as bad as breakage below the ear. Generic Seeds counts the number of plants in the plot that have broken below the ear.

POPULATION - This should be whatever you thinned the plot to. If you chopped out too many plants when thinning or the seed did not germinate well, there may be a difference and it should be recorded in the notebook ASAP. If you have to, you can count the number of plants in the plots again just prior to harvest.

Information needed for compilation of Yield and Statistics

  1. Record the Plot# corresponding to Harvest sample #1. If possible, standardize the first plot to be harvested as PLOT #1 (Rows 1 and 2).
  2. Record the direction of travel of the combine. Was the field harvested in a clockwise or counterclockwise direction?
  3. Were there any deviations in the harvest pattern such as filler plots in the last range being removed prior to the harvesting of the rest of the field? Did the combine start at a different range on one side of the field?
  4. Specify the plot length.
  5. Specify the row width (E.g. 22",30",36")
  6. Every plot number should have an Entry # that can be used as a key for grouping the material to do averages and the one-way Analysis of Variance (ANOVA).

Yield Trial Harvest

Mechanical - Combine

A 3 row K2 Gleaner that was converted to work with plots, was used to harvest the Yield Trials. The conversion consisted of diverting the seed harvested from a plot up to a bin where it was weighed, moisture recorded, possibly a sample taken from the scale dump and the dumped corn augered into the large combine harvest bin. {C}The Combine driver would harvest a plot and then stop. The grain would come up into the bin and the weight and moisture shown on a monitor and were recorded. He would wait until all the seed from the plot was in the plot bin by listening to the sound and noting how fast the scale reading was changing. When the sound was loud and like the sound of the last few kernels of popcorn bouncing around and the scale started to go down, he would press a button to record the data in the electronic data collection system and then press another button to dump the scale. Then the next plot was harvested. {C}The electronic data collection system recorded the data on a paper tape similar to a cash register tape as well as to an electronic memory. The memory was downloaded into a PDA and transferred to the office computer. The advantage of this system is the time saved from having to type all the data into a computer as well as avoiding transcription errors. {C}Unfortunately, the circuit board developed a crack that could not be fixed because the company went out of business and the research company no longer could afford to buy a new one. {C}After the electronic data collection system failed, a manual system was used. A second person sat in the combine and recorded the weight and moisture on a data collection sheet with preprinted Harvest #'s and 2 blank columns for the weight and moisture. This person also placed a Harvest # tag on the top of the sample pulled out by the combine operator. The Harvest# tag was prepared before starting. It was just a cardboard tag about 4 in. x 1 in. with a number stamped on one end and a hole punched in the other end. The numbers were from 1 to the number of plots being harvested. The harvest tags were strung in order through the punched hole onto the bottom wire of a coat hanger that was hung in the combine. The numbers face the data recorder. The tag was ripped off the hanger and put on top of the sample just before the combine operator put the sample into the small red mesh bag. The sample was obtained by placing a pan below the scale and then dumping the scale. This sample was put into a red mesh bag together with a harvest # tag. Every round the samples were dropped off and a second moisture test was done on each sample as a check using a portable moisture tester or the Dickey John moisture tester. The reason for the double check is that the moisture tester on the combine is most accurate only for one of three possible settings and is usually set on the middle setting. This might be a specific range of say, 20% to 30% moisture. If the sample is above or below that range the recorded value may be incorrect. The Harvest# data sheet should be formatted with half a round or one full round per page so the page can be given to the person doing the second moisture test together with the samples. There should be a second Moisture column to record the sample moisture taken with the portable tester or Dickey John moisture tester. Try to order the columns in the order that they will be filled. If the monitor in the com bine is read from top to bottom, list the moisture then the weight. If it is easier to remember the weight first, then put the weight column before the moisture column. 2 half rounds on one page help make sure things are in order since the bottom of the page is the end of a half round (last or first range). Sometimes wet corn would get stuck coming into the bin and mix with corn from other plots changing their moisture reading. The wet corn had to be cleared from the top of the bin and dumped before a new plot was harvested. So design fields to avoid this problem when possible by putting late material in a separate field.

A video showing a Wintersteiger plot combine harvesting two rows of a plot can be found on YouTube here:

( )

Pattern of Harvest (Combine)

The combine operator first removes the headlands and guard rows. Then work begins from plot #1 whenever possible.

