99 Palaeont. afr., 23, 9~104 (1980) BONE COLLECTING BY STRIPED HYAENAS, Hyaena hyaena, IN ISRAEL by J.D. Skinner Institute for Nature Conservation Research, Department of Zoology, Tel-Aviv University and *Mammal Research Institute, University of Pretoria, Pretoria 0002, South Africa S. Davis Department of Zoology, Hebrew University of Jerusalem G. Ilani Nature Reserves Authority, Israel ABSTRACT Differences in bone collecting behaviour of three species of hyaena and porcupines are dis­ cussed. Observations on feeding behaviour of striped hyaenas are described as well as their habit of carrying pieces away particularly if feeding cubs at maternity dens. At one maternity d~n near Arad the floor of the main cavern was littered with bones which covered an area of 40 m2. Of this 2,0 m2 was sampled and found to contain 267 bones and bone fragments from no fewer than 57 individuals , mainly of domestic species such as camel, donkey, caprovines and dogs. CONTENTS Page INTRODUCTION. .......................................... ... .. ... ....................... ... ........ ... ........ ...... ... ....... ..... ..... ...... ...... 99 PROCEDURE ............................................ .. .......... ... .. ... .. .. ... .. ............. ... .. ... ... ....... .. ... ... ............... .... ....... ... 99 RESULTS .................................................................................................................................................... 100 DISCUSSION .......................................................... ................................................................................... 101 CONCLUSION ........................................................................................................................................... 103 ACKNOWLEDGEMENTS .... ... ........................ ...... ... ..... ................... .. ... ... ........... ... .................. ........... ... ... 1Q3 REFERENCES ........................................................................................................................................... : 103 INTRODUCTION Whether or not hyaenas accumulate bones at den sites has long been a subject of dispute. Some of the confusion has undoubtedly arisen from dif­ ferences in behaviour between the three species of hyaena and from differences in behaviour of the same species in different geographic regions. Thus, Hughes (1954a, b) and Dart (1957) concluded that hyaenas are not important accumulators of bones, rather feeding where they find their food. On the other hand, Sutcliffe (1970) describes bone accu­ mulations in East Africa which he ascribes to Cro­ cuta, and Kruuk ( 1972) alleges that Crocuta does not expressly carry food back to the den to feed its young although they frequently carry objects back to their dens to chew. More recently Bearder ( 1977) reported bone remains in six Crocuta dens in the eastern Transvaal and thereby demonstrated the definite involvement of this species in accumu­ lating bone fragments. Recent studies (Skinner 1975, Mills and Mills 1977, Mills 1978, Owens and Owens 1978) have shown that brown hyaenas, Hyaena brunnea, carry food items away from the site of discovery either to cache them or more frequently to feed their young in maternity dens. Preliminary stl,ldies of striped hyaenas, Hyaena hyaena, in East Africa (Kruuk 1976) have indicated that they are similar in this • Address for all correspondence and reprints. regard. The present paper describes bone accumu­ lations at a cave in Israel which have been attri­ buted to Hyaena hyaena, and some observations on the feeding behaviour of this species. PROCEDURE During studies on Hyaena hyaena in Israel in De­ cember 1977 and January 1978, the opportunity arose to visit at least three maternity dens and four other lairs of this species. The dens were all sit­ uated in thejudean and Negev deserts. The most