This project is part of the "Proyecto Atlapetes", an iniative to protect the species and its habitat through scientific, education, public engagement, and other activities.  

B.Sc. Mauricio Díaz (to the left) mounting an audiomoth recorder on a small tree. These tiny audio devices record all sounds during a time interval and extent (for example, record for a minute every 5 minutes during a week). We used 20 recorders that were moved around to cover 80 sites in order to figure out which habitat patches were occupied by the endangered Montañerito Paisa (Atlapetes blancae)

The mitochondrial genome structure of Coeligena hummingbirds follows the typical organization of birds: 22 tRNAS (pink), two rRNAs (red), 13 protein-coding genes (PCGs; blue) and the control region (grey). Coding sequences (CDS) are in yellow. Substitutions among the three genetic groups C. b. consita (orange), C. h. tamai (blue) and the southern group (green) are represented in the inner circles (singletons and intrapopulation variant sites are not shown). Grey boxes indicate the three non-synonymous substitutions found among the genetic groups, the box with black edges indicates the only substitution involving a change between amino acids with different functional features.

Male C. bonapartei and male C. helianthea.

Ectotherms, whose body temperature is regulated by external energy sources, are expected to be regulated by the interaction among microclimatic dynamics, physiology, morphology and behavior (Kearney & Porter 2019). Although mechanistic models exist that couple various highly parameterized equations, here we attempt a correlative approach to predict daily activity budgets based on an ecological niche model calibrated using microclimatic data. 

Remote sensing imagery is available for most places on earth at high spatial (< 1km) and temporal (<1 month) resolution. Here, Dr. Carlos Yusty and collaborators made use of Sentinel images to derive continuous variables to use in occupancy models for birds. We show that these continuous variables (gradient surface model) capture more landscape heterogeneity and represent a potential improvement for describing organismal responses to landscape variation. 


64. The evolution of sexually dimorphic traits in ecological gradients: an interplay between natural and sexual selection in hummingbirds.

Cover image

A male (left) and female (right) of the Violet-tailed Sylph (Aglaiocercus coelestis) from the rainforest in Colombia. Hummingbird sexually dimorphic traits vary with elevation and habitat structure. Photo credit: Cristian Camacho. See the article ‘The evolution of sexually dimorphic traits in ecological gradients: an interplay between natural and sexual selection in hummingbirds’ by Beltrán et al., published in Proc. R. Soc. B

Currently, some of the most used sources to infer broad-scale patterns of distribution are the Global Biodiversity Information Facility, the IUCN species' distribution maps, and citizen science projects like Ebird and iNaturalist. We expect that as more and more data accumulate in each of these sources, that they will converge and reveal consistent patterns. That was not the case for patterns of threatened vertebrates on islands - as you can see on the figure to your left -, reflecting deficiencies and gaps of information in either source. Another call for caution about the use of these, definitely important, databases. 

In recent years, there has been a push for recognizing historical inequalities, advantages and disadvantages of different communities practicing scientific activities. One of such activities that is of primordial importance in biology is fieldwork. There are many situations, where fieldwork is conducted by international teams, or even groups of biologists coming from different places within the same country. In this paper, we compiled a set of guidelines to promote fair and healthy collaborations among all members of such teams.

Making conservation recommendations is not easy. Often, the complexity and the amount of information considered can obscure the reasons behind each decision. On top of that, making conservation recommendations implies a high responsibility, which can be partly balanced with an explicit and reproducible exercise.  In this study, we share a simple method to identify potential conservation drainages, using various criteria focused on the protection of a potential flagship species, Brycon hennii

Diglossa brunneiventris @tochedemonte

Atlapetes blancae @tochedemonte

Two iconic birds of the Altiplano de Santa Rosa de Osos, Diglossa brunneiventris vuilleumieri - a disjunct population from D. b. brunneiventris in southern Peru and Bolivia whose evolutionary relationships are still debated -, and Atlapetes blancae, a highly restricted endemic species thought extinct until recently. 

