Killing Technique of North American Badgers Preying on Richardson’S Ground Squirrels

Color profile: Disabled Composite Default screen

2109

NOTE

Killing technique of North American badgers preying on Richardson’s ground squirrels

Gail R. Michener and Andrew N. Iwaniuk

Abstract: Carcasses of 13 Richardson’s ground squirrels (Spermophilus richardsonii) cached during autumn by North American badgers (Taxidea taxus) in southern Alberta, Canada, were inspected to determine the capture and killing technique. Regardless of prey size (251–651 g) or torpor status (normothermic or torpid), badgers killed ground squirrels with a single grasping bite directed dorsally or laterally to the thorax. The canines and third upper incisors of badgers generally bruised the skin without puncturing it, but caused extensive hematomas on the thoracic musculature and penetrated between the ribs, with associated breakage of ribs and hemorrhaging in the thoracic cavity. Internal organs and bones other than ribs were usually not damaged. Thoracic bites, rather than nape or throat bites, are used by several mustelids, including North American badgers, when capturing small prey (<10% of the predator’s mass). Résumé : Les carcasses de 13 Spermophiles de Richardson (Spermophilus richardsonii) trouvées dans des caches de Blaireaux d’Amérique (Taxidea taxus) dans le sud de l’Alberta, Canada, ont été examinées dans le but de déterminer les techniques de capture et de mise à mort utilisées par les blaireaux. Indépendamment de la taille des proies (251– 651 g) ou de leur état de torpeur (normothermie ou torpeur profonde), les blaireaux tuent les spermophiles d’une seule morsure dorsalement ou, latéralement, au thorax. Les canines et les troisièmes incisives supérieures du blaireau font généralement des ecchymoses sur la peau sans la perforer, mais causent des hématomes importants sur la musculature thoracique et pénètrent entre les côtes; les animaux ainsi blessés ont souvent des côtes cassées et des hémorragies dans la cage thoracique. Les organes internes et les os autres que les côtes ne sont ordinairement pas endommagés. Les morsures au thorax plutôt qu’à la nuque ou à la gorge sont utilisées par plusieurs espèces de mustélidés, y compris le Blaireau d’Amérique, au cours de leurs chasses aux petites proies (<10 % de la masse du prédateur). [Traduit par la Rédaction] 2113

Introduction Note typical musteline manner, with bites that penetrate the nape or throat, but kill small mammals and birds (<10% of the Relatively little is known about the killing technique of predator’s mass) by damaging the heart and major blood North American badgers (Taxidea taxus, family Mustelidae), vessels with a crushing bite to the thorax that often does not largely because they are nocturnal fossorial hunters that involve penetration of the skin. Given that North American capture, then immediately consume, most prey underground badgers are large (usually >6000 g) relative to the size (Messick 1987; Michener 2000). The most commonly of common mammalian prey such as ground squirrels reported prey-killing techniques of mustelids, particularly (Spermophilus spp., usually <500 g), pocket gophers mustelines, are dorsal bites directed to the nape of the neck (Geomys spp., usually <250 g), and mouse-sized murid ro- or the occiput of the skull and ventral bites to the throat dents (Microtus spp. and Peromyscus spp., usually <50 g) (Heidt 1972; Byrne et al. 1978; Rowe-Rowe 1978; Vargas (Lindzey 1982; Marti et al. 1993), badgers may likewise di- and Anderson 1998, 1999). The killing technique can, how- rect killing bites to the thorax rather than the nape or throat. ever, vary with prey size. For example, Ben-David et al. Occasional reports of aboveground captures of ground squir- (1991) observed that marbled polecats (Vormela peregusna) rels and pocket gophers by North American badgers, either kill large mammalian prey (>50% of predator’s mass) in the during daylight hours in natural habitats (Sawyer 1925; Schwab 1978) or at night in captivity (Lampe 1976), provide Received May 1, 2001. Accepted September 12, 2001. limited information on killing techniques. Only Lampe (1976) Published on the NRC Research Press Web site at reported the target of the killing bite, which he noted to be at http://cjz.nrc.ca on December 4, 2001. about the center of the body of a pocket gopher. As part of a long-term ecological study of Richardson’s ground squirrels 1 G.R. Michener. Department of Biological Sciences, (Spermophilus richardsonii) (Michener 1998, 2000), we had 4401 University Drive, University of Lethbridge, Lethbridge, a unique opportunity to inspect ground squirrels captured AB T1K 3M4, Canada. A.N. Iwaniuk. Department of Psychology and Neuroscience, and cached by badgers and thereby assess killing technique. 4401 University Drive, University of Lethbridge, Lethbridge, AB T1K 3M4, Canada. Methods 1Corresponding author (e-mail: [email protected]). The presence of predators was noted on a near daily basis

