Paraguay Released 6 Pumas Into Dead Scrubland —What Happened to the Deer Population Shocked Everyone

Añadido: 2026-05-06

[00:00:00]In the winter of 2019, a field biologist named Rodrigo Villarreal was crouching in the dry scrub of northwestern Paraguay reading data from a camera trap he had set 3 weeks earlier near a dry creek bed in the buffer zone of Defensores del Chaco National Park.

[00:00:17]He expected to see footage of peccaries, maybe a fox.

[00:00:21]What he found instead made him [music] sit back on his heels and stare at the screen for a very long time.

[00:00:27]The footage showed not one, but three Pampas deer moving through a stand of [music] thorny quebracho scrub that had been grazed nearly to the soil just 12 months earlier.

[00:00:38]They were moving slowly, grazing as they went.

[00:00:42]But they were there.

[00:00:43]And they were moving differently than he had ever seen deer move in that landscape before.

[00:00:48]They were not lingering. They were scanning constantly.

[00:00:51]One would feed for a few seconds, then lift its head, turn, and move.

[00:00:56]The others followed.

[00:00:57]It was not the behavior of deer in a place without predators. Something had changed the texture of fear in that landscape. Something was watching.

[00:01:05]Something that had not been there in a very long time.

[00:01:09]This is the story of how six pumas were restored to a dead and overgrazed scrubland in the Paraguayan Chaco.

[00:01:15]How the deer population that everyone expected to collapse instead restructured itself from the inside out, and how a patch of eroded, compacted, seemingly exhausted land began [music] to pull itself back from the edge of permanent degradation in ways that no ecological model had predicted.

[00:01:35]To understand why this matters, >> [music] >> you need to know what was lost.

[00:01:40]The Gran Chaco is the second largest forest on Earth after the Amazon.

[00:01:45]It stretches across roughly 100 million hectares of South America, covering parts of Paraguay, Argentina, Bolivia, and Brazil.

[00:01:54]In Paraguay alone, the Chaco covers the entire western half of the country.

[00:01:59]It is a place of extremes.

[00:02:02]Temperatures can swing between -5°C and 50°C in the same year.

[00:02:09]Rain comes in violent floods, then vanishes for months.

[00:02:13]The thorny forest grows dense and impenetrable in wet years and sheds its leaves like a fever patient in dry ones.

[00:02:21]It is not a gentle place, but it is one of the most biologically dense dry forests on the planet.

[00:02:26]Before large-scale ranching arrived, [music] the Chaco supported a staggering web of life.

[00:02:32]The Pampas deer, known in Guarani as the guazú ti, moved across the open savannas in herds that indigenous communities had hunted and revered for centuries.

[00:02:43]The marsh deer, the largest deer species in all of South America, navigated the seasonally flooded lowlands, capable of strong swimming and following water levels across vast distances.

[00:02:57]The Chacoan peccary, a species so isolated from its closest relatives that biologists once believed it extinct, [music] rooted through the dry forest in family groups.

[00:03:06]Above all of them, maintaining the balance that made this abundance possible, was the puma.

[00:03:12]Puma concolor, the broadest-ranging land mammal in the entire western hemisphere, and in the Chaco, the top predator in the system.

[00:03:22]The puma did something that no rancher or farmer or government planner ever valued or measured.

[00:03:27]It created what ecologists now call a landscape of fear.

[00:03:31]Prey animals, particularly deer, do not simply eat wherever food is available when a large predator is present in the landscape.

[00:03:39]They choose where to eat based on where it is safe to eat. They avoid dense cover where a puma could conceal itself.

[00:03:47]They move constantly, never stripping any one patch of vegetation down to bare ground. They graze in short bursts and then move on, always listening, always scanning.

[00:03:57]The result is a distributed pattern of herbivory that the land can absorb.

[00:04:02]The plants recover. The soil holds. The system breathes.

[00:04:08]Pumas were not wiped out in the Chaco overnight. It happened over decades.

[00:04:12]Cattle ranching expanded steadily through the 1970s, the 1980s, and the 1990s, converting native dry forest into pasture at a rate that researchers using Landsat satellite imagery later calculated at nearly 1% of the Paraguayan Chaco every single year.

[00:04:32]By 2012, 27% of the entire Paraguayan Chaco had been converted from native forest >> [music] >> to grassland, the equivalent of 44,000 square kilometers, predominantly to create new pastures for cattle.

[00:04:49]And Paraguay's portion of the Gran Chaco suffered the highest deforestation rate of any country in the region.

