Royal Society Publishing

A test of the oxidative damage hypothesis for discontinuous gas exchange in the locust Locusta migratoria

Philip G. D. Matthews, Edward P. Snelling, Roger S. Seymour, Craig R. White


The discontinuous gas exchange cycle (DGC) is a breathing pattern displayed by many insects, characterized by periodic breath-holding and intermittently low tracheal O2 levels. It has been hypothesized that the adaptive value of DGCs is to reduce oxidative damage, with low tracheal O2 partial pressures (PO2 ∼2–5 kPa) occurring to reduce the production of oxygen free radicals. If this is so, insects displaying DGCs should continue to actively defend a low tracheal PO2 even when breathing higher than atmospheric levels of oxygen (hyperoxia). This behaviour has been observed in moth pupae exposed to ambient PO2 up to 50 kPa. To test this observation in adult insects, we implanted fibre-optic oxygen optodes within the tracheal systems of adult migratory locusts Locusta migratoria exposed to normoxia, hypoxia and hyperoxia. In normoxic and hypoxic atmospheres, the minimum tracheal PO2 that occurred during DGCs varied between 3.4 and 1.2 kPa. In hyperoxia up to 40.5 kPa, the minimum tracheal PO2 achieved during a DGC exceeded 30 kPa, increasing with ambient levels. These results are consistent with a respiratory control mechanism that functions to satisfy O2 requirements by maintaining PO2 above a critical level, not defend against high levels of O2.

1. Introduction

Many insects at rest breathe using the discontinuous gas exchange cycle (DGC): an alternating pattern of gas exchange and breath-holding. This behaviour is known to occur among at least five [1], and possibly as many as seven [2], insect orders. It is characterized by the sequential repetition of three phases: a ‘closed phase’ when the spiracles are shut and tracheal PO2 drops continuously while PCO2 rises; a ‘flutter phase’ when the spiracles rapidly open and close releasing minimal CO2 while admitting some O2 to the tracheal system; and an ‘open phase’ when both CO2 and O2 move freely between the atmosphere and tracheal system. Large fluctuations in tracheal PO2, varying from near atmospheric (approx. 20 kPa) to severely hypoxic (2–5 kPa), are a peculiar characteristic of this pattern. The potential adaptive significance of the DGC has led to much debate, with suggestions that it is adaptive to reduce oxidative damage (oxidative damage hypothesis [3]), to reduce respiratory water loss (hygric hypothesis, [4]), or as a consequence of neural downregulation affecting respiratory control (neural hypothesis, [5]). The hygric and neural hypotheses both assume that the low tracheal PO2 during the closed and flutter phases occurs inadvertently, either as a consequence of protracted spiracle closure reducing respiratory water loss, or due to a change in the behaviour of the insect's respiratory control system. The oxidative damage hypothesis is based on the assumption that the DGCs primary function is the production of a low tracheal PO2 during the closed phase that is then maintained during the flutter phase [3,68].

Oxygen is a toxic molecule, capable of damaging tissues through the production of reactive oxygen species (ROS) [9]. According to the oxidative damage hypothesis, DGCs evolved as a mechanism to reduce tracheal PO2 periodically when the insect's metabolic rate is low, thereby lowering ROS production. The respiratory behaviour of moth pupae supports this idea, as they regulate their tracheal PO2 at approximately 5 kPa during the flutter phase of their DGC, even in hyperoxic atmospheres up to 50 kPa [3]. While the hygric and neural hypotheses do not require the maintenance of a low PO2 during DGCs, the question remains: is a low and constant tracheal PO2 during the flutter phase a universal feature of DGCs displayed by adult insects? Pupae are physiologically atypical, as they represent a life stage characterized not only by a very low metabolic rate [10], but also reduced neurological functions [11].

To determine whether tracheal PO2 during the flutter phase of a DGC is independent of ambient PO2 in adult insects, we implanted fibre-optic oxygen optodes into the tracheal systems of adult locusts exposed to hyperoxic, normoxic and hypoxic atmospheres.