J:\cjz\cjz79\cjz-11\Z01-169a.vp Tuesday, December 04, 2001 11:10:17 AM Color profile: Disabled Composite Default screen

2110 Can. J. Zool. Vol. 79, 2001

Fig. 1. Representative examples of trauma to hibernating Richardson’s ground squirrels (Spermophilus richardsonii) captured by North American badgers (Taxidea taxus). (a) A pair of bruises, inflicted by upper canines without skin penetration, on the inner dermal surface of the left thorax. The distance between the midpoints of the 2 bruises is 35 mm. (b) Hematomas (darkly pigmented areas) on the left and right thoracic cage of the same individual (mass = 444 g) as in a.(c) Dermal hematoma (darkly pigmented area located dorsally on reflexed skin) and 3 puncture wounds in the left thoracic musculature (indicated by toothpicks penetrating these wounds) of another individual (mass = 421 g).

during a 14-year study (1987–2000) of a population of Richardson’s likely inflicted damage, skulls of 2 North American badgers (adult ground squirrels. The study population was located 5 km east and female PMA75.17.2 and adult male PMA86.8.11) obtained from 1 km south of Picture Butte in southern Alberta, Canada (49°52′N, the Provincial Museum of Alberta were matched against visible 112°43′W, elevation 870 m asl), on a 1.4-ha area surrounded by damage. cultivated fields and a farmyard into which both ground squirrels and predators intruded (Michener 1996). All ground squirrels were individually identified with a numbered tag in each ear. From 1988 to 1999, some ground squirrels (<10% annually) were individually Results fitted with a temperature-sensitive radio transmitter attached to a 3–5 mm wide collar, then located twice daily in their sleeping Of Richardson’s ground squirrels cached by North Ameri- chambers or hibernacula to monitor body temperature (Michener can badgers, a minority (31% of 16 radio-collared animals 1992). and 27% of 11 non-radio-collared animals) exhibited exter- Although direct observations of hunting were infrequent, over- nal signs of trauma, in all cases limited to inconspicuous night hunting by badgers was evident the following morning from blood stains either beside 1 nostril (n = 3), on the thorax the appearance of freshly disturbed soil beside large (≥20 cm diame- (n = 4), or on the throat (n = 1). Necropsies confirmed that ter) tunnels. All aboveground mounds of soil created by badgers puncture wounds penetrating the skin were uncommon (found were searched for carcasses of ground squirrels. Additionally, when in 2 of 13 animals). In contrast, bruises and subcutaneous the signal from a radio-collared ground squirrel was displaced to a hematomas were visible on the inner dermal surface of 11 new underground location coincident with badger excavation near the squirrel’s last location, the radio collar was disinterred. From ground squirrels (Fig. 1). Dermal bruising corresponded in 1990 to 1999, 27 Richardson’s ground squirrels were found cached location to trauma of underlying thoracic musculature. singly either underground (n = 11 radio-collared animals) or above Large hematomas were present on the thoracic muscula- ground (n = 5 radio-collared and 11 non-radio-collared animals; ture of all 13 necropsied ground squirrels, usually (n = 10) Michener 2000). Carcasses cached underground by badgers were on both the left and right sides (Fig. 1b). Additionally, all retrieved for examination 0–11 days (mean = 2 days) after caching. carcasses exhibited discrete, roughly circular wounds to the Most carcasses (24 of 27) were cached in September–November, thoracic musculature, with at least 1 wound penetrating into when the majority of ground squirrels were hibernating (Michener the thoracic cavity through the intercostal muscles (Fig. 1c). 2000). All 27 carcasses were examined for external signs of badger- Ten ground squirrels had broken ribs. Numbers of thoracic inflicted injury. Four carcasses cached in autumn 1992 (1 non- radio-collared and 3 radio-collared) and 9 cached in autumn 1998 wounds and numbers of broken ribs did not differ signifi- (all radio-collared) were stored at –20°C for subsequent necropsy. cantly between male and female ground squirrels, between In July 1999, carcasses were thawed to room temperature, weighed torpid and nontorpid ground squirrels, or between ground to the nearest 1 g, then inspected for trauma to the skin, thoracic squirrels killed by the badger hunting in 1992 and the badger and peritoneal walls, internal organs, and skeleton. We recorded hunting in 1998 (Table 1). Neither the number of thoracic the identity of broken bones and the number and location of bruises, wounds nor the number of broken ribs was correlated with hematomas, and muscle wounds. To assess which teeth had most mass (range 251–651 g) of the 13 cached carcasses (Spearman’s

