OUCC Proceedings 13 (1991)

Kenyan Lava Tube Caves

Proc. 13 Contents.

OUCC Proceedings Index

OUCC Home Page

The 1985 Oxford University Expedition to the Lava Tube Caves of Mount Suswa, Kenya.

Graham Stone


The maasai name for Kenya's Mount Suswa is Ol Doinyo Onyoke, which means "old red man", and aptly describes the low, wide profile of this dormant volcano. The mountain lies only 60km to the west of Nairobi in the Rift Valley, south of the Aberdare escarpment. Mount Suswa has a double caldera; the outer crater is between eleven and twelve miles across and only partially complete. Most of the area within this outer caldera is a wide flat meadow in which the maasai graze their cattle. At the centre of the mountain is a second crater which is partially filled with a volcanic plug 5km across. The plug is separated from the rim of the inner caldera by a deep moat, and has hardly been explored.

On the northeastern rim of the outer caldera is an area covered with layers of lava from the volcano's most recent eruptions. This area, together with the Chyulu hills in Tsavo, is well known for its tube caves, extending for many hundreds of metres beneath the surface of the lava flows. Both sets of caves have been the subjects of Oxford University expeditions. Lava tube caves are formed when rapidly advancing tongues of lava are halted for long enough for the exterior of the flow to form a hard crust. If further lava flows into the upper part of the flow, pressure builds up inside the congealed shell, eventually bursting through the tip. If the molten core of the flow escapes, the shell remains to produce a tube cave.

The roof of the Kenyan tube caves is of varying thickness, and usually cools into blocks. In places the roofs collapse into the cave below, creating an entrance. If the surface topography allows, rainwater runs into the collapse hole, carrying sediment into the rocky talus formed by the collapsing roof. The soil that collects in the collapse holes is protected from the extremes of temperature and humidity that occur on the surface, and supports a variety of plants and animals not found elsewhere in the lava fields. The caves have been explored for many years by the Cave Exploration Group of East Africa (CEGEA), and are known to be home for leopards and baboons.

The principal aims of this expedition were as follows:

  1. To describe, photograph, map and record the species of plants and animals found in and around the caves, concentrating particularly on the collapse holes.
  2. To study the movement of various small animals within the caves, concentrating on the way in which the most abundant ones use the entrances as a pathway between the external and cave environments.
  3. To study the variation in physical factors (such as humidity and temperature) in and outside the caves in an attempt to explain some of the patterns in animal and plant species which we find.

Internal Features of the Caves

Lava tube caves exhibit a remarkably wide range of features, certainly rivalling the more common limestone cave despite a far simpler and more rapid mode of formation. These features have never been described in detail, either in the Suswa caves or in lava tubes elsewhere.

The cave floors, when not covered by acidic soil or bat guano up to 2m thick, show intricate, deeply etched complexes of lava ripples, like a solidified river. Occasionally these ripples become lava ropes where the mobile lava stream has pulled its congealing surface into ridges. These often have latticed crystalline interiors. Such features are particularly pronounced in the least accessible caves. The roofs of the caves are equally dramatic. In places large areas of the roof have fallen to the floor, sometimes breaking through into a lower cave system or blocking the tube completely. They leave considerable gaps in the otherwise semi-circular cave roofs, often revealing the perfectly straight cooling cracks that formed as the lava solidified 240,000 years ago. Extremely sharp stalactite patches also occur on the roofs, notable not for their size (never more than 12cm in length) but for the exact regularity of their distribution.

Superimposed on these and other features there is sometimes a shiny melt glaze, the result of gas fusion effects as the incandescent lava evacuated its tube. The remarkable colours of the lava, ranging from brilliant orange through every shade of red to grey brown and blue, make the caves places of great beauty, especially where the vivid sunlit green of the vegetation in a drop hole is visible in the distance.

The Microclimate of the Caves and Collapse Holes

The microclimate of the caves depended on the diameter of the cave, the distance from the cave entrance, and the form of the cave entrances. Caves with two large entrances often had a breeze running from one end to the other. Wide open caves showed much greater fluctuations in temperature and humidity than those with more tortuous entrances. 30m into the largest tunnels the temperature and humidity were constant at 16.5° C/72%, despite variation in the collapse holes between 5-25° C and 46-98%. Closed caves, and those in the lower cave levels have a constant temperature and humidity of 21.5° C/98%.

The microclimate of the drop holes is especially interesting, and goes a long way towards explaining their great zoological and botanical diversity. Although not as stable as the caves themselves, the temperature/humidity fluctuations were still not as great as those in the surrounding bush, which was hot and dry during the day and very cold and humid at night. In contrast, the drop holes never dried out, nor did they become very cold; thus they formed a perfect temperature environment for many species that would not otherwise be found on Suswa.

