Current Projects

Peruvian Andes    Monteverde    Amazon    Southern Appalachians    Florida

Back to Home

Manu Biosphere Reserve, Department of Cuzco and Madre de Dios, Peru

Avian diversity in the Southeastern Peruvian Andes
Collaborators: Chris Merkord

          A primary goal of my dissertation research is to understand the spatial patterns of species diversity in tropical Andean landscapes.  Within the Manu Biosphere Reserve, our research team has used auditory-visual counts and mist nets to survey an expansive elevational gradient of 2600 meters, ranging from foothill forests at 800 meters (2,625 feet) to forest treeline at 3400 meters (11,150 feet).  I am now using these census data to quantify and describe patterns of avian richness and species turnover with elevation, and to assess the degree to which  landscape heterogeneity found at similar elevations, generated by variation in topographic complexity and recovery from landslide disturbances, contribute to the overall diversity of bird faunas.  This aspect of species turnover has been largely ignored in previous studies of elevational gradients, perhaps due to sampling limitations in difficult terrain, but it is nevertheless important for understanding levels of patchiness and habitat specialization of species within their ranges.
          These data serve multiple purposes for other research projects, for example, serving as training and test data for species distribution models, including those that utilize remote sensing data, for modeling species' response curves along elevational gradients, and for comparisons with diversity patterns of other taxonomic groups (see  below). My collaborator, Chris Merkord, maintains the Kosnipata Ecosystem Project website, which describes some of this work and other ongoing projects in the Manu Biosphere Reserve.

Cloud forest after a mid-afternoon rainstorm in Manu at 1700 m.

Eutoxeres condamini is found in the lower elevations of Manu National Park. We might expect this species to be specialized to few flowering plant genera based on its bill morphology. Photo by Zach Peterson.
Hummingbird Specialization to Flowering Plants in the Andes
Collaborators: Mark Bush, Doug Levey, Scott Robinson, Honey Whitney

The highest diversity of hummingbirds is found in the Andes Mountains in South America, where species often have elevational distributions of a few hundred meters. Despite their well-studied specialization to flowering-plants, no one has examined how relationships between hummingbirds and flowering plants may constrain hummingbird species’ elevational distributions. One enormous advantage of hummingbirds is that they leave a record of which flowers they have visited through the accumulation of pollen on their bills and heads, thus making costly studies that follow hummingbirds as they move through the environment unnecessary.  In our project, we used the Neotropical Pollen Database develolped by the Palynology lab of Mark Bush to identify pollen samples collected from bills of captured hummingbirds along an elevational gradient in Manu, and we are now compiling a list of flowering plants visited by each of 38 species of hummingbirds. These data, combined with data on the occurrence of those hummingbird  and flowering-plant species along the same gradient, will be used to determine whether range limits in hummingbirds specialized to few flowering-plant species coincide with limits in the distribution of their flowering plants, whether hummingbirds with broad distributions change their use of flowering-plant resources with elevation, and whether specialist hummingbirds are constrained to narrower elevational ranges compared to generalists.

To collect pollen we rub a small slice of prepared gelatin over the hummingbird's bill and forehead. This gelatin is then melted and fixed on a microscope slide under a glass coverslip. Photo of Ocreatus underwoodi by Zach Peterson.

Andean bird-vegetation associations: using remotely sensed RADAR data for distribution modeling and biodiversity mapping
Collaborators: Chris Merkord, Norma Salinas, Sassan Saatchi, Miles Silman, Mark Williams

The study of biodiversity in tropical montane forests has long been plagued by an inability to access and sufficiently sample the heterogeneity of these complex landscapes.  Birds are among the most well studied taxa, yet even the most detailed studies of avian distributions in tropical mountains describe elevational ranges by limits of occurrence along single transects, or estimated abundance based on limited spatial sampling.  For birds, we know that vegetation structure and complexity is an important predictor of species richness and is used in habitat selection for individual species, but ground-based vegetation sampling is time-consuming, spatially limited, and sometimes down right dangerous in steep mountainous terrain. Advances in active remote sensing now allow detailed, quantitative measures of vegetation to be gathered over thousands of square kilometers.  This could be enormously useful for understanding patterns of avian diversity and distributions. 
This project will combine our data of tropical bird abundances along the elevational gradient in Manu with interferometric Synthetic Aperture Radar (inSAR) data for the entire study area.  Using variables derived from radar data as indicators of vegetation structure (e.g., canopy height, above-ground biomass), we can generate landscape-scale predictions for bird species richness, variation in species composition (beta diversity), and individual species’ distributions for this Andean landscape, including population size estimates for the best modeled species.  Remote sensing and GIS applications will be used to construct maps of spatial patterns of avian diversity for the montane regions of Manu National Park.