The impacts of precipitation gradients  on biodiversity have been far less studied than the impacts of elevation gradients. 

Here we characterize betadiversity along a precipitacion gradient in the Cauca river canyon in Colombia, and found that, in general, tree species turnover is best explained by changes in precipitation and not by dispersal limitation.  


Hoverflies or flowerflies are an enormous group of flies from the family Syrphidae. They are found worldwide but reach their peak diversity in the neotropics. It is hypothesized that ancestors that colonized the noetropics were preadapted to cold conditions, and therefore, high elevation communities might be expected to be more diverse. 

Hoverflies are also important pollinators worldwide and particularly in Europe, they have been used as bioindicators of habitat quality. In this study led by Augusto Montoya under the guidance of Martha Wolff, we evaluated the possibility of using hoverflies in Colombian paramos as bioindicators. 

Manakins are one of the most iconic neotropical birds. Usually, males have conspicuous plumage coloration and other  ornamentation that is used during displays in leks. On top, a male Striolated Manakin (Machaeropterus striolatus) from Maceo, Antioquia, Colombia. 

The White-bearded Manakin (Manacus manacus) and the Golden-collared Manakin have been considered the same lineage in the past. They have a broad distribution encompassing a variety of habitats and climates. Here we characterize morphological variation in relation to habitat and climate variation and also as a function of distance from a known  hybrid zone in northern Antioquia. There are morphological differences between these two lineages as well as within Manacus manacus, mostly following elevational gradients. 

Enceladus is one of Saturn's many moons. It is the most reflective object in the solar system due to its icy surface. Below the icy crust, there is a liquid ocean with hydrothermal vents, a mixture thought unique to Earth. The presence of organic compounds and hydrothermal vents opens the possibility of life in this moon's under-ice sea. Here we explore this possibility using ecological niche models and population growth models. 

My first astrobiological attempt led by Laura Tenelandia, in collaboration with professors Pablo Cuartas and Jorge Zuluaga, all from Universidad de Antioquia! Although applied in the most simple way possible, this paper envisions the potential use of niche models to predict life in extraterrestrial environments. 

Diego Beltrán's first MSc chapter presenting evidence that rates of plumage color evolution in both sexes and in most feather patches are associated with rates of speciation in humminingbirds, a long-held expectation following the hypothesis that sexual selection in relation to mate selection may facilitate the speciation process. 

Figure above shows phylogenetic relationships within the Trochilidae family and branches are colored according to the speciation rate (logarithmic scale). Black lines across tips indicate the major clades within the family and the histograms represent the distribution of speciation rates for each clade. The vertical black hatched line indicates the mean speciation rate for the family as reference. For example, the Bees' clade shows the highest speciation rates, while the Topazes show the slowest. 

Ana Maria's undergraduate research project achieved with the help of awesome international collaborators Jorge Pérez, Ryan Terrill and Jessica Oswald.  

Here we add molecular and climatic niche evidence supporting the hypothesis that the two disjunct populations of Diglossa brunneiventris are not sister lineages but each seems to be related more closely to other populations within Diglossa humeralis

More research (sampling and analyses) is needed to resolve the evolutionary relationships within this complex.

First chapter from Daniel Valencia's masters thesis project about the ecology and conservation of the Sabaleta. Here we modeled the niche and geographic distribution of the Sabaleta (Brycon henni; see a very large individual below) using point locality data (red dots in figure), with predictor variables generated for the river network that include variables that represent the average conditions upstream. We also compare the potential geographic distribution as predicted from this model, to the one used by the IUCN to make decisions about the conservation status of species. 

The inclusion of additional information as offsets (expert maps and elevation limits) can often improve predictions of species distribution models. In this sequence from right to left, the inclusion of both variables leads to a reduction in the fraction of the study area predicted as present, and an improvement in the Area Under the Curve.  

The Reserva Natural Sanguaré is a private reserve 

Daisy's master thesis project about the occupancy and diel activity patterns from three cooccurring species in a dry forest reserve in the Colombian Atlantic coast.  