significant maternity den was discov­ ered in 1974 by a Bedouin goat herd. It was sited some 3,5 km south-west of Arad about 200 m above the valley floor and some 100 m from the top of the plateau. There are at least two en­ trances. The first is through a round hole nearly half a metre in diameter which opens out after some ten metres into a large, totally dark cavern about 100 m 2 and over 1,5 m from floor to ceiling at its highest point. Leading off this cavern are five short passages into small caverns and at least one long passage to another large cavern some 20 m away on a higher level. Another exit from this sec­ ond cavern seemed to be used primarily as a latrine. The Arad maternity den was occupied by a mother and cubs at the time of this investigation, and the main cavern was reserved as an eating place. Hundreds of bone fragments, including 100 skulls, covered the floor of this cavern over an area of 40 m 2 • Because we were reluctant to disturb this extraordinary deposit we sampled two 1 m 2 areas (A and B) 10m apart, and the bones were taken to the Hebrew University for identification. We also made nightly observations on the feed­ ing behaviour of Hyaena hyaena using mounted and hand-held floodlights to which the hyaenas were habituated and floodlights at feeding sites estab­ lished by the Nature Reserves Authority at En Gedi and Sde Boqer. Five dead donkeys were placed one at a time at one of these feeding sites. There was competition with other carnivores at both sites. At En Gedi at least four individual hyaenas were identified plus a pack of six wolves Canis lupus, and about 20 foxes Vulpes vulpes. At Sde Boqer the site was visited by at least three individ­ ual adut and three subadult hyaenas plus three wolves, about six foxes, a caracal, Felis caracal, and at least 20 griffon vultures. RESULTS Striped hyaenas feed one at a time except the subadtilt siblings which may feed together. A don­ key carcass is opened at the anus and the flesh on the buttocks devoured. The ribs are "scissored" using the carnassial shear. After two nights of eat­ ing the first donkey was completely dismembered, devoured or carried away. All that was left was the earless head, the masseter muscle on the one side and a few neck vertebrae and the pelvic girdle stripped of flesh, the latter being the only remain­ ing skeletal part after three nights. Although no exact measurements were made, it was estimated that adult hyaenas would consume 7-10 kg of soft flesh at a "meal", and subadults, which were only about 5,0 kg lighter, each consumed about 5,0 kg. Usually after consuming as much soft flesh as they wanted, a carcass would be slowly dismem­ bered as each hyaena tore off limbs or large pieces and carried them away. It took an adult hyaena 30 minutes of intense effort to tear off a hind leg from an adult donkey. Three individual hyaenas be­ tween them removed both hind limbs and tore out the stomach of the third donkey and carried them away. The head and forequarters were all that re­ mained after one night, and the following day the vultures devoured the remaining soft flesh and left only the head for the hyaenas. A characteristic of all the dens visited was that one area was used exclusively for feeding. It was here that the bones had accumulated, and these consisted mostly of cranial and mandibular pieces, metacarpals, hoof and horn sheaths and bone flakes from domestic ungulates. At no den was there a very large accumulation of bone fragments; this mostly ranged from 20-30 bone fragments per den. The hyaenas occupying the den at Arad had re­ cently brought in a domestic goat, Capra hircus. The rest of the sample (tables 1 and 2) contained 267 bones and bone fragments that represented at least 57 individuals, mainly camel (38 %) , donkey (25 %), caprovines (18 %) and dogs (16 %). Al­ though not weighed and measured, a large propor­ tion of the bone-fragments