Ventilation of the burrow of the Southern Hairy-Nosed Wombat, Lasiorhinus latifrons, in bushland near the Murray River at Morgan in South Australia

This article is copyright, and its use should be referenced as shown:

Wood G, 1997, "Ventilation of the burrow of the Southern Hairy-Nosed Wombat, Lasiorhinus latifrons, in bushland near the Murray River at Morgan in South Australia" published electonically on the Internet at

Abstract Introduction Materials and Methods Results Discussion Acknowledgements References

Wombats are large fossorial mammals which seek refuge in burrows. Within these burrows we found the diurnal fluctuations in temperature and humidity are dampened. However, wombat burrows do not receive enough air by diffusion to offset hypoxia-hypercapnia and replenish the burrow. We examined the airflow over and through the burrows, concluding that  they was continuous airflow within the burrows. Factors considered to generate the airflow were: changes in  air pressure due to airflow over the warrens; random aspects of burrow openings to optimise the chance of wind capture; and convective airflow caused by a continuous temperature gradient between inside and outside burrow temperatures.

There are obvious advantages for an animal in being fossorial. Worldwide it has been documented that burrows ameliorate temperature and humidity on surface ( Vogel et al. 1973; Wilson and Kilgore 1978). However, the trade off is that the burrows can become hypercapnic and hypoxic with respect to earth’s free atmosphere. The composition of gas in a burrow  results  from interplay between the metabolism of resident animals, bulkflow of gas from the opening and diffusion of gas within the airspaces in the soil (Maclean 1981). It has been well documented that many burrowing animals and birds have a low metabolic rates, an obvious correlate of producing lower levels of respiratory gas production (Boggs and Kilgore 1983). Wombats, like other small mammals have increased  tolerance to elevated carbon dioxide levels within a burrow, conferred by adaptations to a hypoxic environment (Baudinette 1974). In previous analysis of gas exchange in burrows it is clear that diffusion alone from the burrow opening can not provide sufficient oxygen for respiration yet fossorial mammals can remain in burrows for extended periods (Wilson and Kilgore 1978). Olszewski and Skoczen (1965), recognised this apparent paradox and demonstrated that there was bulk flow of gas within burrows, and that the air flow in the burrows of the mole Talpa europaea was related to the speed of air over the ground.
Vogel et al. ( 1973) also addressed this ambiguity in a study of gas exchange in the burrows of the North American Prairie dog. In these burrows the action of wind travelling over the elevated burrow openings was described by Bernoulli’s principle , where pressure differences  were generated by variations in velocity which induced gas movement. Vogel coined the term “viscous entrainment” to describe the phenomena of  gas movement within a burrow as bulk flow due to the entrainment of the gas particles. Vogel et al. (1973) demonstrated the efficacy of the two principles by the use of smoke bombs in the field.
In the present study we examined the hypothesis that viscous entrainment contributes to ventilation of the burrows of the Southern Hairy-nosed Wombat  Lasiorhinus latifrons. The animal is a large burrowing mammal, and is known to occupy complex burrow systems (Triggs 1996).

Materials and Methods
Study Area
Scott’s Creek near Morgan, South Australia off the Fiddlers Green Road (340 02’, 1390 40’). We examined two  warren systems with numerous openings. Evidence of fresh digging and faecal deposits around some burrow entrances indicated recent activity. Burrow openings were  orientated, numbered and mapped. Height and width (at widest point) of burrow openings were measured and the presence and aspect of excavated soil mounds was noted. One burrow opening from each warren was selected for more detailed  measurement.

Temperature and humidity
Inside and outside burrow temperature was measured using an Omega Infrared Spectrophotometric Temperature Sensor (Omega  Engineering  Stanford C. T.) This instrument had been calibrated against a standardised hotplate assuming an emissivity of  0.99. Data loggers (Tiny Talk temperature and humidity data loggers Gemini Data loggers U.K.)  were programmed to record humidity and temperature every fifteen minutes for 24 hours. The temperature and humidity data loggers were placed inside 500ml ventilated plastic jars. One of each type was  positioned approximately one metre inside and adjacent to the burrow entrances. Temperature data loggers were calibrated against standard mercury and glass thermometer traceable to a national standard. Humidity data loggers were calibrated at 3 points against 3 standard chemical solutions (Solomon 1958).Soil temperature profiles were taken from a series of 5 Copper-Constantan Thermo couples, inserted at 10 centimetre depths in soil within a 40 metre profile.Another 3 thermocouples  were placed at the opening. One, 2 metres inside, beneath shaded vegetation above the burrow and one on the soil surface adjacent to the burrows.

