ArticlePDF Available

Randi E, Alves PC, Carranza J, Miloevi-Zlatonovi S, Sfougaris A, Mucci N. Phylogeography of roe deer (Capreolus capreolus) populations: the effects of historical genetic subdivisions and recent nonequilibrium dynamics. Mol Ecol 13: 3071-3083

Authors:

Abstract and Figures

We sequenced 704 mitochondrial DNA (mtDNA) control-region nucleotides and genotyped 11 autosomal microsatellites (STR) in 617 European roe deer (Capreolus capreolus) samples, aiming to infer the species' phylogeographical structure. The mtDNA sequences were split in three distinct haplogroups, respectively, named: Clade West, sampled mainly in Iberia; Clade East, sampled mainly in Greece and in the Balkans; and Clade Central, which was widespread throughout Europe, including the eastern countries and Iberia, but not Greece. These clades might have originated in distinct Iberian and Balkanic refuges during the penultimate or the last glaciations. Clades East and West contributed little to the current postglacial mtDNA diversity in central Europe, which apparently was recolonized mainly by haplotypes belonging to Clade Central. A unique subclade within Clade Central grouped all the haplotypes sampled from populations of the Italian subspecies C. c. italicus. In contrast, haplotypes sampled in central and southern Spain joined both Clade Central and Clade West, suggesting that subspecies C. c. garganta has admixed origin. STR data support a genetic distinction of peripheral populations in north Iberia and southern Italy, and show the effects of anthropogenic disturbance in fragmented populations, which were recently reintroduced or restocked and not may be in mutation-drift equilibrium. Roe deer in central Europe are mainly admixed, while peripheral populations in north Portugal, the southern Italian Apennines and Greece represent the remains of refugial populations and should be managed accordingly.
Content may be subject to copyright.
Molecular Ecology (2004)
13
, 3071– 3083 doi: 10.1111/j.1365-294X.2004.02279.x
© 2004 Blackwell Publishing Ltd
Blackwell Publishing, Ltd.
Phylogeography of roe deer (
Capreolus capreolus
)
populations: the effects of historical genetic subdivisions
and recent nonequilibrium dynamics
E. RANDI,
*
P. C. ALVES,
J. CARRANZA,
S. MILO
S
EVI
C
-ZLATANOVI
C
,
§
A. SFOUGARIS
and N. MUCCI
*
*
Istituto Nazionale per la Fauna Selvatica (INFS), Ozzano Emilia (BO), Italy,
CIBIO/UP, Centro de Investigação em Biodiversidade
e Recursos Genéticos Universidade do Porto, and Faculdade de Ciências do Porto, Portugal,
Cátedra de Biología y Etología, Facultad
de Veterinaria, Universidad de Extremadura, Cáceres, Spain,
§
Institute of Biology, Faculty of Science, University of Kragujevac,
Kragujevac, Serbia and Montenegro,
Department of Agriculture, Crop Production and Rural Environment, University of Thessaly,
Volos, Greece
Abstract
We sequenced 704 mitochondrial DNA (mtDNA) control-region nucleotides and geno-
typed 11 autosomal microsatellites (STR) in 617 European roe deer (
Capreolus capreolus
)
samples, aiming to infer the species’ phylogeographical structure. The mtDNA sequences
were split in three distinct haplogroups, respectively, named: Clade West, sampled mainly
in Iberia; Clade East, sampled mainly in Greece and in the Balkans; and Clade Central,
which was widespread throughout Europe, including the eastern countries and Iberia, but
not Greece. These clades might have originated in distinct Iberian and Balkanic refuges
during the penultimate or the last glaciations. Clades East and West contributed little to the
current postglacial mtDNA diversity in central Europe, which apparently was recolonized
mainly by haplotypes belonging to Clade Central. A unique subclade within Clade Central
grouped all the haplotypes sampled from populations of the Italian subspecies
C. c. italicus
.
In contrast, haplotypes sampled in central and southern Spain joined both Clade Central
and Clade West, suggesting that subspecies
C. c. garganta
has admixed origin. STR data
support a genetic distinction of peripheral populations in north Iberia and southern Italy,
and show the effects of anthropogenic disturbance in fragmented populations, which were
recently reintroduced or restocked and not may be in mutation–drift equilibrium. Roe deer
in central Europe are mainly admixed, while peripheral populations in north Portugal, the
southern Italian Apennines and Greece represent the remains of refugial populations and
should be managed accordingly.
Keywords
:
Capreolus capreolus
, European roe deer, glacial refuges, microsatellites, mitochondrial
DNA control region, phylogeography, postglacial recolonization
Received 13 February 2004; revision received 4 June 2004; accepted 4 June 2004
Introduction
Natural factors have shaped the main genetic subdivisions
in animal species with widespread distributions in Europe
during Quaternary climate changes (Hewitt 2000). More
recently, deforestation, the spread of agriculture, hunting
and other kinds of human-induced disturbances have
deeply affected the size, structure and dynamics of natural
populations (Maehr
et al
. 2001; Harris
et al
. 2002). This is
especially apparent in species of large mammals, which
experienced dramatic fluctuations during the last few cen-
turies, and that are either fragmented in threatened declining
populations (i.e. many large carnivores; Breitenmoser 1998),
or are very common, but intensively managed for hunting
purposes (i.e. ungulates; Rhodes & Smith 1992). Genetic
equilibrium cannot be attained over short time scales
in fluctuating populations, and nonequilibrium factors
could affect the estimates of genetic diversity, gene flow, or
Correspondence: Ettore Randi. Istituto Nazionale per la Fauna
Selvatica, Via Cà Fornacetta 9, 40064 Ozzano Emilia (BO), Italy.
Fax: + 39 051796 628; E-mail: met0217@iperbole.bo.it
3072
E. RANDI
ET AL.
© 2004 Blackwell Publishing Ltd,
Molecular Ecology
, 13, 3071– 3083
divergence (Hey & Machado 2003). Understanding the
dynamics of geographical structuring in perturbed popu-
lations is important for evolutionary and conservation
biology, but it is challenging, because equilibrium and
nonequilibrium factors are not easily disentangled.
The roe deer (
Capreolus
, Artiodactyla, Cervidae) includes
two species, the smaller European
C. capreolus
(distributed
in western Europe), and the larger Siberian
C. pygargus
(distributed in Asia and eastern Europe) roe deer (Danilkin
1996; Randi
et al
. 1998). The European roe deer is widespread
across the continent (Fig. 1), with populations distributed
in former Mediterranean glacial refuges (Sierra de Cadiz in
southern Spain, southern Apennines in Italy, southern
Balkans), and in northern regions (Scandinavia) or mountain
ranges (the Alps and Pyrenees) that were glaciated until the
Holocene (
c
. 10 000 years ago). Roe deer are highly adaptable,
living in broad-leaved forests, ecotonal strips and agri-
cultural areas, in mountains or in lowland regions (Danilkin
1996). However, in Scandinavia the roe deer is absent beyond
the Arctic circle, suggesting that habitats where there is
permafrost could limit its northward distribution. Hence,
predictably, cyclic expansion of permafrost and the con-
comitant contraction of forests in central Europe, could
have forced roe deer populations to retreat to southern
refuges, while deglaciations and the expansion of forests during
the Holocene have certainly fostered roe deer’s coloniza-
tion of the Alps, Pyrenees, central Europe and Scandinavia.
Deforestation and over-hunting led to decline and
eradication of local roe deer populations, particularly in
central and south Iberia (Boutin 1990; Aragón
et al
. 1995a), in
the western Italian Alps and Apennines (Perco & Calò
1994), and in Greece (Adamakopoulos
et al
. 1991). After
World War II some populations began to expand naturally,
following the expansion of forests in mountains, or colon-
izing agricultural lowlands (Tellería & Virgós 1997; Albanis
et al
. 2000). Population decline was successfully hindered
by reintroductions, which were often carried out using
nonindigenous roe deer. All extant populations in the
western Italian Alps and many populations in the northern
Apennines were reintroduced or restocked in the last few
decades (Perco & Calò 1994).
The roe deer is massively hunted and managed in all
European countries, except in Greece and Portugal. A pre-
requisite for adequate conservation and management plans
is the correct definition of the taxonomic rank of populations
(Moritz 1994). However, subspecies distinction in roe deer
is uncertain. Nowadays
C. capreolus
is considered a monotypic
species (Danilkin 1996), although two subspecies were
described in the past: the Italian roe deer
C. c. italicus
(Festa
1925), distributed in southern Italy, and the Spanish roe
deer
C. c. garganta
(Meunier 1983), from central–southern
Spain. The identity of the Spanish subspecies is doubtful
(Aragón
et al
. 1995b), but Lorenzini
et al
. (2003) suggested
a genetic differentiation between Andalusian and northern
roe deer populations in Spain. Populations of
C. c. italicus
showed unique mitochondrial DNA (mtDNA) haplotypes
(Randi
et al
. 1998; Vernesi
et al
. 2002), distinctive micro-
satellite genotypes (Randi & Mucci 2001; Lorenzini
et al
.
2002), and skull morphometry (Montanaro
et al
. 2003).
Patterns of geographical diversification in roe deer are
poorly known. Published research used limited geographical
sampling, and maternally inherited mtDNA sequences
(Randi
et al
. 1998; Wiehler & Tiedemann 1998; Vernesi
et al
.
2002). In this study, we have investigated mtDNA and
simple tandem repeats (STR) genetic diversity in roe deer
populations across the species distribution range, including
the nominate species and both subspecies. We aimed to:
(i) assess whether maternal mtDNA and biparental STR
markers allow description of concordant population sub-
divisions and processes, and identification of extant genetic
units; and (ii) infer the consequences of natural and anthro-
pogenic factors in shaping the observed population structure.
Materials and methods
Sample collection
A total of 773 tissue samples were collected and stored
at
20
°
C in 95% ethanol. They were taken from 44 local-
ities in Europe (Fig. 1a), including natural, reintroduced or
restocked populations (Table 1). All of these samples
were used in phylogenetic analyses. Smaller sample sizes
in population genetic analyses were avoided by pooling
samples (which were not significantly differentiated) in
three distinct groups: north Spain (Galicia and Asturias),
Germany (Westerwald and Bayreuth) and south Italy
(Castelporziano, Gargano and Orsomarso). Samples from
central (Madrid region,
n
= 6) and south Spain (Cadiz;
n
= 5), from Novi Be
c
ej — Novo Milo
S
evo (Serbia;
n
= 6),
and a population that was reintroduced from Slovenia
in Val Susa (Italy, Torino;
n
= 3) were not used. Thus, in
population genetic analyses there were 36 distinct groups,
each one with
n
10 (Table 2). Sample no. 42 was collected
from the original location of the Italian roe deer subspecies
Capreolus capreolus italicus
(Castelporziano, a Mediterranean
forest close to Rome). Samples no. 43 and no. 44 were col-
lected from the other two Italian roe deer populations surviv-
ing in protected areas in southern Italy (Parco Nazionale
del Gargano, Monti dell’Orsomarso; Fig. 1). Throughout
this paper this group will be known as ‘
C. c. italicus
’. Samples
no. 4 and no. 5 were collected within the distribution range
of
C. c. garganta
in central (Madrid region) and southern
(Cadiz) Spain.
