https://doi.org/10.5061/dryad.1g1jwsv2t

All data sets were collected from a genus of parasitoid wasps (Lysiphlebus fabraum species group) that chemically mimic their aphid hosts to avoid ant aggression. We chemically characterized wasps reared on different aphids and then assessed in two behavioral experiments how aggressive aphid-tending ants reacted to these wasps and to positive control wasps (Aphidius colemani) that don't mimic their aphid hosts.

The first dataset (cross_rearing_data) resulted from an experiment where we reared four wasp lines on three different host aphids. We extracted and analyzed (using gas chromatography-mass spectrometry) the cuticular hydrocarbons of wasp and aphid samples and presented the concentrations of all detected compounds in this table.

The second dataset (ant_experiment_acquired_CHC) stems from a behavioral experiment where we exposed wasps from one genetic line (collected on Aphis hederae in the field) reared on three different host aphid species to ants that were tending one of the different host aphid species. The aim was to assess if CHC components that were acquired from the host aphid influence the ant attack behavior towards the wasps. The data shows how aggressive ants reacted to wasps in different aphid backgrounds. We recorded attack events and antennation times (that are classified as friendly behavior) using NOLDUS observer software.

The third dataset (ant_experiment_genetic_CHC) was generated in a similar experiment as the previous. Here, we presented wasps of four different genetic lines that developed on the same host species to ants that were all tending the same aphid background. We again recorded how aggressive ants reacted to assess if genetic components of the wasp cuticular hydrocarbons affect ant aggression towards the wasps.

Description of the data and file structure

The chemical data (cross_rearing_data) is structured in the following way:

• first column: sample identifier
• second column: type of insect
• third column: aphid species (or host aphid species that a wasp sample was reared on)
• fourth column: wasp line: identifiers for the four different wasp lines (CV17-84, IL07-64, IL09-348, IL09-554 (order corresponds to wasp lines 1-4 in the published manuscript))
• fifth column: wasp origin: species on which a wasp line was collected in the field
• sixth column: wasp group: short identifier that combines wasp line and the species a wasp was reared on (used for plotting the data); block: the experiment was conducted in two different blocks
• all other columns: chemical compounds from shortest to longest retention time (values are concentrations: ng/ul)

The data from the first behavioral experiment (ant_experiment_acquired_CHC) is structured in the following way:

• first column:sample identifier
• second column: ant colony in which sample was introduced (1-4)
• third column: a treatment identifier which is a combination of the next two columns
• fourth column: aphid background in which experiment was conducted (Afc: Aphis fabae cirsiiacanthoidis, Ag: Aphis gossypii, Au: Aphis urticata)
• fifth column: host aphid in which wasp has developed or A_col (Aphidius colemani, which was our positive control).

All following columns represent what we counted and measured during the experiment. Each sample was observed for max 10 minutes or until the wasp sample was removed by the ants.

• sixth column: total time (seconds) that ants antennated a wasp sample
• seventh column: the number of antennation events
• eighth column: the number of any type of ant contact
• ninth column: whether a wasp got attacked by ants or not (binary response)
• tenth column: the day of the experiment (1-12)
• eleventh column: the duration of the observation (max. 600s if no attack occurred)
• twelfth and thirteenth columns: start and stop time of the observation.

The data from the second behavioral experiment (ant_experiment_genetic_CHC) is structured in a similar way as the previous table.

• first column: sample identifier; second column: experiment (1 or 2, since this experiment was replicated)
• third column: wasp sample identifier within experiment
• fourth column: wasp line: four different Lysiphlebus lines (CV17-84, IL07-64, IL09-348, IL09-554) and A.colemani control wasps)
• fifth column: aphid background in which sample was presented to ants (ag: Aphis gossypii, afc: Aphis fabae cirsiiacanthoidis, au: Aphis urticata)
• sixth column; day of the experiment (.1 if sample stems from first experiment and .2 if it stems from repetition)
• seventh column: ant colony in which sample was presented (1-5).

All the following columns are measurements and counts from the experiment.

• eighth column: duration of observation (max. 600 seconds)
• ninth and tenth columns: start and stop time of each observation
• eleventh column: mean duration of antennation events
• twelfth column: total duration of all antennation events together
• thirteenth column: number of antennation events
• fourteenth column: if wasp got attacked by ants or not (binary, 1=yes or 0=no)

Sharing/Access information

All data was generated in laboratory experiments in our research facilities and are only publicly available on Dryad.

Code/Software

The code included in this submission was generated using the statistic software R (version 4.0.3) and was written to analyze and visualize the three data sets for the publication of the manuscript of the same name. All packages necessary to run the code are listed within the code. The first part of the code analyzes the chemical data set and the second part the results from the two different behavioral experiments. The code is annotated and it should be clear which analysis the different parts execute.
An additional file contains the license (GNU GENERAL PUBLIC LICENSE) under which the code is published.