Phocoena sinus · vaquita · “little cow”

The world’s most endangered marine mammal always looks like it’s smiling.

Dark rings circle its eyes and its lips curve up at the corners, but the vaquita has little to smile about. The world’s smallest porpoise lives only in one corner of Mexico’s Gulf of California, where fishing nets have drowned so many that the 2025 survey saw only seven to ten.

The species

The vaquita is the smallest porpoise alive and one of the smallest cetaceans (whales, porpoises, dolphins) of any kind. A full-grown female is up to about 1.5 metres long, roughly the height of a 12-year-old, and weighs about as much as a large dog (somewhere between a Labrador and a Great Dane, or 30 to 55 kilograms).

Plenty of porpoises have striking markings (the spectacled porpoise is a good example), but the vaquita’s face is its own. Dark patches ring its eyes like an eyepatch, and a dark line runs along its lips, which gives it a face you won’t mix up with any other.

They are shy, they surface briefly and quietly, and they have never been kept alive in captivity. There are photos of living vaquitas, but they’re taken from a distance as the animals briefly surface, and most close-ups are of stranded, dead animals. Nobody has the sharp photos or video that exist for, say, killer whales. The illustrations on this page are drawn from the few good pictures there are.

Scientific side-view illustration of a vaquita
  • 1.5 m the longest adult females, nose to tail
  • ~ 20 years, the estimated maximum lifespan

7–10

different vaquitas seen during the 2025 survey, including at least one calf. This is a count of the animals observers saw in a small search area, so it is not a population estimate. The last survey to estimate the whole population was in 2015, and the most recent estimate, for 2018, came from acoustic monitoring.

The vaquita’s home is tiny

The vaquita has never been recorded anywhere but the northern end of the Gulf of California. Since 2017, most of the ones anyone has seen or heard were in one small part of it, and recent surveys keep finding a few just outside it.

  1. 1 of 6 · The eastern Pacific

    The vaquita lives at the top of the Gulf of California.

    This view takes in about a quarter of the planet. No vaquita has ever been recorded outside that one sea.

  2. 2 of 6 · The Gulf of California

    The Gulf runs about 1,200 km between Baja California and mainland Mexico.

    Many whales and dolphins live along its length. The vaquita stays in the shallow northern end.

  3. 3 of 6 · The historical range

    The shaded area is everywhere vaquitas have been recorded.

    NOAA puts it at about 8,408 km², roughly the size of Puerto Rico. The dots are records from Mexico’s national biodiversity database, dated 1960 to 2008. They show where vaquitas were found then, not where they live today.

  4. 4 of 6 · Until 14 October 2026

    Gillnets were banned inside the solid outline.

    The ban became permanent in 2017. It covered about 11,100 km² of the northern Gulf. The smaller outline inside it is the Vaquita Refuge, 1,841 km². The dotted line is the biosphere reserve that protects the whole region.

  5. 5 of 6 · 15 October 2026

    Mexico cut the gillnet ban area by 86%.

    New rules took effect on 15 October 2026. Gillnets are now banned only inside the solid outline, 1,562 km², which closely follows the Vaquita Refuge. The dashed line shows where the ban used to end. Outside the new line, fishers can use gillnets the government has authorized.

  6. 6 of 6 · The Zero Tolerance Area

    Nobody may fish in the turquoise rectangle, the 225 km² Zero Tolerance Area.

    Mexico closed this area to fishing in 2020. Most vaquitas seen or heard since 2017 were inside it. Small boats may now cross it along the narrow channel. In the hatched area beside it, diving is the only fishing allowed. Surveys still find some vaquitas outside it. In September 2025, underwater recorders picked them up as far out as the edge of the refuge.

Making sense of the numbers

  • Most detections come from a tiny sliver of the Gulf

    The Gulf of California covers somewhere between 160,000 km² and 177,000 km² of water, depending on the source. Most detections since 2017 fall inside 225 km² of it, which is less than a sixth of one percent.

  • The area the vaquita relies on most is a short stretch of sea

    The Zero Tolerance Area is a tilted box about 22 km along its top edge and 10 km down its side. Were it road instead of sea, you could drive the length of it before your coffee went cold.