Laboratory methods
Total DNA was extracted following Gerloff
et al
. (1995).
The entire mtDNA control-region was successfully amplified
PHYLOGEOGRAPHY OF ROE DEER POPULATIONS
3073
© 2004 Blackwell Publishing Ltd,
Molecular Ecology
, 13, 3071– 3083
Fig. 1 (a) Distribution of roe deer (Capreolus
capreolus) in Europe and sampling locations.
The continuous black lines indicate the
approximate southern limits of the main
distribution range of roe deer. Fragmented
populations in Iberia and southern Italy
are indicated by dark areas and sampling
locations. The interrupted lines indicate the
southernmost limit of permafrost at the last
glacial maximum. The white areas indicate
the Fennoscandian, Pyreneean and Alpine
ice caps at the last glacial maximum. The
locations of the sampled populations are
indicated by dots and arrows, and num-
bered as in Table 1. (b) Proportion of mtDN
A
clades in regional groups of sampled roe
deer populations. The diagrams indicate
the proportions of mtDNA control region
haplotypes belonging to Clade West, Central,
East and ‘C. c. italicus’ in groups: 1 = Portugal;
2 = north Spain (Galicia, Asturias); 3 = centra
l
and south Spain (Segovia, Toledo, Cadiz);
4 = France; 5 = Denmark; 6 = Sweden; 7 =
Germany; 8 = western Italian Alps (Cuneo,
Torino); 9 = central Italian Alps (Sondrio,
Brescia, Lecco); 10 = eastern Italian Alps
(Belluno, Treviso, Asiago, Vicenza); 11 =
Serbia, Montenegro, Kosovo; 12 = Greece;
13 = northwestern Apennines (Savona); 14
= northeastern Apennines ( Bologna, Firenze,
Forlì, Cesena); 15 = populations of C. c. italicus
(Castelporziano, Gargano, Orsomarso); 16
= south Tuscany (Grosseto, Siena); 17 = nort
h
Tuscany (Arezzo, Pistoia, Massa Carrara)
and Modena.
3074
E. RANDI
ET AL.
© 2004 Blackwell Publishing Ltd,
Molecular Ecology
, 13, 3071– 3083
by polymerase chain reaction (PCR) in 728 roe deer samples
using the external primers LcapPro and HcapPhe (Randi
et al
. 1998). Automatic sequencing of 704 nucleotides from
the left and right sides of the mtDNA control region was
performed, using the two PCR primers and the internal
primers Lcap362 and Hcap493 (Randi
et al
. 1998). Sequences
were aligned with a control region sequence of European
roe deer (access no. Z70318; Douzery & Randi 1997), and
unique haplotypes were identified using
collapse
1.0
(D. Posada; http://bioag.byu.edu/zoology/crandall_lab/
programs.htm). From each of the 36 sampled groups 617
roe deer were randomly selected; these were genotyped by
Table 1 Sampling locations of roe deer (Capreolus capreolus) used in this study
Region Locality* Sample size Origin of populations†
Portugal 1 — Northern Portugal 27 Natural
Spain 2 — Galicia 19 Natural
3 — Asturias (Oviedo, Leon) 30 Natural
4 — Madrid (Segovia, Toledo) 6 Natural
5 — Cadiz 5 Natural
France 6 — Tolouse 18 Natural
Germany 7 — Westerwald 10 Natural
8 — Bayreuth 10 Natural
Denmark 9 — Kaloe 12 Restocked
Sweden 10 — Uppsala 10 Natural
Serbia 11 — Backi MonoStor 10 Natural
12 — Novi Kne6evac 20 Natural
13 — Ada Becej 27 Natural
14 — Novi Becej, Novo MiloSevo 6 Natural
15 — Deliblatska PeScara 14 Natural
16 — Stragari (Kragujevac) 30 Natural
17 — Petrovac na Mlavi, Svilajnac 25 Natural
18 — Severni Kucaj 17 Natural
19 — Negotin 15 Natural
20 — Stara planina 27 Natural
Kosovo 21 — Kosovo (PriStina) 13 Natural
Montenegro 22 — Montenegro (Berane) 11 Natural
Greece 23 — Epirus (Zagori) 16 Natural
Italy Alps 24 — Belluno 12 Natural
25 — Treviso 14 Natural
26 — Asiago 22 Natural
27 — Vicenza 20 Natural
28 — Sondrio, Brescia 15 Natural, restocked
29 — Lecco 19 Natural, restocked
30 — Cuneo 19 Natural, restocked
31 — Torino (Val di Susa) 3 Reintroduced
Italy Apennines 32 — Alessandria 14 Reintroduced
33 — Savona 43 Reintroduced
34 — Massa Carrara 42 Natural, restocked
35 — Modena 15 Natural, restocked
36 — Pistoia 20 Natural, restocked
37 — Bologna, Firenze 27 Natural, restocked
38 — Forlì, Cesena 14 Natural, restocked
39 — Arezzo 19 Natural, restocked
40 — Siena 43 Natural
41 — Grosseto 25 Natural
Italy C. c. italicus 42 — Castelporziano 15 Natural
43 — Gargano 8 Natural
44 — Orsomarso 3 Natural
*Geographic locations of the sampled populations are mapped in Fig. 1(a).
†Populations are ranked as: ‘natural’ for populations that were never officially restocked (undocumented restocking cannot be excluded);
‘restocked’ for populations that were officially restocked; ‘reintroduced’ for populations that were completely reintroduced in an area
where indigenous roe deer were recently extirpated.
PHYLOGEOGRAPHY OF ROE DEER POPULATIONS
3075
© 2004 Blackwell Publishing Ltd,
Molecular Ecology
, 13, 3071– 3083
PCR amplifications (annealing temperature 55
°
C) of 11
microsatellites: NVHRT16, NVHRT21, NVHRT24, NVHRT71
(Roed & Midthjell 1998), BMC1009, OarFCB304 (Talbot
et al
. 1996), ILSTS058, ILSTSO11 (Kemp
et al
. 1995), mcM505
(Hulme
et al
. 1995), OarH51 (Pierson
et al
. 1994), and RT1
(Wilson
et al
. 1997). The PCR products were analysed auto-
matically using an ABI 3100 sequencer and the programs
genescan
3.7 and
genotyper
2.1 for microsatellites, or
sequencing analysis
3.7 and
seqscape
1.1 for sequences.
Details of laboratory protocols are available upon request.
Table 2 Estimates of gene diversity at mitochondrial DNA (mtDNA) control region and 11 microsatellite loci (STR) in roe deer
Group*
Region and
population
mtDNA† STR†
(ni/nh)hkθ(s)niHOHEA
1N. Portugal 27/10 0.83 (0.04) 5.4 (2.7) 5.3 (2.4–11.2) 16 0.48 0.57 4.7
2N. Spain 46/10 0.82 (0.04) 6.9 (3.3) 3.5 (1.7–7.2) 21 0.60 0.65 6.5
3France 18/11 0.94 (0.04) 4.1 (2.1) 11.0 (4.6–26.8) 13 0.61 0.79* 6.4
4Germany 20/13 0.91 (0.05) 4.5 (2.3) 15.0 (6.5–35.5) 10 0.67 0.72 5.6
5Denmark 12/6 0.80 (0.09) 4.8 (2.5) 4.1 (1.4–11.5) 10 0.55 0.69 4.4
6Sweden 10/4 0.71 (0.12) 2.7 (1.6) 1.9 (0.6 6.1) 10 0.60 0.64 4.1
7Serbia 11 10/8 0.96 (0.06) 7.5 (3.8) 16.4 (4.9–59.0) 10 0.75 0.72 5.0
8Serbia 12 20/9 0.80 (0.09) 4.5 (2.3) 5.7 (2.4 –13.2) 11 0.58 0.66 5.0
9Serbia 13 27/14 0.94 (0.02) 6.1 (3.0) 11.0 (5.3–22.8) 12 0.72 0.71 5.9
10 Serbia 15 14/7 0.85 (0.07) 6.3 (3.2) 4.9 (1.8–12.9) 12 0.64 0.59 4.4
11 Serbia 16 30/20 0.96 (0.02) 7.6 (3.6) 25.1 (12.3–51.9) 14 0.70 0.71 6.0
12 Serbia 17 25/18 0.97 (0.02) 7.0 (3.4) 27.4 (12.5–62.3) 10 0.59 0.68 5.4
13 Serbia 18 17/13 0.96 (0.03) 8.2 (4.0) 23.4 (9.064.6) 12 0.64 0.70 5.6
14 Serbia 19 15/13 0.97 (0.04) 7.0 (3.5) 43.1 (13.7–149.0) 13 0.64 0.67 5.8
15 Serbia 20 27/17 0.94 (0.03) 6.9 (3.4) 18.7 (9.0–39.5) 11 0.54 0.60 5.3
16 Kosovo 21 13/10 0.92 (0.07) 5.1 (2.6) 18.0 (6.2–55.7) 10 0.69 0.60 3.4
17 Montenegro 22 11/9 0.96 (0.05) 7.0 (3.6) 20.7 (6.3–73.8) 10 0.74 0.69 4.8
18 Greece 23 16/4 0.44 (0.14) 3.0 (1.6) 1.4 (0.4–3.9) 24 0.68 0.53* 4.4
19 Italy Alps 24 12/6 0.76 (0.12) 5.0 (2.6) 4.1 (1.4–11.5) 11 0.67 0.74 5.5
20 Italy Alps 25 14/4 0.39 (0.16) 2.4 (1.4) 1.5 (0.4–4.4) 12 0.70 0.70 5.0
21 Italy Alps 26 16/4 0.35 (0.15) 1.2 (0.8) 1.4 (0.4–3.9) 22 0.65 0.58 4.7
22 Italy Alps 27 20/4 0.28 (0.13) 2.3 (1.3) 1.2 (0.4–3.4) 19 0.67 0.66 4.9
23 Italy Alps 28 15/10 0.94 (0.04) 7.0 (3.5) 12.0 (4.6–31.9) 15 0.71 0.75 6.4
24 Italy Alps 29 19/5 0.62 (0.10) 4.5 (2.3) 1.9 (0.7–4.9) 18 0.68 0.73 5.9
25 Italy Alps 30 17/4 0.65 (0.09) 4.0 (2.1) 1.3 (0.4–3.7) 19 0.69 0.72 6.2
26 Apennines 32 14/4 0.49 (0.15) 2.4 (1.4) 1.5 (0.54.4) 11 0.67 0.70 5.1
27 Apennines 33 27/4 0.62 (0.07) 2.8 (1.5) 1.0 (0.3–2.8) 43 0.61 0.65 6.6
28 Apennines 34 23/3 0.42 (0.10) 2.3 (1.3) 0.7 (0.2–2.1) 42 0.52 0.53 4.7
29 Apennines 35 15/4 0.60 (0.11) 5.1 (2.6) 1.4 (0.54.1) 15 0.65 0.67 5.1
30 Apennines 36 10/3 0.30 (0.18) 3.3 (1.9) 1.0 (0.3–3.6) 20 0.70 0.68 5.1
31 Apennines 37 26/4 0.62 (0.06) 3.3 (1.7) 1.1 (0.3–2.9) 27 0.56 0.59 4.4
32 Apennines 38 13/3 0.59 (0.12) 2.5 (1.4) 0.9 (0.2–2.9) 14 0.67 0.60 3.7
33 Apennines 39 18/3 0.31 (0.13) 0.5 (0.4) 0.7 (0.2–1.4) 19 0.56 0.58 4.1
34 Apennines 40 43/5 0.49 (0.08) 1.5 (0.9) 1.2 (0.5–3.0) 33 0.57 0.66 6.5
35 Apennines 41 24/4 0.71 (0.04) 3.2 (1.7) 1.1 (0.4–3.0) 25 0.69 0.73 6.0
36 C. c. italicus 24/4 0.56 (0.09) 1.6 (1.0) 1.1 (0.4 –3.0) 23 0.62 0.63 4.4
Average 19.7 0.97 7.9 63.7 17.2 0.63 0.77 14.1
(SD) (8.4) (0.01) (3.7) (52.7–76.9) (8.8) (0.11) (0.11) 14.1
For group composition, see: Table 1 and Materials and Methods.
