Advertisement

European Journal of Forest Research

, Volume 131, Issue 4, pp 1127–1138 | Cite as

Genetic variation and divergence in Scots pine (Pinus sylvestris L.) within its natural range in Italy

  • P. Belletti
  • D. Ferrazzini
  • A. Piotti
  • I. Monteleone
  • F. Ducci
Original Paper

Abstract

Twenty-one populations of Scots pine sampled over the entire Italian range of the species were analysed for genetic variation scored at nine nuclear SSR markers. The main aim of the work was to find genetic features useful for conservation management, namely allelic composition, gene diversity and differentiation. High levels of intra-population variability were scored. The only population sampled in the Apennines gave the lowest values, confirming the genetic erosion undergone in the Scots pine remnants in this area. A low level of genetic variability was also scored for populations from the Po valley and hills of Piedmont. Most genetic diversity was found within populations, while only a small amount occurred among them (F ST = 0.058). Both Bayesian clustering and sPCA analysis showed a East–West subdivision, notwithstanding the unclear position of populations from the Po valley. The population from the Apennines was always clearly separated from the others. The results are discussed in terms of post-glacial recolonisation, as well as for defining genetically homogeneous regions for Scots pine in Italy. The management of genetic resources could benefit from the identification of such ‘gene zones’, thereby avoiding the use of non-local reproductive material for plantations, which can represent one of the most important reasons for failure of reforestation. In addition, the assessment of the biogeographic genetic structure by neutral markers is a prerequisite for disentangling the influence of selectively neutral and non-neutral processes on the distribution of adaptive genetic variability.

Keywords

Genetic differentiation Genetic variation Glacial refugia Regions of provenance Scots pine SSR markers 

Introduction

Genetic erosion is one of the most serious threats for the survival of forest ecosystems worldwide. A high level of variability is essential to supply populations with strong adaptability. This is particularly important for forest species, which consist of individuals with long life cycles and no possibility of migrating to more favourable sites (Palmberg-Lerche 2001; Toro and Caballero 2005). Genetic erosion can be strongly enhanced by habitat fragmentation and marginality in the species’ range (Jump and Peñuelas 2006; Eckert et al. 2008). However, the expected ‘genetic signal’ depending on increasing isolation and decreasing population size is often undetected, as fragmentation may have occurred relatively recently and fragments may contain large remnant populations, and the longevity of many tree species can delay the loss of genetic variability (Kramer et al. 2008). In addition, the historically underestimated long-distance dispersal capability of forest trees can also maintain a high connectivity among widely isolated stands (Robledo-Arnuncio and Gil 2005; Nathan 2006; Williams 2010). Conversely, climate warming can strengthen the loss of genetic variability due to fragmentation, causing a marked decline in growth and survival of marginal southern populations of temperate tree species, as demonstrated in European Pinus sylvestris populations by Rubiales et al. (2008) and Reich and Oleksyn (2008).

Knowledge of the level and distribution of genetic variation is of the utmost importance in providing information for the conservation and the management of genetic resources. Furthermore, genetic analysis based on molecular markers can increase our understanding of the historical processes that led to the present distribution of a species (Petit et al. 2003). These markers can provide us with appropriate means to obtain information on the genetic structure of populations, as well as to analyse the distribution of within-species variability (Pautasso 2009). Such data provide important insights into preservation and restoration programmes, indicating areas of high genetic diversity and geographic limits for seed collection and delimiting the scale at which conservation should be planned (Escudero et al. 2003).

Scots pine (P. sylvestris L.) is the most widespread European conifer tree, and its natural range extends from the Arctic Circle in Scandinavia down to central Spain and central Italy and from western Scotland to eastern Siberia. In Southern Europe and Asia Minor, isolated occurrences are confined to the mountain zone (up to 2,200 m in altitude in the Balkans and Spain and 2,700 m in the Caucasus). In Italy, the species is widely spread throughout the Alps, and some relic populations can be found in the northern Apennines, in the hilly areas of Piedmont (north-western Italy) and in the upper part of the Po valley (Pignatti 1982). The present distribution is highly influenced by human activities. In the Alps, Scots pine has often been substituted by other species, namely Norway spruce and black pine, while the other populations are currently regressing, mainly due to recruitment limitation, competition with other forest trees and as a consequence of rural depopulation and the abandonment of woodland management (Camerano et al. 2008).

The complex biogeographic history of Scots pine in Europe has been extensively studied (Sinclair et al. 1999; Soranzo et al. 2000; Cheddadi et al. 2006). Recent work has shown the presence of a previously unknown glacial refugium in Northern Europe during the last glaciation (Naydenov et al. 2007; Pyhäjärvi et al. 2008) and of several small putative refugial areas in Southern Europe that gave rise to geographically limited ‘interglacial refugia’ (Cheddadi et al. 2006). Among the southernmost populations, those in Spain, Balkan and Turkish have usually been considered in broad-scale phylogeographic studies (Sinclair et al. 1999; Soranzo et al. 2000; Naydenov et al. 2007; Pyhäjärvi et al. 2008), whereas Alpine and Apennine populations have received less attention (Scalfi et al. 2009). In the few studies where they were extensively sampled, it was demonstrated that Italian populations shared a common mitochondrial haplotype and that they are different from the surrounding Austrian, Swiss and French alpine populations (Cheddadi et al. 2006). Cheddadi et al. therefore hypothesised a common origin for Italian Scots pine populations from a refugial area in southern Italy, even though the results by Puglisi and Attolico (2000), Labra et al. (2006) and Scalfi et al. (2009), obtained with more polymorphic genetic markers (respectively allozymes, ISSRs and SSRs), showed a marked differentiation between Alpine and Apennine populations, suggesting different recolonisation routes for the two mountain chains. The preservation of genetic resources of highly fragmented Apennine populations therefore appears to be a high priority.

At the population level, genetic structure and gene flow patterns have been studied in Spanish populations from the northern Meseta by Robledo-Arnuncio and Gil (2005) and Robledo-Arnuncio et al. (2005), and show high within-population genetic diversity and extremely high levels of pollen flow over long distances (5% longer than 30 km). On the other hand, small relic Scots pine populations from the Apennines had significantly lower genetic diversity, and are differentiated from those in the Alps, possibly as a consequence of progressive isolation since the early Holocene and their origin from different glacial refugia (Labra et al. 2006; Scalfi et al. 2009). Scalfi et al. (2009), however, hypothesised a possible different role of gene flow via-pollen and seed in determining this genetic differentiation.

In this study, we surveyed the genetic variability of 21 Scots pine populations throughout the species distribution in the Italian peninsula, using nine highly informative nuclear microsatellite (nSSR) markers. Our main aim was to assess the levels and distribution of genetic variability of this conifer by intensively sampling the entire Italian range of the species and to investigate the presence of any cryptic genetic structure shaped by postglacial recolonisation which went undetected in previous studies based on organellar markers. We also discussed our results in the light of the European Directive 105/1999, with particular emphasis on the preservation and restoration of Scots pine genetic resources in the Alps and Apennines.

