are inland 'wolf---ungulate systems influenced by … inland 'wolf---ungulate systems...
TRANSCRIPT
fccll",'u,:'" API,/I('<J{{'.JII,\'. 20ll). 2010, pp. 2Sl 262 V ~t)IQ h~' Ih~ r'.COlog·l~a.l S('iCldV _)r Am~nca
Are inland 'wolf---ungulate systems influenced by marine subsidies of Pacific salmon?
LAY';E G, AD..nI5,U SliM, D. FARLEY,l CRAIG A. STRICKER,3 DOM:>JIC 1. DEVIMA,I,6 GRETCHE'-' H, RorTLER,'
DEN~J5 C. MILLER,4 AND ROBERT O. RuJ
'L'.S. Geofo{:ica/ Survey, Afaskn Sdencc Center, 4210 Unil'ersil)' Drirt:, Anchon/!!<'. Alaska 9'J50R USA "Afas/w Dep{//Im~nl oFFish (!lid G{}m~, 333 Raspbern' Road, Anchorage, Alask(J 99518 USA
)U S G<'oiogica! Sliney. S/llb!(' 1,lO!ope Laboratory, Dem'l'r F(~dern/ Center. Denw'r, Col(Jwdo S0225 USA JCnriboli Ai,. Sel" ice, 1446 Han.\' WII)', Fnir!JclIIks Afe/ska 99709 USA
,1/'Isll',,/('I. Wolves (('Oll£.' lupus) in ~orlh America an: considc:red oblig:nc pn:<!;Hor, of lIngulal(;~ \vith (\Iher food resource~ playing lillie role in wcl!l" population dynami.:s or wolfprey rel<1tiOlls, Hcw,:ever, ~p;1wnjng Pacitic sallllon (OI1('orlr)'lU'lIS spp.) clre common [broughoU! wolf range in nOrlhwesrcrn ~orth America and Ill")' providc a marine subsidy affecting inland wall' ungulak food webs far from thc coast. Wc conducted stable-isotope analyses for nilro~~ell and carbon to evaluatc Ihe culliribution of ,almon 10 dieLs of wolves in Denali Nari(;nal Park and Preserve, l200 river-krn from tidewater in iuterior Alaska. lISA. We <lllalyzed bone collagen /i'om 73 wolves cquipped with radio collars dnring 1986·2002 and ev,llualed estimates of salmon in their dicts relatiw 10 the availability of salmon and ungulates within their home ranges. We compared wolf densilie,~ and ungulate: wolf ralios among regions with differing salmon and ungulate availability to assess subsidizing effects of salmon on these wolf· nngnlarr systcrm, Wolves III the northwcstern flats orthe study area had ,Ieeess 10 spawning salmon bill low ungulate availabiiily and conSl.lmcd more salmon (17% :!:: 7''lc (mean:: SD)) than in upland regio1lS, \vherc ungulates were ~ixfold more ahundam and wolve~
did or did not haw salmon spawrllng areas wllhin their home ranges (go;, :!:: 6Clc and J% :~ YiL rcspcCfivcly). Wolves were only 17'10 less abundant on the northwestern fLus compared to tbe remainder of the stndy area. cven Ihough ungulale dcnsi lic~ were 78%- lower. \Vc eSlimiHcd that hiomass from fall runs of c'hUIll (0. kera) and coho (0. kisulch) salmon on the north.vestenl fiats was comparahle to the ungulate biomass rhere, and the contribntic)lI of salmon to wolf diels was similar to estimates reported for coastal wolves in somheast Alaska. Given the nhiquitous consumpTion of salmon hy wolves on the northwcstcrn I'lms and the ahundanc;e ()f salmon there, we conclude thai wolf numbers in this region were enhanced hy Ihe allochthollous suhsidy provided hy salmon and discuss implications for wolf-ungulate rdations.
Ke.l f1'ord, allochthonous subrldw". op{HJrl'111 comr"'I;C;OI1 Canis lupus; l~!3c, f/5N: Det/ali ,'v{f(iollal [',Irk and /'rese!'Ve, Alaska USA; f)w/'i",,-(kl'iw!d "ulrienls; Oncorhyndlus spp.; Pac//7c salmoll, predel/or· pr(,l' te!uilt}//.,·: lJ.JlgulaLtJ,)·, l"'I)/~('j
lions. even [hougb \volves arc known III comUlllC ii wide arrayal' other prey \Peterson <md Ciucci 2003).Gaining insights into the rclauonships bdw,;.:n WOh'C5 Recently, Pacific :;almon (Oncorhvn('hus spp.) lwve been(Cani,j IUpIIS: sec Plate I) and their prey is essenlia I 10 recognized as an important die! constituenl for coastalguide the complex majlagemcnl of woh-es, a$ well a, wolf populmiollS (Szepanski el al. 1999, DarimOllt andinform the contentious sciemifie and public debates l'n Reimchen 2002_ Darimont et 'll. :;003, 2008). AldlOugh the role of wolves in eeosYStelTls they inhahit (National the poknlial for salmon 10 he impOrlanl 10 inlandResearch Council 1997, FritlS el a1. 2003), Throughout wolves has bcen suggested (Darimont and Reimchen"iorth America, wolves arc view.:d as obligate predalOrs 2002, Darimon! el 31. 2008), salmon Ilave largely beell of llllglllatcs with olher prey cOlltributing little W ",volf ignored as a rood source for inland wol\'<'s (Peterson :inddietS (Pcl.:rsclll and Ciucci 2003). T)llls, liltle effort has Ciucci 2003). been expended 10 consider the role of non-ungulate prey
rvto~t liswe of individual ~aim(Jn accrues duringill the dynamic5 and ecosystem effects of wolf pOpUI,l
growth al sea and tbis marine-derived biomass i..: delivered to freshwater syslems where salmoll spawn
M,lIH1,cnpt reC:~I,,~d 31 July 20(J~: r~vi,ed 17 Apnl 2009: and die (Willson et ~1. 20(4). Considerable altennOn has accepted 5 !vlay 200t) Correwondmg Edllol: J. J. Millspaugh.
been givcn 10 the marine-deriv<'d l.1tltriem input), E-mail: ladam,;(ii:·usg,.gov (, Pre"ent addres,: Alnska Department of Fish and Game, provided by an~ldromol.1s Pacific salmon to freshwater
I~OlJ Glenn Highway, Suite 4, Palmer, Alaska 99645 US1\. <lnd terreSlrial e.:osyslems (Cederholm et al. 1999, Gende
251
I D..~ol()glcal Applh':<l.lJ~"\n!'LA YN[ G. ADAMS 1:1' At.
\\.... 1 :!o, N'-l
('{ at. 2004. Willson et al. 2004. Merz and Moyle 2006) and the imporlance of salmon to a variety of terrestrial carnivores (Ben-David el al. 1997a. D, Hilderbrand d ,II. 1999, Szepan~ki t:t al. 1999. Belan! et al. 2(06). [nlh1eilct:S of ''''Imon-derived nlilrients h,lW heen investigil1eLl almosl entjrcly in coastal or near-coastal « 100 river-km from tidewater) rep:ions; we are aware of only (lne published account on lhe topic at greater dislances from the Ocean (Bcl'lnt et al. 2(J()6). However. Pacitic sallTlon ar(' ~e3.sonally abundant and widely distributed far inland. for example, the Yukon Rivcr Basin, the largest drainage within the North American range of Pacitic 'almon. cncompasses 845000 km 2 of Alaska, Yukon Terriwry, and northern British Columbia (Brabet~ et at. 2(00). On average. over 2.5 million :;almon mig-rete each snmrner into tbe Yukon River ,ystein, di,per,i ng to spewning areas a" far as }OOO km by river from tidewater (Jomt Technical Committee of [h", Yukon River US/Canada Pilud [.fTC] 20[)7u, b).
Marine ecosy':tems me substan lialt}' more prod lIctive than lerrestriaJ ec\),y,tcrns at high latitudes US 7(t N; Gross Ct 'Ii. 191-:8. Erlandson c( aL 20(7): thus. Pacific salmon CLH\SlitU\c a nutrient subsidy onginating in 11
prod llctive marine environ men t with great [)otential to
infiuenee comparatively depauperille terrestrial food webs [hal are lar inland.
\Vhere spawning salmon occur. wolves would be expected to consume them extensIvely because. as a meat 'ouree. they are.: (1) available for several month" during spawning in Sllmmer and fall and as earnnn long afler; (2) predictably clumped and locally abundam at spa\vning an~a,; and 0) less risky or costly \0 acquire lban danger,)l1s or t1eet ungulate prey. Further, ill Alasb ancl the YLI kon Territory, ungulale densities are quit<? low o"er vast regions (Gasaway et aL 1992); thus salmon could provide a particularly irnporwnt food rewurce for wolves in this ponioll of their North American range.
AnadronlOus salm')11 are an example of Ilulrient transpon across ecosjsiem boundaries: the impliealions of such alloc-hthonOl15 subsidies for the 5Lrueture and dynillTlics of recipient rood webs ar~ \vell recognized (Polis et at. 1997, HllXd e! at. 2004. Loreau ancl Holt ~0()4). If 'iufhcient in magniiUde. ecosyslem subsidies to prcdat"rs elJlllll1nnl} re,ult in increased pred"lor lIumher~ and iJlcrl~ased predation pres5ure on resident prey through a rop-dowll process am110golls 10 i1pparenl competition (Holt 1977, Polio et al. J9(7). This OUTcome is particularly likely if residem prey are rar", (Polis el al. 1997. Esics et al. 20(1) or [he predator exhibits il strong lIutrler;ca: respon5e to incrensed food availability (Estes Cl 81. 20(1). Alternatively, predation on resident prey can be relaxed if the predator feed, m,tinly all tbe a1lochtbl1nom resource (Huxei ct al. 200·n. As a con,\"qucnce ,)1' either outcome, alJocb thonom subsidies commonly lead to f,lOcl ",,"eb dyTlillTlics that are mC(lll,i51enl wilh mod~l, based only on local resource and C,)[ISUmer conditions (Polis ct al. 1997).
