Juvenile Grouse in the Diet of Some Raptors

j. RaptorRes. 38(3):243-249 ¸ 2004 The Raptor ResearchFoundation, Inc.

JUVENILE GROUSE IN THE DIET OF SOME RAPTORS

VITALI REIF 1 AND RISTO TORNBERG Departmentof Biology,University of Oulu, P.O. Box3000, FIN-90014 Oulu, Finland

KAUKO HUHTALA Tehtaanpuistokatu3, FIN-85310 Sievi,Finland

ABSTRACT.--Grouse(e.g., Tetrao)constitute a significant part of the diet for some raptors. Especially grousechicks are important for severalavian predatorsincluding speciesthat normally prey upon small mammals. To evaluate the impact of avian predation on grouse, we need to know which raptors are hunting which speciesand when juvenile grouse suffer from the heaviest predation. Becausegrouse chicks are difficult to identify in prey samples based on morphological characteristics,we made an attempt to addressthis problem by measuringthe humerussize of grousechicks found in prey remains of Northern Goshawks(Accipiter gentilis), Eurasian Sparrowhawks(Accipiter nisus), Common Buzzards (Buteobuteo), and Northern Harriers (Circuscyaneus). Then, we plotted humerus sizesfrom prey remains againstgrowth curvesof chicksof Capercaillie (Tetra0urogallus), Black Grouse (Tetra0tetrix), Willow Grouse (Lagopuslagopus), and Hazel Grouse (Bonasabonasia). We found that the size of grousechicks in the diet wasbest explained by the massof raptors, but not the fledging date of their young. Ranges of grouse size in the raptor diets were overlapping,suggesting that all four raptor specieshunt grouse chicks at about the same dates. Pressureof the avian predator assemblageon juvenile grouse does not appear to be uniform; smallergrouse species sufibr from heavierpredation during a longer period than larger grouse. KEYWORDS:Northern Goshawk; Accipiter gentilis; Eurasian Sparrowhawk; Accipiter nisus; Common Buzzard; Buteo buteo; NorthernHarrier, Circus cyaneus;grouse,, predation; Finland.

LAGOPODOS JUVENILES EN LA DIETA DE ALGUNAS AVES RAPACES R•SUMEN.--E1lagopodo (Tetrao)constituye una parte significativaen la dieta de algunasrapaces, en especial los pichones del lagopodo son importantes para muchos depredadores de aves incluyendo especiesque normalmente se alimentan de pequefiosmamiferos. Para evaluar el impacto de la depre- daci6n sobre el lagopodo,necesitamos saber cualesrapaces esfftn cazado a cualesespecies y cuando los juveniles del lagopodoseven afectadospor la mayor depredaci6n.Debido a que los pichonesde lagopodo son dif/cilesde identificar en los restosde presasbasados en caracteristicasmorfo16gicas, hici- mos un ensayopara resolverel problema mediante la medici6n del tamafio del numero de los pichones de lagopodoencontrados en los restosdel azor (Accipitergentills), el gavil/tneuroasi/ttico (Accipiternisus), el ratonero comfin (Buteobuteo), y el aguilucho nortefio (Circuscyaneus). Posteriormente ploteamos el tamafio del numero de los restos de presasversus las curvas de crecimiento de, los pichones de Tetrao urogallus,Tetrao tetrix, Lagopus lagopus, y Bonasabonasia. Encontramos que el tamafio de los pichonesde lagopodo en la dieta se entiende mejor a partir de la masa de las rapacesy por la fecha del emplu- mamiento de los juveniles. Los rangos del tamafio del lagopodo en la dieta de las rapacesse sobre pusieron, lo cual sugiereque las cuatro especiesde rapacescazan pichones aproximadamenteen las mismasfechas. La presi6n del ensamblajede los depredadoresaviares en losjuveniles de lagopodo, es aparentementeuniforme; las especiesmenores de lagopodosufren una mayor depredaci6ndurante un periodo mas largo que las especiesde lagopodo de mayor tamafio. [Traducci6n de Cfisar Mfirquez]