ni = number of genotyped individuals; nh = number of observed mtDNA haplotypes; h = haplotype diversity; k = average pairwise
sequence divergence; θ(s) = 2Nem (Ne = effective population size; m = mutation rate of the haplotype), computed from the number of
segregating sites (Tajima 1983); HO = observed heterozygosity; HE = unbiased expected heterozygosity; A = mean number of alleles per
locus. Standard deviations (SD) are in brackets.
*Significant (P < 0.001) departures from Hardy–Weinberg equilibrium, as estimated using FIS. Critical levels in simultaneous tests of
significance were adjusted using Hochberg’s procedure (Legendre & Legendre 1998: p. 18), as implemented in adjusted-p-values
(P. Legendre; http://www.fas.umontreal.ca/biol/legendre).
3076
E. RANDI ET AL.
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 3071– 3083
Analyses of the mtDNA sequences
Phylogenetic trees were reconstructed using mega 2.1
(Kumar et al. 2001; http://www.megasoftware.net/), with
the neighbour-joining procedure (Saitou & Nei 1987) and
Tamura and Nei’s TN93 genetic distance model (Tamura
& Nei 1993), which is appropriate to describe the evolu-
tion of control region sequences. Trees were rooted using
homologous control region sequences of Siberian roe deer
(Douzery & Randi 1997), and support for the internodes
was assessed after 10 000 bootstrap resampling steps
(BP; Felsenstein 1985). The complete alignment (n = 161
haplotypes) was analysed and a reduced data set (n = 81
haplotypes) was obtained by excluding all singletons (i.e.
the 80 haplotypes that were sampled only once).
Networks are better suited than phylogenetic methods
to infer haplotype genealogies at the population level
because they explicitly allow for extant ancestral sequences
and alternative connections (Bandelt et al. 1999). We used
both the complete and reduced alignments with the
median-joining network procedure (Bandelt et al. 1999),
implemented in network 3.1.1.1 (http://www.fluxus-
technology.com/). We used the ‘star contraction algo-
rithm’ in network to collapse star-like clusters to a smaller
number of representative single sequences, with the aims:
(i) to display the inner structure of the network; and (ii)
to identify those star-like clusters that represent events of
demographic expansion (Forster et al. 2001).
Haplotype diversity (h), average pairwise nucleo-
tide substitutions (k), nucleotide diversity (π) and ana-
lysis of molecular variance (amova; Excoffier et al. 1992)
with Φ-analogues of Wright’s (1965) F-statistics, were
computed using arlequin (Schneider et al. 2002; http://
anthropologie.unige.ch/arlequin). amova was performed
first using only haplotype frequency differences among
populations, then including sequence divergence among
haplotypes (estimated by TN93 distances). Mismatch dis-
tributions were analysed using the sudden expansion model
(Rogers & Harpending 1992), and goodness-of-fit tests (sum
of squared deviations, SSD; Harpending’s raggedness
index, R; Schneider & Excoffier 1999) of the observed to
the estimated mismatch distributions were computed.
Assumption of neutrality of mutations was tested by
Tajima’s D (Tajima 1989) as implemented in dnasp 3.99
(Rozas et al. 2003). dnasp was used also to estimate θ(s) =
2Nem (Ne = effective population size; m = mutation rate of
the haplotype), computed from the number of segregating
sites (Tajima 1983).
Microsatellite diversity within and among populations
Commonly used estimates of genetic diversity (hetero-
zygosity H, and number of alleles A) were computed for each
locus and population using genetix 4.03 (Belkhir et al. 2001;
http://www.University-montp2.fr/genetix/genetix.htm).
Deviations from Hardy–Weinberg equilibrium for each
locus and each population, and across loci and populations,
were assessed using genepop 3.2a (Raymond & Rousset
1995). arlequin was used to compute amova with FST
analogues. Patterns of differentiation were visualized by
factorial correspondence analysis (CA; Benzecri 1973) of
population multilocus scores computed using genetix.
Results
Mitochondrial DNA sequences
The mtDNA alignment (728 individuals, 704 nucleotides,
outgroups excluded), showed 161 haplotypes (GenBank
accession nos AY625732–AY625892), defined by 70 poly-
morphic sites including 69 substitutions (62 transitions and
seven transversions) and two insertions/deletions. Mito-
chondrial DNA diversity was high in roe deer (Table 2),
which showed on average one distinct haplotype over 4.5
individuals (728/161 = 4.5). Haplotypes were distributed
evenly, each one having a frequency lower than 9% in the
total sample. Haplotype diversity was high (h = 0.971
± 0.002, standard deviation), but nucleotide diversity
(π = 0.011 ± 0.006) and pairwise divergence (k = 7.89 ± 3.67)
were small, suggesting that roe deer populations had
historically large effective size (Ne), but that extant mtDNA
lineages originated recently. Tajima’s neutrality test, com-
puted from the number of segregating sites, was negative
(D = 0.52) and not significant (P > 0.10).
Mismatch analyses supported a pattern of demographic
expansion. The goodness-of-fit tests were not signific-
ant (SSD = 0.0041, P = 0.20; R = 0.0051, P = 0.66). The main
expansion event occurred at τ = 8.6 (lower bound τ = 4.3,
upper bound τ = 15.3; a level = 0.05), and involved a popu-
lation change from an initial \ = 1.06 (0.00–7.16), to a final
θ = 139.06 (19.69–535.66). However, the mismatch plot was
bimodal (not shown), with peaks at τ = 7 and 11, suggest-
ing the expansion of two distinct mtDNA clades.
Mitochondrial DNA diversity was higher in roe deer
sampled from eastern countries (Serbia, Montenegro and
Kosovo), than elsewhere in Europe (Table 2). Values of θ(s)
in samples from central–northern Europe, eastern Europe,
Italian Alps and Apennines were significantly different
(P = 0.0036; Friedman’s test). In particular, θ(s) was signifi-
cantly higher in samples from Serbia, than in samples from
the Italian Alps (P = 0.0063) and Apennines (P = 0.0007;
paired t-test). Eastern European roe deer also showed the
highest values of haplotype diversity, which was signi-
ficantly different from h in the Alpine (P = 0.0037; paired
t-test) and Apennine samples (P = 0.0001), but not signi-
ficantly different from the other sampled populations in
Europe (P = 0.1834). Genetic variation was significantly
partitioned among the 36 groups, with ΦST = 0.44 (estimated
PHYLOGEOGRAPHY OF ROE DEER POPULATIONS 3077
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 3071– 3083
using amova and TN93 distances; P = 0.00000), or FST = 0.26
(using F-statistics computed from haplotype frequencies;
P = 0.00000). The contribution of genetic distances increased
the F-value, suggesting that genetically distinct mtDNA
haplotypes are distributed in different geographical loca-
tions, and that, on average, haplotypes sampled within the
same location are more similar to each other than to those
found in other locations.
Phylogenetic relationships among the mtDNA haplotypes
Neighbour-joining trees, computed with the complete or
the reduced alignments and TN93 distances, grouped all
the European roe deer haplotypes in a clade (supported by
BP = 100%), which was split into two main haplogroups
(BP < 50%; Fig. 2 shows the neighbour-joining tree com-
puted with the reduced data set of 81 haplotypes). A first
haplogroup, thereafter named Clade East, was composed
mainly by haplotypes sampled from Serbia, Montenegro
and Kosovo, and included all the haplotypes sampled in
Greece, plus six haplotypes collected in the Italian Alps
and northern Apennines, and in Germany. These haplotypes
did not form distinct subclades, but were nested within
clusters of Serbian haplotypes.
A second haplogroup was split into two distinct line-
ages, respectively, named Clade West and Clade Central
(Fig. 2). Clade West included sequences that were sampled
only in Portugal and Spain, except for a subclade with
three haplotypes that were collected in the central Italian
Alps (sampling locations no. 28 and no. 29 in Fig. 1). Haplo-
types in Clade Central were widespread in central and
north Europe and Italy, as well as in east Europe (including
Serbia, but not Greece) and in west Europe (including
Spain and Portugal). A distinct subclade joined all the
haplotypes from populations of Capreolus capreolus italicus
(Castelporziano, Gargano and Orsomarso), as well as other
samples collected from neighbouring central and southern
Apennine localities.
Average interhaplotype sequence divergence was small
(1.1% TN93 distance), the number of haplotypes (161) was
greater than parsimony informative sites (41), and strong
bootstrap support was not expected in neighbour-joining
trees (Smouse 1998). Nevertheless the mtDNA alignment
contained a significant phylogenetic signal, as indicated by
the analysis of 100 000 random trees (generated by paup*;
Swofford 2002), which showed a skewed distribution of
tree length, and a significant value of the statistics g1 = 0.16
(P = 0.01; see Table 2 in Hillis & Huelsenbeck 1992).