Overview of Departments of Cusco and Madre de Dios, Peru.  Yellow and blue swaths indicate flight paths for remotely sensed radar data. The yellow swaths, collected in July 2009 by Earth Data, Inc., cover the montane portion of the Manu region.

Understanding evolutionary relationships of Andean birds through ectoparasitic chewing lice
Collaborators: Julie Allen, Gustavo Londono, David Reed, Aaron Savit, Judit Ungvari-Martin, Jason Weckstein

          An assortment of ectoparasites live on birds, among which are chewing lice specialized to the wing feathers and body.  These parasites often track the evolutionary histories of their hosts, though this occurs in differing degrees of cospeciation.  Such relationships give us the opportunity to study microevolutionary processes in bird hosts at regional spatial scales, such as dispersal, gene flow, and population divergence, as well as macroevolutionary patterns of host-parasite coevolution in species groups distributed across biogeographic provinces.  As part of our research efforts in Manu, we have collected ectoparasites from captured birds using a feather dusting technique.  With these data, we are now joining forces with bird lice experts and collections in the Field Museum of Chicago to investigate micro- and macroevolutionary patterns in Amazonian and Andean birds.  

More to come soon!

This Crowned Chat-Tyrant (Ochthoeca spodionota) was feeding nestlings near treeline at 3400 meters. Perhaps due to this constraint, she didn't have time to preen feathers, leaving them susceptible to chewing lice. The barbs on her chest contour feathers have been completely eaten, leaving only the rachii.  


Monteverde Cloud Forest, Tilaran Mountain Range, Costa Rica

Cloud forest habitats are moisture-saturated, resulting in moss-laden tree trunks and unmatched epiphyte diversity and abundance.  Our study shows that range boundaries of numerous bird species coincide with the lower limit of cloud forest along the steep biophysical gradients of the Pacific slope.  
Endemism, beta-diversity, and consequences of climate change in Central American cloud forests
Collaborators: Anna Ciecka, Matthew Gasner, William Haber, Robert Lawton, Kerry Rabenold

          The leeward Pacific slope of the Tilaran Mountains offers a unique opportunity to examine the effect of steep biophysical gradients on the organization of biodiversity.  A central feature of this gradient is the "cloud forest margin" which is characterized by major shifts in vegetation structure and punctuated species turnover.  This mountain range also exhibits extraordinarily high levels of endemism, harboring numerous species limited to Central America and an even greater proportion of species limited to the Costa Rican - Panamanian highlands.  The combination of climate warming and regional deforestation are expected to generate serious warming and drying effects on the cloud forests of this region, adding to the need for urgent and precisely placed conservation efforts.               
          Our research of the bird and tree communities of this landscape spans a decade, in which time we have quantified the diversity of avian and tree communities and the correlative effects of temperature and moisture on species composition along the gradient.  We have found that many narrowly endemic birds tend to be numerically rare, in addition to being habitat specialists, creating a "syndrome of rarity" for many cloud forest inhabitants.  Our yearly censusing of the bird community now allows us to assess yearly variation in abundances of cloud forest species and to provide critical data for analysis of population viability for endemic species.     

This Gray-breasted Wood-Wren (Henicorhina leucophrys
) replaces the White-breasted Wood-Wren
(H. leucosticta) in the Tilaran mountains of Costa Rica.  
Photo by Rosalbina Butron.
Elevational species replacements: Testing the Interspecific Competition Hypothesis in Neotropical Birds