En este trabajo, la estudiante de pregrado Laura Hoyos escucho cientos de grabaciones para anotar cantos o vocalizaciones de búhos y evaluar una posible técnica para la detección automatizada.

El búho del Chocó (Megascops centralis) es una especie poco conocida, cuya distribución es restringida y aparentemente fragmentada. Su canto es trémulo  y hermoso como pueden escuchar al final de este video realizado por Diego Calderón (#COLOMBIAbirding). 

White Mantled Barbet (Capito hypoleucus)

Two endemic and lovely species from the northeastern slope of the Central Andes in Colombia, the white mantled barbet to the left and the sooty ant-tanager to the right. They can both be found in the buffer zones of various hydroelectric dams frequent in this region. In this article, we characterized the occupancy and detection of these two species.

Sooty Ant Tanager (Habia gutturalis)

Here, Danny Zapata describes the adaptive and neutral genetic diversity across most of the geographic range of a recently described endemic and critically endangered species (Lara et al. 2012) from the Cauca river canyon in Colombia (map to the left, species' illustration below).

To the right, a nest of this species. As many other wrens, the Paisa wren hangs its nest from a stingy plant (the stingy hairs or trichomes can be seen on the picture and believe us, they are stingy). 

How colors evolve in hummingbirds is an exciting question that we have been interested in for a while. The evolution of structural coloration may not only depend on the vagaries of female preferences, but also on the particularities of the barbule nanostructure that interferes with light. In this awesome paper, Chad Eliason investigates why structural colors in hummingbirds are so flexible and compares them to the evolution of structural colors in other birds (ducks).  

Picture on the right shows the phylogenetic relationships among the species included in this study, and the color reflectances of each species' gorget, one of the feather patches involved in mate displays. 


Many species remained hidden in lush forests far away from any human settlement. This is not the case of the Antioquia Brush Finch. For many years this species was hidden less than an hour away from Medellín in an area full of cattle ranches. In 2018, the species was rediscovered by Rodolfo Correa (second from left to right in the right picture). In the picture also Thomas Donegan (the person who described the species in 2007 based on museum specimens), Blanca Huertas (curator at the Natural History Museum of London and the person the species was named after), and Andrea Lopera, another colleague in this project. 

Beautiful mural painted by Biografos (an art collective) in one of the rural schools in San Pedro de Los Milagros. 

A detailed characterization of sexual dichromatism in Coeligena helianthea. Another great job from an undergraduate student, Juliana Sosa.

Figure on the left shows male and female C. helianthea and a more detailed picture of the barbules on a sample feather from the rump, upper back, crown, gorget, and belly. Below you can see sample reflectance spectra from each feather. 

Nice collaboration with Sayra Espindola and Ella Vásquez on approaches to study invasive species. Figure 1 in the article (left) summarizes the approach, where a represents environmental space based on combinations of variable A and B. The fundamental niche of a hypothetical species is represented in d, while b and c correspond to the current environment available at the native and the invaded (non-native) ranges, respectively. The realized (occupied) niche is represented by (e) at the native range and (f) at the invaded range. Indices of niche shift occur inside the central dotted grey area, which corresponds to the analog conditions between the native and invaded ranges. There are areas in (d) that are not currently available, but are areas of potential distribution if an environmental change occurs. (i.e., areas invaded but outside of the native distribution). Non-analog conditions but comprising a potential range to be invaded, because it is within the fundamental niche, are indicated with red (native range) and blue (invaded range) broken lines.

The paramo is one of the most extremely harsh and peaceful ecosystems found in the neotropics. Here we report the avifauna registered during short expeditions to the paramos on the map on the right. Below, a picture taken from paramo del Sol in Urrao. To the left, two of the species found (Grallaria nuchalis above and Uropsalis segmentata below).

Plant - frugivore interactions in a forest fragment in Urabá, Colombia. Small and abundant bird frugivores play central roles in these networks. Great field work by Sergio Montoya, an undergraduate student.

