Major disturbance events can have large impacts on the demography and dynamics of animal populations. Hurricanes are one example of an extreme climatic event, predicted to increase in intensity due to climate change, and thus expected to be a considerable threat to population viability. However, little is understood about the underlying demographic mechanisms shaping population response following these extreme disturbances. Here, we analyze 45 years of the most comprehensive free-ranging nonhuman primate demographic dataset to determine the effects of major hurricanes on the variability and maintenance of long-term population fitness. For this, we use individual-level data to build matrix population models and perform perturbation analyses. Despite reductions in population growth rate mediated through reduced fertility, our study reveals a demographic buffering during hurricane years. As long as survival does not decrease, our study shows that hurricanes do not result in detrimental effects at the population level, demonstrating the unbalanced contribution of survival and fertility to population fitness in long-lived animal populations.

The demographic dataset of Cayo Santiago rhesus macaques was shared by the Caribbean Primate Research Center, University of Puerto Rico.

This folder contains all the required data and codes to reproduce the main study results. Codes were built in R version 3.5.1. These codes use packages “popbio”, “tidyverse”, “ggplot2”, “scales”, and “gamm4” for demographic data management and analysis. The data contains the following:

1. Morcillo_et_al_nonhurricane_transitions.csv contains every individual transition during non-hurricane years; including age and stage at time t and fate or stage at time t+1.
2. Morcillo_et_al_hurricane_transitions.csv contains every individual transition during hurricane years; including age and stage at time t and fate or stage at time t+1.
3. Morcillo_et_al_nonhurricane_bootstrap.csv contains the bootstrap transition matrix for a non-hurricane year.
4. Morcillo_et_al_hurricane_bootstrap.csv contains the bootstrap transition matrix for a hurricane year.
5. Morcillo_et_al_annual_matrices.RData contains annual transition matrices from 1973 to 2018.
6. Morcillo_et_al_gamm_model.RData contains the generalized additive mixed top model output

Specifically, the codes provide calculations for:

1. Mean matrix populations models for non-hurricane and hurricane years from individual-level transition data.
2. Eigen analysis for matrix models
3. Life table response experiment
4. Bootstrap and Kolmogorov-Smirnov test
5. Generalized additive mixed model
6. Figures