  • The vaquita’s world gets smaller at every step

    The vaquita’s whole known range, 8,408 km², is about the size of Puerto Rico. The refuge inside it, 1,841 km², is a little larger than Greater London. The area where most of them are still heard, 225 km², is less than half the size of Lake Tahoe.

The same facts, without the map
  • The vaquita is found only in the northern end of the Gulf of California, Mexico. It has never been recorded anywhere else.
  • NOAA’s historical-range polygon, referenced to the 2022 IUCN assessment, covers about 8,408 km². NOAA’s prose estimate is smaller again, at roughly 3,934 km² east of San Felipe. They are different products and should not be added together.
  • Three Mexican boundaries in force overlap that range. They are the Vaquita Refuge (1,841 km², DOF 2018), the Zero Tolerance Area (225 km², ) and the area where gillnets are banned (1,562 km², ). The term “exclusion area” doesn’t identify any one of them.
  • The IUCN’s 2022 Red List assessment found that nearly all detections since 2018 had been inside the Zero Tolerance Area (). Surveys since then have found vaquitas outside it. Half the 2024 sightings were in the area just beyond its edge (), and in 2025 a fifth of the sightings were outside the area the concrete blocks protect, with acoustic detections (93 encounters at 24 of 66 sites in September) reaching the edge of the Vaquita Refuge ().
  • Mexico published new fishing rules on 14 October 2026, and they took effect the next day. The area where gillnets are banned shrank from about 11,100 km² to 1,562 km². The Zero Tolerance Area stayed the same, but small boats may now cross it along a narrow channel. Beside it, a new Special Use Area allows diving only ().
  • The nearest ports are San Felipe, Golfo de Santa Clara and Puerto Peñasco.

The threat

One thing kills vaquitas: gillnets. A gillnet is a wall of fine mesh hung in the water column, invisible to a porpoise using echolocation the way a pane of glass is invisible to a bird. They swim into it, they cannot surface, and they drown.

Most of the nets in the vaquita’s range are set illegally for totoaba, a large fish whose dried swim bladder sells for extraordinary sums in an overseas market. Nobody has ever fished for the vaquita itself, and every one caught in a net is bycatch.

The cause is a trafficking market and enforcement that has not held. The fishing communities of the Upper Gulf have lived with a shrinking legal livelihood and a violent illegal trade operating alongside it.

Illustration of a vaquita rising towards the surface with a trail of bubbles
  • 1 cause of death that matters — entanglement in gillnets
  • 2015 the last year a survey estimated the whole population
  • 7–10 vaquitas seen on the 2025 survey, including at least one calf. A minimum count from a small search area.

The timeline

It took thirty-four million years of toothed whales branching apart to arrive at this animal. People have known it exists for seventy-six of those years, and in that time it went from a skull on a beach to the seven to ten animals the 2025 survey managed to see.

The story has two halves that move at very different speeds. Every card carries the study or the law it came from.

Act one · deep time

The vaquita has been a small population for about 300,000 years.

Each mark is a branching point in the family tree, dated from DNA and fossils.

What’s the shaded part?

The shaded stretch is the last 150 years. To fit 34 million years on one line, this scale is squeezed, so the further left you go, the more time each step covers. On an even scale, those 150 years would be thinner than a hair. Act two, below, zooms into them.

  1. 34 million years ago

    Toothed whales begin to diversify

    The lineage that leads to every dolphin, porpoise and sperm whale alive starts splitting apart. Nothing that looks like a porpoise exists yet — this is simply how far back the family goes.

    A model estimate from DNA, calibrated against fossils. It is not an observed date. ()

  2. ~19.8 million years ago

    Dolphins and porpoises part company

    At the root of the group, oceanic dolphins split from the branch that will later divide into porpoises and the beluga–narwhal line. Everything after this is porpoise prehistory.

    95% interval 18.8–20.8 million years. Porpoises as a family did not exist yet at this point. ()

  3. ~11.4 million years ago

    The porpoise family begins

    The porpoise branch separates from the one carrying beluga and narwhal. The study anchored this date partly on Salumiphocaena, a fossil porpoise from California about 10 to 11 million years old, which it treated as the oldest known.