J:\cjz\cjz79\cjz-11\Z01-169a.vp Tuesday, December 04, 2001 11:10:26 AM Color profile: Disabled Composite Default screen

Note 2111

Table 1. Effect of sex, body temperature (assessed by radiotelemetry) at the time of capture, and year of capture on the number of wounds in the thoracic musculature and the number of broken ribs for Richardson’s ground squirrels (Spermophilus richardsonii) killed and cached in autumn by North American badgers (Taxidea taxus). No. of thoracic No. of n wounds* broken ribs* Sex Females (mass = 284 ± 23 g)* 4 3.8 ± 2.1 2.5 ± 2.4 Males (mass = 476 ± 93 g)* 9 4.6 ± 2.6 3.4 ± 2.7 Body temperature Normal 2 4.0 ± 1.4 3.0 ± 0.0 Torpid 10 3.9 ± 2.1 2.6 ± 2.2 Year 1992 4 6.0 ± 2.6 3.8 ± 3.8 1998 9 3.6 ± 1.9 2.9 ± 2.1 Note: Body masses of female and male Richardson’s ground squirrels differed significantly (t = 3.98, P < 0.01). Numbers of thoracic wounds and broken ribs did not differ significantly with sex, body temperature, or year (Mann–Whitney U tests, P > 0.75 for all comparisons except number of wounds between years, for which P > 0.10). *Mean ± SD.

