Evidence of low levels of genetic diversity for the Phytophthora austrocedrae population in Patagonia, Argentina

Authors

  • M. L. Vélez,

    Corresponding author
    1. Protección Forestal, Centro de Investigación y Extensión Forestal Andino Patagónico (CIEFAP), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Universidad Nacional de la Patagonia San Juan Bosco (UNPSJB), Esquel, Argentina
    Search for more papers by this author
  • M. P. A. Coetzee,

    1. Department of Genetics, Forestry and Agricultural Biotechnology Institute (FABI), University of Pretoria, Pretoria, South Africa
    Search for more papers by this author
  • M. J. Wingfield,

    1. Department of Genetics, Forestry and Agricultural Biotechnology Institute (FABI), University of Pretoria, Pretoria, South Africa
    Search for more papers by this author
  • M. Rajchenberg,

    1. Protección Forestal, Centro de Investigación y Extensión Forestal Andino Patagónico (CIEFAP), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Universidad Nacional de la Patagonia San Juan Bosco (UNPSJB), Esquel, Argentina
    Search for more papers by this author
  • A. G. Greslebin

    1. Protección Forestal, Centro de Investigación y Extensión Forestal Andino Patagónico (CIEFAP), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Universidad Nacional de la Patagonia San Juan Bosco (UNPSJB), Esquel, Argentina
    Search for more papers by this author

Abstract

Phytophthora austrocedrae is a recently discovered pathogen that causes severe mortality of Austrocedrus chilensis in Patagonia. The high level of susceptibility of the host tree, together with the distribution pattern of the pathogen, have led to the hypothesis that P. austrocedrae was introduced into Argentina. The aim of this study was to assess the population structure of Paustrocedrae isolates from Argentina in order to gain an understanding of the origin and spread of the pathogen. Genetic diversity was determined based on amplified fragment length polymorphisms (AFLPs). In total, 48 isolates of Paustrocedrae were obtained from infected A. chilensis trees, representing the geographical range of the host. Four primer combinations were used for the AFLP analysis. Of the 332 scored bands, 12% were polymorphic. Gene diversity (h) ranged from 0·01 to 0·03; the Shannon index (I) ranged from 0·01 to 0·04. A high degree of genetic similarity was observed among the isolates (pairwise S values = 0·958–1; 0·993 ± 0·009, mean ± SD). A frequency histogram showed that most of the isolate pairs were identical. Principal coordinate analysis using three-dimensional plots did not group any of the isolates based on their geographical origin. The low genetic diversity (within and between sites) and absence of population structure linked to geographic origin, together with the aggressiveness of the pathogen and the disease progression pattern, suggest that Paustrocedrae might have been introduced into Argentina.

Introduction

Austrocedrus chilensis (ciprés de la cordillera, cypress) is an endemic tree in the Cupressaceae, found in southern Argentina and Chile. It is the most widely distributed species of the small number of conifers found in southern Argentina. Austrocedrus chilensis is found across 140 000 ha, in a wide variety of ecological niches, and in different soil types. It grows between 36° 30′ and 43° 35′S on the eastern, and between 32° 39′ and 44° S on the western, slopes of the Andes (Veblen et al., 1995). In Argentina, the tree grows in a 60–80 km-wide strip along the Andean foothills, across a broad moisture gradient (170 cm rainfall year−1 in the west to 50 cm year−1 in the east). In the west, A. chilensis can be found either in mixed stands with Nothofagus spp. or in pure stands on drier sites. In the north, A. chilensis trees grow in forests mixed with Araucaria araucana. They also grow in open, xeric forests or in isolated clumps at the limit of the Andean forest and the Patagonian steppe, acting as a barrier against the advancing desert. Austrocedrus chilensis is important not only because of its ecological functions but because of the high quality of its wood and its aesthetic value.

High levels of A. chilensis mortality were first detected in 1948 on Victoria Island (Nahuel Huapi National Park) in Neuquén Province. The disease was later detected during 1953 in a stand located near a forest nursery c. 150 km from Victoria Island, in Chubut Province. Since then, mortality has extended to almost all of the range of A. chilensis in Argentina, where the disease is referred to as ‘mal del ciprés’ (Greslebin & Hansen, 2010).

