Jukuri, open repository of the Natural Resources Institute Finland (Luke) All material supplied via Jukuri is protected by copyright and other intellectual property rights. Duplication or sale, in electronic or print form, of any part of the repository collections is prohibited. Making electronic or print copies of the material is permitted only for your own personal use or for educational purposes. For other purposes, this article may be used in accordance with the publisher’s terms. There may be differences between this version and the publisher’s version. You are advised to cite the publisher’s version. This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): Hernán Capador, Berit Samils, Juha Kaitera and Åke Olson Title: Genetic evidence for sexual reproduction and multiple infections of Norway spruce cones by the rust fungus Thekopsora areolata Year: 2020 Version: Published version Copyright: The Author(s) 2020 Rights: CC BY 4.0 Rights url: http://creativecommons.org/licenses/by/4.0/ Please cite the original version: Capador, H, Samils, B, Kaitera, J, Olson, Å. Genetic evidence for sexual reproduction and multiple infections of Norway spruce cones by the rust fungus Thekopsora areolata . Ecol Evol. 2020; 10: 7389– 7403. https://doi.org/10.1002/ece3.6466. Ecology and Evolution. 2020;10:7389–7403.  |  7389www.ecolevol.org 1  | INTRODUC TION Rust fungi are obligate parasites, of plants, that belong to the order Pucciniales of the phylum Basidiomycota. It is a large group with about 7,000 species belonging to this order. Rust fungi colonize plants from fern to higher plants and cause diseases of economic significance in agriculture, horticulture, and forestry. These fungi have complicated life cycles with up to five different spore stages. The life cycles often involve alternation between two unrelated host plant species to be complete (heteroecious life cycle). However, some species can complete their life cycle on a single host plant (autoecious life cycle). Although many rust species require two Received: 17 January 2020  |  Revised: 4 May 2020  |  Accepted: 6 May 2020 DOI: 10.1002/ece3.6466 O R I G I N A L R E S E A R C H Genetic evidence for sexual reproduction and multiple infections of Norway spruce cones by the rust fungus Thekopsora areolata Hernán Capador1 | Berit Samils1 | Juha Kaitera2 | Åke Olson1 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2020 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 1Department of Forest Mycology and Plant Pathology, Swedish University of Agricultural Sciences, Uppsala, Sweden 2Natural Resources Institute Finland, University of Oulu, Oulu, Finland Correspondence Åke Olson, Department of Forest Mycology and Plant Pathology, Swedish University of Agricultural Sciences, Almas Allé 5, Box 7026, 750 07 Uppsala, Sweden. Email: ake.olson@slu.se Funding information Petterson-Grebbe Foundation; Svenska Forskningsrådet Formas, Grant/Award Number: 2017-01489 Abstract Rust fungi are obligate parasites, of plants, with complex and in many cases poorly known life cycles which may include host alteration and up to five spore types with haploid, diploid, and dikaryotic nuclear stages. This study supports that Thekopasora areolata, the causal agent of cherry-spruce rust in Norway spruce, is a macrocyclic heteroecious fungus with all five spore stages which uses two host plants Prunus padus and Picea abies to complete its life cycle. High genotypic diversity without population structure was found, which suggests predominantly sexual reproduc- tion, random mating and a high gene flow within and between the populations in Fennoscandia. There was no evidence for an autoecious life cycle resulting from ae- ciospore infection of pistillate cones that would explain the previously reported rust epidemics without the alternate host. However, within cones and scales identical multilocus genotypes were repeatedly sampled which can be explained by vegetative growth of the fertilized mycelia or repeated mating of mycelium by spermatia of the same genotype. The high genotypic diversity within cones and haplotype inference show that each pistillate cone is infected by several basidiospores. This study pro- vides genetic evidence for high gene flow, sexual reproduction, and multiple infec- tions of Norway spruce cone by the rust fungus T. areolata which expands the general understanding of the biology of rust fungi. K E Y W O R D S aecia, cherry-spruce rust, cone rust, forest pathology, life cycle, Norway spruce, Picea abies, seed orchard 7390  |     CAPADOR et Al. unrelated host plant species to complete their life cycles, rust spe- cies usually have a narrow and specific host range (Cummins & Hiratsuka, 2003). The different spore and life stages include haploid, diploid, and dikaryotic nuclear stages. The reproduction mode of dif- ferent species of rust fungi varies from strict clonal to obligate sex- ual while many species utilize both sexual and asexual reproduction. The obligate nature and complex life cycles make rust fungi difficult to study, and therefore, the life cycle of many species is often poorly understood. Conifers are a major component of boreal and alpine forests that cover large areas of northern hemisphere. These boreal forests are of great ecological and economic importance for many countries. Norway spruce [Picea abies (L.) Karst] is one of the most economically important conifer for the forest industry in Europe and among the most widely planted species. In Sweden and Finland, more than 200 and 100 million Norway spruce seedlings, respectively, are planted in managed forests annually (Haapanen, Jansson, Bräuner Nielsen, Steffenrem, & Stener, 2015; Himanen, 2016). About half of these seedlings originate from plant nurseries which use genetically im- proved seeds from seed orchards planted with superior plants from breeding programs. However, mass production of improved seeds for P. abies is usually lower than the demand (Haapanen et al., 2015), mainly due to irregular flowering and cone production between years and occurrence of pests, that is insects and fungal diseases (Kaitera, Hiltunen, Kauppila, Pitkäranta, & Hantula, 2014). In Finland, good seed crops have repeatedly suffered from cherry-spruce rust caused by Thekopsora areolata (Fr.) Magnus (Kaitera, 2013), a rust fungus which grows in P. abies cones and shoots and reduces seed viability up to 10-fold (Kaitera & Tillman-Sutela, 2014). One of the main strategies to increase the production of im- proved seeds is by intensive seed orchard management (Haapanen et al., 2015). However, cherry-spruce rust imposes a challenge in disease control due to the complexity of its life cycle and repro- duction biology. Thekopsora areolata has a 2-year long life cycle with five different spore stages and host alternation between Picea spp. and Prunus spp. (Figure 1) (Gäumann, 1959; Kuprevich & Tranzschel, 1957). In the spring, monokaryotic basidiospores infect P. abies pistillate cones, where mycelium grows through the axis of the developing cone and forms spermogonia composed of several flexuous hyphae and spermatia on the outer surface of scales. Rust spermatia and flexuous hyphae from genetically distinct individual fuse and develop into dikaryotic mycelium after anastomosis. The resulting heterokaryon will in turn form aecia, globoid structures composed of a thick-walled peridium that bears inside chains of dikaryotic aeciospores. Aecia usually crowd the adaxial and often the abaxial surface of the majority of cone scales within months after fusion, but break and release spores in the spring the next few years (Figure 2) (Kaitera & Tillman-Sutela, 