Materials and methods

Plant material

Twenty-one native populations of Scots pine were chosen within the natural distribution range of the species in Italy (Fig. 1). Four of them (BOS, VEZ, PAS and CAS) are located in the hilly areas of Piedmont, two (TIC and OLG) grow in the upper part of Po valley and another (CRO) is found in the northern Apennines. All the others are distributed along the entire Italian Alpine region. Table 1 summarises names and locations of the populations analysed. Most of the populations occur as mixed forests, with the exception of populations CAR, FEN and SAV which are pure stands of Scots pine. In other locations accompanying species vary according to altitude and latitude: English oak, hornbeam and wild cherry (TIC), pubescent oak, flowering ash and juniper (CRO), sessile oak and European ash (OLG, MAS, GAR), sessile oak, black locust, elm and maples (BOS, VEZ, PAS, CAS), black pine (DOG), Norway spruce (VAL, SIU, BRU, COR and CLA), silver fir (TOC), larch (SAR) and mountain pine (BRI). Among the populations sampled, CRO, CAR, FEN, OLG, VAL, SIU and BRU are registered in the Italian National Book of Seed Stands for Scots pine, selected for their phenotypic characteristics and health status (Morandini and Magini 1975).
Fig. 1

Geographical distribution of the 21 populations of Scots pine analysed in this study

Table 1

Details of site characteristics of Scots pine populations from Italy which were sampled for the study

Code

Population

Geographical region

Location

Average elevation (m a.s.l.)

Annual (and summer) rainfall (mm)

CRO

Vezzano sul Crostolo

Apennines

44°31′N, 10°31′E

450

945 (174)

CAR

Carpe

Western Alps

43°55′N, 7°47′E

1,000

1,100 (260)

BRI

Carnino Briga Alta

Western Alps

44°18′N, 7°43′E

1,200

1,170 (378)

BOS

Bossolasco

Hills of Piedmont

44°32′N, 8°03′E

750

877 (145)

VEZ

Vezza d’Alba

Hills of Piedmont

44°45′N, 8°00′E

300

660 (125)

FEN

Fenestrelle

Western Alps

45°02′N, 7°03′E

1,450

900 (300)

SAV

Savoulx

Western Alps

45°05′N, 6°40′E

1,300

396 (102)

PAS

Passerano Marmorito

Hills of Piedmont

45°05′N, 8°05′E

225

554 (156)

CAS

Casalborgone

Hills of Piedmont

45°10′N, 8°00′E

500

673 (142)

SAR

Sarre

Western Alps

45°43′N, 7°15′E

1,300

698 (174)

TOC

Toceno

Western Alps

46°10′N, 8°32′E

1,100

698 (459)

TIC

Ticino

Po Valley

45°30′N, 8°39′E

250

1,020 (234)

OLG

Olgelasca

Po Valley

45°44′N, 9°11′E

350

1,635 (587)

MAS

Val Masino

Western Alps

46°09′N, 9°34′E

300

458 (165)

GAR

Valvestino Garda

Eastern Alps

45°45′N, 10°35′E

800

685 (263)

VAL

Valda

Eastern Alps

46°12′N, 11°16′E

950

750 (350)

SIU

Alpe di Siusi

Eastern Alps

46°33′N, 11°33′E

1,200

800 (400)

BRU

Brunico

Eastern Alps

46°48′N, 11°56′E

1,100

692 (360)

COR

Cortina

Eastern Alps

46°32′N, 12°18′E

1,200

1,100 (383)

CLA

Claut

Eastern Alps

46°16′N, 12°31′E

600

1,186 (436)

DOG

Val Dogna

Eastern Alps

46°26′N, 13°19′E

1,000

908 (365)

Data on annual and summer rainfall and on average annual temperature were inferred for a period of at least 5 years

Twenty-four adult non-adjacent trees were chosen at random in each population. Since some individuals did not show reliable nSSRs banding patterns, they were excluded from the analysis. The number of these individuals varied among populations: from zero up to six. Consequently, the total number of analysed individuals was 449. Needles collected from the trees were stored at −20°C until DNA extraction was carried out.

Molecular analysis

Frozen needles (100 mg of tissue) were powdered in liquid nitrogen, and genomic DNA was extracted using the QIAGEN® DNeasy plant mini kit, according to the manufacturer’s protocol. The concentration of each sample was adjusted to 20 ng/μl.

Twelve simple sequence repeat markers (SSR) were selected from the published literature and tested on our plant material (Table 2). While the SPAC and SPAG series consisted of primers specific for P. sylvestris (Soranzo et al. 1998), the PtTX series included primers originally designed for Pinus taeda, but they also proved to be as useful as the markers selected for P. sylvestris (Elsik et al. 2000; Auckland et al. 2002; González-Martinez et al. 2004).
Table 2

Descriptive statistics for the twelve microsatellite loci considered for the study

Locus

Repeat motif

Primer sequences (5′ → 3′)

Number of alleles

Molecular weight range (bp)

SPAC 11.4a

(AT)5(GT)19

TCACAAAACACGTGATTCACA

GAAAATAGCCCTGTGTGAGACA

38

130–170

SPAC 11.5a

(AT)8(GT)19-(TA)11

TGGAGTGGAAGTTTGAGAAGC

TTGGGTTACGATACAGACGATG

No amplification

SPAC 11.6a

(CA)29(TA)7

CTTCACAGGACTGATGTTCA

TTACAGCGGTTGGTAAATG

76

103–220

SPAC 11.8a

(TG)16

AGGGAGATCAATAGATCATGG

CAGCCAAGACATCAAAAATG

25

123–181

SPAC 12.5a

(GT)20(GA)10

CTTCTTCACTAGTTTCCTTTGG

TTGGTTATAGGCATAGATTGC

62

116–202

SPAG 3.7a

(TC)45

GTTAAAGAAAATAATGACGTCTC

AATACATTTACCTAGAATACGTCA

No scorable bands

SPAG 7.14a

(TG)17(AG)21

TTCGTAGGACTAAAAATGTGTG

CAAAGTGGATTTTGACCG

64

174–252

PtTX 3032b

(GAT)35(GAC)3GAT(GAC)8-(GAC)6AAT(GAT)6

CTGCCACACTACCAACC

AACATTAAGATCTCATTTCAA

151

254–572

PtTX 3107c

(CAT)14

AAACAAGCCCACATCGTCAATC

TCCCCTGGATCTGAGGA

22

144–175

PtTX 3116c

(TTG)7-(TTG)5

CCTCCCAAAGCCTAAAGAAT

CATACAAGGCCTTATCTTACAGAA

67

100–276

PtTX 4001d

(CA)15

CTATTTGAGTTAAGAAGGGAGTC

CTGTGGGTAGCATCATC

31

197–231

PtTX 4011d

(CA)20

GGTAACATTGGGAAAACACTCA

TTAACCATCTATGCCAATCACTT

21

230–284

aSoranzo et al. (1998)

bElsik et al. (2000)

cElsik and Williams (2001)

dZhou et al. (2002)

Polymerase chain reaction (PCR) amplifications were performed using a Perkin Elmer GeneAmp® PCR System 9600 thermal cycler. The protocol was slightly modified, according to the presence of a fluorochrome (IR-Dye 700 and IR-Dye 800) attached to each forward primer. Each amplification reaction contained 1X reaction buffer (Promega), 2.5 mM MgCl2, 0.2 mM dNTPs, 0.5 μM of each primer, 0.65 U of GoTaq DNA Polymerase (Promega), approximately 10 ng genomic DNA, and deionised water to a total volume of 13 μl.

