There is a number I cannot stop turning over. The breeding population of cod left in the Firth of Clyde is estimated at around 20 tonnes. Not 20 tonnes landed in a good year, or 20 tonnes swimming somewhere offshore – 20 tonnes in total: the entire spawning stock of a fish that once made this corner of Scotland one of the finest angling and commercial grounds in Britain. You could fit it in three builders’ skips.

As a marine biologist, I was taught to think of the sea as productive but finite – a thing to be husbanded, not mined. The Clyde tells a harder story than most. Within living memory this was somewhere people travelled to fish: charter boats running out of Clyde towns like Largs and Rothesay, tackle shops, B&Bs, all the quiet economy that grows up around water full of fish. Then the cod went, and most of that went with them. What unsettles me is not only how far the stock has fallen, but how stubbornly it refuses to climb back, despite the measures meant to help it.

To understand why, you have to look at what else is being caught there. The Clyde’s commercial fishery today is built largely on Nephrops – the langoustine on a restaurant menu, the Dublin Bay prawn in a fishmonger’s window, the scampi in a basket by the bar. It is one of Scotland’s most valuable seafood resources, and there are two very different ways to catch it.

The first is bottom trawling: mobile gear dragged across the seabed, capable of landing high volumes with a crew of two. It is efficient, and it is indiscriminate. It disturbs the seabed it crosses, and it catches what it is not aiming for – including juvenile cod. The best available science on the Clyde stock points to bycatch in the Nephrops trawl fishery as the dominant source of cod fishing mortality since 2005. The cod are not, for the most part, being targeted. They are being caught on the way to something else, and dying as a footnote in another fishery’s accounts.

The second way is creeling: baited static pots set on the seabed and hauled back later. It is slower, and it is far more selective. It barely touches the wider seabed, it does not sweep up juvenile fish in the same way, and – crucially – live creel-caught langoustine fetches a far higher price than trawled product bound for processing. In the Clyde, creeling accounts for only around eight per cent of the fishery. The lower-impact, higher-value method is the minority partner, by a wide margin.