SELECTED LIFE HISTORY ASPECTS AND HABITAT USE BY MERRIAM'S WILD TURKEYS IN OREGON

Transcription

1 SELECTED LIFE HISTORY ASPECTS AND HABITAT USE BY MERRIAM'S WILD TURKEYS IN OREGON by Robert Scott Lutz A THESIS submitted to Oregon Stte University in prtil fulfillment of the requirements for the degree of Doctor of Philosophy Completed September 4, 1987 Commencement June 1988

2 AN ABSTRACT OF THE THESIS OF Robert Scott Lutz for the degree of Doctor of Philosophy in Wildlife Science presented on September Title:Selected life historyaspestnd hbitt use by. Merrim's wild turkey in Oregon.-. Abstrct pprove& Redcted for privcy "--nhn Hbitt use nd reproductive success by Merrim's wild turkey (Melegris gllopvo merrimi) were studied in Wsco County, Oregon, during This low density turkey popultion selectively used forested cover types chrcterized by vriety of structurl fetures, species, nd ge clsses. They voided monotypic forested cover types, used non-forested cover types in proportion to their vilbility, nd never used forested cover types with structure simplified by logging ctivities. Reproductive success nd nesting hbitt were described for 25 Merrim's wild turkey hens. A higher proportion of dults thn yerlings ttempted to nest (75 vs. 25%) nd were more successful (90 vs. 25%). Nests were locted in 5 of 14 vilble cover types; only thinned, mixed conifer ws used more thn it occurred. Thinned, mixed conifer stnds contined 52% of the nests, of which 82% were successful. Two mesures of vegettive screening (shrub density nd visul obstruction t 0.5 m) were significntly greter t sites used for nesting thn t rndom sites. Chrcteristics of trees nd cover types in which Merrim's wild turkeys roosted were described for 36 roosts used by hens during winter, spring, nd summer. Four of 14 cover types were used for roosting; only mture, mixed conifer ws used significntly more thn it occurred on the study re. Sesonl differences in roost

3 chrcteristics (number of trees/site, cnopy cover, slope, nd DBH) were ttributed to flock size nd composition. I suggest tht complexity both within nd mong cover types is n importnt component of hbitt for Merrim's wild turkeys, which should be considered in the evlution of potentil relese sites nd in hbitt mngement plns for existing popultions. I recommend creting nesting hbitt by thinning pole-size stnds in the Dougls-- fir (pseudotsug menziesij) zone nd leving slsh in stnds. Additionlly, dult rther thn yerling hens should be relesed in trnsplnt opertions.

4

5 APPROVED: Redcted for privcy Prof ssor of Wildlife Eco1/bgy in chrge of mjor Redcted for privcy Hed of Deprtment of Fisheries nd Wildlife Redcted for privcy Den of Gr te Schoo11 1 Dte thesis is presented September Typed by Robert Scott Lutz

6 ACKNOWLEDGEMENT Funding for this work ws provided by Oregon Deprtment of Fish nd Wildlife nd Mount Hood Ntionl Forest. Addtionl support ws provided by Oregon Stte University Computer Center nd the Confederted Tribes of Wrm Springs. these gencies for their ssistnce. I thnk the numerous people from A specil thnks to R. Denney for his continued support of the project nd for integrting me into the 'rel world' of wildlife mngement. I thnk my mjor professor, Dr. J. A. Crwford, for his support nd guidnce during our numerous trvils. I lso thnk my grdute committee, Dr. R. G. Anthony, Dr. W. H. Emminghm, Dr. W. C. Krueger, nd Dr. G. L. Zimmermn, for their helpful suggestions. Severl friends contributed to my growth--r. BrndtMiller, B. nd T. Hobugh, K. nd S. MilesIverson, G. Miller, K. Nelson, nd E. Rexstd--I thnk y'll. C. Heth provided importnt ssistnce with field work. I lso thnk my fellow grdute students for their suggestions nd comrdery. I thnk my fmily for their support nd understnding during my eduction. I'm especilly indebted to my der friends, P. nd J. Gober, for their cre.

7 TABLE OF CONTENTS Pge CHAPTER I INTRODUCTION 1 CHAPTER II REPRODUCTIVE SUCCESS AND NESTING HABITAT OF MERRIAM'S WILD TURKEYS IN OREGON 2 ABSTRACT 2 METHODS 4 RESULTS AND DISCUSSION 6 Nesting Success 7 Nesting Hbitt 8 Nesting Success in Reltion to Hbitt 9 Nest Site Chrcteristics 9 MANAGEMENT IMPLICATIONS 10 CHAPTER III SEASONAL USE OF ROOST SITES BY MERRIAM'S WILD TURKEY HENS AND HENPOULT FLOCKS IN OREGON 13 ABSTRACT 13 MATERIALS AND METHODS 15 RESULTS AND DISCUSSION 17 Cover Type Use 17 Roost Tree Chrcteristics 18 Sesonl Differences 19 CHAPTER IV PREVALENCE OF PDXVIRUS IN A POPULATION OF MERRIAM'S WILD TURKEY IN OREGON 25 ABSTRACT 25 MATERIALS AND METHODS 26 RESULTS AND DISCUSSION 27 CHAPTER V HABITAT USE BY MERRIAM'S WILD TURKEY IN OREGON 29 ABSTRACT 29 MATERIALS AND METHODS 31 RESULTS AND DISCUSSION 35 CONCLUSIONS 40 CHAPTER VI CONCLUSION 46 CHAPTER VII LITERATURE CITED 48

8 LIST OF TABLES Tble II.1 Cover types vilble nd used for nesting by rdioequipped Merrim's wild turkeys, Wsco County, Oregon, Pge Comprison of structurl chrcteristics of Thinned, Mixed Conifer stnds used for nesting by Merrim's wild turkey hens with rndomly smpled (vilble) Thinned, Mixed Conifer stnds, Wsco County, Oregon, III.1 Descriptions of forested cover types, percentge of study re occupied by ech cover type, percentge of roosts (N.36) in the cover types used by Merrim's wild turkeys in southern Wsco County, Oregon, Chrcteristics of roost cover types nd trees used by Merrim's wild turkey during three sesons in southern Wsco County, Oregon, V.I Annul nd sesonl mortlity rtes for 4 ge nd sex groups of Merrim's wild turkeys, Wsco County, Oregon, V.2 Annul selection of cover types used by Merrim's wild turkeys, Wsco County, Oregon, V.3 Size (h) of sesonl home rnges for 4 ge nd sex groups of Merrim's wild turkey, Wsco County, Oregon, V.4 Use (7)) of cover types in sesonl home rnges by ge nd sex groups of Merrim's wild turkeys, Wsco County, Oregon, V.5 Structurl fetures of cover types used nd selected by Merrim's wild turkey, Wsco County, Oregon,