The tetraonidsconstitute a significantportion in (e.g., Northern Goshawk [Accipitergentilis], Gyrfal~ the diet of some birds of prey. For some of them con [Falcorusticolus], Golden Eagle [Aquila chrysae- tos]) small game are commonly-taken prey (Huht- ala et al. 1996, Tornberg 1997, Sulkavaet al. 1998, E-mail address: [email protected] Nielsen 1999). For others, such as those that feed

243 244 REIF ET AL. VOL. 38, NO. 3 mainly on small mammals (e.g., Common Buzzard time period this occurred. We compared the ap- [Buteo buteo], Northern Harrier [Circus cyaneus], pearance and the growth of grouse chicks in the EurasianEagle-Owl [Bubo bubo] , Ural Owl [ Strixur- field with the size of juvenile grousein the diet of alensis]), the tetraonids may become an alternative thesefour avian predatorsduring the late nesting prey in poor vole (Microtusspp.) years (Angelstam period in July,when mostprey remainsaccumulate et al. 1984, KorpimSki et al. 1990, Redpath and in raptor nests.

Thirgood 1999, Reif et al. 2001), when they can be MATERIAL AND METHODS an important sourceof food to support breeding. Depending on their size and hunting capacity, The prey material was collected in Finland in the fol- lowing areas:Oulu (goshawk,sparrowhawk, harrier), Ku- raptors kill both juvenile and adult grouse (e.g., P. usamo (goshawk), and Central Ostrobothnia (harrier, Sulkava1964, S. Sulkava1964, Tornberg 1997, Sul- buzzard). Most of the material was collected in Oulu kava et al. 1998, Reif et al. 2001). There are few (65ø00'N, 25ø30'E) and Central Ostrobothnia (64ø00'N, quantitative studiesregarding raptor predation on 24ø00'E). These areas represent coastal lowland with many small•ivers and lakes.Roughly 67% of the area is grouse chicks (S. Sulkava 1964, H6glund 1964, coveredby mosaicof spruce (Piceaabies) and pine forests Gronnesby and Nygfird 2000, Thirgood et al. (Pinus sylvestris)mixed with birch (Betula pubescens). 2000), in part because the latter are difficult to About half of the forested area includes bogs, a large Identify to the specieslevel in prey remains. Five number of them being drained. Cultivated areaswith set- grouse species (Capercaillie [Tetrao urogallus], tlements are situated mainly in river valleys.Kuusamo (66ø00'N,29ø00'E) is a rolling highland area, the highest Black Grouse [Tetrao tetrix], Hazel Grouse [Bonasa hilltops reaching 450 m above the sea level. Area is also bonasial, Willow Grouse [Lagopuslagopus], and characterizedby large lakes.Coniferous forests comprise Ptarmigan [Lagopusmutus]) found in Fennoscan- more than 70% of the area, with fewer bogs than the d•a come in a variety of sizesfrom 350 g (Hazel previous area. Due to lower population density,the area cultivated is less in Kuusamo than in the Oulu-Ostro- Grouse) to 4000 g (Capercaillie cock). The growth bothnia area. rate of their chicksvaries correspondingly (e.g., Se- The prey remains analyzedwere collected during dif- menov-Tian-Shansky1959, Lindfin 1981, Klaus et ferent years between 1966-2003 (Table 1). Prey •emains al. 1990). Large