The role of the subelytral cavity in water loss in the flightless dung beetle , Circellium bacchus ( Coleóptera : Scarabaeinae )

Circe!!ium bacchus is a flightless telecoprid (ball-rolling) dung beetle, endemic to the afrotropical region, where it is found in a few restricted populations in the eastern Cape of South Africa. Its apterous condition and large size (mass ranges from 6 to 12 g) are considered to be adaptations to a semi-arid habitat. This beetle is active in the sun for long periods, walking between widely scattered dung pats, thus is under selection pressure to reduce water loss. C. bacchus has eight spiracles on each side of the body. The metathoracic spiracle and six abdominal spiracles open into the sub­ elytral cavity, which is closed. The mesothoracic spiracle is the largest and most exposed, occurring ventrally in the membrane con­ necting the prothorax and mesothorax. When at rest a cyclic form of respiration, known as discontinuous gas exchange cycle, is used by C. bacchus, releasing a burst of carbon dioxide approximately once an hour when the spiracles open for about 33 minutes. Flow-through respirometry was used to measure water loss from the thorax (being the head, prothorax and mesothorax) and elytral case (containing the metathorax and abdomen) separately. The total water loss of C. bacchus could be divided up as 65% cuticular water loss from the thorax, 35% cuticular water loss from the elytral case, 4% respiratory water loss from the thorax and no measurable respiratory water loss from the elytral case. 1.51 pg of water is lost for every pl of C02 emitted during respiration in the thorax. Thus, the main avenue for both respiration and respiratory water loss is via the mesothoracic spiracles, suggesting that the primary function of the subelytral cavity is not to reduce respiratory water loss.


INTRODUCTION
The success of terrestrial insects is attributed to their ability to cope with a loss of water to the surrounding air and a limited and often an unpredictable supply of water (Hadley, 1994a).To cope with the evaporation of water, insects have developed physiological and morphological resistance to desiccation.The subelytral cavity, an airfilled space between the abdomen and sealed elytra, which is found in many arid-adapted beetles, has been considered an important structure which helps these bee tles reduce their rate of water loss (Cloudsley-Thompson, 1975).Because the abdominal spiracles open into the cavity it maintains air at high humidity, thereby reducing the water lost during respiration (Ahearn, 1970;Zachariassen, 1991).This gave rise to the current hypothesis that the subelytral cavity evolved primarily to conserve water.Both Nicholson et al. (1984) and Zachariassen (1991) recorded low total evaporative water loss from the elytral cases of tenebrionid beetles (Onymacris p!ana and Phrynoco!uspetrosus respectively).However, these findings have always been confounded by the difficulty of deter mining the exact source of the water.Only recently, due to improved equipment, has it become possible to accu rately separate respiratory water loss from cuticular tran spiration.
The aim of this study was to estimate the differences in cuticular and respiratory water loss from the elytral case and anterior region of an arid-adapted beetle.In beetles, the first pair of spiracles occurs on the mesothorax, with the next seven to nine pairs on the metathorax and abdominal segments, covered by the elytra.Therefore the elytral case encloses the third thoracic segment and the entire abdomen.Two previous studies have managed to divide a tenebrionid beetle into sections and measure the water loss from each section (Cooper, 1983;Zachariassen, 1991).However, due to their experimental set-up, they were unable to measure respiratory water loss directly.Bosch et al. (2000) measured respiratory water loss directly in the beetle, Omorgus radu!a, but did not determine the separate contributions of the thorax and abdomen.Zachariassen et al. (1987) suggested that the respiratory water loss is the dominant component of water loss in tenebrionid and carabid beetles from arid habitats in East Africa.However, Hadley (1994b) has shown that respira tory water loss rates may be lower than the cuticular water loss rates.Nevertheless, any reduction in respira tory water loss will help reduce overall water loss.A cyclic form of respiration referred to as the discontinuous gas exchange cycle (for reviews see Kestler, 1985;Lighton 1994Lighton , 1996;;Wasserthal, 1996) is one means by which beetles may be able to reduce respiratory water loss.The discontinuous gas exchange cycle is a cyclic discontinuity in external gas exchange which typically consists of three periods (Miller, 1981;Kestler, 1985).These are; a closed (C) period where the spiracles are shut, which prevents both 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 diffu sion and convection, but little CO2 or water vapour is lost.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.The rapid unloading of CO2 should minimise the time the spiracles are open and therefore reduce water vapour loss.Thus all periods are considered important to reduce respi ratory water loss.
For this study I chose an arid dwelling beetle which possesses a subelytral cavity and uses discontinuous gas exchange cycle for respiration.Circellium bacchus (F.) (Scarabaeinae: Canthonini) is a flightless telecoprid (ball rolling) dung beetle, endemic to the afrotropical region, where it is found in a few restricted populations of the eastern Cape of South Africa (Chown et al., 1995).Telecoprid dung beetles detach a portion of dung from a drop ping, roll it some distance away from the source and then bury it, or place it in a grass tussock.C. bacchus is assumed to be a poor competitor against other large, flying telecoprids, and is therefore confined to drier habi tats.This beetle is active in the sun for long periods, walking between widely scattered dung pats, thus is under selection pressure to reduce water loss.Duncan and Byrne (2000), found an association in dung beetles between the aridity of the habitat and respiration pattern, with C. bacchus being an extreme example of the discon tinuous gas exchange cycle.
The apterous condition and large size (mass ranges from 4 to 12 g) of C. bacchus are considered adaptations to an arid habitat.Fortunately its large size also makes division of the beetle easier and therefore measurement of its respiratory water loss possible.C. bacchus has eight spiracles on each side of the body.The metathoracic spiracle and six abdominal spiracles open into the subelytral cavity, which is closed.The mesothoracic spiracle is the largest and most exposed, occurring ventrally in the membrane connecting the prothorax and mesothorax, behind the coxal cavities of the prothoracic legs.In this position it opens into the "waist" between the thorax and abdomen.By dividing the beetle at its "waist", I was able to measure the water loss rates from the elytral case and the anterior region of the beetle separately and simultane ously.Thus the route of respiratory water loss and contri bution to total water loss could be determined.

