Pollen collection by the western honey bee and common eastern bumble bee foraging in a common landscape and applications for agri-environment schemes
Data files
Feb 06, 2025 version files 216.26 KB
-
alfalfa_crop.csv
1.15 KB
-
corn_crop.csv
154 B
-
dataset_hectare.csv
209 B
-
dataset_pollen_raw.csv
138.03 KB
-
dataset_resource_raw.csv
43.61 KB
-
pollen_resource_full.csv
24.93 KB
-
README.md
8 KB
-
soybean_crop.csv
176 B
Abstract
Agricultural landscapes often provide an impoverished environment for bees given their limited plant and pollen diversity. Agri-environment schemes (AES) such as flower strips have been developed to improve the quality of the agricultural environment for bees but their efficacy varies with their composition and, for specific pollinators, with the value of the available plant species. This study provides a detailed report of the pollen collection patterns of two bee species, the western honey bee (Apis mellifera L.) and the common eastern bumble bee (Bombus impatiens Cresson), over their foraging season. We compared the floral constancy, pollen richness and diversity of the two bee species and the pollen morphotypes of bee-collected pollen in relation to resource availability. The honey bee was more flower constant while the bumble bee collected a greater family level diversity of pollen. While both bee species collected similar resources over their entire foraging season, the preferred morphotypes in given surveys differed between bee species. Neither bee species collected resources based on their availability but indicated patterns of preference and avoidance. We discuss how such knowledge can inform on the composition of AES to best sustain these pollinators in more impoverished depauperate agricultural landscapes.
https://doi.org/10.5061/dryad.tdz08kq88
Description of the data and file structure
Data were collected during a field experiment in 2016 at a research station in southern Wisconsin. Pollen sacs were collected from honey bees and bumble bees at each of 3 sites at the research station, with one hive from each bee species at a site. Pollen was sampled at 5 discrete time periods between June and September, with 3 sampling days during each period. Researchers moved among sites each day during a period to randomize sampling from hives at each site. Wildflower resources were sampled at 4 discrete time points from June to September, roughly 1 month apart at each of the 3 sites. Within a site there were 5 100m transects, with wildflowers sampled in 10 quadrats per transect. Corn and soybean crops were sampled at 1 time point during bloom, with 3 samples taken from each of 3 fields. Alfalfa was sampled from 6 fields over 3 dates.
Files and variables
The supplementary file is a Word document that provides a short description of the RMD file which was uploaded to Zenodo
File: dataset_hectare.csv
Description:
The hectares of each crop and wildflower area at the research station, along with their relative proportions
Variables
- focal_crop: the local crop being considered
- focal_crop_hectares: number of hectares allocated to a crop
- total_cult_hectares: total hectares across all crops
- focal_crop_farm_prop: proportion of the hectares allocated to a crop
File: dataset_resource_raw.csv
Description:
The wildflower survey data. Values of NA indicate that no flowering plants were found inside the quadrat/sample
Variables
- year: the year of sampling
- survey: The survey which sample occurred (June, July, August, or September)
- site: The site which is being sampled
- transect: The transect within a site which is being sampled
- sample_number: The quadrat within a transect which is being sampled. Range from 1 - 10. If multiple flowering plant species are identified in a quadrat, there will be 2 rows for the quadrat.
- family: Plant family identified in the quadrat
- morphotype: The pollen morphotype associated with that plant family, genera, or species
- genus: Plant genus identified in the quadrat
- species: Plant species identified in the quadrat
- cluster_count: The number of flower clusters identified in the quadrat for a given plant taxon. This could be individual flowers (e.g., Caryophyllaceae), number of heads (e.g., Asteraceae), or number of racemes (e.g., Fabaceae).
- sample_type: Whether it was among crop edges and roadsides (wild), near a native flower strip planting (native plants), or fallow alfalfa fields (alfalfa wild).