2. Material and methods

Adult migratory locusts Locusta migratoria (810 ± 72 mg) were reared at the University of Adelaide. Locusts were maintained in plastic terraria, at 33 ± 1°C, approximately 30 per cent RH, 12 L:12 D cycle and ad libitum food. They were fasted 24 h prior to gas exchange measurements and weighed to 0.1 mg (AE163, Mettler, Greifensee, Switzerland) immediately before each experiment. Flow-through respirometry was performed at 22–23°C using a LI-820 CO2 analyser (Li-COR Biosciences, Lincoln, NE, USA) and gas mixes produced by mass flow controllers (Aalborg 0–500 ml min−1 and 0–1000 ml min−1) connected to cylinders of compressed O2 and N2. Gas mix composition was verified using a calibrated Oxzilla II oxygen analyser (Sable Systems, Nevada, USA).

Two groups of locusts were measured: a control group and a treatment group. Individuals in the control group were placed within 10 ml syringe barrels and gas exchange was measured over a 2–8 h period following 1 h of acclimation. They were measured in normoxia (n = 9) using outside air scrubbed of CO2 and water vapour, as well as hypoxia (7 kPa; n = 4) and hyperoxia (40 kPa; n = 4). Briefly, the gas mix was regulated at 350 ml min–1 using a mass flow controller, passed through the syringe containing the resting locust, through a small column of desiccant, and into the CO2 analyser. The analogue outputs of the mass flow controller and CO2 analyser were recorded at 1 s intervals to a computer with a PowerLab data acquisition system and LabChart software (ADInstruments, Bella Vista, NSW, Australia).

To measure tracheal PO2 and gas exchange pattern simultaneously, treatment locusts (n = 3) were affixed by their thorax to the lid of a rectangular 138 ml respirometry chamber using melted dental wax (Ainsworth Dental Company Pty. Ltd., Marrickville, NSW, Australia). This positioned the locust right-way-up within the chamber. A 25 gauge hypodermic needle was used to pierce the dorsal surface of the prothorax through a 5 mm hole in the lid of the chamber. The air sac below this incision was then punctured using the hypodermic needle. A micromanipulator (World Precision Instruments, USA) was used to insert a flat-tipped, 140 µm diameter oxygen optode connected to an oxygen meter (TX3, PreSens GmbH, Germany) through the hole in the respirometer's lid, and into the locust's air sac. Polyvinylsiloxane dental impression material (President, Coltène Whaledent, Switzerland) was used to fill the hole in the lid of the respirometry chamber and create an air-tight seal between the optode and the locust's cuticle. Following this, a gas mixture of 21.3 kPa PO2 was passed through the chamber at 450 ml min−1 and into the CO2 analyser. A minimum of 1 h was given before measurements began. Locusts were exposed to gas mixes containing 5.1, 10.1, 15.2, 21.3, 30.4 and 40.5 kPa PO2.

All mean values are presented ± 95% confidence interval. The effect of different PO2 treatments on DGC duration was analysed using Kruskal–Wallis one-way analysis of variance with Dunn's post hoc test. Differences in mean DGC duration between control and optode-implanted locusts were analysed using unpaired t-tests. While direct comparisons could be made of mean DGC duration between control and optode-implanted locusts at 40.5 and 21.3 kPa, it was necessary to combine 5.1, 10.1 and 15.2 kPa DGC data from the optode-implanted locusts to compare against control locusts measured at 7.1 kPa. All statistical analyses were carried out using GraphPad Prism v. 5 software (GraphPad Software, La Jolla, CA, USA; electronic supplementary material available online).

3. Results

Locusts displayed both continuous and discontinuous gas exchange patterns. During continuous ventilation, tracheal PO2 matched ambient levels from hypoxia across hyperoxia, increasing with a slope of 0.94 (figure 1). There was no significant difference between mean DGC duration of optode-implanted and control locusts (p > 0.05). The overall mean DGC duration for all treatments was 29.8 min. The only significant effect of ambient PO2 on mean DGC duration was found between 21.3 kPa (33.2 ± 4.1 min) and 40.5 kPa PO2 (18.3 ± 2.4 min; p < 0.05). The reduced duration in 40.5 kPa is attributable to a decrease in the closed and open durations and the virtual elimination of the flutter phase (table 1 and figure 2). The lowest tracheal PO2 during a DGC occurred at the transition from the closed to the flutter phase (figure 2). This minimum PO2 varied between 3.41 ± 0.43 kPa in normoxia, dropping slightly to 3.13 ± 1.52 and 1.21 ± 0.22 kPa in atmospheres between 15.2 and 5.1 kPa, respectively (figure 1). In hyperoxia, however, the minimum tracheal PO2 increased with ambient PO2 with a slope of 1.38, rising to 17.66 ± 1.56 kPa in 30.4 kPa O2, and 30.10 ± 1.21 kPa in 40.5 kPa O2 (figure 1).