Gas analysis
Gas samples were withdrawn from previously undisturbed burrows using minimum dead space plastic tubing inserted to between 1.5 to 2.0 metres into the burrow. The dead space was removed via a 3 way tap and the sample was subsequently drawn into a 10ml ground glass syringe. The samples were analysed in a Amaetek C-3A oxygen analyser in a Amatek C- D3A CO2  analyser (Amatek Corporation Pennsylvania USA). The output from the instruments was digitalised and the readings taken from a Sable Systems (Las Vegas USA ) software system.

Airspeed above the surface and at the burrow openings was measured with a Heat-Thermister Ananometers (T.A. 3000 Airflow Developments  Ontario  Canada). Airflow patterns within the burrow were visualised by using a smoke generator (A.B. Regin VENTAX. Smoke Emitter, SWEDEN.)

The two warrens (Warren 1 and 2; Map 1) observed, were sited on gently sloping ground and consisted of medium sized burrows (Triggs 1996). Patchy vegetation was apparent throughout both warrens. The burrow openings were generally symmetrical around a horizontal axis and randomly orientated with respect to each other. Each burrow opening had a single mound of excavated soil with one edge sloping towards the base of the opening. The height of these mounds was variable but small and in keeping with the smaller mound seen in medium sized burrows referred to in Triggs (1996). The aspect of each mound was also varied and did not form a pattern favouring one direction. The orientation of each burrow opening is shown on map 1. Tracks and scats indicated that Warren 2 had been visited within the observation period (Triggs 1996).The burrow openings were generally oval shaped and show some size variation. Average height was 37.5 cms and 38 cms, and average width was 39.5 cms and 44.9 cms for warren 1 and 2 respectively ( Table 1 ).

  Warren 1 (n=6)   Warren 2 (n=8)  
Height cms Width cms Height cms Width cms
46 48 49 74
mean (stdev)
37.5 (6.68) 39.5 (8.09) 38 (7.01) 44.87 (13.11)
30 26 28 31

Table 1: Height and width.of  wombat burrow openings of warren 1 and 2 at widest aspect.

The temperatures inside the burrow were significantly different compared to temperatures adjacent to the burrow openings  (F ratio=240.38; P<0.0001; DF=1). Average temperatures  inside were 13.950C and 12.980C and outside were  28.350C and 26.960C for warren 1 and 2 respectively ( Table 2). There was no significant difference between warrens. (F ratio= 1.44; P=0.2419; DF=1).


  Warren 1 (n=6)   Warren 2 (n=8)  
  Temperature in Temperature out Temperature in Temperature out
15.8 32.1 14.6 33.6
mean (stdev)
13.95 (1.18) 28.35 (2.8) 12.98 (1.26) 26.96 (4.54)
12.6 24.6 11.7 22.60

Table 2: Temperatures taken with an Infra red temperature sensor inside and outside wombat burrow openings of warren 1 and 2.

Diurnal temperature and humidity changes
The temperature within the burrow over the temperature during the datalogger timing period was more consistent that outside temperatures.

Figure 1 shows that the lowest temperature was at approximately 6.00 am. Outside temperature was 2.5 0C   near both warrens. Warren 1 and 2 burrow internal temperatures were similar (10.8 C 0  and 60 C). The humidity within the burrow over the same period was also more consistent than outside humidity.

Figure 2, shows that the humidity outside adjacent to warren 1 burrow was higher (>100%) than warren 2 burrow  (approximately 90%). Inside both burrows humidity was similar overnight but in warren 1 humidity rose to over 90% at dawn whereas humidity in warren 2 rose slightly but remained under 90% .

Soil profile temperatures were relatively constant over the range of soil depths measured and was most ameliorated relative to surface temperatures at 40cms deep (Figure 3). 

Diurnal temperature over a soil depth profile of 40 cm

Figure 3: Diurnal temperature in 0C over a soil depth profile of 40 cm.

