Data from: Increases in predation favor evolutionary shifts in behavioral plasticity in Trinidadian killifish
Data files
Mar 21, 2025 version files 14.02 GB
-
Predator_Cue_Videos_Sample.zip
14.02 GB
-
PredatorCueTrials2ndGenFish.csv
6.70 KB
-
PredatorCueTrialsWildFish.csv
33.93 KB
-
README.md
4.19 KB
Abstract
Behavioral plasticity is expected to be favored in risky environments, such as when prey species coexist with predators, because prey must alternate between fitness related foraging/mating behaviors and antipredator behaviors that enhance survival. We compared behavioral plasticity in Trinidadian killifish that are found in sites with and without predators. We quantified aggressive and antipredator behaviors via a mirror assay in second-generation lab-reared and wild-caught killifish before and after exposure to predator cues. We compared two types of aggression including: overt aggression (ramming, biting, lunging, tail-slapping) and display aggression (spine arching, bending into an s-shape, and opercular flaring). We additionally compare the amount of time the fish spent frozen as a proxy for anti-predator behavior. We show clear differences in plasticity between populations with and without predators. Killifish from sites with predators decreased overt aggression in response to exposure to predator chemical cues. Killifish from sites that lack predators showed a higher degree of canalization in aggression presumably due to the high competition they face in those locales. Interestingly, wild fish from sites without predators retain the ability to detect the predator cue and respond by decreasing overt aggression and increasing time spent frozen, though to a lesser degree compared to the fish from sites with predators. Our results support the expectations that development in a risky environment favors evolutionary changes in predator-mediated behavioral plasticity.
https://doi.org/10.5061/dryad.xwdbrv1pt
Description of the data and file structure
These data represent efforts to determine how predator presence influences evolution of behavioral plasticity in Trinidadian killifish (Anablepsoides hartii). There are .csv files corresponding to our second-generation lab-reared fish and wild-caught killifish. There is also a .zip folder with a subset of the behavioral videos taken (DRYAD does not allow more than 300 GB of data to be uploaded and the total of these videos is over that). Video file names correspond to the fish name (column: Name) in the .csv files. All data were collected in 2023. We performed a mirror-aggression behavioral assay pre and post addition of a predator odor cue. Individual aggressive behaviors are listed (ramming, biting, etc.) in counts as well as time spent frozen (an antipredator behavior) in seconds during the assay. Please note that in the PredatorCueTrialsWildFish.csv file any blank cells in the column titled “weightg” are missing values because we experienced issues with our scale and were unable to weigh those fish (see also the “notes” column). The “PreTrialTime” and “PostTrialTime” columns each have an empty cell per fish because there could only be one time for each trial for each fish which are repeated throughout the file. Some fish are missing the “PreTrialTime”, and “PostTrialTime” because the time was not noted for those trials and the camera’s clock was inaccurate.
Files and variables
File: PredatorCueTrials2ndGenFish.csv
Description: Second-generation lab-reared fish sample
Variables
- River: Watershed of origin
- Pop: Population of origin representative of predator presence or absence (HP=high predation, KO=killifish only)
- ID: Arbitrary name given to track individuals
- Name: Arbitrary name given to track individuals
- Sex: M=male; F=female
- TLmm: Total length (mm)
- Wgtg: Weight (g)
- trial: Pre and post addition of the predator odor cue (1=pre-trial, 2=post-trial)
- Freeze.time.sec: Time spent frozen (sec)
- Ram.instances: Counts of ramming against the mirror with the head
- Bite.instances: Counts of biting the mirror
- Lunge.instances: Counts of darting quickly towards the mirror
- Tailbeat.instances: Counts of slamming the side body/tail against the mirror
- Sbend.instances: Counts of bending the body into an s-shape
- SpineArch.instances: Counts of raising or hunching of the back
- OpFlare.instances: Counts of operculum flaring
- OvertAggression: Sum of ram, bite, lunge, and tailbeat counts (ie., attempts to fight)
- DisplayAggression: Sum of s-bending, spine arching, and operculum flaring (ie., posturing/signaling)
File: PredatorCueTrialsWildFish.csv
Description: Wild-caught fish sample
Variables
- River: Watershed of origin
- Pop: Population of origin representative of predator presence or absence (HP=high predation, KO=killifish only)
- ID: Arbitrary name given to track individuals
- Name: Arbitrary name given to track individuals
- Caught: Date fish were caught
- Videoed: Date fish were assayed
- Lengthmm: Total length (mm)
- LNTL: log natural transformed total length
- Weightg: weight (g)
- Sex: m=male, f=female
- Notes: notes
- PreTrialTime: time at start of pre-trial (before predator cue was added)
- PostTrialTime: time at start of post-trial (after predator cue was added)
- Trial: 1=pre-trial, 2=post-trial
- Ram.instances: Counts of ramming against the mirror with the head
- Bite.instances: Counts of biting the mirror
- Lunge.instances: Counts of darting quickly towards the mirror
- Tailbeat.instances: Counts of slamming the side body/tail against the mirror
- Sbend.instances: Counts of bending the body into an s-shape
- SpineArch.instances: Counts of raising or hunching of the back
- OpercularFlare.instances: Counts of operculum flaring
- Overt.instances: Sum of ram, bite, lunge, and tailbeat counts (ie., attempts to fight)
- Display.instances: Sum of s-bending, spine arching, and operculum flaring (ie., posturing/signaling)
- FreezeTime.sec: Time spent frozen (sec)
The following data sets were collected from second-generation lab-reared and wild-caught Trinidadian killifish (Anablepoides hartii) in 2023. They represent a predator cue experiment in which fish were assayed for aggression and antipredator behaviors before (pre-trial) and after (pre-trial) introduction of a predator odor cue to determine how predator presence influences behavior. The data are in long-format to facilitate repeated measures analyses.