ORIGINAL PAPER

Castanea sativa : genotype-dependent recoveryfrom chestnut blight

Marin Ježić & Ljiljana Krstin & Igor Poljak & Zlatko Liber &

Marilena Idžojtić & Marija Jelić & Jasenka Meštrović &

Marko Zebec & Mirna Ćurković-Perica

Received: 20 November 2012 /Revised: 6 September 2013 /Accepted: 23 September 2013 /Published online: 10 October 2013# Springer-Verlag Berlin Heidelberg 2013

Abstract In Lovran (coastal Croatia), a unique forest/orchardof evenly mixed grafted marrons and naturally growingnongrafted sweet chestnut trees exists. This old chestnut pop-ulation has been devastated by chestnut blight, caused by anaggressive introduced pathogenic fungus, Cryphonectriaparasitica . However, initial observations indicated recovery ofnaturally growing chestnut trees in that area, mediated byCryphonectria-associated hypovirus (Cryphonectria hypovirus1 (CHV-1)). Such recovery was not observed on grafted trees.Genotyping both, we confirmed the clonal origin of the graftedones—marrons. No significant difference was observed betweenfungal strains isolated from naturally growing trees and the onesfrom marrons regarding fungal vegetative compatibility types orthe prevalence of CHV-1. A strong correlation was observedbetween the types of canker: active/deep-expanding versushealing callus or superficial necrosis and the absence or presenceof CHV-1 in the fungal isolates, sampled from naturally growingtrees (Spearman rho 0.686, p value 7.81×10−5, Kendall tau0.686, p value 5.18×10−7). Such correlation was not observedon marrons (Spearman rho 0.236, p value 0.235, Kendall tau0.236, p value 0.084), because, unexpectedly, active/deep-

expanding cankers were often associated with hypovirulent fun-gal isolates. These data indicate that the lack or unequal distribu-tion of naturally occurring hypovirulence were not the cause ofsubstantial marron decay in Lovran. Ecological and age-dependant differences were ruled out because all sampled treesare growing in close proximity and are of similar age. The resultsimply that the marron genotype is especially vulnerable and itsability to recover is limited even when the hypovirulent strain ofthe fungus is present in the canker.

Keywords Biological control .Cryphonectria parasitica .

Genotyping . Hypovirus . Marron . Recovery

Introduction

Sweet chestnut (Castanea sativa Mill.) is a widely spread andimportant multipurpose tree in the Mediterranean area, usedfor its wood, fruit, honey, and tannin. It is also a valuablespecies in ecosystems and landscapes. In countries like Italy,France, Switzerland, and Spain, there is a long tradition ofgrowing marrons, i.e., sweet chestnut cultivars obtainedthrough selection and propagated by grafting, some of whichhave been studied and genotyped (Gobbin et al. 2007; Martinet al. 2009, 2010a, b, 2011), utilizing primers for microsatellitemarkers developed by Buck et al. (2003), Marinoni et al.(2003), and Gobbin et al. (2007).Marrons are grown primarilyfor the production of large quality fruit. In Croatia, sweetchestnut grows in forests in the continental mountainousregions, in Istria, as well as on the islands Krk and Cres(Medak and Perić 2007). Marrons have been planted onprivate properties on the eastern slopes of Učka, in the vicinityof Lovran, located in the bay of Kvarner near Rijeka, forhundreds of years. They were cultivated as orchards insidenatural stands by grafting on the shoots from the stump.However, no scientific data exist whether they stem from a

Communicated by A. Kremer

Electronic supplementary material The online version of this article(doi:10.1007/s11295-013-0667-z) contains supplementary material,which is available to authorized users.

M. Ježić : Z. Liber :M. Ćurković-Perica (*)Division of Biology, Faculty of Science, University of Zagreb,10000 Zagreb, Croatiae-mail: mirna.curkovic-perica@biol.pmf.hr

L. Krstin :M. Jelić : J. MeštrovićDepartment of Biology, University of J. J. Strossmayer in Osijek,31000 Osijek, Croatia

I. Poljak :M. Idžojtić :M. ZebecDepartment of Forest Genetics, Dendrology and Botany,Faculty of Forestry, University of Zagreb, 10000 Zagreb, Croatia

Tree Genetics & Genomes (2014) 10:101–110DOI 10.1007/s11295-013-0667-z

single clone or a collection of cultivars (Idžojtić et al. 2010).The marrons from this region were exported as far back as theseventeenth century and were, along with olives, vine andcherries, important agricultural products in the region forcenturies (Medak et al. 2009). The production and export ofmarrons peaked in the nineteenth century but have droppedconsiderably since. One of the reasons for this decline inproduction has been the spread of chestnut blight disease.

Chestnut blight is a severe disease responsible for thedevastation of chestnut stands in North America and Europe.The causal agent of the disease is the ascomycete fungusCryphonectria parasitica (Murrill) Barr. This pathogen isone of the most aggressive invasive species. It penetrates thehost through wounds in the bark and grows through and underthe bark into the cambium (Heiniger and Rigling 1994). Theresulting cankers progressively enlarge, girdling branches andtrunks, before finally killing the cambium. This deprives thetree beyond the canker of nourishment and causes the infectedbranches and trunks to wilt and eventually die. The firstsymptom of the disease is a small orange-brown area on thebark that develops in a sunken canker and/or where the barksplits. Orange stromata break through the bark of the canker,and asexual fruiting bodies, as well as perithecia, are formedin most cases (Heiniger and Rigling 1994).

C . parasitica originates fromAsia, where it evolved togetherwith its chestnut hosts, Castanea crenata Siebold et Zucc. andCastanea mollissima Blume. These species are therefore resis-tant to infection (Graves 1950; Huang et al. 1998). The funguswas accidentally introduced into North America fromAsia at thebeginning of the twentieth century and devastated nearly all ofthe American chestnut (Castanea dentata /Marshall/ Borkh.)stands (Anagnostakis 1987). In Europe, the disease was firstnoticed on sweet chestnut in Italy in 1938 (Biraghi 1946). FromItaly, it rapidly spread to most of the chestnut-growing areas inEurope (Robin andHeiniger 2001). In Slovenia, it was noticed in1951 on sweet chestnut and the oak treeQuercus petraea (Matt.)Liebl. and its hybrids with Quercus robur L. and, later, onQuercus pubescens Willd. (Jurc 2002). In Croatia, the diseasewas first noticed in the western, coastal region of Croatia, inLovran County in 1955 (Kišpatić 1956; Halambek 1988) andthen spread through the other chestnut-growing regions in Cro-atia during the following decades (Halambek 1988; Novak-Agbaba et al. 2000). Until recently, chestnut blight has causedpartial or complete decay of a large number of marron trees inLovranCounty and has continuallymade the establishing of newplantations extremely difficult since active/deep cankers quicklyappear on newly grafted trees. Chestnut blight has also infectedthe surrounding naturally growing chestnuts (Medak and Perić2007; Idžojtić et al. 2010).