2014). These aeciospores infect Prunus leaves, where uredinia with dikaryotic urediniospores are formed. Urediniospores reproduce clonally and can spread through re-infection of Prunus leaves during the same season until autumn. Thereafter, the fungus overwinters in fallen Prunus leaves where the fungus develops telia and teliospores. Basidia and basidiospores will germinate from the overwintered teliospores as the result of karyo- gamy and meiosis, and basidiospores are disseminated in the spring to infect Picea (Gäumann, 1959). A better understanding of the complex life cycle of T. areolata will provide new insights in the rust fungus' biology, which are es- sential for development of seed orchard management strategies. The loss of a part of the life cycle is a common repeated and in- dependent event in rusts (Ono, 2002). The pine rust Cronartium flaccidum (Alb. & Schwein) G. Winter is heteroecious alternat- ing between two-needle pines and various herbaceous plants, while species of the Peridermium genus (syn. Endocronartium), P. pini (Willd.: Pers. Lév.) and P. harknessii (J.P. Moore), are au- toecious rusts spreading from pine to pine (Hansen, Lewis, & Chastagner, 2018; Kaitera, Hiltunen, & Samils, 2012; Kaitera, F I G U R E 1   Life cycle of Thekopsora areolata. F I G U R E 2   A Picea abies cone and a cone scale infected by Thekopsora areolata. (a) A diseased P. abies cone infected by T. areolata. (b) The rust grows throughout the scale of a cone forming aecia (arrow).      |  7391CAPADOR et Al. Hiltunen, & Hantula, 2015). Today, C. flaccidum and P. pini are considered as two forms of the same species, Cronartium pini (Willd.) Jørst. (Hantula, Kasanen, Kaitera, & Moricca, 2002; Samils, Ihrmark, Kaitera, Stenlid, & Barklund, 2011), but in inoculation experiments aeciospores of the heteroecious C. pini (syn. C. flac- cidum) have been shown to infect only alternate hosts and aecio- spores of the autoecious C. pini (syn. P. pini) only Pinus spp. and not vice versa (Kaitera & Nuorteva, 2008). It has been hypothesized that short-cycling in rusts might be a strategy in northern latitudes to survive in the absence of alternate hosts under harsh condi- tions (Savile, 1953). For T. areolata, it has been hypothesized that aeciospores from infected spruce cones might re-infect spruce flowers or developing young cones directly, that is having an au- toecious reproduction (Kaitera, 2013; Kaitera et al., 2014; Kaitera & Tillman-Sutela, 2014). The hypothesis is based on the observa- tion of severe rust damage without the presence of the alternate host P. padus (Kaitera, Tillman-Sutela, & Kauppi, 2009). An autoe- cious life cycle would explain rust epidemics in remote areas in the north, where Prunus trees are either rare or lacking. If spruce infection with aeciospores was possible, it would mean that inocu- lum was often available close to the spruce flowers and developing cones. Infected cones remain attached in tree canopy for several years, and germination experiments have shown that aeciospores remain viable for at least 4 years (Kaitera & Tillman-Sutela, 2014). In such a case, removal of all infected cones from a seed orchard would greatly reduce the risk of disease outbreaks and improve the health and quality of seeds. Another consequence would be that T. areolata would reproduce clonally on spruce in addition to the clonal reproduction with urediniospores on Prunus spp. Direct studies of the reproduction mode in rust fungi are dif- ficult, but indirect inferences can be made with help of molecular markers as shown in other fungal systems. For instance, the use of microsatellite or simple sequence repeat (SSR) markers with a correct sampling strategy can help to improve knowledge in many areas of fungal population biology (Dutech et al., 2007; Lim, Notley- McRobb, Lim, & Carter, 2004) such as population differentiation (Ali et al., 2014), genotypic and genetic diversity (Barres, Dutech, Andrieux, Halkett, & Frey, 2012), heterozygosity and life-history characteristics like mode of reproduction (Danies et al., 2014), and host selection (Berlin, Djurle, Samils, & Yuen, 2012). Molecular markers have also proved to be useful in population genetic stud- ies of rusts causing diseases on trees, for example, on Cronartium spp. (Hamelin, Beaulieu, & Plourde, 1995; Samils et al., 2011), Melampsora spp. (Barres et al., 2008; Samils, Lagercrantz, Lascoux, & Gullberg, 2001), and Austropuccinia psidii (G. Winter) Beenken (Sandhu, Karaoglu, Zhang, & Park, 2016). In this study, we investigate the reproduction mode and popula- tion genetic structure of the rust fungus T. areolata in Fennoscandia using a hierarchical sampling strategy and recently developed micro- satellite markers (Capador, Samils, & Olson, 2018). We analyze the genetic and genotypic variation of T. areolata across different hierar- chical levels in P. abies seed orchards and stands as well as genotype distribution in cones. Additionally, we apply haplotype inference methods to predict the most likely haplotype in the previous step of the life cycle. The combined results are used to make inferences on the mode of reproduction and epidemiology of T. areolata. 2  | MATERIAL S AND METHODS 2.1 | Partially nested hierarchical sampling Location level: Picea abies cones with aecia were collected from 7 dif- ferent locations in Sweden, Norway, and Finland (Figure 3a). At each location, 30 cones were collected, from which one scale with aecia per cone and one aecium per scale were randomly sampled. At tree level: A more extensive sampling was made at the seed orchard in Ålbrunna (Sweden), where 100 cones with aecia were sampled from 6 different trees at a distance of ca. 20 m–600 m from each other (Figure 3b). From each cone, one scale with aecia and one aecium per scale were randomly sampled (Figure 3b). At cone level: 10 cones with aecia were randomly sampled from two locations in Sweden (a-1 and a-4) and split longitudinally to select 10 scales across each cone, from which 10 aecia per scale were randomly sampled (100 aecia per cone) (Figure 3c). At scale level: 3 individual scales with aecia from cones from two different locations in Sweden (a-1 and a-4) were thoroughly sampled (ca. 40 aecia per scale; Figure 3d). 2.2 | Single aecium genotyping Cone scales with aecia were dipped for 45 min in a solution of 30% H2O2, with a drop of Tween 80 and rinsed three times in deion- ized water. Thereafter, a single aecium was picked with the help of a hooked needle. Because an aecium has been shown to contain a bulk of genetically identical dikaryotic aeciospores, each aecium was treated as one individual (Capador et al., 2018; Rodriguez-Algaba, Sørensen, Labouriau, Justesen, & Hovmøller, 2017). Rodriguez- Algaba et al., 2017). DNA was extracted from each aecium following the protocol of Capador et al. (2018). The samples were genotyped with 8 polymorphic microsatellites markers; Tha9, Tha61, Tha91, Tha92, Tha96, Tha105, Tha136, and Tha137, and microsatellite amplification and scoring was performed as formerly described (Capador et al., 2018) except for a subset of samples which were analyzed with DreamTaq (Thermo Fisher) instead of PIR00 (Sigma). 2.3 | Data filtering Due to the partly-nested sampling strategy and to prevent bi- ases caused by uneven population sizes, the dataset (in GenAlEx format) was curated by removing uninformative loci (minor allele frequency < 5%) and individuals with missing data in more than one loci. Before further analyses, the dataset was divided in four smaller datasets according to its hierarchical level (Figure 3) using the R package poppr 2.8.2 (Kamvar, Tabima, & Grünwald, 2014). 