appeared to be over 50 g. There was no evidence of porcupine gnawing on a large sample of randomly selected bones. TABLE 1 Bones collected from the Arad cave from 2,0 m2 of a 40m2 area covered by bones. The areas sampled, A and B, were of equal size, and the earth beneath was sampled to a depth of 20 em, but no identifiable bones or rodent bones were found in this. AREA A SPECIES LISTING BY ANATOMICAL PARTS Camel: Left side Right side Skull I and 1 maxilla (palate + both dental arcades) Mandible 4 6 Scapula 3 Humerus proximal Humerus shaft Humerus distal Humerus complete Radius proximal Radius shaft Radius distal Radius complete Metacarpus proximal Metacarpus shaft Metacarpus distal Metacarpus complete Acetabulum Femur proximal Femur shaft Femur distal Femur complete Tibia proximal Tibia shaft Tibia distal Tibia complete Metatarsus proximal Metatarsus shaft Metatarsus distal Metatarsus complete Calcaneum Astragalus Phalanx 1 Phalanx 2 Phalanx 3 Shafts of camel size 3 2 2(1 = u* ) I 3 3 I I 3 3(£) 5 5 4 4 8 4 2 2 3 2 ( I = u ) I I I I I 2 I (u) I 4 ( l=u) I 2 2 I I 1 (f) I 1 axis + 2 cervical vertebrae Bos: Mandible Humerus distal Donkey: Skull I Maxilla 1 complete palate with both maxillary arcades (6 . max. teeth + 4 mand. teeth) Mandible Radius complete Femur distal Femur shaft Femur proximal Tibia distal ·Tibia complete Tibia shaft Metacarpus complete Metatarsus complete Phalanx 3 *Epiphysis: u = unfused f= fused 2 2 2 1 2 3 I 4 4 2 (I= horse?) 6 (2=horse?) ( l=u) 2 6 I =horse Goat/ibex/sheep: Horns 16 (1 only with core inside) + 2 cores without horn Complete palate with both arcades 2 Single max. arcades 3 3 Mandible ( + post. parts) 9 5 Acetabulum 3 Humerus Astragalus Metatarsus shaft Metatarsus complete (goat) Ibex d' skull Gazelle d' skull Dog: Skull I Mandible Maxilla/ palate - complete Scapula Humerus distal Humerus proximal Humerus shaft Radius complete Pelvis complete 7 vertebrae Hystrix: 2 2 2 I I palate with both maxillae Pig: Teeth- few Radius proximal Man: mandible AREA B Camel: Mandible Scapula Humerus proximal Humerus shaft Humerus distal Humerus complete Radius proximal Radius shaft Radius distal Radius complete Metacarpus proximal Metacarpus shaft Metacarpus distal Metacarpus complete 2 3 Acetabulum 2 Femur proximal Femur shaft Femur distal Femur complete Tibia proximal l(u) Tibia shaft Tibia distal Tibia complete Metatarsus proximal 3 Metatarsus shaft 4 Metatarsus distal 1 Metatarsus complete I Calcaneum Astragalus (with maxilla) I (without maxilla) 4 I (=I pelvis girdle) I ( I complete lower jaw) 5 ( + I palate with both dental arcades) 2 2 2 I 3 2 4 2 2 8 7 I I 2 (u) + IR(f) 1 Phalanx 1 2 Phalanx 2 Phalanx 3 Shafts of camel size 14 Cervical vertebrae I Donkey: Mandible Humerus Radius complete Radius shaft Radius proximal Femur distal Femur shaft Tibia distal Tibia shaft Tibia complete Acetabulum Metacarpus proximal Metacarpus distal Metacarpus shafts I 3 3 I I 2 2 I 5 Metatarsus proximal 3 Metatarsus distal 2 Metatarsus shafts 2 Metatarsus complete 2 Calcaneum Astragalus Phalanx I Phalanx 3 3 3 2 5 I 2 2 2 2 3 4 ( I = complete) 3 I 2 3 3 3 2 4 max teeth shafts- n/o + I rib Capra: Horns 13 Maxilla Mandible 2 with cores + 2 horn cores i.e. total of 4 cores Scapula Humerus distal Tibia Acetabulum Radius proximal Dog: Skull complete Maxilla Mandible Acetabulum Fox: Maxilla Gazelle: Horn 2 3 2 5 I I I I I 8 5 DISCUSSION - +2 shafts - + I shaft 6 6 1 101 Although predation by spotted hyaenas has been discussed at length during the past fifteen years (Eloff 1964, Kruuk 1972), brown and striped hyaenas are regarded as ineffective hunters that are capable of capturing only the smallest of prey (Mills 1978, Owens and Owens 1978, Kruuk 1976, Skinner and Ilani in press). This difference in hunting ability may be due to the fact that Crocuta are much larger than Hyaena (Skinner 1976, Skin­ ner and Ilani in press) and to social behaviour be­ cause Crocuta live, hunt and feed in clans (Kruuk 1972), whereas Hyaena forage alone or at most in family groups (Kruuk 1976, Mills 1978, Owens and Owens 1978, MacDonald 1978, Skinner and Ilani in press). This social behaviour rna y be re­ lated to brain capacity (Oboussier 1979), since Crocuta has a larger brain and evolved later than Hyaena (Oboussier pers. comm.). However, there are probably other differences in social behaviour 102 TABLE 2 A quantitative faunai iist of mammalian species present in the Arad cave.