The Vegetation of the Cave Area

A considerable period of time was devoted to sampling the main vegetation in the different cave area habitats. The major plants were used as indicators of distinctive zones, which were quite evident at a distance. The open ground around the collapse holes was dominated by at least seven species of grasses, above which grew Bull's Horn Acacia and Camphor bush (Tarchonanthus camphoratus). The increase in diversity of shrubs and trees around the collapse holes was fantastic, especially around the larger collapses (reaching 200m in diameter). Here we found isolated large trees, including fig, olive and juniper.

The collapse holes can be divided into open and steep-side types. The steep-sided collapses were dominated by a palm-like plant (Dracaena sp.). Almost always the sight of this plant indicated the location of a collapse. Other plants grew below the Dracaena cover, but the soil was generally deep, rich and relatively bare. In the more open collapses the crater bottoms were dominated by a tall shrub (Tavenna sp.). At least seventy species not found in the surrounding bush grow in the collapses.

Mammals of Mount Suswa.

Despite our presence for eight weeks, large mammals regularly came into the camp area. The most regular visitors were a herd of giraffe, and occasionally gazelles. A skull in a cave and numerous piles of droppings led us to suspect that suni, tiny antelopes, were using the area. We were lucky enough to see these beautiful creatures on several occasions at dawn or dusk. The rim of the outer crater to the south of the cave area was also home to a small herd of eland - the largest antelope, and notoriously shy. Despite continued hunting by the maasai, we sighted lions on two occasions, and found their tracks less than 100m from the camp. Leopard tracks were found, and we also found the remains of kills which had been dragged into caves. Our rubbish pit attracted nocturnal visits by white tailed mongooses and zorillas. The most exciting visitor was a genet, a shy catlike animal.

We had intended to mistnet bats, but decided only to work on the invertebrates associated with bat colonies in the cave. These areas of the caves are rarely visited - characterised by a nauseating smell of ammonia, clouds of tiny flies, and a general atmospheric debris of parasites and dust. These, together with intense humidity and the spongy covering of bat guano found on all rock surfaces, were enough to deter most casual visitors.

The cave mouths and collapse holes provided a focus for many animals. Most noticeable were baboons (Papio anubis) which use the collapses in the evenings and mornings as a social centre and feeding ground. There appears to be only a single troupe in the area, which moves throughout the cave area during the day. One large collapse has been christened the baboon parliament by the CEGEA, and has been a focal point for this baboon society for many years.

Systematic work on mammals concentrated on small mammals. 21 groups of 10 live traps were set throughout the cave area over the study period. We caught 104 animals in 11 species, including 3 species of gerbil, squirrels and mole rats. Different species were characteristic of different habitats within the cave area. Gerbils were found only in the open grasslands, squirrels and others only in the collapse holes. Apart from a shrew and a genet, no mammals were caught in the caves. The absence of any tracks in the fine sediment of the cave floors suggested that very few small mammals venture into the caves at all. The cave entrances are often pitted with many cat-like footprints, indicating that the caves are used as lairs. Civets and genets leave dung middens in the caves, and one cave was in use as a hyena den. The shrew would not have aroused much interest had it not fallen into an insect trap 30m below the surface in a second level cave 500m from the nearest entrance. How it got there is uncertain, but it seems likely that its arrival there was accidental and that it survived by eating the abundant cave crickets and beetles. The shrew was found beneath a bat colony, and it is unlikely that it would have survived in the more barren areas of the caves.

Barn Owls on Mount Suswa

At several sites in the cave entrances we found large piles of owl pellets containing the undigested remains of insects and vertebrates. Identification of these remains indicated what the birds were eating, and by comparison with the results of the small mammal study, where they were making most of their kills. The owl pellet middens reach a considerable depth, and allow an investigation of how owl diets have changed over time. Mount Suswa's barn owls live mainly on gerbils and mole rats, residents of the open bush. Rodents characteristic of the denser collapse hole vegetation were rare in the pellets. Mount Suswa owls also eat a lot of big crickets and dung beetles, which in some pellets made up 20% of the total mass of remains. In many cases it was possible to identify the insect to species from the pieces of cuticle in the pellets. European barn owls, in contrast, eat very few invertebrates.

Insects in the Caves and Collapse Holes.

Insects were captured either in pitfall traps, or on cards coated with a sticky compound (hyvis) where the cave floor was too rocky to allow pitfall trapping. We also used small portable mercury vapour lamps surrounded by shallow trays of dilute alcohol. Over 8000 insects and other invertebrates were collected. It became clear that the different habitats had their own characteristic invertebrate faunas. The following is a summary of general results.