7392  |     CAPADOR et Al. F I G U R E 3   Partially nested sampling strategy. (a) Location level: 1. Bredinge (Öland, Sweden), 2. Söregärde (Småland, Sweden), 3. Ås (Akershus, Norway), 4. Ålbrunna (Uppland, Sweden), 5. Rörby (Uppland, Sweden), 6. Domsjöänget (Västerbotten, Sweden), and 7. Muhos (Northern Ostrobothnia, Finland). (b) Tree level: Individual trees sampled in Ålbrunna (A-F). (c) Cone level: Cones were split longitudinally and infected scales were taken from 3 cm segments (red horizontal dashed lines) from both sides of the central axis (red vertical dashed line). (d) Scale level: Arrow indicates one aecium. 1 2 43 5 6 7 100 km 100 m 10 mm 1 cm (a) (b) (c) (d) A B C DE F 4 3 2 1 L R      |  7393CAPADOR et Al. At the location level, a maximum of 30 cones per location and one individual per cone was included. At the tree level, 100 cones in Ålbrunna were included with a maximum of one aecium per cone. At the cone level, only those cones with 100 aecia sampled per cone were included. At the scale level, each scale genotyped thoroughly was analyzed separately. Tree, cone, and scale levels were curated separately for uninformative loci (minor allele fre- quency < 10%) and missing data to improve the accuracy of clonal- ity estimates. 2.4 | Genotypic diversity Genotypic diversity and population structure estimates were cal- culated at each hierarchical level using the R package poppr 2.8.2 (Kamvar et al., 2014). Genotypic diversity was estimated using the Shannon-Wiener diversity index (H; Shannon, 2001) which gives the number of multilocus genotypes (MLGs) observed in relation to the number of samples analyzed and with the corrected Simpson index (λ; Simpson, 1949). The function Psex was used to estimate the probability of finding one MLG multiple times by chance (and not by clonal reproduction) at the location level according to Parks, and Werth, (1993) and Arnaud-Haond, Duarte, Alberto, and Serrão (2007). 2.5 | Clonality The standardized Index of Association (rbarD) was used as a tool to detect clonal reproduction across hierarchical levels in R pack- age poppr 2.8.2 (Kamvar et al., 2014). rbarD estimates the degree of linkage between markers and tests to what extent individuals that are the same at one locus are more likely than random to be the same at other loci (Agapow & Burt, 2001). p values for rbarD (P.rbarD) were computed based on a 999 one-side permutation test (Kamvar et al., 2014). Furthermore, a conservative Psex value was calculated in RClone 1.0.2 to account for possible departures from Hardy–Weinberg equilibrium (Arnaud-Haond et al., 2007). 2.6 | Population structure Population differentiation at all hierarchical levels was tested with analysis of molecular variance (AMOVA) (Excoffier, Smouse, & Quattro, 1992). The differentiation was calculated based on the Euclidean distance matrix at each level, and variation within samples could be interpreted as heterozygosity, because it measured differ- entiation between alleles at each loci across each sample. p values were computed based on 999 permutations (Excoffier et al., 1992; Kamvar, Brooks, & Grünwald, 2015; Kamvar et al., 2014). To identify possible population structure, discriminant analysis of principal com- ponents (DAPC) was used at the location hierarchical level. DAPC combined a principal component analysis (PCA) with a discriminant analysis (DA) to weight variability between previously defined populations more than variability within them (Jombart, Devillard, & Balloux, 2010). Additionally, Fst was calculated with GenAlEx (Peakall & Smouse, 2006) to estimate pairwise differences between populations. To analyze further the population structure, we used the program STRUCTURE 2.3.4 that uses a Bayesian approach to assign individuals into groups (clusters) based on genetic similar- ity (Pritchard, Stephens, & Donnelly, 2000). STRUCTURE was run by varying the number of clusters (K) from 2 to 5. The admixture model assuming no linkage between the loci and without a priori information on populations was applied. For each K, we made 3 re- peated simulations with a burn-in period of 500,000 iterations of the Markov Monte Carlo Chain (MCMC) and a run length of 500,000 MCMC iterations. Results were compiled using Structure Harvester (Earl & von Holdt, 2012) and bar plots constructed using the pro- gram Distruct (Rosenberg, 2004). 2.7 | Haplotype inference Haplotypes were inferred using the Excoffier-Laval-Balding (ELB) al- gorithm implemented in Arlequin 3.5 (Excoffier & Lischer, 2010) for the cone and scale hierarchical levels. This algorithm reconstructed unknown gametic phases by performing iterations and phase up- dates made on the basis of haplotype frequencies in the population and linkage between neighboring loci. In this case, the order of the markers was unknown, but they were assumed to be largely unlinked (Capador et al., 2018); therefore, any given order was irrelevant in this case. The value γ = 0.1 was used since very low recombination was expected in the analyzed hierarchical levels (see discussion), α = 0.01 was used to account for the possibility of finding a yet unob- served haplotype, and parameter ε = 0.1 was used for microsatellite data. The resulting best haplotypes for each individual were format- ted with and visualized in R. 3  | RESULTS In total, 951 single aecia were collected from 7 locations in Fennoscandia and across 4 hierarchical levels; location, tree, cone, and scale. A strict filtering to remove genotypes with missing data in more than one loci resulted in a total of 520 successfully genotyped individuals with 7 polymorphic microsatellite markers (marker Tha109 with an MAF < 0.5% was considered uninforma- tive). For analyses of the population structure, 30 cones were col- lected from seven locations (populations) in Fennoscandia, and samples consisted of one aecium per cone. The genotyping suc- cess was between 30% and 83% (Table 1). The mean number of alleles (Ne) per population varied between 2.75 and 4.25 with the highest Ne and mean allele richness (Ar), as well as the highest number of private alleles (N pri) being found in the Söregärde pop- ulation (Sweden; Table 1). The lowest Ne, Ar, and N pri were found in the Domsjöänget population (Sweden; Table 1). The Simpsons 7394  |     CAPADOR et Al. diversity index (λ) and Shannon-Wiener diversity index (H) were highest in Muhos (Finland) and lowest in Ås (Norway) while gene diversity (Hexp) was highest in Söregärde and lowest in Bredinge (Sweden; Table 1). 3.1 | Random mating test for clonality The number of MLGs equaled the number of individuals in all popu- lations except in Bredinge where two individuals had the same MLG. In total, only three MLGs were sampled multiple times over the whole dataset (Table 1). To test the probability that these identical MLGs originated from distinct sexual events, and thus, were differ- ent genets and not clones/ramets of the same MLG, a test was per- formed using a conservative Psex estimation. Test results showed that only MLG 47 (found once in Ålbrunna, twice in Bredinge and once Söregärde) had a low probability (p = 3.3e−10) to be found by chance four times, while the other two are most likely a result of common alleles. To estimate the mode of reproduction of T. areo- lata and the probability of random mating, rbarD was calculated and the probability of observing its value was estimated with 999 permutations (Table 1). The rbarD values were low for all popula- tions, and the p values were nonsignificant (Table 1). In all locations, the hypothesis of random mating and no linkage between markers was accepted. 