Recent recovery of the Italian wolf population: a genetic investigation using microsatelLites

11 Pages·2001·6 MB·English

by  M Scandura

JAN 0 9 2002
Mammalian
Biology
Zeitschrift für Säugetierkunde
Editor IMPRESSUM
Publisher: Urban & Fischer Verlag GmbH & Co. KG,
Deutsche Gesellschaft für Säugetierkunde BranchOfficeJena,P.O.Box100537,D-07705Jena,Germany;
Phone: ++49(0)3641/626-3, Fax: ++49(0)3641/62 65 00,
(German Society of Mammalogy) E-mail:[email protected]
Advertisementbookings: Enquiriesshould bedirectedto
Urban&FischerVerlag,AdvertisingManager:SabineSchröter,
Editorial Office P.O. Box100537, D-07705Jena,Germany;
Phone:++49(0)3641/626445, Fax:++49(0)3641/626421.
D. Kruska, Kiel • P. Langer, Giessen eAfdfveecrttiviesiantgprreasteenst:.TDhiestrpirbicuetiLoinstafnrdomsuJbasncruiaprtyio1n,a2g0e0n1cyi:s
ForcustomersintheGermanspeakingcountries: Please
turn to your bookshop or to Urban & Fischer Verlag,
Advisory Board AboserviceundVertrieb: BarbaraDressler,Löbdergraben14a,
S. Halle, Jena FDa-x0:7+7+4439(J0e)n3a6,41G/e6r2ma6n4y;43,PhEo-nmeai:l:++k.4e9r(n0s)t@3u6r4b1a/n6fi2sc6h4er4.4d,e
G. B. Hartl, Kiel Forcustomersin allothercountries: Pleasecontactyour
bookshoporTheNaturePublishingGroup, LindaStill, Brunei
R. Hutterer, Bonn Road,Basingstoke,RG216XS,UK;Phone:++44(0)1256/3026
29, Fax:++44(0)1256/47 61 17, E-mail: [email protected]
E. Kalko, Ulm SubscribersintheUSAcancalltollfreeon++18007473187.
Termsofdelivery (2001):1volumeconsistingof6issues.
P. Lüps, Bern Subscriptionrate(2001):DM498.00*/GBP200.00.Preference
W. Maier, Tübingen price for personal subscribers: DM 150.00*/GBP 60.00.
Allpricesquoted aboveareexclusive postage. *Suggested
H. Schliemann, Hamburg list price. Price are subject to change without notice.
Personalsubscriptionordersmustbeaddresseddirectlytothe
S. Steinlechner, Hannover publisherUrban&FischerinJenaresp.toTheNaturePublishing
Group. Subscriptions at a personal rate mustinclude the
G. Storch, Frankfurt recipient's name and mailing address. An institutional or
Suchentrun Wien businessaddressisonlyaccepted,iftheinstituteorfirmhas
F. k, asubscriptionaswell.Personalsubscriberscanpaybypersonal
F. TriIlmich, Bielefeld ccohnevqeureteodractretdhietccuarrrdemntenetxicohnaendgeberlaotew.. USdollarpricesare
Weacceptthefollowingcreditcards:VISA/Eurocard/Master-
card/AmericanExpress.(Pleasenotecard-No.andexpirydate.)
Deutsche GesellschaftfürSäugetierkunde Bankingconnections:
For customers in the German speaking countries:
Living Past Presidents PostbankLeipzig, Konto-Nr. 149249903 (BLZ86010090).
D. Starck, Frankfurt (1957-1961, 1967-1971) FDoerutcsucshteomBearnskiJnentah,eKootnhteor-Nrc.ou3nt9r0ie7s6:5U6K(BPLosZt8O2f0f7i0ce00G0i)r;o
H. Frick, München (1972-1976) AccountNo.5192455.Fülldetailsmustaccompanythepayment.
Subscriptioninformation:Ifnotterminatedthesubscription
M. Rohrs, Hannover (1977-1981) willbeeffecteduntilrecalled.Ifno cancellationisgivenuntil
October, 31stthedeliveryofthejournalwillbecontinued.
