Skip to main content
Dryad logo

Comparative analysis of phenotypic plasticity sheds light on the evolution and molecular underpinnings of locust phase polyphenism

Citation

Song, Hojun; Foquet, Bert; Castellanos, Adrian (2021), Comparative analysis of phenotypic plasticity sheds light on the evolution and molecular underpinnings of locust phase polyphenism, Dryad, Dataset, https://doi.org/10.5061/dryad.dz08kprwz

Abstract

Locusts exhibit one of nature’s most spectacular examples of complex phenotypic plasticity, in which changes in density cause solitary and cryptic individuals to transform into gregarious and conspicuous locusts forming large migrating swarms. We investigated how these coordinated alternative phenotypes might have evolved by studying the Central American locust and three closely related non-swarming grasshoppers in a comparative framework. By experimentally isolating and crowding during nymphal development, we induced density-dependent phenotypic plasticity and quantified the resulting behavioural, morphological, and molecular reaction norms. All four species exhibited clear plasticity, but the individual reaction norms varied among species and showed different magnitudes. Transcriptomic responses were species-specific, but density-responsive genes were functionally similar across species. There were modules of co-expressed genes that were highly correlated with plastic reaction norms, revealing a potential molecular basis of density-dependent phenotypic plasticity. These findings collectively highlight the importance of studying multiple reaction norms from a comparative perspective.

Methods

Details about the data collection and processing are found in the Supplementary Information file.

Usage Notes

Files can be found here: https://doi.org/10.5061/dryad.dz08kprwz

Supplementary Data File 1: Differentially expressed genes for the head tissue of piceifrons, and their annotation.

This table lists all differentially expressed genes for the head tissue of piceifrons, discovered by either DEseq2 or edgeR, and their annotation by Blast2GO. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were given for each contig. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 2: Differentially expressed genes for the thorax tissue of piceifrons, and their annotation.

This table lists all differentially expressed genes for the thorax tissue of piceifrons, discovered by either DEseq2 or edgeR, and their annotation by Blast2GO. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were given for each contig. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 3: Differentially expressed genes for the head tissue of americana, and their annotation.

This table lists all differentially expressed genes for the head tissue of americana, discovered by either DEseq2 or edgeR, and their annotation by Blast2GO. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were given for each contig. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 4: Differentially expressed genes for the thorax tissue of americana, and their annotation.

This table lists all differentially expressed genes for the thorax tissue of americana, discovered by either DEseq2 or edgeR, and their annotation by Blast2GO. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were given for each contig. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 5: Differentially expressed genes for the head tissue of cubense, and their annotation.

This table lists all differentially expressed genes for the head tissue of cubense, discovered by either DEseq2 or edgeR, and their annotation by Blast2GO. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were given for each contig. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 6: Differentially expressed genes for the thorax tissue of cubense, and their annotation.

This table lists all differentially expressed genes for the thorax tissue of cubense, discovered by either DEseq2 or edgeR, and their annotation by Blast2GO. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were given for each contig. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 7: Differentially expressed genes for the head tissue of nitens, and their annotation.

This table lists all differentially expressed genes for the head tissue of nitens, discovered by either DEseq2 or edgeR, and their annotation by Blast2GO. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were given for each contig. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 8: Differentially expressed genes for the thorax tissue of nitens, and their annotation.

This table lists all differentially expressed genes for the thorax tissue of nitens, discovered by either DEseq2 or edgeR, and their annotation by Blast2GO. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were given for each contig. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 9: Differentially expressed genes for all head tissues using the piceifrons transcriptome as reference, and their annotation.

This table lists differentially expressed genes for the head tissues of piceifrons, americana, cubense and nitens, using the piceifrons transcriptome as reference. After aligning all filtered reads for the different species to the piceifrons head transcriptome, differentially expressed genes were calculated with both DEseq2 and EdgeR for each species separately. Any gene that is differentially expressed in at least one species, was included in the table. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were listed. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 10: Differentially expressed genes for all thorax tissues using the piceifrons transcriptome as reference, and their annotation.

This table lists differentially expressed genes for the thorax tissues of piceifrons, americana, cubense and nitens, using the piceifrons transcriptome as reference. After aligning all filtered reads for the different species to the piceifrons thorax transcriptome, differentially expressed genes were calculated with both DEseq2 and EdgeR for each species separately. Any gene that is differentially expressed in at least one species, was included in the table. For both DEseq2 and edgeR, the Log2FoldChange (Log2FC) and adjusted p value (padj) were listed. Additionally, gene annotation was obtained with Blast2GO, and the following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 11: Enrichment of Gene Ontology terms of differentially expressed genes, for each species and tissue combination.

