METSÄNTUTKIMUSLAITOKSEN TIEDONANTOJA 829, 2002 FINNISH FOREST RESEARCH INSTITUTE, RESEARCH PAPERS 829, 2002 Proceedings of the IUFRO Working Party 7.02.02 Shoot and Foliage Diseases, Meeting at Hyytiälä, Finland, 17-22 June, 2001 Edited by Antti Uotila Vellamo Ahola VANTAA RESEARCH CENTRE METSÄNTUTKIMUSLAITOKSEN TIEDONANTOJA 829, 2002 FINNISH FOREST RESEARCH INSTITUTE, RESEARCH PAPERS 829, 2002 Proceedings of the IUFRO Working Party 7.02.02 Shoot and Foliage Diseases, Meeting at Hyytiälä, Finland, 17-22 June, 2001 Edited by Antti Uotila Vellamo Ahola VANTAA RESEARCH CENTRE 2 Uotila, A. and Ahola, V. (eds.) 2002. Proceedings of the lUFRO Working Party 7.02.02 Shoot and Foliage Diseases, Meeting at Hyytiälä, Finland, 17-22 June, 2001. Finnish Forest Research Institute, Research Papers 829. 201 p. ISBN 951-40-1809-5, ISSN 951-40-1809-5. Keywords: forest trees, pathogens, endophytes, rDNA-ITS sequences, Gremmeniella, Sphaeropsis, Sirococcus, Botrytis, Raffaelea, Phytophthora, Tubercularia, Flammulina, Septoria, Neofabraea, Phomopsis, Phacidium, Guignardia, Cryptocline Publisher: The Finnish Forest Research Institute, Vantaa Research Centre. Cover photos: Leaf discoloration observed on Picea jezoensis inoculated with Ceratocystispolonica (photo up on the left, Keiko Kuroda). Needle disease on Taxus baccata caused by Cryptocline taxicola: The conidia of the fungus on the ripe acervuli which rupture the epidermis (photos on the right, Alfred Wulf). Spherical fruit bodies develop beneath the epidermis (photo down on the left, A.W.). Orders: 1. Hyytiälä Forestry Field Station, Fin-35500 Korkeakoski Tel. +358 3 3355 111 Fax. +358 3 3355 555 E-mail: antti.uotila@helsinki.fi 2. The Finnish Forest Research Institute, Library, P.0.80x 18, Fin-01301 Vantaa Te1.+358 9 857 051 Fax. +358 9 8570 5582 E-mail: kirjasto@metla.fi Editors' address: Hyytiälä Forestry Field Station, as above. © Authors and Finnish Forest Research Istitute Hakapaino Oy Helsinki 2002 3 Contents Preface 6 General forest pathology Gerard Adams 8 Inoculating trees in nature: a problem of environmental interactions KeikoKuroda 17 The mechanism of tracheid cavitation in trees infected with wilt diseases Giorgio Maresi, Paolo Ambrosi and Paolo Capretti 24 Impact of some crown diseases in Trentino woods Scleroderris canker Gaston Laflamme 30 Taxonomy of the genus Gremmeniella, causal agent of scleroderris canker Timo Kurkela and Antti Uotila 35 Size of ascospores in the A and B types of Gremmeniella abietina Seppo Nevalainen and Ulla Mattila 39 Factors affecting the risk of Gremmeniella abietina infection in Scots pine stands Jesper Witzell 50 Formation and growth of stem cankers caused by Gremmeniella abietina on young Pinus contorta Heikki Nuorteva, Timo Kurkela and Antti Uotila 57 Needle size and needle nutrient contents of Scots pine after Gremmeniella abietina infection and green pruning Tero Tuomi virta 59 The genomes of dsRNA viruses of Gremmeniella abietina The shoot diseases of conifers Giorgio Maresi, Paolo Ambrosi, Andrea Battisti, Paolo Capretti, Roberto Danti, Elisabetta Feci, Stefano Minerbi and Stefania Tegli 60 Pine dieback by Sphaeropsis sapinea in Northern and Central Italy 4 Wang Tian Fu and Antti Uotila 68 Observations of Sirococcus conigenus in Finland Anu Pulkkinen, Juha Heiskanen, Risto Rikala and Raija-Liisa Petäistö 75 Effect of growth stage and microclimate on Botrytis cinerea outbreak in Picea abies container seedlings: Preliminary results The diseases of broadleaved trees Yoshihiro Takahata and Takefumi Ikeda 79 Changes of xylem pressure potential in Quercus serrata saplings inoculated with Raffaelea sp., a possible causal fungus of oak mortality in Japan Arja Lilja and Risto Rikala 83 Phytophthora cactorum on silver birch seedlings Marcus B. Jackson and Robert W. Stack 90 Tubercularia ulmea canker: effect of pathogen, host, and cultural practice Norio Sahashi 98 Frequently isolated endophytes from the Japanese beech -endophytic fungal composition, their temporal variations in colonization rate and distribution in Northern Japan Leena Syrjälä and Marja Poteri 107 Effect of increased carbon dioxide and ozone on leaf spot pathogens of birch Ilona Szabö 113 Occurrence, host range and impact of leaf pathogen fungi on hardwoods in Hungary Maryna Sukhomlyn 120 Flammulina velutipes (curt). Fr. Sing, population structure Yasuaki Sakamoto, Yuichi Takikawa, Yuko Takao and Katsuhiko Sasaki 121 Bacterial canker of Maackia amurensis var. buergeri - Occurrence and pathological anatomy Yasuaki. Sakamoto, Yuichi Takikawa, Yuzou Sano, Atsushi Kato and Katsuhiko Sasaki 131 Watermark disease of willows in Japan - Occurrence and pathological anatomy Gerald E. Weiland, JoAnne C. Stanosz and Glen R. Stanosz 145 Responses of juvenile poplars to inoculation with Septoria musiva can predict their long-term canker disease resistance Risto Kasanen, Timo Kurkela and Jarkko Hantula 146 Neofabraea populi in Scandinavia 5 Nutrient balancies and diseases Hans Anglberger, Erhard Halmschlager, Peter Hietz and Jutta Mattanovich 152 Effect of fertilisation on the resistance of spruce seedlings to shoot blight caused by Sirococcus conigenus Hans Anglberger, Monika Sieghardt, Klaus Katzensteiner and Erhard Halmschlager. 158 Nutritional status of mature Norway spruce related to infection by Sirococcus shoot blight Martti Vuorinen 163 Growth disturbances in spruce forest on burn-beaten areas Needle diseases of conifers Mursel Catal, Gerard C. Adams and Gary A. Chastagner. 164 Detection, identification and quantification of latent needlecast pathogens and endophytes in symptomless conifer foliage by PCR and Dot-blot assays Jose Marmolejo 179 Some interesting fungi found on pine litter of three pine species from Nuevo Leon, Mexico Jennifer Kidder, JoAnne C. Stanosz and Glen R. Stanosz 184 Phomopsis occulta as a landscape pathogen of Douglas-fir Per Hansson 185 Some recent findings concerning Phacidium infestans in Northern Sweden Shun-Ichiro Miyashita and Toshihiro Yamada 189 Phylogenetic analysis of Guignardia cryptomeriae based on rDNA-ITS sequences Jarkko Hantula, Michael Mtiller, Rauni Valjakka and Aki Suokko 192 Diversity of endophytic fungi of single Norway spruce needles and their role as pioneer decomposers Michael Milller, Jarkko Hantula and Anna-Maija Hallaksela 197 Black yeast-like endophytes of Norway spruce Alfred Wulf and Leo Pehl 198 Needle disease on Taxus baccata caused by Cryptocline taxicola 6 Preface Forest pathologists specializing in the study of shoot and foliage diseases have met regularly over the last 30 years to report and discuss their findings on the biology and management of important diseases impacting forest and plantation trees. The first big meeting of this group was in Syracuse New York (USA) in 1983 with 66 scien tists participating. At that meeting the main topic of discussion was the damage that was occurring at that time as a result of Scleroderris canker in North America, Eu rope and Japan. Although there has been extensive research on all aspects of this disease since that first meeting of our group, we learned at this meeting of the current severe epidemic of Scleroderris canker in Sweden occurring in 2001. This underscores not only the need for continued research, but also the importance of environmental factors that can affect the incidence and severity of this disease, limiting our ability to prevent damage in spite of all of our knowledge. Over the years our working party has evolved from a meeting primarily focused on Scleroderris canker to one that now brings together scientists researching a wide range of shoot, stem and foliage diseases from geographically different areas of the world. This interchange of experiences, information and ideas among scientists is important and is the strength and value of our working party to participants. During our meeting on the field trips we became familiar with the local diseases and forestry practices at Juupajoki and Orivesi communities. The region around Tampere is called Pirkanmaa where forestry and forest industries are a very important sector of the economy. Even the roots of the Nokia mobile phone company are in forest industries in the Pirkanmaa area. Nokia is the town close to Tampere. This meeting was held at the Hyytiälä Forestry Field Station of Helsinki University. The role of the field stations is very important in teaching forest sciences. Teaching forestry science has been carried out for the last 90 years at Hyytiälä. Finnish forest ers believe that Hyytiälä is the heart of their education. The field station is not only about teaching and research but provides an atmosphere for food, sports, arts and sauna. It was intended that by having the meeting here, that our working party expe rienced the real Hyytiälä spirit! Antti Uotila The Organiser of the meeting 7 The participants of the meeting at Hyytiälä. 8 Inoculating trees in nature: a problem of environmental interactions Gerard Adams Department of Plant Pathology, Michigan State University, East Lansing, Michigan 48824- 1312, USA. E-mail: gadams@msu.edu Abstract The study of forest pathogens is frequently hindered by problems in statistical analysis of field trials because host susceptibility and pathogen virulence interact strongly with environmental effects that stress trees differently in different years. ANOVA tests are confounded by significant genotype x environment interactions that preclude mean separation and ranking tests. The Addi tive Main Effects and Multiplicative Interaction Model is introduced as a method of obtaining valuable information from the complexity introduced by statistical interactions. Many of the common pathogens of trees are particularly recalcitrant organisms for studying the host pathogen system in nature. This is not due to difficulties in isola tion or cultivation of the microorganisms, rather the difficulties are in reproducing the disease symptoms on the host. A basic paradigm in forest pathology is that many pathogens are virulent only on a host suffering stress caused by one factor or multi ple factors over time. Environmental factors have profound effects on the incidence and severity of disease symptoms on trees. The environmental factors that predispose trees to disease are complex and vary with site. Perhaps the best known of the factors causing stress are drought and insect pests that vary seasonally and spatially. Further compounding the complexity of environment on host susceptibility is the variable response of differing host geno types and host species to a stress. Completing Koch's Postulates with a fungal patho gen on a tree species is not straightforward. A review of older literature shows the surprising fact that many initial inoculation trials yielded only 50-70% successful infection with disease expression. Would today's reviewers accept the claim of com pletion of the Postulates with such low success levels? Environmental interactions can have so profound an effect on virulence of a pathogen that the pathogen becomes labeled as weak or opportunistic. Labeling for est pathogens as weak opportunists has damaged the study of forest pathology in recent decades as the label has been used as a reason to reject grant proposals seek ing funds to study the organisms. A compounding factor in the study of forest patho gens is the more recent revelations that many fungal pathogens are endophytic resi dents in waiting. Well-studied examples of forest pathogens that are endophytic and opportunistic yet cause severe disease and great economic damage when interacting 9 with environmental stresses, include: Botryosphaeria, Sphaeropsis, Armi Ilaria, Phomopsis, and Cytospora. Inoculating trees with such pathogens may yield isolate ratings of avirulence, low virulence, and highly variable virulence, but a reaction of high virulence is un likely. For example, to successfully reproduce a canker symptom on a tree with a Cytospora pathogen we find it necessary to inoculate in the fall in the Northern Hemisphere which allows the pathogen to colonize the host when dormancy has disabled the host's defense response. Additionally, we mechanically wound and freeze damage the cambium to aid infection and colonization. In regions without a winter period of plant dormancy we have been unable to induce disease without adding a different stressing agent such as drought stress. Knowledge that the virulence of a pathogen is altered by interactions of envi ronment on the host has become well established in forest pathology. Additionally, it is well established that host genotypes (i.e., provenances) respond variably to envi ronmental factors. The virulence of a pathogen that can infect several host genotypes and host species will interact in complex ways to quantitative and qualitative differ ences in environmental stresses. Dealing with environmental and genotype interactions in inoculation trials in nature remains a major problem in forestry research. The standard method for re moving environmental interactions from inoculation trials has been to isolate the plants in an environmentally controlled glasshouse or growth chamber. Unfortu nately, inoculations of seedlings do not always reproduce the type of disease symp toms characteristic of trees in nature. More importantly, inoculations removed from nature do not supply information on the behavior of a pathogen in nature. In particu lar, little can be learned about how the virulence of a pathogen will differ under seasonal changes. The genotype of the pathogen is another complicating factor of significant concern to forest pathologists. Differences among isolates of a pathogen in relative virulence on a host can be the major concern in studies of hypovirulence or of breeding for resistance. To predict the success of hypovirulence in establishing biological control of a forest pathogen, the investigator needs to identify the relative virulence among isolates over several seasons, and over several host genotypes. If the pathogen can infect more than one host species, then the relative virulence among isolates should be examined over different host species, and over several seasons. It is conventional practice in Agronomy to evaluate genotypes a minimum of three years in replicated field trials. In Agronomy, statistical ly significant genotype x environment interactions complicate field trials, frequently. Field trials examining hypovirulence are likely to have significant interactions among isolate genotypes and season (environment) and host genotypes (tree replications or species) and sea son. The problem for forest pathologists becomes the statistical analysis of data collected from such field trials. The analysis becomes unusually complicated be cause of the great variability in disease severity or isolate virulence due to the strong genotype x environment interactions; isolate genotype x seasonal effects (environ mental stresses) and host genotype (individual trees or different species) x seasonal effects. In fact, because forest pathologists often are working with a "weak" patho 10 gen, analysis is further complicated. Interaction of seasonal effects with host geno type often has a more extreme influence on isolates of lower virulence because rela tive virulence can swing from avirulence in one season to high virulence in another. Such variation in statistical analysis causes significant nonadditivity in an ANOVA test when comparing isolates of low virulence that exhibit greater variability with isolates of high virulence that are more stable (consistently high in virulence). An example of such a comparison includes the study of hypovirulent and virulent iso lates of Cryphonectria parasitica. In an ANOVA test, the investigator seeks highly significant differences be tween treatments, generally. Significant differences permit mean separation tests that result in the ability to rank the treatments as differing in response. In a field trial, the ANOVA test will reveal whether the isolates are different in virulence, whether the hosts are different in susceptibility, and whether the seasons are different in stress ing the hosts. But when the ANOVA test also reveals that statistically significant interactions occurred between isolate genotype x environment or host genotype x environment, it is not legitimate in mathematica to separate means or rank means. Thus, the experiment fails to answer the questions of the investigator. It is the usual case in forest pathology that field trials and even glasshouse trials have statistically significant interactions in the ANOVA test. This situation has hindered forest pathol ogy for decades. What is most needed for the study of the pathogen/host/environment interac tion of forest pathology is a method of separating and ranking treatment means de spite having statistically significant interactions in the ANOVA test. For example, we should be able to calculate the mean virulence of an isolate (an adjusted mean) averaged over more than one year, despite the interactions. Additionally, a measure of interactivity is needed for ranking hosts and pathogens in trials extending over years. For example, it would be worthwhile to know that two hosts may be similar in relative susceptibility, but one might radically change in susceptibility in response to stresses while the other may express a consistent level of susceptibility under the changed level of stress. Knowing the level of interactivity response of a host or isolate in each year is also an aide in identifying the season in which the stress occurred, or in identifying the nature of the stress. With such methods available, the mean separation test would answer the investigators questions on the treatment ef fects. Additionally, the interactivity component would supply a value in relation to the manner in which each treatment responses with environment. Today, these objectives are obtainable with a relatively new method of statis tical analysis, yet no forest pathologist has published an application of the method to our knowledge. The method uses ANOVA combined with principal component analy sis (PCA). It is called the Additive Main Effects and Multiplicative Interaction Model (AMMI) (Gauch 1992). The result of employing the AMMI model includes the cal culation of adjusted means that are more accurate than the observed means due to partitioning of the effect of interactivity and apportioning the interaction to specific means. In addition, mean separation and ranking tests can be performed in a math ematically legitimate way on the adjusted means. Let us explain the use of the AMMI modeling with an example of a typical 11 forest pathogen, Sphaeropsis sapinea and a field inoculation trial on various Pinus species over two years. The objective of the trial is to compare the virulence of many isolates. The data are from an experiment of ours in which we sought to uncover hypovirulent strains. Seasonal environmental factors, such as water stress (Blodgett et al. 1997 a, b; Swart and Wingfield 1991) and pine spittlebug (Haddow and Newman 1942), have been documented as modifying disease development by S. sapinea on Pinus species. Therefore, most pathogenicity studies of S. sapinea on pine have been confounded by significant interactions between year and isolate (Swart and Wingfield 1991), between trial and isolate (Blodgett and Stanosz 1997), and between Pinus species and isolate (Blodgett and Stanosz 1997). The significant interactions have been com mon in field inoculation trials as well as in greenhouse trials under controlled cli matic conditions (Blodgett and Stanosz 1997). These interactions demand closer study. The relative virulence among 30 isolates was evaluated in field trials by inoculation of three species of Pinus over two years, year one (-yrl) and year two (- yr2): Pinus sylvestris-yrl, P. sylvestris-yrl, P. nigra-yr\, P. nigra-yrl and P. resinosa yr2. The experimental design for evaluating virulence on several Pinus species was a randomized complete block design with two factors and five replications (five trees of each Pinus species). The main plots consisted of the Pinus species-years and the subplots consisted of the fungal isolates being tested. The resulting data was analyzed statistically as a two-way factorial ANOVA in the normal fashion. The ANOVA additive model for the two-way factorial was the usual equation [l]: U. sr = u+a. + B +O. +B. where U. is the virulence (isolate by Pinus species-year by " l ~s is isr isr v J . replication), ji is the overall mean, a. is the fungus isolate deviation, P s is the Pinus species-year deviation, 0. s is the interaction (non-additive residual), and 8. sr is the error term. The ANOVA table (additive model) of the randomized complete block design is presented below. Table 1. The ANOVA reveals no significant differences among the replicate trees, which is good. Pinus sylvestris-yrl, P. sylvestris-yrl, P. nigra-yrl, P. nigra-yr2 and P. resinosa-yx2 differ highly significantly in their susceptibility to S. sapinea which is Source Degrees of Sums of Mean Freedom Squares Squares F Value Probability Total 749 16677.8 Replication 4 (r-1) 18.2 4.5 0.6 A. Pinus species-year 4 (a-1) 4814.2 1203.5 170.4 0.01 B. Isolates 29 (b-1) 4159.7 143.4 20.3 0.01 A x B Interaction 116 (a- 1 )(b- 1) 3475.5 30.0 4.2 0.01 Error 596 (ab-1 )(r-1) 4210.3 7.1 12 expected. Isolates differed highly significantly in their relative virulence which is desired. Unfortunately, highly significant interactions occurred between isolate and Pinus species-years. These interactions prevent statistical separation and ranking of the means of relative virulence of isolates over the combined Pinus species or years. Therefore, we cannot make valid conclusions on the mean virulence of any isolate. Additionally, the interactions prevent separation of the means of the susceptibility of Pinus species over all isolates in a year, or over combined years. Therefore, we could not make valid conclusions on the mean susceptibility of any tree species. With this standard ANOVA, the considerable effort and expense of conducting the field trials was wasted and usually the result would have been that the data was discarded. Fortunately, all the value of the field trials, and more, can be uncovered with our use of AMMI analysis to study (model) the effects on virulence of the interaction between isolate (genotype) and Pinus species-year (environment). AMMI is a sig nificant statistical resource for understanding genotype x environment interactions (Gauch 1992). The pattern in the interactivity of the isolates is studied in detail by means of principal component analysis of the interaction sums of squares in AMMI analysis. The analysis provides a score (interactivity score) for quantifying the sen sitivity of the response of an isolate or of a tree species to the year, in the host pathogen-year interaction. AMMI analysis is useful in several unique ways. It helps to model interactions, to estimate virulence more accurately, and to permit better selections of hypovirulent isolates. The technique and its application are reviewed by Gauch (1992) and Gauch and Zobel (1996). AMMI analysis combines ANOVA and principal component analysis (PCA) into a single analysis with both additive and multiplicative parameters. The additive part of the AMMI model uses ordinary ANOVA that leaves the non-additive re sidual, the interaction. The multiplicative part of the AMMI model uses PCA to decompose the interaction into PCA linear axes. Decomposing the interaction al lows a major portion of the interaction to be accommodated in a calculation of a predicted response, for example, the predicted virulence of an isolate. AMMI analy sis revealed that the first principal component (PCA 1) accounts for the predictable pattern in the interaction between isolate and Pinus species-year (genotype x envi ronment) in our data set, whereas, PCA 2 or PCA 3 might be more important in other data sets. The following statistical model adapted from Gauch and Zobel (1996) and Eisenberg et al. (1996) is used to obtain a new and more accurate estimate of the mean virulence of each isolate, and similarly a new mean susceptibility of each Pinus species of the combination Pinus species-year over all isolates. The equation [2] is U. =u+a. + (3 + £ A, v. 5 where U. is the predicted virulence of an isolate 1 1 is " l ' s n n' in sn is r over Pinus species-year, |_i is the overall mean, a. is the fungus isolate effect, P s is the Pinus species-year effect, A, n is the singular value for principal component n, y. is the fungus isolate vector of scores, 5 s is the Pinus species-year vector of scores, and n = 1 for the axis of PCA 1. The AMMI ANOVA table (Additive and Multiplicative Model) of the randomized complete block design is presented below. The F test and Probability are not a component of AMMI ANOVA. 13 Table 2. The treatment mean square is found to contain 8.43% of the noise related to the experimental design. The AMMI analysis recovers the pattern in its AMMII model and relegates the noise to a discarded residual in order to increase the predic tive accuracy and thus supply predicted means of virulence for each isolate. The relative virulence of each isolate is predicted on each Pinus species each year sepa rately. The predicted virulence represents a calculation of the virulence of the iso late on the Pinus species after removing the predictable error due to the influence of the year's effect. Discarding a residual SS of 8.43% of the treatment sums of squares causes the AMMI adjusted (predicted) means to differ from the unadjusted means by a root mean square of 23.93% of the unweighted grand mean. Results of an AMMI analysis are presented on a type of graph called a biplot, similar but not identical to graphs used in other PCA plots. On the biplot one axis is the adjusted mean virulence and the other axis is the sensitivity of response to the interactivity. The y-axis represents the distribution of interactivity scores of the iso lates and Pinus species-year, the y and the 8 in equation 2. The virulence of isolates near the zero line (Fig. 1) is characterized as being very much more stable or consist ent over Pinus species-years (i.e., less interactive) than those further away. One as pect of this type of graph that is not intuitively clear is the negative values on the interactivity axis (y-axis). The negative value does not have a biological meaning, only a mathematical meaning, therefore, negative values do not have less interactivity. The biplot is informative because it shows both main (additive) and interaction (multiplicative) effects for both genotype and environment. Displacement along the abscissa indicates differences in the additive effects; displacement along the ordinate indicates differences in multiplicative effects. The estimate of the predicted virulence and interactivity of an isolate appears on the biplot (■); and also the susceptibility of a Pinus species in a year to the mean virulence of all the isolates combined appears on the biplot (A). The AMMI calculations and the drawing of the biplots are carried out with the help of the MATMODEL program (Gauch 1993). For each isolate (■) the x-axis shows a predicted virulence expressed over the combined five Pinus species-years. For example, the mean virulence for all the isolates of morphotype A, except A+l47, is higher than the mean virulence of all of the Source Degrees of Sums of Mean Freedom Squares Squares Total 749 16677.817 22.267 Treatments 149 12449.317 83.552 A. Pinus species-year 4 4814.153 1203.538 B. Isolates 29 4159.657 143.436 A X B interaction 116 3475.507 29.916 iPCA 1 32 1862.989 58.218 Residual 84 1612.518 19.197 Error 600 4228.500 7.048 14 individual isolates of morphotype B, except 8+466. The isolates of lowest mean virulence for each morphotype can be selected readily for further study of potential hypovirulence. This selection will be more accurate in prediction of an isolate's virulence over the different Pinus species and over different seasons than would selection based on inoculation trials in glasshouses under controlled climatic conditions. Thus, AMMI analysis provides for a major improvement in the quality of research. Additionally, AMMI analysis has supplied a standard error of the means (SEM) among isolate means. The SEM is 0.53. This value can now be used legitimately for mean separation tests. Our trial and effort are saved! We can now make statistically sound conclusions on which isolate is significantly different in mean virulence than another. The predicted virulence of an isolate expressed over three Pinus species and two years (combined as five Pinus species-years) are in Table 1 where they are ranked by Duncan's multiple range test, P = 0.05, Least significant difference (LSD) value of 1.475, standard error 0.53. For the Pinus species-year (A) the x-axis shows a measure of the susceptibility of a Pinus species to the mean virulence of all the isolates combined. For example, the mean virulence of all the isolates combined is greatest for P. sylvestris in the first year and in the second year, therefore that species is the most susceptible. Conversely, P. nigra in the second year was the least susceptible while P. nigra and P. resinosa were equal in relative susceptibility in the first year. AMMI analysis has also provided valuable information on the sensitivity of the different Pinus species differs in response to changes in seasonal effects. For example, P. nigra exhibits a relative stability in its interactivity with different isolates and year. Pinus sylvestris is highly interactive and will react vigorously with different isolates depending on the year (Fig. 1). Figure 1. 15 Additionally, AMMI analysis has supplied a SEM among means of the Pinus species-years. The SEM is 0.22. This value can now be used legitimately for mean separation tests. We can now make statistically sound conclusions on which Pinus species-year is significantly different than another in mean susceptibility to the mean virulence of all the isolates combined. The Pinus species-years can be ranked in the following order: P. sylvestris-yrl, 9.48 (A); P. sylvestris-yr2, 5.70 (B); P. nigra-yrl, 4.09 (C); P. resinosa-yrl, 3.21 (D); and P. nigra-yr2,2.25 (E). The letters in brackets following the means are the DMRT rankings at P = 0.05, LSD value of 0.602, stand ard error 0.22. In conclusion, in our pathogenicity tests on three species of Pinus highly sig nificant statistical interactions occurred between isolate virulence, Pinus species, and year. Pine species-year had a profound impact on virulence. The pattern in the interactions was revealed by PCA of the interaction SS of the ANOVA. The interactivity analysis was used to apportion interaction to specific isolates to im prove the accuracy of the estimates of virulence. Estimates of the relative virulence of isolates were predicted over five different Pinus species-years. Isolates were ranked in virulence and interactivity using the AMMI model. This model permitted mean separation tests of the relative virulence among isolates over the combined Pinus species-years. The virulence and interactivity of isolates was summarized in a biplot graph of adjusted means to facilitate selection of isolates that had potential for hypovirulence. The AMMI method of using PCA within an ANOVA test is a new and reveal ing mathematics for the study of the interaction of pathogenic fungi with their host plants in changing annual environments. AMMI analysis permits the study of the natural system with S. sapinea, rather than moving the host-pathogen system out of the natural environment to reduce the interactions. AMMI analysis provides an interactivity score for quantifying the sensitivity of the response of an isolate or a tree species to the year, in the host-pathogen-year interaction. AMMI analysis also allows for the statistical separation and ranking of means of relative virulence of each isolate over the combined Pinus species-years, as well as, the separation of means of the susceptibility of a Pinus species to the mean virulence of all the isolates combined. General experience with agricultural data has found that predicted yield esti mates from an AMMI model are as accurate as treatment means based on two to four times as many replications (Gauch and Zobel 1996). Routinely, AMMI models are more accurate in predicting yield than are the unadjusted data from yield trials be cause of genotype X environment interaction (Gauch and Zobel 1996). AMMI analysis and the resulting biplot is most interesting when the interactivity can be associated with a causal factor such as drought or insect infestation. We hope that this research report will encourage further use of AMMI analysis in pathological studies. References Blodgett, J. T., Kruger, E. L. and Stanosz, G. R. 1997 a. Effects of moderate water stress on disease development by Sphaeropsis sapinea on red pine. Phytopathology 87: 422-428. 16 Blodgett, J. T., Kruger, E. L. and Stanosz, G. R. 1997b. Sphaeropsis sapinea and water stress in a red pine plantation in central Wisconsin. Phytopathology 87: 429-434. Blodgett, J. T. and Stanosz, G. R. 1997. Sphaeropsis sapinea morphotypes differ in aggressiveness but both infect nonwounded red and jack pines. PI. Dis. 81: 143-147. Eisenberg, B. E., Gauch, H. G., Zobel, R. and Kilian, W. 1996. Spatial analysis of field experiments: fertilizer experiments with wheat (Triticum aestivum) and tea (Camellia sinensis). In: Genotype by environment interactions. Ed. by Kang, M. S., Gauch, H. G. Boca Raton, FL: CRC Press, pp. 373-404. Gauch, H. G. 1993. MATMODEL version 2.0. AMMI and related analysis for two-way data matrices. Ithaca, NY: Microcomputer Power. Gauch, H. G. 1992. Statistical analysis of regional yield trials: AMMI analysis of factorial design. New York, NY: Elsevier. 278 pp. Gauch, H. G. and Zobel, R. W. 1996. AMMI analysis of yield trials. In: Genotype by environment interactions. Ed. by Kang, M. S., Gauch, H. G. Boca Raton, FL: CRC Press, pp. 85-122. Haddow, W. R. and Newman, F. S. 1942. A disease of the Scots pine (Pinus sylvestris L.) caused by the fungus Diplodia pinea Kickx. associated with the pine spittle-bug (Aphrophora parallela Say.), Part I. Trans. Roy. Canad. Inst. 24: 1-17. Swart, W. J. and Wingfield, M. J. 1991. Seasonal response of Pinus radiata in South Africa to artificial inoculation with Sphaeropsis sapinea. PI. Dis. 75: 1031-1033. 17 The mechanism of tracheid cavitation in trees infected with wilt diseases Keiko Kuroda Kansai Research Center, Forestry and Forest Products Research Institute Momoyama, Fushimi, Kyoto 612-0855, Japan. E-mail: keiko@afifrc.go.jp, http://www.fsm.afifrc.go.jp/Pathology/keiko-eng.html Abstract In trees infected with pine wilt disease, extensive embolism occurs in tracheids. The enlargement of a cavitated and dysfunctional xylem induces a water deficit and mortality in trees. This phenomenon was observed in a wilt disease in oak trees caused by Raffaelea sp. To confirm an assumption that this wilt mechanism is fundamentally similar in several wilt diseases, inoculation experiments were conducted on larch and spruce trees. In Larix kaempferi trees inoculated with Ceratocystis lalicicola, acoustic emission from the trunks increased before the initiation of apex wilting. This indicates that abnormal dehydration from tracheids was occurring in the infected trees. In the trees that rapidly developed symptoms, xylem dysfunction had progressed widely. The leaves of Picea jezoensis were partially discolored about one month after inoculation with C. polonica. The fungal hyphae were elongating radially through the ray tissue. Before the symptoms were visible, areas that were cavitated and dysfunctional appeared in the xylem. When the outer annual rings became dysfunctional, the sap ascent decreased significantly, and the leaves began to discolor. The dysfunctional areas were much wider than the areas plugged with resin. The fungal activity promoted secondary metabolism in ray parenchyma cells and is responsible for the cavitation. A similar mechanism that promotes embolism and inhibits the refilling of water in pine wilt also occurs in the wilt diseases of larch and spruce trees. Keywords: Pine wilt, embolism, sap ascent, Raffaelea, Picea, Larix, Ceratocystis Introduction Pine wilt is caused by the nematode Bursaphelenchus xylophilus and is one of the most serious tree diseases in Japan. From the end of 1980s, many oak forests on Honshu Island have been damaged by Raffaelea sp. Ceratocystis spp. are killing spruce and larch trees in northern Japan. All the pathogens of these diseases are vectored by beetles. The leaves of the trees become discolored about three weeks after inoculation with the pine wood nematode (Kuroda et al. 1988). Usually, healthy trees do not wilt very easily, even when the supply of water is stopped for a certain period. Why these pathogens can kill trees so effectively is a question that remains to be answered. Furthermore, the cause of the swift death of the apical and cambial cells of a huge tree does not yet have a theoretical explanation. 18 In healthy trees, xylem sap ascends spirally in pine trunks. In trees infected with pine wilt, dehydrated areas emerge in xylem prior to leaf discoloration (Kuroda et al. 1988). Such dysfunctional areas enlarge, and the tree wilts within one or two months. It is clear that some kind of a system rapidly excludes the xylem sap from the tracheids (Kuroda 1989, 1991). In the case of oak trees infected with Raffaelea sp., the discoloration of sapwood and blockage of water are well underway before the appearance of other symptoms of wilt in the summer months (Kuroda 2001, Kuroda and Yamada 1996, Takahata and Ikeda 2001). The mechanism of the symptom development of wilt described here (Fig. 1) is based on research into the pine wilt and oak mortality caused by Raffaelea sp. When microorganisms attack a tree, a protection strategy is set in motion. A synthesis of secondary metabolites starts in the tree tissue. With this reaction, cavitation or dis coloration occurs in the xylem. Tracheids and vessels become dysfunctional and are filled with air. Unfortunately, the secondary metabolites are ineffective in their ef forts to protect the tree against the pine wood nematode and Raffaelea sp. As the pathogen is widely distributed, protection occurs in many places in a tree, and, there fore, xylem dysfunction is widely spread. The sap ascent is extensively reduced, and the tree dies because of a water deficit. I considered that these findings might be applicable to other wilt diseases. To check this possibility, wilt processes in larch and spruce trees were monitored after inoculation with the pathogens Ceratocystis lalicicola and C. polonica, respectively. Investigations were made with techniques involving Acoustic Emission (AE) to detect embolism and cavitation in tree trunks (Kuroda 1995, Kuroda and Kuroda 2000), a dye injected into tree trunks to trace the course of sap ascent, and anatomical analyses to check cytological reactions of host cells and hyphal distribution. Figure 1. Schematic illustration of wilting mechanisms hypothesized for the wilting disease of pine and oak trees. 19 Materials and methods Japanese larch On July 21, 1998, the pathogen of wilt disease Ceratocystis lalicicola Redfern et Minter* was inoculated on the lower trunks of seven-year-old Larix kaempferi Sarg. trees. The features of the inoculated and control trees are cited in Table 1. At the start of the experiment, AE transducers (140 kHz) were attached to the lower trunks of the inoculated and control trees. This technique is useful for monitoring the embolism in living trees without cutting (Kuroda and Kuroda 2000). In the experiments on pine wilt, the extremely high rate of AEs was confirmed and reflected abnormal dehydra tion and dysfunction of the tracheids (Kuroda 1995). AE events were monitored until the harvest. The trees were checked for symptoms every week until the end of November. Yezo spruce On July 21, 1999, the pathogen Ceratocystis polonica (Siem.) C. Moreau** was inoculated into the trunks of seven-year-old Picea jezoensis (Sieb. et Zucc.) Carr. trees. Inoculated trees were harvested at one-week intervals. Table 2 shows the har vest dates and symptom development. AE transducers were attached onto the lower trunks prior to the inoculation. To visualize the water conduction, a dye solution was injected at the base of the trunks (1% aqueous acid-fuchsin) at harvest (Kuroda et al. 1988). The main stem was then dissected and processed for an anatomical analysis. Microscopic observations were made to check the reaction of the tree tissues to the fungal activities. * Pathogen: Isolated by T. Yamaguchi from the blue-stained xylem of aL. kaempferi in the experimental forest at the Hokkaido Research Center, Forestry and Forest Products Research Institute. ** Pathogen: YCC-115. Isolated by Y. Yamaoka from the gallery wall of P. jezoensis in the Tokyo Univesity Forest in Hokkaido. Results and discussion Wilt of Japanese larch Two weeks after infection, the trees' apices began to droop (Trees No. 4 and 5 in Table 1, Fig. 2). Leaf yellowing and defoliation followed at three weeks after inocu lation. These are characteristic symptoms of the disease. Figure 3 shows the fre quencies of AEs detected around the time that the apices began to droop, which was the initial symptom. The typical pattern of AEs in healthy trees is indicated by the control data. In healthy trees, the AE pattern increases rapidly when the transpiration rates are high as a result of normal cavitation in the tracheids. There was no AE event at night. In inoculated trees, the frequency of AE was normal until August 24, and then it increased from August 25 just before the apices began to droop. 