Skip to main content

Changing patterns of nest predation and predator communities along a tropical elevation gradient


Gomez, Juan Pablo; Londoño, Gustavo (2023), Changing patterns of nest predation and predator communities along a tropical elevation gradient, Dryad, Dataset,


Tropical montane communities host the world’s highest beta diversity of birds, a phenomenon usually attributed to community turnover caused by changes in biotic and abiotic factors along elevation gradients. Yet, empirical data on most biotic factors are lacking. Nest predation is thought to be especially important because it appears to be common and can change selective pressures underlying life history traits, which can alter competitive interactions. We monitored 2,538 nests, 338 of which had known nest predators, to evaluate if nest predation changes along a tropical elevational gradient. We found that nest predation decreased with elevation, reflecting the loss of lowland predators that do not tolerate colder climates. We found different “super” nest predators at each elevation that accounted for a high percentage of events, suggesting that selection pressures exerted by nest predator communities may be less diffuse than has been hypothesized, at least for birds nesting in the understory.


Field Sites description and Data collection

Nesting data were collected through fieldwork conducted between 2008 and 2014 along the elevational gradient of the Kcosñipata Valley, located on the eastern slope of the Andes in southeast Peru (Figure 1). Between August and December each year, we visited four study sites, the Pantiacolla Lodge (Lowlands; -12.6420288, -71.2392228; 412 m), the Tono River (Foothills; -12.9562228 , -71.4816398; 950–1,000 m), the Cock-of-the-Rock Lodge (Mid elevations; -13.055389, -71.546806; 1100-2000 m) and the Wayqecha research station (Highlands; -13.175028, -71.587389; 2,300-3,100 m). At each study site, crews of 4–6 volunteers searched for nests six days per week, ten hours per day. All study sites are connected by continuous forest, representing one of the last elevational gradients with continuous forest between 300 and 3100 m remaining in the Andes. For additional details about the study sites, see Londoño et al. (2015) and Sanchez & Londoño (2016).

At each study site, crews of 4–6 volunteers searched for nests six days per week, ten hours per day. A unique plot of 10 to 15 hectares was assigned to each volunteer. Upon discovery, nests were monitored using one of three techniques, largely based upon immediate availability of equipment; nonetheless, all species and all elevations were monitored with the three techniques in similar proportions. 1) A thermocouple connected to a U-12 HOBO data logger (Onset Computer Corporation, Pocasset, MA, USA, was placed inside the nest to record nest activity by measuring temperature every minute. 2) A Reconyx PC85 or R60 Rapidfire professional color IR camera trap (Reconyx, WI, USA, was placed on a tripod located ~50–100 cm from the nest and covered with vegetation to reduce its visibility. Cameras were programmed to take 10 photos with every movement at the nest, and simultaneously one photo every minute to capture slow-moving predators such as snakes. Cameras remained at the nest until it became inactive due to predation or nestling departure. 3) Visual (in-person) checks every three days during the egg stage and daily (or sometimes every other day) during the nestling stage, until the nestlings were preyed upon or fledged from the nest. At the end of the monitoring period, each nest was assigned one of three possible fates: failed, successful or preyed upon. Failed nests were those with eggs that did not hatch or with nestlings that died for reasons other than predation. Successful nests were those with active and fully feathered nestlings that disappeared between one visit and the next. Preyed-upon nests were identified through camera trap images or, if the nest was monitored using a thermocouple or in-person visits, through discovery of a destroyed nest or an empty nest at a stage when fledging was not a possibility. We acknowledge that these techniques may bias estimates of predation rate, but any such biases are likely homogeneous across sites, allowing us to infer the shape of the relationship between elevation and nest fate. Although we were able to recognize predation events using all three monitoring methods, predators were identified only for those nests monitored using camera traps. Predators were identified to species when possible. In cases in which we were unable to identify a predator to the species level, we classified it to the lowest taxonomical rank possible and assigned it to a morphospecies. Predator identities to the species or morphospecies level were used to estimate the change in predator diversity and predator turnover along the elevational gradient. For analyses to determine how the most common types of predators may change with elevation, predator species or morphospecies were placed into one of seven predator categories: invertebrates, reptiles, birds, marsupials, rodents, primates, and other mammals.

Usage notes

Microsoft Excel


National Science Foundation, Award: DEB1120682