Network analyses and geographical distribution of the
mtDNA haplotypes
Networks computed using the complete or reduced
data set were very reticulated, but concordantly split the
haplotypes in three main groups, corresponding to Clade
East, West and Central. Three rounds of star contraction
with the star contraction algorithm in network, per-
formed using the entire data set and a mutational distance
radius δ = 5, allowed the initial 161 sequences to be col-
lapsed into a final set of 96 sequences, which were used to
construct a median-joining network. The star-contracted
network (Fig. 3) supported the distinction of the three main
clades, collapsed the Alpine haplotypes in Clade West in
a single sequence (H56), collapsed the nine haplotypes of
C. c. italicus’ in three sequences (H35, H42 and H15), and
supported the existence of a distinct mtDNA clade in
Iberian roe deer joining Clade Central, which included the
other Iberian mtDNA haplotypes.
The geographical distribution of the haplotypes is
synthesized in Fig. 1(b). Pie diagrams showed that popu-
lations sampled in Europe have different compositions,
and that control-region clades have restricted geographical
distributions. Haplotypes of ‘C. c. italicus’ were sampled only
in neighbouring locations in central Apennines (locations
34, 35, 36, 40 and 41 in Tuscany and Emilia-Romagna; see
Fig. 1), however, well beyond (stippled line in Fig. 1b) the
putative southern distribution of C. c. italicus (continuous
line in Fig. 1b).
Microsatellite diversity and geographical population
structure
All the microsatellite loci were polymorphic, showing an
average of 14.1 alleles per locus in the total sample (n =
617), and from 3.4 alleles (in roe deer from Kosovo) to 6.6
alleles (in a reintroduced population sampled from location
33) in the local samples. The difference in the average
number of alleles per locus in the total or in the local
samples indicates that distinct alleles are differentially dis-
tributed in the sampled populations. The total sample was
not in Hardy–Weinberg equilibrium (P < 0.000; multilocus
multipopulation test), confirming that populations are
genetically distinct. The local populations were in Hardy–
Weinberg equilibrium, excluding samples from France
and Greece (P < 0.001). Heterozygosity and the average
number of alleles per locus did not show any detectable
geographical trend (Table 2). However, microsatellite vari-
ability was significantly partitioned among the 36 groups
(FST = 0.16; P < 0.0000; amova).
The roe deer sampled from the different regions were
distributed in distinct areas of the CA plot (Fig. 4), suggest-
ing geographical subdivisions. Roe deer sampled from
eastern Europe and from the Italian Alps were partially
admixed in the central part of the CA plot, while samples
collected from the geographical peripheries (central and
southern Apennines, north Iberia) were distinct, plotting,
respectively, towards the left and lower sides of the dis-
tribution. This plot shows that the central cluster of samples
3078 E. RANDI ET AL.
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 3071– 3083
Fig. 2 Neighbour-joining tree of roe deer mtDNA control region haplotypes, computed using TN93 genetic distances and a reduced
alignment of 80 haplotypes (singletons excluded). The trees were rooted using homologous Siberian roe deer sequences (position of the roo
t
indicated by an asterisk). The main clades are labelled and bootstrap percentages are indicated at the main internodes. The table on the righ
t
shows the geographical distribution of these haplotypes (PT = Portugal; ES = Spain; FR = France; DE = Germany; DK = Denmark;
SE = Sweeden; SR = Serbia, Montenegro, Kosovo; GR = Greece; ItAl = Italian Alps; ItAp = Italian Apennines; Ita = C. c. italicus).
PHYLOGEOGRAPHY OF ROE DEER POPULATIONS 3079
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 3071– 3083
from Serbia, Kosovo and Montenegro includes also the roe
deer from Greece and the samples collected from the two
populations that were reintroduced in the western Italian
Alps (Alessandria and Savona) using roe deer from eastern
Europe or eastern Alps. Roe deer from central and south
Apennines plot towards the left side of the CA, differenti-
ating mainly along FC-I (which explains 13.01% of the total
genetic variation), while roe deer sampled from the north-
ern Apennines (Forlì, Cesena, Bologna, Firenze, Arezzo)
plot towards the upper right part of CA. Samples collected
from C. c. italicus and neighbouring areas also plot towards
the left side of the CA. Samples collected from western (Iberia,
France) and northern Europe (Sweden) plot towards the
lower right part of the CA, differentiating along FC-II
(which explains 11.38% of the total genetic variation.
Discussion
Global phylogeographical patterns in roe deer
Phylogenetic trees and networks identify three main
mtDNA haplogroups, which could have originated in
Iberia (Clade West and perhaps Central) or in the Balkans
(Clade East). Clade West apparently contributed little to
the current mtDNA diversity in central Europe, which
is mainly because of the widespread distributions of
Clade East and Central. Some Clade West haplotypes were
sampled in the Italian Alps, suggesting a postglacial
colonization route from Iberia towards the Mediterranean
coasts, or the presence of ancestral haplotypes in the
Alps. The restricted distribution of Clade East haplotypes
supports the existence of an eastern glacial refuge, from
which, however, roe deer did not disperse extensively
westward. In contrast, the widespread distribution of
Clade Central haplotypes in the Balkans, central and north
Europe, Apennines, Alps (although some haplotypes could
have been translocated in the western Alps), and Iberia,
supports the existence of distinct refugial populations that
contributed extensively to the recolonization of Europe.
These findings indicate that roe deer dispersed in
Europe from multiple refuges, not from a single western
Mediterranean refuge, as previously suggested (Wiehler &
Tiedemann 1998).
Phylogenetic trees (Fig. 2) and networks (Fig. 3), and
the mismatch distributions showed instances of two main
expansions, respectively, at τ = 11 and τ = 7 mutations.
Assuming, in a very simple way, that divergence time is
T = τ/2µ, with µ = λg (i.e. the rate of annual substitution
per haplotype λ = 0.04–0.08 × 106 × 750 nucleotides, per
generation time g = 3 years; Randi et al. 1998), coalescent
times can be estimated to range from 122 000 to 244 000 years
for the oldest, and from 78 000 to 156 000 years for the most
Fig. 3 Star-contracted median-joining networ
k
computed using network (with default
options and all sites equally weighted),
and the entire data set of 161 haplotypes.
Haplotypes are indicated by numbers, and
circles are not proportional to frequencies.
The main clades preserved after star-contractio
n
are labelled.
3080 E. RANDI ET AL.
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 3071– 3083
recent expansions, respectively (in agreement with Randi
et al. 1998; Vernesi et al. 2002). These dates, which might be
underestimated because of recurrent mutations at hyper-
variable control region sites, could identify two waves of
population expansion, respectively, during the penultimate
(c. 200 000 years ago), and the last (c. 130 000 years ago)
interglacials. Extensive mtDNA and STR genetic diversity
suggests that roe deer populations did not suffer any recent
global strong bottleneck. The negative value of Tajima’s
neutrality test, which might reflect the effects of population
histories, can result from postglacial population expansions.
During glacial periods, permafrost and Arctic tundra
ecosystems were widespread in central Europe down to a
latitude of 45° (Andersen & Borns 1997). Temperate animal
populations were forced to survive in fragmented south-
ern refuges, following the range shifts of broad-leaved and
Mediterranean forests. A prevalent eastward roe deer
colonization of central Europe correlates with patterns
of broad-leaf forest expansion from the east, as inferred
both from fossil pollen and molecular data sets (Petit et al.
2003). Alternatively, divergent mtDNA lineages could have
evolved in Iberia, thereafter spreading towards central
Europe and Italy following the postglacial colonization
routes that were documented in other species (Taberlet
et al. 1998). Additional sampling in France and Germany is
needed to assess the distribution of Iberian haplotypes in
central Europe, and locate the geographical origin of Clade
Central in Iberia or in the Balkans. The original popula-
tions in western Alps and Apennines have been eradicated
and were recently reintroduced using roe deer from east-
ern Alps and east Europe. Hence, only museum samples
could provide support to the existence of southwestern or
eastern colonization routes.
Local diversification and subspeciation in peripheral
populations
Phylogenetic trees and networks identify a clade of haplo-
types with restricted geographical distribution in the central
and southern Apennines, within the range of Capreolus
capreolus italicus. The star-contraction analysis preserved
this clade, thus identifying an event of demographic
expansion of phylogenetically related populations that are
currently distributed in the central and southern Italian
Apennines. The average mtDNA sequence divergence
among haplotypes in clade ‘C. c. italicus’ is low (d = 0.36%),
indicating a recent coalescence. The origin of this clade is
the likely consequence of population isolation/expansion
in a southern Italian refuge during the last glacial maximum/
early Holocene. Mitochondrial, STR and morphometric
data (Randi et al. 1998; Randi & Mucci 2001; Lorenzini et al.
2002; Vernesi et al. 2002; Montanaro et al. 2003) indicate
that these vicariant populations evolved distinctive dia-
gnostic traits, and support the validity of subspecies C. c.
italicus Festa (1925). However, mtDNA haplotypes of clade
C. c. italicus’ were detected also in roe deer sampled in
north Tuscany and Emilia-Romagna, north of the putative
distribution limits of the Italian subspecies, where they are
admixed with roe deer bearing haplotypes belonging to
Clade Central (Fig. 1b).
The original distribution of C. c. italicus is unknown,
because roe deer in the Apennines were already largely
eradicated before World War II. Molecular and morpho-
metric data now suggest that populations of C. c. italicus
persisted not only in the protected areas of Castelporziano,
Gargano and Orsomarso (Fig. 1a), but also in southern
Tuscany, and probably in small remnant populations
Fig. 4 Factorial correspondence analysis
(CA) of population multilocus scores com-
puted using genetix. Multilocus scores are
plotted in the bivariate space defined by the
first two factorial components (CA-I and
CA-II). Locations of individuals sampled
from the main geographical regions are
labelled. Sampled groups are numbered as
in Fig. 1 and Table 1. The dotted ellipses
were drawn by hand to indicate popu-
lations sampled in the Balkans and in the
Italian Alps. The arrows indicate the
approximate linear geographical distances
between populations plotting, respectively,
towards the upper left or lower right
corners of the CA space.
PHYLOGEOGRAPHY OF ROE DEER POPULATIONS 3081
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 3071– 3083
along the ridge of the Apennines, in Tuscany and Emilia-
Romagna (locations nos 34, 35, 36, 40 and 41; Fig. 1a). Alter-
natively, or concomitantly, roe deer bearing ‘C. c. italicus’
mtDNA haplotypes might have recently expanded their
range towards northern Tuscany and Emilia-Romagna. In
these areas, roe deer populations are currently admixing,
as a result of reintroduction and restocking operations,
thus threatening the integrity of C. c. italicus.
Roe deer in central and southern Spain showed
haplotypes belonging to two distinct and distantly related
clades (Figs 2 and 3), which originated either in Clade
West or Central. In the southern Spanish population from
Cádiz we found three mtDNA haplotypes: no. 158, no. 159
(linked to Clade Central), and no. 160 (belonging to Clade
West). Moreover, haplotype no. 158 was shared with
populations in central Spain (Segovia, Burgos). If Clade
Central originated in eastern Europe, roe deer in Spain
could have admixed origins, and monophyly of subspecies
garganta would be not supported by molecular data.
