Thirty-three years of capuchins, and a forest that won't sit still

The 2014 to 2016 dry season at Lomas Barbudal was among the worst on record. Rainfall on the Pacific slope of Costa Rica fell well below normal in step with one of the strongest El Niño events of the modern era. Capuchin monkeys in the reserve died. Sofía Carrera and her colleagues, working from blood and faecal samples collected in the years before and during the drought, would publish in Science Advances in 2025 that the white-faced capuchins (Cebus imitator, the monkey most people in Costa Rica call the mono cara blanca) most likely to survive the drought were the ones whose stress response had been strongest: a steeper rise in glucocorticoids during the preceding milder droughts predicted survival of the severe one. The drought was a selection event.

One year later, the same long-term study released a second paper. Odd Jacobson, Margaret Crofoot, Genevieve Finerty, Susan Perry and Brendan Barrett, working with thirty-three years of demographic and movement data from twelve neighboring capuchin groups, asked a different question. The survival study had asked who lives and who dies; this one asked who pushes whom around. Their answer, published in Nature Ecology & Evolution on 6 May 2026, is that climate extremes do not just kill animals; they reshape the rules by which the survivors compete, and which kind of group gains ground. The capuchin population at Lomas Barbudal is restructuring itself in response to the climate it has, and the climate it is getting.

Between them, the two papers make a single point: at Lomas Barbudal, El Niño and La Niña sort capuchin life directly. The Pacific climate cycle decides who survives a drought and which troop pushes into its neighbors' range, reaching all the way down to the individual animal.

What the new paper found

The textbook expectation for a fruit-eating monkey in groups is straightforward. A bigger group strips a fruiting tree faster than a smaller one, so a bigger group should have to walk further every day to find the next ripe tree. The idea, written down by the primatologist Richard Wrangham in 1980 and refined by Carel van Schaik in 1983, is called the ecological constraints model, and it has been used for forty years to explain why most primate species settle around a characteristic group size that the local environment can carry. Lomas Barbudal has twelve groups, ranging from five to forty animals with an average of about nineteen, so the prediction can be tested directly.

The data are extraordinary. Researchers identified 335 individual capuchins by face and followed them for 4,952 hours of feeding observation between 2006 and 2021. Between 2009 and 2020, an observer walked with each group carrying a handheld GPS that logged the troop's path every few minutes, and every thirty seconds in the later years. Vegetation greenness, the basic indicator of how much leaf and fruit a forest patch is producing in any given month, came from satellite imagery (Landsat) processed in Google Earth Engine. Drought severity came from a global climate dataset that closely tracks the El Niño signal on the Pacific slope of Costa Rica. The behavioral and movement data were then run through statistical models that, for the first time, treated each pair of capuchin groups as a relationship in its own right, the same way social scientists analyze pairs of households or co-workers.

The first finding contradicts the textbook. Larger capuchin groups do eat less fruit per individual, the within-group cost is real and measurable, but they do not walk further on any given day. Instead, they hold larger total home ranges and visit any given location less often. They expand their range across longer timescales, rotating through less-depleted patches over months and years rather than chasing fruit over hours and days. The mechanical solution to within-group competition is slow geographic spreading, not faster daily marching.

The second finding is about who pushes whom. Using the pair-of-groups model, Jacobson and colleagues asked: when two neighboring groups share territory and the size of the overlap shifts over time, which of the two drove the shift? In the pairs where one group had gained at least five animals on the other and their overlap had widened sharply, that larger group was the one pushing in, eighty-four percent of the time. Larger groups encroach on smaller neighbors. Smaller groups appear to give up range. There is no equivalent class of cases in which small groups push back into larger neighbors' territories.

The third finding sets the first two against each other and lets the climate tip the balance. Being in a big group costs each member fruit; it also wins the group ground from its neighbors. In ordinary years those two roughly cancel, so groups of every size persist side by side. In the most extreme years, the dry seasons of strong El Niños and the wet seasons of strong La Niñas, the fruit cost grows much steeper: each member eats far less, even as neighbors cede ground. In the in-between, counter-seasonal years, when a dry season runs wetter than expected or a wet season runs drier, the fruit cost nearly disappears and the territorial advantage of being big takes over. Climate variability does not just stress capuchin groups uniformly. It changes which kind of competition dominates, and therefore which kinds of groups gain or lose ground.

