Does Variable-Rate
Seeding Boost
Sunflower Yield?
Reading time — 15 min
45 and 101-hectare field experiment
Usevalad Henin
Usevalad is an expert in GIS and agricultural chemistry. He has been developing precision farming tools since 2013. He is also the co-founder of OneSoil.
Last year, I conducted a few experiments to determine how effective variable-rate seeding is. The first experiment I'd like to share with you was in sunflower fields in Ukraine. In this article, I'll discuss how variable-rate planting impacted the germination rate in chestnut soils. And most importantly, whether or not we were able to increase the crop's yield through variable-rate seeding.
For a quick read, jump to my conclusions and recommendations.
For a quick read, jump to my conclusions and recommendations.
No one field is completely homogeneous in terms of crop yield. In one area, you can reap 12 tons of grain per hectare every year, while you only get 4 in another one. The higher an area's yield is over multiple years, the more productive it is.
It stands to reason that if you increase the seeding rate for these areas, their yield will grow, too. And, on the other hand, if you decrease the seeding rate in low yielding areas, the yield won't change.
This approach is called variable-rate seeding. In theory, it cuts down on the number of seeds you use while increasing the yield.
Last year, I decided to find out if this is true. To do so, I conducted several experiments in Ukraine involving spring crops: corn, soybeans, and sunflowers. Through a series of essays, I'll talk about how the experiments went and share the results. We'll start with the first experiment: sunflower fields in Ukraine's Kherson Region.
The optimal seeding rate for sunflowers is highly debated. Farms usually stick to the rate recommended by their local seed supplier. I decided to sow sunflower seeds at a varied rate to see how it would impact crop yield and whether or not I could cut back on the number of seeds I used.
I conducted the experiment on Agroproduct, LTD's fields in the Kherson Region. I used two fields with different moisture levels. One field was irrigated, while the other was dryland, or land without artificial irrigation.
It stands to reason that if you increase the seeding rate for these areas, their yield will grow, too. And, on the other hand, if you decrease the seeding rate in low yielding areas, the yield won't change.
This approach is called variable-rate seeding. In theory, it cuts down on the number of seeds you use while increasing the yield.
Last year, I decided to find out if this is true. To do so, I conducted several experiments in Ukraine involving spring crops: corn, soybeans, and sunflowers. Through a series of essays, I'll talk about how the experiments went and share the results. We'll start with the first experiment: sunflower fields in Ukraine's Kherson Region.
The optimal seeding rate for sunflowers is highly debated. Farms usually stick to the rate recommended by their local seed supplier. I decided to sow sunflower seeds at a varied rate to see how it would impact crop yield and whether or not I could cut back on the number of seeds I used.
I conducted the experiment on Agroproduct, LTD's fields in the Kherson Region. I used two fields with different moisture levels. One field was irrigated, while the other was dryland, or land without artificial irrigation.
Variable-rate seeding is a precision farming method that uses different (varying) amounts of seeds in different parts of the field. We use variable-rate seeding on the notion that the field has varying soil make-up, nutrient content, moisture levels, and, as a result, varying crop yields.
Throughout the experiments, the OneSoil team helped farmers analyze and plan fieldwork. The farmers selected the seeds and equipment.
Productivity zones
I delimited the fields into productivity zones, which served as the basis of my experiment. I delimited the zones using data on vegetation I got from satellite images and data from the actual crop yield. Agroproduct was kind enough to share them.
Field 1 productivity zones (irrigated)
45.5 ha area. The field was irrigated using a center-pivot system. Chestnut soils with loamy soil texture. Homogeneous relief of the field with an elevation difference of about 1−2 meters.
Field 2 productivity zones (dryland)
101 ha area. Chestnut soils with loamy soil texture. This field’s relief is less homogeneous. The elevation difference here is around 5 meters.
Hypotheses
I formulated two hypotheses. In one, I posited that crop yield would grow in high-productivity areas of the field when we increased the seeding rate. There are fewer nutrients in low-productivity areas of the field, so we could decrease the seeding rate without it affecting the crop yield. It would, however, help me save on the number of seeds I used. That was the first hypothesis.