20 Table 1. Symptoms in 7-year-old Larix kaempferi trees inoculated with Ceratocystis lalicicola. H: No symptom, D: Apex drooping, Y: Yellowing, F: Defoliation, M: Mortality, R: Recovery Table 2. Symptoms on 7-year-old Picea jezoensis trees inoculated with Ceratocystis polonica. H: No symptom, D: Discoloration, ED: Extensive discoloration, M: Mortality (part), R: Recovery, Felling In pine wilt disease, it has been confirmed that cavitation occurs extensively in periods of high AE rates. The cavitated tracheids were not refilled with water. The increasing pattern of AE frequencies in larch trees resembles pine wilt, although the rates were lower. Therefore, the abnormal embolism and cavitation were assumed to have occurred in the periods of high AE rates in the trunks of infected larch trees. The shoots drooped because of a lack of water caused by the extensive embolism in the trunks. Wilt of Yezo spruce The shoots of the spruce trees did not droop. Instead, their needles became discolored as an initial symptom (Fig. 4) about three weeks after the inoculation. In spruce trees, AE events did not increase as drastically, probably because the cavitation pro Date Week Tree No No.l No.4 No.5 No.7 No.9 No. 10 No.6 Jul-14 0 Inoculated Control Jul-21 1 H H H H H H H Jul-27 2 H D&F D D D D H Aug-01 3 D Y Y Y Y&F Y H Aug- 18 5 F F F M M F H Nov-27 4 m R M M M M R H Date Week Tree No No.2 No.9 No.! 3 No.l No.5 No.3 No.4 No.10 No.6 Jul-21 0 Inoculated Controls Jul-28 1 H H H H H H H H H Aug-04 2 H H H H H H H Aug- 11 3 - H D H D D H Aug- 18 4 - - D D D H Sep-0 1 6 - D D H Sep- 15 8 ED D H May-02 (Next year) M R 21 gressed gradually. As a result of dye absorption from the trunk bases, functioning tracheids were stained red. In the inoculated trees, white parts without staining were wider than the controls even before symptom development. Tracheids in those areas were filled with air and had become dysfunctional. Blockage of sap ascent was ex tensive by the start of leaf discoloration (Fig. SA). This severe dysfunction caused water shortage and induced leaf discoloration. Ray tissue in sapwood consists of living cells. Before symptom development, hyphae distributed through ray tissues, some resin canals, and a few tracheids. Ver tical distribution was very narrow and limited. Radial elongation was deep and reached the depth of several annual rings. Ray parenchyma cells invaded by hyphae were necrotic. Ray parenchyma cells produced secondary metabolites as a reaction before Figure 2. Drooping symptom observed on Larix kaempferi inoculated with Ceratocystis lalicicola. Figure 3. High-frequency AEs detected from a trunk of L. kaempferi on July 26 and 27. 22 Figure 4. Leaf discoloration observed on Picea jezoensis inoculated with Ceratocystis polonica. Figure 5. Blockage of xylem sap ascent in P. jezoensis indicated by dye injection at the start of leaf discoloration (A) and normal water conduction in the control tree (B). necrosis, and the cell contents looked yellow. Cambial necrosis was just restricted to inoculated sites and was not serious. All these indicators showed that this fungus is not effective to kill cambium and that the cambial lesion in narrow area does not cause wilt. On the other hand, dysfunction and desiccation occurred in the xylem rapidly and widely. The activity of the fungus was effective to stop water conduc tion. It was significant that the dysfunctional area was much wider than the area of resin occlusion. Resin occlusion is not an essential factor for dysfunction, although it may also be effective to block sap ascent. 23 Conclusion Distribution of two Ceratocystis species in larch and spruce trees was both slower and narrower than in the case of pine wood nematode; however, these fungi were effective to stop sap ascent. Complete stop of sap ascent is due to extensive cavita tion (or embolism) of water conduits that cannot refill with water. This is the cause of wilt in these two wilting diseases. Ceratocystis spp. promoted secondary metabo lism in parenchyma cells. As assumed in the case of pine wilt, substances with low surface tension and a hydrophobic character may promote embolism and inhibit refilling with water in the wilting diseases of larch and spruce caused by Ceratocystis spp. Acknowledgements I wish to thank Dr. T. Yamaguchi, Hokkaido Research Center, Forestry and Forest Products Research Institute, and Dr. Y. Yamaoka, Tsukuba University, for providing isolates of two Ceratocystis species. References Kuroda, K. 1989. Terpenoids causing tracheid-cavitation in Pinus thunbergii infected by the pine wood nematode (Bursaphelenchus xylophilus). Jpn. J. Phytopathol. 55: 170-178. Kuroda, K. 1991. Mechanism of cavitation development in the pine wilt disease. Eur. J. For. Path. 21: 82-89. Kuroda, K. 1995. Acoustic emission technique for the detection of abnormal cavitation in pine trees infected with pine wilt disease. International symposium on pine wilt disease caused by pine wood nematode (Beijing, China). Proceedings 53-58. Kuroda, K. 2001. Responses of Quercus sapwood to infection with the pathogenic fungus of a new wilt disease vectored by the barkbeetle Platypus quercivorus. J. Wood Science 47 (in press). Kuroda, K. and Kuroda, H. 2000. Detection of embolism and acoustic emissions in tracheids under a microscope: Incidence in diseased trees infected with pine wilt, in "New Horizons in Wood Anatomy," ed. by Y.S. Kim, Chonnam National Univ. Press Kwangju, Korea, 372- 377. Kuroda, K, Yamada, T., Mineo, K. and Tamura, H. 1988. Effects of cavitation on the develop ment of pine wilt disease caused by Bursaphelenchus xylophilus. Jpn. J. Phytopathol. 54: 606-615 Kuroda, K. and Yamada, T. 1996. Discoloration of sapwood and blockage of xylem sap ascent in the trunks of wilting Quercus spp. following attack by Platypus quercivorus. J. Jpn. For. Soc. 78: 84-88 (in Japanese with English summary). Takahata, Y. and Ikeda, T. 2001. Changes of xylem pressure potential in Quercus serrata saplings inoculated with Raffaelea sp., a possible causal fungus of oak mortality in Japan. Proceed ings of lUFRO working party 7.02.02 Shoot and Foliage Diseases Meeting (Hyytiälä, Fin land). 24 Impact of some crown diseases in Trentino woods Giorgio Maresi, Paolo Ambrosi and Paolo Capretti G. Maresi, P. Ambrosi, U.O. Foreste, Istituto Agrario (lASMA), S. Michele a/Adige (Trento), Italy. E-mail: maresi@ismaa.it P. Capretti, DIBA - Dipartimento di Biotecnologie Agrarie, Universitä - Firenze, Italy. Abstract The Forest Trees Damages Monitoring (FTDM) has been carrying out in the Trentino woods since 1990, in collaboration with the Forest Services of the Autonomous Province of Trento, as a tool to support the naturalistic management of forests. In this monitoring program, data were obtained on several crown pathogens. Among them, few seem effective to produce economic damage as like as Lachnellula willkommii. Snow fungi proved to be dangerous only in some well defined topo graphic situation. In the case of Lirula nervisequia and some rust fungi, environmental conditions of some sites proved optimal for attacks but no negative effects were observed on the evolution of infected woods. Others pathogens as Sirococcus strobilinus, Lirula macrospora and Apiognomonia errabunda can be occasionally harmful to individual trees, however they are not able to produce sensible damages to the forest. Occasionally outbreaks of crown parasites were reported (attacks by Mycosphaerella laricina and Meria laricis were recorded on larch and Rhizosphaera kalkhoffii on Picea abies). The damage were strictly related to particular site conditions and were delimited in space and time. More effective, and related to climatic conditions, appeared the impact of Sphaeropsis sapinea and Cenangium ferruginosum on Pinus nigra artificial stands. The possibil ity to use these fungi as bio-indicators of host-stress conditions or of wrong forest management choices is discussed. Introduction Trentino is one of the most forested provinces of Italy having 345.000 hectares of woodland which cover more than half of the Trento's Province land surface. This mountain area is located in the Northern-east part of the Alps and it is characterised by woodland environment extremely varied because of different climatic conditions, rock formations and soil features. All the alpine tree species, with a clear predomi nance of Norway spruce {Picea abies), are present together with a minor but inter esting presence of some Mediterranean ones. Total wood mass in the high forests is estimated around 50x10 6 m 3 and species that determine this biomass are P. Abies (62%), Larix decidua (16%), Abies alba (12%), Pinus sylvestris and Pinus nigra (6%), Fagus sylvatica (3%) Pinus cembra (1%). Woods have a strong economical importance not only for their primary production but also because they are the main landscape features, playing a fundamental role in this touristic area. Since the 50's, silviculture in Trentino aims to enhanced the naturalistic wood land management. The goal is to conserve or restore the natural characteristics of the 25 woodland which are specific to each environment. The naturalistic management points to maintain a continuos woodland cover and to obtain everywhere it is possible a natural regeneration. The management is pursued by means of the detailed "Wood land Management Plans" which are drawn up for every woodland public property and generally revised every ten years. On this occasion, a census is made of the trees, their development in time is evaluated and decisions are made on appropriate treat ments. In this contest the presence of parasites and pests can play a fundamental role influencing or modifying the evolution of wood and nullifying the choices of the forest manager. Knowing and understanding the pathogens role is the target of the Forest Trees Damages Monitoring (FTDM) which has been carrying out since 1990 by the Forest Research Unit of IASMA in collaboration with the Forest Services of the Autonomous Province of Trento (Ambrosi and Salvadori 1998). Material and methods Forest Trees Damages Monitoring (FTDM) Method is based on five distinct phases: observations in the forest, report of the problems, diagnosis, transmission and elaboration of data. The Forest Service personnel who are specially trained to make the job carry out observations in the forest within a well-organised network (10 forest districts with 45 stations). The ratio of the forest area and personnel is 2350 ha per individual. The personnel note down observations on five different forms during pre-de termined periods of the year. Each of them helps in grouping the various pathologies and other types of damage that manifest during a known period of the year. The form contains a series of information regarding topography of the affected area (forest plan compartment, exposition, altitude, etc.), type of damage and intensity (number of trees affected, defoliation, total affected growing stock, etc.). When the forest damages are not recognised in field, the Forest Unit of the Institute is informed, field survey and sample examination performed and diagnosis defined. All the diagnosis are completed with indications on management treatment to check the problem and are referred to the Woodland Management Plan compart ments. Data is maintained in a database, related to a Geographic Information System (GIS) based on the woodland management plans compartments; elaboration and conclusions are made available to the organisations in the territory, forest districts and plots that are involved in the study. For some diseases, specific and more detailed forms can be prepared by the Forest Unit and performed by the Forest Service personnel, previously advised and instructed. A yearly report compiled by the Forest Research Unit of the ISMAA contains all the causes of damage, pests, pathogens and relationship with weather conditions. 26 Other data The values of climatic parameters are collected in several meteorological stations, placed in Trentino Province. Results and discussion In more than ten years of continuos monitoring, several parasites were observed and described in the Trentino woods. The list of the observed pest and pathogens in 2000 is reported in Table 1. Some parasitic fungi may cause economic damage and have a direct impact Table 1. List of the observed pests and pathogens in Trentino woods during the year 2000. Pathogens were listed following the host species. DAMAGE AGENTS HOST TREE PESTS Ips typographus Picea excelsa Tomicus minor e piniperda Pinus nigra e sylvestris Ips acuminatus Pinus sylvestris Coroebus florentinus Quercus spp. Rhynchaenus fagi Fagus sylvatica Thaumetopoea pityocampa Pinus nigra e sylvestris Zeiraphera griseana Larix decidua Coleophora laricella Larix decidua Caoptilia syringella Fraxinus ornus Yponomeuta evonymella Prunus padus Haematoloma dorsatum Pinus nigra Heterarthrus cuneifrons Acer pseudoplatanus PATHOGENS Chrysomyxa rhododendri and C. abietis Picea excelsa Heterobasidion annosum Picea excelsa, Larix decidua Rhizosphaera kalkhoffii Picea excelsa Sphaeropsis sapinea Pinus nigra e sylvestris Cenangium ferruginosum Pinus nigra e sylvestris Lachnellula willkomii Larix decidua Meria laricis Larix decidua Mycosphaerella laricina Larix decidua Lirula nervisequia Abies alba Herpotrichia parasitica Abies alba Cryphonectria parasitica Castanea sativa Phytophthora cambivora Castanea sativa Ophiostoma novo-ulmi Ulmus spp. Other damage Defoliation Ostrya carpinifolia Wilting Alnus viridis 27 on forest ecosystem evolution. Among them Heterobasidion annosum, Ophiostoma novo-ulmi and Lachnellula willkommii. The last one is able to produce damage on young regeneration of Larix in some areas. This parasite, killing Larix young trees, allows the persistence and the predominance of Norway spruce and stops the evolu tion of forests to mixed woods. Chestnut blight is also present in Trentino's chestnut stands but the disease impact is reduced by the clear prevalence of healing cankers due to hypovirulent isolates. On the contrary Dutch elm disease caused the disap pearance of Ulmus glabra and induced foresters to define a sanitation program in order to save the few surviving Elm trees. Other fungi are present in forest on small scale areas. Their impact appear effective although geographically localised. Among them Phacidium infestans Karsten and Herpotrichia juniperi (Duby) Petrak caused some damage on natural regenera tion of P. cembra and P. Abies particularly at high altitude where snow pockets can stand. Herpotrichia parasitica Rostrup and Lirula nervisequia (DC.) Darker are common on A. alba trees but they produce evident symptoms only in few sites: their effectiveness in damaging the growth of infected trees or in influencing the estab lishment of A. alba regeneration is not yet confirmed. Chrysomyxa rhododendri (DC.) de Bary is reported at timberline in several valleys but the infected surface varied considerably in the different years, probably in relation to the weather conditions during the infection period at the beginning of the growing season. Only recently survey started to estimate the rust influence on the spruce regeneration at the timberline. Several foliar and needles pathogens and rusts are regularly observed in woods but resulted harmful generally only on single trees and on limited parts of the crowns. Their presence is often unnoticed. Among these parasites: Sirococcus strobilinus Preuss is present on cones and shoots and Lirula macrospora Darker on needles of P. Abies, Gremmeniella abietina (Lagerb.) Morelet may colonise shoots and needles on P. cembra and P. sylvestris and G. laricina (Ettlinger) Schlapfer-Bernhard is occa sionally found on shoots and needles of larch. Both the fungi were not associated to evident damage till now. In the last years sudden and widespread outbreaks of some foliar disease were reported in all the Trento Province. During the summer 1999, in many larch woods severe foliar defoliation by Mycosphaerella laricina (Hartig) Neg were observed. The pathogen is common in central Europe but scarcely considered or unnoticed until now in Italy probably for the contemporary presence of the crown insects Coleophora laricella and or Adalges laricis. Damage was noticed, mainly on mixed woods, at around 1000 m a.s.l. of elevation on the southern slopes of valleys. Fungus caused needle necrosis and early defoliation starting from the beginning of Septem ber. The climatic conditions of summer 1999, characterised by relatively high tem peratures and high humidity during the conidial dispersal period (August) was the main factor involved in the epidemics (Maresi et al. 1999). At the end of winter 1999-2000, in several valleys it was observed an anoma lous reddening of Picea abies related to the presence of Rhizosphaera kalkhoffii Bubäk on necrotic needles. Trees showed complete necrosis of 1-3 years old nee dles. Damages were concentrated on the bottom of small valleys, generally above 28 1200 m. a.5.1., on the exposed part of the crown and on trees placed at the margin of forests. Symptoms were related to climatic stress conditions during winter, enhanced by the longer vegetative season in 1999, characterised by rain and temperatures higher than usual (Maresi et al. 2001). During July 2000, in some mixed larch forests, plants showed a complete crown yellowing, showing tips of needles first brown then whitened. On infected foliage it was detected Meria laricis Vuill. The fungus was not described before in Trentino although was probably undetected. Its pathogenic role appears limited. After the sudden appearance the disease stopped and no more effective damages were observed. Lower temperatures and heavy precipitation during July may be the pre disposing factors for the attacks (Maresi et al. 2001). Climatic condition during the last few years and particularly several spring summer drought periods induced spreading of Sphaeropsis sapinea in Pinus nigra stands which in the past were planted in poor soil areas of the Province. The attacks of the parasite were sometimes associated with Cenangium ferruginosum. Starting from 1995 damage were so heavy (more than 11000 cubic meters were felled during 1998) that the survival of some stands resulted very difficult (Maresi et al. 1999). Both the fungi can be considered as bio-indicator of decline of "pioneer" pine stands which are at the end of their life cycle (Ambrosi et al. 2000). Silvicultural manage ment is needed to reduce impact of damage and to improve natural evolution of stands towards more resilient broadleaf woods. Conclusion Forest Tree Damages Monitoring proved to be an useful tool both to collect data on pathogens and to support the managing choices. From the recorded data it is possible to combine information on the presence of pathogens, site conditions and transfer them on maps to use in the Woodland Management Plans. The improvement of a GIS dedicated to forest health problems is one of the future goals. In this contests knowledge on the role of several "minor" parasites in forest ecosystem resilience must be enhanced both by a better mapping of their presence and spread and working on their relationship with the eco-physiological status of trees. Most of the recent outbreaks, including those by minor or unknown pathogens, appear strictly related to weather conditions particularly favourable to the micro organisms or to climatic factors that induced stress on hosts trees as in the case of S. sapinea attacks or as supposed for the observed wilting of Alnus viridis (Maresi and Ambrosi 1999). In this contest fungal pathogens may be considered as bio-indicators both of wrong past silvicultural choices and-or climatic changes that can modify forest sta bility and evolution. A better knowledge on these topics is desirable. 29 References Ambrosi, P. and Salvadori, C. 1998. Monitoraggio fitopatologico delle foreste quale strumento per la gestione selvicolturale: otto anni di applicazione in Trentino. Monti e Boschi IL, 5: 9- 12. Ambrosi, P., Minerbi, S. and Maresi, G. 2000. II deperimento di alcune pinete in Trentino-Alto Adige: fattori biotici ed abiotici coinvolti. Atti del II Congresso SISEF "Applicazioni e prospettive per la ricerca forestale italiana" Bologna, 20-22 ottobre 1999. Ed. Bucci G., Minotta G. and Borghetti M. pp 441-446. Maresi, G. and Ambrosi, P. 1999. Nuovi disseccamenti di ontano verde: prime osservazioni in Trentino. Sherwood, Foreste ed alberi oggi, n. 47 (5, n.7): 39-42. Maresi, G., Ambrosi, P. and Capretti P. 2001. Sulla presenza di Meria laricis Vuill. nei lariceti trentini. (in collaborazione con P. Ambrosi eP. Capretti) Monti e Boschi LII, 1: 18-22. Maresi, G., Ambrosi, P. and Capretti, P. 2000. Attacchi di Mycosphaerella laricina (Hartig) Neg. in lariceti del Trentino. Monti e Boschi LI, 1: 27-31. Maresi, G, Ambrosi, P., Angeli, F. and Capretti, P. 2001. Arrossamenti delle chiome e Rhizosphaera kalkhoffii Bubäk su Picea abies in Trentino. Monti e Boschi LII, 3-4: 19-23. Maresi, G., Ambrosi, P., Confalonieri, M. and Capretti, P. 1999. Disseccamenti da Cenangium ferruginosum e Sphaeropsis sapinea nelle pinete trentine. Monti e Boschi L, 2: 35-41. 30 Taxonomy of the genus Gremmeniella, causal agent of scleroderris canker Gaston Laflamme Natural Resources Canada, Canadian Forest Service, Laurentian Forestry Centre, 1055 du P.E.P.S., P.O. Box 3800, Sainte-Foy, Quebec, Canada GIV 4C7. E-mail: glaflamme@cfl.forestry.ca Abstract The conifer disease known as Scleroderris canker is caused by ascomycetes of the genus Gremmeniella. The taxonomy of these pathogens has undergone many changes from the rank of order down to the rank of species, since the first description of the teleomorph in Sweden under the name Crumenula abietina Lagerberg, in 1913. A second species found in Switzerland on larch, C. laricina Ettlinger, was described in 1945. In 1975, three serovars or races were recog nized for the species G. abietina. In 1989, a reassessment of the genus was made based on results of morphological, cultural and biochemical studies. Thus, in addition to the var. abietina, repre senting the three serovars, the var. balsamea was proposed for the pathogen infecting spruce and balsam fir in Canada. New information on the epidemiology of these fungi infecting different hosts, in combination with results of sequencing the isolates representative of the host range of the disease, provides a better understanding of this disease on its respective hosts. A review of the genus Gremmeniella is presented, including a proposal for several new species. Keywords: Scleroderris canker, Gremmeniella, taxonomy Introduction Pathogen fungi belonging to the genus Gremmeniella cause serious diseases of coni fers in Asia, Europe and North America (Donaubauer 1972, Karlman et al. 1994, Laflamme and Lachance 1987, Ohman 1966, Setliffet al. 1975, Yokota 1984). In the genus Gremmeniella, morphological characteristics are not very helpful for micro scopic identification and overlaps are common between measurements of specimens from different hosts (Petrini et al. 1989). The diagnostic of the disease on pine was limited to the identification at the genus level which was believed for quite a long time to be monospecific with G. abietina. This is why the pathogen identification at the species level was not considered to be a problem even if fungal taxonomy is a key element in the understanding of a pathosystem. For example damage caused by the so called European race, which can kill large trees, is very different from that caused by all other Gremmeniella types. Symptoms of the disease are quite well known, starting in early spring with brown coloration at the base of the previous year's needles, followed by a shoot 31 blight in the summer (Laflamme 1991). Cankers can be produced a few years later on twigs and on the main stem. The English name of the disease comes from the fungus'scientific denomination Scleroderris lagerbergii, used in the late 1960s when the disease became an epidemic in western Europe, in eastern North America and in Japan. This disease has been found on several conifer species of the northern hemi sphere but the most severe damage has been recorded on pine. Classification from the order to the genus These Ascomycetes were classified in the order Helotiales and in the family Helotiaceae by Dennis (1968); these fungi are now ranked in the order and family Leotiales and Leotiaceae by Hawksworth et al. (1995). Gremmeniella, the genus denomination, is now widely accepted, but this was not always the case; the denomi nation went through several changes over the last century. When discovered in 1913 by Lagerberg in Sweden, this pathogen was first included in the genus Crumenula. Forty years later, Gremmen decided to integrate this fungus into the genus Scleroderris. Several fungal taxonomists did not agree with this last change and from 1969 to 1971 three different genera were proposed by three different authors. Schläpfer Bernhard (1969) proposed integrating this fungus into the genus Ascocalyx. At the same time, Morelet (1969) and Reid (in Dennis 1971) created two different new genera, respectively Gremmeniella and Lagerbergia. Gremmeniella was the valid name because it was published first. But it took some time before everyone agreed on that name because Ascocalyx was proposed again (Miiller and Dorworth 1983). Based on a study using chemical, biochemical, cultural and morphological informa tion on these fungi, Petrini et al. (1989) supported the name Gremmeniella. Table 1. List of genera used for the teleomorph denomination during the last century for the causal fungi of Scleroderris canker. Species in the genus Gremmeniella Morelet (1969) created the new genus Gremmeniella with G. abietina as the type species and the only species of the genus. Crumenula laricina was then moved to the genus Encoeliopsis. A few years later, Dorworth and Krywienczyk (1975) recog nized three pathological races based on serology: the North American race, the Eu Crumenula abietina 1913 Scleroderris 1953 Ascocalyx 1969 Gremmeniella 1969 (Lagerbergia) 1971 Ascocalyx 1983 Gremmeniella 1989 32 ropean race and the Asian race. These races or serovars do not have any taxonomic value. Petrini et al. (1989) studied isolates and their corresponding herbarium speci mens of Gremmeniella spp. from pine, spruce and balsam fir as well as specimens of Ascocalyx laricina from larch and of Ascocalyx abietis from balsam fir. For that study, morphological, cultural and chemical characteristics were used. All the speci mens from larch were very distinct from the specimens of all other hosts but had the characteristics of the genus Gremmeniella. Therefore, a new combination was cre ated, G. laricina (Ettlinger) Petrini et al. The species Ascocalyx abietis Naumov, a saprophyte on fir, was retained as it is distinct from G. abietina. All the specimens on the other hosts had overlaps in their morphological measurements and the authors could not create new species. Then two varieties were recognized: the variety abietina for the type species including the three serovars, and the variety balsamea for the specimens from spruce and balsam fir from Canada. New information on the disease and on the fungi give a better understanding of the genus Gremmeniella. It was believed that the European race was more viru lent than the other two races on all pine species, but that was not true. In fact P. banksiana and P. contorta are strongly resistant to the European race (Laflamme and Blais 2000, Laflamme et al. 2000, Simard et al. 2000) and probably other pine spe cies are too. The isolates from pine, balsam fir and spruce in North America show host specificity reactions (Laflamme et al. 1996). Phylogenies based on sequencing show that the three races of the type species are not closely related (Hamelin and Rail 1997). The Asian race is closely related to the variety balsamea. The diver gence between G. abietina and G. laricina is as great as between the varieties balsamea and abietina. Finally, in the variety balsamea, it is possible to differentiate isolates from fir and spruce by their colour in culture on PDA. In Europe, the so-called European race of G. abietina has been divided into three entities. The Fennoscandian amplitype (Hamelin et al. 1996) is the equivalent of the small tree type in Sweden (Hellgren 1995) or type B in Finland (Uotila 1983). The second one, the Alpine amplitype, is restricted to the Alps and damage caused by the pathogens is similar to the previous one and to the North American race in North America. The third one, the European amplitype, is the equivalent of the large tree type in Sweden (Hellgren 1995), type A in Finland (Uotila 1983) and the European race in North America. The species G. juniperina L. Holm & K. Holm has not been studied. Discussion From this review, it is becoming more and more evident that the genus Gremmeniella is a complex of several species. When Morelet (1980) studied the type species Crumenula abietina, he did not have access to the type specimen collected by Lagerberg on Picea abies. As the type specimen was lost, he had to propose a neotype, which was another specimen that Lagerberg collected on Pinus sylvestris. But the first specimen on spruce could have been the European amplitype and the one on Scots pine could have been the Fennoscandian amplitype. If the first Lagerberg speci men could be found, further observations could clarify this point. 33 For the past several years, we have studied Gremmeniella on fir and spruce in Canada (Lafiamme 1988 a, 1988b). We are now confident that we are dealing with two different species and a paper is in preparation to describe them. We did not have enough material from Japan on Abies sachalinensis but it seems possible that the Asian race could prove to be a new species after further studies. Finally, more observations should be done on isolates from G. juniperus. Conclusion In short, the complex of suggested species is as follows: 1. G. abietina (or a new species?) found mainly on Pinus spp. = NA race = small tree type = type B = Finoscandian amplitype = Alpine amplitype, which is very similar to the Fennoscandian one 2. G. abietina (or a new species?) found mainly on Pinus spp. = European race = large tree type = type A = European amplitype 3. G. n.sp. 1 on Abies balsamea in Canada = G. abietina var. balsamea 4. G. n.sp. 2 on Picea spp. in Canada was included in G abietina var. balsamea 5. G. (new species?) on Abies sachalinensis in Japan = Asian race 6. G laricina on Larix spp. Is the species in Europe different from the one in North America? 7. G juniperina on Juniperus: to be further studied. References Dennis, R.W.G. 1968. British Ascomycetes. Verlag von J. Carmer, Stuttgart, Germany. Dennis, R.W.G. 1971. New or interesting British microfiingi. Kew Bulletin 25: 335-374. Donaubauer, E. 1972. Distribution and hosts of Scleroderris lagerbergii in Europe and North America. Eur. J. For. Pathol. 2: 6-11. Dorworth, C.E. and Krywienczyk, J. 1975. Comparisons among isolates of Gremmeniella abietina by means of growth rate, conidia measurement, and immunogenic reaction. Can. J. Bot. 53: 2506-2525. Hamelin, R.C. and Rail, J. 1997. Phylogeny of Gremmeniella spp. based on sequences of the 5.8S rDNA and internal transcribed spacer region. Can. J. Bot. 75: 693-698. Hamelin, R.C., Lecours, N., Hansson, P., Hellgren, M. and Laflamme, G. 1996. Genetic differen tiation within the European race of Gremmeniella abietina. Mycol. Res. 100: 49-56. Hawksworth, D.L., Kirk, P.M., Sutton, 8.C., and Pegler, D.N. 1995. Ainsworth & Bisby's Dic tionary of Fungi. Bth8 th ed. CAB International. 34 Hellgren, M. 1995. Comparison of Gremmeniella abietina isolates from Pinus sylvestris and Pinus contorta in terms of conidial morphology and host colonization. Eur. J. For. Pathol. 25: 159- 168. Karlman, M., Hansson, P. and Witzell, J. 1994. Scleroderris canker on lodgepole pine introduced in northern Sweden. Can. J. For. Res. 24: 1948-1959. Laflamme, G. 1988 a. Scleroderris canker on balsam fir. Can. J. Plant Pathol. 10: 367. Laflamme, G. 1988b. Description et distribution du chancre scleroderrien sur Picea mariana (MiII.)B.S.P. Eur. J. For. Pathol. 18: 230-239. Laflamme, G. 1991. Scleroderris canker on pine. For. Can - Quebec Region. Information Leaflet LFC 3 (revised 1991). Laflamme, G. and Blais, R. 2000. Resistance of Pinus banksiana to the European race of Gremmeniella abietina. Phytoprotection 81: 49-55. Laflamme, G., Blais, R., Bussieres, G. and Mallett, K. 2000. Resistance to Gremmeniella abietina, European race, in Pinus contorta. Can. J. Plant Pathol. 22: 187. Laflamme, G. and Lachance, D. 1987. Large infection center of Scleroderris canker (European race), in Quebec province. Plant Dis. 71: 1041-1043. Laflamme, G., Ylimartimo, A. and Blais, R. 1996. Host preference of two Gremmeniella abietina varieties on balsam fir, jack pine, and black spruce in eastern Canada. Can. J. Plant Pathol. 18: 330-334. Morelet, M. 1969. Un discomycete inopercule nouveau. Bull. Soc. Sci. Nat. Archeol. Toulon. 183,9. Morelet, M. 1980. La maladie ä Brunchorstia. I. Position systematique et nomenclature du pathogene. Eur. J. For. Pathol. 10: 268-277. Mtiller, E. & Dorworth, C.E. 1983. On the discomycetous genera Ascocalyx Naumov and Gremmeniella Morelet. Sydowia 36: 193-203. Ohman, J.H. 1966. Scleroderris lagerbergii Gremmen: the cause of dieback and mortality of red and jack pines in upper Michigan plantations. Plant Dis. Rep. 50: 402-405. Petrini, 0., Petrini, L.E., Laflamme, G. and Ouellette, G.B. 1989. Taxonomic position of Gremmeniella abietina and related species: a reappraisal. Can. J. Bot. 67: 2805-2814. Schläpfer-Bernhard, E. 1969. Beitrag zur Kenntnis der Discomycetengattungen Godronia, Ascocalyx, Neogodronia und Encoeliopsis. Sydowia 22: 1-56. Setliff, E.C., Sullivan, J.A. and Thompson, J.H. 1975. Scleroderris lagerbergii in large red and Scots pine trees in New York. Plant Dis. Rep. 59: 380-381. Simard, M., Rioux, D. and Laflamme, G. 2000. Formation of ligno-suberized tissues in Pinus banksiana: response to infection by the European race of Gremmeniella abietina. Can. J. Plant. Pathol. 22: 192. Uotila, A. 1983. Physiological and morphological variation among Finnish Gremmeniella abietina isolates. Comm. Inst. Forestalls Fenniae 119 12p. Yokota, S.T. 1984. Pathogenicity and host range of races of Gremmeniella abietina in Hokkaido. In Manion, PD. (Ed.) Proc. Int. Symp. on Scleroderris canker of conifers, Syracuse, N.Y., June 21-24, 1983. pp. 47-53. 35 Size of ascospores in the A and B types of Gremmeniella abietina Timo Kurkela and Antti Uotila T. Kurkela, Finnish Forest Research Institute, Vantaa Research Centre, RO. Box 18, 01301, Finland. E-mail: timo.kurkela@metla.fi A. Uotila, University of Helsinki, Hyytiälä Forestry Field Station, Hyytiäläntie 124, 35500 Korkeakoski, Finland. Abstract To find morphological differences, ascospores of A- and B-types of Gremmeniella abietina were measured and compared. A significant difference between A- and B-types was found in the length of ascospores. Width and roundness of the spores did not differ or varied irregularly between samples and both types. However, the frequency distributions in the length measurements were overlapping too much to be useful as a taxonomic key. If still used, to get reliable data at least one hundred spores have to be measured. Introduction The fungus Gremmeniella abietina, which causes scleroderris canker of pines and some other conifers, is known to be a complex of several geographical strains or races. Dorworth and Krywienczyk (1975) suggested to divide the fungus into North American, European, and Asian races, based on physiological and immunological characteristics. By comparing morphology and electrophoretic patterns, Petrini et al. (1989), segregated G. abietina into two varietis, G. abietina var. abietina (includ ing European, North American, and Asian races) and G. abietina var. balsamea which occurs on Picea and Abies in Quebec. Later Petrini et al. (1990) found also some evidence for the existence of one more variety occurring on Pinus cembra at high altitudes in the Alps and on P. sylvestris in the boreal forests of northern Europe. DNA-tests conducted in Sweden by Hellgren and Högberg (1995) revealed that G. abietina had two ecologically different strains (LTT and STT) which seemed to have reflection with A and B types of the fungus described by Uotila (1983, 1992) in Finland. Finally, this ecotypic variation was found to correlate in Europe with three distinct DNA amplification profiles (Hamelin et al. 1996), from which two, northern European and Alpine types were adapted to long-lasting snow and the third type was spread throughout Europe. In further studies Hamelin and Rail (1997) showed that two varieties, G. abietina var. abietina and G. abietina var. balsamea are divergent at species level. Results from hybridization experiments allowed Uotila et al. (2000) to conclude that A and B biotypes are genetically different populations and should be described as separate species as soon as sufficient information is available. 