          The phenomenon of elevational replacements between closely related species is well documented in tropical birds, especially in the Neotropics.  It is assumed that interspecific competitive interactions underlie these replacements, preventing the coexistence of strong competitors.  In his study of the Vilcabamba elevational gradient, Terborgh (1975) concluded that such competitive interactions could limit the elevational ranges for up to 30% of the montane species detected.  These conclusions were based primarily on distribution patterns of congeners along single gradients, as well as evidence for range expansion along mountain slopes where one species of the congener pair was absent.  Until now, experimental evidence to support the interspecific competition hypothesis in tropical mountains has been lacking.  
          My project used interspecific playback experiments to test for territorial behaviors between species in their contact areas along elevational gradients.  Such aggressive territorial interactions between species, if
detected, would support the hypothesis that interspecific competition maintains these range boundaries in montane species exhibiting elevational replacements.  Using recorded songs of target species, I conducted a series of these playbacks (control, congener, and conspecific) to birds holding territories in replacement zones between congeners, and at increasing distances moving away from these zones, where interspecific encounter rates decline precipitously.  I have conducted these playback experiments in both the Tilaran Mountains, Costa Rica, and in Manu National Park, Peru.
          Results from the Tilaran mountains of Costa Rica show that species respond aggressively to congener songs in playback experiments where species come into contact along the gradient.  As one moves away from the replacement zone, however, responses to congener songs grow weaker, and at distances well within the elevational ranges of these species, responses to congener songs cannot be distinguished from responses to negative control playbacks. While species tested in Costa Rica showed aggressive interactions with congeners at range boundaries, the strength of interspecific aggression varied between species and among genera, suggesting that some species could be behaviorally dominant.  We emphasize that interspecific aggressive interactions at range boundaries, and especially asymmetries in interactions could pose constraints on species as they attempt to shift along with changing regional climate regimes.

More to come soon!


White Sands Forest of the Northern Amazon, Iquitos, Peru

Using community phylogenetics to understand suboscine assemblages in heterogeneous Amazon forests
Collaborators: Julie Allen, Jessica Oswald, Judit Ungvari-Martin



Southern Appalachian Mountains, Tennessee and North Carolina

Resilience of spruce-fir bird communities to invasion of the exotic Balsam Woolly Adelgid
Collaborators: Anna Ciecka, Kerry Rabenold



University of Florida Campus, Gainesville, Florida

A nesting Northern Mockingbird (Mimus polyglottos) attacking a "repeat offender" threatening its nest.  Photo by Aaron Spalding.
Mockingbirds quickly learn to recognize individual humans
Participants: Monique Hiersoux, Doug Levey, Gustavo Londono, John Poulsen, Scott Robinson, Christine Stracey, Judit Ungvari-Martin

          After working with Northern Mockingbirds for several months, my lab mates, Gustavo Londono and Christine Stracey, began to talk about their mockingbirds becoming more and more aggressive toward them, as they returned periodically to monitor the nests they found.  What's more is they were not acting more aggressively toward other passersby--this seemed to be directed at them.  One day, Gustavo happened to walk by a nest he was monitoring on campus, when one of the mockingbirds picked him out of a crowd of ordinary college students on the sidewalk and dive-bombed at his head, to the shock of the many bystanders! These observations spurred our graduate and undergraduate research groups in the Robinson and Levey labs to organize a formal investigation of mockingbirds' ability to recognize us as individual humans.  Our study, published last year in Proceedings of the National Academy of Sciences USA, received nationwide press coverage.  The take home message is that mockingbirds have a spectacular ability to remember threats to their nests, much more than we have given them credit for!

Check out this example of recognition ability by mockingbirds, and see if you can tell which person is the most threatening to this pair of birds!  And here are links to articles in the press that covered our study: ScienceNOW,, CBS News

Ecological pressures shaping bird song characteristics
Collaborators: Michele Feole, Doug Levey, Puja Patel, Marcela Salazar, Christine Stracey

          Another ongoing project focuses on the effects of urban noise on bird song characteristics.  Several studies have now shown that birds can adapt several aspects of their songs in order to transmit a better signal in noisy environments.  In urban areas, birds may sing louder (increasing amplitude) to overcome increased ambient noise, sing at different times of day, or even change the pitch (or frequency) of the songs that they sing.  It is this final characteristic that we have studied in the Northern Mockingbird.  Two undergraduate students, Puja Patel and Michele Feole, have dedicated many hours to collecting field recordings of singing mockingbirds in and around Gainesville, spanning noise environments that range from bustling rush-hour traffic at major urban intersections to quiet pastures in the countryside.  So far we have found that mockingbirds in urban areas indeed sing at a higher pitch, adjusting the minimum frequencies of their songs to avoid competing with the low frequency rumble of urban road noise.  Non-urban mockingbirds, by comparison, maintain their songs at lower frequencies, which can transmit easily through a less noisy environment.  This project continues to offer multiple layers of interesting questions because of the complicated singing behaviors of mockingbirds to mimic birds in their environment.  Find out more on our work with mockingbird song along an urbanization noise gradient here.


Back to Home