Silver eel meta-analysis

About

This study investigates the silver European eel (i.e. the downstream migrating eel) (Anguilla anguilla L.) seaward migration behaviour on a continental scale, i.e. Europe.

Fish migration is a poorly understood phenomenon. Traits like migration speed and timing have evolved to reach spawning and growing habitats in time under the prevailing conditions. Hence, these traits are likely population and/or region specific. For instance, European eels growing in Norway need to migrate a longer distance to the spawning area, presumably in the Sargasso Sea, than conspecifics growing in Portugal. Consequently, they will either migrate faster or earlier than the eels from Portugal. Undoubtedly, migration barriers interfere with fish migration leading to delays. To restore the historically low abundances of diadromous fish populations, we urgently need a better fundamental understanding of their migration. This project will focus on the key hypothesis that the migration speed and timing are population- and region-specific and coordinated in the life cycle of European eel; this hypothesis is subdivided in three subhypotheses:

Migration speed and timing of European eel vary according to 1. distance between foraging and spawning habitats within the species distribution range; 2. size and sex; 3. the presence and types of migration barriers.

For this meta-analysis, data from 19 projects/locations and 9 countries have been centralized: Belgium (4), Denmark (1), France (3), Germany (1), Lithuania (1), Norway (1), Portugal (1), The Netherlands (4), UK (3)

Project structure

Data

Data last updated on 23-11-2021

• /raw:

  • raw_detection_data.csv: dataset containing the raw detection data (gitignored due to its large size; see https://doi.org/10.5281/zenodo.15260539)
  • eel_meta_data.csv: dataset containing the meta-data on the tagged eels
  • deployments.csv: dataset containing the station names and positions of the receivers from ETN
  • /stour:
    • stour_data.csv: dataset containing the raw detection data from the River Stour, '2013_stour' project
    • stour_eel_meta.csv: dataset containing the eel meta-data from the River Stour, '2013_stour' project
    • stour_deployments.csv: dataset containing the deployment meta-data from the River Stour, '2013_stour' project
  • /nedap_meuse:
    • /raw_nedap_data: folder containing the raw NEDAP tracking data from the River Meuse in The Netherlands. The data is separated in files with English and Dutch headers.
    • nedap_meuse_eel_meta.csv: dataset containing the eel meta-data from the River Meuse under the NEDAP tracking system in The Netherlands
    • nedap_meuse_deployments.csv: dataset containing the deployment meta-data from the River Meuse under the NEDAP tracking system in The Netherlands

• /interim:

  • deployments.csv: dataset containing the station names and positions of the receivers from ETN, Stour project and NEDAP project
  • eel_meta_data.csv: cleaned dataset containing the eel meta data from ETN and Stour project
  • detection_data.csv: cleaned dataset containing the detection data from ETN and Stour project
    • successful_migrants_final_detection.csv: file with last detections of successful migrants (= eels considered to escape to the sea)
  • /receivernetworks: folder containing the generated files with the receiver networks per project that serve as input to calculate the distances between detection stations following the waterbodies in another GitHub repo (https://github.com/inbo/fish-tracking/tree/master/scripts/receiverdistanceanalysis)
  • /residencies: folder containing the residency datasets, calculated via the smooth_eel_tracks.R code
  • /speed: folder containing the speed datasets, calculated via the calculate_speed.R code
  • /migration: folder containing the datasets with records flagged as migration or not, calculated via the identify_migration.R code

• /external:

  • release_locations_stations.csv: file with the release locations and the abbreviated release station names.
  • /distance_matrices: folder containing the distance matrices with the distances between detection stations per receiver network (matrices are created at https://github.com/inbo/fish-tracking/tree/master/scripts/receiverdistanceanalysis).
  • station_order.csv: file containing the stations upstream the release location. This file is needed in calculate_speed.R
  • habitats.csv: file with the classification of the habitats linked to the detections stations, i.e. freshwater, tidal freshwater, tidal brackish, marine barrier and marine.
    • eels_wrs.csv: the number of water regulating structures (wrs) eels need to pass to reach the sea and the water regulating impact score of each water body.
  • project_geographical_location.csv: latitude of the different projects

