The study measures age-specific reproduction, lifetime reproductive success (LRS),
rate-sensitive fitness, and survival in the nematode Caenorhabditis elegans
across three generations (P0, F1 and F3) following early-life larval (L1) starvation.
Two experimental factors are crossed:

• Parental treatment – whether the parents experienced larval (L1) starvation
  (Larval Starvation / L1) or were reared under standard food conditions
  (Control / Con).
• Offspring treatment – the same starvation vs. control manipulation applied to
  the focal (offspring) generation.

All analyses were performed in R. The repository provides a complete,
reproducible workflow from the raw CSV files to the figures and model-output tables
reported in the paper.

---

Description of the data and file structure

The deposit contains two compressed folders and an R script:

data/       # Raw data as CSV files (input to the analyses)
scripts/    # R analysis scripts
run_all.R   # Master script that runs every analysis in order

Decompress data.tar and scripts.tar to access the folders.

There are two kinds of data file:

1. Reproduction files – one row per replicate plate, with daily offspring
   counts in columns D1, D2, … Each column is the number of live offspring
   counted on that day of the reproductive period.
2. Survival files – individual worm survival records, one row per worm (except
   P0GLMM.csv, which is in long format with one row per worm-per-day).

Treatment coding conventions

In both schemes the letters mean C / c = Control and L / l = Larval (L1)
starvation. The unambiguous Parental Treatment and Offspring Treatment columns
are also provided

Missing values

Missing values are coded as NA or left as an empty cell. In survival files,
Death_date, D1, Day and Age can be NA for individuals that were lost or
never scored. In reproduction files, days with no observation are blank/NA. Dates
are in DD/MM/YYYY format.

Cause-of-death codes (survival files)

Only worms that died of natural causes contribute a death event
(Event / status = 1). All other codes denote censored individuals
(Event / status = 0) removed for reasons unrelated to natural ageing:

---

Files and variables

Data files

File: data/P0repro.csv

Age-specific reproduction of the parental (P0) generation. One row per plate
(60 plates; 30 per treatment).

• Plate — plate / replicate identifier
• Treatment — Control or Larval Starvation
• D1–D6 — number of live offspring counted on days 1–6 of the reproductive period (integer counts)

File: data/P0Surv.csv

Individual survival records for P0 worms, one row per worm (200 worms). Input to the
Cox mixed-effects survival model.

• Treatment — Con (control) or L1 (larval starvation)
• Plate — plate identifier
• Worm — worm identifier within a plate
• Death_date — date the worm died or was censored (DD/MM/YYYY)
• Cause — cause code (see cause-of-death table above)
• D1 — start date of observation, i.e. day 1 (DD/MM/YYYY)
• Age — age in days at death/censoring (Death_date − D1)
• Event — 1 = died (natural death), 0 = censored
• Day — survival time in days (equal to Age)

File: data/P0GLMM.csv

Long-format daily survival records for P0 worms (one row per worm-per-day; 3,198
records). Input to the binomial survival GLMM (status ~ Treatment + Age).

• (first, unnamed column) — row index
• Worm — worm identifier
• Age — observation-day index for that worm (1, 2, 3, …)
• Treatment — Con or L1
• Plate — plate identifier
• Death_date — date of death (DD/MM/YYYY)
• Cause — cause code (all D in this file)
• D1 — start date (DD/MM/YYYY); may be NA
• Day — day value used to reconstruct the survival timeline; may be NA (filled forward in P0.R)
• status — 0 = alive on that day, 1 = dead

File: data/F1.csv

Age-specific reproduction of the F1 generation. One row per plate (120 plates; 30 per
treatment combination).

• Plate — plate identifier
• Parental Treatment — Control or Larval Starvation (grandparental/parental exposure)
• Offspring Treatment — Control or Larval Starvation (F1 exposure)
• D1–D5 — number of live offspring counted on days 1–5 (integer counts)
• Treatment — combined code in Offspring(Parental) format: C(C), C(L), L(C), L(L)

File: data/F3.csv

Age-specific reproduction of the F3 generation (transgenerational). One row per plate
(120 plates; 30 per treatment combination).

