Migration and Mortality of Alberta Ferruginous Hawks

The Condor891169-174 0 The Cooper Ornithological Society 1987

MIGRATION AND MORTALITY OF ALBERTA FERRUGINOUS HAWKS

JOSEFK. SCHMUTZ Department of Biology Universityof Saskatchewan,Saskatoon S7N 0 WO, Canada

RICHARD W. FYFE Environment Canada, Canadian Wildhfe Service,4999-98 Avenue,Edmonton T6B 2X3, Canada

Abstract. Since 1967, 2,444 FermginousHawks (Buteo regalis) have been bandedin southeasternAlberta and 80 recoveriesare examined.The majority of AlbertaFerruginous Hawkswinter in Texas.Upon departingfrom their grasslandbreeding range in September or October,they exhibit a strongtendency to remain in habitat broadly categorizedas grassland.As a result,the hawksmix little with the westerlypopulations frequenting desert shrubhabitat. While in Texas,Ferruginous Hawks are commonnear black-tailedprairie dog (Cynomysludovicianus) colonies regardless of the land use in the surroundingarea. Survivorshipschedules based on band returnssuggest that 66% of the hawksdie during their firstyear. Key words: FerruginousHawk; gene flow; migration;survivorship; grassland: winter ecol-

INTRODUCTION and depart from their breeding range in Alberta Animals may move for many reasons including at a time when the Richardson’s ground squirrel finding food, finding mates, avoiding predators (Spermophilus richardsonii) are entering estiva- and becauseof aggressive interactions. Most an- tiomhibemation (Schmutz et al. 1979). imals disperseat least a short distance from their Although the migratory behavior of Ferrugi- natal to breeding site. Some also exhibit long nous Hawks has not been proven heritable, we distance migrations (see Swingland and Green- assumethat suchheritability exists and therefore wood 1983 for a review). Migration representsa that migration behavior can be influenced by nat- long distance movement followed by a return. ural selection. In this analysis we suggesta pat- The ultimate causeof migration is an avoidance tern of migration which we think is exhibited by of temporarily adverseecological conditions, but Alberta Ferruginous Hawks in general. Since we proximate factors(e.g., food supply, weather) may basethis analysis on reportsof hawks which were trigger the onset and small-scale direction of either dead, sick or injured, we assume that this movement. Migratory movements of most bird subset of individuals behaved in a way similar populations are highly consistent, suggestinga to those not reported. genetic basis to this behavior. Migratory rest- METHODS lessnesshas been shown to have a genetic basis (Berthold and Quemer 1981), and since migra- From 1967 to 1980, 1,9 15 Ferruginous Hawks tory restlessnessis positively correlated with dis- were banded under a permit held by R.W.F. and tance migrated, the latter is assumed to be her- from 1983 to 1985, 529 were banded under a itable also. permit held by J.K.S. The bands were applied In this paper we describe the migratory habit when the nestlings were more than half grown. of the Ferruginous Hawk (Buteo regalis) banded Most banding was done on a 3,000 km2 area in Alberta. This population inhabits the most southeastof Hanna, Alberta but somehawks were northern portion of this species’ range consisting banded throughout southeasternAlberta. Two of of sparsely treed grassland east of the Rocky 74 hawks recovered were banded in southwest- Mountains and south of parkland habitat. The em Saskatchewan. Once a report of a banded hawks rely largely on ground squirrels for prey hawk was received, a questionnaire was sent to the finder requesting additional information on precise location, habitat and condition of the ’ ’ Received2 1 May 1986.Final acceptance13 August hawk when found (freshly killed, partly decayed, 1986. fully decayed).

[I691 170 JOSEF K. SCHMUTZ AND RICHARD W. FYFE

0 512 month, . 13-36 months : 0 s37 months :..: ../ :. .. .

FIGURE 1. Latitude and time of year of recoveries of Ferruginous Hawks. Dates for hawks reported as freshly killed were used directly. From 5 to 45 days dependingon location and time ofyear were subtracted from the finding date of eight hawks reported as partly decayed.