Phytophthora austrocedrae (Pythiales, Peronosporomycetes, Straminipila) was recently described as a new species and is believed to be the primary cause of A. chilensis mortality in Patagonia (Greslebin & Hansen, 2010). The disease is generally referred to as Austrocedrus Root Disease (ARD). Phytophthora austrocedrae is considered to be the main component in the complex pathology that leads to ‘mal del ciprés' (Greslebin et al., 2007; Greslebin & Hansen, 2010). The pathogen resides in clade 8 of the Phytophthora phylogenetic tree of Cooke et al. (2000). Phylogenetic analysis of the nuclear and mitochondrial genes has shown that Phytophthora syringae and Phytophthora obscura are the closest relatives of Paustrocedrae (Greslebin et al., 2007; Robideau et al., 2011; Grünwald et al., 2012). Among these species, Paustrocedrae is homothallic and is characterized by a combination of very slow growth, semipapillate, non-caducous and non-proliferating sporangia, oogonia with amphigynous antheridia, and low (17·5°C) optimal temperature for growth (Greslebin et al., 2007).

Phytophthora austrocedrae affects A. chilensis trees of all ages and is present across the native growth range of the trees in Argentina (Greslebin & Hansen, 2010). Symptoms of ARD include chlorosis, a wilting of the foliage and root rot. Trees, and especially seedlings, can die rapidly, in which case the foliage changes from chlorotic to a red colour. Some infected trees die slowly, first becoming chlorotic followed by a progressive defoliation and eventual tree death after several years (Filip & Rosso, 1999). Studies have shown that the disease originates in the root systems, and is associated with poor drainage and high moisture levels in soils (Havrylenko et al., 1989; Filip & Rosso, 1999; La Manna & Rajchenberg, 2004). Additional symptoms include necrotic lesions on the roots and stems that can affect the entire breadth of the phloem. The pathogen is able to colonize the xylem, which is evident by superficial discoloration of woody tissue and the presence of hyphae blocking rays and tracheids in xylem sections (Greslebin & Hansen, 2010; Vélez et al., 2012).

A possible mechanism to explain the decline of A. chilensis caused by Paustrocedrae has been presented by Vélez et al. (2012). Inoculation studies on 2-year-old saplings with Paustrocedrae led to a progressive and significant reduction in net photosynthesis, stomatal conductance and stem-specific hydraulic conductivity. The substantial negative impact on plant physiology was mainly attributed to the extensive death of the bark and cambium tissue, resulting in a disruption of phloem transport. Additionally, blockage of xylem transport by hyphal colonization, presence of resinous plugs, and death of xylem ray parenchyma, leads to a loss of hydraulic conductivity. Involvement of effectors secreted by Paustrocedrae, or host-derived responses to the pathogen, have also been suggested (Vélez et al., 2012).

The geographic origin of Paustrocedrae is unknown, although the impact of the pathogen suggests that it has been introduced into Argentina. Until recently, Paustrocedrae was not known from any region outside of Argentina, or on any other host than A. chilensis. However, in 2011, the pathogen was reported as the causal agent of root disease and mortality on Chamaecyparis nootkatensis in a public park in East Renfrewshire, UK (EPPO, 2011). This was followed by a report describing a damaging disease in rare native juniper bushes in the Upper Teesdale National Nature Reserve in the UK (Forestry Commission Great Britain, 2012). The source of Paustrocedrae could not be determined, but it was assumed that the pathogen had been introduced into these areas.

The presence of Paustrocedrae outside Argentina has raised concern regarding the origin of the pathogen. Because of its limited geographic range, and the high level of susceptibility of its primary host, it has been hypothesized that Paustrocedrae was introduced into Patagonia. However, the population genetic structure of the pathogen has not been established, precluding a more definitive view of this question. The aim of this study was, therefore, to assess the genetic diversity of Paustrocedrae in Argentina. The study was also undertaken to gain knowledge that would inform disease management strategies.

