Pleiotropy (multiple phenotypic effects of single genes) and epistasis (gene interaction) have key roles in the development of complex phenotypes, especially in polymorphic taxa. The development of discrete and heritable phenotypic polymorphisms often emerges from major-effect genes that interact with other loci and have pleiotropic effects on multiple traits. We quantified gene expression changes during ontogenetic colour development in a polymorphic insect (damselfly: Ischnura elegans), with three heritable female morphs, one being a male mimic. This female colour polymorphism is maintained by male mating harassment and sexual conflict. Using transcriptome sequencing and de novo assembly, we demonstrate that all three morphs downregulate gene expression during early colour development. The morphs become increasingly differentiated during sexual maturation and when developing adult colouration. These different ontogenetic trajectories arise because the male-mimic shows accelerated (heterochronic) development, compared to the other female morphs. Many loci with regulatory functions in reproductive development are differentially regulated in the male-mimic, including upstream and downstream regulators of ecdysone signalling and transcription factors potentially influencing sexual differentiation. Our results suggest that long-term sexual conflict does not only maintain this polymorphism, but has also modulated the evolution of gene expression profiles during colour development of these sympatric female morphs.

This dataset contains data for gene expression analyses, reproductive failure data and maturation time data for the common bluetail damselfly Ischnura elegans. The data for gene expression analyses includes a transcriptome (Trinity.fasta) and gene-level counts (counts.txt). These files stem from RNA-Seq data collected from whole bodies of field-caught individuals and individuals raised in outdoor mesocosms. Raw RNA-Seq reads are available through NCBI’s Sequence Read Archives (BioProject # PRJNA624641). The transcriptome, based on all 83 samples, was assembled in TRINITY after quality control and trimming. Gene-level counts for the 27 samples in the experimental design (Exp_design.csv) were then obtained by mapping the reads on the transcriptome and clustering transcripts based on sequence and expression similarity in CORSET.

After excluding lowly expressed transcripts, transcripts were blasted against the non-redundant database in NCBI. Blast hits were then mapped to Gene Ontology (GO) terms using Blast2GO. The resulting file contains transcript-level annotations (blast2go_mapping_20180521_1147.txt). Gene ontology terms were then extracted for each gene cluster and are also included in this dataset (AnnotationsByCluster.txt).

The gene-level counts (counts.txt), experimental design (Exp_design.csv) and gene-level ontology terms (AnnotationsByCluster.txt) are used in differential expression analyses, vector correlation analyses and GO enrichment analyses presented in this study. All these analyses can be reproduced using the R Markdown script IschnuraColourDevelopment.Rmd. This study also examines in more detail a subset of differentially expressed genes. The GO names and descriptions of this gene subset (DEG_Gos_expanded.tsv) are also included in the dataset. This file is also used in the R Markdown script to identify genes with gene regulatory functions among the subset of interest.

Reproductive failure data (Rep_Failure.csv) come from field-collected individuals that were caught while mating and brought into an indoor laboratory for ovipostion. Eggs were then photographed and counted. Reproductive failure data were collected between 2001 and 2017 in 18 natural populations of I. elegans in Southern Sweden. Finally, colour maturation data (Morph_Maturation.csv) was collected from field-caught females which were raised in outdoor mesocosms in the summers of 2015-2017. Females were marked with wing colour codes and monitored daily to determine the duration of colour transitions. These two data files are also used in the R Markdown script for this study.