• Plate — plate identifier
• Parental Treatment — Control or Larval Starvation
• Offspring Treatment — Control or Larval Starvation
• D1–D6 — number of live offspring counted on days 1–6 (integer counts)
• Treatment — combined code in Offspring(Parental) format: C(C), C(L), L(C), L(L)

File: data/F1_survival.csv

Individual survival records for F1 worms, one row per worm (400 worms).

• Parental Treatment — Con or L1
• Treatment — offspring treatment: Con or L1
• Plate — plate identifier
• Worm — worm identifier within a plate
• Death_date — date of death/censoring (DD/MM/YYYY)
• Cause — cause code (see table above)
• D1 — start date, day 1 (DD/MM/YYYY)
• Age — age in days at death/censoring
• Event — 1 = died, 0 = censored
• TreatmentCode — combined code in parental+offspring format: cc, cl, lc, ll

File: data/F3_survival.csv

Individual survival records for F3 worms, one row per worm (400 worms). Columns as in
F1_survival.csv, plus:

• Day — survival time in days (equal to Age)

File: data/Transgen.csv

Combined F1 + F3 age-specific reproduction, used for the transgenerational analysis
(240 rows: 120 F1 plates + 120 F3 plates).

• Plate — plate identifier
• Generation — F1 or F3
• Parental Treatment — Control or Larval Starvation
• Offspring Treatment — Control or Larval Starvation
• D1–D6 — number of live offspring counted on days 1–6 (integer counts)
• Treatment — combined code in Offspring(Parental) format: C(C), C(L), L(C), L(L)

Code / script files

All scripts live in scripts/ and use the here package for portable file paths,
so they read the data from data/.

• P0.R — P0 age-specific reproduction (spline GLMM), rate-sensitive fitness (Leslie-matrix dominant eigenvalue), LRS, and survival GLMM.
• P0_Plots.R — P0 reproduction / LRS / fitness figures.
• P0Surv.R — P0 Cox mixed-effects (coxme) survival model, Kaplan–Meier curve and forest plot.
• F1.R — F1 reproduction, fitness, LRS and survival analyses.
• F1Plots.R — F1 figures.
• F1Surv.R — F1 Cox mixed-effects survival model and Kaplan–Meier curves.
• F3.R — F3 reproduction, fitness, LRS and survival analyses.
• F3Plots.R — F3 figures.
• F3Surv.R — F3 Cox mixed-effects survival model and Kaplan–Meier curves.
• transgen.R — Transgenerational (F1 vs F3) reproduction analysis using Transgen.csv.
• WormSimulation.R — Simulation of the dynamics of two life-history strategies in a worm population.
• GraphExport.R — Assembles and exports the main and supplementary figures (SVG) to plots/Final/Export/.
• ModelOutputs.R — Helper functions that turn the fitted models into formatted tables and export them as .docx to model_output/.
• run_all.R — Master script (in the repository root) that runs all of the above in order.

---

Code / software

Software: R (version ≥ 4.2). RStudio recommended.

R packages used across the scripts:

here, tidyverse (dplyr, tidyr, ggplot2, purrr, readr, stringr),
glmmTMB, DHARMa, car, splines, emmeans, easystats (performance),
patchwork, gt, survival, survminer, coxme, AICcmodavg

Install any missing packages with, e.g.:

install.packages(c(
  "here", "tidyverse", "glmmTMB", "DHARMa", "car", "emmeans",
  "performance", "patchwork", "gt", "survival", "survminer",
  "coxme", "AICcmodavg"
))

To reproduce the full analysis:

1. Download the deposit and keep the data/, scripts/ and run_all.R structure intact.

2. Open the folder in R / RStudio (with the working directory set to the folder root).

3. Run:

   source("run_all.R")
   

This runs the P0, F1, F3 and transgenerational analyses and the simulation, fits the
survival models, builds the figures, exports the model-output tables to
model_output/. Figures are written
to plots/Final/Export/.

---

Access information

Authors: Isaac Harris¹, Elizabeth M. L. Duxbury¹, Tracey Chapman¹,
Simone Immler¹, Alexei A. Maklakov¹

¹ School of Biological Sciences, University of East Anglia, Norwich Research Park,
Norwich NR4 7TJ, UK

Contact: Isaac Harris — Isaac.Harris@uea.ac.uk