The combine operator harvests plot#1 then waits for the seed to come into the plot bin and be recorded and possibly sampled. Then the driver moves ahead to the next plot that will be in the next range behind the range of plot#1. The combine operator continues to move in the same direction until reaching the headlands at the other end of the field. Then the operator begins harvest again on the other side of the field going in the opposite direction. If the combine operator always starts at PLOT#1, then the match of PLOT#'s to Harvest#'s will be the same regardless of whether PLOT#1 is on the right or left.

Manual Harvest

If the location is small, or too far away to make transporting the combine there profitable, a manual method of harvest can be used. A data collection station is set up midway along the length of the field on one side. Here the preharvest data is recorded and cobs from a plot are put into a bucket and weighed and a sample of seed is sliced offf the cob with a plane and a moisture sample taken.

Preharvest counts are taken for the plant population, # of broken stalks, lodging etc.

For the harvesting, the workers are given belts with hooks on the left and right sides. Small mesh bags are hung on the hooks. Cobs from each plot are put into the mesh bag(s) and picked up by another person who takes them to the data collection place. Tags are put into the mesh bags with the PLOT# or the plots are done in numerical order one by one. The cobs are weighed with a simple spring scale and a sample of seed is planed from the cob for a moisture test of the seed.

You might find some of Robert W. Jugenheimer’s notes in CORN: IMPROVEMENT, SEED PRODUCTION, and USES to be of interest regarding manual harvest. It involved a large amount of number crunching.

Getting the Data Together in the Computer

If you have an automated data collection system, getting the harvest data into the computer is probably not a problem. Otherwise, you need to get all the Harvest data sheets for the Yield trial field together and get the data entered manually into the computer. I entered the harvest data and then used a match of harvest numbers to plot numbers to match the harvest data to the notebook data (see Linking the Harvest Number and Plot Number below).

First, I entered only the second moisture number and the plot weight. Usually I entered all of one trait at a time going down the sheet rather than 2 traits per harvest number. Then I added a list that matches the harvest numbers to plot numbers to the left of the harvest data. Keep all the data per harvest number together in one row in as many columns as observed traits, then sort by plot number. Then I added the harvest data to the notebook data. Lastly, I added the yield column and inserted a formula into the cells of the yield column to compute the yields. For in-house tests one could also add a column for the performance index or any other index you want to use to sort by.

Linking the Harvest Number and Plot Number

Manual Method (tedious so avoid if possible)

The Yield Trial plots were not harvested in Plot Number order. This saved wear and tear on the combine's expensive clutch and avoided excessive movement backing up and moving forward for each plot.
The headlands, side guard rows and last range filler plots were removed prior to the start of harvest/data collection.
The combine started from one corner of the field and moved straight up through the ranges to the other end. Then the combine moved to the other side of the field and harvested the plots going through the ranges in reverse order.
The combine then moved to the other side in an overall kind of circular pattern.
Usually, the first plot was Plot #1 but there could be access problems that required starting from a different corner.
Be sure to note down exactly where the combine started and if any ranges or filler plots were removed prior to data collection.
I used the Planting Plan and Excel's Datafill/Series Fill to match the Plot numbers to the Harvest numbers.
  • Using Excel's datafill/sequence completion feature, expand the Planting Plan across a worksheet to show all plot numbers using datafill.Note that plot numbering might not be sequential for all ranges.
  • Use a macro or manually add a blank column between each column of plot numbers.
  • Pretend to be the combine and use datafill/sequence completion to fill the empty columns to the left of the plot numbers with the Harvest numbers in their proper sequence going up on one side and down on another.Note that there might be filler plots removed prior to data collection and alter harvest numbering appropriately.
  • Strip off (cut) the Harvest and Plot numbers in pairs to create a two column list of Harvest and plot numbers down the worksheet.
  • Attach Harvest number data to this list
  • Sort the list by Plot number
  • Use the sorted list to match Harvest Number and Plot Number data.

