Hermaphrodites have lower metabolic rates than gonochores
Data files
Oct 28, 2025 version files 6.27 MB
-
JarvisMarshall_mr_herm_data_code.zip
6.26 MB
-
README.md
4.15 KB
Abstract
Hermaphroditism, where an individual can reproduce as both male and female, offers some clear reproductive advantages. Simultaneous hermaphroditism guarantees that every mature adult can mate with another – a particular advantage when opportunities to mate are scarce. Despite this potential benefit, hermaphroditism is relatively rare in animals. This paradox has long involved an energetics argument: hermaphrodites require more energy to fuel two reproductive roles instead of only one, which favours the evolution of separate sexes. However, this argument has never been tested. Here, we compare resting metabolic rates between hermaphrodites and gonochores across 536 species of marine invertebrates, spanning 11 phyla. Our analyses, which control for body size, environmental temperature, motility, and phylogeny, contradict predictions from classic theory: instead of requiring more energy than gonochores, hermaphrodites require ~27% less energy on average. These findings overturn a 150-year-old argument that hermaphroditism is rarer in animals because it is more costly, and highlight the need to reconsider the role of energetics in the evolution of sexual systems.
Description: Explanation of data files and instructions for analyses
from 'Hermaphrodites have lower metabolic rates than gonochores’
Author: George C. Jarvis
Date: Friday 17 October 2025
- General note: These data were set up for analyses in R.
Specifically, files were organized so that readers could download
the zip folder (‘JarvisMarshall_mr_herm_data_code.zip’), unzip it, open the R
project folder (‘mr_herm.Rproj’), and reproduce all of the analyses
and figures from the project folder (‘JarvisMarshall_mr_herm_data_code’) as
their working directory. The code should run fine on both Mac and PC
computers (I originally ran this on a PC), but feel free to email me
at gjarvis [at] uchicago.edu if you have any questions or issues running
the script.
-
Notes on subfolders within unzipped folder ‘JarvisMarshall_mr_herm_data_code’:
2a) ‘data_and_tree’ – contains the files for raw data
(mr_herm_data_v10.xlsx) and phylogenetic tree
(herm_tree_congruified_536_spp.nwk [Newick file]) for analyses and
figures.2b) ‘functions’ – contains .R scripts for functions used in the
analyses.2c) ‘scripts’ – contains the R notebook file to analyze the data and
generate the figures in the MS ('herm_code_v4.Rmd').2d) 'figures' - folder where figures from will be exported to. Note
that for some figures, I uploaded them into Powerpoint, and edited
them there.
-
Notes on files within ‘data_and_tree’ subfolder:
3a) Raw data are located in file 'mr_herm_data_v10.xlsx' within the
sheet 'raw_data'. "NA's" represent missing data (e.g., in cases
where there is no resolved taxonomic level for 'order' for a
species) or cases that are not applicable (e.g., the type of
hermaphroditism for species that have separate sexes:
'herm_type_sim_seq' = 'NA'). Explanations for column headings in the
raw data sheet, and the references for raw data are located in the
'metadata' and 'reference list' sheets, respectively. NOTE: I used
the "raw_data" sheet in this .xlsx file to run all analyses and
generate all figures for the MS.3b) I've also included .csv files for each of the 3 sheets in the
.xlsx file mentioned above: 'mr_herm_raw_data_v3.csv' (raw
phenotypic data); 'mr_herm_phylum_comparison_v3.csv' (data comparing
species distributions between our dataset and that of the World
Register of Marine Species [WoRMS]); 'mr_herm_metadata_v3.csv'
(metadata, describing column headings in the
'mr_herm_raw_data_v3.csv' and 'mr_herm_phylum_comparison_v3.csv'
files); 'mr_herm_raw_data_v3.csv' (raw phenotypic data) and
'mr_herm_data_references_v3.csv'. These files are in the folder
'xlsx file as separate csv files', within the 'data_and_tree'
folder.3b) The phylogenetic tree ('herm_tree_congruified_536_spp.nwk') has
already been time-calibrated with the 'congruify.phylo' function in
the package 'geiger' v. 2.0.11. For more information on this
process, see: Eastman JM, Harmon LJ, Tank DC. 2013 Congruification:
Support for time scaling large phylogenetic trees. Methods Ecol.
Evol. 4, 688–691. (doi:10.1111/2041-210X.12051)
-
Steps to reproduce analyses and figures:
4a) Open the R project file (‘mr_herm.Rproj’) project file within
the unzipped ‘JarvisMarshall_mr_herm_data_code’ folder.4b) Within the R project, open the .Rmd file, ‘herm_code_v4.Rmd’. The
code should run fine from there.4c) I included the output figures in the 'figures'
folder.
-
General notes on R notebook file:
5a) The names of the files are as they appear in the .Rmd file
('herm_code_v4.Rmd'). In other words, if you change the name of a
file (e.g., raw data, function) you will need to correspondingly
change the names of the file as they are called in the .Rmd file.5b) There are some packages that I had to download manually in the R
console (e.g., package 'ggtreeExtra'). In those cases, I left
instructions for how to download the packages in the .Rmd file.