The effect of hypoxia on fish schooling

Abstract

Low-oxygen areas are expanding in the oceans as a result of climate change. Work carried out during the past two decades suggests that, in addition to impairing basic physiological functions, hypoxia can also affect fish behaviour. Given that many fish species are known to school, and that schooling is advantageous for their survival, the effect of hypoxia on schooling behaviour may have important ecological consequences. Here, we review the effects of hypoxia on school structure and dynamics, together with the mechanisms that cause an increase in school volume and that ultimately lead to school disruption. Furthermore, the effect of hypoxia generates a number of trade-offs in terms of schooling positions and school structure. Field observations have found that large schools of fish can exacerbate hypoxic conditions, with potential consequences for school structure and size. Therefore, previous models that predict the maximum size attainable by fish schools in relation to oxygen levels are also reviewed. Finally, we suggest that studies on the effect of hypoxia on schooling need to be integrated with those on temperature and ocean acidifications within a framework aimed at increasing our ability to predict the effect of multiple stressors of climate change on fish behaviour.This article is part of the themed issue 'Physiological determinants of social behaviour in animals'.

Keywords: climate change; environmental stressor; fish behaviour; fish physiology; group behaviour; group living.

Figure 1.

Metabolic rate and scope as a function of water temperature and the modulatory effect of hypoxia. Hypoxia reduces the maximum metabolic rate (MMR) of individual fish but it does not affect the standard metabolic rate (SMR). The reduction in the MMR induced by hypoxia causes an overall reduction in the MS. Largely based on Pörtner & Farrell [13].

Figure 2.

The hydrodynamic benefit of individual fish swimming in a school. (a) The scheme of a school of fish swimming in a diamond formation based on the hypothesis by Weihs [71]. (b) The propulsive motion of the two fish swimming within the diamond formation (grey filled) generates a reverse von Karman vortex street which can be exploited by the downstream individual (black filled) if it is positioned between the vortex streets produced by the two individuals swimming ahead. From Liao [68] with permission from Royal Society Publishing.

Figure 3.

The hydrodynamic advantages of swimming in a school. (a) Positions of the focal fish (red dot) relative to its closest neighbour. For positive values, the focal fish was ahead of its closest neighbour; at a value of zero, the focal fish was swimming next to its closest neighbour; and for negative values, the focal fish was swimming behind its closest neighbour. (b) Reduction in the TBF (%), compared with swimming alone, for the focal fish when swimming in different positions in a school. Bar values refer to the position represented in panel a. (c) Reduction (in %, compared with solitary fish) in the estimated metabolic rate (MO₂) of a focal fish swimming in various positions relative to its closest neighbour in the school. Values are mean ± s.e. From Marras et al. [78] with permission from Springer.

Figure 4.

Behavioural observations of schooling striped surfperch exposed to a progressive hypoxia protocol. (a) Swimming speed, (b) spontaneous turning rate, (c) angular correlation. In hypoxia, speed and turning rate are significantly reduced and angular correlation is significantly increased. Further data (not shown) suggest no effect of hypoxia on school polarity. (d) (i, ii, iii) Schematic of the angular correlation of turning behaviour (r) and the net displacement (s) of swimming individuals over an arbitrary period of time. High values of r (ii, iii) indicate uniform turning behaviours, while low values of r (i) represent random and erratic turning behaviours. (d) (iv, v) The polarity of the school is assessed by the value of circular variance (Φ). High values of (Φ) indicate a school with low polarity, while low values of (Φ) indicate a highly polarized school. From Cook et al. [90] with permission from Elsevier.

Figure 5.

Schooling can cause a reduction in oxygen level. (a) Schematic of field data showing that oxygen levels tend to decrease along the axis of motion of the school as a result of the oxygen consumption by the fish at the front of the school. Oxygen levels decreased from about 7.2 mg l⁻¹ at the front of a 150 m long school, to about 5.2 mg l⁻¹ at the rear of the school. This corresponds to about 28% reduction in oxygen levels. From McFarland & Moss [65]. (b) Reduction in oxygen levels within a school of mullet swimming at 30 cm s⁻¹. The relationship between the distance from the front of the school (X) and its total length (length) and the ratio between the oxygen levels measured at a certain distance from the front of the school (C) and the oxygen levels outside the school (C₀) is shown. Data are based on McFarland & Moss [65] and the lines are based on the model by McFarland & Okubo [103]. Relative position within the school ranges from 0 (front of the school) to 1 (back of the school). When X > length, the effect of oxygen diffusion induces a quick recovery of the oxygen levels to the values measured ahead of the school. Panel (a) from McFarland & Moss [65] reproduced with permission from the American Association for the Advancement of Science (AAAS). Panel (b) redrawn from McFarland & Okubo [103].