9 SELECTED LIFE HISTORY ASPECTS AND HABITAT USE BY MERRIAM'S WILD TURKEYS IN OREGON CHAPTER I INTRODUCTION An understnding of hbitts utilized by wildlife species is necessry for effective mngement of the resource. A fundmentl concept of wildlife mngement is tht nimls utilize hbitt in mnner tht optimizes their survivl nd productivity. Hbitt selection my occur on brod scle (e.g. nnul use of n re by the entire popultion) or my be quite specific (e.g. use of prticulr hbitt by one segment of the popultion during prticulr time). Informtion bout hbitt selection provides insight bout the wys niml popultions re mintined nd, in turn, my be used by mngers to enhnce existing popultions or to select sites for introductions of new popultions. In this work, I exmined hbitt selection by popultion of Merrim's wild turkey (Melegris gllopvo merrimi) which were introduced into Oregon from birds obtined in Colordo, New Mexico, nd Arizon in 1961 (Mce 1965). My gol ws to better understnd hbitt use nd selection by these birds on both brod nd specific bses. To provide context for understnding hbitt selection, I mesured number of ttributes of the popultion.

10 2 CHAPTER II REPRODUCTIVE SUCCESS AND NESTING HABITAT OF MERRIAM'S WILD TURKEYS IN OREGON Abstrct: Reproductive success nd nesting hbitt were described for 25 Merrim's wild turkey (Melegris gllopvo merrimi) hens in northcentrl Oregon during A higher proportion of dults thn yerlings ttempted to nest (75 vs. 25%) nd were more successful (90 vs. 25%). Nests were locted in 5 of 14 vilble cover types. Only thinned, mixed conifer ws used significntly more thn it occurred. Thinned, mixed conifer stnds contined 52% of the nests, of which 82% were successful. Two mesurements of vegettive screening (shrub density nd visul obstruction t 0.5 m) were significntly greter t sites used for nesting thn t rndom sites. We recommend creting nesting hbitt by thinning polesize stnds in the Dougls fir (Psuedotsug menziesii) zone nd leving slsh in stnds. Additionlly, dult rther thn yerling hens should be relesed in trnsplnt opertions.

11 3 Quntittive mesures of nesting hbitts nd nest success of Merrim's wild turkeys cn provide informtion useful for mnging nesting hbitt. Similrly, knowledge of nest-site chrcteristics nd nesting hbitt selection cn be used by mngers for evluting potentil relese sites. Limited informtion is vilble on productivity nd nesting hbitt for popultions of this subspecies. Low nesting rtes nd reproductive success of yerling hens my hve limited reproduction of Merrim's turkeys in New Mexico (Lockwood nd Sutcliffe 1985). Nesting hbitt hs been described s lightly cutover res in mixed conifer stnds in New Mexico (Jones 1981), ponderos pine (Pinus ponderos) in Arizon (Phillips 1980) nd South Dkot (Petersen nd Richrdson 1975), nd white ok (Quercus grryn)-dougls fir in Wshington (Mckey 1982). None of these studies, however, ddressed nesting success in reltion to hbitt selection. Accordingly, the objectives of this study were to determine reproductive success, nest-site selection, nd reproductive success in reltion to vilble hbitt, nd to describe nest site chrcteristics of Merrim's wild turkeys in Oregon. Oreg. Dep. Fish nd Wildl. nd Mount Hood Ntl. For. provided funding for the project. During ll phses of the project, R. R. Denney nd D. P. Longrie were most helpful. We thnk R. G. Anthony, W. C. Krueger, nd T. F. Hensly for their mnuscript reviews nd C. A. Heth for help with collection of field dt. This is Tech. Pp of the Oreg. Agric. Exp. Stn.

12 4 METHODS This study ws conducted in southern Wsco County, Oregon, in 1981 nd Merrim's wild turkeys were introduced into this re in The 135,141-h study re encompssed portions of Mount Hood Ntionl Forest, the Oregon Deprtment of Fish nd Wildlife White River Gme Mngement Are, the Confederted Tribes of Wrm Springs Reservtion, nd djcent privte holdings. Study re boundries were defined by the movements of rdio-mrked turkeys. The re ws locted primrily within the Dougls fir nd ponderos pine forest zones described by Frnklin nd Dyrness (1973). Specific cover types were defined ccording to the dominnt vegettion nd lnd use prctices nd were delineted from LANDSAT stellite imgery. Ok Woodlnd (24.6%) ws the most bundnt of the 14 cover types present; other mjor cover types included Ponderos Pine (17.8%), Mture Mixed Conifer (14.2%), nd Ponderos Pine-Dougls Fir-Ok (12.5%). Elevtion rnged from 610 to 1067 m. Turkeys were trpped from 1 Jnury to 5 April ech yer with modified wlk-in trp (Ligon 1946) consisting of 4 welded wire pnels (1 x 3 m) nd net wire (10 x 10 cm) top. Age nd sex of birds were determined by the presence or bsence of brring on the 9th nd 10th primries nd brest fether colortion (Lrson nd Tber 1980). Trnsmitters, weighing 90 g nd with n opertionl life of months, were ttched with bckpck hrness of rubber tubing covered by brided nylon. Trnsmitters contined n ctivity sensor tht indicted when the bird hd been sttionry for minutes. Hen turkeys were locted 3 times weekly throughout the breeding seson to

13 5 scertin cover types used for nesting nd to determine times of initition of incubtion nd htching. One bird rdioed in 1981 survived nd nested in Clutch size ws determined by counting eggs in the nest fter incubtion begn. A nesting ttempt ws defined s the construction of nest nd lying of 1 egg. A successful nest ws one in which 1 chick htched. Hbitts within which hens nested were clssified into 14 cover types. Structurl chrcteristics of vegettion within cover types used for nesting were smpled with 3 circulr plots, ech with 10m rdius nd 25 m prt. One plot ws locted t the nest site; the other 2 plots were plced t rndomly selected compss berings 25 m from the nestsite. The number of individuls of ech woody species nd distnce between the plot center nd nerest woody species were recorded for the shrub, understory, nd overstory strt to provide vegettion dispersion index. Cover vlues for grsses, forbs, nd understory nd overstory trees were estimted t 4 rndomly selected loctions within ech of the 3 plots. A smpling frme, 0.5 m in dimeter, ws used to estimte ground cover nd sighting tube (Jmes nd Shugrt 1971) ws used for cnopy cover. An index of visul obstruction of the vegettion ws obtined by trversing 3 10m trnsects within ech plot nd recording contcts with vegettion t 0.5 nd 1.5 m heights. Index vlues were verged to obtin plot vlues. Weights of slsh were estimted with photoseries key (Mxwell nd Wrd 1976). If nesting cover type ws used in significntly greter