We employed stable-isoLOpe analyses to assess consumption of salmon by Ivolves in Denali National Park and Prcscrve (DNPP) in central Al<lska, 1200 km vi.! lhc Yukon Rivcr and its tribtttaries from the Bcring: Sea coast Stable isotopes hav<' bcen used witll 5l1CCCS3 to apportion dicts bClwecn marine and lcrrestrial sources owing to predictable differences in Iso'()pic COlTlposilion of malenals of marine or terrestlial origin (Chisholm el aL 1082. Gannes et 'II. 1998. Kelly 2(00). Further, because stablc Isotope approachcs estimate contributions of food i{ems that are assimilated into consumer lissues, biases relative 10 prey size and digeslibility inherent in otber methods used to estimate wolfcliets are not an issue (Hilderbrand el a1. 1996, G,tnnes et aL 19%)
Research on wolf population dymuTIles in DNPP dUring 19~62002 (f\:kch ct 'il. 1998; L. G. r\d::Ul1~.
unpublished data) provided a unique 0pP(lr!unity to
investigatc thc usc of salmon by wolves rcJa/lve iO tbc distribution of spawning salmon and unguliiks on the l;lndscape. 111 addilion to documenling Vi()lf distriblilion and abundance via radiotelemelry, skelelal remains of [!losi wolves [lUll were equipped with radio collars and sllbsequently died in or ncar thc silldy area were regularly archived. Measurement of stable carbon (e) and nitrogen (N) isotope raiios of bone collagen allowed for appraisal of lhe long-term assimila tion of food resources by these wolves. Bone collagen is remodeled at a rate such that isolope measurements i1re thought 10
retiect diets m tcgrated over a perjod ranging from a few years (Ganncs et ai. I~98. Uochcren~ and Drucker 2007 i to an individuHl's en tin: life (Bnitlnc ei a!. 2005). Becau~e we knew the spatial distribution of home ranges
or ""'ol"e$ we sampled, we could a%ess the contribution of salmon 1O wolf diets relative to the availability of salmon ,1nd ungulates across tbe Denali landocape. finally, with data on the sizes and home range'S ofneHrly all wolf packs in {be study area during this 16-year period. wc cl)uld cvalllHtc evidence of effects of salmon availability on wolf abundance. We hypothesized Ihal where ungulate <1bundanec was low ;md salm,)TL 'vere widdy available. salmoll wotdd be well-represented m diets of wolves and tha t wolf numbers would be higher {han expected from ungulate availability alone.
MfTHODS
Studvarea
The study area (15400 k1ll 2 ; 63" N, ISl o W) included
the portion of DNPP nonh of the Alilska Range creSt <liGD In in elevation (Fig, I). Popul<ltions of wolves and ung\llaleS were liltle affected by hunHJn ha rI'c,j, wnhill om :;mdy area (~·1e....'h et 81. 1998\. During 1986 2002, wolf densi tics averaged 5.3 and (, 9 wolves' I000 krn~ in mid-March and early OCl0ber, respeclJvcly (Mcch <:t al. 1998: L G. Adams, unpublished clara). Wohes were c11~tribuled throughout [he area in pach that averaged 6.1 and 7.6 wolves (2-:~9 wolves/pack) in lat(' winter and fall, respectively, and maintained
253 JanuHry 20iO SAL!I..lQN AND INLAND WOH-UNGULATE SYSTEMS
,,
4 ..... •• •
•:.:~ .l (
..-1 ' -I!, ~;;:,;:
<,.:';",; ':1 ,.~i
r:==-'"-'"_ .....' .. ' '! r----j Denali NPr ,: ~
o SII;<Iy .,"', -"
o ~C ?O N
"'L~A
F!u. I. Study ,Irea 1'01' invesligations or wolr population dynamIcs during 1986-2002. Den'lli National Park dno Pre;"rve (~I'P). Ala.'ka. USA. Regiona! ungul,ltc dellSll1es (light gny. 70 moo"e cquivltlenl.\ilOOO km 2
; dark gray, 320 nlOO~e eqlliHI~ntsjiOOO km\ known and smpecLed salmon "pawning area, (dark gray stream ,,~glllenls). and rcsultmg ~aL~gories or IIngu!ale and "tl",on availabilit.y (I. salmon pre:>elll. Jow unguldlc density 2, salmon present, high ungula.le density; anO 3, no ~~dnj01i4 hlgh ung.ulale den.~ii:) dTc dCPlCkd
distInct. but uvcrl"pping. home range'> (f\:]ech el al.
1<)98' L. G. Adams. uJ1[Juhlnhed dal,}). Moose (AIN'S
<'(ft'f.,). cariho\! (Ralir;d~'r larQlidw) , and Dall's sheep (0\'/.1' dalii) constlluted the ungul<Jic prey <Ivailab!c to
\V()lve~ in DNPP. tn early winter. (ca. ! 0lo\'ember) "·2000 moose OCe\lrrc(] wilhin the: sludy arca. at
del1.~ilies of 200 moose! I000 km1 in the mountains and
fo,Jlbills and 50 moose:'IOOO km" in tbc northwestern lowland sprl1c,~ fiats (U.S. National Park Service [NPS].
till/Jubli.l'hcd flwfl!lstTiprs). The study area encompassed most ,)['lhe range of Ihe Denali caribou herd (averaging
~300 canbou In autumn during the study: Adams ~O()5)
<mc! the '['onIOn" carib'/ll herd's (~1000 caribou: NPS. 1.J/ll'uhll.l'hed /'I/()I1U.~UiPbl rang.e overlapped the western
penphery. Canbol1 maillly uliliLed opcn habjlal~ in Lhe mountain~ ,md foothills. with limlLed sc~{sonal usc of
forested 100... land~ (L G. Adams. ul1puhli.\·h"d dala).
ApplOximalely 200U DalJ\ shcep ('ccuITed primarily in the eastern moun l<tins of tbe: study arca (NPS. unpub
lished nW/1/c.I'cnjus). (jiven eadl species' nl1mber~ and distribmion throughOllt the year. ungulat.es were sub'\wnlldlly TI1nre abundant in [he open. uplalld habirals along the /\la~ka lbn~e and ea~Lern foothills of Ihe
sr.udy arca (.<20 moose cqul\'alcnts; 1000 km 2: [ moose
equivalent"" I Ino\)'>e. J canb'lli. or (> sheep. follOWing KClih [19831 and hiler [J989J) than ill Ihe k)",land,
spnlee·co\('rcd flal,; tn {he nortlw"~~1 l70 moo~e
c£jllivalctHs.'IOOO km l: Fig j). Wl,lvc-; also o:x<l~i\lnally
preyed on heal'er (Casto/' c<1nw!e!1,i\'), snowshoe hare,;
(Lepu5 anwric<7l1us). :'\rc:ic ground squirrels (Spamophilus pa/'ryu). hoary marmots (Mannola m!(g{i/o). and various hirds tMech el al. J998). hll'lher. wolVe'; in the study area have been observed consuming salmon on occasion (Mcch ct a1. 1998; D. C. Miller. per,l'(J!<a!
obSfrvai ;(ms).
Three species of salmon occurred within the siudy Mea. Chinonk salnwn (0. (shall'}'lscha). Ihe large'll of the threc species (averaging 8.5 kg vs. 3.1 and ~ 7 kg for
chum [0. kl'l<1] and coho [0. lcisut<hJ. l'especll\'<,ly: Ab"ka Department of Fish and Game jADH3) 200(». dmvcd first In carly July followed shortly by a slimmer run of chum :;almon in Illid-July. A ;;e::ond run \lfL'llum
salmon reached the area in hLl~ August and spawned
into "!ovcmbcr. Coho salmon arrived last iT! a run
slightly later hut overlapping the fall chum rutL
The Kantishna River drained most of the study area. (7::l%); the Nenana River aud the Swifi Fork of lhc Kuskokwim River drained tbe eastern and westelll margins. and all three river :;ysterns $upponed spawnlllg
salmon. Whereal> information 011 the rnagni!ud,~ of salmon funs within lhe Denali region was limited, there wa, 5ubl>l<ll1lial information on fall dum sa)m'ln in ih.: Kanllshna Rl\'cr sys(('m. Since 1974, COlln\~ of:,pawning
fail churn~ ,11 Toklat. Spnngs, 15 kill north or the study
I L::"oi~)gJ\.~aj Applll:'<.ii1~'"In~LA Y1\'[ G. ADAMS lOT AL.
Vol ;0, N\l
C1T~<1 (Fig. i), luve been cond\lcred annually to index pupuliltion lTell(h (Hue et al. 2(06). DLLring \999-·20(15. {"<1l1 chum popuiallOn e~limate~ in the Kantishna drainage were derived annually via mark-recapture rnethods, ranging from 21 SOO [0 107700 salmon (Ckary ,md lLlm:i;r,lkl 20(6). Hased Dn these data. we eSI imated thai an ,lWI ilcle l1{ -, n 000 fall chums spawned annually during 1986·-2002 within the Knntishna River drainage and annual runs may have varied by more than an ()rder uf magmillde, from 21000 10 2-40000 salmon. Lillie mformation existed on abundance of the other three SCilmon pl)p\l!ariollS in Ihe Kanllshna system. Coho were pr"bably second in Humber 10 fall chunl', <1.verag1l1g ,lbClut 5000 annually (1'. M. Cleary. pi'I'sol/({1 WiflmUlli
<"(./liMI). Chnwok. alld summer ehunH were [es, common, WIth tiJdl <lWr<:lging .s2000 salmon p~r :veilr (Eiler et al. ~004, .fTC 2007a; H. M. Borba, persollal mnml!lnicalion; l. R. Spene;;!'. per.\O/lu{ ("rlTlllnullir alioll). All three salmon ~pcc'ics al;;o spawned in the S,vifl Fork and Nt'nand River portic)lls of Lhe stuely ilrea, but salmon llulflbers there were lI11krwwn.
vVo[f distrihwioll and ,,/JundaIlCI!
