Data from: Lek habitat selection by sympatric manakin species in Northwestern Ecuador
Data files
Nov 17, 2025 version files 43.51 KB
-
fcat-boundary.cpg
5 B
-
fcat-boundary.dbf
1.57 KB
-
fcat-boundary.prj
145 B
-
fcat-boundary.shp
22.04 KB
-
fcat-boundary.shx
108 B
-
lek-habitat-data_pca-variables.csv
3.32 KB
-
lek-habitat-data_raw.csv
7.59 KB
-
README.md
8.73 KB
Abstract
Habitat selection plays a fundamental role in determining community structure and species coexistence, although the role played by sexual selection in shaping settlement patterns is less well understood. Manakins (Pipridae) are a Neotropical family of lekking birds that exhibit similar behavioral ecology across species, both in terms of resource use and dependence on elaborate visual signaling for mate attraction, yet they differ in the form of their sexually selected displays and ornaments. We characterized and compared the spatial dispersion and habitat attributes of lek sites for four species of sympatric manakins in the Chocó region of northwestern Ecuador to test two primary hypotheses for lek habitat selection. First, the interspecific hotspot hypothesis predicts that ecologically similar species should position leks in locations where females are likely to be encountered (e.g., resource-rich patches, topographic channels), and thus lek sites of different species should cluster in geographic space. In contrast, the signal enhancement hypothesis proposes that species should establish leks in habitats with ambient light or structural properties optimal for the transmission or perception of species-specific mating signals, and therefore species should segregate in environmental space due to their distinct displays and plumage ornaments. We found that leks of the four species did not cluster in geographic space and were generally associated with distinct environmental characteristics. Moreover, when species’ leks did overlap in geographic or environmental space, they segregated vertically in terms of display perch height. These findings are consistent with the idea that landscape-level lek site selection by sympatric manakins may be shaped primarily by sexual display optimization, rather than resource-related mechanisms related to their shared ecology. Finally, this study also flags the local population of Masius chrysopterus as a species of potential conservation concern due to its distinct and restricted lek habitat preferences.
https://doi.org/10.5061/dryad.ncjsxkt4c
Study synopsis
The goal of this study was to gain insight into the mechanisms driving habitat partitioning among sympatric species subject to strong sexual selection. Using a community of manakins in northwestern Ecuador, we tested two competing hypotheses for the factors driving lek habitat selection: (1) the interspecific hotspot hypothesis, which predicts ecologically similar species to establish leks in similar locations due to similar patterns of resource use or female movement among species; and (2) the signal enhancement hypothesis, which predicts species to establish leks in environmentally distinct sites to optimize the production or transmission of species-specific sexual signals.
Similar to a study conducted by Loiselle et al. (2007) in the Ecuadorian Amazon, we found no evidence of spatial clustering of manakin leks at our study site, and species’ lek sites segregated in multidimensional space based on environmental characteristics. These results run counter to the idea that ecologically similar species converge in their spatial placement of lek sites and instead suggest that males of different species select distinct locations and habitat types to optimize sexual signaling. This study replicates key findings of the Loiselle et al. (2007) study, while also extending our knowledge of these patterns to the Chocó biogeographic zone, an understudied region of northwest Ecuador of high conservation priority. Importantly, our results raise conservation concerns about the local population of the golden-winged manakin (Masius chrysopterus) due to its restricted lek habitat preferences.
Description of the files and data structure
The files contained herein contain environmental data collected from lek sites of four species of manakins in northwestern Ecuador: red-capped manakin (Ceratopipra mentalis), velvety manakin (Lepidothrix velutina), golden-winged manakin (Masius chrysopterus) and white-bearded manakin (Manacus manacus). The first file ("lek-habitat-data_raw.csv") contains the raw data, while the second ("lek-habitat-data_pca-variables.csv") contains a subset of the variables reformatted for use in dimensionality reduction analyses (e.g., PCA, LDA).
The columns in "lek-habitat-data_raw.csv" are defined as follows:
- Observer: The initials of the person collecting the data.
- Site: The plot where the observation was recorded; either "FCAT" (i.e., the Fundacion para la Conservacion de los Andes Tropicales Reserve) or "Bilsa" (i.e., Bilsa Biological Station)
- Location: The specific locality within site where the observation was recorded.