raptor speciesmight prey relatively and pellets were collected from the nestsafter fledging little on grouse chicks generally (e.g., Golden Ea- once a year. Prey remains for goshawkswere collected alsofrom the area surroundingthe nest during the post- gle; Sulkava et al. 1998) or mainly on chicks of fledging period ('after the young leave the nest, but stay large grouse species(e.g., Northern Goshawk;S. in the nesting area). The total number of collected sam- Sulkava1964, Tornberg 2001). However,the latter ples varied among the speciesas follows: goshawks,32 soon escapepredation of small raptor species,such (and 14 from nestareas); buzzards, 74; sparrowhawks,14; harriers,7. We assumedthat most collectedprey remains as Eurasian Sparrowhawks(Accipiter nisus) and the were accumulatedin the nests shortly before fledging, Northern Harriers. becausefemales often clean their nestswhen they rear To assessthe predation impact of the whole small chicks (Dement'ev and Gladkov 1951, pets. ob- communityof raptorson different grousespecies, serv.). we need to know the variation in the sizeof grouse To determine the size of grouse chicks in the diet we measured the length of hu•neri. The humeri of juvenile chicks in the diet of different avian predators. In grousewere distinguishedfrom the adults' by the degree our studyarea, goshawks,sparrowhawks, buzzards, of ossification.There are four grouse speciespresent in and harriers hunt small game on a regular basis. the studyarea: Willow Grouse, Black Grouse,Capercail- The aim of this studywas to determine how these lie, and Hazel Grouse. However, the bones of grouse chickswere not sortedby speciesbecause they could not predators partition this common food resource-- be identified with certainty. When two opposite humeri grouse chicks. We used a large existing prey-re- of the same size were found, we counted this material as mains collection of the Zoological Museum, Uni- one a•fitnal. versityof Oulu, and published data. We expected We compared the size of grouse chicks (expressedin that the selectionof juvenile grouseby these pred- humerus length) found in the diet of each raptor at the end of its nestling period (and post-tledgingperiod for ators would depend: (1) on the size of the raptor the goshawk)with sizesof grousechicks of different spe- that generallyaccords with the sizeof the prey and, cies derived from growth curves (Fig. 1). Given the scar- (2) because of rapid growth of grouse chicks and city of grouse remains in many nests, we were not able temporal variations in breeding times of the rap- to consider each nest as a unit for analysis;therefore, each grouse chick was consideredan experimental unit tors, on the date of capture. In addition, we at- instead.Fledging dates of raptorswere calculatedby add- tempted to assesswhich speciesof grouseare most ing the length of the nestling periods (Cramp and Sim- vulnerable to raptor predation and during which mons 1980) to hatching dates. We calculated hatching SEPTEMBER2004 JUVENILEGROUSE IN RAPTORDIETS 245