3.2 | Population structure and variation Of the total genetic diversity, most of the variation was attributed to the aecium (83%) and between aecia (16%) while location and coun- try explained <1% of the variation found (Table 2). The results of AMOVA indicated that population differentiation was insignificant between countries or locations (p > .05). To reveal any structure of T. areolata populations at country and location level, a discriminate analysis of principle components (DAPC) was performed. Samples from different locations and countries overlapped each other with a small separation between them (Figure 4). As an exception, the Söregärde population differed slightly from the other populations. In addition, Pairwise Fst values were calculated to further analyze the differentiation between T. areolata populations from different loca- tions (Table S1) as well as a Bayesian cluster analysis using software STRUCTURE (Figure S1) but no significant differences were found. 3.3 | Genotype and diversity parameters for tree, cone, and scale The aecial samples from different cones from six trees in Ålbrunna showed high genotypic diversity, and the number of samples equaled the number of MLGs found in all trees (Table 3). The Simpsons diver- sity index (λ) and Shannon-Wiener MLG diversity (H) ranged from 1.1 to 2.9 and 0.67 to 0.94, respectively, reflecting the increasing number of samples. The gene diversity was highest in the trees AL-A and AL-D (Hexp = 0.4) and lowest in the trees AL-E and AL-F. TA B L E 1   Population diversity parameters for the populations of Thekopsora areolata Parameters Ålbrunna (SE) Bredinge (SE) Rörby (SE) Söregärde (SE) Domsjöänget (SE) Ås (NO) Muhos (FI) Total Fennoscandia N 22 18 12 14 11 10 25 112 MLG 22 17 12 14 11 10 25 107 Ne 3.88 3.250 3.88 4.250 2.75 3.00 4.125 6.875 N pri 3 2 3 6 1 3 3 NA Ar 3.10 2.69 3.42 3.54 2.52 2.89 3.12 H 3.091 2.813 2.485 2.639 2.398 2.303 3.219 4.644 λ 0.955 0.938 0.917 0.929 0.909 0.900 0.960 0.990 Hexp 0.386 0.333 0.460 0.479 0.370 0.391 0.392 0.401 rbarD −0.033 0.028 0.015 −0.016 −0.001 −0.029 0.011 −0.004 p.rbarD .946 .155 .352 .677 .466 .725 .312 .662 Abbreviations: Ar, Mean allele richness; FI, Finland; H, Shannon-Wiener Index of MLG diversity (Shannon, 2001); Hexp, Nei's unbiased gene diversity; MLG, multilocus genotypes; N pri, number of private alleles; N, number of individuals; Ne, mean number of alleles;NA, Not analysed; NO, Norway; p.rbarD, p-value for rbarD; rbarD, standardized index of association; SE, Sweden; λ, Simpson's index. TA B L E 2   The distribution of variance within and between locations in Thekopsora areolata Hierarchical level Variance (%) p value Phi value Between countries −0.149 .559 −0.001 Between locations, within country 0.672 .186 0.006 Between aecia, within location 16.384 .001** 0.165 Within samples 83.092 .001** 0.169 **Significant at p < .001.      |  7395CAPADOR et Al. Contrastingly, at the cone and scale level identical MLGs were found repeatedly within the cones and scales (Table 3). Cone BR_202 con- tained three MLGs among the 14 aecia sampled (Table 3). However, in cone Br_231 each sampled aecia had a unique MLG (Table 3). In general, there was large variation in the number of MLGs per cone but in the scales the number of different MLGs was low (Table 3). To estimate the mode of reproduction of T. areolata and the probabil- ity of random mating at the tree, cone, and scale hierarchical levels, rbarD was calculated (Table 3). In five of the trees in Ålbrunna, the null hypothesis of random mating and no linkage between markers was accepted while in tree AL-A a low but significant (p = .046) link- age among markers was found (Table 3). For the cones, the null hy- pothesis was rejected (p < .01) at two out of ten cones and at both scales analyzed (Table 3). The AMOVA indicates that no significant variation was found be- tween trees but variation occurred between aecia within trees (19%) and within aecia (82%, p < .001; Table 4). However, at the cone level, 19% (p < .001) of the variation could be attributed between cones and 83% (p < .001) within aecia while <1% of the variation was found between aecia within cones (Table 4). F I G U R E 4   Discriminate analysis of principle components of Thekopsora areolata. Samples from each location are marked with distinct colors. Ålbrunna Bredinge Rörby Söregärde Domsjöänget Ås Muhos DA eigenvalues 7396  |     CAPADOR et Al. 3.4 | Haplotype inference and pedigree analysis within cones and scales At the cone and scale levels repeated MLGs were usually dom- inant in number (Figures 5 and 6). Most cones were dominated by one to three MLGs (Figure 5). The repeatedly sampled MLGs were usually located on more than one scale and throughout the cone (Figure 5). The samples from cone BR_202 represented al- most exclusively (12 out of 14) by MLG 145 (Figure 5). The number of MLGs per scale for samples AL_88_1, AL_14_1, and BR_221_1 were 4, 2, and 6, respectively (Table 3). Two of the scales AL_88_1 and AL_14_1 had one dominating MLG, while in BR_221_1 two of the MLGs occurred several times (Figure 6a). Aecia of the domi- nant MLGs were spatially clustered, which also was true for the unique MLGs (Figure 6a). Haplotype inference was used at the cone and scale level to trace back the most likely haploid genotypes (i.e., haplotypes) of the homokaryotic basidiospores that colonized the cone, and their prob- able mating events resulting in the observed heterokaryotic aecia. According to the haplotype inference, the dominant aeciospore MLG shared haplotypes with some of the other MLG in the cone in several cases (Figure 7, e.g., AL_69, AL_89, BR_202, BR_231). For some other cones, less haplotypes were dominant, and the present MLGs seemed to have arisen through combinations with single, less represented haplotypes (Figure 7: AL_6, AL_37, BR_255). According to their predicted haplotypes, fertilization events between few haploid individuals could explain the genotypic diver- sity observed. In two of the scales just one mating event at each scale gave rise to dominating aecia (Figure 6b). In scale BR_221_1, the most frequent MLG was monomorphic at all loci (Figure 6, scale 221, MLG 480). The MLGs 342, 393, and 492 in the cones AL_88, BR_213, and BR_255, respectively, were also monomorphic at all loci according to the haplotype inference (Figure 7). 4  | DISCUSSION Details of the reproduction and epidemiology of many rust fungi are lacking, mainly due to the difficulties to study them because of their biotrophic nature and complex life cycle. In this study, we used popu- lation genetics tools to clarify features of the life cycle of Thekopsora areolata. We provided genetic evidence for high genotypic diver- sity, sexual reproduction, and multiple infection of pistillate cones Hierarchical level N MLGs H λ Hexp rbarD p.rbarD Tree AL_A 6 6 1.792 0.833 0.403 0.177 .046* AL_B 3 3 1.099 0.667 0.383 −0.346 .999 AL_C 8 8 2.079 0.875 0.394 0.006 .409 AL_D 6 6 1.792 0.833 0.407 −0.079 .884 AL_E 18 18 2.890 0.944 0.356 −0.039 .911 AL_F 17 17 2.833 0.941 0.367 0.003 .419 Cone AL_6 20 11 2.398 0.909 0.403 0.087 .060 AL_37 12 10 2.303 0.900 0.358 −0.051 .867 AL_46 20 9 2.197 0.889 0.381 0.319 .001** AL_69 23 10 2.303 0.900 0.413 0.084 .060 AL_89 19 10 2.303 0.900 0.247 −0.019 .551 BR_202 14 3 1.099 0.667 0.248 0.250 .179 BR_213 24 15 2.708 0.933 0.334 −0.016 .650 BR_231 18 18 2.890 0.944 0.413 0.035 .149 BR_285 33 18 2.890 0.944 0.376 0.020 .234 BR_255 20 12 2.485 0.917 0.304 0.125 .035* Scale AL_88_1 26 4 0.484 0.213 0.237 0.669 .001** AL_14_1 24 2 0.173 0.080 0.152 NA NA BR_221_1 18 6 1.303 0.630 0.107 0.399 .001** Abbreviations: H, Shannon-Wiener Index of MLG diversity (Shannon, 2001); Hexp, Nei's unbiased gene diversity; MLG, Number of multilocus genotypes; N, number of samples; NA, Not analysed; p.rbarD, p-value for rbarD; rbarD, standardized index of association; λ, Simpson's index. *Significant at p < .05. **Significant at p < .001. TA B L E 3   Genotypes and population diversity parameters for the samples of Thekopsora areolata at tree, cone, and scale levels      |  7397CAPADOR et Al. of Norway spruce by T. areolata as well as detailed information of spread of fungal genotypes within cones and scales. The rust T. areolata represents one population in Fennoscandia, with no evidence for any significant substructure or genetic differ- entiation between any of the populations analyzed. There is a large overlap between the populations in the DAPC plot, although the Swedish population Söregärde showed some difference from the other populations. However, pairwise Fst analysis did not support dif- ferentiation between the Söregärde population and any of the other populations in Fennoscandia. The difference found in the population from Söregärde can be explained by the high number of private al- leles unique to this population. The reason why this population has higher number of private alleles than the other populations analyzed remains unanswered and needs additional studies to be resolved. The lack of population structure is likely to be maintained through a high level of gene flow between the populations, which is not sur- prising due to the nature of urediniospores as robust spores capable of dispersing long distances, even across continents (Brown, 2002). One similar example is Melampsora epitea Thümen on willows which is genetically highly diverse with small geographic differentiation in Sweden, probably due to high dispersal ability of urediniospores (Samils, Stepien, Lagercrantz, Lascoux, & Gullberg, 2001). Likewise, populations of M. larici-populina Kleb. from European poplars show low genetic differentiation (Barres et al., 2008). In addition to uredin- iospores, aeciospores can be disseminated long distances leading to low geographical differentiation among C. pini populations in Finland, Sweden, and southern Europe (Hantula et al., 2002; Hantula, Niemi, Kaitera, Jalkanen, & Kurkela, 1998; Samils et al., 2011). Furthermore, the wide distribution of the both hosts of T. areolata in northern Europe increases the chances for long-distance dispersed spores to infect susceptible Prunus and augments genotypic diversity, since sexual reproduction and recombination occur on this host. At each seed orchards and stands, the number of MLGs equaled the number of samples in all populations except in Bredinge (Sweden) where one MLG was sampled twice. In addition, two more MLGs were found in more than one population. Analyses of the likelihood of finding the same genotype twice suggest that MLG 47 was not likely to be found by chance, and this suggests that MLG 47 is a true clone (Arnaud-Haond et al., 2007). If correct, this result would imply that aeciospores of T. areolata are able to infect directly from pine to pine without passing the alternate host as well as spread over large distances since clone MLG 47 was found in three different seed or- chards. However, further studies are needed to confirm this single case. It cannot be excluded that individuals within MLG 47 are not true clone mates but appear to be due to limited resolution of the markers used. The high genotypic diversity, lack of population structure, and the low level of genetic linkage between markers were found in this study are consistent with random mating, suggesting that sexual re- production is regularly occurring in T. areolata in Fennoscandia. In other rusts like M. larici-populina, similar evidence has been found on populations known to reproduce sexually (Barres et al., 2012), where sexually derived basidiospores with unique genotypes are the main drivers of genotypic diversity, and they are also responsible for the observed linkage equilibrium (Rodriguez-Algaba, Walter, Sørensen, Hovmøller, & Justesen, 2014). In the cones, the genotypic diversity was lower with both unique and identical MLGs found within the same cone, and devi- ation from random mating was observed in a few cases even after clone correction (Table 4). Furthermore, we showed that there is usually one or a few dominant genotypes per cone. The dominating genotypes are present on different scales both along the central axis as well as on neighboring scales at the same level (Figure 5). These genotypes often share haplotypes as indication of family relationships in the same cone (Figure 7), which corroborates the deviation from random mating after clone correction. This pattern in MLG distribution sheds light into the reproduction mode of T. areolata: Firstly, usually more than two haplotypes per cone were inferred (Figure 7) indicating that multiple basidiospore infections can take place in the same cone. However, as only a few dom- inant MLGs were found throughout the cones, the basidiospore infections were local and vegetative spreading occurred only after the dikaryotization event had occurred. This is in line with Kuprevich and Tranzschel (1957), who reported that mycelium of T. areolata grows through the axis of the cone. Secondly, only a few genotypes dominated in scales and shared haplotypes with less common MLGs (Figure 6). The pattern we observed in cones and scales was similar to other rust pathosystems. In Cronartium, many aecia within lesions had different MLGs (Kasanen, Kaitera, & Hantula, 2000; Samils et al., 2011), and diversity of Puccinia gram- inis Pers., P. coronata Corda, and P. striiformis var. tritici Westend was high within aecial clusters in Berberis vulgaris L. (Berlin, Samils, & Andersson, 2017; Rodriguez-Algaba et al., 2017). A study of the white pine blister rust, Cronartium ribicola show that 70–75% of the genetic variability is found between aecia within cankers while the rest is found between canker within site or between sites (Hamelin, 1996). It has been hypothesized that the high variability TA B L E 4   The distribution of variance within and between tree and cone hierarchical levels in Thekopsora areolata Hierarchical level Variance (%) p value Phi value Between trees, within Location −0.8 .710 −0.007 Between aecia, within trees 18.7 .001** 0.186 Within aecia 82.1 .001** 0.179 Clone corrected at cone levela  Between cones, within location 18.7 .001** 0.187 Between aecia, within cone −1.5 .691 −0.018 Within aecia 82.5 .001** 0.174 aAt this level, only cones with more than 100 aecia genotyped were included. **Significant at p < .001. 