Exploiting animal personality to reduce chronic stress in captive fish populations

Abstract

Chronic stress is a major source of welfare problems in many captive populations, including fishes. While we have long known that chronic stress effects arise from maladaptive expression of acute stress response pathways, predicting where and when problems will arise is difficult. Here we highlight how insights from animal personality research could be useful in this regard. Since behavior is the first line of organismal defense when challenged by a stressor, assays of shy-bold type personality variation can provide information about individual stress response that is expected to predict susceptibility to chronic stress. Moreover, recent demonstrations that among-individual differences in stress-related physiology and behaviors are underpinned by genetic factors means that selection on behavioral biomarkers could offer a route to genetic improvement of welfare outcomes in captive fish stocks. Here we review the evidence in support of this proposition, identify remaining empirical gaps in our understanding, and set out appropriate criteria to guide development of biomarkers. The article is largely prospective: fundamental research into fish personality shows how behavioral biomarkers could be used to achieve welfare gains in captive fish populations. However, translating potential to actual gains will require an interdisciplinary approach that integrates the expertise and viewpoints of researchers working across animal behavior, genetics, and welfare science.

Keywords: personality fish; quantitative genetics; selection; stress; welfare.

Figure 1

(A) Illustration of how repeated observations (points) on different individuals (colors) can be used to partition among- from within-individual variation in a population. The horizontal dashed line shows the population mean, and horizontal solid lines the mean for each individual. Solid vertical lines show the extent to which individuals differ (on average) from the population mean. Dashed vertical lines show how an individual's observations vary from their own average value (i.e., the within-individual or residual variance). (B) Observations (points) and individual-specific means (horizontal lines) for duration of time in the middle (inner) zone for 24 zebrafish (Danio rerio) observed over six open field trials. Five representative individuals have been colored to highlight their means and observed values. The adjusted repeatability of this behavior within the study was estimated at 0.56 from a mixed model analysis. Data redrawn from (60) with permission. Danio rerio silhouette by Josefine Bohr Brask and used under Creative Commons Licence BY-NC-SA 3.0 (https://creativecommons.org/licenses/by-nc-sa/3.0/).

Figure 2

Example from a recent study of stress-related behavior and glucocorticoid response in Trinidadian guppies (Poecilia reticulata) by Houslay et al. (67). (A) Individuals from a pedigreed population were observed repeatedly in open field trials (OFTs). Blue track shows a summary of individual movement over the trial period; red rectangle indicates the overlaid zoning (inner or middle zone vs. outer zone). (B) Individuals were also measured repeatedly for their free circulating cortisol response to a mild stressor of handling and isolation. (C) Points show (predicted) genetic deviations from the overall means for time in the middle (x axis) and ln-transformed cortisol (y axis) of this pedigreed population. Horizontal and vertical error bars around each point show standard errors on these estimates. The black line shows the regression line, calculated as cov(y,x)/var(x) from the genetic variance-covariance matrix estimated in a multivariate animal model. Data redrawn from Houslay et al. (67) with permission. Photograph by T. Houslay. P. reticulata silhouette by Ian Quigley used under Creative Commons Licence BY-NC-SA 3.0 (https://creativecommons.org/licenses/by-nc-sa/3.0/).

Figure 3

Among-family variation in boldness in juvenile guppies. (A) Using a simple scototaxis assay for anxiety like behavior juvenile guppies are individually transferred to cylinder and then released into a 5 × 5 cm arena equally divided into black and white sections. A camera placed above allows video tracking and measurement of boldness. Here we define a simple boldness score from a 300 s observation period as time over light background –time over dark background (such that higher values indicate greater preference for light background). (B) Individual observations (grey points) on 149 fish from 24 families (of 1–33 offspring, mean brood size = 12) aged 2–14 days (with a of 6 days). Fish consistently prefer the dark background (negative boldness scores), with dashed line at boldness = 0 corresponds to no preference. However, mean behavior also varies significantly among-families (black points and associated error bars depict family means with 95% CI; tested conditional on age, F_23,124 = 1.71, P = 0.034) consistent with genetic variation for boldness.