14 proportion thn it occurred on the study re (Neu et l. 1974), vegettion chrcteristics t nest sites were compred with rndomly 6 smpled (vilble) sites within the sme cover type. Smpling of vegettion in cover types vilble for nesting turkeys ws identicl to the smpling done in cover types used for nesting except tht rndomly locted point ws used s the focl point for smpling. We used t-tests (Nie et l. 1975:267) to compre vegettive chrcteristics between nesting sites nd rndom sites nd between clutch sizes of dults nd yerlings. Chi-squre tests (Sokl nd Rohlf 1969:552) for goodness of fit were used to test for differences in expected nd observed numbers of nests in cover types. If difference ws detected, confidence intervl estimtes (Neu et l. 1974) were clculted for ech cover type to identify which contributed to the difference. The 95% confidence level ws considered significnt. RESULTS AND DISCUSSION Of 113 birds trpped during 1981 nd 1982, 43 were femles (12 ds nd 31 yrls). All dult hens nd 20 yerling hens were equipped with rdios; 25 of these (12 ds nd 13 yrls) survived to the breeding seson nd produced 16 nests, of which complete informtion (initition dte, clutch size, nd fte) ws recorded for 15. Nests of 6 unmrked birds were discovered during field ctivities nd were used in descriptions of nest site chrcteristics. All of these nests, however, were bndoned by the hens presumbly becuse of disturbnce by the observer. Willims et l. (1969) lso noted tht

15 7 turkeys bndoned nests following disturbnce. Therefore, only dt from rdio-equipped birds were used to describe clutch sizes nd nesting success. Nesting Success Adult nd yerling hens differed in the proportion tht ttempted to nest, in nest success, nd in incubtion behvior. All 12 dult hens ttempted to nest, but only 4 of 13 yerlings did so. We hve no explntion for the low level of nesting ttempts mong yerlings. Seemingly, the reduced propensity of yerlings to nest is confined to Merrim's subspecies mong the rces of turkeys. Lockwood nd Sutcliffe (1985) reported tht 75% of dult nd pproximtely 8% of yerling Merrim's turkeys in New Mexico ttempted to nest. Willims et l. (1976) found tht nesting ttempts, clutch size, nd percent of eggs htching did not differ between the 2 ge clsses of Florid wild turkey (M. _ osceol) but dults were more persistent t renesting. Likewise, Porter (1979) reported tht 889 of yerling femles of mixed rces (1gl. merrimi X silvestris) ttempted to nest in Minnesot. In ddition, dults were more successful t nesting thn yerlings. Nine of 12 dults nested successfully, compred with only 1 of 4 yerling hens. None of the unsuccessful hens ttempted to renest, lthough Ligon (1946) noted tht Merrim's turkeys in New Mexico renested redily. In New Mexico Jones (1981) reported tht renesting rtes rnged between 25 nd 100%, wheres Lockwood nd Sutcliffe (1985) noted tht only dult hens ttempted to renest. Men clutch size (_11 = 15) ws 10.9 eggs (SD = 1.7). Men clutch

16 size of dults (x = 11.1, SD = 1.8, N = 11) ws not significntly 8 different from tht of yerlings (x = 10.3, SD = 1.5, N = 4). Young htched from 28 My to 4 June. Clutch sizes nd htching times were similr to those reported by Mckey (1982) in Wshington. Behvior during incubtion my be n importnt fctor cusing differentil nest success between dults nd yerlings. Activity sensors on the rdio trnsmitters indicted tht dult hens reduced ctivities during the 1st week of incubtion, wheres nesting yerling hens were ctive through the 2nd week. This difference presumbly resulted from lower ttentiveness t the nest by yerlings. Higher levels of ctivity by yerlings t the nests possibly incresed vulnerbility of nests nd hens to predtors. One hen ws killed while wy from the nest by predtor nd 2 nests were destroyed by predtors. Nesting Hbitt Turkeys nested in 5 of the 14 vilble cover types (Tble 1). Cover types used for nesting were described for 21 nests, of which 19 (90%) were in the Dougls fir stnds nd 2 were in the Ponderos Pine stnds. Generlly, hens used erly serl stges within both stnd types; 17 of the nests in the Dougls fir were in selectively logged (e.g., slvge sles) or thinned stnds. Eleven of the 15 rdioequipped birds nested in Thinned, Mixed Conifer stnds within Dougls fir stnds, 1 nested in 5-yer-old powerline right-of-wy, nd 1 nested in n undisturbed stnd of Dougls fir-ponderos Pine-Ok. The other 2 hens nested in selectively logged stnds of Ponderos Pine.

17 9 Only the Thinned, Mixed Conifer cover type ws used more frequently (P < 0.05) thn it occurred on the study re (Tble 1). These stnds hd been reduced to pproximtely 50% of their pre thinning bsl re (trees verged cm nd 616 trees/h) nd hd n bundnt mount of slsh (limbs, tree tops, nd other woody debris) on the ground. Nests commonly were locted in or djcent to ccumultions of slsh. Thinned, Mixed Conifer stnds contined n estimted metric tons/h of slsh. Ffolliot et l. (1977) reported high use by turkeys of stnds where slsh ws bundntly vilble. Petersen nd Richrdson (1975), Jones (1981), nd Phillips (1982) lso noted tht nests often were ssocited with pine slsh. Nesting Success in Reltion to Hbitt Thinned, Mixed Conifer stnds contined 52% (11 of 21) of ll nests nd 82% (9 of 11) were successful. Only one of the other 10 nests, distributed mong 4 cover types (Tble 1) ws successful. These results indicted tht Thinned, Mixed Conifer stnds were the most secure hbitt for nesting, which my ccount for selective use of this cover type by hens. Nest Site Chrcteristics Vegettion t nest sites in Thinned, Mixed Conifer ws compred with vegettion t 19 rndomly locted sites in the sme cover type. Vegettion from nests in other cover types ws not compred similrly becuse of indequte smple size (10 nests in 4 cover types). Nest sites in Thinned, Mixed Conifer typiclly hd greter density of shrubs nd trees in the understory nd fewer trees in the overstory

