Evaluation of Techniquesfor Capturing Common Ravens in Southwestern Idaho

K.A. Engel• Environmental Services Department Pacific Power and Light Company 920 SW 6th Avenue, 380 PFFC Portland, Oregon 97204

L.S. Young Snake River Birds of Prey Research Project U.S. Bureau of Land Management 3948 Development Avenue Boise, Idaho 83705

Introduction

ate the relative effectiveness of these capture tech- corax) have captured them using several tech- niques and interpret our results with respect to the niques,esearchers includingstudying rocketCommon nets (MahringerRavens (Corvus 1970, ecology of Common Ravens in our studyarea. Dorn 1972, Kerttu 1973, Stiehl 1978), drop-in traps (Coldwell 1967, Stiehl 1978, • Magoun pers. comm.), Study Area and Methods padded leghold traps (J. Hughes pers. comm.), single- end Havahart traps (Schwan and Williams 1978), a Our studyfocused on ravensroosting communally on a handheldnet gun (R. Ambrosepers. comm.),and a box 124 km segmentof 500 kV transmissionline in south- trap fitted with a manually tripped, trap door (B. western Idaho. Heavily grazed shrubsteppevegetation Heinrich pers. comm.; Table 1). From April 1985 to (West 1983) and agriculture dominate this area. November1986, we employedmost of the abovetech- Topographicrelief increasesfrom east to west as the niques and several new ones to capture Common transmission line crosses the Snake River Plain and Ravens in southwesternIdaho. In this paper, we evalu- enters the foothills of the Owyhee Mountains. Annual

Table 1. Successof techniquesused to captureCommon Ravens in this andother studies.

Capture Totaltrap Ravens Capturerate Study Technique season(s) time(days) captured (days/raven)

Coldwell(1967) Drop-intrap Winter <180 750 <1 Mahringer(1970) Rocketnet Allseasons 38 148 <1 Dorn(1972) Rocketnet Summer • 26 <2 Kerttu(1973) Rocketnet Winter-Spring <30 21 <1 Schwanand Williams (1978) Havaharttraps Fall • 11 <5 Stiehl(1978) Drop-intrap Winter-Spring 101 99 1 Stiehl(1978) Rocketnet Spdng&Fall 159 6 27 Youngand Engel (1988) Legholdtraps AllSeasons 180 23 8 Youngand Engel (1988) Rocketnet Spdng 18 1 18 Ambrose(pers. comm.) Netgun Winter 2 2 1 Heinrich(pers. comm.) Boxtrap Winter 4* 91 <1 Hughes(pers. comm.) Legholdtraps Winter 30 36 <1 Magoun(pers. comm.) Drop-intrap Winter 40-50 43 <1

*This trap was pre-baited7 to 42 daysprior to eachof the 4 capturedays.

]Currentaddress: Department of WildlifeEcology, University of Wisconsin, Madison, Wisconsin 53706.