To study their distribution on the winter range, J.K.S. counted Ferruginous Hawks and other raptors seen on a road transect in southwestern New Mexico and northern Texas. The route was from El Paso to Seminole, north around Ama- FIGURE 2. Locationofrecovered Ferruginous Hawks rillo and back to El Paso via Clovis and Roswell, is shown in relation to banding area and wintering range of this species.The solid line outlines the entire New Mexico. This count took place from 3 to 7 wintering rangeofthis speciesbased on 1,532 sightings November 1985. Hawk location, activity, hab- of FerruainousHawks during 37.172 hr of observation itat type, and land use were recorded. To mini- on Audibon ChristmasBird Co&s from 1980to 1983. mize a potential bias in visibility imposed by The brokenline denotesthe “concentrated”winter range and includes counts where more than one hawk was differences in vegetation or terrain, the search seen per 100 km travelled. In Mexico, Ferruginous was concentrated within 200 m of the road. A Hawks were seen on few counts totalling only 112 hr total of 246 raptors were recorded on 1,392 km of observation and hence a concentratedrange is not travelled. shown. The entire range shown there representssight- ings and recoveries combined. RESULTS AND DISCUSSION In someraptor populations, adults remain in their Young also appear to migrate on their own in breedingarea year-round and thesemay be joined Alberta. After fledging, during the secondhalf of by others from other areas (Mebs 1964). Some July and all of August 1985, 10 adults and 20 Ferruginous Hawks are present all year in the young were recorded on a study area south and southern part of their breeding range (Olendorff east of Hanna, Alberta, whereas only four adults 1973, Smith and Murphy 1978) but whether these and no young were seen in September and Oc- were migrants or summer residents was un- tober. One breeding pair was observed on what known. was presumed to be their previously used nest tree on 16 October 1985. One young of this pair TIMING OF MIGRATION was found partly decayedin Texas on 22 October In Alberta, apparently all individuals departed 1985. The locations of recoveries reported as from their northern part of the breeding range freshly killed or partly decayed also suggestthat by late October. Young which were marked in most young Ferruginous Hawks leave Alberta in Utah, departed independently of nest mates and August (Fig. 1). Later, during November through their parents (Woffinden and Murphy 1983). February only one (5%) of 19 individuals was FERRUGINOUS HAWK MIGRATION 171 found north of Colorado and Kansas. Few were TABLE 1. Number of Ferruginous Hawks recorded reported in March through June but the three on Audubon Christmas Bird Counts (American Birds) in 1980 to 1983. Only states with counts of one or adults, 36 months old or older, were found dis- more FerruginousHawks during at least 100 hr of ob- tinctly further north than four younger individ- servation over four years are included. The number of uals. This small sample suggeststhat subadults FerruginousHawks recovered are also given. do not return to their natal area or they spend Totaltime REOV- less time there than adults. Age at first breeding stateor CO”“tw (hr) cries is three years in this speciesbased on a sample of seven known-age adults livetrapped at nests. Kansas 15.0 568 3 Nebraska 7.5 227 6 Colorado 6.5 2,943 4 Wyoming 5.6 108 3 MIGRATION ROUTES New Mexico 5.4 