Variation in discontinuous gas exchange cycles

Each of the periods in a single DGC can be characterized by several parameters, including duration, gas exchange rate, and volume of gas exchanged. Moreover, the entire cycle can be characterized by its frequency, mean peak height of the O-periods, proportional contributions (volume, duration) of each of the periods, and by the overall respiratory quotient (RQ = CO2/O2) if both gases are measured. Because CO2 is the most straightforward of the gases to measure for small insects in a flow-through system (Lighton 1991b), rates and volumes of each of the periods often refer solely to CO2. However, where both CO2 and O2 are measured (Lighton 1988b), rates and volumes can refer to either of the gases, and short term RQs (i.e. for a single period, and also known as respiratory exchange ratios, RER) can be calculated. Measurements of CO2, O2, and H2O are also sometimes made (Hadley and Quinlan 1993), although combined measurements of CO2 and H2O are more common (Lighton 1992; Lighton et al. 1993b; Quinlan and Lighton 1999; Chown and Davis 2003). In the latter case, molar ratios of CO2 to H2O loss can also be calculated for each of the periods.

Because characterization of DGCs in adult insects, and the effects of mass, temperature, and VCO2 on them, have been the focus of much of the recent literature, mean values for the parameters characteristic of each period and of the DGC as a whole (for one or more temperatures) are generally provided in each investigation. Whether these summary statistics are valuable for understanding the reasons for interspecific differences in DGC (such as the apparent utility of an extended F-period for conserving water—see Lighton 1990; Davis et al. 1999; Duncan and Byrne 2000; Chown and Davis 2003), depends fundamentally on the extent of the variation of the DGC both within and among individuals, compared to that among species. Some recent reports have suggested considerable variation both within and among individuals (Lighton 1998; Chown 2001), indicating that within- and among-individual variation might be almost as large as, or sometimes larger than, that found between species. Clearly, there is much merit in investigating the partitioning of variance in gas exchange parameters, from the within-individual to among-species level, in several monophyletic groups (see Chown 2001 for further discussion). To date, only a single investigation has provided a careful quantification of the contribution of within-and among-individual variation to total variation in the parameters of the DGC (Marais and Chown 2003). In that study it was shown that repeatability (= among-individual variation) in a cockroach species characterized by at least four different gas exchange patterns (see Fig. 1.2) is high and generally significant, and that size-correction prior to estimates of repeatability should not be undertaken.

The work by Marais and Chown (2003) has also provided an indication of the correct level at which statistical analyses of the variation in gas exchange parameters should be undertaken. In several studies, investigations of relationships between various characteristics of the DGC are undertaken using measurements for each cycle of a gas exchange trace for a given individual. Thus, if seven individuals are investigated, and the parameters of four cycles are measured in each individual, the sample size is given as 28. If within-individual variation in DGC parameters was much greater than that between individuals, then this procedure might not provide any cause for concern. However, Marais and Chown (2003) demonstrated that variation within individuals is quite low, and this also seems to be most commonplace in the published literature. Therefore, statistical analyses undertaken using single cycles as independent data points are flawed. Not only will the degrees of freedom for the analysis be overestimated, leading to an increase in the likelihood of Type I statistical errors, and hence erroneous conclusions, but also estimates of the variability of gas exchange cycles will be confounded. Therefore, either mean values for each of the parameters for each individual animal should be calculated and used in analyses (Lighton and Wehner 1993; Lighton and Garrigan 1995), or variation within and among individuals should be fully explored.


Temperature, body mass, and metabolic rate are the three most important factors influencing the DGC and each of its periods, although metabolic rate is clearly also influenced by temperature and body mass. An increase in temperature usually results in an increase in metabolic rate (Section 3.4.6). In all of the species examined to date, this temperature-related increase in metabolic rate is accompanied by an increase in DGC frequency. However, in some species, such as adult Camponotus vicinus (Hymen-optera, Formicidae) (Lighton 1988a), several species of Pogonomyrmex harvester ants (Quinlan and Lighton 1999), the fire ant Solenopsis invicta (Vogt and Appel 2000), adult carabid beetles (Duncan and Dickman 2001), and several lepidopteran pupae (Buck and Keister 1955; Schneiderman and Williams 1955), O-period volume declines with an increase in temperature. In others, such as the ant C. bicolor (Lighton and Wehner 1993), several species of dung beetles (Davis et al. 1999), two species of Phoracantha (Cerambycidae) beetles (Chappell and Rogowitz 2000), and termites (Shelton and Appel 2001b), O-period volume does not change (Fig. 3.12).

At present, it is not clear what mechanisms are responsible for these changes in frequency and volume (Lighton 1988a, 1996). An increase in temperature results both in a decline in CO2 solubility and a decline in haemolymph pH, so affecting haemolymph buffering capacity (Lighton 1996). Thus, the effect of temperature is not restricted to a change in metabolic rate. The change in buffering capacity and increased metabolic demand presumably lead to an increase in DGC frequency and a decline in O-period volume. However, the

0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 VCO2 (mlh-1)

Figure 3.12 Relationships between mean VCO2 and DGC frequency (mean ± SE, circles) and O-period emission volume (mean ± SE, squares) in Scarabaeus westwoodi (Coleoptera, Scarabaeidae).