Recent population changes, disturbance and
nonequilibrium effects
Deforestation and over-hunting led to the eradication of
roe deer populations in Spain, Italy and Greece. After
World War II, natural expansion or local reintroductions
led to the reconstitution of extant populations, which are
often expanding and admixing, sometimes after prolonged
isolation in fenced or protected areas. The roe deer is subject
to strong hunting pressure, which affects abundance and
distribution and may modify population structure and
reproductive behaviour (MiloSevi5-Zlatanovi5 et al. 2003).
Thus, the demographic structure of some roe deer popu-
lations in Europe fluctuated recently, and probably these
populations are not in genetic equilibrium.
Roe deer sampled from different areas of the Apennines,
spanning a range of c. 300 km, are genetically more distinct
than roe deer sampled from across all Europe, which span
a range of more than 4000 km (Fig. 4). These results sug-
gest that genetic drift as a result of recent fragmentation,
isolation and bottleneck greatly inflated the observed
genetic distances among roe deer populations distributed
in the Apennines. Microsatellites support the phylogeo-
graphical patterns described by the mtDNA sequences,
that is differentiation of peripheral roe deer populations
in north Iberia and south Italy vs. greater admixture of
populations sampled in central Europe and in the Balkans.
However, nonequilibrium features warn against the use of
pairwise population distances to infer mtDNA or micro-
satellite population trees (Lorenzini et al. 2002; Vernesi et al.
2002). Bottlenecks and founder events may result in the
loss of genetic variability, but they can also result in a rapid
differentiation between populations, through random
sampling of alleles at polymorphic loci.
Expectations from the southern refuges phylogeograph-
ical model (Hewitt 2000) suggest that genetic diversity
should be greater in refugial populations. However, roe
deer in Italy could have lost genetic diversity because
of population decline and drift, and are the less variable
among the studied samples. Quite surprisingly, popula-
tion no. 33 that was reintroduced in Liguria showed the
highest number of alleles (A = 6.6). However, this popula-
tion originated from roe deer from three different regions
that were released in 1952 (two males and one female from
Tuscany), in 1959 (two males, four females and three
juveniles from Slovenia), and in 1974 (six roe deer from
Trentino, eastern Alps). Admixed populations can be
variable, particularly if they expand rapidly and retain most
of their original genetic diversity. Moreover, recently founded
populations can show high values of allele number and
heterozygosity, which result from the sampling process
(Tarr et al. 1998).
Conclusions
Roe deer in Europe showed a complex pattern of popu-
lation structuring, which probably results from historical
vicariance in southern glacial refuges (as described
mainly by mtDNA findings), and subsequent population
admixture as a result of natural secondary contacts or
of anthropogenic disturbance (as described mainly by
STR data). Maternal and biparental markers used in this
study described concordant patterns of population struc-
ture, that is population admixture in the central Europe
vs. greater diversification towards the peripheries. As
expected, the STR markers were more informative in
describing the consequences of the most recent population
events (isolation, expansion, translocations, current dis-
persal rates). In accordance with current taxonomy (Danilkin
1996), roe deer in most of Europe should belong to a single
panmictic population. In contrast, peripheral roe deer
populations in north Portugal, southern Italian Apennines
(and perhaps Greece) may represent the remains of late
glacial refugial populations that should be preserved
and not artificially admixed with other geographical
populations.
Acknowledgements
We deeply appreciated the collaboration of everybody who
helped in sampling collection: M. Had6i-Pavlovic (National Park
Derdap’), T. Radosavljevic (Hunters Association ‘Stragari’) and
K. Laslo (Hunters Association ‘Becej’) in Serbia; A. Giannakopoulos
and H. Goumas in Greece; M. Paganin and G. Stefani (Asiago),
E. Bonavetti (Val Camonica), P. Gatti (Lecco), G. Meneguz (Torino
and Alessandria), S. Mattioli, V. Guberti and E. Raganella (Bologna),
C. Matteucci (Forlì), M. Ferri (Modena), M. Apollonio (Arezzo),
L. Orlandi, S. Nicoloso and DREAM (Tuscany), L. Bonfigli (Massa
Carrara), A. Scappi (Grosseto, Siena), I. Farronato (Vicenza),
3082 E. RANDI ET AL.
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 3071– 3083
V. Gabardo (Treviso), B. Morriconi and Hunters Association of Alta
Valtellina (Sondrio), G. Benvenuti (Firenze), A Drovandi (Pistoia),
M. Cattani (Marradi), M. Meacci (Arezzo), S. Spanò and A. Marsan
(Savona), B. Audino (Cuneo), L. Pedrotti (Bormio) in Italy; W.
d’Oleire-Oltmanns in Germany. Most of the Portuguese samples
were provided by BTVS-ICN (Wild Animal Tissue Bank, Portu-
guese Conservation Institute). V. Piorno and G. Pajares provided
Galician and Asturian samples, respectively. We thank the officials
of the Serbian Hunters Association for field and data acquisition
assistance. We thank J.G. Martínez and J.L. Fernández for com-
ments to the manuscript. Funding were provided to E.R. by INFS,
and Ministero dell’Ambiente, Direzione Conservazione della
Natura.
References
Adamakopoulos P, Adamakopoulos T, Bousbouras D et al. (1991)
Les grands mammiferes de Grece (Carnivores et Artiodactyles):
situation actuelle, repartition, habitat. Les especes menacees,
perspectives de protection. Biologia Gallo-Hellenica, 18, 107–126.
Albanis K, Galanos F, Boskos L, eds (2000) Criteria and Indicators for
the Sustainable Forest Management in Greece. Greek Ministry of
Agriculture, Athens.
Andersen BG, Borns HW Jr (1997) The Ice Age world: An Intro-
duction to Quaternary History and Research with Emphasis on
North America and Europe During the Last 2.5 Million Years.
Scandinavian University Press, Oslo.
Aragón S, Braza F, San José C (1995a) Socioeconomic, physio-
gnomic, and climatic factors determining the distribution pattern
of roe deer Capreolus capreolus. Spain. Acta Theriologica, 40, 37–43.
Aragón S, Braza F, San José C (1995b) Características morfológicas
de los corzos (Capreolus capreolus) de las sierras de Cádiz-
Málaga. Doñana Acta Vertebrata, 22, 51–64.
Bandelt H-L, Forster P, Rohl A (1999) Median-joining networks
for inferring intraspecific phylogenies. Molecular Biology and
Evolution, 16, 37–48.
Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F (2001)
GENETIX 4.02, Logiciel Sous Windows TM Pour la Génétique Des
Populations. Laboratoire Génome, Populations, Interactions,
CNRS UMR 5000, Université de Montpellier II, Montpellier.
Benzécri JP (1973) L’analyse Des Donnèes: T. 2, L’analyse Des Cor-
respondances. Dunoud, Paris.
Boutin JM (1990) Le statu du chevreuil (Capreolus capreolus) au
Portugal et les recherches effectuées sur cette espèce. Bulletin
Mensuel de l’Office Nationale de la Chasse, 143, 35–37.
Breitenmoser U (1998) Large predators in the Alps: the fall and rise
of man’s competitors. Biological Conservation, 83, 279–289.
Danilkin A (1996) Behavioural Ecology of Siberian and European Roe
Deer. Chapman & Hall, London.
Douzery E, Randi E (1997) The mitochondrial control region of
Cervidae: evolutionary patterns and phylogenetic content.
Molecular Biology and Evolution, 14, 1154–1166.
Excoffier L, Smouse PE, Quattro JM (1992) Analysis of molecular
variance inferred from metric distances among DNA haplo-
types: application to human mitochondrial DNA restriction
data. Genetics, 131, 479–491.
Felsenstein J (1985) Confidence limits on phylogenies: an
approach using the bootstrap. Evolution, 39, 783–791.
Festa E (1925) Il capriolo dell’Italia centrale (Capreolus capreolus
italicus sbsp. nova). Bollettino del Museo di Zoologia e Anatomia
Comparata Dell’università di Torino, 7, 1–2.
Forster P, Torroni A, Renfrew C, Rohl A (2001) Phylogenetic star
contraction applied to Asian and Papuan mtDNA evolution.
Molecular Biology and Evolution, 18, 1864–1881.
Gerloff U, Schlotterer C, Rassmann K et al. (1995) Amplification of
hypervariable simple sequence repeats (microsatellites) from
excremental DNA of wild living Bonobos (Pan paniscus). Mole-
cular Ecology, 4, 515–518.
Harris RB, Wall WA, Allendorf FW (2002) Genetic consequences
of hunting: what do we know and what should we do? Wildlife
Society Bulletin, 30, 634–643.
Hewitt G (2000) The genetic legacy of the Quaternary ice ages.
Nature, 405, 907–913.
Hey J, Machado CA (2003) The study of structured populations —
new hope for a difficult and divided science. Nature Reviews,
Genetics, 4, 535–543.
Hillis DM, Huelsenbeck JP (1992) Signal, noise, and reliability in
molecular phylogenetic analyses. Journal of Heredity, 83, 189–195.
Hulme DJ, Smith AJ, Silk JP, Redwin JM, Beh KJ (1995) Poly-
morphic sheep microsatellites at the McM2, McM131, McM135,
McM136, McM140, McM200, McM214, McM373, McM505,
McM507 and McM512 loci. Animal Genetics, 26, 369–370.
Kemp SJ, Hishida O, Wambugu J et al. (1995) A panel of poly-
morphic bovine, ovine and caprine microsatellite markers.
Animal Genetics, 26, 299–306.
Kumar S, Tamura K, Jakobsen IB, Nei M (2001) MEGA2: Molecular
Evolutionary Genetics Analysis Software. Arizona. State Univer-
sity, Tempe, AZ.
Legendre P, Legendre L (1998) Numerical Ecology, 2nd edn. Else-
vier, Amsterdam.
Lorenzini R, Lovari S, Masseti M (2002) The rediscovery of the
Italian roe deer: genetic differentiation and management
implications. Italian Journal of Zoology, 69, 367–379.
Lorenzini R, San José C, Braza F, Aragón S (2003) Genetic differen-
tiation and phylogeography of roe deer in Spain, as suggested
by mitochondrial DNA and microsatellite analysis. Italian Jour-
nal of Zoology, 70, 89–99.
Maehr DS, Noss RF, Larkin JL, eds. (2001) Large Mammal Restora-
tion. Island Press, Washington, DC.
Meunier K (1983) Das Spanish Reh. Cited in Lorenzini R, Lovari S,
Masseti M (2002) The rediscovery of the Italian roe deer: Genetic
differentiation and management implications. Italian Journal of
Zoology, 69, 367–379.