Jan.-Mar. 1989 North American Bird Bander Page 5 Evaluation of Techniquesfor Capturing CommonRavens (cont.) precipitation averages 20 cm (U.S.D.I. 1979), most of Sixteen food items, most of which were carrion, were which occursfrom late autumn to early spring. used for bait (Table 3). Ring-necked Pheasant (Phasianus colchicus) and black-tailed jack rabbit We attempted to trap and transmitter-equip ravens (Lepuscalifornicus) were the most commonbaits used, from communal roosts on the transmission line in order followedby whole or quartered calves(Bos taurus) and to describe their movements and habitat use. Traps deer mice (Peromyscusmaniculatus). Cagedlive domes- were set near roost sites in areas where ravens had tic mice (Mus musculus), various cereal grains, and recently been observedforaging. Padded #1 single and chicken(Gallus gallus) eggswere alsooccasionally used #2 double flat-spring leghold traps (Woodstream as bait. Manufacturing Co., Lititz, Penn.), a drop-in trap (Kalmbach 1939), and rocket nets (Dill and We patterned our drop-in trap after that describedby Thornsberry 1950) were the most commontechniques Stiehl (1978). The trap was baited with carrion, usually employed (Table 2). We also attempted to capture a whole or quartered calf. The drop-in trap was always ravens using bal-chatri traps (Berger and Mueller pre-baited for several days before 1 or 2 live decoy 1959), noosecarpets (Anderson and Hamerstrom 1967), ravens were placed inside. Two 9 x 18 m rocket nets a handheld net gun (Coda Enterprises, Inc., Mesa, and a net launcher equipped with a 12 x 12 m net were Ariz.), and a net launcher (Coda Enterprises, Inc., used in areas where terrain allowed. Rocket net loca- Mesa, Ariz.). tions were always pre-baitedfor several days before a net was installed. Once a net was installed, we usually Leghold traps appeared to be the most effective tech- waited at least 2 days beforearming and manning it to nique and, therefore, were employedmost frequently allow ravens to adjust to its presence.Bal-chatri traps (Table 2). Leghold traps (#2's only) were prepared by were either circular or Quonsetshaped and baited with weakeningthe springswith vise grip pliers and spray- one-three live domestic mice. Noose carpets consisted ing them with fiat black paint. Strips of 9.5 mm foam of 10 crn squares of plastic hardware cloth equipped rubber were cut to fit the leading edge of the jaws and with numerous3.6 kg-test, brown, monofilamentnoos- secured with spirally wrapped, black vinyl, electrical es. Noosecarpets were either set around baits or on fre- tape. Traps were chained together in groups of quently used perches. We also attempted to capture two-four. Proceduresfor setting legholdtraps were sim- ravens using a net gun equippedwith a 3.7 m square ilar to those describedby Harmata (1984). Traps were net. placed in shallow depressions excavated in the soil around a bait. Squares of fiberglass insulation (7 x 7 Ages of captured ravens were determined by palate cm) were placed under trigger pans to prevent soilfrom color accordingto criteria describedby Kerttu (1973). collecting beneath them and to increase pan tension. To assesstrapping effort, we consideredtraps set before Traps were then coveredwith finely sifted soil, usually dawn to be open 30 min before sunrise and traps set collectedon site to ensure that it closelymatched the until dark to be closed at sunset. For leghold traps, 1 surroundingsubstrate. Two to 17 traps were used in a trap-hour refers to a set of traps open for 1 hour. Pre- singleset, dependingon trap and bait sizes.A set was baiting dayswere not includedin captureeffort calcula- consideredto be the group of traps associatedwith a tions. singlebait. Results

Table 2. Capture rates for techniquesused to cap- Trapping Success.During 1,430 trap-hours, 24 ravens ture CommonRavens, April 1985-October1986. were captured:23 with legholdtraps and i with a rock- et net (Table 2). Sixteen of these were adults •2 years Trap-hours/ old), sevenwere subadults•1 and <2 years old), and Technique Trap-hoursRavens captured* raven onewas a youngof the year (•1 year old). Legholdtraps 1015(71%) 23(15A, 75, 1Y) 44.1 Of the 16 items used to bait legholdtraps, 8 were suc- Drop-intrap 234(16%) 0 cessfulin luring at least i raven into capture: pheas- Rocketnets 141(10%) 1 (1A) 141.0 ant, rabbit, adult cattle, mice, and lamb carcasses,plus Bal-chatritraps whole or quartered calves and live mice (Table 3). & noosecarpets 22 (2%) 0 Successrates for legholdtraps did not differ significant- Netgun ly amongthe eight baits with which at least one raven & netlauncher 18(1%) 0 was caught (log-likelihoodratio, G = 2.29, df = 7, P > 0.90). Successrates for leghold traps differed signifi- Total/Mean 1430 24 59.6 canfly amongseasons (log-likelihood ratio, G = 9.17, df = 3, P < 0.05; Table 4); successwas highest during win- *A = adult; S = subadult;Y = young of the year. ter and spring, and lowest during summer.

Page 6 North AmericanBird Bander Vol. 14, No. 1 The response of ravens to baits was inconsistent and Table 3. Capture rates for baits used with leghold unpredictable. Ravens seemed less apt to approach traps to capture CommonRavens, April 1985 - larger bait items suchas cattle,jack rabbits,and pheas- October 1986. ants, and frequently observedother animals feed before they approachedeven pre-bait carcasses.Smaller baits, Bait Traphours Ravens Traphours/ suchas mice, seemedmore readily approachedif ravens captured raven located them.