Materials and methods

Isolate selection and growth conditions

Isolates of Paustrocedrae were obtained from the culture collection of the Centro de Investigación y Extensión Forestal Andino Patagónico (CIEFAP). These isolates were collected during several surveys of declining A. chilensis stands in Patagonia, Argentina (Table 1). Sampling and isolation methods followed the protocols of Greslebin et al. (2007) and Greslebin & Hansen (2010). Forty-eight isolates from 19 stands were selected for analysis, reflecting the geographical range of A. chilensis mortality known at the time of the study, with a distance of c. 550 km between the two most distant locations (Fig. 1).

Table 1. Origin of Phytophthora austrocedrae cultures used in this study
SiteIsolateProvinceLocationDescriptionDate of isolation
1

Phy-255

Phy-263

Neuquén

40°9′45·79″S

71°20′44·88″ W

Lanin National Park, Cte. Díaz hill, San Martín de los AndesMay 2008
2Phy-308Neuquén

40°29′58·4″ S

71°21′6·39″W

Lanin National Park, Filo-Huaum lakeMay 2009
3

Phy-256

Phy-257

Phy-258

Neuquén

40°40′34·95″S

71°18′41·62″W

Lanin National Park, road from Confluencia to TrafulMay 2008
4

Phy-298

Phy-304

Phy-314

Neuquén

40°39′48·024″S

71°22′20·71″W

Lanin National Park, various stands of A. chilensis in the surroundings of TrafulMay 2009
Phy-299

40°40′9·66″S

71°21′6·39″W

Phy-300

Phy-305

Phy-312

40°40′0·007″S

71°21′36·12″W

5Phy-290Neuquén

40°58′37·03″S

71°30′59·02″W

Nahuel Huapi National Park, Puerto Totora and Bella Vista Hill, Victoria IslandJan 2009
Phy-292

40°57′41·26″S

71°31′52·25″W

6Phy-286Río Negro

41°13′44·32″S

71°25′11·12″W

Nahuel Huapi National Park, Gutierrez LakeOct 2008
7

Phy-276

Phy-278

Phy-279

Phy-281

Phy-318

Chubut

42°0′26·98″S

71°32′11·27″W

Epuyén, Golondrinas, Reserva Forestal del INTA

Sep 2008

May 2009

8

Phy-203

Phy-205

Phy-209

Phy-211

Phy-213

Phy-215

Phy-338

Chubut

42°46′29·97″S

71°32′11·27″W

Los Alerces National Park, Braese stream

Oct 2005

Jan 2010

9

Phy-219

Phy-221

Phy-223

Phy-225

Phy-232

Chubut

42°48′26·8″S

71°38′58·9″W

Los Alerces National Park, Quebrada del León streamOct 2005
10

Phy-243

Phy-244

Chubut

42°53′5·21″S

71°35′52·58″W

Los Alerces National Park, Las Rocas camping areaApr 2006
11Phy-294Chubut

43°9′55·8″S

71°42′18·4″W

Los Alerces National Park, seccional Río GrandeNov 2008
12

Phy-234

Phy-237

Chubut

43°7′34·36″S

71°33′46·10″W

Trevelin, Aldea Escolar, Estación Experimental INTAJan 2006
13

Phy-191

Phy-195

Phy-201

Phy-309

Chubut

43°12′55·5″S

71°32′50·9″W

Futaleufú, Trevelin, Rio Grande Valley, ‘La 106’ property

Sep 2005

Jun 2009

14

Phy-238

Phy-239

Chubut

43°11′39·7″S

71°28′23·2″W

Futaleufú, Trevelin, Río Grande valley, Nant y Fall fallsJan 2006
15

Phy-271

Phy-273

Chubut

43°34′31·11″S

71°41′10·64″W

Corcovado, Momberg propertyAug 2008
16

Phy-267

Phy-270

Chubut

43°33′51·97″S

71°38′54·83″W

Corcovado, Underwood propertyAug 2008
Figure 1.

Map showing the origin of Phytophthora austrocedrae isolates from Austrocedrus chilensis in forests of Patagonia. For location details, see Table 1.

