I spent three days last month in Cockburn Sound, Western Australia, watching divers plant Posidonia australis seedlings into bare patches of sand that used to support continuous meadows before the 1970s industrial dredging campaigns. The work is slow, expensive, and — based on the numbers I’ve been pulling together — finally starting to show results worth talking about.

Seagrass restoration has been one of those marine conservation areas where ambition has consistently outrun outcomes. For decades, the standard story was that restored plots failed within two to three years, often without anyone publishing the failure data. That’s been changing, and 2026 is a useful checkpoint for assessing where things actually stand.

The Cockburn Sound numbers

The Western Australian Seagrass Restoration Network released their consolidated 2025 monitoring report in February, and a few figures jumped out. Survival of Posidonia transplants at the five-year mark across the Cockburn Sound trial sites is now sitting at 47 percent. That sounds modest until you compare it to historical baselines from the 2000s, where five-year survival rarely exceeded 15 percent.

What changed? A few things, layered together. Substrate preparation got more rigorous — divers now spend significant effort preparing the seabed before planting, removing drift macroalgae and confirming sediment depth. Seedling sourcing shifted from rhizome fragments toward fruit collection and laboratory germination, which produces more genetically diverse stock. And site selection has gotten brutal. Projects now reject candidate areas that don’t meet specific light, current, and depth criteria, even when those areas seem otherwise suitable.

The result is fewer hectares attempted but a much higher proportion succeeding. From a carbon storage perspective, this matters. Mature Posidonia meadows in this region sequester carbon in sediment at rates that can exceed many terrestrial systems, and the long-term storage stability is significantly better than forest carbon because the buried material doesn’t burn.

Eastern coast efforts have been harder

Restoration work in NSW and Queensland has had a rougher run. Zostera and Halophila species in the eastern estuaries are dealing with a different set of pressures — agricultural runoff, recurrent flood events that smother seedlings with sediment plumes, and warming events that have stressed even established meadows.

The Sydney Institute of Marine Science published interim results from their Botany Bay restoration trial in late 2025, and the picture is mixed. Initial establishment looks reasonable, but two of the three pilot sites lost significant cover during the February 2025 flood event. This isn’t a failure of the technique. It’s a question of whether the surrounding watershed conditions can support seagrass at all under the current climate trajectory.

I want to be honest about this. Some of the sites we’re trying to restore probably can’t be restored in any meaningful way without changes to upstream catchment management. Planting seagrass into water that will repeatedly receive heavy sediment loads is a Sisyphean exercise.

The role of remote sensing and AI in monitoring

One genuinely useful development is how monitoring has scaled. Drone-based multispectral surveys, combined with machine learning classification of meadow extent and condition, are allowing project teams to track changes across hundreds of hectares without sending divers to every plot. The CSIRO’s seagrass mapping toolkit, released in updated form in late 2025, is being used by most major Australian projects now.

For organizations setting up these monitoring pipelines, working with an AI consultancy that understands both marine data and the limits of automated classification has become more common. Getting the model to distinguish seagrass from macroalgae from bare sand under varying water clarity is a non-trivial problem, and off-the-shelf tools tend to underperform in turbid Australian conditions. The teams that have invested in proper validation pipelines are getting useful data. The ones that trusted vendor demonstrations are mostly redoing their analyses.

The Posidonia clone discovery and what it implies

A reminder for context. The 2022 discovery in Shark Bay of what may be the largest single clonal organism on Earth — a Posidonia australis meadow covering around 180 square kilometers — changed how some restoration practitioners think about genetic strategy. The Shark Bay clone is genetically uniform across an enormous area, suggesting that under the right conditions, single genotypes can persist and expand for thousands of years.

This has implications for restoration. Some teams have shifted toward planting fewer, well-chosen genotypes in conditions matched to their tolerance profile, rather than attempting to maximize genetic diversity for its own sake. The debate isn’t settled, and there are good arguments on both sides. But the conversation has gotten more sophisticated than it was even five years ago.

What I’m watching for the rest of 2026

Three things, mainly. First, the next round of UNESCO marine heritage assessments due later this year will include several Australian seagrass-relevant sites, and the assessment language tends to drive funding. Second, the carbon credit market for blue carbon projects is finally starting to price seagrass restoration credits at levels that make commercial-scale work viable, though the methodology disputes are ongoing. Third, I’m watching to see whether the Cockburn Sound results hold at the ten-year mark. Five-year survival is encouraging. Decadal persistence is the real test.

The honest summary is this: we now know how to restore some seagrass in some places. Whether we can do it fast enough, and at the scale needed to compensate for ongoing losses, remains an open question. Worth watching closely.