    95% interval 10.0–14.0 million years; estimates differ between studies. ()

  4. ~5.4 million years ago

    The seven living porpoises trace back to this point

    The first split inside the modern family. Everything from the finless porpoises of Asian coasts and the Yangtze to the vaquita in the Gulf of California descends from this one ancestor.

    95% interval 4.2–6.9 million years. Dated from mitochondrial DNA; a nuclear family tree can differ. ()

  5. ~2.4 million years ago

    The vaquita branch splits off

    Its line separates from the ancestor of two southern porpoises, the spectacled and Burmeister’s. The authors’ most likely explanation is that colder water during an ice age let an ancestor from the south cross the equator, and its descendants stayed isolated in the north.

    95% interval 1.7–3.2 million years. A lineage split is a proxy for becoming a species, not an exact moment. ()

  6. ~300,000 years ago

    Cut off in the Gulf of California

    The genome shows a population that has stayed small since about this time. The study’s authors think a warming climate at the end of an ice age left it cut off in the newly expanded, shallow upper Gulf, where it has lived ever since.

    The genome shows a small population for about the last 300,000 years. The isolation is the authors’ hypothesis, which they tie to the end of an ice age about 130,000 years ago. ()

  7. ~315,000 years ago

    Homo sapiens appears, at almost the same moment

    The oldest fossils of our own species come from Jebel Irhoud in Morocco, dated to roughly 315,000 years ago. Within the resolution of these dates, the vaquita’s long run as a small population was beginning at about the time we were becoming ourselves. The two lines have run in parallel ever since.

    Fossil date, ±34,000 years. Species origins are gradual; this is the oldest evidence, not a birthday. ()

  8. ~25,700 years ago

    Around 2,800 breeding animals

    Long before people arrived, the vaquita was already rare. About 25,700 years ago its breeding population fell from about 4,500 to about 2,800, and it has stayed small since. That turns out to be good news — it carries far less damaging inbreeding than a recently-crashed species would.

    Estimate 2,807, 95% CI 1,832–3,982. This is effective population size, a genetic measure of breeders that is always smaller than a headcount. ()

  9. 1950 and 1958

    A skull on a beach near San Felipe

    Kenneth Norris picked up a skull on a beach near San Felipe in March 1950. Eight years later he and William McFarland described it as a species no one had recorded before: Phocoena sinus, the vaquita.

    This is when science met the animal. ()

Act two · since we met it

It took us less than one lifetime to bring it to the edge of extinction.

Most actions were only taken after the collapse of the population.

What the marks mean
  • Whole-population estimate
  • Minimum individuals seen — a different measurement
  • Law or protected area
  • Status change or alert
  • Science and surveys
  1. 1920

    About 5,000 vaquitas

    5,015 Whole-population estimate

    No one was counting in 1920. This figure comes from a population model worked backwards from later data — the best available picture of the species before industrial gillnets reached the Upper Gulf.

    The median of a model reconstruction (Jaramillo-Legorreta 2008), not a survey. No range is published with it. ()

  2. Science

    18 March 1950

    A skull on the shore of Punta San Felipe

    Kenneth Norris found a vaquita skull on the north-east shore of Punta San Felipe. It became the specimen the species was described from. ()

  3. Science

    February 1958

    Phocoena sinus enters the record

    Norris and McFarland described the animal from three skulls. Nobody had yet seen a living one. ()

  4. Science

    1961

    The first mention of bycatch

    Three years after being named, the species is already being reported as accidental catch in local fisheries. ()

  5. Law

    1 August 1975

    Mexico closes the totoaba fishery

    The fishery had already collapsed. Totoaba nets were the biggest killer of vaquitas — and the illegal version of that fishery is still the biggest killer today. ()

  6. Status

    28 June 1979

    Highest international trade protection

    Twenty-one years after being named, the vaquita is in the strictest category the trade convention has. ()

  7. Status

    9 January 1985

    Listed under the US Endangered Species Act

    Listed under the name then used for it, the cochito. Effective 8 February 1985. ()