Pheasant 236(23%) 4 59.0 Groupsof nearby ravens were always disruptedby the Rabbit 229(23ø/ø) 4 57.3 capture of a conspecific.Ravens respondedto such cap- Calf 179(18%) 4 44.8 tures by flushing, vocalizing,and usually leaving the Deadmice 107(11%) 5 21.4 area promptly. Livemice 88 (7"1o) 2 44.0 Discussion Chickeneggs 28(3%) 0 Cow 22(2%) 2 11.0 Our overall trapping successwas poor compared to Lamb 16(2%) 1 16.0 other studies (Table 1). Drop-in traps and rocket nets Grain 16(2%) 0 were effective in most studies which employedeither; Squirrel 15(1%) 0 however,the drop-in trap failed to capture ravens in our Badger 14(1%) 0 study, and successusing a rocket net was extremely Quail 14(1%) 0 low. The net gun, which captured two ravens in two Chicken 7 (1%) 0 attempts in Alaska (R. Ambrosepers. comm.),was also Horse 2 (<1%) 1 2.4 unsuccessfulin our study. Even padded legholdtraps, Elk 2 (<1ø/o) 0 our most effective technique, were considerably less Dog 1 (<1%) 0 effective than in the i other study which employed them (J. Hughes pers. comm.). Similar to this study, however,most other studiescaptured ravens primarily Total 1015 23 44.1 during winter and spring.Only three often studiescap- tured ravensduring other times of year (Table 1). Raven Responseto Trapping. Ravens appearedwary of changeswe introduced into their environment. Visible We believeour relativelylow successwith legholdtraps capture deviceswere usually approachedindirectly or was not due to inadequate concealment of traps. One not at all. The drop-in trap occasionally lured one to would expect,if this was the case,that we would have three ravens to the ground next to the trap, but none captured primarily less-experienced,juvenile ravens. ever entered. Ravens only occasionallyfed on baits However,most of our captureswere adults. We suspect associated with rocket nets or the net launcher, but our poor trapping successwas due to a combinationof even then, no more than two ravens were observedfeed- four factors: interference by other species during ing simultaneously.Attempts to use the net gun failed, legholdtrapping, a history of raven persecutionin the becausewe were unable to approachwithin firing range studyarea, prior experiencewith coyote(Canis latrans) (10 m) of ravens. Ravens maneuvered around noose car- control measures, and an abundant, year-round, food pets and did not land on bal-chatri traps. supply. Legholdtrapping was sometimesinterfered with by non-target captures.Coyotes, and domesticdogs and Table 4. Seasonalcapture rates for legholdtraps cats were attracted to baits and sometimes either used to capture Common Ravens,April 1985 - removed baits or were captured in traps. Ravens were October 1986. also frequently accompaniedby Black-billed Magpies (Pica pica) which fed more readily on carcassesthan Trap Ravens Traphours/ ravens did. Consequently,magpies were often captured Season hours captured raven first and in front of nearby ravens. The response of ravens toward captures of other specieswas similar to Winter(Dec- Feb) 63 (6%) 3 21.0 their responsetoward capturesof conspecifics. Spring(Mar- May) 254(25%) 11 23.1 Ravensare commonlyshot in our study area (U.S.D.I., Summer(Jun- Aug) 385(3•'/o) 4 96.3 F &W.S., unpubl. data) which may explain their appar- Autumn(Sep- Nov) 312(31%) 5 62.4 ent warinesstoward trapping activities.Knight (1984) observedstronger avoidance behavior by ravensnesting Total/Mean 1015 23 44.1 in areas of high persecutionthan by those nesting in