The spread of C. parasitica threatened to destroy Europeanchestnut stands, as it already had in the USA. During the 1950sand 1960s, healing was observed, and hypovirulent fungal strains,providing natural biological control of the disease, were isolated

from healing cankers (Grente 1965). Consequently, due to theappearance of hypovirulent C. parasitica isolates, many chestnutstands spontaneously recovered from the disease (Heiniger andRigling 1994). In Croatia and Slovenia, all attempts to stop thespread of chestnut blight during the 1950s and 1960s failed, andthe disease devastated many chestnut stands. Research on thedisease stopped until 1978, when callusing cankers were noticedin chestnut stands in continental Croatia (Halambek 1986). Sincethen, these chestnut forests have slowly recovered.

Hypovirulent strains are infected by a fungal virus from thefamily Hypoviridae , which causes a reduction in fungal viru-lence and sporulation (hypovirulence) (Elliston 1985; Choiand Nuss 1992). Cryphonectria hypovirus 1 (CHV-1) is thebest-characterized hypovirus and the only hypovirus detectedin Europe to date (Allemann et al. 1999; Sotirovski et al. 2006;Krstin et al. 2008; Robin et al. 2010). The spread of CHV-1has been limited by the fungal diversity, i.e., the high numberof different vegetative compatibility (vc) types, and by sexualreproduction, because the virus is not spread via sexual spores(Anagnostakis 1988). The vegetative compatibility betweenthe hypovirulent and virulent C . parasitica individuals en-ables anastomosis of the hyphae and transfer of the virus frominfected to uninfected mycelium (Cortesi et al. 2001). Fungalstrains infected with hypovirus may transfer CHV-1 double-stranded RNA (dsRNA) via anastomoses to virulent ones thatpreviously caused the active/deep canker (Anagnostakis andDay 1979). As a consequence, callus cankers (healing cankers)are formed, and the tree may recover from the disease (Grenteand Sauret 1969). Superficial cankers (necroses) form whenhypovirulent strains infect the trees (Biraghi 1953). To summa-rize, three basic types of cankers have been identified: activecankers—deep, expanding, potentially lethal cankers—whichgrow on the bark and in the cambium and are initiated in thepresence of virulent strains of the fungus; callus or healingcankers which are formed after the virulent strain of fungushas been transformed to a hypovirulent one due to CHV-1infection; and nonlethal superficial necroses which develop ifa hypovirulent strain infects a tree (Heiniger and Rigling 1994).

Screening C . parasitica from ten naturally growing sweetchestnut populations throughout Croatia revealed that CHV-1,subtype I (Krstin et al. 2008; Perica et al. 2009), is present in allpopulations but is more widespread in continental than coastalregions. In the Lovran County, where the marrons are grown,the symptoms of chestnut blight on the grafted marron trees areworse than on the naturally growing sweet chestnut trees,although they grow literally only a few meters apart from eachother. This raises question whether this is the consequence ofdifferent prevalences of the hypovirus and C . parasitica vctypes on marrons as opposed to naturally growing chestnuts, ormarrons may be less capable of recovering in the presence ofhypovirulent fungus than naturally growing trees.

The aims of this study were, therefore, to (1) genotype themarrons and the natural chestnut population growing in the

102 Tree Genetics & Genomes (2014) 10:101–110

same area, (2) describe the diversity of C . parasitica and theprevalence of hypovirulent isolates of the fungus on marronsand in the natural population, and (3) compare the ability ofthe marrons and the naturally growing chestnuts to recoverfrom the blight, in the presence of hypovirulent fungus in theirnatural habitat.

Materials and methods

Sampling

To reduce sampling bias and the effects of ecological andphysiological parameters, characteristics of the study areawere taken into account. In Lovran County, both naturallygrowing sweet chestnut and grafted marron trees grow at analtitude between 400 and 700 m above the sea level in mixedforest/orchard. Marrons grow exclusively around Lovran,whereas naturally growing trees are also found on the nearbymountain Učka (Fig. 1). Between 2009 and 2011, bark andleaf samples have been collected from 44 nongrafted chestnutsand 57 grafted trees for chestnut genotyping andC . parasiticaisolation. Both, nongrafted and grafted chestnut trees fromLovran were genotyped first to ensure that the trees for thestudy were correctly assigned to two groups (nongrafted, i.e.,naturally growing chestnuts, and grafted trees, i.e., marrons).However, results presented in this study encompass 26nongrafted (naturally growing) and 26 grafted (marron) treesgrowing in Lovran County for which all analyses were suc-cessfully performed (chestnut genotyping, isolation of purefungal colonies from the bark, vc typing, and CHV-1 detec-tion). Other samples which we have been able to analyze onlypartially (i.e., we have not been able to isolate pure C .parasitica culture, or chestnut genotyping was not completedwith all primer pairs) have been excluded. In Lovran, naturallygrowing chestnut trees and marrons grow close together, only10–20 m apart, in the forest/orchard. Therefore, possibledifferences in ecological parameters that might skew the anal-ysis are reduced. Naturally growing trees and marrons in anarea of 3.4 km2 were sampled, taking the age of the trees intoaccount by selecting them according to tree trunk diameter(Meštrović and Fabijanić 1995). Such sampling reduces thepossible physiological effects of tree age on the disease pro-gression. All the grafted trees had diameters at breast height ofat least 30 cm, and therefore, only naturally growing trees withdiameters larger than 30 cm were sampled. According to thegrowth and yield tables for the chestnut in the Lovran region,this meant that all the sampled trees were at least 70 years old(Meštrović and Fabijanić 1995) and were already growing atthe time of initial infection. The bark samples were taken fromthe first accessible canker on the tree, and in the case ofmarrons, above the graft line.