The PCR conditions varied for different primers and were adjusted for the presence of the forward labelled ones: for P. sylvestris-specific primers, the conditions included an initial step of 3 min at 94°C, followed by five cycles of touchdown consisting of 94°C for 30 s, 65°C for 30 s Δ↓ 1°C (SPAC 11.4 and SPAC 11.5) or 60°C for 30 s Δ↓ 1°C (SPAC 11.8, SPAC 12.5, SPAG 3.7 and SPAG 7.14), 72°C for 1 min, and subsequent 25 cycles of amplification consisting of 94°C for 30 s, 60°C for 30 s, 72°C for 1 min and a final extension of 72°C for 10 min. Touchdown PCR was not necessary for SPAC 11.6, where the protocol consisted of a first step at 95°C for 5 min, and 35 cycles (denaturation at 94°C for 1.5 min, annealing at 55°C for 1.5 min and elongation at 72°C for 1.5 min), followed by 10 min at 72° C. For P. taeda primers, PCR profiles consisted of an initial step of 5 min at 94°C, followed by 20 cycles of touchdown [94°C for 1 min, 59°C for 30 s Δ↓ 0,5°C (PtTX 3032) or 55°C for 30 s Δ↓ 0,5°C (PtTX 3116) or 60°C for 30 s Δ↓ 0,5°C (PtTX 3107, PtTX 4001 and PtTX 4011), 72°C for 1 min], followed by 20 cycles of amplification at 94°C for 1 min, annealing temperature for 1 min, 72°C for 1 min, and a final extension of 72°C for 3 min.

The forward sequence of each primer pair was labelled with a fluorescent dye at its 5′ end: IR-Dye 800 for SPAC 11.6, SPAC 12.5, SPAG 7.14, PtTX 3107 and PtTX 4001, IR-Dye 700 for SPAC 11.4, SPAC 11.8, SPAC 11.5, SPAG 3.7, PtTX 3032, PtTX 3116 and PtTX 4011.

Electrophoresis and detection of PCR products were carried on a 6%, 25-cm-long, 0.25-mm-thick, denaturing polyacrylamide gel using a sequencer (model DNA 4200 Sequencer LI-COR Biotechnology). Gels were run at 2,000 V in TBE 1X buffer, for 1–3 h, depending on the product sizes. Two different size standards (50–350 and 50–700 bp) were also run on each gel to enable accurate scoring of individual bands. Data were collected by e-Seq software (DNA Sequencing and Analysis Software), and all the size scores were visually checked.

Data processing

Allele frequencies and within-population genetic diversity parameters (mean number of alleles per locus, A; mean number of private alleles per locus, P a; effective number of alleles per locus, N e; observed and expected heterozygosity, H O and H E, respectively) were estimated using GenAlEx v.6 software (Peakall and Smouse 2006). Allelic richness, based on a minimum sample size of 16 gene copies (A r16), was calculated using FSTAT (Goudet 1995). Genotypic disequilibrium between pairs of loci was tested at the single population level and across all populations, with Fisher’s exact test using Arlequin software (Excoffier et al. 2005).

Fisher’s exact test using the Markov Chain algorithm (Guo and Thompson 1992) was used to assess deviations from Hardy–Weinberg equilibrium for each population and each locus, and where significant deficiencies of heterozygotes from Hardy–Weinberg expectations were found, the presence of a relatively high frequency of null alleles was suspected (Pemberton et al. 1995). Loci with high frequencies of null alleles were identified by estimating null allele frequencies for each locus and each population, using the software Micro-Checker (Van Oosterhout et al. 2004). In further analysis, we eliminated problematical loci with high null allele frequency from the data set, using only loci with < 0.19 null allele frequencies. This value has been considered as a threshold over which significant underestimate of H E due to null alleles can be found (Chapuis et al. 2008). Where possible, we employed analyses that were robust in the presence of null alleles (Chapuis and Estoup 2007; Chapuis et al. 2008), including STRUCTURE (Pritchard et al. 2000; Falush et al. 2003), ordination methods (sPCA, Jombart et al. 2008), and the Mantel test on chord distance (DC, Cavalli-Sforza and Edwards 1967). In particular, the inbreeding coefficients (F IS) were calculated taking into account the estimated null allele frequencies using the program INEst and running the individual inbreeding model (IIM) with a Gibbs sampler of 105 iterations (Chybicki and Burczyk 2009).

FreeNA was used to compute the value of Weir’s (1996) estimators of F statistics to analyse the population genetic structure of the overall samples. In particular, F ST was calculated in order to estimate the proportion of the total genetic variation due to the differentiation among populations. Genetic differentiation between populations was estimated computing a pairwise F ST. FreeNA applies the ENA correction method to correct efficiently for the positive bias induced by the presence of null alleles on the F ST estimation (Chapuis and Estoup 2007).

The genetic structure of the populations was explored using Bayesian clustering and spatial principal components analysis (sPCA). Bayesian clustering was performed with the software STRUCTURE (Pritchard et al. 2000). The program uses a Markov chain Monte Carlo (MCMC) algorithm to cluster individuals into populations on the basis of multilocus genotypic data. Individual multilocus genotypes are first assigned probabilistically to genetic clusters (K) without considering sampling origins. Admixed or hybrid individuals can be identified as they will have a fraction of their alleles derived from each genetic cluster. The program was run setting a burn-in period of 105 followed by 5 × 105 iterations, and using the admixture ancestry model and the correlated allele frequency model, given the low F ST and the high genetic connectivity typical of forest tree populations. Posterior probabilities of K were calculated from the means of 20 runs for each value of K ε {1, … 10} and the optimum K determined using the method of Evanno et al. (2005).

The sPCA is a spatially explicit multivariate method recently developed by Jombart et al. (2008), to investigate the spatial pattern of genetic variability using allelic frequency data of individuals or populations. It takes spatial information directly into account as a component of the adjusted model, focusing on the part of the variability that is spatially structured. This analysis does not require data to meet the Hardy–Weinberg expectations or linkage equilibrium to exist between loci. The sPCA yields scores summarising both the genetic variability and the spatial structure among individuals (or populations). Global structures (patches, clines and intermediates) are disentangled from local ones (strong genetic differences between neighbours) and from random noise. Neighbouring sites were defined by building a connection network based on Delaunay triangulation. The existence of global and local structuring was tested using the multivariate Monte Carlo tests implemented, as the sPCA procedure, in the adegenet package for R (Jombart 2008; R Development Core Team 2009).

A Mantel (1967) test was applied to the matrices of pairwise chord distance and log-transformed geographical distance between populations (natural logarithm scale) to assess isolation-by-distance, namely the model under which genetic differentiation between populations is the result of drift. Chord distance (Cavalli-Sforza and Edwards 1967) for each pair of populations was calculated using the INA correction described in Chapuis and Estoup (2007) with FreeNA. The test of significance for Mantel test was carried out on 9,999 permutations of the data. Mantel test was executed on the entire data set and on clusters detected by STRUCTURE and sPCA analyses.