Freshwater fish are very interesting study systems since their habitats are represented by a fluvial network with direction and flow. This leads to very general expectations about the distribution of diversity, some of which Juliana Herrera (an undergraduate student!) evaluated in this article. 

The origin of mountain streams are characterized by low but highly endemic fish communities, including representatives of the Astroblepus genus (image from Wikipedia). 

How do colors distribute among coocurring species? A fascinating question that still keeps me up at night. This is a great contribution from Hugo Gruson and a wonderful team from France.

Figure on the left represents the distribution of hues from gorget and upper back feathers of all hummingbird species (15) present at one site in Colombia in relation to the background hue (B)

Want an example of phenotypically distinct species with a very recent origin? These two (or maybe more) species are one of our favorites :)! On the left, an adult male of Coeligena bonapartei, and on the right, an adult male of Coeligena helianthea (Photos by Diego Torres).


This book comprises an ongoing study of the avifauna along the Cauca River Canyon in the department of Antioquia, where the second largest river in Colombia is being dammed. Map shows some of the localities visited and the steepness of the terrain.

The always amazing Royal Flycatcher! One of the species present at the humid end of the canyon. 

Scanning (a-d) and transmission (e-f) electron microscopy pictures from feathers of the gorget of Anna's hummingbird Calypte anna. The barbules in this brightly colored feathers have a flat reflective portion (a-b) that contains ovoid melanin platelets (c) infiltrated with air pockets (d-f)  that look highly organized from a transversal cut . These structures interfere with light to produce the remarkable iridescent coloration. 


Profile of a section of the Central Andes in Colombia where the northern limit of the distribution for the Great sapphirewing (Pterophanes cyanopterus) lies. White dots represent observations of the species and the two most northern points are the first reports for Antioquia.

Model that describes the probability that one species cooccurs with a focal species is maximum at short to intermediate phylogenetic distances. Species that are very closely related or very distantly related tend to not cooccur.

Climatic suitability through time for Hyperolius cinnamomeus during the last 120000 years.Thin black lines show the country boundaries including the current coastline. Surfaces indicate the climatic suitability of each area through time and how the coastline has changed. Not too long ago (~11000 years ago) the island of Bioko was connected to the mainland, thus allowing the movement of organisms including frogs. 

Boa imperator in Uraba region, Antioquia, Colombia. 

Estimated niches of the two Boa lineages in Central America. The climatic space is defined by the first two principal components that account for  % of the total variation in climate in the study region. Solid and dashed lines represent the climates available within the accesible area for each lineage (100 and 50%, respectively).  

We modelled the climatic niche and potential distributions of three mosquito species from the genus Anopheles that are the three primary vectors of Malaria in Colombia. Figure on the right shows the climatic spaces occupied by the species and identifies two unoccupied regions that are characterized by different NDVI signatures through the year. 


Alternative strategies to define the species pool for a given community (in this example, the community is represented by the red dot) and the different blue intensities represent the similarity in temperature, precipitacion, or the euclidean or a cost-distance from places in geography to the community of interest. The species occurring in these areas could be alternative hypotheses of the species pool. Lessard et al. 2016.

Hypothetical representation of the niches of the various lineages within the genus Nesillas that each inhabits a different island in the Malagasy region. Surprinsingly, to me at least,  each island represents an almost completely different combination of environments. This could partly explain how the colonization of an island could eventually lead to the formation of new lineages, which could afterwards become a cyclical process. 


Simplified diagram representing the different levels of biological organization and how functional traits can be described in them. Lopez et al. 2015. 


Left: Map of Colombia and the department of Antioquia (lower inset) showing all localities with at least one bird specimen in the MUA collection. Morales-Rozo et al. 2014.

Right: Tangara nigroviridis, one of the ~1000 species of birds present in Antioquia. 

Conceptual framework integrating evolutionary and ecological approaches to studying mechanisms contributing to local patterns of diversity along elevational gradients. Diversification rate and time of lineage persistence (panel I) combine to generate potential patterns of diversity along the gradient (panel II). Colonization can also influence potential patterns of diversity, but for visual clarity it is not shown. Ecological filtering acts to produce the observed local richness pattern (panel III) which can take various forms.