Chestnut genotyping

Using the OmniPrep (G Bioscience) commercial extraction kit,DNAwas isolated from fresh leaf tissue according to the man-ufacturer's instructions. The concentration of DNA was deter-mined with NanoDrop (Thermo Scientific) and adjusted to10 ng μl−1. For the PCR, 50 ng of DNAwas used as a templatein a reaction that contained 1× PCR buffer, 1.5 mM MgCl2,200 μM dNTPs (Promega), 5 μM forward and reverse primers,and 2.5 U Taq polymerase. Ten pairs of primers were used forthe microsatellite loci in this study: CsCAT01, CsCAT02,CsCAT03, CsCAT04, CsCAT06, CsCAT14, CsCAT16,CsCAT17 (Marinoni et al. 2003), EMCs15 (Buck at al. 2003),and OAL (Gobbin et al. 2007) with fluorescently labeled for-ward primers. After the initial denaturation at 94 °C for 2min, 35three-step cycles followed: denaturation at 94 °C for 30 s; an-nealing for 45 s using different annealing temperatures (Ta) foreach primer pair, as suggested byMarinoni et al. (2003), Buck atal. (2003), and Gobbin et al. (2007); and elongation at 72 °C for90 s, and after which, a final elongation step at 72 °C for 8 minwas added. Successful amplification was verified by agarose gelelectrophoresis, and amplicons obtained by the PCRwere sent tothe Genescan Service ofMacrogen (Seoul, Korea). The resultingchromatograms were analyzed for allele sizes using Peak Scan-ner v.1.0. The genotyping of individuals was followed by prin-cipal coordinates analysis (PCoA) analysis to reveal the relation-ship between the grafted marrons and the nongrafted trees fromthe natural population.

Canker assessment, vc-type diversity of C . parasitica ,and prevalence of CHV-1

The state of the canker on the trees previously genotyped[active/deep (aggressive, expanding on the chestnut bark andin the cambium), callus (healing of the canker by the forma-tion of a callus which partially or completely enclosed thewound), and necrosis (superficial form growing slowly on thebark)] was assessed on site from the collected bark sample.Most of the trees had quite a few cankers visible, but only oneper tree was collected, as it has been done in previous studies(Breuillin et al. 2006; Milgroom et al. 2008; Krstin et al. 2008,2011) because, even in population with predominant sexualreproduction, clones of the fungus more often occur close toeach other (Milgroom and Lipari 1995). Furthermore, cankerswere collected only from trunks of the trees which were morethan 70 years old with diameter of at least 30 cm, because thedevelopment and progression of active/deep cankers mightalso depend on the diameter of the tree at the site of infection;progression of active/deep cankers on branches might beaffected by the size of their diameter.

Since the trees are old with wide trunk and thick bark, thecankers were assessed as active/deep if deep wounds withclear necrosis of the underlying tissue were visible on the

Tree Genetics & Genomes (2014) 10:101–110 103

trees. Cankers where the callus growth partially or completelyisolated the wound were assessed as “healing cankers—calli,”and cankers where only slight superficial peeling of the barkhad occurred were assessed as superficial necroses. The as-sessment of each canker was always done the same way andby the same person.

A piece of bark (6×8 cm) was taken from the margin of eachcanker on each nongrafted and grafted tree. Before sampling, alltools were sterilized in 96 % ethanol. Isolation of C . parasiticafollowed the procedure of Cortesi et al. (1996). The fungalcultures were maintained on potato dextrose agar (PDA, Biolife)in Petri dishes as described in Hillman et al. (1990). All C .parasitica isolates were assessed for the presence of hypovirususing culture morphology. For the assays, the isolates weregrown on PDA plates at 20–22 °C in the dark for 7 days,followed by incubation under daylight on the laboratory benchfor 7 days. Under these conditions, hypovirus-free C . parasiticaisolates typically produce yellow-orange mycelium with abun-dant conidiation, while C . parasitica isolates infected withCHV-1 remain white with no or low conidiation (Bisseggeret al. 1997). Isolates were also tested for the presence of CHV-

1 dsRNA using two methods. Each isolate was grown on PDAwith a cellophane overlay for 6 days at 25 °C in the dark(Allemann et al. 1999). The culture was stripped from thecellophane overlay, transferred to a 2-ml reaction tube, andlyophilized. The dried mycelium (40 mg) was ground to a finepowder with a steel ball using the Tissue Lyser (Qiagen). Totalnucleic acids were extracted, and hypoviral dsRNAwas purifiedusing cellulose CF-11 chromatography, as described inAllemann et al. (1999) and Krstin et al. (2008), or HPLC(Perica et al. 2009).

The fungal isolates were assessed for vegetative compatibil-ity according to the merging/barrage response (Anagnostakis1988; Bissegger et al. 1997; Cortesi et al. 1998). Strains werepaired approximately 3 mm apart and 5 mm from the edge ofthe Petri plate. European vc type testers from EU-1 to EU-64were used (Cortesi et al. 1998). Six pairings were made perPetri plate (9 cm in diameter), with each plate containing 25 mlof PDA. Petri plates were incubated at 25 °C for 7 days in thedarkness, followed by 7 days of incubation in daylight at roomtemperature. Vegetative compatibility was scored after 7 and14 days. Merging of the two colonies indicates a compatible

Fig. 1 Geographical location ofLovran in Croatia. Chestnutforest/orchard is shaded in red

104 Tree Genetics & Genomes (2014) 10:101–110

pairing, i.e., the isolates have the same vc type, whereas abarrage line between the colonies indicates an incompatiblepairing (Cortesi et al. 1998). Each vc test for each isolate wasdone in triplicate, and only if the replications were identical wasthe isolate assigned to the particular vc type.

Statistical analyses

To explore the diversity of the native, naturally growing chest-nut population, several analyses were performed. Allelic fre-quencies (pa) for all alleles were determined, and the expectedand observed heterozygosity was calculated with Identity 1.0(Wagner and Sefc 1999). Gametic disequilibrium among pairsof loci was tested within each population using Fisher exacttests in GenePop 4.0 software (Raymond and Rousset 1995;Rousset 2008). To assess whether the multilocus linkage dis-equilibrium is a phenomenon characteristic of the naturallygrowing chestnut population in Lovran (i.e., whether the pop-ulation is isolated or, for whatever reason, not reproducingsexually), the index of multilocus linkage disequilibrium rdwas estimated with Multilocus 1.3 software (10,000 permuta-tions) (Agapow and Burt 2001). The index of multilocus link-age disequilibrium is actually a modified index of association,which is insensitive to the number of loci considered. Theaverage heterozygosity was estimated using Dispan. To testfor genotypic disequilibrium, FSTAT 2.9.3.2 (Goudet 1995)was used. PCoA was performed with GenAlEx 6.3 to assessthe relation of Lovran marrons with other naturally growingchestnuts from the same region, especially to find out if marronsdiffer significantly from the naturally growing trees in the studyarea (Peakall and Smouse 2006).