rank correlation, rS = 0.17, P > 0.55, and rS = –0.24, P > the bite perpendicularly over the squirrel’s dorsum. On the 0.40, respectively). other 10 squirrels, tooth contacts were asymmetrical, located The number of wounds (mean ± SD, with the range in pa- dorsolaterally on one side and ventrolaterally on the other rentheses) was 2.5 ± 1.7 (1–6) on the left thorax and 1.8 ± side, indicating that the bite was angled laterally across the 1.2 (0–4) on the right thorax, for a total of 4.3 ± 2.4 (1–9) squirrel’s body. For all 13 ground squirrels, a single bite di- thoracic wounds per animal (n = 13). Of wounds that pene- rected to the thorax was sufficient to inflict the observed trated through the intercostal muscles, most (36 of 42) were wounds. located posterior to the 5th rib and anterior to the 11th rib, For 31 radio-collared Richardson’s ground squirrels killed with the modal penetration point between the 8th and 9th by badgers, we excavated the area around the capture site to ribs (13 punctures). Comparison of the size and site of punc- determine how badgers obtained access to prey. Badgers tun- ture wounds with skulls of 2 badgers indicated that the larg- neled underground for 37–270 cm (mean ± SD = 152 ± est puncture wounds were inflicted by the upper and lower 55 cm, n = 15 systems with relevant measurements) to reach canines, with smaller punctures attributable to the canine- the wall of 27 chambers and dug downward for 27, 28, 36, like third upper incisors. and 49 cm to reach the roof of 4 chambers. Three torpid For the 10 Richardson’s ground squirrels with broken ribs, animals were presumably clawed out of the hibernaculum the number of broken ribs (mean ± SD, with the range in pa- because the only entrance into the chamber was a ground- rentheses) was 2.0 ± 0.9 (1–4) on the left thorax and 2.1 ± squirrel-sized tunnel ≤10 cm in diameter. One of the latter 2.2 (0–8) on the right thorax, for a total of 4.1 ± 2.1 (1–9) squirrels was necropsied; it exhibited more dermal bruising per animal. Most breakages (37 of 41) involved the 5th than other necropsied ground squirrels. through 11th ribs, and most ribs broke near the dorsal or Radio collars were not damaged on any of 16 Richard- ventral end (23 and 7, respectively) rather than near the mid- son’s ground squirrels radiotelemetered at the time of cap- point. The only additional osteological damage was to the ture and caching by badgers in 1990, 1992, 1993, or 1998. scapula of 1 squirrel that had been bitten on the anterior tho- Of 26 radio collars inspected after the squirrel had been rax. eaten by a badger, either on the night of capture or after All 13 Richardson’s ground squirrels exhibited hemor- cache retrieval (Michener 2000), 73% were undamaged but rhaging in the thoracic cavity, and 3 animals also had blood 5 had tooth prints in the collar and 2 had broken collars, in- in the peritoneal cavity. In no animal did thoracic hemor- dicating that tooth contact with the radio collar was more rhaging appear to have resulted from laceration of the heart, likely during consumption than during capture of ground though 2 squirrels had damaged lungs. Peritoneal hemor- squirrels (Fisher’s exact test, P = 0.03). rhaging in 1 squirrel was associated with a gash on the liver On 8 occasions badgers were observed transporting a Rich- that coincided with a puncture penetrating the abdominal ardson’s ground squirrel in the mouth from the capture site skin, but the other 2 animals had intact peritoneal organs. In to another site at which the animal was subsequently cached 1 of the latter squirrels, blood had leaked into the peritoneal or eaten. Usually distance and poor light prevented assess- cavity from the thorax via a wound piercing the diaphragm. ment of how the badger was holding the carcass, but photog- Based on the locations of bruises, hematomas, and punc- raphy was possible for 2 transports. Both squirrels were held tures, tooth contact by badgers was confined to the thorax on at midcarcass so that the forequarters and hindquarters drooped 9 squirrels, the thorax and throat on 2 animals, and the tho- to each side of the badger’s mouth. Of 12 necropsied radio- rax and adjacent peritoneum on 2 animals. Tooth contacts collared animals, 1 was cached within the ground squirrel’s were symmetrically positioned in nearly equivalent bilateral burrow system, 1.3 m from the hibernaculum, whereas the locations on 3 animals, indicating that the badger had aimed other 11 were transported above ground for a distance of

J:\cjz\cjz79\cjz-11\Z01-169a.vp Tuesday, December 04, 2001 11:10:26 AM Color profile: Disabled Composite Default screen