MiloSevi5-Zlatanovi5 S, Cerani5 A, Stamenkovi5 S (2003) Eco-
geographic constraints in current game management of the roe
deer (Capreolus capreolus L.) in Serbia. Xxvith International IUGB
Congress, September 1–6, 2003, Abstracts, p. 72, Braga, Portugal.
Montanaro P, De Marinis AM, Riga F, Focardi S, Toso S (2003)
Variabilità craniometrica in alcune popolazioni italiane di
Capriolo. In: Atti IV Congresso Italiano Di Teriologia (eds Prigioni C,
Meriggi A, Merli E). Hystrix, 14, 169–170.
Moritz CC (1994) Defining ‘evolutionarily significant units’ for
conservation. Trends in Ecology and Evolution, 9, 373–375.
Perco F, Calò CM (1994) The status of roe deer in Italy. In: Proceed-
ings of the 2nd European Roe Deer Meeting. Brixen, South Tyrol/
Italy, October 27–30. (ed. Wotschikowoki U), pp. 43–49. Wild-
biologische Gesellschaft, Munich.
Petit RJ, Aguinagalde I, de Beaulieu J-L et al. (2003) Glacial refugia:
hotspots but not melting pots of genetic diversity. Science, 300,
1563–1565.
Pierson CA, Ede AJ, Crawford AM (1994) Ovine microsatellites
at the OarHH30, OarHH51, OarVH54, OarCP88, OarCP93,
OarCP134 loci. Animal Genetics, 25, 294–295.
PHYLOGEOGRAPHY OF ROE DEER POPULATIONS 3083
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 3071– 3083
Randi E, Mucci N (2001) Genetic structure and origin of roe deer
populations in Italy. In: Proceedings of the 3rd Roe Deer Inter-
national Symposium (Abstracts), Tredozio (BO), Italy. (ed. Focardi S),
p. 5. INFS, Bologna.
Randi E, Pierpaoli M, Danilkin A (1998) Mitochondrial DNA
polymorphism in populations of Siberian and European roe
deer (Capreolus pygargus and C. capreolus). Heredity, 80, 429
437.
Raymond M, Rousset F (1995) Genepop Version 1.2.: population
genetics software for exact tests and ecumenicism. Journal of
Heredity, 86, 248–249.
Rhodes OE Jr, Smith MH (1992) Genetic perspectives in wildlife
management: the case of large herbivores. In: Wildlife 2001:
Populations (ed. McCullogh DR, Barrett RH), pp. 985– 996. Elsevier
Science, London.
Roed KH, Midthjell L (1998) Microsatellites in reindeer, Rangifer
tarandus, and their use in other cervids. Molecular Ecology, 12,
1773–1776.
Rogers AR, Harpending H (1992) Population growth makes
waves in the distribution of pairwise genetic differences. Mole-
cular Biology and Evolution, 9, 552–569.
Rozas J, Sanchez-DeI, Barrio JC, Messeguer X, Rozas R (2003)
DnaSP, DNA polymorphism analyses by the coalescent and
other methods. Bioinformatics, 19, 2496–2497.
Saitou N, Nei M (1987) The neighbor-joining method: a new
method for reconstructing phylogenetic trees. Molecular Biology
and Evolution, 4, 406–425.
Schneider S, Excoffier L (1999) Estimation of demographic
parameters from the distribution of pairwise differences
when mutation rates vary among sites: applications to human
mitochondrial DNA. Genetics, 152, 1079–1089.
Schneider S, Roessli D, Excoffier L (2002) ARLEQUIN Ver 2.0. A Soft-
ware for Population Genetics Data Analysis. Genetics and Biometry
Laboratory. Department of Anthropology and Ecology. Univer-
sity of Geneva, Geneva, Switzerland.
Smouse P (1998) Tree or not tree? Molecular Ecology, 7, 399–412.
Swofford DL (2002) PAUP*: Phylogenetic Analysis Using Parsimony,
Version 4.0b2a. Sinauer Associates, Sunderland, MA.
Taberlet P, Fumagalli L, Wust-Saucy A-G, Cosson J-F (1998) Com-
parative phylogeography and postglacial colonization routes in
Europe. Molecular Ecology, 7, 453–464.
Tajima F (1983) Evolutionary relationships of DNA sequences in
finite populations. Genetics, 105, 437–460.
Tajima F (1989) Statistical method for testing the neutral mutation
hypothesis by DNA polymorphism. Genetics, 123, 585–595.
Talbot J, Haigh J, Plante Y (1996) A parentage evaluation test in
North American elk (wapiti) using microsatellites of ovine and
bovine origin. Animal Genetics, 2, 117–119.
Tamura K, Nei M (1993) Estimation of the number of nucleotide
substitutions in the control region of mitochondrial DNA in
humans and chimpanzees. Molecular Biology and Evolution, 10,
512–526.
Tarr CL, Conant S, Fleischer RC (1998) Founder events and vari-
ation at microsatellite loci in an insular passerine bird, the Laysan
finch (Telespiza cantans). Molecular Ecology, 7, 719–731.
Tellería JL, Virgós E (1997) Distribution of an increasing roe deer
population in fragmented Mediterranean landscape. Ecography,
20, 247–252.
Vernesi C, Pecchioli E, Caramelli D, Tiedemann R, Randi E,
Bertorelle G (2002) The genetic structure of natural and reintro-
duced roe deer (Capreolus capreolus) populations in the Alps and
central Italy, with reference to the mitochondrial DNA phylo-
geography of Europe. Molecular Ecology, 11, 1285–1297.
Wiehler J, Tiedemann R (1998) Phylogeography of the European roe
deer Capreolus capreolus as revealed by sequence analysis of the
mitochondrial control region. Acta Theriologica Suppl., 5, 187–197.
Wilson GA, Strobeck C, Wu L, Coffin JW (1997) Characterization
of microsatellite loci in caribou Rangifer tarandus, and their use
in other artiodactyls. Molecular Ecology, 6, 697–699.
Wright S (1965) The interpretation of population structure by
F-statistics with special regards to systems of mating. Evolution,
19, 395–420.
This study is part of a long-term project on phylogeography,
population and conservation genetics of the roe deer. E. Randi is
head of conservation biology and genetics at INFS. N. Mucci
is a laboratory technician at INFS. A. Sfougaris is a lecturer in
Ecosystem Management at the University of Thessaly, and co-
ordinator of research projects on the ecology and management of
mammalian and avian species. J. Carranza is head of a research
group on Behavioural Ecology and Population Management at
the University of Extremadura in Spain. S. MiloSevi5-Zlatanovi5
is a lecturer in Systematics and Comparative Morphology of
Chordates at University of Kragujevac and coordinator of research
projects on the genetics, ecology and management of cervids in
Serbia and Montenegro. P.C. Alves is a biologist at CIBIO/UP,
University of Porto, Portugal, and is interested in the ecology and
conservation genetics of Iberian mammals.
... Below species level, the roe deer harbors an overall high genetic diversity, as revealed by biochemical and molecular results (c.f. Hartl et al. 1991;Randi et al. 2004;Lorenzini and Lovari 2006). Molecular information, in particular, told us much about historical dynamics and contemporary demography of populations, their connectivity, as well as the uptake of interactions with the environment, within a so-called landscape genetics framework, that is, the integration of landscape ecology and population genetics. ...
... Recently, molecular phylogeographic investigations revealed that the European roe deer harbors high genetic variation and a complex population structure across the entire range. Data from different mitochondrial and nuclear DNA-based studies (Randi et al. 2004;Lorenzini and Lovari 2006;Lorenzini et al. 2014) concordantly suggest the existence of four major genetic lineages of roe deer in Europe (Fig. 3). One central lineage of mitochondrial haplotypes is widely distributed across the whole European continent (from northern Iberia to Scandinavia, down to Germany and central Alps), Ukraine, and Crimea. ...
... A third Southern Iberian (or Western, cf. Randi et al. 2004) lineage comprises the roe European Roe Deer Capreolus capreolus (Linnaeus, 1758) deer populations from central-southern Spain and Portugal, that is, the south-western limit of the species' distribution, where the subspecies C. c. garganta was proposed (see below). A fourth lineage, or alternatively a subclade within the Central main lineage (cf. ...
Chapter
This comprehensive species-specific chapter covers all aspects of the mammalian biology, including paleontology, physiology, genetics, reproduction and development, ecology, habitat, diet, mortality, and behavior. The economic significance and management of mammals and future challenges for research and conservation are addressed as well. The chapter includes a distribution map, a photograph of the animal, and a list of key literature.
... During the Last Glacial Maximum (LGM, 26-19 Ka BP;Clark et al. 2009), roe deer survived in a large refugial area stretching from the Iberian Peninsula and southern France, to the Apennine Peninsula and the northern parts of Italy, the Balkan and the Carpathian regions, the northern shores of the Black Sea, as far as the Caucasus Mts. (Barros et al. 2020;Lorenzini et al. 2014;Plis et al. 2022;Randi et al. 2004;Sommer et al. 2009). After the LGM roe deer spread from refugia to most European countries and even to the northern parts of the continent, including the then existing Doggerland-a land bridge between the British Isles, Scandinavia, and mainland Europe. ...
... After the LGM roe deer spread from refugia to most European countries and even to the northern parts of the continent, including the then existing Doggerland-a land bridge between the British Isles, Scandinavia, and mainland Europe. In recent centuries roe deer have been affected by hunting, which probably caused extinctions of local populations (e.g., in southern parts of the British Isles in the sixteenth century: Baker and Hoelzel 2013; in Scandinavia in the nineteenth century: Randi et al. 2004). The local extinctions of Finnish roe deer during the Little Ice Ages of the seventeenth and eighteenth centuries suggest that climate fluctuations continued to impact demography of the species throughout the Holocene (Pulliainen 1980). ...
... Although roe deer is widely distributed species in Europe, not many studies have evaluated its genetic diversity on a range-wide scale. Previous analyses of incomplete mitochondrial DNA (D-loop or cytochrome b) revealed three clades among the European roe deer (Lorenzini et al. 2014;Plis et al. 2022;Randi et al. 2004) with a further subdivision into haplogroups, often geographically separated (Baker and Hoelzel 2014;Barros et al. 2020;Gentile et al. 2009;Mucci et al. 2012;Plis et al. 2022;Tsaparis et al. 2019). Studies which included the eastern part of the continent, revealed a wide area, where hybridization between Siberian and European roe deer led to introgression of the Siberian mtDNA into European roe deer populations (Lorenzini et al. 2014;Markov et al. 2016;Matosiuk et al. 2014;Olano-Marin et al. 2014;Zvychaynaya et al. 2013). ...