Automated method

This would involve use of a small program manipulating arrays, a mathematical equation or an algorithm. Possibly EXCEL's vlookup could be used. One could easily create a Plot Number array and a Harvest Number array and then find matches in each array corresponding to the same coordinates. Arrays use RAM so if one had a HUGE array, one might use the algorithm instead to calculate matches. A graphical output of the arrays would be very desirable as well. First one could output just the arrays of numbers in a 2D form and then later as rectangular numbered plots.
Currently I only have crude algorithms for a field that has Plot #1 in the lower left corner and the combine starts at Plot #1.It can't handle cutting out filler plots yet but this would probably be done using math from computer graphics and maybe transformation matrices and group theory.
  • To get a list of the Plot-Harvest pairs would just involve looping through the equations and printing the results to a worksheet or a text file.
  • Given the dimensions of the field/block, find the matching Harvest number given a Plot Number.
Columns is the width of the field/block in terms of the number of plots wide (left to right). Note: Need a better name
Ranges is the number of Ranges.
PlotNumber is the Plot Number for which the matching Harvest Number is desired
Plotnum_x is the column number or x-coordinate of the Plot Number.
Plotnum_y is the range number or y-coordinate of the Plot Number. Note: if you name the variable Ploty in Python there is a name conflict.
HarvestNumber is the matching Harvest Number in the Harvest Number Array.
Obtain the values for Columns, Ranges and PlotNumber from the user with an input prompt.
Plotnumy = Int ((PlotNumber - 1)/Columns) + 1
if Plotnumy mod 2 = 1, then // the plots in the range are numbered left to right for odd numbered ranges
Plotnumx = PlotNumber - Columns*(Plotnumy - 1)
Plotnumx = ((PlotNumber - 1) mod Columns) + 1
if Plotnumy mod 2 =0, then // the plots in the range are numbered right to left for even numbered ranges
Plotnumx = Columns * Plotnumy + 1 - PlotNumber
Plotnumx = Columns - ((PlotNumber - 1) mod Columns + 1)
If Plotnumx <= int((Columns + 1)/2), then // left of up/down divide/split. Harvested UP
HarvestNumber = 2*Ranges*(Plotnumx - 1) + Plotnumy
else: // right of split (int(Columns +1)/2)), harvesting DOWN ranges.
Harvest# = 2*Ranges((Columns - Plotnumx + 1)) - Plotnumy + 1
  • E.g. Given Columns = 5, Ranges = 6, PlotNumber = 21, find the Harvest Number
Plotnumy= Int((PlotNumber - 1)/Columns) + 1
Plotnumy = Int((21 - 1)/5) + 1
Plotnumy = 5
Plotnumy mod 2 = 1
Plotnumx = ((PlotNumber  - 1) mod Columns) + 1
Plotnumx =(21 - 1) mod 5 + 1
Plotnumx = 1
((Columns+1)/2) = 3
Plotnumx < 3
HarvestNumber = 2*Ranges*Plotnumx - 2*Ranges + Plotnumy
HarvestNumber = 2*6*1 - 2*6 + 5
HarvestNumber = 5
  • Find the Plot Number given the Harvest Number.
Columns is the number of plots wide.
Ranges is the number of Ranges. (Numbered from bottom to top of page)
HarvestNumber is a Harvest Number
Harvestnumx is the column number or x-coordinate of the Harvest Number.
Harvestnumy is the range number or y-coordinate of the Harvest Number.
The Column Index (n) is the sequential order of a column in the harvest sequence.
Harvestnumx is the actual positional order of the column from left to right.
n = (Int (Harvest# - 1)/R) +1
if n mod 2 = 1, then // this is a column that was harvested UP the ranges from bottom to top of the page (ODD)
Harvestnumx = (n+1)/2
Harvestnumy = [(HarvestNumber - 1) mod Ranges] + 1
If n mod 2 = 0, then // this is a column on the right side of the up/down split so it is an even number and was harvested DOWN the ranges
Harvestnumx = Columns + 1 - n/2
Harvestnumy = Ranges - [(HarvestNumber - 1) mod Ranges]
if Harvestnumy mod 2 = 1, then
PlotNumber = Columns * ( Harvestnumy - 1) + Harvestnumx
if Harvestnumy mod 2 = 0, then
PlotNumber = Columns*Harvestnumy + 1 - Harvestnumx

Yield Calculation

The company has reported yield results in both kg/ha and bu/ac. The basic formula is just to calculate the amount of corn there would be if the sample was dried to 15.5% and then to keep the dimensions correct, divide by the area of the plot. (Since division is the inverse of multiplication, multiply by the inverse of the plot size in the units of area used. ) Two formulas have been used. I have used a dried moisture of 15.5% here but due to the rounding feature in VP-Planner, a value of 15 will give the same results. Formula 1 (kg/ha)

Dryweight in kg / plot size in ha
Dryweight * inverse of plotsize in ha.

For a plot that is 1/1000 of a hectare:

(Plotwt - (Plotwt * ((Moist -15.5)/100)))*1000

Plotwt is the Plotweight measured in kg. Moist is the recorded percentage of moisture. E.g. 21% entered as 21. The factor of 1000 is the inverse of the plot size measured in ha. The plot was 1/1000 ha (.001 ha).