MATERIAL AND METHODS
Circellium bacchus adults were collected from the Addo Ele phant National Park (33E30'S, 25E41'E) in the Cape Province, South Africa, with permission from the National Parks Board because of the conservation status of this species.The beetles were housed in 10l bins half filled with soil, in an insectary at 25°C with a 14 : 10 h light:dark cycle.They were fed fresh cow dung weekly, and survived well for several months under labo ratory conditions.
A flow-through respirometry system was used to measure CO2 and H2O emission in inactive beetles at room temperature (23°C).The individual beetle to be tested was placed in a per spex respirometry chamber.The chamber was separated into two halves, together having a volume of approximately 250 ml.Both halves were made from a block of perspex with an internal cavity of approximately 40 mm diameter drilled out for the ani mal.Four holes at each corner allowed the two halves to be bolted together.The chamber halves were divided using a sheet of latex (dental dam 0.013-0.018cm thick, The Hygenic Corpo ration) thus separating the respiratory chamber into two air-tight compartments.A small hole was cut into the latex sheet which formed a collar around the "waist" of the beetle.In this configu ration the anterior part of the beetle (head, prothorax, meso thorax and forelegs) was in one compartment and the posterior part (elytral case, middle legs and hindlegs) was in the other.Each compartment could be sampled or ventilated through two stainless steel pipes attached to the respirometry equipment (Fig. 1).The entire configuration was tested for leaks, and the latex sheet was replaced for each measurement.
The flow-through system was set-up for each compartment separately, with each having its own source of air, flow con troller and gas analyser.Room air, scrubbed of CO2 and H2O vapour by a Drierite/Ascarite column, was drawn through the compartment at 50 ml min-1 (controlled by a calibrated Supelco flow meter) and then into either a Licor CO2 analyser (LI-6251) or Licor CO2 /H2O analyser (LI-6262).The LI-6262 is a dual wavelength infrared absorbance analyser capable of measuring CO2 and H2O simultaneously to a resolution of 0.1 p.p.m. and 0.002 kPa, respectively.Bev-A-Line low permeability tubing was used to connect the components, and the distance from each compartment to the respective analyser was kept to the minimum and was identical.Readings of the volume of CO2 and/or H2O emitted were taken every 5 s (unless noted other wise) and recorded using computerised data acquisition soft ware (Datacan V, Sable Systems).Measurements were made on individual beetles that were weighed to ± 0.1 mg (Precisa 160A balance).The beetles were measured for a minimum of 6 hours in the dark.They were occasionally observed to ensure that they remained quiescent during sampling.After measurements, the beetles were weighed again.
Baseline drift of the analysers was corrected during analysis from measurements at the beginning and end of each trial with the respirometry chamber empty.All measurements were cor rected to Standard Temperature and Pressure.The CO2 record ings were converted to rate of CO2 production (CO2) in ml h-1 and H2O data were converted to water loss rate in mg h-1.The discontinuous gas exchange cycle (DGC) characteristics (as given in Table 1) were calculated as follows.The DGC fre quency (= burst frequency) is calculated by determining the number of peaks of CO2 per second, and the DGC duration is one complete cycle (ie.closed, flutter and burst period).The mean rate of CO2 or H2O emission is the mean value over sev eral complete discontinuous gas exchange cycles.To measure the emission volume, integration of the area under the curve against hours was performed.
Data are represented as means ± standard deviation.Sample size (n) is indicated in the text as either representing individual beetles or in the case of gas exchange characteristics and burst period water loss, three to ten discontinuous gas exchange cycles per beetle.Regression analysis was done by the least squared method.