Article
The European roe deer (Capreolus capreolus) is one of the most numerous and widespread ungulate species in Europe, which has complicated the assessment of its genetic diversity on a range-wide scale. In this study, we present the mitochondrial DNA control region (mtDNA CR) genetic diversity and population structure of roe deer in Europe based on the analyses of 3010 samples, which were described as European roe deer individuals. Our analyses revealed two main diversity hotspots, namely Eastern and Central Europe. We proposed that these hotspots result from the Siberian roe deer (C. pygargus) mtDNA introgression and the secondary contact of mtDNA clades, respectively. Significantly lower values of genetic diversity (nucleotide and haplotype diversity) were recorded in the peripheral areas of the species' range, including the southernmost parts of the Last Glacial Maximum (LGM) refugial areas. Roe deer population in Europe consists of 2-3 genetic groups according to SAMOVA, and 15-16 clusters identified by GENELAND. The main driver of roe deer population structure in the eastern parts of the continent has been introgression of mtDNA of C. pygargus. Spatial genetic analyses revealed a complex structure of roe deer on a pan-European scale, which presumably results from post-glacial recolonization of the continent from various parts of a large LGM refugial area by different roe deer mtDNA clades and haplogroups.
... So far, phylogenetic studies on European roe deer covered a large part of their contemporary range and indicated three main mitochondrial DNA (mtDNA) clades-Central, Eastern, and Western (Baker & Hoelzel, 2014;Lorenzini et al., 2014;Randi et al., 2004;Tsaparis et al., 2019). The Central clade is the most common throughout Europe, the Eastern one is restricted mainly to the Balkans, and the Western clade occurs only in the Iberian Peninsula (Gentile et al., 2009;Lorenzini et al., 2014;Mucci et al., 2012;Randi et al., 2004). ...
... So far, phylogenetic studies on European roe deer covered a large part of their contemporary range and indicated three main mitochondrial DNA (mtDNA) clades-Central, Eastern, and Western (Baker & Hoelzel, 2014;Lorenzini et al., 2014;Randi et al., 2004;Tsaparis et al., 2019). The Central clade is the most common throughout Europe, the Eastern one is restricted mainly to the Balkans, and the Western clade occurs only in the Iberian Peninsula (Gentile et al., 2009;Lorenzini et al., 2014;Mucci et al., 2012;Randi et al., 2004). ...
... There were no roe deer studies done in northern (Finland), eastern (Belarus, Estonia), central (Slovakia, Czech Republic), and southeastern (Croatia, Slovenia) parts of the continent. The most comprehensive phylogeographical analyses, done by Lorenzini et al. (2014) and Randi et al. (2004) focused mostly on identifying the origin of mtDNA clades and on disentangling the relationships among them. So far, there had been no attempts to determine the location of the contact zones between them. ...
Article
Full-text available
To provide the most comprehensive picture of species phylogeny and phylogeography of European roe deer (Capreolus capreolus), we analyzed mtDNA control region (610 bp) of 1469 samples of roe deer from Central and Eastern Europe and included into the analyses additional 1541 mtDNA sequences from GenBank from other regions of the continent. We detected two mtDNA lineages of the species: European and Siberian (an introgression of C. pygargus mtDNA into C. capreolus). The Siberian lineage was most frequent in the eastern part of the continent and declined toward Central Europe. The European lineage contained three clades (Central, Eastern, and Western) composed of several haplogroups, many of which were separated in space. The Western clade appeared to have a discontinuous range from Portugal to Russia. Most of the haplogroups in the Central and the Eastern clades were under expansion during the Weichselian glacial period before the Last Glacial Maximum (LGM), while the expansion time of the Western clade overlapped with the Eemian interglacial. The high genetic diversity of extant roe deer is the result of their survival during the LGM probably in a large, contiguous range spanning from the Iberian Peninsula to the Caucasus Mts and in two northern refugia.
... So far, phylogenetic studies on European roe deer covered a large part of their contemporary range and indicated three main mitochondrial DNA (mtDNA) clades-Central, Eastern, and Western (Baker & Hoelzel, 2014;Lorenzini et al., 2014;Randi et al., 2004;Tsaparis et al., 2019). The Central clade is the most common throughout Europe, the Eastern one is restricted mainly to the Balkans, and the Western clade occurs only in the Iberian Peninsula (Gentile et al., 2009;Lorenzini et al., 2014;Mucci et al., 2012;Randi et al., 2004). ...
... So far, phylogenetic studies on European roe deer covered a large part of their contemporary range and indicated three main mitochondrial DNA (mtDNA) clades-Central, Eastern, and Western (Baker & Hoelzel, 2014;Lorenzini et al., 2014;Randi et al., 2004;Tsaparis et al., 2019). The Central clade is the most common throughout Europe, the Eastern one is restricted mainly to the Balkans, and the Western clade occurs only in the Iberian Peninsula (Gentile et al., 2009;Lorenzini et al., 2014;Mucci et al., 2012;Randi et al., 2004). ...
... There were no roe deer studies done in northern (Finland), eastern (Belarus, Estonia), central (Slovakia, Czech Republic), and southeastern (Croatia, Slovenia) parts of the continent. The most comprehensive phylogeographical analyses, done by Lorenzini et al. (2014) and Randi et al. (2004) focused mostly on identifying the origin of mtDNA clades and on disentangling the relationships among them. So far, there had been no attempts to determine the location of the contact zones between them. ...
Article
Full-text available
To provide the most comprehensive picture of species phylogeny and phylogeography of European roe deer (Capreolus capreolus), we analyzed mtDNA control region (610 bp) of 1469 samples of roe deer from Central and Eastern Europe and included into the analyses additional 1541 mtDNA sequences from GenBank from other regions of the continent. We detected two mtDNA lineages of the species: European and Siberian (an introgression of C. pygargus mtDNA into C. capreolus). The Siberian lineage was most frequent in the eastern part of the continent and declined toward Central Europe. The European lineage contained three clades (Central, Eastern, and Western) composed of several haplogroups, many of which were separated in space. The Western clade appeared to have a discontinuous range from Portugal to Russia. Most of the haplogroups in the Central and the Eastern clades were under expansion during the Weichselian glacial period before the Last Glacial Maximum (LGM), while the expansion time of the Western clade overlapped with the Eemian interglacial. The high genetic diversity of extant roe deer is the result of their survival during the LGM probably in a large, contiguous range spanning from the Iberian Peninsula to the Caucasus Mts and in two northern refugia.
... Below species level, the roe deer harbors an overall high genetic diversity, as revealed by biochemical and molecular results (c.f. Hartl et al. 1991;Randi et al. 2004;Lorenzini and Lovari 2006). Molecular information, in particular, told us much about historical dynamics and contemporary demography of populations, their connectivity, as well as the uptake of interactions with the environment, within a so-called landscape genetics framework, that is, the integration of landscape ecology and population genetics. ...
... Recently, molecular phylogeographic investigations revealed that the European roe deer harbors high genetic variation and a complex population structure across the entire range. Data from different mitochondrial and nuclear DNA-based studies (Randi et al. 2004;Lorenzini and Lovari 2006;Lorenzini et al. 2014) concordantly suggest the existence of four major genetic lineages of roe deer in Europe (Fig. 3). One central lineage of mitochondrial haplotypes is widely distributed across the whole European continent (from northern Iberia to Scandinavia, down to Germany and central Alps), Ukraine, and Crimea. ...
... A third Southern Iberian (or Western, cf. Randi et al. 2004) lineage comprises the roe European Roe Deer Capreolus capreolus (Linnaeus, 1758) deer populations from central-southern Spain and Portugal, that is, the south-western limit of the species' distribution, where the subspecies C. c. garganta was proposed (see below). A fourth lineage, or alternatively a subclade within the Central main lineage (cf. ...
... Recently, detailed research has been carried out into the species' phylogeography (Vernesi et al., 2002;Randi et al., 2004;Lorenzini and Lovari, 2006;Royo et al., 2007;reviewed and discussed in Sommer et al., 2009). Combined patterns of genetic and palaeontological data suggest several regions in Iberia, southern France, Italy and the Balkans as well as the Carpathians and/or Eastern Europe as Pleistocene glacial refuge areas (Sommer et al., 2009). ...
... With regard to intraspecific systematics, the European roe deer is most often regarded as a monotypic species throughout Europe but recently attention has been focused on the Iberian and Italian roe deer where different studies yielded unequivocal evidence of substantial genetic substructuring (Lorenzini et al. 2002Randi et al., 2004;Lorenzini and Lovari, 2006;Royo et al., 2007). In Italy, the central and southern indigenous populations are genetically distinct irrespective of the molecular marker analysed, which is in line with their sometimes being classified as C. c. italicus. ...
Chapter
This book considers a number of problems posed by ungulates and their management in Europe. Through a synthesis of the underlying biology and a comparison of the management techniques adopted in different countries, the book explores which management approaches seem effective - and in which circumstances. Experts in a number of different areas of applied wildlife biology review various management problems and alternative solutions, including the impact of large ungulates on agriculture, forestry and conservation habitats, the impact of disease and predation on ungulate populations and the involvement of ungulates in road traffic accidents and possible measures for mitigation. This book is directed at practising wildlife managers, those involved in research to improve methods of wildlife management, and policy-makers in local, regional and national administrations.
... garganta (Randi et al., 2004;Groves & Grubb, 2011). ...
Thesis
Le mouvement individuel agit directement sur la valeur adaptative d’un animal en impactant sa survie et sonsuccès reproducteur. Il participe de façon importante à moduler la dynamique des populations et lefonctionnement des écosystèmes. Il a été montré que la variation spatiotemporelle des ressources et durisque conduit un animal à se déplacer entre patchs, pour accéder à des ressources disponibles en évitant lesrisques. En parallèle, il a été montré que la fidélité spatiale apporte de nombreux bénéfices aux individus,tels que de pouvoir mémoriser la localisation et la disponibilité des ressources et les exploiter de façonoptimale, pouvoir effectuer une trajectoire plus efficace et ainsi échapper plus facilement aux prédateurs.Ces bénéfices les amènent à occuper un domaine vital stable pendant au moins une saison. Cependant, ladistribution des ressources dans l’environnement n’est pas stable tout au long d’une saison. Cette étude viseà comprendre les mécanismes de mouvement à l’échelle intra-saisonnière, en analysant des trajectoiresprovenant de localisations GPS de 251 chevreuils et de 113 cerfs à travers la quasi-totalité de leur aire derépartition en Europe, puis en focalisant les analyses sur 105 individus d’une populations de chevreuils dansun paysage hétérogène. J’ai pu mettre en évidence que les deux espèces ne sont pas réellement sédentaires etn’occupent pas un domaine vital stable durant une saison donnée, mais adoptent une tactique de mouvementintra-saisonnière basée sur l’occupation de plusieurs domaines vitaux sous-saisonniers (sfHR), appeléetactique de multi-range. Les individus sont capables d’ajuster cette tactique, en termes de nombre et desurface des sfHR, ainsi que de distance parcourue entre sfHR, pour répondre au mieux à la variationspatiotemporelle de la distribution des ressources. Mon travail de recherche offre de nouvelles perspectivesdans l’étude du mouvement animal, notamment puisque les variations dans la tactique de multi-range sontliées aux changements ou à l’hétérogénéité de l’environnement. Cela pourrait ainsi permettre aux espècesadoptant cette tactique de faire face aux changements globaux.Mots clef
... Mediterranean peninsulas (Iberia, Balkan peninsula, and Italy) acted as glacial refugia for many European taxa and still represent interesting concentrations of genetic diversity and unicity, with the highest percentage of endemic taxa in Europe [1][2][3][4]. Divergence of genetic lineages in Mediterranean peninsulas is apparent even for large sized species (e.g., the roe deer Capreolus capreolus, the brown bear Ursus arctos, and the grey wolf Canis lupus [5][6][7]), thus being much stronger and more definite in small mammals (Rodentia and Eulipotyphla), which remained isolated for long times from other European conspecific populations [2]. Accordingly, the majority of endemic mammal species in Italy (i.e., 8 out of 10 endemic species [2,3]) belongs to the group of "small mammals" [8]. ...