Table 1. Juvenilegrouse in the prey-remaincollections at raptor nestsin Finland.

No. OF TOTAL SAMPLES NO. OF MEAN •WITHJUV. juv. HUMEra GROUSE GROUSE IN A SPECIES HUMERI HUMERI SAMPLE SE YEARS OF COLLECTION

Northern Goshawk (nest) 21 40 1.9 0.18 1967, 1970-73, 1975, 1978, 1980, 1984, 1987-90 Northern Goshawk (nest area) 9 18 2.0 0.44 1994-96, 1998, 2001 Common Buzzard (nest) 14 22 1.6 0.29 1997, 1984, 1986-91, 1996 EurasianSparrowhawk (nest) 13 22 1.7 0.16 1985, 1988-91, 1995 Northern Harrier (nest) 7 19 2.7 0.62 1966, 1971, 1976-77 dates by estimatingthe age of chicksfrom wing lengths using a curve calculatedfrom material of the Zoological based measurement data on known-aged birds and a Museum, University of Oulu (Power model: y = nonlinear-regressionmodel: y = x3 - 0.002x'• + 0.325x 4.8104X042•4,/2 = 0.962, N = 16, P< 0.001;Fig. 1). Given - 0.534 (Tornberg unpubl. data). The wing-growthpat- the lack of data for newly-hatchedchicks, the growth tern has been found to be fairly similarfor raptorsweigh- curves start at the age of 10 d. The hatching time of ing around 1 kg (Kenward et al. 2000). For the spar- grouse was obtained from literature (von Haartman et al rowhawk, we used a growth curve developed by Moss 1963-72, Lindfin 1983, Marjakangasand Ttrm515 1997), (1979). We found the followinghatching dates (in Julian P. Helle (pers. comm.), and our own observations.Van- days)for different birds of prey:goshawk, 154.18 + 0.67 ation of hatching dates is available only for Black Grouse (SE; 2 June; N = 146); buzzard, 151.94 + 0.44 (31 May; studied by radiotelemetry in Kainuu, 200 km eastof Outu N = 282); harrier, 167.77 + 1.01 (16 June; N = 52); (Marjakangasand Ttrm515 1997, Marjakangasunpubl. sparrowhawk,172.58 + 0.52 (21 June; N = 90). data). The mean hatching date for 126 Black Grouse The growth patternsof grousechicks (in terms of body broods in 1991-95 was 165.8 _+ 0.421 Julian days (14 mass) were taken from Lindan (1981) for Capercaillie, June). We assumedthat the correspondenttime spanof from Klaus et al. (1990) for Black Grouse, and from Se- hatching period wassimilar for all other grousespecies menov-Tian-Shansky(1959) for Hazel and Willow grouse. Becausesample sizesdiffered among the raptor spe- For the latter two species,the growth patternswere given cies, and were relatively small for all but goshawks,only with calendar dates without exact age reference; there- variations in mean grouse size in the diet (Fig. 2) were fore, we aged those Hazel and Willow grouse with given tested statistically.Correspondence of the size ofjuvende massesaccording to our own observationsand by com- grouse in the diet to the size of different grousespecies paring their mass/ageratios to that of BlackGrouse. The in the wild wasanalyzed only visuallyon the graphs (Figs body massof chickswas convertedinto humerus length 2-4). Capability of the raptors to kill juvenile grouse of different size (i.e., the whole size range; Fig. 4) was derived basedon singlespecimens of maximal and minimal 140 size. In order to find the time period when juvenfie grousesuffer from the heaviestpredation, we plotted d•- 120 mension zones on graphs that correspondedto the full ranges of grouse size taken in raptor diets. On the time •'100 scale, these "predation zones" begin at the point when g any of the grouse speciesreaches the minimal size hunt- • 8o ed by the raptor (Figs. 3a-d). The growth curves of grouse species plotted over the zones indicate when • 6o grouse grow in and out of the "predation zones." Th•s analysiswas done based on the assumption that the full '• ,40- range of grouse size in raptor diet does not change through the summer. Becauseall these raptors are ca- 20- pable of killing even adults of the smallestgrouse, the "predation zones" are not limited on the right side (to- 200 400 600 800 ,000 ,2'00 ,4'00 &o ,800 ward the end of the summer). These zones,when plotted together indicate an overlappingarea in which the s•ze Mass (g) of grousechicks makes them vulnerableto predation by

ß Museum Data all four raptor species(Fig. 4). • Model ------95% Conhdence L•mits RESULTS Figure 1. Model employed to estimatejuvenile grouse The size of juvenile grouse (expressed as hu- humerus size based on mass. merus length) found in the raptor nests differed 246 REIF ET AL. VOL. 38, NO. 3

grouse speciesin the wild at the same time. The smallest chick (the minimal value of humerus size) lOO Goshawk post-fledglng.•,,• was found in the sparrowhawknests (Fig. 3b). Ac- 8o GoshawkBuzzardJ cording to the regressionmodel, this chick'smass was 84 g, which correspondsto the age of about 10 d for Capercaillie,20 d for Black Grouse, 25 d for Willow Grouse and Hazel Grouse. Grouse •o chicks taken by goshawksin the nestling period

20 were clearly larger than chicks of any grouse spe- Jun 20 Jun 30 Jul 10 Jul 20 Jul 30 Au cies (Fig. 2). Sparrowhawksdelivered chicks to the Date nest around the size of the Black Grouse, which were larger than Willow and Hazel grouse chicks Capercailhemale but smaller than female Capercaillie chicks.Chicks • • - Capercadhefemale killed by buzzardswere closestto the size of Black Black Grouse male Grouse chicks, but were smaller than female chicks ..... Black Grouse female