1,821 1 Of 74 Ferruginous Hawks reported dead, 13 died Mexico 5.4 112 5 during the first four months of life within 150 Oklahoma 5.1 901 2 Nevada 4.9 143 km of their banding location. The remaining 6 1 Texas 4.9 5,624 15 locations are shown in Figure 2. The results in- Arizona 4.7 4.220 1 dicate that the migration route and winter dis- Utah 4.2 ‘708 tribution of Alberta Ferruginous Hawks are dis- Washington 3.7 164 tinctly nonrandom in comparison to the total South Dakota 3.2 407 4 California 2.9 18,530 population. While Ferruginous Hawks can be Oregon 2.0 298 found wintering anywhere in the western U.S., Montana 1.8 164 2 Alberta hawks occupied only the midwestern Missouri 1.4 144 statesand most wintered in Texas. On Audubon Christmas Bird Counts (American Birds 1980 to 1983), Ferruginous Hawks were more abundant natural selectionwas operating on Alberta hawks south of about 43”N latitude. The pattern was to cause a change in the areas used. corroborated by a preponderance of Alberta To further compare the east-westdistribution hawks recovered south of Wyoming and South of recoveries of Ferruginous Hawks banded in Dakota during November through February. North Dakota (Gilmer et al. 1985) with those of Thus, while those from Alberta segregatethem- this study we considered all recoveries south of selves from the remainder of the wintering pop- Wyoming (41”N latitude). Of 22 recoveries of ulation further west, their distribution north and North Dakota hawks, 11 (50%) were west of a south more closely reflectsthat of the population line extending the eastern boundary of the Texas as a whole. The preponderance of Alberta hawks Panhandle (1Oo”W longitude). Of 42 hawks origin the midwestem statesin winter doesnot mere- inating from Alberta, 38 (9 1O/o) were recovered ly reflect east-westdifferences in abundance. Fer- west of 1OO”Wlongitude (x2 = 13.18, P < 0.001). ruginous Hawks were common in California but Therefore,North Dakota hawkswere more widely no individuals banded during this study were scatteredeast and west than Alberta hawks and found there (Table 1). Only six Ferruginous the latter overlap in significant numbers only with Hawks were reported from northern Mexico. Our the western segment of the North Dakota pop- data may underestimate the importance of this ulation. Alberta hawks, which have a more region as a wintering area becausereporting rates northerly origin, do not “leap-frog” on migration may differ. over those from North Dakota as is the case in Prior to this study, Ferruginous Hawks were some other raptors (Newton 1979). There was banded in Alberta by Salt (1939). Despite a 30 no significant difference in degreeslatitude be- year lag between studiesand a changein breeding tween the southernmost % of the Alberta (n = distribution in the interim (Schmutz 1984), there 23) and the North Dakota populations (n = 13; was no difference between the two studies in the Mann-Whitney U = 163, P = 0.428). distribution of recoveries.The average longitude Despite an east-west difference in winter dis- of 15 of Salt’s (1939) U.S. recoverieswas 104”W tribution between Alberta and North Dakota and of48 in this study 103”W. Other raptors have hawks, both populations preferentially occupied shifted their wintering locations during less than habitat broadly categorized as grassland (Ham- 30 years’ time (Newton 1979) suggestingthat no mond Inc. 1982). Actually, 83% of 135 Ferru- 172 JOSEFK. SCHMUTZ AND RICHARD W. EYFE

