Simulating larval dispersal of portunid crabs

This repository stores all code/files used for simulation and analysis of particle tracking experiments described in Hewitt et al., 2022 and Schilling et al. 2022. These experiments were run on the UNSW High Performance Cluster (Katana). Briefly, these studies simulate larval dispersal of Giant Mud Crab (Scylla serrata), Blue Swimmer Crab (Portunus armatus) and Spanner Crab (Ranina ranina) to assess population connectivity along the east Australian coast.

Particle tracking is conducted in Python using PARCELS, all output is then processed in R.

Jobs are submitted using the .pbs scripts in the Simulations folder, and are separate for each species/direction combination. The workflow includes multiple job submissions and follows this order:

1. Submit the particle tracking job with the appropriate .pbs. These are named in a species_direction.pbs format, so to run backwards simulations for Giant Mud Crab you'd submit gmc_backwards.pbs to Katana. Depending on the direction and species these jobs get split up in different ways (to manage memory/computation time), unless you intend to change the structure of the workflow this shouldn't matter.
2. Once particle tracking simulations are complete we need to introduce some biology to our particles (e.g., growth rates, settlement). This is achieved by running the R script portunid_particle_processing.R which uses a range of functions stored in the R folder. This script is set up to do different things for different species/directions, so we need to submit the appropriate _processing.pbs (prefixed by our species_direction combination) to tell this script which species we're working on. So, to continue our backwards Giant Mud Crab analysis we'd submit gmc_backwards_processing.pbs
3. No we're ready to bring all of the output together (remember the splitting up in step 1). We do this by submitting the appropriate _join.pbs, prefixed by species_direction (e.g., gmc_backwards_join.pbs).
4. Now we have our fully processed particle trajectories - ready for analysis. The analysis described in Hewitt et al. (2022) is carried out by submitting the R script portunid_particle_analysis_plotting.R which is called via portunid_particle_plotting.pbs (note this runs the analysis for both Giant Mud Crab and Blue Swimmer Crab). All code for the analysis described in Schilling et al. (2022) can be found in the BRAN_2020 folder.