One piece of the paper is a method other projects can borrow. The statistical tool the team built, a Social Relations Model, was designed to study relationships between individual animals; Jacobson and colleagues rebuilt it to compare whole groups instead. Any project that has followed several groups for many years can now run the same who-encroaches-on-whom analysis. Santa Rosa, La Pacífica, La Selva and Taboga already have, or could readily collect, the data it needs.

Why a thirty-three year record exists in Bagaces

The Lomas Barbudal Biological Reserve covers about 2,645 hectares in the cantón of Bagaces, on the eastern edge of the Tempisque basin. It was created in 1986 by Decreto Ejecutivo 16849-MAG, primarily to protect populations of native bees and the dry-forest community that supports them, and is administered today by the Sistema Nacional de Áreas de Conservación through the Área de Conservación Arenal Tempisque. The reserve sits inside a working agricultural landscape: Hacienda Pelón, Brin d'Amor, and the community of San Ramón de Bagaces border it on three sides. Its current management plan was published in 2014.

Susan Perry of the University of California, Los Angeles, with Joseph Manson and Julie Gros-Louis, founded the Lomas Barbudal Monkey Project at the reserve in 1990 to study social intelligence in white-faced capuchins. Habituation of the first study group, called Abby's Group, began around 1990. The project spent about five years based at the Max Planck Institute for Evolutionary Anthropology in Leipzig in the early 2000s, then returned to UCLA, and later added a parallel home at the Max Planck Institute of Animal Behavior in Konstanz under Margaret Crofoot. Funding has come piecemeal from the U.S. National Science Foundation across five major grants between 1996 and 2021, the National Geographic Society across five grants between 2006 and 2018, the L.S.B. Leakey Foundation, the Templeton World Charity Foundation, the Wenner-Gren Foundation, the University of Michigan, the University of Konstanz Excellence Cluster, the Wild Capuchin Foundation, and the Alexander von Humboldt-Stiftung. The dataset that powers the 2026 paper exists because that funding was successfully renewed enough times to outlast several presidential administrations and most of the careers of the original research assistants.

Linda Fedigan of the University of Calgary started the Santa Rosa Primate Project, on the same dry forest but inside the Área de Conservación Guanacaste, seven years earlier, in 1983. Santa Rosa tracks three monkey species: the same white-faced capuchin, the mantled howler (Alouatta palliata, the mono congo), and the black-handed spider monkey (Ateles geoffroyi, the mono colorado or mono araña). The project's first census paper, by Fedigan and her colleagues in 1985, ran in the Costa Rican journal Brenesia. The four decades of work since then have produced two synthesis volumes (Fedigan and Jack in 2012, Melin and colleagues in 2020), a body of papers on capuchin demography (Campos and colleagues in 2014 and 2015), a 2017 multi-site collaboration that linked seven monkey species' demography to climate variation, and a 2018 synthesis chapter that established the basic signature of El Niño cycles on Santa Rosa rainfall.

The third active long-term capuchin project is younger. Capuchinos de Taboga, based at Reserva Forestal Taboga south of Cañas, was established in 2017 by Elizabeth Tinsley Johnson, Marcela Benítez, Thore Bergman and Jacinta Beehner, and tracks the same species in a more fragmented, more agricultural landscape than either Lomas Barbudal or Santa Rosa. Taboga gives the network something it had been missing: a comparison site with similar climate and the same capuchin species but a more disturbed habitat matrix. Together, the three projects are the closest thing Costa Rica has to a national capuchin observatory.