Investigations 11f wn]f population dynamics al DN'"?P have heen descrihed in detail elsev.:here (Meeh cr al. 1998) In brid, rh~ dislribllll(1n and sizes of wolf packs \v·cre l1loniwred by "quipping two 10 three wolves per p'lCk. widl radio collars via helicopler darting and then locating them approxima lely every twn weeks from lighl aircra(l, vVllh aJdinonal obstrvcHloJiS In lale Seplcmbt:rcady October ,HHi r-,'larch t() obtain fall dnd late-winter pack COlllHS, respectively. Home range.... ()f wolf packs were delefmined by the minimum convex polygl)n method with localions accumulated over tWI),year periods t,) ensure adequate sample sizes (Burch et al. 2005). We categorized wolves $ampl~d lor isotope dnaly,<:s relative fl) salmon and ungulate availability within home ranges or their packs. \Ve c,)mpared pack Sil.6 in fall (ca. I October) and lale winter (ca. 15 rVIMCh). home range sizes of wolf packs. \"lhin-pack wolf (kllsities (pack size/horne range si/e), and ungu· l.ltt:: \\011' r,llios relaLi\T to salmon and ungul,llc dl,;LrilJutlon in the sllldy area. We limited analyses of h(ln1e range size;; 10 those based on ~40 radioloeations Lo redUCe sample size effects on horne rang;e estimate,; (Burdl d al. 2005). \\'C used within-pack densilies ratber .hiln densilies calculated for regions lJf the ~l\ldy area because Yhe numbers of' packs within regions were often smdll enough to bias density estimates (fewer \han six pack,;; Bu rch cL al. :W05), and average wilhin-pack den,,:li;;;, were strongly correlated wi{h population-wide e,imHHes ()( \\011' density acros,; l\onh AmcncCln slUdics (,. .~. (1.95,11 = 30, P <: O.OOL data from Fuller el al. ~t)(I.l: 1(,5· 174)
Sample colic! [iOf)
During I986-100L we obtained samples of hone from 73 wolves lhat were equipped Wilh radio c:ollar,; and
monilOred as pan of the DNPP research and that died Hl
or near the ~tlldy area. TheSe wolve.> were 29 months lJld ""hen illltially ..:apture<l and "vere radiotral:ked Cor all average of 761 days (range 19-2665 days). We caLegoriled eilch wolf based on whether ~pa\Vning salmon w~re
likely Lo oc:cur within the home range of iis P'ICk. and whether ils pack. lived predomillantly on the nOrlhwe,iern flats where ungulate abundance was low. yieldmg ihl-ee g;roups: (1) salmoll pre;;tl1i, low ungulate density, (2) s<:llmon present, high ungulate demilY; and (,») no salmon, high ungulate density (Fig. 1).
The isot,Jpic composilion of ungulate prey v.·as determined from blood samples coikcled in iall (late Seplemher-eariy November) and late winter (mid- lO bie March) during studie,; of lTloose and caribou in D!', PI' (Adams and Dale 1998: L. G Adams, unpuh
lished data) and blood samples and lmir collected in lllldMarch from Dall's shecp 60 km eaSt of our study area (Arthm 2003), Red blood cells from fall and late win t~r
were assumed to represent isotopic composi tilJn of these ungulillc,s during the ihree lTIllnths prior Lo colleCiion (mid-summer to fall and mid- !o latC' "","iTHtr for fall and late-winter collections, respectively; Hi ldcrbrund et al. 1996, Ben-David et al. 20(1). whereas sheep hair sarnple;; provided estimate, during summer and fall \~ben the hair was prodUced (Hilderbrand el al. 1996, Danrnont and Reirnchell 2002). {SolOpe compOSition l)f "i<:llmon \\-as determined from recenLly spawned bll chums collected ill rnld-OctlJber from rhe Toklat Spnngs ;,pawning area. All lIngul,He and ,(lImon sarnpks were kepI frozen until processed for lWlOPIC analyses.
Sample preJ!arariou lIud Isolopic ana/ysi.,
Wolf bones were initially cleaned of soft tissue" ill hOI water and detergent, then broken inw smill! Chunks. washed in a weak detergent solution, tboroughly nnsed In deiClnized waler. and dried. Next. s,Hnplcs were grOlllld in a cryomJlI and lipids \\"de eXlral'kd by washing three timcs In a 1: I dllowfoT"rn: melluilol mixture. Approximately 0,5 g of rcsultmg powder was washed Lhree times with 10 rnL 0.25 mol/L HCl and alrdried. Collagen was cxtraded fmrn hone as described by ChIsholm et aL (1983), then freeze-dried and gnlllTld. Red blood cen,; collected I'rom ungulaie prey weI'.:' freeze-dried, then ground lo a fine powder. Sheep hair samples were washed in a mild detergent solutioll. )insed several limes in dislilled waler. washed three Limcs in a :2:) chloroform: methanol solvent.. allowed to air-dry, then Anely chopped. Salmon tissue samples were freczedried, Iben ground LO a fille powder in a cr','omil!. Approximalely 2.0 mg of' each l)f' the resulting sampll', ""ere loaded III 1() Lin capsules fur Isowpic nnaly,e;..
Prepared samples were analyzed fM stable C and 1'\
IsoLPpes by continuous-flow isotope ratio TlMSS "pee' lronl<?lry uSlllg illl elemental ana!yler cNlpled llJ a rnd~'
speclrllrneter \Fry l'L al. 1992). Results are reponed in 0 notation as deviations in parts per t!wu,;and (%,,) relalive to a sLandard (Vienna PeeDee Belemnite [VPD(3) and air
255 hnUM~C 2010 SALMON A!\D INLAND \VOl.F-l.'NGl'LATE SYSTEMS
PI,'H. 1 i\ w<.lf(Can!< IUp'I,I) in Denali '1aLlofJal Park ,111<1 Pr~scf\'e, :\Jasb (USA). PhoLo credit: Tom Mell;L (,1 S Na[inllJ.i Park Savicc
gellelally enrich I"sues in 13C ~lld ,.''') (Hilderbrand <::1
aL 1996. Kelly 2000) To account for diet tissue
dlscrimi;lJUon, we used values for enrichment from diet
where X is 1'c or ")\; and R i~ LIl" appropriate i.lC: 12C to bll'ocl cells in caplive red foxes ( VU/(ieI vulpes; +0,6%"
or I'N,HN raLio. Analytical sequences included labora· and +2.6%" for 6 1·,C and 8 1'N, respeClively; Roth and
wry standanb, and reproducibility of rcs,lIts \Vas Hobson 2000) and addcd adjustmcllts for enrichmcnt wt:
generally beller Lhan ±0,2%Q based on repealed analyses noled from wolf red blood c-:Jls to bone collagen ( ;·2J}%" of standards and samples. and 'j.(). 1I 'Yo". respectively; L. G. Adams. unpuhlijh<?d
l'olOpic r~tios cOnlInonly chc\nge as dictary constitu do iii); Lhus. prey signatures were :ldj\lStcd by -;-3,2%" and
ent, 'Ire assimil'iled via differential retention of isolOpe;, -'-3.4%0 1.0 account for diet 10 bone collagen disGrtminfl
dUring mcraho]isln and tlssue syn Lhesis: these processes ti;m of I 'C and i5 N. respectively,
'f.WI I I \V"l]' di~r <lnal)'l" (ill C and 0 IS1\; ihml bone eolla~en dnd ~,"Ilmated ,,,Imon (OI7('ol'flJ.'J}('u.\ ';1'1'.) consllmpllOn by Wl»)VCS
(C01j/~' lupu.\) (dt!o~g()li7Cd by !he- p.re.,ence or clbsC'Bce of' ~rawnlllg salmon and ungulate abundance (low or h!gh) wIthIn t!l(']!
h'll11C rang'"~, O"n<lh N'llional Park ,lnd Prt>,;ervt. AJo,LJ. LJSA, 1'186·2002
8J 1(' t~(JL)J 8J)~ \'Yo,,) Salmon )11 diet ('l,.)
Wolr' grIm!, /I Mcnn SD Range Mean SD R<lnge Menn SI) Range
!) Salmon, 10\\ I.IIl~ulate -'-' ··20 I 054 ·21 2 III ·191 7.1 0.80 I) I 8Ci 17 :.3 g J4 2) S,tlll1on, high ul;gulate 29 -197 0.57 -20 S 10 -18.7 6.[ 0.64 5.1-7.1 8 5.7 0-· hi \) High ung\llak 1)111y 17 -19,8 067 -21.210-18.7 54 0.45 48 6.3 3 }Jl 0 III
Cnmblned n -I\>9 0.61 - 21 2 to -18. 7 6,3 0.94 4 8·,'j \) 10 X2 0 14
I L':Oiugil.:dl Apph"::d tl~lnsLA YNE G. ADAMS E'l AL
V,,1. "n, Nu
18
16
14
12
n
~
.?- 10
'r
\ .. B •
, • 0 eo 6· c~~c ..
+~h~'P 0
Moose
2 .
-20 -19 ~18 -11 -16
013e (%o)
Fir;. 2 Value. or 6J\' J.nd 0,51\ frorn b,mc colldgell llf'
wolv", (ll -"- 73) Denali NilTio!ut Purk and Preserve. Aia"ka. W'oh~$ ,·,ere calc:gOl'ued by salmon ,-.nd ungulatc: hvail,lbilily II Ilhin tlKlr h0rne rang.cs: solid circles, gmup I ("dmon, low unglll.th~ abundancc): gray circles, group 2 (salmon. high iJniul.lIe ,Iblll'danl'e), .Ind open circles. gl'oup 3 tno s.lhnon. Ill';'!; "nguJ,lIt- .1huf!lL"1Ct.l 1\-t<:Ull values:!:. SI.) arc ind,c,ttcd for uIl!lld.ILc, prey dlHi ~rllmon cr~bk 2) adlusted to account 1'01' lldphic e·nrlL,hrncnl (..,.1.~%. and T.14%" lor 6 f'C Hnd lifSN,
r~~~rt'c l i"ely)
Dala Clnalysi.\
\Ve <"mploy~d K neart'st neighbor~ rillldomizaLion Le,ls tKNNRT; RO$ing eL al. 1998) Lo e\'aJual~
aifferenee~ in isotopic composition among the three groups of wolves. sea.;onal isotope values for ungulate prey, ;).nd iSOiOpic signatures among the three ungulate pr,';- and s<llmoJl. IkcalJ~c of diffcrcnc<;,s Iwtea in sen"oi\dl 1,0l0PIC composition of moo,e and canhou. we dl\;ra:,red se,),onal vallh:, to derive a YCilI-round CS{Jrnal~ j~'r each ungnlaLe ,pecics
Became we were primarily interested in rhe proporlion of salmon in assimilaLCd wolf diets and differences between N isotope ratios of salmon and ungula tes were subsl.antially greHler than t.hose of C (sec Resuils) , cOllpled wilh serious concerns recently rai$ed regarding fhe usc orc isotope raLios in estimating diels (Feliceui ci al. 2(03), we used a simple dual-source mi)\ing m<1Clel (Hobson el al 2000) based Oll N isolope values to eStimalc the salmon contnhurion to (he dlel (If each wolf we sal11pled:
WhCIC p,.,) was tbe proponion 01' a wolfs diet derived from salmon, o:sN wlr was the isowpie value for thai wol.t'. o'<N UllO was Lhe combined isotopic valu~ fOt" ,til llnguJales in the diet. and 61SN,a' wa~ the isotopic valuc of salmon. To determine Lhe ungulale end member. o"N"n'l' we used the proportions of ungulate bib observed while radiotracking or ~llowlral'king wokes during 1986 1993 (Meeh el ,11. 1991\; L G Adarm. /.II/published dala) as ,ipprOXIlJlati(ln~ of the proporllon· al biomass of each species consumed by lht.' woives (Mech el al. 1998). Because the composition and abundance of ungulates available in the northwi':stcrn flats differed hum the remainder or our study ruea, we estimaled lht' ungu]are composition of wolf diets separately for each region (northwestern llal$, 39% caribou, 61% moose [II = 145 kills]; rClnainder, J9{r, caribou, 43% moose. 180/.· sheep [1/ "" 371 kills]). We assumed thu L prey other Iha n lIllgu]aks alld sulmon (bea.ver. snowshoc hares, Arctic ground squint'ls. hoary marmols, etc.) mad.:: lip a ,mall portion Lli woll dieLs and were' slmrlar !sOWpie<lli y lo lhe ungu!cll'S ISzepanski el al. 199'). Lrt.on and Hobson 2005). Small ne'g-::llive estimate, of salrnc1n consumption tgreater lhan or equal ll) .. ·3.7%; 11 = 9) were considered to
maica[e no sa!rnon In Lhe diet of lhuse wolve, (Phillips 200 I J. We tested for differences 111 s~llmOI1 consumption among the three wolf groups with one-way Al\OVA proccdures. For all ksts, we consickrcd P ~ 0.05 10 bc indicalili~ of a significant result.
