Plant breeders wanting to fast track how they identify and develop suitable grass varieties for commercial release are using techniques to help speed up the process – one well known to anyone in the livestock sector, the other more familiar to surveyors and autonomous vehicle makers.
The research work is part of Pastoral Genomics, a joint funding initiative including investment from Beef + Lamb NZ, DairyNZ, Dairy Australia, AgResearch and seed companies AgriSeeds and PGG Wrightson Seeds with matching investment from Ministry of Business, Innovation and Employment.
The process involves DNA profiling to predict desirable commercial traits in forage plants. It is now well founded in livestock breeding.
Traditional methods would have involved waiting for a plant to be tested in the field for three years, selecting and breeding from it then seeking out the offspring exhibiting the desirable commercial traits.
“The work we presented on at this year’s Grasslands Conference was really about proof of concept,” lead researcher Brent Barrett said.
“We chose a model trait, in this case ryegrass flowering and selected using genomics for that.
“Our work showed genomic selection was accurate across multiple populations of grasses and across generations. This was after being tested across two generations of grasses.”
The technique delivers a double win for researchers, speeding up the selection cycle and increasing the number of progeny they can look at, therefore, the intensity of the selection process.
“Genomics allows us to breed two generations in a year versus one every three years using traditional breeding so that offers a sixfold increase in the speed of creating higher merit varieties.
“With the traditional method you are also limited by the number of plots or plants that can be accurately observed.
“Genomics allows us to scale those numbers up – all you need is a seedling, its DNA sample and match to the DNA profile you are looking for.”
Researchers are sticking to the base trait expected from ryegrass in the Forage Value Index – its drymatter yield.
Barrett is optimistic his team will have early results on the field progeny tests for grass dry matter yield by this time next year. They will be the physical proof the plants selected genomically deliver the higher yields the genomics predicted.
“We can use the same approach to help breeders select plants with greater persistence, lower environmental footprint, higher quality or any other trait of value.”
“This work is conceptually simple, comparing a plant’s DNA to some desired commercial data points on the DNA, but does require some intense and powerful computing capacity.”
Sitting alongside the genomics is another technology-intensive piece of kit being developed by a group led by Dr Kioumars Ghamkhar.
It is a self-propelled plant scanner with the capacity to travel 1km an hour, scanning hundreds of plants in research plots and identifying those exhibiting the greatest drymatter yield.
At the machine’s heart is LIDAR (light imaging detection and ranging) technology which effectively makes the machine a super fast and accurate plant technician capable of scanning thousands of plants a day to measure their drymatter.
“It is essentially a robotic eye that trundles along.
“It collects data with a high level of accuracy, with greater than 85% match to actual drymatter yield, which is encouragingly high.”
For researchers the LIDAR machine is a dream come true.
Rather than spending hundreds of hours cutting pasture samples to weigh, dry and measure, the machine does it in a fraction of time.
“And because it does not involve cutting the plant you can see how that plant behaves in a real grazing environment. How the plant responds over post-grazing recovery can be measured by grazing the trial then putting the machine back in to accurately identify what plants recover the quickest.”
The use of the technology for grass yield measurement is the most accurate work Barrett knows has been done anywhere in the world.
In future deeper artificial intelligence (AI) applications will evolve as the machine collects greater amounts of data and has the ability to learn more about sample characteristics.
One ability that has Barrett and his team particularly excited is the potential ability of LIDAR to detect a plant’s response to stress and subsequent stress recovery.
“This opens up the possibility for looking at other traits that relate to efficiency of growth, such as finding cultivars that are water-efficient and climate-ready.
“Work done already shows there can be a twofold difference between plants in their water efficiency but it is not a trait that has been selected before, simply because of the work, time and cost to do so. With genomics and LIDAR, that is now possible.”
Barrett acknowledges compared to animal breeding or even the gains in maize breeding, grasses have gained efficiencies relatively slowly, at a global rate of only 3-4% a decade.
“But now, with animals being more efficient and effectively capable of generating the same protein with fewer of them over time, these techniques enable us to also offer pasture plant varieties that can keep up with those higher producing, hungrier animals.”
