Data from: Functional composition and dominance drive seed removal in ant assemblages of Atlantic forest fragments
Data files
Sep 30, 2026 version files 22.49 KB
-
README.md
6.35 KB
-
seed_removal_plot.csv
4.14 KB
-
site_landscape.csv
752 B
-
species_abundance_site.csv
7.74 KB
-
species_traits.csv
3.51 KB
Abstract
This dataset contains field data on ant assemblages and artificial seed removal collected in 16 Atlantic Forest fragments in the municipalities of Una, Belmonte, Canavieiras, and Mascote, Bahia, Brazil. It includes four tables: (1) plot-level seed removal counts (small, medium, and large seed classes) together with landscape covariates (forest cover, edge density) and site coordinates (n = 80 plots, 16 sites × 5 plots); (2) site-level geographic location and landscape metrics within a 1 km buffer (n = 16 sites); (3) ant species abundance per site in long/tidy format (n = 268 non-zero records); and (4) functional trait data (habitat preference, seed-related functional role, mandible length, body length) for 55 ant species/morphospecies identified by a specialist in Neotropical Formicidae. The dataset supports analyses of how ant functional composition and dominance relate to seed removal in fragmented tropical forest landscapes and can be reused for further studies on ant community ecology, seed dispersal/predation, and landscape effects on ecosystem function in the Atlantic Forest. No human subjects or personally identifiable information are involved. A full data dictionary is provided in the accompanying README file.
Associated manuscript: Ecological Entomology, manuscript ID 26-0199-EEN
Authors: Juliana S. Carneiro, Emanuelle L. S. Brito, Eliana Cazetta, Elmo B. A. Koch, Michaele S. Pessoa
Study area: 16 Atlantic Forest fragments in the municipalities of Una, Belmonte, Canavieiras, and Mascote, Bahia, Brazil
Files
seed_removal_plot.csv— seed removal data at the plot level (n = 80: 16 sites x 5 plots)site_landscape.csv— geographic location and landscape metrics per site (n = 16), corresponding to Table S1 of the manuscriptspecies_abundance_site.csv— ant abundance per species per site, long/tidy format (n = 268 non-zero records)species_traits.csv— functional trait data per ant species/morphospecies (n = 55)
Data collection summary
Voucher specimens: deposited in the ant collection of the Myrmecology Laboratory, acronym CPDC, Executive Commission of the Cocoa Plantation Plan (CEPLAC), Ilhéus, Bahia, Brazil; curated by J. Delabie; complete material lot registered under accession number #5946.
Data dictionary
seed_removal_plot.csv
One row = one plot (5 plots per site, 16 sites, n = 80).
| Column | Description |
|---|---|
| site | Site code (matches site in site_landscape.csv and species_abundance_site.csv) |
| plot | Plot identifier within site (site_plot) |
| cover | Forest cover (%) in a 1 km buffer around the site (site-level value repeated across its plots; see site_landscape.csv for the authoritative site table) |
| edge | Edge density (m/ha) in a 1 km buffer around the site (site-level value repeated across its plots) |
| x | Projected coordinate (UTM, meters; datum WGS84, zone 24S) |
| y | Projected coordinate (UTM, meters; datum WGS84, zone 24S) |
| ric_total | Total ant species richness recorded in the plot |
| abun_total | Total ant abundance recorded in the plot |
| small_seed | Number of small seeds removed during the 60-minute observation period |
| medium_seed | Number of medium seeds removed during the 60-minute observation period |
| large_seed | Number of large seeds removed during the 60-minute observation period |
site_landscape.csv
One row = one site (n = 16). Corresponds to Table S1 of the manuscript.
| Column | Description |
|---|---|
| site | Site code (e.g., 46_1, 78_0); matches site in seed_removal_plot.csv and species_abundance_site.csv |
| municipality | Municipality in Bahia, Brazil, where the site is located |
| longitude | Geographic longitude, decimal degrees (WGS84) |
| latitude | Geographic latitude, decimal degrees (WGS84) |
| FC_percent | Forest cover (%) in a 1 km buffer around the site |
| ED_percent | Edge density (m/ha) in a 1 km buffer around the site |
species_abundance_site.csv
Long/tidy format. One row = one species recorded at one site (zero counts omitted).
| Column | Description |
|---|---|
| site | Site code (matches site_landscape.csv and seed_removal_plot.csv) |
| species | Ant species or morphospecies name |
| abundance | Number of individuals recorded |
species_traits.csv
One row = one ant species/morphospecies (n = 55).
| Column | Description |
|---|---|
| species | Ant species or morphospecies name |
| habitat_preference | Forest dependent / Forest independent |
| seed_related_role | Functional role in seed removal (e.g., Active seed disperser, Seed predator, Incidental role, No role) |
| mandible_length_mm | Mandible length (mm) |
| body_length_mm | Body length (mm) |
License
Data are shared under a CC0 waiver (no copyright reserved), per Dryad requirements.
Contact
Michaele S. Pessoa — michaelepessoa@gmail.com
Seed removal assays were conducted during the morning period, when seed removal by ants is known to be highest (accounting for approximately 60% of total removal), and sampling effort is optimized relative to afternoon or nocturnal periods. At each of the 16 forest fragments, five observation stations were monitored simultaneously, on the same day, by two calibrated observers, avoiding confounding effects associated with diel activity segregation among ant species and ensuring comparable sampling effort across all fragments. Artificial seeds (small, medium, and large size classes) were monitored for 60 minutes; all ants interacting with the seeds were collected and their behavior recorded. A seed removal event was defined as the moment an ant removed a seed from the observation arena; ants that interacted with seeds without removing them (e.g., cleaning behavior) were also collected and recorded.
Ants were preserved in 70% ethanol, sorted, examined under a stereomicroscope, and identified to the lowest reliable taxonomic level based on morphological characters, using a reference collection of Formicidae from the Atlantic Forest, relevant taxonomic literature and identification keys, and direct comparison with reference specimens. All specimens were identified by a specialist in Neotropical Formicidae. Taxonomic nomenclature was standardized according to Bolton's online catalogue of the ants of the world (Bolton, 2026), accessed in February 2026. Voucher specimens were deposited in the ant collection of the Myrmecology Laboratory (acronym CPDC), Executive Commission of the Cocoa Plantation Plan (CEPLAC), Ilhéus, Bahia, Brazil; the complete material lot generated by this study is registered under accession number #5946.
Landscape metrics (forest cover and edge density within a 1 km buffer around each site) were derived from land-cover classification and are reported per site in the site_landscape table.