RLSULfS
Isotopic values of bone collagen ... "ried widely ,JI1IOl1g
the 73 wolves we sampled (Table I, Fig. 2). hill a clear paucrn emerged whclI st.raulicd by their home rang.:: locations relative to salmo!) and ungulale availability. Wolves belonging to group I tsalmon prc~enl. lo\\ ungulate density) e)\hibited isoLOpe ratios Ihal were diffcn:nlfroUl those of groups 2 <Ind 3 (KNNRT, P < 0.00 I). primarily because l)f 01sN values lhaL averaged > )')(., higher (Tablc I). Isotope sign<11 nrcs llf wolves in groups :2 (salmon present, high ungulate demiLy) and J (no salmon, iugh ungillate dcnsliy) did nOl differ slgnIiicantly (KNNRT P = 0.07h, Table I), hut group
257 JilnUM:>' 2010 SALMON AND INLAND WOLF-UNGULATE SYSTLV1:S
'J ·\ULL 2. ]soil)pe 'Ign,ltures (mcan "!: SD) for ungul,Il<:S and salmon il.vallable ~o wolvc~ In Denali Natioil<l.1 Park and Prcser~c,
Ala~b
Allnual e~;jmate-
:i"mple type .jlld pcriod col!(,<;ICI\ Ii 1)"C ,,1~N K nearcst neq;hbors tcstt o13e oj,\! ("':.Hibou
"'all Rile (late Sepfearly 1'-0\') 30 -21.0 -: {J.33 3.0 ,. U.51 P <: (l.(JO I - 22.6 :- 0 31 2;)", 0 56 Lare ',.\'l1HeJ R BC (lTI1d i h,te Mar) 28 -22.3 :!: 030 "2 7 ± 0.60
'vf o"s~
Fall RBe \e~.r1y No~j 29 24.9 -+ 0.63 0.7 ± 0.84 P < 0.001·24.8 ± 053 I!::: 080 Laic wll\ler RBe (mid-Mar) 30 -246 :': 040 1.4 ± 0.75
Sheep
Late w,lller R BC (llllll-!\1cu) 21 -24 I -+ lUI I 5 :t 0.72 P '" 0.267 -24.2 ~ 0.39 1.5 Y (, 59 HalT (mid·lVl,ll) 15 ··242 ... OA9 I 5 ., 0.33-
S:liit;on
Sk.!!l·'Y'''i,L·bone (Od) 27 207 .,. I 01 17.7 ... - 0.88
Viii," ~c':J'I)nal i),)lOP~ r;lLll" were a,eraged [,) arri\l' ~l <til annual c,r!m,ll<: I'or each ungularc 'peelc,. RBC ,1,lnd, I'N red blnod ..:C'~ lS
., Se,h,JIlai i'OI<JPI~ ,"Iu,', w~re cnlntHred \old K il~Jre,t ilt.ghbor~ tc,l, de,crJb~d by RO$ing ei al (1998)
~ wolves had isotope value., that ",'ere generally
inkintecli:He betWeen gnHlp, 1 and 3 (Fig. 2j.
The isotope sign81 ure.> of the three ungulate 5pecies
and 'i.llmon differed ~ignijicantly in all pairWise com
parbons (Kl";~RT. P < 0.001; Tahle 2. Fig. 2). Salmon isowpic \·alues dilTcred SubSl'llltid.l!y from the ungulate
"alLIes, c'xeeeding uO\ribl)u by I.'}'}';" and 9(J%o IlJr /'i1~C
<inti i;"\.I. respectively (rabk 2). h,imal.c, uf the salmon ('onlrib\llioll to diets or
indiVidual wolves vari"d widely. ranging from () to .14%
of Iheir !.)ng-,crm aSSlmilatcd dlelS (Table I. Fig. 3). ESlimates of ,<t!mon consumplion varied ~ignificilntly
among the three wolf g.roup~ (F~.7u :::. 36.4, P <: O.OU I) and were consistent with expectations from prey
availabIlity: groLlp I wolve, avcraged nearly thrc~ times
a~ nlu<.:h salmon in 1heir dicl, as wolves in [be olbcr
group> l'ombined (Tahle 1). All group 1 wolve, had
Ii "N \'<lIne, lIldicaliv(' of salmon eLJn~umption 2:l\%. whereas nem Iy half (14 of 29) of the group 2 wolve, fell
belo\\" dUll leyel and only Oll~ wolf in grollp J exceeded
H% (hg. 3). The ungulatc endmember.:; «)I~t\,,,~) we calculated I"nr regl,m~ Cif low and high ungulaL~
abundanec wcrc idcl1lical (I.H%,,). Lhwi variation in
eslimates of salmon ill lh..: diets of individual wolves re~ulted entirely from the variation in their I;I.'N values.
Ungulate dcnsilic~ were 78% lower in the northwest
em tlal~ compared tu the remainder of our :study area,
but withln·pack wolf den~itje" were reduced by only
about 17% (Tnb1l' J). A~ a result, ralios of ungulates Lo
wolvc,> differed \\idcly bClwc.:n Ihe t\....o regiom ll2 and
4·f nH'''~e cquivalcnl> '\~olf 1I1 rcgj()ns of low and high
ungul:Hc abundilncc, re\pccl1vclyj. [)ifrererH:<'~ ill wilh
In-pack wolf' dcn-;ilic~ between the two legion~ resulted
from combined cll"eets or slightly smaller pack sizes (9%
and 5% in [,lIJ and 10lc winter. respectively) and slightly
larg~r holtl~ ranges (9%) for wolves mhilbiling the low
ungulate area Crable 3).
DrSCUSSIO:>J
Althotlgh the wolves we st udicd lived ~ 1200 river km from tile coast, Pacific salmon were UlJli7ed to varymg
degr~~s by wolves throughout the Dena!i eCOSY'lelTl. In
particular, salmon cOl1tribured mo~t t(1 diets of W()I\(~S
where sa!nwl1 were abundant and ungulates ccclJrre,1 iiI low densities: all these wolves had N isowpc ratiO,
mdlcallve of salmon consliLUling 2:R% of the:r diet and
salmon made up .2:20% of the dIet for one-thtrd of them.
Givcn that bone collagen provided isotopic \'alues thaI wert inlegratcd ovcr ill kaSI a few ycars (Ganncs el al.
'998. Bochcrclls and Drucker 2007). thesc wolves included higher proportions of salmon in ,belr diels
during. somc years.
Isotopic values for wolves iuhabiting Del1dh's nOrlhwestern ftms were similar 10 tho,e reported by Szepan~ki
el HI. (1999) for wolvcs in coastal ;;otitheasl Alaska, ,md eSlimates of salmon in wllif diet' \Ver~ nearly Identical 117o/~ in lhi~ study \-s. !80/,. for '>oulheasl .Alaskan
vl-olve,). As noted by Szepanski el al. (]\)99), eoa~lal
\v()lws had sevcral olher tn~rine foods aV;lilal->k: [0 llt,'ln,
incltiding harbor "e~Js (PllOta Filulillll), various marine
manmwJ carcasses, anadrornnus euldchon Sllldl (Fha· Il'ichlhys pacijlcU,\). and marine lnvcncbrales. \\·hi\:h wer~ not accounted for in their analyses of wolf diet;,. Thus, coastal wolves probably utilizcd other marinc foods and [he estimated contribUlion of \almon t,) thc'ir dkl., Vias likcly inHaled. Given that salmon was Ihe only marine-deri \-ed !(wd rcsutlrec <I vail able 10 [kna"
wlllves. ,aIm on con,umpLIon by wulves in the rwrthwestern region of DNPP may h<lve a<.:tual1~· exceeded that of lhe coastal \wl ves ,;tudied by Szepanskl ei ,11. (199'1).
Wolves wi lh home ranges in v.-hich ungulales were more ~bundanL ul.llizecl ,almon less on average th~n
thos~ Wilh few ungulales available. However, SIX of 17 wolves we sampled wi lh no spawning salrnon wlthin lheir homc ranges exhibited OI.I N values indica.tive of
l [(D!\l~l("al App:luLth r:,LA yt\ L C; ADAMS ["]" AL
\"pj ~Q, [\0,1
40
U) ro
30 ro ~
0'> E; C ~ co Q. 20 <n
.<2 (f) '.fJ (1)
::5 " I'J' ro
I o L.. .
o 4-7 8-11 12-15 1&-·19 2().·-23 24-27 28,-·31 32-3.')
Salmon in diet (%)
b
, .J,..
II;
.-,.
"'~,, ., . •
~
Iii
II ? 3
't ., .... ~ 'po.
. t
, , /- .. ,. , J." ~ , ,
, I
Wolf diet group a _ 1. salmon. low ungulales \11 - 27)
C~::::J 2, ,almon, lligh ungulates (n" 29) r-----J 3 no salrnorl, lugh ungulates (n = llj
~ I
II
_J
------_._----~
FIG. 3. Pl';::vaicncc ofoalmon in lhedids or .....ohe' from Denali National Park and Preserve. AJ,lska(a) relatIve to IXriXntage of wolves ill e~ch group wll.n acc<:~~ to sa.JOloJ) >pa"rHng areas and ungulale ;nail;< biJilY a nd (b) as dbLribut.ed 011 I.he laJ)chc.lpC (circle, 11IdlC'lil': pef,-:eJ)taf'~ or ",lImon ill diel' lJco, solid ..... hite: 3d %. ,olid hlack), NUll1b<:red areas in pand (b) correspond Lo "'nil' dler group, l-} K I1(W. 11 and ~uspecled saimOll spawning areas arc depicted as hlack stream segm~nts.
