Optimism towards price negotiations, the role of innovative technology, and new tools for technical challenges were all highlighted during the recent CUPGRA potato conference. CPM joined delegates to find out more.
“I’d suggest growers are in the best position to sit down and negotiate fair and sustainable pricing that I’ve seen during the past 20-30 years.” -ANGUS ARMSTRONG
By Mike Abram
Potato growers are currently in a much stronger position to negotiate fair prices, according to Angus Armstrong, former chief executive officer of Greenvale AP.
Speaking at the 35th annual CUPGRA potato conference, he suggested security of supply is a key battleground for packers and processors to secure good, reliable growers for their raw material requirements.
This is because volatile markets during Covid, higher costs largely attributable to the Ukraine war, plus soil damage after wet harvests deterred landlords from renting out for potato production, has seen growers leave the sector.
“Consequently, I’d suggest growers are in the best position to sit down and negotiate fair and sustainable pricing that I’ve seen in the past 20-30 years,” he said. “The crop is in demand and there shouldn’t be any over-supply.”
To take advantage of the situation, he stressed growers require a solid understanding of their true cost of production including storage costs, associated weight loss, and a realistic margin. “On the back of good accurate costings you don’t commit to contracts that aren’t viable.”
PROFESSIONALISM
Strive to be in the top quartile – or at least the top half – of growers for your main customer, he added. “Every buyer wants to feed off good quality crops grown by professional growers.
“Communication is key; get input from your buyer but be proactive. Set review dates to discuss what’s working in your business, their business, how you’re aligned and what can be better. Don’t leave it to chance,” he urged.
James Green, group director of agriculture for G’s Growers, provided insights for how potato growers could improve their businesses by drawing on an example from outside the sector. His pointers included investing in infrastructure and technology to improve efficiency and reduce labour costs, especially the use of automation and data-driven systems to optimise yields and reduce waste.
G’s has made significant investment in digital agriculture, for example, to improve productivity as costs grew without much increase in selling prices, explained James. In fact, 10 years of development began with an initial question of how to have more crop available for customers to remove some of the peaks and troughs in supply.
Working with Microsoft Research in Cambridge led to G’s initially developing its ‘Ice CAM’ model, using planting date, temperature and sunlight to predict harvest dates for its iceberg lettuces – helping to manage shortages or surpluses for the all-year harvest.
“Part of the forecast isn’t just when, but how many, and we soon realised what we thought was in the field, wasn’t,” said James.
That led to work with Cranfield University counting lettuces with drones. “A 3-4% difference in establishment on 50M iceberg lettuces adds up through the season.”
Next came managing size using precision fertilising technology based on a 30x30cm grid. “We can apply per plant fertiliser where we treat only the smaller plants and not the big ones, which has driven up to a 50% saving in nitrogen and created a more uniform crop.”
James added that the latest project for G’s uses AI and machine learning to understand and manage more factors that affect uniformity and quality. “Data is power. Hang onto yours because it’s valuable and if you’re not collecting it, start, because the simplest data can be powerful when you see the patterns within it,” he said.
New technologies using data are also starting to gain traction directly within the potato industry. Dr Joseph Mhango, a senior lecturer in applied data science at Harper Adams University, highlighted how combinations of machine learning, remote sensing and AI could improve the accuracy of growth models used to predict yields by dynamically adjusting parameters based on real-time data.
For example, ground cover estimates could be improved using AI analysis of drone images to overcome biases in traditional methods which rely on setting thresholds for soil and plants, he explained.
Ground cover is used as a proxy for light interception within the growth model but could be replaced altogether by AI models that accurately predict absorbed radiation from satellites, he added.
A combination of machine learning and radar is helping to predict 50% emergence dates which is useful for initial yield predictions, estimating chronological age and managing irrigation for scab control, said Joseph.
“Using crop models for deciding harvest dates requires accurate dry matter prediction, but, while most models use a general conversion factor, the range of dry matter concentration in the field is wide enough to create significant errors.
“Whereas we can use machine learning to dynamically predict dry matter concentration to understand how fresh weight develops over time with respect to inputs.”
Other techniques such as ground penetrating radar could be used to assess tuber size and distribution without undertaking test digs. That technology might also be useful for precision de-stoning, he said.
It’s technology that could be used to help address the challenge of stone content within the Innovate-UK funded Potato-LITE project, noted Mac McWilliam, the project lead from PepsiCo.
The four-year cross-sector project is focusing on how regenerative agricultural practices, particularly lower intensity cultivation, can be implemented in the potato crop. One of the project outputs will be to use the data to create a decision support framework where factors such as end use, movement date and therefore risk of bruising, soil type and stone content perhaps determined by technology, are fed in to guide cultivation strategies, explained Mac.
Another decision support tool was introduced at a CUPGRA conference workshop, which is looking at the latest potato cyst nematode research. With a working title of ‘PCN Pro’, a model has been developed by PCN Action Scotland to replace AHDB’s PCN calculator. Its aim is to incorporate some of the features of successful Dutch tool NemaDecide, such as a large variety database and a cost benefit analysis for different scenarios.
Users input a starting population for each PCN species, potato variety, rotation length, any treatment, region, a start or planting date and an end date for how long to run the model, explained Anglia Ruskin University’s Dr Marcus Bellett-Travers.
“What it calculates is the active PCN population for each species in the soil over time as influenced by management practices, and the impact on yield.”
DEVELOPMENT PROSPECTS
Currently trained on mostly Scottish data, for both calculations it assumes average weather for the region. Future iterations could allow the use of more location-specific data while the impact of cover crops on control is another factor likely to be added.
Demonstrating the model, Marcus showed how varieties with different tolerance or resistance to Globodera pallida or Globodera rostochiensis affected PCN populations over time, predicted yield impact, and the influence of other management practices.
Earlier in the workshop, James Hutton Institute researcher Dr James Price explained his latest research on the genetic basis of varietal resistance to PCN. “It’s more complicated than just being two species – there are also different pathotypes within each.”
A pathotype is a population that can interact differently with a host plant, usually exhibiting different virulence, he explained, which is important when considering varietal resistance particularly for G. pallida.
While typically there’s a single score for potato variety resistance, that could vary depending on what pathotypes are present. For example, Lanorma, which has a resistance rating to G. pallida of six, showed good reductions in egg numbers after being grown in a Potato Partnership (TPP) trial in England but had the opposite result in a Scottish PCN Action trial.
The discrepancy in performance was likely due to the presence of different PCN pathotypes with PA2 and PA3 identified in the English trial, while it’s suspected that PA1 was present in Scotland, said James. “That six score could be for PA2, PA3 but might only be a two for PA1, for example.”
Such inconsistency underscored a necessity to also know the specific pathotypes in a field when selecting a variety for PCN management. To that end, the Hutton Institute is developing simple low-cost PCR tests which could distinguish between different pathotypes, concluded James.
This article was taken from the latest issue of CPM.
For more articles like this, subscribe here.
Sign up for Crop Production Magazine’s FREE e-newsletter here.