  8. Status

    1986

    First IUCN assessment

    The species enters the Red List as Vulnerable, the lowest of the three threatened categories. By 1990 it is listed as Endangered. ()

  9. Law

    10 June 1993

    Upper Gulf biosphere reserve

    Mexico protects the region and bans capture of vaquita and totoaba inside it, indefinitely. ()

  10. Status

    1996

    One step from extinct in the wild

    Ten years after the first assessment, the vaquita is moved to the highest risk category, where it has stayed ever since. ()

  11. 1997

    567 vaquitas — the first real count

    567 Whole-population estimate

    Three vessels, purpose-built survey design, the entire known range. This is the number every later estimate is measured against, and the last time the species could be described in hundreds.

    95% CI 177–1,073. The bar shows that range. ()

  12. Science

    1997

    Mexico and the US count the whole range

    Three vessels, the entire known range, 567 animals. Every number since is compared with this one. ()

  13. Law

    8 September 2005

    A refuge in the western Upper Gulf

    The first protected area drawn around the vaquita itself rather than the region. It was enlarged in 2018. ()

  14. 2008

    245 left

    245 Whole-population estimate

    Eleven years later, a survey that watched and listened together found well under half as many.

    95% CI 68–884. Wide, but the direction is not in doubt. ()

  15. 2015

    About 100, from the last survey of the whole population

    100 Whole-population estimate

    Published first as roughly 60, later revised upward to about 100 once animals seen alive in 2017 and 2018 were taken into account. No survey since has estimated the whole population.

    Posterior mean 100, median 93, 95% CrI 47–191. Supersedes the original estimate of about 60 (median 59, 95% CrI 22–145). ()

  16. Law

    10 April 2015

    Two years, no gillnets

    An emergency suspension across a large polygon of the northern Gulf, paired with compensation for fishers. The population kept falling while it was in force. ()

  17. 2016

    About 30

    33 Whole-population estimate

    Projected forward from the 2015 survey using acoustic detection rates, which fell by about half (a mean of 49%) between 2015 and 2016.

    Mean 33, median 27, 95% CrI 8–96. A projection from acoustics, not a new survey, made from the original 2015 estimate of about 60 before it was revised to about 100. ()

  18. Law

    30 June 2017

    The ban is made permanent

    Mexico permanently prohibits gillnets for small vessels in the northern Gulf. Since then, the problem has been enforcing it. ()

  19. Science

    October–November 2017

    VaquitaCPR stops after a death

    An international team tried to hold vaquitas safely in human care. The first animal was released within hours; the second, an adult female, died. Captures were suspended and will not be tried again. ()

  20. 2018

    Nine

    9 Whole-population estimate

    The last formal whole-population estimate anyone has published, made eight years ago.

    Mean 9, median 8, 95% CrI 6–19. ()

  21. 2019

    7 to 15 individuals seen

    7–15 Minimum individuals seen — a different measurement

    From here the record changes shape. Instead of estimating a population, surveys count how many different animals they can positively identify in a small search area. The experts judged three calves the most likely number among them.

    Plausible range 7–15, most likely about 11; 84% belief of 9–13. Minimum unique individuals seen in the survey area, not a population estimate; the two series are different measurements. ()

  22. Law

    24 September 2020

    A no-fishing core, in law

    A 225 km² box where no fishing of any kind is allowed, inside the wider prohibition area. Most detections since have been inside it or just beyond its edges. ()

  23. 2021

    5 to 13 individuals seen

    5–13 Minimum individuals seen — a different measurement

    Healthy animals, calves again, in the same small core area — under continuing gillnet pressure.

    Plausible range 5–13, with 7 and 8 judged equally most likely; 78% belief of 6–10. Minimum seen, not a population estimate. ()

  24. Law

    July–October 2022

    193 hooked blocks on the seabed

    The Navy laid a grid of concrete blocks designed to snag illegal nets. 216 more followed in 2024, and vessel presence in the area dropped sharply. ()

  25. 2023

    8 to 13 individuals seen

    8–13 Minimum individuals seen — a different measurement

    Sixteen sightings, 61 acoustic detections, one or two calves — and a sharp drop in illegal fishing inside the Zero Tolerance Area.