..... Wdlow Grouse of Capercaillie and larger than Willow Grouse

..... Hazel Grouse chicks.The sizeof grousechicks taken by harriers was close to the size of Black Grouse, and larger II• Mean_+1 SE than that of Willow Grouse chicks.Finally, grouse Figure 2. Mean size of humeri of grousechicks in the chicksfound in goshawkdiet during the post-fledg- diet of raptors in relation to fiedging dates of raptors. ing period were about the size of female Caper- The growth curves of humeri of different grousespecies calllie, but smallerthan male chicksof Capercaillie are also presented (see Materials and Methods). Crosses and larger than Black Grousechicks (Fig. 2). show relationship of fledgling periods (horizontal bars) The lowestsize limit of the "four-speciespreda- to humeri lengths of grouseremains (vertical bars). For Northern Goshawks _nvocrosses are shown; (l) late nest- lion zone" (i.e., when grousechicks were hunted hng period (about fledgling date) and (2) data collected by all the raptor species) was 45-mm humerus, dating the post-fledgingperiod. which was limited by the goshawk'sdiet (Fig. 4), and correspondsto the age of about 20 d for Cap- ercaillie, 30 d for Black Grouse, 35 d for Willow significantlyamong the species(one-way ANOVA, Grouse, and 40 d for Hazel Grouse.The upper size F = 11.64, P < 0.001). The grouse chicksin the boundary of the zone (65 mm humerus) is limited prey remains of goshawkswere larger than those by the harrier's diet and ca. correspondsto 30-d- from the nestsof other raptors (Tukey test: spar- old males and 35-d-oldfemales of Capercaillie,45- rowhawks, P < 0.001; buzzards, P < 0.05; harriers, d-old Black Grouse, 65-d-old Willow Grouse and P < 0.001' Fig. 2, Table 2). The largest grouse older than 80-d (adult-size) Hazel Grouse. chicks were also found among the goshawkprey, DISCUSSION whereas sparrowhawkskilled the smallestgrouse (Fig. 2, Table 2). During the post-fiedgingperiod, The sizeof grousechicks in prey remainsof rap- the goshawkskilled larger grouse chicks (• length tor speciesdiffered significantly.However, varianc- = 68 mm) than during the late-nestlingperiod (62 es of sizeswithin each classoverlap considerably ram; t = -2.26, df = 56, P < 0.05). Using stepwise- (Fig. 4). Thus, we were not able to determine ex- hnear regression,we found that the size of the actlywhich grousespecies each raptor wastaking. grousechicks in the diet wasexplained by the mass We found that the goshawktook relativelylarger of the raptor female (F = 14.5, B = 0.0116, R2 -- grousechicks than sparrowhawks(Fig. 2). This was 0.26, P < 0.001), but not by the date of fiedging clearly related to the sizeof the raptors (Table 2). or mean mass of the male. In the diet of goshawks,grouse chickscomprised We plotted variationsof grousesize in the raptor 7% in June, 24% in July, and 41% in Augustby dietsagainst the growth curvesof grouseaccording number (Tornberg 1997). This can be explained to the time when the prey remains were presum- by the growth of the different grouse species, ably accumulated(Figs. 3a-d). This enablesvisual which makes them more profitable as a food item analysisof the correspondencebetween the grouse toward late summer.In the beginning of July,only size in the diet and the size of chicks of different the Capercailliechicks seem to reach the size that SEPTEMBER2004 JUVENILEGROUSE IN RAPTORDIETS 247

a 12o ......

oo Nestling Post-fledgling• v .1oo- 8o r---T:-

2c 20 Jun 20 Jun 30 Jul 10 Jul 20 Jul 30 Aug 9 Jun 20 Jun 30 Jul 10 Jul 20 Jul 30 Au0 9 Date Date Figure 4. Overlap of size range of humeri of grouse chicks in the diet of raptors. The empty boxes are the size range (correspond to the shaded boxes in Fig. 3) and the growth curves of grouse speciesas in Fig. 2. The overlap (the shaded zone) indicates the time period when all the raptor speciescan hunt the grouse of given size range.

i , .' , -•...,:-,s'-.--':' is suitable for the goshawk,while in August chicks

Jun 20 Jun 30 Jul 10 Jul 20 Jul 30 Aug 9 of the smallest grouse species, Hazel Grouse, are Date also large enough to be taken (Fig. 3a). This fits with S. Sulkava's (1964) and Tornberg's (2001) findings that goshawkshunt chicks of Capercaillie 120...... and Black Grouse relatively more frequently than lOO smaller grouse species.However, the mean size of juvenile grouse found in goshawknests was larger than that of availableBlack Grouse and Capercail- lie chicks. This could be due to feeding young

