18  / SEEDWORLD.COM  SEPTEMBER 2026
mining which genetics will perform best 
across thousands of acres and countless 
environmental conditions.
She says testing remains essential, but 
it also requires enormous investments of 
time and resources.
“Testing in many environmental sce­
narios is the most expensive thing we do,” 
Honold says. “It also takes a long time to 
really characterize and make sure that we 
know exactly how our seeds and traits are 
going to behave.”
That reality is driving increased inter­
est in predictive breeding tools that allow 
researchers to evaluate performance 
before seeds ever reach a grower's field.
“How we use those resources, that’s 
where AI comes in,” she says. “And then, 
how we superpower those decisions and 
that data is again where AI comes in.”
Digital twin technology combines 
environmental, weather, geographic and 
genetic data to simulate performance 
across a wide range of conditions.
“It’s a set of models that take in [...] cli­
mate information, geographical data, but 
also weather data and all these different 
inputs,” Honold explains. “It basically simu­
lates performance on any given acre.”
The concept has roots in other 
industries, but agriculture is a unique 
challenge because researchers are 
trying to model biological systems and 
environmental variability simultaneously.
“For us, it’s much more complex 
because it's environmental scenarios. 
We're simulating nature,” Honold says. 
“But it’s incredibly effective because we 
The Race to 
Accelerate Genetic Gain
New tools, predictive models and systems thinking are 
helping breeders move innovation faster while preparing 
crops for a more complex future.
By Aimee Nielson, Seed World U.S. Editor
PLANT BREEDERS HAVE always 
chased the same goal: genetic gain.
Every improvement in yield potential, 
disease resistance, stress tolerance or 
agronomic performance traces back to a 
breeder's ability to identify better genet­
ics and move them into farmers' fields. 
What has changed is work speed.
Advances in AI, predictive modeling, 
gene editing and data analytics help 
breeders evaluate genetics, understand 
environmental interactions and make 
decisions earlier in the breeding process.
Bayer North America breeding lead 
Julia Honold believes the convergence of 
predictive breeding, AI and gene editing 
could create a major jump in genetic gain.
“The expectation is that the rate of 
gain continues increasing,” she says. 
“There were paradigm shifts, like when we 
first started using hybrids, and you could 
see the gain just shot up. This is going to 
be the next one.”
For breeders, the next leap forward 
is not about replacing experience and 
intuition. It is about combining them with 
tools that help researchers evaluate more 
possibilities than ever before.
A New Era of Prediction
At its core, plant breeding remains 
remarkably simple.
“Breeding at its core is about taking 
diverse variation and then selecting the 
best version of that,” Honold says. “How 
we’ve done that over the past two dec­
ades has greatly evolved.”
The challenge has always been deter­

View this content as a flipbook by clicking here.