Plotsize of 1/1000 ha, Moisture:18.9%,Plotwt: 7.95 kg

(7.95 - (7.95*((18.9 - 15.5)/100)))*1000
= (7.95 - (7.95*(.034)))*1000
= (7.95 - (.27))*1000
= 7.68 * 1000
= 7680.0 kg/ha

NOTE : If the plot size is changed, the area factor must be changed to reflect this.

Formula 1 can be used for bu/ac with proper conversion factors. The weight has usually been recorded in kg (because there are fewer digits to record) and then converted to lbs in a separate spreadsheet column.

A 1/1000 ha plot for 30 in. rows

Row width = .76 m (30 in.)
Plot length = 6.6 m
2-row plot = plot width of 2 * .76m = 1.52 m
Plot area = 6.6m x 1.52m = approximately 10 square meters
1 hectare = 10,000 square meters
10 square meters = 1/1000 hectare

A plot of two rows spaced .76m (30") apart and about 6.6 m long, has an area of about 1/1000 of a hectare.

Formula 2 (bu/ac)

This formula is from pg. 43 of the GREAT LAKES HYBRIDS NOTEBOOK
GLH is an American company so the size of a bushel is different than in Canada.

This formula uses a conversion factor for SHELLED CORN of 110.465

(100 - moisture) x (110.465) x (lbs of grain) / (row length in feet) / (row width in inches) / (number of rows)

E.g. 2-row plot,Row width: 30 in., Row length: 17.4 ft, moisture: 18.9% {C}plotwt: 7.95 kg = 17.49 lbs

(100 - 18.9)*110.465*17.49/17.4/30/2 {C}= 150.1 bu/ac

If nothing was ever changed from year to year you could use

(100-moist)*lbs of grain*0.1058

but this formula tells you nothing about where 0.1058 came from, what the plot size was or what to do if the plot size is changed. It mixes numbers used to get the dry weight in with numbers used for the plot size. It also assumes the plotwt is recorded in lbs. It's just 110.465/17.4/30/2 = 0.1058

The best way to make conversions is to multiply by a factor of 1 and cancel dimensions. This can be done for chains as long as needed.

E.g. Convert 30 in. to meters

30 in x (2.54 cm/1 in.) x (1 m/100 cm) = .762 m

{C}Converting kg and lbs

1 kg = 2.2 lbs
1 lb = .454 kg

Converting Yield

1 kilogram/hectare (kg/ha) = .8922 (.9) pounds/acre
1 pound/acre = 1.121 (1.1) kilograms/hectare


(bu/acre x test weight(industry standard is 56lb/bu))/0.892=kg/ha

Converting hectares and acres

1 hectare = 2.5 acres
1 acre = .4 hectares

Converting cm, inches, meters, feet

2.54 cm = 1 in.
100 cm = 1 m
12 in. = 1 ft
30 in. = .76 m

Lbs and bushels (See also the appended Grain Yield Conversion page) {C}56 lbs shelled corn = 1 U.S. bushel (A U.S. bushel is a volume of 35.24 litres)

Since a bushel is a volume rather than a weight, the Plotwt in kg will have to be converted to lbs and then the pounds converted to bushels using 56 lbs shelled corn = 1 U.S. bushel. This testweight or density of 56 lbs/bu isn’t always true. It can be as low as 48 lbs/bu for one variety and 60 lbs/bu for another variety. See the note I added to the appended Grain Yield Conversion page.

60 lb/bu = 67.248 kg/ha
56 lb/bu = 62.723 kg/ha
48 lb/bu = 53.801 kg/ha
32 lb/bu = 35.867 kg/ha

{C}Converting Imperial Bushels to U.S. Bushels

1 Imperial Bushel = 36.36879 litres
1 U.S. dry bushel = 35.24 litres

{C}Plot size

width x length
Row width* number of rows * plot length
Take 1/2 row width on left plus full row width(s) plus half row width on right
17 ft. 5 in. = 17.4 ft. = 1/1000 of an acre

Square feet to acre conversion

43,560 square ft in one acre

2 - 30 in. rows = 15+30+15 =60 in. = 60 in. * 1 ft/12" = 5 ft. One plot is 5 ft wide x 17.4 ft. long = 87 square ft. 43560 square feet/87 square feet = 500.689 The plot is approximately 1/500 th of an acre.

The moisture and calculated dry yield figures are reported to one decimal place.

E.g. 21.2,6082.4

The plot weight is reported to 2 decimal places.

