Data and code from: Eco-evolutionary feedback to a pesticide worsens the impact of a pesticide switch in a pivotal freshwater non-target species
Data files
Sep 21, 2026 version files 167.13 KB
-
genotype_analysis.R
34.90 KB
-
genotype_data.csv
88.94 KB
-
H0.stan
355 B
-
H1.stan
457 B
-
population_analysis.R
30.68 KB
-
population_analysis.stan
2.15 KB
-
population_data.txt
5.09 KB
-
prior_tune.stan
728 B
-
README.md
3.83 KB
Abstract
Pesticide exposure can drive adaptive evolution of non-target populations living near agricultural areas, however, such adaptation may entail costs. Pesticide switches are common in pest management, which result in non-target populations being sequentially exposed to different pesticides. We assessed to what extent exposure to an insecticide (chlorpyrifos) affects genotype composition of populations of a non-target species, Daphnia magna, and subsequent responses of these to a second pesticide exposure. We designed a two-phased experiment, in which D. magna populations were, in a selection phase, exposed to either a chlorpyrifos or control treatment, and the resulting populations were subsequently exposed to chlorpyrifos, malathion (same mode of action as chlorpyrifos), deltamethrin (different mode of action from chlorpyrifos), or no pesticide. Exposure to chlorpyrifos in the selection phase resulted in changes in genotype composition and a reduction in genotype diversity compared to control populations. In the second phase, selection background strongly influenced population development. In line with adaptive evolution, chlorpyrifos-selected populations achieved higher maximum and average densities as well as higher population growth rates than control populations when subsequently exposed to chlorpyrifos and malathion, two pesticides with the same mode of action. They also performed better in the absence of a pesticide. When exposed to deltamethrin, however, chlorpyrifos-selected populations showed lower average and maximum population densities than control populations. Our results highlight an eco-evolutionary feedback in which evolution in response to a pesticide worsens the effect of a pesticide shift in a non-target species, making them more vulnerable to common agricultural practice.
Dataset DOI: 10.5061/dryad.bzkh189qx
Files and variables
File: genotype_analysis.R
Description: R script for analyses of genotype frequencies at the end of each experimental phase
File: H1.stan
Description: Stan script for the probabilistic model corresponding to the alternative hypothesis (corresponding to clonal selection), needed for the Bayes factor analysis.
File: H0.stan
Description: Stan script for the probabilistic model corresponding to the null hypothesis (corresponding to no clonal selection / random sampling), needed for the Bayes factor analysis.
File: population_analysis.stan
Description: Stan script for the main population density analysis featuring the hierarchical Gaussian process approach.
File: population_data.txt
Description: Dataset containing population densities of Daphnia magna in the different treatments tracked throughout the experiment
Variables
- Tank_name: Population name
- pre-exposure: treatment received during the selection phase - CTL (control) or CPF (chlorpyrifos)
- exposure: treatment received during the second exposure phase - CTL (control), CPF (chlorpyrifos), MAL (malathion) or DTM (deltamethrin)
- replicate: replicate number
- tank: population identifier
- date: date of data collection
- population_density: population density
File: population_analysis.R
Description: R script for analyses of population densities
File: prior_tune.stan
Description: Stan script to determine the optimal length scale prior used in "population_analysis.stan".
File: genotype_data.csv
Description: dataset containing microsatellite genotyping results for eight microsatellite markers (combined in multiplex M01) on genotyped individuals collected from each experimental population at the end of each experimental phase
Variables
- Sample_code: Population name
- Pre.exposure: treatment received during the selection phase - CTL (control) or CPF (chlorpyrifos)
- Exposure: treatment received during the second exposure phase - CTL (control), CPF (chlorpyrifos), MAL (malathion) or DTM (deltamethrin)
- Replicate: replicate number
- M01_B###_A/B: Identified alleles for the different loci
- Concatenated_alleles: sequence of the different identified alleles for each genotyped individual
- Genotype_ID: genotype identifier. Animals with the same genotype identifier originate from the same clonal line.
Code/software
Population density analyses:
The model used was implemented using the probabilistic programming language Stan and the rstan v.2.21.2 package in R v.4.0.3. We used the tidybayes v.2.3.1 package to visualize posterior distributions.
Genotype frequency analyses:
We performed a Bayesian multinomial test on the observed genotype counts using the ‘multibridge’ v1.2.0 package in R v4.3.1. We implemented the null and alternative model in the probabilistic programming language Stan using the ‘rstan’ package v2.32.3 and computed the Bayes factors using the ‘bridgesampling’ package v1.1-2 in R v.4.3.1.
The two R scripts are fully independent of one another and can be run in any order. Each is self-contained and should be run from top to bottom.
The script "population_analysis.R" performs the analysis of population densities over time. It reads the csv file "population_data.csv" and requires the Stan models "prior_tune.stan" and "population_analysis.stan".
The script "genotype_analysis.R" performs the analysis of genotype frequencies. It reads the csv file "genotype_data.csv" and requires the Stan models "H0.stan" and "H1.stan".
The experiment was divided into two phases: a selection phase, and a second pesticide exposure phase. During the selection phase, populations of Daphnia magna were exposed to either a model insecticide chlorpyrifos, or to a control treatment for three weeks. At the end of this selection phase, a set of animals were collected from each population for genetic analyses. During the second pesticide exposure phase, each of the populations from both treatments of the selection phase (chlorpyrifos-selection or control-selection) were divided into new populations and expose to one of four treatments: chlorpyrifos, malathion, deltamethrin or control. During this phase, population densities were collected twice a week for three weeks. At the end of the second phase, animals were collected and preserved for genetic analyses. The animal samples collected were used to determine genotype frequencies at the end of each experimental phase. For this, we used microsatelite markers structured in one microsatellite multiplex.