Predator size structure fails to alter nonconsumptive effects in streams
Data files
Nov 04, 2025 version files 668.37 KB
-
Enclosure_experiment_code_9-26-25.R
48.33 KB
-
Enclosure_experiment_data_10-15-25.csv
617.18 KB
-
README.md
2.86 KB
Abstract
Predator population size structure varies over space and time, mediating the top-down, consumptive effects of predators on ecosystems. Yet the role of predator size variation in governing nonconsumptive predator effects has received little targeted research attention. We manipulated stonefly (Acroneuria abnormis) predator size structure and feeding ability and measured effects on the benthic invertebrate prey community in a headwater stream. Field enclosures retained stonefly predators but allowed smaller prey to emigrate as a behavioral avoidance response. Stoneflies caused a ~30% reduction in total prey abundance regardless of whether or not they could feed, indicating a major role for nonconsumptive effects in determining the overall predator effect. This pattern was consistent across two different stonefly predator size structures with equivalent total biomass, as well as most prey responses measured at both the community level and the individual taxon level. Our study demonstrates that stonefly predators cause a community-scale nonconsumptive effect and suggests that predator biomass, rather than predator size structure, might determine the strength of this effect.
Dataset DOI: 10.5061/dryad.905qfttzs
Description of the data and file structure
We ran an experiment testing the effects of stonefly predator size structure (small vs. large) and mouthpart gluing (unglued vs. glued) on prey community structure in streams. Our experiment crossed these factors resulting in 4 unique treatments and we additionally conducted a no-stonefly control. We considered effects of experimental factors on aspects of prey community structure (measured through abundance data) as well as prey size structure (measured using prey head width data).
Files and variables
File: Enclosure_experiment_data_10-15-25.csv
Description: Data are in a CSV format and are viewable in MS Excel. Each row in this data set represents a single prey individual that was sampled from an experimental enclosure. Some prey head width data are missing due to damaged samples and, therefore, encoded as NA.
Variables
- block: The temporal block
- start: The day the block began
- end: The day the block ended
- start_temp: The stream temperature at the start of block (degrees Celsius)
- end_temp: The stream temperature at the end of block (degrees Celsius)
- start_flow: The current velocity at the start of block (feet per second)
- end_flow: The current velocity at the end of block (feet per second)
- enclosure: A numerical code for the particular enclosure used
- unit: The number of the experimental unit
- replicate: The replicate number of each treatment
- individual: The individual insect prey number
- general_taxon: The general taxon of prey (e.g., Order)
- taxon: The more specific taxon of prey (e.g., a genus)
- conversion_rate: Conversion used to measure prey head capsule width under the microscope
- width_marks: The number of width marks on the microscope scale bar
- head_width: Individual prey head capsule width in mm
- size: Stonefly size treatment
- glued_or_unglued: Stonefly mouthpart gluing treatment
- treatment: Treatment as a 5-level factor
- predator: Whether the treatment included a stonefly predator or not
Code/software
R script (Enclosure_experiment_code_9-26-25.R) was written and run using R version 4.3.0. The R script runs all the analyses and creates all figures presented in the published article. The package versions used in running this script are:
ggplot2 - ‘3.4.2’
dplyr - ‘1.1.2’
lme4 - ‘1.1.33’
lmerTest - ‘3.1.3’
multcomp - ‘1.4.24’
emmeans - ‘1.8.6’
pbkrtest - ‘0.5.2’
scales - ‘1.2.1’
vegan - ‘2.6.6.1’
DHARMa - ‘0.4.6’
reshape2 - ‘1.4.4’
tidyverse - ‘2.0.0’
ggvegan - ‘0.1.999’
devtools - ‘2.4.5’
cowplot - ‘1.1.3’
Rmisc - ‘1.1.3’
ggsci - ‘3.2.0’