- field_crop_primary: The type of crop field the transect is adjacent to. If NA, the transect is not adjacent to a crop field
- field_acreage_primary: The acreage of the adjacent crop field
- average_flower_number: Average number of flowers per cluster of a plant species in a sample (unused in published analysis)
- flower_count: The estimated total number of flowers of that plant species in a sample (unused in published analysis)
File: dataset_pollen_raw.csv
Description:
The raw data pollen identification from bee collected pollen sacs
Variables
- species: The bee species considered (hb = honey bee, bb = bumble bee)
- period: The sampling period starting in June and continuing through September (1-5)
- day: The day within a sampling period
- site: The site where sampling occurred. 1 hb and 1 bb hive are located at each site
- bee: The bee from which pollen was sampled from on a given day. Bee numbers do NOT correspond across days. A bee that had a pollen sac with more than one pollen morphotype will have more than 1 row
- rank: The rank order of most abundant to least abundant pollen in a multi pollen morphotype pollen sac collected from a bee
- count: The number of pollen grains counted using visual microscopy within a pollen pellet
- morphotype: The pollen morphotype identified
- family: The plant family from which the pollen morphotype is associated with
- prop: The proportion of each pollen morphotype in a pollen sac (count / total count)
- total_count: The total pollen grains counted within a pollen sac
File: pollen_resource_full.csv
Description:
This dataset is a compilation of the pollen collection data and resource availability data. It is used to compare the proportion of each pollen morphotype collected across all bees in a colony (a bee species at a site) with the available resources. The variable names that start with A_ to H _ represent the proportion of a pollen morphotype based on which resources are included. All of these columns include wildflowers, and some combination of alfalfa, corn, and soybean. Alfalfa was included during June, July, and August, Corn and Soybean were only included in July for the analyses in this study.
Variables
- species: The bee species identifier
- survey: The resource survey (June, July, August, September). For the bee collected pollen, which had 5 periods, periods 2 and 3 were pooled into the July resource survey for these analyses.
- morphotype: The pollen morphotype being compared
- prop_morph: The proportion of a pollen morphotype in bee collected pollen
- A_Wildflower: Resources that only include wildflowers (resource variable for analyses during September).
- E_Alfalfa_Corn: Resources that only include wildflowers, alfalfa, and corn
- F_Alfalfa_Soybean: Resources that only include wildflowers, alfalfa, and soybean
- B_Alfalfa: Resources that only include wildflowers and alfalfa (Resource variable for analyses during June and August)
- D_Corn: Resources that only include wildflowers and corn
- C_Soybean: Resources that only include wildflowers and soybean
- G_Soybean_Corn: Resources that only include wildflowers, soybean, and corn
- H_Alfalfa_Soybean_Corn: Resources that include all categories; wildflowers, alfalfa, soybean, and corn (Resource variable for analyses during July)
File: soybean_crop.csv
Description:
Number of soybean clusters per quadrat
Variables
- field: The field identifier from where samples were gathered
- quadrat: The sample number within a field. Each quadrat was 1m2
- ttl_racemes_quadrat: The total racemes estimated within the quadrat
File: corn_crop.csv
Description:
The number of corn stalks per quadrat
Variables
- field: The field identifier from where samples were gathered
- quadrat: The sample number within a field. Each quadrat was 1m2
- total_stalks: The total corn stalks estimated within the quadrat
File: alfalfa_crop.csv
Description:
The number of alfalfa flower clusters per quadrat
Variables
- date: The date which sampling occurred
- field_number: The field identifier from where samples were gathered
- species: The plant species being counted. These are all* M. sativa* (alfalfa)
- sample_type_number: The sample number for which an Alfalfa plant was found
- cluster*count: *The number of racemes counted on an Alfalfa plant within a sample. There can be multiple plants within a sample, which is why there can be multiple rows per sample type number variable.
Code/software
The data files were processed using R software. The data processing was performed with an RMD script broken into chunks, and with a short description of each chunk placed immediately prior to the chunk.
The libraries needed, and for which import codes can be found in the first chunk:
dplyr
psych
ggplot2
tidyr
abdiv
vegan
purrr
ggpubr
stringr
econullnetr
Access information
Other publicly accessible locations of the data:
- None
Data was derived from the following sources:
- None
This dataset was collected as part of a field experiment in 2016. Corbicular pollen loads of honey bees and bumble bees were collected, along with floral resource availabilty from June through September. The contents of pollen loads were idenfied using visual microscopy to determine the frequency of flower constancy, along with the richness and diversity of pollen collected by bee colonies. In addition, the abundance distribution of flower resources was compared to that of bee collected pollen to examine more preferred and less preferred pollen types for these two bee species throughout the foraging season. Floral abundance was sampled seperately for wildflowers, alfalfa, corn, and soybean, so there will be a seperate dataset for each.
Data was processed using R, and an annotated RMD script can be found here https://doi.org/10.5281/zenodo.14738621