Buzz pollination: Investigations of pollen expulsion using the discrete element method data
Data files
Nov 15, 2024 version files 7.25 MB
-
DEMBuzzPollinationData.zip
7.24 MB
-
README.md
6.85 KB
Abstract
This dataset captures the particle dynamics within vibrating anthers, specifically focusing on pollen expulsion patterns under buzz pollination conditions. Using the discrete element method (DEM) within STAR CCM+, we conducted a parameter sweep across various vibrational frequencies to simulate natural pollination processes. The dataset includes time-resolved metrics: the number of particles in the system over time, particle-particle collisions, and particle-wall collisions. These variables provide a comprehensive view of how vibrational parameters affect pollen ejection. This data is valuable for studies aiming to understand mechanical pollen transfer in plants and could inform agricultural practices for optimizing pollination efficiency. The dataset emphasizes the role of specific vibrational frequencies and collision interactions in achieving effective pollen dispersion from the anther.
https://doi.org/10.5061/dryad.0p2ngf29q
Description of the data and file structure
This dataset was collected as part of a numerical study investigating pollen expulsion dynamics within vibrating anthers during buzz pollination. Using the discrete element method (DEM) in STAR CCM+, simulations were conducted across a range of vibrational frequencies and amplitudes to examine how mechanical stimuli impact pollen release and distribution. The data captures time-resolved metrics, including particle counts, particle-particle collisions, and particle-wall collisions, to analyze the factors influencing effective pollen dispersal. The results aim to improve understanding of pollination mechanics, with potential applications in ecological and agricultural contexts.
Files and variables
File: DryadRepositoryFiles.zip Directory Structure:
- ParticlesOverTime
- Contains data on the number of particles in the system over time.
- Subdirectories include
NonPoricidal,Poricidal, andPoricidalPC(Particle Count variations). - Example files:
150HZ_0.2.csv: Data for 150 Hz frequency with a 0.2 amplitude.
- PP_Interaction (Particle-Particle Interactions)
- Tracks the count of particle-particle collisions over time.
- Files are named by frequency and amplitude, e.g.,
550HZ_0.2PP.csv.
- PW_Interaction (Particle-Wall Interactions)
- Tracks the count of particle-wall collisions over time.
- Files are named by frequency and amplitude, e.g.,
900HZ_1PW.csv.
Variable Definitions:
- Time: The simulation time at each data point (in seconds).
- Particle_Count: The number of particles remaining in the anther model over time.
- PP_Collisions: The cumulative count of particle-particle collisions at each time point.
- PW_Collisions: The cumulative count of particle-wall collisions at each time point.
Units of Measurement:
- Time: Seconds (s)
- Particle Count, PP_Collisions, PW_Collisions: Count (dimensionless)
File Naming Convention:
Files are named using the format [Frequency]HZ_[Amplitude].csv, where:
- Frequency: Vibration frequency used in the simulation (e.g.,
150HZ,550HZ). - Amplitude: Vibration amplitude applied in the simulation (e.g.,
0.2,1), units of millimeters.
Missing Values:
- Any missing values in the data files are indicated by blank cells.
Required Software
- MATLAB (R2023a or later)
- Description: MATLAB is used to load, process, and analyze the dataset. Although no specialized toolboxes are strictly required, the Curve Fitting Toolbox can be useful for advanced analysis.
- Alternative: For users without MATLAB, Octave (an open-source alternative) may be used to run simpler parts of the scripts, though compatibility with more complex functions may vary.
- MATLAB Workflow:
- The provided MATLAB scripts automate data analysis across the dataset, including loading CSV files, calculating expulsion rates, and visualizing particle interactions.
- Directory Setup: Place all CSV files in their respective subdirectories (e.g.,
ParticlesOverTime,PoricidalData,NonPoricidalData) within the same working directory as the MATLAB scripts. - Execution: Running each script sequentially will analyze particle expulsion characteristics and interaction counts, generating scatter plots, surface plots, and summary data.