36 The aim of this study was to find out possible morphological differences in the size of ascospores between A and B biotypes of G. abietina. Materials and methods A-type samples were from southern Finland and the B-types were from both south ern and northern Finland (Table 1). The type of each sample was tested with DNA markers (Hantula and Mtiller 1997). Spores from mature apothecia were allowed to drop to microscopic objective glasses on which they were mounted. Only specimens with mature apothecia were chosen for the study. We concluded that apothecia were mature, if high number of spores were released shortly (in few hours) after moisten ing. Ascospore images were obtained in microscope with a digital camera using 10x40 magnification. Image analyser (Leica Qwin) was used to measure the dimensions of ascospores. In this system one pixel was equivalent to 0.21 jam. The software meas ured automatically the area, length, width, perimeter, and roundness of the spores. Roundness was determined with the following equation: ANOVA was used for statistical treatment of the data. Results Ascospores representing B-type were longer and slightly wider than those of A-type (Fig. 1, Table 2). The differences in both dimensions were statistically significant (pO.OOl). In the average length of spores the difference was 3.7 (im. That caused significant difference also in the area and roundness. The width differed only very slightly, 0.2 (am. The spores were the longest in Solo B, which differed significantly also from other B type specimens (p<0.05). The appearance of spores was also vari able. In some specimens the ascospores had rounded ends while in some the ends were much more acute. However this difference was not possible to reveal by round ness measurements. This character differed between some specimens but appeared to vary quite much in both types. Table 1. The specimens of Gremmeniella abietina used in the study and the number of spores measured. , Perimeter 2 Roundness = 4x7ix Area x 1.064 Specimen Type Location Coordinates Number Solo B Inari, Solojärvi N68°49\ E26°47' 104 Arctic circle B Rovaniemi, Arctic circle N66°33', E25°38' 84 Siika B Ruovesi, Siikakangas N61°52', E24°10' 120 Huikko A Juupajoki, Huikko N61°51',E24°19' 77 Valkea A Ruovesi, Valkeajärvi N61°52', E24°16' 117 37 Figure 1. The ascospore length and width in A and B type specimens. Table 2. Mean values of the characteristics of ascospores. Discussion We found spore size difference between A and B types of G. abietina, which could be a useful clue for describing these two types as distinct taxa. However, the difference could hardly be used as an identification key, since the frequency distributions are too much overlapping on the scale. The maturity of apothecia could also be a limit ing factor when one is using spore size as distinguishing character. We do not know when the spores in an ascus reach their maximum size. Is it when they are ready to be released, or is it long before that phase. Usually spores are measured from smashed apothecia on microscopic glasses, in which case they could be released by an exter nal force prior to maturity. The number of spores measured is also important, if we try to distinguish these two types using spore sizes. It seems that at least 100 spores have to be measured for obtaining reliable data, instead of 5-10 that is common practice in identification (Hawksworth 1974). If we like to describe A and B types as different species, we will meet another difficulty. We have to separate the new proposed species also from other related Strain Area, |im 2 Length, nm Width, nm Roundness A 50.2 14.6 4.9 1.9 B 61.4 18.3 5.1 2.3 38 strains, such as North American strain and alpine strain in Central Europe, both occurring on pines, and other ones occurring on firs in North America and Japan. For describing a new species in this complex, more reliable data have to be obtained. It has been shown genetically (Hamelin et al. 1996, Hamelin and Rail 1997, Uotila et ai. 2000) that G abietina should be divided to several distinct species. However, to establish new species in this complex is very difficult until the evidence obtained with tests using DNA-markers is not accepted at the same level as morphological characteristics. Therefore, we still need morphological evaluations of all strains of G. abietina. In the present study, we found hints that some morphological differences could be detected between these strains. References Dorworth, C. E. and Krywienczyk, J. 1975. Comparisons among isolates of Gremmeniella abietina by means of growth rate, conidia measurement, and immunogenic reaction. Canadian Jour nal of Botany 53: 2506-2525. Hamelin, R. C., Lecours, N., Hansson, P., Hellgren, M. and Laflamme, G. 1996. Genetic differen tiation within the European race of Gremmeniella abietina. Mycological Research 100: 49- 56. Hamelin, R. C. and Rail, J. 1997. Phylogeny of Gremmeniella spp. based on sequences ofthe 5.8S rDNA and internal transcribed spacer region. Canadian Journal of Botany 75: 693-698. Hantula, J. and Miiller, M. M. 1997. Variation within Gremmeniella abietina in Finland and other countries as determined by random amplified microsatellites (RAMS). Mycological Research 101: 169-175. Hawksworth, D. L. 1974. Mycologist's handbook. Commonwealth Mycological Institute, Kew, Surrey. 231 pp. Hellgren, M. and Högberg, N. 1995. Ecotypic variation of Gremmeniella abietina in northern Europe: disease patterns reflected by DNA variation. Canadian Journal of Botany 73: 1531- 1539. Petrini, 0., Petrini, L. E., Laflamme, G. and Ouellette, G. B. 1989. Taxonomic position of Gremmeniella abietina and related species: a reappraisal. Canadian Journal of Botany 67: 2805-2814. Petrini, 0., Toti, L., Petrini, L. E. and Heiniger, U. 1990. Gremmeniella abietina and G. laricina in Europe: chracterization and identification of isolates and laboratory strains by soluble pro tein electrophoresis. Canadian Journal of Botany 68: 2629-2635. Uotila, A. 1983. Physiological and morphological variation among Finnish Gremmeniella abietina isolates. Communicationes Instituti Forestalis Fenniae 119: 12 p. Uotila, A. 1992. Mating system and apothecia production in Gremmeniella abietina. European Journal of Forest Pathology 22: 410-417. Uotila, A., Hantula, J., Väätänen, A. K. and Hamelin, R. C. 2000. Hybridization between two biotypes of Gremmeniella abietina var. abietina in artificial pairings. Forest Pathology 30: 211-219. 39 Factors affecting the risk of Gremmeniella abietina infection in Scots pine stands Seppo Nevalainen and Ulla Mattila Finnish Forest Research Institute, Joensuu Research Centre, RO. Box 68, 80101 Joensuu, Finland. E-mail: seppo.nevalainen@metla.fi Abstract Field inventory data from the temporary plots of the Bth8 th National Forest Inventory (NFI) in south ern Finland, digital elevation models, modelled air pollution deposition and models for monthly temperature and rain were used in this study. Gremmeniella abietina was the most frequently identified cause of damage in Scots pine stands in southern Finland. The disease was clearly spatially clustered. On mineral soils, disease frequencies were highest on paludified plots or on the most fer tile plots. On peatland plots, the disease was most common on drained, transformed peatland plots. Naturally regenerated stands were affected more than artificially regenerated stands. The proportion of diseased plots increased with an increase in stand density and in absolute elevation in mineral soils. The relative altitude of the plot had a stronger influence. The risk of G. a. infestation was modelled with various techniques. CART (classification and regression trees) analysis was used for data exploration. Already at this stage, it was possible to recognize groups of very high or very low risk. 77.5% of the diseased stands were correctly classified. Stand density, drainage stage, relative elevation, regeneration method, division be tween mineral soil/peatland plots and weather factors were the most important classifying factors. A preliminary risk model with multi-level logistic regression gave support to the importance of transitional stages after drainage on peatland, rich site type on mineral soil and, most of all, stand density. Keywords: Gremmeniella abietina, Pinus sylvestris, Finland, national forest inventory, stand density, peatlands, topography, climate, risk modelling Introduction The ascomycete fungus Gremmeniella abietina (Lagerb.) Morelet is the causal agent of a canker and die-back disease of Pinus and many other coniferous genera. The disease has been known in Europe for more than a century. In southern parts of Finland severe outbreaks have been reported in the mid-1970'5, the early 1980's and in 1988 (Aalto-Kallonen and Kurkela 1985, Nevalainen and Yli-Kojola 1990). In southern Finland, it is the most common fungal disease in forests dominated by Scots pine (Pinus sylvestris L.). According to the Bth8 th National Forest Inventory (NFI), disease symptoms were recorded in 10.5% of Scots pine stands, on more than 690 000 hectares during 1986-1992. 40 Several environmental factors may affect the risk of forest stands to infection by G. abietina. Planting of poorly adapted provenances is one of the most important reasons for severe epidemics of Scots pine in Europe, although physiogenic diseases are also important in this contex (Dietrichson, 1968, Uotila 1985, Roll-Hansen 1972). When pines are of suitable origin, epidemics can be favoured by meteorologi cal factors and site conditions. Epidemics usually begin after cold and rainy growing seasons (Aalto-Kallonen and Kurkela 1985, Uotila 1988); and on the other hand, warm growing seasons can slow down the spread of the disease (Karlman et al. 1994). Frost damage is an important reason for the differences in susceptibility to G. abietina (Dietrichson 1968). Steep valleys, north-facing slopes and kettle holes are among the most vulnerable sites (Dorworth 1974, Uotila 1988, Sairanen 1990). Sus ceptibility of pines is increased in shaded or dense stands (Read 1968, Niemelä et al. 1992). Some epidemics have occurred on areas with considerable pollution (Bragg and Manion 1984), but in Scandinavia the effects of air pollution have not been confirmed with experiments or field observations (Kaitera et al. 1995, Vuorinen and Uotila 1997). Most of the information about the factors that can affect the risk of infection come from experiments or small scaled field studies. Experiments are normally aimed at testing one specific hypothesis at a time. Local field studies, on the other hand, have most often been conducted in known diseased centres. Thus, they may overem phasize some predisposing factors. Only large-scaled, statistically sound samples make a simultaneous examination of the several predisposing variables and their relative importance conceivable. The Finnish National Forest Inventory (NFI) has produced information on forest resources over large areas of the country for more than 70 years. The Eighth NFI was the first to include detailed information on health of forests, including diseases and pests. The NFI dataset allows for the simultaneous comparison of several environmental and silvicultural factors affecting the occur rence of diseases and pests. It also provides raw data for risk modelling. Further more, it is possible to combine other spatial data and/or models with the NFI field measurements. The purpose of this study is to describe the occurrence Gremmeniella abietina in relation to several environmental, site and silvicultural factors, based on a large representative field sample, and to develop models for estimating the susceptibility of Scots pine stands. Material and methods The Bth8 th National Forest Inventory (NFI) was a systematic field sampling of the forest resources covering the entire area of Finland. Field plots were arranged in detached 21-plot clusters, called tracts, the plots forming a half-square. In southern Finland the distance between the tracts was 8 km x 7 km (south-north x west-east directions) and the distance between the plots was 200 metres. Altogether 18999 Scots pine dominated stands (without recent cuttings) on forest land, from the years 1986-1992, in southern Finland were included in this study. Each "stand" was actually the forest compartment in which the centre point of the plot was situated. 41 Forest damage was assessed in every plot situated on forest land. By defini tion, on forest land, the mean annual increment of wood should be at least 1 m 3 /ha. Forest damage was assessed using three different codes: the symptom, the cause and the apparent severity of the damage (damage degree). Only stand level damage as sessments were used in this study. Only one, the primary damage, was recorded for each stand. Description of the variables and codes used for assessing forest damage in this study are presented in Table 1. In the analysis, the disease degrees were re corded into three classes: o=no Gremmeniella infection l=slight infection 2=at least moderate infection (disease degree more than 1). About one hundred variables de scribing the stand and site conditions were recorded in the NFI. A more detailed description of the inventory system and of the most important variables assessed in the field in the NFI can be found in Nevalainen (1999). In addition, some external data was merged to the NFI dataset. A nationwide elevation data set, provided by the National Land Survey of Finland, was used to study the effects of elevation. The digital elevation model (DEM) was calculated from the contour lines of the basic map by triangulation network interpolation onto a grid model. The original grid cell size was 25 m x 25 m, but in this study the DEM was mean filtered to a 1 km x 1 km grid. This grid was used to compute the elevation of the 1 km x 1 km cells in the tract area and the relative elevation of the plot, i.e. the absolute elevation minus tract area elevation. The average sulphate and total nitrogen deposition on the sample plots was based on a model applied at the Finnish Meteorological Institute. The long-range deposition in the model calculations was obtained from the EMEP model (Tuovinen Table 1. Description of the variables and codes used for assessing forest damage in this study. Description Codes Visual symptom of 0) no damage 1) dead standing tree 2) fallen tree or standing injury stem broken below the crown 3) decay 4) stem or root damage within 1 m from the stem 5) broken or dry top (in the upper half of the crown) 6) other crown malformations 7) defoliation 8) discoloration of the needles or leaves Letters A-F are used for symptoms 1-6, if the injury is older than five years Primary cause of injury 0) unknown l)wind 2) snow 3) other meteorological or soil factors 4) competition between plants 5) harvesting scars 6) other human activity 7) voles 8) moose 9) Tomicus sp. 10) Other insects 11) Cronartium sp. 12) Gremmeniella abietina 13) Other fungi Importance (degree) of 0) slight damage, symptoms observed, but the damage does not the damage reuce the silvicultural quality of the stand 1) moderate, the stand quality is reduced by one class 2) severe, the stand quality is reduced by more than one class 3) complete, artificial regeneration is required 42 et ai. 1994), and the domestic deposition was based on the HAKOMA (Finnish inte grated acidification) model (Johansson et al. 1990). The chemical composition of till at depths of 0.5-2 m was obtained from the data of the Geochemical Atlas of Finland (Koljonen et al. 1992). In this study, con centrations of calcium, magnesium and potassium were computed for each field plot of the NFI by the Geological Survey of Finland. Climatic data were generated using the models of Ojansuu and Henttonen (1983). The monthly mean temperature, effective temperature sum and precipitation sums for each plot were computed for the period of ten years prior to the inventory year, and as a grand mean for the 30-year normal period (1960-1990). Classification and regression trees (CART) analysis (Breiman et al. 1984, Steinberg and Colla 1997) was used for data exploration, e.g. for selecting the vari ables to be included in logistic regression analysis, and for determining the cut points of continuous variables. A preliminary risk model was built using the MlwiN software (Goldstein et al. 1998). It can be assumed that the binary responses of the stands (occurrence of damage) are correlated within for example different spatial levels, year level, meas urement team level etc. In modelling the probability of damage, ignoring these cor relations leads to underestimated standard errors of estimated regression coefficients. The problem can be mitigated by the introduction of random variables to the model, estimating the residual variances at different hierarchical levels of the data (Goldstein 1995). Therefore, a multilevel logit-model was used in the model construction. The model can be expressed as where y.jk is the observed response of stand i in cluster j measured by team k, 7T.jk is corresponding response probability, x. jk are explanatory variables of the model, |3 ijk are the coefficients for the explanatory variables anduuk.k and vk are normally distrib uted errors at cluster and measurement team levels, respectively. The predictors for the fixed parts of the probability models were selected stepwise based on the Wald test (Hosmer and Lemeshow 1989), in addition to the CART procedure mentioned earlier. Results Relationships between the disease and some environmental and silvicultural factors Stand density, here expressed as basal area (m 2/ha) was most clearly correlated with the disease degree (0-2), and the relation was almost linear up to 28 m 2/ha (Table 2, Fig. la). The dummy variable "artificial regeneration" correlated negatively with the disease degree, indicating more disease in naturally regenerated stands (Table 2). logit(7t ijk )= ln(7t ijk / (l-7i ijk)) = P oljk+P luk x l]jk +.. ,+Pmjk Xnijk + Ujk + vk y ijk ~Binomial (niA) 43 The disease was more common onpeatland than on mineral soil plots. Drained peatlands were most liable to Gremmeniella infection, but the difference between undrained (natural state) peatland and mineral soil plots was not significant. The more the original peatland site type had changed after drainage, the more common was the disease. The correlations between the two drainage stages that had changed the original type the most and the disease degree were very significant, although mineral soils and peatlands were treated together. The fertility of the site correlated positively with the disease degree. Paludification (in mineral soil plots) as well as the variable indicating the need for drainage showed a positive correlation with the disease degree (Table 2). Damage due to Gremmeniella abietina increased slightly with an increase in absolute elevation (Fig. lb). However, the relative altitude of the plot compared to the elevation of the tract area was most strongly correlated with the disease. The disease was most common (and severe) in the plots situated lower than the mean elevation of the tract area (Fig. lc, Table 2). The modelled total S- and N-deposition were not correlated with the disease. The modelled total concentration of the base cations K, Mg and Ca in the till were slightly positively correlated. Of the monthly weather factors 10 years prior to infection, only rain sum during the summer months (June-August) correlated significantly with the disease degree. On the other hand, almost all the variables depicting the difference between 30 years normal period and the period 10 years prior to field survey, were signifi cant. Of these, the difference in rain during spring quarter (March-May) and during the most probable infective period (May-June), were the most significant (Table 2). Classification of the disease stands with CART was not very successful. 77.5% of the diseased stands were correctly classified as diseased with CART. The corre sponding number in the 'healthy' plots group was 66.0%, which gave the overall accuracy of 67.3%. Despite this fact, the CART-analysis was able to separate some groups with a very high or a very low disease frequency out of the data. For exam ple, the disease frequency was as high as 25.0% within the 776 plots which shared the definitions "stand density >12.5 m 2/ha and drainage stage something else than undrained mineral soil". On the other hand, the disease frequency was only 3.8 % within the 299 plots sharing the definitions "stand density <12.5 m 2 /ha and drainage stage undrained mineral soil, drained mineral soil or recently drained peatland". The most important variables (based on the relative importance of variables as primary splitters) varied slightly in different CART runs. In general, the most important were stand density (expressed as stand basal area), stand age, peatland (0/ 1), the regeneration method, drainage stages, the need for drainage, difference in winter temperature, autumn temperature, winter temperature and the relative eleva tion of the plot. 44 Table 2. Non-parametric correlations (Spearman's rho) between the disease degree (0-2) and some silvicultural and environmental variables. Significant correlations in each variable group are printed in bold. Variable Coefficient Significange Number of plots Proportion of pine -.026 .000 18999 Stand age .177 .000 18999 Density (basal area) .217 .000 18998 Artificial regeneration (0/1) -.112 .000 18999 Mean elevation in tract area (1 x 1 km pixels) .055 .000 18956 Elevation of the plot .022 .003 18999 Relative elevation of the plot -.097 .000 18956 Total S-deposition -.004 .573 18999 Total N-deposition -.004 .593 18999 Aluminium in till .016 .034 16950 Base cations (K, Ca, Mg) in till .019 .015 16950 Peatland plot (0/1) .117 .000 18999 Site needs drainage (0/1) .103 .000 18359 Palufied mineral soil (0/1) .045 .000 13682 Transition stage after drainage (peatlands) .073 .000 5317 Drained mineral soil (0/1) -.001 .915 18999 Recently drained peatland (0/1) .006 .385 18999 Transforming drained peatland (0/1) .092 .000 18999 Transformed drained peatland (0/1) .073 .000 18999 Undrained peatland plot (0/1) .003 .699 18999 Site type (l=richest) -.053 .000 18999 Grove or grove-like site type (0/1) .029 .000 18999 Fresh site type (0/1) .026 .000 18999 Dryish site type (0/1) -.006 .397 18999 Dry or poorer site type (0/1) -.047 .000 18999 Effective temperature sum -.016 .028 18379 Spring rain (March-May) 10 yrs prior survey .008 .278 18999 Spring temperature (March-May) -.010 .185 18999 Infective period rain (May-June) -.006 .381 18999 Summer rain (June-Aug) .037 .000 18999 Summer temperature (June-Aug) -.015 .037 18999 Autumn rain (Sept-Nov) .009 .218 18999 Autumn temperature (Sept-Nov) -.009 .204 18999 Winter rain (Jan-Feb) .020 .006 18999 Winter temperature (Jan-Feb) -.002 .799 18999 Difference in spring rain (30 yrs-10 yrs) -.040 .000 18999 Difference in spring temperature -.030 .000 18999 Difference in infective period rain -.039 .000 18999 Difference in summer rain .009 .230 18999 Difference in summer temperature -.016 .031 18999 Difference in autumn rain .030 .000 18999 Difference in autumn temperature -.027 .000 18999 Difference in winter rain -.022 .002 18999 Difference in winter temperature -.033 .000 18999 45 Figure 1. Schematic presentation of the relationships between the disease and some explanatory variables. The curves are smoothed in SPSS-interactive graphics using local linear regression, normal kernel and the bandwidth multiplier 1 a) stand density (expressed as stand basal area m 2 /ha) b) the elevation of the plot (m) c) relative elevation of the plot (plot elevation - elevation of the tract area, dm) d) precipitation sum during June-August for 10 years before field survey (mm). A preliminary risk model for the susceptibility of Scots pine stands The hierachial multilevel logit-model gave results that were quite similar than the ones reported in the earlier paragraph. However, several of the explanatory variables used earlier (Table 2) were dropped out of the model. For instance, the absolute elevation of plot was not included in the model. According to our preliminary model, the probability of damage was higher -In dense stands (density expressed as stand basal area here) -In stands situated lower than the mean altitude in the tract area (7xB km) -In areas, where it has rained much during May-June 10 years before the field sur vey (compared to a long-time average) -In drained peatlands (compared to undrained peatlands and mineral soils) -In grove-like, fresh, and dryish stands (compared to poorer site types) (Table 3). 46 On the other hand, the probability of damage was lower on artificially regen erated stands on fresh forest sites (compared to artificially regenerated dryish and poorer sites and to naturally regenerated fresh sites (Table 3). Statistically signifi cant residual variation was found on cluster level (a 2 =1.114, s.e. 0.077) and on measurement team level (variance a 2 =0.328, s.e. 0.127). The effects of fixed variables are shown in terms of population averaged odds ratios, which are calculated by taking the exponent of the estimated parameter value of the fixed part of the model (odds ratio = exp(b)). The population averaged odds ratio approximates, for example, how much the ratio of the frequency of damaged stands to undamaged stands changes when the fixed variable increases by one unit. For example, ratio of the frequency of damaged stands to undamaged stands is 1.07 times higher when the basal area of the stand increases by one m 2/ha. Accordingly, ratio of the frequency of damaged stands to undamaged stands is exp(l 0*0.068) = 1.97 times higher when the basal area of the stand increases by ten m 2 /ha (Table 3). Table 3. Estimated parameter coefficients and odds ratios for the preliminary risk model. Discussion In terms of sampling errors, the National Forest Inventory produces accurate re sults. However, this data set has many disadvantages regarding monitoring of forest diseases. The most important of these is that the field work covered two or three administrative areas (forestry board districts) per year, and thus, when epidemic dis eases are concerned, the assessments from different years are not directly compara ble. Further, only one injury per stand (the economically most important) was re corded, which leads to an underestimation of the overall occurrence of a disease. The authors think, however, that the most important stand damage (in an economic sense) was recorded reliably in the routine NFI inventory. Variable Estimated s.e. Odds ratio coefficient ((3) expP Relative elevation -0.014 0.002 0.98 (1 m) Difference in infective period rain -0.020 0.006 0.98 (1mm) Stand density (basal area) 0.068 0.004 1.07 (1 m2/ha) Recently drained peatland 0.813 0.174 2.25 (0/1) Transforming drained peatland 0.927 0.070 2.53 (0/1) Transformed drained peatland 0.834 0.099 2.30 (0/1) Grove & grovelike site 0.393 0.131 1.48 (0/1) Fresh site type 0.431 0.103 1.54 (0/1) Dryish site type 0.301 0.091 1.35 (0/1) Stand age 0.006 0.001 1.01 (1 year) Artificial regeneration -0.240 0.117 0.78 (0/1) Intercept -5.068 0.299 47 77% of the residual variation in the preliminary logit model occurred on sam pling cluster (tract) level. This could be a result of the cluster-wise systematic sam pling. In our opinion in merely indicates that the disease itself is clustered (Nevalainen 1999). Residual variation on measurement team level includes partly variation on larger spatial level than cluster and variation on year level. It, however, also suggests that in practice, the teams have not assessed the stands as damaged or undamaged with similar criteria. As to Gremmeniella abietina, the field inventory cannot have any data the occurrence of various races of the fungus. Although the age structure of affected stands and the disease symptoms refer to type A (or LTT), in the present study the presence of the B-type (STT) (Uotila 1983, Petäistö et ai. 1996) cannot be com pletely ruled out. Gremmeniella abietina occurred more frequently in naturally regenerated stands than in artificially regenerated stands, in contrast with the earlier findinds of e.g. Kallio et ai. (1985). This is due to the higher proportions of slight damage in natu rally regenerated stands. On the poorest forest types on mineral soils, the proportion of the disease has been higher in artificially regenerated stands than on naturally regenerated stands (Nevalainen 1999). The effect of the estimated climatic factors area partly questionable, due to the effect of elevation, which was included in their computation. The difference be tween the grand mean of 30 years and 10 years should remove artifacts like these. Then, however, almost all the weather variable differences became significant. These factors should be studied more closely in the future. The effect of topography was not studied in detail in this paper. However, the effect of e.g. aspect of the slope and microtopography around the plot were not as important as the relative elevation of the plot (as compared to the mean elevation in the tract area) (Nevalainen 2001). In our vision, disease risk models will be incorporated into forestry planning systems, although at the moment it is still unclear, how this should be done. Maybe they should form a part of "natural process models" (birth, growth and death mod els). References Aalto-Kallonen, T. and Kurkela, T. 1985. Gremmeniella disease and site factors affecting the condition and growth of Scots pine. Communicationes Instituti Forestalis Fenniae 126: 1- 28. Bragg, R.J. and Manion, P.D. 1984. Evaluation of possible effects of acid rain on Scleroderris canker of red pine in New York. In: Manion, P.D. (ed.). Scleroderris canker of conifers: proceedings of an International Symposium on Scleroderris Canker of Conifers, held in Syracuse, USA, June 21-24, 1983. Forestry sciences;l3. M. Nijhofl7W. Junk: 130-141. Breiman, L., Friedman, J., Olshen, R. and Stone, C. 1984. Classification and Regression Trees. Pacific Grove: Wadsworth. Dietrichson, J. 1968. Provenance and resistance to Scleroderris lagerbergii Gremmen (Crumenula abietina lagerb.). The international Scots pine provenance experiment of 1938 at Matrand. Meddelelser fra Det Norske Skogforsoksvesen 25: 398-410. 48 Dorworth, C.E. 1974. Epiphytology of Scleroderris lagerbergii in a kettle frost pocket. European Journal of Forest Pathology 3: 232-242. Goldstein, H. 1995. Multilevel statistical models. Edward Arnold, London. Goldstein, H., Rasbash, J., Plewis, 1., Draper, D., Browne, W., Yang, M., Woodhouse, G, and Healy, M., 1998. A user's guide to MlwiN. Version 1.0, January 1998. London: Multilevel models project, Institute of Education, University of London. Hosmer, D. W Jr., Lemeshow, S. 1989. Applied Logistic Regression, John Wiley and Sons, Inc., New York. Johansson, M., Kämäri, J., Pipatti, R., Savolainen, 1., Tuovinen, J.-P. and Tähtinen, M. 1990. Development of an integrated model for the assessment of Acidification in Finland. In: Kauppi, P., Anttila, P. and Kenttämies, K. (eds.) Acidification in Finland. Springer-Verlag, Berlin - Heidelberg: 1171-1194. Kallio, T., Häkkinen, R. and Heinonen, J. 1985. An outbreak of Gremmeniella abietina in central Finland. European Journal of Forest Pathology 15: 216-223. Kaitera, J., Fedorkov, A., Jalkanen, R., Krutov, V. and Tsvetkov, V. 1995. Occurrence of Gremmeniella abietina damage on Scots pine along a pollution gradient from Monchegorsk nickel smelter to western Lapland. European Journal of Forest Pathology 25 (1): 13-23. Karlman, M., Hansson, P. and Witzell, J. 1994. Scleroderris canker on lodgepole pine introduced in northern Sweden. Canadian Journal of Forest Research 24 (9): 1948-1959. Koljonen, T., Gustavsson, N., Noras, P. and Tanskanen, H. 1992. Sampling, analysis and data processing. In: Koljonen, T. (ed.). The Geochemical Atlas of Finland. Part 2: Till. Geologi cal Survey of Finland. Espoo: 16-27. Nevalainen, S. 1999. Gremmeniella abietina in Finnish Scots pine stands in 1986-1992 - a study based on the National Forest Inventory. Scandinavian Journal of Forest Research 14: 111- 120. Nevalainen, S. 2001. The incidence of Gremmeniella abietina in relation to topography in south ern Finland. Manuscript, submitted to Silva Fennica. Nevalainen, S. and Yli-Kojola, H. 1990. The occurrence of abiotic and biotic damage and its relation to defoliation (needle loss) of conifers in Finland (1985-1988). In: Kauppi, P. et al. Hapro report on acidification in Finland. Springer Verlag: 561-582. Niemelä, P., Lindgren, M. and Uotila, A. 1992. The effect of stand density on the susceptibility of Pinus sylvestris to Gremmeniella abietina. Scandinavian Journal of Forest Research 7: 129- 133. Petäistö, R.L., Uotila, A., Hellgren, M., Kaitera, J., Tuomainen, J. and Kajander, E.O. 1996. Two types of the European race of Gremmeniella abietina can be identified with immunoblotting. Mycologia 88: 619-625. Ojansuu, R. and Henttonen, H. 1983. Kuukauden keskilämpötilan, lämpösumman ja sademäärän paikallisten arvojen johtaminen Ilmatieteen laitoksen mittaustiedoista. Summary: Estima tion of the local values of monthly mean temperature, effective temperature sum and precipi tation sum from the measurements made by the Finnish Meteorological Office. Silva Fennica 17(2): 143-160. Read, D.J. 1968. Some aspects of the relationship between shade and fungal pathogenity in an epidemic disease of pines. New Phytologist 67: 39-48. Roll-Hansen, F. 1972. Scleroderris lagerbergii: Recistance and differences in attack between pine species and provenances. European Journal of Forest Pathology 2: 26-39. Sairanen, A. 1990. Site characteristics of Scots pine stands infected by Gremmeniella abietina in Central Finland. I. Mineral soil sites. Acta Forestalia Fennica 216. 27 pp. Steinberg, D. and Colla, P. 1987. CART-Classification and Regression Trees. San Diego, CA: Salford Systems. Tuovinen, J.-P., Barret, K. and Styve, H. 1994. Transboundary acidifying pollution in Europe: Calculated field and budgets 1985-1993. The Norwegian Meteorological Institute, Oslo. WEP/MSC-W Report 1. 258 p. 49 Uotila, A. 1983. Physiological and morphological variation among Finnish Gremmeniella abietina isolates. Communicationes Instituti Forestalls Fenniae 119. 12 pp. Uotila, A. 1985. Siemenen siirron vaikutuksesta männyn versosyöpäalttiuteen Etelä-ja Keski suomessa. Summary: On the effect of seed transfer on the susceptibility of Scots pine to Ascocalyx abietina in southern and central Finland. Folia Forestalia 639: 12. Uotila, A. 1988. Ilmastotekijöiden vaikutus männynversosyöpätuhoihin. Summary: The effect of climatic factors on the occurrence of Scleroderris canker. Folia Forestalia 721: 1-23. Vuorinen, M. and Uotila, A. 1997. The effect of acid rain treatments on the susceptibility of Pinus sylvestris to Gremmeniella abietina. European Journal of Forest Pathology 27 (2): 125-135. 50 Formation and growth of stem cankers caused by Gremmeniella abietina on young Pin us contorta Jesper Witzell Department of Silviculture, Swedish University of Agricultural Sciences, 901 83 Umeä, Sweden. E-mail: jesper.witzell@ssko.slu.se Abstract From 1990-1995, the formation and growth of stem cankers caused by Gremmeniella abietina (Lagerb.) Morelet on Pinus contorta Dougl. ex Loud. var. latifolia Engelm. was studied in three stands in northern Sweden. The stands were planted in 1976-1980. The total number of cankers on 756 trees that were individually followed increased from 233 to 477 during the five-year period. With 42.0% of the cankers, the pathogen entered through or from the base of diseased branches, and 33.6% through visually undamaged bark. Most of the cankers were within 100 cm of the ground. In one of the three areas, the cankers were evenly distributed within 180 cm of the ground. The frequency of cankers facing north exceeded those facing south. The average vertical length of cankers had increased, 55.6% of cankers had increased their percentual coverage of the stem girth; 13.8% had fully girdled the stem. Introduction In a large-scale study of damage to commercial plantings of P. contorta in northern Sweden 1987-1991, we found that the frequency of stem cankers caused by G. abietina increased during the seasons 1989-1991 (Karlman et al. 1994). North of latitude 64°N, 20% of the trees had cankers, of which 7% were severe (Karlman et al. 1992). The symptoms of G. abietina on P. contorta in northern Sweden are similar to those of the North American race of the fungus, with canker formation and produc tion of apothecia on trees 1-3 m high (Laflamme 1991). Investigations of genetic variation in G. abietina in northern Sweden recorded, however, only the Northern amplitype, identical to type B (Uotila 1983) or small tree type (Hellgren and Högberg 1995), of the European race in P. contorta (Hamelin et al. 1996, Hansson et al. 1996). The aim of this study was to describe the formation and growth of cankers caused by G. abietina on P. contorta in northern Sweden, focusing on canker devel opment on younger trees, 10-15 years old. Materials and methods From 1990-1995, the formation and growth of stem cankers caused by G. abietina was studied on P. contorta in the following localities of northern Sweden. In total, 51 756 trees were included in the study. 1. Grannäs (lat. 64°56'N, 510 m. a.5.1.). The area is located on an east-facing slope originally regenerated by P. abies. The stand forms a provenance test contain ing the most northern provenances available of P. contorta, the majority originating north of lat. 62°N in the Yukon Territory, and thus treated as a homogenous material. It was planted with P. contorta in 1976 after patch scarification. A total of 226 trees were included in the study. 2. Allejaur (lat. 65°54'N, 500 m. a.5.1.). The area is located on a slope with a slight east exposure that was originally regenerated by P. abies. It was planted with P. contorta, provenance Rusty creek (63°30'N) in 1980. A total of 385 trees were studied. 3. Akkajärvi (lat. 66°55'N, 425 m. a.5.1.). The area is located on a north ex posed slope originally regenerated by P. abies. It was planted with P. contorta, prov enance Rusty creek (63°30'N) in 1980. One-hundred forty-five trees were exam ined. Tree height was measured and the trees were examined each year for shoot blight and cankers caused by G. abietina. Every year, each canker was measured for vertical length and horizontal width. For each canker, the pathogen's entrance to the stem was noted. The aspect of each canker in terms of compass direction (N-E, E-S, S-W and W-N) was recorded. The occurrence of apothecia of G. abietina were also recorded. Results and discussion The frequency of trees with shoot blight caused by G. abietina increased in Allejaur and Akkajärvi from 1990 to 1995. However, both sites showed a slight decrease in the frequency of damaged trees during 1992 due to trees assigned as moderately (<50% crown damaged) or lightly damaged (a single branch damaged) in 1991 and assigned as recovered or lightly damaged in 1992 (Fig. 1). In Grannäs, the number of damaged trees was constant during the period. The frequency of severely damaged trees and trees killed by G. abietina increased at all three sites during the period (Fig. 1). The total number of cankers increased from 233 to 477 during the five-year period. The number of cankered stems increased in Allejaur and Akkajärvi, but was constant in Grannäs. However, the number of multi-cankered stems increased at all three sites during the period (Fig. 2). In Allejaur and Akkajärvi, 18% of the stems investigated had cankers in 1990 (Fig. 2). This was in accordance with our earlier investigations in northern Sweden, where the P. contorta plantations severely damaged by G. abietina in 1988 showed a high frequency of stem cankers in 1989 (Karlman 1989, Karlman et al. 1994). How ever, in Grannäs 36% of the investigated trees had cankers in 1990 (Fig. 2). This was probably a result of the somewhat older and higher stand in Grannäs compared to Allejaur and Akkajärvi. 52 Figure 1. The percentage of trees with shoot blight caused by Gremmeniella abietina during 1990-1994. Killed or dying trees are shown by black, severely damaged trees (>50% of crown damaged) by stripes, moderately damaged trees (25-50% crown damaged) by dots and lightly damaged trees (a single branch damaged) by white. Figure 2. Percentage of trees with stem cankers caused by Gremmeniella abietina during 1990- 1994. Trees with one canker are shown in white, those with two cankers by sparse dots, those with three cankers by stripes, those with four cankers by dense dots, and trees with five cankers are shown in black. 53 Figure 3. The percentage of cankers caused by Gremmeniella abietina in a) Grannäs, b) Allejaur and c) Akkajärvi, with respect to height above ground (cm) and year of formation. Cankers formed in 1990 are shown in black, cankers formed from 1991 to 1994 are shown in stripes, and cankers formed in 1995 are shown in white. In 1990, the average height of the four years older trees in Grannäs was almost twice that in Allejaur and Akkajärvi. During the winter of 1988, with a 1.5-1.6 m snow cover, twice as thick as average, in areas between latitudes 64°N and 66°N in northern Sweden (Karlman 1993), the trees in Grannäs were high enough to be bent over into the snow, providing highly favourable temperature and moisture condi tions for the fungus (cf. Laflamme 1991, Yokotaetal. 1974, Tanaka 1988, Marosy et al. 1989). 