TABLE 2. Countsof FerruginousHawks on roadtransects according to habitatand land usein westernTexas and southeasternNew Mexico. The total distancetravelled was 1,392 km between3 and 7 November 1985. Countsof Red-tailedHawks are shownfor comparison.Percentages are in parentheses.

Cultivation Cultivation (z-90%) and Prairie Shrubland Grassland grassland Dryland Irrigation dog town Total FermginousHawks 3 (4) 2 (3) 9 (13) 11 (15) 24 (33) 23 (32) 72 Hawks/l 00 km 0.68 2.60 5.17 4.64 5.49 79.30 5.17 Red-tailedHawks 17 (29) 9 (15) 2 (3) 13 (22) 16 (27) 2 (3) 59 Hawks/100 km 3.88 11.67 1.15 5.49 3.66 6.90 4.24

ginous Hawks banded within grassland habitat prey comprised 8% (Olendorff 1973, Lokemoen (Fig. 2) in Colorado (Harmata 198 l), North Da- and Duebbert 1976, Schmutz et al. 1980, Gilmer kota (Gilmer et al. 1985), and Alberta (Salt 1939, and Stewart 1983). In desert shrub habitat the this study) were recovered again in grassland. converse was true, lagomorphs comprised 50% This habitat clearly has been altered by agricul- and sciurids only 7% of prey (Smith and Murphy tural practices but the results suggestthat this 1978). It is probable that the most effective hunt- land is still important to the hawks either because ing tactics are different for lagomorph versus ro- of resourcesprovided, becauseof traditional mi- dent prey and that hawks of the subpopulations gration routes or both. Alberta hawks appear to differ in hunting behavior and morphology. Lo- be concentrated east of the Rocky Mountains cal adaptations that may exist, may be preserved along the western portion of grassland in the by reduced gene flow which is evident between southern U.S. (Fig. 2). It appearsthat the hawks subpopulations. do not migrate in a straight line to western Texas but rather veer easterlyalong the mountain range HAWK DISTRIBUTION IN TEXAS and then south. Only three (13%) of 24 recoveries The hawks were more widely scattered during north of New Mexico and Oklahoma were on winter than their natal nestswere in Alberta. The the west side of a line directly connecting the nests of 16 (84%) of 19 individuals that died center of the banding area with western Texas during November to February, were located in (Binomial Test P < 0.001). If the route taken an area of about 840 km2. On their wintering had been direct, 50% of the recoveries should grounds, 84% of these same hawks were found have been on either side. It is probable therefore, over an area 440 times larger. that appropriate habitat servesas a guide to the Ferruginous Hawks show a strong preference migrating hawks in addition to a general south- for grassland habitat during the nesting period ward direction that may be under genetic influ- (Schmutz 1984) and hence it is not surprising ence. This may also explain why Alberta Fer- that this preference is also evident outside the ruginous Hawks were concentrated along the nesting period. In many parts of their range, a western edge of the belt of grassland habitat. well documented decline in population size has Ferruginous Hawks banded in the desert shrub been attributed to a loss of nesting habitat owing habitat of the western U.S. move southward also to cultivation of grassland (Olendorff 1973, and appear to frequent desert habitats in winter Howard and Wolfe 1976, Gilmer and Stewart (Thurow et al. 1980, Perkins and Lindsey 1983, 1983, Houston and Bechard 1984). However, Woffinden and Murphy 1983). It is conceivable during November 1985, Ferruginous Hawks were that the populations which occupy the different abundant in intensively cultivated and irrigated habitats are subtly different ecologically and may areas in Texas (Table 2). In view of their pref- representdifferent “ecotypes.” We compared the erence for grassland during the nesting period, percentageof prey items used during the nesting the high number of Ferruginous Hawks found in period in an effort to approximate the degree of intensively cultivated areas was surprising. difference in the use of food by Ferruginous The