Severe and widespread reductions in nighttime activity of nocturnal moths under modern artificial lighting spectra
Data files
Dec 03, 2025 version files 109.32 MB
-
Activity_AdditionalData.csv
15.31 MB
-
Activity_MainData.csv
77.75 MB
-
Analysis_Script.R
78.78 KB
-
BayesMod_withoutspecies.Rdata
6.26 MB
-
BayesMod_withspecies.Rdata
9.89 MB
-
LightSpectra.csv
20.57 KB
-
README.md
5.39 KB
-
Species_size.csv
764 B
-
supermatrix.proteins.trimmed.fa.treefile
1 KB
Dec 11, 2025 version files 108.37 MB
-
Activity_AdditionalData.csv
15 MB
-
Activity_MainData.csv
77.11 MB
-
Analysis_Script.R
79.73 KB
-
BayesMod_withoutspecies.Rdata
6.26 MB
-
BayesMod_withspecies.Rdata
9.89 MB
-
LightSpectra.csv
20.57 KB
-
README.md
5.30 KB
-
Species_size.csv
764 B
-
supermatrix.proteins.trimmed.fa.treefile
1 KB
Abstract
Artificial lighting has many negative impacts on nocturnal insects, from harmful phototaxis to disruption of feeding and reproduction. Although comparatively poorly-explored, many detrimental outcomes could be associated with changes in activity. We investigated the effects of different types of light-emitting diode (LED) lighting on activity in multiple wild-caught moth species in three different families. While behaviour under natural conditions varied among species, artificial lighting strongly and largely consistently suppressed activity. In our main experiment, testing 843 moths of 23 species, we found that white LEDs at 10 lx illuminance depressed activity by 85% on average relative to natural nighttime illumination, and even purportedly less harmful amber lighting had similar impacts at the same intensity. There were no differences between the effects of broad-spectrum LEDs and combinations of narrowband LEDs that produce equivalent light for human vision. Collection methods, using light traps or hand-catching with nets, did affect activity in some species, with implications for future research. Finally, further experiments found significant activity suppression overall under lighting at 1 lx, for white LEDs especially, with some species affected even by skyglow levels of white light. Substantial inhibition of activity under multiple streetlight-relevant LEDs suggests potential for widespread impacts on moth populations.
Dataset DOI: 10.5061/dryad.jq2bvq8ps
Description of the data and file structure
Last updated: 24/11/2025
Study summary: This study tests the effect of artificial light treatments on activity in multiple species of nocturnal moths. Key analyses for the main experiment include 1) differences in overall activity between control and lit treatments, 2) Bayesian phylogenetic modelling of overall activity across treatments, 3) models testing differences in activity patterns between species in the control conditions, and across species between lighting treatments. In addition, smaller datasets are analysed to test 4) impacts of lower light levels on activity patterns in a smaller number of species, and 5) effects of different capture methods.
All analyses carried out in R v.4.3.2.
Files and variables
File: Analysis_Script.R
Description: R script containing code for all analyses & plots presented in the paper
File: Activity_MainData.csv
Description: Compiled results of all video analyses for the main experiment; each row represents a frame from an analysed video.
Variables
- Date: Date each trial started
- Moth_ID: Individual ID for each moth in the experiment
- Light_type: Type of light used - White, Broad Amber, Narrow Amber, RGB White, RGB Amber, Natural (Control)
- Light_intensity: Illuminance (in lux) of light treatment - 0, 0.1, 1, 10
- Expt_Start: Starting time for each experiment
- X: X coordinate of moth position in frame
- Y: Y coordinate of moth position in frame
- Xcm: X coordinate of moth position in frame, converted to cm (using diameter of funnel top as scale bar)
- Ycm: Y coordinate of moth position in frame, converted to cm (using diameter of funnel top as scale bar)
- OnFunnel, OnLid, OnBase: Variables indicating whether moth was resting on the sides of the chamber, its lid, or its base respectively. 1 indicates presence of moth, 0 absence.
- zone1 - zone12: dicates whether position of moth overlaps with each of 12 background segments lining the sides of the experimental chamber. 1 = presence, 0 = absence
- zone0: Indicates whether moth position overlaps with no background segments (ie. on base, or on lid over base). 1 = position does not overlap with any segments, else 0
- Time: Time at each frame
- Species: Moth species
- Treatment: Combination of light type and intensity, defining light treatments used for analysis
- NightDate: Date object for date of start of trial
- RealDate: Date object for actual time at each frame (experiments take place overnight, so begin on one date and end on the next)
- DateTime: Date & time object combining date and time at each frame
- PlotDate: Dummy variable facilitating plotting of experiments run on different days along the same axis
- Family: Moth family (Erebidae, Geometridae, Noctuidae)
- CaptureMethod: Method of moth collection - trap (light trap) or net (butterfly net at night)
- Location: Location of moth collection - Campus (university grounds) or Garden (private gardens)
- FloorQuarter: Time in 15 minute intervals, to allow correspondance with weather station records
- Temperature: Temperature, in °C, at the time of each frame (averaged over 15 minute intervals), from weather station on university grounds
- Humidity: Humidity, in %RH, at the time of each frame (averaged over 15 minute intervals), from weather station on university grounds
File: Activity_AdditionalData.csv
Description: Data for additional experiments with lower light levels
Variables
- Most variables as above for main dataset.
- Frame: Number of video frame analysed
- Time_Seconds: Time since start of recording, in seconds
File: LightSpectra.csv
Description: Unprocessed spectra for lights used in the experiments
Variables
- Wavelength: Wavelength (in nm) of each measurement, from 300-700nm
- White, PC amber, Narrow amber, RGB amber, RGB white: Irradiance, in W/m2/nm, measured for each light type. PC amber = Broad amber LED
File: Species_size.csv
Description: Forewing length information (in mm) for species in the main experiment, from Waring & Townsend (2018)
Variables
- Species: Moth species (all 23 in main dataset)
- ForewingLength_min, ForewingLength_max, ForewingLength_mean: minimun and maximum forewing length, from Waring & Townsend (2018), and calculated mean
File: supermatrix.proteins.trimmed.fa.treefile
Description: Phylogenetic tree for 21 species in main experiment
Files: BayesMod_withspecies.Rdata, BayesMod_withoutspecies.Rdata
Description: Saved outputs of MCMCglmm models (see code for details)
Code/software
All analyses carried out in R v.4.3.2, see Analysis_Script.R for all code.
Package versions: tidyverse_2.0.0, patchwork_1.2.0, ggpattern_1.1.3, rptR_0.9.22, lme4_1.1-35.1, lattice_0.21-9, multcomp_1.4-25, glmmTMB_1.1.10, DHARMa_0.4.6, MuMIn_1.47.5, AICcmodavg_2.3-3, lubridate_1.9.3, pavo_2.9.0, moonlit_0.1.0, phytools_2.3-0, geiger_2.0.11, phylobase_0.8.12, phylosignal_1.3.1, ape_5.7-1, MCMCglmm_2.36, emmeans_1.10.0, tidybayes_3.0.7
Changes after Dec 3, 2025:
A small number of moths in the original dataset were incorrectly labelled as moving in the first frame of their trials; to fix this problem, the "movement " columns have now been removed from the raw data, and the movement calculation has been placed within the analysis script, so that the process for calculating movement between frames is explicit, and the resulting data are correct.