54 In Allejaur and Akkajärvi, 97.9% and 95.5% respectively of the cankers were formed within 100 cm of the ground (Fig. 3). In Grannäs, however, the cankers were evenly distributed between 180 cm of tree height and the ground (Fig. 3). Most cankers above 100 cm of tree height were formed in 1990 or earlier. Cankers formed between 1991-1995 were mostly situated below the height of normal snow depth. These results support Dorworth (1973), who found that 95% of Gremmeniella-in fected branches were within 0.8 m of tree height, and thus covered by snow during winter. Cankers were formed seven different ways (Table 1). Cankers appeared around the base of diseased branches, as also found by Dorworth (1973). On P. resinosa, he found that the fungus advanced along the branch approximately one internode per year. Once the fungus had entered the bole, a canker was formed. However, G. abietina can also infect through injuries resulting from snow bending the branch (Kurkela 1981). Cankers occur between the whorls of branches, and form around mechanical damage as well as on visibly undamaged bark (Table 1). According to Roll-Hansen and Roll-Hansen (1973) and Kurkela (1981), bark damage caused by frost and ice formation or by snow bending, may serve as an entrance for the pathogen directly into the bole. Furthermore, Kurkela (1984) suggests that visible damage to the bark is not always necessary for successful infection. Further experimentations are, how ever, necessary to clarify the importance of bark damage as a way of entrance for the pathogen. The frequency of cankers facing north (58.8%) exceeded those facing south (41.2%) (p<0.05). Most cankers were formed on the north side of the stem; 35.5% in the N-E sector (p<0.05). This was in accordance with Kurkela and Norokorpi (1979), who observed that formation of cankers was somewhat more frequent on the NW-E sectors compared to the SE-W sectors of the stems, but found that the position of the canker in terms of compass direction did not affect its size. Dorworth (1973) found that the frequency of infected branches facing north exceeded those facing south by approximately 25%. In 1995, apothecia by G. abietina were present in 64.6% of the cankers. Twenty five per cent of the cankers were formed in the final year of the study, why apothecia could not be recorded in those cankers. The high canker development in 1995 might be a result the cold and wet weather in northern Sweden during the autumn of 1993 (Anon. 1993, cf. Kaitera and Jalkanen 1992). For most cankers, however, fruit bod Table 1. Percentage of cankers with regard to different ways of entrance for the pathogen (values with different letters are significantly separated at 0.05 level) Entrance Percentage Infected branch 42.0 A Undamaged bark 33.6 A Other/unknown damage 21.2 B Vole damage 1.8 C Infected lateral stem 0.5 C Moose damage 0.5 C Mechanical damage 0.4 C 55 ies of G. abietina were recorded in affected branches adjacent to the cankers. No pycnidia was found in the cankers. The average vertical length of the cankers increased during the period. For cankers formed in 1991-1995, the average vertical length increment the first year after formation was 85.5 mm, and from year two 23.4 mm/year. The fast growth during the first year might be explained by limited chemical defence and absence of callus tissue in the bole during the first winter after infection. During the following years, the fungus must exceed the mechanical and chemical defence of the tree in order to expand (Blanchette 1992, Manion 1991, Shigo 1984). For 55.6% of the cankers, the average coverage of stem circumference in creased from the first measurement to that of 1995; 13.6% of the cankers had fully girdled the stem. For 44.4% of the cankers, the average coverage of stem circumfer ence decreased or was constant between the first measurement and 1995; 10.1% of the cankers were completely occluded in 1995. For 40% of the cankers, the average coverage decreased or was constant between the first measurement and that of 1995. Some of the cankers were completely occluded in 1995. Kurkela (1981) also re corded healing of cankers, but found it possible for the fungus to survive in the wood of eight-year-old, completely occluded cankers. There might be a risk that the patho gen will start to grow again in periods of weather conditions favourable for the fun gus. Occluded cankers can also hide severe damage to the wood, substantial resin exudation, fibre aberrations, bark pockets and compression wood, though some of the occluded cankers are hardly visible on the trunk surface (Karlman and Witzell 1991). Gremmeniella-damaged wood has a considerably higher basic density com pared with undamaged wood (Ahlqvist et ai. 1996). This can partly be explained by the very high extract content of the former, and by the occurrence of reaction wood. The use of trees with larger contents of Gremmeniella-damagQd wood leads to seri ous processing problems. Thus, such wood should be sorted out and classed as low grade raw material (Ahlqvist et ai. 1996). Acknowledgements Prof. Margareta Karlman initiated this study. Financial support was received from the Swedish Forest Research Foundation and the Swedish Environmental Protection Agency. References Ahlqvist, 8., Karlman. M. and Witzell, J. 1996. Gremmeniella-infected Pinus contorta as raw material in the production of kraft pulp. Eur. J. For. Path. 26: 113-121. Anon. 1993. Väder och vatten. Ärets väder 1993. Swedish Meteorological and Hydrological In stitute, Norrköping. 7 pp. (In Swedish.) 56 Blanchette, R. A. 1992. Anatomical Responses of Xyleme to Injury and Invasion by Fungi. In: Defence Mechanisms of Woody Plants Against Fungi. Ed. by Blanchette, R. A. and Biggs, A. R. Heidelberg: Springer-Verlag. pp. 76-95. Dorworth, C. E. 1973. Sequence of Formation and Resin Content of Scleroderris Canker. Can. J. For. Res. 3: 161-164. Hamelin, R. C., Lecours, N., Hansson, P., Hellgren, M. and Laflamme, G. 1996. Genetic differen tiation within the European race of Gremmeniella abietina. Mycol. Res. 100: 49-56. Hansson, P., Wang, X-R., Szmidt, A. E. and Karlman, M. 1996. RAPD variation in Gremmeniella abietina attacking Pinus sylvestris and Pinus contorta in northern Sweden. Eur. J. For. Path. 26:45-55. Hellgren, M. and Högberg, N. 1995. Ecotypic variation of Gremmeniella abietina in northern Europe: disease patterns reflected by DNA variation. Can. J. Bot. 73: 1531-1539. Kaitera, J. and Jalkanen, R. 1992. Disease history of Gremmeniella abietina in a Pinus sylvestris stand. Eur. J. For. Path. 22: 371-378. Karlman, M. 1989. Gremmeniella-ett orostecken. Skogen (2): 56-59. (In Swedish.) Karlman, M. 1993. The Gremmeniella disease situation on lodgepole pine in northern Sweden. In: Pinus contorta - from untamed forest to domesticated crop. Proceedings. lUFRO, 1992 Umeä. Ed. by Lindgren, D. Dept.of Forest Genetics and Plant Physiology, Swedish Univer sity of Agricultural Sciences, Umeä. Rapport 11: 335-349. Karlman, M., Hansson, P. and Witzell, J. 1994. Scleroderris canker on lodgepole pine introduced in northern Sweden. Can. J. For. Res. 24: 1948-1959. Karlman, M. and Witzell, J. 1991. Faran inte över för contortan. Inre skador döljs ofta. Skogen (12): 26-28. (In Swedish.) Karlman, M., Witzell, J. and Hansson, P. 1992. Skadeläget i praktiska kulturer med Pinus contorta inorra Sverige planterade 1974-81. Resultat frän ären 1987-91. Dept. of Silviculture, Swed. Univ. Agric. Sei, Umeä. Arbetsrapport nr 62. 58 pp. (In Swedish.) Kurkela, T. 1981. Canker and die back of Scots pine at precommercial stage caused by Gremmeniella abietina. Folia Forestalia 485, 12 pp. (In Finnish with English summary.) Kurkela, T. 1984. Factors affecting the development of disease epidemics by Gremmeniella abietina. In: Scleroderris Canker of Conifers; Proceedings of an International Symposium, Syracuse, N.Y., USA, June 21-24, 1983. Ed. by Manion, P. D. pp. 148-152. Kurkela, T. and Norokorpi, Y. 1979. Pathogenicity of Scleroderris lagerbergii, Lachnellula pini and L. flavovirens and their cankers on Scots pine. Commun. Inst. For. Fenn. 97(1): 1-16. Laflamme, G. 1991. Scleroderris canker on pine. Information Leaflet LFC 3. Forestry Canada, Quebec region, Sainte Foy, Quebec, Canada. 12 pp. Manion, P. D. 1991. Tree disease concepts. 2 ed. Englewood Cliffs, N.J. USA: Prentice-Hall, Inc. pp. 182-206. Marosy, M., Patton, R. F. and Upper, C. D. 1989. A conducive day concept to explain the effect of low temperature on the development of Scleroderris shoot blight. Phytopath. 79: 1293-1301. Roll-Hansen, F. and Roll-Hansen, H. 1973. Scleroderris lagerbergii in Norway. Hosts, distribu tion, perfect and imperfect state, and mode of attack. Medd. Norske Skogsforsoksvesen 30: 442-459. Shigo, A. L. 1984. Compartmentalization: A conceptual framework for understanding how trees grow and defend themselves. Ann. Rev. Phytopathol. 22: 189-214. Tanaka, K. 1988. The present status of Scleroderris canker of Sachalin-fir in Hokkaido, Northern Japan, and its control. In: Recent Research on Scleroderris canker of conifers. Proc. lUFRO Working Party 52.06-02 Ljubljana, Yugoslavia, September, 1986. pp 124-133. Uotila, A. 1983. Physiological and morphological variation among Finnish Gremmeniella abietina isolates. Commun. Inst. For. Fenn. 119. 12 pp. Yokota, S., Uozumi, T. and Matsuzaki, S. 1974. Scleroderris canker of Todo-fir in Hokkaido, Northern Japan. Eur. J. For. Path. 4: 65-74. 57 Needle size and needle nutrient contents of Scots pine after Gremmeniella abietina infection and green pruning Heikki Nuorteva, Timo Kurkela and Antti Uotila H. Nuorteva, T.Kurkela, Finnish Forest Research Institute, P.0.80x 18, 01301 Vantaa, Finland. E-mails: Heikki.Nuorteva@metla.fi, Timo.Kurkela@metla.fi A. Uotila, University of Helsinki, Hyytiälä Forestry Field Station, Hyytiäläntie 124, 35500 Korkeakoski, Finland. E-mail: auotila@hyytiala.helsinki.fi Abstract Scots pine trees may have several long-term foliar responses to defoliation caused by Gremmeniella abietina or green pruning. Two years after pruning treatments the needles were longer, heavier and contained significantly more nutrients (per individual fascicles) than the needles of the unpruned trees. Even 10 years after Gremmeniella infection needles of the affected trees were still slightly longer and contained more boron and calcium than the needles of the healthy control trees. On the other hand, the foliar contents of few other elements and dry weight of the individual fascicles were lower in the diseased trees than in the control trees. Introduction Scleroderris canker (Gremmeniella abietina (Lagerb.) Morelet) may rapidly reduce the size of the living crowns of Scots pine (Pinus sylvestris L.) by killing their shoots and the whole branches during the epidemics. When this type of defoliation concen trates on the lower parts of the crown of the trees, it may alter the chemical compo sition of the healthy needles which have grown after the needle loss in the remaining upper and green part of the crown (Nuorteva and Kurkela 1998). In our previous study (Nuorteva and Kurkela 1993) we have described altered nutrient concentra tions in healthy needles of Scots pine trees which had been seriously defoliated few years earlier by Gremmeniella abietina or artificial green pruning of the lower branches. In this study we report about the possible compensatory changes in the length and dry weight of those same upper-crown needles of the affected trees (Nuorteva and Kurkela 1993). We were also interested, whether there were any changes due to defoliation in actual amounts of needle nutrients in situ (eg. per needle or fascicle). 58 Material and methods The needle material (youngest needles, uppermost lateral top shoot) of this experi ment was collected during winter from six Scots pine stands in southern Finland. Four of the stands (tree age 25-30 years) had suffered from Gremmeniella abietina 5-10 years and in two healthy stands part of the trees had been pruned 1-2 years earlier (tree age 15-20 years). Twenty sample trees per stand (10 diseased or pruned and 10 control trees with an unaffected crown) were selected based on the length of the living crown. The crown ratio (length of the green crown / tree height) of the diseased or pruned trees were on average over a half smaller than in the control trees. The length and dry weight of 100 needle fascicles per each sample shoot were measured (a total of 12 000 fascicles). In addition, we calculated the actual needle nutrient contents per individual needle fascicles based on the earlier analyses of foliar nutrient concentrations (for more detailed description of the sample trees and the chemical analyses see Nuorteva and Kurkela 1993). Results and discussion The needle length was slightly increased both in the diseased (+7%) and pruned trees (+24%) as compared to the needles of the adjacent control trees with an unaf fected crown. The dry weight of the individual needle fascicles in the pruned trees were also higher (over 20%) than those of the unpruned trees. However, in the dis eased trees, the dry weight of the fascicles were still lower (-10%) than in the control trees in spite of slightly longer needles of the affected trees. The amounts of foliar boron and calcium per needle fascicles were increased both in the diseased and pruned trees. Other element contents (per fascicle) were in the diseased trees mainly lower (Fe, Mg, K, C and H) and in the pruned trees higher (N, S, P, K, Mn, Cu, Zn, Na, C and H) than in the healthy undefoliated trees. According to this experiment, reduction in the living crown by Gremmeniella abietina or green pruning induces long-term changes in the needle length and weight of recovering Scots pine trees. Further, we noticed significant changes in the amount of nutrients per individual needles (=nutrient content), and that these changes were larger, expressed in percentages, than those of the concentrations in the foliage. References Nuorteva, H. and Kurkela, T. 1993. Effects of crown reduction on needle nutrient status of scleroderris-canker-diseased and green-pruned Scots pine. Can. J. For. Res. 23: 1169-1178. Nuorteva, H., Kurkela, T. and Lehto, A. 1998. Rapid living crown reduction caused by Gremmeniella abietina affects foliar nutrient concentrations of Scots pine. Eur. J. For. Path. 28: 349-360. 59 The genomes of dsRNA viruses of Gremmeniella abietina Tero Tuomivirta Finnish Forest Research Institute, PL 18, 01301 Vantaa, Finland. Abstract Two fairly different, independently occurring, virus particles were found from plant pathogenic fungus Gremmeniella abietina. GaRV-L particle had one -6000 by long dsRNA molecule and GaRV-MS particle had three dsRNA molecules (-2000 bp, -1900 by and -1300 bp) in its genome. The sequence of GaRV-L 1 was determined and analysed. GaRV-L 1 belongs to the Totiviridea family according its nucleic acid sequence. The genome of GaRV-L 1 contained two large open reading frames which code for putative coat protein and putative RNA dependent RNA polymerase. GaRVMS belongs to the Partitiviridae family. The viruses discussed here seem not have any effect on the pathogenicity of their hosts. 60 Pine dieback by Sphaeropsis sapinea in Northern and Central Italy Giorgio Maresi, Paolo Ambrosi, Andrea Battisti, Paolo Capretti, Roberto Danti, Elisabetta Feci, Stefano Minerbi and Stefania Tegli G. Maresi, P. Ambrosi, UO Foreste - Istituto Agrario (lASMA), S. Michele a/Adige (Trento), Italy. E-mail: giorgio.maresi@ismaa.it A. Battisti, P. Capretti, E. Feci, S. Tegli, DIBA - Dipartimento di Biotecnologie Agrarie, Universitä - Firenze, Italy. R. Danti, IPAF - Istituto per la Patologia degli alberi forestali, CNR, Firenze, Italy. S. Minerbi, Ripartizione Foreste - Provincia Autonoma di Bolzano, Italy. Abstract Pine dieback, shoot blight and needle disease associated with Sphaeropsis sapinea are quite com mon symptoms on Pinus spp. in both natural stands and plantations, after a relatively mild winter followed by a diy summer. During the last few years damage was quite severe in Northern and Central Italy, particularly on Pinus nigra but also on P. halepensis, P. pinea, and P. sylvestris. Fungal isolates were analysed with both morphological and molecular methods and were assigned to the type A. The role of an insect, the cone bug Gastrodes grossipes, as epidemiological factor in the spreading of conidia, was evidenced. Weather patterns and site conditions seemed the most effective factors in enhancing the symptom development. The fungus was quite common in pine forest and stands and it can be considered as a bio-indicator of natural decline of pine plantations which are at the end of their "pioneer role". Keywords: Sphaeropsis sapinea, Gastrodes grossipes, Pinus, fungal population, epidemiology Introduction Sphaeropsis sapinea (Fr.) Dyko & Sutton is a parasitic fungus known and studied in Italy since the early '9OO (Petri 1913, Petri et Adani 1916, Goidanich 1933). It is considered responsible of many symptoms and may affect different parts of the plant. The fungus has been reported on various host species mainly from the genus Pinus but also on Pseudotsuga and, recently, Cupressus (Frisullo et al. 1997/ The most affected hosts are Pinus nigra (Fig. 1) and Pinus pinea, the latter especially at cone level (Capretti 1956, Maresi et al. 1999, Vagniluca et al. 1995, Tainter and Baker 1996). During the last century, considering the occurrence and damage by S. sapinea on different host species and environmental conditions in Italy, many attempts were done in order to differentiate the fungus and "new species" or varieties were re- 61 ported. The organism was described as: - Sphaeropsis ellisii Sacc. (Petri 1913) - Sphaeropsis necatrix Petri and Adani (1916) - Sphaeropsis ellisii var. cromogena (Goidanich 1933). Later all these "species and varieties" were included in S. sapinea (Fr.) Dyko & Sutton. Leonello Petri (1913) described a disease of shoots on young trees of Pseudotsuga douglasii planted nearby Pinus sylvestris affected by the fungus. As the conidial morphology and size did not correspond to those reported on pines the or ganism was described as a "new species": Sphaeropsis ellisii Sacc. (Petri 1913). A different causal agent was considered responsible of Pinus pinea cones' disease, the so-called "Malattia delle pine pagliose" Sphaeropsis necatrix (Petri et Adani 1916). The fungus was associated with intense necrosis on immature cones and seed abortion, with economic damage due to the commercial use of seeds. Petri et Adani (1916) judged S. necatrix different from S. ellisii mainly because of co nidial morphology. One of the fungus' peculiarities (to cause blue stain of colonised wood) in duced Gabriele Goidanich (1933) to describe a new variety as Sphaeropsis ellisii var. cromogena. The fungus produces old hyphae with pigmented substances and colonises both medullar rays and tracheids. Recently, as a consequence of heavy damage, particularly in northern Italy (Maresi et al. 1999), the main epidemiological factors such as fungal characters, insect association and environmental conditions related to S. sapinea disease have been evaluated. Materials and methods i) Studies on the fungal pathogen have been focused on population variability de tected by both morphological and molecular methods. The fungus was collected from different hosts species and localities. Conidia were checked according to the Figure 1. Pycnidia of Sphaeropsis sapinea on a cone of Pinus nigra. 62 criteria suggested by Wang et al. (1986) and Palmer et al. (1987) with both optical and scanning electronic microscope (SEM). DNA profiles from a selection of 9 Italian isolates collected mainly from Pinus nigra, P. sylvestris and P. pinea have also been compared utilising RAPD markers produced by DS9, DSIO and DSI9 primers following Smith and Stanosz (1995). Electrophoretic patterns have been compared with those of typical referenced A and B isolates kindly provided by G Stanosz. ii) As regard to the dissemination of the conidia, the possibility that the fungus colo nises the seed cone of the pine hosts (Vujanovic et al. 2000), producing a consider able quantity of pycnidia, suggested to assess the presence of the fungus on P. nigra cones and evaluate the role of cone insects as potential dissemination agents in a stand of P. nigra of Monte Morello, Firenze, Italy. The cone bug Gastrodes grossipes (De Geer) (Heteroptera Lygaeidae) (Fig. 2) was the most frequent insect species and appeared to be a good candidate. Four hundred open cones were collected from 40 trees and inspected for the presence of both the fungus and the insect in spring 2000. The presence of the fungal conidia on the insect body was ascertained by a washing technique already established for a cone bug attacking cypress cones and dissemi nating fungal conidia (Battisti et al. 1999). Each living insect had been washed in 400 ml sterile water with 1% detergent (Tween 80), shaking for 1 min at 40 Hz. The presence of conidia in the suspension was assessed at the microscope on a slide carrying 10 ml of the suspension. An insect was considered free of conidia when five replicates were negative. iii) Forty-one P. nigra plots, all severely infected by S. sapinea in 1998 in Trentino (Northern Italy), have been considered to analyse site characters. Data were obtained from Woodland Management Plans in which woodland property was subdivided in homogeneous compartments: for each of them about 120 different environmental parameters were recorded every ten years following standards adopted in all the Figure 2. Gastrodes grossipes on Pinus nigra cone scales. 63 Trentino woods. In this context, the following characters were considered: facing, morphology, slope, soil type, soil depth and soil humidity (Maresi et al. 1999). Weather patterns were obtained from temperature and rainfall data collected at S. Michele all'Adige: the monthly mean of temperature and total monthly rainfall of the period 1997-2000 were compared with the mean of the period 1983-1995. Results and discussion i) S. sapinea isolates have been collected from different localities and host species, mostly from P. nigra but also from P. halepensis, P. pinea, P. sylvestris; characteris tics of the fungal isolates are reported in Table 1. Conidia of these samples varied within 34-41 x 10-15 (am in size, their surface at SEM observation was smooth while the mycelium was white to grey-green on PDA media. When compared with molecular makers all the isolates tested were assigned to the type A by the presence of characteristic bands (respectively 600 bp with DS9, 1600 bp DslO and 350 bp with DSI9 RAPDs primers). ii) The cones collected in the stand of Monte Morello were highly infected with both the fungus and the insect (Table 2). G. grossipes is a common insect which inhabits mature and open cones of pines but feeds on needles and shoots. The association between the cone bug G grossipes and the pathogenic fungus S. sapinea in cones of P. nigra appears to be occasional, as the insect exploits the cones only for shelter during its whole development. A possible advantage for the insect would follow from the extended persistence of the fungus-infected cones on the branches of P. nigra trees, which has been observed in sampled stands. The high frequency of in sects loaded with conidia of S. sapinea may suggest to consider their potential role as fungus' dissemination agents. Nymphs can spread the fungus at tree level from the cones to the pine needles where they feed, whereas adults after exploiting open Table 1. Selection of Sphaeropsis sapinea isolates utilised for fungal characterisation. ' Smith and Stanosz, 1995 Site location and Latitude and longitude Host species Length x width Type 1 province (mean fim) Prissiano (Bolzano) 46°.33' -1 l°.ll' Pinus sylvestris 38.7 x 12.4 A Valternigo (Trento) ON O 0 1 o O P. sylvestris 40.2 x 12.8 A Alassio (Savona) 44°.11' - 8°.10' P. halepensis 34.7 x 10.8 A Monte Senario (Firenze) 43°.54' - 11°.20' P. nigra 41.2 x 14.7 A Artimino (Firenze) 43°.47' - 11°.22' P. nigra 38.3 x 14.4 A Firenze 43°.46' - 11°.15' P. nigra 34.2x 11.1 A S. Rossore (Pisa) 43°.43' - 10°.20' P. pinea 34.2 x 14.8 A S. Severino (Macerata) 43°.14'- 13°.ll' P. nigra 40.6 x 14.4 A S. Cataldo (Lecce) 40°.23' - 18°.18' P. halepensis 39.1 x 11.2 A 64 cones can be responsible for the dissemination of the conidia among trees. Adults of Gastrodes are able to fly over distances of several kilometres in alpine valleys, reach ing also the upper timberline (Nägeli 1934). These findings may suggest to include the insect-fungus association in the management of the disease in the stands of P. nigra. iii) Although the fungus is quite common in Italy, the occurrence of heavy damage and death of adult trees by S. sapinea can be considered as a indicator of pine de cline. This was particularly undertaken after field observation in Trentino (Northern Italy). Here the disease was more intense on sunny slopes, poor soil and dry condi tions. Most of the damage (76.0%) was registered on southern and western facing slopes; 83.0% on superficial soils and 98.0% on well drained and dry soils. As showed in Table 3 all the damaged stands vegetated in site where drought problem can be easily enhanced by morphological and soil characteristics. Analysis of meteorological data emphasised a heavy reduction of rainfall in February and March both in the years 1997 and 1998 when compared with the pe riod 1983-1995. Adrought period was also recorded in the autumn of 1997. In the same year the temperature from January to March was higher than in the reference period (Figs. 3, 4). Following these climatic conditions, an evident and dramatic increase of damage was recorded in 1998. However in 1999 and 2000 the damage was lower probably due to an increase of rainfall and lower temperature. Table 2. Results of the analysis of 400 Pinus nigra mature cones from the stand of Monte Morello Firenze, Italy. Table 3. Characters of 41 damaged stands in Trentino as obtained from Woodland Management Plan following the standard classification adopted by the Forest Service of the Province. Cones with the fungus Cones with insects Cones with fungus and insects Insects loaded with the fungus 41.7% 42.9% 24.4% 58.2% Facing 22% N; 2% E; 39% S; 37% W Morphology 30% bottom of valley; 48% slope; 7% terrace; 15% edge Slope 15% flat; 80% steep; 5% very steep Soil type 58% brown soil; 42% rendzina soil Soil deep 83% superficial; 17% medium deep Soil humidity 55% arid; 43 % dry; 2% wet 65 Figure 3. Mean of monthly precipitation data for the periods 1983-1995,1997 from S. Michele a 1 Adige (Trento). Figure 4. Monthly average of temperature of S. Michele a/Adige (Trento) for the period 1983- 1995 and 1997. Conclusion S. sapinea is a quite common fungus in pine forest in Italy. It can be easily detected on cones even when there is no visible damage on shoots. P. nigra was the most affected tree species especially in the North of the country, but P. sylvestris was also badly affected in South Tyrol and P. halepensis in Southern Italy (Danti and Capretti 1997). In these stands the disease was effective in changing the wood composition and in leading the evolution of the stand towards a mixed wood in which Pinus species are absent or secondary (Ambrosi et al. 2000). 66 Water stress may pre-dispose trees to the attack of the fungus at crown level and tree mortality can be often observed. In this context, Blodgett et al. (1996) and Paoletti et al. (2001) after inoculation tests showed that the damage was higher on water stressed trees of P. halepensis and other Pinus species. All the randomly collected isolates analysed in this study belong to the morphotype "A" which is considered the more pathogenic (Smith and Stanosz 1995). The differences observed in conidial size suggest that a certain variability among the isolates may exists. A contribution to the epidemiology of the fungus was given by the observation on the role of G. grossipes in dissemination of conidia. References Ambrosi, P., Minerbi, S. and Maresi, G. 2000. II deperimento di alcune pinete in Trentino-Alto Adige: fattori biotici ed abiotici coinvolti. Atti del II Congresso SISEF "Applicazioni e prospettive per la ricerca forestale italiana" Bologna, 20-22 ottobre 1999 (Bucci G., Minotta G. and Borghetti, M. ed.): 441-446. Battisti, A., Roques, A., Colombari, F., Frigimelica, G. andGuido, M. 1999. Efficient transmission of an introduced pathogen via an ancient insect-fungus association. Naturwissenschaften, 86: 479-483. Blodgett, J. T., Kruger, E. L. and Stanosz, G. R. 1996. Effects of moferate water stress on disease development by Sphaeropsis sapinea on red pine. Phytopathology, 87: 422-428 Capretti, C. 1956. Diplodia pinea (Desm.) Kickx agente del disseccamento di varie specie del gen. Pinus e di altre conifere. Annali Accademia Italiana di Scienze Forestali, 5: 171-202. Danti, R. and Capretti, P. 1997. Shoot blight of Pinus halepensis Mill. In the Italian peninsula. In lufro WP 7.02.02 meeting, "Foliage, shoot,stem diseases" Quebec City, May 25-31, 1997. Ed. by Laflamme, G., Berube, J.A. and Hamelin, R. C. Quebec city: Natural resources Canada, Canadian Forest Service, Laurentian Forestry Centre: 103-107. Frisullo, S., Bruno, G., Lops, F. and Sparapano, L. 1997. Un nuovo agente di cancro del cipresso in Italia. Petria7: 141-158. Goidänich, G. 1936. Le alterazione cromatiche parassitarie del legname in Italia. Bollettino della R. Stazione di Patologia Vegetale 15: 442-470. Maresi, G., Ambrosi, P., Confalonieri, M. and Capretti, P. 1999. Disseccamenti da Cenangium ferruginosum e Sphaeropsis sapinea nelle pinete trentine. Monti e Boschi L, 2: 35-41. Nägeli, W. 1934. Ueber Biologie und Verbreitung der beiden Langwanzen Gastrodes abietum Bergr. und Gastrodes grossipes De Geer. Mitt. Schweiz. Aust. Forstl. Vers., 18: 193-280. Palmer, M.A., Stewart, E.L. and Wingfield, M.J. 1987. Variation among isolates of Sphaeropsis sapinea in the North Central United States. Phytopathology, 77: 944-948. Paoletti, E., Danti, R. and Strati, S. 2001. Pre and post inoculation water stress affects Sphaeropsis sapinea canker lenght in Pinus halepensis seedling. Forest Pathology 31: 209-218. Petri, L., 1913. Disseccamenti dei rametti di Pseudotsuga douglasii Carr. prodotto da una varietä di Sphaeropsis ellisii Sacc. Annales Mycologici, 11,3: 278-280. Petri, L. and Adani A. 1916. Malattia dei coni del" Pinus pinea". Ann. R. Acc. Agric. Torino, 59: 1-23. Smith, D.R. and Stanosz, G.R. 1995. Confirmation of two distinct population of Sphaeropsis sapinea by in the North Central United States Using RAPDs. Phytopathology, 85: 699-704. Tainter, F.H., and Baker, F.A. 1996. Principles of Forest Pathology. John Wiley & Sons, Inc. NY, 805 pp. 67 Vagniluca, S., Goggioli, V. and Capretti, P. 1995. Cankers and shoot blights of Pinus pinea in Italy. In: Shoot and Foliage diseases in Forest Trees. lUFRO Meeting Vallombrosa, Firenze, Italy, June 6-11, 1994: 284-286. Vujanovic, V., Arnaud, M. and Neumann, P.J. 2000. Susceptibility of cones and seeds to fungal infection in a pine (Pinus spp.) collection. Forest Pathology, 30: 305-320. Wang, C.G., Blanchette, R.A. and Palmer, M.A. 1986. Ultrastructural aspects of the conidium cell wall of Sphaeropsis sapinea. Mycologia, 78: 960-963. 68 Observations of Sirococcus conigenus and its pathogenicity Wang Tian Fu and Antti Uotila University of Helsinki, Hyytiälä Forestry Field Station, Hyytiäläntie 124, 35500 Korkeakoski, Finland. E-mail: antti.uotila@helsinki.fi Abstract The occurrence of Sirococcus conigenus on spruce was observed in Southern Finland. In addition the pathogenicity tests were established on spruce and pine seedlings or pine germlings. Sixty samples of the fungus were collected mainly from spruce cones. The fungus was also found on spruce shoots earlier infected by Chrysomyxa abietis or wounded by frost. The fungus was iso lated from pine germlings growing below spruces. When pine seedlings were grown in petri dishes the point inoculation of conidia produced 36% infection. In spruce seedlings the inoculation pro duced 14% infection. It seemed that Sirococcus is quite weak pathogen. Special conditions are needed for infections. Introduction Sirococcus conigenus (DC.)P. Cannon&Minter, syn. S. strobilinus Preuss, (Desm.) Petrak, Ascochytapiniperda Lindau, A. parasitica (Hart.) Rostr., was first reported by Hartig (Hartig 1894). It could cause shoot blight to spruce (Kujala 1950, Suther land et al. 1981, Halmschlager et al. 1998), seed-borne disease (Motta et al. 1996), and damping off and dieback on larch (Motta et al. 1994, Harrison 1997). This spe cies occurs widely in Europe and parts of Canada and USA. In Britain it has been recorded on cones of lodgepole pine (Ellis and Ellis 1985). Conidia are fusiform, two celled, 12-16x3 jxm (Ellis and Ellis 1985), 12-15 x 3.5 (J.m (Kujala 1950) in dimension. Mature pycnidia, dark-brown to black, usually multilocular appearing on the dead bark of shoots, needles or cones. Hyphae are colourless or whitish, sparsely forked. S. conigenus can infect young seedlings in the early summer at the base or the middle of the current shoots, where the needles turn brown and shrank, and soon fall. The disease tends to spread upwards from lower needles towards the ends of the shoots and some movement may also take place downwards into the shoots of previ ous year's growth, leading to dealth of some of side shoots (Hartig 1894). For a time the shoots tips remain green, but later wilt, droop and finally die, often curling over and assuming the shepherds'crooks (Funk 1972). Most cases spring or autumn frosts can significantly promote the fungus to infect shoots (Hartig 1894). Temperature and humidity are important factors to the infection. A daily temperature cycle of 16- 69 21° C could create the great infection (Wall and Magasi 1976). The aim of this study was to introduce some basic characteristics of Sirococcus conigenus. The pathogenicity of the fungus was studied by doing artificial inocula tions. Materials and methods Isolates The fungus was isolated from symptomatic shoots, seedlings and spruce cones. In summers 1996 and 1997 over 60 samples of this fungus on diseased seedlings, spruce cones and frosted shoots were collected in southern part of Finland (Table 1). Tablel. List of collected samples of Sirococcus conigenus. Life cycle of Sirococcus conigenus in Vitro On June 10 th , 1997, the single spore isolation method (Korhonen and Hintikka 1980) was used to isolate conidium of Sirococcus conigenus to MA petri dish. Eight Petri dishes were left in the growth chamber (Termaks 5260) of light (Airam 15W, 35-X) at 15° C, and another 8 petri dishes were stored in the growth chamber in darkness at the same temperature. Observation was carried out every second day. The infection experiment of Sirococcus conigenus in Vitro The origin of pine seed (Pinus sylvestris) was seed orchard No. 246 (Tammela, Äijoenniitty) and it was collected in the autumn of 1992. Spruce seeds (Picea abies) were from the seed orchard No. 179 (Inkoo, Svartbäck) collected at same time. The seeds were stored in the cold storage at 5°C, in the dark. In vitro tested seed germi nation was 95% of pine and 96% of spruce. Location Geographic coordinates Collection time Host Samples Juupajoki 24°17'E and 61°50'N June 5-7, 1996, July 1997, July 1998 Spruce cones and shoots, pine germlings 30 Kuru 23°30'E and 61°45'N June 13, 1996 Spruce cones 2 Orivesi 24°00'E and 61°30'N June 12, 1996 Spruce cones and shoots 8 Orivesi 24°19'E and 61°48'N July 1, 1997 Pine germlings, spruce cones 8 Ruovesi 24°15'E and 61°50'N July 1, 1996 Spruce shoots and cones 10 Savonlinna 29°30'E and 61°50'N July, 1997 Spruce cones 2 70 Pine seeds were sown in 10 Petri dishes of malt agar on May 19 th in the trans fer chamber, which have been disinfected by 30% of H 2 0 2 for 30 minutes, then rinsed 3 times by using sterile water and placed the seeds on sterilised filter paper to get rid of water. In each petri dish 8 pine seeds were sown under aseptic condition. Then Petri dishes were wrapped by polyethylene and stored in the growth chamber at 15° C in light. The inoculation was carried out on June 17 th , 5 petri dishes for inoculation and another five for control. The average germination was 88%, and no seedborn damp ing off was detected. Following stage we designed a point inoculation with those seedlings in the petri dishes. The inoculum was prepared by increasing sterilised water on petri dish where the fungus has produced already pycnidia and conidia. The suspension was mixed with sterile glas bar. A small drop of suspension was transferred to needles or stem of the seedlings growing in Petri dishes. The suspension was transferred so that it was not dropped to the agar. After inoculation Petri dishes were kept in growth cabin at 15° C and light. Inoculation onto spruce seedlings On June 12 th , 1997,480 spruce seeds disinfected by H2 O 2 30% and 480 undisinfected spruce seeds were sowed in 24 containers (<))l3cm x 10cm) filled with peat (TURVE BIOLAN.AV). The initial idea was to test whether this spruce had seed-borne dis ease, then they could be inoculated as tested objects by using inoculum Sirococcus conidial suspension. After six week growth 60 seedlings were treated as irrigating distilled water; 120 seedlings were sprayed with conidial suspension and left in the open area (where the day temperature changed naturally), and 60 seedlings inoculated by same co nidial suspension were grown in the greenhouse (the average temperature was around 26° C). Before inoculation, on July 17 th the seedlings were irrigated. To ensure the moisture was high enough the containers were put in plastic bags one week after inoculation. The conidial suspension was produced by adding distilled water into Sirococcus culture in Erlenmeyer flask and shook the bottle until the liquid became turbid, then transferred the suspension (160 000 conidia/ml) into houseplant atom izer. In each plastic bag packed one container of seedlings, sprayed the suspension onto seedlings until liquid drops existed on needles, then sealed the plastic bags immediately with rubber strings and kept them in such condition for one week, and then took off the plastic bags and incubated the seedlings as normal treatments. The fresh, surface sterilised spruce cones were inoculated with S. conigenus conidia. The cones were incubated in plastic boxes in light and 15° C. 71 Figure 1. Sirococcus conigenus black pycnidia formed on a dead Norway spruce shoot, cone scales and aphid gall. Table 2. The dimensions of conidia, hyphae and pycnidia. a : 150 conidia obtained from the pycnidia formed on aphid galls of Norway spruce shoot. b : 100 hyphae were measured from a single strain growing in malt agar. 0 : 150 pycnidia obtained from aphid galls of Norway spruce shoots. Figure 2. Conidia of Sirococcus conigenus (x 400 ) stained with anilin blue. Conidia" Length (nm) Width (|im) Hyphae b Diameter ((im) Pycnidia 0 Diameter (mm) Mean 11.9± 1.7 4.4 ± 07 5.3 ± 1.2 0.3 ±0.1 Max. 15 5 10 0.5 Min. 7.5 2.5 2.5 0.1 72 Results Characteristics of the fungus Hyphae were long and sparsely forked, colourless, thickness 5 pm. Pycnidia was dark-brown to black on spruce cones or shoots, 0.1 to 0.5 mm in dimension (Fig. 1). In wet conditions ripened pycnidia opened irregular exits and released spores. Co nidia, colourless, 2-celled, 12 x 3.3 pm, spindle-like or fusiform (Fig. 2, Table 2.) Conidial production of the fungus in Vitro After five days incubation hypha grew in petri dishes in light and darkness. After one week hypha composed mycelia in dark and light. It seemed that light had no influence to the growth of S. conigenus. However two days later some mycelia in light chamber started to form pycnidia and released conidia. But in dark chamber pycnidia were not developed and no conidia were formed. The fungus could com plete its asexual life cycle within two weeks in Vitro (light, 15° C). A micro-point inoculation to pine seedlings in Vitro Eight days after inoculation the first symptoms appeared. Firstly diseased needles or stems shrank and wilted. The infected tissues started to rot, the shoot tops were bending, and finally the seedlings died. Whitish and woolly mycelia were growing on the diseased tissues and spreading towards the healthy tissues. Soon brown pycnidia of S. conigenus emerged on the diseased seedlings and conidia were formed on the pycnidia (Fig. 3). It takes one week time for the fungus to kill a germling. The mean infection rate was 36%. No infections were in control seedlings. Inoculation to spruce seedlings and cones After one week incubation symptoms appeared. On needle tips or in the middle of needles appeared brown points, extended towards stems, then the tissues of the stem and needles wilted and died. But the infection frequency was not high, the average infection was only 14%. In the greenhouse the temperature during the incubation was 22-32° C (accordance with the daily data recording in Hyytiälä Forestry Field Station of Helsinki University), the fungal infection was 13.5%, however the natural temperature at same moment was concentrated at 14-26° C, and the infection was 14.6% (Table 3). Two weeks after inoculation pycnidia were developed in all inoculated spruce cones (Fig. 4). One week later the conidial mass was liberated from pycnidia. 73 Table 3. The number of diseased seedlings in experiments. The symptoms of control seedlings were caused by unknown factors. Figure 3. Disease symptoms on Scots pine seedlings in the point inoculation experiment. Note the woolly mycelia and pycnidia of S. conigenus on the needles and stems. Figure 4. Pycnidia of S. conigenus on spruce cones. Left: Black pycnidia occurred on scales of the cone. Right: Uninoculated control. Group Mean-extreme Seedlings Symptomatic temperature (°C) % Control 14-26 134 4.5 Open area 14-26 303 14.6 Greenhouse 22-32 120 13.5 74 Discussion Sirococcus conigenus is common fungus on spruce in Finland (Kujala 1950). On Norway spruce it was found on shoots, cones and aphid galls. On pine it killed germlings in shady conditions. The fungus could cause seed born damping off (Motta et al. 1993), but this was not detected in one tested seedlot in this study. The infec tion seem to be successful in moist and shady conditions. In Austria the fungus has caused damage in foggy zone spruce forest (Halmschlager et al. 1998), which sup port the importance of moisture as predisposing factor. It seemed that S. conigenus is weak pathogen and predisposing factors as moisture, shading, frost, nutrient imbal ances or infection of Chrysomyxa abietis are needed for successful infections on spruce shoots. In Austria S. conigenus destroys the spruce stands suffering magne sium deficiency (Halmschlager et al. 1998). In spruce cones the fungus was present regularly and in these infections no special predisposing is needed. The dimensions of conidia were almost same size as mentioned in literature (Ellis and Ellis 1985, Kujala 1950). Kujala (1950) collected the fungus also from Pseudotsuga menziesii and Abies concolor. The possible genetic variation of the fungus is unknown. Propably intra spe cies genetic variation is detected if the isolates from different geographic origins are compared. This kind of research could be important with this pathogen due to broad distribution area and several host species in Europe and North America. References Ellis, M. B. and Ellis, J. P. 1985. Microfiingi on land plants: An identification handbook. Croom Helm Ltd. 180 p. Funk. A. 1972. Sirococcus shoot-blight of Western hemlock in British Columbia and Alaska. PI. Dis. Reptr. 