H.-J. Kuhn, Göttingen (1982-1986) Copyright:Thearticlespublishedinthisjournalareprotected
E. Kulzer, Tübingen (1987-1991) byCopyright. Allrightsarereserved. No partofthejournal
maybereproducedinanyformwithoutthewrittenpermission
U. Schmidt Bonn (1992-1996) ofthepublisher.Thisincludesdigitalisationandanyfurther
H. G. Erkert, Tübingen (1997-2000) telreacntsrmoinsisciCoonmpouftdiinggi,talliikseesdavmiantge,ricaolpyfirnog,mptrhiinstijnoguronraeLle(cotnrloinniec
oroffline).Authorizationtophotocopyitemsforinternalor
personaluseofspecificclients,isgrantedbyUrban&Fischer,
Managing Committee forLibraries and otherusers registered with the Copyright
ClearanceCenter(CCC)TransactionReportingService,provided
President thatthebasefeeof$15.00percopyispaiddirectlytoCCC,
R. Schröpfer, Osnabrück 0212/26R6o$se1w5o.o00d/0D.rive, Danvers, MA01923, USA, 1616-5047/
Thisconsentdoesnotextendtocopyingforgeneraldistribution,
foradvertising orpromotionalpurposes, forcreating new
Board Members collectiveworks orforresaleand otherenquiries. In such
cases,specificwrittenpermissionmustbeobtainedfromthe
H. Frädrich, Berlin publisherUrban&Fischer.
J. Ganzhorn, Hamburg @TypAessettingandprinting,binding: druckhausköthenGmbH
ofvol. 61, number 1 (1996) the paperusedin
G.B. Hartl, Kiel thispublicationmeetstherequirementsofANSI/NIS0
D. Kruska, Kiel Z39.48-1992 (PermanenceofPaper).
Printed in Germany ?2001 Urban & Fischer Verlag
P. Lüps, Bern
Fordetailedjournalinformation see ourhome page:
A. Wöhrmann-Repenning, Kassel http://www.urbanfischer.de/journals/mammbiol
Mamm. biol. 66 (2001) 321-331 Mammalian Biology
© Urban & FischerVerlag
http://www.urbanfischer.de/journals/mammbiol Zeitschriftfür Säugetierkunde
Original investigation
Recent recovery of the Italian wolf population:
a genetic investigation using microsatellites
By M. Scandura, M. Apollonio and L. Mattioli
Department of ZooLogy and Anthropology, Um'versity of Sassari, Sassari, Italy and WildLife Management Unit,
ProvincialAdministrationofArezzo,Arezzo,Italy
ReceiptofMs. 20. 09. 2000
AcceptanceofMs. 31. Ol. 2001
Abstract
Genetic differentiation within the Italian wolf population was investigated by microsatellite analy-
sis of38 individuals from 4 distinct sampling sites ofthe currentwolfränge throughoutthe penin-
sula. Asetof6 microsatellite loci was used, which showed a high levelofpolymorphism and a com-
bined probability of identity ranging from 10"4 to 10"6. The overall DNA variability detected was
considerable and only slightly lower than that found for North American grey wolves.
The two largest Italian subpopulations taken into consideration, Tuscan Apennines and central-
southern Apennines, proved moderately divergent, and data are consistent with a derivation of
the western Alps subpopulation from the former, while the latter showed close similarity to the
western coast subpopulation. Gene flow was relatively high across the Italian population and the
presence ofisolation by distance was supported by our data, as measures ofgenetic distance were
consistent with geographical distribution of sampling sites. High levels of divergence were found
between Italian and other European samples. These findings suggest that, despite their absolute
mtDNA monomorphism, Italian wolves have preserved a high nuclear DNA heterogeneity and a
well-defined genetic identity. A further enlargement of ränge, which can be expected on the basis
ofextensive wolfdispersal, might canceltheir historicalisolation in a few decades, thus favouring
a genetic exchange with the east European gene pool.
Key words: Canis lupus, microsatellites, variability, population structure, Italy
Introduction
Since the end of the 19th Century large pre- have led to the restoration of more favour-
dator populations have declined in Italy able environmental conditions and to in-
due to progressive habitat disruption and creased protection of these species. Due to
to direct persecution by humans. As a con- such improvements, in the last 30 years an
sequence, different species approached ex- important predator species, the wolf (Canis
tinction (i.e., brown bear, wolf, and lynx). lupus), has increased in number and en-
Changes in human activities, wildlife and larged its ränge (Francisci and Guberti
wood management, and public opinion 1993; Boitani and Ciucci 1993; Meriggi
1616-5047/01/66/06-321 $15.00/0.