7398  |     CAPADOR et Al. M LG .2 81 M LG .2 88 M LG .1 82 M LG .2 79 M LG .2 80 M LG .2 84 M LG .2 87 M LG .3 06 M LG .3 07 M LG .4 29 M LG .4 31 0 2 4 6 8 10 12 0 2 4 6 8 10 12 0 2 4 6 8 10 12 1 2 3 4 5 L R Side S eg m en t C ou nt C ou nt C ou nt 0 2 4 6 8 10 12 C ou nt C ou nt 0 2 4 6 8 10 12 C ou nt C ou nt 6 1 2 3 4 L R Side S eg m en t 69 M LG .1 84 M LG .3 82 M LG .3 90 M LG .1 72 M LG .1 83 M LG .2 08 M LG .3 83 M LG .3 85 M LG .3 88 M LG .3 89 1 2 3 4 L R Side S eg m en t 37 M LG .2 72 M LG .3 6 M LG .5 5 M LG .6 7 M LG .6 9 M LG .2 19 M LG .2 22 M LG .2 24 M LG .2 71 M LG .3 31 1 2 3 4 L R Side S eg m en t 46 M LG .2 41 M LG .2 43 M LG .2 37 M LG .1 71 M LG .2 30 M LG .2 40 M LG .3 63 M LG .3 71 M LG .3 72 1 2 3 4 L R Side S eg m en t 89 M LG .3 39 M LG .3 73 M LG .1 32 M LG .3 35 M LG .3 36 M LG .3 37 M LG .3 41 M LG .3 42 M LG .3 53 M LG .4 96 0 2 4 6 8 10 12 C ou nt 0 2 4 6 8 10 12 1 2 3 4 L R Side S eg m en t 202 M LG .1 45 M LG .1 55 M LG .2 06 1 2 3 4 L R Side Side S eg m en t 213 M LG .3 63 M LG .3 92 M LG .2 31 M LG .1 05 M LG .1 51 M LG .1 78 M LG .2 29 M LG .2 32 M LG .2 35 M LG .2 38 M LG .2 42 M LG .2 54 M LG .3 66 M LG .3 93 M LG .4 32 1 2 3 L R S eg m en t 285 M LG .6 3 M LG .3 26 M LG .3 28 M LG .2 1 M LG .3 5 M LG .3 7 M LG .4 7 M LG .6 4 M LG .6 6 M LG .7 3 M LG .8 8 M LG .1 17 M LG .2 11 M LG .2 27 M LG .2 49 M LG .3 25 M LG .3 27 M LG .3 29 0 2 4 6 8 10 12 Count 12345      |  7399CAPADOR et Al. is due to single different cross-fertilization events (fusion of sper- matia or dikaryotization) in the same spermogonium. Berlin et al. (2017) showed that MLGs at each aecial cluster of Puccinia shared at least one haplotype. This strengthens the hypothesis of aecium formation due to single cross-fertilizations between different flexuous hyphae from the same spermogonium (identical haploid genotype) and different spermatia (different haploid genotypes), resulting in different MLGs within the same aecial cluster (Berlin et al., 2017). However, this study showed that even within scales and cones, aecia shared more than one common haplotype, which implied spermatization with spermatia originating from more than one different spermogonia. Most of the aecia sampled were heterozygotic; however, a few samples diverged from this pattern by being homozygous for all marker loci, which could be indicative of self-fertilization (Figure 6, scale 221, MLG 480). However, the relative few markers used in this study suggest that the probability to find homozygous genotypes by chance with the allele frequencies present in Ålbrunna and Bredinge is high (p > .05). In general, it has been assumed that cross-fertiliza- tion is needed for dikaryotization and that insects play a key role on the transport of spermatia to cross-fertilize spermogonia (Naef, Roy, Kaiser, & Honegger, 2002). This has been supported by ex- periments in rust fungi Uromyces pisi (Pers.) Schrot, where insects were attracted by the honeydew matrix (in which the spermatia are F I G U R E 5   Frequency and distribution of MLGs across cones. Only samples with known location in the cone are included in figure. L and R represent samples from left and right side of the cone, and the segment represents the longitudinal position in the cone. For each cone, MLGs found more than once are presented in specific colors (the same color represents the same MLG in both the circles and bars, within one cone). The size of the circles represents the number of individuals with identical MLG. Gray circles represent MLGs found only once. F I G U R E 6   Fine scale genotyping and haplotype inference of Thekopsora areolata in three different scales. (a) Location of genotyped aecia within the scale. (b) Haplotype inference pedigree for the genotyped aecia. IH, inferred haplotype, MLG, multilocus Genotype. The color of the circles over the aecium in (a) corresponds to the color of the MLG in b). White asterisks denote MLGs homozygote for all loci. 12 4 12 2 12 8 12 9 1 2 88 16 5 16 6 2 1 14 48 0 48 3 48 2 48 5 48 7 49 0 3 5 2 1 221 MLG IH MLG IH MLG IH (b) (a) * 7400  |     CAPADOR et Al. F I G U R E 7   MLG and haplotype inference pedigrees in ten different cones. Haplotype inference was used to infer the haplotypes (IH) of each multilocus genotype (MLG) at the cone level. In color, MLGs repeated more than once in each cone, White asterisks denote MLGs homozygote for all loci. 28 1 28 8 18 2 27 9 28 0 28 4 28 7 30 6 30 7 42 9 43 1 3 4 11 1 6 27 2 36 55 67 69 21 9 22 2 22 4 27 1 33 1 3 1 6 37 24 1 24 3 23 7 17 1 23 0 24 0 36 3 37 1 37 2 4 5 3 8 12 46 18 4 38 2 39 0 17 2 18 3 20 8 38 3 38 5 38 8 38 9 5 7 1 6 2 69 33 9 37 3 13 2 33 5 33 6 33 7 34 1 34 2 35 3 49 6 4 5 3 6 89 14 5 15 5 20 6 2 4 202 36 3 39 2 23 1 10 5 15 1 17 8 22 9 23 2 23 5 23 8 24 2 25 4 36 6 39 3 43 2 4 3 1 2 17 8 16 213 4 5 10 3 10 4 10 7 11 5 11 6 12 1 13 3 14 3 17 7 20 2 20 5 21 8 25 3 34 6 34 7 43 3 3 1 8 18 19 24 231 49 4 11 3 10 0 33 2 99 10 1 10 2 33 0 48 9 49 1 49 2 49 7 6 5 9 11 3 4 8 255 63 32 6 32 8 21 35 37 47 64 66 73 88 11 7 21 1 22 7 24 9 32 5 32 7 32 9 2 3 19 4 21 1 10 11 285 MLG IH IH IH IH IH MLG MLG MLG MLG MLG IH IH IH IH IH MLG MLG MLG MLG * * * *      |  7401CAPADOR et Al. embedded) and were required for the formation of the dikaryon (Pfunder & Roy, 2000). Additionally, it has been shown that either presence of insects or manual cross-fertilization among different spermogonia resulted in successful aecia formation, while caged and self-fertilized spermogonia formed a significantly lower number of aecia (Pfunder & Roy, 2000). We found identical homozygous MLGs in four cones and one scale analyzed, which suggests that the ho- mozygotic MLGs in T. areolata could be a result of self-fertilization between genetically identical spermatia of the same genet or lack of resolution due to the low number of SSR markers and the uneven allele distribution. This interesting observation needs further inves- tigation to confirm or discarded the presences of self-fertilization in T. areolata. An autoecious life cycle for T. areolata through aeciospore re-in- fections on Picea has been suggested as one possible explanation for the repeated infections found in isolated seed orchards with no telial hosts within or in close proximity of the plantation (Kaitera, 2013; Kaitera et al., 2014; Kaitera & Tillman-Sutela, 2014). Such autoe- cious replication should result in identical fungal genotypes spread among cones like in the case of the autoecious rust, P. pini (Hantula et al., 1998; Samils et al., 2011). Our data support that T. areolata is a macrocyclic heteroecious rust, which utilizes two host plants and all five spore stages as suggested earlier (Sato & Takahashi, 1970). Actually, it has been reported that aeciospores can germinate in young shoots without forming other structures (Kuprevich & Tranzschel, 1957), but this seems to be a rare finding since results of artificial inoculations have been negative (pers. comm.) and the biological mechanisms how aeciospores could colonize cones are un- known. However, there is evidence of colonization of Picea shoots by T. areolata as reported by Hietala, Solheim, and Fossdal (2008) who found the fungus in 100 symptomatic young seedlings of P. abies by quantitative PCR (Hietala et al., 2008). Among other rusts, the autoecious stem rust on Scots pine, P. pini, is able to infect the host by aeciospores, grow and sporulate from spermogonia and aecia in host tissues systematically for years (Kaitera & Nuorteva, 2008). In the case of P. pini, it is unknown, if spermogonia are nonfunctional or the spermatization takes place by self-fertilization. If autoecism exists in T. areolata, it is probably rare and insignificant or occurs at low frequency. In the case of T. areolata, successful inoculations on the host by aeciospores could confirm autoecism of the rust. Much biological and genetic evidence is still needed to fully understand the mode of reproduction of T. areolata in P. abies. 