,almon cunLribuLlTlg 4·-10'1;, (If lheir diets. Wolves are j1i1Ck ral1g~ Innlts, well wilhll\ dblancc~ regularly kno"'n to OCc:JldOn~~ll\' n1Hk~ long·disLance forays lrav~kd by wolvcs (Mech and 130ll"nJ 2n(3) 0111;.!de of I hell' Ic:rntone:" Ircsl'assJ(lj!. on neighbOring Fa!1 rUI1~ of chum :ifld cohu Iikd~' contributt:d 1110::': ill
",,<'H packs (Fuller It)~9, 1\1c<:h cl ;II. I\In), :wcl fi",: of wolt· diet.s in Ihe Denali regIOn. The\e salmon were lhes,~ wolve" were radiO-localed on such i'otay" wilhin a ~lIbSI,lOli:.llly m,)r<: ahundanl Ulan ,ummer-spawnmgIc\, kilomclcr~ of spawning areas (L. G. Adams, dmm and chinook. Also. unlike summer-nli1 salmon. unpllhhs!7cd dura). Spawning areas would be aLtractive those spawning in fall congregate in stream reaches that to trespassing \Volve> becduse of Lhe predictable commonly remain ice-free througbout the winter dllC Lo availabilily of easlly obla.ined meal Lhat could outweigh upwdling of groundwater (JTC l007h). Retemion or rhe risk~ associ,Hed wilh deleel ion by residenL wolve~ s,llmon carcasses in Denali slreams was probably very (J'...1cch and 130itani 2003). Allhough spawnmg areas did high because of bigh sinuosity. "iOl)dy debns. an,1 nOl C:-'Isl in several pi!ck home rang<:s in our study area. reduced base flows during winlcr (Cederbolm el (11. 1989. ,pawrllilg c\1'c:as occurred \\ ilhin 20 kin of lhe,e wolf Gende el al. 2(04). Further. decornpo.';itlol1 of earca,scs
259 ~AU'\'WN AND INLAND \VOU-Ui\GULATE SYSTEMS
r.\HIX J. Popul,ltiol! chdnlclc-ristics of wolvcs inh~lhitjng lcglon, of Dendll Na.llonal Park 'dnd Pr~sci \'C \I'llh tOil ungul.,k ,d:>\lnddlh.'(; <JllJ spawning siJlmt)!) (grpup I, 7(1 nh10S~ ~qlllvalenls:lllOO km~) \'s. high ungul<tle abundance (gI(H'p., 2 ,I11J 3, .120 rn(Jo'~ cqul\olel1i,'W(jI) krn") during 1986-2002
Gr,JUp I GruL\[ls :: dlHJ 3
Charat:lerisllc n ivtean
Pack sm: (woJ\.(':;/p;tck)
1',1'1 (I t.kl) 67 ' , L,ne IIllller (I') v1;\I-) 60 6()
Ilo~Yl;: ri.tng.t :=.itc (krn2 : 53 15(,7
Wilh:n ·p,LcK density (,\ 0 h (',.d oon krn"l
~"i:dl (I Oct) 49 58 Li, l~ II in Lcr ( 15 ,\-Jar) 53 44
:V,,!('. The ;:',1l11ple ;:'12". I., p,LCk-vears On~ m0()~C eqllll·a.knt. ( 10X~I)
of fish d:.ing in idte fall would be inhibilcd by cold ,V!IlICr lcmperatllres. Thus 1~jll-rtlll salmon were probably available to wolve, to SOllte degree throughotll \vinLcl.
Fall ;<Limon runs of chum lInd coho constitute c,Jnsidcwbk biomass o;:nlering lho;: sludy area. Given lhe Il1Clgnirude of fall salmon runs lillcllhe thstribuliol1 or' :;pawnillg IMoiral wnbUi tho: [<'antishna drainage, we eslil1li\tcd thal s'Llmon hiomass on the northwe>lcrn flal~
of the: slndy ~\.rea dveraged approximately 150 Mg ,Innu,llly, equivalent to ".. 70 moose/lOOn km" lavcrage moose 350 kg: rranZll1ann Cl al. 19 7 1::; NPS, 1.I/1j)ub!!shed m<1nuscripts) or equivalent to the ungliialc blOnldSS tbere Jn the remainder of the sludy area. ~alrnon constilulcd a small fi'aeuon or-the, availablc prey biomas, hecall~e or markedly higher ungulate ahundance and limited s:llmon spuwning bahital (Fig I).
Wolr Jbund"n~\:: ill the northwestern flais was only ~llghllr lower than in the remainder of thc study arc.. , ,;\'en ll·h.Jllgh unglilates occurred al suhSlilnlially lo\"er dcmillc~ V>/c .::oncludc that wolf abundanCe in thi\ region WeH enlMnced a~ a rc,uil of the al!ochthonous subsid) provided by salmon in that: (I} wolvcs are known to exhibit ,I srrong numcrical rc~ponse 10 prcy availabihiy (Fuller 1l)~9. ruller c1 aL 20(3); (2) salmon provide a food resour<:e equal in magnitude 10 ungulate ab\lndance; ;llld (.'\) the u~<~ of 'aImon by woh'es residing in lhe ,l[c',t Wih llbiquitc1l1S, accounting for [7'}'<) of their diels on average. Marine subsidies have' been sh()wn to 1I1cr';dse abund,lllcc of other tcrrestrial predmors across :J wide :.irray of laXH inclnding spiders (Poli, and Hurd 1996). lizards (Polis and Hurd 19961, Arctic fox.:s (A [ope.\" [ag(J/ills: Ro th lOU]). coyole~ (Cal1i~' [omJJ!s; Ro'e and Polis 1(98), and hrown beM' (Cr.llis urc(os;
Hilderbrand el aL 1999). I ncr,'"sed ahundanee of ~ub~ldizc'd eOil,mmcrs gcner
ally rC~llits Il1 incrc<Jsed predation pressure 011 local prey rcsourc"s ((loll 'lnd Lawton 1<)94, Pvlis d al. 1(97),
pill'1 ieul.!! Iy when pr~y ~ub~idies Gceur 'It high levels (Hll\eJ and McCann 1998, Esl~S cl al. 200J). Although ~<llln(}n co11SlilUlCd il ~i?ahle pOllinn of the diel for wolvc, \Jll the nnnbwestem Aars, un~uLltes still c:om-
Sf) Range n Mean SD Range
.1,93 U9
8121
2-23 2 13
482-4005
137 129
122
80 63
1420
5 :;0 ·UI
759 I
2 29 2 2~
226-4417
, 74 235
11-166 i 2 Il 3
116 119
7.2 52
4 tJS 3.27
06· 20.2 or, 13 ~
I moo,e. ;1 ~M:b()u. or n sheep. rolk'''dng K~ilh (In]) and Fulkr
prised an e;;timaled 83 ck of Ihe wolf (hel~, inchcalive of substilntial predation pressure on tmgulales, parflcularly givcn the low ratios of ungulates to wolves that occurred there. With information presented hcrc (wolf diel composition, wolf densilies, a11lJ ungulatc (knsi,ic.,) and estimates llf consumption rales for \\'(llves (PelerS,m and CiUCl~1 20(3). we apprm.inwled wima predalillll rare, (15 October-30 April; rhe period from when estimates of rno,)se and caribou ahundanee werc denvcd to immediately prior to rhe ,Ulllual llngulare hirth pllls('). exprc<;scd .1S lhe proporlion of iTJoose eqlllvalents In cach rcgillll of om siudy Mca lhal welT c<Jn,umed by wo)vc~. With thc cons.:rvctlivc assull1ptioll that ,lil ,ali non cllnsumplion by wolves occurred during Ihls period (differences 111 predatlon ralcs he tween regiolls increased with lower salmon conSllmptioll dunng the period), estimatcd pred<:nion rates on ungulate, in the northv.·eSlnn Aats were approximately lhree tjme~
hight'!' than those in Ihe remainder of our sludy areCl (!9Sh \-S. 6c,-i. respeclively). These eSlimale~ are reaS()J1able gJ\'en SlJrvi\'al patLern, of moose and canbou HI
DNPP (L. G Adams. Impllhlishcd data). \fore nnpo!lan!!y, while the valLle, oflilesc r;;\tes will vary depending on the consumption fa Ie used to calculate them, Ihe
relative difference bct\-vcen the two niles result soitl) from Ihe dirrerence~ in wolf and ungulale abundance and ,~allnon clmtribULion to wolf dieb in the two rcgi()]]~
of our study area_ Lhus providing slrong evidc:nce I hal wolf predation rates were markedly higher Oil ungUlate, on Lhe northwcstern flats Further, given [he low ratio of ungulales La wolves on the nonhwestern flat,. salmon would have io con,litulc -40% of annual w(Jjf dict~
Ihere for predmion rak, on ungulales 10 be sImilar in the upland regiolls of the siudy area
\Ve conclude lhal sal1110n as all allochthonous sllb~)dy
(c) wolves probably contributed 1.0 low ungulate denSities obscrved in nonhwestcnJ DNPP. Moose cOll1pri~cd
most of the ungula le biomass III the northwestern ILlts and occurred at densil1es approaching the [llwesl in NOrLh America (Gasaway et al. 1992, Mes,~ICr 1994) while exhibiting body size and rerroduetl\-c eharacleriSlics indicative of little null; ti0113 I conslraint ,'m the,r
&~ological Applkal!or,l;;260 LA YNE G, ADAMS ET AL V,'l ~(), No 1
populaliOll dYl1<llIllCS (Boenje ei al. 2007; L. G. Adams, lIlIPUf.>lishC'd daro). Further, moose populatiom with 11100se: wolf ratio, <20 generally OC,'lLr al very low dcnsiues or exhibit popuiaUOJl decline, compilred 10 ureas where mo,)'>e: \Vol!' ralio, arc higher (Gilsaway et ,II J Yl:i3) \\'ilh bn'wn bcurs anc! f\mcncan hlal:k bears (t .. , .1 iI I" <1111t'f!( Will.') in the system. rn,),)se populalInlls are <:\'en lJl,lno >[toug]y lirnited by the combined predation dfecis (Van Balicnhcrghe 1987. Ballard and Van Bi\JJenberghe 19(7), 'vVe spe:;ulmoO Lhat Pacific sillnwn may playa similar role in other regiollS of northwestern ~orth America where spawning Sillman are abundant. and tlngnl~le densities are low.
