Field trials are woven into the history and future of agriculture. They’ve been heavily used in formal plant breeding for the advancement of agricultural practices, products, and ideas.
Equipped with the tools to gather and interpret field trial data, you can now use agricultural field trials as a powerful way to determine what inputs and agronomic practices work best on your land. And you don’t need to be a research and development crop researcher or data scientist to glean insights from these trials.
There are five key steps to use this scientific practice as an effective on-farm, boots-on-the-ground practice: identify the why, design the trial, gather data, interpret results, and draw conclusions. With these five steps, you can be better equipped to conduct a well-designed on-field experiment.
1. Identify the “why”
Beginning with the end in mind is critical for successful agricultural field trial design. You need to identify the purpose of your on-field experiment and define the research question you want to answer. What do you want to learn from your agricultural experiment?
Good field research questions can be answered with a “yes” or “no” and will directly impact your farming operation or bottom line. Answering these questions will give you the information you need to make decisions about your agronomic practices, which inputs your crop needs, and how you can increase yield, reduce cost, and improve sustainability.
A few examples of good research questions could be:
- “Would applying fungicide in-furrow in corn at planting increase yield?”
- “Does corn variety X yield higher than corn variety Y in well-drained soils?”
- “Does the use of a bio-stimulant increase corn yield?”
The research question will then help you determine what controls or treatments are needed.
The treatment group receives the change that you want to see the results of – such as a drought-resistant variety, an earlier planting date, or a new insecticide.
The control or check group is identical in all ways to the treatment group except they don’t receive the treatment. The control is used to compare the results of the treatment. Control plots and treatment plots together are referred to as a block.
The research question you select should provide a clear line of sight to the equipment and resources needed to complete the trial. As you plan out the research question, think through how the results can be measured, if the available test site is appropriate for the research question, and how the field history might impact the outcome of the study.
Keep in mind that field trials can be conducted on a single farm or across many locations.
2. Design the trial
Any on-farm experiment requires a good field research design. While few fields out there fit the “flat and square” description, even slight variations in slope, fertility, and soil type can impact the integrity of field research.
To control for field variability, field trial design should account for variation. Fields should be sectioned according to field characteristics. Using replication and randomization in the on-farm field trial design can reduce result bias and account for field variability as well.
Replication means repeating field trial treatment blocks of the control or check and the treatment, often four to six times, across the field. This provides a large quantity of data. But, it may not minimize result bias, depending on the field characteristics. For example, a field whose northern end consistently yields higher means the northernmost treatment in each block would always yield higher.
Randomization can be used in combination with replication to overcome field variability. The treatment blocks are placed randomly throughout the field as opposed to in a pattern, which removes any preference for one treatment over another.
Research designs set the on-farm field trial up for success, both in execution and in data analysis. Research designs should include both replication and repeatability. Conditions within blocks should be similar but can differ from block to block.
Some common agricultural field trial designs include paired comparisons, randomized complete block designs, and split-plot designs.
A paired comparison is a field trial design for comparing any pair of treatments, such as two different fertilizer rates or crop varieties. Blocks include one plot of each treatment, placed randomly within the block. The block is replicated across the field, typically four to six times. A paired comparison is a type of randomized block design, but because it only involves one treatment and one control, statistical analysis of the results is more simple.
For a comparison of three or more treatments, a randomized complete block design is a good choice. It involves a block that includes all treatments and an untreated check in a randomized order within the block. The block is repeated at least four times across the field.
Interactions between treatments can be studied through a split-plot design, where main treatments have sub-treatments applied to them. Split-plot design can also be used when one of the treatments requires additional replication. This field study research design requires a larger area and additional management due to its complexity.
3. Gather data
Data collection is an essential part of your on-farm experiment. Take note of what data you need to collect before you begin your on-farm research.
You may already record things like planting, application, harvest dates, varieties and population(s) planted, and moisture.r Collecting your normal data in a consistent manner as well as capturing additional crop condition and growth data is critical. Data collected could include notes or photographs. Information such as node or pod counts, pest pressure, or storm damage will help you better interpret the study results after the crop is harvested.
Yield data is a vital dataset to capture for your on-field experiment. Ensure weight is measured from a calibrated scale, and moisture and test weight are also captured, if applicable. Consider capturing data from the center rows of each plot to minimize the impact of potential treatment drift.
Your on-farm research plan should include data management and storage processes. Cloud-based technologies and telematics make data collection and storage automatic, reducing the opportunity for error and creating efficiency.
4. Analyze and Interpret Experiment Results
Interpreting data collected during field trials can feel overwhelming. But analysis is what makes data actionable, and it requires going beyond a simple comparison of treatment averages.
Statistical analysis helps determine if there’s a significant difference between treatments – meaning the results aren’t due to chance or variability within the field.
When an experimental is designed for statistical analysis – such as one of the agricultural field trial designs above – the Least Significant Difference (LSD) can be used to determine whether the results are likely to occur again in the future because they are due to the treatment. The LSD is based on a probability level that indicates how certain you can be that you’re correct. If the averages of the two different treatments differ by more than the LSD value, you can be confident that the result will likely occur again in the future.
You don’t have to be a statistician to have confidence in your on-field experiment design and results. Today, there are powerful agronomic tools and software available at your fingertips to streamline this process. Using technology to manage data, including storing it and sharing it with farm advisors, can also make the interpretation of field trial results simple.
5. Draw Conclusions and Determine What’s Next
Think back on the initial research question. What conclusions can you draw from your agricultural experiment? Based on your findings, you might decide to change production practices or that it just didn’t pencil out. Either way, the results are valuable to understand field performance and management outcomes.
You might be ready to make a decision for next year based on what you found. Or, you might still have questions or doubts. To build confidence, you can conduct the same on-farm field study over multiple seasons or in different fields to ensure the results weren’t tied to that year or that location.
It’s also worth thinking about how to enrich your dataset. Is there a neighbor you could partner with to expand the trial? What about an aggregated, legacy dataset that you could compare your results to? Knowing if the results are repeatable both on your farm and across a broader geography can be helpful when deciding to implement a new practice as a result of an on-field experiment.
You may be interested in:
Trial and Error: Navigating Agricultural Trials with Biologicals
CROs Can Count on Agronomic Data Analytics Tools
Agronomic Field Trial Compliance and Reporting with Advanced Tools
Agronomic Field Trial Design and Management Made Easy with Agmatix
The five steps for field trial success are building blocks for better decision-making on the farm. A well-conducted agriculture experiment can unlock yield potential and build confidence in the return on investment of new production practices.
But, the process doesn’t have to be difficult. Agmatix’s holistic field trial software is a user-friendly platform designed to help you scientifically plan and manage on-field experiments. You can design your trial on-map with hundreds of treatment combinations. You can set up status updates for your field trial and easily communicate with those involved in the trial.
Once your trial is complete, data collection for agronomic trials makes it easy to use big data to create value for your farm through operationalizing insights. You can standardize legacy data or combine agronomic data to dig deeper into information that was once difficult to piece together. With the whole picture in mind, field data management can take your on-farm field trials to the next level. You can be confident in the next best step to farm for the future.