Finally, we used the program Bottleneck v.1.2.02 (Piry et al. 1999) to test for recent population bottlenecks. A Wilcoxon’s sign rank test was used to compare expected heterozygosity from Hardy–Weinberg equilibrium with predicted heterozygosity at mutation-drift equilibrium, on the basis of the observed allele number (Piry et al. 1999). The program was run under a two-phase model of mutation (TPM) that generally fits microsatellite evolution better than either pure stepwise or infinite allele models (Di Rienzo et al. 1994). One thousand simulations were performed for each sample based on a TPM consisting of 90% single-step mutations and 10% multistep changes.

Results

Allelic diversity of microsatellite loci

Ten out of the 12 tested microsatellites (SPAC 11.4, SPAC 11.6, SPAC 11.8, SPAC 12.5, SPAG 7.14, PtTX3032, PtTX3107, PtTX3116, PtTX4001 and PtTX4011) showed reliable banding patterns with clear and reproducible bands. On the contrary, SPAC 11.5 did not amplify in any sample, and SPAC 3.7 failed to amplify the DNA in the majority of samples. The latter two markers were therefore excluded from the analysis. SPAC 11.8 was also excluded due to the high frequency of possible null alleles (0.57). The presence of null alleles was also suspected for 6 out of the remaining 9 loci, with frequencies ranging from 0.07 (PtTX3032) to 0.19 (SPAC 11.6).

The nine selected microsatellites were highly polymorphic and generated a total of 532 alleles (range, 21–151), with a mean number of ~48 alleles per locus. It was also possible to detect 137 private alleles, that is, present only in one population. The population frequency of private alleles was on average 0.036 and ranged between 0.021 and 0.208. The distribution of these alleles among loci ranged from 43 (PtTX3032) to 1 (PtTX3107). PAS (17 alleles) was the population with the highest number of private alleles, followed by TOC (14 alleles) and COR (10 alleles), and three populations (BOS, CLA and GAR) presented only two private alleles each.

Genetic variation within populations

Statistics on the genetic diversity within populations are given in Table 3. A high level of intra-population variability was found, since on average, more than 14 alleles per locus were observed (A = 14.45). CRO (the only populations from the Apennines) showed the lowest value of genetic diversity (A = 10.4, N e = 5.4, A r16 = 6.781, and H E = 0.754). The populations belonging to the Po valley (TIC and OLG) also showed a low variability, as well as the populations from the hills of Piedmont (BOS, VEZ, PAS, and CAS), even though the genetic impoverishment was less pronounced. Except for these particular populations (which, in most cases, are small, isolated and/or on the edge of the main range), only BRI, GAR, and CLA showed a more or less marked reduction in genetic diversity. This pattern was confirmed for all genetic diversity parameters considered.
Table 3

Statistics of genetic variation within Scots pine populations at nine microsatellite loci

Population

A

Ne

Ar16

Pa

HO

HE

FIS

CRO

10.4

5.4

6.781

0.333

0.603

0.754

0.037

CAR

15.9

10.4

9.525

1

0.714

0.886

0.030

BRI

12.3

9.3

8.636

0.333

0.651

0.846

0.036

BOS

12.4

8.0

8.353

0.222

0.724

0.829

0.022

VEZ

14.6

9.4

8.939

0.778

0.670

0.852

0.032

FEN

16.0

10.2

9.391

0.556

0.664

0.845

0.052

SAV

18.4

12.2

10.003

1

0.665

0.877

0.023

PAS

14.1

10.1

9.830

1.889

0.576

0.874

0.049

CAS

13.9

8.7

8.601

0.778

0.645

0.858

0.031

SAR

16.4

10.9

9.641

0.889

0.698

0.876

0.029

TOC

17.6

11.0

9.637

1.556

0.641

0.850

0.025

TIC

12.9

6.9

7.713

0.778

0.703

0.822

0.016

OLG

11.2

7.2

8.049

0.444

0.621

0.811

0.047

MAS

16.3

10.6

9.363

0.889

0.703

0.855

0.049

GAR

13.1

8.2

8.398

0.222

0.661

0.808

0.028

VAL

15.3

10.6

9.793

0.333

0.772

0.872

0.022

SIU

15.7

9.7

9.137

0.889

0.676

0.854

0.028

BRU

16.0

9.8

9.326

0.444

0.630

0.865

0.021

COR

14.0

10.0

9.635

1.111

0.724

0.872

0.024

CLA

12.8

8.1

8.771

0.222

0.684

0.828

0.044

DOG

14.0

9.8

9.304

0.556

0.642

0.863

0.042

Overall mean

14.45 (0.46)

9.37 (0.41)

8.992

0.724

0.670 (0.012)

0.847 (0.007)

0.033 (0.001)

A, mean number of alleles per locus; N e, effective number of alleles per locus; A r16, allelic richness based on a minimum sample size of 16 gene copies; P a, mean number of private alleles per locus; H O, average observed heterozygosity; H E, average gene diversity or expected heterozygosity; F IS, average inbreeding coefficient, calculated taking into account the estimated null allele frequencies. Values in parenthesis are standard errors

Despite the fact that some of the investigated populations are characterised by a low genetic diversity, no evidence for recent bottlenecks was found. In fact, for all the populations, HW heterozygosity and expected gene diversity at mutation-drift equilibrium did not differ significantly.

The probability that two randomly sampled alleles in a given population were not the same was higher than 84% (mean H E = 0.847), whereas the observed heterozygosity (mean H O = 0.670) was lower than expected. The difference, that determines a significant positive value for mean inbreeding coefficient, is mainly due to non-random mating and null alleles. Recalculating the inbreeding coefficients, taking into account the frequencies of null alleles, we found that deviations from the Hardy–Weinberg equilibrium were low (F IS ranging from 0.052 in FEN to 0.016 in TIC, with a mean of 0.033). None of estimated F IS was significantly different from zero (confidence interval calculated through INEst estimates of F IS overlapped zero in all populations).

Genetic differentiation among populations

Most of the genetic diversity was found within populations, while a small amount of the variability occurred among populations (F ST = 0.058, CI: 0.037–0.081). The F ST values per locus ranged from 0.024 (SPAC 11.6) to 0.12 (PtTX 4011), and there were no obvious differences between the P. sylvestris and the P. taeda sets of markers. The genetic divergence between populations was further investigated by computing a pairwise F ST matrix. Multilocus F ST values varied between 0.015 (CAR and SAR) and 0.141 (CRO and VEZ). The population from the Apennines (CRO) was always clearly separated by the others. Almost all pairwise F ST values were significantly greater than zero, confirming the presence of a slight, although significant, amount of population structuring in Italian Scots pine (results not shown).

