Experimental evidence for the desynchronization of ecosystem dynamics by global change
Data files
Oct 14, 2025 version files 551.25 KB
-
cv_ScriptDataAnalyses.R
13.48 KB
-
DATAexpe1.Rdata
315.27 KB
-
DATAexpe2.Rdata
220.73 KB
-
README.md
1.77 KB
Nov 03, 2025 version files 3.69 MB
-
cv_Script_FINAL.R
36.42 KB
-
DATAexpe1.csv
1.81 MB
-
DATAexpe1.Rdata
318.94 KB
-
DATAexpe2.csv
1.30 MB
-
DATAexpe2.Rdata
224.28 KB
-
README.md
1.91 KB
Abstract
Understanding the coordinated response of ecosystems is fundamental to better apprehend ecosystem dynamics under environmental changes. Here, we conducted two independent mesocosms experiments to study how warming, nutrient enrichment, predator overexploitation, and their combination desynchronize freshwater ecosystems by studying high-frequency measurements of oxygen dynamics and of their cyclicity at different time scales (sub-daily to monthly). Nutrient enrichment desynchronized the whole dynamic of oxygen relative to control mesocosms and desynchronized oxygen cycles at all time scales, likely due to its effect on primary production. Warming and overexploitation only desynchronized oxygen cycles, especially at short-term, reflecting alteration of the diel rhythm of oxygen dynamic. Lastly, the combination of nutrient enrichment and warming dampened desynchronization, whereas the effects of nutrient enrichment were accentuated when combined with overexploitation. This first empirical evidence of global change impact on ecosystem synchrony opens important avenues to identify the destabilizing mechanisms of ecosystem functioning at uninvestigated spatio-temporal scales.
https://doi.org/10.5061/dryad.4qrfj6qn3
Description of the data and file structure
The data were collected using high frequency sensors in two independant experiments: Experiment 1 conducted in mesocosms and Experiment 2 conducted in experimental ponds.
Files and variables
File: DATAexpe1.Rdata
Description:
One row is one measurement of oxygen saturation.
The dataset includes five columns:
- Mesocosm: the id of the mesocosms
- Hour: is the date with a format POSIXct with yyyy-mm-dd HH:MM:SS
- Treat: is the treatment type that can be Control, Warming, Nutrients, and Combined
- O2mean: is the measure of oxygen saturation, hourly aggregated
- Day: Day format of the date
File: DATAexpe1.csv
Description:
Same as "DATAexpe1.Rdata" but in format ".csv".
File: DATAexpe2.Rdata
Description:
One row is one measurement of oxygen saturation.
The dataset includes five columns:
- Mesocosm: the id of the experimental pond
- Hour: is the date with a format POSIXct with yyyy-mm-dd HH:MM:SS
- Treat: is the treatment type that can be Control, Overexploitation, Nutrients, and Combined
- O2mean: is the measure of oxygen saturation, hourly aggregated
- Day: Day format of the date
File: DATAexpe2.csv
Description:
Same as "DATAexpe2.Rdata" but in format ".csv".
Code/software
The scripts were coded with R 4.3.2, and the packages used in the scripts are: stringr, dplyr, tidyr, ggplot2, readxl, lubridate, lme4, sjPlot, effects, dtw, zoo, WaveletComp, pracma
The R codes can be found in the file "cv_Script_FINAL.R".
They are organized to be reproducible for Experiment 1 and Experiment 2.
- Prepare data
- Cross-correlation analyses
- Wavelet analyses
- Plots
- Supplementary analyses
High frequency monitoring of oxygen during several months in two independant mesocosm experiments.
Changes after Oct 14, 2025: Changes have been done after a review process focused on data and scripts. Analyses remain the same, but the script has been better annotated to better guide future users. In addition, ".csv" files are now provided (they are exactly the same as the ".Rdata").