RESULTS
It was not possible for activity to be monitored continu ously during the measurements.However, during inter mittent visual observations while beetles were recorded, I noted that even slight movements were accompanied by noticeable changes in the pattern of CO2 emission from the thorax (Fig 2).In all experiments the CO2 emission from the thorax was measured to ensure that the data came from inactive beetles and any measurements showing activity were excluded from the analysis.
The trace of CO2 and H2O emission from the same beetle (mass = 7.296 g) with separate recordings of the anterior part, referred to as thorax (consisting of head, pro-and meso-thorax regions), and elytral case (meta thorax and abdomen) is shown in Fig. 3.All three beetles showed a similar pattern and the results are summarised in Table 1.There is a clear discontinuous gas exchange cycle pattern from the thorax similar to that recorded by Duncan and Byrne (2000) for whole specimens of C. bac chus (Fig. 3A), with a definite burst and closed periods but no or limited flutter period.In Table 1, the interburst period refers to the region between two CO2 bursts ie. between the burst period, as there was no distinct closed and flutter periods (Fig. 3A).During the burst period of the CO2 emission trace, a corresponding increase in water loss occurs.These bursts of H2O loss are the respiratory water loss from the mesothoracic spiracles while open, and the recording of water loss between these bursts cor responds to the cuticular water loss as the spiracles are closed during this period.(I have assumed that in the interburst period there is only uptake of O2 and no loss of CO2 and H2O; see Kestler, 1985).In Fig. 3B, the CO2 emission from the elytral case is less than that of the thorax and no distinct cycles can be seen.The CO2 trace does, however, show some periodic increases but there is no corresponding increase in water loss.The random fluctuations in the water loss trace can be attributed to instrument noise.Thus for the elytral case it was not pos sible to separate respiratory water loss from cuticular Fig. 3. Water loss (upper trace, broader line) and CO2 emission (lower trace) from a 7.296 g Circellium bacchus at 23°C while motionless.A -recording taken from thorax region (ie.head, prothorax and mesothorax).B marks the burst period and I marks the interburst period of the discontinuous gas exchange cycle; B -recording taken from the elytral case (ie.metathorax and abdomen) of the same beetle.water loss.65.1% of the total water loss and 71.3% of total CO2 emission came from the thorax (Table 1).
Fig. 4 shows that on opening the mesothoracic spiracles, three events occurred instantaneously viz.CO2 emitted, H2O lost and O2 is taken up.The same event was not observed for the elytral case, thus expired air is not leaving through a single aperture above the anus in an intermittent pattern.
By combining the water loss rate (mg h-1) from the thorax and elytral cavity (Table 1), the average total water loss rate for the three beetles (mean mass of 7.195 g ± 2.81) in this study is 10.1 mg h-1 ± 4.25, which compares well with that found by Chown et al. (1995) for C. bac chus.In their study of water loss rates using the gra vimetric method they found an average rate of 14 mg h-1 ± 1.9 (range of 8.1-34.8)from beetles of average mass 6.489 g ± 0.332.
elytral case): 64% of total cuticular water loss is from the thorax while 36% is from the elytral case.Respiratory water loss could only be measured from the thorax and contributed 4% to total water loss, but 6% of thoracic water loss.