[00:04:58]During the early 2000s, Paraguay was losing an estimated 220,000 hectares of Chaco forest every year.

[00:05:07]When the forest went, the pumas followed it. They required large home ranges, in some populations as much as 500 square kilometers for a single adult. They required prey.

[00:05:19]As habitat fragmented and the deer and peccary populations that sustained them shrank, pumas found themselves increasingly cornered. When a cornered puma kills a calf, the rancher shoots the puma. It was a straightforward of disappearance.

[00:05:34]The large cats retreated into the remaining protected areas. In the unprotected scrubland beyond the park boundaries, they were functionally gone.

[00:05:42]And without pumas, the deer behaved differently, very differently.

[00:05:47]The turning point came not from a government initiative or a major international funding body. It came from a small team working under the Wildlife Conservation Society program in Paraguay, led by biologist Laura Villalba, who had spent years studying jaguar and puma movement patterns across the northern Chaco landscape and building relationships with the ranching [music] families whose properties bordered the protected area.

[00:06:12]Villalba had spent enough time following GPS collar data and analyzing camera trap footage [music] to understand something that was not yet showing up in any official conservation report.

[00:06:23]The scrubland outside the park was not just losing predators, it was losing its structure.

[00:06:30]The vegetation was being hammered down by deer that had lost their fear.

[00:06:36]In the absence of pumas, Pampas deer populations in the degraded buffer zones had expanded beyond what the habitat could sustainably support.

[00:06:45]They were not numerous in the absolute sense. There were no massive herds, but they were concentrated.

[00:06:51]They returned to the same patches of remaining native vegetation day after day, grazing without urgency, without the behavioral pressure that would have kept them moving.

[00:07:02]The quebracho saplings were being browsed before they could establish.

[00:07:07]The ground cover was compacted.

[00:07:09]During the brief rains, water sheeted across the surface rather than soaking in.

[00:07:14]The dry creek beds that had once held standing water weeks after rain were drying out in days.

[00:07:21]The land was entering a slow spiral.

[00:07:24]Villalba proposed something that drew significant skepticism from her colleagues and outright hostility from some of the ranching community.

[00:07:34]She wanted to relocate a small number of pumas from a population that had established itself inside Defensores del Chaco [music] National Park, where the animals were monitored and had shown stable breeding, into two carefully selected zones in the degraded scrubland buffer.

[00:07:52]Six animals, four females and two males.

[00:07:57]The goal was not to restore the Chaco to some imagined pristine state. The goal was to restore fear.

[00:08:05]The models were not encouraging.

[00:08:07]Standard predator reintroduction models suggested that six pumas operating in a fragmented and partially degraded landscape would likely disperse, lose contact with each other, and fail to establish a functioning believed the animals would drift toward ranches and trigger exactly the kind of retaliatory killing [music] that had caused the problem in the first place.

[00:08:30]One livestock association formally objected to the proposal, arguing that the project would cost ranching families in the region economically with no measurable benefit to them whatsoever.

[00:08:44]Villalba and her team spent 18 months before the release doing something that most predator reintroduction projects skip entirely.

[00:08:52]They worked directly with 17 ranch families whose properties bordered the release zones, not to persuade them, to listen to them first.

[00:09:02]The ranchers explained which of their grazing areas were most vulnerable. They explained where water sources were located, where calving happened, and where previous puma encounters had occurred. That information shaped the final placement of the release zones.

[00:09:17]Two ranching families who had initially objected eventually agreed to allow camera traps on their properties in [music] exchange for being among the first to receive data from those cameras. They were not converted, but they were included.

[00:09:34]The actual translocation took place in two stages. The first group of four animals in July of 2017, [music] and the second pair in March of 2018.

[00:09:45]Each animal was fitted with a GPS collar programmed [music] to transmit location data four times per day.

[00:09:52]The biologists selected [music] individuals based on age, health status, and crucially, the absence of previous >> based on age, health status, and crucially the absence of previous exposure to livestock as a prey source.

[00:10:10]Younger animals unfamiliar with cattle were chosen over older ones that had already learned that calves were easier to take than peccaries.

[00:10:18]The release [music] zones were not random. Both areas had been mapped extensively for remaining native vegetation, water availability, and prey density.

[00:10:28]The team had used 2 years of camera trap data to establish baseline measurements for deer behavior before the release.

[00:10:36]They counted how long individual deer spent [music] at a feeding site before moving.