In order to verify the identity of the isolates collected, DNA sequence comparisons were made for a subset of the isolates using the internally transcribed spacer regions (ITS1 and ITS2) and the cytochrome oxidase I gene (coxI). These gene regions were amplified and sequenced for isolates 13-Phy-309, 15-Phy-271 and 16-Phy-270 (GenBank accession numbers ITS: JX121857, JX121856, JX121855; coxI: JX448319, JX448318, JX448317, respectively). DNA was extracted as outlined below. The ITS region was amplified using PCR with the primers ITS4 and ITS5 (White et al., 1990); the coxI gene was amplified with primers FM 83 and FM 84 (Martin & Tooley, 2003). PCR reaction mixtures were the same for both regions. The mixture comprised dNTPs (0·25 mm each), MgCl2 (2·5 mm), PCR buffer, 0·1 μm of each primer, DNA (50–100 ng) and Taq DNA polymerase produced at the Forestry and Agricultural Biotechnology Institute (2·5 U). The PCR conditions for both regions were identical and followed those described by Martin & Tooley (2003). PCR products were separated on a 1% agarose gel stained with GelRed and visualized under UV light. PCR products were purified using a MSB Spin PCRapace purification kit (Invitek). DNA sequencing reactions were done using a BigDye Terminator v. 3.1 cycle sequencing kit (ABI) and sequences were determined on an ABI 3100 DNA automated sequencer. The regions were sequenced in both directions with the primers used for PCR amplification. Contigs were assembled and edited with the staden software package (Staden, 1996). DNA sequences were compared to those available in GenBank using blastn.

Four isolates of Psyringae were included in this study for comparative purposes. These isolates were obtained from soil samples collected in forests located in different regions of Patagonia using soil baiting. The isolates were: syr-40, obtained in Corcovado (near site 15); syr-157, from Los Alerces National Park (near site 9); syr-259, from Lanin National Park (near site 3); and syr-319, from Futaleufú, Río Grande valley (near site 13; Fig. 1).

DNA extraction

Isolates were maintained on clarified tomato juice agar (TA) at 16°C in the dark (Greslebin et al., 2007). To produce mycelium for DNA extraction, tomato juice broth was inoculated with agar plugs cut from the actively growing margins of fresh cultures. After 30 days’ growth at 16°C, the mycelium was harvested under sterile conditions, washed with sterile DNase/RNase-free water and centrifuged to discard the supernatant. Washed mycelium was transferred to Eppendorf tubes and DNA was extracted using the procedure described by Möller et al. (1992) with minor modifications. Mycelium was ground in liquid nitrogen, then 600 μL extraction buffer (200 mm Tris–HCl pH 8, 150 mm NaCl, 25 mm EDTA, 50 μg proteinase K, 100 μg RNase A) was added. This mixture was homogenized at 4 m s−1 for 20 s using a Fastprep FP120 (QBIOgene) homogenizer. The mixture was frozen in liquid nitrogen and then incubated for 60 min at 60°C. After incubation, 140 μL 5 m NaCl and 65 μL 10% CTAB were added, vortexed and incubated at 60°C for 10 min. One volume of chloroform/isoamyl alcohol (24:1) was added to the mixture, mixed gently and incubated at 4°C for 30 min. Samples were centrifuged at 11 700 g for 20 min at 4°C, the supernatant was transferred to new Eppendorf tubes and 0·55 volumes of isopropanol were added, followed by centrifugation at 11 700 g at room temperature. The supernatant was discarded and the pellet washed with 70% cold ethanol, at 6200 g for 5 min at 4°C. This step was repeated, the pellet dried and dissolved in 50 μL DNase/RNase-free water. Successful DNA extraction was confirmed by electrophoresis (1% agarose gel stained with GelRed and visualized under UV light). The DNA quality and concentration was determined using a NanoDrop ND1000 spectrophotometer (NanoDrop Technologies Inc.). The final concentration was adjusted to 20 ng μL−1.