The other primate species each have their own long-term anchor. Kenneth Glander began studying the mantled howlers of Hacienda La Pacífica, in Cañas, in the early 1970s; Margaret Clarke and her colleagues continued the demographic record into the 2000s, including documenting the population's response to the canal-construction deforestation between 1990 and 1994. The howlers of La Selva Biological Station, on the Caribbean slope, were censused first by Kathryn Stoner in 1992 and again by Christopher Johnson and his colleagues in 2022, a thirty-year repeat that found the population had grown rather than declined. Sue Boinski's work on the Central American squirrel monkey (Saimiri oerstedii) at Manuel Antonio in the late 1980s and early 1990s remains the empirical foundation for what is known about that species's social organization, and is followed by a body of Costa Rican-authored monitoring work from the Instituto Internacional en Conservación y Manejo de Vida Silvestre at the Universidad Nacional, including Daniela Solano and Grace Wong's 2009 habitat-and-population paper on the southern Osa subspecies. None of these other anchors quite matches the spatial-temporal density of the capuchin record, which is part of what makes the Jacobson and Carrera papers possible at all.

The forest the capuchins live in

Guanacaste's tropical dry forest is one of the most threatened forest types on Earth. Liz Miles and her colleagues mapped the world's remaining tropical dry forests in 2006 and found the Mesoamerican forest among the most depleted and least protected anywhere. The Guanacaste-Tempisque-Arenal complex of national parks and reserves that Lomas Barbudal sits inside is one of the largest unbroken stretches remaining on the Pacific coast of the Americas. The basic rhythm of this forest, mapped most thoroughly in Gordon Frankie, Alfonso Mata and S. Bradleigh Vinson's 2004 synthesis volume, alternates a dry season starting in December with a wet season starting in May; the dry months historically ran four to five and have stretched toward six in recent decades. Most of the trees in the canopy are deciduous: they drop their leaves when the rains stop, the same way an oak in temperate climates drops leaves in autumn, in order to lose less water. Food, water and shade for animals concentrate, in those months, in the evergreen strip of forest along the streams and rivers (the gallery forest).

That rhythm is bent by El Niño and La Niña, the two phases of the climate cycle that dominates Pacific Costa Rica year to year, and its grip on the dry forest is direct. Mauricio Vega-Araya at the Universidad Nacional, reading satellite measurements of how green the vegetation is from year to year, has tied the forest's productivity to the cycle's swings, and Silja Hund and colleagues found that in an extreme El Niño the streamflow and groundwater recharge of Guanacaste's Potrero-Caimital watersheds fall by roughly sixty percent. The extremes do the damage. The veranillo, the short dry pause that normally breaks the wet season for two or three weeks in July and August, is where the worst recent drought began: Talia Anderson and her colleagues found that the 2015 to 2019 drought, the driest stretch in Central America's instrumental record, traced to the failure of exactly those July and August rains, with warming deepening the soil-moisture deficit. And the trend runs one way: Christian Birkel of the Universidad de Costa Rica, projecting the country's climate under a moderate (RCP4.5) emissions pathway, put warming at roughly 2.6 °C by 2040, enough to push the boundaries of Costa Rica's tropical life zones.

The forest is not just losing water in the bad years. Daniel Janzen and Winnie Hallwachs, working with more than four decades of caterpillar and insect-trap observations inside the Área de Conservación Guanacaste, reported in Proceedings of the National Academy of Sciences in 2021 that the dry forest's caterpillar fauna had declined catastrophically across decades. They called themselves the "insectometers." For a primate population whose annual food budget includes a substantial fraction of invertebrates, the insectometer signal is part of that budget shrinking.

Fire compounds the picture, though not in lockstep with drought. Daniela Vargas-Sanabria and her colleagues at the Universidad Nacional and the Tecnológico de Costa Rica have mapped where wildfire vulnerability concentrates across the Área de Conservación Guanacaste, tracing it to geomorphology and land cover. Gabriela Jones Román and Benjamín Álvarez Garay reconstructed two decades of ACG fire history from 1997 to 2017 in the Universidad Nacional's Perspectivas Rurales, and found that the years of greatest fire impact did not coincide with El Niño events, a reminder that ignition and land use, not drought alone, govern how much of the forest burns.