We classified each canker found on a tree as “active/deep,”“callus,” or “necrosis,” based on visual inspection on site.Furthermore, after dsRNA extraction, each fungal sample wasdescribed as “orange” (virulent, absence of dsRNA) or “white”(hypovirulent, hypoviral dsRNA present). The vc type diversityof C . parasitica was determined using the Shannon diversityindex (H), as described by Anagnostakis et al. (1986), and thegenotypic diversity index (G). H was calculated as H =−∑pilnpi, in which pi is the frequency of the i th vc type, and Gwas calculated as G =1/∑pi2, where pi is the frequency of thei th genotype. Evenness (e) was calculated as H /lnNH. Theindex E 5 was used as recommended by Grünwald et al.(2003), G −1/eH−1.

To test whether the canker type (active/deep versus necrosis orcallus) was associated with the presence of CHV-1 in the fungalisolate (virulent or hypovirulent) in naturally growing trees andmarrons, two nonparametric indices of association, Spearmanrho and Kendall tau, were employed. Fisher exact tests wereperformed to assess the recovery of cankers, i.e., number ofhealing cankers. For this purpose, the data were rearranged,and the Mann–Whitney U test was performed to test whetherthe distribution of virulent and hypovirulent (CHV-1-infected)

isolates differed among the different vc types from fungal col-lections obtained from naturally growing trees and marrons.

Results

Genetic diversity of the chestnut trees in the study areain Lovran

Between three (on loci CsCAT14 and EMCs15) and nine (onlocus CsCAT03) different alleles were found among the tenloci included in our study. With the primer combination usedin our experiments, 54 different alleles could be distinguishedamong naturally growing chestnuts. The allelic frequencies ofnaturally growing chestnut population from Lovran as well asits expected and observed heterozygosity are given in Table 1.The average heterozygosity was 0.641±0.055. The genotypicdisequilibrium was explored with GenePop, and not all of thetested loci pairs were found to be in equilibrium (Supplemen-tary material, Table S1). The index of the multilocus linkagedisequilibrium, rd, was very low and significantly differentfrom zero, 0.039, p value 0.006.

Our analysis uncovered a very diverse natural chestnut popu-lation. We observed 26 different genotypes in 26 naturally grow-ing trees included in our study. In the contrast, all the grafted treeshad the same genotype (Supplementary material, Table S2).PCoA showed a homogenous population of chestnuts, with themarron genotype falling inside the main cluster (Fig. 2). This wasimportant because it allowed us to group the chestnuts unambig-uously into “naturally growing” (nongrafted or “wild”) or “mar-rons” (grafted trees, one specific sweet chestnut cultivar), since thepossibility existed that different genotypes were used as grafts.

Analysis of the C . parasitica population from Lovran

All the trees in the Lovran region are infectedwithC . parasitica .Each canker on a tree was classified as active/deep, callus, ornecrosis, based on a visual inspection on site. Almost all theobserved calli were completely closed (the callusing was com-plete). The one exception being in the process of wound closingon one of naturally growing trees was also designated as callus.C . parasitica isolates were characterized as orange (virulent,absence of dsRNA) or white (hypovirulent, hypoviral dsRNApresent) after dsRNA isolation, as shown in Table 2. The vc typediversity of the chestnut blight fungus, described by the Shannondiversity index (H), evenness (e), genotypic diversity (G), andE5, was similar for the isolates from naturally growing chestnutsand marrons (Table 3). No statistically significant differenceswere observed between the two chestnut groups (naturally grow-ing trees and marrons) regarding the abundance of different vctypes and the percentage of white hypovirulent isolates (Tables 3and 4). The most striking difference, however, was the almostcomplete lack of healing cankers or necroses on marrons, even

Tree Genetics & Genomes (2014) 10:101–110 105

though a significant number of white isolates were isolated fromthem (Table 2). The morphology of the isolate was found to beassociated with the type of canker on naturally growing trees(Spearman rho 0.686, p value 7.81×10−5, Kendall tau 0.686, pvalue 5.18×10−7), but not on marrons (Spearman rho 0.236, pvalue 0.235, Kendall tau 0.236, p value 0.084). Even more, inour initial sample, we have observed eight healing cankers (calli)among 44 naturally growing trees and two among 57 marronsand performed the tests on this larger sample, as well (Spearmanrho 0.950, p value 3.263×10−20, Kendall tau 0.950, p value 1,725×10−17 for naturally growing trees; Spearman rho 0.179, pvalue 0.264, Kendall tau 0.179, p value 0.100 for marrons). Ineither case, there was no correlation between canker type andmorphology of the C . parasitica isolate in marrons, while thecorrelation was observed in the naturally growing tree sample.Therefore, further results are presented for isolates for whichcomplete set of analyses was available. No association betweenthe vc type and the presence or absence of CHV-1 was observed(Table 3). Fisher exact tests revealed thatmarrons aremuchmorelikely to have active/deep cankers than naturally growing trees(p value=0.00082). Furthermore, active/deep cankers withwhite isolates are much more frequent on marrons than onnaturally growing trees, whereas healing cankers and necrosesare much more likely on naturally growing trees (p value=0.00017) (Table 4).

Table 1 Diversity of ten analyzedmicrosatellite loci in naturally growingchestnuts from Lovran

Locus Allele(bp)

Frequency Expectedheterozygosity(He)

Observedheterozygosity(Ho)

Fixationindex(F IS)

CAT01 193 0.442 0.697 0.808 −0.139206 0.019

208 0.038

214 0.135

216 0.019

218 0.058

222 0.288

CAT02 207 0.115 0.763 1.000 −0.292211 0.058

215 0.250

221 0.019

225 0.250

231 0.308

CAT03 197 0.019 0.833 0.962 −0.135213 0.154

225 0.154

227 0.231

233 0.135

241 0.212

245 0.019

253 0.038

255 0.038

CAT04 213 0.654 0.521 0.500 0.059**221 0.192

235 0.038

237 0.115

CAT06 158 0.173 0.748 0.769 −0.009164 0.096

172 0.404

177 0.019

180 0.038

182 0.212

194 0.058

CAT14 133 0.558 0.587 0.731 −0.227141 0.173

150 0.269

CAT16 126 0.365 0.670 0.923 −0.361132 0.250

143 0.365

147 0.019

CAT17 130 0.038 0.717 0.692 0.055138 0.019

142 0.019

146 0.346

152 0.327

154 0.019

160 0.231

Cs15 83 0.038 0.240 0.192 0.219

Table 1 (continued)