Amplified fragment length polymorphism (AFLP) analysis

Of the 48 Paustrocedrae isolates, five were duplicated in the AFLP analysis to serve as internal controls and to ensure reproducibility. AFLP analysis was performed following the protocol of Vos et al. (1995) with minor modifications. Restriction digestion of genomic DNA (100 ng) was done using EcoRI and MseI. The resulting restriction fragments were ligated to the corresponding enzyme-specific oligonucleotide adapters (Vos et al., 1995). Preselective amplifications were performed with zero-base-addition EcoRI and MseI adapter-specific primers, using the PCR conditions described in De Vos et al. (2007). Successful restriction, ligation and preamplification were confirmed by gel electrophoresis (1% agarose gel stained with GelRed and visualized under UV light). The product of the preamplification step was diluted 1:10 in TE buffer (10 mm Tris–HCl pH 8, 0·1 mm EDTA) and used as the template for subsequent amplifications. Final selective amplifications were made using EcoRI and MseI primers with two-base-additions. Before the final amplification with all the samples, a screening test with five isolates (13-Phy-191, 8-Phy-205, 1-Phy-255, 16-Phy-267 and 5-Phy-290) and 16 primer combinations (E-AC, E-AT, E-AA and E-TT, in combination with each of M-AC M-CC, M-AG and M-TG) was done. The EcoRI primers were labelled with the infrared dyes, IRDye 700 or IRDye 800 (LI-COR). PCR conditions were: 13 cycles of 10 s at 94°C, 30 s at 65°C with a decrease of 0·7°C per cycle, 1 min at 72°C; followed by 23 cycles of 10 s at 94°C, 30 s at 56°C, 1 min at 72°C with an increase of 1 s per cycle; and a final elongation step of 1 min at 72°C.

Amplified fragment length polymorphism (AFLP) fragment analysis was performed on a model 4200 LI-COR automated DNA sequencer. Parameters for electrophoresis were: 1500 V, 35 mA, 35 W, 45°C, motor speed 3 and signal filter 3. Electrophoresis pre-run time was set to 30 min and run time to 4 h. Digital gel images obtained from the LI-COR system were analysed and scored manually. Based on the clarity of the resulting fingerprinting profiles (allowing unambiguous scoring), reproducibility and maximal variability obtained between the isolates, four primer combinations (E-AC/M-AC, E-AA/M-TG, E-AC/M-CC, E-CC/M-CC) were selected and used to evaluate the larger group of isolates. The AFLP procedure was replicated from the initial restriction digest step for all study isolates.

Each clearly resolved AFLP band was assigned a number based on its migration distance. For each isolate, the presence or absence for all monomorphic and polymorphic bands was scored visually. The presence (1) or absence (0) matrices of each primer combination and a matrix that included all data were analysed using popgene v. 1.31 (Yeh et al., 1997) to assess the genetic diversity based on Nei's gene diversity index (h; Nei, 1973) and Shannon's information index (I; Lewontin, 1972). Values of these indices for each set of primers were compared using one-way anova. In order to analyse the degree of genetic similarity between isolates, a pairwise distance matrix for each species was generated using the Jaccard, Dice and simple matching coefficients of similarity as calculated in infostat v. 2011 (Di Rienzo et al., 2011). Because isolate grouping was shown to be highly similar among the different coefficients, only results from Dice's index (Sj) are presented. Dice's coefficient measures the proportion of shared AFLP markers between each pair of isolates, while correcting for the dominance of the AFLP data by disregarding shared absence of bands. Distance dendrograms were then constructed using the infostat program. Phytophthora syringae was used as the out-group to root the dendrogram. The binary matrix was also analysed in paup, v. 4.0b10 (Swofford, 2002) to obtain bootstrap values at the nodes for a tree generated from Nei and Li genetic distances (1000 bootstrap replicates). To test possible grouping of the isolates by provenance and according to variables (band profiles), the binary matrix was also analysed in a principal coordinate analysis using infostat.

