Short summary of the study

Controlling for phylogeny and adult body mass, we identified the trade-off between neonate mass and litter size across the family Mustelidae. Using models of trait evolution on the phylogeny, we revealed lineage-specific directional selection in sea otters favoring larger neonate mass and smaller litter size relative to adult mass, which deviates from the evolutionary patterns observed for each trait across the Mustelidae. Furthermore, we revealed lineage-specific directional selection in sea otters favoring larger total litter mass relative to adult mass, as well as larger neonate mass even after excluding the effect of the trade-off against litter size. Lineage-specific directional selection in sea otters favoring larger neonate mass and litter mass cannot be solely attributed to the prolonged gestation period.
Our findings suggest that sea otters have undergone distinctive evolutionary changes in the trade-offs among life history traits.

Data description

The phylogenetic tree, data and code files used to conduct the analyses in the article with the above-mentioned title are stored in this repository.

Phylogenetic tree files

The phylogenetic trees in Newick format are provided in the following files. These files were used when running the codes described in the file entitled "PGLS_R_code_Sea_otter_lifehistory_evolution.txt" in R version 4.2.2.

(1) phylo1_Sea_otter_lifehistory_evolution.txt
This file contains the primary tree of the mustelid species used in this study, as shown in Figure 1 in the article.

(2) phylo2_Sea_otter_lifehistory_evolution.txt
This file contains the alternative tree of the mustelid species used in this study, as shown in Supplementary Figure S1 in the article.

Data and code files used to conduct phylogenetic generalized least squares (PGLS) analyses

(3) Dataset_Sea_otter_lifehistory_evolution.csv
This file contains the dataset of the traits used when running the codes described in the file entitled "PGLS_R_code_Sea_otter_lifehistory_evolution.txt" in R version 4.2.2. The variable descriptions are as follows:

• Species: species scientific names, corresponding to those in the phylogenetic trees given in the file entitled "Phylogeny_phylo1.txt" and "Phylogeny_phylo2.txt"
• Adult_mass: Adult body mass (g)
• Neonate_mass: Neonate body mass (g)
• Litter_size: Litter size
• Litter_mass: Litter mass, which was calculated as neonate mass multiplied by litter size
• Gestation_period: Gestation period (days)

(4) PGLS_R_code_Sea_otter_lifehistory_evolution
This file contains the codes to run the following PGLS models (A-N) in R version 4.2.2.

(A) PGLS model used to examine the relationship between adult mass and neonate mass using the primary tree. The results are shown in Table 1A in the article.

(B) PGLS model used to examine the relationship between adult mass and litter size using the primary tree. The results are shown in Table 1B in the article.

(C) PGLS model used to examine the relationship between adult mass and litter mass using the primary tree. The results are shown in Table 1C in the article.

(D) PGLS model used to examine the relationship between adult mass and gestation period using the primary tree. The results are shown in Table 1D in the article.

(E) PGLS model used to examine the trade-off relationships between neonate mass and litter size after controlling for adult body mass using the primary tree. The results are shown in Table 2A in the article.

(F) PGLS model used to examine the relationship between neonate mass and gestation period while controlling for adult body mass using the primary tree. The results are shown in Table 2B in the article.

(G) PGLS model used to examine the relationship between litter mass and gestation period while controlling for adult body mass using the primary tree. The results are shown in Table 2C in the article.

(H) PGLS model used to examine the relationship between adult mass and neonate mass using the alternative tree. The results are shown in Supplementary Table S2A in the article.

(I) PGLS model used to examine the relationship between adult mass and litter size using the alternative tree. The results are shown in Supplementary Table S2B in the article.

(J) PGLS model used to examine the relationship between adult mass and litter mass using the alternative tree. The results are shown in Supplementary Table S2C in the article.

(K) PGLS model used to examine the relationship between adult mass and gestation period using the alternative tree. The results are shown in Supplementary Table S2D in the article.

(L) PGLS model used to examine the trade-off relationships between neonate mass and litter size after controlling for adult body mass using the alternative tree. The results are shown in Supplementary Table S3A in the article.

(M) PGLS model used to examine the relationship between neonate mass and gestation period while controlling for adult body mass using the alternative tree. The results are shown in Supplementary Table S3B in the article.