322 M. Scanduraetal.
and Lovari 1996). Wolves in the Italian en- ent studies carried out on North American
vironment play a key role in wild commu- wolf populations demonstrated the effec-
nities, being the only well-distributed large tiveness ofmicrosatellite loci as amolecular
mammals preying mostly on wild ungulates tool for assessing population structure para-
(Mattioli et al. 1995). Düring the last two meters (Roy et al. 1994; Forbes and Boyd
centuries the Italianpopulation became iso- 1997), genotyping animals for reintroduc-
latedfrom otherEuropeanpopulations, due tion programmes (Garcia-Moreno et al.
to the extirpation of the species throughout 1996), evaluating relatedness among indivi-
the Alps (Cagnolaro et al. 1974). The duals (Smith et al. 1996), and estimating ge-
northern border ofits ränge initially moved netic Variation following natural coloniza-
southwards, towards the central regions, tions (Forbes and Boyd 1996).
and the presence was restricted to the less The aims of the present study were: 1) to
accessible areas of the Apennines and to a reconstruct the dynamics ofthe Italian wolf
wooded area along the Tyrrhenian coast recovery by comparing the genetic pattern
(Cagnolaro et al. 1974; Zimen and Boitani of different subpopulations; 2) to evaluate
1975). In 1973, the number of wolves inha- nuclear DNA diversity among and within
biting the Italian peninsula was estimated sub-populations; 3) to estimate the degree
to be approximately 100 individuals (Zimen of gene flow among different areas. Wolves
and Boitani 1975), after which the popula- fromhistorical "stronghold areas" andfrom
tion recovered, reaching an estimated size recent colonized regions were sampled for
of400-500 individuals (Boitani 1992). comparison.
The demographic recovery of the species
was accompaniedby anorthward expansion
ofits ränge, which during the last ten years Material and methods
led to the recolonization of the western
Alps, and to the consequent appearance of Sample description and collection
the wolfin France (Breitenmoser 1998).
The effects of population decline and the Italian wolfsamples were collected from four re-
subsequent ränge expansion in the genetic AgiRons-(TFiugs.c1a)n: Apennines, which probably repre-
diversity ofthe Italianwolfpopulationwere sented the northern border of the Italian wolf
studied by allozyme and mitochondrial ränge along the Apennines for half a Century
DNA (mtDNA) analyses (Randi et al. (Cagnolaroet al. 1974);
1993; Wayne et al. 1992; Randi et al. 1995; AB - Central-southern Apennines, the part of
Vilä et al. 1999; Randi et al. 2000). Randi Italy where the species has always been present
and co-workers (1993), using a set of 40 al- in historictimes;
lozymes over a sample of 38 wolves, found VC - Alta Maremma, where the presence ofthe
a level ofpolymorphism and heterozygosity species was always recorded during the last Cen-
comparable to that of larger North Ameri- tnueryw,bSuettttlheemeanrtesawoafsbcrheaerdaicntgerpiazcekdsbyfoclolnotwienduobuys
can populations (Kennedy et al. 1991). On complete orpartialeradication (illegalkilling);
the other hand, mtDNA consensus se- FR - Alpes Maritimes, France, originating in the
quences revealed the presence of a single early 1990s from individuals moving across from
haplotype in all the sampled Italian wolves. Italy (Randi et al. 2000). Its present size is ap-
This apparent contradiction probably re- proximately20-30units, ofwhich dispersingindi-
sults from the different inheritance Systems, viduals are colonizing new areas of the western
based exclusively on female mtDNA trans- Alps.
Seventeen tissue samples came from illegally
mission.
killed wolves, recovered before 1992 in central
In order to investigate the actual effects of
and southern Italy and obtained from Prof.G.B.
the population bottleneck and fragmenta- Hartl (University of Kiel, Germany). Thirteen
tion on the genetic structure and variability specimens are from the northern Apennines,
ofwolves, nuclear genetic markers, e.g. mi- mostly the province of Arezzo: 12 (3 tissues and
crosatellites, are more informative. Differ- 9 hairs). supplied by the Provincial Administra-
RecentrecoveryoftheItalian wolfpopulation 323
'SL
FR
SP
<0
500km
Fig. 1. GeographicalLocationofsamplingsites (solidgrey) and presentdistribution rängeofwolves (hatching).