5  | CONCLUSIONS In this study, we have clarified several parts of the life cycle of the fungi T. areolata, the causal agent for cherry-spruce rust, which is a serious problem in Norway spruce seed orchards. Thekopsora areo- lata has a heteroecious life cycle, and the majority of infections of spruce cones are a result from basidiospore infections and mating. The main objective of this study was to investigate the population genetics and reproductive biology of the rust fungus T. areolata. The analysis together with a hierarchical sampling showed apparent ran- dom mating and common sexual events. Moreover, it showed that the rust populations in the locations sampled belonged to the same metapopulation with no obvious genetic structure. Additionally, it highlighted the complex mixed mode of reproduction of T. areolata in P. abies cones with multiple infections and vegetative spread in the cone and scale after fertilization. Overall, these findings showed high genotypic diversity and high levels of gene flow in T. areolata in Fennoscandia. Although no genetic differentiation is found be- tween locations, probably due to long-distance spread by aecio- or urediniospores, local disease pressure is most likely supported by P. padus trees in close proximity of the seed orchards and should be eradicated if possible. ACKNOWLEDG MENTS We would like to thank Maria Jonsson for technical assistance with the DNA extraction and genotyping and Halvor Solheim (Norwegian Forest Research Institute) and Curt Almqvist (Forest Research Institute of Sweden) for assistance in cone collections. The compa- nies Bergvik skog, Holmen, Sveaskog, and Södra skogsägarna are ac- knowledge for providing access to their seed orchards. This project was supported by Formas; Grant 2017-01489 and the Petterson- Grebbe foundation. CONFLIC T OF INTERE S T Authors declare no conflict of interest. AUTHOR CONTRIBUTIONS Hernán Capador: Conceptualization (supporting); data curation (lead); formal analysis (lead); investigation (supporting); methodology (supporting); project administration (supporting); validation (equal); visualization (lead); writing–original draft (equal); writing–review and editing (supporting). Berit Samils: Conceptualization (equal); data curation (supporting); formal analysis (supporting); funding acquisi- tion (supporting); investigation (supporting); methodology (equal); project administration (supporting); software (equal); supervision (equal); validation (equal); writing–original draft (supporting); writ- ing–review and editing (supporting). Juha Kaitera: Conceptualization (supporting); investigation (equal); project administration (support- ing); supervision (supporting); validation (supporting); writing–origi- nal draft (supporting); writing–review and editing (supporting). Åke Olson: Conceptualization (lead); data curation (supporting); formal analysis (supporting); funding acquisition (lead); investigation (equal); methodology (equal); project administration (lead); resources (equal); supervision (lead); validation (supporting); visualization (supporting); writing–original draft (lead); writing–review and editing (lead). DATA AVAIL ABILIT Y S TATEMENT Information on sample location and microsatellite genotypes are available at Dryad https://doi.org/10.5061/dryad.2ngf1 vhk9 ORCID Åke Olson https://orcid.org/0000-0001-8998-6096 7402  |     CAPADOR et Al. R E FE R E N C E S Agapow, P.-M., & Burt, A. (2001). Indices of multilocus linkage dis- equilibrium. Molecular Ecology Notes, 1, 101–102. https://doi. org/10.1046/j.1471-8278.2000.00014.x Ali, S., Gladieux, P., Leconte, M., Gautier, A., Justesen, A. F., Hovmøller, M. S., … de Vallavieille-Pope, C. (2014). Origin, migration routes and worldwide population genetic structure of the wheat yellow rust pathogen Puccinia striiformis f.sp. tritici. PLoS Path, 10(1), e1003903. https://doi.org/10.1371/journ al.ppat.1003903 Arnaud-Haond, S., Duarte, C. M., Alberto, F., & Serrão, E. A. (2007). Standardizing methods to address clonality in popula- tion studies. Molecular Ecology, 16(24), 5115–5139. https://doi. org/10.1111/j.1365-294X.2007.03535.x Barres, B., Dutech, C., Andrieux, A., Halkett, F., & Frey, P. (2012). Exploring the role of asexual multiplication in poplar rust epidemics: Impact on diversity and genetic structure. Molecular Ecology, 21(20), 4996–5008. https://doi.org/10.1111/mec.12008 Barres, B., Halkett, F., Dutech, C., Andrieux, A., Pinon, J., & Frey, P. (2008). Genetic structure of the poplar rust fungus Melampsora la- rici-populina: Evidence for isolation by distance in Europe and recent founder effects overseas. Infection, Genetics and Evolution, 8(5), 577– 587. https://doi.org/10.1016/j.meegid.2008.04.005 Berlin, A., Djurle, A., Samils, B., & Yuen, J. (2012). Genetic varia- tion in Puccinia graminis collected from oats, rye, and barberry. Phytopathology, 102(10), 1006–1012. https://doi.org/10.1094/ PHYTO -03-12-0041-R Berlin, A., Samils, B., & Andersson, B. (2017). Multiple genotypes within aecial clusters in Puccinia graminis and Puccinia coronata: Improved understanding of the biology of cereal rust fungi. Fungal Biology and Biotechnology, 4(1), 3. https://doi.org/10.1186/s4069 4-017-0032-3 Brown, J. K. M. (2002). Aerial Dispersal of Pathogens on the Global and Continental Scales and Its Impact on Plant Disease. Science, 297(5581), 537–541. https://doi.org/10.1126/scien ce.1072678 Capador, H., Samils, B., & Olson, Å. (2018). Development of microsat- ellite markers for Thekopsora areolata, the causal agent of cherry spruce rust. Forest Pathology, 48, e12413. https://doi.org/10.1111/ efp.12413 Cummins, G. B., & Hiratsuka, Y. (2003). Illustrated genera of rust fungi (3rd ed.). St. Paul, MN: APS Press. Danies, G., Myers, K., Mideros, M. F., Restrepo, S., Martin, F. N., Cooke, D. E. L., … Fry, W. E. (2014). An ephemeral sexual population of Phytophthora infestans in the Northeastern United States and Canada. PLoS One, 9(12), e116354. https://doi.org/10.1371/journ al.pone.0116354 Dutech, C., Enjalbert, J., Fournier, E., Delmotte, F., Barrès, B., Carlier, J., … Giraud, T. (2007). Challenges of microsatellite isolation in fungi. Fungal Genetics and Biology, 44(10), 933–949. https://doi. org/10.1016/j.fgb.2007.05.003 Earl, D. A., & von Holdt, B. M. (2012). STRUCTURE HARVESTER: A web- site and program for visualizing STRUCTURE output and implement- ing the evanno method. Conservation Genetics Resources, 4, 359–361. https://doi.org/10.1007/s1268 6-011-9548-7 Excoffier, L., & Lischer, H. E. L. (2010). Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources 10, 564–567. https://doi. org/10.1111/j.1755-0998.2010.02847.x Excoffier, L., Smouse, P. E., & Quattro, J. M. (1992). Analysis of molecu- lar variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction data. Genetics, 131(2), 479–491. https://doi.org/10.1007/s0042 4-009-0730-7 Gäumann, E. (1959). Die Rostpilze Mitteleuropas. Beiträge zur Kryptogamenflora der Schweitz (Vol. 12). Bern, Germany: Buchdruckerei Buchler and Co. Haapanen, M., Jansson, G., Bräuner Nielsen, U., Steffenrem, A., & Stener, L.-G.-S. (2015). In L.-G. Stener (Ed.). The status of tree breeding and its potential for improving biomass production - A review of breeding activ- ities and genetic gains in Scandinavia and Finland, (1–56). Gävle: Gävle Offset. Hamelin, R. C. (1996). Genetic diversity between and within cankers of the white pine blister rust. Phytopathology, 86, 875–879. Hamelin, R. C., Beaulieu, J., & Plourde, A. (1995). Genetic diversity in populations of Cronartium ribicola in plantations and natural stands of Pinus strobus. Theoretical and Applied Genetics, 91(8), 1214–1221. https://doi.org/10.1007/BF002 20932 Hansen, E. M., Lewis, K. J., & Chastagner, G. A. (2018). (Ed.), Compendium of conifer diseases (2nd ed., p. 232). St. Paul, MN: APS Press. Hantula, J., Kasanen, R., Kaitera, J., & Moricca, S. (2002). Analyses of genetic variation suggest that pine rusts Cronartium flaccidum and Peridermium pini belong to the same species. Mycological Research, 106, 203–209. https://doi.org/10.1017/S0953 75620 1005457 Hantula, J., Niemi, E. M., Kaitera, J., Jalkanen, R., & Kurkela, T. (1998). Genetic variation of the resin-top fungus in Finland as determined by random amplified microsatellites (RAMS). European Journal of