Article
Full-text available
The Italian peninsula represented one of the main glacial refugia during climatic oscillations of the Pleistocene, currently being a biodiversity hotspot. In this work, we analysed for the first time the genetic diversity of harvest mouse populations in Italy, and we compared them with those of the rest of Eurasia. Mitochondrial cytochrome-b gene was amplified from 12 samples from throughout the Italian range. We recorded a very low genetic diversity, in line with the rest of the harvest mouse range. In the comparative phylogenetic tree, Northern Italy samples clustered together as a sister group of the rest of Europe, whereas those from Central Italy clustered with Central Europe samples. Harvest mice have recently conquered Southern Europe, i.e., possibly at the start of the Holocene. The global genetic homogeneity might be due to accidental human-mediated introductions or to the sharp decline of the habitat of the harvest mouse, which may in turn have caused severe bottlenecks in the populations of this small rodent.
... Additional studies found "no significant relationship between population size and levels of heterozygosity" (Bezemer et al., 2019). The fact that the distribution of expected GD is not what one finds in nature has been termed "Lewontin's paradox" (Buffalo, 2021;Ellegren & Galtier, 2016) and is increasingly being considered in empirical studies (Pearse et al., 2006;Poissant et al., 2005;Randi et al., 2004;Valente et al., 2017) and theoretical models (Brandvain & Wright, 2016;Carroll et al., 2019;Evans et al., 2007;Kramer & van der Werf, 2010). The importance of population size and isolation ("insularity") to GD thus remains uncertain and variable in nature. ...
Article
Full-text available
We conducted a quantitative literature review of genetic diversity (GD) within and among populations in relation to categorical population size and isolation (together referred to as “insularity”). Using populations from within the same studies, we were able to control for between‐study variation in methodology, as well as demographic and life histories of focal species. Contrary to typical expectations, insularity had relatively minor effects on GD within and among populations, which points to the more important role of other factors in shaping evolutionary processes. Such effects of insularity were sometimes seen—particularly in study systems where GD was already high overall. That is, insularity influenced GD in a study system when GD was high even in non‐insular populations of the same study system—suggesting an important role for the “scope” of influences on GD. These conclusions were more robust for within population GD versus among population GD, although several biases might underlie this difference. Overall, our findings indicate that population‐level genetic assumptions need to be tested rather than assumed in nature, particularly for topics underlying current conservation management practices. We conducted a quantitative literature review of genetic diversity (GD) within and among populations in relation to categorical population size and isolation (together referred to as “insularity”). Contradictory to typical expectations, we found that insularity had relatively minor effects on GD within and among populations, which points to the more important roles of other factors in shaping evolutionary processes. Our findings indicate that population‐level genetic assumptions need to be tested rather than assumed in nature, particularly for topics underlying current conservation management practices.
Article
Full-text available
In 2021, the International Union for Conservation of Nature (IUCN) introduced a novel method for assessing species recovery and conservation impact: the IUCN Green Status of Species. The Green Status standardizes recovery using a metric called the Green Score, which ranges from 0% to 100%. This study focuses on one crucial step in the Green Status method—the division of a species’ range into so-called “spatial units”—and evaluates whether different approaches for delineating spatial units affect the outcome of the assessment (i.e., the Green Score). We compared Green Scores generated using biologically based spatial units (the recommended method) to Green Scores generated using ecologically based or country-based spatial units for 29 species of birds and mammals in Europe. We found that while spatial units delineated using ecoregions and countries (fine-scale) produced greater average numbers of spatial units and significantly lower average Green Scores than biologically based spatial units, coarse-scale spatial units delineated using biomes and countries above a range proportion threshold did not differ significantly from biologically based results for average spatial unit number or average Green Score. However, case studies focusing on results for individual species (rather than a group average) showed that, depending on characteristics of the species’ distribution, even these coarse-scale delineations of ecological or country spatial units often over- or under-predict the Green Score compared to biologically based spatial units. We discuss cases in which the use of ecologically based or country-based spatial units is recommended or discouraged, in hopes that our results will strengthen the new Green Status framework and ensure consistency in application.
Article
Full-text available
In recent years, much effort has been devoted to defining the attractive goal of sustainable forest management by means of principles, criteria and indicators in the attempt to overcome the difficulties in putting it into practise. Results pertaining to principles and generic features are broadly similar on the international scale and provide a good starting point for establishing practical standards. However, the similarity appears to apply only to the conceptual level since any attempt to define forest management problems in concrete terms tailored to the particular attributes always leads to defining and developing new, specific criteria and indicators. The purpose of this article is to describe the path followed in Italy to define sustainable management. An outline of the structural attributes of Italian forests is provided, followed by a description of the procedures that led to the definition of criteria and indicators. The main initiatives implemented are compared so as to identify critical aspects and results that are helpful in defining the criteria and indicators for forests and a number of methodological conclusions are then drawn.
Article
The recently-developed statistical method known as the "bootstrap" can be used to place confidence intervals on phylogenies. It involves resampling points from one's own data, with replacement, to create a series of bootstrap samples of the same size as the original data. Each of these is analyzed, and the variation among the resulting estimates taken to indicate the size of the error involved in making estimates from the original data. In the case of phylogenies, it is argued that the proper method of resampling is to keep all of the original species while sampling characters with replacement, under the assumption that the characters have been independently drawn by the systematist and have evolved independently. Majority-rule consensus trees can be used to construct a phylogeny showing all of the inferred monophyletic groups that occurred in a majority of the bootstrap samples. If a group shows up 95% of the time or more, the evidence for it is taken to be statistically significant. Existing computer programs can be used to analyze different bootstrap samples by using weights on the characters, the weight of a character being how many times it was drawn in bootstrap sampling. When all characters are perfectly compatible, as envisioned by Hennig, bootstrap sampling becomes unnecessary; the bootstrap method would show significant evidence for a group if it is defined by three or more characters.
Article
Episodes of population growth and decline leave characteristic signatures in the distribution of nucleotide (or restriction) site differences between pairs of individuals. These signatures appear in histograms showing the relative frequencies of pairs of individuals who differ by i sites, where i = 0, 1, .... In this distribution an episode of growth generates a wave that travels to the right, traversing 1 unit of the horizontal axis in each 1/2u generations, where u is the mutation rate. The smaller the initial population, the steeper will be the leading face of the wave. The larger the increase in population size, the smaller will be the distribution's vertical intercept. The implications of continued exponential growth are indistinguishable from those of a sudden burst of population growth Bottlenecks in population size also generate waves similar to those produced by a sudden expansion, but with elevated uppertail probabilities. Reductions in population size initially generate L-shaped distributions with high probability of identity, but these converge rapidly to a new equilibrium. In equilibrium populations the theoretical curves are free of waves. However, computer simulations of such populations generate empirical distributions with many peaks and little resemblance to the theory. On the other hand, agreement is better in the transient (nonequilibrium) case, where simulated empirical distributions typically exhibit waves very similar to those predicted by theory. Thus, waves in empirical distributions may be rich in information about the history of population dynamics.
Article
We have amplified and sequenced 679 nucleotides of the mitochondrial DNA control-region in 45 Siberian (Capreolus pygargus) and European (C. capreolus) roe deer from two localities in Russia and seven in Italy. Average interspecific sequence divergence was 4.9%. Six different haplotypes were found in Siberian roe deer, and 14 haplotypes in Alpine European roe deer. A population of the endemic Italian subspecies C. c, italicus was monomorphic bearing a single haplotype with one unique nucleotide deletion and a fixed transversion. Phylogenetic relationships among haplotypes indicated that the two species were separated with 100% bootstrap support, and there were two distinct population clusters within each species. These clusters correspond to different geographical locations of the samples: Siberian roe deer were subdivided into west Siberia (Kurgan region) and east Siberia (Amur region), and European roe deer were subdivided into an eastern and a western Alpine group. Average sequence divergence among conspecific populations was 1.2%. Calibrations of evolutionary rates of the different domains of the control-region suggest that Siberian and European roe deer speciated about 2-3 million years ago, and haplotype diversity within species was generated during the last 150 000-370 000 years. Geographical structuring of sequence variability in roe deer allows us to identify historical and recent intraspecific population differences, including the effects of human disturbance. The genetic peculiarities of the endemic Italian subspecies C. c. italicus call for careful conservation of its surviving populations.
Chapter
Changes in population characteristics are usually thought to result from demographic processes or responses to habitat fluctuations with no consideration of possible genetic factors. Populations are often considered to be genetically homogeneous over short distances, invariant over time periods normally considered in population studies, and in genetic equilibrium among different sex-age groups. However, wildlife populations are genetically structured in space, and often have significant differences in gene frequencies over distances that can be easily traversed by individual animals. In addition, genetic characteristics of populations often change annually, and differ among animals of various ages and sexes. For example, white-tailed deer (Odocoileus virginianus) can show significant differences in gene frequencies between animals taken over distances as little as 5 km, between sexes, and among fetuses, fawns, and older animals. Thus, wildlife populations are spatially structured, temporally dynamic, and genetically diverse units. These characteristics have implications for the design of management programs and the interpretation of changes or differences in biological resources over space or time. Correlations of functionally important characteristics with an animal’s genotype, or multilocus heterozygosity, are often observed under field conditions. These correlations suggest that certain types of changes in wildlife populations may be at least partially due to genetic causes. This hypothesis, as well as the general need to conserve the genetic variability of wildlife populations, suggests the necessity of incorporating genetic considerations into management plans and of evaluating genetic explanations for observed changes in wildlife populations. Currently available techniques allow the collection and use of genetic information for natural wildlife populations. We discuss examples of how to use genetic information in wildlife management programs.