tion of Arezzo (Tuscany) or by the Corpo Fore- et al. 1996). Amplifications were carried out in
mM mM
stale dello Stato, derived from animals dying in 20ulvolume, containing 10 Tris-HCl, 50
wthaespceorliloecdte1d99i1n-1t9he99f,iewlhdedrueraisngonaesuhravierysianmptlhee eKaCclh, d2NmTPM, 0M.5gCulni2t,s0T.a1q%DTNriAtonPoXl-y1m0e0r,as2e00(PuroM-
Foreste CasentinesiNationalPark.Threesamples mega), 5pmol ofeachprimerand \-Aul ofDNA
of the Alta Maremma population were obtained Solution. An initial denaturation step at 94°Cfor
from the Veterinary Service of Volterra (Pisa) 3min was followed by 35 cycles of amplification
and one hair sample was collected in a natural each of 45 sec at 92°C, 45sec at the annealing
preserve near Volterra. For the Alpine popula- temperature (55-58°C) and 60 sec at 72°C. The
tion, 4 tissue samples were provided by Valiere reactionterminatedwithapolymerizationstep at
Nathaniel (University of Grenoble, France). In 72°Cfor5min.
addition, ten wolf samples from Asturias, Spain Inordertoverifythe successfulproduetsofsingle
(SP) andthree from Slovakia (SL) were analysed PCRs, 5 ul ofreactionSolutionwere run on a2%
forcomparison. agarose gel (Biorad) containing ethidium bro-
mide, and the presence of correctly amplified
fragmentswasdetectedbycomparingtheirlength
DNA
DNAIsolation and amplification with a sizemarker.
Single alleles were sized by running denatured
Proteinase K digestion and phenol/chloroform PCR produets through capillary electrophoresis
Standard protocols were used for genomic DNA in an ABIPRISM310 automatic sequencer (Per-
isolation from tissues. Nuclear DNA was ex- kin-Elmer).
tracted from hairbulbs eitheraccordingto Higu-
chi et al. (1988) or by Chelex isolation (Walsh
et al. 1991). Samples were genotyped for five di- Statisticalanalysis
nucleotide (AC)n Polymorphie microsatellite loci
and one tetranucleotide locus, previously charac- By combining alleles at each locus, individual
terizedin dog (Ostranderet al. 1993;Francisco genotypes were obtained. The GENEPOP soft-
324 M. Scanduraetal.
wäre programVersion 3.2a (Raymond and Rous- where the second term represents the mean PAS
set 1995) was usedto calculate allele frequencies calculated over all combinations between the z'th
and to test data sets for deviation from Hardy- and the ;'th subpopulation genotypes. Mean dis-
Weinbergequilibrium(HWE) aswellasforgeno- tance values were also computed among indivi-
typic linkage disequilibrium. Because many rare duals of a single sample, in order to evaluate in-
HWE
alleles were present, departures were also tra-group homogeneity. To eliminate the bias
tested "with pooling" (Hartl and Clark 1989), originating from different degrees of sample
by grouping genotypes into three classes: homo- homogeneity, mostly due to different breadths of
zygotesforthemostcommonallele,common/rare sampling areas, a new matrix was extrapolated,
allele heterozygotes, and all othergenotypes. averaging the differences between inter- and in-
Observed heterozygosity (HQ) and unbiased ex- tra-populationdistance values:
mpeactteeddfhoerteearoczhysguobspiotpyul(aHtei)on(.NePiro1b9a7b8i)liwtyeroefgesetni-- j^, _ (Dasij -Dasa) + (Dasg -Dasg)
otypeidentitywas obtainedusingthe formula
Furthermore, Nei's unbiased genetic distance
(Nei 1972) was computed by BIOSYS-2 Software
/' i j>i (Swofford and Selander 1997) between all sub-
wherep; andpj are the frequencies ofthe z'th and populationpairs.
;'thallelesatagivenlocus (PaetkauandStrobeck Multilocus F-statistic was calculated by GENE-
1994). Single locus probabilities were combined POP, estimating the FST coefficient for each pair
to obtain the total probability over all 6 loci, as- of samples and for all Italian samples, according
suming independence of different loci, as sup- to Weir and Cockerham (1984). Inorderto eval-
ported by the microsatellite linkage map in the uate gene flow among subpoputations, the same
domestic dog (Mellersh et al. 1997). program allowed the effective number of mi-
InordertoevaluatethelevelofgeneticVariation, grants per generation (Nm) to be estimated on
H
the e estimated for the overall Italian popula- the basis of the private allele model (Slatkin
tion (N=38) over five of the six examined loci 1985; Barton and Slatkin 1986). Thereafter a
(109, 123, 204, 250, and 377) was compared with Statistical analysis was carried out using the
values recalculated for North American popula- ISOLDE program, in the GENEPOP package,
tions over the same loci using published data performingMantel's tests (1000permutations) to
(Roy et al. 1994;Forbes andBoyd 1997). highlightthepossiblepresenceofisolationbydis-
Allelic andgenotypicdifferentiationswere evalu- tance in the Italian population. For this purpose,
raätendgefo(rAeRa,chABp,opuVlCa,tioanndpaFiRr)w,itahnidn tthheenItwaleiraen DasASme,Naesiu'rseusnbofiagseendetdiicstdaincvee,rgaenndceFSaxnwderceomcphaorseedn
pooled and compared with the two other Eu- withgeographic distance.