E.g. 6.85,7.65,4.80,6.00

An example of using FORMULA 1 to obtain bu/ac when the plotwt data was recorded in kg.

Plot size of 1/500 th of an acre
so multiply the bushels obtained for 1/500 ac. by 500 to get results for one acre.
Moisture: 18.9%
Plotwt: 7.95 kg


((PLOTWT(lbs)-(PLOTWT(lbs)*((Moist-15)/100)))/56)* AREA CONVERSION FACTOR
1) Convert the plotwt in kg to lbs in a separate spreadsheet column
2) Use the lb plotwt in the formula for the yield
3) The area conversion factor is 500 since the plot is 1/500 ac.
= (17.49-(17.49*.034))*500
= ((17.49-.595)/56)*500
= (16.89/56)*500
= .30 * 500
= 150 bu/ac

or 16.89 DRYWEIGHT(lbs) * 8.93 (500/56=8.92) = 150.8 bu/ac


This is one area that the company really needs some help with. It could be learned in only a few months but hiring or consulting a person with extensive experience would be a wise decision. The company was just number grinding with a black box called MSTAT. In CSS 590, the preferred statistics program is called SAS.

There are three main areas where statistics are of benefit.

  1. The design of yield trials and any experiments the company may do.
  2. The analysis of these tests particularly when it involves decisions regarding selection based on only a small amount of data. The king of Agricultural Statistics is R.A. Fisher and his methods are the standard methods.However, an understanding of Bayesian Analysis as taught by Edwin Jaynes in Probability Theory: The Logic of Science would be useful in the future.
  3. The use of phenotypic ratios in genetics.

It’s a bit difficult to find material on experimental design that is written in plain concrete English. Often the material is too general and theoretical to see how to apply it specifically to the company's work. A good specific work that is easy to read and understand is Driscoll's "Plant Sciences " Ch. 27-31 pp. 183–200 and Appendices 2-3.Unfortunately it's out of print.

27 Experimental Design
28 Analysis of Variance
29 Completely Randomized Design
30 Randomized Complete Block
31 Factorial Experiment
Appendix 1. F table (simplified)
Appendix 3. t table (simplified)

Plant Sciences: Production,Genetics and Breeding

by Colin J. Driscoll
ISBN 0-13-677048-7
SB 123.D75 1990
Part V Experimental Design and Statistical Analysis
Ch. 27-31
Copyright 1990 Ellis Horwood Ltd. (Currently out of print)

The Maize Yield trials were usually a Randomized Complete Block design with 2 or 3 replicates. Prior to 1995 there were often tests with 6 replicates with CV's usually of around 11% to 12%. Three or more replicates at several locations would be better than 2 replicates in one location but there have been budgetary reasons given by many companies for decreasing the number of replicates. An ANOVA was done on each test primarily only to determine a C.V. for the yield and moisture. This isn’t the proper way to use ANOVA. To begin with, ANOVA is only valid for tests with 3 or more replicates.

Yield reports grouped the data by variety (entry) giving for each variety, the yield for each plot and an average yield for the variety.

The company has not been very interested in extreme accuracy. Averages have been good enough and often from only 2 2-row plots. A C.V. value of 15% was used to decide if the trial was ok or not. Any trial with a CV over 15% was considered a bad trial and the data was considered invalid. Fisher's Least Significant Difference LSD was not used but it's commonly used by researchers and the company will use it in the future.

Any introductory statistics text has information on ANOVA, Randomized Complete Block Experimental Design, the use of F tests and Fisher’s Least Significant Difference as well as other multiple comparison procedures. University libraries also have numerous texts on Experimental Design, however, the textbook information is often presented with too many bells and whistles and for some other field of work.I recommend Driscoll’s chapters listed above (PLANT SCIENCES) as an introduction. I also recommend the use of a computer as the number crunching gets very involved.


MSTAT was used to obtain CV's for the moistures and the yields in each test. The data had to be formatted to an ascii (.PRN) file with only the numbers to be used and no text or headings.
MSTAT had a nice print out of the ANOVA and the averages for each variety. This was redundant, however, because the averages were computed and printed out in a different format for the final report.
EXCEL has an ANOVA function but I've never tried it on a test with several replicates.
A way needs to be found to replace MSTAT. It has many of the functions one might want but the interface, proprietary file format, and lack of customizability were strikes against it.