[00:10:40]They measured how frequently deer returned to the same vegetation patches.

[00:10:44]They tracked the percentage of quebracho seedlings [music] being browsed before they could reach a height of 1 m.

[00:10:50]These were their control measurements, the baseline against which they would measure everything that came next.

[00:10:55]For the first 8 months, almost nothing happened, or rather, nothing happened in the way that anyone who was not paying very close attention would have noticed.

[00:11:04]>> [music] >> The pumas established themselves. They moved, they hunted, they avoided the ranch infrastructure with a consistency that surprised even the biologists. The GPS data showed [music] clear learned avoidance of areas near buildings and cattle pens. One female appeared to be using a seasonal creek channel as a travel corridor, moving [music] along it at night and sheltering in dense thorn scrub during daylight hours.

[00:11:28]>> [music] >> One of the males established a home range that overlapped almost exactly with the area the biologists had predicted as the highest quality remaining habitat. [music] But the deer the deer were the story.

[00:11:44]>> [music] >> By the end of the first year, the camera trap footage began to show something that Rodrigo Villarreal, who was running the field monitoring, described to his team as behavioral restructuring.

[00:11:56]The deer that had spent the previous 2 years grazing for an average of 22 minutes >> [music] >> before moving on from any given patch were now spending an average of less than 8 minutes. That number is not a rounding error. It is not measurement noise. It represents a fundamental change in how those animals were experiencing the landscape.

[00:12:18]Fear had returned. Here is the key insight that changed everything.

[00:12:23]The deer were not being eaten into submission. In the first year after the puma release, the team documented only 11 confirmed puma kills in the monitored zones. A predation rate well within the range that healthy deer populations can sustain without decline. The pumas were not controlling the deer by killing them, they were controlling the deer by existing.

[00:12:44]By being present in a way that the deer could detect through scent, through tracks, and through occasional direct observation at a distance, the pumas had reimposed a spatial logic on how the deer used the landscape. Scientists had predicted a modest behavioral shift.

[00:13:01]What they found instead was a cascade.

[00:13:04]By the end of the second year, vegetation monitoring plots were recording something that stopped the team in its tracks.

[00:13:11]Quebracho seedling survival [music] in the release zones had increased by 41% compared to the pre-release baseline and compared to matched [music] control plots outside the release area.

[00:13:23]That is nearly a doubling of the rate at which woody vegetation was successfully [music] establishing itself.

[00:13:30]The deer were still there. They were eating, but they were moving.

[00:13:35]The plants they've been suppressing were no longer being hit repeatedly by the same animals returning to the same spots.

[00:13:42]Then came the soil data.

[00:13:44]Soil compaction measurements taken using standardized penetrometer readings at fixed points across the monitoring grid showed a measurable reduction in surface compaction at the most heavily grazed sites within the release zones.

[00:13:59]This finding was not expected. The team had not included soil compaction as a primary outcome variable.

[00:14:07]They had measured it as a secondary indicator largely out of habit. But when the numbers came back, they showed that the redistributed grazing pressure was allowing the soil surface to recover its structure in patches.

[00:14:21]Crust formation, which had been sealing the surface and preventing water infiltration, was breaking down in areas where deer had stopped concentrating.

[00:14:30]Water infiltration changed next. The dry creek beds that had been losing standing water within days of rain were holding water longer, not by weeks, but by enough days that a second round of vegetation growth was occurring in those riparian margins.

[00:14:45]Where deer had been concentrated, the creek banks had been trampled bare, and the channels were eroding.

[00:14:52]Where deer had dispersed, root systems were stabilizing the banks.

[00:14:56]Let me put that in context.

[00:14:58]Six pumas relocated at a total cost that the program estimated at roughly equivalent to 1/5 the annual operating cost of a water storage project the regional development agency had been proposing as a solution to the same water retention problem were doing the work of an engineering intervention, not as well, not as quickly, but in a self-sustaining way free of maintenance costs running entirely on peccaries and pampas deer 24 hours a day every day of the year.

[00:15:29]William Ripple, a [music] professor of ecology at Oregon State University who has spent decades documenting trophic cascades triggered by large carnivores, has written that top predators function as ecosystem architects, reshaping the structure of entire biological communities through a web of direct and indirect effects that extend far beyond the animals they kill.

[00:15:53]His research on pumas in Zion National Park showed that when cougars were effectively removed from a canyon by the presence of large numbers of human visitors, mule deer densities increased and their browsing behavior intensified leading to the collapse of cottonwood recruitment along stream banks, increased erosion, and declining populations of reptiles, amphibians, and insects.