The second, or inverse, hypothesis stemmed from a different assumption. Perhaps the plants in high-productivity parts of the field would compete for sunlight instead of nutrients, while low-productivity areas of the field would be evenly depressed.
In that case, if we lowered the seeding rate in high-productivity areas, fewer plants would yield a crop bigger or the same size as they would with a high seeding rate. Conversely, we would notice an increase in crop yield in low-productivity areas of the field when increasing the seeding rate. That’s the inverse hypothesis.
The second, or inverse, hypothesis stemmed from a different assumption. Perhaps the plants in high-productivity parts of the field would compete for sunlight instead of nutrients, while low-productivity areas of the field would be evenly depressed.
In that case, if we lowered the seeding rate in high-productivity areas, fewer plants would yield a crop bigger or the same size as they would with a high seeding rate. Conversely, we would notice an increase in crop yield in low-productivity areas of the field when increasing the seeding rate. That’s the inverse hypothesis.
Prescription map
I put together a prescription map for equipment. I used productivity zones to determine the amount of seeds to use in different parts of the field and to compile prescription maps.
I had three sunflower planting rates for each productivity zone. Low and high rates aimed at testing my two hypotheses, while a moderate rate represented the farm's usual application rate.
I had three sunflower planting rates for each productivity zone. Low and high rates aimed at testing my two hypotheses, while a moderate rate represented the farm's usual application rate.
Prescription map for Field 1
The farm's usual application rate was 60,000 per hectare. I increased the seeding rate in high-productivity areas and lowered it in low- and moderate-productivity areas. As such, the seeding rate varied between 40,000 and 64,500 seeds per hectare.
Prescription map for Field 2
The average seeding rate in this region is 45,000 seeds per hectare. For this experiment, I used rates between 33,000 and 45,000 seeds per hectare. In other words, I decreased the seeding rate in all productivity zones in this field.
To test the second hypothesis, I placed two test strips in both fields. I seeded the first one at a moderate rate. For the second strip, I lowered the seeding rate in high-productivity areas and increased it in low-productivity areas.
Limiting factor
I identified a limiting factor in low- and moderate-productivity zones. Doing so was necessary to correctly interpret the experiment results and to understand what impeded plant development.
Based on my experience from previous experiments, I know that these factors can be a lack of moisture caused by features of the terrain or the soil’s texture, anomalous acidity values, low organic nutrient content, or salinity.
In order to identify the limiting factor, I analyzed the field’s terrain and the soil’s brightness using equipment data and satellite images, respectively.
Based on my experience from previous experiments, I know that these factors can be a lack of moisture caused by features of the terrain or the soil’s texture, anomalous acidity values, low organic nutrient content, or salinity.
In order to identify the limiting factor, I analyzed the field’s terrain and the soil’s brightness using equipment data and satellite images, respectively.
According to the law of limiting factors, the most important factor for a plant is the one that most significantly deviates from the optimal value. If crop yield falls sharply in one part of the field, I’d first have to examine the limiting factor to understand the reasons why.
Soil brightness for Field 1
Relief map for Field 1
I compared the first field's relief map with its soil brightness from March 2019, when it was tilled. The image shows soil brightness fall in the low-lying area. It's only 1 meter lower than the rest of the field area. Moisture usually amasses in low-lying areas, which makes the soil darker and rich in nutrients. Therefore I assumed that these areas may see salinity, which was confirmed after I scouted the field.
I then superimposed the soil brightness maps over productivity zones. Low-productivity areas aligned with areas with decreased soil brightness. As such, I concluded that Field 1's limiting factor was salinity.
I then superimposed the soil brightness maps over productivity zones. Low-productivity areas aligned with areas with decreased soil brightness. As such, I concluded that Field 1's limiting factor was salinity.