In the final report, all of the data for a variety was listed and averaged. The report was made from a Spreadsheet Notebook page stripped to essentials into which the observed data had been entered and to which the harvest data was then matched, added, and a row added for the averages. The row for the averages of each of the varieties and a blank space or two below it was added using a macro that inserted a preformatted averaging row at the proper place.
NOTE that there is a redundancy here since the MSTAT ANOVA print-out calculates and shows entry averages as well.The problem is that the averages are also wanted in a different format in the final report.

Harvest Index

A harvest index is a numerical value that is used by the breeder to assist in the selection of lines to advance based on harvest data. A simple harvest index is to divide the yield of each variety by its moisture content at the time of harvest. This may be referred to as the Performance Index, but you may also see it referred to as just the INDEX.

  • Performance Index (PI) = Yield (bu/ac) / % moisture (P.I. to 3 decimal places)
  • Adjusted PI = ((100 - % broken stalks) x (yield)) / % moisture
The adjusted PI is the grain yield on standing plants
  • Kg/Ha% or Kg%= (Yield of variety / Yield of Standard 1) x 100
  • INDEX% or IND% = (Index of variety / Index of standard 1) x 100
  • COMBINED = Kg/Ha% + INDEX%
  • PERCENT = Avg Index Variety / ((Avg Index Standard 1 + Avg Index Standard 2) / 2) x 100 (to one decimal place)

These last 4 indices were used on old reports of advanced line and standard trials where two standards were used such as P3979 and P3984. Standard 1 was P3984 and standard 2 was P3979.The averages report for each test showed the average INDEX (P.I.) for each variety, and was sorted by PERCENT values. The data pages for these tests included an INDEX calculation for each plot and an average INDEX for each Entry. Standard 2 was almost always at the top of the list. Anything of interest was between the two standards. By circling the yield value, even if the percent value was low, note was made of any variety with a yield greater than the highest yielding standard. Note was also made of varieties with average stalk breakage of greater than 4.5 plants by x'ing out the breakage value. Also noted were varieties with an average plot population of less than 50 by circling the population value. Moistures below 12.0 on the Breeder's report were marked by the computer with an asterisk and the moisture value and the weight value was x'ed in pencil. In one test this involved a variety with one plot at 17.6 and a the second plot at 5.5 giving an average of 11.5*. The average plot weight for the variety was 1.9 kg which was low.In this same test, standard 2 was on top as usual but had an average moisture of 11.4*. One plot was 12.0 the other was 10.9. Average plot weight was 2.9.

Reports for a test prior to 1995 consisted of two parts. {C}Part 1) A listing of all of the data for each plot grouped by variety with averages for the data on each variety and an index value based on standard 1. In standard trials there were 24 plots per test.With 2 replicates per variety. This was printed out on two pages showing 12 varieties per page.

Part 2) This was the report used by the breeder for making decisions. It was one page showing all of the averages for the 24 varieties including the average index value as well as a PERCENT value. The data was sorted by percent values with the highest percent value at the top.

Yield Trial Report

The final Yield Trial Report consists of 3 sections. Section one is a summary of what was done and of the environmental conditions during the yield trial. Section two is the formatted Data Report. Section 3 is the Statistical Analysis print out and any comments regarding the interpretation of the data. There are numerous Yield Trial Reports available on the internet. I have included with these notes 2 pages from the 2002 Dryland Corn Performance Test entitled AGRONOMIC and TEST INFORMATION: GRANGER that is a good example of the kind of information that would be in section one of your report. A more formal and complete report is the 5 page PERFORMANCE OF CORN HYBRIDS IN LOUISIANA, 1999 by the Louisiana Agricultural Experiment Station (LAES) included with these notes for an idea of the format for sections 1 and 3 of the report. In addition to this report there would be appendices with the data and statistical analysis computations. Note that unlike the LAES, you will not be including a recommendations section for your clients.

{C}Section 1

Section 1 is a 1 page cover letter prepared with a word processing program.
In this cover letter be sure to include the following data:
0) Particulars of the test.
  1. ) The date the Yield Trial was planted
  2. ) The date the Yield Trial was harvested
  3. ) The rates and form of N,P and K applied to the Yield Trial Field
  4. ) ??? A very brief summary of how the plots were prepared
  5. ) The herbicide applied and the rate used
  6. ) Information regarding the environmental conditions during the test
  7. ) Statement of any mistakes made resulting in incomplete or misleading data for a plot
  8. ) Explanation of any unusual results


Cornland, CANADA
ATTN: Plant Breeder, Mr. I.M. Pioneer

Dear Mr. Pioneer

Please find enclosed the results of the 2004 BIG YELLOW SEED CO. Yield Trials at Location1 and Location2 in Somebody’sprovince, Canada.

