Testing Yourself

There is much talk about making sure that products you use are independently tested and the results available for groundsmen/greenkeepers. Whilst I fully endorse this view point, there is no greater verification of whether a product or treatment is suitable for you than by testing them on your own turf.

No testing ground can reproduce the levels and intensity of wear, timing of use, exact conditions and all the other variable inputs that you, as the turf manager, put into your turf surfaces.

This article is written to provide some guidelines on how you can gain meaningful results that are relevant to your playing surfaces. In particular, I want to show two methods that you can use to compare treatments of different products.

First, some guidelines:

1. Before the experiment starts:

Be clear in your mind on why the test is being conducted in the first place. It is very important that you start a trial with clearly defined objectives and purposes.

Decide on how the trial will be run. You may want to carry out plot trials that measure many treatments and require the use of statistics to produce accurate data. Or, you may just need a simple comparison between one treatment and another.

Approach your trials with an open mind. Many experiments are made invalid because of a bias towards a certain result, which affects how the data is collected and interpreted.

2. Materials and Methods

List the materials to be used (products, sprayer, etc.) and the methods that will be used to implement the test (sprayer settings, product rates, evaluation methods).

Decide what is actually to be measured, e.g. colour, growth rates, density or playing characteristics. Bear in mind that measurements can be subjective, e.g. how do you measure difference in green if you are colour blind? Avoid too ambitious measurements that require expensive recording data, e.g. infra red measuring of leaf chlorophyll content. Keep it to practical levels.

Concentrate on the most important reason for choosing a particular product or treatment. The more variables you introduce the less likely you will be able to determine the response of your turf. Keep everything consistent with your normal turf management, and then introduce the one change you would like to make on test areas.

3. Once the experiment has started:

Begin recording observations for each product or process being tested. Observations can be descriptions of visual characteristics ("excellent", "good", "fair" are all visual descriptions), numerical ratings (weights of clippings, electrical conductivity readings, etc.) or relative ratings (subjective performance estimates of quality, such as a 1-9 turf quality rating system).

Without good record keeping, the effort put into a testing programme will be wasted, because you will have no way of remembering how and why your results were obtained.

Use a notebook to record your Objectives, Materials and Methods, observations and Discussion, as well as any other thoughts you have about product performance, the reaction of golfers to management practices, or any difficulty you experience handling or applying a material. Date each entry, and take notes carefully and legibly! Remember - excess information is always better than insufficient information, so don't be stingy with your words. If you are able to take photos, tape them inside your experimental log; these can be invaluable in summarising your results. If you are a good record keeper, you'll find that your notebooks will hold their value for years to come - in resolving disputes about which practice or technique is best, where or how a product was applied, or the history of a problem area of turf.

4. At the end of the experiment:

Review your notes and write a discussion, or summary of your findings, why you think these observations occurred, how valid you think the test was, and how the information will be used. Consult with colleagues and staff members to bring out different interpretations and ideas than your own. Ask how will you incorporate these results into your management programme? Are there any follow-up tests that might be useful?

These four components are essential to any testing programme. If you omit any of them, you will find it difficult to determine what actually happened, and your time will be wasted.

Experimental Design

Replication - Never compare one treatment with another without replication. Trials work is about comparing the average (or mean) results of one set of same treatments with the mean of another set of different treatments.

Despite our best efforts, the turf on a green or fairway is usually not exactly the same. There are differences in microclimate, moisture, turf quality and a host of other factors that result in variability that is beyond the control of the researcher. Without replication (repeating a treatment in two or more locations), this type of variability can lead us to draw the wrong conclusions from a trial.

In order to be sure that the differences observed during an experiment are the result of a treatment, and not simply due to differences in the quality of the turf across the test area, each treatment should be repeated, or replicated, in three different areas. In most cases, three replications should be sufficient to separate out the good from the lousy treatments.

Control Plot - Once you have selected an area within a green, tee, fairway or rough that you want to apply a new treatment or product to, you will also need to select an adjacent area to serve as the non-treated plot. The non-treated plot should be managed exactly the same way as your treated area, with one exception. That is, you should not subject the non-treated area to the product or practice that you are evaluating.