56: 645-647. Halmschlager, E., Neumuller, A. and Anglberger, H. 1998. Sirococcus conigenus - a fungus contributing to shoot dieback and top dying of Norway spruce in Austria. Proceedings of the "7th International congress of Plant Pathology", August 08-16, 1998, Edinburgh, Scotland. Hartig, R. 1894, Text-book of Disease of Trees, English ed, trans L.W. Someville, revised and edited H. Marshall Ward, Macmillan, London. Korhonen, K. and Hintikka, V. 1980. Simple isolation and inoculation methods for fungal culture. Karstenia 20: 19-22. Kujala, V. 1950. Über die kleinpilze der Koniferen in Finnland. Communicationes Instituti Forestalis Fenniae 38(4): 1-121. Motta, E., Annesi, T. and Balmas, V. 1996. Seedborne fungi in Norway spruce: testing methods and pathogen control by seed dressing. Eur. J. For. Path. 26: 307-314. Sutherland, J. R., Lock, W and Farris, S. H. 1981. Sirococcus blight: a seed disease of container grown spruce seedlings in Coastal British Columbia forest nurseries. Can. J. Bot. 59: 559- 562. Wall, R. E. and Magasi, L. P. 1976. Environmental factors affecting Sirococcus shoot blight of black spruce. Can. J. For. Res. 6: 448-452. 75 Effect of growth stage and microclimate on Botrytis cinerea outbreak in Picea abies container seedlings: Preliminary results Anu Pulkkinen, Juha Heiskanen, Risto Rikala and Raija-Liisa Petäistö A. Pulkkinen, Pitkäkoskentie 12, 75500 Nurmes, Finland. J. Heiskanen, R. Rikala, R. Petäistö, Finnish Forest Research Institute, Suonenjoki Research Station, 77600 Suonenjoki, Finland. E-mail: raija-liisa.petaisto@metla.fi Introduction Grey mold (Botrytis cinerea) is a very common fungus, which can have saprofytic, pathogenic or endofytic character. This fungus causes damages also in forest tree nurseries especially on spruce (Picea abies) and birch (Betula pendula). Most of the spruce seedlings in Finnish nurseries are grown in containers in greenhouses. This growing method seems to favour the outbreak of the grey mold disease compared to the growing of bareroot seedlings outdoors. The microclimate inside container seed ling canopy is supposed to be the main reason. The aim of our work was to study possible changes in the resistance of P. Abies seedlings to B. cinerea and in microlimate factors within seedling canopy during the growing season and interactions between them. The final purpose was to find possi bilities to predict the outbreak of the disease and to control it. Experiments P. Abies seedlings were sown in spring into containers filled with light peat growth medium. They were for two greenhouse experiments (Exp. 1 and 2) and for a growth chamber experiment with simulated growing season (Exp. 3). Sensors for tempera ture, relative humidity (RH), surface moisture and light intensity were set to meas ure conditions near the surface of the growth medium in containers (Fig. 1). Daily mean temperature was commonly within 10 and 25° C, relative humidity (RH) above 60% and vapour pressure deficit (VPD) within 0.2 and 0.95 kPa during the growing season. Temperatures below 10° C, RH above 85-90% and VPD below 0.2 kPa were recorded from the beginning of September. Inoculations with B. cinerea were made on the seedlings when they were 1,2, 3,4, sor 6 month-old. Conidia in water suspension (conidia density 1 x 10 6 /ml, 3.2 ml/seedling) were sprayed on the seedlings. Four seedling trays (with 36 inoculated and 36 uninoculated seedlings in each tray) were used on each inoculation time in 76 each experiment. In Exp. 3, seedlings were put into a dew chamber after inocula tions (VPD close to zero). In Exp. 1, seedlings stayed in the greenhouse and in Exp. 2 seedlings were moved into a small plastic house, where the surface of seedlings was kept continuously wet by using atomizer nozzles. The seedlings were checked daily during the 11-12 day-period after inoculations to find symptoms caused by the mold. Figure 1. Measurement sensors used in the experiments. Results Microclimate was affected by the growth of the seedlings and especially by the clo sure of the canopy causing e.g. decreasing saturation deficit (data not shown). The first symptoms in the seedling needles were small spots with color changes. In the new needles, the whole needle might change color and decay, while in the older needles a clear dead strip in the middle part of the needle was common (Fig. 2). The speed of disease outbreak was measured with the duration. The effect of seedling growth stage and microclimate on the disease outbreak was considered using the duration when 50% of seedlings was diseased (days to 50% diseased = DD J0 ) (Fig. 3, Fig. 4). In the beginning of the growing season the disease broke out slowly, DD50 was more than 7 days in all the three experiments. Nearly the same situation was at the end of the growing season in Exp. 1 and 2 but not in Exp. 3. The conditions in the dew chamber (Exp. 3) were rather optimal for the disease outbreak, e.g. the satura tion deficit was continuously very low. So, it seemed that at the end of the growing 77 season conditions close to optimal are needed for the disease outbreak. In the middle of the growing season (2 to 5 months) the seedlings got disease more quikly, DD 50 was only 1-3 days, except in the inoculation time 3 in Exp. 1. An additional experi ment in the dew chamber showed clear symtoms in 3-month-old seedlings after only a couple of hours from the inoculation. Further experiments and data handling are to be carried out concerning the grey mold outbreak in tree nurseries. Figure 2. Grey mold symptoms in the needles of the seedling. Figure 3. Cumulative proportions of the seedlings having the grey mold outbreak after different inoculations (1-6) in Exp. 1. 78 Figure 4. Duration in days when 50% of the seedlings had the grey mold outbreak. 79 Changes of xylem pressure potential in Quercus serrata saplings inoculated with Raffaelea sp., a possible causal fungus of oak mortality in Japan Yoshihiro Takahata and Takefumi Ikeda Y. Takahata, Kansai Research Center, Forestry and Forest Product Research Institute, Kyoto 612-0855, Japan. E-mail: amellea@affrc.go.jp T. Ikeda, Department of Forest Science, Kyoto Prefectural University, Kyoto 606-8522, Japan. Abstract From the end of the 1980s, mass mortality of Quercus spp. has increased in Japan. All dead trees show numerous entry holes bored by the oak pinhole borer, Platypus quercivorus (Coleoptera: Platypodidae). Anew species of Raffaelea has been isolated from both the beetle's body and the tree tissue attacked by this insect. After mass attacks by P. quercivorus, oak trees exhibit rapid wilt symptoms. In order to study the pathogenicity of Raffaelea sp. and evaluate the changes in the water status of the Quercus trees after a fungal infection, Q. serrata saplings (5-7 years old) were inoculated with Raffaelea sp., and some water physiological parameters were measured in 1996, 1998, 1999 and 2000. The transpiration rate and stomatal conductance to water vapor were greatly reduced in the saplings after the inoculation. Midday and predawn shoot xylem pressure potentials were higher after the inoculation than those in the control saplings. These potentials suddenly and rapidly declined in some of the inoculated saplings. The length of the periods before the decline in the xylem pressure potential was highly variable. Not all saplings died as a result of the inoculation. In the 2000 experiment, five of 16 trees died. Raffaelea sp. was reisolated from the dead saplings. Raffaelea sp. showed ability to reduce the water conductance of the xylem and kill Q. serrata saplings. These results support the hypothesis that the fungus causes mass mortality of the Quercus, which has been increasing in recent decades in Japan. Introduction From the end of the 1980s, mass mortality of oak trees (Quercus spp.) has increased on the Honshu and Kyushu islands of Japan (Ito and Yamada 1998). The dead trees are deciduous and evergreen oak species, mainly Quercus crispura, Q. serrata, Q. salicina, and Q. acuta. All dead trees showed numerous entry holes bored by the oak pinhole borer, Platypus quercivorus (Coleoptera: Platypodidae). This insect was as sumed to be an ambrosia beetle, but its associated (ambrosia) fungi were not known. Similar mass mortalities of oaks had occurred from the 1930s (Ito and Yamada 1998), but no forest pathological research was conducted at that time. Recently, however, a new species of Raffaelea was isolated from the bodies of P. quercivorus (body sur face and mycangia) and tree tissues attacked by this beetle (Ito et al. 1998, Kubono and Ito, under submission). This fungus was proposed to be the cause of the recent 80 mass mortality of oak trees. Indeed, according to inoculation tests, the pathogenicity of this fungus against Q. crispura was suggested (Ito et al. 1998). However, it is still not clear whether Raffaelea sp. has pathogenicity against other Quercus species in Japan. After the mass attacks by P. quercivorus, oak trees exhibited rapid-wilt symp toms. We speculated that oak trees inoculated with Raffaelea sp. would show rapid changes in their water status if the fungus is a causal agent of this mortality. In order to inquire the pathogenicity of Raffaelea sp. and evaluate the changes in the water status of the Quercus trees after a fungal infection, Q. serrata saplings were inoculated with Raffaelea sp., and some water physiological parameters of water relations were measured. Materials and methods The inoculation tests were conducted from late June to early July in 1996, 1998, 1999 and 2000. Saplings of Q. serrata (5-7 years old) growing at an experimental nursery at the Kansai Research Center, Kyoto, Japan (35°1' N, 135°44' E) were used for inoculation. The saplings were the same age for each year. The Raffaelea sp. isolate for inoculation was NA9. This isolate was isolated from the bark of an unidentified Quercus (Q. crispura or Q. serrata) tree. At each inoculation area, bands and plugs of bark were removed, and the mycelium of the isolate NA9 grown in a rice bran and wheat bran medium was placed on the wound. Thereafter, bands and plugs were replaced, and the inoculated area was sealed with parafilm and cloth adhesive tape to avoid contamination and desiccation (Fig. 1). Three types of control were used to evaluate the effects of wounding because the wound in these experiments was potentially harmful to the saplings. Type 1 con trol saplings were inoculated with a sterile medium. Type 2 control saplings were wounded as inoculated saplings, and no treatment was done on the type 3 control saplings. Figure 1. Diagram of the method of an inoculation. 81 To evaluate the water status of saplings, the transpiration rate, stomatal conductance to water vapor, and midday and predawn shoot xylem pressure potentials were measured. The transpiration rate, stomatal conductance to water vapor, and midday xylem pressure potentials were measured around noon on sunny days. On the following days, predawn xylem pressure potentials were measured. The transpiration rate and stomatal conductance were measured with a portable gas exchange system (Shimadzu SPB-H3), and xylem pressure potentials were measured with a pressure chamber (Soilmoisture Equipment Co.). These parameters were measured one or two times before inoculation and four to eleven times after inoculation. The intervals of measurements varied from one day to about two weeks. The last measurement was conducted by late September in each year. Results In 1996 and 1998, no sapling died after inoculation (Table 1). Two of 12 inoculated saplings died in 1999; five of 16 inoculated saplings died in 2000; however, none of the control saplings died during these two years. Raffaelea sp. was re-isolated from the tissues of some dead saplings. The length of the time periods before the death of the inoculated saplings was highly variable. In each year, after inoculation, the means of the transpiration rates and sto matal conductance to water vapor dropped greatly, except in the type 3 control sap lings, which were untreated. The type 1 and 2 controls recovered values equivalent to those of the type 3 control saplings by the end of the experiments. Although the values of the living inoculated saplings also recovered, more time was required than for the type 2 and 3 controls. The values did not recover in the dead inoculated saplings. During 2-7 weeks after inoculation, midday and predawn shoot xylem pres sure potentials in the inoculated, type 1 and 2 control saplings were higher than those of the type 3 control saplings. However, in some of the inoculated saplings, midday and predawn shoot xylem pressure potentials dropped suddenly and rapidly to extraordinarily low values, followed by death within several days. Table 1. Survival of Q. serrata seedlings inoculated with Raffaelea sp. * In 1996, plugs of bark were not removed ** Mean value ± standard deviation 1996* 1998 1999 2000 No. No. No. No. No. No. No. No. Treatment trees Dead trees dead trees dead trees dead Raffaelea sp. 9 0 12 12 2 16 5 Sterilized medium 3 Barked KB■ 3 KBfl 3 Mean diameter at ground level (cm) 2.8 ±0.59** 4.2 ±0.86 3.8 ±0.48 2.4 ± 0.44 82 Discussion The results in which some inoculated saplings wilted after inoculation and Raffaelea sp. was re-isolated from dead saplings show that this fungus can kill Q. serrata saplings. Because inoculation of Raffaelea sp. caused closure of stomata and reduc tion of shoot xylem pressure potentials, it was thought that the fungus was related to xylem dysfunction of Q. serrata saplings. These results suggest that Raffaelea sp. has pathogenicity against Q. serrata. The rapid progress of wilt in some of the inocu lated saplings in these experiments was similar to that observed in mature oak trees in mass mortality In conclusion, the results of this research support the hypothesis that this new Raffaelea species is the causal agent of the recent Quercus mass mortality in Japan. References Ito, S. and Yamada, T. 1998. Distribution and spread of the mass mortality of oak in Japan. J. Jpn. For. Soc. 80(3): 229-232. (In Japanese with an English summary.) Ito, S., Kubono, T., Sahashi, N. and Yamada, T. 1998. Associated fungi with the mass mortality of oak trees. J. Jpn. For. Soc. 80(3): 170-175. (In Japanese with an English summary.) Kubono, T. and Ito, S. Raffaelea quercivorus sp. Nov. associated with mass wilt of Japanese oak and the ambrosia beetle (Platypus quercivorus) submitted to Mycoscience. 83 Phytophthora cactorum on silver birch seedlings Arja Lilja and Risto Rikala A.Lilja, Finnish Forest Research Institute, Vantaa Research Center RO. Box 18, 01301 Vantaa, Finland. E-mail: arja.lilja@metla.fi R. Rikala, Finnish Forest Research Institute, Suonenjoki Research Station, 77600 Suonenjoki, Finland. Abtract In 1991, Phytophthora cactorum was first isolated from necrotic stem lesions of nursery seed lings of Betula pendula in Finland. Inoculations of birch stems with this Oomycetous Chromista in class Pseudofungi, resulted in necrotic lesions identical to those on birch seedlings in nurseries. In this study we monitored the effect of P. cactorum infection on the development of con tainer-grown, silver birch seedlings in nursery and after outplanting. In nursery stem lesions af fected the height growth of birches, the shoot height of seedlings was related to the disease sever ity. Asymptomatic seedlings were taller than the diseased seedlings and the shortest seedlings were those with stem leasons covering over half their stem diameter. After outplanting stem le sions did not affect seedling mortality or on the number of leader shoot changes. The height growth of seedlings in the reforestation site did not decrease with the disease rating. The diffenrences in the shoot heights related to the disease rating almost disappeared. It is still too early to conclude that silver birch seedlings will recover totally from P. cactorum infection although the differences in shoot heights present in the nursery between disease and apparently healthy seedlings have after two growing season in the field reduced. Keywords: Stem lesions, Betula pendula Introduction Since the early 1980's the production of container-grown Betula seedlings in Fin land has increased from 5 to 85% (Finnish Statistical... 2000). The use of green houses with controlled temperature and light, low -humified Sphagnum peat as the growth medium, and controlled irrigation and and fertilization in greenhouse and later on outdoor areas has resulted in good seedling growth in a shorter time. How ever the rapid seedling growth and high seedlings growing density may favor fungal diseases. Numerous problems such as stem lesions and cankers have occurred during nursery production and subsquent outplanting of silver and pubescent birch seed lings (B. pendula Roth and B. pubescens Ehrh.). Several fungi cause these lesions. The most extensively studied in Finland is Godronia multispora J. W. Groves which affects both natural birch saplings (Kurkela 1974) and nursery seedlings (Petäistö 1983). Other fungi isolated from birch stem lesions in nurseries include Fusarium 84 avenaceum (Corda, Fr.) Sacc., Alternaria spp. and Botrytis cinerea Pers. ex Nocca & Balb (Petäistö 1983, Lilja et ai. 1997). Anisogramma virgultorum (Fr.) Theiss. & Sydow, Syn. Plogwrightia virgultorum (Fr.) Saccardo also causes lesions on young silver birch seedlings in clearcut areas (Kujala 1942). In 1991, Phytophthora cactorum (Lebert & Cohn) was first isolated from necrotic stem lesions of silver birch seedlings in Finland (Lilja et ai. 1996). Inocula tions of birch stems with this Oomycetous Chromista in class Pseudofungi (Baldauff et al. 2000) resulted in necrotic lesions identical to those on birch seedlings in nurs eries (Lilja et al. 1996, Hantula et al. 1997, 2000). The aim of the present study was to monitore the effect of P. cactorum infec tion on the development of container-grown, silver birch seedlings in nursery and after outplanting on a reforestation site. Material and methods During the rainy summer of 1998 P. cactorum commonly affected birch seedlings in several Finnish forest nurseries. The following spring diseased and healthy silver birch seedlings were collected from two nursery fields in mid-Finland. In total 480 seedlings were collected. Half of the seedlings had been grown in peat pots, and the other half in hard plastic containers. The stem-lesion disease severity on each seed ling was assessed using a scale of Ito 4 where: 1= no lesion, 2 = lesion < 5 mm 2 , 3 = lesion > 5 mm 2 , but not covering over half of the stem diameter, and 4 = lesion spread over half of the stem diameter, but not girdling the stem (Fig. 1). On 4 May, after one-week storage at + 4°C, the seedlings were outplanted in a field in Southern Finland. The field (on former agricultural land had been ploughed one week earlier) was divided to 12 blocks. The seedlings grown in peat pots were planted to blocks 1-6 (experiment 1) and those grown in hard plastic containers were planted in blocks 7-12 (experiment 2). There were 10 seedlings in each of the four disease categories in both experiments in each block. There was 1 meter between rows and seedlings within rows and the rows were randomized within each block. A fence was built around the field to exclude moose. To avoid vole damage each seed ling was protected with a treeshelter. The shoot height of each seedling was meas Figure l.The stem-lesion disease severity on each seedling was assessed using a scale of 1 to 4 where: 1= no lesion, 2 = lesion < 5 mm 2 ,3 = lesion > 5 mm 2 , but not covering over half of the stem diameter, and 4 = lesion spread over half of the stem diameter, but not girdling the stem. 85 ured after planting and later on 22 September in 1999 and on 5 September in 2000. On these latter two dates dead seedlings and those in which a new leader had formed were also counted. On 21 June, 2000, 50% glyfosate (Roundup Bio, Monsanto, USA) was ap plied to weeds around each birch seedling using a Weedwiper 40 (Lasco, USA). Results Nursery experiment Stem lesions affected the height growth of birch seedlings both in peat pots and in hard plastic containers. In both cases the shoot height of seedlings was related to the disease severity. Asymptomatic seedlings were taller than the diseased seedlings and the shortest seedlings were those were lesions had spreaded over half of the stem diameter, but had not girdled the stem (Fig. 2). The effect of disease severity on the shoot height was more evident on seedlings grown in peat pots than on seedlings grown in hard plastic containers with seedlings in peat pots being about 40% taller than the seedlings in hard plastic containers. Outplanting experiment 1 After outplanting the healthy seedlings grown in peat pots in the nursery grew less than diseased seedlings, especially during the second growing season (Fig. 2). Height growth of seedlings increased with disease severity. Growth of seedlings with stem leasons covering over half their stem diameter grew more than healthy control seed lings or seedlings with smaller stem lesions. However, after the first summer dis eased seedlings were still shorter than the asymptomatic controls. After the second growing season category 4 seedlings were the shortest, but the differences in the shoot heights related to the disease rating at time of outplanting had almoust (Fig. 2). Stem lesions did not affect seedling mortality or on the number of leader shoot changes after outplanting. All seedlings were alive in September, 1999. After the second growing season in the field the seedling mortality (mean ± SE) for disease categories 2-4 was .7 ± 1.7, 3.3 ±2.1 and 11.7 ± 6%, respectively (Fig 3). In 1999 the percentage of seedlings which developed a new leader was 5.0 ± 2.2, 10 ± 3.6, 11.7 ± 4 and 3.3 ± 2.1 for disease categories 1-4. The corresponding percentages in 2000 were 8.3 ± 3.1, 5 ± 2.2, 8.3 ± 4 and 13.3 ± 4.9 (Fig. 4). Outplanting experiment 2 The height growth of control seedlings and diseased seedlings grown in hard plastic containers in the nursery did not differ after outplanting (Fig. 1). After the first grow ing season in the field, differences in shoot heights among seedlings rated in differ ent categories at the time of planting were still visible, however, after the second growing season these differences had almost diseapeared as in the experiment 1. 86 Stern lesion severity did not affect seedling mortality after outplanting. During the first growing season 1.7 ± 1.7% of the seedlings in category 2 died while 3.3 ± 2.1% died in category 4. After the second growing season the mortality of seedlings were according to the disease rating 1-4 :11.6±5.4, 16.7± 8, 10±4.5 and 16.7±4.9 %, respectively (Fig. 3). The percentage of seedlings which changed the leader shoot were 1 = 3.3 ± 2.1, 2 =3.3 ± 2.1, 3 = 1.7 ± 1.7 and 4 = 5 ± 2.2 in 1999. The corre sponding percentages in 2000 were 6.7 ±3.3, 15 ± 5.6, 8.3 ± 4.8 and 5 ± 2.2 (Fig. 4). Figure 2. The height of seedlings at the time of planting and growth after outplanting. Seedlings grown in peat pot in nursery (Experiment 1). Seedlings grown in hardplastic containers in nursery (Experiment 2). 1= no lesion, 2 = lesion < 5 mm 2 , 3 = lesion > 5 mm 2 , but not covering over half of the stem diameter, and 4 = lesion spread over half of the stem diameter, but not girdling the stem. Figure 3. The mortality of silver birch seedlings after outplanting. Seedlings grown in peat pot in nursery (Experiment 1) Seedlings grown in hardplastic containers in nursery (Experiment 2). 1= no lesion, 2 = lesion < 5 mm 2 , 3 = lesion > 5 mm 2 , but not covering over half of the stem diameter, and 4 = lesion spread over half of the stem diameter, but not girdling the stem. 87 Figure 4. The percentage of silver birch seedlings, which changed leader shoot after outplanting. Seedlings grown in peat pot in nursery (Experiment 1) Seedlings grown in hardplastic containers in nursery (Experiment 2). 1= no lesion, 2 = lesion < 5 mm 2 , 3 = lesion > 5 mm 2 , but not covering over half of the stem diameter, and 4 = lesion spread over half of the stem diameter, but not girdling the stem. Discussion In this study nursery the stem lesion caused by P. cactorum decreased the height growth of birch seedlings. This is accordance with the previous studies with nursery seedlings (Lilja et ai. 1996). After outplanting the height growth of seedlings did not decreased with the disease rating. The diffenrences in the shoot heights related to the disease rating which were present at time of outplanting had almoust diseappered in two years in the field. Vigorous growth of healthy seedlings in hard plastic containers in nursery might explain the poor growth of control seedlings after outplanting in experiment 1. The seedlings were too tall compared with the size of container cavity. After planting the rooting process might have been less successfull because of large transpiration area. Today the production of container seedlings to a large extent have replaced production of bareroot seedlings in Finland (Finnish Statistical... 2000). The envi ronmental factor on which soil-borne Phytopthoras have a obligatory dependency is moisture (Grove et al. 1985, Erwin and Ribeiro 1996). The microclimate within birch stands in containers is not easy to keep dry and well aerated, because regular irrigation is needed. During last tree years excess rainfall in midsummer, in a phase when coverplastics have been removed from greenhouses, has been found to create favourable conditions for severe P. cactorum infection. The clonality of P. cactorum on strawberry inside Europe shows that it easily can be spread on infected seedlings (Hantula et al. 1997, 2000). It is also commonly brought into apple (Malus domestica Borkh) orchards on seedlings or rootstocks 88 (Harris 1991). A Phytophthora sp. (Brasier et ai. 1999) causing widespread alder (Alnus glutinosa (L.) Gaertn.) mortality on riparian populations (Cech and Brandstetter 1999, Gibbs et ai. 1999, Streito et ai. 1999) is also supposed to be spread on nursery seedlings. The potential for Phytophthoras to survive and spread after outplanting de pends on that how these water depedent pathogens can adapt on new environmental conditions. Phytophthora spp., including P. cactorum, causing root rot on Douglas fir (Pseudotsuga menziesii (Mirb.) Franco) caused mortality in forest, but surviving trees regenerated healthy roots even though Phytophthora persisted in old lesions (Hansen et al. 1980). In this study stem lesions did not effect on mortality of birch seedlings after outplanting. And even those lesions healed well which have spread over half of the stem diameter in the nursery. Usually under suitable conditions, high moisture and cool temperatures, new Phytophthora inoculums are producted continously when lesions produce sporangia and zoopsores which can be splahed by rain (Erwin and Ribeiro 1996). The summer 1999 was exceptional dry and warm in the South part of Finland. There was no need to control weeds before summer 2000. Dry growing season in 1999 and weed control in 2000 perhaps kept the relative humidity of air around the seedlings under the level optimal for the pathogen growth and new infections. It is still too early to conclude that silver birch seedlings will recover totally from P. cactorum infection although the differences in shoot heights present in the nursery between disease and apparently healthy seedlings have after two growing season in the field reduced. Acknowledgments We thank Ms Ritva Vanhanen and Pentti Kananen for technical assistance as well as Dr. Jack Sutherland for his critical reading of the manuscript. The work was sup ported by a research grant from Metsämiesten säätiö. References Baldauf, S. L., Roger, A. J., Wenk-Siefert, I. and Doolittle, W. F. 2000. A kingdom-level phylogeny of eukaryotes based on combined protein data. Science 290: 972-977. Brasier, C. M., Cooke, D.E.L. and Duncan, J.M. 1999. Origin of a new Phytophthora pathogen through interspecific hybridization. Proc. Nat. Acad. Sci. U.S.A. 96: 5878-5883. Chech, T. L. and Brandstetter, M. 1999. Report on the current situation of Phytophthora decline in Austria. Forstschutz Aktuell. No. 23-24: 16-19. Erwin, D. C and Ribeiro, O. K. 1996. Phytophthora Diseases Worldwide. APS Press. St. Paul, Minnesota. ISBN 0-89054-212-0. Finnish Statistical Yearbook of Forestry, 2000. Finnish Forest Research Institute. SVT. Agriculture, forestry and fishery 2000: 14. Gummerus, Jyväskylä. ISBN 951-40-1752-8. Gibbs, J. N., Lipscombe, M. A. and Peace, A. J. 1999. The impact of Phytophthora disease on riparian populations of common alder (Alnus glutinosa) in Southern Britain. European Journal of Forest Pathology 29: 39-50. 89 Grove, G. G., Madden, L. V., Ellis, M. A. and Schmitthenner, A. F. 1985. Influence of temperature and wetness duration on infection of immature strawberry fruit by Phytophthora cactorum. Phytopathology 75: 165-169. Hansen, E. M., Roth, L. E, Hamm, P. B. and Julius, A. J. 1980. Survival, spread, pathogenicity of Phytophthora spp. on Douglas-fir seedlings planted on forest sites. Phytopathology 70:422- 425. Hantula, J., Lilja, A. and Parikka, P. 1997. Genetic variation and host specificity of Phytophthora cactorum isolated from Europe. Mycological Research 101: 565-572. Hantula, J., Lilja, A., Nuorteva, H., Parikka, P. and Werres, S. 2000. Pathogenicity, morphology and genetic variation of Phytophthora cactorum from strawberry, apple, rhododendron and silver birch. Mycological Research 104: 1062-1068. Harris, D. C. 1991. The Phytophthora diseases of apple. Journal ofHortic Science 66: 513-544. Kujala, V. 1942. Plowrightia virgultorum koivun tuholaisena. Metsätaloudellinen Aikakauslehti 59: 127-128. Kurkela, T. 1974. Godronia multispora Groves (Helotiales) and its pathogenicity to Betula verrucosa Ehrh. and B. pubescens Ehrh. Karstenia 14: 33-45. Lilja, A., Rikala, R., Hietala, A. and Heinonen, R. 1996. Stem lesions on Betulapendula seedlings in Finnish forest nurseries and the pathogenicity of Phytophthora cactorum. European Journal of Forest Pathology 26: 89-96. Lilja, A., Lilja, S., Kurkela, T. and Rikala, R. 1997. Nursery practices and management of fungal diseases in foret nurseries in Finland. A review. Silva Fennica 31,1: 79-100. Petäistö, R-L. 1983. Rauduskoivun versolaikut taimitarhalla. Abstract: Stem spotting of birch {Betulapendula) in nurseries. Folia Forestalia 544: 1-9. Streito, J. C., Villartay, G. Tabary, F. and de Villartay, G. 1999. Une nouvelle espece de Phtophthora s'attaque a l'aulne. PhytomaNo. 519: 38-41. Strouts, R. G. 1981 .Phytophthora diseases of trees and shurbs. Arbocultural Leaflet, Department of the Environment, UK. No. 8. 16 p. 90 Tubercularia ulmea canker: effect of pathogen, host, and cultural practice Marcus B. Jackson and Robert W. Stack M. B. Jackson, Dept. of Plant Science, North Dakota State University, Fargo, ND 58105 USA. R. W. Stack, Dept. of Plant Pathology, North Dakota State University, Fargo, ND 58105 USA. E-mail: rstack@ndsuext.nodak.edu Abstract Several systemic fungicides were tested as treatments for stem cankers caused by Tubercularia ulmea on Ulmus pumila, Eleagnus angustifolia, and Gleditsia triacanthos. None was able to prevent infection but benomyl most consistently reduced canker size. Further work is warranted. Keywords: chemotherapy, Nectria cinnabarina, tree wound dressings Introduction In windbreak and other tree plantings in North Dakota, USA, cankers caused by Tubercularia ulmea Carter have caused considerable damage to Ulmus pumila L. (UP), Elaeagnus angustifolia L.(EA), and Gleditsia triacanthos L. (GT) tree plantings (Dooling 1973, Sengpiel 1977, Walla and Stack 1987, Walla and Stack 1988). In this otherwise treeless region of windswept prairies, planted trees are constantly subject to stress by wind, drought, winter cold, summer heat, and attack by defoliating in sects (Stack et al. 1990). Tubercularia ulmea was identified as causing a stem canker disease of Ulmus pumila by Carter in Illinois, USA (Carter 1947). We identified the teleomorph of T. ulmea as Nectria cinnabarina Tode ex Fries (Sengpiel and Stack 1985). That would make T. ulmea synonymous with T. vulgaris', however, we continue to use the anamorph name T. ulmea for the present because our strains of this fungus differ somewhat from typical T. vulgaris. Stack et al. (1990) applied Manion's disease spiral concept to the decline of Siberian elm windbreaks on the northern Great Plains. They cited T. ulmea and B. hypodermia as contributing factors in decline of U. pumila. Since most facultative canker fungi require a weakened host, maintaining tree vigor is extremely important in prevention of canker diseases. Drought and freezing stress have been recognized as two of the main environmental factors in canker disease development (Schoeneweiss 1981). Treatment of pruning wounds was suggested for many years, based on a long accepted belief that painting wounds could prevent decay. Shigo and Shortle (1983) 91 disproved that assumption; however, their findings in regard to decay have been misinterpreted by some to mean that wound dressings are of no value under any circumstance. Wound dressings are still recognized as valuable for prevention of canker diseases (Bedker and Blanchette 1983, Riecken 1985). Researchers continue to search for fungicide-containing dressings that will reduce the effects of some canker fungi. Rosenberger et al. (1983) found that neither 6.0 g/ml captafol nor 0.3 g/ml benomyl were effective in controlling twig dieback of apple caused by N. cinnabarina. Bedker and Blanchette (1984) found that N. cinnabarina cankers on GT treated with 0.1 g/ml benomyl were smaller than those not treated. Benomyl has been shown to reduce Nectria galligena Bres. canker on apple (Malus pumila Mill.) (English et al. 1979). Riecken (1985) showed that a wound treatment containing 3% thiophanate-methyl was effective in reducing N. cinnabarina canker development. Peterson and Helmer (1992) demonstrated that applying a wound treatment containing 20 g/1 imazalil would reduce N. cinnabarina penetration in several species. The studies reported here were done to determine the variability in patho genicity of T. ulmea isolates in North Dakota, and to investigate the potential value of wound dressings to prevent T. ulmea infections in species often planted in ND. Materials and methods Seventeen T. ulmea isolates from seven host species were tested for pathogenic vari ability on EA and UR Ten wound dressings were applied in field tests to determine effectiveness in preventing T. ulmea canker development. Isolation and inoculum production Dead stems and cankered bark with sporodochia were collected, cleaned with dis tilled water, and induced to sporulate by incubating for 24 hours over moist filter paper. Conidia were removed from sporodochia and cultured on PDA. After two to four days, colonies were transferred to new PDA plates. After six weeks, sporodochia developed on the agar surface. Isolates were transferred to fresh media approxi mately every 60-90 days. Colonized wheat kernel inoculum was used throughout these tests and was prepared by placing sterilized wheat kernels onto the surface of T. ulmea cultures for two to five weeks. When thoroughly colonized, kernels were removed and applied to the wounds. Pathogenicity tests Seventeen T. ulmea isolates were collected. Greenhouse-grown UP and EA trees were inoculated with T. ulmea isolates in two experiments. Isolates confirmed as T. ulmea were tested for the ability to cause disease by inoculating plants in the green house; each test had three replicate inoculations of each isolate on UP and EA. An 92 inverted V-notch wound 5 mm by 8 mm was used in greenhouse tests. A colonized wheat kernel was placed between the bark and the wood in the inverted v-notch wound. The bark flap was pressed against the kernel, wrapped in parafilm®, and covered with aluminum foil. Cankers were measured at 3 wk or 5 wk after inocula tion. Wound dressings Wound dressings were tested in field studies to determine if they would suppress T. ulmea cankers in a natural environment. Several commercially available fungicides were used; these included benomyl, fenarimol, propiconazole, tebuconazole, thia bendazole, and thiophanate-methyl. Other products were a commercial tree paint ("Ortho") and alcohol-based shellac. Choice of wound dressings was based on green house trials not reported here (Jackson 1997). For tests on UP and EA, trees approximately 30 years old were used. Sixty stems 5-8 cm diam were selected at random for each host. A 27 mm disk of bark and vascular cambium was removed from each stem with a hole saw and two seeds of wheat kernel inoculum were placed in opposite sides of the wound. Two isolates were used on each host. The wounds were immediately dressed with one of eight treatments at 1% active ingredient or left untreated. Fungicides were carried in dor mant oil except benomyl and thiophanate-methyl which were in water. Stems were wrapped with a 30 cm x 30 cm piece of plastic food wrap. Non-inoculated controls consisted of wounded stems covered with plastic wrap. Three replicate inoculations for each treatment x isolate combination and control were done. Canker length was determined at four months after inoculation for EA or 6 months for UP. For GT, methods were similar except that 20-year-old trees were used, the wound was 10-mm diam, the dressings were applied immediately at 5% active in gredient, four inoculum kernels were used per wound, and canker length was meas ured at two months after inoculation. Disease evaluation and data analysis To determine the extent of cankering, the bark on each inoculated and control stem was cut and removed above and below the inoculation point. Canker length meas urements were taken above and below the wound and measurements combined. For the pathogenicity tests, the two experiments for each host were combined for analy sis. For the tests of wound dressings, similar measurements of canker length were taken. The isolate by treatment interaction was not significant in any of the experi ments so the results for the two isolates for each host were combined for analysis. 