Learning to handle flowers increases pollen collection benefits for bees but does not affect pollination success for plants
Data files
Oct 14, 2024 version files 318.34 KB
-
AnalysisData_All-Trials-Summary-Data_20_8-4-2024.csv
27.77 KB
-
AnalysisData_Handling-Time-Traces-EXAF_5-24-2023.csv
74.87 KB
-
AnalysisData_Handling-Time-Traces-MIGU_5-24-2023.csv
54.18 KB
-
AnalysisData_Handling-Time-Traces-PHCA_5-24-2023.csv
87.96 KB
-
AnalysisData_Handling-Time-Traces-SOEL_5-24-2023.csv
55.12 KB
-
AnalysisData_Naive-Bees_Efficiency_8-4-2024.csv
6.41 KB
-
README.md
12.03 KB
Abstract
Cooperation and conflict are common in plant-pollinator interactions. Flowering plants often entice pollinators to visit by offering floral food rewards, thereby facilitating pollination. However, pollinators such as bees can learn to improve their collection of floral rewards (such as pollen), changing how they interact with the flower’s reproductive organs, which together could reduce pollination success. Consequently, complex flowers that slow pollinator learning might benefit the plant. Yet how pollinator learning and flower complexity interact to affect pollination success is unknown. We therefore asked how differences in complexity of four flower types (Phacelia campanularia, Exacum affine, Solanum elaeagnifolium, and Erythranthe guttata) affected learning by pollen-foraging generalist bumble bees (Bombus impatiens) and how learning affected pollen collection and pollen deposition on these flowers. We found that bees generally learned how to efficiently handle more complex flower types more slowly. Bees that required more visits to become efficient foragers collected less pollen, with no effect on pollen deposition. Except for the simplest flower type, learning also involved development of motor routines unique to each flower type. Experienced bees overall collected more pollen, but individual differences in motor routines did not affect pollen collection. Conversely, individual differences in motor routines affected pollen deposition, but there was no overall effect of experience. Thus, even though learning overall benefits the bee, it does not alter female (and potentially male) fitness benefits for the plant. We discuss potential reasons for these patterns and consequences for bee behavior and flower evolution.
README: Learning to handle flowers increases pollen collection benefits for bees but does not affect pollination success for plants
https://doi.org/10.5061/dryad.gtht76hwk
Description of the data and file structure
We divided painted flower-naïve bees into four treatments, with two sub-treatments each. A minimum of three colonies were represented per treatment. Treatments differed by plant species used: 20 cut flowers of a given species were spaced 7 cm apart in a 5 x 4 Cartesian grid design on the arena wall. We systematically alternated assignment of bees to each treatment to control for effects of time and day on behavior. Sub-treatments differed in terms of whether a bee was initially flower-naïve (‘naïve’) or had previously been tested (‘experienced’): each bee therefore received two consecutive trials with the same plant species, with the naïve and experienced trial separated by approximately 24 hours. After its naïve trial, the bee was returned to its colony, but after its experienced trial, the bee was permanently removed from the colony, to facilitate a colony to recruit more foragers.
To initiate a behavioral trial, flowers were set up and a single painted flower-naïve worker bee was gently captured from the foraging arena using a 40 dram vial (Bioquip) and immediately released in the center of the test arena following Russell et al. (2017). Prior to release, we visually confirmed the absence of pollen on the bee’s body. To estimate how much experience was required to reach an asymptote in flower handling time, we ran several long practice trials (120 visits) and estimated that an asymptote was reached less than 50% through each practice trial; data from these practice trials was not used in any subsequent analyses. We therefore terminated each naïve behavioral trial after 70 visits (or earlier if the bee stopped visiting flowers for 5 minutes) to ensure that bees had learned and to avoid bees depleting flowers of pollen rewards or filling their pollen baskets completely. For a subset of flowers and trials we extracted pollen from the anthers to confirm that pollen had not been depleted after a trial. We terminated each experienced behavioral trial immediately once the bee had made the same number of visits as in its naïve trial. To terminate a trial, we turned off the overhead arena lights and captured the bee in a vial. After terminating a trial, the complete pollen load from one pollen basket (i.e., corbicular pollen load) was carefully removed from the bee, and the anthers and styles of all visited flowers were removed. The pollen load, anthers, and styles were stored separately in 70% ethanol for pollen counting (styles and anthers for each trial were separately pooled into microcentrifuge tubes). The arena was cleaned with distilled water and fresh paper towels thoroughly between trials.
