Data from: The Red Queen unveils the sexual and mating strategies of flowers
Abstract
The Red Queen Hypothesis states that sex evolved as an adaptation of hosts against their parasites, allowing them to produce offspring protected by rare combinations of defence-genes. It has been hypothesized that plant species under higher pressure by short-lived natural enemies should invest more heavily in outcrossing. Here, we test the interspecific prediction that plant species associated with a higher richness of insect herbivores – used as a proxy for a higher, more diverse and consistent herbivory pressure over geographic space and evolutionary time – should exhibit sexual and mating traits leading to outcrossing. We predict that plant species associated with a richer insect fauna will have a higher probability to be auto-incompatible, dichogamous, heterostylous, dimorphic, dioecious, and allogamous. Furthermore, they are predicted to exhibit a higher pollen-ovule ratio, to be more dependent on pollinators or wind for cross-fertilization, and to rely mostly on sexually-generated seeds for reproduction rather than on asexual means of reproduction.
In order to perform such comparative tests, we assembled published insect-plant interaction records and information on key reproductive floral traits linked to the inbreeding-outcrossing gradient for 1884 native angiosperm species from Germany. The analyses were performed with generalized linear models and phylogenetic models using species richness of insect herbivores as the single explanatory variable, but also with plant geographic range size and plant height as covariates.
We showed that plant species consumed by a higher species richness of insect herbivores have a higher probability to be auto-incompatible, dichogamous, dimorphic, dioecious, and allogamous. In addition, they had a higher pollen-ovule ratio, were more dependent on pollinators or wind for cross-fertilization, and relied mostly on sexually-generated seeds for reproduction rather than also use asexual reproduction. Model fitting improved considerably after taking into consideration the phylogenetic structure of the data and the results remained relatively consistent after controlling for geographic range size and plant height.
Synthesis. Although the conventional wisdom is that floral traits of plants evolved in concert with their mutualistic pollinators, here we showed that several key sexual and mating traits of plants, which modulate their outcrossing strategy, evolved in response to the pressure exerted by their antagonistic insect herbivores. The Red Queen Hypothesis emerges as a unifying theory which helps to explain the sexual diversity of angiosperms, with consequences for genetic and phylogenetic diversity, animal-plant network structures, and plant-based ecosystem services.
https://doi.org/10.5061/dryad.73n5tb3c9
Date: 08 / 04 / 2026
The files deposited here contain public data (CC0), codes, outputs, and the phylogeny used in the manuscript, The Red Queen unveils the sexual and mating strategies of flowers, authored by Carlos Roberto Fonseca, Martin Brandle, and Martin M. Gossner, published in Journal of Ecology (2026).
The material provided here is organized in four folders: data.zip, scripts.zip, outputs.zip, and phylo.zip.
The material should be organized in an R-Project, named Project_RedQueen_JEcol_CRFonsecaetal, with four folders: 1) data with the subfolders raw and processed, 2) scripts, 3) outputs with the subfolders figures and tables, and 4) phylo.
FOLDER DATA.ZIP
The folder data.zip has four CC0 data files:
File name: HerbPlantInteractions_FinalList_2024_07_13
File type: csv
Delimiter: semicolon (;)
Variables:
- Order - Original order classification of the HerPlantInteraction database.
- Family - Original family classification of the HerPlantInteraction database.
- Host - Original species name of the HerPlantInteraction database.
- Count of Herbivores - Number of herbivores associated to each plant species.
- Neobiota - A check variable which indicates whether species are non-native (1) or native (0) to Germany. It should always be 0, since only natives from Germany were selected.
- InRLD - A check variable which indicates if the species is the the German Red List (1) or not (0). All are present.
- OnlyRLD - A check varible which indicates that the species is only in the German Red List.
- Order_Kew - Valid plant order according to Kew Gardens.
- Family_Kew - Valid plant family according to Kew Gardens.
- Host_Kew - Valid plant species according to Kew Gardens.
- Kew - A check variable to verify if the taxonomy changed (kew), according to Kew Gardens classification, or remained unaltered (ok)
Observations: In Count of Herbivores, zeros represent plant species that had no record of associated insects in the literature.