18 10 this obstruction nd my hve provided conceling cover nd nest. Jones (1981) reported tht horizontl cover for the hen (0-0.3 m) t Merrim's wild turkey nests ws consistently dense nd tht cover ws provided by logging slsh nd shrubs. MANAGEMENT IMPLICATIONS Adult hens, which composed only 28% of ll hens trpped, contributed the bulk of the reproductive output. Accordingly, we suggest tht in popultions tht re declining or experiencing poor reproductive performnce, some mesure of protection should be fforded to hens. Eithersex hunting, which typiclly is permitted in the fll, should be prohibited in these circumstnces. When trnsplnting Merrim's wild turkeys, mngers should relese dult hens rther thn yerlings to increse the chnces for successful reproduction during the 1st yer fter relese. We lso suggest thinning of Dougls fir stnds in res scheduled for restortion or introduction of Merrim's wild turkeys. Nesting hbitt could be creted by thinning polesize (10-20 cm) stnds to tree densities of 616/h. Becuse bundnt slsh cover ws chrcteristic of nest sites, we suggest tht slsh should not be burned or removed in thinning nd logging opertions in mixed coniferous stnds. Abndonment of nests by the 6 unmrked hens confirmed the sensitivity of hens to disturbnce t nest sites nd indicted tht humn ctivities in the key cover type used for nesting (Thinned, Mixed Conifer) should be minimized during April, My, nd erly June.

19 11 Tble II.1. Cover types vilble nd used for nesting by rdio equipped Merrim's wild turkeys, Wsco County, Oregon, Cover type Avilble(%) Used(%) (N = 21) Ok 24.6 Ponderos Pine Mture Mixed Conifer 14.2 Ponderos PineDougls firok Ponderos PineOk 9.5 Thinned, Mixed Conifer Young, Mixed Conifer Grin Fields 4.6 Rngelnd 0.6 Old Clercuts 0.6 Irrigted Pstures 0.3 Mixed Deciduous 0.2 Recent Clercuts Shelterwoods 0.1 Use of cover type significntly greter thn expected CP < 0.05).

20 12 Tble Comprison of structurl chrcteristics of Thinned Mixed Conifer stnds used for nesting by Merrim's wild turkey hens with rndomly smpled (vilble) Thinned, Mixed Conifer stnds, Wsco County, Oregon, Chrcteristic Avilble 16) Used Cy. = 11) x SD x SD Grss cover (Z) <5 <5 Forb cover (7) <5 <5 Shrub Density (no./h) , Dispersion (m) Understory Density (no./h) Dispersion (m) Cover (7) <5 <10 Over story Density (no./h) dispersion (m) cover (7) Obstruction of vision t 0.5 m m Significntly different from expected, P <0.05.

21 13 CHAPTER III SEASONAL USE OF ROOST SITES BY MERRIAM'S WILD TURKEY HENS AND HENPOULT FLOCKS IN OREGON Abstrct: Chrcteristics of trees nd cover types in which Merrim's wild turkeys (Melegris Rllopvo merrinri) roosted were described for 36 roosts used by hens during winter, spring, nd summer in north centrl Oregon from 1981 to Four of 14 cover types were used for roosting; only mture mixed conifer ws used significntly more thn it occurred on the study re. Number of roost trees/site, cnopy cover, slope, nd DBH were roost chrcteristics tht differed mong sesons. Ponderos pines (Pinus ponderos) (I DBH =67.8 cm) were used for roosting most frequently in winter nd spring, wheres smller CI DBH.47.7 cm) Douglsfirs (Psuedotsug menziesii) were used most often in summer. Averge number of trees/site rnged from 3 in spring to 6 in winter. Roosts were locted more frequently on steep slopes during winter, likely n outcome of the restricted distribution of mture mixed conifer cover type in winter hbitt. In summry, lthough hen poult flocks roosted in the sme cover type during winter, spring, nd summer, sesonl differences in roost chrcteristics were ttributed to flock size nd composition.

22 14 Roost sites of Merrim's wild turkeys hve been described for severl loctions within the originl rnge of this subspecies. Ponderos pine trees were most commonly reported s roosting hbitt in Colordo (Hoffmn 1968) nd Arizon (Boeker nd Scott 1969), but Dougls-fir nd white fir (Abies concolor) in mixed-conifer stnds were used by turkeys in New Mexico (Jones 1981). were used most often for roosting in these res. Lrge, mture trees Merrim's wild turkeys hve been introduced into severl western sttes; however, little reserch hs been conducted on cover types or trees used by turkeys for roosting in these res. Mckey (1984) found tht Dougls-fir cover type (10 of 17) nd trees were used most commonly for roosting from Februry through August in Wshington, nd Jons (1966) reported the ponderos pine community ws most frequently used for roosting hbitt in Montn. Previous reserch focused most often on the tree or trees used for roosting; little work hs been reported on cover types in which roost trees were locted. In ddition, generlly only 1 tree mesurement (DBH) ws tken, no informtion ws collected on the ge nd sex of birds using roosts, nd most reserch described roosts from only 1 seson or combined sesons. The purpose of this project ws to better understnd roosting hbitt used by dult hen nd hen/poult groups of Merrim's wild turkeys in Oregon, where the birds were introduced in 1961 (Mce 1965). Our objectives were to describe cover types used for roosting, to describe chrcteristics of trees selected for roosting within

23 cover types, nd to test for differences in sesonl use of roost 15 sites. Adult hens were selected for study becuse dults, lthough they composed only 28% of the femle segment of the smple in our study, contributed 91% of the broods (Crwford nd Lutz 1984). Informtion bout these cover types nd roost tree chrcteristics my be useful to lndmngers interested in mnging res for Merrim's wild turkeys. We thnk R. Anthony nd W. Krueger for reviewing the mnuscript. This work ws supported by funding from Oregon Deprtment of Fish nd Wildlife (ODFW) (PRW-79R-2) nd Mount Hood Ntionl Forest (MHNF). We thnk R. Denney, ODFW, D. Longrie, MHNF, nd T. Luther (Wrm Springs Indin Reservtion) for their coopertion. C. Heth ided in collection of roost site dt. This is Technicl Publiction 7828 of the Oregon Agriculturl Experiment Sttion. MATERIALS AND METHODS Twelve dult hens, 2 in 1981 nd 10 in 1982, were trpped from Jnury to April nd equipped with rdio trnsmitters (Telonics Inc, Mes, Arizon) on the White River Gme Mngement Are, Mount Hood Ntionl Forest, nd djcent privte lnds in southern Wsco County, Oregon. Nine of the 12 hens successfully produced broods. We ttempted to find roosts twice monthly by selecting 1 rdioequipped hen nd determining the roost loction 1 hr fter sunset. different hen during ech roost finding ttempt to insure We selected tht ll hens contributed to roost loctions. Cover types on the 135,141h study re were mpped from LANDSAT stellite imgery (Murry 1981).