The 2004 BIG YELLOW SEED CO. Yield Trial was planted May 5, 2004 and harvested on October 5, 2004.

Nitrogen was applied to the field prior to planting at a rate of ___ lbs per acre in the form of broadcast Urea pellets.

Fertilizer was applied at planting at the following rates:

N: _____
P: _____
K: _____

The plots were thinned to 60 plants per 2-row plot.

Herbicide "X" was applied at a rate of ______

Environmental conditions were excellent for growing corn this year. {C}______ mm of rainfall and _______ heat units were received during the growing season.

You will notice that the population rates of Plots #65-#70 are less than 50 plants per plot. Unfortunately, due to circumstances beyond our control, these plots were in a low spot in the field and were drowned. Also, due to a cultivation error, data is not available for Plot #275.

Generic Test Co. apologizes for the inconvenience caused by the unavailable data and is refunding to you not just the contracted price for each affected plot but the price of all other plots of affected entries in the test. Corrective measures have been taken to prevent these errors occurring in the future and Generic Test Co. hopes that you will consider allowing us to serve you in the coming year.

Joe Manager
Generic Test Co.

NOTE: I have shown only N,P, and K rather than just N or a specific nitrate formula, P2O5 and K2O because the form of the fertilizers and the method of application can be different. There are many forms of nitrogen, for example, that can be applied with numerous pros and cons for each. We would have preferred multiple applications of nitrogen in various forms (liquid, granular, foliar application, side-banded, ammonium nitrate, ammonium phosphate, etc.) during the growing season but there was a high probability that rainfall would keep equipment out of the field so it was decided to broadcast Urea (46-0-0) and place midrow an (N-P-K) granular fertilizer at planting time. Some local agricultural producers applied anhydrous ammonia which is inexpensive but there are often problems related to applicator nozzles clogging with dirt that made its use for research work somewhat questionable if we could not ensure uniform placement.

There are also other elements such as boron, zinc, sulphur or maybe even vitamin E that one can apply with the fertilizer. I have forgotten the exact formula for the bulk side-banded fertilizer placed with the seed at planting time but an actual (N,P,K) example would be nice. E.g. 20-20-20 or 18-51-20). Perhaps someone might contribute a formula being used in practice. I have one from a farm magazine article somewhere. This is part of the Agronomist/Field worker hat. A rule of thumb is to apply a pound of nitrogen for every bushel of desired yield.

☀[possible inclusion here of maize growing dates and fertilizer application for Missouri farm

Source: Farm Journal

Special Features article by Del Deterling in the Feb. 2008 issue.

Jerry Cox

Yield: 205 bu/ac - 227 bu/ac

Location: SE Missouri

Planting: last week of March,

full season, 118 day - 120 day Yield Gard Hybrids,

Amplify -L seed treatment

80% ridge or no-tilled silt loam 220 lb N/ac

planting: 4-5 gal 3-18-18 per ac + 1 gal 0-0-25-175 at planting 8 gal - 10 gal 10-34-0, 8-10 gal 32% liquid N, 3-4 gal 0-0-25-175/ac, 2" over and 2" down

4 pints boron (deficient Boron and Zinc)

sidedressed at 3" - 4" tall with 50 gal. 32% N at 12", apply 2 gal. 3-18-18 plus 1/2 pint each of boron and zinc

A week before taselling, fly on 40-50 lb N urea

Headline fungicide

Harvest begins Aug. 20

grain moisture in low 20% range

{C}SECTION 2 {C}Yield Trial Data Summary

The basic format for this summary is to group the data by entry # in numerical order showing the data obtained from each replicate of the entry and averages for that entry.

Each page should have a header showing the company name and year of the Trial, date planted, location, and date harvested. {C}Information identifying varieties is on the left and the data is on the right.

A possible ordering of the data column headings is: {C}Entry #, Plot #, Company name, Variety name/Pedigree, Vigor, Tassel, Silking, Stalk breakage above the ear, root lodging, population, Moisture,Plot Weight, test weight and Yield @ 15.5% in Kilograms per hectare or Bushels per acre. In-house reports may be preceded by an averages page containing only the averages for each entry and harvest indices that rank the varieties from best to worst. This is followed by pages grouped by entry # showing the raw data for each replicate of an entry and the averages for the entry. Unusually low populations or above average stalk breakage can be indicated in red on reports. In-house data reports may have different columns and possibly a harvest index column as well.

In addition to the formal Data Summary Report, one may be asked to include copies of the Field Notebook.