ropean populations (SP and SL). Two different
approacheswereusedtoestimatethelevelofdif-
ferentiation between samples by GENEPOP: a
Fisherexacttestwasperformedtotestthe homo- Results
geneity of allelic distributions across populations
(Raymond and Rousset 1995), whereas a log- A total of 51 wolves was genotyped at six
likelihood(G)basedexacttestwasusedforgeno- microsatellite loci. Allthe loci showedpoly-
typic differentiation (Goudet et al. 1996). The morphisminthefourItaliansubpopulations
significance level was always established using investigated, except for locus 377 in the Al-
Bonferroni's criterion for multiple tests. In both pine samplewhere alleleAwasfixed. Allele
cases, anunbiasedestimateofthep-valuewasob- H
tained, associated with the null hypothesis of frequencies are givenintable 1. Average e
identical distribution acrosspopulations. ranged from 0.505 ±0.106 to 0.680±0.038,
Amatrixwascreatedcontainingtheproportionsof while the probability ofidentityvariedfrom
shared alleles (Pas), over the six loci, for all pair- 1/1700 for the Alpine sample to less than 1/
wise comparisons of sampled individuals, as de- 100000 for the central-southern Apennines
scribedinBowcocketal.(1994).Inordertoobtain subpopulation (Tab. 2). In three out of four
a measure of divergence among populations, PAs Italian samples, average He had alowerval-
vtahelupeasirwweirseedaivstearnacgeedvalouveerDApoSpwualsatciaolncuplaaitresdaasnd udeuetthoanlitmhieteodbosuetrbvreededoinneg.(NTeabg.at3i),vepoFssivbally-
is
(1-PaS,;) ues confirm this possibility (meanFis over 6
Recentrecoveryofthe ItaLian wolfpopulation 325
Table 1. Allele frequency distributions at 6 microsatellite loci in wolf samples (AR - Tuscan Apennines; AB -
Central-southernApennines;VC-Alta Maremma: FR-AlpesMaritimes:SP-Spain;SL-Slovakia).
AR AB VC FR SP SL
1nnic 1HO
A U.11D nu.nununu nu.nununu nu.nununu Un.11nUnU nU.nUnUnU
DD n ^3R U.Dlo nU.7/cD;Un nU.^DnUnU nU.j3D^Un nJ.cD^UnUn
rL n n3R n n^Q nKJ.11l?.Z^j nu.nununu nu.nununu nU.nUnUnU
nu Un.nU3jRO U.1/U nu.nununu nU.J371^D; nU.'Afnunu nu.nununu
p nU.1imA-7/ nU.1lL?cJ; U.1C.D un.nununu nu.nununu
pr nu.nununu un.nununu nu.nununu nu.nununu nu.nununu nU.^DnUnU
G 0.000 0.000 0.000 0.000 0.150 0.000
Inmc.1?3
nA n ?An n 3^3 nu.nununu nu.nununu nu.nununu nu.nununu
RD 0.000 0.059 0.125 0.125 0.000 0.000
r U.1UU 0.206 0.375 0.000 0.278 0.667
nu nu.nuRonu nu.nununu 0.000 0.250 nU.nU^DfUi nu.nununu
p n nnn n n?Q nu.nununu nU.C?.^DnU n 3RQ nU.3J3J3J
pr 0.520 0.324 0.500 0.375 0.167 0.000
G 0.000 0.000 0.000 0.000 0.111 0.000
H 0.000 0.029 0.000 0.000 0.000 0.000
nn ?04
I i<;
rA\ 0.000 0.000 0.000 0.000 0.350 0.000
RD Un.nU3JRO 0.000 0.000 0.000 0.250 1.000
rV- Un.3JnURO nU.J37/^D nU.J37/^D nu.o3unun nu.nununu
n 0.000 0.088 0.000 0.000 0.000 0.000
pL 0.654 0.382 0.500 0.625 0.050 0.000