In addition to effed; on the ]ong·lenn numerical relalionships between Ivolves and ungulate~. Pacine sairnlHI probably irnokc: substantial variability in \VolfL1n~L1lale inlen:cllOl1s rnlm year Il) YCdr. Rew)"ns of fall churn tu the K,lnll.,hna drainage: vilned fiH:fold dunn~
1'J'N·20(15 (( blry and Hamazaki 200h) and indoOx c,'lInU of fall chum ,ainll1n have vaned by more than an order llfm;Jgnitudeduring the bst 35 years (Buc el aL 2006). TIm,;, lh~ i:t.vailahility ,11" salmon in northwestern DNPP bas probahly varicd from less than one-third to marc (han three limes Ihe ungulate biomass there. Faclors influencing annual run sjze~ for Pacinc salmon are cOlnplex and include oceanographic and climatic patterns over the 1\l1i"lh P;'lcific (Down ton ,end Miller 1998, Finney et al. 2002), salmon harv~'ts on the high ~Ci1' :lnd wi [hill J ivcrine system'> (J-h)lder :l.nd Seneca!Alhl<'chl 19\18. SdJlmiler ct al. 2005. Jrc 2007a). and <'-;cap..:mcnl or salmon 10 spawning. areas in previau.> }'('M, (i-Julder and Scnecal-Albii:.'chl 1998. .fTC 20070.) Thib. farLOr~ ;lSsocl,lled WiTh di~t<Hl1 lllarine envlr(Hl· menls and the compiex management of sa]rnrm fisherieS likely influel1l'c tile snbsidj1.ing effects of salmon l)[I
iuland wnlf-ungulate systems, Currenr understanding and management of wolf
ungulate ,ysrcms is based on the assumplion that effects of olher prey re,ourL'CS are minimal (Mes,ier 1994, .\-lcc!l dud PelerS,HJ 2(03). That assurnptic'n may he rC,bllf!<l hk wllere ullgulale pr~y ar~ a hundant and allcrtlaiivc prey make lip a ,mall prop,lrlion (if Lhe prey hl,'Hna~s ll" ii7('d by 1',.-oIH;S, HoweVcl, Ll1I"\ltlghoUl the Clirrenl range of wol"'e~ Jl1 "orlh Amcrka, !cl\v-densiLY wolf-ungulate ~ysrcms are \juite common (Gasaway el al 1992, Fuller lOt aL 20(3) and the,e sy,tems are prone LO bc InAuenced by llon-ungul;He alternative prey (Dale el al. J994), For example. Mech C~007) recently reponed Ibat (rends of wolf ahundance on Ellesmere Island in thoO Canadian Arctic Archipelago were corre);]ted with Arctic hMe (Lepus i1l'clic1.l.I) <:lhundance. hUl nOl with lhat l'f Jllll:iko'< (Ol·ibm· IY/oscharus), (he predominant l!1lg:UI<HI' in the regi<)n
OUI' li!ldJn~s indlc'alc th;H Prrci[ic ,ahnl)ll are an impon"nt f00J <;our,;e [(,,. wolvo:;, well beyond coaswi <lr~<1S dnd can prll\'lcJe a substantial marine lIltluenc<: un wcllf--pre} system, far inland where ungulates ,wellr at low densiucs. HCC,HIS<' salmon an.' allOChthonous sub.,;i
dies in these \....olf-prey systems, tl\ey can reduce ungulate: wolf ratios lhrough numeri(;ill responses or wolvcs to :;almon ,wailahilily and heighten predation pressure 011 resident ungulates. Funher, the ahun(i<lncc elf spawning ~almOI1 varies widely among years, ..Leiding suh,Wnl1dl \·ariabihty 11\ Ihe realized effects of thl> subsidy on local woif.-ungulate communities, Thus, the >pawning migrations of PaCIfic salmon provide ;) previollsly lIllrt'cognized and dynamic {'on neCl.!on between inland wolf lIngulate ,;onuuumtics in norlhweSI<.'1"[1 North America and distanl marine ecosystems,
Th" ," udy "as I'unded by ih~ Nall'lnal P,Lrk SerVL~e and U S Ger.,llJglcfll Survey. This rroJed l'ould nol have been r(l~;;ll'k
wilhvul the dlorl; of T J. ?v1eicr ,mu J W. Burdl. who mcl.;cul()u~lv u~hived bone Samples l'fllJl1 ,rudy woh·cs 111<11 di~d dUring i9H6-199}. long. bd()f~ W~ bcc<1me ,nt~re,led :n e'"illl,lling .',llmoll consllnlpl.wll by wolves \Ve thunk f' "..I Cleary and R R. Holder for col1ecl!ng salmoll tis;,lle s,lmplcs S 1\:1 Anilur gr,lCIously supplied lh~ bl00d and hair sampk~ or Dal!", sheep C. I< Bern, C. L Kester, C Klcckncl, and E ShOCn,l! as,isled ~·ith laholatory 8n~iyscs J L Bcl~nL
M Ben-Da\'ld. B. W Dale, J. A. Esit.,. S M Gcndc. l D Mcch, T, J M.:icr, anli J. W. Testa provided useful cnmmcnt~
011 prevIOus drafls of this mft.nu,;clipl.
LITU\.-\ rl.lt" CiTeD
Ad;uns, L. G. 2005. Effects or male·rnal ch(l.rac(ensl,,-,s and cltmallc I'ari,nion on blflh rna;Se; or AI,lsk,ln c,1I'lbou JOUI rlai of :Vlammalogy &6:506-513
Ad,um, L G, Jnd B W, Dale 1l)\IH I<eprodltctl\c IJcrl'Ol' rrlllrlC:C of rem,lk AJ,hkan Larib')lI .h>tlnHlI of Wrldhl"c \1~n.lgen~cnl 62:11$'1 1195
ADFG [Ala;;ka Departmeni of Fish and G:i1n~J. 2nOli 20lJn J\laska (,;0 mmc.rCI,lf ~ahn('\11 hdr\.c::~t~ ex\·~c,;ci vc..;!ue.':l. ~y.,\\ ....v c( acHy. Sl~lle ,lk n.=,:gen\nl{}'11l!ii~h·\~ilm0n c(dch\,'.d:bltlsh ..~ci
UfJ~~wsi pdf:: Arthur. S. r,,1. 2003. Inlerrelatiol1,hlp~ 01' O~li shLcp and
predators m the ccntral Alaska Rangt Federal Aid in Wildlife ReslOrall(,n Research Final Performance Report. I July 1998 30 Junc 2t)03. Granls W-23-2 10 W-3:l-1. Pr<)jecr 6 )3 Al,ls1<a Departmcnt of FISh a nd Game, JUrl~a\,. Alaska, USA
B~lli\rd, \Y, 8., :Jnd V. V~n Ba.lknberghe 1997. Pr~dalor prey rciatiClnshlps. Page; 247-273 iii A W l'ranzln'11111 and C. C SChWM\Z, edllClr!> [eology and ll1dn,lg"men: of the l'onh Amcncilil moose. Slnll!l;VI1l.1I1 1n,(1I\111011 Pre,;, \V.-,s!lll1g· lon, D.C. USA.
Ad,lnl . .1 L, K KI(:lland, E. H. Follmanll, and L G. Adams 2006. Illlcr'peciJic resource paflHlon;ng in ~Ylllpa(f1L ,,-[sids E,ol\lgi~al Apl'lical:on, J 6:233., 2343
Ben·D,meL. M _ R. Vol, FlylJn, and D. M Schell. t997(1 Annu,t! :Jnd seasou~1 changes in did, of marlens' ""idenl:e from stable Isotope analySIS. Oecologla III 28U-291.
Ben·Da,id. M" T. A I-Ianh~y. D R. Klein. <lnd D. M. Schell 19'171>, Seasonal changcs in dlcts ofcoasl,t1 and rivenne mink thc TOle (If "pa",ning Pacitic salmon C,lllfidi"l1 .Jol,rnal 01 Z,Jo!ogy 75:803 811
B~I1-Ddv:d. M" F: Sho.:hat, and L. G Adam" 200t. UlJill) or ~Ldble !~OlOPC i_JI1al}f;l~ III "luo:J'ylng {'Otrlgtng ecniogy dt" h~rbIH)re, <",llllple, f,OIl} rnoo,e and carihou Ake, 17 4::1-·1""'
l3(lc!1crcns, }J • and [) G Orltcker 2o(}7 Carbollate ,lilhk iSNopes' {erresm'll teeth and bOllc,; P,tge, J()9 31' In S A LiI,IS. cdllor CncvclopcJia or QllMtenMrv l"C,CdTCh [lsc\·t~r,
Oxr0/d. LI\.
SALMON AND INLAND WOU:-UNGLLAl'E 5\5'"( [M~ 2(, \
B(),~rtle. RD. K. k K~llic, C T. Semon. \1. A. Keech. O. D Y"llng. B. W. D.~k. L Ci Acldm,.•l!!U A. R Aderman ~007
R"n~ing Ald,kd moose nulrllion: SIgnals l,) begin IjDer~1
i-lDL1crk" h.lt'vesl,. Journal 'If Wildlife Milndg~me\l\ 71: 1494-15fJ6
Brabct~. T. P. 8. W~ng, and R. H. :"'lcadc :WOO. Lnvironmenml and hydrologic overview of Lbe Yukon RIver [jasin. A.I'bk.l ~nd CanaJ:1. WaLcr Rcsources Invest,gJti0n RfpN( 99-",20-\. U S. GCOIOgiC,1 Survcy. Anchorage. A.laska, USA.
Br'lunc. II lVI., r-.:.. :\ I·Jobson. 'InLi 13. J M,done. 2()()5. RC!,ional differ~nces in coliagen slable i,0topc and lissu,~
trac~ element PIOI',jC~ III populatons of 1,)Ilg-lailcd dnck brced,ng in lhe Canadian Arclic. SCience of (he Tmal El\\ irOI\:11<'I\\ 346 I 56-1 6~
Bll~. t· .I. B. M Borba. und D J. Berg,rrOlrl 200b Yukon Rlwr f,.l!\ ;:]lUrn ~alrnon st,)ck ,latus and r.t1! seaSe'!) s,lJlnon risherles. a r~p()n to the AlasLl [jo~rd 01 I",sherrcs. Special Pubb::~llJon (J6~J6. /\lask t\ I)cpa.rt.mC'nr or Fl~b and Gan1c.