Results

In total, after analysis in duplicate, 332 clearly resolved AFLP bands were visually evaluated from the digital images generated using four different primer combinations (Table 2). The banding patterns for each of the five duplicated isolates, which served as internal controls, were the same, indicating intra-assay AFLP reproducibility. Of the 332 bands that were scored, only 40 (12%) were polymorphic. The total number of bands scored per primer combination ranged from 65 (E-CC/M-CC) to 97 (E-AC/M-AC). The lowest percentage of polymorphic loci (8·2%) was generated by the E-AC/M-AC primer combination, whilst the E-AC/M-CC primer combination yielded the highest percentage (16·2%) of polymorphic loci (Table 2). Twenty-four rare alleles (i.e. present in five or fewer isolates) were identified and none of these could be associated with the region or stands from which the isolates were collected.

Table 2. Total number of amplified fragment length polymorphism (AFLP) bands, proportion of monomorphic and polymorphic bands, gene diversity (h) and Shannon index (I) obtained with each primer pair combination
AFLP primer combinationTotal no. of bandsMonomorphic bandsPolymorphic bandsPolymorphic bands (%) h a I
  1. a

    Gene diversity according to Nei (1973).

E-AC/M-AC978988·20·010·02
E-AA/M-TG96861010·40·010·01
E-AC/M-CC74621216·20·010·03
E-CC/M-CC65551015·40·030·04
Total3322924012·00·020·03

Genetic analysis of the bands yielded gene diversities (h) that ranged from 0·01 to 0·03. The Shannon index (I) ranged from 0·01 to 0·04 (Table 2). Comparison of the values obtained from the four different primers sets using anova showed that they were not statistically different (h,= 0·065; I,= 0·092).

A high degree of genetic similarity was found for the P. austrocedrae isolates. The pairwise similarity (Sj) values were close to 1, ranging from 0·958 to 1 (0·993 ± 0·009; mean ± SD) for the majority of comparisons (Fig. 2). The largest distances were obtained between isolates 8-Phy-203, 16-Phy-270 and 15-Phy-271 (0.04), all originating from Chubut Province (8-Phy-203 from Los Alerces National Park, and 16-Phy-270 and 15-Phy-271 from two sites geographically close to each other in Corcovado; Fig. 1). Isolates obtained from the most distant locations (Corcovado and San Martín de los Andes) yielded high similarity values (0·969–0·998). Isolates originating from the same stand, in general, yielded similarity values (Sj values: 0·993 ± 0·006) that were comparable with those from different stands (Sj values: 0·993 ± 0·004).

Figure 2.

Distribution of Dice's coefficient of similarity (Sj) for Phytophthora austrocedrae. Frequency histogram showing the distribution of all pairwise Sj for the 48 isolates (= 1128). Frequency indicates percentage of comparisons with a given Sj value; Sj values of 1 indicate pairs with 100% similarity.

Dendrograms based on similarity indices from each individual AFLP primer combination were concordant and differed only slightly (data not shown). Data from the four primer combinations were, therefore, concatenated to generate a single dendrogram (Fig. 3). The isolates resided in one large group that was supported by a 100% bootstrap value (Fig. 3). Within this group, 45 isolates grouped in a single cluster (92% bootstrap support), while three isolates (16-Phy-270, 15-Phy-271 and 13-Phy-309) grouped separately but with no bootstrap support (54%; Fig. 3). DNA sequence comparisons for these three isolates with sequences on GenBank showed high similarity (maximal identity from 99 to 100%, E-values <6e−177) to Paustrocedrae and therefore were assumed to be Paustrocedrae. These three isolates derived from stands relatively close to each other in terms of the geographic distribution of the host. Stands in Corcovado (sites 15 and 16) were very close to each other, whilst ‘La 106’ property in Rio Grande Valley (site 13) is about 60 km from those in Corcovado. Although there seemed to be a geographic pattern in this case, the bootstrap did not support this partition; in addition to this, other isolates obtained in the same stands grouped with isolates collected in other distant stands. This is the case for isolates 13-Phy-191, 13-Phy-195 and 13-Phy-201 from ‘La 106’ property, and 15-Phy-273 and 16-Phy-267 from the two stands in Corcovado (Fig. 3). Among the 45 isolates that grouped together, a subgroup of five isolates from two different stands located c. 90 km from each other could be distinguished (89% bootstrap support; Fig. 3). This subgroup was, in turn, subdivided into two according to isolate provenance (90% bootstrap support). The subgroup of five isolates was obtained from plants that were relatively close to each other (stands with small areas); therefore an effect caused by sampling could have occurred. However, as stated above, there does not seem to be geographic partitioning because other isolates from the same sites (13-Phy-309 and 8-Phy-209, 8-Phy-211, 8-Phy-213, 8-Phy-215, 8-Phy-338) grouped apart and mixed with isolates obtained in other stands and sites (Fig. 3). Thus, there was no overall clear partitioning of genetic diversity that corresponded to geographic origin. In the principal coordinate analyses, three-dimensional plots did not resolve any clear grouping of isolates on the basis of geographical origin (Fig. 4).