nu.nununu 0.000 0.000 0.000 0.050 0.000
H 0.000 0.000 0.125 0.000 0.000 0.000
Locus250
A 0.077 0.206 0.000 0.286 0.500 0.667
B 0.115 0.147 0.375 0.286 0.150 0.000
C 0.000 0.029 0.000 0.000 0.000 0.000
D 0.692 0.324 0.500 0.429 0.150 0.000
E 0.000 0.000 0.000 0.000 0.000 0.333
F 0.077 0.294 0.125 0.000 0.150 0.000
G 0.000 0.000 0.000 0.000 0.050 0.000
H 0.038 0.000 0.000 0.000 0.000 0.000
Locus377
rA\ nU.RÖA4f0i n A71 Un.RDnUnU 11.nununu nU.nU^DnU nU.3J3J3J
pu nU.1111Dc; Un.11i/fu\ Un.^3nUnU 0.000 0.050 0.000
p 0.038 0.088 0.000 0.000 0.200 0.167
>1J 0.000 0.000 0.000 0.000 0.000 0.333
K nu.nununu nu.nuRoRo 0.000 0.000 0.000 0.000
NIN nu.nununu nu.nununu nu.nununu 0.000 0.050 0.000
nu nu.nununu nU.1J.7/fDi nu.nununu nu.nununu nu.Atnunu Un.11Ufi/7
p nu.nununu nu.nununu 0.000 0.000 0.050 0.000
nu.nununu nu.nununu 0.000 0.000 0.200 0.000
Locus2158
A 0.000 0.059 0.000 0.000 0.000 0.000
nD 0.000 0.000 0.000 0.000 0.000 0.167
c 0.042 0.059 0.000 0.000 0.000 0.000
D 0.458 0.324 0.000 0.500 0.000 0.667
E 0.000 0.000 0.000 0.000 0.444 0.000
F 0.000 0.088 0.250 0.000 0.111 0.000
G 0.083 0.265 0.375 0.375 0.000 0.000
H 0.083 0.000 0.000 0.000 0.000 0.167
I 0.333 0.147 0.375 0.000 0.000 0.000
J 0.000 0.059 0.000 0.000 0.000 0.000
L 0.000 0.000 0.000 0.125 0.444 0.000
326 M. Scanduraetal.
Table 2. Expected heterozygosity (number ofalleles in parentheses) and probability ofidentity in Italian wolf
samples (AR-TuscanApennines;AB-CentraL-southernApennines;VC-Alta Maremma; FR-AlpesMaritimes).
Heterozygosity ProbabilityofIdentity
Locus AR AB VC FR AR AB VC FR
109 0.621 (5) 0.562 (4) 0.406 (3) 0.594 (3) 0.197 0.236 0.388 0.248
123 0.640 (4) 0.723 (6) 0.594(3) 0.719 (4) 0.182 0.125 0.248 0.130
204 0.476 (3) 0.566 (3) 0.594(3) 0.469 (2) 0.357 0.277 0.248 0.392
250 0.494(5) 0.744(5) 0.594(3) 0.653 (3) 0.282 0.110 0.248 0.194
377 0.269 (3) 0.701 (5) 0.500 (2) 0.000 (1) 0.555 0.128 0.375 1.000
2158 0.663 (5) 0.785 (7) 0.656 (3) 0.594 (3) 0.170 0.076 0.193 0.248
AULoci 0.527 0.680 0.557 0.505 3.4xlO"4 8.7xlO"6 4.3xlO"4 6.1xlO"4
Table3. Genetic Variation at 6 microsatellite loci and deviation from Hardy-Weinberg equilibrium. N, sample
size; A, mean number ofalleles per locus; H0, observed heterozygosity (direct count); He, Hardy-Weinberg ex-
pected heterozygosity; HWE, significance of chi-square test for Hardy-Weinberg equilibrium without pooling;
HWEp, significance ofchi-square testfor Hardy-Weinberg equilibrium with pooling; SE, Standard error; n.s., not
significant.
Subpopulation/Population N A(SE) H0 He(SE) HWE HWEp
AR-TuscanApennines 13 4.0 (0.4) 0.579 0.527 (0.061) n.s. n.s.
AB -Central-southern Apennines 17 5.0 (0.6) 0.676 0.680 (0.038) n.s. n.s.
VC-Alta Maremma 4 2.8 (0.2) 0.792 0.557 (0.036) n.s. n.s.
FR-Alpes Maritimes 4 2.7 (0.4) 0.652 0.505 (0.106) n.s. n.s.