93 Results Pathogenicity tests In all, 17 T. ulmea cultures were evaluated by inoculation to EA and UP. In two experiments, all of the T. ulmea isolates tested caused cankers on both hosts. A few isolates were present in only one of the two experiments. There was no relationship between the source (host) of the isolate and lesion size on either test host (data not shown). A combined analysis using data from both experiments and both hosts showed that isolates differed significantly in the size of cankers produced (Fig. 1). Canker size was taken as a measure of generalized disease causing ability = pathogenicity. Figure 1. Pathogenicity of T. ulmea isolates to U.pumila and E. angustifolia. Values are means of two experiments and two hosts. Canker length determined at 3 or 5 weeks after inoculation with T. ulmea by placing a colonized wheat kernel into a stem wound in greenhouse grown saplings. Fisher's F-protected Least Significant Difference (FLSD) @(p=.os) = 33.0. Wound dressings Wound dressings were tested in field trials to determine if they could reduce or prevent cankers caused by T. ulmea. In the UP and EA tests, cankers developed from all wounds in the inoculated controls. Cankers also developed from many non-in 94 oculated wounds as well probably because of the abundance of natural inoculum in this nursery. The effect of the wound dressings on canker size at 6 months following inocu lation with T. ulmea in the field test on UP is shown in Figure 2. The canker size caused by the two isolates was not different so only the combined values are shown. None of the treatments reduced canker size compared to the inoculated control but benomyl and Ortho tree paint did have cankers shorter than the non-inoculated con trols. The effect of the wound dressings on canker size at four months following inoculation of EA in the field test is shown in Figure 3. Again, the canker size caused by the two isolates of T. ulmea did not differ. In this test, two treatments, benomyl and thiophanate-methyl, reduced canker length compared to the inoculated control. Two additional treatments, thiabendazole and propiconazole, differed significantly from the non-inoculated control. The effect of the wound dressings on canker size at two months following inoculation of stems on the 20 yr old GT in the field test is shown in Figure 4. The canker size caused by the two isolates of T. ulmea differed but there was not a sig nificant isolate by treatment interaction so the measurements for the isolates were combined. In this test, no treatments differed from the inoculated check although the benomyl treatment showed a strong trend toward reduced canker length. Despite variation, some of the wound dressings did reduce canker develop ment; benomyl, in particular, tended to be effective in reduce canker length. None of the dressings, however, was really effective in preventing infection, even though they were applied at an optimum time. Figure 2.. Effect of wound dressings on canker size in U. pumila inoculated with T. ulmea in field tests. Canker length was determined at 6 months after inoculation with T. ulmea by placing colonized wheat kernels into a wound on stems of 30 year old trees. Mean of two isolates. Fisher's F protected Least Significant Difference (FLSD) @(p=.os) = 15.2. 95 Figure 3. Effect of wound dressings on canker size in E. angustifolia inoculated with T. ulmea in field tests. Canker length was determined at 4 months after inoculation with T. ulmea by placing colonized wheat kernels into a wound on stems of 30 year old trees. Fisher's F-protected Least Significant Difference (FLSD) @(p=.os) = 25.2. Figure 4. Effect of wound dressings on canker size in G. triacanthos inoculated with T. ulmea in field tests. Canker length was determined at 2 months after inoculation of stems of 20 year old trees with T. ulmea by placing a colonized wheat kernels into a stem wound. Fisher's F-protected Least Significant Difference (FLSD) @(p=.05) = 12.9. Discussion Pathogenicity tests We observed during the pathogenicity tests that T. ulmea sporodochia occurred more often on killed stems than on live, cankered stems, suggesting that saprophytic growth of T. ulmea favored sporulation. 96 Cankers that caused wilting of tops beyond the inoculation site were much larger than cankers that did not cause wilting. A major factor appeared to be the rate at which the cankers grew in width across the stem. It appeared that some T. ulmea isolates caused more extensive lateral cankering even though their cankers were not as long as some other isolates which caused extensive growth lengthwise along the stem. Unfortunately, the numbers of plants were not sufficient to form any conclu sion about this. Additional study might show whether isolates differ in how they form cankers. Wound dressings Cankers developed from many non-inoculated wounds. This nursery with its many 30 yr old trees had a long history of broken branches and tops from storms, etc. While many old cankers had been removed prior to this study, some probably re mained and provided natural inoculum of several potential pathogens - including T. ulmea. When combined with a severe wind/rain storm that occurred in this nursery 10 days after the wound treatments were done, that natural inoculum proved more than sufficient to infect the control wounds. Wound closure Index. On EA and UP, small but significant reduction in width of wounds (wound closure) occurred in some treatments. Those treatments that showed reduced canker development also tended to show wound closure. Since the two measures, canker length and wound closure, appear correlated, the latter was not analysed further. On GT, an additional comparison was made of stem wounds dressed with only petroleum oil (not shown in Figure 4). These developed larger cankers than the in oculated controls, suggesting this product was toxic to freshly cut tissue. The choice of oil as a carrier in the EA and UP trials was probably unfortunate, and may have interfered with any potential activity of those fungicides, especially propiconazole, that were carried in it. The variable and somewhat inconsistent results noted in this study of wound dressings is similar to some of the results reported by others. In our case we believe the large coefficient of variation noted in analysis of these experiments indicates a high level of experimental error. This variability may have been due to differences in the individual host plants (grown from seed collected in the wild), to changes in the T. ulmea isolates over time, or the environment, or more likely a combination of all these. Reducing the variation in future studies will be needed to better discriminate among treatments. This might be done by using clonal trees and improving maintainence of cultures, and substantially increasing the degree of replication within experiments. As was seen in previous research (Bedker and Blanchette 1984, Riecken 1985, Peterson and Helmer 1992), none of the chemicals tested was curative. The reduced cankering produced by benomyl and possibly by thiophanate-methyl and propiconazole were nevertheless encouraging and warrant further work. When com bined with improved cultural practices, application of benomyl could be an impor tant component of an integrated program for the control of cankers caused by T. ulmea. 97 References Bedker, P.J. and Blanchette, R.A. 1983. Development of cankers caused by Nectria cinnabarina on honey-locust after root pruning. Plant Dis. 67: 1010-1013. Bedker, P.J. and Blanchette, R.A. 1984. Control of Nectria cinnabarina cankers on honey-locust. Plant Dis. 68: 227-230. Carter, J.C. 1947. Tubercularia canker and dieback of Siberian elm (Ulmus pumila L.). Phytopathology 37:243-246. Dooling, O.J. 1973. Cankers in North Dakota windbreak plantings, survey and evaluation. Insect and Disease Report #73-10. USDA Forest Service Northern Region, Missoula, MT. English, H., Dubin, H.J. and Schick, F.J. 1979. Chemical control of European canker of apple. Plant Dis. Rep. 63:998-1002. Jackson, M.B. 1997. Influence ofhost and chemical treatments on cankers caused by Tubercularia ulmea. M.Sc. Thesis, North Dakota State Univ., Fargo, ND. 91 p. Peterson, J.L. and Helmer, D.B. 1992. A wound treatment system to suppress cankers and wood rot in trees. J. Arboric. 18: 155-160. Riecken, I. 1985. Experiments on the prevention of the coral spot disease (Nectria cinnabarina). J. Plant Dis. Protect. 92: 516-529. Rosenberger, D. A., Burr, T.J. and Gilpatrick, J.D. 1983. Failure of canker removal and postharvest fungicide sprays to control Nectria twig blight on apples. Plant Dis. 67: 15-17. Schoeneweiss, D.F. 1981. Infectious diseases of trees associated with water and freezing stress. J. Arboric. 7: 13-18. Sengpiel, H.W. 1977. Tubercularia canker: a problem on highway plantings in North Dakota. Internal publication, North Dakota State Highway Department. Bismarck. Sengpiel, H.W. and Stack, R.W. 1985. Tubercularia canker of Russian-olive in North Dakota. (Abstr.) Phytopathology 75: 1367. Shigo, A.L. and Shortle, W.C. 1983. Wound dressings: results of studies over 13 years. J. Arboric. 9: 317-329. Stack, R.W., Krupinsky, J.M. and Walla, J.A. 1990. Decline of Siberian elm on the Great Plains. (Abstr.) Phytopathology 80: 1064. Walla, J.A. and Stack, R.W. 1987. An evaluation of the condition of twelve common windbreak species in North Dakota. Proc. North Dakota Acad. Sci. 41: 55. Walla, J.A. and Stack, R.W. 1988. Tubercularia canker of honey-locust in North Dakota. Plant Dis. 72: 734. 98 Frequently isolated endophytes from the Japanese beech -Endophytic fungal composition, their temporal variations in colonization rate and distribution in Northern Japan Norio Sahashi Kyushu Research Center, Forestry and Forest Products Research Institute (FFPRI), Kurokami, Kumamoto, 860-0862, Japan. Email: sahasi@affrc.go.jp Abstract To determine the dominant fungal endophytes of the Japanese beech (Fagus crenata) and to monitor their isolation frequency, we isolated fungi from symptomless organs of beech including leaves, petioles, and current and old (1- to 5- year-old) twigs after surface sterilization. Of the thirteen fungal taxa obtained, three were isolated most often. An unidentified species of Discula and an unidentified sterile fungus, Lb, were isolated frequently from leaves, and an unidentified species of Phomopsis was isolated most frequently from twigs. The isolation frequency over the growing season varied for the two dominant fungal species in the leaves, Discula sp. and Lb. These two species had similar patterns of isolation, even in petioles and current-year twigs, although isola tion frequencies of a given species varied with organs. An organ-specific distribution of the fungal species in the host plant was apparent. The three fungal species noted above were considered to be the dominant endophytes of the Japanese beech. In addition, to confirm whether the two endophytic fungi in the beech leaves are übiquitous at different sites, leaves of the beech were collected monthly during the vegetation period at five sites in the Tohoku district, the northeastern part of Japan. Two dominant endophytic fungi were Discula sp. and Lb. This result strongly suggests that these two fungi are generally associated with leaves of the Japanese beech at different sites. Spores of Discula sp. were released for a very short time in late May, just after the disappearance of the snow cover on the forest floor. These spores may be important for the infection of newly flushing leaves. Keywords: Fagus crenata, fungal endophyte, isolation frequency Introduction Endophytic fungi, which induce asymptomatic infections of healthy plant tissues, have been the subject of many studies. These studies, which have focused on endo phyte communities in woody plants mainly in Europe and in North America, have covered such subjects as methods of detection, taxonomy, species composition, dis tribution at a variety of scales, biological, ecological and physiological aspects, and interactions and mutualistic symbiosis between endophytes and host plants (e.g., 99 Bills 1996, Carroll 1986, 1988, 1995, Petrini 1986, 1991, 1996). In Japan, although there have been some reports on endophytic fungi in conifers (Hata and Futai 1993, 1995, 1996, Kaneko et al. 1997), little is known about the species composition of endophytic fungi in deciduous broad-leaved trees. Moreover, limited information is available on temporal changes in the endophyte assemblages. Clarification of the fungal species composition, variations on a temporal scale of endophyte isolation, and their distribution among different sites is an important step in understanding the biology, ecology and physiology of endophytes and host-endophyte interactions. The objectives of this study were (I) to isolate fungal endophytes from various organs of the Japanese beech (Fagus crenata Blume), (II) to monitor their temporal variations in isolation success and (III) to confirm whether the endophytic fungal composition of the beech are übiquitous at the different sites in the Tohoku district, Northern Japan. Materials and methods Isolation of fungal endophyte Several organs of beech which included healthy and disease-symptomless leaves, current- and 1- to 5-year-old twigs, and petioles (Fig. 1) were collected twice a month in 1995 (for old twig, from may, 1995 to April, 1996) at Tazawako. Symptomless leaves were also collected monthly from five different sites (Fig. 2) in 1996. These samples were cut into small pieces, surface sterilized and incubated on potato dex trose agar (PDA) at 15° C for 3 weeks or more. The isolation frequency (IF) of a single endophyte taxon was calculated by the following formula: where, N. and N t are the number of segments from which the fungus was isolated and the total number of segments cultured, respectively. Figure 1. Organs of Japanese beech used for isolation of fungal endophyte. IF=N./N x 100 1 t 100 Figure 2. Map of sample collection sites for isolation of endophytic fungi from beech leaves. Black circles indicate main five collection sites. Numbers in parentheses show the stand age (years) in each experimental sites. Detection of air-borne spore of Discula sp. To confirm whether spores of Discula sp. are dispersed as air-borne inocula and, if dispersed, to determine the dispersal time of spores, spore trapping experiments using Petri-plates containing PDA was carried out in one of the main experimental sites, Tazawako. Ten trapping plates were exposed weekly for 1, 3, 5 and 10 min. at 1 m above ground under the canopy in the beech forest from mid-April to August. The plates were taken to the laboratory within 2 h, incubated at 15° C for 2 weeks and the number of Discula sp. colonies were calculated. Snow depth at the same site was also recorded under the canopy inside the forest. Results and discussion Endophytic fungal composition Discula sp. and an unidentified isolate, Lb, which had Ascochyta-like colonies, were frequently found in the leaves, and Phomopsis sp. was frequently found in the twigs after surface sterilization. In early May, just after expansion of the current leaves, no fungal isolates were detected in the leaves. Also, no fungal isolates were detected in leaflets flushing from terminal buds. Discula sp. was detected for the first time on May 25 and continued to be isolated with a relatively high frequency until July 11 101 (Fig. 3A). The isolation frequency of the fungus then decreased gradually until Au gust 7, and was 10 to 15% until the end of September. In early October, at the end of the growing season, the isolation frequency of this fungus increased slightly. On the other hand, the fungus Lb was first isolated in late June and the isolation frequency increased rapidly and kept a high incidence (over 80%) until the end of the growing season (Fig. 3B). Thus, the isolation profiles of Discula sp. and Lb appeared to be different. In addition to these two dominant fungal species, Phomopsis sp. was iso lated at a low frequency (3 to 13%) from early June to the end of the experiment, although there was a small peak in late July (Fig. 3C). From the old (1- to 5-year-old) twigs, only one fungal species, Phomopsis sp., was isolated frequently (Fig. 4). From early May 1995 to late April 1996, it was isolated from more than 85% of the segments. Discula sp., which was one of the largest components of the fungal assem blage in the leaves, and Phomopsis sp., was isolated frequently in the newly devel oping current-year twigs (Fig. 5). Phomopsis sp. was isolated frequently throughout the growing season, but the isolation frequency of Discula sp. gradually decreased, as it did in the leaves (Figs. SA, C). In petioles, the fungi frequently isolated were Discula sp. and Lb, which were also isolated from the leaves, and Phomopsis sp., which was frequently isolated from in the twigs (Fig. 6). The patterns of variation in the isolation frequency of Discula sp. and Lb were similar in the petioles and the leaves, although isolation frequency of a given species slightly differed among organs (Figs. 6A, B). The results of the present study also showed an organ specific distribution of fungal species in a given host species. These three fungal species were considered to be the dominant endophytes of the Japanese beech. Distribution of the endophytes at different sites Two fungal taxa, Discula sp. and an unidentified sterile fungus, Lb were also fre quently isolated from the beech leaves in all five sites surveyed (Figs. 2, 7). These two fungi accounted for the great majority of the isolates while other fungi were infrequently or rarely isolated as reported for conifers by Carroll and Carroll (1978). At most sites, Discula sp. first appeared in May, had a peak isolation frequency in June and then gradually decreased. On the contrary, the fungus Lb was first isolated in late June to July and the isolation frequency increased until the end of the vegeta tion period at all sites. The results on major endophyte composition and on temporal pattern of isolation frequency were consistent with the results at Tazawako in 1995. These results strongly indicate that infections of beech leaves by these two fungi are übiquitous in the Tohoku district and these two fungi are generally associated with leaves'of the Japanese beech at different sites. 102 Figure 3. Isolation frequency of endophytic fungi over the growing season from healthy beech leaves. (A) and (B) isolation frequency of two frequent endophyte, Discula sp. and sterile Lb. (C) less frequent endophyte Phomopsis sp. +++, significant difference at 0.001 level with sampling date. Figure 4. Isolation frequency of the endophytic fungi Phomopsis sp. (solid bar), Discula sp. (shaded bar) and a sterile Lb (open bar), over one year from healthy twigs. ++, significant difference at 0.01 level with sampling date for Phomopsis sp. 103 Figure 5. Isolation frequency of endophytic fungi Discula sp. (A), sterile Lb. (B) and Phomopsis sp. (C), over the growing season from healthy current-year twigs. +++ andNS, significant difference at 0.001 level and no significant difference with sampling date, respectively. Figure 6. Isolation frequency of endophytic fungi Discula sp. (A), sterile Lb. (B) and Phomopsis sp. (C), over the growing season from petioles. ++, + and NS, significant difference at 0.01 and 0.05 levels and no significant difference with sampling date, respectively. 104 Figure 7. Isolation frequencies (percentage of total 150 segments) of the two main endophyte, Discula sp. and the sterile Lb., in beech leaves collected at different experimental sites from May to October. Spore dispersal of Discula sp. No spore dispersal of Discula sp. was detected in the presence of a snow cover. Dispersal of spores was detected for a very short time in late May with the peak of spore dispersal occurring just after the disappearance of snow covering the forest floor (Fig. 8). This suggests the hypothesis that Discula sp. overwinters on fallen leaves under the snow and sporulates just after snow melt. It is most likely that dispersed spores play an important role in infecting beech leaves, as suggested by Johnson and Whitney (1992) and Toti et ai. (1993) for endophytes of black spruce and European beech, respectively, because current leaves of beech were fully ex panded at this time. The observation that no infected young rolled leaflets within winter buds were observed also supports this idea. Further studies should be made to clarify the ecology and host-endophyte in teraction for a better understanding of the role of endophytes in forest ecosystem. 105 Figure 8. Detection of air-borne spores of Discula sp. at Tazawako beech forest. Acknowledgements I thank K. Hata, T. Kubono, Y. Miyasawa, S. Ito and S. Kaneko for their help and valuable discussion and comments. References Bills, G. F. 1996. Isolation and analysis of endophytic fungal communities from woody plants. In Endophytic fungi in grasses and woody plants; Systematics, ecology and evolution. Edited by S. C. Rediin and L. M. Carris. APS Press, St. Paul, Minnesota, pp. 31-65. Carroll, G. C. 1986. The biology of endophytism in plants with particular reference to woody plants. In Microbiology of the phyllosphere. Edited by N. J. Fokkema and J. van den Heuvel. Cambridge Univ. Press, Cambridge, UK. pp. 205-222. Carroll, G. C. 1988. Fungal endophytes in stems and leaves: from latent pathogen to mutualistic symbiont. Ecology 69: 2-9. Carroll, G. C. 1995. Forest endophytes: Pattern and process. Can. J. Bot. 73 (Suppl. 1): 51316- 51324. Carroll, G. C. and Carroll, F. E. 1978. Studies on the incidence of coniferous needle endophytes in the Pacific Northwest. Can. J. Bot. 56: 3034-3043. Hata, K. and Futai, K. 1993. Effect of needle aging on the total colonization rates of endophytic fungi on Pinus thunbergii and Pinus densiflora needles. J. Jpn. For. Soc. 75: 338-341. Hata, K. and Futai, K. 1995. Endophytic fungi associated with healthy pine needles and needles infested by the pine needle gall midge, Thecodiplosis japonensis. Can. J. Bot. 73: 384-390. Hata, K. and Futai, K. 1996. Variation in fungal endophyte populations in needles of the genus Pinus. Can. J. Bot. 74: 103-114. Johnson, J. A. and Whitney, N. J. 1992. Isolation of endophytes from black spruce (Picea mahana) dormant buds and needles from New Brunswick, Canada. Can. J. Bot. 70: 1754-1757. Kaneko, S., Sakamoto, Y. and Kiyohara, T. 1997. Biological characteristics of Cryptosporiopsis abietina on Hinoki cypress and its antagonistic effect to other microorganisms. Mycoscience 37: 391-399. Petrini, O. 1986. Taxonomy of endophytic fungi of aerial plant tissues. In Microbiology of the phyllosphere. Edited by N. J. Fokkema. and J. van den Heuvel. Cambridge Univ. Press, Cambridge, UK. pp. 175-187. Petrini, 0. 1991. Fungal endophytes of tree leaves. In Microbial ecology of leaves. Edited by J. H. Andrews and S. S. Hirano. Springer-Verlag, New York. pp. 179-197. 106 Petrini, O. 1996. Ecological and physiological aspects of host-specificity in endophytic fungi. In Endophytic fungi in grasses and woody plants; Systematics, ecology and evolution. Edited by S. C. Rediin and L. M. Carris. APS Press, St. Paul, Minnesota, pp. 87-100. Toti, L., Viret., 0., Horat, G. and Petrini, 0.1993. Detection of the endophyte Discula umbrinella in buds and twigs of Fagus sylvatica. Eur. J. For. Path. 23: 147-152. 107 Effect of increased carbon dioxide and ozone on leaf spot pathogens of birch Leena Syrjälä and Marja Poteri Finnish Forest Research Institute, Suonenjoki Research Station, Juntintie 40, 77600 Suonenjoki, Finland. E-mails: leena.syrjala@metla.fi, marja.poteri@metla.fi Abstract We studied how increased carbon dioxide and ozone affect birch - leaf spot pathogens - interac tion. 9-year-old ozone sensitive (clone 80) and ozone tolerant (clone 4) silver birch (Betulapenduld) clones were fumigated with elevated (2 x ambient) concentrations of C0 2, ozone and combined C0 2 +0 3 in open-top chambers. In Experiment 1 we monitored the effect of treatments on the development of spots in naturally infected leaves during August-September in 2000 by photo graphing and image analyzing the leaf spots. In Experiment 2 we studied indirect effects of the gases on disease development by exposing clone plantlets to fumigations for 5 weeks, after which we inoculated the plantlets with Marssonina betulae -leaf spot fungus and incubated them in a growth chamber. In Exp. 1 both ozone and C0 2 increased the diseased leaf area (DLA) of clone 4. In clone 4 the increasing effect of fumigations was more than twice higher than in clone 80, though ozone also increased the DLA of the latter. The disease caused by M. betulae increased with increasing leaf age. Ozone enhanced this leaf age effect. In Exp. 2 all fumigation treatments in creased the DLA of clone 4in young and fully expanded leaves, and ozone and C0 2 +0 3 increased it in old leaves. In clone 80 the DLA decreased in all young and fully expanded fumigated leaves, but in old leaves ozone increased the DLA even more than in clone 4. In most cases C0 2 fumigated together with ozone decreased the effect of ozone. Keywords: ozone, carbon dioxide, Betula pendula, Marssonina betulae, leaf disease Introduction The effect of ozone and C02 on pathogens varies between fungus and host species. The effect may be direct or indirect through altered leaf physiology and structure. Ozone has induced, decreased or had no effect on the germination of fungus spores, sporulation and on the growth of mycelium depending on fungus species (Hibben and Stotzky 1969, Treshow et al. 1969, Beare et al. 1999). On mycelium ozone has caused abnormal characters (Hibben and Stotzky 1969). It has also increased the number of germtubes produced (Beare et al. 1999). Ozone may raise host resistance to pathogens through activation of phenylpropanoid pathways and lignin formation (Booker and Miller 1998), but on the other hand plant injuries caused by ozone may favoure some necrotrophs (Manning and Tiedemann 1995). Ozone also accelerates leaf senescence which may have different kind of effects on biotrophs and necrotrophs. Majority of the studied both necrotrophic and biotrophic fungi have gained benefit 108 from increased C02 (Manning and Tiedemann 1995) through changes in leaf nutri ent and water content. There are, however, only a few studies about the effects of ozone and C02 on the diseases of deciduous trees. Very little is known especially about the effects of combined C0 2 +0 3 fumigations on diseases. The aim of this work was to study how increased carbon dioxide, ozone and combined gases affect birch - leaf spot pathogens - interaction. Materials and methods Plant materials used in tests were twenty 9-year-old (year 2000) ozone sensitive (clone 80) and twenty ozone tolerant (clone 4) silver birches (Betulapendula) grow ing at an experimental site at the Suonenjoki Research Station, and 160 micropropagated clonal plantlets. Fungus material which was used in an inoculation experiment (Exp. 2) was Marssonina betulae (Lib.) Sacc. -leaf spot fungus (pure culture provided by Prof. Timo Kurkela; the origin of the isolate was Suonenjoki). Naturally occurring leaf spot fungi in the leaves of experimental trees were utilized for the screening of dis ease development in trees (Exp.l). Experimental trees were fumigated with elevated (2 x ambient) concentra tions of C02, ozone and combined C02 +0 3 in open-top chambers at the experimen tal site during May-September 1999 and 2000. There were two kinds of controls: an 'outside control' without a chamber around the tree and a 'chamber control' with a tree enclosed in the chamber but without any gas treatment. Each of the five treat ments had four replicate trees of both clones. Experiment 1 - Monitoring We monitored the effect of the gas treatments on the development of leaf spots in naturally infected leaves during August-September in 2000. Five leaves with spots in each experimental tree were chosen (20 leaves/treatment), and the test leaves which were left attached to the trees, were photographed with a digital camera three times (7.8., 24.8. and 5.9.2000). The development of spots (amount of spots, % of leaf area) was analysed from the photographs by using a ColAn™-image analyses computer program. Experiment 2 - Inoculation experiment To study the indirect effect of the gases on the disease development we exposed 160 clone plantlets to the five different treatments at the experimental site during June- July 2000. Totally 144 plantlets were placed in the chambers (two plantlets of clone 4 and two plantlets of clone 80 in each chamber) and as outdoor controls were left 16 plantlets (eight plantlets of clone 4 and eight plantlets of clone 80). After 5 weeks we moved the plantlets into a greenhouse and inoculated them with Marssonina betulae. Of the 160 exposed plantlets 140 plantlets were used in the inoculations. Three healthy 109 leaves in each plantlet were sprayed with M. betulae spore suspension (120 000 spores/ml; about 0.8 ml for each leaf) on the abaxial side of a leaf, and one leaf was sprayed with water only to control the effect of inoculation treatment and the amount of natural infection. After spraying the leaves were enclosed in a plastic bag for 5 days to keep the moisture content high for ensuring the infection. Inoculated plantlets were then incubated for 4 weeks in a growth chamber. At the end of the experiment, the test leaves were detached and photographed with a digital camera. The amount of spots (% of leaf area) was analysed by using a ColAn- program. Results Experiment 1 In the monitorings some leaves fell down before last monitoring. Those leaves were usually the most spotted ones. The amount of diseased leaf area (DLA) of fallen leaves is not included in the DLA of monitoring 3. The higher is the number of fallen leaves the higher should also the DLA be, however. In clone 4 the DLA increased most in ozone and C0 2 fumigated leaves (Table 1). The number of fallen leaves was also high in both treatments (7 and 8, respec tively). Also several C0 2 +0 3 fumigated leaves fell down before last monitoring. The DLA, when the small number of fallen leaves is also taken into account, was the smallest in both controls. In clone 80 the highest amount of disease was in ozone fumigated leaves, when fallen leaves are also included. Next highest number of fallen leaves and DLA Table 1. Diseased leaf area, %, in the naturally infected leaves of C0 2 , ozone and C0 2 +0 3 fumigated clone 4 and clone 80 silver birch trees in three different monitorings (n=2o/ fumigation treatment). Column 'fallen leaves' indicates the number of before monitoring 3 fallen leaves, which are not included in the diseased leaf area of monitoring 3. Diseased leaf area, % Fumigation Chamber Outside treatments control C02 03 C02+03 control Clone 4 Monitoring 1 0.76 1.2 1.07 0.85 0.26 stdev 0.93 1.16 1.14 0.95 0.21 Monitoring 2 3.09 5.53 4.62 4 1.15 stdev 5.19 5.9 4.85 5.36 1.91 Monitoring 3 5.01 6.15 7.31 4.86 2.99 stdev 5.87 9.49 4.79 7.47 4.33 Fallen leaves 3 8 7 8 2 Clone 80 Monitoring 1 0.53 0.69 0.41 0.53 0.27 stdev 0.91 1.54 0.52 0.63 0.2 Monitoring 2 2.3 1.5 1.48 1.61 0.86 stdev 2.91 2.03 1.5 2.3 0.69 Monitoring 3 3.01 2.95 2.97 4.01 2.86 stdev 2.59 4.72 2.97 5.1 1.75 Fallen leaves 5 3 7 0 4 110 was in chamber control leaves. The amount of disease was smaller and quite similar in C02+0 3, C0 2 and outside control leaves. Experiment 2 In the control leaves of the inoculation experiment the amount of disease was small (about 0.1% of leaf area) in all other treatments except outside control, where the amount of disease in the control leaves was as high as in Marssonina -inoculated leaves. The severity of disease increased with an increasing leaf age in each treatment and in both clones. The results were therefore interpreted in three different classes of leaf ages: Agel - young, expanding leaves, leaves numbers 1 - 4 from the top of the plantlet; Age 2 - adult, fully expanded leaves, numbers 5-7 from the top of the plantlet; and Age 3 - old leaves, numbers 8-11. In clone 4, in Agel leaves the highest DLA (1.4%) was in C0 2 treated leaves (Fig. 1). In the other treatments the DLA was 0.1 -0.6%. In Age 2 leaves the amount of disease was the highest (1.2 %) in ozone treated leaves and the DLA was slightly higher than in Agel leaves in all other treatments except in C0 2 treated and in out side control leaves. In Age 3 leaves the DLA was twice as high than in younger leaves in outside control, and 4-9 times higher than in Age 2 leaves in the other treatments. In ozone treated leaves the amount of disease was highest, 4.8% and in C02 +0 3 treated leaves next highest, 4.4%, when in chamber control it was 3.0%. Figure 1. Diseased leaf area (DLA), %, caused by Marssonina betulae, in the different leaf age classes of silver birch clonal plantlets 4 and 80 exposed for 5 weeks with C0 2, ozone or C0 2 +0 3 before the inoculation. Bars indicate STDEV. 111 In clone 80, in Age 1 and Age 2 leaves the DLA was highest in chamber control and outside control leaves (0.8 - 1.9%). In Age 3 leaves the DLA was twice as high in ozone treated leaves (6.0%) than in chamber control (2.9%). In C0 2 treated leaves the amount of disease was nearly the same than in chamber control, and in combined gases treatment and outside control slightly lower. Discussion In the monitorings the amount of disease was in both clones almost the same in chamber control, but in outside control it was slightly lower in clone 4. In the inocu lation experiment the DLA caused by M. betulae was smaller in clone 4 than in clone 80 in both controls in young and fully expanded leaves, and in outside control in the old leaves. In the earlier studies made with the same birch clones and Pyrenopeziza betulicola Fuckel -leaf spot fungus, but without any gas treatments, clone 4 has been more resistant to P. betulicola than clone 80 (unpublished data). According to the results of monitorings clone 4 was more affected by the fu migation treatments than clone 80. The DLA and/or the number of fallen leaves in the fumigation treatments was more than twice higher in clone 4 than in the other clone. In the inoculation experiment all fumigation treatments increased the DLA in clone 4 in the young and fully expanded leaves, and ozone and combined gases treatment increased it in the old leaves. In clone 80 the DLA was lower in all young and fully expanded fumigated leaves than in chamber control leaves, but in old leaves the ozone treatment increased the DLA in clone 80 even more than in clone 4. In the fumigated plantlets ozone increased the DLA caused by M. betulae in clone 4 in fully expanded and old leaves, and in clone 80 in old leaves. Also the amount of natural infection and the development of leaf spots was increased in ozone fumigations in both clones. When the clone plantlets were fumigated with ozone combined to C0 2 , the DLA was lower than in ozone alone fumigated leaves in the fully expanded and old leaves of clone 80, and in the young and fully expanded leaves of clone 4. In the monitorings the DLA was lower in both clones when ozone and C02 were fumigated together than alone. Thus C02 seemed to reduce the dis ease increasing effect of ozone. The results of Exp. 1 and Exp. 2 were quite similar. In the monitorings the direct and indirect effects of the fumigations were not possible to separate. The ef fects were similar, however, to the indirect effect in the inoculation experiment. The increasing effect of leaf age on leaf spot disease caused by M. betulae was clear. Leaf age effect varies between host and fungus species. In Populus spp. fully expanded leaves, which were at the state of sink-source transition, were most susceptible to biotrophic fungi (Coleman 1986). Young leaves may be more resistant to fungi, be cause they contain more some harmful chemical compounds than older leaves, like phenols (Coleman 1986, Cline and Neely 1984). On the other hand, young leaves may be more susceptible to fungi, because the leaf structure of soft expanding leaves is not yet as resistant as that of the mature leaves (Spiers and Hopcrofit 1984). In our results ozone fumigation increased leaf age effect. Pre-fumigation of Populus trichocarpa x Populus balsamifera with ozone has also increased significantly the 112 size of lesions caused by Marssonina tremulae on old leaves, but has decreased it significantly on young leaves (Beare et al. 1999). The strong leaf age effect was apparent also in the absence of 03, with larger lesions occurring on older leaves. This effect was enhanced by the ozone. Acknowledgements The staff working in the Open Top Chamber -fumigation field at Suonenjoki Re search Station in summer 2000 is acknowledged. This work was part of Finnish Global Change Research Program (FIGARE) and funded by Finnish Academy of Sciences. References Beare, J.A., Archer, S.A. and Bell, J.N.B. 1999. Marssonina leafspot disease of poplar under elevated ozone: pre-fumigated host and in vitro studies. Environmental pollution 105: 409- 417. Booker, F.L. and Miller, J.E. 1998. Phenylpropanoid metabolism and phenolic composition of soybean (Glycine max (L.) Merr.) leaves following exposure to ozone. Journal of Experimental Botany 49: 1191-1202. Cline, S. and Neely, D. 1984. Relationship between juvenile-leaf resistance to anthracnose and the presence of juglone and hydrojuglone glucoside in black walnut. Phytopathology 74: 185-188. Coleman, J.S. 1986. Leaf development and leaf stress: increased susceptibility associated with sink-source transition. Tree Physiology 2: 289-299. Hibben, C. R. and Stotzky, G. 1969. Effects of ozone on the germination of fungus spores. Can. J. Microbiol. 15: 1187-1196. Manning, W. J. and Tiedemann, A. V. 1995. Climate change: Potential effects of increased atmospheric carbon dioxide (C0 2 ), ozone (0 3 ) and ultraviolet-B (UV-B) radiation on plant diseases. Environmental Pollution 88: 219-245. Spiers, A.G. and Hopcroft, D.H. 1984. Influence of leaf age, leaf structure and frequency of stomata on the susceptibility of poplar cultivars to Marssonina brunnea. Eur. J. For. Path. 14: 270- 282. Treshow, M., Harner, F. M., Price, H. E. and Kormelink, J. R. 1969. Effects of ozone on growth, lipid metabolism, and sporulation of fungi. Phytopathology 59: 1223-1225. 113 Occurrence, host range and impact of leaf pathogen fungi on forest trees in Hungary Ilona Szabö University of West- Hungary, Institute of Forest- and Wood Protection, H-9400 Sopron, P. O. Box. 132, Hungary. E-mail: szaboi@emk.nyme.hu Abstract The results of a recent inventory of leaf and shoot diseases of forest trees in Hungary are pre sented. During the last few years field investigations were executed in the main forest regions of the country, leaves with symptoms were collected, the pathogens identified and occasionally iso lated in pure culture. All species of forest trees and shrubs presenting leaf symptoms were sur veyed. About 150 species of leaf pathogenic fungi were identified on more than one hundred tree and shrub species. Occurrence of several leaf pathogens has been recorded for the first time in Hungary: Asteromella tiliae, Cristulariella depraedans, Dothistroma septospora, Drepanopeziza sphaeroides, Glomerella miyabeana, Hadrotrichum dryophylum, Meria laricis, Microsphaera vanbruntiana, Oidium carpini, Phaeocryptopus gauemannii, Phloeospora associata, Phyllosticta concentrica, Phyllosticta globulosa, Phyllosticta hypoglossi, Phyllosticta minima, Septogloeum carthusianum, Thedgonia ligustrina, Tubakia dryina, Uromyces laburni. A. tataricum and A. saccharinum are first reported as hosts of Cristulariella depraedans and Quercus pubescens as host of Tubakia dryina in Europe. Keywords: leaf pathogen fungi, forest trees Introduction The rate of forestation in Hungary amounts to about 19%. The majority of the forests lies under 350 m altitude in 9 main forest regions in continental, subatlantic and submediterranean climate conditions. As species composition the hardwoods are dominating (85.5%), represented primarily by oaks, black locust, beech and hybrid poplars. Among the conifers mainly Scotch and Austrian pines have a considerable forest surface. The leaf diseases of hardwoods caused by fungi do not occur generally in mass of economic importance in stands, although epidemics of certain species as Microsphaera alphitoides Griffon et Maubl. on oaks, Melampsora rusts and Marssonina leaf disease of hybrid poplars are to be observed regularly. On conifers Sphaeropsis blight and Dothistroma red band disease are the economically impor tant fungal leaf diseases. Although a few species cause sensible damages for the forestry, the whole spectrum of the leaf fungi of forest trees is of interest in phy topathological aspect. 