[00:16:19]The cougar had not been killed, it had simply moved away, and the entire riparian system had shifted in response to that absence.

[00:16:28]The Paraguay data fit that same architecture precisely, but in reverse.

[00:16:33]The predator had returned, and the system was reading it.

[00:16:37]Villarreal reviewed the 2-year data set and found something he had not looked [music] for. Maned wolves, a species whose long legs and specialized diet made it a poor competitor with deer for most [music] resources, had increased their use of the release zones after the puma reintroduction.

[00:16:53]>> [music] >> The working hypothesis was that redistributed deer grazing was opening up the ground layer sufficiently [music] for the wolves to access the small mammals and insects they preferred.

[00:17:05]A single change at the top of the food chain was producing effects [music] three trophic levels down. The system was rebuilding itself, not the way engineers rebuild [music] things, the way nature rebuilds things. Slowly, then suddenly.

[00:17:20]This is not a simple story, and it would be dishonest to tell it as one.

[00:17:25]Ramon Galliano has run a cattle operation on a property that borders the western edge of the release zone since his father established it in the 1980s.

[00:17:35]He did not oppose the puma program publicly, but he has been consistently clear about what it has cost him.

[00:17:41]In the first year after the releases, he documented four confirmed calf losses that his veterinarian attributed to puma predation.

[00:17:50]Four calves is not a number that appears alarming in a conservation report.

[00:17:55]It is not alarming to anyone who does not understand what calves are worth to a small operation running on thin margins in the dry season of northwestern Paraguay.

[00:18:05]Galliano received compensation through a fund the conservation program had established, and he acknowledges that the process was straightforward. He still watches the fence line with a different kind of attention than he did [music] before.

[00:18:18]There is a harder version of this tension that no compensation fund fully resolves.

[00:18:24]The communities that bore the economic costs of the [music] Chaco's original destruction, the small ranching families who were squeezed out of the Eastern Atlantic Forest region after Paraguay enacted its zero deforestation law in 2004 and pushed into the Chaco [music] as the last affordable land frontier, are not the communities that designed the global demand for Paraguayan beef that drove that destruction in the first place.

[00:18:47]Asking them to absorb the costs of a restoration program that benefits a broader ecological commons they did not damage feels to many of them like an injustice with a conservation branding.

[00:18:59]Villalba acknowledges this directly. Her team has argued internally and in published reports that predator restoration programs without robust livelihood alternatives for affected ranching families are not just politically fragile, they are ethically incomplete. The science of the cascade is real. The benefits to the landscape is measurable, and the burden of that benefit should not fall disproportionately on the people who are already carrying the most weight.

[00:19:28]Return to that dry creek bed in the winter of 2019.

[00:19:33]Return to Rodrigo Villarreal watching three pampas deer move through a stand of thorny quebracho that had been scraped nearly bare 12 months before. The deer he is watching are alive.

[00:19:47]The pumas that changed their behavior are alive. The seedlings pushing up through cracked clay in the beds where the deer have stopped concentrating It is a system that has remembered something it was not designed to forget.

[00:20:00]Here is a detail that almost never makes it into the conversation about predator reintroduction in South America.

[00:20:08]The Chaco has one of the highest deforestation rates on the planet. It is losing native forest faster than almost any other dry forest ecosystem on Earth.

[00:20:18]And the loss is not only driven by [music] chainsaw and fire.

[00:20:21]It is driven by the slow, steady pressure of ungulate populations that have no biological incentive [music] to move, no reason to fear, no predator imposing the spatial discipline [music] that has governed their behavior for millions of years.

[00:20:37]Restore the puma and you do not just [music] restore one species.

[00:20:41]You restore the logic of the entire system. [music] What surprised you most about this story? Was it the speed of the behavioral change? The soil data that nobody was expecting?

[00:20:52]Or the realization that six animals, weighing perhaps 60 kg each, could shift the hydrology of a landscape?

[00:21:01]Leave that [music] in the comments. It matters that people who do not work in ecology start having this conversation.

[00:21:07]Because the next [music] decision about a buffer zone in the Chaco will be made by people who are reading the economics, not the camera trap data.

[00:21:16]If you are not subscribed yet, now is the time to subscribe.

[00:21:21]We cover stories about restoration and the people who are fixing what is broken.

[00:21:26]Hit the subscribe button. See you in the next one.