A new initiative designed to improve alfalfa producers’ access to precision management tools could boost the crop’s popularity and increase production, industry officials say.
Data collection has started for two new alfalfa artificial intelligence (AI) decision management tools across Ontario, Quebec, Nova Scotia and Manitoba.
The project aims to create the data base needed to develop decision making and assessment tools that could make alfalfa more competitive.
Why it matters: Unlike the annual cash crop sector, forages lack a database of information on best management practices and mitigation protocols to improve yields, nutritional values and increase winter survival rates.
In July, the Canadian Forage and Grassland Association (CFGA) received approximately $2 million in funding from Agriculture and Agri-Food Canada’s Agriculture Strategic Priorities Program to develop two AI tools. One identifies the agronomic, climatic and soil-related factors affecting alfalfa yield and nutritional value, and another assesses alfalfa’s winter survival rates.
The forage sector hasn’t kept pace with other crops in terms of capitalizing on satellite imagery and smart agriculture-type applications, said Cedric MacLeod, CFGA executive director.
That has hampered the sector’s ability to maximize the genetic potential of seed.
“Dr. Leduc – Maxime – basically took it upon himself to develop this proposal and bring together various partners, and very large and detailed projects, to try and explore this and ultimately get some decision management tools into the hands of the growers,” said MacLeod, in reference to the McGill professor leading the project.
The CFGA grant application was likely successful because it addressed a gap in service measured against the loss of forage and grassland within Canada’s agricultural landscape.
Forage accounts for 72 million acres across Canada and is a crucial ingredient for several other sectors.
MacLeod said without tools like those available for annual crops, which help to maximize yield and profits, the forage sector will continue to lose forage acres because it is less competitive.
“Anything we can do to push yield to be more efficient, more effective, raise the net revenue per acre on forage and try to keep us neck-and-neck with the annual cropping sector, the more forages we’re going to keep on the landscape,” he said. “Right now, that’s one of our main goals.”
Leduc said the forage sector lacks a central database like ones compiled for the annual crop sector or the one Lactanet Canada has for the dairy industry.
“If you need to develop a model, you need diversified data – data collected in different green conditions, with different fertilizer applications, and so on,” Leduc said. “At the same time collecting data for a new project is very expensive. It’s why in this project, we collaborate for both.”
He said there are 40 field advisors or agronomists collecting data across Quebec, Ontario, Nova Scotia and Manitoba on 80 farms for yield data and 245 measuring stem count to assess winter survival.
Numerous factors play into that process, including assessing plant biotics, chemical and physical soil factors along with weather — was there snow or no snow, and how did that impact winter kill?
“We have to build the algorithms to train the satellites to understand what it is that they’re seeing, so that we can make management decisions on how to correct actions or how to keep doing the right thing,” MacLeod said.
But ground-truthing that level of data is costly.
While $2 million is significant, Leduc said, the actual cost of the two AI tools will be closer to $4 million considering the value of universities, scientists, agronomists, private companies and producer partnerships.
The weakness of the Queen of Forages
“Alfalfa is the queen of forage in Canada, but she has a weakness, and this weakness is, as a perennial plant, it’s susceptible to winter kill,” said Leduc.
Examples that could trigger loss are nutrient deficiencies in the soil, poor cutting management or soil pH that falls outside of the ideal 6.5.
“All these small factors could have an impact on winter survival,” said Leduc. “The problem for the agronomist and producer is they don’t know which one is the real factor, and they don’t know if they have a problem or not for winter survival.”
The winter survival tool will assess risk regarding survival over winter and consider whether alfalfa management must be adjusted or confirm winterkill is simply due to weather.
The yield and nutritive value tool will take a sample analysis and interpret the cost of correcting soil fertility or restoring pH. It will also assess impact on future forage quality and estimate tonnage losses per acre if no action is taken.
As slick as AI and satellite technology is, MacLeod said, the key is identifying what’s happening in the field from an imagery perspective and then ground-truthing what the satellites see to create effective tools.
MacLeod said alfalfa is a complex crop to grow. Unlike corn, soy or wheat, alfalfa is a four-year crop that is fertilized, cut and regrown three or four times per season.
“There’s a lot of complicated pieces that go along with forage, and to build a program like this,” said MacLeod.
The project is expected to build infrastructure to get information from all sources of data collection over the next two years. Those will be used to create algorithms, followed by beta testing.
“Bottom line, it gets very expensive to do ground-truthing, so the better able we are to use satellite imagery and use AI to tell what’s happening in the field, and help us make management decisions, the better able we will be to give cost-effective management suggestions to growers in the future,” said MacLeod.