Forest Pathology, 28(6), 361–371. https://doi. org/10.1111/j.1439-0329.1998.tb011 990.x Hietala, A. M., Solheim, H., & Fossdal, C. G. (2008). Real-time PCR-based monitoring of DNA pools in the tri-trophic interaction between Norway spruce, the rust Thekopsora areolata, and an opportunistic ascomycetous Phomopsis sp. Phytopathology, 98(1), 51–58. https:// doi.org/10.1094/PHYTO -98-1-0051 Himanen, K. (2016). Maaottelu Kuusen Siemenviljelyksillä. Metsätieteen Aikakauskirja, 3–4, 175–180. https://doi.org/10.14214 /ma.5709 Jombart, T., Devillard, S., & Balloux, F. (2010). Discriminant analysis of principal components: A new method for the analysis of geneti- cally structured populations. BMC Genetics, 11(1), 94. https://doi. org/10.1186/1471-2156-11-94 Kaitera, J. (2013). Thekopsora and Chrysomyxa cone rusts damage Norway spruce cones after a good cone crop in Finland. Scandinavian Journal of Forest Research, 28(3), 217–222. https://doi.org/10.1080/02827 581.2012.727024 Kaitera, J., Hiltunen, R., & Hantula, J. (2015). Cronartium rust sporulation on hemiparasitic plants. Plant Pathology, 64, 738–747. https://doi. org/10.1111/ppa.12291 Kaitera, J., Hiltunen, R., Kauppila, T., Pitkäranta, M., & Hantula, J. (2014). Fruiting and sporulation of Thekopsora and Chrysomyxa cone rusts in Picea cones and Prunus leaves. Forest Pathology, 44(5), 387–395. https://doi.org/10.1111/efp.12114 Kaitera, J., Hiltunen, R., & Samils, B. (2012). Alternate host ranges of Cronartium flaccidum and Cronartium ribicola in northern Europe. Botany-Botanique, 90(8), 694–703. https://doi.org/10.1139/B2012 -039 Kaitera, J., & Nuorteva, H. (2008). Inoculations of eight Pinus spe- cies with Cronartium and Peridermium stem rusts. Forest Ecology and Management, 255, 975–981. https://doi.org/10.1016/j. foreco.2007.10.005 Kaitera, J., & Tillman-Sutela, E. (2014). Germination capacity of Thekopsora areolata aeciospores and the effect of cone rusts on seeds of Picea abies. Scandinavian Journal of Forest Research, 29(1), 22–26. https://doi.org/10.1080/02827 581.2013.844851 Kaitera, J., Tillman-Sutela, E., & Kauppi, A. (2009). Seasonal fruiting and sporulation of Thekopsora and Chrysomyxa cone rusts in Norway spruce cones and alternate hosts in Finland. Canadian Journal of Forest Research, 39(9), 1630–1646. https://doi.org/10.1139/X09-070 Kamvar, Z. N., Brooks, J. C., & Grünwald, N. J. (2015). Novel R tools for analysis of genome-wide population genetic data with emphasis on clonality. Frontiers in Genetics, 6, 1–10. https://doi.org/10.3389/ fgene.2015.00208      |  7403CAPADOR et Al. Kamvar, Z. N., Tabima, J. F., & Grünwald, N. J. (2014). Poppr: An R pack- age for genetic analysis of populations with clonal, partially clonal, and/or sexual reproduction. PeerJ, 2, e281. https://doi.org/10.7717/ peerj.281 Kasanen, R., Kaitera, J., & Hantula, J. (2000). The genetic composition of Peridermium pini and Cronartium flaccidum cankers on Scots pine as revealed by two multi-allelic loci. Forest Pathology, 30, 221–230. https://doi.org/10.1046/j.1439-0329.2000.00209.x Kuprevich, V., & Tranzschel, V. (1957). Cryptogamic Plants of the USSR (I). Moskva-LeninGrad: Academy of Sciences of the USSR, Moscow- Leningrad, USSR: . Komarov Institute of Botany. Lim, S., Notley-McRobb, L., Lim, M., & Carter, D. A. (2004). A compari- son of the nature and abundance of microsatellites in 14 fungal ge- nomes. Fungal Genetics and Biology, 41(11), 1025–1036. https://doi. org/10.1016/j.fgb.2004.08.004 Naef, A., Roy, B. A., Kaiser, R., & Honegger, R. (2002). Insect-mediated reproduction of systemic infections by Puccinia arrhenatheri on Berberis vulgaris. New Phytologist, 154(3), 717–730. https://doi. org/10.1046/j.1469-8137.2002.00406.x Ono, Y. (2002). The diversity of nuclear cycle in microcyclic rust fungi (Uredinales) and its ecological and evolutionary implica- tions. Mycoscience, 43(5), 421–439. https://doi.org/10.1007/s1026 70200062 Parks, J. C., & Werth, C. R. (1993). A study of spatial features of clones in a population of bracken fern, Pteridium aquilinum (Dennstaedtiaceae). American Journal of Botany, 537–544. https:// doi.org/10.1002/j.1537-2197.1993.tb138 37.x Peakall, R., & Smouse, P. E. (2006). GENALEX 6: Genetic analy- sis in Excel. Population genetic software for teaching and re- search. Molecular Ecology Notes, 6(1), 288–295. https://doi. org/10.1111/j.1471-8286.2005.01155.x Pfunder, M., & Roy, B. A. (2000). Pollinator-mediated interactions be- tween a pathogenic fungus, Uromyces pisi (Pucciniaceae), and its host plant, Euphorbia cyparissias (Euphorbiaceae). American Journal of Botany, 87(1), 48–55. https://doi.org/10.2307/2656684 Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of pop- ulation structure using multilocus genotype data. Genetics, 155, 945–959. Rodriguez-Algaba, J., Sørensen, C. K., Labouriau, R., Justesen, A. F., & Hovmøller, M. S. (2017). Genetic diversity within and among aecia of the wheat rust fungus Puccinia striiformis on the alternate host Berberis vulgaris. Fungal Biology, 121, 541–549, https://doi. org/10.1016/j.funbio.2017.03.003 Rodriguez-Algaba, J., Walter, S., Sørensen, C. K., Hovmøller, M. S., & Justesen, A. F. (2014). Sexual structures and recombination of the wheat rust fungus Puccinia striiformis on Berberis vulgaris. Fungal Genetics and Biology, 70, 77–85. https://doi.org/10.1016/j. fgb.2014.07.005 Rosenberg, N. A. (2004). Distruct: A program for the graphical display of population structure. Molecular Ecology Notes, 4, 137–138. https:// doi.org/10.1046/j.1471-8286.2003.00566.x Samils, B., Ihrmark, K., Kaitera, J., Stenlid, J., & Barklund, P. (2011). New genetic markers for identifying Cronartium flaccidum and Peridermium pini and examining genetic variation within and between lesions of Scots pine blister rust in Sweden. Fungal Biology, 115(12), 1303–1311. https://doi.org/10.1016/j.funbio.2011.09.009 Samils, B., Lagercrantz, U., Lascoux, M., & Gullberg, U. (2001). Genetic structure of Melampsora epitea populations in Swedish Salix vimina- lis plantations. European Journal of Plant Pathology, 107(4), 399–409. https://doi.org/10.1023/A:10112 70315251 Samils, B., Stepien, V., Lagercrantz, U., Lascoux, M., & Gullberg, U. (2001). Genetic diversity in relation to sexual and asexual reproduc- tion in populations of Melampsora larici-epitea. European Journal of Plant Pathology, 107(9), 871–881. https://doi.org/10.1023/A:10131 21809990 Sandhu, K., Karaoglu, H., Zhang, P., & Park, R. (2016). Simple sequence repeat markers support the presence of a single genotype of Puccinia psidii in Australia. Plant Pathology, 65(7), 1084–1094. https://doi. org/10.1111/ppa.12501 Sato, H., & Takahashi, L. (1970). Notes on the Japanese rust fungi. VI. Inoculation experiments of Thekopsora areolata (Fr.) Magnus, a cone rust of Picea spp. in Japan. Transactions of the Mycological Society of Japan, 11(3), 109–112. Savile, B. D. O. (1953). Short-season adaptations in the rust fungi. Mycologia, 45(1), 75–87. https://doi.org/10.1080/00275 514.1953. 12024251 Shannon, C. E. (2001). A mathematical theory of communication. ACM SIGMOBILE Mobile Computing and Communications Review, 5, 3–55. https://doi.org/10.1145/584091.584093 Simpson, E. H. (1949). Measurement of diversity. Nature, 163, 688. https://doi.org/10.1038/163688a0 SUPPORTING INFORMATION Additional supporting information may be found online in the Supporting Information section. How to cite this article: Capador H, Samils B, Kaitera J, Olson Å. Genetic evidence for sexual reproduction and multiple infections of Norway spruce cones by the rust fungus Thekopsora areolata. Ecol Evol. 2020;10:7389–7403. https://doi.org/10.1002/ece3.6466