Cyclic gas-exchange in the Chilean red cricket: inter-individual variation and thermal dependence

In animals, energy metabolism represents the `intensity' of functioning in a given situation. The rate of metabolism (MR), usually measured by the rate of consumption of O₂ or production of CO₂, represents the in vivo energy metabolism of an individual animal. Such an informative state variable has long attracted the interest of physiologists,who have measured the rate of metabolism in thousands of animals in the search for adaptive patterns (McNab,2002). In a single record, whenever the MR pattern has consistently deviated from a relatively constant time-series with non-periodic fluctuations, investigators have wondered about a mechanistic description and have mostly found an adaptive explanation. For instance, sudden reductions in metabolic rate over days or weeks in small birds and mammals were demonstrated to represent periods of physiological dormancy (i.e. torpor and hibernation),which have adaptive significance as an energy-saving strategy (e.g. Barclay et al., 2001). Total interruptions in MR in several invertebrates were suggested to be adaptations for desiccation (Alpert, 2006). Drastic increases in MR capacity during cold exposure in small endotherms were shown to be a significant adaptation for enduring winter (e.g. Klaasen et al., 2002). Sharp increases in invertebrate MR after feeding were shown to be a direct consequence of digestive energy costs, with adaptive significance for nutrient absorption (Bradley et al.,2003).

The complex patterns of gas exchange in some insects are an exception:their adaptive significance is still unknown. The respiratory patterns of insects have been classified into three main kinds: continuous gas exchange,cyclic gas exchange (CGE) and discontinuous gas-exchange cycles (DGC)(Gibbs and Johnson, 2004). CGE in insects is similar to other animals in that a continuous trace of CO₂ production is recorded, with minor variation due to activity and/or random noise. CGE is characterized by periodic bursts of CO₂production, followed by valleys of low CO₂ production. In contrast,DGC is characterized by bursts of CO₂ production interspersed by zero exchange. It has been shown that these phases are the response to successive openings of the spiracle (Chown and Nicolson, 2004 and references therein). Hence, the peaks of CO₂ production are named the O- (`open') phase, and the valleys are known as the C- (`closed') phase. In some cases, before to the O-phase a small CO₂ release is detected, which is known as the F- (`flutter')phase, referring to intermittent openings of the spiracle(Chown and Nicolson, 2004).

Interest in CGE and DGC in insects has been considerably unbalanced in terms of insect diversity. According to Chown et al.(Chown et al., 2006), the great majority of studied species are beetles and other holometabolous insects. The interesting view presented by Marais et al.[(Marais et al., 2005), p. 4505] suggests that few species exhibit DGC, and all of these also present CGE and continuous gas exchange. According to this phylogeny, DGC appears to be a recent acquisition in insects, having emerged independently at least five times in distantly related groups. One of these appearances is within the orthoptera clade, mainly represented by orthopters(Marais et al., 2005). Here we describe for the first time the cyclic respiratory patterns of a member of the Stenopelmatidae family (king crickets, Jerusalem crickets, wetas and long-horned grasshoppers), the Chilean red cricket Cratomelus armatus, and its inter-individual variation together with its thermal dependence.

Five hypotheses for the evolutionary origin of CGE and DGC have been proposed; in brief: (1) the `hygric hypothesis' (CGE and/or DGC represent an adaptation to xeric environments, limiting evaporative water loss), (2) the`chthonic-hygric' hypothesis (CGE and/or DGC represent an adaptation to hypoxic/hypercapnic environments, at the same time limiting evaporative water loss), (3) the `oxidative damage' hypothesis (CGE and/or DGC represent a strategy to reduce the oxidative damage during periods of reduced oxygen demand), (4) the `emergent property' hypothesis (CGE and/or DGC represent a non-adaptive epiphenomenon of the interaction of two feedback system regulating gas exchange with minimal demand) and (5) the `strolling arthropod'hypothesis (CGE and/or DGC are an adaptation to increase the frequency of spiracle closure to reduce the risk of parasitic infestation) (for details,see Chown et al., 2006), but none of these hypotheses appear to have received consistent experimental support and the debate seems to be focussed on the first two(Chown and Nicolson, 2004; Gibbs and Johnson, 2004; Marais et al., 2005).