Soil brightness for Field 2
Relief map for Field 2
The second field’s soil brightness was about the same. The terrain here was less homogeneous, with an approximately 5-meter difference in elevation. I also noted that on slopes and higher parts of the field, there is less organic matter content than in even and low-lying parts of the field, where the soil is brighter. After comparing the relief map with the productivity zones, it became apparent that slopes and higher grounds correspond to low-productivity zones.
That’s how I concluded that the limiting factor for the second field is its terrain, which affects moisture distribution, organic matter content, and, at the end of the day, field productivity.
That’s how I concluded that the limiting factor for the second field is its terrain, which affects moisture distribution, organic matter content, and, at the end of the day, field productivity.
Germination rate
I evaluated the germination rate. To do so, I used detailed field maps we got using a drone. Agroproduct commissioned Kherson Zemproekt, LLC, to create the maps. We then agreed with the company CropSaver on calculating the number of seedlings in the fields. I grouped these values into 10×10-meter squares and correlated them with actual sowing rates for each square.
I ended up with the germination rate as a percentage. This came in handy when looking at whether the germination rate depends on the seeding rate.
I ended up with the germination rate as a percentage. This came in handy when looking at whether the germination rate depends on the seeding rate.
Germination rate map for Field 1
The germination rate for areas with a seeding rate of 40,000/ha ranged from 78% to 93%. In areas with a seeding rate of 64,500/ha, this was 60−66%.
Germination rate map for Field 2
The seeding rate was 33,000/ha in the second field for high-productivity zones. The average germination rate here is 82%. With a seeding rate of 45,000/ha, I got a germination rate of 77%.
The germination rate was 82% for the low-productivity area, with a seeding rate of 33,000/ha. At a seeding rate of 45,000/ha, the germination rate was 77%.
The germination rate was 82% for the low-productivity area, with a seeding rate of 33,000/ha. At a seeding rate of 45,000/ha, the germination rate was 77%.
What does that mean? The germination rate doesn’t depend on productivity zones; it drops when the seeding rate is increased.
Sunflower yield analysis
I analyzed the crop yield. I did this using yield maps from onboard computers and homogeneous zones I put together manually.
Yield map for Field 1
Received from the onboard computer
Yield map for Field 2
Received from the onboard computer
Interpreting yield maps
How do we interpret the maps? The maps show us that the yield doesn’t change for test strips in areas with varying seeding rates and the same productivity. At first glance, this indicates that the seeding rate doesn’t affect the sunflower yield.
This method worked great when I needed to make a visual assessment. However, in that case, it's easy not to notice important relationships. That's why I analyzed the sunflower yield by homogeneous zones, too.
It's important to understand why this task should be done by humans instead of being entrusted to equipment. Combine harvester yield data is always noisy. This noisy data comes from anomalous yield values at a specific point.
If a combine skids to a halt while harvesting, we'll most likely see abnormally low values. We see the opposite — abnormally high values — when unloading grain.
If a combine follows a tramline, we'll see low values again. We can't automatically filter out this noisy data. That's why processing the data manually is more reliable.
It's important to understand why this task should be done by humans instead of being entrusted to equipment. Combine harvester yield data is always noisy. This noisy data comes from anomalous yield values at a specific point.
If a combine skids to a halt while harvesting, we'll most likely see abnormally low values. We see the opposite — abnormally high values — when unloading grain.
If a combine follows a tramline, we'll see low values again. We can't automatically filter out this noisy data. That's why processing the data manually is more reliable.
Several homogeneous areas of Field 1 (irrigated). I delineated 150 of these areas. The area of each one is 0.25−0.5 ha.
What does it all mean? The area's yield is noted at the center of the homogeneous areas. It's important to keep track of whether there is a pattern of yield increase or decrease relative to the seeding rate in those areas.
The seeding rate is different on the two test strips. If there's a pattern, we can analyze the data later on, draw conclusions, and make decisions. If there's no pattern, then the conclusion is simple: the seeding rate didn't impact the yield.
The seeding rate is different on the two test strips. If there's a pattern, we can analyze the data later on, draw conclusions, and make decisions. If there's no pattern, then the conclusion is simple: the seeding rate didn't impact the yield.