:\ncho"<Lg~, A!;JSK". USA. ILrch,.I. B . L G Auam'. E H Ft.dltnanrJ, and [' ;\ Rc,slad
lOPS. lvu!ua\iol1 e\j' wolf ,knsIly e,tlmallon J'rOin r~diOlc
krn~lry dnta. WildliJ'e SIxlely Bulklltl 3~: 1225 12.>6. C,~(k,hdim, C 1.. D B HOl1~r')ll. J) L C,)I~, and W. J
Sc,lri('lI 19H hHe (II' (\)1\,) sahnoll ((}l/cor/n,nc:!III,' kllwch) ~(ln.~aS50~~ In ~p\t\V1H.ng stream."), Journ<ll or Fis.he:rics and .:\qluti~ Sncllcc 46: 1J47· 13SS
Ccrkrholm. C J. \-1 J) Klll)l~. T t\·lurou, <lnd A. Sibutani. t01J9 Paulic S(ll,nnri carcaS$l;;"s: cs~cnllal L()n(ribl1(ion~ of nutnent." and energy I~)r ,Otqu;\lic ~lI'ld (el re~trial cc{)~ystetns
f· ishencs 24: 6 15 CllI'h,'lm.13 S, D E Nd~on, K A. Hob,on. H P ScllwdrcL
<1,;<1 11,,1 Knyr :98] ell bon lsolope n)e;hUr~menl \echniqu", 1c1 r bone eoll.!g~n: l\(lies l\H Ih~ ar<:hcologlst ,1o!il'IMI ,,1' Arcil(ol;'gICJl 5<:1"11<:'; !\):35S 360.
(hi'h'lhn. B. S.. O. E. Neisoll, dn(] H I' Schw:ln.:%. 1982 SLabJ(>cLrbon 1~(1[ore rall<» a\ a nleJ~ure 01' n1d.rrne \'"r~u~
lern'.'lrial prc1tem III anclenl dlel'. SCiCll<:C 216:1 131·J 132 Cb,ry. PM, Jnd T. HarniJZ:Jki. 2{)06 Fall chum ':iimon
m~~rk rC(;iJpltlfC <lbllJldanc(:' e~linlallOn on the T~nH.na and KJlJtlSh!ld RlvCJ'>. 200~. Fi~heries Da\,: Serie:; N\1111btl 06~67.
A!a:-..ka Dcpartnlul1r (lr FI::.h and (Janie, r\nthOLtge1 Aillska l
U~A.
D,lit. B. W .. L. G. Ad:\lm, and R. T Bowy~r 199-\. funcllOn,t1 r~Sp,)IBe of w(llve:: preying 011 barrell-gf<1ulld caribou 1I1 it
lTIlIll.ipk pr~J "C."Sy';Ictll. Journal or Anirn:ll Ecology ("64~652.
D,lI·lmonl. C T.. P. C Paqll<~I, and r E ReHl1cil~n 2008. Sp"Y>lling "i1I1l():l d:srupl llOpillc couphng between wol\'<:s and t:ngu)al(' pr::>' 1[\ ,oasla! Brrlish Cnlurn b\a. FllOlvkd Ccnl!'.l1 [,cology ~: I, I:'
O'\I·(111nJlI. C T .. 'tnd 'f. E. Rcirnchcn. 2002 Inlr,l·b~ir "aDlc i~I)~OP~' dn(dy~i~ IInp11':') ~e~·tsulli.d ~hlfllO ~:dn1ull III gra.y wolf dle'i Cd.lladLl;1 Journ:;! d Z(l01()g~! 80 J638-IM2.
DJrJmonl.. C. 1 .·l L Re'mellen .•llld P C [\ler".lel. 2003. Foraging behJ.vlou:' hy gray ~\'ol\'e~ on s<Jlrnon slrc{Jnls in cO'I,LJi Br;rlsh COllJll1bl~ C"nad,an Journal of Zoology RI: ~-t()-J53
Oownlon. IItI. W., and K. A Milkr 1<)93 Rd,l\lon,hip~
kl\;'~en Alaskan ,almon c<teh and Nonh Pacdk c1im,ll~ on inleIMI!lU,ri 111d Lnt'~rdeC,ldJI l;me scales Canddlan Journal or l':shc,'ie> a!Jd AquatIC Science 55'2255·-2265
Ellel. J Il .. T. R. Spencer. .1. J. PC!!.l. M M M~sud<l, ~nd R R. HolLie, 200·1 Dlslnblllioll al\d mov~rncnl patterns oC ~Il!rIOOk sfLlnl()1) n~t.uriling to [he Yuh.on Ri"er hnsin In 200n 10C,2 Tedllllcul \'!cmorHndulY! NMl:-'S-Al:-'SC-14~
i\,LlJonal Oceanic anlJ :\IIlIOsrh~n,; Adm;nistlaIIOI1. SeaHle. \V{~~lhtllglo;1. USA.
[lliIHdson. J M, M H (jnham. H. J BOJlrque, D Corbett, J ...\ [,Ie" ane.! R S SlCncck. loti? The kelp hi~h\Vay
lJyp:.Hhes1s: ~nann~ c..::~)logy, the (;()a~lal mignnioll [hl::or1', <lnd
the pcopllng of the Amcriea, .JOLlrn,t1 or lslanLi and Clast:li ArclH:J)[ogy 2:161-L 74
Esles. J.. K Cro0ks, ~nLl R. Holl. 200J. Pred3tor" e~ol.)glcill
rok or. Pages 857-87!\ In S Lcvin. edilor Encyclopcdl.t or biodiversity. Acaucmlc Prcss, Siln D;eg0. C.ahforrna. USA.
Felicetti. L. A. C C. Schwam:, R. 0. Rye, M A He.roldson. K. A. Gun I.hc'·, O. L. Philhps, and C. T Robbins. 200J. u~c
of sulfllr and nitrogen s[able isotope~ to determine the Importancc of whitebark pin" nuts lO Yello\<stonc grinl)' bcars. Canadian Journal of Z00logy 8 L:763-770.
hnncy, B. P. I. Gregory·Caves, M. S. Y Oougla" and J. P. Smol. 20lJ2.. fishcnes and produem·'1[y J1l the l1orihea,ICI n Pacine O~call ovcr the PUSt 2)()(J yeilr~ N,lture 41 () 729·- ?) 1.
h<lrlDnalln, /\ W. R E LeResehe. R ..;, R<l.u,ch ..1I,d.1 L Oldcmewr. 197i1, Alaskan J'IW(lse rnedS"r~lllenl> ilnd 1'~lghls
dnd rnc()UJ"'~lncnl \t~ighl rcldtl()n~hlr~ C.\)Il(!dt.lfl J('llrn,d 01 Zool<)g.y 56 2.9x ~06
f rich. S H .. R O. SlcpheJ1~oll. R D Hayc" und L BOlI,Il\! 2003 \\'(llves rLlld human, J'age~ 2~9 .~! (, In l. ]) Mcch and l,. Boilara. ~UllOr, Wolves. beh<lVlor. cco!ugy, ,Did COilscr\,atlon Ulllv~rsiLv (If ChlC3f!O Prt::)~) Cll1Ci:H!O~ llljll~J1~. USA
~I·V. 13., W. Brand: f J MC1~,c1l, K Tholke': 3nd R. (,arnlL 1992 AUIOJ1\:He.! aHaIY'I~ syslem ror C()uplcd ol'e '\.lld 61.'''l measurcrncllt,. An31yucal Chcrllistry 64:288 291.
fulkr, T. K 1989. PopUlaLJOll dynamiCs of wohe, In northc~nlrd.llv[lnncsoti'l Wildlik i\'lnnograph, lOS'!, ,"
Fuller. T K .. L. D. Mech. and J F Cochrane 200, W"JJ pOpuiHlwn dymlnJl<:s l\lge~ lill 191 //I L. D. M~ch .ll1d I FlOI1;lIH, ~dljors. Wolve,: D~hclvl(,r. ~('[1lug}. and <:UI1,"rV.I· liOIl Ulll\'CfSily of ChIcago Press. Chicago, 1IIInOI~, l·S ..\.
Gannes. L Z. C. Martlllcz del RIO. 'llld P. Ko<:h. 1998. NalUJ'd!
3bui\[hrnc~ \'Jriat!ollS In stable isoi<)peS <tnd \helr potel1!i,l{ lJ~e, in dl\irnal rhy,lo!ogical ewing) Comparallv~ Flloch<:mIstr} and Physi(llogy 119A:725 7}7.
Gasaway, W. c., R D. Boerlje, D. V. Cir"llga~rd, D. G. Kellcyhouse, R. O. Sl~ph~nsoil. and D. G. Larsen. i 992. The wk of pr;:u;;.!I,m In limiting moosc 01 low ,knsiliBs m AJa.~ka
and Yukon ,lnd impJiCillions for l'Onsel'vUlion. Wildlife MODogr:lpns 120: 1-59.
Gasa\\·~y. W. c., R. O. Slcph::n,oll. J. L Davis. P [ K Shepherd, and O. E. Bnnis. 1983 In,errclallonsllJp,; of \+olvt" prey. and nMill III mlenol Ala~k". Wildlrr~ M0nographs 84:1-50.
liend~. ,) M , T P. QUltlfi. \f. F Willson. R HelnLz. ,In<1 1 \1 Sroll 2004. Md.gnilllde ,Ind fale or s,i1mon·derr,t'd null ;Cll<> ond fl1ergy in iJ CUJ,I,d slredll) ~u.lSy>lell) Journlll ('I Fresh'Aater Ecolog> 19: l49- 160.
(ira,.,. M. R. R M Coleman. and R.. i\'l \'1CDO'>3il I~~~
AqU(lllc pwduc!!v!\Y alld the e\'ohlLJon or diildrolllou' lhh mlgrauon. S<:lelic~ 239:1291--129 .•.
Hilderhrand. G V.. S. D. hlll~)'. C "r R0Dhllls. T A. H,lIlley. K. r IlllS, and C Servhl.'cn. ]Q96. USC l1f sl",hk imlopcs 1(1
determine dic[s or liv,ng and cxtmct bears. Canadl,\11 Journal of Zo,)logy 74:2080-2088.
Hildclor,lnd, G. V., C. C Sch\vartz. C. T. Robbin" M. L JJlCf'by. T. A Hanley. S. M. Anh\\f. Jnd C SelVh,~cll )<)99. The importance of mea:. parlicularl\' )almon. \0 body ))/.,. populatlOlI producli\IlY. "nd cl)J]scrv<Llion of NCIl th .timer ieall brO\HI bears Canadia.n Journal of L0010gy 77'132 l iii
H"hon, K ,\. B ,'- McLellan, and.1 G. \Vc,ods. 21)()() C~in~
'Iilblc carbon (ii"C) d.l\d nitrogcn (o"Ni iso(opcs lO lnk'r \rophlc rciat'on~hips among bl.!ck (lnd gnal'! bears in tk uppcr Cllhunbm Rivct' basin, Br:\ish Columbia Canadi3n JournaL of Zoology 78: I3J2-I3W.