Directory: This file should be saved as: data/raw/HerbPlantInteractions_FinalList_202_07_13.csv.
File name: Data_Pollen_ovule_ratio
File type: csv
Delimiter: semicolon (;)
Variables:
- Species - valid plant species name.
- PO_ratio - mean pollen-ovule ratio per plant species. It is calculated as the arithmetic mean of the values found by different sources (PO_ratio_by_source). This is the value used in all analyses.
- PO_ratio_by_source - pollen-ovule ratio recorded by different authors.
- source - the full reference of each value shown in the column PO_ratio_by_source.
Observation: This file has no missing value.
Directory: This file should be saved as: data/raw/Data_Pollen_ovule_ratio.csv.
File name: Data_OtherSources_Height
File type: csv
Delimiter: semicolon (;)
Variables:
- Species - valid plant species name.
- Height_m - plant height in meters.
- Reference: literature or site from where the data were obtained.
Observation: Missing data is represented as NA.
Directory: This file should be saved as: data/raw/Data_OtherSources_Height.csv.
File name: Data_OtherSources_Heterostyly
File type: csv
Delimiter: semicolon (;)
Variables:
- Genus - plant genus.
- Species - valid plant species name.
- Polimorphism - 0 [Homostylous], 1 [Heterostylous] (binary variable).
- Hercogamy - containing the classes distilus, distylous, dystilous, heterostilic, heterostylous, homostylous, tristilous, tristylous among others.
- Reference: literature or site from where the data were obtained.
- doi: doi identifier.
- Obs - general observations.
Observation: Missing data is represented as NA.
Directory: This file should be saved as: data/raw/Data_OtherSources_Heterostyly.csv.
FOLDER SCRIPTS
The folder scripts contains all R scripts used for data analyses.
The scripts are ordered according to the structure of the paper.
Scripts 00 organizes the insect data-set, generate the phylogeny, analyse the taxonomic distribution of the plants, describe the insect richness distribution and analyse the geographic distribution of species.
Scripts 01 to 10, analyse each one of the ten sexual and mating traits considered in the article.
Scripts with letter a perform GLM analyses shown in Table S1 and produce the partial figures of Fig. S5-S6.
Scripts with letter b perform phylogenetic analyses shown in Table 1 and produce the partial figures of Fig. 2 and Fig. S6.
IMPORTANT: run scripts with letter a before scripts with letter b.
Scripts 11 integrates Fig. 2 and Fig. S5.
Scripts 12 and 13 perform the analyses controlling for geographic range and plant height, respectively.
Script 14 perform phylogenetic analyses controlling for geographic range and height simultaneously.
Script 15 evaluate the phylogenetic signal of all variables. Scripts 11 to 15 should be run only after all scripts from 0 to 10 were run.
IMPORTANT: run Scripts 11-15 after run all scripts from 0 to 10
The following scripts are available:
Script00a_Insect_dataset
Script00b_Phylogeny
Script00c_Plant_taxonomic_distribution
Script00d_Insect_richness_distribution
Script00e_Geo
Script01a_Self-incompatibility
Script01b_Self-incompatibility_phylo
Script02a_Dichogamy
Script02b_Dichogamy_phylo
Script03a_Heterostyly
Script03b_Heterostyly_phylo
Script04a_SexDimorphy
Script04b_SexDimorphy_phylo
Script05a_SexualSystems
Script05b_SexualSystems_phylo
Script06a_BreedingSystem
Script06b_BreedingSystem_phylo
Script07a_PollenOvuleRatio
Script07b_PollenOvuleRatio_phylo
Script08a_InsectDependence
Script08b_InsectDependence_phylo
Script09a_WindDependence
Script09b_WindDependence_phylo
Script10a_ReproductiveStrategy
Script10b_ReproductiveStrategy_phylo
Script11a_FigS5_Integration_GLM
Script11b_Fig02_Integration_phylo
Script12_Geographic_range_control
Script13_Height_control
Script14_Phylo_Height_Geo_control
Script15_Phylogenetic_signal
FOLDER PHYLO
The folder phylo has the file Tree_Insect.tre with the phylogenetic tree of the plant species which was generated by the Script00b_Phylogeny with the help of V.Phylomaker (see Material and Methods).
