Trait matching affects the probability of nectar robbing in plant-pollinator networks
Data files
Nov 03, 2025 version files 12.30 MB
-
Cameras_data_Ecuador.txt
1.28 MB
-
Interactions_data_Ecuador.txt
9.79 MB
-
Plant_traits.txt
1.21 MB
-
README.md
9.79 KB
Nov 10, 2025 version files 12.32 MB
-
Cameras_data_Ecuador.txt
1.28 MB
-
Interactions_data_Ecuador.txt
9.79 MB
-
Plant_traits.txt
1.21 MB
-
README.md
9.79 KB
-
script.R
20.86 KB
Abstract
Mutualistic interactions support many ecological functions, including pollination. Interactions are, however, vulnerable to cheaters, species that benefit from interactions without providing anything in return. Nectar robbing, where the nectar is depleted but the flower is not pollinated, is a well-known example of cheating and is often observed in pollination networks. Further, pollinating birds often switch between legitimate (i.e., mutualistic) and nectar-robbing flower visits. In this study, we quantify how widespread nectar robbing is at high elevations in the northern Andes using interactions recorded with time-lapse camera traps. Additionally, we assess the importance of two trait-based mechanisms in explaining legitimate versus nectar robbing flower visits by birds: trait complementarity, measured as the continuous difference between bird bill and flower tube lengths, and trait barrier, which is a binary assessment of whether a species pair is physically able to interact based on the length difference between the flower tube and the bird bill. Nectar robbing occurred in 7% of the interactions we sampled, and the specialised flowerpiercers (Diglossa; Thraupidae) relied on this technique at higher frequencies than hummingbirds (Trochilidae). We further observed that the use of nectar robbing was strongly driven by the trait barrier: nectar robbing happened mostly when the bill of the bird was shorter than the flower tube. This suggests that legitimate flower visits are the favoured foraging strategy for nectarivorous birds, and that robbing is used mostly to feed on otherwise inaccessible resources. These results suggest that nectar robbing is an important, yet overlooked, characteristic of tropical bird pollination networks.
Dryad DOI: https://doi.org/10.5061/dryad.rn8pk0pqx
## DOWNLOADING THE DATA
To allow reproducibility, the data and the R script are provided.
## R SCRIPT
All the data preparation and analysis is contained in a single script "script.R".
## DATA
Interactions_data_Ecuador.txt contains data on birds recorded by time-lapse cameras from eighteen sites along an elevation gradient in Pichincha, Ecuador (from which only three sites are considered in the present study). Each row corresponds to an independent interaction between a bird and a plant species. This data set is linked to Cameras_data_Ecuador.txt through the column “waypoint”.
Variables:
waypoint Unique camera ID.
site Site (transect) name.
date Date of the picture (YYYY-MM-DD).
time Time when the picture was recorded (HH:MM:SS).
bird_species Latin name of the bird species detected according to GBIF taxonomy. Can be NA if the species was not identified.
bird_genus Bird genus according to the GBIF taxonomy.
bird_family Bird family according to the GBIF taxonomy.
bird_order Bird order according to the GBIF taxonomy.
bird_sex The sex of the bird, if specified with “m” for male, “f” for female, or “NA” if it was not specified.
interaction Whether the bird was recorded interacting with the flower or not; "yes" for pictures where the bird was seen engaging with the flower; "no" for when the bird in the picture was not engaging with the flower; "NA" if it was not specified.
feeding_activity The way the bird interacted with the flower if specified; “hoverflying” for bird interacting with flowers while flying; “perching” for bird interacting with flowers while perched; "perching,hoverflying" for a combination of both; “no_feeding” if the bill of the bird was not touching the flower, or “NA” if it was not specified.
nectar_robbing The foraging behavior of the bird in terms of legitimate feeding, where the reproductive parts of the plant touched the bird, or nectar robbing where the bird pierced a hole in the flower base or used an existing hole to extract nectar; “yes” indicates nectar robbing, where the bird used a hole at the base of the flower to extract nectar; “no” indicates a legitimate interaction; “maybe” indicates possible nectar robbing that could not be conclusively determined from the picture; “thief” indicates nectar robbing without the use of a hole (e.g., feeding from nectaries of Marcgraviastrum gigantophyllum); “not_interacting” applies when the bird was visible in the picture but not observed interacting with the flower; and “NA” is used when no information was specified. Note that in data from Ecuador, “no” was also used in cases where the robbing status could not be fully determined because the bird was positioned behind the flower, meaning “no” may overlap with “NA” in these instances.
folder Folder where the picture is stored (for internal use).
filename File name of the picture (for internal use).
Cameras_data_Ecuador.txt contains data on each “waypoint” (i.e., sampled plant), including plant species, recording period, and location. This data set is linked to Interactions_data_Ecuador.txt through the column “waypoint”.