24 16 Study re boundries were bsed on totl nnul loctions becuse of considerble overlp of res used by rdio-mrked hens throughout the yer. Physiogrphic (percent slope, distnce to nerest ridge, nd spect) nd vegettive chrcteristics were used to describe roost sites. Mesurements of individul roost trees included dimeter t 1.4 m (DBH), percent cnopy closure, cnopy height, nd height to lowest living nd ded limbs. Percent slope nd ll height mesurements were estimted with clinometer. Percent cnopy closure ws estimted with sighting tube (Jmes nd Shugrt 1971) t 4 rndomly selected points within 3 m of roost trees. Presence of stumps within the roost site ws used s evidence of logging ctivity. Percentges of forested nd non-forested cover types within 0.33 km of the roost site were determined from eril photos with grid of 50 rndomly distributed points (Mrcum nd Loftsgrden 1980). This re round the roost ws within the dily movement rnge of hen-poult flocks (Crwford nd Lutz 1984). Numbers of droppings t bses of trees were used to clssify mount of use (light=<10 droppings, moderte=10-50 droppings, hevy=>50 droppings). Sesons of use were winter (16 November-28 Februry), spring (1 Mrch-31 My), summer (1 June-15 September), nd fll (16 September-15 November). Chi-squre for goodness of fit (Sokl nd Rohlf 1969:552) were used to test for differences between expected nd observed numbers of roosts in ll cover types. If difference ws detected, confidence intervl estimtes (Neu et l. 1974, Byers et l. 1984) were clculted for ech cover type to identify cover types tht contributed to the difference. One-wy nlysis of vrince (Nie et

25 17 l. 1975, ) ws used to test for differences in physiogrphic nd vegettive chrcteristics of roosts mong sesons. If there were differences in chrcteristics nd the ssumption of homogeneity of vrince ws met, Student-Neumn-Keuls multiple comprison tests (Nie et l. 1975,428) with hrmonic mens (Hull nd Nie 1981,307) were used to identify which sesons differed. The 95% confidence level ws ccepted s significnt. RESULTS AND DISCUSSION A totl of 39 roost sites ws found. Twelve roost sites from ech of 3 sesons (winter, spring, nd summer) were nlyzed for sesonl differences. Fll ws not included in the nlyses becuse only 1 roost ws locted. Fll roosts were difficult to locte becuse of frequent, long movements by hens during tht seson. Two roosts were used during more thn 1 seson (spring & fll) nd were excluded from the sesonl nlysis. The sex nd ge composition of groups t roosts differed mong sesons. Spring roosts were used exclusively by dult hens. Adult hens nd poults occupied summer nd fll roosts. In winter, roosts were used by dult hens nd yerlings of both sexes. Cover Type Use Fourteen cover types were clssified. on the study re were forested (Tble 1). Eleven of the cover types Non-forested hbitts (rnge, irrigted lnds, grin) occupied less thn 6 percent of the study re.

26 18 Four of the 11 forested cover types on the study re were used for roosting. Mture mixed conifer ws selected for roosting in ll sesons (92%-winter,75%-spring,59%-summer) nd ws the only cover type used significntly more thn it occurred on the study re (Tble 1). Ponderos pine, ponderos pine/dougls-fir/ok, nd ponderos pine/white ok conifer cover types lso were used for roosting. Cover djcent to roost sites ws primrily (1 807,rnge=58-96) forested (Tble 2). Scott nd Boeker (1977) found tht djcent forest cover ffected use of roost sites nd suggested mximum distnce of 90 m between roost sites nd forested cover. Low elevtion sites (<1000 m) on our study re hve 80-yer history of smll-scle, selective logging ctivity (Willims 1978:13); however, only 7.5% of the roost sites hd been logged, in contrst with dt from New Mexico (Jones 1981) where 46% of the roost sites hd been logged. These contrsting results indicted rther brod rnge of response by Merrim's wild turkeys to logging ner roost sites. Roost Tree Chrcteristics Within the mture mixed-conifer cover type, hen-poult flocks used lrge (DBH =59.3 cm), dominnt trees most frequently for roosting. Increment borings indicted tht trees of this dimeter were >300 yers old. Merrim's turkeys roosted exclusively in either ponderos pine (62.3%) or Dougls-fir (37.7%) trees. DBH of roost trees in Oregon (59.3 cm) ws mong the highest reported. Men DBH in other res rnged from 45 cm in Montn (Jons 1966) nd New Mexico (Jones 1981) to 64.5 cm in Arizon (Boeker nd

27 19 Scott 1969). Differences in sizes of trees used for roosting my be function of sizes vilble, but neither we nor these uthors determined DBH of trees not used for roosting by turkeys. Although these results indicted regionl differences in sizes of trees used for roosting, ll roosts typiclly were composed of lrge, mture trees in ech of the res. Sesonl Differences Very few trditionlly used roost sites were identified; only 2 sites were used during more thn 1 seson nd 82% of ll sites were lightly used. All sites used in spring were used lightly, wheres 79% of winter nd 67% of summer roost sites were used lightly. In contrst, Jones (1981) found tht ll (N=17) winter roost sites were repetedly used (i.e., t lest moderte use), but only 1 of 22 summer roost sites ws used more thn once. Number of roost trees/site (F=11.9,P < 0.001), DBH (F=7.4,P <0.001), cnopy cover (F=11.9,P<0.001), nd slope (F=7.6,P<0.001) differed mong the 3 sesons (Tble 2). An verge of 2.9 (1-10) trees composed spring roosts, wheres winter nd summer roost sites consisted of 5.4 (1-9) nd 5.9 (1-10) trees, respectively. were in multilyered, mture, mixedconifer cover types. Roosts Number of trees used during the sesons ws probbly relted to turkey flock size. Spring flocks often were composed of <5 individuls nd summer nd winter flocks of >30 individuls. There ws similr trend mong roost sites in ponderos pine in Colordo, lthough they were lrger nd verged 17 trees/site in winter nd 6 trees/site for spring nd

28 20 summer combined (Hoffmn 1968). Jones (1981), in New Mexico, found 8 trees/site in winter nd 2 trees/site in summer. DBH of summer roost trees ws smller thn the DBH of winter nd spring trees. Summer roost sites often were locted in swlog (Bker 1950) sized trees in the mture mixed conifer cover type. Once poults were cpble of flying, they used low tree limbs to move into the tree cnopy. type. Often swlog sized trees offered brnch structure of this Brnch structure provided by this size Douglsfir nd Ponderos pine ws likely n importnt chrcteristic of roosts used by hens with young poults in summer. Cnopy cover of winter nd spring roosts ws similr nd verged 20%, wheres summer roosts verged 43%. High cnopy cover reflected the incresed use of smller sized trees s roost trees during summer. Mckey (1984) reported men cnopy cover of 73.5% (ll sesons combined) t roosts in Wshington. The high cnopy cover from Mckey's study likely ws relted to use of Douglsfir trees for roosting by turkeys. Mckey proposed tht cnopy cover my be importnt for therml regultion in winter by roosting birds. Slope t spring nd summer roosts ws less thn t winter roosts. In ddition, winter roost sites were found lmost exclusively (93%) on northfcing slopes. During other sesons, ll spects were used. Steep, northfcing slopes t winter roosts probbly reflected the restricted distribution of mture mixed conifer stnds t lower elevtions where turkeys were most commonly found during winter. Most mture conifer stnds in less rugged terrin t these elevtions hd been logged within the lst 80 yers.