Fisher exact tests were performed in Blast2GO by comparing the Gene Ontology terms in the set of differentially expressed genes to the ones in the whole transcriptome. Results obtained for the four different species and two different tissues were all pooled together in a single table. The first three columns contain the full name (GO name), ID (GO ID) and category (GO category) of each GO term. For each species and tissue combination, the table contains the False Discovery Rate associated with a certain GO term (FDR), the amount present in the differentially expressed genes (# in DE genes) and in the whole transcriptome (# overall), and last the percentage of genes that were annotated with that specific annotation term within the set of all DE genes (% in DE genes) or within the full transcriptome (% overall).

Supplementary Data File 12: WGCNA module membership for each gene in the head tissue.

Modules of co-expressed genes for the head tissue were obtained with WGCNA. For each contig, the module for which it obtained the highest membership value was listed in the column ‘Module’. As an estimate of the importance of the particular contig in explaining the variation of phenotypic traits, a gene significance (GS) value and an associated p value was calculated for each tested behavioural trait. Additionally, the module membership (MM) of each gene was calculated for each module, in addition to the associated p value. Finally, the table contains the gene annotation of Blast2GO for each gene. The following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 13: WGCNA module membership for each gene in the thorax tissue.

Modules of co-expressed genes for the thorax tissue were obtained with WGCNA. For each contig, the module for which it obtained the highest membership value was listed in the column ‘Module’. As an estimate of the importance of the particular contig in explaining the variation of phenotypic traits, a gene significance (GS) value and an associated p value was calculated for each tested behavioural trait. Additionally, the module membership (MM) of each gene was calculated for each module, in addition to the associated p value. Finally, the table contains the gene annotation of Blast2GO for each gene. The following outputs were given: the gene description, based on BLAST hits (Gene description), contig length (Length), the Expect value associated with the best BLAST hit (E value), the mean similarity at the nucleotide level to its BLAST hits (Similarity), the Gene Ontology IDs (GO IDs) and Gene Ontology names (GO names), and finally the annotation confidence level, for which ‘interpro’ refers to sequences with a hit to the interpro database, ‘blasted’ refers to sequences with at least one BLAST hit to the arthropod nr database, ‘mapped’ was given to sequences for which gene ontology information could be retrieved from its BLAST hit(s), and ‘annotated’ was only given to sequences for which the mapped GO terms also scored above the default thresholds for annotation.

Supplementary Data File 14: Archive (zip) file containing raw data from behavioural assays for all four species of Schistocerca (Behaviour_dataset.csv)

Supplementary Data File 15: Archive (zip) file containing raw colour data (jpg images and coordinates) for all four species of Schistocerca for running the program patternize (folder: Raw_colour_data_and_coordinates)

Supplementary Data File 16: Archive (zip) file containing raw data from morphological measurements for all four species of Schistocerca (Morphology_dataset.csv)

Supplementary Data File 17: Archive (zip) file containing: (1) Text file describing R scripts for running edgeR and DESeq2 using SARTools (Code_DE_analyses.txt) [raw reads used for these analyses are included in Supplementary Data File 18]

Supplementary Data File 18: Archive (zip) file containing: (a) Raw data (raw count data and phenotypic trait data) for all four species of Schistocerca for running the program WGCNA (folder: Raw_data_for_DE_WGCNA)

Supplementary Data File 19: Archive (zip) file containing: (a) FASTA file containing the alignment used for building the phylogeny of hexamerins (hexamerin.fasta); (b) Phylip file containing the alignment used for building the phylogeny of hexamerins (hexamerin.phy); (c) Newick file containing the maximum likelihood tree from the phylogenetic analysis of hexamerins (hexamerin.tre)

Supplementary Software File 1: Archive (zip) file containing RMarkdown file used for analysing behavioural reaction norm data (Rmarkdown_behaviour.Rmd)

Supplementary Software File 2: Archive (zip) file containing Html file describing codes used for analysing behavioural reaction norm data (Rmarkdown_behaviour.html)

Supplementary Software File 3: Archive (zip) file containing RMarkdown file used for running patternize to analyse colour reaction norm data (Rmarkdown_colour.Rmd)

Supplementary Software File 4: Archive (zip) file containing Html file describing codes used for running patternize to analyse colour reaction norm data (Rmarkdown_colour.html)

Supplementary Software File 5: Archive (zip) file containing RMarkdown file used for analysing morphological reaction norm data (Rmarkdown_morphology.Rmd)

Supplementary Software File 6: Archive (zip) file containing Html file describing codes used for analysing morphological reaction norm data (Rmarkdown_morphology.html)

Supplementary Software File 7: Archive (zip) file containing RMarkdown file used for running WGCNA (Rmarkdown_WGCNA.Rmd)

Supplementary Software File 8: Archive (zip) file containing Html file describing codes used for running WGCNA (Rmarkdown_WGCNA.html)

Funding

National Science Foundation, Award: IOS-1253493

TAMU-CONACYT, Award: 2015-027(S), 2018-012-2

U.S. Department of Agriculture, Award: Hatch Grant TEX0-1-6584

Texas A&M University / Consejo Nacional de Ciencia y Tecnología, Award: 2015-027(S)