DISCUSSION
Several functions have been suggested for the subelytral cavity, one being to help reduce respiratory water loss (Cloudsley-Thompson, 1975).For this to occur, the abdominal spiracles which open into this cavity should be the main site for gas exchange.The assumption being that when the spiracles are open to allow gas exchange to occur the respiratory passages would be exposed to humid air and lose less water through diffusion.In this study, using C. bacchus, I have shown that for the poste rior part of the body, including the elytral cavity, respira tory water loss cannot be separated from cuticular water loss, and that respiratory water loss is probably negligible as less than one third of the expired air leaves from this site (Duncan and Byrne, manuscript submitted).This sug gests that the spiracles inside the subelytral cavity are not the major site of gas exchange.
Table 2 shows the results from studies on beetles in which the respiratory water loss rates from the elytral case were estimated separately from the thorax.In this study I attempted to measure respiratory water loss directly, whereas in the other three studies it was meas ured using indirect methods, which may partially explain some of the differences.In Onymacris plana, Nicolson et al. (1984) found the cuticular water loss rate from the ely tral case was the same as that measured from the tip of the abdomen.They assumed that expired air leaves through a single aperture above the anus and by measuring water loss rates from this area they would determine the contri bution of elytral respiratory water loss and cuticular water loss to total water loss rates.However, they were unable  Ahearn (1970), 'from Cooper (1983), 4from Zachariassen (1991).
to simultaneously measure whether CO2 was being elimi nated through this aperture.
In the thorax, an increase in CO2 emission during the interburst period resulted in a corresponding increase in water loss [linear regression: water loss rate (mg g-1 h-1) = 0.36 + 853.9 CO2 emission (mg g-1 h-1), r2 = 0.73, P < 0.001].Thus, there could be a small but undetectable amount of respiratory water loss occurring from the ely tral case when there is an increase in CO2 emission.
In C. bacchus the water loss rates were less from the elytral case (Tables 1 and 2).However from this study it was not possible to ascertain whether; the presence of humid air in the subelytral space, a more regular surface structure and possibly less surface area, or a difference in cuticular permeability (Loveridge, 1980) in this region is responsible for the difference measured.The contribution of the elytral cavity to total water loss in C. bacchus was less than half, similar to that found in O. plana by Nicolson et al. (1984).
Estimates of the contribution of respiratory water loss to total water loss in beetles have ranged from 6.5% in the trogid beetle, Omorgus radula (Bosch et al., 2000), to 70% in the tenebrionid beetle, Phrynocolus petrosus (Zachariassen, 1991).The use of discontinuous gas exchange cycles in respiration could lower respiratory water loss (Kestler, 1985) and the relative contribution of respiratory transpiration in insects that are known to exhibit discontinuous gas exchange cycles ranges from 1.9% to 13% (Hadley, 1994b).The value estimated for C. bacchus (4.12%) is within this range.
By using discontinuous gas exchange cycles in the mesothoracic spiracles, C. bacchus is reducing the amount of time the respiratory passages are exposed to dry air.The opening of spiracles is due to the build up of CO2 in the haemolymph (Harrison, 2001), but the closure may have been induced to reduce water vapour loss.Although the mesothoracic spiracles are the most exposed in C. bacchus, in their situation behind the coxal cavities of the prothoracic legs they do not open directly to out side air and some water vapour may be retained.
Another means of reducing respiratory water loss is to increase CO2 emission efficiency.This has been used by C. bacchus as the ratio of water lost to C 02 emitted during the burst period is markedly lower than that reported for other arid-dwelling insects.In C. bacchus 1.51 pg of H2O is lost for every pl of CO2 (Table 1) com pared to ants; 4.8 in Cataglyphus bicolor (Lighton, 1994), 3.25 in Pogonomyrmex rugosus alates (Lighton et al., 1993), 5.37 in P. rugosus workers and 5.29 in P. occidentalis workers (Quinlan & Lighton, 1999) and approxi mately 5 in the trogid beetle Omorgus radula (Bosch et al., 2000).Respiratory water loss in C. bacchus is kept to a minimum as indicated by the small increase in peak burst water loss (Table 1).

CONCLUSION
In C. bacchus the rate at which water is lost through respiration has been reduced, such that it is only a small component of total water loss.This arid-adapted beetle does not seem to use the subelytral cavity to reduce respi ratory water loss but uses discontinuous gas exchange cycles through the thoracic spiracles with a high C 02 emission frequency.Thus the hypothesis that the subely tral cavity evolved primarily as a water conservation adaptation is questioned.

Fig. 1 .
Fig. 1.Diagram of the experimental set-up, showing the beetle divided into an anterior part and posterior part and flow through system for each part separately.(See text for details).

Fig. 2 .
Fig. 2. Carbon dioxide emission from thorax region of a 9.677 g Circellium bacchus at 23°C, showing the effects of movement on the discontinuous gas exchange cycles.

Fig. 4 .
Fig. 4. Recordings of CO2 emission (upper trace), water loss (middle trace) and O2 uptake (lowest trace) taken every 0.06 seconds from the thorax region in Circellium bacchus.The delay in the O2 recording has been corrected (the O2 analyser was further downstream of the beetle).The water loss trace only indicates the increase above the cuticular water loss.The dotted line indicates the opening of the spiracles.

Table 1 .
Relative contribution of thorax and elytral case to water loss in Circellium bacchus(mean mass = 7.192 g ± 2.54, n = 3)

Table 2 .
Relative contribution of cuticular and respiratory tran spiration to total water loss.The thorax region refers to the head, prothorax and mesothorax; the elytral case refers to the metathorax and abdomen sections of the beetles.