Total(Italian population) 38 6.0 (0.7) 0.664 0.644(0.040) n.s. n.s.
SP-Spain 10 4.8 (0.5) 0.683 0.693 (0.023) n.s. n.s.
SL-Slovakia 3 2.3 (0.4) 0.389 0.435 (0.097) n.s. n.s.
Table4. Comparison ofgenetic Variation at 5 microsatellite loci among different wolf populations and related
canid species (N, samplesize;A, mean numberofalleles perlocus±Standarderror; He, Hardy-Weinbergexpected
heterozygosity±Standarderror).
a recomputedfromSinglelocusfrequenciesdata.
Species/Population N A He Reference
Ca/775lupus
Italy 38 5.4±0.4 0.619±0.039 thisstudy
Spain (Asturias) 10 5.2±0.5 0.713±0.013 thisstudy
Canada (NorthwestTerriton'es) 24 6.0±1.2 0.714±0.063 Rovetal. (1994)a
Canada (Alberta) 20 5.2±0.6 0.709±0.027 ROYetal. (1994)a
USA (Montana) 66 5.0±0.4 0.659±0.055 Forbesand Boyd (1997)a
USA(Yellowstone NationalPark) 31 5.4±0.7 0.686±0.074 Forbesand Boyd (1997)a
Canissimensis 42 2.6±0.7 0.167±0.011 Gottellietal. (1994)a
Canislupusf.familiaris 40 6.8±0.8 0.714±0.075 Gottellietal. (1994)a
lociforthe overall Italianpopulationequals (FR) and a strongly fluctuating population
-0.039), indicating breeding among non-re- (VC) showed a marked excess of heterozy-
latives. Departures from the Hardy-Wein- gotes, although statistically not significant.
berg equilibrium for the different samples Linkage disequilibrium for each pair ofloci
proved not significant at all loci. However, was confirmed by a probability test analysis
as shown in table 3, a recent colonized area in GENEPOP.
RecentrecoveryoftheItaLian wolfpopulation 327
Referring to the restricted analysis (over 5 FSt values accounted for the proportion of
loci), mean He for the Italian sample was total Variation due to diversity between
0.619±0.039, with a mean number of al- samples. Overall, FSt for the whole Italian
A
leles per locus (A) of 5.4±0.4. compari- population was 0.053, a low value consider-
son with other wolf populations and with ing that a value of zero expresses the iden-
two canid species (Ethiopian wolf, Canissi-
mensis, and domestic dog, Canis lupus f.fa-
miliaris) is shown in table4.
Levels of differentiation among subpopula-
tions and populations were detected by
comparing allelic and genotypic frequency
distributions across loci. Significant levels
of divergence within the Italian ränge were
obtained only from the comparison be-
tween AR and AB samples (allelic data:
Fisher exact test, p = 0.00044; genotypic
data: G-test p = 0.00042). On the other
hand, the whole Italian population showed
both a genic and a genotypic statistically
significant divergence from the Spanish Fig. 2. Upgma phenogram of wolf populations based
on Nei'sunbiasedgeneticdistances.
and the Slovakian samples (allelic data:
Fisher exact test, p<0.001; genotypic data:
Table 5. Pairwise genetic distances by Shared Alleles
G-test p<0.001). Single-locus comparisons
(Das)
allow discrimination of subpopulations due
AR
to the presence of private alleles. Both Subpopulation AR AB VC
and AB subpopulations showed exclusive
AR-TuscanApennines
alleles at different loci. All the alleles but
H AB-Central-southern 0.049
one (allele at locus 204) present in the
ValCso.sOamnpltehebeolthoenrghtoantdh,etAhBe FsRubspoapmupllaetihoans AVFCpRe--nnAAillntpaeessMMaarreimtmimaes 00..006720 00..006819 0.162
alleles present in both AR and AB popula-
tions, except for three alleles at locus 123,
otinveelye,xcalnudsifvoertaollAelRe Lanadttlwoocutso2A1B58,,raebsspeenct- TIotnaaabllli)eana6.nsduNbeppiao'ipsruwulinasbteiiaoFsnSTes-dv(agal)euneeastnid(cabadomivsoetnagdnicaeEguo(rnboaeplle)oawanmdwioaonglg-f
in other Italian samples. Mean DAS (Pro- populations (b).
