In 2009, nine Iberian lynx were released from their transport boxes at dawn in a scrubland valley in Spain’s Córdoba province. The animals were bred in captivity, and the release site had not seen the species in decades. Roughly half of similar captive-bred animals died within their first year in the wild, and with fewer than 100 lynx remaining, mistakes were costly. Spain had already invested more than 70 million euros in the species’ recovery.

Four months later, field ecologist Francisco Palomares reviewed camera footage from the valley. The lynx were present, but another expected species was absent. Red foxes had disappeared from the monitored grid squares entirely. This was not the result of predation. It was the first sign of a larger ecological pattern.

The Iberian lynx is smaller than a German shepherd, heavier than a house cat, and covered in spotted rather than striped fur. It hunts at dawn and dusk and lives only on the Iberian Peninsula. Fossil records show continuous presence in the region for about 10,000 years. In the 1940s, researchers counted about 15 distinct subpopulations across Spain and Portugal. The species had survived ice ages, droughts, and the fall of empires. Then a government price list changed its trajectory.

In 1953, Spain’s Ministry of Agriculture under Francisco Franco issued the Alimañas law, a pest control decree. It designated species the state wanted removed from the countryside: wolves, foxes, eagles, kites, vultures, and the Iberian lynx. Each carried a bounty paid upon presentation of the carcass. A wolf brought 15 pesetas. A pregnant wolf brought 20. The Iberian lynx was priced at three pesetas and 75 céntimos, about a quarter of a wolf’s value, and less than a day’s wages. The state’s classification was clear: the lynx was not a threat but an inconvenience.

Records from that campaign show the scale of the culls. About 2,000 wolves, 55,000 foxes, 5,000 eagles, 6,000 hawks, and 1,000 vultures were eliminated. Lynx deaths were not counted separately. The rural Spain of the 1950s was poor and reliant on subsistence farming, and bounty systems were common across Europe at the time. What made Spain’s version destructive was not the principle but the scope and the fact that it targeted a species with no margin left.

Spain granted the lynx legal protection in 1973, and Portugal followed in 1974. By then, 15 subpopulations had collapsed into two. By 2002, those two held just 94 animals, with about 25 breeding females. Twenty-nine years of protection had not reversed the decline, only slowed it.

In 2019, a hunter in Castilla-La Mancha shot a collared female lynx named Nova, an individually monitored animal with years of research history. He claimed he mistook her for a fox. The court ordered him to pay 100,800 euros in compensation and banned him from hunting for three years. The contrast with the 1953 bounty price is stark: the same species, the same country, seven decades apart.

The bounty law explains only part of the collapse. The deeper cause was the loss of prey. Most large predators are opportunistic. Wolves switch from deer to wild boar; bears switch from salmon to berries. The Iberian lynx lacks this flexibility. Rabbits make up 85 to 95 percent of its diet, and a breeding female needs one to three rabbits per day to raise kits. When rabbits vanish, lynx vanish within one or two breeding seasons.

In 1952, myxomatosis reached Spain. The virus had been introduced elsewhere in Europe deliberately to control rabbit populations on farmland. Within three years, it killed more than 95 percent of European rabbits across the Iberian Peninsula. The disease arrived in 1952; the pest control law came in 1953. Over the next decade, Spain paid bounties for killing a predator while a viral epidemic eliminated its only prey. No one coordinated the two events. That is precisely why they worked so well together.

In the late 1980s, rabbit hemorrhagic disease delivered another blow to recovering populations. In the 2000s, a new variant called RHDV2 appeared, one that killed kits, including those vaccinated against the earlier strain. Mortality rates in affected populations ranged from 70 to 90 percent.

Field data from the remaining lynx strongholds was grim. In Doñana National Park, rabbit numbers fell by more than 80 percent between 2012 and 2013. In Andújar Natural Park, the decline was about 75 percent between 2010 and 2013. Studies tracking both species found a clear correlation: a 60 to 70 percent collapse in rabbits corresponded to a 65 percent decline in lynx breeding success. The two species are tightly linked, not loosely associated.

The rabbit is not just a food source. In Mediterranean scrubland, it functions as an ecosystem engineer. Its burrows aerate the soil, and its grazing maintains the mosaic of dense scrub and open ground that dozens of vertebrate species depend on, including the Spanish imperial eagle, the red kite, and the stone curlew. Remove the rabbit and the landscape itself begins to change.

Spain understood most of this by the 1990s and responded with funding, programs, and engineers. European Union LIFE program money began flowing in the mid-1990s. The El Acebuche breeding center opened in Doñana, and protocols were developed for a difficult species whose siblings often fight, sometimes to the death, in their first weeks. By 2009, the program was producing captive-born kits in reliable numbers. The hardest technical problem had been solved in theory. Then the animals were released, and many died.