Four primate species, four climate sensitivities

Costa Rica is home to four native monkey species. Each has a different way of making a living, and each will respond to climate swings differently from the white-faced capuchin that the Jacobson paper describes. The differences matter because protecting the dry forest as a whole is not the same task as protecting capuchins. Ruth Wiederholt and Eric Post, in a 2010 study of four ateline primate genera (howler, spider, woolly and muriqui monkeys) across the Neotropics, showed that El Niño cycles synchronise the population trends of the more frugivorous species; capuchins were not part of that analysis, and the fact that other populations move together does not mean each species is equally exposed.

The white-faced capuchin (Cebus imitator), the species the Jacobson paper studies, was recognized as a separate species from the South American white-faced capuchin (Cebus capucinus) only in 2012, when Jean Boubli, Anthony Rylands and colleagues showed that the two had diverged genetically about two million years ago. The IUCN Red List, the global standard for species risk assessment, lists Costa Rica's white-faced capuchin as Vulnerable, with a population in continuing decline (the 2021 assessment was led by Kim Williams-Guillén and colleagues). Costa Rica's own national list, issued by SINAC in 2021 (Resolución R-SINAC-CONAC-008-2021) under Decree 40548-MINAE (the Regulation to the Wildlife Conservation Law) and the Wildlife Conservation Law 7317, classifies the species as having a "población reducida" (a reduced population). The capuchin is omnivorous: fruit makes up roughly half to three-quarters of its annual diet, the rest invertebrates, small vertebrates and other plant matter. Of the four Costa Rican monkeys, it is the species that has shown the clearest documented population response to recent drought, both in the Carrera 2025 stress-and-survival paper and in Fernando Campos and colleagues' 2015 paper on climate and demography from Santa Rosa.

The mantled howler (Alouatta palliata, the mono congo) sits at the other end of the diet spectrum. Howlers can fall back on leaves when fruit fails, an option that capuchins do not really have. Katharine Milton and Jacalyn Giacalone's comparative work on Barro Colorado Island in Panama, published in 2014, found that capuchins and howlers reacted differently to the same severe drought events, with capuchins showing more measurable physical stress. Costa Rican work tells a complementary story. Gabriela Jones Román, Cinthia Villalobos Suárez and Rose Marie Menacho Odio's 2021 review in the Universidad Estatal a Distancia's Biocenosis identifies the leading national-scale threats to howlers as habitat loss, fragmentation and electrocution rather than climate per se, with the country's electrical grid documented killing more than seven thousand wild animals in the 2018–2019 monitoring period, nearly a thousand of them monkeys. Climate is one stressor among several for this species; for capuchins it is closer to center stage.

The black-handed spider monkey (Ateles geoffroyi) sits closest to the climate edge. Spider monkeys eat almost only ripe fruit, live in loose societies that split and rejoin across very large territories (a system primatologists call fission-fusion, like a school of fish that breaks into smaller schools and merges again), and need an unbroken canopy to move through. Filippo Aureli and Colleen Schaffner have followed a spider-monkey community in Sector Santa Rosa since 2000. Costa Rica classifies the species as in danger of extinction, on the national list issued under Wildlife Conservation Law 7317. The combination of canopy dependence and ripe-fruit specialization makes the spider monkey the most directly exposed of the four species to the kind of climate-driven fruit famines that Stephen Wright and colleagues documented in 1999, in which the strongest El Niño events propagated through Neotropical mammal populations as a wave of starvation.

The Central American squirrel monkey (Saimiri oerstedii, the mono tití) is the country's only monkey species found nowhere else, restricted to the Pacific lowlands of Costa Rica and adjacent western Panama. There are two recognized forms: a northern one in the Quepos and Manuel Antonio area, with a 2006 estimate by Sue Boinski of roughly 1,500 individuals, and a southern one in the Osa Peninsula and Golfo Dulce, where troops can run up to about seventy strong. The IUCN lists the species as Endangered: its remaining range covers less than 3,500 square kilometers and is broken into pieces, and habitat loss continues. Most of the Costa Rican-authored monitoring has come through ICOMVIS at the Universidad Nacional under Joel Sáenz, Eduardo Carrillo and Grace Wong, with Daniela Solano and Wong's 2009 paper on habitat in Osa as the standard reference. The squirrel monkey is the only one of the four species for which the Pacific dry forest is not the core habitat. Its range sits in the Pacific moist-forest belt to the south, and an intensified climate regime of the kind Birkel projects would push the dry forest's northern boundary southward into squirrel-monkey range.