2112 Can. J. Zool. Vol. 79, 2001

14–133 m (mean ± SD = 64 ± 41 m, based on the straight- enabling them to grasp prey with just sufficient force to ef- line distance between capture and cache sites) before cach- fect killing with minimal piercing of the skin. We do not ing. Neither the number of thoracic wounds nor the number know whether a controlled bite was also used on Richard- of broken ribs was correlated with the distance these 11 ani- son’s ground squirrels that were consumed immediately on mals were carried above ground by the badger (rS = –0.08, capture or whether badgers selectively cached only those P > 0.75, and rS = 0.40, P > 0.20, respectively). squirrels that had been minimally damaged on capture. We assume that badgers usually began eating Richardson’s ground squirrels from the head as reported by Hetlet (1968) and Discussion Lampe (1976) for badgers eating golden-mantled ground For some mustelines, particularly when killing relatively squirrels (Spermophilus lateralis) and plains pocket gophers large prey, the initial bite may be followed by additional (Geomys bursarius) and by Heidt (1972), Llewellyn (1942), bites, manipulation of the prey with the feet, and reorienta- Powell (1978), and Rowe-Rowe (1978) for a variety of other tion of the prey to achieve a killing bite on the neck or throat mustelids. The lack of damage to most radio collars in- (Miller 1931; Heidt 1972; Powell 1978; Rowe-Rowe 1978; spected after the badger had eaten the ground squirrel sug- Vargas and Anderson 1998). In contrast, our necropsies of gests that radiocollars usually fell free with no tooth contact Richardson’s ground squirrels captured and cached in au- by the badger, further implying that North American badgers tumn by North American badgers indicated that killing was exercise bite control during both killing and consumption of achieved by means of a single grasping bite directed to the Richardson’s ground squirrels. thorax. The pressure of the badger’s canines and third upper incisors, which generally bruised the skin without puncturing Acknowledgements it, caused extensive hematomas on the thoracic musculature, penetration wounds between the ribs, often accompanied by We thank D. Charge, D. Michener, and D. Pankhurst for breakage of ribs, and hemorrhaging in the thoracic cavity. assistance with recovery of carcasses cached underground, When marbled polecats kill small prey (<10% of the preda- B. Weimann, Provincial Museum of Alberta, for the loan of tor’s mass), the initial bite to the thorax likewise serves as badger skulls, and S. Pellis for comments on the manuscript. the killing bite, usually by crushing the thorax with the mo- This research project was supported by the Natural Sciences lars (Ben-David et al. 1991). and Engineering Research Council of Canada through an op- Whereas long-tailed weasels (Mustela frenata) and black- erating grant to G.R.M. and a postgraduate scholarship to footed ferrets (Mustela nigripes) use a nape bite when hunt- A.N.I. ing ground-dwelling sciurids above ground and a throat bite when capturing the same prey species in confined spaces References underground (Byrne et al. 1978; Vargas and Anderson 1998), Allen, D.L. 1938. Notes on the killing technique of the New York North American badgers did not direct bites to either the weasel. J. Mammal. 19: 225–229. nape or throat when killing Richardson’s ground squirrels Anderson, D.C., and Johns, D.W. 1977. Predation by badger on underground. Of the Richardson’s ground squirrels we in- yellow-bellied marmot in Colorado. Southwest. Nat. 22: 283– spected, most were in torpor when captured underground by 284. badgers, and so would have been in a curled posture, with Ben-David, M., Pellis, S.M., and Pellis, V.C. 1991. Feeding habits the dorsum uppermost, and unresponsive to the proximity of and predatory behaviour in the marbled polecat (Vormela peregusna the predator. The concentration of traumatic injuries on the syriaca): I. Killing methods in relation to prey size and prey be- thorax and the lack of tooth marks on the radio collars of haviour. Behaviour, 118: 127–143. cached Richardson’s ground squirrels indicate that the tho- Byrne, A., Stebbins, L.L., and Delude, L. 1978. A new killing rax was the intended target of killing bites by badgers. Mar- technique of the long-tailed weasel. Acta Theriol. 23: 127–131. bled polecats likewise target the thorax if prey are small or Heidt, G.A. 1972. Anatomical and behavioral aspects of killing and nondefensive (Ben-David et al. 1991), and striped polecats feeding by the least weasel, Mustela nivalis L. Arkansas Acad. (Ictonyx striatus) and American martens (Martes americana) Sci. Proc. 26: 53–54. sometimes use a thoracic bite on small prey (Rowe-Rowe Hetlet, L.A. 1968. Observations on a group of badgers in South 1978; Spencer and Zielinski 1983). Badgers occasionally Park, Colorado. M.S. thesis, Colorado State University, Fort Collins. capture large ground-dwelling squirrels such as yellow-bellied Lampe, R.P. 1976. Aspects of the predatory strategy of the North marmots (Marmota flaviventris) (Verbeek 1965; Anderson American badger, Taxidea taxus. Ph.D. thesis, University of and Johns 1977), but whether the target of the killing bite Minnesota, St. Paul. shifts from the thorax to the throat or nape with increasing Lindzey, F.G. 1982. Badger Taxidea taxus. In Wild mammals of North America: biology, management, and economics. Edited by prey size or greater prey defensiveness, as occurs in marbled J.A. Chapman and G.A. Feldhamer. Johns Hopkins University polecats (Ben-David et al. 1991), is not known. Press, Baltimore. pp. 653–663. Killing bites of badgers resulted in minimal piercing of Llewellyn, L.M. 1942. Notes on the Alleghenian least weasel in the skin of Richardson’s ground squirrels, a pattern that has Virginia. J. Mammal. 23: 439–441. also been reported for some other mustelids (Allen 1938; Long, C.A. 1965. Functional aspects of the jaw-articulation in the Vargas and Anderson 1998). Long (1965) noted that the North American badger, with comments on adaptiveness of tooth- well-developed temporal and masseter muscles of North Amer- wear. Trans. Kans. Acad. Sci. 68: 156–162. ican badgers permit a powerful bite. Nonetheless, badgers Marti, C.D., Steenhof, K., Kochert, M.N., and Marks, J.S. 1993. evidently can control this bite when hunting prey that are Community trophic structure: the roles of diet, body size, and small (251–651 g in this study) relative to the badger’s size, activity time in vertebrate predators. Oikos, 67: 6–18.

