- Global dataset and code for inferring fish metabolic rates using phylogenetic structural equation models (PSEMs).
This repository contains the code, data processing pipeline, and analyses associated with the paper Filling metabolism data gap in marine fish reveals dual pace-of-life and reproductive strategy axes. The study assembles a global dataset of 18,129 marine fish species and uses phylogenetic structural equation models (PSEMs) to infer missing metabolic traits, particularly resting metabolic rate (RMR). Results highlight the relationship between metabolism and life-history strategies (pace-of-life syndrome), with implications for predicting species’ responses to climate change.
Marine BENEAT (Marine.Beneat@ifremer.fr),
Alaia MORELL (amorell@ifremer.fr),
Fabien MOULLEC (fabien.moullec@umontpellier.fr),
Nicolas BARRIER (nicolas.barrier@ird.fr),
Yunne SHIN (yunne-jai.shin@ird.fr),
Bruno ERNANDE (Bruno.Ernande@ifremer.fr).
Responsible of data collection, codes development and online upload: Marine BENEAT.
The repository is split into three main directories: The first one is about the dataset construction, the second about data analysis, and the third one is the .csv file with the 18,129 species trait values estimated using the inference process. The folders 01- and 02- contain an Input folder (for 02-: the input data is the output of the folder 01-), a Script folder, and an Output folder.
The dataset construction folder (01-) has two steps. Scripts 01, 02, 03, : collection of the temperature- and mass-specific respiration data. Script 04 : inference process. The inference process is used as input of the data analysis.
The analysis folder (02-) both produces the cross-validation plots for all the SEM tested, and produces the analysis plots for the selected SEM model.
The dataset folder (03-) is the raw output from the inference process using the selected SEM (mechanistic 1 (b)). Some of the trait values are log-transformed according to the method described in the corresponding paper.