Variables:
waypoint Unique camera ID.
site Site (transect) name.
plant_species Latin name of the plant species according to the GBIF taxonomy. If the plant was not identified at the species level, an ID of the unknown species is written, such as “Bromelia sp1”.
plant_genus Plant genus according to the GBIF taxonomy.
plant_family Plant family according to the GBIF taxonomy.
plant_order Plant order according to the GBIF taxonomy.
start_date Date when the camera started recording (YYYY-MM-DD).
end_date Date when the camera finished recording (YYYY-MM-DD).
start_time Time of day when the camera started recording (HH:MM:SS).
end_time Time of day when the camera finished recording (HH:MM:SS).
duration_sampling_hours Total hours of sampling, calculated as the time difference between start and end time. This time duration only considers daylight hours, as the cameras were programmed to turn on at dawn and off at dusk (hours).
duration_from_pics Approximate total hours of sampling obtained from the time difference between the first and last interaction pictures, considering the hours of daylight. This field can be useful if the information from duration_sampling_hours is missing (hours).
longitude_waypoint Longitude of the waypoint associated with the camera (WGS 84, EPSG:4326).
latitude_waypoint Latitude of the waypoint associated with the camera (WGS 84, EPSG:4326).
elevation_waypoint Elevation of the waypoint associated with the camera (m above sea level).
time_waypoint The date and time when the waypoint was recorded in the GPS (YYYY-MM-DD HH:MM:SS).
camera_height_cm Height above the ground where the camera was positioned (meters).
camera_flowers_count Number of open flowers in the frame. Not available for data from Ecuador.
camera_problem Categorical classes indicating whether there was a problem with the camera ("no" - no problem with the camera; "yes" - there was a problem with the camera, "maybe"- not clear whether the camera recorded as expected, "flower_problem" - the flower was covered by a leaf or flowers closed or felt during the sampling, "NA" - if it was not specified).
Plant_traits.txt contains morphological data for most of the species featured in the Cameras_data_Ecuador.txt data set. The data has been sampled on the same sites where interactions have been recorded. The dataset also contains data from Brazil and Costa Rica, but these have been left out in the present study.
Variables:
plant_species Latin name of the plant species according to the GBIF taxonomy. If the plant was not identified at the species level, an ID of the unknown species was written, such as “Bromelia sp1”
plant_genus Plant genus according to the GBIF taxonomy.
plant_family Plant family according to the GBIF taxonomy.
plant_order Plant order according to the GBIF taxonomy.
taxonomic_level Level of taxonomy associated with identification.
flower_symmetry Corolla symmetry and whether it can be divided into two or more identical or mirror-image parts (actinomorphic; asymmetric; zygomorphic).
flower_type Type of flower (bilabiate; campanulate; funnelform; hypocrateriform; ligulate; tubular).
flower_shape Functional characterization of the flower or inflorescence shape (Bell; Brush; Dish; Flag; Gullet; Spurred; Tube).
anther_position Orientation of the stamens with respect to the flower main axis (dorsal; radial; semiradial; ventral).
flower_position Emergence of the flower on the stem (axillar; caulinar; terminal).
petals_fusion Petals fused or not fused (dialipetalous; gamopetalous).
anther_dehiscence Physical mechanism of anther dehiscence and release of pollen grain (longitudinal; pore; transverse; rimose; valvate).
colored_part Organ involved in attracting the pollinator (corolla, bracts, leaves) was determined through the color palette (bract; flower; leaf).
tube_length Length of the floral tube corresponds to the distance, in a straight line, from the base of the corolla to the corolla opening, without taking into account curvatures or protuberances (cm).
stamen_length Measurement of the length of the stamen from the base of the calyx or hypanthium to the apical part without including the thecae, regardless of whether or not it arises from the base of the corolla or calyx (cm).
style_length Length of style from base to the tip of the style (cm).
corolla_opening_lateral Width of the corolla in its lateral plane, at the opening of the floral tube, in species with zygomorphic flowers (cm).
corolla_opening_dorsal Width of the corolla in its dorsal plane, at the opening of the floral tube (cm).
corolla_curvature_central Corolla curvature measured at the center of the flower as k=1/R. Curvature k of the fitted circumference, measured as 1 divided by the radius (cm-1).
sexual_system Gender distribution in the flower or plant (cosexual; dioecious; monoecious).
distyly_occurrence Presence of floral polymorphism comprising two different floral phenotypes, long- and short-styled flowers (yes; no).
sexual_maturation Timing of maturation of male and female reproductive organs in cosexual flowers (homogamy; protandry).
flower_orientation Position of the flowers with respect to the Cartesian plane (erect; eorizontal; pendulous; erect_horizontal; pendulous_horizontal; various).
resupination Orientation of zygomorphic flowers during development resulting in a twist upside down of petals (yes; no).
site Site (transect) identity.
indiv_ID Plant individual identity number.
flower_ID Flower identity number.
country Country identity (Brazil; Costa-Rica; Ecuador).
AVONET Supplementary dataset 1 birdlife.csv contains morphological, ecological, and geographical data for all bird species in the world, obtained from museum specimens. Only “Beak.Length_Culmen” is used in the present study for nineteen bird species sampled in the study area. This data set has been compiled by Tobias et al. (2022). It can be accessed through this link: https://figshare.com/s/b990722d72a26b5bfead?file=34480856 DOI: 10.111/ele.13898
Changes after Nov 3, 2025: I added the R script file, which I apparently forgot to submit.