    Mean 10.6; 76% probability of 8–13. The 2025 report’s harmonized comparison restates it as 67% probability of 8–12. Minimum seen in the study area. ()

  26. Status

    7 August 2023

    The IWC issues its first Extinction Alert

    The International Whaling Commission had never issued one before. It named the vaquita and said replacing gillnets remains the only thing that will work. ()

  27. 2024

    6 to 8 individuals seen

    6–8 Minimum individuals seen — a different measurement

    Healthy animals including a juvenile, but no newborn confidently identified. Only about an eighth of the 2015 search area was covered.

    Mean 7.62; 75% probability of 6–8. Explicitly not a whole-population count. ()

  28. September 2025

    7 to 10 individuals seen

    7–10 Minimum individuals seen — a different measurement

    One or two calves. Many detections came from north-west of the protected core — outside the box where enforcement is absolute.

    Mean 8.52; 67% probability of 7–10. Minimum seen in the study area, not the total population. ()

  29. Status

    March 2026

    CIRVA warns about the proposed boundaries

    The recovery committee said the proposed boundaries leave a shallow area to the north-west without concrete blocks. Vaquitas were detected there in 2025. It asked for blocks to go in straight away. ()

  30. Law

    14 October 2026

    Mexico shrinks the gillnet ban area

    New rules replace the 2020 ones. Gillnets are now banned across 1,562 km², down from about 11,100 km². The Zero Tolerance Area stays, with a channel for boats to cross it. ()

 Population estimate ·  Animals identified in surveys (not comparable)

Can the vaquita recover, and how long would it take?

Suppose every gillnet came out of the Gulf of California tomorrow. The vaquita could recover, but vaquitas have very few young, so getting their numbers back would take somewhere between 70 and 100 years, and that clock only starts once the nets are gone.

How long it could take

Vaquitas have very few young. A female can first breed somewhere between three and six years old and has one calf at a time, and scientists aren’t sure whether she rests a year between calves or has one every year. That one unknown changes the answer by decades, so here are both.

How long until there are as many vaquitas as in 1997 (567), starting from about ten today
  • If mothers rest a year between calves (what scientists have long thought) about 100 years · around 2130

  • If mothers have a calf every year (the most hopeful case) about 70 years · around 2095

Either way it’s somewhere between seven and ten vaquita generations. A child born this year would be about 70, or over 100, before the vaquita was back to where it was in 1997.

What these estimates assume

  • Every gillnet is gone this year, and no more vaquitas die in them.
  • We start from about ten vaquitas, the top of the likely range from the 2025 survey.
  • The population grows at the best-case rate scientists have published for vaquitas, every single year.
  • Vaquitas still find each other to breed, even when there are very few of them.
  • Nothing else goes wrong, like a heatwave, a disease or a year when food runs short.

Why the real answer is probably slower

Real life is messier than any estimate, and some of those things will go wrong. So treat these as the fastest the vaquita could come back. The Gulf will need protecting for decades after the first signs of recovery.

Can ten be enough?

Their genes are in better shape than you’d expect

When a species is down to a handful of animals, the usual worry is inbreeding. Closely related parents pass on hidden genetic faults, and calves don’t survive. Vaquitas have always been rare, though, and over tens of thousands of years most of those faults were weeded out. When scientists read the genomes of twenty vaquitas, they found very little hidden damage.

They then used those genomes to simulate the next fifty years many times over. If no more vaquitas die in nets, the vaquita survives in 94 of every 100 runs. If nets keep killing them, even at a fifth of the rate of recent years, the vaquita is more likely to disappear than survive.

Good genes don’t protect against everything. Heatwaves, food running short and a changing sea aren’t part of these models, and nobody has measured how they would affect the vaquita.

94 in 100 simulated futures where the vaquita is still here after fifty years, if no more die in nets

Frozen cells can’t bring them back

In 2017, scientists grew living cells from two female vaquitas and froze them at the Frozen Zoo in San Diego. Those cells are valuable for research, and they’re a big part of why we know so much about vaquita genetics.