In the present work, we have studied the respiratory patterns of the Chilean red cricket Cratomelus armatus, a fossorial species that inhabits humid forests in the Southern hemisphere. According to the`chthonic-hygric' hypothesis, we predict that this species will exhibit DGC and/or CGE in order to survive the hypercapnic/hypoxic environments that it inhabits. To be a target of natural selection, a trait should exhibit inter-individual variation or repeatability. Consequently, we predict that the metabolic patterns of this species will exhibit significant repeatabilities(see also Chappell and Rogowitz,2000; Marais et al.,2005; Chown et al.,2006).

The Chilean red cricket Cratomelus armatus Blanchard is a robust endemic cricket that lives in central-south Chile, in humid forests. It is found below dead logs, where it builds galleries in moist soil with its strong mandibles, feeding on a great variety of vegetal and animal material (i.e. it has an omnivorous diet). It is of solitary and nocturnal habits, exhibiting conspicuous aggressiveness and cannibalism when living together, especially individuals of the same sex and size. Adults have small (non-functional) wings and do not have a stridulating mechanism. According to Angulo(Angulo, 2001), the growth of the Chilean red cricket consists of five instars with considerable variation in adult size (ca. 20–30 mm long, 1.5–2.5 g), males being smaller than females.

We captured 35 crickets ranging from 0.2 to 2 g (representing all stages)in Valdivia (39°48′S, 73°14′W) and housed them individually in plastic cages filled with 3 cm of humid litter. Crickets were fed daily with rabbit chow and water ad libitum, and maintained at 18°C under a natural photoperiod.

Our respirometry system was similar to that described elsewhere(Lighton and Turner, 2004; Chappell and Rogowitz, 2000). In brief, CO₂ production was measured continuously using an infra-red CO₂ analyzer (LI-COR LI6262, Lincoln, NV, USA) capable of resolving differences of 1 p.p.m. of CO₂ in air. The analyzer was calibrated periodically against a precision gas mixture. Although there was almost no drift between calibrations, we performed baseline measurements before and after each recording. Flow rates of dry, CO₂-free air(ambient air scrubbed free of water vapour and CO₂ using a Drierite/soda lime column) were maintained at 100 ml min^–1±1% by a Sierra mass flow controller (Henderson, NV, USA). We used cylindrical Sable Systems (Las Vegas, NV, USA) precision metabolic chambers (8 ml) covered by metal paper, associated with an AD-1 linear activity unit(Sable Systems) to record movement of the crickets by monitoring fluctuations in reflected infrared light at ca. 900 nm. Additionally, crickets were visually monitored at intervals of ca. 5 min during measurements. Each cricket was measured over at least a 1 h period. Each record was automatically transformed by a macro program recorded in the Expe Data software (Sable Systems), in order to (1) correct the 8-s lag introduced by the distance between the analyzer and the chamber and then to match the activity record with the V̇_CO2record, (2) to transform the measure from p.p.m. to ml CO₂h^–1, taking into account the flow rate, (3) to eliminate the first 5 min records (300 samples). To evaluate the effect of ambient temperature on V̇_CO2 and CGE,metabolic rate was measured at 15°C, 20°C and 25°C (using three independent groups of crickets). To determine inter-individual variation in V̇_CO2 and CGE(i.e. its time-consistency or repeatability), we repeated the measurements at 20°C, 15 days later (with the same crickets).

Given that all metabolic variables were correlated among them, the effect of ambient temperature on metabolic variables was assessed by multivariate analysis of covariance (MANCOVA) using body mass M_b as covariate, and Tukey post-hoc test in each response variable. To fulfil ANOVA assumptions, data were log_e-transformed. Repeatability was computed by the intraclass correlation coefficient (τ) by calculating the between- and within-individual variance component from one-way ANCOVAs(M_b as covariable). Then, τ=(between individuals variance component)/(within individual variance component). In some cases(ventilatory period, O-phase and interburst), several measurements were taken within each record, which were included as a nested factor. Hence, in addition to the between- and within-individual variance components, for these variables we also computed the between records-within individual variance,component.