Figure 3.

Dendrogram generated from AFLP distance matrix (1–Sj) for 48 Phytophthora austrocedrae isolates obtained along the geographical range of Austrocedrus chilensis. The first number of each isolate name indicates its site of origin (Table 1). Groups of isolates with zero branch length comprise samples with identical AFLP genotypes. Values at the branches are bootstrap values. Phytophthora syringae was used as the out-group.

Figure 4.

Principal coordinate analysis generated in infostat from the binary matrix obtained from AFLP patterns of 48 Phytophthora austrocedrae isolates obtained along the geographical range of Austrocedrus chilensis. The first number of each isolate name indicates its site of origin (Table 1).

Discussion

This is the first study to consider the genetic structure and diversity of Paustrocedrae in Argentina. Results revealed high levels of within-species genetic similarity and no evidence of partitioning of genetic diversity among the collection sites. This suggests that the pathogen represents a single population with low heterogeneity in Patagonia and a possible explanation is that Paustrocedrae might have been introduced into Argentina. The fact that isolates in this study were collected from across the broad distribution of A. chilensis, during different times of the year and in different years, adds credence to this view.

The fact that the only known host of Paustrocedrae in Argentina exhibits high levels of susceptibility is consistent with the behaviour of an introduced pathogen, and the results obtained in this study indicate that there is a possibility that Paustrocedrae might be exotic to the region. Generally, exotic species, especially near-obligate or obligate pathogens such as Phytophthora spp., undergo population bottlenecks when they are introduced into new areas (Goodwin, 1997). Therefore, lower genetic variation is expected for these populations in comparison to those where they occur naturally. Moreover, endemic species exhibit population genetic structures that reflect the presence of geographic or environmental barriers, or a combination of both, to gene flow (Goodwin, 1997). In this study, a low genetic variation and absence of geographical differentiation was observed for isolates of Paustrocedrae, supporting the notion that it might be introduced into Argentina.

Another plausible explanation for the finding of the low genetic diversity in the pathogen population is that Paustrocedrae could have recently evolved from another species. With regard to this, Psyringae and Pobscura are the closest relatives of Paustrocedrae, by evaluation of nuclear and mitochondrial genes (Greslebin et al., 2007; Robideau et al., 2011; Grünwald et al., 2012). Only Psyringae has been previously isolated in Patagonia (Greslebin et al., 2005). More studies are needed to investigate this possibility.

Previous studies on Phytophthora spp. using AFLP analysis, and where pathogens have been suggested to be introduced into new areas, have reported similar levels of genetic diversity. The present study found that the percentage of polymorphic loci was slightly lower for Paustrocedrae than was reported for Phytophthora nemorosa in forests in the USA (Linzer et al., 2009) or for Phytophthora pinifolia in Chile (Durán et al., 2010), but was slightly higher than obtained for Phytophthora quercina in Europe (Cooke et al., 2005) or Phytophthora pseudosyringae (Linzer et al., 2009) in the USA and Europe. These species are homothallic, with the exception of Ppinifolia that is sterile. The very low levels of genetic diversity and the lack of geographic structuring that is unexpected for endemic organisms, including those that are able to reproduce homothallically, allowed the authors to hypothesize that they had been introduced. Although the AFLP primers used in these studies are not the same as those used in this present work, the relative number of polymorphic loci can be compared, because it is a parameter to indicate the proportion of variable loci. Therefore, a similar conclusion that Paustrocedrae has been introduced into Patagonia could be justified.