Following the method of Evanno et al. (2005), the Bayesian clustering results obtained with STRUCTURE indicate that K = 2 clusters represents the most likely representation of the overall genetic structure that we analysed (Fig. 2). We found that most individuals from Western populations (VEZ, PAS, CAS, SAV and TOC) clearly belong to cluster 2, whereas eastern populations and CRO (the Apennine population) are primarily composed by individuals from cluster 1. Some admixed populations were detected among both Western populations (FEN and SAR) and eastern populations (DOG, COR, SIU and VAL). Although populations belonging to the pedo-climatic region of the Po valley (TIC and OLG) are geographically closer to Western populations, their individuals are predominantly assigned to cluster 1. This East–West subdivision was clearly shown also by sPCA analysis (Fig. 3). The existence of such ‘global structure’ (sensu Jombart et al. 2008) was demonstrated using both Delaunay triangulation (t max = 0.0631, P < 0.05) and Gabriel graph (t max = 0.0672, P < 0.05) for building the connection network.
Fig. 2

STRUCTURE results. On the top, the relationship between K (number of inferred clusters) and Ln(K) and DeltaK, respectively. On the bottom, the probability of belonging to each of the two inferred clusters, according to the method by Evanno et al. (2005) for each of the individuals

Fig. 3

On the left, the geographic representation of the connection matrix based on the Delaunay triangulation. On the right, the geographic distribution of the first positive sPCA scores

The correlation between genetic diversity, expressed as Cavalli-Sforza and Edwards (1967) chord distance for pairs of populations, and the logarithm of distances expressed in km did not show the typical pattern of isolation by distance and did not suggest a strong relationship between the two factors. Mantel’s test was not significant when performed on the entire data set (P = 0.094), as well as when performed separately on the two clusters detected by sPCA analysis (P = 0.669 in eastern populations and P = 0.482 in Western populations excluding CRO).

Discussion and conclusion

The aim of this study was to assess the level and the distribution of genetic variation of Scots pine throughout its natural range in Italy, in order to get fundamental knowledge that can be applied for plant propagation and genetic resources conservation. The results could be slightly biased due to the limited number of individuals sampled per populations. Although a larger database (more than 50 individuals per population) would be preferred to obtain stronger data on genetic variation at the level of polymorphic loci (Nei 1978), phylogeographic studies using nSSR markers often are based on less than 30 individuals per populations (e.g. Williams et al. 2007; Ferrazzini et al. 2008; Bagnoli et al. 2009; Scalfi et al. 2009; Bai et al. 2010). Kalinowski (2005) demonstrated that some genetic distances, such as F ST, showed limited sampling variance at loci characterised by a high mutation. He also showed that increasing the number of loci, instead of increasing the number of individuals, is an effective way to improve the precision of measures of genetic differentiation. Miyamoto et al. (2008) showed by resampling simulations that an accurate estimate of genetic diversity (H E) can also be achieved with small samples (less than 30 plants per population) genotyped at nSSRs. On the other hand, larger sample sizes are needed in order to obtain more accurate estimates of allelic richness, although the statistical technique of rarefaction can compensate for sampling disparities and allow for meaningful comparisons among populations. In general, in the present paper, we used population genetic indexes and techniques that minimise possible estimation biases caused by small sample size.

The populations analysed showed a considerable amount of genetic diversity, as estimated by means of variation scored at nine nuclear microsatellite loci. The observed number of alleles per locus (A) ranged from 10.4 to 18.4, with an average per population of 14.45, and the average gene diversity (H E) was as high as 0.847. The high degree of observed diversity is not surprising since it has been recognised for a long time as a peculiar characteristic of woody plants (Hamrick et al. 1992). Furthermore, species such as Scots pine, which do not have a strong habitat specificity and are almost continuously distributed, are expected to have more within-population diversity than those with strong habitat preference and a scattered distribution. Data from this analysis are coherent with those reported in the literature: for example, Robledo-Arnuncio et al. (2005) estimated nSSR loci values of 23.0 and 0.923 for A and H E respectively. However, this authors used only three loci, but if we reference our data with these markers we obtained values of 19.4 (A) and 0.885 (H E).

The Apennine population (CRO) showed lower values for all the calculated genetic variation indices. This confirms the genetic erosion undergone by this population, most likely as a consequence of isolation and limited population size. Lower values of genetic diversity in Apennine populations, compared with Alpine ones, were already recorded by Scalfi et al. (2009) at nSSR loci and by Puglisi and Attolico (2000) and Labra et al. (2006), using different genetic markers, respectively, isozymes and ISSRs. Similar patterns were also observed with reference to populations from the Po valley (TIC and OLG). As suggested earlier for CRO, isolation and habitat fragmentation could be the reasons for the low genetic diversity observed in TIC and, to a lesser extent, OLG, especially after the reduction of the Scots pine range in the Po valley due to intensive use of land for both agriculture and urbanisation purposes. These non-alpine populations are currently regressing, mainly due to problems of seed dispersal, competition with other forest trees and as a consequence of rural depopulation and abandonment of woodland management (Camerano et al. 2008; Regione Piemonte and Regione Valle d’Aosta 2008). Their preservation is therefore a primary goal: it is needed to maintain genetic diversity as well as the forest landscape. Particular attention should be addressed to the population PAS. It is a very small population, where only a few dozens of individuals still survive, threatened by the competition with other species (namely black locust). This population showed the highest number of private alleles (17), confirming an ongoing genetic isolation process which could cause the local extinction of the population over a short period of time.

The overall level of genetic diversity arising from the differentiation between populations found in this study (F ST = 0.058) is moderate, but higher than that previously observed in the same species in other European countries (Müller-Starck et al. 1992). In the Scandinavian region, for instance, values of F ST ≤ 0.02 were found in populations of Scots pine studied with different markers (Karhu et al. 1996). Allozymes gave a F ST = 0.03 between Sweden and Siberian populations (Wang et al. 1991) and a G ST = 0.021–0.046 among many European stands (Prus-Glowacki et al. 1993, 2003). The F ST value found in our study is interesting, especially if we consider the relatively small geographic distances between the Italian populations. A low level of genetic differentiation among populations is common in conifers. They maintain most of their variation within populations (Hamrick et al. 1992), which can be explained by the mainly allogamous mating system, and by the high gene flow rate favoured by their dispersal strategy and widespread diffusion (Petit and Hampe 2006; Piotti et al. 2009; Williams 2010). Mantel’s test, performed in order to check the presence of IBD between the populations studied, was not significant, and thus, IBD was apparently not a mechanism shaping the present distribution of genetic variability.

The analyses of genetic differentiation and the most likely population clustering according to STRUCTURE and sPCA analyses indicate that the global structure detected separate gene pools for the eastern and the Western Alps. Despite the higher similarity of the Apennine population with the Eastern Alps populations, rather than with closer populations from the Western Alps, we found a generally high differentiation between populations from the two mountain chains. These two results seem to exclude the hypothesis of a common postglacial origin of Italian P. sylvestris populations based on the shared RFLP and nad 1 intron mitotypes (Sinclair et al. 1999; Cheddadi et al. 2006; Labra et al. 2006). Our results support the hypothesis of an Apennine glacial refugium (see for instance Puglisi and Attolico 2000), but with no evidence of a remarkable expansion from the Apennines into the southern slope of the Alps, as hypothesised by Cheddadi et al. (2006). The P. sylvestris Apennine population (CRO) is genetically distinct from those in the Western Alps, although a certain level of admixture exists within BOS and CAR populations. Similar results have been recently obtained by Piovani et al. (2010), studying Abies alba populations. These authors also found that Apennine populations are genetically differentiated from populations from the Western Alps. Recent studies based on the analysis of stratigraphic records of pollen, stomata and macrofossils in northern Italy showed that, although the southern slope of the Alps was extensively glaciated during the last glacial maximum (LGM), conifer and several broadleaved tree species survived in the Po plain and along the south-eastern Alpine border (Vescovi et al. 2007). It has also been recently demonstrated that P. sylvestris survived the LGM in the Euganean Hills (north-eastern Italy), a hilly area 50–60 km south of the maximum extent of the last Alpine glaciation (Kaltenrieder et al. 2009). An early presence (ca 15,000 BP) of P. sylvestris after LGM was also signalled by Finsinger et al. (2006) in the Western Alps, at Lago piccolo di Avigliana (Piedmont). Our results depict a scenario where populations from the Western Alps, Eastern Alps and the Apennines originated from at least three different refugia, with possible contact zones between the Western and Eastern Alps, colonisation/expansion routes at the latitude of the TIC and TOC populations, and between the Western Alps and the Apennine populations in southern Piedmont (BOS and CAR populations). In addition, our study presents further evidence of the possible past genetic connection of Western Alps and Apennine populations through the Po plain, a vast area (ca 200 km large) where P. sylvestris was widespread during the early Holocene (Labra et al. 2006, Scalfi et al. 2009).