In this manuscript, we developed a set of metrics, the specific overrepresentation score (SOS) and the geographic node divergence (GND) score, which together combine ecological and evolutionary patterns into a

single framework and avoids many of the problems that characterize community phylogenetic methods in current use. The figure shows an example of GND scores for the New World flycatchers. The colour scale and symbol sizes are pro-

portional to GND for each node. Nodes labeled A through F correspond to major distributional shifts in flycatcher assemblages. For example, node labeled C has the highest GND score and this node separates the tyrant flycatchers

(Tyrannines and Myiarchines) from the Fluvicolines. The tyrant flycatchers

are primarily distributed in tropical lowland forests but extend to the surrounding savannas and southern North America; whereas the Fluvicolines inhabit colder and drier environments, and extend to the subarctic zones at the pole-

ward tips of South and North America.

The 30% highest and lowest quantiles (high and low) of the three beta diversity dimensions (taxonomic, phylogenetic and trait beta diversity) are summarized through boxplots for four of the best predictors of betadiversity (euclidean distance, cost distance, annual precipitation and elevation) . The dashed line represents the median value of all comparisons in the entire data set. Elevation was one of the best predictors, showing that the highest betadiversity values occurred among communities that differed in ~2000m of elevation whereas the lowest betadiversity values occurred among communities that had differences below 1000 m of elevation.  

The Atrato river mouth in the gulf of Uraba is a mosaic of habitats, full of interesting birdlife including the endemic sooty-capped puffbird (Bucco noanamae).


Population genetic structure of Pseudoeurycea leprosa.. The distribution of genetic variation throughout the Trans-Mexican Volcanic Belt (TVB) can be partly explained by distance and partly by landscape. Barplot below and pie charts represent probabilities of individual assignment to demes.

Geographic predictions of an ecological niche model for the Crested Drongo with the recorded localities for this species (present) and their predicted distribution during the last glacial maximum (LGM) and last interglacial (LIG). The stability map refers to localities that have maintained suitable throught these periods. 


The Phylogenetic Species Variability index (PSV) is a metric of how related coocurring species are. This metric as shown in the figures has a negative relationship with elevation, indicating that coocurring species at higher elevations are more closely related.  R-squared values (rsqu) and the slope (s) of the linear regression are shown on the lower right corner.

The distribution and abundance of bird species in tropical ecosystems has been thought to be influenced heavily through biotic interactions because of high diversity in stable environments. Nonetheless, the study of competition in the field is tough endeavor even for sessile organisms. We outline a series of independent studies, each tackling long-standing questions by utilizing or combining modern techniques and resources that were largely unavailable in previous decades.  Collectively, these approaches offer rich potential for evaluating the strength of competition between interacting species and clarifying its effect on the structure of communities at local and regional scales.

Despite the expectation that high elevation hummingbird communities should have lower diversity of traits because they represent closely related lineages and phenotypic evolution shows a conservative behaviour, we do not see this pattern in observed communities. Either the time scale at which we measure phenotypic evolution is off or we are missing some other key factor.   


We found no evidence of temperate lineages displaying wider elevational ranges than tropical ones. Nonetheless,  temperate lineages exhibited larger temperature ranges than tropical ones. 

The relationship between phylogenetic structure and elevation changes depending on the clades used. On the left we can see that for hummingbird species belonging to the Coquette clade, species coocurring at higher elevations are more closely related than species that cooccur at lower elevations. On the other hand, if we do the same with species belonging to the Emerald clade (figure on the right), the opposite pattern appears. This might reflect different colonization histories among clades but also highlights potential complications in the interpretation of phylogenetic strutcture patterns. 


The figure on the right shows the distribution of species from clades that cooccurr more than expected by chance in the phylogeny, in geography and in environmental space. Nodes highlighted in the phylogeny indicate which clades those species belong to and the same colors are highlighted in geography and in environmental space. The message is that these lineages have particular geographic and environmental distributions. This could indicate adaptations shared across lineages from a common ancestor and also that these species should be under intense competition where they cooccur. 