(N) PGLS model used to examine the relationship between litter mass and gestation period while controlling for adult body mass using the alternative tree. The results are shown in Supplementary Table S3C in the article.

• Packages and software

R version 4.2.2
R Development Core Team. 2022. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. https://www.R-project.org.

ape package in R
Paradis E, Schliep K. 2018 ape 5.0: an environment for modern phylogenetics and evolutionary analyses in R. Bioinformatics 35, 526–528.

nlme package in R
Pinheiro J, Bates D, DebRoy S, Sarkar D, R Core Team. 2019 nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1–141.

Code files used to conduct an approach that combined simulation-based likelihood and approximate Bayesian computation (ABC)

The following files contain codes to run Markov chain Monte Carlo (MCMC) computations using the approximate likelihood based on the simulation of trait evolution. The codes were written in the C language, and the computations were performed using a MacPro (OS 10.6.7, 2 × 2.93 GHz Quad-Core Intel Xeon).
The parameter descriptions displayed when these codes are run are as follows:

• MRCA: the trait value for the most recent common ancestor of all species, denoted as θ0 in the article.
• drsl_OTT: the effect of lineage-specific directional selection in the branches of Lutrinae except for the terminal branch leading to the sea otter, denoted as drslOTT in the article.
• drsl_SEA: the effect of lineage-specific directional selection in the terminal branch leading to the sea otter, denoted as drslSEA in the article.
• mu_0: the background evolutionary rate in a Brownian motion (BM)-like model across the phylogeny, denoted as μ0 in the article.

The parameters of the evolutionary models described in the article were estimated from the posterior distributions of the parameters obtained by running five independent MCMC chains.

(5) Neonate_mass_phylo1_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of neonate mass relative to adult mass using the primary tree. Parameter estimation results are summarized in Table 3A in the article.

(6) Litter_size_phylo1_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of litter size relative to adult mass using the primary tree. Parameter estimation results are summarized in Table 3B in the article.

(7) Litter_mass_phylo1_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of litter mass relative to adult mass using the primary tree. Parameter estimation results are summarized in Table 3C in the article.

(8) Gestation_period_phylo1_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of gestation period relative to adult mass using the primary tree. Parameter estimation results are summarized in Table 3D in the article.

(9) Neonate_mass_corrected_for_litter_size_phylo1_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of neonate mass corrected for adult mass and litter size using the primary tree. Parameter estimation results are summarized in Table 3E in the article.

(10) Neonate_mass_corrected_for_gestation_phylo1_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of neonate mass corrected for adult mass and gestation period using the primary tree. Parameter estimation results are summarized in Table 3F in the article.

(11) Litter_mass_corrected_for_gestation_phylo1_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of neonate mass corrected for adult mass and gestation period using the primary tree. Parameter estimation results are summarized in Table 3G in the article.

(12) Neonate_mass_phylo2_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of neonate mass relative to adult mass using the alternative tree. Parameter estimation results are summarized in Supplementary Table S4A in the article.

(13) Litter_size_phylo2_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of litter size relative to adult mass using the alternative tree. Parameter estimation results are summarized in Supplementary Table S4B in the article.

(14) Litter_mass_phylo2_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of litter mass relative to adult mass using the alternative tree. Parameter estimation results are summarized in Supplementary Table S4C in the article.

(15) Gestation_period_phylo2_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of gestation period relative to adult mass using the alternative tree. Parameter estimation results are summarized in Supplementary Table S4D in the article.

(16) Neonate_mass_corrected_for_litter_size_phylo2_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of neonate mass corrected for adult mass and litter size using the alternative tree. Parameter estimation results are summarized in Supplementary Table S4E in the article.

(17) Neonate_mass_corrected_for_gestation_phylo2_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of neonate mass corrected for adult mass and gestation period using the alternative tree. Parameter estimation results are summarized in Supplementary Table S4F in the article.

(18) Litter_mass_corrected_for_gestation_phylo2_MCMC_code.c
This file contains the code used to estimate the parameters in the analysis of lineage-specific directional selection in the evolution of neonate mass corrected for adult mass and gestation period using the alternative tree. Parameter estimation results are summarized in Supplementary Table S4G in the article.