View this table:
Table 1.

Effect of ambient PO2 on the mean closed, flutter and open-phase durations (± 95% CI) of DGCs displayed by adult locusts.

Figure 1.

Intratracheal PO2 (± 95% CI) in locusts breathing continuously (open circles) and the minimum tracheal PO2 in locusts displaying DGCs (filled circles), during exposure to hypoxia and hyperoxia. Tracheal PO2 in Atticus atlas pupa during the flutter phase (filled diamonds) are included for comparison (data from Hetz & Bradley [3]). The solid black line indicates unity.

Figure 2.

Discontinuous gas exchange pattern in a 900 mg locust exposed to hypoxic, normoxic and hyperoxic atmospheres, showing CO2 release (V̇co2, black lines) and tracheal PO2 (grey lines). Horizontal dashed lines indicate ambient PO2.

4. Discussion

The minimum tracheal PO2 in adult locusts displaying DGCs is dependent on ambient PO2 in hyperoxia (figure 1). This is in contrast to a tracheal PO2 maintained at low levels in hyperoxia, as expected by the oxidative damage hypothesis [3]. Why then do moth pupae display a consistently low tracheal PO2 in hyperoxia during their flutter phase, when locusts do not? The answer lies in the relative tracheal and body fluid volumes of the two insects. Current research indicates that spiracles flutter open and closed in response to internal hypoxia and open fully in response to high CO2 [12,13]. Owing to its high solubility, most of the CO2 produced during the closed and flutter phases of the DGC is sequestered in the insect's body fluids, while the bulk of the O2 used during the closed phase is obtained from air in the tracheal system [14,15]. Therefore, a flutter phase and its associated low and stable PO2 can occur only if the tracheal O2 store is depleted to a hypoxic threshold before CO2 accumulation in the body fluids initiates an open phase. This is determined by the ratio of the insect's O2 store (tracheal system volume) to its CO2 sink (body fluid volume and buffer capacity).

Moth pupae possess a tracheal system that contains approximately 140 µl of air per gram of body mass, while a locust's tracheal system holds approximately 310 µl g−1 (for references, see the electronic supplementary material). As such, moth pupae have approximately half the volume of O2 on which to draw during the closed phase of their DGC relative to their CO2 sink (body fluid volume; proportional to mass) compared with a locust. Additionally, moth pupae have higher whole body buffer capacity (75 mmol HCO3 pH−1 kg−1) than those of a locust (28–43 mmol HCO3 pH−1 kg−1; electronic supplementary material) and, therefore, have a greater capacity to store CO2 relative to pH change. These differences mean that a pupa in hyperoxia can deplete its tracheal O2 reserve to approximately 5 kPa during the closed phase of its DGC and transition to the hypoxia-initiated flutter phase before it has accumulated sufficient CO2 to initiate an open phase. This accounts for the consistently low tracheal PO2 displayed by moth pupae in hyperoxic atmospheres. In contrast, a locust's substantial O2 reserve in hyperoxia means that its tracheal PO2 cannot drop to the hypoxic flutter threshold during the closed phase before excessive accumulation of CO2 initiates an open phase. This ensures that the locust's tracheal PO2 remains close to ambient levels (figures 1 and 2). We conclude that the apparent independence of tracheal PO2 in moth pupae can be explained as the by-product of a comparatively small tracheal volume coupled with a large CO2 buffer, and the interaction between a hypoxic threshold initiating the flutter phase and a high CO2 threshold initiating the open phase.


This research was supported by the Australian Research Council (project DP0879605).

  • Received February 12, 2012.
  • Accepted March 9, 2012.


View Abstract