114 Material and method Selected forest compartments representing all forest types and tree species were surveyed annually for occurrence of leaf disease symptoms in each mean forest re gion of the country. The frequency of the symptoms was estimated and pathological material collected. The field investigations were carried out at the start and middle to late growing season to reach both the early and the late symptoms of foliage diseases. The collected material was examined for identification of causal pathogens. The identification was done on fresh material in case of presence of fungal fructifications. For stimulation of the sporulation wet chamber method was used for a few days after surface sterilisation of the plant material with 2 g/1 active chlorine containing NaOCl solution. In case of lack of fructifications the pathogens were isolated from symptomatic tissues on PDA and identified on the cultural characters and occasionally sporulation in the pure culture. Results and discussions The identified pathogens, caused symptoms and importance of the diseases are dis cussed in order of main wood species for the forestry in Hungary. Deciduous trees Oaks are the most important forest trees in Hungary as occupied territory and grow ing stock as well. The main species are the sessile (Quercus robur), pedunculate (Q. petraea) and Turkey (Q. cerris) oaks covering together about 33% of the forest sur face. The downy oak (Q. pubescens) and the introduced red oak (Q. rubra) have a low rate of 1-2%. The powdery mildew caused by Microsphaera alphitoides is the most frequent foliage disease of oaks. It is übiquitous especially on sessile oak, not rare on pedunculate and downy oaks. Mycrosphaera hypophylla Nevodovskij also was found rarely on Q. robur. Apiognomonia quercina (Kleb.) Höhn. (anamorphe Discula quercina AVestend./Arx) was found on sessile, pedunculate and downy oaks. As endophyte species its symptoms are not common in forest stands. Sometimes occur together with powdery mildew on sessile oak or frequent in association with foliar galls on other oaks. Septoria quercicola (Desm.) Sacc. was epidemical on pedunculate oak in 1999 in some regions of SW Hungary. It causes small, 1-3 mm, round brown foliar spots. Microstroma album (Desm.) Sacc. was found on peduncu late and Turkey oaks. Its characteristic white colonies appear on the lower surface of the leaves. On the upper surface confluent yellow, later necrotic points can be ob served. The disease was epidemical in some places but without considerable effect on the healthy status of the trees. Tubakia dryina (Sacc.) Sutton caused symptoms on pedunculate, downy and Turkey oaks. On Q. petraea the round, 5-15 mm sized, brown spots appeared around the suck points of Phylloxera aphids. On Q. pubescens the disease was epidemical in some places, and the round, concentric zoned brown spots extended finally on large leaf surfaces. The pycnidia were observed in concen 115 trie circles epi- and hypophyllous. On Turkey oak small, 2-4 mm, round brown spots could be observed without fungal fructifications even after wet chamber incubation. The causal agent was identified after the isolation of the fungus, based on the cul tural characters. In culture the fungus produces abundantly pyenidia with character istic microconidia. T. dryina is associated with leaf spots of different oak species in Europe and Northern America. In Europe it is known primarily on Q. petraea (Butin 1996), than on Q. rubra and Q. cerris (Belisario 1993). It is frequently isolated as an endophyte of Q. petraea and Q. robur (Halmschlager pers. Com.). Its occurrence on downy oak has not been reported yet, so this is the first report of this pathogen on Q. pubescens. Further identified foliar pathogens Hadrotrychum dryophylum Sacc. on pedunculate, Phloeospora associata Bubäk on downy, and Phyllosticta globulosa Thiim. on sessile oak proved to be rare, each were found at one occasion only. These three species as well as Tubakia dryina have been first recorded in Hungary (Szabo 2000). Beech (Fagus sylvatica) has an amount of 6.3% in the forest surface. The anthracnose disease caused by Apiognomonia errabunda (Rob.) Höhn. (anamorphe Discula umbrinella /Berk et Br./Sutton ) is not rare in Hungary. In some years it is epidemical in stands. In the nurseries it is common also. Frequent is associated with galls (Mikiola fagi) or other insect damages (Rhynchaenus fagi). Other foliage fungi of beech appeared rarely, for example the powdery mildew Phyllactinia guttata (Wallr ex Fr.) Lev. late summer, without considerable importance. Black locust (Robinia pseudoacacia), is one of the most important forest trees, having a rate of nearly 21%. Phloeospora robiniae (Lib.)Höhn. is the most frequent foliage pathogen of black locust (Szabo 1993). Generally spread, it causes brown, necrotic spots and early defoliation, especially of the upper crown. Microsphaera pseudacaciae (Marczenko) U. Braun occurs mostly on young plants in nurseries and on sprouts, without importance in stands. The sporadically found Ectostroma robiniae (uncertain taxon) causes small, 1-3 mm, round black stromata without fungal fructifications. The isolation essay also was unsuccessfully. Hornbeam (Carpinus betulus) is wide-spread admixed forest tree species. Its rate amounts nearly 6%. Asteroma carpini (Lib.)Sutton is the most common foliar fungus of hornbeam. Its symptoms appear on senescent leaves late summer. The other, less frequent observed leaf pathogen is Monostichella robergei (Desm.)Höhn. This one is more pathogen, causes extended, grey-light brown leaf necrosis. Its oc currence is not general, but local appears epidemically, frequent in association with foliar damages caused by drought. Oidium carpini Foitzik, the recently described powdery mildew of hornbeam (U. Braun 1995) was found in 1999 for the first time in Hungary. The rate of ash species (Fraxinus excelsior, F. ornus, F. pennsilvanica) is about 2%. The appearance of their leaf diseases was not frequent in the years of investiga tions. The followings are worthy of mention: powdery mildew Phyllactinia fraxini (DC.)Homma, than Phomopsis pterophila (Nits.) Died. and Phoma macrostoma Mont, on F. excelsior, Phyllosticta omi Bubäk on F. ornus. The Phloeospora state of Mycosphaerella fraxini Niessl. was found for the first time in Hungary on F. ornus. Maples (Acer campestre, A. platanoides, A. pseudoplatanus, A. tataricum, A. negundo) occur in all type of forest having a rate of about 1% of the forest surface. 116 The white leaf spot of maples caused by Cristulariella depraedans (Cooke)Höhn. was found first in Hungary in 1992 on Acer saccharinum (Szabo 1994), then in 1998-99 on A. campestre, A. pseudoplatanus and A. tataricum. The pathogen is known in Europe preponderantly on A. pseudoplatanus (Butin 1981), but has also been reported on A. platanoides and mentioned on A. campestre among the 21 different woody and herbaceous hosts (Lang 2000). Its occurrence on A. tataricum and A. saccharinum in Europe is first reported in this paper. Tar spot disease caused by Rhytisma acerinum (Pers.)Fr. is generally spread, especially on A. pseudoplatanus but it is not rare also on A. campestre and A. platanoides. The powdery mildews of maples, Sawadaea tulasney (Fuck.)Homma and S. bicornis (Wallr. ex Fr.)Homma were found frequent on maples. The first is frequent on A. campestre, A. negundo, A. platanoides, A. pseudoplatanus and A. tataricum, the second on A. campestre, A. platanoides and A. pseudoplatanus. Phyllosticta minima (Berk et Curt.)Underw. et Earle is a foliar pathogen recorded newly in Hungary (Szabö 1997 a, 1997b). It is a true Phyllosticta with typical conidia bearing gelatinous appendage and spermatia of Leptodothiorella type (Van der Aa 1973). In Germany was found rarely on A. pseudoplatanus (Wulf 1994). In Hungary occurs on A. pseudoplatanus, A. campestre and A. negundo. Phyllosticta aceris (Lib.)Sacc. and P. negundinis Sacc. et Speg. were found on Acer campestre and A. negundo respecivelly. Mycosphaerella latebrosa (Cooke)Schroet. (anamorphe Phloeospora aceris /Lib./Sacc) causes small, point shaped leaf spots, premature yellowing and falling of the leaves. It was found on A. campestre and A. pseudoplatanus. Diplodina acerina (Pass.)Sutton was found on A. pseudoplatanus and A. tataricum, Didymosporina aceris (Lib.)Höhn. was epidemical on A. campestre in 1995, Discula campestris (Pass.)Arx occurs sporadically on A. campestre. The elm species (Ulmus glabra, U. laevis, U. minor, U. procera, U. pumila var. arborea) have a low rate of less than 1%. Mycosphaerella ulmi Kleb. (anamorphe Phloeospora ulmi /Fr. ex Kunze/Wallr.) is generally spread on all elm species. The pink mass of conidia spread from hypophyllous acervuli is characteristic. Platychora ulmi (Schleich. et Duval)Petr. is wide-spread biotrophic pathogen on U. minor and U. pumila var. arborea. Its black stromata are remarkable on the upper surface of the leaves. Hybrid poplars, especially Populus x euramericana are important cultivated trees in Hungary, their rate amounts to 6,6%. The autochthonous poplar species {Populus alba, P. x canescens, P. tremula, P. nigra,) cover nearly 3% of the forest area.On hybrid poplars Melampsora larici-populina Kleb. is wide spread. M. allii populina Kleb. was found less frequent in stands, sporadically on P. nigra in flood sites. On white poplars M. pinitorqua Rostr. and some other Melampsora species occur, but the confirmation of their identification carried out on the uredospore morpholpgy only is necessary. The most frequent occurring Drepanopeziza species on hybrid poplars is D. punctiformis Gremmen (anamorphe Marssonina brunnea / Ellis et Everh./Magn.). D. populi-albae (Kleb.)Nannf. (anamorphe Marssonina castagnei /Desm. et Mont./Magn.) is also not rare on white and grey poplars in lowland forests. D. populorum (Desm.)Höhn. (anamorphe M. populi /Lib./Magn.) was found only sporadically on P. nigra. Venturia macularis (Fr.) E.Muller et Arx 117 (anamorphe Pollaccia radiosa /Lib./ Bald. et Cif.) is frequent on white and grey pop lars as on aspen causing not only leaf disease, but also top and bark necrosis of shoots. Its impact is of considerable importance in nurseries. Mycosphaerellapopuli (Auersw.)Schrot. (anamorphe Septoria populi Desm.) causes small 2-3 mm sized, dark limited spots. It is frequent especially on Populus nigra 'ltalica'. Asteroma frondicola (Fr. ex Ficinus et Schubert) Morelet is frequent in flood plain causing extended, roundish, grey leaf spots with concentrically arranged, epiphyllous acer vuli on P. alba and P. x canescens. More then ten Salix species occur naturally in Hungary. The investigations covered mostly the tree shaped willows, Salix. alba, S. fragilis (about 1% of the forest surface) and S. caprea. Melampsora rusts occur frequently on willows (M allii-salicisalbae Kleb., M. caprearum Thtim., M. euonymi-caprearum Kleb. M. galanthi-fragilis Kleb. etc.) Their identification on uredospores morphology only needs confirmation. Drepanopeziza salicis (Tul. et C.Tul.)Höhn (anamorphe Monostichella salicis /Westend./Arx) is frequent, especially on S. fragilis, causing round dark brown, confluent epiphyllous spots. Drepanopeziza sphaeroides (Pers.)Höhn (anamorphe Marssonin salicicola /Bres./Magn.) occurs on weeping willow (S. alba var. vitellina pendula). It causes spotting and premature yellowing of leaves, top necrosis and bark crusting of shoots. Glomerella miyabeana (Fuckel)Arx (anamorphe Colletotrichum gloeosporioides Penz.) was found on aspen and white willow. These two latest pathogens were identified for the first time in Hungary in 1991 (Szabo 1992). Venturia species (V. chlorospora/Ces./Aderho\d) were not found frequent either in willow stands or in riparian associations. Phyllosticta salicicola Thtim. causes sporadically brown spots on S. alba cultivars. Common alder (Alnus glutinosa) is cultivated on ca 2% of the forest surface. It occurs also in riparian associations. Leaf blister caused by Taphrina tosquinetii (Wesend.)Tul. is not rare on young trees and on sprouts. T. sadebeckii Johanson was found on riparian alder of different age. Further leaf pathogens as Gnomoniella tubiformis (Tode)Sacc. (anamorphe Asteroma alneum /Pers. ex Fr./Sutton) and Discula sp. were found sporadically. Limes (Tiliä cordata, T. platyphyllos, T. tomentosa) amount less than 1% of the forest surface. Cercospora microsora Sacc. is a frequent leaf fungus of T. cordata causing small, 1-2 mm, dark spots, premature yellowing and falling of the leaves. It occurs also on T. tomentosa but more rarely and the symptoms are less intensive. Apiognomonia tiliae (Rehm)Höhn (anamorphe Discula sp.) appears early after bud ding, causing leaf- and shoot anthracnose. It is frequent on T. cordata, sporadic on T. tomentosa. At the end of the growing season the conspicuous symptoms of leaf blotch caused by Asteromella tiliae (Rud.)Butin et Kehr appear frequently, especially on T. platyphyllos, but also on the two other linden. Formerly the symptom was connected to Asteroma tiliae Rud.. The pathogen was recently correct described and its con nection to Didymosphaeriapetrakiana Sacc. demonstrated (Butin et Kehr 1995). Its occurrence in Hungary has been recently recorded (Szabo 1997b). On silver birch (Betula verrucosa) the powdery mildew Phyllactinia guttata (Wallr. ex Fr.)Lev. is frequent late summer, autumn as loose, hypophyllous texture. Discula betulina (Westend.)Arx causes brown spots and premature leaf fall. The grey, round spots of Venturia ditricha (Fr.)Karsten (anamorphe Fusicladium betulae 118 Aderhold) appear late summer. Birch rust (Melampsoridium betulinum /Pers./Kleb.) is epidemical in some years, as it was in 1999. Kabatiella apocrypta (Ellis et Everh.)Arx proved to be a general spread übiquitous leaf pathogen. It was found in leaf necrosis on many different forest trees (Acer, Alnus, Betula, Cerasus, Fagus, Populus, Salix etc.). The shrub layer of the deciduous forests is rich in species. More then 50 leaf pathogen fungi were identified on about 30 shrub species. Only a few fungi from among the first recorded ones in Hungary are enumerated there: Microsphaera vanbruntiana Gerard on Sambucus racemosa, Phyllosticta hypoglossi (Mont.)Allesch. on Ruscus aculeatus, Septogloeum carthusianum Sacc. on Euonymus europaeus, Thedgonia ligustrina (Boerema)Sutton on Ligustrum vulgare, Uromyces laburni (DC.)Fuckel on Laburnum anagyroides. Conifers Pinus sylvestris and P. nigra are planted on poor sites and in lowland forestations. Their rate amounts to about 13%. Several pathogens cause important leaf- and shoot diseases of these tree species. In nurseries and young forestations Lophodermium seditiosum Minter, Staley et Millar causes needle cast of seedlings and young trees. L. pinastri (Schrad.)Chev. is also present as week pathogen of aged needles. Sphaeropsis sapinea (Fr.)Dyco et Sutton causes shoot blight especially on P. nigra, but it is present also on P. sylvestris. This disease appeared in mass during 1980s causes serious damages especially in association with other factors as drought, poor nutrition and some week pathogens as Cenangium ferruginosum Fr.. Mycosphaerella pini E.Rostrup (anamorphe Dothistroma septospora /Dorog./Morelet)) was first iden tified in 1991 (Szabö 1997 c). Late 1990s the epidemic and damages caused by this fungus have been generalized in young forests of P. nigra (Koltay 2001). Other iden tified needle pathogens of pines are Sclerophoma pithyophila (Corda)Höhn, Lophodermella conjuncta Darker, Coleosporium species and some week pathogens in old needles as Cyclaneusma niveum (Pers.)DiCosmo, Peredo et Minter and Cytosporapinastri Fr.. The leaf pathogen of larch (Larix decidua) is Mycosphaerella laricina (Hartig)Neger. It causes intensive leaf fall of young trees some years. In the nurser ies Meria laricis Vuill. has been identified first time in Hungary (Szabo 1999). The spruce (Picea abies) needles are sporadically attacked by Lirula macrospora (Hartig)Darker. The shoots and young needles of spruce and larch some times are infected by Botrytis cinerea Pers. and Sclerophoma pythiophila. Douglas fir (Pseudotsuga menziesii) is affected by Phaeocryptopus gauemannii (Rhode)Petrak and Rhabdocline pseudotsugae Sydow. P. gauemannii was newly recorded in Hungary. In the last years it seems to become generalized in Western Hungary. On the needles of different conifers in arboretums and Christmas tree planta tions some further pathogens were identified as Chrysomyxa abietis (Wallr.)Unger and Rhizosphaera kalkhoffii Bubäk on blue spruce, Lirula nervisequia (DC)Darker and Cytosporafriesii Sacc. on fir. The old needles of Taxus baccata were infected by 119 Phyllosticta concentrica Sacc.. This fungus has been reported recently in Hungary (Szabö 1997 a). Acknowledgement This research was supported by the OTKA grant T 025173. References Aa, H.A. van der 1973. Studies in Phyllosticta I. Studies in Mycology No. 5. Belisario, A. 1993. Ist report of Tubakia dryina on Quercus cerris. Plant Disease 77 (6) 647. Braun, U. 1995. The Powdery Mildews (Erysiphales ) of Europe. Gustav Fischer Verlag Jena Stuttgart New York. Butin, H. 1981. Die Weissfleckigkeit des Bergahorns - eine 'neue' Blattkrankheit. Allgem. Forstz. 36, 327-328. Butin. H. and Kehr, R. 1995. Leaf blotch of lime associated with Asteromella tiliae comb. Nov. and the latter's connection to Didymosphaeriapetrakiana. Mycol. Res. 99 (10) 1191-1194. Butin, H. 1996. Krankheiten der Wald- und Parkbaume. Georg Thieme Verlag Stuttgart New York. Koltay A. 2001. Incidence of Dothistroma septospora (Dorog.)Morlet in the Austrian pine (Pinus nigra Arn.) stands in Hungary and results of chemical control trials, (in Hungarian) Növenyvedelem 37 (5) 231-235. Lang, K. J. 2000. New hosts of Cristulariella depraedans. For. Path. 30: 117-120. Szabö, I. 1992. Fungi causing leaf spot symptoms and shoot destruction on willows, (in Hungar ian) Növenyvedelem 28 (7-8) 295-300. Szabö, I. 1993. On the leaf-spot disease of locust trees (Robinia pseudoacacia L.). (in Hungarian) Növenyvedelem 29 (11) 527-529. Szabö, I. 1994. Conidial fungi causing leaf diseases of broad-leaved forest trees and shrubs. 40th Plant Protection Days, Budapest, 22-23. Feb. 1994. p. 132. (in Hungarian). Szabö, I. 1997 a. Phyllosticta species on woody plants in the Arboretum of the University of Sopron. 43th Plant Protection Days, Budapest, 24-25. Feb. 1997. p. 126. (in Hungarian) Szabö, I. 1997b. Some foliage necrosis causing Coelomycetes on broad leaves forest trees and shrubs in the surroundings of Sopron, Hungary. Acta Phytopathologica et Entomologica Hungarica 32 (1-2) 69-78. Szabö, I. 1997 c. Occurrence of Dothistroma septospora (Dorog)Morelet in black pine planta tions. (in Hungarian) Erdeszeti Lapok 132 (2) 44. Szabö, I. 1999. On the fungi causing needle cast of larch. 45th Plant Protection Days, Budapest, 23-24. Feb. 1999. p. 124. (in Hungarian) Szabö, I. 2000. Fungi causing leaf diseases of oaks. 46th Plant Protection Days, Budapest, 22-23. Feb. 2000. p. 120. (in Hungarian) Wulf, A. 1994. Pilzbedingte Blattkrankheiten an Ahorn. Schriften aus der Forstl. Fac. Uni. Göttingen und Niedersachs. Forstl. Versuchsanst. Band 116. J.D. Sauerlander's Verlag Frankfurt am Main. 120 Flammulina velutipes (curt). Fr. Sing, population structure Maryna M. Sukhomlyn Biological faculty of Donetsk National University, Schorsa str. 46, Donetsk 83050, Ukraine. E-mail: marina@bio.donetsk.ua Abstract The first three stages of life cycle of wood-eating Basidiomycetes: germination of basidiospores, haploid homokaryotic mycelium and fertile mycelium, in which the nuclei are associated as spe cialized heterokaryon - dikaryon are very important in the plan of research of infection biology of fungi, protective mechanisms of plant and of the increases it resistance to diseases. The study of population structure of separate species fungi enables to understand mechanisms of its develop ment in wood. With this purpose we conduct a research of the genetic status of fruit bodies of fungus Flammulina velutipes, which developed on same plant - host. The distribution of the factors of compatibility at the fruit bodies of F. velutipes assumes initial development of one dikaryon and availability of numerous spore infection. The comparison of the genetic status of fruit bodies assumes distribution of infection with spores, development homokaryotic mycelium and limited distribution of clones. Rather high factor of outbreeding is marked. The researches of joint cultivation dikaryons, monokaryons, both di- and monokaryons in pure culture have allowed to trace features of vegetative compatibility at this species. 121 Bacterial canker of Maackia amurensis var. buergeri - Occurrence and pathological anatomy Yasuaki Sakamoto, Yuichi Takikawa, Yuko Takao and Katsuhiko Sasaki Y. Sakamoto, K. Sasaki, Hokkaido Research Center, Forestry & Forest Products Research Institute (FFPRI), Sapporo 062-8516, Japan. E-mail: yasusaka@ffpri.affrc.go.jp Y. Takikawa, Faculty of Agriculture, Shizuoka University, Shizuoka 422-8529, Japan. Abstract A new disease of Maackia amurensis var. buergeri was recently found on the northern island of Hokkaido, Japan. Affected trees were heavily damaged and had cankers on both trunks and branches. After natural infection, a series of swellings on the bark surface developed longitudinally. These swellings burst and coalesced to become long cankers with exposed woods. The causal pathogen was isolated and characterized as Pseudomonas syringae on the basis of laboratory tests. Patho genicity of the bacterium was confirmed by inoculation into the host. Hence, we propose that the disease be designated "bacterial canker of M. amurensis". Anatomical observations to elucidate the process of the canker formation were carried out. First, abnormalities appeared near the cambial zone. Xylem formation was suppressed, hyperplasia of irregular ray parenchyma cells occurred, then the swellings appeared. The irregular cells increased in number; after that, the surface of the swollen bark burst open. The ruptured swellings coalesced to form long irregular cankers. Keywords: Bacterial canker, Maackia amurensis var. buergeri, Pseudomonas syringae, anatomy, canker formation Introduction Maackia amurensis Rupr. et Maxim var. buergeri Schn. (hereafter referred to as M. amurensis) is a leguminous deciduous tree. In Japan, it is mainly found on the is lands of Hokkaido, Honshu and Shikoku. The wood of M. amurensis exhibits a beau tiful combination of yellowish white sapwood and dark brown heartwood and is very durable. Recently, a major outbreak of a new canker disease of M. amurensis was reported in two plantations at Kawayu in Teshikaga city (eastern Hokkaido) and Higashiyama in Furano city (central Hokkaido). Some trees were also affected in natural forests in Kimobetsu city (southern Hokkaido), Chitose and Sapporo city (central Hokkaido), in the garden of "Yotei seinen no mori" Forest Park in Makkari city (south Hokkaido) and in roadside trees in Sapporo and Tokoro (eastern Hokkaido). The bacterium was isolated from the affected tissues and its pathogenicity to M. amurensis was confirmed. The bacterium was characterized as Pseudomonas syringae by its bacteriological characteristics. 122 This report deals with the description of the symptoms, the isolation of the pathogen and its characterization and pathogenicity tests, then proposes the name of this disease as "bacterial canker of M. amurensis" caused by P. syringae (Sakamoto et al. 2000). There is also an anatomical study of the disease development (Sakamoto 1999). Symptoms Symptoms were observed regularly from 1993 to 1997 in the Higashiyama stand. Cankers were observed on trunks and branches alike. On trunks, they grew longitu dinally (Fig. 1). The initial symptom developed from late spring to early summer as a series of longitudinal swellings (approximately 3 x 1.5 cm) on the surface of the bark, which were interspersed with slightly swollen bark (hereafter referred to as SSB). These swellings gradually burst open and coalesced to form longitudinal cracks (Figs. 2a, b). These long cankers on the trunks sometimes spread horizontally and girdled the trunks. The tissues of the affected inner bark were observed to crack and break off easily. Finally, brownish black, dead inner bark tissues tore and fell off, exposing the xylem tissues (Fig. 3). Figure 3. Transverse view of the canker with exposed wood. Figure 1. Serious canker of M. amurensis (photographed in Higashiyama). 123 Figure 2. Development of natural symptoms in Higashiyama stand. Arrows indicate the same site a. Photographed at 10 August 1993. b. Photographed at 6 July 1995. Material and methods Isolation of the pathogen Small pieces of inner bark tissue (approximately 5x5x5 mm) were excised from swollen, water-soaked areas in the swellings and from the margins of cankers. Each piece was macerated in 5 ml of sterile peptone water (1%). The resulting suspen sions were streaked on plates of nutrient agar (NA: Eiken E-MC01) and incubated at 20° C. After 2-3 days, single colonies were re-streaked on fresh NA plates to ensure purity. Characterization of the pathogen The pathogen was characterized based on their colony morphology, cell morphology and bacteriological characteristics (Sakamoto et al. 2000). 124 Pathogenicity Pathogenicity tests were performed on 7-year-old seedlings of M. amurensis in 1994. Holes were bored into the xylem of the trunks of seedlings with a cork borer (5 mm in diameter), then filled with a mass of bacterial cells from NA plates and sealed with vinyl tape. The vinyl tape was removed 2-3 weeks after inoculation. Anatomy Eighteen affected and five trees were selected. Samples of the SSB, the swellings, the margins of cankers and the healthy tissues (approximately 1.5x1.5x1.5 cm, in cluding phloem and xylem) were fixed with FAA (formalin: acetic acid: 50% etha nol = 5:5:90). After washing under running water for 3-4 hours, they were dehy drated in ethanol series and embedded in 10% celloidin (Toyokuni Chemical Co.). Transverse, tangential and radial sections of 20-25 (im in thickness were cut on a sliding microtome. They were stained with safranin 0(1% solution in 50% ethanol) and fast green (0.1% solution in 95% ethanol), and were then examined under a light microscope. Results Isolation of the pathogen The same bacterium was regularly isolated from all the affected freshly tissues. The colonies were 1-2 mm in diameter on NA after 3 days at 20° C, with an entire margin, and were circular, convex, smooth and glistening. Characterization of the pathogen The current isolates were Gram-negative, non-sporing, straight rods, motile with one to three flagella, produced a white to cream glistening growth, metabolized glu cose oxidatively, did not grow at 40° C, nor did they accumulate poly-B-hydroxy butyrate granules. Other results are shown in Table 1. Pathogenicity The inoculated inner bark tissues began to proliferate vigorously in the vicinity of the inoculation points after approximately two weeks. The bark became cracked and the symptoms continued to proliferate until the end of September 1994. At the end of April 1995, the symptoms started to develop again. Several swelling (Fig. 4a) and noticeable cankers (Fig. 4b) were observed. 125 Table 1. Characteristics of the current isolates and reference isolates. a = Author's data, b = Data from Takikawa et al. (1989). + = positive, - = negative, K = alkaline reaction, D = digestion, w= weak reaction * = No visible change was observed, V= Variable results among the isolates. Characteristics Isolated strains P. syringae pv. syringae a) P. syringae pv. myricae a) P. syringae pv. tremae P. syringae pv. actinidiae b -1 Fluorescent pigment on King's B medium - + - - Hydrolysis of Aescrin - + - - - Starch - - - - Tween 80 + + + + + Levan production V + + - + Activity of Tyrosinase - - + + - Urease - + - - Lecithinase - - - - - Reaction in purple milk .*) KD /) .*) Kw Gluconate oxidation - - - - H2S production - - - - Indole test - - - Liquefaction of Pectin - - - - Gelatin - + - - - Reducing substances from sucrose + + + + + V.P.-M.R. test - - - - - Arginine dihydrolase - - - - - Utilization of Glucose + + + + + Sucrose + + + + + D-Galactose + + + + + Fructose + + + + + D-Ribose + + + + + D-Arabinose - - - - - Lactose - - - - Maltose - - - - - L-Arginine + + + + + Gluconate + + + + + Citrate + + + + + L-Malate + + + + + L-Tyrosine - - - + Formate - - - - - Propionate - - - - - - - - - 126 Figure 4. Symptoms of bacterial canker of M. amurensis on inoculated seedling. a. Swellings (big arrows) which developed after artificial inoculation (date of inoculation: 23 June 1994; date of photography: 22 June 1995). Small arrow indicates the inoculation site. b. Canker which formed after artificial inoculation (date of inoculation: 15 July 1994; date of photography: 31 August 1995). Arrow indicates the inoculation site. Anatomy In the tissues of the SSB far from the swelling, the cambial zone was clearly ob served. The most remarkable difference between healthy tissues and SSB was a sup pressed lignification of the xylem tissues in inner areas of the cambial zone. In these areas, vessel formation was also suppressed (Figs. 5, 6). In the SSB near swelling, the cambial zone became fuzzy or disappeared (Fig. 7). In the swellings, hyperplasia of irregular-shaped parenchyma cells was com monly observed. Masses of bacteria were observed in intercellular spaces of the irregular ray cells. The irregular-shaped parenchyma cells increased in number with the development of the disease until the cambial zone had completely disappeared. In the phloem tissues, the bacteria degraded the parenchyma cells, and then the bac terial lesions (the areas where the mixture of degraded parenchyma cells and bacte ria was observed) became more and more numerous (Fig. 8). These areas corre spond to the stained spots in the macroscopic transverse views of the swellings (Fig. 9). The irregular-shaped parenchyma cells increased in number with degradation, the bacterial lesions got bigger and bigger, then the phloem tissues got thicker and thicker. 127 Figure 5. Transverse view of the healthy tissue. Arrows indicate the cambial zone. Figure 6. Transverse view of the SSB far from the swelling. Arrows indicate the cambial zone. Figure 7. Transverse view of the SSB near the swelling. Arrows indicate the cambial zone. 128 Figure 8. Transverse view of the swelling. Arrows indicate the bacterial lesions. Figure 9. Transverse view of the swelling. Arrow indicates the stained spot. Figure 10. Transverse view of the ruptured swellings. W indicates the wound periderm and B indicates bacterial lesion. 129 In the tissues of ruptured swellings, wound periderms were formed to protect the living cells in the cortex and phloem. However, bacterial lesions expanded to the cortex and phloem with the development of the disease, and coalesced with the wound periderm (Fig. 10). The coalesced tissues were observed to crack and break off easily. Discussion Based on the bacteriological characteristics mentioned above, the current isolates should be included in the genus Pseudomonas (Palleroni 1984). Other characteris tics of the current isolates were very similar to the characteristics of P. syringae pathovars that were pathogenic to woody plants, such as P. syringae pv. syringae (Tables 1, Young 1991), P. s. pv. myricae (Tables 1, Ogimi and Higuchi 1981), P. s. pv. dendropanacis (Ogimi et al. 1988 a), P. s. pv. tremae (Tables 1, Ogimi et al. 1988b), P. s. pv. actinidiae (Tables 1, Takikawa et al. 1989) and P. s. pv. daphniphylli (Ogimi et al. 1990). Therefore, the isolates were classified as a member of P. syringae (Sakamoto et al. 2000). Bacterial canker of M. amurensis has not been previously reported in the world. Hence, the authors proposes the common name "bacterial canker of M. amurensis'''' caused by P. syringae (Sakamoto et al. 2000). The results of the current research clarified the anatomical characteristics of the development of the disease as follows. The first anatomical abnormalities of the disease appeared near the cambial zone as the suppression of normal xylem forma tion. Then, the hyperplasia began to form irregular-shaped parenchyma cells, and caused several swellings and S SB on the bark. The irregular cells increased in number with degradation by the bacteria, and the surface of swollen bark burst. The cracks of the swellings coalesced to form long irregular cankers. This study revealed the pathogens and mechanisms of disease development of the canker disease. It presents many important suggestions for pathological research of other bacterial cankers and galls. References Ogimi, C. and Higuchi, H. 1981. Bacterial gall ofYamamomo (Myrica rubra S. et Z.) caused by Pseudomonas syringae pv. myricae pv. Nov. Annals of the Phytopathological Society of Japan 47: 443-448 (in Japanese with English summary). Ogimi, C., Higuchi,H. and Takikawa, Y. 1988 a. Disease of Kakuremino (Dendropanax trifidus Mak.) caused by Pseudomonas syringae pv. dendropanacis pv. Nov. Annals of the Phytopathological Society of Japan 54: 296-302. (In Japanese with English summary.) Ogimi, C., Higuchi, H. and Takikawa, Y. 1988 b. Bacterial gall disease of Urajiroenoki (Trema orientalis BL.) caused by Pseudomonas syringae pv. tremae pv. Nov. Journal of the Japanese Forestry Society 70: 441-446. (In Japanese with English summary.) Ogimi, C., Kubo, Y., Higuchi, H. and Takikawa, Y. 1990. Bacterial gall disease of Himeyuzuriha (Daphniphyllum teijsmanni Z.) caused by Pseudomonas syringae pv. daphniphylli pv. Nov. Journal of the Japanese Forestry Society 72: 17-22 (in Japanese with English summary). 130 Palleroni, N.J. 1984. Pseudomonas Migula 1894. In: N.R. Krieg and Holt, J.G. (eds.) Bergey's Manual of Systematic Bacteriology Vol. 1. William & Wilkins Company, Baltimore, p. 141- 199. Sakamoto, Y. 1999. Anatomy of Bacterial Canker on Maackia amurensis var. burgeri. Journal of Forest Research 4: 281-285. Sakamoto, Y., Takikawa, Y., Takao, Y. and Sasaki, K. 2000. Bacterial canker of Maackia amurensis var. buergeri caused by a putative Pseudomonas syringae. Forest Pathology 30: 19-28. Takikawa, Y., Serizawa, S., Ichikawa, T., Tsuyumu, S. and Goto, M. 1989. Pseudomonas syringae pv. actinidiae pv. Nov.: The causal bacterium of canker of kiwifruit in Japan. Annals of the Phytopathological Society of Japan 55: 437-444. Young, J.M. 1991. Pathogenicity and identification of the lilac pathogen, Pseudomonas syringae pv. syringae van Hall 1902. Annals of Applied Biology 118: 283-298. 131 Watermark disease of willows in Japan - Occurrence and pathological anatomy Yasuaki Sakamoto, Yuichi Takikawa, Yuzou Sano, Atsushi Kato and Katsuhiko Sasaki Y. Sakamoto, K. Sasaki, Hokkaido Research Center, Forestry & Forest Products Research Institute (FFPRI), Sapporo 062-8516, Japan. E-mail: yasusaka@ffpri.affrc.go.jp Y. Takikawa, Faculty of Agriculture, Shizuoka University, Shizuoka 422-8529, Japan. Y. Sano, Faculty of Agriculture, Hokkaido University, Sapporo 060-8589, Japan. A. Kato, FFPRI, Tsukuba Norin P.O. Box 16, Tsukuba 305-8687, Japan. Abstract The watermark disease of willows (Salix spp.) was identified in Japan for the first time, in the northern island of Hokkaido. During spring and summer, the leaves and branches suddenly wilted, and the affected trees sometimes died. When affected branches or trunks were cut, a distinct, watery zone stained reddish brown or brownish black (watermark) was observed in the sapwood. The pathogen was isolated from the watermark and identified as Erwinia salicis. Inoculation tests confirmed its pathogenicity to willows. The characteristics of the watermark were investigated. In the watermark, some of the ray parenchyma cells caused necrosis. Dye-injection tests revealed that water conduction did not take place in the watermark. Therefore, the non-conductive water mark in sapwood may be analogous to discolored wood, and its formation and expansion in the affected sapwood may be the reason for the wilting symptoms. Soft X-ray photography and cryo scanning electron microscopy revealed that the watermark had a high level of moisture (wetwood). Osmotic potentials ((f>n )of the watermark were substantially lower than those in healthy sapwood. 13 C-NMR spectroscopy revealed that E. salicis produced levan in the watermark, which is sus pected to be a causal agent of low (j)7t. Vessel-ray parenchyma and interfibre pit membranes were often damaged and absent. They could serve as effective pathways for the accumulation of water in the watermark. Keywords: Watermark disease, Erwinia salicis, willows (Salix spp.), discolored wood, bacterial wetwood Introduction Watermark disease is one of the most serious diseases on willows (Salix spp.) and has been reported in the U.K. (Day 1924, Dowson 1937), the Netherlands (Lindeyer 1931) and Belgium (Rijckaert et al. 1984). The causal bacterium was first named as Bacterium salicis by Day (1924), but the name was later changed to Erwinia salicis (Day 1924) Chester 1939 (Chester 1939). In August 1993, a serious shoot blight, wilt and dieback was found on willows in natural forests around the mountainous area of Mount Taisetsu in central Hokkaido. 132 Currently, this disease is observed on three local Salix species: S. bakko Kimura, S. sachalinensis Fr. Schm. and S. kinuyanagi Kimura in the natural forests of the moun tainous area of Mt. Taisetsu (Kamikawa, Kamisihoro, Rubeshibe and Oketo) and in the natural forests of the upper part of the Nissho-Pass (Hidaka). The development of the wilting symptoms in the leaves and shoots of this disease suggests a progressive inhibition in its ability to conduct water. The watermark (affected wood tissues) had an abnormally high level of mois ture compared to the surrounding regions (Sakamoto and Sano 2000). This kind of wood tissue is called wetwood (Panshin and de Zeeuw 1980). Several kinds of bac teria in the xylem tissues are suspected to be associated with wetwood in many cases (e.g., Hartley etal. 1961, Ward and Pong 1980, Ward andZeikus 1980), and the term "bacterial wetwood" has been applied to such kinds of wetwoods (Murdoch and Campana 1983). This report deals with symptoms of the disease, isolation and characterization of the pathogen and establishment of its pathogenicity (Sakamoto et al. 1999). Wood anatomy, water conductivity and physiological and chemical properties of the water mark are also studied (Sakamoto and Sano 2000, Sakamoto and Kato, submitted). Symptoms Symptoms were observed regularly from 1993 to 1998 in the mountainous area near Mt. Taisetsu. During spring and summer, the leaves on some branches suddenly wilted and turned reddish brown, but remained on the trees for a time. On some trees almost all the branches were affected. Some branches and trunks died and then be came leafless. Sometimes recovery shoots developed on the affected branches but they often became affected later. Seriously affected trees died within 4 to 5 years (Fig. 1). The internal symptom is called a 'watermark'. It was found in the sapwood of affected branches and trunks as a distinct, watery zone stained reddish brown or Figure 1. Dieback of S. sachalinensis (photographed in Kamikawa). 133 brownish black. It forms a circumferential stain (Fig. 2), sometimes covering almost the entire transverse section in seriously affected trees. Bacteria tended to ooze from watermark on cutting. On exposure to air, the watermark quickly turned dark brown or black and a dark brownish-black liquid (mixture of bacterial ooze and metabolic substances in watermark) was exuded. Material and methods Isolation of the pathogen Small pieces of the watermark (approximately 5x5x5 mm) were macerated in 5 ml of sterile peptone water (1%). The resulting suspensions were streaked on plates of nutrient agar (NA: Eiken E-MC01), and incubated at 20° C. After 4-6 days, single colonies were re-streaked on fresh NA plates to ensure purity. Identification of the pathogen The pathogen was identified based on their colony morphology, cell morphology and bacteriological characteristics (Sakamoto et al. 1999). Pathogenicity Pathogenicity tests were performed on 2-year-old seedlings from trees of S. sachalinensis and S. kinuyanagi in 1998. One-year-old branches were pricked to the xylem with a needle through a drop of bacterial suspension (approximately 10 9 cfu/ mL). Figure 2. Circumferential watermark stain (arrows) in S. bakko (collected in Kamikawa). H indicates heartwood. 