This phylogeny is used in all phylogenetic analyses.
FOLDER OUTPUT
The folder outputs contains two folders, figures and tables, where outputs are deposited.
OUTPUT/FIGURES
It contains jpg files of the main figures of the article, including Fig.1, the integrative figure Fig.2a-i (phylogenetic analyses), as well as individual figures of all sexual and mating traits.
Fig1_SexDifferentiation.jpg
Fig2_Integration_3_3.jpg
Fig2a_Self_incompatibility_phylo.jpg
Fig2b_Dichogamy_phylo.jpg
Fig2c_Sex_dimorphism_phylo.jpg
Fig2d_Sexual_system_phylo.jpg
Fig2e_Breeding_system_phylo.jpg
Fig2f_Pollen_ovule_ratio_phylo.jpg
Fig2g_Insect_dependence_phylo.jpg
Fig2h_Wind_dependence_phylo.jpg
Fig2i_Reproductive_strategy_phylo.jpg
It also contains JPG files of all figures of the supplementary material:
FigS1_Plant_Order_distribution.jpg
FigS2_Plant_Family_distribution.jpg
FigS3_Plant_Genus_distribution.jpg
FigS4_Histogram_Insects.jpg
FigS5_Integration_3_3.jpg
FigS5a_Self_incompatibility.jpg
FigS5b_Dichogamy.jpg
FigS5c_SexualDimorphism.jpg
FigS5d_Sexual_system.jpg
FigS5e_BreedingSystem.jpg
FigS5f_PollenOvule.jpg
FigS5g_Insect_dependence.jpg
FigS5h_Wind_dependence.jpg
FigS5i_Reproductive_strategy.jpg
FigS6_Heterostyly.jpg
Fig_SI_geo.jpg
Furthermore, it contains a PDF figure of the plant phylogeny used in all phylogenetic analyses and a JPG file of the graphical abstract:
Data_S1_Phylogeny.pdf
GraphicalAbstract.jpg
OUTPUT/TABLES
The folder tables contains the tables (.csv) for the rank distribution of Order, Family and Genus.
Table_Plant_Family_distribution.csv - with the column Family (character), showing the plant families, and N_Family (numeric), showing the number of species in each plant family.
Table_Plant_Genus_distribution.csv - with the column Genus (character), showing the plant genera, and N_genus (numeric), showing the number of species in each plant genus.
Table_Plant_Order_distribution.csv - with the column Order (character), showing the plant orders, and N_Order (numeric), showing the number of species in each plant order.
The structure of the analyses
In Fonseca et al (2026, J. Ecol), the following response variables were considered:
(a) Self-incompatibility (binary; [0] self-compatible, [1] self-incompatible)
(b) Dichogamy (binary; [0] synchronous, [1] asynchronous)
(c) Heterostily (binary; [0] homostylous, [1] heterostylous)
(d) Sexual dimorphism (binary; [0] monomorphic, [1] dimorphic)
(e) Sexual system (ordinal; [0] hermaphrodite, [1] andromonoecy and gynomonoecy, [2] diclinous monoecy, [3] androdioecy, gynodioecy and trioecy, [4] dioecy)
(f) Breeding system (ordinal; [0] autogamy, [1] facultative autogamy, [2] mixed mating, [3] facultative allogamy, [4] allogamy)
(g) Pollen-ovule ratio (continuous; pollen number / ovule number; log10-scale)
(h) Insect dependence (ordinal; [0] rarely [1] often and [2] always insect-pollinated)
(i) Wind dependence (ordinal; [0] rarely, [1] often and [2] always wind-pollinated)
(j) Reproductive strategy (binary; [0] vegetative and seed reproduction, [1] reproduction by seed)
The explanatory variable is:
Richness of insect herbivores (i.e., number of insect species feeding on a given host species), provided here in data/Data_Insects.csv
The covariate variables are:
Geographic range size - estimated by the area of occupancy of the plants obtained from the FloraWeb, expressed as numbers of TK25 quadrats occupied by a plant species across Germany (www.floraweb.de).