By applying the concept of nothing to your test in this way, you will be able to use the non-treated control as a yardstick to measure any improvement (or damage) that results from the treatment under evaluation. If you fail to include a non-treated plot in your test, you really have no way of knowing how well the new product or practice is performing. For this reason, you should always be wary of manufacturer's data that doesn't include a non-treated plot.

Carrying out simple trials work

We shall now look at two different methods of carrying out simple trials work. The first is a simple method of comparing the application of a treatment on turf with not applying the treatment.

1. Where plywood is king

Plywood is one of the most effective and easy ways of creating "instant" non-treated plots. Let's say you want to see the effects of using a particular fertiliser or pesticide. To make instant non-treated plots, simply place three or four pieces of plywood on top of a few of the treated areas, just before you are ready to begin applying the test product.

When you treat with the product, the areas covered by the plywood will remain untreated.

After application, mark each of the plywood's four corners with turf paint. You can now remove the plywood, and the turf paint will allow you to locate the non-treated areas so that you can make your observations.

If you observe the treated and non-treated areas daily for several days (recording your observations as described below), you should be able to determine whether the application improved, decreased, or had no effect on turf quality.

If there is no visible difference between the treated and non-treated areas, the product probably doesn't have much effect on that area of turf, and you will have saved yourself time, effort and expense by avoiding an unnecessary application. If, on the other hand, the treated area looks better than the non-treated area, then you can treat with the confidence that the product will produce the desired effects.

Plot Trials

Whilst simple, using plywood does not answer the question of 'what is the most effective application rate?', or 'how would this product compare with another product?'. The answer is to set up a series of trial plots and apply different treatments to make simple comparisons.

Plot Size

Most research trials are conducted using small plots, usually all placed within one green, one tee or one fairway. But what size? The smallest plots that I would recommend for on-site testing are 1.5 metres x 2 metres. For most small plot work, this is a convenient size for using a knapsack sprayer or not too big for applying granules by hand.

Another method is to treat one half of a green or fairway and not the other, but this will require replication on three other separate greens or fairways.

Randomisation, or rolling the dice

The use of a randomised design helps us to properly arrange the treatments in the test area, so that variability is minimised, e.g. a dry area on one side of the test area. There are several choices available to us in how we arrange the different treatments in each replicate, so that the test plots are arranged to factor out the influence of that dry area on the results.

Problems can arise if the treatments are arranged in the same order in each replicate, as shown above (assume that treatments 1 and 2 represent two different fertilisers, and treatment 3 represents a non-treated check; the dry area is indicated by the shaded gray area).

In Fig 2. treatment 1 is receiving an unfair amount of pressure, because the dry spot is concentrated in the treatment 1 plots. Using this design, would you be able to tell whether the poor performance of treatment 1 is due to the negative effects of fertiliser 1, or is it due to the our having placed treatment 1 plots where soil conditions are dry? You have no way to find this out, using a non-randomised design.

In contrast, treatments can be arranged randomly as in Fig 3 above. In this case, the randomisation has been done correctly, and the negative effect of the dry area is more evenly spread over all of the treatments, giving you a fairer look at the performance of each treatment.

Trials work, whilst being great fun, can take up large amounts of time, and a busy turf manager needs meaningful data on which to base decisions on where and when to spend his/her employer's money. The use of the Plywood Sheet method will give very quick results, and simple plot trials help in comparing combinations of treatments.

If you are unfamiliar with trialing, practice by setting out a simple trial of a known application, e.g. a fertiliser at different rates, as in Fig.5.

Finally - keep and open mind, be precise in application of treatments, replicate treatments, record everything and discuss the results with colleagues before making a final conclusion and/or recommendation.

About the author: Andrew Turnbull BSc (Hons), Dip. RSA, Cert Ed. Owner of AllTurf Management, and Managing Director of The Great Lawn Company Ltd. Contact allturfman@ntworld.com