Jan.-Mar. 1989 North American Bird Bander Page 7 Evaluation of Techniquesfor Capturing CommonRavens (cont.) areas with low persecution. Goodwin (1976) also Literature Cited observedexceptional corvid wariness toward changesin their environmentin areas where they had been perse- Anderson,R. IC, and F. Hamerstrom. 1967. Hen decoys cuted. aid in trapping cockprairie chickenswith bownets and noosecarpets. J. Wildl. Manage. 31:829-832. Ravens are also occasionallycaptured in legholdtraps Angell, T. 1978.Ravens, crows, magpies, and jays. Univ. set around large carcassesintended for coyotesand are WashingtonPress, Seattle. commonly flushed or shot at coyote bait stations Berger, D. D., and H. C. Mueller. 1959. The bal-chatri: (U.S.D.A., Anita. Damage Control pers. comm.). Given a trap for the birds of prey.Bird-Banding 30:18-26. nearly a century of exposureto these activities,ravens Coldwell, C. 1967. Raven banding in Nova Scotia. may have learned to exercisecaution toward large car- Inland Bird-BandingAssoc. News 39:68-69. casses. Kilham (1985) also observed raven wariness Dill, H. H., and W. H. Thornsberry. 1950. A cannon-pro- toward baits in New Hampshire and attributed such jected net trap for capturing waterfowl. J. Wildl. behaviorto prior experiencewith coyotetrapping. Manage. 14:132-137. Dom, J. L. 1972. The CommonRaven in JacksonHole, Characteristics of the food supply for ravens in our Wyoming.M.S. Thesis,Univ. Wyoming,Laramie. study area may partially accountfor the apparent dis- Goodwin, D. 1976. Crows of the world. Cornell Univ. interest ravens often showedtoward baits. Agricultural Press, Ithaca, N.Y. and natural food sourcestogether provide an abundant, Harmata, A. R. 1984. Bald Eagles of the San Luis perennial, food supply for ravens in our study area. Valley, Colorado: their winter ecologyand spring Consequently,our study area supportsbeth exception- migration. Ph.D. Diss., Montana State Univ., ally large nesting(U.S.D.I. 1979) and year-round,roost- Bozeman. ing populationsof ravens (Youngand Engel 1988). We Kaitabach, E. R. 1939. The crow in its relation to agri- believe the continuousexistence of a rich food supply culture. U.S. Dep. Agric. Farm. Bull. No. 1102. decreasedthe oddsof a raven selectinga bait for food. 2 lpp. Kerttu, M. E. 1973. A•ing techniquesfor the Common There appear to exist area-specificdifferences in raven Raven (Corvus corax principalis Ridgeway). M.S. behavior(Knight 1984) that may prove crucial to meet- Thesis, Michigan Technol.Univ., Houghton. ing the objectivesof a study which requirestheir cap- Kilham, L. 1985. Sustainedrobbing of American Crows ture. We recommendthe use of paddedleghold traps for by CommonRavens at a feedingstation. J. Field. capturing ravens in areas where they appear exception- Ornithol. 56:425-426. ally wary of visiblecapture devices. Based on our study Knight, 1• L. 1984. Responsesof nesting ravens to peo- and the results of previous studies, we also recommend ple in areas of different human densities. Condor the use of carrion and live mice for baiting ravens, and 86:345-346. suggestthat trapping effortsfor ravens be concentrated Mahringer,E. B. 1970. The populationdynamics of the during the winter and spring when successis likely to Common Raven (Corvus corax principalis be higher. Ridgway)on the Baraga Plains, UAnse, Michigan. M.S. Thesis, Michigan Technol.Univ., Houghton. Acknowledgments Schwan,M. W., and D. D. Williams. 1978. Temperature regulation in the Common Raven of interior This work was part of a cooperative study between Alaska. Comp. Blochem.Physiol. 60A:31-36. Pacific Power and Light Company's Environmental Stiehl, R. B. 1978. Aspects of the ecology of the Services Department and the U.S. Bureau of Land Common Raven in Harney Basin, Oregon. Ph.D. Management's Snake River Birds of Prey Research Diss., Portland State Univ., Oreg. Project.We gratefully acknowledgethe assistanceof A. U.S. Department of the Interior. 1979. Snake River Brody,1• Bowman,D. Ames, IC Gonzalez,IC Hegstad, birds of prey specialresearch report. U.S.D.I., Bur. and M. Ichisaka, whose hard work allowed us to cap- Land Manage., Boise,Id. ture enoughravens to meet the objectivesof our study. West, N. E. 1983. Western Intermountain Sagebrush We also thank C. Gray and V. Drain who donatedtime Steppe. Pp. 351-374, in N. E. West, ed. Ecosystems and equipment to help us trap ravens. J. Roppe, IC of the world, part 5: temperate deserts and semi- Steenhof,M. Kochert, and S. Wilder providedguidance deserts.Elsevier Publ. Co., Amsterdam. and support throughout the study. B. Heinrich, M. Young, L. S., and IC A. Engel. 1988. Implications of Kochert, R. Knight, J. Marks, IC Steenhof, and S. communalroosting by CommonRavens to opera- Temple critically reviewed drafts of this manuscript. tion and maintenance of Pacific Power and Light Company'sMalin to Midpoint500 kV transmission line. PacificPower and Light Co., Portland, Oreg.

Page 8 North AmericanBird Bander Vol. 14, No. i