Plant height (m) - extracted from TRY and LEDA databases.
ADDITIONAL DATA SOURCES
Most response variables used in our study were collected in online databases which do not provide a CC0 license waiver. However, the data can be obtained under request in the following databases:
BiolFlor, a comprehensive database on biological and ecological traits of the German flora (Kuhn, I., Durka, W., & Klotz, S. (2004). BiolFlor - a new plant-trait database as a tool for plant invasion ecology. Diversity and Distribution, 10, 363-365. http://doi.org/10.1111/j.1366-9516.2004.00106.x).
- Self-incompatibility - containing the original classes: self-incompatibel, self-compatibel, +- self-compatibel, +- self-incompatibel.
Bioflor Data should be saved as: data/raw/Data_Bioflor_Self-incompatibility.csv (comma-delimited). - Pollen vector - containing the classes: cleistogamy, geitonogamy, insects, pseudocleistogamy, selfing, slug, water, wind. Additionally, it provided the column Abundance - containing the classes: always, at failure of outcrossing, never, often, possible, rare, the rule, unknown.
Bioflor Data should be saved as: data/raw/Data_Bioflor_Vector.csv (comma-delimited). - Breeding system - allogamous, autogamous, automixis, facultative allogamous, facultative autogamous, mixed mating.
Bioflor Data should be saved as: data/raw/Data_Bioflor/Breeding_system.csv (comma-delimited). - Dichogamy - with the classes: homogamous, pronounced protandrous, pronounced protogynous, protandrous, protogynous, slightly protandrous, slightly protogynous.
Bioflor Data should be saved as: data/raw/Data_Bioflor_Dichogamy.csv (comma-delimited). - Type of reproduction - with the classes: by seed and vegetatively, by seed/by spore, mostly by seed, rarely vegetatively, mostly vegetatively, rarely by seed, vegetatively.
Bioflor Data should be saved as: data/raw/Data_Bioflor_Type_of_reproduction.csv (comma-delimited).
TRY, a global database of curated plant traits (Kattge, J., et al. (2020). TRY plant trait database - enhanced coverage and open access. Global Change Biology, 26, 119-188. https://www.try-db.org/TryWeb/Home.php.
- Plant height vegetative (Trait 3106) - Height in meters.
- Flower number of pollen per ovule (Trait 210) - number of pollen divided by the number of ovules in a plant population.
- Flower sex (Trait 213) - containing the classes: Apomiticism, Hermaphroditism, Andromonoecium, Gynomonoecium, Monoecious, Androdioecism, Gynodioecism, Trioecism, Dioecism, among others. From this variable, the variable sexual dimorphism can be derived.
- Flower heterostily (Trait 209) - flowers with styles and filaments with different length occurring in different plant individuals.
The TRY Data, containing all variables, should be saved as: data/raw/TryData.txt (comma-delimited).
SIMON-PORCAR (2024)
- For heterostily, the main source is Simon-Porcar et al. (2024). Convergent evolutionary patterns of heterostyly across angiosperms support the pollination-precision hypothesis. Nature Communications 15: 1237. https://doi.org/10.1038/s41467-024-45118-0.
The Simon-Porcar data, available in the supplementary material should be saved in: data/raw/Data_PolimorphicGenera_Simon_Porcar.csv (comma-delimited).
LEDA, a database of life-history traits of the Northwest European flora (Kleyer, M. et al. (2008). The LEDA Traitbase: a database of life-history traits of the Northwest European flora. Journal of Ecology, 96, 1266-1274. https://doi.org/10.1111/j.1365-2745.2008.01430.x).
- Canopy height - with the variable maximum CH (m) - maximum canopy height in meters.
Leda Data should be saved as: data/raw/Data_Leda_Canopy_height (comma-delimited).
- Distribution - geographic range size estimated by the area of occupancy of the plants obtained from the FloraWeb, expressed as numbers of TK25 quadrats occupied by a plant species across Germany (www.floraweb.de).