- Python (version 3.9 or later)
- Packages: For users without MATLAB, the following Python packages can replicate much of the analysis:
- Pandas: For reading and handling data from CSV files.
- NumPy: For performing numerical operations on datasets.
- Matplotlib: For visualizing data with plots.
- Python Workflow:
- Load the CSV files into Pandas dataframes, and use NumPy to replicate calculations such as normalization and binning. Matplotlib can create visualizations for particle count, particle-particle, and particle-wall collisions over time, similar to the MATLAB outputs.
- Packages: For users without MATLAB, the following Python packages can replicate much of the analysis:
- Spreadsheet Software
- LibreOffice Calc or Google Sheets: Both provide an easy way to inspect or perform basic operations on the raw CSV data files before deeper analysis.
Provided Scripts
MATLAB Scripts:
The following MATLAB scripts perform a complete analysis of the dataset, each tailored to a specific aspect of particle dynamics:
- LinearRegionAnalysisVelAccJerk.m
- Purpose: This script analyzes the linear region of pollen particle expulsion rates across frequency and amplitude sweeps, capturing metrics like jerk, velocity, and acceleration.
- Functions:
- Loads files from
ParticlesOverTimeto analyze poricidal, pseudo-poricidal, and non-poricidal configurations. - Calculates slopes, velocities, accelerations, and jerks to identify expulsion characteristics.
- Produces scatter plots comparing expulsion rate against jerk, velocity, and acceleration for each configuration.
- Loads files from
- Outputs: Saves visual plots and, optionally, summary data to CSV.
- PoricidalInteractionCounts.m
- Purpose: This script calculates and visualizes particle-particle (PP) and particle-wall (PW) interactions within poricidal anthers.
- Functions:
- Loads data from the
PoricidalDatadirectory and calculates maximum interaction counts. - Generates 3D surface and 2D contour plots to visualize interaction frequencies and compare PP and PW interactions.
- Loads data from the
- Outputs: Creates visualizations for PP and PW interactions and interaction fractions across different frequencies and amplitudes.
- PseudoPoricidalInteractionCounts.m
- Purpose: Similar to
PoricidalInteractionCounts.m, this script performs the same analysis for pseudo-poricidal anther configurations. - Functions:
- Loads data from
NonPoricidalDatafor pseudo-poricidal configurations. - Calculates PP and PW interactions and produces visual comparisons using 3D surface and 2D contour plots.
- Loads data from
- Outputs: Generates similar plots for interaction counts and comparisons as the poricidal script.
- Purpose: Similar to
Workflow: Ensure the relevant CSV files are in the correct folders within the MATLAB working directory. Each script iterates over all relevant files in its directory, processing the data and producing visualizations and summaries for expulsion rates or interaction counts. The default directory structure of the zip file is the proper format.
Access information
Data was derived from the following sources:
- This dataset was generated solely through simulations in STAR CCM+ and has no external data sources.
This dataset was generated using discrete element modeling (DEM) in STAR CCM+, simulating pollen dynamics within vibrating anthers to study pollen ejection patterns under buzz pollination conditions. The simulation applied a parameter sweep across vibrational frequencies to observe pollen particle behavior and collision dynamics within an idealized anther model.
Key variables recorded in the raw dataset include:
- Number of Particles in the System Over Time: Tracking the total pollen particles within the anther model at each timestep.
- Particle-Particle Collisions Over Time: Documenting interactions between pollen particles as they respond to vibrational stimuli.
- Particle-Wall Collisions Over Time: Capturing contacts between pollen particles and anther walls, which influence pollen ejection trajectories.
The dataset consists of unprocessed, time-resolved data files that retain the original simulation outputs. To assist in analyzing the data, MATLAB scripts are provided, which include functionality for normalizing particle counts, binning collision events, and calculating pollen ejection rates. These scripts support users in performing detailed time-series analyses and comparisons across different vibrational parameters to understand pollen dispersal mechanics.