hawks were least common in shrubland, Hawks in desert versus grassland habitat. In moderately common in grassland, and most grassland habitat sciurid and geomyid rodents common in cultivated areas. The regularity with comprised on average 77% of prey, lagomorph which Ferruginous Hawks were found near FERRUGINOUS HAWK MIGRATION 173 black-tailed prairie dog (Cynomys ludovicianus) TABLE 3. Survival scheduleof FerruginousHawks. ’ colonies (K. Seyffert, pers. comm.; Table 2) sug- Since some recently banded hawks have yet to be re- gests that this rodent represents an important ported only individuals banded from 1967 to 1980 are used here.Z prey speciesduring winter. The distribution of black-tailed prairie dogs (Bart and Grossenhei- Age alive at der 1964) correspondswell with the distribution interval beginning Fraction Number Fraction of grassland habitat (Fig. 2) in the midwestem (Years) of interval surviving dying dying U.S. Prairie dog colonies were always found in O-l 71 1.00 47 0.66 patchesof grasslandin Texas. Ferruginous Hawks 2 24 0.34 6 0.09 frequented these colonies regardlessof the sur- 3 18 0.25 2 0.03 rounding land use. The Red-tailed Hawk (B. ja- 4 16 0.22 5 0.07 maicensis) was also present in all categories of 5 11 0.16 3 0.04 6 8 0.11 2 0.03 habitat and land use but was relatively uncom- 7 6 0.08 1 0.01 mon at prairie dog colonies and most common in grassland (Table 2). ; 52 0.070.03 3 0.040.01 The type of land use alone was a poor predictor 10 1 0.01 0.00 11 1 0.01 0.00 for the presenceof Ferruginous Hawks. The hia- 12 1 0.01 : 0.00 tus between two clusters of recoveries, one in 13 1 0.01 0 0.00 northern Texas and one in western Texas (Fig. 14 1 0.01 1 0.01 2) appearsto reflect an actual gap in distribution. I This survival schedulerepresents the actualmortality profileincluding natural and man-relatedcauses of mortality. No Ferruginous Hawks were seen there on a 185 2An additionalsix recoverieswere received while this manuscriptwas km portion of the route west and north of Sem- in pressand theseare includedin the mortality sectiononly. inole whereas 72 (6.37/100 km) were seen in the Lubbock-Amarillo area. The latter area corre- hawks’ survival than food shortageswhich may sponds with a cluster of recoveries in northern be expectedto be most severe in late winter and Texas (Fig. 2). The reasonsfor FerruginousHawks spring. The survivorship schedule is presented having been infrequent in or absent from the in Table 3. The data suggestthat 66% of the Seminole area are not obvious. The intensity of young died during their first year of life. The cultivation and the type of land use in theseareas oldest individual in this study was in its 14th were similar. Perhaps subtle differences in land year of life. Some individuals lived to at least use or other ecologicalfactors causeddifferences their 15th and 20th year in earlier studies (see in prey density. Houston 1984). Data based on recoveries must be treated with caution in estimating survivor- MORTALITY ship since it is probable that reporting rates differ Of 1,9 15 Ferruginous Hawks banded between for hawks of different ages(Lakhani and Newton 1967 and 1980, 7 1 (3.7%) have been recovered. 1983, Anderson et al. 1985). Young hawks may The remaining nine recoveredwere color banded die of different causesand more often in places in 1984 or 1985. Of the 71 reports of banded frequented by people, introducing a bias into the hawks received, 43 respondents addressed the data. To overcome this shortcoming a study is known or probable causeof death. The causewas presently underway to estimate survival directly not known for 27 (63%) individuals but shooting using color bands readable at a distance. was suspected in eight. Of the remaining 16 hawks, eight (18%) were struck by a vehicle, five ACKNOWLEDGMENTS (11%) were found injured or sick, one (2%) was This study relied on the cooperation of interested killed by an unidentified raptor, one (2%) struck membersof the publicin threecountries. We aregrate- a power line