What an effective population size actually tells you

Every spring the Rare Breeds Survival Trust publishes its watch list, the equestrian press reports it, and somewhere in the third or fourth paragraph a number appears with a parenthesis after it. Effective population size — a measure of genetic diversity, the parenthesis usually says, not a population total. Then the article moves on.

That parenthesis is doing an enormous amount of work, and we have leant on the number ourselves — in the Suffolk Punch's good few years and in the man who counted the Cleveland Bays — without ever stopping to explain it properly. It is worth explaining, because it is the closest thing the rare-breed world has to a single honest summary of a population's health, and because a buyer who understands it will ask better questions than one who does not.

It is not a headcount, and the gap is the point

Effective population size, written Ne, is not how many animals exist. It is the size that an idealised, perfectly-behaved breeding population would have to be in order to lose genetic diversity at the same rate this one is actually losing it.

Everything follows from that framing. The real population is never idealised. Some animals never breed. Some breed once. A small number breed a great deal. The animals are not evenly split between the sexes in the breeding pool, and in almost every horse and dog population they are wildly not, because one stallion or one stud dog can cover a great many mares or bitches and a mare or a bitch cannot reciprocate. All of that unevenness makes a population behave, genetically, as though it were much smaller than it is. Ne is the number that measures how much smaller.

The simplest version of the arithmetic makes the mechanism visible. For a population with a given number of breeding males and females, the idealised estimate is four times the males times the females, divided by the sum of the two. Put a hundred breeding mares and four working stallions into that and the answer is not a hundred and four. It is about fifteen.

Now invert it. Fifty mares and twenty-five stallions — seventy-five animals, far fewer than the hundred and four above — gives about sixty-seven. Three-quarters the headcount, four times the effective size.

That is the whole idea in two sums. The scarcer sex is the ceiling. A breed does not run out of genetics when it runs out of animals; it runs out when it runs out of the animals it is actually willing to breed from, and the count that matters is the count of distinct contributors, not of registrations. Real Ne figures are calculated from pedigree data by rather more careful methods than this — usually from the observed rate of inbreeding across generations, which catches things the simple formula misses — but the formula is honest about where the pressure comes from.

Why fifty is the line

The United Nations Food and Agriculture Organisation treats an effective population size of 50 as a floor. Below it, inbreeding accumulates faster than about one per cent per generation, which is the rate at which the losses stop being theoretical and start showing up as reduced fertility, smaller litters and foal crops, and the quiet re-emergence of recessives that nobody was tracking. The wider conservation-genetics version, the so-called 50/500 rule, adds a second and much higher line: around 50 to avoid inbreeding damage in the short term, around 500 to retain enough variation for a population to adapt over the long one.

Almost no British native equine breed is anywhere near 500. Several are below 50. Read the watch list with that in mind and it stops being a league table and becomes a set of distances from a floor.

What 27 looks like from the outside

The Hackney Horse and Pony is the case that prompted this piece. Its effective population size fell to 27 in 2025, from 52 the year before — halved in twelve months. In the same year the number of dams producing registered progeny fell by 41 per cent, and the number of breeders by 37.5 per cent. The RBST's chief executive called the situation critical, which is not a word that body uses loosely.

There are considerably more than 27 Hackneys in Britain. You can go and watch them; the breed is a fixture of the harness classes and photographs beautifully doing what it was built to do. Nothing about a show ring communicates a figure of 27. That is precisely why the number exists — a breed can be entirely visible and simultaneously be running on a genetic base that a single bad decade would finish.

The breeder figure deserves its own sentence. Losing 37.5 per cent of the people who breed a rare animal is not the same kind of loss as losing 37.5 per cent of the animals, and it is usually worse. Animals can be bred back. A breeder who stops is a set of decisions, a line of mares, a stallion nobody else was using and forty years of knowing which pairings not to make, and none of that is recoverable by anyone registering a foal next season. Every departure also concentrates what remains: fewer people making more of the choices, which is exactly the pressure that pushes Ne down.

The number is a property of the population, not of the animal

This is where it matters for anyone buying.

A hip score belongs to a dog. A five-stage vetting belongs to a horse. A DNA result belongs to the animal that was swabbed, and can be handed over, framed, and linked from a listing — which is most of what we are building GenoVaq to verify. Effective population size belongs to none of them. It is a fact about the whole stud book, computed by a breed society or a conservation body from years of pedigrees, and no individual animal can be clear of it or carry it.

So there is no certificate. There will never be a badge. This is the same wall we ran into writing about the Sealyham's health list, where the governing body puts prioritising genetic diversity above the DNA test on the minimum standards — an instruction with no laboratory behind it — and the same argument we made from the other side in why rare breeds don't breed the carriers out.

What a buyer can do is ask questions that a low Ne makes newly relevant, and none of them are rude:

• How many stallions or stud dogs were used in this breed last season, and how many litters or foals did the busiest one account for? The answer to the second half is the popular-sire effect we wrote about in what popularity costs a breed, and it is the single largest depressor of effective size.
• What is the coefficient of inbreeding for this pairing, and how does it sit against the breed average? What a coefficient of inbreeding actually tells you covers the reading of it. Against a small Ne, the breed average is itself climbing, so a figure that looks unremarkable this year may be the average precisely because everyone is now related.
• Is there anything in this animal's pedigree that is rare in the breed as a whole? In a population under real pressure, an unfashionable line is an asset rather than a compromise — and a breeder who can answer this fluently is telling you they think about the population and not only the mating.
• Has the society or the RBST published a breeding recommendation for this breed, and where does this pairing sit against it?

A breeder working in a Priority breed will almost certainly have opinions on all four. The questions read as informed rather than suspicious, which is the register worth aiming for generally — see questions to ask a breeder before you commit.

The honest limit of what a marketplace can do

We cannot verify an effective population size. Nobody can verify it per animal, because it is not that sort of number, and a platform that offered a diversity badge would be selling a reassurance it had no way to stand behind.

What a marketplace can do is publish the inputs. A coefficient of inbreeding, shown rather than described. A stud's progeny record, so that a buyer can see for themselves how much of a generation one animal has already accounted for — the argument of what a progeny record actually tells you, and the standard set for us by the Clumber Spaniel society's published stud list. Health results with the certificates behind them. Enough of the record, in other words, that the person reading it can do the reasoning that no badge could do for them.

The figure of 27 is not a verdict on the Hackney. It is a description of how much room the people breeding it have left, and it went from 52 to 27 in a single year largely because a number of them stopped. That is a reversible sentence. It reverses through more breeders using more stallions on more mares, deliberately and unfashionably, which is slow, unglamorous work that no marketplace performs and every marketplace depends on.

If you breed a Priority or At Risk breed, sit on a society committee, or would rather we heard the state of your stud book from you than reconstructed it from published figures, the address is info@genovaq.co.uk.

---

Sources

• Emily Bevan, "Fears grow as Hackney numbers described as 'critical' in latest RBST watchlist", Your Horse, 23 April 2026
• Equine breeds — RBST Watchlist 2026-27
• Hackney Horse and Pony — RBST Watchlist
• In situ conservation of livestock and poultry — Food and Agriculture Organisation of the United Nations
• Animal genetic resources — UK Biodiversity Indicators, JNCC

— Rene