J:\cjz\cjz79\cjz-11\Z01-169a.vp Tuesday, December 04, 2001 11:10:27 AM Color profile: Disabled Composite Default screen

Note 2113

Messick, J.P. 1987. North American badger. In Wild furbearer Powell, R.A. 1978. A comparison of fisher and weasel hunting be- management and conservation in North America. Edited by M. havior. Carnivore, 1: 28–34. Novak, J.A. Baker, M.E. Obbard, and B. Malloch. The Ontario Rowe-Rowe, D.T. 1978. Comparative prey capture and food stud- Trappers Association, Toronto. pp. 587–597. ies of South African mustelines. Mammalia, 42: 175–196. Michener, G.R. 1992. Sexual differences in over-winter torpor pat- Sawyer, E.J. 1925. Badger runs down ground squirrels. J. Mam- terns of Richardson’s ground squirrels in natural hibernacula. mal. 6: 125–126. Oecologia, 89: 397–406. Schwab, E.R. 1978. An observation of badger predation on Rich- Michener, G.R. 1996. Establishment of a colony of Richardson’s ardson ground squirrels. Prairie Nat. 10: 60–61. ground squirrels in southern Alberta. In Proceedings of the Fourth Spencer, W.D., and Zielinski, W.J. 1983. Predatory behavior of Prairie Conservation and Endangered Species Workshop, pine martens. J. Mammal. 64: 715–717. Lethbridge, Alta., February 1995. Edited by W.D. Willms and Vargas, A., and Anderson, S.H. 1998. Ontogeny of black-footed J.F. Dormaar. Occas. Pap. No. 23, Provincial Museum of Alberta, ferret predatory behavior towards prairie dogs. Can. J. Zool. 76: Natural History, Edmonton. pp. 303–308. 1696–1704. Michener, G.R. 1998. Sexual differences in reproductive effort of Vargas, A., and Anderson, S.H. 1999. Effects of experience and Richardson’s ground squirrels. J. Mammal. 79: 1–19. cage enrichment on predatory skills of black-footed ferrets Michener, G.R. 2000. Caching of Richardson’s ground squirrels by (Mustela nigripes). J. Mammal. 80: 263–269. North American badgers. J. Mammal. 81: 1106–1117. Verbeek, N.A.M. 1965. Predation by badger on yellow-bellied Miller, F.W. 1931. A feeding habit of the long-tailed weasel. marmot in Wyoming. J. Mammal. 46: 506. J. Mammal. 12: 164.

J:\cjz\cjz79\cjz-11\Z01-169a.vp Tuesday, December 04, 2001 11:10:27 AM