Scripts

• /src:
• download_data.R: Download data from silver eel meta-analysis from ETN database via RStudio LifeWatch server
  • obtain detection dataset raw_detection_data.csv
  • obtain meta-data on tagged eels eel_meta_data.csv
  • obtain meta-data on deployments deployments.csv (station names and positions)
• attach_stour.R: Merge data from the River Stour to the dataset
• attach_nedap_meuse.R: Merge NEDAP data from the River Meuse in the Netherlands to the dataset
• clean_eel_metadata.R: Eel meta-data cleaning by removing irrelevant and redundant eels, columns and make certain column values consistent
• clean_detection_data.R: Data cleaning by filling in missing values and removing false detections
• attach_release.R: Add eel release positions and date-time to detection dataset
• create_interactive_maps.R: Create interactive html widget maps per project
• extract_network.R: Extract receiver networks based on detection data
  • This serves as input to calculate the distance matrices at https://github.com/inbo/fish-tracking/tree/master/scripts/receiverdistanceanalysis
• smooth_eel_tracks.R: Smooths duplicates and calculates residencies per eel per station. Therefore, it calls the following two functions:
  • 9a. get_nearest_stations.R: general function to extract the smoothed track for one eel (via its transmitter ID)
  • 9b. get_timeline.R: function to get the stations which are near a given station (where near means that the distance is smaller than a certain given limit, e.g. detection range).
    • --> Generate residency datasets per project and store them in /interim/residencies
• calculate_speed.R: Calculate movement speeds between consecutive detection stations. Also calculates swim distance, swim time, cumulative swim distance and station distance from source station.
  • 10a. calculate_speed_function.R: function to calculate speed between consecutive displacements; based on a function in Hugo Flavio's actel package
  • 10b. calculate_sourcedistance_function.R: function to calculate the station distance from a 'source' station; based on a function in Hugo Flavio's actel package
    • --> Generate speed datasets per project and store them in /interim/speed
  • 10c. clean_residency_data_functions.R: functions to clean residency datasets of specific projects (false and incorrect detections that came to light via the distance plots.
• identify_migration.R: Code that identifies records as migration based on a speed and distance threshold until the most downstream station. The method also includes a stationary threshold, which defines from which movement distance an eel can be considered non-statinoary.
  • 11a. identify_migration_functions.R: functions to identify migration.
• create_distance_plot.R: Create plots with travelled distance per eel and store as .pdf
• process_migration_data.R: Process migration data by combining the datasets of the different studies and calculate the number of migratory eels per study
• identify_escapement_success.R: Identify migration success as successful escapement to the sea - --> Generate successfulmigrantsfinaldetection.csv interim dataset with final record per eel that successfully escaped to the sea. This will be used in the script `analysemigration_period.R`.
• analyse_escapement_success.R: Analyse the successful escapement to the sea according to migration barrier impact and fishing
• analyse_arrival_at_sea.R: Analyse the last day as (successful) migration, which is the arrival at sea, to size, geographical location and migration barrier impact (i.e. WRS class)
• calculate_migration_speed_overall.R: Calculate the migration speed as the speed between the first and last detection and relate it to geographical position, habitat, size and sex
• calculate_migration_speed_habitats.R: Calculate the migration speed according to habitat type
• analyse_migration_speed_tidal.R: Analyse the migration speed in tidal areas according to size, sex and geographical location
• analyse_migration_speed_nontidal.R: Analyse the migration speed in non-tidal areas according to size, sex, geographical location and migration barrier impact (i.e. WRS class)
• analyse_migration_speed_tidal_vs_nontidal.R: Analyse the migration speed between tidal areas and non-tidal areas

Figures

• /figures:
  • /html_interactive_maps: interactive html widget maps per project
  • /distance_tracks: pdf-files with traveled distances per project +/raw: distance tracks of the raw data +/migration: distance tracks with arrivals records flagged as migratory vs non-migratory based on a specific distance and speed threshold (thresholds in the file titles)