There is an assumption built deep into conservation science: if a species crashes to a tiny remnant population, it is permanently damaged. The genetic diversity it loses in that crash is gone forever — and the species will carry the consequences forever, through inbreeding, reduced adaptability, and vulnerability to disease.
A new study published this week in the journal *Science* suggests that assumption may be wrong — at least for one iconic Australian species.
**The crash**
In the early 20th century, koalas were in freefall. The fur trade had devastated their numbers across southern Australia. In Victoria, the situation was catastrophic: by around 1900, the entire state population had been reduced to an estimated **500 individuals**.
From millions to 500. The bottleneck — the genetic term for a population squeeze that wipes out genetic variety — was severe.
Conservation efforts over the following decades slowly reversed the population collapse. Small numbers of koalas from Victoria were relocated to nearby islands for protection. From those islands, animals were reintroduced to the mainland as populations stabilised. The numbers climbed. By 2020, Victorian koala numbers had recovered to approximately **500,000 animals** — a thousand times the low point.
But the assumption remained: even with half a million koalas, the genetic damage from that 500-animal crash was baked in. The gene pool was shallow. The animals were genetically similar in ways that would matter for their long-term survival.
**What the genomes show**
Researchers at the University of Melbourne, working with an international team, analysed whole-genome data from **418 koalas across 27 populations** in Australia — the most comprehensive koala genomic study ever conducted.
In the Victorian populations — those that descended from the 500 survivors — something unexpected was happening.
Rapid population growth was producing increased **genetic recombination** — the process by which chromosomes shuffle and exchange segments during reproduction. More animals reproducing means more opportunities for new gene combinations to emerge. And as new genetic variants arose, natural selection had more material to work with.
The result: genetic diversity in Victorian koalas is **actively recovering**. The bottleneck created by the crash is being partially repaired — not by time alone, but by the process of rapid reproduction and population expansion.
'Rapid population growth can facilitate genetic recovery,' the researchers wrote. 'Conservation decisions should not rely solely on static measures of genetic diversity.'
**The contrast**
The study also found a sobering counterpoint.
Koala populations in northern Australia — Queensland and New South Wales — had historically maintained higher genetic diversity than their southern counterparts. But those populations are now in decline, facing habitat loss, disease pressure, and fragmentation. And as they shrink, their genomic health is deteriorating: accumulating harmful mutations, losing adaptive potential.
The genetic health of a species, this research suggests, is not fixed at any moment in time. It moves with the population. Recovery — real, functional recovery — is possible.
**What it means for conservation**
The implications extend far beyond koalas. Across the world, conservation programmes manage species that have survived through small population bottlenecks — cheetahs, northern elephant seals, black-footed ferrets, Iberian lynx. For all of them, the question of genetic recovery matters deeply.
This study suggests that for species capable of rapid reproduction, the genetic damage from near-extinction may not be irreversible. Population growth itself can be a form of genetic repair.
For the koalas of Victoria, it means that half a million animals with recovering genomes are writing a different story than the one scientists expected.
The fur trade nearly ended them. Their DNA is still remembering how to be wild. 🐨
*Sources: Science (March 2026) · University of Melbourne · Smithsonian Magazine · Science News · Nautil.us · ScienceDaily*
