Are they two seeds in a pod? Comparing seed rain recovery in grasslands using artificial grass carpets versus sticky traps
Data files
Feb 10, 2026 version files 1.88 MB
-
Carpet_Methods_Code_Dryad.Rmd
12.61 KB
-
Carpet-Methods-Code_Dryad.html
1.79 MB
-
README.md
3 KB
-
seed_recovery_all.csv
82.54 KB
Abstract
Premise: Seed dispersal is a critical process for plant community assembly; however, natural rates of seed arrival are rarely quantified compared with other assembly mechanisms, especially in herbaceous communities.
Methods: Here we compare the utility of artificial grass carpet squares (“artificial grass”) for capturing seed rain with classic “sticky trap” methods. We placed paired sticky traps and artificial grass squares in two grassland ecosystems, added known numbers of seeds of multiple species to each trap, and recovered seeds at one-week, one-month, and two-month intervals.
Results: We found that both trap types lost seeds through time at similar rates, but each trap type had advantages and disadvantages. Overall, sticky traps retained more seeds and measured primary dispersal, but recovering seeds was difficult and hindered by debris stuck to the traps. Alternatively, artificial grass traps measured effective dispersal as more seeds were lost through time to secondary dispersal and granivory, but recovered seeds could be handled easily and retained for long-term storage and germination.
Discussion: We encourage the broad adoption of seed rain studies to improve links between theoretical and empirical community ecology. Both sticky traps and artificial grass traps are useful in measuring seed rain in grasslands but vary in the types of information they provide.
Dataset DOI: 10.5061/dryad.j9kd51ct2
Description of the data and file structure
Seed recovery data was collected from two trap types (artificial turf and sticky traps) at two sites (Colorado and Missouri) over the course of one week, one month, and two months.
Files and variables
File: Carpet_Methods_Code_Dryad.Rmd
Description: .Rmd code to run analyses and create figures, and produce an .html file to display results.
File: seed_recovery_all.csv
Description: Data from the seed recovery experiment
Variables
- site: Location where the experiment was conducted (mo = Missouri, co = Colorado)
- block: Block randomization factor and random effect (1-10)
- plot: The specific plot that seed recovery was measured at
- time_month: How long the traps were left in the field (0.25 = 1 week, 1 = 1 month, 2 = 2 months)
- treatment: the type of seed arrival trap used (carpet = artificial turf, sticky = sticky trap, post-sticky = sticky trap sat in the field for a month, then had seeds distributed on it).
- species: The plant species code for the seed species we used in the experiment. Plant codes are the first three letters of the genus name and the first three letters of the species name. Running the full code here provides the full names of the species.
- number_recovered: The number of seeds recovered for each species.
- number_original: The number of seeds of each species placed on each trap
- number_herbivorized: The number of seeds that showed evidence of herbivory
- notes: any sort of note we might need to take.
- pct_recovered: the percent of recovered seeds (number recovered/number original)
- mean_1seed: the average weight of one seed of that species in mg
- sd_1seed: the standard deviation of seed weight of on seed of that species in mg
- number_lost: the number of seeds removed in the experiment (number_original - number_recovered)
File: Carpet-Methods-Code_Dryad.html
Description: an .html print out of the code that runs all analyses and figures.
Code/software
The file "Carpet Methods Code_Dryad.Rmd" runs all the data in R version 4.3.3 and R Studio version 2024.12.0+467.
This file describes all the working directories used to create the output files. They are also listed below:
library(tidyverse)
library(RColorBrewer)
library(viridis)
library(cowplot)
library(glmmTMB)
library(ggeffects)
library(emmeans)
The basic protocol includes:
- Reading data in
- Creating exploratory figures (Figures 2 and 3 in the manuscript)
- Running analyses and creating Figure 4 and 5
- Running analyses for the supplemental materials
Access information
Other publicly accessible locations of the data: