Metabolic rate and respiratory gas-exchange patterns in tenebrionid beetles from the Negev Highlands, Israel

Beetles belonging to the family Tenebrionidae represent an important part of the invertebrate fauna in deserts. They are able survive under even the harshest conditions as a result of physiological and behavioural adaptations and occupy certain ecological niches in desert habitats. The ecology of tenebrionid beetles living in the Ramon erosion cirque, at the southern boundary of the Negev Desert Highlands, which is a dry area that contains a sharp precipitation gradient with a mean annual precipitation from 100 mm in the northwest to 56 mm in the southeast, all occurring in winter, has been studied by Krasnov and Shenbrot (1996, 1997). They found that soil texture, productivity and vegetation architecture, but not topography, had profound effects on the spatial structure of the tenebrionid community.

In addition, they observed that there were two distinct spatial guilds of tenebrionid species. One guild was composed by species that preferred sandy-gravel soils and more ‘xeric’ habitats, and the other guild consisted of species that preferred clay soils and more ‘mesic’ habitats. Furthermore, some species occurred only in ‘xeric’ habitats, others only in ‘mesic’ habitats, although there were also species that occurred in both ‘xeric’ and ‘mesic’ habitats (see Table 1). There are four temporal phases of the beetle community, with the diversity of beetle communities being maximal in summer when the mean temperature is 34°C (Krasnov and Ayal, 1995). Most of these summer-active species are crepuscular and nocturnal and use burrows, the undersides of stones, rock fissures and plant cover to hide during the day (Ayal and Merkl, 1994). Only two summer species, both belonging to the genus Zophosis, were exclusively active during the day.

In contrast to the Negev Desert, many Namib tenebrionid beetles are diurnal during the summer months. The aim of this study was to investigate the respiratory patterns and metabolic rates of the Negev Desert tenebrionid beetles and to compare them with those of the Namib Desert tenebrionid beetles. This will help to explain what physiological adaptations enable some species to be active during the day and cause others to adjust their activity to prevent dessication. Seely and Mitchell (1987) and Krasnov and Ayal (1995) have suggested that the activity patterns of tenebrionid beetles are influenced by factors, e.g. physiological and reproductive adaptations, other than simple avoidance of high temperatures.

A physiological adaptation that would enable beetles to forage at higher temperatures would be to reduce their rate of dessication. In a hyper-arid desert such as the Negev, there is a significant selection pressure on water economy. The diet of tenebrionid beetles, all kinds of litter and dead arthropods, is generally low in water, so these beetles employ various strategies to limit desiccation, e.g. physiological resistance (for reviews, see Edney, 1977; Hadley, 1994a), burying below the surface (Seely et al., 1988) or adjusting their surface activity (Edney, 1971).

An important means of physiological resistance to desiccation is to reduce the rate of water loss. There are two major components of water loss: the transcuticular component that takes place across the cuticle and the respiratory component that accompanies the exchange of respiratory gases through the spiracles. Although few studies have been performed in which respiratory water loss rates can be separated from overall water loss rates, this route of water loss is in some cases an important component of total water loss (Hadley, 1994b). From a detailed study of the tenebrionid beetle Phrynocolus petrosus, occurring in an arid area in East Africa, Zachariassen (1991) concluded that respiratory water loss was the dominant component of water loss.

Insects may be able to reduce respiratory water loss by using a cyclic form of respiration referred to as the discontinuous gas-exchange cycle (DGC) (for a review, see Kestler, 1985; Lighton, 1994, 1996; Wasserthal, 1996). Several beetle species have been shown to exhibit this form of respiration, e.g. carabid beetles (Punt et al., 1957; Kestler, 1971), Namib Desert tenebrionid beetles (Bartholomew et al., 1985; Louw et al., 1986; Lighton, 1991), Scarabaeidae dung beetles (Davis et al., 1999; Duncan and Byrne, 2000) and a trogid beetle (Bosch et al., 2000). Most, but not all, of these species occur in arid habitats. The importance of the DGC for arid-dwelling arthropods is under debate (see Kestler, 1985; Hadley, 1994b).

The DGC is a cyclic discontinuity in external gas exchange that typically consists of three periods (Miller, 1981; Kestler, 1985). There is a closed (C) period, during which the spiracles are shut and which reduces respiratory water loss and gas exchange. Oxygen levels in the tracheae drop while CO₂ is largely buffered in the tissues and haemolymph. This is followed by the flutter (F) period, during which slight opening of the spiracles on an intermittent basis allows some normoxic O₂ uptake through the spiracles by diffusion and convection, but little loss of CO₂ or water vapour. These rapid spiracular movements (‘fluttering’) and short micro-openings are characteristic of the flutter period. Because of bulk movement of air into the insect, the minimised opening area and the low partial pressure differences, diffusive loss of CO₂ and water vapour is reduced. The final period, the CO₂ burst (B) period, is triggered when the accumulation of CO₂ from respiring tissues causes some or all of the spiracles to open widely in closed–flutter–open cycles or triggers ventilation as a bout of pumping movements in closed–flutter–ventilation cycles. The rapid unloading of CO₂ should minimise the time the spiracles are open and therefore reduce loss of water vapour. Thus, all periods are considered important in reducing respiratory water loss.

A study of the respiratory patterns of summer-active Negev Desert tenebrionid beetles with different diel activity patterns would enable us to determine whether there was any physiological adaptation that enables a few species to be active during the hot midday hours. Table 1 shows the habitat associations and ecological characteristics of the study species. Although all these species live under desert conditions, some are able to use the more xeric areas whereas others inhabit the more mesic areas. The study species are distributed between ancient and more modern tribes in the classification of Kaszab (1982). A comparison with the respiration patterns of the Namib Desert tenebrionid beetles (Louw et al., 1986; Lighton, 1991), of which several species are diurnal, would further our knowledge of the taxonomic scope of the occurrence of DGCs within the family Tenebrionidae and help in explaining the mechanistic basis of the DGC.

The nine beetle species (Table 1) were collected in Mitzpe Ramon, Israel, during July 2000. The nocturnally active beetles were collected by means of pitfall traps, the two diurnally active species were collected by hand while they were running about and the crepuscular/nocturnal species was collected from the hillocks around bushes. The beetles were housed in the laboratory in 10 l fish tanks with a layer of sand. They were fed dog food, and the occasional piece of apple and water was provided. The beetles were used in experiments within 4 weeks of collection and were released at their collection site after measurements.

A flow-through respirometry system was used to measure CO₂ emission in inactive beetles at 25°C. Room air scrubbed of CO₂ and H₂O vapour, by a Drierite/Ascarite column, was pumped at 100 ml min^–1 through a respirometry system consisting of a respirometry chamber (20 ml), a calibrated CO₂ analyser (Licor LI6262) and a mass-flow controller (Tylan). To ensure an accurate and constant airflow, a calibrated Supelco flow meter was also used. Readings of the amount of CO₂ produced were taken at 2 s intervals and recorded using computerised data-acquisition software (Datacan V, Sable Systems). Measurements were made on individual beetles weighed to ±0.1 mg using a Mettler balance (Toledo AB54). All beetle respiration was measured in the dark in a water bath maintained at 25°C and verified using a mercury thermometer. To compensate for any circadian rhythm effect, the respiration of the nocturnally active beetles was measured during the day for 2–6 h per beetle. The diurnally active beetles were measured at night for 10–12 h, with samples taken at 5 s intervals. We noted no change in the respiratory pattern if the measurement of these beetles continued during the morning and the beetle remained inactive. All beetles were occasionally observed to ensure that they remained stationary during sampling.

To convert V̇_CO2 to metabolic rate (measured in energy units of W kg^–1), the respiratory quotient (RQ) was assumed to be 0.8, as determined previously for Psammodes striatus (Coleoptera: Tenebrionidae) (Lighton, 1988), which seems a reasonable assumption for the beetles. An RQ of 0.8 gives an energy equivalent of 24.5 J ml^–1 CO₂ (Schmidt-Nielsen, 1980). This value was used to express metabolic rate in terms of energy.

The characteristics of the discontinuous gas-exchange cycle (DGC) (as given in Table 3) were calculated as follows. The mean rate of CO₂ emission in each period (V̇_CO2 in μl g^–1 h^–1) is the volume of CO₂ emitted during that period, obtained by calculating the area under the CO₂ peak and then expressed as a mass-specific value and divided by the period duration. The period duration is the time taken for CO₂ to be emitted during that period. The DGC frequency (=burst frequency) is calculated by determining the number of peaks of CO₂ emission per second, and the period is the duration of one DGC (i.e. a closed, flutter and burst period; C, F and B, respectively, in Fig. 2).

At least three beetles were used from the common species to measure body water content and water loss tolerance. The beetles were kept in a desiccator (0 % relative humidity), which was then placed in an incubator at 28°C. The beetles were individually weighed before placement in the incubator and observed on a daily basis until loss of coordination, when they were re-weighed. To obtain the dry mass, the beetles were placed in an oven at 110°C until they reached a constant mass (Zachariassen et al., 1988).

Unless stated otherwise, data are presented as means ± standard deviation (s.d.). Sample size (N) is indicated in the text either as representing individual beetles or, in the case of gas-exchange characteristics, for 5–90 discontinuous gas-exchange cycles per beetle. Unless noted otherwise, statistical comparisons were made either with Student’s t-tests or by analysis of variance (ANOVA). Significant ANOVAs were followed with the Newman–Keuls multiple-range test. The significance level was set at P<0.05.

It was not possible for activity to be monitored continuously during the measurements. However, during visual observations on a few beetles, we noted that even slight movements were accompanied by noticeable changes in the pattern of CO₂ emission. Thus, we are confident that our measurements are from inactive beetles.

The nocturnally active beetles exhibited an irregular form of continuous ventilation (Fig. 1): they displayed a constant rate of CO₂ emission with little variation from its mean and a coefficient of variation (CV) less than 1 (Table 2). The only exception was Blaps sulcata, which had a CV of 1.82 and exhibited a form of cyclic respiration without the distinct periods that characterise DGCs. The oscillations in CO₂ emission could be due to abdominal pumping (ventilation), which is known to occur in beetles (Kestler, 1971; Bartholomew et al., 1985). We were unable, however, to observe or measure abdominal pumping in this study. The sudden reductions in CO₂ emission may be due to spiracular closing reflexes in response to vibrations in the laboratory, as described by Kestler (1971).

The two diurnally active Zophosis species showed a distinct DGC pattern (Fig. 2) similar to that measured in the Namib Desert tenebrionid beetles (Lighton, 1991). In both these studies, the tenebrionid beetles produced a flutter period consisting of several discrete bursts of CO₂ emission with little or no CO₂ output between these emissions. Because of their regularity, these minibursts may be due to miniature ventilations. The subsequent CO₂ burst period is possibly due to complete opening of the spiracles or to active ventilation. There was no evidence in the CO₂ emission trace (Fig. 2) to suggest that active ventilation was occurring during the burst period. The crepuscular/nocturnal species Pimelia canescens had a distinct DGC but with 0–7 typical micro-openings during the flutter period.

The DGC characteristics of the two Zophosis species were very similar (Table 3). Unfortunately, only two specimens of P. canescens were obtained, so it was not possible to make statistical comparisons among all three species. From the data in Table 3, P. canescens seemed to have longer burst period and closed period durations, resulting in a DGC of longer duration than that in the Zophosis species. There were fewer miniature ventilations during the flutter period. In all cases, DGC frequency was independent of beetle mass, as anticipated (see Lighton, 1991).

The duration of the flutter period was tightly controlled relative to the DGC duration. In both species, a significant positive correlation was found between flutter period duration and DGC duration. The flutter ‘ventilation’ period coefficient (the ratio of slopes of the linear regression of flutter period duration on DGC duration) was 0.80±0.16 (r²=0.77, P=0.02) for Z. complanata and 0.84±0.08 (r²=0.94, P<0.001) for Z. punctata (means ± s.e.m.). As the duration of the flutter period increased, the number of miniature ventilations within each flutter period also increased, and a similar distribution of the number of miniature ventilations was found in the two species (Fig. 3). The frequency of CO₂ bursts within the flutter period was similar: Z. complanata had a frequency of 19.4±5.7 mHz (408 cycles) and Z. punctata had a frequency of 23.4±6.3 mHz (381 cycles) A significant positive correlation was found between V̇_CO2 (ml CO₂ h^–1) and the rate of flutter period CO₂ emission (μl g^–1 h^–1) for both species. A large proportion of total CO₂ was emitted during the flutter period. The flutter period CO₂ volume/burst period CO₂ volume ratio was 0.48±0.17 in Z. complanata and 0.29±0.14 in Z. punctata.

The duration of the burst period was shorter than that of the flutter period and more similar to the duration of the closed period (Table 3) in both species. The burst period duration, however, was independent of DGC duration. A significant positive correlation was found between V̇_CO2 (ml CO₂ h^–1) and mass-specific volume of CO₂ emitted during the burst period.

Very small amounts of CO₂ were emitted during the closed period (Table 3). Only in Z. complanata was there a positive correlation between the duration of the closed period and DGC duration.

The standard metabolic rate for each beetle species was calculated from the mean V̇_CO2 (ml h^–1) during stable continuous gas exchange or while the beetle exhibited a regular discontinuous gas-exchange pattern (in the case of Zophosis species and P. canescens). All the beetle species had a similar metabolic rate (Table 2), implying a common resting state.

Grouping all the species data, we found that V̇_CO2 (ml CO₂ h^–1) scaled allometrically with mass, V̇_CO2=0.179M^1.06, where M is mass in g (r²=0.8, P<0.001). For the three species exhibiting DGCs, the flutter period and burst period CO₂ emission volumes showed a significant relationship with body mass within a species and at the interspecific level. Grouping the data for these three species, the burst period CO₂ emission volume scaled with M^1.20 and the flutter period CO₂ emission volume scaled with M^1.44.

To determine whether these tenebrionid species have lowered their metabolic rate to survive in more arid environments, we compared our values with those obtained by Lighton and Fielden (1995) for a large data set of arthropods. Our data for the individual beetles from nine species gave the allometric relationship MR=1216M^1.06 (where M is mass in g and MR is metabolic rate in μW) compared with the equation MR=906M^0.825 obtained by Lighton and Fielden (1995). This suggests that metabolic rates do not seem to have been reduced in the more arid environment. The differences between the two data sets could be due to unrecognised activity effects causing higher metabolic rates and because Lighton and Fielden (1995) used a Q₁₀ value of 2.5 to correct metabolic rate to a temperature of 25°C. This is not always a valid assumption, as has been shown by Duncan and Dickman (2001). Without a larger data set of insect metabolic rates, we can only suggest that the Negev Desert beetles may have elected to be active when the water vapour pressure gradient was reduced, i.e. at night, or to use other adaptations to enable them to be active at high temperatures rather than lower their metabolic rates.

An important finding of this study is that only the diurnally active Negev Desert tenebrionid beetles used DGCs and had a gas-exchange pattern similar to that found by Lighton (1991) in 10 species of tenebrionid beetle from the Namib Desert. These tenebrionid beetles are the only beetle species measured to date to produce discrete bursts of CO₂ (miniature ventilations) during the flutter period. Miniature ventilations have previously been observed in the locust Locusta migratoria migratorioides (Hustert, 1975). In all the other beetle species, e.g. carabid beetles (Punt et al., 1957), Scarabaeidae dung beetles (Davis et al., 1999; Duncan and Byrne, 2000), a trogid beetle (Bosch et al., 2000) and wood-boring beetles (Chappell and Rogowitz, 2000), the flutter period consists of a gradual increase in the amount of CO₂ emitted that results in a large release of CO₂, the burst period (see Fig. 2, DGC in P. canescens). In many of the above investigations, the flutter and closed periods were indistinguishable.

The reduction in respiratory water loss is thought to result from the diffusive loss of CO₂ and water vapour being restricted to discrete cyclic events (Miller, 1981). Thus, the burst period (high respiratory water loss) of the DGC should be short relative to the closed and flutter periods (low respiratory water loss). In Zophosis species, the duration of the burst period is less than a quarter of the duration of a complete gas-exchange cycle (Table 3). For a desert species, we would have expected the closed period to be longer than the burst period, as found in the Namib Desert tenebrionid beetles (Lighton, 1991). Yet, in our study, the flutter period was 3–5 times longer than the closed period (Table 3), whereas it is only twice as long in the Namib Desert tenebrionids. In Z. complanata, 48 % of total CO₂ output occurred during the flutter period (29 % for Z. punctata), similar to the value found by Lighton (1991) for nine tenebrionid species, in which 24 % of total CO₂ release occurred during the flutter period (47 % for Epiphysa arenicola).

In the Zophosis species, the flutter period plays an important role in the modulation of metabolic rate with body size. The increase in metabolic rate was accommodated by increasing the rate at which CO₂ was emitted during the flutter period. Long and conspicuous flutter periods appear to be common in arid-adapted arthropods, e.g. the ant species Cataglyphis (Lighton and Wehner, 1993) and a trogid beetle (Bosch et al., 2000). The flutter period is, therefore, an important component of the gas-exchange cycle in arid-adapted insects and, in this study, the flutter period duration was found to be tightly coupled with DGC duration.