Seed transfer and the plantation of genetically improved material have often been used in the management of Scots pine forests. As a result, most European countries have a mosaic of native stands and plantations of heterogeneous origins (Savolainen and Yazdani 1991; Mason and Alia 2000). The conservation and management of genetic resources may benefit from the identification of genetically homogeneous regions since the effect of introgression of non-local material to local genetic stocks can be reduced by the use of suitable propagation material. Since 1999, the European Council Directive 1999/105/CE regulates the forest reproductive material market and transfer in Europe. The Italian Government has implemented this directive, with the Decree No. 386/2003. One of the most important features of the acts is the definition of region of provenance as ‘the area or group of areas subjected to sufficiently uniform ecological conditions in which stands or seed sources showing similar phenotypic or genetic characters are found’. The identification of these areas plays a basic role for a rational management of activities linked with forest tree propagation, including afforestation and in situ genetic preservation.

Forest trees show clear phenotypic adaptation to environmental gradients at multiple spatial scale (Savolainen et al. 2007), even though in common garden experiments, local provenances were often outperformed by others [see Reich and Oleksyn (2008) for a thorough re-analysis of many climate gradient experiments on wide-ranging P. sylvestris populations]. The importance of gene zones by neutral molecular markers in the identification of adaptive/evolutionary potential sources has been debated for long. The consistency between neutral molecular marker-based gene zones and classification of provenances based on growth traits has been proved in Norway spruce (Bucci and Vendramin 2000). However, the identification of adaptive/evolutionary potential sources requires the description of functional genetic diversity as well as of structural genetic diversity (Bastien and Alia 2000). Neale and Kremer (2011) claimed that in the near future, forest tree genomics will provide easy-to-use diagnostic tools to assess the distribution of adaptive genetic variability along climatic gradients. To this purpose, it will be mandatory a preliminary assessment of background levels of population structure by neutral markers to understand the influence of selectively neutral processes on the distribution of adaptive genetic variability (Grivet et al. 2009; Eckert et al. 2010). Our neutral molecular data suggest the possible presence of three regions of provenance in the investigated area, but common garden experiments as well as landscape genomics studies are needed to examine the distribution of adaptive variation in populations of Scots pine in northern Italy. In addition, we found some admixed populations that are recommended for further research to investigate the reasons for their peculiar genetic characteristics. Human activities and gene flow between different genetic lineages are the most likely causes of these ‘intermediate types’ even though, in particular for populations at the periphery of the investigated area, a certain level of admixture with other gene zones cannot be excluded.

Notes

Acknowledgments

The research was supported by the project ‘RI.SELV.ITALIA’ (task 1.1, Biodiversity and production of forest reproductive material) from the Italian Ministry for Agriculture, Food and Forestry Policy. Thanks are also due to the Italian Ministry of Education, Universities and Research, which partially granted the work (local research funds). Thanks are also due to colleagues from ‘Translations Group’ for the language revision of the manuscript (http://www.translation-group.com/).