Rate of gas flow
External wind velocity fluctuated throughout the vertical profile reflecting the gusty breezes of the day. There was no clear relationship between wind velocity and height above the ground. Nevertheless, the wind speed at ground level was 1.5m.sec-1 ( Fig 4). 

Wind profile of burrow

Figure 4:  Wind profile of  burrow 1 in warren 2 in m.sec-1 over a height range of 260 cms.

Air flow was recorded through the centre of all burrows of both warrens, The maximum speed (1 m.sec-1) and the minimum speed (0.1 m.sec-1) was in warren 1 (Table 5). There was a positive correlation for wind velocity between  burrow centre and that of the top and slope for both warrens with the exception of the centre and slope in warren 1.

  Warren 1 (n=6)     Warren 2 (n=8)    
Gas flow (m/sec)
Centre Top Slope Centre Top Slope
1.0 3.0 3.0 0.7 2.2 2.4
mean (stdev)
0.34 (0.32) 2.14 (0.74) 2.0 (0.76) 0.32 (0.2) 1.1 (0.91) 1.07 (0.85)
0.1 1.0 1.0 0.2 0.4 0.4
  0.49 -0.58   0.72 0.84

Table 5: Gas flow in m.sec-1 across the top and mound slope and also flow through the centre of the burrow.

Gas Analysis
Compared to warren 1; the oxygen levels were lower in warren 2 burrows and carbon dioxide levels were higher in warren 2 burrows by one order of magnitude (Table 6).

Warren 1     Warren 2    
Hole No
% O2
% CO2
Hole No
% O2
% CO2
no result for the vial

Table 6: Oxygen and carbon dioxide levels in gas samples removed from burrows in warren 1 and 2.

Airflow patterns within burrows
Three burrows were traced with the smoke generator. Warren 1, burrow 1 was inoculated followed by burrow 2. Within two minutes smoke plumes were emitted from  six burrow openings ( Number;  2,3,4,5,6,7 ) and smaller wisps emerged from small holes of eroded burrow ceilings between opening 3 and 7. Warren 2, burrow 6 conducted smoke to opening 8 and burrow 1 conducted smoke to opening 2.
There was some backflow of smoke from the burrow openings where the smoke was introduced. It was not possible to conclude that the smoke and/or air exchange along the linked burrows, was unidirectional.

The diurnal and seasonal extremes in temperature and humidity experienced at Scott’s Creek can present the wombat with challenges to their thermoregulation. The fossorial habit of wombats has offset the threats these extremes pose. Within the two warrens examined in this survey, there was no doubt that these burrows ameliorated temperature and humidity changes for the inhabitants (Figs 1&2). Wombats have the potential to be several metres from external air, as burrows can be several metres long (Triggs 1996). However, Wilson and Kilgore (1978) have shown that where an animal is more than three body lengths from the opening, external air is no longer available for respiration. Whereas oxygen can be replenished to burrows through pores in the soil, Wilson and Kilgore (1978) concluded that for a large mammal, diffusion alone does not supply enough oxygen to prevent hypoxia-hypercapnia. Gas diffusion is unlikely to be a sole air source for an animal the size of a wombat, therefore another type of ventilation  must be acting.