Locus Allele(bp)

Frequency Expectedheterozygosity(He)

Observedheterozygosity(Ho)

Fixationindex(F IS)

86 0.096

92 0.865

OAL 298 0.635 0.511 0.654 −0.261302 0.0.38

306 0.0.38

310 0.288

**p value of <0.01, significant

Co

ord

. 2

Coord. 1

Principal Coordinates

Fig. 2 Principal coordinates analysis of the chestnut population fromLovran. Naturally growing chestnuts are represented with white squares ,and marron genotype is shown in black . The first axis contributes with26.41 % and the second with 19.31 % to the total genetic variation

106 Tree Genetics & Genomes (2014) 10:101–110

Discussion

The genetic diversity and characterization of chestnut cultivars inEurope have been investigated in several studies in recent years(Gobbin et al. 2007; Martin et al. 2010a, b, 2011). Buck et al.(2003) and Marinoni et al. (2003) described the microsatellitemarkers that can be used in chestnut population studies, andMarinoni (2003) genotyped 20 chestnut cultivars, finding

different cultivars genetically quite diverse. Martin et al.(2010a, b, 2011) also genotyped several chestnut cultivars aswell as naturally growing trees from Italy, Spain, and Greece, asdid Gobbin et al. (2007) for trees from Switzerland. In manycases, they found synonymies and homonymies, but numerouscultivars were shown to be true clones. Among the wide varietyof chestnut cultivars these studies confirmed in Europe, Gobbinet al. (2007) included seven wild (naturally growing) treesamples in their study, which clustered with different chestnutcultivars, suggesting that different cultivars were establishedfrom different, and rather varied, chestnut genotypes.

Until our study, it was not known whether Lovran marronswere one single clone or several different ones, and their rela-tionship to marron cultivars grown in other countries still needsto be clarified. Our study, in which ten primer pairs (Buck et al.2003; Marinoni et al. 2003; Gobbin et al. 2007) were used,revealed that the marrons from Lovran are, in fact, a single clonewith no variations between individuals. In contrast, the naturallygrowing trees interspersed in orchards in Lovran were found tobe very diverse, as can be expected for a naturally growingpopulation, with diversity indices comparable to the populationsfrom Spain, Italy, and Greece (Martin et al. 2010a, b, 2011).Until now, no data have been collected on how different C .sativa cultivars recover from C . parasitica infection after thehypovirulence has been introduced in a particular region. It isknown that American C . dentata and European C . sativa aresusceptible to blight infection, whereas Asian species like C .crenata and C . mollissima tolerate it (Kubisiak et al. 1997).Efforts have therefore been directed to breeding hybrids of C .

Table 2 Morphology and canker type of C . parasitica isolates from Lovran

Isolate morphology Naturally growing (wild) Marrons (“grafted”)

Canker type

Active/deep Callus Necrosis Total Active/deep Callus Necrosis Total

Orange (virulent) 10 0 0 10 15 0 0 15

White (hypovirulent) 4 5 7 16 10 0 1 11

Total 14 5 7 26 25 0 1 26

Table 3 Vegetative compatibility types, diversity, and evenness of theC .parasitica population from Lovran

vc type Naturally growing (wild) Marrons (grafted)

Isolate morphology

Orange(virulent)

White(hypovirulent)

Orange(virulent)

White(hypovirulent)

EU1 0.15 0.07 0.23 0.15

EU2 0.11 0.22 0.19 0.04

EU12 0.00 0.15 0.04 0.12

EU13 0.04 0.07 0.04 0.00

EU15 0.04 0.00 0.00 0.04

EU17 0.07 0.07 0.08 0.08

Total 0.41 0.58 0.58 0.43

Shannon diversityindex (H)

1.617 1.532

Evenness (e) 0.838 0.772

Shannon diversityindex (H)

1.595

Evenness (e) 0.854

Genotypicdiversity (G)

4.44 3.98

Evenness (E5) 0.853 0.821

Genotypicdiversity (G)

4.35

Evenness (E5) 0.854

Table 4 Morphology and canker type ofC . parasitica isolates in Lovranrearranged for the Fisher exact test

Naturallygrowing (wild)

Marrons(grafted)

Active/deep cankers 14 25

Calli and/or necroses 12 1

Active/deep canker with white(hypovirulent) fungus morphology

4 15

Calli and/or necroses with white(hypovirulent) fungus morphology

12 1

Tree Genetics & Genomes (2014) 10:101–110 107

dentata and C . mollissima resistant to chestnut blight (Diskinet al. 2006), in the USA, where attempts to biologically controlchestnut blight disease have almost completely failed (Milgroomand Cortesi 2004). The resistance of some hybrids to C .parasitica (Anagnostakis 1992; Kubisiak et al. 1997) was testedto find out more about the mechanisms and genetics involved inthe Asian species' tolerance to the disease. In Europe, effortshave also been made to breed tolerant hybrids of European andAsian chestnuts. The work done in Switzerland in the previousdecades (Bazzigher and Miller 1991) explored the chestnuts'susceptibility to the disease, using stocks which were, mostprobably, also hybrids of European and Asian chestnuts' species.In Europe, dissemination of hypoviruses also appears to be aquite effective way to control chestnut blight biologically (Robinand Heiniger 2001; Sotirovski et al. 2011).

In all chestnut populations near Lovran, previously screenedfor C . parasitica , naturally occurring CHV-1 is widespread(Krstin et al. 2008, 2011). The diversity of the C . parasiticapopulation from Lovran is comparable to that of other C .parasitica populations from Croatia and Slovenia, with theproportion of hypovirulent isolates even higher than in the nearbypopulations studied (Cres, Istria, Ostrovica) (Krstin et al. 2008,2011). Lovran is also the place where chestnut blight was firstobserved in Croatia more than five decades ago (Kišpatić 1956).Hypovirulent isolates were observed more than three decadesago (Halambek 1986). These scientific data indicate that C .parasitica population in Lovran is the oldest in Croatia and thathypovirulent isolates have been present here for decades.