They can’t replace the animals, though. Cloning a vaquita would need egg cells and a surrogate mother from a closely related species, no whale, dolphin or porpoise has ever been cloned, and no vaquita has ever survived in human care. Cells from two animals couldn’t rebuild a population anyway.

None of this recovery can begin while gillnets are still in the water, so every year they stay adds another year to the wait.

Where these numbers come from

The growth rates are the published best cases: about 4% a year if mothers rest a year between calves, and roughly 6% if they have a calf every year (). A vaquita generation is about 10 years (), and the breeding ages and calving intervals come from and .

The estimates assume that best-case growth every single year, starting from ten animals: about 103 years to reach 567 at 4%, and about 69 at 6%. They are rounded to about 100 and about 70 because nobody can predict a population this precisely.

Sources

What’s being done to save the vaquita?

Mexico and scientists from around the world have been working to protect the vaquita for more than thirty years. Some of that work is now clearly helping the animals that are left, and some of it never stopped the nets.

Latest · 15 October 2026 Mexico’s new fishing rules cut the gillnet ban area by 86%.

Since 15 October 2026, gillnets are banned across 1,562 km², down from about 11,100 km² (). CIRVA, the scientific committee that advises Mexico on the vaquita, had warned that the new boundaries leave a shallow area to the north-west without concrete blocks. Vaquitas were seen and heard there in 2025 ().

What’s working

  • 409 concrete blocks snag nets where the vaquitas live.

    Mexico’s Navy has laid 409 hooked concrete blocks on the seabed where the last vaquitas live, 193 in 2022 and 216 more in 2024, and they catch any gillnet set there. CIRVA says the blocks are the only measure so far proven to protect vaquitas.

  • More than 90% fewer boats fish in the core area.

    Fishers now mostly keep out of the Zero Tolerance Area, the 225 km² no-fishing zone set up in 2020. Sea Shepherd’s radar counted over 90% fewer boats there in each fishing season from 2023 to 2025, compared with before the blocks went in, with joint Navy patrols helping.

  • The vaquitas there look healthy, and they’re still having calves.

    Surveys keep finding young vaquitas. In 2025, drone footage showed adults and young in good condition and feeding undisturbed, and a female the team calls Frida, known by her bent dorsal fin, had her second calf since 2023.

  • Underwater microphones hear vaquitas nobody can see.

    Underwater detectors pick up vaquitas’ clicks when nobody can see them. They tracked the decline from 2011, and in 2024 and 2025 they showed vaquitas using habitat north-west of the protected core, which is why CIRVA wants that habitat protected too.

What hasn’t worked

  • Gillnet bans on paper didn’t keep the nets out of the water.

    Mexico banned gillnets across the northern Gulf for two years in 2015, then permanently in 2017. The vaquita kept declining. In 2026, officials told CIRVA that the 2020 rules had been impossible to enforce. On some days that January, radar still picked up more than 40 boats possibly net fishing in the refuge. On 15 October 2026, new rules cut the area where gillnets are banned by 86%.

  • The switch to safer fishing gear is still unfinished.

    Since 2008, Mexico has offered fishers compensation and gear that doesn’t trap porpoises. According to NOAA this may have slowed the decline without stopping it, and CIRVA still lists the move to vaquita-safe gear as unfinished work.

  • Catching vaquitas to protect them ended when one died.

    In 2017 an international team, VaquitaCPR, tried to move vaquitas into human care until the nets were gone. The first animal caught was so stressed she was released within hours, the second, an adult female, died, and captures will not be tried again.

  • Money from smuggled totoaba keeps the nets coming back.

    The nets are set for totoaba, a large fish whose dried swim bladder is smuggled to China and sold for large sums. Illegal fishing for totoaba has carried on since Mexico closed the fishery in 1975.

What you can do

Become a vaquita supporter

You’ll hear what’s happening in the Gulf, and we’ll tell you when there’s a petition to sign or a letter to send.

  • News from the Gulf, when there’s something worth telling you
  • A heads-up when there’s a petition to sign or a letter to send
  • The dates of the next Save The Vaquita Week