April - May 2025 Vol 46 No 2

NDVI maps for targeted product applications in cotton

Emma Ayliffe, director of Summit Ag in Lake Cargelligo, NSW, is a former Young Farmer of the Year and the 2021 Chris Lehmann Trust Cotton Achiever of the Year. She and her partner manage a 10,000 hectare mixed cropping enterprise, which also serves as a testing ground for new ideas, including the three uses of NDVI discussed in this article.

A major motivation for Emma’s approach is the need to better utilise the wealth of data available. “We have all these tools and data just sitting there. What’s the point of having it if we’re not using it? The goal was to find new ways to analyse and add value to the data to deliver results for growers,” she explains.

Another key factor is the need for aerial product applications. “With cotton, about 90 per cent of what we do is aerial, so we can’t rely on technologies like tractor-mounted cameras. We need to see large sections of the farm and have large zones because it takes around 50 metres for a plane to adjust to a rate change.”

NDVI images are collected every 3–5 days via SataMap, helping guide decisions, but Emma stresses the importance of regular field scouting. “We check these crops every week, so we know what’s happening on the ground. NDVI is a valuable tool, but we always confirm what we see by walking the paddocks.”

Growth regulators

Cotton is an indeterminate crop, meaning it will continue vegetative growth unless actively managed. Growth regulators are used to control this growth, redirecting the plant’s energy towards boll development.

In the region around Griffith, Hillston and Lake Cargelligo, cotton often experiences a cold start to the season, with temperatures regulating early growth. However, when temperatures rise quickly, crops can exhibit uneven growth. Certain areas, particularly those with good soil or high nutrition may experience vigorous growth, while other areas with more challenging soils conditions lag behind.

Growth regulators are used to even out the crop growth. Regulators are often applied as a blanket rate or as a rough estimate of application areas based on a paddock check.

“Typically, an agronomist or farmer goes out into the paddock and looks for visual height differences as an indicator,” Emma says. “But it’s impossible to see the entire paddock at once.”

According to Emma, this method presents two significant downsides: “You end up putting growth regulator on areas of the paddock that don’t need it, slowing growth unnecessarily. You also underdose areas that require a heavy application because you don’t want to over-regulate the rest of the crop. Overall, this leads to inefficient use of resources, wasted money, and uneven crop development.”

NDVI imagery can more accurately identify high-growth areas and apply growth regulators where they’re needed most. For example, Figure 1a is a NDVI image of a paddock with hand-drawn growth regulator application rates. This was made while Emma was in the paddock, checking the imagery aligned with crop condition. Differences in growth are largely due to soil variation, including patches of dispersive soil, and waterlogging.

Figure 1b shows the variable rate (VR) application map created from the NDVI data. This map was sent to the plane for aerial spraying. Fifteen days after application (Figure 1c) there was a noticeable evening out of growth across the paddock, and 45 days after application (Figure 1d) the crop shows fairly uniform growth.

The benefits of using NDVI for targeted growth regulator applications extend beyond cost savings. Direct savings on product are not always substantial, but the overall impact on crop health, yield, and farm operations can be significant. “It’s the flow-on effects where we see the real benefits,” Emma said.

One of the primary advantages is more consistent boll development, which contributes to a more uniform crop and easier harvest. Emma says, “Uniformity is king. 2022 was not a good growing season, with the average yield in the areas around eight bales, while this paddock (Figure 1d) yielded 11 bales.”

Other benefits are:

• Improved defoliation efficiency: When it’s time to defoliate (discussed in more detail later in this article), a uniform crop allows growers to use cheaper defoliation products;
• Timely winter planting: Cotton fields with inconsistent maturity can delay the harvest and impact subsequent crop cycles. “We often pick cotton and then plant wheat straight after. If the cotton crop drags on, we miss the opportunity for timely winter planting”; and,
• Reduced compaction: If picking cotton when it’s wet, there are more compaction issues. If the crop is slow, your harvest is late, which means more chance of rain.

Targeted weed control

Weed pressure in cotton paddocks can vary significantly. Growers usually wait until the defoliation stage to apply herbicide, as the dense cotton canopy makes it difficult for chemicals to penetrate to the weeds below. However, variation in crop growth in a paddock means some areas develop higher weed pressure than others. This variation can be targeted with a strategy Emma calls ‘reverse NDVI’.

At the end of the season, lower NDVI areas indicate low cotton biomass and incomplete canopy closure, making them prime targets for herbicide application. “That’s where we know we can get coverage on the weeds,” Emma says. “The plants haven’t closed the canopy, so we can effectively get the herbicide down to address the problem.”

Areas with a fully closed canopy often have lower weed pressure due to the competition from the cotton crop itself.

This targeted approach is particularly useful in areas where cotton growth has been stunted due to factors such as waterlogging, or in replant areas. For example, floods in 2023 impacted crop growth and brought huge/serious weed pressure on Emma’s farm, including windmill grass, barnyard grass and milk thistle. Figure 2 shows stunted crops in the foreground and healthier crops in the background. The crops in the foreground did not improve over the season.