Floraweb data, containing only the variables Species_Name (character) and Distribution (numeric) should be saved as: data/processed/Data.Geo.csv (comma-delimited).
The Project_RedQueen was developed in:
R version 4.4.0 (2024-04-24 ucrt) -- "Puppy Cup"
Copyright (C) 2024 The R Foundation for Statistical Computing
Platform: x86_64-w64-mingw32/x64
METHODS
Our dataset focused on flowering plants from Germany, where insect-plant interaction associations have been recorded for centuries with the contribution of naturalists, botanists and entomologists and this knowledge has been organized as large publications and online datasets. To test the Red Queen predictions, we compiled species-level data on ten key reproductive traits that can modulate the inbreeding/outcrossing level of plants as well as information on plant-insect herbivore interactions to estimate the richness of insect herbivores associated with each plant species.
Ten response variables were considered: (a) Self-incompatibility (binary; [0] self-compatible, [1] self-incompatible), (b) Dichogamy (binary; [0] synchronous, [1] asynchronous), (c) Heterostily (binary; [0] homostylous, [1] heterostylous), (d) Sexual dimorphism (binary; [0] monomorphic, [1] dimorphic), (e) Sexual system (ordinal; [0] hermaphrodite, [1] andromonoecy and gynomonoecy, [2] diclinous monoecy, [3] androdioecy, gynodioecy and trioecy, [4] dioecy), (f) Breeding system (ordinal; [0] autogamy, [1] facultative autogamy, [2] mixed mating, [3] facultative allogamy, [4] allogamy), (g) Pollen-ovule ratio (continuous; pollen number / ovule number; log10-scale), (h) Insect dependence (ordinal; [0] rarely [1] often and [2] always insect-pollinated), (i) Wind dependence (ordinal; [0] rarely, [1] often and [2] always wind-pollinated), (j) Reproductive strategy (binary; [0] vegetative and seed reproduction, [1] reproduction by seed). Notice that all variables were ordered from a higher probability of selfing to a higher probability of outcrossing. The plant traits were mainly extracted from the databases BiolFlor, a comprehensive database on biological and ecological traits of the German flora (Kühn et al., 2004. http://doi.org/10.1111/j.1366-9516.2004.00106.x), LEDA, a database of life-history traits of the Northwest European flora (Kleyer et al., 2008. https://doi.org/10.1111/j.1365-2745.2008.01430.x), and TRY, a global database of curated plant traits (https://www.try- db.org/TryWeb/Home.php). For pollen-ovule ratio, we complemented the data by performing literature search. In some analyses, we used geographic range size and plant height as covariates. Geographic range size was estimated by the area of occupancy of the plants obtained from the FloraWeb, expressed as numbers of TK25 quadrats occupied by a plant species across Germany (www.floraweb.de). Plant height (m) was extracted from TRY and LEDA data bases. A full description of the methods related to the plant traits can be found in the original article.
In order to build the insect-plant association dataset, we departed from the plant species list contained in the taxonomic reference database for Germany (Jansen & Dengler, 2008. urn:nbnhebis:30:3-449555). The species richness of insect herbivores associated with each plant was obtained by searching literature and databases. The initial species list of insects for the compilation was obtained in the Fauna Europaea web service (https://fauna-eu.org/), considering both phytophagous and xylophagous records, except for Coleoptera that was obtained from the Palaearctic Catalogue (https://weevil.myspecies.info/content/palaearctic-catalogue). Records of interactions were found on monographs, papers and online resources for host-plant records. Insect records for a given host were accepted if they consumed any plant part as a larva or in the adult phase, such as leaves, stems, branches, trunk, roots, flowers and seeds. Valid records included only plants identified until the species level. Infra-species plant categories were amalgamated and hybrids were excluded. Plant names were standardized according to the World Flora Online (2024. http:// www.worldfloraonline.org). The complete list of literature sources used in this compilation is found in Appendix S1 of the original article.
The material deposit here includes an R project with the data and R scripts for cleaning up and organize the data and perform all analyses.