Trials were video recorded to enable using an event logging program (BORIS v.8.19.3) to measure the handling time per flower visit (from three legs on the flower until physical contact with the flower ended), the latency to begin collecting pollen after landing on the flower (from three legs on the flower until the bee began scrabbling or buzzing on the anthers), and the duration of each flower visit during which bees used easily distinguishable categories of pollen collecting handling routines (Figure 3; see Results for a description). We also recorded whether the bee had collected pollen on a given flower visit. We failed to video-record the naïve trial for three bees and this data was therefore unavailable for behavioral analyses.
Assessment of pollen collection and transfer
To examine whether bee foraging experience affected pollen collection, the corbicular pollen load from each trial was stored in separate microcentrifuge tubes with 1 mL of 70% ethanol. We counted pollen in two or three 10 μL aliquots using a haemocytometer (Hausser Scientific, Horsham, PA) at 400× or 100× (Leica DM 500) to arrive at an estimate for the total volume. Estimated pollen counts were rounded to the nearest whole number. To evaluate pollen deposition, the pollen grains on flower stigmas was enumerated. We acetolyzed styles from flowers (following Dafni 1992) pooled by trial, condensed samples by centrifugation to 40 μL, and counted resuspended pollen in two 10 μL aliquots; if we counted zero grains, we counted conspecific grains in all four aliquots.
Files and variables
File: AnalysisData_All-Trials-Summary-Data_20_8-4-2024.csv
Description:
Variables
- BeeID: Unique ID for the bee
- ColonyID: Unique ID for the colony
- Species: Plant species 4 letter abbreviation (1st two letters of genus, 1st two letters of species epithet); EXAF = Exacum affine; PHCA = Phacelia campanularia; ERGU/MIGU = Erythranthe guttata; SOEL = Solanum elaeagnifolium
- Treatment: One of two treatments: experienced and naive
- TimeTotal: total time in seconds spent on flowers in a trial
- StigPol: Number of grains of pollen on the stigmas in a trial
- CorbPol: Number of grains of pollen in the pollen baskets of a bee in a trial
- StigPolperVis: Number of grains of pollen on the stigmas in a trial divided by number of flower visits made in that trial
- CorbPolperVis: Number of grains of pollen in the pollen baskets of a bee in a trial divided by number of flower visits made in that trial
- NumVisits: Number of flower visits made in the trial
- MeanPropMotRoutine: Mean average percent of time spent on the flower, across all visits for a trial, using the original or modified motor routine
- VisitsToCriterion: Number of visits to meet the 80% learning criterion to successfully extract pollen (8 of the last 10 flower visits)
- LearnSlope: Slope of the Wright learning curve for handling time
- LearnIntercept: Intercept of the Wright learning curve for handling time
- AntherLearnSlope: Slope of the Wright learning curve for time to discover the anthers after landing on the flower
- AntherLearnIntercept: Intercept of the Wright learning curve for time to discover the anthers after landing on the flower
- SpeciesDayOne: Plant species for the naive bee; EXAF = Exacum affine; PHCA = Phacelia campanularia; ERGU/MIGU = Erythranthe guttata; SOEL = Solanum elaeagnifolium
- LearnSlopeDayOne: Slope of the Wright learning curve for handling time, naive bee
- LearnInterceptDayOne: Intercept of the Wright learning curve for handling time, naive bee
- AnthLearnSlopeDayOne: Slope of the Wright learning curve for time to discover the anthers after landing on the flower, naive bee
- AnthLearnInterceptDayOne: Intercept of the Wright learning curve for time to discover the anthers after landing on the flower, naive bee
- SpeciesDayTwo: Plant species for the experienced bee; EXAF = Exacum affine; PHCA = Phacelia campanularia; ERGU/MIGU = Erythranthe guttata; SOEL = Solanum elaeagnifolium
- LearnSlopeDayTwo: Slope of the Wright learning curve for handling time, experienced bee
- LearnInterceptDayTwo: Intercept of the Wright learning curve for handling time, experienced bee
- AnthLearnSlopeDayTwo: Slope of the Wright learning curve for time to discover the anthers after landing on the flower, experienced bee
- AnthLearnInterceptDayTwo: Intercept of the Wright learning curve for time to discover the anthers after landing on the flower, experienced bee