References

  1. Auckland LD, Bui T, Zhou Y, Shepherd M, Williams CG (2002) Conifer microsatellite handbook. Corporate Press, RaleighGoogle Scholar
  2. Bagnoli F, Vendramin GG, Buonamici A, Doulis AG, González-Martínez SC, La Porta N, Magri D, Raddi P, Sebastiani F, Fineschi S (2009) Is Cupressus sempervirens native in Italy? An answer from genetic and palaeobotanical data. Mol Ecol 18:2276–2286PubMedCrossRefGoogle Scholar
  3. Bai WN, Liao WJ, Zhang DY (2010) Nuclear and chloroplast DNA phylogeography reveal two refuge areas with asymmetrical gene flow in a temperate walnut tree from East Asia. New Phytol 188:892–901PubMedCrossRefGoogle Scholar
  4. Bastien C, Alia R (2000) What might be useful measures of genetic variability for adaptive traits within populations of Scots pine? Invest Agr Sist Recur For 9:97–110Google Scholar
  5. Bucci G, Vendramin GG (2000) Delineation of genetic zones in the European Norway spruce natural range: preliminary evidence. Mol Ecol 9:923–934PubMedCrossRefGoogle Scholar
  6. Camerano P, Gottero F, Terzuolo P, Varese P (2008) Tipi forestali del Piemonte, 2nd edn. Regione Piemonte, Blu EdizioniGoogle Scholar
  7. Cavalli-Sforza LL, Edwards AWF (1967) Phylogenetic analysis: models and estimation procedures. Am J Hum Genet 19:233–257PubMedGoogle Scholar
  8. Chapuis M, Estoup A (2007) Microsatellite null alleles and estimation of population differentiation. Mol Biol Evol 24:621–631PubMedCrossRefGoogle Scholar
  9. Chapuis M, Lecoq M, Michalakis Y, Loiseau A, Sword GA, Piry S, Estoup A (2008) Do outbreaks affect genetic population structure? A worldwide survey in Locusta migratoria, a pest plagued by microsatellite null alleles. Mol Ecol 17:640–3653CrossRefGoogle Scholar
  10. Cheddadi R, Vendramin GG, Litt T, François L, Kageyama M, Lorentz S, Laurent JM, De Beaulieu JL, Sadori L, Jost A, Lunt D (2006) Imprints of glacial refugia in the modern genetic diversity of Pinus sylvestris. Glob Ecol Biogeogr 15:271–282Google Scholar
  11. Chybicki I, Burczyk J (2009) Simultaneous estimation of null alleles and inbreeding coefficients. J Hered 100:106–113PubMedCrossRefGoogle Scholar
  12. Di Rienzo AA, Peterson C, Garza JC, Valdes AM, Slatkin M, Freime RB (1994) Mutational processes of simple-sequence repeat loci in human populations. Proc Natl Acad Sci USA 91:3166–3170PubMedCrossRefGoogle Scholar
  13. Eckert CG, Samis K, Lougheed SC (2008) Genetic variation across species’ ranges: the central-marginal hypothesis and beyond. Mol Ecol 17:1170–1188PubMedCrossRefGoogle Scholar
  14. Eckert AJ, Bower AD, Gonzalez-Martinez SC, Wegrzyn JL, Coop G, Neale DB (2010) Back to nature: ecological genomics of loblolly pine (Pinus taeda, Pinaceae). Mol Ecol 19:3789–3805PubMedCrossRefGoogle Scholar
  15. Elsik CG, Williams CG (2001) Low-copy microsatellite recovery from a conifer genome. Theor Appl Genet 103:1189–1195CrossRefGoogle Scholar
  16. Elsik CG, Minihan VT, Hall SE, Scarpa AM, Williams CG (2000) Low-copy microsatellite markers for Pinus taeda L. Genome 43:550–555PubMedGoogle Scholar
  17. Escudero A, Iriondo JM, Torres ME (2003) Spatial analysis of genetic diversity as a tool for plant conservation. Biol Conserv 113:351–365CrossRefGoogle Scholar
  18. Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software structure: a simulation study. Mol Ecol 14:2611–2620PubMedCrossRefGoogle Scholar
  19. Excoffier L, Laval G, Schneider S (2005) Arlequin (version 3.0): an integrated software package for population genetics data analysis. Evol Bioinform 1:47–50Google Scholar
  20. Falush D, Stephens M, Pritchard JK (2003) Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164:1567–1587PubMedGoogle Scholar
  21. Ferrazzini D, Monteleone I, Belletti P (2008) Small-scale genetic diversity in oneseed hawthorn (Crataegus monogyna Jacq.). Eur J For Res 127:407–414CrossRefGoogle Scholar
  22. Finsinger W, Tinner W, Vanderknaap W, Ammann B (2006) The expansion of hazel (Corylus avellana L.) in the southern Alps: a key for understanding its early Holocene history in Europe? Quat Sci Rev 25:612–631CrossRefGoogle Scholar
  23. González-Martinez SC, Robledo-Arnuncio JJ, Collada C, Díaz A, Williams CG, Alía R, Cervera MT (2004) Cross-amplification and sequence variation of microsatellite loci in Eurasian hard pines. Theor Appl Genet 109:103–111PubMedCrossRefGoogle Scholar
  24. Goudet J (1995) FSTAT (version 1.2): a computer program to calculate F-statistics. J Hered 86:485–486Google Scholar
  25. Grivet D, Sebastiani F, Gonzalez-Martinez SC, Vendramin GG (2009) Patterns of polymorphism resulting from long-range colonization in the Mediterranean conifer Aleppo pine. New Phyt 184:1016–1028CrossRefGoogle Scholar
  26. Guo SW, Thompson EA (1992) Performing the exact test of Hardy-Weinberg proportion for multiple alleles. Biometrics 48:361–372PubMedCrossRefGoogle Scholar
  27. Hamrick JL, Godt MJW, Sherman-Broyles SL (1992) Factors influencing levels of genetic diversity in woody plant species. New For 6:95–124CrossRefGoogle Scholar
  28. Jombart T (2008) adegent: a R package for the multivariate analysis of genetic markers. Bioinformatics 11:1403–1405CrossRefGoogle Scholar
  29. Jombart T, Devillard S, Dufour AB, Pontier D (2008) Revealing cryptic spatial patterns in genetic variability by a new multivariate method. Heredity 101:92–103PubMedCrossRefGoogle Scholar
  30. Jump AS, Peñuelas J (2006) Genetic effects of chronic habitat fragmentation in a wind-pollinated tree. Proc Natl Acad Sci USA 103:8096–8100PubMedCrossRefGoogle Scholar
  31. Kalinowski ST (2005) Do polymorphic loci require large sample sizes to estimate genetic distances? Heredity 94:33–36PubMedCrossRefGoogle Scholar
  32. Kaltenrieder P, Belis CA, Hofstetter S, Ammann B, Ravazzi C, Tinner W (2009) Environmental and climatic conditions at a potential glacial refugial site of tree species near the Southern Alpine glaciers. New insights from multiproxy sedimentary studies at Lago della Costa (Euganean Hills, Northeastern Italy). Quat Sci Rev 28:2647–2662CrossRefGoogle Scholar
  33. Karhu A, Hurme P, Karjalainen M, Karvonen P, Kärkkäinen K, Neale D, Savolainen O (1996) Do molecular markers reflect patterns of differentiation in adaptive traits of conifers? Theor Appl Genet 93:215–221CrossRefGoogle Scholar
  34. Kramer AT, Ison JL, Ashley MV, Howe HF (2008) The paradox of forest fragmentation genetics. Conserv Biol 4:878–885CrossRefGoogle Scholar
  35. Labra M, Grassi F, Sgorbati S, Ferrari C (2006) Distribution of genetic variability in southern populations of Scots pine (Pinus sylvestris L.) from the Alps to the Apennines. Flora 201:468–476CrossRefGoogle Scholar
  36. Mantel N (1967) The detection of disease clustering and a generalized regression approach. Cancer Res 27:209–220PubMedGoogle Scholar
  37. Mason WL, Alia R (2000) Current and future status of Scots pine (Pinus sylvestris L.) forests in Europe. Invest Agr Sist Recur For 9:317–333Google Scholar
  38. Miyamoto N, Fernandez-Manjarres JF, Morand-Prieur ME, Bertolino P, Frascaria-Lacoste N (2008) What sampling is needed for treliable estimations of genetic diversity in Fraxinus excelsior L. (Oleaceae). Ann Sci 65:403–410CrossRefGoogle Scholar
  39. Morandini R, Magini E (1975) Il materiale forestale di propagazione in Italia. Collana Verde n. 34, MAF—Corpo Forestale dello Stato, RomaGoogle Scholar
  40. Müller-Starck G, Baradat PH, Bergmann F (1992) Genetic variation in European tree species. New For 6:23–47CrossRefGoogle Scholar