Number and type of cankers (either active or healing ornecrosis) are usually used in assessing the disease severity inchestnut forests. In human-mediated inoculations, it has beenconsistently shown that virulent types of C . parasitica inducethe formation of active/deep cankers, whereas superficial ne-croses are formed only when hypovirulent strains are inocu-lated on chestnuts (McManus et al. 1989; Hogan et al. 2002).In nature, however, active/deep cankers cannot be unambigu-ously associated only with the virulent type of fungus. Bothvirulent and hypovirulent or only hypovirulent strain(s) cansometimes be found in active/deep cankers.

This happens as a result of highly dynamic nature of thispathosystem. Virulent strain c . parasitica involved in the initialinfection, and formation of an active/deep canker on the chestnutcan become hypovirus-infected at some later point in time afterinitial canker establishment. Furthermore, the mycelium inside acankermight only be partially infectedwith CHV-1 (Hogan et al.2002). The conversion between active/deep form of canker tohealing form (callus) does not occur immediately and is depen-dent of the tree's ability to form a protective wound periderm.Furthermore, from healing cankers and superficial necroses,hypovirulent strains of fungus are isolated, but occasionally,isolation attempts from healed cankers fail to detect ahypovirulent strain of fungus which induced the healing of thecanker (Prospero et al. 2006).

Since equal distribution of CHV-1 in different C . parasiticavc types on both groups of chestnuts in Lovran, marrons andnaturally growing trees, was observed in our experiment, ourpremise was that different canker types would also be similarlydistributed among naturally growing trees and marrons. Howev-er, while all three types of cankers—active/deep, calli, andnecroses—were commonly observed on naturally growing trees,the most striking feature in the marron group of trees was thehigh number of active/deep cankers. Healing calli and necroseswere found on marrons only sporadically. In the bark samplesfrom active/deep cankers onmarrons (whichwere numerous andsignificantly outnumbered superficial cankers), we identifiedmany white, hypovirulent isolates. Since C . parasitica popula-tions nearby are very diverse (Krstin et al. 2008, 2011; Ježić et al.2012), this fact raised the question of whether this difference inhow naturally growing trees and marrons recovered could be aconsequence of the unequal distribution of hypovirulence in thedifferent vc types of the fungus present in Lovran. However, wefound a similar distribution of different vc types on naturallygrowing trees and marrons as well as a similar distribution ofhypovirulent isolates on both groups of the trees. Therefore, thelack of naturally occurring hypovirulence on marrons can beruled out as the reason why the state of grafted trees was clearlyworse than that of naturally growing chestnuts.

It seems that naturally occurring hypovirulence enables therecovery of the naturally growing, nongrafted trees, sincehealing calli develop, but it is unable to reproduce the sameeffect and number of healing cankers on grafted trees. Theunique characteristic of this particular chestnut forest/orchardwith its even mix of naturally growing trees and marronsexcludes the environmental factors as a possible cause forthe inability of marrons to efficiently form healing cankersin the presence of hypovirulent strains of C . parasitica .

Some studies suggest that older trees are less susceptible tochestnut blight as it progresses more slowly on them (Heinigerand Rigling 1994). The chestnut forest/orchard in Lovranconsists mostly of large, old marrons and naturally growingtrees of all ages. In this study, only naturally growing treesolder than 70 years were included, so tree age can also be ruledout as a reason for the poor recovery of marrons. Older agedoes not seem to make marrons more resistant to chestnutblight nor reduce its progression. Since, in this research, anaturally established pathosystem chestnut–fungus–virus wasused, it was impossible to assess the possible physiologicalimpact of grafting itself on recovery of marrons. Therefore,grafting naturally growing scions on naturally growing root-stocks will have to be addressed in future experiments.

It is known that the interaction of the pathogen and a planthost is dependent upon their genotypes, plant age, environmentalconditions, etc. In the case of chestnut blight, the development ofthe disease is dependent upon the complex three-lateral interac-tion between hypovirus, fungus, and chestnut. In previous re-searches (Peever et al. 2000; Sotirovski et al. 2011), the

108 Tree Genetics & Genomes (2014) 10:101–110

development of the chestnut blight in the forest was observedafter inoculation of different C . parasitica strains infected bydifferent viruses on randomly selected trees. However, the out-come of those experiments was inevitably affected by the spreadof naturally occurring fungal strains. Furthermore, in researchesdone so far, the possible contribution of C . sativa genotypes indisease development and recovery was not taken into account.The experimental design we applied is different than thoseapplied before. We compared the response of maroon genotypeand that of natural chestnut population to chestnut blight incondition of well-established naturally occurring hypovirulence.We genotyped the host, characterized C . parasitica population,and measured the prevalence of naturally occurringhypovirulence, proving by statistical methods that populations'characteristics of the fungus (vc type diversity) and hypovirusprevalence were not different between marrons and naturallygrowing trees. An additional confirmation of these results couldbe obtained by inoculating virulent and hypovirulent strains ofC .parasitica on grafted marrons and naturally growing trees ingreenhouse conditions, followed by inoculation of hypovirulentstrains of the fungus in cankers previously induced by virulentstrains of C . parasitica. In that way, it would be possible toassess not only susceptibility to the disease (in the first part of theexperiment) but also recoverability of the different chestnutgenotypes in the presence of hypovirulent strains of C .parasitica in the canker in controlled conditions.

Nevertheless, our results reveal that the lack of recovery ofmarrons probably relates to their unique genotype. Theirrecovery is much more difficult, even in the presence ofhypovirulent strains of the fungus. It is logical to expectdifferent sweet chestnut genotypes to have different recoveryabilities in the presence of hypovirulent C . parasitica , but ithas never been properly demonstrated. We found that, unlikethe naturally growing chestnut population in Lovran, a spe-cific Lovran marron breed of sweet chestnut recovers from thedisease slowly, even in the presence of natural hypovirulence.This has affected the productivity of Lovran's marron or-chards, which dropped significantly after the introduction ofthis pathogen and has not recovered since (Medak et al. 2009).Only a few hundred marron trees have survived.

In conclusion, Lovran marrons, which can be identified bytheir unique genotype, appear to be less able to recover fromchestnut blight than naturally growing trees. Although this breedis greatly valued for its large nuts with their distinctive flavor, itis, unfortunately, muchmore likely to succumb to chestnut blightthan the natural chestnut population. Therefore, human-mediatedbiocontrol utilizing strong CHV-1 strain is necessary to preventthe loss of this unique and much valued marron genotype.

