Coding data
For the 1,716 records included, we recorded information on (1) species; (2) experiments; and (3) references. Descriptions of each category are described below. Each line coded was for a single experiment—a single record could have multiple experiments and thus multiple rows. In total, we had data for 3,315 experiments.
Species
We included phylum, class and species for each experiment. We also recorded developmental mode: planktotrophic (i.e. planktonic, feeding larvae); lecithotrophic (i.e. planktonic, non-feeding larvae); or direct development (i.e. aplanktonic, crawl-away juveniles). We used the package ‘taxize’ in R (v. 4.1.2) to search the Global Biodiversity Information Facility (GBIF) database to identify species names in the dataset that were synonyms—we then referred to each species using just one name (Chamberlain & Szocs 2013).
Experiments
For each experiment, we recorded what traits were measured (Table S1) for each life-history stage included: F0 adult, embryo, larva, metamorphosis, juvenile, and F1 adult (Table S2; Figure 2a). We also classified each experiment into one of three experimental designs that depended on the number of stages included and the unit of replication. The three experimental designs include: a single stage of the life history (‘single stage’); multiple stages using a cohort of individuals (‘cohort longitudinal’); and multiple stages using individuals (‘individual longitudinal’). We also recorded whether the experiment was conducted in the field or laboratory. Studies conducted partially in the laboratory and field were classified overall as field studies because the characterisation of phenotypes almost inevitably required some laboratory work.
References
The reference for each experiment is included.