* A + B COMBINED Minimum Min. No. of No . of Bone counts individuals Bone counts % individuals % A B Camel 49 52 Donkey 24 43 Cow 2 0 Caprovines 31 17 Dog 17 27 Fox 0 1 Hystrix 1 0 Gazelle 1 0 Pig I 0 Man I 0 A B 6 8 101 6 5 67 I 0 2 9 5 48 4 8 44 0 I I 1 9 1 1 0 1 1 0 1 1 0 1 38 14 25 11 1 18 14 16 12 I 1 1 1 I 57 25 19 25 21 or a total of I 140 indi­ viduals over the 40m2 area •Counts not exhaustive but for estimating the minimum numbers of animal species represent~ and for comparing areas A and B. which are not yet fully understood; for example, the Crocuta female is larger than the male, has a well developed penile clitoris and is dominant (Racey and Skinner 1979), but in Hyaena sexual dimorphism favours the male. There are also inter­ esting differences in the behaviour of Crocuta clans in different geographic regions such as East Africa (Kruuk 1972), the Kalahari desert (Eloff 1964), and the eastern Transvaal lowveld savanna (Bearder 1977). These seem to be related to the presence and behaviour oflions. However, Hyaena are best known as scavengers that carry food to a cache or to their dens, partic­ ularly when they have cubs (Harrison 1968, Kruuk 1972, Mills and Mills 1977, Mills 1978, Owens and Owens 1978). The present study con­ firms this behaviour. Not only does Hyaena feed singly (Ilani 197 5, Kruuk 1976, Mills 1978, Skin­ ner and Ilani in press), but, after filling their stomachs, their whole strategy is aimed at dis­ membering the carcass and carrying away large parts. The amount they appear capable of devour­ ing at one meal (7-10 kg) is half that which Crocuta can devour at one time (Bearder 1977). Although Sutcliffe ( 1970) and Bearder ( 1977) support the view that Crocuta does accumulate bones at den sites, it is probable that the accumu­ lations reported by Sutcliffe also resulted from for­ aging activities of Hyaena hyaena, since he made no attempt to separate the activities of various scav­ engers. Morever, Brain ( 1976) has emphasised that any cave which has been open for thousands of years will contain bones brought to it in a var­ iety of different ways. Bearder ( 1977) examined the bones for evidence of porcupine gnawing but found only two examples in a sample of 409. Brown hyaenas are extremely rate in Bearder's study area and probably made no contribution to the bone accumulations he ascribed to Crocuta. On the other hand evidence has shown that Hyaena brunnea usually collects bones at maternity dens (Skinner 197 5, Mills and Mills 1977, Mills 1978, Owens and Owens 1978). Bone collections of 20-30 fragments seen at Hyaena hyaena dens in the present study were very similar to those described by Mills and Mills ( 1977) for Hyaena brunnea. However, it frequently happens that vacant dens are used by other species such as aardvarks, wart­ hogs and porcupines. Porcupines have long been known to accumulate bones in quite large quanti­ ties in their dens in the Kalahari sandveld (Brain 1976). Indeed, Brain suspects that porcupines carry more bones to African caves than any other species, but he does find this behaviour unusual for a vegetarian rodent. Examining the Nossob porcupine bone collections, Brain ( 1976) found that the great bulk (70 per cent) of the l 708 indi­ vidual specimens weighed between 0-50 g and 77 per cent were between l-15 em in length. Further­ more, Brain ( 1976) suggests that bone collecting by