portion ofalleles not shared) within the Ita-
a)
lian population was 0.466, whereas the
mean values of derived Das f°r subpopula- Subpopulation AR AB VC FR
tion pairs ranged between 0.049 and 0.162
AR-TuscanApen- 0.058 0.071 0.051
(Tab. 5). nines
Nei's unbiased distances were lower than AB-Central-southern 0.094 0.022 0.042
0.2 for all the comparisons among Italian Apennines
samples (Tab. 6a, below diagonal), and VC-Alta Maremma 0.081 0.051 0.121
higher than 0.6 for all inter-population pair- FR-Alpes Maritimes 0.054 0.093 0.162
wise comparisons (Tab. 6b, below diago-
nal). In the former case, the minimum va- b)
lues were obtained between AB and VC
samples (0.051) and between AR and FR Population IT SP SL
samples (0.054). On the basis of Nei's un- IT-Italy 0.199 0.262
biased distance, a cophenetic tree may be SP-Spain 0.764 0.181
plotted (Fig. 2). SL-Slovakia 0.851 0.610
328 M.Scanduraetal.
tity of allele frequencies among all subpo- High Hardy-Weinberg expected heterozyg-
pulations. FStvalues obtained for each sub- osity, in comparison with the one observed
population and population pair are shown in Italian samples, may arise from limited
in table6a and 6b, respectively (above di- outbreeding. as confirmed by the negative
agonal). The multilocus estimate of Nm for F value. Breeding between unrelated indi-
Ls
the overall Italian population gave a num- viduals is a common trend in natural wolf
ber of about 2.0 migrants per generation. populations (Smith et al. 1996). neverthe-
which suggests a relevant amount of gene less high levels of induced mortality may
flow among subpopulations. When calculat- enhance the natural turnover ofpack mem-
ing Nm between the two most representa- bers and favour outbreeding.
tive subpopulations, AR and AB, a value The most immediate indicator of genetic
of 1.7 was obtained. differentiation is allele frequency distribu-
A positive correlation was found, using tion. A significant level of divergence
Mantel's test, only between D'AS and geo- among Italian samples was found over all
graphical distance (Spearman rank correla- 6loci only for the AB-AR pair. comparing
tion coefficient, p = 0.038), suggesting the both allelic and genotypic frequencies. As
presence of moderate isolation by distance. microsatellites are particularly sensitive to
No significant correlation was found for allele frequency differentiation, such differ-
Nei's distance and FSt- ences may not be considered on their own
a proof ofgenetic isolation.
Looking at allele frequency distribution.
Discussion generally. the highest was the frequency
across populations, and the widest was spa-
Italian population samples showed a high tial diffusion. Several rare alleles were pre-
intra-group diversity. as both He and Pid sent. with occasional local specificity.
were relevant within each subpopulation. Examining the presence of Single alleles,
Although the Standard errorwas sometimes AR and AB samples showed a moderate
considerable, due to the small sample size, level ofdiversity due to the presence ofpri-
the Awas high for most loci even in small vate alleles (5 for AR and 10 for AB). VC
samples. Mean heterozygosity over 5 loci and FR samples appear compatible with a
proved very close to the values obtained possible derivation from the other two sub-
for North American populations (Roy et populations, but they also have exclusive al-
al. 1994; Forbes and Boyd 1997). and also leles. The allele H at locus 204. present in
the mean number of alleles per locus was the VC sample, was never found before in
completely comparable. This agrees with al- wolfindividuals. whereas it proved the pre-
lozymic data, whose level ofheterozygosity valent allele in dog samples (data not
A
for the Italian population was found to be shown). possible wolf dog hybridization
relatively high (Randi et al. 1993). event among the ancestors ofthe female in-
Comparing Italian wolf with a related spe- dividualpresenting such an allele cannot be
cies population. Ethiopian wolf (Gottelli excluded. On the contrary, allele L at locus
et al. 1994), which went through prolonged 2158 in the Alpine sample was found in
isolation, avast difference inheterozygosity other European wolves (e.g. in Spain) and
calculated over the same loci is evident. might be present also at low frequency in
All samples fitted Hardy-Weinberg expec- the Italian population. but it was not de-
tations, whether the x2-test was performed tected in this work as a consequence ofthe
with or without pooling. Excess of hetero- limited sample size.
zygotes in strongly fluctuating or recently Both DAS and Nei's unbiased distance val-
colonized subpopulations. VC and FR, ues confirm the expected origin of the Al-
proved high but not significant. This may pine subpopulation from the northern
be due to random assembling of founder Apennines. The allelic pattern of the VC
genotypes occupying new territories. sample, combined with measures of dis-