Breeding, registration and seed certification can no longer operate as isolated steps in an era of continuous data and rapid innovation.

The most important change coming to plant breeding may not be faster science. It may be a completely different relationship with time.

For most of its history, breeding has operated according to a stubborn biological clock. Brian Rossnagel, the celebrated barley and oat breeder, captures that reality perfectly when he talks about choosing parents for a cross. Get that decision wrong, and you may not know for 10 or 15 years.

That has shaped virtually everything around plant breeding. Programs are built for long horizons; variety registration evaluates innovation after years of developments, certification protects the identity of the finished product, and investment decisions are made knowing the payoff could be more than a decade away.

But what happens when the biology remains slow while everything surrounding it becomes fast? Luke Dojack’s lentil research offers a glimpse of that future. UAV imagery, computer vision, soil sensors, root imaging, microbiome data and artificial intelligence are giving breeders ways to detect patterns that would once have taken years of field observation to recognize.

AI cannot make a lentil plant grow faster. It can, however, help identify a bad bet much sooner. That distinction could transform the economics of breeding.

The greatest cost in a breeding program is not simply the money spent developing successful varieties. It is the time, land, labour and opportunity consumed by thousands of candidates that ultimately fail. If new tools can eliminate poor prospects earlier, their biggest contribution may not be shortening the biological cycle, but freeing breeders to place more bets.

That changes the policy discussion. We often talk about plant breeding investment as though the goal is to put more money into the existing system, but perhaps the larger opportunity is to redesign the system around a new reality: breeding is becoming a data business built on top of biology. That means the organizations surrounding plant breeding may have to reconsider what they actually do.

Variety registration cannot simply be the gate at the end of the process. It may need to become part of a continuous evidence system that begins much earlier. Seed certification cannot remain only a mechanism for confirming varietal identity. In an increasingly digital marketplace, it could become the trust architecture connecting genetics, production records, seed movement and performance information.

Public research institutions cannot be treated merely as places that release varieties. They are also the source of long-term datasets, hard-to-commercialize crop research and scientific capacity that the private sector cannot easily reproduce.

Private investment cannot be expected to operate on 15-year timelines when the rules surrounding commercialization, intellectual property and market access remain uncertain. This is where the comments from Seeds Canada’s Curt Baldwin and Canadian Seed Growers’ Association President Shannon Bieman become especially important. Both are grappling, in different ways, with institutions created for stability that must now learn to operate in an age of acceleration.

That will require more than digitizing existing paperwork or shaving a few months from an approval. It requires asking a more uncomfortable question: if we were building Canada’s seed system today, knowing what we know about AI, genomics, climate volatility and data, would we build the one we have? Probably not. We would likely design something more connected and less sequential. Breeding, testing, registration and certification would not function as separate stages in a long relay race; instead, data would move between them continuously. Evidence collected during development could follow a variety through registration, seed production and commercial use.

The seed industry often describes this process as a pipeline. That metaphor may now be holding us back. Pipelines are linear. They move one thing in one direction through a fixed series of gates. The future looks more like a network. Breeders, seed growers, regulators, farmers and researchers will create and exchange information throughout the life of a variety. The competitive advantage will belong to systems that can turn that information into better decisions without compromising trust.

Rossnagel’s 15-year bet will not disappear. Biology will continue to demand patience. But patience should belong to the plants, not the institutions around them. Canada’s opportunity is not simply to breed faster, but to build a seed system that learns faster than the world around it changes.