The physiological basis of hypoxia avoidance behaviours in fish

Abstract

Aerobic respiration is an essential component of vertebrate life. However, fish occupy an aquatic environment that possesses low and fluctuating O2 conditions as a result of numerous physical, chemical and anthropogenic forcings. An understanding of how fish respond both physiologically and behaviourally to low O2 conditions (hypoxia) is therefore an important component of fish biology. Kingfish (Seriola lalandi) demonstrate a moderate ability to tolerate hypoxic conditions, and show no apparent changes in behaviour during progressive hypoxia exposures until they experience inescapable and severe hypoxia at a 4 kPa oxygen partial pressure (PO2). This change in behaviour during inescapable hypoxia involves a marked agitation response and is associated with elevated levels of physiological stress (including cortisol, lactate and glucose). Interestingly, kingfish show no changes in swimming behaviour (i.e. gait and speed) or evidence of physiological stress when exposed to severe (4 kPa), but escapable hypoxia. With no evidence of physiological stress upon exposure to escapable hypoxia it is suggested that they are not required to change their behaviour or avoid severely low O2 conditions if they are able to mitigate hypoxic stress with regular forays into normoxic environments. The suggestion that physiological stress may initiate hypoxia avoidance behaviours is supported by evidence of snapper (Pagrus auratus) avoiding hypoxia at 3.1 ± 1.2 kPa PO2, an oxygen level that is below their critical O2 partial pressure (Pcrit) of 5.8 ± 0.6 kPa PO2 at 18°C. Indeed, avoidance only occurs when snapper experience significant anaerobic and glycolytic stress. However, this stress and perturbation is not associated with physiological fatigue because metabolic fuel reserves were not depleted, identifying that snapper can tolerate severely low O2 conditions for a limited period of time. In one investigation, snapper showed a significant and subtle reduction in swimming speed prior to avoiding hypoxic conditions. This energy-saving response may have helped snapper minimise the physiological challenge of low O2 residence, but since this response was not seen in two other studies this behavioural response is only deemed to have minor importance to the behavioural low O2 avoidance strategy of snapper. Ultimately, these results identify that physiological limits underpin the behavioural low O2 tolerance of fish and that the physiological stress experienced upon approaching these physiological limits may initiate hypoxia avoidance behaviours. The physiological state of snapper was manipulated experimentally to identify whether the hypoxic tolerance and associated avoidance behaviours of snapper are determined by aerobic and anaerobic physiological characteristics of the species. The aerobic capacity of snapper was modified using a haemolytic agent (phenylhydrazine) that induced anaemia and their response was compared against a sham control. The haemolytic anaemia treatment consistently reduced the maximum metabolic rate and aerobic scope at all levels of water PO2 investigated which raised the Pcrit limit from 5.3 ± 0.4 kPa to 8.6 ± 0.6 kPa. This physiological modification also resulted in the behavioural threshold for hypoxia avoidance shifting from 2.9 ± 0.5 kPa in untreated normocythaemic fish, to 6.6 ± 2.5 kPa in anaemic individuals. Each treatment therefore avoided hypoxia just below their Pcrit, and were subject to equivalent levels of physiological stress (plasma lactate and glucose). Using a different form of experimental manipulation, snapper preconditioned to hypoxia over 6-12 weeks at O2 levels between 10.2-12.1 kPa show improvements in their hypoxic tolerance. This increased tolerance was characterised by hypoxia preconditioned fish avoiding lower PO2 levels than normoxia preconditioned fish (3.3 ± 0.7 vs 5.3 ± 1.1 kPa, respectively) without displaying greater perturbations of lactate, glucose and blood pH. The increased tolerance shown by hypoxia preconditioned fish was not associated with changes in Pcrit, which were unchanged at ~ 7 kPa, as were standard, routine and maximum metabolic rates. Moreover, blood O2 carrying capacity was not increased as expected because haemoglobin concentrations were decreased following hypoxia-acclimation. Acclimation to hypoxic conditions only resulted in a few subtle physiological changes that appeared to underlie a significant improvement in low O2 behaviour. The PO2 at which haemaglobin was half saturated (i.e. the P50) was elevated unexpectedly, perhaps facilitating Hb-O2 off-loading to tissues. The respiration of cardiac mitochondria as measured in situ in permeabilised fibres also showed improved O2 uptake efficiencies. However, changes in aerobic function were not the only drivers of behavioural adjustment as the proportion of the anaerobic metabolic enzyme lactate dehydrogenase was increased relative to the aerobic marker enzyme citrate synthase in heart and skeletal red muscle. By exposing fish to two contrasting experimental manipulations it has been clearly identified that physiological changes can significantly influence the hypoxia avoidance behaviours of fish. Throughout this thesis snapper consistently showed a characteristic release of lactate from the cell into the primary circulation prior to avoiding low O2 conditions. Whether this release of lactate acts as an indicator of cellular low O2 stress capable of motivating the hypoxia avoidance behaviours of fish was therefore investigated. It was demonstrated that the artificial infusion of lactate into the primary circulation caused fish to spend a significantly greater proportion of time in low O2 (5 kPa) than shaminfused individuals which strongly avoided the same low O2 conditions. These unexpected results raise the possibility that circulating lactate might actually assist aerobic metabolism in patchy low O2 conditions, a finding that contrasts markedly to those of other studies that suggest lactate triggers emergency behavioural responses in both vertebrates and invertebrates (De Wachter et al., 1997; Portner et al., 1994). Despite my attempts, the physiological trigger of hypoxia avoidance behaviour was not identified. This remaining unknown provides an intriguing avenue for future research. The Pcrit thresholds of 5.8 ± 0.6 and 7.0 ± 0.6 kPa PO2 at 18 and 21°C, respectively, identify the environmental survival limits of snapper when exposed to severe and acute hypoxic challenges. These O2 levels represent the lethal limits for snapper and do not consider the sub-lethal effects that will be experienced at less severe levels of hypoxia. I therefore consider that the future success of snapper and other coastal marine fish species of New Zealand would depend upon their ability to undergo geographical shifts away from waterways that experience low O2 conditions.