Source: Physiological and Biochemical Zoology, Davis et al., 72, 555-565. © 1999 by The University of Chicago. All rights reserved. 1522-2152/1999/7205-98157S03.00

0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 VCO2 (mlh-1)

Figure 3.12 Relationships between mean VCO2 and DGC frequency (mean ± SE, circles) and O-period emission volume (mean ± SE, squares) in Scarabaeus westwoodi (Coleoptera, Scarabaeidae).

Source: Physiological and Biochemical Zoology, Davis et al., 72, 555-565. © 1999 by The University of Chicago. All rights reserved. 1522-2152/1999/7205-98157S03.00

maintenance of constant volumes with changing temperature is more difficult to explain. Presumably, species in which volumes remain constant must either be capable of modulating the O-period CO2 setpoint, or of offsetting temperature-related changes in haemolymph buffering capacity. Davis et al. (1999) suggested that the former is the case in the dung beetles they investigated, although their conclusion was not experimentally verified. It has been suggested that there might be an adaptive advantage to the temperature-independence of O-period emission volumes because DGC frequency might not increase at as high a rate as might otherwise be expected, and therefore the exposure of the tracheal system to dry air might be reduced (Lighton 1994). However, Davis et al. (1999) found no interspecific differences in temperature-related modulation of the DGC despite the fact that the species were collected across a broad range of habitats (mesic to xeric). At present, the limited number of investigations of modulation of frequency of the DGC and O-period volume with changes in metabolic rate make any general conclusions unwarranted. Moreover, where temperature has not been used to alter metabolic rate, it has been shown that an increase in metabolic rate is accompanied by both an increase in DGC frequency and an increase in O-period volume (Lighton and Berrigan 1995).

Body size

The scaling of components of the DGC has been investigated in several species. In general, O-period CO2 emission volumes and VCO2 have similar interspecific scaling exponents, resulting in mass independence of DGC frequency (Lighton 1991a; Davis et al. 1999; Chappell and Rogowitz 2000), a situation which is very different to that found in vertebrates (Peters 1983). Scaling of the other components of the DGC tends to vary between taxa, with rates tending to scale positively with body size, durations showing no relationship with size, and the scaling of F and C-period emission volumes being variable in their significance. Fewer investigations of intraspecific scaling have been undertaken. In those species where body mass variation is low, relationships tend to be insignificant as might be expected (Bosch et al. 2000), whereas in other species, where there is a reasonable range in size, both DGC frequency and O-period emission volume increase with mass (Lighton and Berrigan 1995). Insufficient attention has been given to the relationships between intra- and interspecific scaling exponents of the DGC and its characteristics though there are some taxa where this could be readily done, such as size-dimorphic scarab beetles (Emlen 1997) and size-variable harvester ant workers (Lighton et al. 1994). Likewise, to date, only a single study has investigated the effects of VCO2 on the characteristics of the DGC while accounting for the effects of both mass and temperature. Chappell and Rogowitz (2000) found that DGC frequency, O-period emission volume, and O-period peak rate increased with increasing VCO2, whereas the combined CF period duration declined. The former findings are in keeping with those of Lighton and Berrigan (1995).

Proportional duration of the periods Several studies have suggested that differences in the proportional durations of the DGC periods (especially the F-period), and the ways in which these proportional durations change with temperature might reflect adaptations of the DGC to various environmental conditions, particularly water availability (Lighton 1990; Davis et al. 1999; Bosch et al. 2000; Duncan et al. 2002a, but see also Lighton et al. 1993a). These studies have all relied on calculating the relative durations of each of the DGC periods. Lighton (1990) argued that because the sum of the constants of the regression relationships between total DGC duration and durations of each of the periods is likely to be close to zero, these proportional durations can be calculated as the exponents of the least squares linear regressions of the durations of the C, F, and O-periods, on DGC duration, respectively. These ventilation phase coefficients have subsequently been used in several investigations (Lighton 1991a; Duncan and Lighton 1997), although it has now been demonstrated that they might give rise to considerable errors in estimating the proportions of the DGC occupied by each of the phases (Davis et al. 1999). This latter finding applies to any investigation in which the slope of a regression equation is used to determine a proportion. The slope of the regression equation provides an estimate of the change in the dependent variable with a given change in the independent variable. This change can be thought of as a proportion only if the intercept of the relationship is zero. If the intercept is a non-zero value or if the relationship is non-linear, this proportion changes systematically with a change in the independent variable (as has long been appreciated for Qi0 values—see Cossins and Bowler 1987).

Was this article helpful?

0 0

Post a comment