Environment

Google AI and CSIRO sequence 600 giant kelp specimens to support restoration of Tasmania’s declining underwater forests

By Kelly Lippke · November 22, 2026 · 11:51 AM · 5 min read
Google AI

Timed to the opening of COP31, Google and Australia’s national science agency CSIRO dropped results from an AI-assisted genomics project aimed at restoring giant kelp forests along the Tasmanian coast. The collaboration produced two high-quality reference genomes for Australian giant kelp and sequenced 600 wild specimens from 40 coastal sites — a genetic dataset researchers hope will lay the groundwork for large-scale replanting efforts.

Google and CSIRO publish AI-assisted kelp genomics results at COP31

The timing wasn’t accidental. Google and CSIRO chose the opening of COP31 to release findings from a multi-year partnership applying AI and genomics to one of Australia’s most urgent marine conservation problems. The centerpiece: two high-quality reference genomes for Australian giant kelp, now publicly available as a foundation for conservation genetics research.

Reference genomes give scientists a baseline — a complete genetic map to compare individual specimens against. Without them, large-scale genomic analysis simply isn’t possible. Publishing two is a meaningful first step, not a footnote.

Sequencing hundreds of specimens and identifying meaningful genetic variation across all of them isn’t something traditional lab methods handle efficiently.

The project falls under Google Australia’s Digital Futures Initiative, which funds tech-driven solutions to environmental and social challenges. Making the results open aligns with the broader push at COP31 for science-backed, collaborative approaches to climate adaptation.

Ocean warming has reduced Tasmanian kelp populations to 5% of historic levels

Giant kelp forests are in serious trouble globally. Rising ocean temperatures have disrupted the cold, nutrient-rich conditions these ecosystems depend on, triggering mass die-offs across multiple continents — and Tasmania’s situation is among the most documented and most severe.

Populations there have fallen to just 5% of their historic numbers. That’s not a gradual decline — it’s a near-collapse, and it happened within living memory. The forests that once lined the Tasmanian coast provided habitat for fish, invertebrates, and countless other species. Much of that is gone.

To understand what’s left — and what might be saved — CSIRO researchers collected specimens from 40 sites along the coast. That geographic spread was deliberate. Sampling widely increases the chances of capturing the full range of genetic variation still present in wild populations, including any traits that might confer tolerance to warmer water. At 5% of historic levels, there’s little margin for error, and restoration has to be targeted and data-driven.

DeepVariant AI tool adapted to analyze giant kelp DNA across 600 specimens

The analytical challenge here is scale. Sequencing hundreds of specimens and identifying meaningful genetic variation across all of them isn’t something traditional lab methods handle efficiently. That’s where AI came in.

Researchers trained a version of DeepVariant — Google‘s open-source AI tool originally built for human genomics — specifically on giant kelp DNA. That adaptation matters because kelp genetics differ enough from human genetics that a generic model wouldn’t cut it. Training on kelp-specific data is what made the analysis credible.

Once adapted, DeepVariant processed all 600 wild specimens. The result is a genomic dataset at a scale that would’ve been impractical using conventional sequencing approaches alone. The data maps genetic diversity still present in wild Tasmanian kelp populations and flags specimens carrying traits linked to heat tolerance — the individuals that survived as temperatures climbed, and whose genetics are exactly what restoration scientists want to work with.

Genetic data and cross-breeding studies aim to produce heat-tolerant kelp for replanting

Genomic data alone doesn’t replant a forest. The next step is turning what the sequencing revealed into actual specimens that can survive in warming waters.

Research partners are combining the genomic findings with lab-based cross-breeding studies — identifying heat-tolerant individuals in the wild population, then selectively breeding them to propagate those traits at scale. It’s a conservation genetics approach, using DNA analysis to guide which specimens get replanted rather than picking randomly. Diversity matters here for the same reason it matters in any ecosystem: a genetically uniform population is more vulnerable to disease, environmental shifts, and future warming.

Giant kelp forests do considerable ecological work. They protect coastlines from wave energy, provide complex habitat for marine species, and underpin the food webs that coastal fisheries depend on. Restoring them isn’t just about the kelp.

What this project adds to marine conservation science

Giant kelp forests rank among the most productive marine ecosystems on the planet. A single kelp frond can grow several inches a day under the right conditions, and the underwater canopies they form support fish, sea urchins, crustaceans, and dozens of other species while buffering coastlines against erosion and storm surge.

Their decline is driven primarily by ocean warming linked to climate change — making them both a casualty of the problem COP31 is trying to address and a potential indicator of restoration progress.

Conservation genetics, using DNA analysis to inform replanting decisions, is still an emerging discipline in marine environments. Most applications so far have focused on coral reefs. Applying the same logic to kelp at the scale this project achieved is a notable expansion of the approach — and one of the first large-scale uses of AI genomics tools specifically for kelp conservation.

The two published reference genomes, the 600-specimen dataset, and the adapted DeepVariant model are now available for other researchers to build on. That open availability may prove to be the project’s most durable contribution.

Kelly Lippke
Kelly Lippke

Kelly is an experienced writer with 15 years of experience exploring the big stories that shape our world, from tech breakthroughs and space exploration to climate, energy, and the fascinating quirks of science. She has a talent for turning complex ideas into sharp, memorable insights that stay with readers long after they’ve finished reading.

Kelly Writer
Kelly Lippke

Kelly is an experienced writer with 15 years of experience exploring the big stories that shape our world, from tech breakthroughs and space exploration to climate, energy, and the fascinating quirks of science. She has a talent for turning complex ideas into sharp, memorable insights that stay with readers long after they’ve finished reading.