On-farm trials can be an effective tool for evaluating new practices or products in the vineyard (Veseth et al. 1999). While universities and government laboratories around the country provide information on viticulture research trials, some of this information may not be readily applicable to varying local conditions (Wuest et al. 1995). Designing on-farm trials does not have to be complicated; however, basic components need to be included in every trial in order to properly interpret the results (Koenig et al. 2000; Ketterings et al. 2012).
This guide provides an overview of the basic concepts and principles that are important in conducting research on farms. These principles are also applicable to the successful implementation and interpretation of vineyard trials.
This guide also provides information on the kinds of data to collect and when and how to collect them, since data collection and interpretation are integral to drawing informed conclusions on vineyard trial results. This discussion includes examples using specific viticulture on-farm trials and highlights how to organize and interpret trial results. Finally, the guide provides information on how to make decisions based on trial results.
Basic Components of an Experiment
Most experiments have five basic components:
- A control;
- One or more treatment(s);
- Replication of the control and treatments;
- Randomization of the control, the treatments, and their associated replications; and
- Repetition of the experiment.
These components are discussed in detail below.
Control. A “control” is the practice or product that is in current use on the farm. A control provides a baseline or benchmark against which to compare a new practice or product. All experiments need a control. In some cases, a single control is sufficient for multiple treatments; in other cases, more than one control is necessary. Either way, the control needs to be fully integrated in each experiment. It should be applied within the same area that the experiment is conducted and applied at the same time as the experiment.
There are two main types of controls: 1) A positive control, which in the case of on-farm trials is generally the use of a standard practice or product and 2) a negative control, which is generally not doing or applying anything at all. For example, when testing a new fertilizer, a section of the vineyard should be treated with the standard fertilizer as a positive control. An alternative testing method would be not to treat a section of the vineyard at all even if you normally would, which would be considered a negative control.
If space and time allow, both types of controls can be used in an experiment. The use of controls allows for more definitive comparisons to know if the results from using the new practice or product were really due to that practice or product rather than due to other external factors (e.g., differences in weather or a change in irrigation practices).
Treatment. A treatment is a new or different practice or product to be tested. Treatments are used for a variety of things, but in viticulture they are often applied to evaluate new varieties, trellising systems, canopy management practices, irrigation practices, fertilizer regimes, cover crops, pesticides, or the effectiveness of new equipment (for examples see Appendix 1). Experiments can be designed with as few as one treatment or as many as can be physically handled.
However, the more treatments are evaluated, the more input is needed in terms of labor, land, capital, and data collection and evaluation. Sometimes the addition of multiple treatments may also require the addition of multiple controls, thus taking an on-farm trial from manageable to unruly. The most common mistake in many on-farm trials is the desire to evaluate too many treatments in a single experiment.
Replication. Replication of both treatments and controls within an experiment is necessary for appropriate interpretation of the results. Even though a vineyard may be planted to a single clone of a specific grape variety, there can be substantial variation within that vineyard due to differences in soil type or depth, slope, or other site aspects either inherent to the site or constructed, such as the location of vineyard infrastructure.
Consequently, it is important to test both control(s) and treatment(s) in multiple locations within a vineyard to reduce results that might be affected by pre-existing site conditions.