The pattern of V̇_CO2 observed in the Chilean red cricket was cyclic in 90% of the records,(Fig. 1, Fig. 2). During CGE, no cricket exhibited any movement within the chamber other than ventilatory movements,most of which (especially in large individuals) were also detected by the motion infrared device. According to the general pattern of V̇_CO2, in order to characterize CGE, five variables were extracted from the records: the O-phase period or duration (from the spectral analysis), the ventilatory period,average V̇_CO2,O-phase amplitude and interburst amplitude V̇_CO2(Table 1). The only variable that did not exhibit body mass dependence was O-phase duration(Fig. 3).

There were significant effects of ambient temperature on metabolic variables (Wilk's lambda=0.166; F_10,20=2.91; P=0.020). Tukey's post-hoc test revealed that the O-phase duration was inversely affected by ambient temperature(Fig. 4). There were non-significant effects of temperature on ventilatory period(Fig. 4). The amplitude of average and O-phase increased with temperature whereas the amplitude of the interburst changed only between 15° and 20°C(Fig. 5).

Both ventilatory and O-phase periods presented significant repeatabilities(Table 2). Also, the within-record component was significant for the ventilatory period, indicating that this variable is remarkably consistent over time. On the other hand,average, O-phase amplitude and interburst amplitude did not exhibit significant repeatability at the individual level, although the mean V̇_CO2 and the interburst amplitude showed moderate repeatability(Table 2). The amplitude of the O-phase exhibited significant within-individual component of variation,however (Table 2).

Historically, the most studied aspects of king crickets, Jerusalem crickets, wetas, long-horned grasshoppers and other Stenopelmatidae are the population biology (Leisnham and Jamieson,2002; McCartney et al.,2006), taxonomy/phylogeography(Trewick et al., 2000; Morgan-Richards and Wallis,2003; Jost and Shaw,2006), reproduction (Bateman,2000; Gwynne,2004), stridulating mechanisms(Field and Roberts, 2003) and physiology of freezing tolerance (Wharton et al., 2000). Thus it is surprising to find that this is not only the first report of cyclic gas exchange (CGE) in this family, but the first study of any aspect of its metabolic physiology.

The pattern of CGE exhibited by C. armatus in many cases presents the C-, F-, O-phases with considerable intraspecific variation. Actually shifts to continuous gas exchange in the same individuals and even in the same record are common, as in most insects that exhibit DGC or CGE(Marais et al., 2005).

It has been suggested that DGC should be a lot more common in insects than actually reported because low flow rates combined with large metabolic chambers would prevent records falling to zero during the C-phase due to incomplete washout (Gray and Bradley,2006). We believe this is not the case with our data, since the combination of chamber size (8 ml) and flow rate (100 ml min^–1) that we used gives a time constant of 0.08 min. The time required for the partial pressure of CO₂ in the chamber to be diluted to 1% of its peak value is 5 times its time constant (i.e. 24 s)(Gray and Bradley, 2006). Considering that in most cases the duration of the interburst in our CGE records ranged from 5 to 25 min, and in no case did V̇_CO2 reach zero,it is unlikely that C. armatus would be experiencing DGC and we were not capable of detecting it.

The components of CGE that we studied were similar to those reported in previous studies of DGC/CGE in insects(Chappell and Rogowitz, 2000; Marais and Chown, 2003) since although cyclic records were clearly identified, they exhibited considerable variation, from showing the three phases, to only showing peaks and valleys that never reached zero (i.e. the spiracle was never completely closed). The O-phase was characterized by conspicuous ventilatory movements (i.e. the O-phase is predominantly convective) that produced high frequency cycles,which yielded highly consistent results over time across individuals.