Acknowledgments We thank Silvia Dingwall for the final Englishediting. This research was supported by Swiss National Science Founda-tion (SCOPES project IZ7370-12792/1) and the Croatian Ministry ofScience Education and Sport.

Data Archiving Statement Genotypes of the chestnut trees in Lovranare given in Supplementary 847 materials. 848.

References

Agapow PM, Burt A (2001) Indices of multilocus linkage disequilibrium.Mol Ecol Notes 1:101–102

Allemann C, Hoegger P, Heiniger U, Rigling D (1999) Genetic variationof Cryphonectria hypoviruses (CHV1) in Europe, assessed usingrestriction fragment length polymorphism (RFLP) markers. MolEcol 8:843–854

Anagnostakis SL (1987) Chestnut blight: the classical problem of anintroduced pathogen. Mycologia 79:23–37

Anagnostakis SL (1988) Cryphonectria parasitica , cause of chestnutblight. Adv Plant Pathol 6:123–136

Anagnostakis SL (1992) Measuring resistance of chestnut trees to chest-nut blight. Can J Forest Res 22:567–571

Anagnostakis SL, Day PR (1979) Hypovirulence conversion in Endothiaparasitica. Phytopathology 69:1226–1229

Anagnostakis SL, Hau B, Kranz J (1986) Diversity of vegetative com-patibility groups of Cryphonectria parasitica in Connecticut andEurope. Plant Dis 70:536–538

Bazzigher G, Miller GA (1991) Blight-resistant chestnut selections ofSwitzerland: a valuable germplasm resource. Plant Dis 75:5–9

Biraghi A (1946) Il cancro del castagno causato da Endothia parasitica .L'Italia Agricola 7:406

Biraghi A (1953) Possible active resistance ofEndothia parasitica inCastaneasativa. 11th Rep Congr Int Union For Res Org, Rome, pp 643–645.

Bissegger M, Rigling D, Heiniger U (1997) Population structure anddisease development of Cryphonectria parasitica in Europeanchestnut forests in the presence of natural hypovirulence. Phytopa-thology 87:50–59

Breuillin F, Dutech C, Robin C (2006) Genetic diversity of the chestnutblight fungus Cryphonectria parasitica in four French populationassessed by microsatellite markers. Mycol Res 110:288–296

Buck EJ, HadonouM, James CJ, Blakesley D, Russell K (2003) Isolationand characterization of polymorphic microsatellites in Europeanchestnut (Castanea sativa Mill.). Mol Ecol Notes 3:239–241

Choi GH, Nuss DL (1992) Hypovirulence of chestnut blight fungusconferred by an infectious viral cDNA. Science 257:800–803

Cortesi P, Milgroom MG, Bisiach M (1996) Distribution and diversity ofvegetative compatibility types in subpopulations of Cryphonectriaparasitica in Italy. Mycol Res 100:1087–1093

Cortesi P, Rigling D, Heiniger U (1998) Comparison of vegetative compatibil-ity types in Italian and Swiss subpopulations ofCryphonectria parasitica.Europ J Forest Pathol 28:167–176

Cortesi P, McCulloh CE, Song H, Lin H, Milgroom MG (2001) Geneticcontrol of horizontal virus transmission in the chestnut blight fun-gus, Cryphonectria parasitica . Genetics 159:107–118

Diskin M, Steiner KC, Hebard FV (2006) Recovery of American chestnutcharacteristics following hybridization and backcross breeding to restoreblight-ravaged Castanea dentata. Forest Ecol Manag 223:439–447

Elliston JE (1985) Characteristics of dsRNA-free and dsRNA-containingstrains of Endothia parasitica in relation to hypovirulence. Phyto-pathology 75:151–158

Gobbin D, Hohl L, Conza L, Jermini M, Gessler C, Conedera M (2007)Microsatellite-based characterization of theCastanea sativa cultivarheritage of southern Switzerland. Genome 50:1089–1103

Goudet J (1995) FSTAT, version 1.2, a computer program to calculate F-statistics. J Heredity 86:485–486

Graves AH (1950) Relative blight resistance in species and hybrids ofCastanea . Phytopathology 49:1125–1131

Tree Genetics & Genomes (2014) 10:101–110 109

Grente J (1965) Les formes Hypovirulentes d'Endothia parasitica et lesespoirs de lutte contre le chancre du châtaignier. Acad Agric France51:1033–1037

Grente J, Sauret S (1969) L'hypovirulence exclusive phenomene originalen pathologie vegetale. C R Hebd Seances Acad Sci Ser D 268:2347–2350

Grünwald NJ, Stephen BG, Milgroom MG, Fry WE (2003) Analysis ofgenotypic diversity data for populations of microorganisms. Phyto-pathology 93:738–46

Halambek M (1986) Chestnut blight in Yugoslavia. EPPO Bull 16:533–535Halambek M (1988) Istraživanje virulentnosti gljive Endothia parasitica

Murr./And. uzročnika raka kore pitomog kestena (Castanea sativaMill.). Dissertation, Faculty of Forestry, University of Zagreb.

Heiniger U, Rigling D (1994) Biological control of chestnut blight inEurope. Annu Rev Phytopathol 32:581–599

Hillman BI, Shapira R, Nuss DL (1990) Hypovirulence-associated sup-pression of host functions in Cryphonectria parasitica can be par-tially relieved by high light intensity. Phytopathology 80:950–956

Hogan EP, Griffin GJ (2002) Incomplete movement of Cryphonectriahypovirus 1 within a vegetative compatibility type ofCryphonectriaparasitica in natural cankers on grafted American chestnut trees.Forest Pathol 6:331–344

Huang H, Dane F, Kubisiak TL (1998) Allozyme and RAPD analysis ofthe genetic diversity and geographic variation in wild populations ofthe American chestnut (Fagaceae). Am J Bot 85:1013–1021

Idžojtić M, Zebec M, Poljak I, Liber Z, Šatović Z, Vahčić N (2010)Očuvanje genofonda maruna. In: Jasprica N, Pandža M, MilovićM (eds) Treći hrvatski botanički kongres, knjiga sažetaka, Murter,24–26 September 2010, pp 92–93.