The first detection of mortality of A. chilensis trees occurred during 1948 in Piedras Blancas Bay on Victoria Island in Neuquén Province. The subsequent development of the disease is characteristic of an introduced pathogen that has encountered a highly susceptible host grown over an extended area. Victoria Island is known for the introduction of many exotic woody plants from different continents, especially during the 1920s and 1930s (Simberloff et al., 2002). This, together with the fact that the first appearance of A. chilensis mortality occurred in this area, leads the authors to believe that Paustrocedrae was introduced to the island on infected plants. The results of this study support this view.

The recent discovery that Paustrocedrae has caused mortality of C. nootkatensis and J. communis in the UK (EPPO, 2011; Forestry Commission Great Britain, 2012) is relevant to the present study. These tree species belong to the same family (Cupressaceae) as A. chilensis. Chamaecyparis nootkatensis is endemic to the USA and J. communis occurs naturally in a wide distribution from the south Arctic to around 30° N latitude in USA, Europe and Asia, in the northern hemisphere. The source of Paustrocedrae in the UK has not been determined, but it has been suggested that the pathogen was also introduced (EPPO, 2011). Clearly, it is impossible to determine the origin of Paustrocedrae without having isolates from a hypothetical natural host or area of origin. The same situation is true for many Phytophthora spp. such as Pramorum, Ppinifolia, Plateralis, among others, that have unexpectedly appeared in new areas but for which the likely areas of origin are unknown (Hansen et al., 2000; Ivors et al., 2004; Durán et al., 2010). In this regard, global surveys for Phytophthora spp. must continue and, in the case of Paustrocedrae that appears to be a conifer-specific pathogen, such surveys should include areas where related conifers occur naturally. For example, the recent discovery of Plateralis, a serious root pathogen of Chamaecyparis lawsoniana in western North America, infecting Chamaecyparis obtusa var. formosana, a native species in Taiwan (Webber et al., 2012), represents a good example of how the area of origin of a Phytophthora sp. might be found.

The possibility that Paustrocedrae could be native to Patagonia cannot be entirely excluded. The pathogen is a homothallic organism, therefore successful gene flow across the geographic range of the pathogen, together with self-fertile reproduction, could also lead to a population with low genetic diversity. However, this hypo-thesis is not consistent with the high levels of susceptibility of A. chilensis and the disease progression pattern that shows a progression from affected sites to healthy sites (La Manna & Matteucci, 2012).

The high level of genetic similarity found for Paustrocedrae isolates in this study could be accounted for if the pathogen had undergone a host shift from another native plant. However, a host jump following anthropogenic introduction (Anderson et al., 2004; Woolhouse et al., 2005) would be more plausible when one considers the ecology and epidemiology of other Phytophthora spp. (Hansen et al., 2000; Ivors et al., 2004; Brasier et al., 2005; Prospero et al., 2007; Linzer et al., 2009; Durán et al., 2010) and characteristics of emerging infectious diseases of plants in general (Anderson et al., 2004). It has been shown that host shifts are more likely when the original host and the new host are phylogenetically closely related (Gilbert & Webb, 2007). In this regard, at least nine conifer species, phylogenetically related to A. chilensis, have been introduced into Victoria Island where the disease was first reported and Paustrocedrae is found. The areas of origin of these trees would be good targets for surveys aimed at discovering the origin of Paustrocedrae.

Acknowledgements

This work was supported by Ministerio Nacional de Ciencia, Tecnología e Innovación Productiva (Argentina) and South Africa through an International Cooperation Program (MINCYT_DST SA/10/02) and Agencia Nacional de Promoción Científica y Técnica (ANPCyT, FONCyT) of Argentina (PICTO 36776, PICT 08-0579). Administración de Parques Nacionales of Argentina and landowners of cypress forests are acknowledged for allowing work on their properties. The authors also thank members of the Tree Protection Co-operative Programme (TPCP) and the NRF/DST Centre of Excellence in Tree Health Biotechnology (CTHB), South Africa, for financial support.

Ancillary