Data and code from: Prey depletion by a predator guild suggests spatial differences in competitive ability, but not prey partitioning, consistent with functional trade-offs
Data files
Jul 27, 2026 version files 2.51 MB
-
arthropod_abundances_for_publication_120423.csv
810.18 KB
-
arthropod_biomass_for_publication_072426.csv
48.18 KB
-
individual_biomasses_for_publication_120423.csv
1.64 MB
-
lizard_competition_size_groups_for_publication_120423.R
4.97 KB
-
lizard_competition_taxon_groups_script_for_publication_120423.R
2.89 KB
-
lizard_competition_total_biomass_for_publication_120423.R
1.74 KB
-
README.md
4.56 KB
Abstract
Understanding the mechanisms by which competing species coexist is fundamental for explaining the distribution and abundance of organisms. Many coexistence mechanisms require tradeoffs in the ability to deplete shared resources, but resource depletion by competing species is rarely measured, especially in communities of mobile predators. Among competing predators, differences in diet and habitat use are commonly observed, and to promote coexistence, these differences must be associated with tradeoffs in the ability to suppress different prey types (prey partitioning) or to suppress prey in different habitats (spatial differences in competitive ability), respectively. We hypothesize that these two mechanisms are important for competition among three arboreal lizards common in Hawaii. We test these mechanisms using experimental monocultures to measure prey suppression, and a consumer-resource model parameterized with trait and demographic data from the same experiment. We found that species differed in spatial patterns of prey suppression, supporting spatial differences in competitive ability, but species did not differ in which prey taxa they suppressed, providing no support for prey partitioning. Patterns of prey suppression were consistent with differences between species in microhabitat-specific attack rates, and including differences in the conversion of prey into offspring allowed model outcomes to better align with species’ relative prey suppression. When using the model to predict competitive outcomes, Phelsuma laticauda and Anolis sagrei are expected to coexist over a wide range of microhabitat availabilities, Anolis carolinensis can coexist with P. laticauda under a wider range of microhabitat availabilities than it can with A. sagrei, and A. carolinensis cannot persist in the presence of both of the other species under any conditions. These predictions are consistent with the change in community composition that has occurred over time in the field: Anolis carolinensis has declined while the other two species have increased in distribution and abundance. Our results also demonstrate that differences in diet do not necessarily translate into resource partitioning of prey, highlighting the importance of measuring resource depletion when understanding exploitation competition.
https://doi.org/10.5061/dryad.xgxd254pb
A guild of mobile predators exhibits spatial differences in competitive ability, but not prey partitioning, consistent with functional trade-offs
Three data sets with their corresponding R scripts for analysis are provided. Each data set includes data on arthropod prey in different microhabitats over time in the presence of different lizard predators in experimental enclosures. The data sets are for prey abundance, prey size, and prey biomass. The scripts include the models and contrasts to test whether spatial heterogeneity in competitive ability and/or prey partitioning are occurring.
Description of the data and file structure and accompanying code
Here are the pairs of data sets and scripts. All code is in R version 4.3.2
1) prey abundance and the prey taxon model:
data: arthropod_abundances_for_publication_120423.csv
column headers:
- taxon: arthropod taxon, typically to order
- Freq: the number of individuals of that taxon observed
- plot: the experimental enclosure the sample was collected from (ranges from 1-7)
- date: date in month/day/year format
- trap: the trap the sample was collected from (ranges from 1-6)
- treatment: which lizard species was present. asag = Anolis sagrei; acar = Anolis carolinensis; plat = Phelsuma laticauda
- type: location of the trap. veg = suspended above the ground, hanging from the vegetation. ground: pitfall trap sunk in the ground.
- trapUnit: combination of date, plot, and trap columns into one variable.
script: lizard_competition_taxon_groups_script_for_publication_120423.R
This R script uses arthropod counts to perform the primary set of analyses, where arthropod abundances are grouped by taxon. Analyses of abundance grouped by size class and analyses of total biomass are in separate scripts. It uses the following packages: lubridate, glmmTMB, and emmeans.
2) prey size and the prey size model, as well as predicted total biomass from individual measurements
data: individual_biomasses_for_publication_120423.csv
column headers:
- date: date in month/day/year format
- plot: the experimental enclosure the sample was collected from (ranges from 1-7)
- trap: the trap the sample was collected from (ranges from 1-6)
- type: location of the trap. veg = suspended above the ground, hanging from the vegetation. ground: pitfall trap sunk in the ground.
- trt: treatment. Which lizard species was present. asag = Anolis sagrei; acar = Anolis carolinensis; plat = Phelsuma laticauda
- lengthMM: length of arthropod in mm
- widthMM: width of arthropod in mm
- biomassCalc: biomass calculated using taxonomy-specific regressions of length and width
- taxon: arthropod taxon, typically to order
script: lizard_competition_size_groups_for_publication_120423.R
This R script performs analyses where arthopod counts are grouped by arthropod size class, and also compares patterns of total biomass to what is predicted when biomass is predicted from individual sizes. It uses the following packages: dplyr, glmmTMB, lme4, lmerTest, effects, emmeans
3) total biomass and total biomass model
data: arthropod_biomass_for_publication_072426.csv
column headers:
- date: date in month/day/year format
- plot: the experimental enclosure the sample was collected from (ranges from 1-7)
- trap: the trap the sample was collected from (ranges from 1-6)
- cupID: all arthropods from a trap were put in a small foil weighing boat, which we called a cup, and each cup got a unique ID.
- cupMass: this is the mass in grams of the empty cup
- cupMassBugs: this is the mass in grams of the cup with the arthropods in it
- arthBmass: this is cupMassBugs - cupMass. Subtract the weight of the cup from the weight of the cup filled with arthropods to get the weight of just the arthropods.
- type: location of the trap. veg = suspended above the ground, hanging from the vegetation. ground: pitfall trap sunk in the ground.
- trt: treatment. Which lizard species was present. asag = Anolis sagrei; acar = Anolis carolinensis; plat = Phelsuma laticauda
script: lizard_competition_total_biomass_for_publication_120423.R
This R script uses measurements of total arthropod biomass per trap to test for spatial differences in competitive ability. It uses the following packages: glmmTMB, emmeans.