A comparison of plasma glucose and plasma cortisol as selection markers for high and low stress-responsiveness in female rainbow trout

Abstract

Two groups of female rainbow trout displaying consistently divergent plasma cortisol responses to a 3 h period of confinement were identified following five separate confinement episodes at monthly intervals. High-responding (HRC) and low-responding individuals (LRC) continued to display divergent cortisol responses to confinement up to 21 months after the start of the study (342±34 and 208±21 ng ml⁻¹, respectively, at the final sample; p<0.01). HRC fish were significantly larger than LRC fish throughout the study period (533±13 and 422±10 g, respectively, overall; p<0.001), although significant differences in specific growth rate (SGR) were apparent only at the start of the study. Individual fish were also selected from the same population on the basis of their plasma glucose levels following confinement (high-responding glucose (HRG): 189±6; low-responding glucose (LRG) 121±3 mg dl⁻¹; p<0.001). However, the two selection traits (cortisol and glucose) identified separate subsets of the experimental population. HRG fish were also significantly larger than LRG fish although this difference was not so pronounced as for the cortisol-selected fish. There was no reciprocal relationship between body weight and stress responsiveness; fish selected from the population on the basis of high or low body weight displayed no divergence in either cortisol or glucose responses to confinement. Differences in size and SGR may indicate that HR fish adapted more rapidly to changes in environmental and social factors at the start of the study than LR fish did.

Introduction

Repeated or prolonged exposure of salmonid fish to stressful stimuli can result in reduced growth, impaired reproductive performance and immunosuppression (Barton, 1997). A wide range of responsiveness to stressors has been reported both for strains of salmonid fish (Pickering and Pottinger, 1989; McGeer et al., 1991) and for individual fish within strains (Pottinger et al., 1992). Because stress is an unavoidable component of the finfish aquaculture environment, and because of the adverse effects of chronic stress on performance characteristics, it has been suggested that fish which display a lower degree of responsiveness to stressors may perform better under intensive culture conditions than those which display a higher level of responsiveness (Pottinger and Pickering, 1997). It is also acknowledged that the converse may be true; given the adaptive significance of the stress response, high-responding fish may prove to be at a greater advantage, or low-responding fish at a relative disadvantage, overall. The prospect of selectively breeding fish for reduced responsiveness to stress has been addressed with respect to disease resistance (Fevolden et al., 1993) and overall performance (Pottinger et al., 1994). It is clear from these studies and similar work in poultry (Satterlee and Johnson, 1988), that selection of fish on the basis of stress responsiveness is practicable. However, as yet, no thorough evaluation of the performance of fish selected for divergent stress responsiveness exists. The present study was carried out as one of a number of preliminary elements of a selective breeding programme, the aim of which is to generate lines of rainbow trout (Oncorhynchus mykiss) which are divergent for stress-responsiveness. If successful, this will allow the relative performance of fish with high- and low-responsiveness to stressors to be compared and permit an assessment to be made of which trait, if either, is desirable in fish under aquaculture conditions.

The changes in blood cortisol levels which occur in fish following the perception of a stressful stimuli are widely employed as a reliable index of activation of the neuroendocrine stress response (Barton and Iwama, 1991). Furthermore, cortisol itself is closely linked to many of the adverse consequences of chronic stress including effects on growth (Pickering, 1993), reproduction (Pickering et al., 1987), and the immune system (Pickering and Pottinger, 1989). Consequently, the extent of cortisol elevation following exposure to a standardised stressor has been adopted as a trait of physiological significance on which a selection procedure may be based (Fevolden et al., 1991; Pottinger et al., 1994). Blood glucose levels are also increased in fish following exposure to stressful stimuli. In the short-term, hyperglycaemia is mediated primarily as a consequence of the rapid increase in blood catecholamines which occurs during the first phase of the stress response (Wendelaar Bonga, 1997) although in the longer-term, cortisol-dependent gluconeogenesis may be responsible (van Raaij et al., 1996). The measurement of post-stress blood glucose levels therefore offers an indirect alternative marker of the magnitude of the neuroendocrine stress response. The measurement of blood glucose can be carried out rapidly using portable equipment (Morgan and Iwama, 1997), in contrast to the more sophisticated facilities required for the determination of cortisol. This is an important consideration when assessing the practicality of employing either selection trait within a commercial environment.

The aim of this study was to compare the utility of both plasma cortisol and plasma glucose levels as selection markers for stress responsiveness and to assess the reliability of each as identifiers of consistency in stress-responsiveness with time. Because uncertainty exists regarding the relative benefits of a high or low responsiveness to stressors, individuals with divergent responses were sought to permit comparison.