H.)lder. R. R. and D. Scnec<ll-Alhr"cht. compiler". 199~
Yukon River comprchenslve salmon pl,UI for Alaik.t Alaska Dcp.lrtmcnt of} ish ,llld Gamc. Anchorage. Al,ISk ••, US'\
Holt. R D. 19 7 7. Predation. apparent compcfltion, und [h~
,tructurc of prcy communilies. Theorcllcal Popellalil)!) Biology 12:197· 229
[\.·o!t)g!L~-al /\Pphl.dl,11IO::262 LA YN[ G. AD!\lV1S ET !\L.
\','1 :0. No I
11,)\1. R \.). and J H La'" Ion 1994. The ecological (:l)n~e4lL('nce, "I' ,hMed ndtl''''! el1eml~S Annuill Re'lew 01' [C"hlgy <li1.d SY~[CmJlI'" 25:495 520.
lillxeL Ci R. Hnd K :\·leCann. I'Ng Food web ~t\lt'ihty. rh~
lllnU~nL:<; or I.rophiL; nO\\~ (lr.:l()~~ haJ.)il.d(':-' An)';rJcClll N(Uu· ral"" 152Af,() 46\1
Iluxel, GR., G A Polis. dnd R D. Holt ~ll()4 Allhe fronller or tll~ l111egrari0n 0f (Dod \~eb eC010gy dnd land,cape .;cn!OE;\· P3l1~, 41~ 451 i,l G A Polis. M. E Powcr, and G. R'. '11l!~~1, ctlil0rs I ood w~bs ~r (he Iandoeapl' kvcl. UniverSity (If ChlC<I!!O Pj~SS, Chi,:a\!o. Illinol" USA.
JTC IJIlIli1 :rGChlllCnl C:ommiuee or ll~c Yukon RI'.,a US.T:anIldd P\lnel1 l007a '~{uknll R~\'cr salmon 2006 sca~on
)lIH1mary and 2007 ::'~d,.,on (llLlJo0k. Reglon ..ll Inf0nllauon Hepon \1107.01 Ala,ka Department 01 FIsh and Game. OjV!~h)n Dr C0n:rn("rClaJ Fishcfic~. Anchoragl\. Alusk~\.
USA. ,lTC [Jillnt TechmL,Ll C(l1l1il1Ji.lee oflh~ \'ul-,on River l:S.:Cal1ad:j
P~"ldj 2U«(i/, Inl.r0d"cl!OI1 t,) 'y uk,)Jl K,vu ,aimo;; <hllp:; ) nkl)i1 I"]'vcrpd Ih'L ,:\,.'1 rn sl!.Im.on,:'-1 bou ~ ') u kL)n'J 1\ ~r-::.~ilrnon·
ChU.lD'j
K,ilh. L' n 1%3 f'opuJ'llldll dynamiC' orwohc" Page, 66 77 ill L. "", CdfbYl1, ~dltor V>"ol~e> or Cl\1aJa and Ab$k,l. lheir >LaLus, bilJl()~y ,md 111al1d,gt'1)lenl Rcport S{;'l"It's 45. Canadllll1 WJldLl'e SCT\-i<:~, Edmonton, Albena, Canada.
Kelly. J. F. 2000. St.lblt' i$Ol0pes of corb0n and nitfvgen III t.h~
:--Ita!j ~)r ;.,Hlall dnd ina.mln~·JliaJl 1.T(lphi~ t'cl.)logy C:lnadian .hnl1-pal 1.)1' ZC'I0h)!!y (g. I 2.,
L,):'~JlI M _and R D H"il.2004 SPitl,,)1 110"" 'lnd rcgul,nion 'li"¢C(',Y~I¢m, AIIl~nL.m N,Jiurah,1 16~:606 6J5
i'vkci'l. L D.2nO? Annllal 'lreric "'llf p.lck SI~C related 10 arellc hdi"e numbers Ardi(" 6U'3119-3, I.
M~(;h. 1... D . L G. Ad'lIns. T J. M~ler. J W. BurdI. "n<1 B. W DJ.le 1'1% Th~ "olv,~" 01' Jkll~li. UJ1!~Cr~.ly or 1\·linnc,Ola Pr...~s~, rv11nnc(lpol~~. \·11Ilnc...oL£l. US:\.
Mt'ch. L [), ilnJ L. HOJ!al1J. 2003. \\'011' socllil e<:oI0g~. Pages I ,,'j 01 L D. Mech .1nd L. HOI t(mi, cditor, Wahl'S' hchaVII'l". ecology lind conservation. UllivcrsllY 01' Chicago Pre,s, Cbic1g0, lJ)inoi". USA,
M~ch. L. f) , it,nd R. O. Pelerson. 20l13. \\1011', prey rda\l0J1s. P,lges 131-100 in L. D. Mcch and l. Bo1tanJ. cditors. \\'(lJ\,cs: behavior. ~c(llvgy ,")0 cvn:>er\'Illic'n University or C'hicd~O
1';"'"iS, Chlcdgo. flhnoi'>. USA \-L-rL. J E.. ,mel P B. Moyle 2U06 S~lm0n. ~·Jidlirc, dad Wine:
nlhnn~-dt\n\cd nUlncnl~ 111 hllrnan-dol-lll!1d.l.ed e\:vsyslcm~ or ""ntr,11 C,hlomlil fcO!Ogl<:AI Applleaiil)]\;, 16:~99 ! ()(19.
i\·k"icI.I- Iq<f4 lJn;:ulatc popuimion models wllh prcda\inrJ (I CLle,c sll,dy .....'It), lhl.: Nerth AnH:'J H.,:cEl lTIOOSt' L:<.:<..']og: 75,-~7;"; ~1~~
!'.;',l:(>11,jl RC'C<JILfi COIHlcl 1<)07. \Vohe" b",[r~, ,"ld lhClr [\I'ey 111 f\bsk~: bwlo~I('.i1 <Inti ;(JCl.l! cndlienge" in \VlIJllr~
m~na!!cmcnl. Nationa.l Academv Pr.;:s,. W,lshington DC. USA
Pe[crson. It 0., and P. CiIlCCI. 2003. The ",oil' as a carniv0re P,lg~S 104-130 ill L D. Mech and L. 130itan;, cdil0rs. W(I!ves' beha\·ior. eC0logy 3nd c('n5crv~ti(Jn Unl~w'lty or Chicago Pre,5. U1Jcago, Ilhnois, US'\.
Phillips. D. L 2001. l'vli~lng model, in analyses or diet usi\l!1mllillpic slahl..: lSotopes: a uiliqllc. O~cologm 127: 106 170
Polis, G. A • W. R. Anderwn, ;iIld R D Holl. 1997 Tvwdnl,l11 InLegralion oflandscape and Cood weh ec;ulogy: the dyn,ume, or spalH,lIy 'lIbSldtze([ f00d web.,. Annual Rt'vkw or [e"log) and S)'~ternaIICs 2g:2ii9· .~!6
Poli;,. G. A, 'tnd S. D Hurel. 1996. LI\)klllg l11dnne and l~rre:>lfldJ rON} ",~bs allo,:hlhoI1011, ll1pUl r:olll Ihe 0etan ,UPP0rl:-, h~gh se-conJary produ<:tlvllV oil $o1aH l')Ltncb dnd wdsi<\l blnd communitl';:' American Ndlurallsl 147:396 .., .~23.
Ro'e. 1\·1 D.. and G A. Polis 19')!!. The di,lnhuti,m ,lnd "bundancc or eoyOL~<' the clTect, or allochihonous roocl suo,idlcs rwm rhe sed [CV!c1gy 97:0')R-lOo7
RO"lllg, M. 1', M Herr·DJ.~id, and R.. P Barry i\l\l!! Analysr, nr S\able i,otorc dal".' a K nearest-neighhor" rand0lJ1113tlOl\ IC,!. JOllrnal 01' Wildlife Man,lgc!llcnL 02:3~Ll 388.
Roth. J. D. 2G03. Val"labdny (\1 m3W1~ r~"OllTt:e, ailed' ,ircllc "ox populalion <1yn<WllC' Jvurnal or Anill1,d F.c:()IOi!) 72. 66~ f>7f,
Roth . .I. D., and K A. }-[obwn. 2000. Stable caroon illh!
nJlrogeu lSOlOPI'~ frd,clionatl0n bdl,'ten dl¢1 ilnd U,>\J~ 01 (\lpIIVe i~d [o'\" impllcalwn; 1'01' Chelar) recOlhlrl'C[IO;1 CanAdian Jvurnal 01" Zook)gy 78848-·852.
Sc;hindler, D E.. P. R LeJyill. C. S Br0cL S P. ]o!lmon. IlnJ P. D. QU,lY. 2005 Marille-(krived nutrients, (Oll1ll1CITlal fislJt'I"I(os, iind prOdllctlon or ;alrrwn and bkc alg.le In Ahbk,-(. Ecology 86,3225-323!
S,ep"nski, M. 1\1, M. Ben-J)<lVld, anJ V Van Balicnberghe 1999. t\sse,srnenr ofanadr0n1l!llS salmon resources inlhe Jiet or the AleJ;andcl' Ard11pdago ..... olf u"ing stabk lsolOpe Jmllysis. Oecologia 120:327 _~35.
Urton. E. J. M., anu K A Hobson. 2()()5. lnirapoplilatwn y3J"llliion 111 gray wolf i,owpe (<ibN and S"C) [_>r'l)fiie,' IfnphcatlollS f0r Ihe e(;ology or lllu!vidLll1h. Occol0~ld i4s· 317 326.
V,1n Hallcnherghe, V I()!!'" EITclI, of pred'Hlon on rollO,,: number,. ,I reView I.'/" r(Lenl f\orth AlllCI"ICd Il 'ludle, S',ecdl,h W!ldllfe Re,~al\:h S"I'plt>ment I:,H 1--4(1)
WilhOit, 1\'1. F. S. M. Cicnc.lc. anJ P. A HlSS011 lor)4 Anadronwus tishcs ,lS ~c0J0glc,J1 Iinb b,'iw~en ocean. J"rc,lmal('!". dud lanJ. P"gc, 284-300 It! G A Pok,. M. C. 1"v',,'e1, "nd G. R HllXd. FvoJ webs III lh~ landscJpe level t'nivcrSlly (>]" Chicago, ChIcago, lillnol>, USA