29 21 Among sesons, mture mixed conifer ws the only cover type used selectively for roosting. Although the sme cover type ws used during the 3 sesons, species nd sizes of trees used for roosting differed sesonlly. These differences ppered relted to the ges of birds using the roost. For exmple, dult nd yerling birds used lrger ponderos pine nd Douglsfir trees during winter nd spring thn dults with poults used during summer.

30 Tble III.1. Descriptions of forested cover types, percentge of study re occupied by ech cover type, nd percentge of roosts (Nm36) in the cover types used by Merrim's wild turkeys in southern Wsco Co., Oregon, Cover type Cover type description X % of vilble roosts Oregon White Ok Found t low elevtions ( m) on the estern edge of the study re. Ok ws the dominnt species in the shrub, understory, nd overstory strt Ponderos pine Mture mixed conifer Found primrily on the southern portion of the study re on the WSIR between 600 nd 1000 m. Both overstory nd understory consisted of ponderos pine. At elevtions below 750 m these stnds were restricted to steep north fcing slopes. At these elevtions, the overstory ws composed of Dougls-fir nd ponderos pine. At elevtions bove 750 m, the overstory ws true fir nd Dougls-fir Ponderos pine/dougls- Found s trnsition between ponderos pine/ fir/white ok ok nd the mixed conifer stnds between 750 nd 825 m. Generlly there ws widely dispersed overstory with dense understory of ponderos pine, Dougls-fir, nd ok Ponderos pine/white ok Also found t the estern edge of the study re; often djcent to ok stnds. The overstory consisted of ponderos pine nd ok ws consistently found in the understory

31 Tble III.1. Continued Cover type Cover type description % of vilble roosts Thinned mixed conifer Single-story stnds with Dougls-fir or ponderos pine overstory. These stnds occurred from m nd hd been thinned during the pst 15 yers. Young mixed conifer Often due to selective logging tht hd removed overstory species. The result ws mixture of dense spling to pole size (Dougls-fir, True fir) stnds interspersed with smll open res often seeded to orchrd grss. Old clercuts These hbitts rnged in size from 5 to 75 h nd were the result of logging yers go. Hbitts consisted of either single-story conifer plnttions or dense, multi-lyered shrub communities. Mixed deciduous These stnds were most often in riprin zones nd consisted of dense shrub ptches interspersed with smll, wet medows. Recent clercuts Found primrily in the mixed conifer stnds bove m. These hbitts were logged within the pst 5 yers nd rnged in size from 5 to 15 h. Shelterwoods Generlly locted t the upper elevtions (>900 m) in mixed conifer stnds. The stnds consisted of widely dipersed overstory of mture Dougls-fir or true fir nd well-developed mixed conifer understory. Other Grin fields, rngelnd, nd irrigted pstures Use of cover types significntly greter thn expected (P > 0.05).

32 24 Tble Chrcteristics of roost cover types nd trees used by Merrim's wild turkey during three sesons in southern Wsco County, Oregon, Men (Stndrd Error) Chrcteristics Winter (N =12) Spring (N =12) Summer (N =12) Cover type Trees/site (#) Forested hbitt (%) Ponderos pine (7) Distnce to ridge (m) 5.4(0.4) 2.9(0.3) 5.7(0.5) 76.4(6.4) 87.4(2.2) 77.6(4.7) (5.8) 37.0(4.9) 51.8(4.6) Tree DBH (cm) Cnopy coverge (%) Slope (%) Height to 1st living limb (m) Height to 1st ded limb (m) Crown height (m) 63.9(3.1) 71.7(4.3) 47.7(2.1) 20.7(2.6) 19.1(3.4) 42.6(5.0) 24.2(2.6) 16.2(3.2) 12.0(1.5) 9.8(0.8) 10.3(0.9) 9.7(0.9) 5.2(0.7) 5.4(1.4) 4.2(0.9) 17.2(1.3) 18.5(1.2) 15.2(0.9) Significntly different mong sesons (P <0.05)

33 25 CHAPTER IV PREVALENCE OF PDXVIRUS IN A POPULATION OF MERRIAM'S WILD TURKEY IN OREGON Abstrct: An introduced popultion of Merrim's wild turkeys ws exmined for poxvirus when birds were trpped from Jnury through April in 1981 nd Poxvirus lesions were found in three of 113 (167) turkeys; ll infected birds were immture mles.

34 26 Poxvirus hs been reported from wild turkey (Melegris gllopvo) popultions in nine southestern sttes of the United Sttes (Lthm, 1956; Powell, 1965; Prestwood, 1968; Prestwood et l., 1973; Akey et l., 1981; Dvidson et l., 1985). Dvidson et l. (1985) considered poxvirus one of the most importnt diseses of wild turkeys in this geogrphic region nd noted tht 22% of birds submitted for exmintion were infected. Little informtion is vilble bout the prevlence of poxvirus in wild turkey popultions, especilly in western sttes, lthough Thoms (1964) found one of 330 Rio Grnde wild turkeys (1, gllopvo intermedi) in Texs ws fflicted with poxvirus. Consequently, study ws initited in 1981 to determine the prevlence of poxvirus in n introduced popultion of Merrim's wild turkeys (M. gllopvo merrimi) in Oregon. We cknowledge R. Anthony nd W. Kruegr for providing criticl reviews of the mnuscript. This is Technicl Pper No of the Oregon Agriculturl Experiment Sttion. Funding for this work ws provided by the Oregon Deprtment of Fish nd Wildlife nd the U.S. Forest Service. MATERIALS AND METHODS The study ws conducted in Wsco County, Oregon, on the White River Gme Mngement Are (Oregon Deprtment of Fish nd Wildlife) nd the Mount Hood Ntionl Forest (U.S. Forest Service). Merrim's wild turkeys were introduced into this re in 1961 with stocks from Colordo, Arizon, nd New Mexico (Mce, 1965). From 1 Jnury to 5 April in 1981 nd 1982, 113 Merrim's wild turkeys (16 dult mles, 12