Emma used NDVI imagery (Figure 3) to create a VR glyphosate map, targeting the lower NDVI areas.

In Figure 3, the darker red areas are higher biomass with full canopy closure; there was no point applying glyphosate as the product wouldn’t get through the canopy and there is some weed competition from the canopy. The blue areas in the centre of the field had lasering issues and the crop did not grow well. Granular glyphosate was applied at 1.5 kg per hectare in the blue areas, which covered just under 40 per cent of the paddock. No glyphosate was applied to the rest of the paddock.

Using this process means Emma can tackle weeds earlier than previously, target the weeds earlier so they are easier to kill and there is a less of a weed burden at the end of the season.

Defoliation

Variable rate defoliation is a practice that’s becoming more common in cotton. Defoliation, typically requires at least two passes of defoliation products. In cooler conditions, a third pass is often necessary to ensure complete leaf drop and boll opening.

“We usually end up doing three passes because of the cooler weather, but there’s always a percentage of the paddock that doesn’t need that third pass. Instead of blanket spraying, we now target those areas that actually need the treatment.”

Using NDVI, Emma identifies areas of high leaf retention or where the bolls haven’t opened. This approach is particularly useful in paddocks with varying canopy sizes or replant sections that develop later than the rest of the crop, or areas affected by herbicide drift, where leaf drop or boll opening may be hindered.

“We’re able to just go and target that chemistry exactly where it’s needed. Be it a later section with bolls not open, or a high biomass area that’s got more leaf,” Emma says.

Figure 4 is a satellite image of a paddock before defoliation product has been applied. Lighter areas indicate open bolls; darker are closed and need treating. Figure 5 is the VR defoliation map developed based on Figure 4 and a paddock inspection. Areas with heavier canopies and higher biomass receive higher application rates, while areas with sparse growth or already opened bolls receive less. Figure 6 shows the paddock two weeks after the VR defoliation application, with bolls open across the paddock.

“We worked with a grower to come up with a rate range for Promote to open the bolls. I marked up a few key spots on an NDVI screenshot, sent it off to PCT, and they created a variable-rate map for the plane. This particular crop had received a double dose of fertiliser, resulting in a larger, harder-to-defoliate crop.”

Defoliation chemicals are expensive, with top-tier products costing upwards of $65 per hectare at full rates. Reducing the treated area can lead to significant savings, especially for large-scale growers managing hundreds or even thousands of hectares.

There’s a strong incentive for growers to adopt variable-rate defoliation. Emma says, “The cost savings and productivity gains are too significant to ignore. We already have the NDVI data and monitor it closely, so implementing variable-rate defoliation is straightforward.”

The cost of creating the VR maps ranges from $1.20 to $1.40 per hectare. Emma notes that any initial hesitancy among growers quickly fades once they see the results: “Growers may be hesitant at first, but once they realise the savings, they want it done every year.”

A more uniform crop at harvest leads to more efficient harvesting, and targeted product applications reduce the risk of chemical drift. “There’s a lot of discussion around drift, especially with defoliation chemicals, particularly in the Murrumbidgee Irrigation Area (MIA), where we have other sensitive crops like citrus. The more precise we can be, the better it is for everyone in the district—and we save a few dollars per hectare by using less chemical.”

What’s next?

Emma is now looking ahead to other crops and scenarios where VRT might provide similar benefits. “We’ve probably got a good handle on the obvious use cases in cotton, but now we’re exploring how it could work in other summer crops like corn, sorghum, and sunflowers. We’re also considering rolling this out to irrigated winter crops to see what the benefits could be.”

One of the latest experiments involves using growth regulators in irrigated wheat, targeting high-biomass areas prone to lodging. However, Emma points out, “This is still in its early stages. We’ve only just applied the treatments, and the wheat hasn’t been harvested yet, so we don’t have results yet.”

Another area under investigation is combining NDVI with grid soil maps and field data to identify growth and nutrient issues more precisely. “We’re experimenting with overlaying NDVI data on grid soil tests to spot patterns and identify issues, but it’s still a work in progress.”

For frost, Emma is looking at the change in NDVI before the frost and four weeks after. Emma notes that she can start to see dead heads or frost damaged areas, that are confirmed with a paddock inspection.

Emma also raises the question of whether this technology can be applied to dryland crops over large hectares, where management is often more challenging due to less consistent water availability.

For now, it’s a ‘watch this space’ scenario, as more experiments and trials are underway to see how far the potential of NDVI can be pushed across different crops and conditions.

This article originally appeared in the Summer 2025 edition of Precision Ag News - the flagship publication of the Society of Precision Agriculture Australia. It is reprinted with permission

To join SPAA go to www.spaa.com.au.

---

Thryvon trait to manage thrips Integrated management of glyphosate-resistant weeds in cotton