40 hawks on the nest during post-fledgingperiod or becauseof large age difference between the chicks 2O Jun 20 Jun 3O Jul 10 Jul 2O Jul 3O Aug 9 (the youngeststays on the nest long after the older Date ones have left). For sparrowhawks,grouse are not very important prey (P. Sulkava 1964). The mean size of chicks taken by sparrowhawkswere closestto the size of Black Grouse (Fig. 2). This pattern appears -•1oo- reasonable, because grouse chicks are usually found at the end of the Sparrowhawk's nestling period, which may result from the females' hunt- 60- ing. At that time Black Grouse chicks are ca. the 46 same mass as adult Hazel Grouse, which sparrowhawk females often kill in spring before laying

Jun 26 Jun 30 Jul 10 Jul 20 Jul 30 Aug 9 Date (P. Sulkava 1964). This may have contributed to our finding that raptor female size appeared to be Figure 3. The size range of humeri of grouse chicksin the best predictor for the size of juvenile grouse in the diet of raptors in relation to fledging datesof raptors: the diet. (a) goshawk,(b) sparrowhawk,(c) buzzard, and (d) har- der. The boxes contain the median and the 50% of val- ues falling between 25th and 75th percentiles, the whiskersrepresent the highest and the lowestvalue. The position of the boxes on the date scalecorrespond to the of grouse speciesas in Fig. 2. The shaded boxes are the mean fledgling date and the width of the boxes is 2 x "predation zones" and indicate the whole size range of SE (same as horizontal bars in Fig. 2). The growth curves grouse in raptor diet (see Materials and Methods). 248 REIF ET AL. VOL. 38, NO. 3

Table 2. The mean body massof the avian predators from harrier predation in 10-15 d, chicks of Wil- (Cramp and Simmons 1980) and mean body mass of low Grouse stay in the "four-species predation young grouse in raptor diet (estimatedwith regression zone" more than 30 d and all Hazel Grouse chicks model) during the raptor fledging period. older then 40 d and adults can be preyed upon by all the raptors. Thus, predation pressureon juve- MEAN MASS MEAN MASS nile grouse from the assemblageof the avian pred- OF RAPTORS, OF C,ROUSE• ators would not be even; smaller grouse speciesare G G under heaviestpredation during a longer period MALE FEMALE MEAN SD than larger grouse. However, the correspondence EurasianSparrowhawk 140 260 261 125 of the mean size of grouse chicks in the diet of Northern Harrier 350 530 274 95 raptors at the end of their nestling periods to avail- Common Buzzard 690 860 322 141 able juvenile grousein the wild suggeststhat juve- Northern Goshawk 860 1410 428 146 nile Black Grouse are preyed upon by most of these raptor species--sparrowhawks,buzzards, and harriers (Fig. 2). Although the goshawkcertainly Grouse, especiallyjuveniles, are important alter- takes the largest share of grouse, the other three native prey for the Common Buzzard (Reif et al. predators may have a high cumulative effect, es- 2001), which took on averagelarger grousechicks peciallyin yearsof poor vole abundancewhen buz- than sparrowhawks(Fig. 2). The buzzardis capable zardsand harriers switchto juvenile grouseas their of hunting even adult Black Grouse (Reif et