35 dult femles, 54 immture mles, nd 31 immture femles) were cptured with bited, wlk-in trps nd equipped with numbered leg 27 bnds. All birds were exmined for the presence of poxvirus lesions. If lesions were present, smple of the lesion ws supplied to D.H. Helfer, Oregon Stte University Veterinry Dignostic Lbortory, to verify field dignosis. Lesion smples were fixed in 10% buffered formlin for 24 hr, sectioned t 5 micrometers, nd stined with hemtoxylineosin for histopthologic evlution. Bolinger bodies chrcteristic of poxvirus were observed microscopiclly. Viruses were not isolted. RESULTS AND DISCUSSION During the study, three immture mle turkeys displyed ctive cses of poxvirus infection. One bird ws cptured initilly on 14 Februry 1981 nd no poxvirus lesions were noted. However, when this individul ws retrpped on 17 Mrch 1981, lesions were present on the hed, neck, nd wings. Lesions were present on the leding edge of the wings ner the lul. The right eye ws lmost completely occluded nd the left ws pproximtely 50% occluded. The infection ws sufficiently severe to potentilly impir the survivl of this bird. Another immture mle with poxvirus lesions on the hed nd neck ws cptured on 15 Februry A third immture mle ws cught on 15 Februry 1982: three poxvirus lesions were present on the skin immeditely under the lower mndible. These birds were not retrpped gin, nd ll three birds were members of lrge flock of turkeys (40-60 individuls) tht wintered long the south side of

36 28 White River Cnyon. Previous work with vin pox in gllinceous birds in Oregon (Crwford et l., 1979; Crwford, 1986) indicted possible connection between gme frm opertions nd the occurence of this disese in Cliforni quil (Cllinepl clifornic) nd ringnecked phesnts (Phsinus colchicus). Turkeys in our study were neither rised nor housed t gme frm fcilities, lthough some infected birds my hve been present mong the originl 38 turkeys relesed. It is lso possible the poxvirus ws trnsmitted to turkeys from nother species of wild bird or tht turkeys were infected from domestic poultry. Krstd (1971) noted tht poxvirus strins were often pthogenic for severl species. The study re ws remote nd sprsely populted by humns, but turkeys descended to lower elevtions in winter nd occsionlly forged ner frmsteds. Interestingly ll infected birds in the study were immture mles. Although some studies (e.g. Dvidson et l., 1980) hve shown no difference in prevlence mong sex nd ge groups of gllinceous birds, work by Crwford (1986) reveled tht prevlence of poxvirus in Cliforni quil ws pproximtely one nd one-hlf to three times greter mong immtures thn mong dults. Prevlence of poxvirus in turkeys in our study (2.6%) ws greter thn tht reported by Thoms (1964) for Rio Grnde turkeys (0.37). It did not rech the level reported by Dvidson et l. (1985) who cknowledged tht the prevlence of 22% in the smple of ded or moribund turkeys my not hve reflected the infection rte of the popultions.

37 29 CHAPTER V HABITAT USE BY MERRIAM'S WILD TURKEY IN OREGON Abstrct: Hbitt use nd selection by Merrim's wild turkeys in Wsco Co, Oregon, ws studied during This low density turkey popultion selectively used forested cover types chrcterized by vriety of structurl fetures, species, nd ge clsses. The popultion used monotypic forested cover types less thn expected, used nonforested cover types in proportion to their vilbility, nd voided forested cover types with structure simplified by logging ctivities. The 4 ge nd sex clsses hd lrge sesonl home rnges (x = 1313 h) nd in most instnces used cover types s they were vilble. We suggest tht structurl complexity of vegettion both within nd mong cover types is n importnt component of hbitt for Merrim's wild turkeys, which should be considered in the evlution of potentil relese sites nd in hbitt mngement plns for existing popultions.

38 30 Merrim's wild turkeys (Melegris gllopvo merrimi) were ntive to 4 nd introduced into 10 western sttes (Jons 1966); however, informtion on hbitt use in both ntive nd introduced rnges is sprse. Ligon (1946) described the ntive rnge of Merrim's wild turkey s coinciding with the distribution of ponderos pine (Pinus ponderos) in Colordo, New Mexico, nd Arizon. Most informtion vilble bout hbitt use by Merrim's wild turkeys reltes to roosts nd nests (Hoffmn 1968, Boeker nd Scott 1969, Jones 1981, Lutz nd Crwford 1987). Jons (1966) nd Mckey nd Jons (1982) provided the only quntittive dt on sesonl hbitt use. An introduced popultion of Merrim's wild turkeys in Montn most frequently (>50%) used 2 non-forested cover types, sedge (Crex 11)- blue grm (Boutelou grcillis)-bluestem (Agropyron sp.) grsslnd nd snowberry (Symphoricrpos sp.)-grsslnd ecotone during ech seson (Jons 1966). In contrst, Mckey nd Jons (1982) found introduced Merrim's wild turkeys preferred 2 forested cover types, ponderos pine-ok (Quercus grryn) nd Dougls-fir (Psuedotsug menziesii), during winter, spring, nd summer in Wshington. There is no informtion on sesonl use of hbitts by ge nd sex groups of turkeys. Hbitt use informtion is essentil for mngement of existing popultions nd vluble in evluting sites for possible introductions of turkeys. My objective ws to determine hbitt use nd selection by popultion of Merrim's wild turkeys tht resulted from trnsplnt of 38 birds into Oregon from Colordo, Arizon, nd New Mexico in 1961 (Mce 1965). I lso described ttributes of this

39 31 popultion including sex nd ge structure, density, mortlity rtes, nd home rnge. Funding for the project ws provided by Mount Hood Ntionl Forest (MHNF), Oregon Deprtment of Fish nd Wildlife (ODFW), nd Oregon Stte Univ. Computer Cent. We thnk R. D. Denney (ODFW) nd D. P. Longrie (MHNF) for their ssistnce in dministering the project, nd D. R. Lockwood, New Mexico Gme nd Fish Dept., who provided unpublished dt. R. G. Anthony, W. H. Emminghm, W. C. Krueger, G. L. Zimmermn, nd K. M. Kilbride reviewed erlier drfts of this mnuscript. A. C. Heth ssisted with field work. This is Technicl Publiction 8320 of the Oregon Agriculturl Experiment Sttion. MATERIALS AND METHODS The study ws conducted in southern Wsco County, Oregon, in 1981 nd The 135,141-h study re encompssed portions of Mount Hood Ntionl Forest, the ODFW White River Wildlife Mngement Are, the Confederted Tribes of Wrm Springs Reservtion, nd djcent privte lnds. Study re boundries were delineted by the movements of rdio-mrked turkeys. The re ws locted primrily within the Dougls-fir nd ponderos pine forest zones described by Frnklin nd Dyrness (1973). Specific cover types were defined ccording to dominnt vegettion nd lnd use prctices nd were mpped from LANDSAT imgery. Lutz nd Crwford (1987b) provided descriptions of the 14 cover types on the study re. Turkeys were trpped from 1 Jnury to 5 April ech yer with modified wlk-in trp (Ligon 1946), consisting of 4 welded wire pnels