porcupines (Hystrix africae-australis) "has to do with the wearing down of the incisor teeth rather than with nutrition". As in other rodents the in­ cisors are open-rooted, grow throughout life and through attrition are kept at usable length. While resting, porcupines select objects and gnaw on them for this purpose. Moreover, Brain ( 1976) be­ lieves collecting behaviour has become compulsive, and they bring back more objects than they re­ quire; for example, only 55,2 per cent of the bones ~rom the Nossob collection showed signs of gnaw­ mg. This remarkable behaviour of porcupines is ap­ parently unique amongst other species with similar incisor teeth, even those of similar size and weight such as Lepus or Procavia. Indeed, the necessary at­ trition takes place through the wear of upper against lower incisors as well as when gnawing food. Furthermore, the porcupines in Israel, Hys­ trix indica, do not gnaw bones (Mendelssohn pers. comm.): In fact Mendelssohn ascribes bone chew­ ing to phosphorus deficiency. It is well known that the veld in southern Africa is deficient in phosphorus, particularly in the sandy areas of the Kalahari savanna (Du Toit and Bisschop 1929, Du Toit, Malan, Louw, Holzapf~l and Roets 1932, 1935a, b, Henrici 1928) and that this results in osteophagia which in turn caused widespread cattle deaths from botulism as early as the eighteenth century in Namaqualand (le Vail­ lant 1796 cited by Henning 1949). Throughout this region phosphates are given to domestic live­ stock to compensate for this deficiency. The reason why porcupines do not die of botulism can be as­ cribed in part to Brain's ( 1976) observation that no more than 15 bone fragments in the Nossob col­ lection showed appreciable traces of fattiness. In other words porcupines show a preference for weathered bones. It is therefore not surprising that the bones in the Arad cave accumulation were not gnawed by porcupines. Although porcupine quills and bones were present on the cave floor, these were prob­ ably from individuals eaten by Hyaena, as Hyaena are known to kill porcupines when the opportunity arises (Skinner and II ani in press). There are poss­ ibly three reasons why porcupines have not been a factor in the Arad collection. First, it is not a phosphorus deficient area (Mendelssohn pers. comm.), and osteophagia hardly occurs in herbi­ vores in those areas. Secondly, there may be spe­ cies differences in porcupine behaviour, or compul­ sive collecting as suggested by Brain (1975) may have evolved in a phosphorus deficient area. Thirdly, the Arad cave is very high and may be in­ accessible to porcupines. The Arad cave has been used by Hyaena as a den, perhaps exclusively for maternity purposes, for many years, perhaps for hundreds of years. This does not preclude the possibility that other 103 agents may also have accumulated bones there. Nevertheless, it is a unique deposit because the bone accumulations have largely been protected from weathering as a result of the narrow tunnel openings to the cave. Moreover, it has been pro­ tected from incidental scavenging and vultures be­ cause of its size and seclusion. Only carnivores seeking such seclusion would find it. CONCLUSION We suggest that the Arad collection is largely, perhaps exclusively, that of Hyaena and we believe that other agents are unlikely to have been respon­ sible. Primitive man is unlikely to have entered a dark cavern through such a narrow tunnel. "Civi­ lised" man in this region was and still is so suspi­ cious of hyaenas (Harrison 1968) that if he had known of the cave he would probably have avoided it. Leopards have been absent from this area in living memory, and porcupines are not im­ plicated. The collection reflects the present large mammal composition of the surrounding area, and the collection is probably the remains of individ­ uals scavenged by hyaenas and brought to the den to feed