Geographic prediction of niche models trained on two data sets: Historical (localities collected at the beginning of the 20th century) and Modern (localities collected at the beginnning of the 21st century) for a chipmunk in Yosemite National Park, California, USA. We used this data set to evaluate niche model transfer in time, both from the past to the present and viceversa. 

Distribution of wet forest in the Australia Wet Tropics in green. Dashed line represents the buffer used in palaeomodelling analyses. Sampling localities used for modelling analyses relative to the  modelled distribution of suitable habitat in the current environment for Lampropholis robertsi (B) and L. coggeri (C). Arrows point to regions of over-prediction.


How did all these wonderful color combinations evolve? What are the evolutionary mechanisms behind the production of color diversity in hummingbirds? (*we do not answer all these fabulous questions in this paper)

We present a molecular based hypothesis of the phylogenetic relationships among species in the genus Coeligena. This is an interesting group of hummingbirds all associated to the Andes and having both sexually monochromatic and dichromatic lineages. 

Using ecological niche models and climate surfaces developed for Ecuador we predict the potential geographic distributions of all Grallaria species and identify what parts of the distribution have been affected by deforestation. Patterns of deforestation are very different between the Amazon and the Andes and their consequences are also likely variable. 

Patterns of phylogenetic structure for hummingbird communities across Ecuador. Hummingbird communities inhabiting humid lowland forests (Amazon and Choco) are comprised of evenly distributed lineages across the phylogeny, while communities at very high elevations or in very dry areas (southwestern Ecuador)  are usually comprised of close relatives. This pattern can be interpreted by various mechanisms including ecological filtering, niche conservatism and sequence of colonization events. 


Geographic patterns of contraction, expansion and stability of mammal ranges during the 20th century using two alternative climate reconstructions (Anusplin and PRISM). We identified areas of incongruence between these two climate data sets and highlight the importance of considering climatic uncertainty in niche modeling. 

Changes in elevation ranges detected throughout the 20th century in Yosemite National Park for a group of 28 species of small mammals. Despite the expectation of upward shifts due to climate change, many species showed idiosyncratic changes. 


Ornithological collections are an invaluable resource for biodiversity knowledge. It is imperative that biologists as well as other professionals understand the importance of collections and support their maintenance. There are vast regions in Colombia that have been very poorly sampled. An example are dry forests and their associated fauna. These ecosystems have almost dissapeared because of multiple economic pressures and we have no record of the diversity they hold. 

Atalotriccus pilaris, a common dry forest flycatcher.  


We developed continuous surfaces for four climate variables (Monthly Precipitation, Monthly Mean, maximum and minimum Temperature) for the world terrestrial areas based on a global dataset of climate stations. These surfaces were used to generate a set of 19 bioclimatic variables that are thought to be relevant for the distribution of biodiversity. All these surfaces are frelly available from the Worldclim


We took advantage of a series of climate surface developed for Ecuador and a well curated database of presence localities for all species within the genus Grallaria in Ecuador to evaluate the performance of models based on different predictor variables including remotely sensed variables such as NDVI. 


4. Temporal variation in the diet of black curassows (Crax alector, Cracidae). Cracid ecology and conservation in the new millennium.

Male Crax alector taking a dust bath. 

We provide a thorough review of the diet of this cracid based on close follow up of a small number of individuals. Curassows consume a wide variety of food items but a few can be identified as staple foods. These items also vary considerably depending on the season and the stage of the life cycle. 

We studied the use of space of a family of Mitu salvini during eight months of the year. We identified changes in habitat use related to the distribution of food resources and also to the reproductive stage. 

Family of Mitu salvini foraging through the forest. 


During our undergrad, we had the fortune to visit this wonderful place within one of Colombia's least known National Parks: PNN Tinigua and Serrania de La Macarena. Here we compiled the list of all bird species observed in the area. 

We describe the nests, eggs and chicks of previously little known antbird species.