Data from: Site-level field of view is associated with altered anti-predator responses in farming damselfish
Data files
Sep 08, 2025 version files 94.06 KB
-
anti_predator_data_beh_ecol.csv
40.12 KB
-
raw_viewshed_data.csv
28.30 KB
-
README.md
6.16 KB
-
refuge_data_antipred.csv
16.73 KB
-
Rugosity_sites.csv
1.32 KB
-
viewshed_data.csv
1.43 KB
Abstract
The three-dimensional (3D) structure of habitats influences how prey detect and respond to predators, but the specific roles of different aspects of structural complexity remain poorly understood, particularly in coral reef ecosystems. We used 3D models of three Caribbean reef sites to quantify three structural metrics at site-level: field of view (the extent of observable area), refuge density (density of holes), and rugosity (reef surface roughness). We then observed the anti-predator behaviour of damselfish, parrotfish, and wrasses at each site. Refuge density was similar across sites, likely reflecting long-term regional loss of fine-scale complexity in the Caribbean. Territorial damselfish showed species-specific responses to habitat structure, especially in relation to field of view. Stegastes adustus, for example, exhibited shorter flight initiation distances (FIDs) at the site with the highest field of view, consistent with expectations from optimal escape theory. In contrast, wrasse and parrotfish species showed little variation in behaviour across sites, though larger individuals tended to have longer FIDs and flight distances. While rugosity is widely used as a proxy for reef complexity, our results suggest that field of view may be more strongly associated with differences in anti-predator behaviour, particularly in damselfish. These findings highlight the need to assess multiple dimensions of habitat structure, as even closely related species may exhibit distinct behavioural adaptations to their 3D environment.
1. Paper Citation
Boon, J. S., Stratford, J. E., Lynch, J., Yesson, C., Keith, S. A., Exton, D. A., & Field, R. (2025). Site-level field of view is associated with altered anti-predator responses in farming damselfish. Behavioral Ecology
2. Abstract
[see above]
3. Originators
James S Boon
4. Contact information
James S Boon
Location: School of Geography, University Park, University of Nottingham, UK
Email: boony9@hotmail.co.uk
5. Date of data collection
June to August 2022 and 2023
6. Geographic location of data collection
Utila, Bay Islands, Honduras
7. Funding sources
Natural Environment Research Council
ACCESS INFORMATION
1. Licenses/restrictions placed on the data
None
2. Data derived from other sources
No
DATA FILES AND VARIABLES
1. anti_predator_data_beh_ecol.csv
This .csv file contains the raw anti-predator response data collected in the field. In the file are the following columns:
• Site: The dive site where anti-predator response data was collected.
• Date: The date the data was collected (day/month/year).
• Dive_start_time: The time we entered the water and began both the dive for data collection.
• Group: Broad taxonomic category of the focal fish used in the anti-predator response experiment (e.g., damselfish, parrotfish, wrasse).
• Species: The specific species of the focal individual
• Length: Estimated total length of the individual (cm).
• Group_size: Estimated number of conspecifics present (NA represent "Not applicable" due the individual not being in a group).
• SD: Starting distance, which is the distance between the model predator and the focal individual when the approach began (cm).
• FID: Flight initiation distance, which is the distance between the model predator and the focal individual when escape behaviour was initiated (cm).
• Escape_response: General category of escape behaviour after the model predator was presented (e.g., flee, evade, take shelter/refuge).
• Distance_refuge: For individuals that took refuge, the distance from the initial flight point to the refuge (cm; NA denotes "Not applicable" due to the individual not fleeing in to a refuge).
• Distance_fled: For individuals that fled into open water, the distance travelled before slowing down (cm; NA denotes "Not applicable" due to the individual not fleeing in to open water)
• Refuge_type: Type of refuge used if the focal individual sought one (e.g., live coral, rock, dead coral; NA denotes "Not applicable" due to the individual not fleeing in to a refuge).
• Investigator: The person who took the measurements for that specific experiment.
2. refuge_data_antipred.csv
This .csv file contains the raw data from the refuge analysis. In the file are the following columns:]
• Site: Name of the dive site around Utila.
• Quadrat_number: The quadrat (replicate) used in the refuge analysis.
• Hole_size: The predefined refuge size being counted.
• Hole_count: The number of refuges of that size.
3. raw_viewshed_data.csv
This .csv file contains the raw data from the field of view/viewshed analysis. In the paper, we report the mean values, but here you will find the raw measurements for each field of view/viewshed taken on every transect at each site. In the file are the following columns:
• Site: Name of the dive site around Utila.
• Transect: Name of the individual 3D transect where the cross section and measurements were taken.
• Angle: The angle to the highest topographical point from a set point.
• minus_90: The angle value minus 90°, showing how much of the view is unobstructed.
• Proportion_visible: The proportion of the 90° field that is visible. This is the value used to determine field of view/viewshed.
4. viewshed_dates.csv
This .csv file contains the mean data from the field of view/viewshed analysis. In the file are the following columns:
• Site: Name of the dive site around Utila.
• Transect: Name of the transect where the field of view/viewshed measurements were taken.
• Viewshed_mean: Mean field of view/viewshed value for each transect (calculated from the Proportion visible column in the raw_viewshed_data.csv file).
• n: Number of individual data points used to calculate the mean (verified against the Proportion visible values in raw_viewshed_data.csv).
5. Rugosity_sites.csv
This .csv file contains the raw data from the rugosity analysis. In the file are the following columns:
• UniqueID: Name of the transect where the rugosity measurement was taken.
• Site: Name of the dive site around Utila.
• Rugosity: Rugosity value for each transect.
CODE SCRIPTS
1. Behav_ecol_2025_antipred_code.R
The R script used for the paper’s statistical analysis includes comparisons of structural metrics between sites and analyses of anti-predator behaviours for each taxon across sites.
SOFTWARE VERSIONS
R v. 4.2.3
Loaded packages and version:
• tidyverse: 2.0.0
• ggpubr: 0.6.0
• ggdist: 3.3.2
• ggpp: 0.5.7
• grid:base
• brms: 2.21.0
• tidybayes: 3.0.6
• bayesplot: 1.11.1
• brmstools: 0.5.3
• emmeans: 1.10.2
• ggeffects: 1.6.0
• DHARMa: 0.4.6
• car: 3.1-2
• rstatix: 0.7.2
• report: 0.5.8