40 32 (1 X 3 m) nd net wire (10 X 10 cm) top. Age of birds ws determined by the pttern of brring on the 9th nd 10th primries nd sex determined from brest fether colortion (Lrson nd Tber 1980). I recognized 2 ge ctegories of birds: subdults (8-20 months old) nd dults (> 20 months of ge). The subdult ctegory ws bsed on previous work tht indicted Merrim's subspecies <2 yers rrely were involved in breeding ctivities (Lockwood nd Sutcliffe 1985, Lutz nd Crwford 1987). Trnsmitters, weighing 90 g nd with n opertionl life of months, were ttched with bckpck hrness of rubber tubing covered by brided nylon. Trnsmitters contined n ctivity sensor tht indicted when bird remined sttionry for min. Turkeys were relocted 3 times weekly throughout the yer to scertin cover types used. I delineted 4 sesons: winter, 16 November-28 Februry; spring, 1 Mrch-31 My; summer, 1 June-31 August; fll, 1 September-15 November. If bird died within 1 week of cpture, I ssumed tht the deth ws result of cpture stress nd the bird ws not used to estimte hbitt use or home rnge. Only birds tht survived n entire seson were used to estimte sesonl home rnge sizes, but ll birds tht survived >1 week were used in the cover type nlysis. Use of cover types ws determined from 2377 loctions nd bsed on 49 birds in winter, 32 birds in spring, 23 birds in summer, nd 21 birds in fll. Home rnge estimtes were bsed on 32 birds during winter, 23 in spring, 22 in summer, nd 21 in fll. Seventeen birds were not nlyzed becuse of rdio filures (11) or cpture stress (6).

41 33 Structurl chrcteristics of vegettion within cover types were smpled with 3 circulr plots, ech with 10-m rdius nd 25 m prt. Cover vlues for grsses, forbs, nd understory nd overstory trees were estimted t 4 rndomly selected loctions with ech of the 3 plots. A smpling frme, 0.5 m in dimeter, ws used to estimte ground cover, nd sighting tube (Jmes nd Shugrt 1971) ws used to estimte cnopy cover. Indices of visul obstruction of the vegettion were obtined by trversing 3 10-m trnsects within ech plot nd recording contcts with vegettion t 0.5 nd 1.5 m heights. Popultion ttributes were described with ge nd sex rtios determined from winter trpping. Additionlly, indices to density nd mortlity were clculted. The density index ws clculted with the Lincoln-Petersen method (Lincoln 1930) from mles hrvested during the spring hunting seson s 'recptures' to clculte the mle popultion. The mle popultions ws used with winter sex rtios to extrpolte totl popultion size. Leopold (1944) observed tht winter flock size incresed with density, so we lso used winter counts s density index. An index to mortlity ws clculted (Porter 1978) with dt from rdio-mrked turkeys. The minimum convex polygon (Mohr 1947) nd hrmonic men trnsformtion (Dixon nd Chpmn 1980) methods were used to estimte sesonl home rnges for the 4 ge nd sex groups (dult nd subdult mles nd femles). Home rnges were clculted with the computer progrm MCPAAL (Stuwe nd Blowhowik 1986). Anlysis of vrince (SAS Institute Inc. 1985:57) ws used to compre differences mong sesonl home rnge sizes within ge nd sex groups of turkeys. Simple

42 34 correltion (SAS Institute Inc. 1985:269) ws used to determine the reltionship between sesonl home rnge size nd sesonl mortlity rtes. Bonferroni simultneous confidence intervls (Neu et l. 1974, Byers et l. 1984) were used to test for differences between expected nd observed use of the cover types. Expected vlues corresponded to the reltive re of the cover type nd observed vlues were the percentge of loctions of rdioequipped turkeys in the cover type. Four ctegories of use were defined: 1) cover type used more thn would be expected by chnce (high use or selection), 2) cover type used in proportion to its vilbility (proportionl use), 3) cover type used but less thn would be expected by chnce (low use), nd 4) cover type not used (voidnce). Cover type utiliztion ws determined 1) for ll birds combined on the entire study re nd 2) by sex nd ge groups within sesonl home rnges. Preliminry nlysis reveled no differences in cover types used in home rnges clculted with minimum convex polygon nd hrmonic men techniques, hence we used only results of the minimum convex polygon. To better understnd use of cover types, vegettion structure (e.g. % cover, density) within cover types ws investigted by comprison of loction sites with rndom (vilble) sites. A rndom smple of loctions (N =337) ws used for the vegettion structure utiliztion nlysis. Rndom sites were smpled in ech cover type to provide "vilble" vegettion structure. Differences in vegettive structure were tested with multivrite nlysis of vrince

43 35 (MANOVA;Hull nd Nie 1981:1) by comprison of chrcteristics between loction sites used by turkeys with rndomly smpled (vilble) sites. RESULTS AND DISCUSSION Popultion Attributes I trpped nd bnded 113 birds which included 28 dults (12 femles, 16 mles) nd 85 subdults (31 femles, 54 mles); 66 of these birds were equipped with rdio-trnsmitters (12 dult femles, 20 subdult femles, 15 dult mles, nd 19 subdult mles). Spring density estimtes were 0.1 birds/km for 1981 nd 0.4 for 1982, which were substntilly lower thn densities reported from New Mexico (Ligon 1946) nd Montn (Jons 1966). The increse in density indices between yers on our study re ws supported by winter flock counts (ODFW, Unpubl. dt) tht indicted 3-fold increse between 1981 nd Averge nnul mortlity rte for the popultion ws 53% (Tble 1), which ws within the rnge of 50 to 60% reported by Wischnofske (1960) nd Jons (1966). Adult mles hd the highest nnul mortlity rte (61%) nd dult femles hd the lowest (39%). Cuses of mortlity included predtion (70%), legl hrvest (11%), ccidents (7%), nd poching (4%). Home Rnge Size Home rnges were lrgest in spring (2575 h), intermedite in fll (1535) nd summer (1509 h), nd smllest in winter (636 h). (Tble 2). Subdult femles in spring hd the lrgest home rnges (4213 h) nd dult mles in winter hd the smllest home rnges (347