Captive-born lynx died at a rate of about 52 percent after release. Wild-born lynx translocated between sites died at about 29 percent. The gap was not about prey availability. Release sites were chosen partly for rabbit density. The difference was physiological and behavioral. Captive-raised animals were metabolically adapted to controlled feeding schedules and predictable stress. They knew how to hunt rabbits but not how to live in the wild.

In 2007, feline leukemia virus swept through the Doñana population, then the largest wild lynx group in the world. The virus killed roughly two-thirds of infected animals. Between 2006 and 2011, the disease accounted for about 38 percent of recorded lynx deaths. Just as captive breeding began producing a surplus, disease was dismantling the very population those animals were meant to reinforce.

Rabbit management ran in parallel: artificial burrows, shrub replanting, vaccinations, and translocations. Results were local, inconsistent, and fragile. One burrow complex thrived for two seasons before emptying after an outbreak. One translocated population failed due to unsuitable vegetation.

By the time the LIFE Iberlince and Lynx Connect projects reached full scope, Spain and the EU had committed over 70 million euros. The lynx remained critically endangered. The failure was not a lack of effort. Budgets were real, science was serious, and staff were competent. The failure was structural, and it only became visible on cheap infrared cameras in a valley in front of a man who had gone there to film cats.

By the mid-2000s, wild Iberian lynx existed in only two places: Doñana in the southwest and Sierra Morena in the south-central mountains. Both were monitored, managed, and slowly stabilizing. They were also isolated from each other and from any other suitable habitat by highways, farmland, reservoirs, and above all, the absence of sufficient rabbit density in the areas between them. Two populations without genetic exchange do not constitute a viable species. They are two clocks about to stop at slightly different rates.

The decision was made to expand rather than reinforce. Instead of adding animals to the two remaining populations, the program established new ones in areas that had not seen lynx for decades. Sites were evaluated on four criteria: rabbit density, habitat suitability, cooperation of landowners, and distance from busy roads. Two sites passed first: Guadalmellato in Córdoba and Guarrizas in Jaén, both within the broader Sierra Morena range, both free of lynx for a generation.

In 2009, captive-bred animals were placed in transport boxes in Andalusia and driven to release points. Each Iberian lynx has a unique coat pattern that functions like a fingerprint, allowing researchers to identify individuals by camera without recapture. GPS collars tracked movement, and trail cameras were placed about every 500 meters across the release area.

At dawn, the boxes were opened one by one. For a second or two, nothing happened, because a lynx that had spent its entire life in an enclosure does not necessarily register that the wall is gone. Then it moved, low and fast, disappearing into the scrub in less than three seconds. The people standing there did not follow. From that moment, the experiment ran itself in terrain they did not control, under conditions no one could dictate.

The immediate behaviors matched expectations. Males ranged widely and irregularly, testing boundaries. Females settled faster, establishing home ranges within two to three weeks and maintaining them. Then the trail camera data began showing something the teams were not looking for.

Palomares had spent years working on Iberian carnivores. The work involves walking a grid through scrubland, pulling memory cards from weatherproof boxes attached to trees every 500 meters, and returning to a screen to review thousands of frames. Most frames are empty: moving grass, a branch, a heat shimmer. The job is mostly disciplined attention to nothing for hours at a time, which is why absence was recorded before presence.

In early 2010, while reviewing four months of images from squares where collared lynx had established territories, he noticed that red foxes were not appearing. Not fewer, not later at night, but not at all. Fox observation rates dropped sharply. Egyptian mongoose observations in the same squares also declined. Those mesopredators had not been hunted or eliminated. Nine lynx could not kill enough foxes in four months to produce that curve.

The foxes had detected the presence of an apex predator through scent, kill sites, and disturbance patterns left by a territorial cat. They had left before the lynx could reach them. In the weeks following the fox withdrawal, active rabbit burrows began appearing in the same grid squares that had previously had none.

No one had planned this sequence. No budget line existed for it because no one had proposed it. The nine animals had been released to live and, with luck, to breed. Whether they bred that first year was almost secondary. The valley had begun reorganizing itself around them before a single kitten was born there.

The mechanism behind this change appears to be a process ecologists call mesopredator release. When an apex predator is removed, mid-sized predators expand to fill its ecological space. In Mediterranean scrubland, the dominant mesopredators are the red fox and the Egyptian mongoose. Both eat rabbits, but differently from the lynx. The lynx is a stalk-and-ambush hunter that takes adults at low density. The fox digs. The mongoose enters burrows, targets kits, and applies continuous pressure that pushes colonies to disperse rather than breed.

In the absence of an apex predator, that mid-tier layer removes far more rabbit biomass than a handful of lynx would, and it removes it from the reproductive side. Where lynx had been absent, fox and mongoose activity consistently exceeded natural densities. The mid-tier had expanded to fill the gap at the top. Then the lynx returned.

Research published by Palomares and colleagues in Biological Conservation documented what the cameras were showing across multiple release sites. In areas where lynx populations became established, red fox and Egyptian mongoose abundance dropped by up to tenfold compared to lynx-free areas. The mechanism was largely behavioral rather than lethal. Lynx kill foxes and mongooses opportunistically, but the dominant effect was avoidance. Smaller predators recognized that an animal capable of killing them controlled the area, so they left or became cautious enough that their hunting efficiency dropped.