cubs. ACKNOWLEDGEMENTS T~is research was carried out while one of us U.D. Skinner) was visiting Israel as an exchange scientist, and we thank the National Council for Research and Development of Israel and the South African Council for Scientific and Industrial Re­ search for their generous support. In addition, we wish to thank the following institutions for logistical support, hospital­ ity and accommodation: Department of Zoology, Tel-Aviv University; Director Nature Reserves Authority, Israel; and the Society for the Protection of Nature in Israel. We would like to express our appreciation to Professors A. Zahavi and H. Mendelssohn for their interest and encouragement. REFERENCES BEARDER, S.K. ( 1977). Feeding habits of spotted hyaenas in a woodland habitat. E. Ajr. Wildl.j., 15, 163-280. BRAIN, C.K. (1976). Some criteria for the recognition of bone collecting agencies in African caves. Berg Wartenstein Symposium No. 69. New York, Wenner-Gren Foundation DART, R.A. ( 1957). The Osteodontokeratic culture of Australo­ pithecus africanus. Transv. Mus. Mem., 10, 1-105. DU TOIT, P.J. and BISSCHOP,J.H.R. (1929). The breeding of cattle in phosphorus deficient veld No. I. 15th Rep. Dir. Vet. Serv ., 1059-1159. DU TOIT, P.J., MALAN, A.J., LOUW,J.G., HOLZAPFEL, C.R. and ROETS, G.C.S. (1932). A study of the mineral con­ tent and feeding value of natural pastures in the Union of South Africa (First Report) 18th Rep. Dir. Vet. Serv. 525-557. ---- ( 1935a) Idem. (Second Report) Onderstepoort]. Vet. Sci. 2, 607-648. ---- ( 1935b) Idem. (Third Report) Onderstepoort]. Vet. Sci. 5, 215-270. ELLOFF, F.C. (1964). On the predatory habits of lions and hyaenas. Koedoe, 7, 105-112. HARRISON, D.L. (1968). The mammals of Arabia. Vol. II. Carni­ vora, Artiodactyla, Hyracoidea. 273-279. London, Ernest Benn. HENNING, M. W . ( 1949). Animal diseases in South Africa. J ohan­ nesburg, Central News Agency Ltd. HENRICI, M . ( 1928). The phosphorus content of the grasses of Bechuanaland in the course of their development. 13th and 14th Rep. Dir. Vet. Ed. and Res. , 1077-1208. HUGHES, A.R. (1954a). Hyaenas versus Australopithecines as agents of bone accumulation. Am. ]. phys. Anthrop., 12, 467-486. ---- (1954b). Habits of hyaenas. S. Afr.J. Sci., 51, 156-158. ILANI, G. (1975). Hyaenas in Israel. Israel-Land and Nature. Oct., 10-18. KRUUK, H. ( 1972). The spotted hyaena, a stuqy of predation and so­ cial behaviour. Chicago University Press. ---- ( 1976). Social behaviour and foraging of the striped hyaena (Hyaena vulgaris). E. Afr. Wildl.J.,l4, 91-111. MACDONALD, D.W. (1978). Observations on the behaviour and ecology of the striped hyaena Hyaena hyaena in Israel. Is­ rael]. Zoot. MILLS, M.G.L. (1978). Foraging behaviour of the brown hyaena (Hyaena brunnea, Thunberg, 1820) in the southern Kalahari. Z. Tierpsychol. 48, 113-141. MILLS, M .G.L. and MILLS, M .E.J. (1977). An analysis of bones collected at hyaena breeding dens in the Gemsbok National Park. Ann. Transv. Mus., 30, 145-155. OBOUSSIER, H. (1979). The evolution of the brain and phy­ logenetic development of African Bovidae. S. Afr.]. Zoot., (in press). 104 OWENS, M.J. and OWENS, D. (1978) , Feeding ecology and its influence on social organization in brown hyaenas (Hyaena brunnea, Thunberg) of the central Kalahari desert. E. Afr. Wildl.j. , 16, 113-135. RACEY, P.A. and SKINNER, J.D. (1979) . Endocrine aspects of sexual mimicry in spotted hyaenas Crocuta crocuta. ]. Zool. London, 187, 315-321. SKI 1 1ER, J.D . ( 1976). Ecology of brown hyaena Hyaena brunnea in the Transvaal with a distribution map for southern Africa. S. A Jr.]. Sci ., 72, 262-269. SKINNER, J.D. and ILANI, G. (in press). The striped hyaena Hyaena hyaena of the J udean and Negev deserts and a comparison with the brown hyaena Hyaena brunnea. Israel ]. Zool. SUTCLIFFE, A.J. (1970). Spotted hyaena: crusher, gnawer, digester and collector of bone. Nature, Lond., 227, 1110-1113.