al. alternative prey. 2001). Therefore, at the end of nestlingperiod the We acknowledgethat there were possiblebiases maximal size of grouse in the buzzard'sdiet most in our data. Becauseof the scarcity of grouse re- likely was limited by available chicks in the field mains, we had to use each grouse chick found in (Fig. 3c). Thus, during the post-fledgingperiod, the nest as a unit for the analysis.Multiple grouse buzzards probably hunt larger grouse chicks too. remains in one nest cannot be considered inde- Scantydata for harriers alsosuggested that grouse pendent, becausemore than likely they were taken were a fairly prominent part in their diet during by the same hawk. Furthermore, in some samples poor vole years (Thirgood et al. 2000, R. Tornberg the sizesof grousebones were very close to each and K. Huhtala unpubl. data). The mean sizesof other, which may be due to the fact that the prey chicks found in the harrier's diet corresponded belonged to the same brood (i.e., once found, the best to those of Black Grouse chicks (Fig. 3d). whole brood mayhave been killed by the samerap- However,they could be also large chicksof Willow tor). However, because the numbers of grouse Grouse, becausethe habitat used by both species chicks found per nest were low (Table 1), we be- was similar (Redpath and Thirgood 1999). lieve that these limitations have not resulted in se- Basedon the sizeranges of grousein the diet of rious pseudoreplication.Moreover, we suggestthat these four raptors, in the course of development, the small sample sizes of the data we used most grouse chicks undergo predation pressureby all likely affected the evaluation of hunting capabili- raptor species studied. The smallest chicks are ties of raptors only as an underestimation of max- preyed upon by sparrowhawksonly (Fig. 3b). Later imum and minimum sizesof grouse. Becausethe on, while growing, they reach the size suitablefor upper limit of the "four-raptor predation zone" is harriers and buzzards (Figs. 3c, 3d). Finally,juve- defined by the size of juvenile grouse in the har- nile grousesuffer from the heaviestpredation (by riet's diet, underestimation of maximum grouse all fbur raptors) when they reach the size hunted sizesin other raptors would not have an affect on by goshawks(Fig. 3a) and thus fall within the determining the heaviestpredation period. "four-speciespredation zone" (Fig. 4). Chicks of In conclusion, we have found that several birds d•fferent grousespecies become vulnerable to gos- of prey share by size a common resource, grouse hawk predation at different ages(i.e., the youngest chicks, during their nestling periods. Juvenile were Capercaillie and the oldest were Hazel grousewere vulnerableto different avian predators Grouse). The time periods when chicks were un- in the course of their development. Our data clear- der predation pressureby all four raptor species ly indicate that the impact of predation of the also varied according to the growth patterns in whole raptor community on grouse needs to be grouse (Fig. 4). While juvenile Capercaillieescape considered when examining predation on grouse. SEPTEMBER2004 JUVENILEGROUSE IN RAPTORD•ETS 249