The wombat burrows observed in this survey were all continuously ventilated, although since there was evidence that only one of these burrows was occupied, it is unlikely that all of these burrows were aerated by the “piston effect” of wombat activity giving rise to air movement. Instead, three possibilities for mechanisms of ventilation are possible. The aspect of these warrens may assist air flow in these burrows in that they appear on a slope which would assist convection of air.  While increased airflow may be an outcome of the design, these warrens have probably not been specifically sited for the sole purpose of ventilation. Placement of warrens on sloping ground is not unusual for wombats as there is less danger from flooding from surface runoff, although  wombats generally tend to exploit natural weaknesses in varied types of  terrain which do not always include slopes (Triggs 1996). Where there is minimal wind there may be other factors acting for the gas flow occurring in these burrows.  Air flow may be maintained within the burrow on windless days by convective gas movement as a result of differences in air density, a phenomenon which has been documented in mole burrows by Olszewski and Skoczen (1965). The significant difference seen between temperatures inside and outside these burrows may generate a gas density gradient that would explain air flow in these circumstances. This phenomenon may also explain continued airflow in the burrows where there was a negative correlation between outer and inner airflow. The airflow within these burrows could be explained by the viscous entrainment  of gasses acting in conjunction with wind action, causing ventilation. However, unlike the mechanisms of airing described for prairie-dog burrows by Vogel et al. 1973, the wombat burrows observed in this study are not ventilated by unidirectional wind-induced bulk flow. The wombat burrow openings would need to be symmetrical around a vertical axis for the operation of  Bernoulli’s principle, and they are clearly not. The generation of viscous entrainment within the burrow can be explained by airflow across the top of the burrow as suggested by the positive correlation between airflow inside with airflow outside the burrows.  The  significance of the negative correlation seen for one burrow within warren 1 is unclear and may result from local variability of wind direction at that time. This presumably indicates that the aspect of that particular burrow mouth was not sited for a positive correlation with wind from that direction. It is likely that the variability in the aspect of these horizontal burrow openings allows an interconnected warren to optimise wind flow from several directions for ventilation. A clearer picture of the significance of wind flow and direction may be gained by extending the number of warrens surveyed and making replications of the flow measurements. Also a directional flow anenometer may allow improved interpretation of the correlations between wind speed and airflow within the burrows.

While smoke entrainment demonstrated  interconnectedness and gas flow between the burrows within warrens, each individual burrow may not be guaranteed optimal ventilation. There was increased carbon dioxide levels seen within the two sets of interconnected burrows of warren 2. It has been documented by Arielli (1978) for mole rats that when a burrow is occupied,  carbon dioxide levels may rise. More burrows were interconnected within warren 1 than warren 2 which may explain lower carbon dioxide levels as a function of airflow within warren 1. Even though the velocity of the air flow through the burrows of  warren 2 was similar to those of warren 1, the increased carbon dioxide levels  may have been due to the presence of a wombat. This burrow had been visited overnight and was probably still occupied when the gas samples were taken. These observations show that wombat burrows are continuously, but not neccessarily uniformly ventilated. Viscous entrainment contributes to ventilation, being largely generated by above ground windflow and to a lesser extent, convective flow. The burrow ventilation may also be optimised by the variable aspects of the burrow openings and enhanced by opportunistic warren site aspects such as slopes.The wombat burrow is a simple and apparently haphazard construction that has many sophisticated functions. It supplies shelter from predators and bushfires while allowing the animals to deal with problems of thermoregulation by cushioning them from the elements. Over evolutionary time these animals have continued to create a ventilated environment which will hopefully guarantee them a safe haven now and into the future.

This project was funded by a grant from the Natural Heritage Trust. I would like to thank my co-workers M. Harris (Fossorial man), A. Kerby, M. Le Duff, and S. Phillips for their able technical expertise in the field and for the data analysis that followed. Thanks also to Professor R.V. Baudinette for his advice in all aspects of this project

Arielli, R. (1978). The  atmospheric environment of the fossorial mole rat (Spalax ehrenbergi): Effects of season, soil texture, rain, temperature and activity Comparative Biochemistry and Physiology 63A  569-575

Baudinette, R.V. (1974). Physiological correlates of burrow gas conditions in the California ground squirrel  Journal of Comparative Physiology  48A 733-743

Boggs, D.F, and Kilgore, D.L. (1983).  Ventilatory responses of the burrowing owl and bobwhite to hypercarbia and hypoxia  Journal of Comparative Physiology  149  527-533

Maclean, G. S. (1981). Factors influencing the composition of respiratory gasses in mammal burrows  Comparative Biochemistry and Physiology  69A  373-380

Olszewski, J. L. and Skoczen, S. (1965) The airing of burrows of the mole Talpa europaea  Linnaeus, 1758  Acta Theriologica Vol. X,11 181-193

Triggs, B. (1996). Common Wombats in Australia   NSW  Press Australian Natural History Series

Vogel, S., Ellington, C. P., Jr. and Kilgore, D. L., Jr. (1973).   Wind-induced ventilation of the burrow of the prairie-dog, Cynomys ludovicianus   Journal of Comparative Physiology  85      1-14

Wilson, K. J. and Kilgore, D. L. Jr. (1978).  The effects of location and design on the diffusion of respiratory gasses in mammal burrows  Journal of Theoretical Biology  71  73-101

Return to Home page