and one (2%) was electrocuted. ful to findersof dead hawkswho respondedwithout beingmade aware of thesignificance of theirindividual The majority of first year deaths occurred dur- courtesies.We aregrateful to manyindividuals in west- ing the first few months of independence. Of 3 1 em Canadawho climbedtrees and cliffsin their spare yearlings, 8 1% had died prior to January whereas time to band FerruginousHawks. H. Armbruster,U. only 46% of hawks older than one year died dur- Banashand L. J. King filed many band reportsand ing this same period (x2 = 4.93, P = 0.027). This collatedrecovery data. We thank S. M. Schmutz,J. C. Davies and G. Holroyd and an anonymousreviewer suggeststhat migration and the establishment of for suaaestionson earlierversions of this manuscrint. a winter home range is more detrimental to the Financialsupport was provided by theCanadian Wiid- 174 JOSEF K. SCHMUTZ AND RICHARD W. FYFE life Serviceand the Alberta RecreationParks and Wild- lationsin northernSouth Dakota. Condor 78:464- life Foundation, the SpecialAreas Board of Hanna and 470. the Univesity of Saskatchewan. MEBS,T. 1964. Zur Biologie und Populationsdyna- mik des MBusebussards(Buteo buteo). J. Omithol. LITERATURE CITED 105:247-306. NEWTON,I. 1979. Population studiesof raptors. Bu- ANDERSON,D. R., K. P. BURNHAM,AND G. C. WHITE. teo Books, Vermillion, SD. 1985. Problems in estimatingage-specific surviv- OLENDORFF,R. R. 1973. The ecologyof nestingbirds al from recovery data of birds ringed as young. J. of urev of northeastern Colorado. U.S. Int. Biol. Anim. Ecol. 54:89-98. Program Grassl. Biome, Tech. Rept. 211. BART, W. H., AND R. P. GROSSENHEIDER.1964. A PERKINS,M. W., AND W. J. LINDSEY. 1983. Nesting field guideto the mammals. Houghton Mifflin Co., studiesof FerruginousHawks in the Ely BLM Dis- Boston. trict, Nevada. N. Am. Bird Bander 8: 106-107. BERTHOLD,P., AND U. QUERNER.198 1. Genetic basis SALT, R. W. 1939. Notes on recoveries of banded of migratory behavior in European warblers. Sci- FerruginousRough-legged Hawks (B. regalis).Bird- ence 212177-79. Banding 10:80-85. GILMER,D. S., AND R. E. STEWART. 1983. Ferrugi- SCHMUTZ,S. M., D. A. BOAG,AND J. K. SCHMUTZ. nous Hawk populations and habitat use in North 1979. Causesof an unequal sex ratio in Richard- Dakota. J. Wildl. Manage. 47:146-157. son’s groundsquirrels. Can. J. Zool. 57:1849-1855. GILMER,D. S., D. L. EVANS,P. M. KONRAD, AND R. SCHMUTZ,J. K., SCHMUTZ,S. M., AND D. A. BOAG. E. STEWART. 1985. Recoveries of Ferruainous 1980. Coexistenceof three speciesof hawks (Bu- Hawks banded in south-centralNorth Dakita. J. teo spp.) in the prairie-parkland ecotone. Can. J. Field Omithol. 56: 184-l 87. Zool. 58:1075-1089. HARMATA,A. R. 1981. Recoveries of Ferruginous SCHMUTZ,J. K. 1984. Ferruginous and Swainson’s Hawks banded in Colorado. N. Am. Bird Bander Hawk abundance and distribution in relation to 6:144-147. land use in southeasternAlberta. J. Wildl. Man- HAMMONDINC. 1982. World atlas. Maplewood, NJ. age. 48:1180-1187. HOUSTON,C. S., AND M. J. BECHARD.1984. Decline SMITH,G. D., AND J. R. MURPHY. 1978. Biology of of the Ferruginous Hawk in Saskatchewan.Am. the Ferruginous Hawk in Central Utah. Socio- Birds 38:166-170. biology 3:79-95. HOUSTON,C. S. 1984. An unusualstory-record 20- SWINGLAND,I. R., ANDP. J. GREENWOOD,[EDS.] 1983. year longevity of FerruginousHawk. Blue Jay 42: The ecologyofanimal movement. ClarendonPress, 99-101. Oxford. HOWARD,R. P., AND M. L. WOLFE. 1976. Range THUROW,T. L., C. M. WHITE, R. P. HOWARD,AND J. improvement practices and Ferruginous Hawks. F. SULLIVAN. 1980. Raptor ecology of the Raft J. Range Manage. 29133-37. River Valley, Idaho. E G & G Idaho, Idaho Falls. LAKHANI,K. H., AND I. NEWTON. 1983. Estimating WOFL~NDEN,N.-D., AND J. R. MURPHY. 1983. Fer- age specific bird survival rates from ring recov- ruginousHawk postfledgingactivities. N. Am. Bird eries-can it be done?J. Anim. Ecol. 52:83-9 1. Bander 8:94-96. LOKEMOEN,J. T., AND H. F. DUEBBERT.1976. Fer- ruginousHawk nesting ecology and raptor popu-