This repository contains public code for the manuscript:

Zou, Heng-Xing, SJ Schreiber, and Volker HW Rudolf. "Stage-mediated priority effects and season lengths shape long-term competition dynamics" bioRXiv (2020). https://doi.org/10.1101/2020.08.14.251926

For any questions, please contact Heng-Xing Zou

Abstract

The relative arrival time of species can affect their interactions and thus determine which species persist in a community. Although this phenomenon, called priority effect, is widespread in natural communities, it is unclear how it depends on the length of growing season. Using a seasonal stage-structured model, we show that differences in stages of interacting species could generate priority effects by altering the strength of stabilizing and equalizing coexistence mechanisms, changing outcomes between exclusion, coexistence, and positive frequency dependence. However, these priority effects are strongest in systems with just one or a few generations per season and diminish in systems where many overlapping generations per season dilute the importance of stage-specific interactions. Our model reveals a novel link between the number of generations in a season and the consequences of priority effects, suggesting that consequences of phenological shifts driven by climate change should depend on specific life histories of organisms.

Code Files

All simulations were run in R version 4.2.1.

• Functions.R contains all functions used for contructing the model and running simulations.
• Analysis.Rmd contains all code for generating figures in the main text and the appendix. Running any portion of this file will automatically run Functins_SpatialBH.R. Note that running all simulations takes a long time; we therefore do not recommend knitting this file. Instead, we have provided SavedEnviron.RData that contains all generated data.

Data

SavedEnviron.RData contains generated data from all simulations in the main text. This is a pre-packaged R dataset that includes all variables in the environment. To load the dataset, simply double click while Functions.R and Analysis.Rmd are opened in RStudio. Loading SavedEnviron.RData will load results from invasion analyses using the following model formulation:

• Stage-mediated interspecific competition (3 data frames): lowintra, midintra and highintra.
• No stage-mediated interspecific competition (3 data frames): xlowintra, xmidintra and xhighintra.
• Both stage-mediated intra- and interspecific competition (3 data frames): vlowintra, vmidintra and vhighintra.
• Periodic arrival time with stage-mediated interspecific competition (1 data frame): var2.
• Periodic arrival time without stage-mediated interspecific competition (1 data frame): xvar2.
• Simulation with \Delta s>5, with stage-mediated interspecific competition (1 data frame): long_midintra.
• Simulation with \Delta s>5, without stage-mediated interspecific competition (1 data frame): xlong_midintra.
• Simulation with adult density dependence, with stage-mediated interspecific competition (1 data frame): adu_dd_midintra.
• Simulation with adult density dependence, without stage-mediated interspecific competition (1 data frame): xadu_dd_midintra.

All above data frames have the following columns:

• inv_i: identify of invading species.
• res_j: identify of resident species
• lambda_ij: long-term growth rate of species i when invading the stable stage distribution of species j.
• lambda_i: long-term growth rate of species i in monoculture
• sensitivity: sensitivity to competition of species i, calculated as in Eqn. 2.4.
• delta_s: the initial stage difference of the simulation.
• n_gens: the season length of the simulation.

Reproducing Figures

• Make sure all files are under the same directory.
• Open Analysis.Rmd, then load SavedEnviron.RData.
• Follow specific instructions in Analysis.Rmd to generate figures in the main text, or change model parameters to produce figures in the appendix.

Latest release: version Jan 2023; uploaded Sep 12 2023