Recovery of reef shark populations invokes anti-predator behaviours in mesopredatory reef fishes on a coral reef
Data files
Mar 06, 2025 version files 168.85 KB
-
Ashmore_data.csv
165.61 KB
-
README.md
3.24 KB
Abstract
Natural experiments where populations of large predators have recovered following management initiatives offer insights into the potential for these animals to structure communities via consumptive and non-consumptive effects on their prey. Ashmore Reef, a coral reef off the coast of Western Australia, provides such an opportunity. Here, reef shark populations have increased significantly since the enforcement of a no-take MPA in 2008. This change has been accompanied by an increase in abundance of medium and large mesopredatory teleosts, but a decline in small mesopredatory teleosts.We explored whether these changes in abundance were accompanied by changes in anti-predator (non-consumptive) behaviours of mesopredators due to an increase in both acute and chronic risk of predation. We analysed videos from Baited Remote Underwater Video Systems (BRUVS) collected prior to enforcement of no-take status in 2004 and after enforcement and shark recovery in 2016 to quantify any changes in anti-predator behaviours of small (<50 cm TL), medium (50-100 cm TL) and large (>100 cm TL) size classes of teleost mesopredators. Comparisons of the effect of chronic and acute risk on the total time teleosts spent in the BRUVS videos, proportional time spent on activities associated with various risks, and foraging intensity were made both within years (acute risk: in videos where sharks were absent compared to present), and between years (chronic risk: 2004 and 2016). Our results indicate that both small and medium-sized mesopredatory fishes exhibit behaviours indicative of risk-induced trait responses (anti-predator behaviours) in the presence of reef sharks, and that these seem to occur as a joint response to the interaction of acute and chronic risks. Our study suggests that the decline of small mesopredatory fishes following the recovery of reef sharks could be due to both the consumptive and non-consumptive impacts of sharks as predators in this system. These results show that both chronic and acute risk seem to play significant roles in shaping behaviours of mesopredators.
https://doi.org/10.5061/dryad.x95x69pvx
Description of the data and file structure
Files and variables
File: clean_data.csv
Description:
Each row represents the observation (from when a fish entered to when it left the field of view) of one mesopredatory teleost. During each observation, the below variables were recorded:
Variables
- genus_species: Genus and species. (Nominal)
- size_class: Size class of each mesopredatory fish species. Each mesopredatory fish species was assigned one of three size classes: ‘small’ (≤50 cm TL), ‘medium’ (50–100 cm TL), and ‘large’ (>100 cm TL). (Nominal, small/medium/large)
- year: Year of BRUVS drop. (Categorial, year)
- shark_in_video: Whether a shark was in the video, yes or no. (Binary, yes/no)
- sec_investigating_circling: The seconds a fish in an observation spent doing the 'investigating and circling' behaviour. (Continuous, time in seconds)
- sec_passing: The seconds a fish in an observation spent doing the 'passing' behaviour. (Continuous, time in seconds)
- sec_feeding: The seconds a fish in an observation spent doing the 'feeding' behaviour. (Continuous, time in seconds)
- sec_total: The total seconds a fish spent in the frame (the total seconds of the observation). (Continuous, time in seconds)
- frac_investigating_circling: The fraction of time a fish in an observation spent doing the 'investigating and circling' behaviour (as compared to the total time of observation, sec total). (Continuous, fraction of total seconds)
- frac_passing: The fraction of time a fish in an observation spent doing the 'passing' behaviour (as compared to the total time of observation, sec total). (Continuous, fraction of total seconds)
- frac_feeding: The fraction of time a fish in an observation spent doing the 'passing' behaviour (as compared to the total time of observation, sec total). (Continuous, fraction of total seconds)
- bout: The number of feeding bouts the observed fish in engaged in. (Discrete, count)
- bite: The number of feeding bites the observed fish in engaged in. (Discrete, count)
- habitat: The habitat type the video was recorded in. (Nominal, reef/near-reef)
- depth: The depth of the BRUVS top. (Continuous, depth in meters).
- log_depth: Log of the depth variable. (Log of 'depth')
- total_time_shark_present_frame: The total time sharks were present during the observation. (Continuous, time in seconds)
- n_frame_total: MaxN (maximum number of fish in the frame) during each observation. (Discrete, count)
- log_n_frame_total: log of n_frame_total. (Log of 'nframetotal')
- video: Unique video ID number for each BRUVS video. (Nominal)
Code/software
For code used to analyse the provided dataset, please view GitHub (https://github.com/andieich/Ashmore_behaviour)
Access information
Other publicly accessible locations of the data:
Data were provided from a subset of Baited Remote Underwater Video System (BRUVS) video recordings surrounding Ashmore Reef in 2004 (n=24) and 2016 (n=24). We analysed BRUVS videos collected prior to enforcement of no-take status in 2004 and after enforcement and shark recovery in 2016 to quantify any changes in anti-predator behaviours of small (<50 cm TL), medium (50-100 cm TL) and large (>100 cm TL) size classes of mesopredatory reef fishes. The dataset comprises behavioural measurements of the behaviours of these reef fishes. For a full description of the methods please see 'Recovery of Reef Shark Populations Invokes Anti-Predator Behaviours in Mesopredatory Reef Fishes on a Coral Reef' (Storm et al., Journal of Animal Ecology, 2025). The dataset has been analysed in R, for the code please see https://github.com/andieich/Ashmore_behaviour