The outcome seems paradoxical but is not. The Iberian lynx does not protect rabbits by leaving them alone. It protects rabbits by making the environment more dangerous for the animals that were preying on them at a faster rate.

Researchers note their own caveats. Camera observation rates measure activity, not absolute abundance, and a fox that becomes more cautious is harder to photograph even if it does not leave. Habitat improvement work was also underway at some sites during the same years, making it difficult to separate effects. That is why Palomares spent a decade studying the pattern rather than publishing in 2010. But the trend held consistently across sites and years.

Spain spent decades treating rabbits and lynx as two engineering problems. The correct intervention turned out to be simpler and far cheaper: return the apex predator and let the hierarchy reorganize itself.

Population counts reflect the pattern. From 94 lynx in 2002, numbers rose to roughly 319 by 2013. A rabbit specialist cannot triple its numbers without the rabbit population growing beneath it. Fox decline, mongoose decline, rabbit rise, lynx rise, in that order, with no management plan directing it.

Two factors in this environment do not respond to predators. One is roads. In 2024, 214 Iberian lynx deaths were recorded across Spain and Portugal, with 162, about 75 percent, caused by vehicle collisions. In 2025, road deaths reached 212, nearly eight out of ten recorded deaths. These are casualties of success. The population is expanding into new territory, and new territory means roads. A lynx home range can extend to 30 square kilometers, and dispersing young travel much farther, crossing infrastructure built without consideration for a cat nobody expected to survive.

Underground passages, fencing, and crossing structures are being installed and work where they exist. The problem is speed. The population is growing faster than engineering can keep pace, and each new subpopulation opens a new set of roads.

The second factor is viral. RHDV2 has not disappeared. In 2013, it caused an 80 percent decline in rabbits in Doñana and a 75 percent decline in Andújar. Castilla-La Mancha now holds roughly 46 percent of Spain’s lynx population, and rabbit numbers there remain vulnerable to the same virus. One outbreak in the wrong place at the wrong time in the breeding cycle could interrupt a recovery that took 30 years to build. The apex predator suppresses the fox. It has no solution for a virus in a burrow.

Hunting estate operators in parts of Extremadura and Andalusia were initially hostile to reintroductions, claiming lynx would compete with paying hunters for partridges and rabbits. Field data has largely contradicted this. Lynx hunt dense-scrub rabbits rather than the open-ground species that hunting estates are built around, and some estates with lynx have reported better rabbit numbers for the reasons the cameras showed. Enforcement still matters. Where enforcement is weak, illegal killing and poisoning continue. Where enforcement is strong, Andalusia cut recorded wildlife poisoning incidents by more than 90 percent over two decades.

Funding is not permanent. The Life Lynx Connect program runs on EU grant cycles, and favorable conservation status requires about 750 breeding females. In 2024, the count was 470. By 2025, it reached 542, still short of the target but the best number ever recorded.

In June 2024, the counts had changed enough for an official ruling. The International Union for Conservation of Nature removed the Iberian lynx from the endangered species category and reclassified it as vulnerable. It is the first cat species in recorded history to be downgraded from endangered as a direct result of conservation intervention. In IUCN terms, the species moved from a very high risk of extinction in the wild to a high risk. That is the entire distance traveled. It took about 20 years and more than 70 million euros.

The numbers behind the new classification: 94 lynx in 2002, about 319 by 2013, and 2,663 across Spain and Portugal by 2025, distributed across a growing number of subpopulations rather than concentrated in two fragile pockets. Those two figures together, 94 and 2,663, represent nearly a 28-fold increase for a species the state once priced at three pesetas and 75 céntimos.

The lesson from the nine lynx in the Guadalmellato valley is the part that can be generalized. Spain spent 50 years managing two problems, the lynx and the rabbit, as if they were separate. Both were necessary, and neither was sufficient, because both treated symptoms in a food web whose structure had been broken at the top. The missing piece was not an external addition. It was a relationship.

The lynx was not just an animal in need of rescue. It was the organism that had kept the mid-tier of that ecosystem in balance for 10,000 years. Its absence had allowed foxes and mongooses to expand into the rabbit community until it could no longer support anything above it. Returning the cat restored a pressure that no burrow-building budget could replicate.

The most important number in this story is not 2,663. It is nine: nine captive-born animals released at dawn into a valley that had forgotten they existed, and that they rearranged in a single season.

The Iberian lynx is not yet saved. Eight out of ten recorded deaths last year were road collisions. The rabbit remains one virus away from another collapse, and breeding females number about 200 short of favorable status. But for the first time since a ministry printed a price list in 1953, the species is no longer declining. It is growing.

After 73 years since the state decided a lynx was worth less than a day’s wages, the animal it priced at nothing turned out to be the one thing the landscape could not function without.