The importance of the flutter period in reducing respiratory water loss has been advocated by Lighton and Garrigan (1995). In the ant Camponotus vicinus, they found that the rate of water loss relative to CO₂ emission during the burst period increased approximately fourfold compared with flutter period levels. This was explained by suggesting that, as a result of spiracular fluttering, less water would be lost during the flutter period than during the burst period, when there is no spiracular control, even though the molar rates of CO₂ and water loss during the flutter period are equivalent. The miniature ventilations during the flutter period should also reduce water loss mainly because the respiratory passages are not continuously exposed to the external atmosphere. The DGCs recorded from P. canescens do not show numerous miniature ventilations during the flutter period (Fig. 2). Thus, respiratory water loss may be greater in these beetles, compelling them to change their activity pattern throughout the year. These beetles are diurnal during the winter months and become nocturnal during the summer months.

From the above data, we can draw some conclusions about the function of the DGC in relation to diel activity. The activity patterns of the beetles are more important than their habitat associations in designating the type of respiration used. Thus, we propose that, in some beetle species, discontinuous ventilation has been used to limit respiratory water loss, enabling them to be active during the hot daytime hours. Being active during the day may enable the small Zophosis species to reduce competition for limited food resources. All the tenebrionid species in this study are well-adapted to desert conditions. The values in Table 4 for body water content, critical hydration levels and water loss tolerance are typical for arid-adapted insects (Hadley, 1994b). It is possible that, because of their diel activity and small size, respiratory water loss is an important component of the water budget of diurnally active species.

Seely and Mitchell (1987) postulated that, if beetles need to minimise dessication by minimising the vapour pressure gradient between themselves and their surroundings then, in a typical desert area where there has been no recent rain, they would elect to be on the surface throughout the night. Six of the nine study species are active only at night. They may either not have a need to use discontinuous ventilation or discontinuous ventilation may not be the ancestral condition. It is interesting to note that the most ancient species in this group (Kaszab, 1982) do exhibit DGCs. Until a detailed phylogenetic analysis is performed and the respiratory patterns of more species of tenebrionid beetle are measured, we cannot be sure whether discontinuous ventilation was the ancestral condition or is a more-derived condition. In his study of the Namib Desert tenebrionid beetles, Lighton (1991) found DGCs in the one nocturnal beetle (Epiphysa arenicola), which inhabited a particularly harsh environment within the desert. If discontinuous ventilation is the ancestral condition, then factors other than being unable to reduce respiratory water loss through the use of DGCs may have induced beetles to be active at night.

Traditionally, it was proposed that the DGC evolved as a mechanism to enhance water conservation (Kestler, 1985; Lighton, 1996). The main alternative hypothesis is that the DGC evolved to facilitate respiration in the hypercapnic and hypoxic environments experienced by arthropods living underground (chthonic hypothesis) (Lighton, 1996). In our study, both the nocturnally active and the diurnally active beetles bury themselves in the soil, preferably into a small sandy hill close to a shrub, where the soil is less compact (B. Krasnov, personal observations). From the study of Louw et al. (1986) in the Namib Desert, we would be surprised if the tenebrionid beetles experienced conditions of hypercapnia and hypoxia. Louw et al. (1986) found that CO₂ diffused very rapidly through a 5 cm column of dune sand, the depth at which the tenebrionid beetles are commonly found. None of our study species inhabits deep burrows where they might experience hypoxic or hypercapnic conditions. If the chthonic hypothesis is valid, there should be no difference in the respiratory patterns between the species since they all burrow into the sand. Furthermore, we would expect the nocturnally active beetles to bury deeper to escape the extreme day time temperatures.

This research was conducted while F.D.D. was visiting the Ramon Science Center. We thank Laura Fielden for her assistance and encouragement and Paul Kestler for kindly commenting on the manuscript. The visit of F.D.D. was supported by the Blaustein Center for Scientific Cooperation Research and the Science Liaison Division of the NRF. Financial support for M.M. was provided by the Israel Ministry of Foreign Affairs (MASHAV program). B.K. received financial support from the Israel Ministry of New Immigrant Absorption, the Israel Ministry of Science, Culture and Sport and the Local Council of Mizpe Ramon. This is publication no.125 of the Ramon Science Center.