JežićM,Krstin L, Rigling D,Ćurković-PericaM (2012) High diversity inpopulations of the introduced plant pathogen, Cryphonectriaparasitica, due to encounters between genetically divergent geno-types. Mol Ecol 21:87–99

Jurc D (2002) An overview of the history of chestnut blight epidemic inSlovenia. Zbornik gozdarstva in lesarstva 68:33–59

Kišpatić J (1956) Rak kestenove kore (Endothia parasitica Anders.).Zavod za zaštitu bilja, Zagreb. Uputstva iz zaštite bilja 19:1–12

Krstin L, Novak-Agbaba S, Rigling D, Ćurković-Perica M (2011) Diver-sity of vegetative compatibility types and mating types ofCryphonectria parasitica in Slovenia and occurrence of associatedCryphonectria hypovirus 1 . Plant Pathol 60:752–761

Krstin L, Novak-Agbaba S, Rigling D, Krajačić M, Ćurković-Perica M(2008) Chestnut blight fungus in Croatia: diversity of vegetativecompatibility types, mating types and genetic variability of associ-ated Cryphonectria hypovirus 1 . Plant Pathol 57:1086–1096

Kubisiak TL, Hebard FV, Nelson CD, Zhang J, Bernatzky R, Huang H,Anagnostakis SL, Doudrick RL (1997) Molecular mapping of re-sistance to blight in an interspecific cross in the genus Castanea .Phytopathology 87:751–759

Marinoni D, Akkak A, Bounous G, Edwards KJ, Botta R (2003) Devel-opment and characterization of microsatellite markers in Castaneasativa (Mill.). Mol Breed 11:127–136

Martin MA, Alvarez JB, Mattioni C, Cherubini M, Villani F, Martin LM(2009) Identification and characterization of traditional chestnutvarieties of southern Spain using morphological and simple se-quence repeats SSR markers. Ann Appl Biol 154:389–398

Martin MA, Mattion C, Cherubini M, Taurchini D, Villani F (2010a)Genetic characterization of traditional chestnut varieties in Italyusing microsatellites (simple sequence repeats) markers. Ann ApplBiol 157:37–44

Martin MA, Mattioni C, Cherubini M, Taurchini D, Villani F (2010b) Geneticdiversity in European chestnut populations by means of genomic andgenic microsatellite markers. Tree Genet Genomes 6:735–744

Martin MA, Mattioni C, Molina JR, Alvarez JB, Cherubini M, HerreraMA, Villani F, Martin LM (2011) Landscape genetic structure of

chestnut (Castanea sativa Mill). in Spain. Tree Genet Genomes 8:127–136

McManus PS, Ewers FW, Fulbright DW (1989) Characterization of thechestnut blight canker and the localization and isolation of thepathogen Cryphonectria parasitica . Can J Bot 67:3600–3607

Medak J, Perić S (2007) Šume pitomog kestena u Hrvatskoj-fitocenološke,ekološke i gospodarske karakteristike. In: Britvec M, Škvorc, Ž (eds),Zbornik sažetaka 2. botaničkog kongresa. Zagreb, 20–22 September2007, pp 89–90.

Medak J, Idžojtić M, Novak-Agbaba S, Ćurković-Perica M, Mujić I,Poljak I, Juretić D, Prgomet Ž (2009) Croatia. In: Avanzato D (ed)Following chestnut footprints (Castanea spp.)—cultivation andculture, folklore and history, traditions and use. Scripta Horticult 9:40–43

Meštrović Š, Fabijanić G (1995) Forest management planning manual.Republic of Croatia, Ministry of Agriculture and Forestry, Zagreb,Croatia, pp 322–323 (in Croatian)

Milgroom MG, Cortesi P (2004) Biological control of chestnut blightwith hypovirulence: a critical analysis. Annu Rev Phytopathol 42:311–338

Milgroom MG, Lipari SE (1995) Spatial analysis of nuclear and mito-chondrial RFLP genotypes in populations of chestnut blight fungus,Cryphonectria parasitica . Mol Ecol 4:633–642

Milgroom MG, Sotirovski K, Spica D, Davis JE, Brewer MT, Milev M,Cortesi P (2008) Clonal population structure of the chestnut blightfungus in expanding ranges in southeastern Europe. Mol Ecol 17:4446–4458

Novak-Agbaba S, Liović B, Pernek M (2000) Prikaz sastojina pitomogkestena (Castanea sativa Mill.) u Hrvatskoj i zastupljenosthipovirulentnih sojeva gljiveCryphonectria parasitica (Murr.) Barr.Radovi Šumarskog instituta Jastrebarsko 35(1):91–110

Peakall R, Smouse PE (2006) GENALEX 6: Genetic analysis in Excel.Population genetic software for teaching and research. Mol EcolNotes 6:288–295

Perica MĆ, Šola I, Urbas L, Smrekar F, KrajačićM (2009) Separation ofhypoviral double-stranded RNA on monolithic chromatographicsupports. J Chromatogr A 13:2712–2716

Peever TL, Liu Y-C, Cortesi P, Milgroom MG (2000) Variation intolerance and virulence in the chestnut blight fungus–hypovirusinteraction. Appl Environ Microbio 66:4863–4869

Prospero S, Conedera M, Heiniger U, Rigling D (2006) Saprophyticactivity and sporulation of Cryphonectria parasitica on dead chest-nut wood in forests with naturally established hypovirulence. Phy-topathology 96:1337–1344

Raymond M, Rousset (1995) GENEPOP (Version 1.2): Populationgenetic software for exact tests and ecumenicism. J Hered 86:248–249

Robin C, Lanz S, Soutrenon A, Rigling D (2010) Dominance of naturalover released biological control agents of the chestnut blight fungusCryphonectria parasitica in south-eastern France is associated withfitness-related traits. Biol Control 53:55–61

Robin C, Heiniger U (2001) Chestnut blight in Europe: diversity ofCryphonectria parasitica , hypovirulence and biocontrol. For SnowLand Res 76:361–367

Rousset (2008) GENEPOP 007: a complete reimplementation of theGENEPOP software for Windows and Linux. Mol Ecol Resour 8:103–106

Sotirovski K, Milgroom MG, Rigling D, Heiniger U (2006) Occurrenceof Cryphonectria hypovirus 1 in the chestnut blight fungus inMacedonia. Forest Pathol 36:136–143

Sotirovski K, Rigling D, Heiniger U, Milgroom MG (2011) Variation invirulence ofCryphonectria hypovirus 1 inMacedonia. Forest Pathol41:59–65

Wagner HW, Sefc KM (1999) IDENTITY 1.0. Centre for AppliedGenetics, University of Agricultural Sciences, Vienna

110 Tree Genetics & Genomes (2014) 10:101–110