Forty five feet down in the Moray Firth, on a seabed of rippled sand that runs flat for miles.
A steel frame settles on the bottom with a canister of mackerel strapped to it and two cameras pointed inward.
Nearby a four legged tower rises out of the sand, carrying a wind turbine 84 of its neighbors also carry.
The frame records for 45 minutes, then comes up.
What it filmed changed with distance.
What it filmed was mostly flatfish.
Why bare steel legs make this a cleaner test than usual
Most offshore foundations get a skirt of quarried rock dumped around the base to stop the current scouring a hole beside them.
That rock is habitat in its own right. It has crevices, it has surface, it holds crabs and small fish, and it is nothing like the sand it was tipped onto.
So when a study finds more fish near a foundation, the honest question is whether the fish came for the structure or for the rock pile at its foot. Most surveys cannot separate the two.
This site can. These turbine jackets were installed without scour protection, driven straight into the seabed on four piles.
Whatever effect shows up here belongs to bare steel standing in sand, which is a narrower and more useful finding than the usual one.
No rock means no confusion about what attracted them.
What the numbers came back as
Abundance and biomass both ran roughly 1.5 to 1.6 times higher 98 feet from a foundation than 787 feet out.
The change was a gradient rather than a wall. Intermediate distances sat between the two ends rather than dropping off sharply at some edge.
Twelve species appeared across the whole set. Two of them accounted for almost everything, with flatfish at 64.2 percent of individuals and haddock at 30.0 percent.
Body length showed no significant difference between near and far for either group, which rules out the simplest explanation that bigger animals hold the better ground.
The work was done in a single summer stretch in July, in daylight, with the whole array sampled at once rather than over a drifting season.
Same fish, closer together, and no larger.
Why flatfish dominating is the awkward part
The reef story people expect goes like this. Hard structure appears, it grows mussels and anemones, and reef associated fish move in to use it.
Flatfish are not reef fish. They are ambush predators that lie on and in sediment, and the sand they specialize in is precisely what surrounds these towers.
A plausible reading is that the structure is not offering them shelter at all. It is concentrating what falls off it, because everything growing on a jacket leg eventually drops to the bottom around it.
That would make the foundation a feeder rather than a refuge, which is a different mechanism with different consequences.
Haddock at 30 percent fits the conventional picture more comfortably, and haddock are also the commercially interesting half of the result.
The steel may be a table rather than a shelter.
What a factor of 1.5 does and does not mean
It is a real effect and it is a modest one, and both halves of that sentence get lost in coverage.
A ratio of 1.5 measured within a few hundred feet says where fish are sitting. It does not say whether there are more fish in the firth than before the towers went in.
Attraction and production are the two candidate explanations, they look identical on a camera, and separating them needs tagging or tissue work rather than video.
The authors say plainly that the wider ecosystem consequences of the effect are not known, which is the correct thing to say and rarely the thing that gets quoted.
The headline count and the distances were covered when the cameras came up.
The species split, the distances and the size result are set out in a marine journal.
Video finds position. It never finds production.
What would have to come next
One summer, one site, daylight only. Winter would be the obvious second look, since both these groups move seasonally.
Bait also biases what turns up. A canister of mackerel calls in what hunts by smell and misses what does not, so the species list is a list of responders rather than residents.
The useful comparison would be the same cameras at a farm that does have rock around its bases, run in the same week with the same gear.
That would finally split the steel from the stone, and nobody has published it.
Years of tracking around turbine legs elsewhere have shown the structures working as reefs, which makes the flatfish result stranger rather than less so.
The site details, the jacket design and the water depth are described by a reference entry.
Twenty four turbine jackets, 96 drops, and the animal that showed up most was not a reef animal.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