File: AnalysisData_Handling-Time-Traces-EXAF_5-24-2023.csv
Description:
Variables
- BeeID: Unique ID for the bee
- ColonyID: Unique ID for the colony
- Species: Plant species 4 letter abbreviation (1st two letters of genus, 1st two letters of species epithet); EXAF = Exacum affine; PHCA = Phacelia campanularia; ERGU/MIGU = Erythranthe guttata; SOEL = Solanum elaeagnifolium
- Treatment: One of two treatments: experienced and naive
- VisitNo: Visit from 1-n that the bee is making
- HandlingTime: Time spent on the flower in seconds
- TimeToAnth: Time to discover the anthers after landing on the flower in seconds
File: AnalysisData_Handling-Time-Traces-MIGU_5-24-2023.csv
Description:
Variables
BeeID: Unique ID for the bee
ColonyID: Unique ID for the colony
Species: Plant species 4 letter abbreviation (1st two letters of genus, 1st two letters of species epithet); EXAF = Exacum affine; PHCA = Phacelia campanularia; ERGU/MIGU = Erythranthe guttata; SOEL = Solanum elaeagnifolium
Treatment: One of two treatments: experienced and naive
VisitNo: Visit from 1-n that the bee is making
HandlingTime: Time spent on the flower in seconds
TimeToAnth: Time to discover the anthers after landing on the flower in seconds
File: AnalysisData_Handling-Time-Traces-PHCA_5-24-2023.csv
Description:
Variables
BeeID: Unique ID for the bee
ColonyID: Unique ID for the colony
Species: Plant species 4 letter abbreviation (1st two letters of genus, 1st two letters of species epithet); EXAF = Exacum affine; PHCA = Phacelia campanularia; ERGU/MIGU = Erythranthe guttata; SOEL = Solanum elaeagnifolium
Treatment: One of two treatments: experienced and naive
VisitNo: Visit from 1-n that the bee is making
HandlingTime: Time spent on the flower in seconds
TimeToAnth: Time to discover the anthers after landing on the flower in seconds
File: AnalysisData_Naive-Bees_Efficiency_8-4-2024.csv
Description:
Variables
BeeID: Unique ID for the bee
ColonyID: Unique ID for the colony
Species: Plant species 4 letter abbreviation (1st two letters of genus, 1st two letters of species epithet); EXAF = Exacum affine; PHCA = Phacelia campanularia; ERGU/MIGU = Erythranthe guttata; SOEL = Solanum elaeagnifolium
NumVisits_HandTimeEfficient: Number of flower visits required to reach an asymptote in flower handling time efficiency
Efficiency_FirstVisitHandTime: Handling time on the first visit in the naive trial as a proportion of the handling time on the first visit in the experienced trial
Slope_EfficientHandTime: Slope of the relative efficiency of a bee during its first trial as a percentage of its maximum efficiency (the mean handling time or latency during its second trial). For all naïve bees we thus calculated initial foraging efficiency for the first flower visit, as well as the rate of improvement in efficiency (slope from linear regressions) until reaching an asymptote in performance (three consecutive visits with ≥ 90% efficiency
NumVisits_AnthLatEfficient: Number of flower visits required to reach an asymptote in efficiency to find the anthers
Treatment: One of two treatments: experienced and naive
TimeTotal: total time spent on flowers in a trial in seconds
StigPol: Number of grains of pollen on the stigmas in a trial
CorbPol: Number of grains of pollen in the pollen baskets of a bee in a trial
StigPolperVis: Number of grains of pollen on the stigmas in a trial divided by number of flower visits made in that trial
CorbPolperVis: Number of grains of pollen in the pollen baskets of a bee in a trial divided by number of flower visits made in that trial
File: AnalysisData_Handling-Time-Traces-SOEL_5-24-2023.csv
Description:
Variables
BeeID: Unique ID for the bee
ColonyID: Unique ID for the colony
Species: Plant species 4 letter abbreviation (1st two letters of genus, 1st two letters of species epithet); EXAF = Exacum affine; PHCA = Phacelia campanularia; ERGU/MIGU = Erythranthe guttata; SOEL = Solanum elaeagnifolium
Treatment: One of two treatments: experienced and naive
VisitNo: Visit from 1-n that the bee is making
HandlingTime: Time spent on the flower in seconds
TimeToAnth: Time to discover the anthers after landing on the flower in seconds
Code/software
R or R studio, R v.4.2.2 (R Development Core Team 2023)
All .csv files are prepared and ready to be analyzed in in the R file
Access information
Other publicly accessible locations of the data:
- N/A
Data was derived from the following sources:
- N/A