  41. Nathan R (2006) Long-distance dispersal of plants. Science 313:786–788PubMedCrossRefGoogle Scholar
  42. Naydenov K, Senneville S, Beaulieu J, Tremblay F, Bousquet J (2007) Glacial vicariance in Eurasia: mitochondrial DNA evidence from Scots pine for a complex heritage involving genetically distinct refugia at mid-northern latitudes and in Asia Minor. BMC Evol Biol 7:233PubMedCrossRefGoogle Scholar
  43. Neale DB, Kremer A (2011) Forest tree genomics: growing resources and applications. Nat Rev Gen 12:111–122CrossRefGoogle Scholar
  44. Nei M (1978) Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583–590PubMedGoogle Scholar
  45. Palmberg-Lerche C (2001) Conservation of forest biological diversity and forest genetic resources. For Genet Resour Bull 29:24–26Google Scholar
  46. Pautasso M (2009) Geographical genetics and the conservation of forest trees. Perspect Plant Ecol 11:157–189CrossRefGoogle Scholar
  47. Peakall R, Smouse PE (2006) GENALEX 6: genetic analysis in excel. Population genetic software for teaching and research. Mol Ecol Notes 6:288–295CrossRefGoogle Scholar
  48. Pemberton JM, Slate J, Bancroft DR, Barrett JA (1995) Nonamplifying alleles at microsatellites loci: a caution for parentage and population studies. Mol Ecol 4:249–252PubMedCrossRefGoogle Scholar
  49. Petit R, Hampe A (2006) Some evolutionary consequences of being a tree. Annu Rev Ecol Evol S 37:187–214CrossRefGoogle Scholar
  50. Petit R, Aguinagalde I, Beaulieu J-L, Bittkau C, Brewer S, Cheddiari R, Ennos R, Fineschi S, Grivet D, Lascoux M, Mohany A, Müller-Starck G, Demesure-Musch B, Palme A, Martin J, Rendell S, Vendramin G (2003) Glacial refugia: hotspots but not melting pots of genetic diversity. Science 300:1563–1565PubMedCrossRefGoogle Scholar
  51. Pignatti A (1982) Flora d’Italia, vol III. Edagricole, BolognaGoogle Scholar
  52. Piotti A, Leonardi S, Piovani S, Scalfi M, Menozzi P (2009) Spruce colonization at treeline: where do those seeds come from? Heredity 103:136–145PubMedCrossRefGoogle Scholar
  53. Piovani P, Leonardi S, Piotti A, Menozzi P (2010) Conservation genetics of small relic populations of Silver Fir (Abies alba Mill.) in northern Apennines. Plant Biosyst 144:683–691CrossRefGoogle Scholar
  54. Piry S, Luikart G, Cornuet J (1999) BOTTLENECK: a computer program for detecting recent reductions in the effective population size using allele frequency data. J Hered 90:502–503CrossRefGoogle Scholar
  55. Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155:945–959PubMedGoogle Scholar
  56. Prus-Glowacki W, Urbaniak L, Zubrowska-Gil M (1993) Allozyme differentiation in some European populations of Scots pine (Pinus sylvestris L.). Genetica Polonica 2:159–176Google Scholar
  57. Prus-Głowacki W, Stephan BR, Bujas E, Alia R, Marciniak A (2003) Genetic differentiation of autochthonous populations of Pinus sylvestris (Pinaceae) from Iberian peninsula. Plant Syst Evol 239:55–66CrossRefGoogle Scholar
  58. Puglisi S, Attolico M (2000) Allozyme variation in natural populations of the Italian range of Pinus sylvestris L. For Genet 7:221–232Google Scholar
  59. Pyhäjärvi T, Salmela MJ, Savolainen O (2008) Colonization routes of Pinus sylvestris inferred from distribution of mitochondrial DNA variation. Tree Genet Genomes 4:247–254CrossRefGoogle Scholar
  60. R Development Core Team (2009) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0Google Scholar
  61. Regione Piemonte, Regione Autonoma Valle d’Aosta (2008) Il deperimento del pino silvestre nelle Alpi occidentali: natura e indirizzi di gestione. Compagnia delle Foreste, ArezzoGoogle Scholar
  62. Reich PB, Oleksyn J (2008) Climate warming will reduce growth and survival of Scots pine except in the far north. Ecol Lett 11:588–597PubMedCrossRefGoogle Scholar
  63. Robledo-Arnuncio JJ, Gil L (2005) Patterns of pollen dispersal in a small population of Pinus sylvestris L. revealed by total-exclusion paternity analysis. Heredity 94:13–22PubMedCrossRefGoogle Scholar
  64. Robledo-Arnuncio JJ, Collada C, Alia R, Gil L (2005) Genetic structure of mountain isolates of Pinus sylvestris L. in a Mediterranean refugial area. J Biogeogr 32:595–605CrossRefGoogle Scholar
  65. Rubiales MJ, Garcia-Amorena I, Alvarez SG, Manzaneque FG (2008) The late Holocene extinction of Pinus sylvestris in the western Cantabrian range (Spain). J Biogeogr 10:1840–1850CrossRefGoogle Scholar
  66. Savolainen O, Yazdani R (1991) Genetic comparison of neutral and artificial populations of Pinus sylvestris. In: Müller-Starck G, Ziehe M (eds) Genetic variation in European populations of forest trees. Sauerländer’s Verlag, Frankfurt am Main, pp 228–234Google Scholar
  67. Savolainen O, Pyhäjärvi T, Knürr T (2007) Gene flow and local adaptation in forest trees. Ann Rev Ecol Evol Syst 38:595–619CrossRefGoogle Scholar
  68. Scalfi M, Piotti A, Rossi M, Piovani P (2009) Genetic variability of Italian southern Scots pine (Pinus sylvestris L.) populations: the rear edge of the range. Eur J For Res 128:377–386CrossRefGoogle Scholar
  69. Sinclair WT, Morman JD, Ennos R (1999) The postglacial history of Scots pine (Pinus sylvestris L.) in Western Europe: evidence from mitochondrial DNA variation. Mol Ecol 8:83–88CrossRefGoogle Scholar
  70. Soranzo N, Provan J, Powell W (1998) Characterization of microsatellite loci in Pinus sylvestris L. Mol Ecol 7:1260–1261PubMedGoogle Scholar
  71. Soranzo N, Alia R, Provan J, Powell W (2000) Patterns of variation at a mitochondrial sequence-tagged-site locus provides new insights into the postglacial history of European Pinus sylvestris populations. Mol Ecol 9:1205–1211PubMedCrossRefGoogle Scholar
  72. Toro MA, Caballero A (2005) Characterization and conservation of genetic diversity in subdivided populations. Phil T Roy Soc B 1459:1367–1378CrossRefGoogle Scholar
  73. Van Oosterhout C, Hutchinson WF, Wills DPM, Shipley P (2004) Micro-checker: software for identifying and correcting genotyping errors in microsatellite data. Mol Ecol Notes 40:535–538CrossRefGoogle Scholar
  74. Vescovi E, Ravazzi C, Arpenti E, Finsinger W, Pini R, Valsecchi V, Wick L, Ammann B, Tinner W (2007) Interactions between climate and vegetation during the Late glacial period as recorded by lake and mire sediment archives in Northern Italy and Southern Switzerland. Quat Sci Rev 26:1650–1669CrossRefGoogle Scholar
  75. Wang XR, Szmidt AE, Lindgren D (1991) Allozyme differentiation among populations of Pinus sylvestris L. from Sweden and China. Hereditas 114:219–226CrossRefGoogle Scholar
  76. Weir BS (1996) Genetic analysis data II. Sinauer Associates, SunderlandGoogle Scholar
  77. Williams CG (2010) Long-distance pine pollen still germinates after meso-scale dispersal. Am J Bot 97:1–10CrossRefGoogle Scholar
  78. Williams DA, Wang YQ, Borchetta M, Gaines MS (2007) Genetic diversity and spatial structure of a keystone species in fragmented pine rockland habitat. Biol Conserv 138:256–268CrossRefGoogle Scholar
  79. Zhou Y, Bui T, Auckland LD, Williams CG (2002) Undermethylated DNA as a source of microsatellites from a conifer genome. Genome 45:91–99PubMedCrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2012

Authors and Affiliations

  • P. Belletti
    • 1
  • D. Ferrazzini
    • 1
  • A. Piotti
    • 2
  • I. Monteleone
    • 1
  • F. Ducci
    • 3
  1. 1.DIVAPRA Plant Genetics and BreedingUniversity of TurinGrugliascoItaly
  2. 2.Department of Environmental SciencesUniversity of ParmaParmaItaly
  3. 3.Agricultural Research CouncilInstitute of ForestryArezzoItaly

Personalised recommendations