This is a print out of the Analysis of Variance (Anova), Coefficients of Variance, LSDs etc and any information pertaining to the statistical analysis of the trial. CV’s greater than 15% are considered to indicate bad experiments whose results are not reliable. In the LAES 1999 Yield Trial Report you will see in the Results and Discussion section that the tests at Alexandria were discarded because the CV’s were greater than 20%.

Yield Trial Appendices

  • A:Yield Trial Terminology

Terminology (Page 1 of 2)


Terminology (Pg. 2 of 2)

  • B:EDGAR: Experimental Design Generator and Randomiser (7 pages)
  • C:LSU AgCenter
  • D:GRANGER TRIAL (2 pages)

Need to find Test design text these pics go with {C}

First randomization attempt


Final Field Map showing Plot and Entry numbers


The Department of Agronomy at the University of Wisconsin-Madison {C}

Fall Field Scouting and Test Plot Evaluation {C}

Methods for Calculating Corn Yields {C}

Breeding Corn for Silage

How Narrow Should Corn Rows Be?

There is a lot of great high-quality information available from the University of Wisconsin but the focus is on agricultural producers. Surf and enjoy.

{C}Yield Trial Reports

There are lots of online Yield Trial Reports

Experimental Design

  • CROP 590:Experimental Design in Agriculture, Oregon State University [7]
see the 'Lectures and Notes' link on the right as well as the other links. They are Powerpoint

and Excel files so you will probably need those programs on your computer.

The instructor for the Crop 590 course is Professor Jennifer G. Kling (Professor/Senior Research, Plant Breeding and Genetics)[]
  • Texas Cooperative Extension Notes

Sample Table of Contents pages from old Yield Trial Notebooks Alpha Lattices

Efficiency of Alpha Lattice Design in Pakistan

Conduct and Management of Field Trials
A Proposed Standard Method for Illustrating Pedigrees of Small Grain Varieties by Purdy,L.H.,Loegering,W.Q.,Konzak,C.F.,Peterson,C.J.,Allan,R.E. CROP SCIENCE, Vol. 8: July–August, 1968,(Pp. 405-406)

Looks like we could lose our jobs to Wall-e soon.

Plant breeder taps latest technology to feed the world

By Tom Fleischman |

October 31, 2018

ICIS-CIMMYT modified Purdy notation in GMSInput.pdf,(Pg. 7 of 14)
Michael Fields Agricultural Institute

Python Programming Project1 

Create a program that creates a randomized list of Yield Trial entries for a random complete block yield trial with 3 replicates 

MSTAT OUT (a facsimile of the output of MSTAT (Michigan State University Statistical Analysis program)

MSTAT Webpage

This is a very old menu driven DOS piece of software for plant breeding experimental design and statistical analysis from the mid 1980's. The menu subprogram labels are abbreviated so one usually needs the documentation for the first few times. E.g. PRLIST means Print List.

Setting parameters for printing and experimental design could be a bit maddening because one was always asked the same questions and in detail each time one used a subprogram. Modern software is recommended instead where parameters are set by means of a parameter setting page rather than by individual yes/no type queries.


  • Robert W. Jugenheimer, "Corn: Improvement, Seed Production, and Uses" Krieger Publishing Company; Reprint edition (February 1986)
    (ISBN 0898746620)

The 1950s FAO version is the first book I could find at a University regarding the work I did but the information I was seeking was never given in detailed form and was only mentioned in a Miscellaneous section as a trivial aside.I found the Reprint edition in a public library but the references to addressograph plates and punch cards are outdated. There are some good photos of some relevant things. This publication has lots of good information and might be a good place to start.

  • Managing Trials and Reporting Data for CIMMYT's International Maize Testing Program 9706480374&q=

  • Manjit S. Kang, "Handbook of Formulas and Software for Plant Geneticists and Breeders", (2003), Food Products Press and The Haworth Press Inc.
    (ISBN 1-56022-948-9) (hard)
    (ISBN 1-56022-949-7) (soft)
    (Dewy Decimal #: QK981.5 .H36 2003)
Table of Contents, Preface and all of Chapter 1 available on internet
Cover page, Table of Contents and Preface (9 pages)

':(Dr. Kang's Handbook of Formulas is NOT what I'm seeking and is more for the Breeder. The reason given for its publication, however, is the same as my own))'

Note the statements regarding software in obsolete DOS form, scattered research software(something different used by different research groups all over the world,(usually a computer science graduate student program often on a different University mainframe in a multiplicity of programming languages and commercial software products)).