As reported in other species (and expected in an ectotherm), the amplitude of the variables that we determined are strongly temperature-dependent(Chappell and Rogowitz, 2000; Terblanche et al., 2004a; Terblanche et al., 2004b). However, there is variation in the duration of the O-period. In most species,a higher ambient temperature induces an increase in the frequency of cycles(or a reduction in the O-period) (Lighton,1988; Quinlan and Lighton,1999; Voght and Appel,2000; Chappell and Rogowitz,2000). In many species, however, there is no dependence of the O-period or volume on temperature (Davis et al., 1999; Shelton and Appel,2001). Our results suggest an intermediate response since the O-period was considerably reduced at the lowest temperature (15°C), but changed little between moderate to high temperatures (20–25°C). It is striking to note that no study has determined the thermal dependence of the ventilatory period in those cases of convective O-phase, especially considering that convective gas-exchange is a necessity for large insects(Kestler, 1985). This variable would be an indication of the effort required by the insect to flush out excess CO₂ during the open phase, and hence could be important for survival. In addition, as with other convective systems such as vertebrate lungs, as the size of the chamber increases (i.e. the volume of the tracheal system), the duration of the cycle is also extended, which is consistent with the significant correlation with body mass that we found.

Although many authors have demonstrated that insect standard metabolic rate is repeatable (Chappell and Rogowitz,2000; Nespolo et al.,2003; Marais and Chown,2003; Terblanche et al.,2004a), we could not do so for three of its components (i.e. average V_CO2, O-phase and interburst amplitudes). However,our low sample size precluded any statistically significant conclusion regarding the absence of repeatability (i.e. repeatabilities <0.4 are not detectable with a sample size of N=7). In contrast, and even with our relatively low sample size, the repeatability of both ventilatory and O-phase duration was high and significant. Our O-period (estimated by spectral density analysis) was comparable to the O-phase duration computed (from visual inspection) in cerambycid beetles (Chappell and Rogowitz, 2000), or the several variants of the O-period reported for a cockroach (Marais and Chown, 2003) using a simple visual test(Marais et al., 2005). In all three cases, repeatabilities were the largest among the computed metabolic variables in each study, which suggests some generality regarding the inter-individual variation in the O-phase and perhaps its potential to respond to natural selection (Dohm,2002; Chown et al.,2006).

According to Chown et al., there are three requisites necessary to demonstrate the adaptive significance of a trait: (1) that it exhibits significant repeatability; (2) that there exists a consistent relationship between the trait and fitness; and (3) that it exhibits high heritability(Chown et al., 2006).

In the present study, we show that the first condition is met, since at least two of the components of metabolism in C. armatus were highly repeatable (see also Chappell and Rogowitz,2000; Nespolo et al.,2003; Marais and Chown,2003). Given that C. armatus is fossorial and lives in humid soils of temperate forests in the Southern hemisphere(Angulo, 2001), our results would support the concept of adaptation to a hypercapnic/hypoxic environment[the `chthonic-hygric' hypothesis (see Lighton, 1996; Chown et al., 2006)]. This statement could be a sign that the putative relationship between the trait and fitness could exist, and hence provide some support for the second requisite to infer adaptation. However, there are actually no data with which to assess the third requisite (i.e. the trait presents high heritability). So it is clear that although the many published studies to date have been useful in characterizing the cyclic mode of respiration in insects and other arthropods,its phylogenetic relationships and its mechanistic causes, further research is needed in order to answer the simple question: is CGE/DGC an adaptive feature of insects?

 
• CGE

  cyclic gas exchange

 
• C-phase

  `closed' phase

 
• DCG

  discontinuous gas-exchange cycle

 
• F-phase

  `flutter' phase

 
• M_b

  body mass

 
• MR

  rate of metabolism

 
• O-phase

  `open' phase

 
• SDE

  spectral density estimation

 
• t

  time

 
• V̇ _CO2

  rate of CO₂ production

 
• τ

  intraclass correlation coefficient

 
• λ

  frequency

This work was supported by `Proyecto Anillos de Ciencia y Tecnología ACT-38' to R.F.N. We thank two anonymous reviewers, Mark Chappell and John Lighton for technical assessment on the analysis and interpretation of our records. Luis Castañeda and Paulina Artacho acknowledge a CONICYT and MECESUP fellowship, respectively.