  1. The habitats outside the caves and collapse holes are divisible into recognisably different types. These habitats are differentiated by types of plant cover. Collapse hole habitats were differentiated by the nature of the collapse itself and by the vegetation growing in it (affected greatly by the size of the collapse hole and the steepness of its sides). They can also be differentiated by the presence or absence of dung or other animal material, as can the deep cave habitats where the main contributors are bats. These isolated food sources are major concentration points for invertebrates.
  2. Pit fall trapping showed that the most productive habitat of all those sampled for invertebrates was beneath a large bat colony deep in a cave, where beetles were by far the dominant insects. A single coffee cup placed beneath the bat colony for 24 hours regularly contained 100 insects. Over the same time period a pitfall in the open bush would be expected to capture only 7.
  3. The diversity of insect species outside the caves is far higher than it is inside them. Even in the cave entrances away from the bat colonies there is a greater variety of species than in the deep cave.
  4. Some insect species associated with the collapse holes penetrate the twilight zone of the caves. These include small crickets caught in the collapse holes (particularly on baboon dung middens). The black beetle P. temnecostatus and small moths closely related to the common clothes moth also entered the caves. The red carabid beetle Plagiopyga cyclogona, a predator, penetrates from the collapse holes deep into the caves.
  5. Some insects characteristic of the deep cave environment are found in lower numbers in the cave entrances. These include the characteristically hunch-backed cave crickets and two beetles - large Pogonobasis sp. and at least one species of small ptinid spider beetle. The bat parasites and two of the deep cave beetle species were never caught away from the bat colonies.
  6. Some insects -most notably several species of butterfly - ssociated with moist and wooded habitats were found in and around the collapse holes and not in the open savannah. The importance of the collapse holes as "islands' for at least some species (and no doubt many yet to be found) is certain.

A Comparison with other Kenyan Caves

In 1971 John Macfarlane published a study of beetles in bat caves at Ithundu, in a much warmer area of Kenya at lower altitude. He found one beetle, Villersia trivialis, in vast numbers. Another beetle (Alphitobius diaperurius) was also very common. This contrasts greatly with the situation at Suswa. Neither of the beetles named above were caught at Suswa, and the commonest beetle in the Mount Suswa caves (Pogonobasis sp.) existed at a much lower density that the Ithundu beetles. The other beetles found in the Mount Suswa caves are also noted from Ithundu, but seem to be far less abundant there.

This result emphasises that even cave environments, so constant in microclimate compared to the exterior, are nonetheless variable in terms of other characteristics important to the animal life they contain. John Macfarlane recently suggested that the higher altitude and cooler climate of Mount Suswa might be an important reason for the differences between the communities of these two cave systems.

A full account of the scientific work of this expedition is available from Graham Stone, The Department of Applied Biology, Imperial College at Silwood Park, Ascot SL5 7PY.

Expedition members

Graham Stone (21, Queen's College, zoology), Paul Sunnucks (22, Queen's College, zoology), Jeremy Lambert (21, Oriel College, geography), Paul Parsons (21, Magdalen College, zoology), Ali Ibrahim (entomologist, National Museums of Kenya), Benson Katimu, Mzungu Mwalimu and Patterson (watchmen and pyromaniacs).


June 28 1985 Flight London-Nairobi.
June 29-July 9 Expedition preparations in Nairobi.
July 10-July 19 First field research period.
July 20-July 23 Resupply in Nairobi and museum work at the National Museums of Kenya.
July 24-August 6 Second field research period.
August 7-9 Resupply in Nairobi.
August 10-29 Final field research period.
August 30-Sept 13 Museum and Herbarium work.
September 14 Flight Nairobi-London.
The expedition worked closely with the National Museums of Kenya in Nairobi and the CEGEA. We hired an ancient landrover for the entire duration of the expedition, and Benson Katimu as a driver and mechanic. Patterson and Mzungu Mwalimu were hired as askaris to prevent raiding of our camp by the maasai or wildlife during the night. We were able to borrow all of our camping equipment, and obtain the use of a truck to carry it all in, at no charge. Because lions are known to live on Mount Suswa we also borrowed a steel prefabricated building. This provided security for Patterson, Mwalimu and Benson when the British members of the expedition returned to Nairobi for supplies. Water was obtained from the satellite tracking station at Mount Margaret 13 miles to the north.


The expedition could not have succeeded without the help and advice we received from a large number of individuals and companies. Our special thanks go to John and Mary Edwards for their hospitality and helpfulness. They gave us a home in Nairobi and tolerated all our scruffiness. We also thank the National Museums of Kenya, the CEGEA and the Mountain Club of Kenya.