Code and data from: Silver spoons, reproduction, and growth catch-up in eastern grey kangaroos
Data files
Mar 05, 2025 version files 138.11 KB
-
Early_reproduction.csv
132.91 KB
-
README.md
5.20 KB
Abstract
Early growth and environmental conditions can shape life-history trajectories. Long-lived iteroparous species with indeterminate growth face growth-reproduction trade-offs for most of their lives. Poor early conditions can delay primiparity and restrict growth, potentially compensated for by faster growth later in life, at the cost of reduced reproduction. We explored the variation in age at primiparity and early reproduction in eastern grey kangaroos, based upon 13 years of monitoring about 100 known-age females. We then examined associations between early reproduction, later reproduction, lifetime growth. We used a modified von Bertalanffy growth function to model the indeterminate growth of females and to test the effects of early reproduction on lifetime growth. Favorable environmental conditions, large size, and condition as a subadult led to earlier reproduction and higher reproductive success at ages 3-5. As females aged, size and condition had diminishing effects on reproductive success. Females with greater early-adulthood reproduction had slightly higher reproduction later in life. We did not detect a growth cost of early reproduction. Large females in good condition favored early reproduction over growth, while those with poor early growth exhibited growth catch-up at the cost of reproduction both early and later in life. As reported for other long-lived iteroparous species with indeterminate growth, female kangaroos prioritize growth over reproduction for much of their lifespan. Eastern grey kangaroos show heterogeneity in early growth and reproductive strategies. Early primiparity and reproduction are constrained by body condition, size, and environmental conditions when females are pre-reproductive subadults.
https://doi.org/10.5061/dryad.ffbg79d5g
Description of the data and file structure
Data files and script's code and vector definitions:
Age certainty: how we aged the female kangaroo
- Age_certainty = 0 - Unknown age
- Age_certainty = 1 - Aged in pouch
- Age_certainty = 2 - First caught as subadult or < 20 kg
- Age_certainty = 3 - First caught when over 20 kg
- Age_certainty = 4 - Aged from molar progression
Pouch young survival: specific to the female's reproductive success
- PY_survival = 0 - Female had no pouch young
- PY_survival = 1 - Females' pouch young died before large pouch young stage (LPY)
- PY_survival = 2 - Females' pouch young died as large pouch young (LPY)
- PY_survival = 3 - Females' pouch young died as young at foot (YAF)
- PY_survival = 4 - Females weaned a young
Early reproduction total: number of LPY a female had per year between ages 3-5 at age 5
- Earlyreproduction_total = 0 - No LPY between ages 3-5
- Earlyreproduction_total = 1 - One LPY between ages 3-5
- Earlyreproduction_total = 2 - Two LPY between ages 3-5
- Earlyreproduction_total = 3 - Three LPY between ages 3-5
Vector definitions: (ellipses indicate a variable placeholder as a prefix or suffix (e.g., 'scaled_...' could represent 'scaled_mass')):
- Prim_age: Age of primiparity
- Following_RS: Reproductive success to LPY stage in the following year
- Current_RS: Reproductive success to LPY stage in the current year
- Veg_per_roo: available vegetation per kangaroo (kg/kangaroo)
- cum_vegperroo: cumulative vegetation per kangaroo, from age 1 to current age
- Kn: Relative condition index (Kn)
- Juvenile_survival: Annual proportion of offspring that survived to weaning from the LPY stage (kangaroos weaned / LPY).
- scaled_...: variable was scaled or standardized (variable - mean of variable / standard deviation of variable)
- ..._age2: variable measured at age 2
- ..._age5: variable measured at age 5
von Bertalanffy growth equation parameters (Vincenzi et al. 2016):
- loginitialsize: logarithm of the initial size of kangaroos at LPY stage for von Bertalanffy growth function (~250mm) parameterized through priors
- logMSP: logarithm of the metabolic scaling parameter
- logEP: logarithm of the environmental parameter
- psi: logarithm of the environmental-behavioral interaction parameter
Vincenzi S, Crivelli AJ, Munch S, Skaug HJ, Mangel M. 2016. Trade-offs between accuracy and interpretability in von Bertalanffy random-effects models of growth. Ecol Appl. [accessed 2024 Dec 10];26(5):1535–1552. https://doi.org/10.1890/15-1177
Files and variables
File: Supplementary Material.docx
Supplementary material word file contains tables for annual environmental data (Table S1) as well as model structures (Table S2), prior specification (Table S3), and outputs (Tables S4-S5). Supplementary figures referred to in the text are also found in this file (Figure S1-S19).
File: Early_reproduction.csv
Variables:
- ID: Unique identifier for each individual kangaroo.
- Year: Year of the capture/measurement.
- Age: Age of the individual in years at the time of capture.
- Age_certainty: Certainty of the age estimate (e.g., known vs. estimated).
- Cohort: Birth year or age class of the individual.
- PY_survival: Survival status of the pouch young (see data description for codes).
- Month: Month of capture/measurement.
- Age_months: Age of the individual in months at the time of capture.
- Mass: Body mass of the individual (kg).
- Leg: Hind leg length of the individual (mm).
- PY_survival_at_least_LPY_2023: Indicator of whether the pouch young survived to at least the LPY stage in all years including 2023. At the time of publication, survival to weaning was unknown in 2023, but this does not impact results, as RS was measured based on survival to the LPY stage at a minimum.
- Yearly_veg_sum: Total available forage in the study area for a given year (kg/ha).
- Density: Estimated population density (kangaroos/ha) for a given year.
- Veg_per_roo: Available vegetation per kangaroo (kg/kangaroo), calculated as yearly available forage divided by population density.
Missing data
All missing data (NA) represent missing data, either because the variable was not recorded that year due to human error, or that individual kangaroo was not captured.
Code/software
The data were analyzed using R (version 4.2.3) via RStudio (version 2023.12.1-402) (R Core Team 2023; Posit Team 2024).
Required software and packages:
- Data manipulation: dplyr, tidyr, stringr, forcats
- Bayesian modeling and diagnostics: brms, tidybayes, bayesplot, bayestestR
- Statistical summaries: broom
- Visualization: ggplot2, gridExtra, patchwork, cowplot, ggh4x
Access information
Other publicly accessible locations of the data:
We monitored a population of eastern grey kangaroos in Wilsons Promontory National Park, Australia, from 2008 to 2023. Vegetation availability was estimated using herbivore exclosures, and population density was assessed through seasonal line-transect sampling and DISTANCE software. Female kangaroos were captured, marked, weighed, and measured, with reproductive status recorded. Birthdates of pouch young were estimated using growth models, and individuals of unknown age were excluded. See article methods for more detail.