I made a similar map with homogeneous areas for the second field. To make it easier to analyze the results, I then calculated the average yield of the homogeneous areas for each productivity zone in the fields used for the experiment.
Irrigated field
Average yield, t/ha
Dryland field
Average yield, t/ha
* part of the field wasn't harvested
* part of the field wasn't harvested
As we can see in the tables, I got approximately the same yield in all productivity zones for all three seeding rates in both fields.
Experiment conclusions and recommendations
The seeding rate didn’t impact sunflower yield in either the dry or irrigated fields. I suppose that this happened due to plant plasticity or the hybrid that I sowed. I’ll continue the experiments in 2020 on even larger scales to identify whether or not there are hybrids that react positively to variable-rate seeding. I’ll keep you up-to-date through our blog!
Plant plasticity refers to a plant’s ability to adapt to various changes in its environment, including in the seeding rate.
Recommendations
Throughout the course of the experiment, I formulated several recommendations for those of you who are already practicing variable-rate seeding or those who are just planning to do it. These recommendations will help you save on the number of seeds you use and better manage crop yield.
1. Test various seeding rates to determine what the optimal seeding rate is for a specific field or hybrid. In this experiment, for example, an increased seeding rate didn't affect yield in either of the fields. But that doesn't mean that the same thing would happen in another field with a different crop.
2. Identify the yield's limiting factor. If you know what impedes plant development, you can at least minimize it or, best-case scenario, eliminate it.
3. Save yield data for each field. They'll help delineate productivity zones for variable-rate planting or applying fertilizers, and to identify the limiting factor.
1. Test various seeding rates to determine what the optimal seeding rate is for a specific field or hybrid. In this experiment, for example, an increased seeding rate didn't affect yield in either of the fields. But that doesn't mean that the same thing would happen in another field with a different crop.
2. Identify the yield's limiting factor. If you know what impedes plant development, you can at least minimize it or, best-case scenario, eliminate it.
3. Save yield data for each field. They'll help delineate productivity zones for variable-rate planting or applying fertilizers, and to identify the limiting factor.
Recommendations
Throughout the course of the experiment, I formulated several recommendations for those of you who are already practicing variable-rate seeding or those who are just planning to do it. These recommendations will help you save on the number of seeds you use and better manage crop yield.
1. Test various seeding rates to determine what the optimal seeding rate is for a specific field or hybrid. In this experiment, for example, an increased seeding rate didn't affect yield in either of the fields. But that doesn't mean that the same thing would happen in another field with a different crop.
2. Identify the yield's limiting factor. If you know what impedes plant development, you can at least minimize it or, best-case scenario, eliminate it.
3. Save yield data for each field. They'll help delineate productivity zones for variable-rate planting or applying fertilizers, and to identify the limiting factor.
1. Test various seeding rates to determine what the optimal seeding rate is for a specific field or hybrid. In this experiment, for example, an increased seeding rate didn't affect yield in either of the fields. But that doesn't mean that the same thing would happen in another field with a different crop.
2. Identify the yield's limiting factor. If you know what impedes plant development, you can at least minimize it or, best-case scenario, eliminate it.
3. Save yield data for each field. They'll help delineate productivity zones for variable-rate planting or applying fertilizers, and to identify the limiting factor.
“
We first heard about OneSoil from our co-workers. We heard the app's name, read up on it online, and started to try it out. We decided to participate in this experiment because we developed a great deal of trust in OneSoil's founders as professionals. We did something like this before, but only in one field. That's why we were keen on trying out variable-rate seeding at a larger scale with the OneSoil team. We're still pondering this experiment's findings, but work is already underway on the next experiment.
Text
Illustration
Want to
conduct an experiment
with OneSoil?
conduct an experiment
with OneSoil?
Try OneSoil Yield now
Any farmer can purchase the program from an agricultural technology dealer in his region. Explore the app's features in a free demo account to take care of any doubts you may have before buying it.
Related Articles
People, fields and technology
Get all new articles straight to your inbox! You can read them even in the field.