Animals frequently defend themselves against predators and parasites using toxins obtained from their diets. Monarch butterflies are a preeminent example of toxin sequestration and gain protection from cardenolides in their milkweed host plants. Although sequestration behavior is well-studied in monarchs, relatively little research has studied genetic variation in sequestration ability. In this study, we use the monarch’s global range expansion to test hypotheses about how cardenolide sequestration has evolved over recent evolutionary history. First, using a reciprocal rearing experiment involving six monarch populations and six associated milkweed host species, we test for whether natural selection has increased cardenolide sequestration in monarch populations reared on their sympatric hosts. Second, we test for whether contemporary species interactions affect sequestration by measuring cardenolides in monarchs from Guam, an oceanic island where bird predators have been functionally absent for approximately 40 years. We find evidence for substantial genetic variation in sequestration ability, although no consistent pattern of enhanced sequestration in sympatric monarch/milkweed combinations. One monarch population (from Puerto Rico) shows strong support for cross-hosts tradeoffs in sequestration ability, with elevated sequestration from two tropical milkweed species (Asclepias curassavica, Gomphocarpus physocarpus) but greatly reduced sequestration from two temperate species (A. syriaca, A. speciosa). Monarchs from Guam show some evidence for reduced cardenolide sequestration in both a cross-island comparison of wild-caught butterflies as well as a population-level comparisons of greenhouse-reared butterflies. Our results suggest that processes involved in toxin sequestration are subject to natural selection and may evolve in response to contemporary changes in species interactions.

This manuscript contains data from two experiments:

1. A common-garden experiment conducted in 2017 and 2018 that involved rearing monarch butterflies on various milkweed species in a greenhouse.
2. Wild-caught butterflies from the Mariana Islands (Guam and Rota) collected in 2015.

The primary data collected is chromatography data representing cardenolides present in milkweed leaf tissue and monarch butterfly wing tissue. For dataset 1, these data were generated by the lead author and processed into interpretable output (.csv files) using proprietary software (Chromeleon). For dataset 2, these data were generated and processed by Dr. Mark Hunter. All analyses were conducted using the resulting .csv files. We also included a number of raw data files (in .txt form) for the purpose of demonstration.

The full repository contains the following files and directories:

1. adult_data.csv : contains information for butterflies reared during experiment 1
2. chromatograms : directory containing .txt files for raw chromatogram readout from experiment 1
3. concatenated_cardenolides.csv : full data matrix of cardenolide peaks across all leaf and wing tissue samples from experiment 1
4. marianas_leaf_cardenolides.csv : data matrix of cardenolide peaks for milkweed leaf tissue collected in the field as part of experiment 2
5. marianas_wing_cardenolides.csv : data matrix of cardenolide peaks for monarch wing tissue collected in the field as part of experiment 2
6. tissue_mass.csv : contains dry mass of tissue used for cardenolide measurement

The scripts directory contains four R scripts that were used for all data analysis and figure/table generation. Those scripts are:

1. cardenolide_chromatograms.R : Script for plotting raw data from cardenolide readout. This script was used to produce Figure 2A and Figure S4.
2. cardenolide_stats.R : Primary analysis script for statistics and most figures/tables.
3. guam_cards.R : Script for processing data from Guam and Rota samples collected in 2015. This script was used to produce Figures 5B and 5C as well Figure S1.
4. maps.R : Script for generating maps used in Figure 1A and Figure 5A 

All analyses were conducted in R version 4.0.3.