- Perch Code: A unique identifier for the perch where the male was observed displaying; formatted as "Genus-Lek#Perch#"
- Lek Code: A unique identifier for the lek site where a male perch was located; formatted as "Genus-Lek#"
- Date Encountered: The date that the habitat data was collected
- Species: The species of manakin (Aves: Pipridae) that was observed displaying at the site; four species in this study are mentioned above
- UTM x: Universal Transverse Mercator x coordinate of the observation; collected with handheld GPS
- UTM y: Universal Transverse Mercator y coordinate of the observation; collected with handheld GPS
- Altitude: Elevation of the observation (meters above sea level); collected with handheld GPS
- Tree: Tree family, genus, or species that the manakin was using as a display perch or substrate; UNK indicates an "unknown" tree taxon
- Tree perch height (m): The total height in meters of the tree or plant that the male manakin was observed using as a display perch/substrate; this was not ultimately used in any analyses
- Perch height (m): The height at which the male manakin was observed producing displays
- #Cecropia within a 10m radius: The number of trees in the genus* Cecropia* located within a 10-m radius of the manakin perch/display substrate
- #DBH>10cm: The number of trees with diameter at breast height (DBH) greater than 10 cm located within a 10-m radius of the manakin perch/display substrate
- #DBH>50cm: The number of trees with diameter at breast height (DBH) greater than 50 cm located within a 20-m radius of the manakin perch/display substrate
- #Rubiaceae in a 10m radius: The number of plants in the family Rubiaceae (a family of plants that produces fruits known to comprise a large portion of the manakin diet) located within a 10-m radius of the manakin perch/display substrate; this was not ultimately used in any analyses
- #Melastomataceae in a 10m radius: The number of plants in the family Melastomataceae (a family of plants that produces fruits known to comprise a large portion of the manakin diet) located within a 10-m radius of the manakin perch/display substrate; this was not ultimately used in any analyses
- Canopy height (m): The height of the canopy (in meters) at the manakin display perch/substrate, as measured with a handheld clinometer
- Forest type: General description of the forest age/type; "Primary" forest is defined as old growth, undisturbed forest; "Secondary" forest is defined as forest that is in the process of regenerating due to clearing for agriculture or ranching (generally < 20 years of age); "Altered" forest is defined as forest that has been selectively logged, such that old-growth trees occur in conjunction with flora characteristic of secondary/regenerating zones
- Densiometer: Whether or not a densiometer measurement was taken at the location; with the densiometer, we take the densiometer in the hands at a constant height of 1.30 m, and the densiometer has a bubble of water to control the level (the mime should be in the center); we separate the device 30 cm from the observer so that the head is not reflected in the concave mirror with a grid of 24 squares. Each frame is imaginarily subdivided into four subframes (the total value is 24x4= 96) and the number of subframes that have light is counted to measure the aperture (i.e. the entrance of light/canopy openness). This reading is made in the four cardinal points (North, East, South and West) with the help of a compass. The numerical result is averaged and the single value is multiplied by the densiometer constant (1.04).
- North: The densiometer measurement in the northward-facing direction.
- East: The densiometer measurement in the eastward-facing direction.
- South: The densiometer measurement in the southward-facing direction.
- West: The densiometer measurement in the westward-facing direction.
- Distance to road (m): The distance between the site and the nearest road, in meters; this was estimated by the observer and not ultimately used in any analyses. Cells with "NA" denote those for which distances to the nearest road were not obtained.
- Distance to water (m): The distance between the site and the nearest stream or other water feature, in meters; this was estimated by the observer and not ultimately used in any analyses. Cells with "NA" denote those for which distances to the nearest water feature were not obtained.
- Present water: Whether a water feature was present/visible from the site where the measurements were taken; this was not ultimately used in any analyses.
For the most part, the renamed columns in "lek-habitat-data_pca-variables.csv" can be easily mapped to the original columns (and their definitions) from the raw file, as given above. However, we clarify a few columns:
- Forest-Type: We numerically coded Primary, Secondary, and Altered forests as 1, 2, and 3, respectively.
- Avg-Densiometer: This value is the result of averaging the North, East, South, and West columns in the raw file and multipying the product by the densiometer constant, 1.04, as described above. This value represents canopy openness, with larger values indicating more open canopies.
- We note that the following columns were ultimately excluded from dimensionality reduction analyses due to their qualitative nature, incompleteness, and/or pertinence to habitat use rather than forest structure: Rubes, Melastomes, Forest-Type, Road-dist, Water-dist, Water-presence
We have also uploaded a shapefile (fcat-boundary.shp) and associated files (fcat-boundary.cpg, fcat-boundary.dbf, fcat-boundary.prj, fcat-boundary.shx) to be used for conducting the spatial analyses of lek site dispersion within the FCAT Reserve, as presented in the manuscript.
Code/Software
To see the R code used to generate results from the above data files, please see the included file titled: "Final_Lek_Habitat_Analyses_02.15.25.R