ACKNOWLEDGMENTS Owls support lhe alternative prey hypothesis?Oikos 58:47-54. We are grateful to Erkki KorpimSkifor the idea of this studyand to SvenJungellfor providingthe prey and ring- LINDEN,H. 1981. Growth rates and early energy require- ing materialßMarkku Hukkanen and the Ringing Centre ments of captivejuvenile Capercaillic, Tetraourogallus of the Natural Museum of Helsinki Universityprovided Finn. Game Res. 39:53-67. the ringing data. Pekka Helle communicated informa- ß 1983. Metsonja teeren muna-ja pesyekoonvaih- tion on the hatching dates of grouse and Arto Marjak- elusta. (Variationsin clutch size and egg size of Cap- angas provided his unpublished data. Mikko Mtnkkt- ercaillie and Black Grouse.) SuomenRiista 30:44-50 nen, Harto Linden, and two anonymous referees gave 1VIagJ^•OAS,A. ^NDL. TOR 'MAIA.1997. Femaleage and valuable comments on the early versionsof the manu- script. The study at various stageswas financially sup- breeding performance in a cyclicpopulation of Black ported by the Centre for International Mobility (CIMO) Grouse Tetrao tetrix. Wildl. Biol. 3:195-203. of the Finnish Ministry of Education, the Jenny and Antti Moss,D. 1979. Growthof nestlingSparrowhawks (Acczp- Wihuri Foundation, and the Environmental Graduate iternisus). J. Zool.Lond. 187:297-314. School of EnviroNet of the University of Oulu (to V. NIELSEN,O.K. 1999. Gyrfalcon predation on Ptarmigan Reif). numerical and functional responses.J. Anita. Ecol.68. 1034-1050. LITERATURE CITED REDPATH,S.M. AND S.J. THIRGOOD.1999. Numerical and ANGELSTAM,P., E. LINDSTR(SM,AND P. WIDI•N. 1984. Role functional responses in generalist predators: Hen of predation in short-term population fluctuationsof Harriers and peregrineson Scottishgrouse moors. J some birds and mammals in Fennoscandia.Oecologia Anita. Ecol. 68:879-892. 62:199-208ß REIF, g., R. TORNBERG,S. JUNGELL,AND E. KORPI 'MAKI. CRAMP,S. ^ND K.E.L. SIMMONS(EDS.). 1980. The birds of 2001. Diet variation of Common Buzzards in Finland the western Palearctic. Vol. 2. Oxford Univ. Press, Ox- supportsthe alternativeprey hypothesis.Ecography 24' ford, U.K. 267-274. DEMENT'EV,G.P. AND N.A. GI•DKOV (EDS.). 1951. Birds SEMENOV-TIAN-SHANSKY,O.I. 1959. On ecologyof tetraon- of the Soviet Union. Vol. 1. Israel Program for Sci- ids. Lapland State Reserve,Moscow, U.S.S.R. entific Translations,Jerusalem, Israel. SULRAV^,P. 1964. Varpushaukanpesimisaikaisesta kSyt- GRONNESBY,S. AND T. NYG•I). 2000. Using time-lapsevid- tSytymisest5ja ravinnosta.Suomen Riista 17:93-105. eo monitoring to studyprey selectionby breedinggos- SULRAV^,S. 1964. Zur nahrungsbiologiedes habichts,Ac- hawksAccipiter gentilis in central Norway. OrnisFenn. cipitergentilis. Aquilo Set Zool. 3:1-103. 77:117-129. --, K. HUHTAI•A, P. RAJALA,AND R. TORNBERG.1998. H•bGLUND,N. 1964.•ber die Ernahrungdes Habichts Changesin the diet of the Golden Eagle Aquila chry- (Accipitergentilis L.) in Schweden.Viltrevy 2:271-328. saetosand smallgame populationsin Finland in 1957- HUHTALA,K., E. PULLIAINEN,P. JussIIA, AND P.S. TUNIS- 96. Ornis Fenn. 76:1-16. KAro.1996. Food niche of the GyrfalconFalco rusticolus THIRGOOD,S.J., S.M. REDPATH,P. ROTHERY,AND NJ. AE- nestingin the far north of Finland as comparedwith BISCHER.2000. Raptor predation and population lim- other choicesof the species.Ornis Fenn. 73:78-87. itation in red grouse.J. Anita. Ecol. 69:504-516. KENWARD,R.E., V. MARCSTR•SM,AND M. KARLBOM. 1993. TORNnERG,R. 1997.Prey selection of the GoshawkAcczp- Post-nestlingbehaviour in goshawksAccipiter gentills: iter gentillsduring the breeding season:the role of the causesof dispersalßAnita. Behar.46:365-370. prey profitability and vulnerability.Ornis Fenn. 74:15- , R.H. PFEFFER,M.A. AL-BOWARD,N.C. Fox, K.E. 28. RIDDLI•;,E.A. BRAGIN, A. LEVIN, S.S. WAI.I.S, AND K.H. --. 2001. Pattern of goshawkAccipiter gentilis preda- HODDER.2000. Setting harnesssizes and other mark- tion on four forest grouse speciesin northern Fin- ing techniques for falcon with strong sexual dimor- land. Wildl. Biol. 7:245-256. phism.J. FieldOrnithol. 72:244-257. VON HAARTMAN, L., O. HILDI•N, P. LINKOLA, P. SOUMALAI- KLAUS, S., H.-H. BERGMANN, C. MARTI, F. MI•LLER, O.A. NEN,AND R. TENOVUO.1963--72. Pohjolan linnut var- V•TOVIC,ANt)J. WIESNER.1990. Die Birkhfihner. Wit- ikuvin. Vol. 4-5. Otava, Helsinki, Finland. tenbergLutherstadt, Ziemsen, Germany. KoRPI•, E., K. HUHTALA, AND S. SULKAVA.1990. Does Received2 December 2002: accepted7 April 2004 the year4o-yearvariation in the diet of Eagle and Ural AssociateEditor: Juan josd Negro