The riparian refuge, and the threat to it

One of the Jacobson paper's dry-season findings is anchored to a particular patch of ground. At Lomas Barbudal, where the streams run down into the Tempisque, the dry months drive the things the capuchins depend on, fruit-bearing trees, water and shade, into the evergreen gallery forest along those streams. Encounters between neighboring groups rise there. Larger groups occupy the best gallery patches, and smaller groups end up at the edges. The gallery forest is, in plain terms, the refuge that carries the population through the bad years. The same pattern shows up across the rest of the dry-forest fauna. Christopher Vaughan and Kelly Weis, in the Revista de Biología Tropical, mapped dry-forest waterhole use across the Pacific slope. Flávio Guimarães Rodrigues, Eduardo Carrillo, Victor Montalvo and their colleagues, using camera traps in the same forests, documented how mid-sized and large mammals converge on the same waterholes and pathways in the dry months. The pattern is regional and not specific to any one species.

Costa Rican law has long recognized that riparian zones need protection. Article 33 of Ley Forestal 7575, in force since 1996, declared áreas de protección on each side of every watercourse: a strip 15 meters wide in rural zones (10 meters in urban zones) where the terrain is flat and 50 meters where it is steep, and prohibited cutting, burning or building inside them. Ley de Aguas 276, dating to 1942, complements the Forestal Law with parallel hydrological protections.

The more direct threat is a reservoir, now under construction, that will flood part of the reserve. PAACUME, the Proyecto de Abastecimiento de Agua para la Cuenca Media del Tempisque y Comunidades Costeras, promoted by SENARA (the Servicio Nacional de Aguas Subterráneas, Riego y Avenamiento), is building an embankment dam on the Río Piedras, a tributary of the Tempisque. The reservoir will cover about 850 hectares, expand the Arenal-Tempisque irrigation district by roughly 18,800 hectares across the cantones of Carrillo, Santa Cruz and Nicoya, and supply those communities with drinking water. Of the 850 hectares, 113 fall inside the Reserva Biológica Lomas Barbudal. Ley 9610, published in 2018, removed those 113 hectares from the reserve's protected status; the Sala Constitucional upheld it, and as of 2025 the project was about a fifth built, with excavations begun in Bagaces.

The opposition to the desafectación rested on the bees, on the legal status of the reserve, and on the procedural obligations of the EIA. The Jacobson and Carrera papers add a behavioral-ecological dimension: the gallery forest the reservoir will flood is the ground the capuchins fall back on in the worst dry years, part of the spatial structure they use to ride out the droughts that Anderson and colleagues expect to keep recurring as warming continues. Guanacaste is building the reservoir because the province is drying out, and the forest it floods is the refuge the capuchins use against the same drought.

The next thirty-three years

The Jacobson paper does not predict the future, and neither does this article. What both can say is that the rules of competition among capuchin groups at Lomas Barbudal are climate-dependent, and that the climate is moving. The other three primate species each carry their own version of the same story, with the spider monkey closest to the edge and the howler farthest from it.

The method carries over to the other long-term sites; their datasets are already structured for it, and the analysis tools are open-source. A national capuchin observatory, in the working sense, already exists in scattered files at three institutions; what it needs is the time and the funding to keep collecting and to start sharing.

What thirty-three years of walking the same trails bought is the ability to see any of this. A few years of data would have shown that monkeys die in a drought. The record showed what a death count misses: the bigger groups take ground from the smaller ones, and the small groups never take it back. When a hard dry season shrinks the forest to the green strip along the rivers, the big groups hold the best of it and the small groups are pushed to the edges.