134 Light microscopy Six affected and three healthy trees were selected. Samples of the watermark and the healthy sapwood (approximately 1.5x1.5><1.5cm) were fixed with FAA (forma lin: acetic acid: 50% ethanol = 5: 5: 90) for one week, and then washed under run ning water for 3-4 hours. Twenty to twenty five |im thick transverse, tangential and radial sections were cut on a sliding microtome. They were stained with safranin O (1% solution in 50% ethanol) and fast green (0.1% solution in 95% ethanol), and were then examined under a light microscope. Some unstained sections were also observed. Dye-injection test Three affected and one healthy tree (for control) were selected. Funnel-shaped plas tic collars were attached to the trunks, and then a safranin O solution (0.2% aqueous solution) was poured in the collars. Four to six holes (8.5 mm in diameter, average depth of 59.1 mm) were bored with an electric drill under the surface of the solution in the collars. About four hours later, the trees were cut down and divided into 1 m in long sections. The distribution of the dye was examined to detect the zone of xylem, which remained conductive. Soft X-ray photography Cylindrical specimens were excised from the trunks of one affected and one healthy tree. The preparation of samples for soft X-ray photography was made according to the methods developed by Sano et ai. (1995). Soft X-ray has low penetrability; it cannot penetrate the zone which has a high level of moisture in the green wood disk. Hence, the zones with high level of moisture shown as dark (black) zone in the photograph. Cryo-scanning electron microscopy Cylindrical specimens were excised from the trunks of one affected and one healthy tree, which were used for soft X-ray photography. The collection of samples was made according to the method developed by Utsumi et al. (1996), and preparations of samples were made according to the procedures described by Sano et al. (1995). The samples were kept frozen and were never melted during the course of this study. The transverse surfaces of the samples were examined under the cryo-SEM (JSM 840 A equipped with CRU-40) to observe the distribution of ice (water in natural condition) in the tissues. Scanning electron microscopy Four affected and two healthy trees were selected, then wood blocks (approximately 135 5x5x5 mm) were taken from the watermark and healthy sapwood. They were freeze dried according to the methods described by Sano and Fukuzawa (1994). After dry ing, the blocks were split radially or tangentially with razor blades, and two halves of each specimens were coated with gold using an ion sputter (FC-1100; JEOL), and observed with a SEM (JSM-5600LV; JEOL) at 10 or 15kV. Osmotic potential Four affected and two healthy trees, which were selected for scanning electron microscopy, were used. Osmotic potentials (7r ) of the liquid exuded from the wa termark and the small disks (approximately 5 mm in diameter, 2 mm in thickness) of unaffected sapwood (from healthy trees) were measured by a dew-point microvoltmeter (HR-33T; WESCOR INC., USA) equipped with a sample chamber (C-52-SF; WESCOR INC., USA). Polysaccharide in watermark 1) Isolation of extracellular polysaccharide (EPS) from bacterial culture The isolate of E. salicis was incubated on the high-sucrose medium developed by Dye (1968) at 25° C for 2 days. The crude EPS slime on the plate medium was col lected and purified. 2) Isolation of polysaccharide from watermark Four affected and two healthy trees, which were selected for scanning electron microscopy, were used. The wood samples (lOg in fresh weight) were excised from the watermark and healthy sapwood. The water extracts of milled samples were collected, purified and measured their dried weight. 3) 13 C-NMR spectroscopy Samples of the EPS fraction from bacterial culture, the polysaccharide fraction from watermark and the healthy sapwood were dissolved in deuterium oxide and sub jected to 13C-NMR spectroscopy using ALPHA-500 (JEOL) operated at 125 MHz. Commercial levan (13-2,6-D-fructofuranan) (WAKO Pure Chemical Industries, Ltd.) was served as the reference standard. Results Isolation of the pathogen The same bacterium was regularly isolated from all the freshly affected tissues. The colonies were 1-2 mm in diameter on NA after 4 days at 20° C, with an entire margin, and were circular, convex, smooth and glistening. 136 Identification of the pathogen The current isolates were Gram-negative, nonsporing, straight rods, motile with peritrichous flagella, produced a transparent to white growth, metabolized glucose fermentatively, and showed negative reactions in the potato soft rot, oxidase activity, and nitrate reduction tests. Other results are shown in Table 1. Pathogenicity The reactions were observed approximately 2-3 weeks after inoculation. Initially a few leaves became light brown and curled, but about 3-4 days later, other leaves also turned brown. Then whole branches wilted, the affected leaves fell off, and the branches died within a month. The watermark stain was observed in transverse sec tions of affected branches. Table 1. Some characteristics of the current isolates and Erwinia salicis. a = The author's data, b = Data from Dye (1968). + = 80-100% isolates positive, - = 0-20% isolates positive. CunHTtmfetEÖ31 E. salicis(h) Characteristics O/F test F F Potato soft rot test - Oxidase activity - - Catalase activity + + Nitrate reduction - - Reaction in purple milk - - Growth factor requirements - - Growth at 36°C - - Levan (Mucoid growth) + + H2S from Cystein + + Casein hydrolysis - - Aesculin hydrolysis + + Gelatin liquefaction - - Acetoin + + Methyl Red test - - Indole test - - Acid production from Glucose + + Fructose + + Lactose - - Maltose - - Utilization of Acetate + + Fumarate + + Benzoate - - Oxalate - - 137 Light microscopy In the unstained tissues of the watermark, contents of the affected ray parenchyma cells were yellow to brown, and some cells lost their contents. In the stained tissues of the watermark, not all the vessels were clogged with tyloses and masses of bacteria, and some of the ray parenchyma cells caused plas molysis and necrosis (Fig. 3). Masses of bacteria formed in the vessels irrespective of the presence of tyloses (Fig. 4). Tyloses were also found in the tissues without the watermark, but the masses of bacteria were only found in the watermark. The ray parenchyma cells adjoining the vessels clogged with bacteria were affected first. Then, the parenchyma cells in the watermark gradually died as the disease progressed. Sometimes, the bacteria were seen in dead ray parenchyma cells in contact with the vessels clogged with the bacteria. In the healthy tissues, no tyloses, no masses of bacteria in vessels and no necrosis of ray parenchyma cells were observed. Dye-injection test In the healthy tree, all the sapwood had conductivity (but only late wood in old sapwood). However, no conductivity was observed in the watermark. In seriously affected trees (the entire transverse section was almost completely covered by the watermark), only part of sapwood in the outer layers was conductive. Soft X-ray photography When comparing the normal black and white photograph and the soft X-ray photo graph of the affected green wood disks, the watermark was observed to be darker than the surrounding sapwood on the soft X-ray photograph taken of the green (fro zen) wood (Figs. sa, b). This indicated that there was a high accumulation of mois ture. There were no black coloured areas (a high level of moisture) in the soft X-ray photograph of the dried disk from the affected wood. Figure 3. Radial view of the ray parenchyma cells in the watermark. Big arrow indicates plasmolysised cell and small arrows indicate necrotic cells. 138 Figure 4. Radial view of the watermark. T indicates tyloses and B indicates masses of bacteria. Figure 5. Soft X-ray photography. a. A normal photograph of the wood disk from the affected tree in green state. Arrows indicate the watermark. b. A soft X-ray photograph of the disk in a in green state. Note that the dark zone (arrows) corresponds with the watermark in a. Cryo-scanning electron microscopy The cryo-scanning electron microscopy revealed that almost all the cells in the wa termark were filled with ice irrespective of the type of cell (Fig. 6). However, ice was seldom present in the lumina of wood fibers in the sapwood of healthy samples. 139 Scanning electron microscopy In the watermark, vessel-ray parenchyma pit membranes were heavily incrusted, damaged, or often absent (Fig. 7). Figure 8 shows a complementary pair of interfibre surfaces in which the pit membranes are absent. In contrast to the watermark, appreciable incrustations were rarely observed in tissues of healthy sapwood. Also, vessel-ray parenchyma and interfibre pit mem branes typically remained intact. Figure 6. A cryo-SEM photograph of the watermark. Note that almost all the cells were filled with ice (water in natural conditions). Figure 7. SEM photographs of the vessel-wall surfaces between vessels and ray parenchyma cells. Pit membranes are often heavily incrusted or absent. 140 Figure 8a, b. SEM photographs of complementary pair of surfaces between wood fibers. Pit membranes are often absent on both faces of pit pairs (arrows). Osmotic potential Substantial decrease in the mean values of jr was observed in the watermark (-0.34 MPa) compared to the values of the unaffected sapwood (-0.04 MPa). Polysaccharide in watermark The mean values of the dried weight of the polysaccharide fraction from watermark and healthy sapwood were 40.0 mg and 31.2 mg, respectively. The signals in the I3C-NMR spectrum of the commercial levan appeared in the range of 60-110 ppm, and these signals completely agreed with the signals in the spectrum of the EPS fraction (arrowheads in Fig. 9). Thus, the EPS fraction was identified as levan. The spectrum of the polysaccharide fraction of the watermark showed the same signals as the EPS fraction and commercial levan, while the spec trum of healthy sapwood did not (Fig. 9). This fact obviously indicated that the watermark contained levan; on the contrary, the healthy sapwood did not. 141 Figure 9. 13C-NMR spectroscopy. a. Commercial levan. b. EPS from the bacterial culture. c. Polysaccharide from healthy sapwood. d. Polysaccharide from the watermark. Arrowheads indicate the signals of levan. 142 Discussion Based on the bacteriological characteristics mentioned above, the current isolates should be included in the Erwinia amylovora group (Bradbury 1970). Other charac teristics of the current isolates corresponded to those of E. salicis (Table 1), and their pathogenicity to Salix spp. was confirmed. Therefore, the current isolates were iden tified as E. salicis. The external and internal symptoms corresponded to the symp toms of watermark disease of willows as previously reported in Europe (Day 1924, Dowson 1937, Smith et al. 1986, Strouts and Winter 1994). Our report (Sakamoto et al. 1999) is the first description confirming the presence of the watermark disease in willows, and its pathogen, E. salicis, in Japan. Although E. salicis was reported to be present in Japan in one textbook (Bradbury 1986), no evidence to support this state ment was provided. The ability to conduct water was not observed in the watermark. The contents of the affected ray parenchyma cells in the watermark were yellow to brown. This suggests that there was an accumulation and oxidation of phenolic compounds. A direct detection test of the phenolic compounds in the watermark was not performed in this study; however, Wong and Preece (1978) reported an increased level of phe nolic compounds in infected willow wood. The present anatomical study showed that the parenchyma cells in the watermark caused plasmolysis and necrosis with the development of the disease. These facts suggest that, in this respect, the watermark in sapwood may be analogous to wound wood 'discolored wood'. Discolored wood is the tissue that forms as a kind of defense reaction in sapwood, with similar proc esses to heartwood formation. The wood is believed to become discolored wood as follows. When xylem tissues in healthy sapwood are attacked by microorganisms or are mechanically injured, the accumulation level of secondary substances and anti microbiological substances, such as phenolic compounds, increasesTT'hen, the pa renchyma cells gradually die. The color of these tissues will turn brown to black brown with oxidation and polymerization of the secondary substances, then lose the ability to conduct water as in normal heartwood (Shigo and Hillis 1973, Bauch 1984, Shigo 1984). Therefore, the non-conductive watermark formation and its expansion in the affected sapwood as the disease progresses may be the reason for the wilting symptoms. This report revealed that masses of E. salicis were only found in watermark and not in surrounding or healthy sapwood. Except for levan, I3 C-NMR spectroscopy revealed that both watermark and healthy sapwood contained the same kinds of polysaccharides (not identified in this study). These facts indicated that the incre ment of dried weight of the polysaccharide fraction of the watermark was due to levan, which was produced by E. salicis. The low (j>n in the watermark can be attributed in large part to levan, which acts as the osmotically active material. Fur thermore, as shown in Figure 4, masses of E. salicis were often found in the lumina of tyloses. This fact indicated that the tyloses were often collapsed and did not block the vessel lumina from water invasion. This study also revealed that the vessel-ray parenchyma and interfibre pit membranes were often damaged. They could also serve as effective pathways for the accumulation of water in the watermark. These data support the concept that levan production decreases n , leading to water accu 143 mulation through those pathways. Studying the more precise mechanism of water inhibition and water accumula tion in the watermark might also provide relevant information to elucidate the mecha nisms of wilt diseases and bacterial wetwood in other tree species. References Bauch, J. 1984. Discolouration in the wood of living and cut trees. lAWA Bulletin new series 5: 92-98. Bradbury, J.F. 1970. Isolation and preliminary study of bacteria from plants. Review of Plant Pathology 49: 213-217. Bradbury, J.F. 1986. Guide to Plant Pathogenic Bacteria. CAB International, Wallingford, p. 88- 89. Chester, F.D. 1939. Genus IV. Erwinia Winslow et al. In: D.H. Bergey, Breed, R.S., Murray, E.G.D. and Hitchens, A.P. (eds.). Bergey's Manual of Determinative Bacteriology, sth edi tion. Williams & Wilkins Company, Baltimore, p. 404-420. Day, W.R. 1924. Watermark disease of the cricket-bat willow. Oxford Forestry Memoirs No. 3:1- 30. Dowson, W.J. 1937. Bacterium salicis Day, the cause of the watermark disease of the cricket-bat willow. Annals of Applied Biology 3: 528-545. Dye, D.W. 1968. Ataxonomic study of the genus Erwinia. I. The "amylovora" group. New Zea land Journal of Science 11: 590-607. Hartley, C., Davidson, R.W. and Crandall, B.S. 1961. Wetwood, bacteria, and increased pH in trees. USDA Forest Products Laboratory Report No. 2215. USDA Forest Service, Washing ton, D.C. Lindeijer, E.J. 1931. Een bacterie-ziekte van de wilg. Tijdschrift v. Plantenziekten 37: 63-67. Murdoch, C.W. and Campana, R.J. 1983. Bacterial species associated with wetwood of elm. Phy topathology 73: 1270-1273. Panshin, A.J. and de Zeeuw, C. 1980. Textbook of wood technology, 4th edition. McGraw-Hill, New York. Rijckaert, C., Van Tomme, R. and Steenackers, V. 1984. The occurrence of the watermark disease of willows (Salix) in Belgium. Mededelingen Van de Faculteit Landbouwwetenschappen Van de Rijksuniversiteit Gent 49: 509-515. Sakamoto, Y., Takikawa, Y. and Sasaki, K. 1999. Occurrence of watermark disease of willows in Japan. Plant Pathology 48: 613-619. Sakamoto, Y. and Sano, Y. 2000. Inhibition of water conductivity caused by watermark disease in Salix sachalinensis. lAWA Journal 21: 49-60. Sakamoto, Y. and Kato, A. Some properties of the bacterial wetwood (watermark) in Salix sachalinensis caused by Erwinia salicis. lAWA Journal (submitted). Sano, Y. and Fukazawa, K. 1994. Structural variations and secondary changes in pit membranes in Fraxinus mandshurica var. japonica. lAWA Journal 15: 283-291. Sano, Y., Fujikawa, S. and Fukazawa, K. 1995. Detection and features of wetwood in Quercus mongolica var. grosseserrata. Trees 9: 261-268. Shigo, A.L. and Hillis, W.E. 1973. Heartwood, discolored wood, and microorganisms in living trees. Annual Review of Phytopathology 11: 197-222. Shigo, A.L. 1984. Trees and discolored wood. lAWA Bulletin new series 5: 99. Smith, 1.M., Dunez, J., Lelliott, R.A., Phillips, D.H. and Archer, S.A. 1986. European handbook of plant diseases. Blackwell Scientific Publications, Oxford, p. 193-194. Strouts, R.G. and Winter, T.G. 1994. Diagnosis of ill-health in trees. HMSO, London, p. 246-247. 144 Utsumi, Y., Sano, Y., Ohtani, J. and Fujikawa, S. 1996. Seasonal changes in the distribution of water in the outer growth rings of Fraxinus mandshurica var. japonica: A study by cryo scanning electron microscopy. lAWA Journal 17: 113-124. Ward, J.C. and Pong, W.Y. 1980. Wetwood in trees: A timber resource problem. USDA General Technical Report PNW-112. USDA Forest Service, Washington, D.C. Ward, J.C. and Zeikus, J.G. 1980. Bacteriological, chemical and physical properties of wetwood in living trees. Mitteilungen der Bundesforschungsanstalt fur Frost-Holzwirtschaft. 131: 133-168. 145 Responses of juvenile poplars to inoculation with Septoria musiva can predict their long-term canker disease resistance Gerald E. Weiland, JoAnne C. Stanosz and Glen R. Stanosz Department of Plant Pathology, University of Wisconsin-Madison, 1630 Linden Drive, Madison, WI, 53706, USA. E-mail: grs@plantpath.wisc.edu Abstract Septoria musiva causes cankers that severely limit production of hybrid poplar in eastern North America. A field trial in 1998 and a greenhouse trial in 1999 were conducted to determine whether a short-term assay of response to stem infection by S. musiva was predictive of long-term field performance (LTFP). Fresh leaf scars of 27 clones (field) and 15 clones (greenhouse), represent ing a range in long-term performance as reflected by canker disease, were inoculated with five mm MEA plugs colonized with S. musiva. Results were measured four months after inoculation in the field trial, and three months after inoculation in the greenhouse trial. Incidence, canker length, and percent of stem circumference affected (girdle) were recorded for each cankered tree. Re sponses were compared with ratings of canker damage from longer-term field studies by logistic regression. Analysis showed that the field assay predicted LTFP better than the greenhouse assay. In general, girdle was the best predictor for LTFP. 146 Neofabraea populi in Scandinavia Risto Kasanen, Timo Kurkela and Jarkko Hantula Finnish Forest Research Institute, P.O. Box 18, 01301, Vantaa, Finland. E-mail: risto.kasanen@metla.fi Abstract The ascomycete Neofabraea populi was identified from two damaged hybrid aspen stands in Southern Finland. Genetic markers from 24 single-ascospore isolates, two isolates from wood tissue and three Norwegian reference isolates of N. populi were compared. No observable genetic variation was found with 32 random amplified microsatellite markers. The identity of five RAMS loci between samples was assured by successful PCR amplification with sequence specific prim ers. The results indicate that the genetic variation within N. populi is extremely low, suggesting prescence of widely spread clone. This might be due to introduction of small founder population of the fungus to Europe. It is also likely that the fungus is homothallic. Keywords: Neofabraea populi, hybrid aspen, genetic variation, introduced species Introduction Neofabraea populi Thompson causes cankers on Populus grandidentata, P. tacamahaca and P. tremuloides and also hybrid aspen (Populus tremula x P. tremuloides) (Thompson 1939, Roll-Hansen and Roll-Hansen 1969). Originally N. populi was described in Northern America (Thompson 1939). The disease had also been noted on several hybrids poplars in Japan since late 1940'5, although it was identified in the end of 1960's (Ito et al. 1969). The affected species included P. alba, P. sieboldii, P. tremula var. davidiana, P. alba x P. sieboldii, P. tremula var. davidiana x P. canescens, P. sieboldii x P. canescens, P. nigra, P. deltoides var. monilifera, P nigra x P. deltoides var. monilifera, P. nigra x P .maximoviczii, P. maximoviczii and P. simonii (Ito et al. 1969). The first damages caused by the fungus in Europe were noted few years after hybrid aspen plantations were established in Norway. Thus it was assumed that the fungus had been accidentally introduced from North America (Roll-Hansen and Roll- Hansen 1969). There are only few documented observations of N. populi in Finland (Semb and Hirvonen-Semb 1968), but it is thought to occur commonly with planted hybrid aspen. Recently, the importance of hybrid aspen as raw material for pulp production has increased. The wood is planned to be grown by short-rotation for estry on reforested agricultural land. Since conditions for spread of N. populi are optimal in dense coppice stands, the fungus is a potential threat for production of hybrid aspen. 147 The aim of this study was to identify the causal agent of canker disease ob served in hybrid aspen stands in Finland. In the disease, 2 nd generation (root suckers) of trees are seriously damaged with cankers and dead bark. Both morphological characters and DNA markers were used. Material and methods Fungal isolates Hybrid aspen stands in Fiskars (South-West Finland) and Pornainen (Southern Fin land) with dense 2 nd generation of trees, were visited in October 2000. Serious can ker damages and dead stems of aspen suckers were observed within both stands. A total of 25 cankers of young hybrid aspen with cankers were cut apart. In Fiskars, fresh cankers with apothecia were frequent and easy to cut from living stems. In Pornainen, no apothecia bearing cankers on living trees were found, because the stand was almost destroyed by the disease and most trees had died during pervious years. In the laboratory, slice of the apothecia bearing bark was cut from each stem, and placed under lid of a petri dish containing water agar. Released single ascospores were identified from the water agar by microscope and isolated with modified pasteur pipette. The single-ascospore isolates were further cultured on 1,5% malt agar. A total of 24 isolates were isolated from this stand. Although apothecia were not found on canker samples collected from Pornainen, it was possible to isolate N. populi directly from the wood material. Pieces of wood were cut from the barrier zone of a canker, surface-sterilized by rinsing in 4% NaOCl, 75% ethanol and finally sterilized water (10 sec each) and incubated on 1,5% malt agar. The morphology of the isolates was compared to single ascospore isolates. In addition to number of saprophytic fungi, also two isolates of N. populi were found. Three isolates were obtained from Halvor Solheim, Norwegian Forestry Re search Institute. They were isolated during the N. populi epidemy in 1960'5. Microscopy Ascospores were identified on the water agar plates under a low power microscope. For microscopic purposes, blocks of water agar carrying spores were cut from the culture, placed on a microscope slide and covered with a coverslip. The slides were placed on a warm plate at 50° C to dry and reduce agar to a thin layer before the spores were photographed. One hundred ascospores were measured from the micro scope slides. DNA markers For DNA isolation the fungal cultures were grown on malt agar covered with cello 148 phane membrane. Nuclear DNA was isolated as described in Vainio et ai. (1998). A total of 36 RAMS (Random Amplified Micro Satellite; Hantula et ai. 1996) markers were PCR amplified by using anchored di- and trinucleotide repeat primers and analysed on 1% agarose gels. RAMS primers used were DBV(CAT) 5 DDB(CCA) 5 , DBD(CT) S C, HBH(GAG) 5 and VDD(TCC) 5 , where B= C/G/T, D=A/G/T, H=A/C/T and V=A/C/G), respectively. Primers for amplification of five sequence character ized amplified regions (SCAR) were designed; individual RAMS markers were cloned, sequenced and new primers were designed based on the sequence informa tion. Results Microscopy Based on the morphology of ascospores, the aspen canker pathogen in Finland was similar to N. populi (Fig. 1). The size of the ascospores was 15-25 x 5-10 (am. The ascospores were oblong-ellipsoid, slightly curved, contents granular, hyaline.lt was also noted that the morphology of the colonies in vitro was highly similar with refer ence isolates with practically no variation. Figure 1. Spores of N. populi. Size of the spores is 15-25 x 5-10 (im. 149 DNA markers Surprisingly, all the Scandinavian isolates were identical according to the used mark ers. No variation was observed within 32 RAMS markers amplified from ascospore isolates, canker isolates or reference isolates (Figs. 2,3). With DBV(CAT) 5 primer it was possible to score seven reproducible markers, with DDB(CCA) 5 six markers, with DBD(CT)5 C eight markers, with HBH(GAG)5 six markers and with VDD(TCC)5 five markers, respectively. Accordingly, the SCAR markers were successfully am plified from all isolates (Fig. 4). Figure 2. RAMS fingerprints amplified from N. populi isolates (Fiskars) with DBV(CAT) 5 primer. Note the similarity between lanes. Leftmost lane is 100 bp molecular weight ladder, where the lower intense band is 600 bp. Figure 3. Similar RAMS fingerprints amplified with DBV(CAT)5 primer. Lanes from left: A 100 bp molecular weight ladder, where the lower intense band is 600 bp, three Norwegian isolates, a isolate from Fiskars and a molecular weight ladder (right). 150 Figure 4. Pairs of SCAR markers amplified from one Norwegian (left band in each pair) and one Finnish isolate (right band in each pair). Distal lanes are molecular weight ladders. Discussion Ascospore morphology The ascomycete N. populi was identified from two damaged hybrid aspen stands in Southern Finland. The size of the ascospores measured in this study (15-25 x 5-10 (im) was close to earlier observations; 16-22x5-6,5 |im (Thompson (1939), 14-22 x 4-6 urn (Ito etal. 1969), 13,2-20,8 x 4-6,5 (Roll-Hansen 1969). Also the slightly curved, ellipsoid shape and hyaline colour of the spores were observed both in this study and previous reports. DNA markers No genetic variation was observed with the 32 RAMS markers, although theoreti cally it would be possible to identify 2 32=4 294 967 296 clones of the fungus by scoring the presence or absence of the markers. Although it is known that no marker system has ultimate resolution of genotypes, it can be considrered highly probable that the fungal isolates studied were actually clonal. The same markers amplified from isolates from 1960's and 2000's suggest also exclusive clonality of the fungus in Scandinavia, since no evidence of mating with different genotypes during last four decades was found. As most of the isolates were isolated from single ascospores, the observations also support the homothallism of N. populi, since mating is usually required to pro duce apothecia. Homothallism was also hypothesized by Thompson (1939), who observed that the fungus produced apothecia in vitro. However, no apothecia or similar structures have been observed within cultures in this study. The extremely low ge netic variation was also in correspondece with the previous theory of introduction of the fungus to Europe, and suggested that the founder population had been only few isolates, perhaps one isolate. However, more isolates from Scandinavia and North America are needed, to obtain more information about the population structure of the fungus within Scandinavia and also the possible introduction of the fungus from North America to Europe. 151 References Hantula, J., Dusabenyagasani, M. and Hamelin, R. C. (1996). Random amplified microsatellites (RAMS) - a novel method for characterizing genetic variation within fungi. European Jour nal of Forest Pathology. 26: 159-166. Ito, K., Chiba, O. and Kondo, H. 1969. Neofabraea canker of poplars in Japan. Bulletin of the Governement Experiment Station, Meguro 225: 31-40. Roll-Hansen, F. and Roll-Hansen, H. 1969. Neofabraeapopuli on Populus tremulax P. tremuloides in Norway. Comparison with the conidial state of Neofabraea malicorticis. Meddeleser fra det Norske Skogforsokvesen 22: 215-226. Semb, L. and Hirvonen Semb, A. 1968. Poppel-barkbrann en ny soppsjukdom i Norge Gartneryrket 58: 582-583. Thompson, G.E. 1939. A canker disease of poplars caused a new species of Neofabraea. Mycologia 31:455-465. Vainio, E. J., Korhonen, K. and Hantula, J. 1998. Genetic variation in Phlebiopsis gigantea as determined with random amplified microsatellite (RAMS) markers. Mycolological Research 102: 187-192. 152 Effect of fertilisation on the resistance of spruce seedlings to shoot blight caused by Siro coccus conigenus Hans Anglberger, Erhard Halmschlager, Peter Hietz and Jutta Mattanovich H. Anglberger, E. Halmschlager, J. Mattanovich, Institute of Forest Entomology, Forest Pathology and Forest Protection, University of Agricultural Sciences, Hasenauerstraße 38, A -1190 Vienna, Austria. E-mail: anglberg@edvl.boku.ac.at P. Hietz, Institute of Botany, University of Agricultural Sciences, Gregor Mendel-Straße 33, A -1180 Vienna, Austria. Abstract The effect of fertilisation with different Mg-fertilisers on the susceptibility of four-year-old Picea abies seedlings to Sirococcus shoot blight was examined in an artificial inoculation experiment. Prior to inoculation the nutritional and the physiological status of the seedlings was assessed by measurements of chlorophyll fluorescence and the contents of proteins, low molecular weight carbohydrates and starch in needles. Four weeks after inoculation shoot dieback was highest on seedlings fertilised with a neu tral Mg-salt (Bittersalz), whereas no significant differences were found between unfertilised seed lings and seedlings fertilised with an alkaline fertiliser derived from magnesite and dolomite and amended with organic compounds (Magnosol). The Bittersalz treatment was also characterised by reduced chlorophyll fluorescence in the last year's needles and by higher contents of glucose and pinitol in the current year needles. The reisolation revealed striking differences between the infection rate of shoots and the expression of shoot blight symptoms, indicating that the infection of current year shoots by Sirococcus conigenus is not necessarily associated with dieback or development of symptoms. Keywords: Sirococcus conigenus, Picea abies, carbohydrates, proteins, chlorophyll fluorescence Introduction The mitosporic fungus Sirococcus conigenus (DC.) P. Cannon & Minter causes shoot blight and seedling mortality on many conifer hosts in the northern temperate zone (Sutherland 1987, Farr et al. 1989). It has been implicated in shoot dieback of pine, spruce and hemlock from the seedling stage (Funk 1972, Illingworth 1973, Redfern 1973, Smith 1973, Magasi et al. 1975, Sutherland et al. 1981, Motta et al. 1996) to mature size (O'Brien 1973, Neumiiller 1994, Maresi and Capretti 1995). In Central Europe the disease mainly affects Norway spruce (Picea abies [L.] Karst.). In Austria, Sirococcus shoot blight has caused severe damage in some loca 153 tions since the early 1980s. Symptoms of crown thinning, dieback of twigs and branches as well as top dying of mature Norway spruce have been observed espe cially in secondary spruce forests on poor and acidified soils additionally impover ished by litter raking and lack of admixed tree species (Halmschlager et al. 2000 a, 2000b, Anglberger and Halmschlager 2000, Jandl et al. 2000). Preliminary results from field experiments indicated that imbalances in tree nutrition, particularly a deficiency of magnesium and calcium, increased the susceptibility ofNorway spruce to Sirococcus shoot blight (Anglberger and Halmschlager 2000, Jandl et al. 2000). To verify these results under controlled conditions, the effect of fertilisation on the resistance of spruce seedlings to Sirococcus shoot blight was tested in an inoculation experiment. Material and methods Three-year-old potted spruce seedlings from a single clone, grown on homogenised and autoclaved mineral soil obtained from a Norway spruce stand severely affected by the fungus, were fertilised in spring 1998 with "Magnosol®" (an alkaline ferti liser derived from magnesite and dolomite and amended with organic compounds) and "Bittersalz®" (a neutral Mg-salt), respectively. A third collective of seedlings remained unfertilised. After one year, 20 seedlings of each treatment were inocu lated during flushing by spraying with a conidial spore suspension of S. conigenus. Inoculations were carried out twice with a nine day interval using conidial suspen sions of two different S. conigenus isolates (8,1 x 10 5 and 1,1 x 10 6 conidia/ml). Ten seedlings from each fertiliser treatment served as control and were sprayed with sterile deionised water only. Following inoculation, seedlings were transferred to growth chambers and placed in polyethylene bags to maintain a high humidity (near 95% relative humid ity). According to Wall and Magasi (1976) the seedlings were incubated under the following day/night conditions: 16 h/8 h; 21°C/16°C; 2000 lx/ 0 lx. Eighteen days after inoculation seedlings were removed from the growth chambers and exposed to open air. Evaluation of shoot blight symptoms was carried out 29 days after inocula tion. Seedlings were examined for the presence or absence of shoot blight symptoms and seedlings with at least one severely damaged shoot were classified as damaged. In order to estimate the rate of fungal infection reisolations were carried out 37 days after inoculation. One shoot, apparently healthy, from each of the 90 seedlings used in the inoculation experiment and 40 shoots from damaged seedlings were cut, sur face sterilised and plated onto malt extract agar (MEA 2%). The nutritional and the physiological status of the host prior to inoculation was assessed on seedlings from each treatment which were not used for the inocula tion. Proteins, low molecular weight carbohydrates and starch were analysed from current year needles (7 replicates per treatment) and chlorophyll fluorescence was measured in current and last year's needles. For the chemical analyses the needles ® Magnosol is a registered trademark of Bodenkalk GmbH, Austria; Bittersalz is a registered trademark of Kali+Salz GmbH, Germany 154 were frozen in liquid nitrogen, lyophilised, ground to = 2 mm and stored at - 30° C. For carbohydrate analyses aliquots of the frozen powder were pulverised in a sample mill (MM2, Retsch) and extracted with hot distilled water (4% w/v). 1 ml of the water extracts was purified using a coupled cation and anion exchange resin (Dowex 50W x 8, 50-100 mesh, H + -form, and Dowex Ixß, 50-100 mesh, COO" form, Fluka), with arabinose added as an internal standard. The neutral fraction (low molecular weight carbohydrates) was analysed by HPLC (Hewlett Packard, Aminex HPX-87 P, 300 x 7,8 mm, Biorad, refractive index detector) (Meister 1998). Starch was degraded to glucose (Winter et al. 1996) which was measured photometrically following Boehringer-Mannheim (1995). Proteins were analysed according to Peterson (1977). Chlorophyll fluorescence (F y /F m ) of dark adapted current and last year's nee dles was measured with a MINI-PAM Photosynthesis Yield Analyser (Walz, Effeltrich, Germany). Reflectance of photosynthetically active radiation in the range of 430 - 680 nm was measured with an SD2OOO spectrometer (Ocean Optics, Dunedin, Florida) on current year needles. Datt (1999) showed that the chlorophyll content correlates with relative reflectance [(r 850 -r 710 )/(r 850 -r 6go )], which was used to obtain a relative chlorophyll content of the spruce needles. Results and discussion Although development of symptoms was very low in general (only 0-12% of the shoots from a seedling were damaged by Sirococcus shoot blight), number of damaged seedlings differed significantly between treatments (x 2 =6.172; p=0.046, Table 1). Number of damaged seedlings was highest in the Bittersalz treatment, whereas no significant differences occurred between the unfertilised seedlings and the Magnosol treatment. Thus seedlings fertilised with Bittersalz appeared more susceptible to shoot dieback. Control seedlings remained free of symptoms. Table 1. Number of damaged spruce seedlings, 29 days after inoculation with a spore suspension of S. conigenus, related to fertiliser treatments (x 2 =6.172; p=0.046). Interestingly, reisolation from inoculated current year shoots indicated a gen erally high rate of infection (Table 2). Although highest from needles of wilted shoots, the fungus also was frequently reisolated from needles and stems of inoculated shoots showing no visible symptoms. These results suggest that development of symptoms is related to the chemistry and physiology of the current year shoots as affected by the treatment with different fertilisers. However, it remains unclear which factor actually results in expression of symptoms. Because length of current year shoots Treatment Magnosol Bittersalz Not fertilised Not damaged 10 3 9 Damaged 10 17 11 155 did not differ at the date of inoculation, differences in disease incidence cannot be explained by varying susceptibility of the seedlings due to different stages of shoot development. Table 2. Reisolation rate (%) of S. conigenus from stems and needles of diseased and healthy looking inoculated current year shoots as well as from shoots of the control variant (n = number of shoots used for reisolation). In contrast to the other treatments seedlings fertilised with Bittersalz showed needle yellowing, especially on older needles, and reduced chlorophyll fluorescence in the last year's needles prior to inoculation, indicating impaired photosynthesis (Table 3). Reduced photosynthetic performance therefore perhaps promoted devel opment of disease symptoms. The fact that the fertiliser had no effect on the chloro phyll fluorescence of the current year needles suggests that plants from all treat ments translocated nutrients to the young needles. The reflectance from needles of Bittersalz-fertilised plants also indicated lower chlorophyll content. The carbohydrate contents in current year needles related to treatments are given in Table 4. Values of the carbohydrate content did not much differ from those obtained by Dermutz (1992) and Schafellner et al. (1996) who investigated the car bohydrate contents of Picea abies growing in young plantations. Furthermore, the contents of glucose and pinitol in the current year needles of the Bittersalz treatment were significantly higher than in the Magnosol fertilised and the unfertilised seed lings, indicating higher contents of soluble carbohydrates. No differences in carbo hydrate and protein contents occurred between the Magnosol treatment and the unfertilised seedlings. The results of this study do not agree with the hypothesis of Funk (1972), who claimed Sirococcus shoot blight as a "low sugar disease", according to Horsfall and Dimond (1957). Seedlings fertilised with Bittersalz were most severely affected by Sirococcus shoot blight in spite of the high carbohydrate contents prior to inocula tion. Since no additional data are available on the dynamics of carbohydrate con tents after inoculation and during incubation of the seedlings in the growth cham bers, an explanation of the relatively strong symptoms in the Bittersalz treatment is difficult. Whatever the biochemical mechanism is, this study showed that the expres sion of symptoms was affected by variations in needle chemistry and physiology resulting from treatments with different fertilisers. Furthermore, fertilisation with magnesium sulphate (Bittersalz) was not appropriate to reduce susceptibility of seed lings, grown on nutrient poor and acidified soils, to Sirococcus shoot blight. Inoculated Control Damaged No symptoms Damaged No symptoms n = 40 O II C n = 0 n = 30 Stems 67.5 % 28.3 % 0,0 % Needles 88.6 % 73.3 % 3.3 % 156 Table 3. Relative estimate of total chlorophyll content and dark fluorescence yield (Fv/Fm) of current and last year's needles due to fertiliser treatment. Table 4. Concentrations (mg/g dry weight) of carbohydrates and proteins in current year needles prior to inoculation, related to fertiliser treatments (means followed by different letters are significantly separated at p<0.05 by ANOVA). Acknowledgements This research was conducted as part of the Special Research Program "Forest Eco system Restoration (SF008)", funded by the Austrian Science Foundation and the Ministry of Agriculture and Forestry. References Anglberger, H. and Halmschlager, E. 2000. Fertilization as a control measure to Sirococcus shoot blight in secondary Norway spruce stands. In: Forest Ecosystem Restoration: Ecological and Economical Impacts of Restoration Processes in Secondary Coniferous Forests. Ed. by Hasenauer, H. Proc. Int. Conf. on Forest Ecosystem Restoration, Vienna, Austria, pp. 29-34. Boehringer-Mannheim. 1995. Methoden der enzymatischen Lebensmittelanalytik - Arbeits anleitung. Boehringer-Mannheim Gmbh. Datt, B. 1999. A new reflectance index for remote sensing of chlorophyll content in higher plants: Tests using Eucalyptus leaves. J. Plant Phys. 154: 30-36. Dermutz, A. 1992. Muster und Dynamik löslicher Kohlenhydrate im Neuaustrieb von Fichten (Picea abies Karst.). Diploma thesis, Univ. f. Bodenkultur, Wien. Farr, D.F., Bills, G.F., Chamuris, G.P. and Rossman, A.Y. 1989. Fungi on plants and plant products in the United States. APS Press, St. Paul. Treatment Control Bittersalz Magnosol Significance (ANOVA) Total chlorophyll 0.492 0.512 0.540 0.0163 (relative values) (±0.013) (±0.004) (±0.012) F /F 1 -year-old v m J 0.767 0.751 0.774 0.041 needles (±0.007) (±0.008) (±0.005) F /F current year v m J 0.769 0.769 0.770 0.993 needles (±0.006) (±0.005) (±0.005) Sucrose Glucose Pinitol Fructose Myo inositol Starch Protein Magnosol 6,64a 22,09 a 30,4 l ab 24,47 a 3,93 a 9,l a 27,72 a Bittersalz 6,09 a 28,19 b 34,42 b 24,35 a 4,42 a 10,5 a 26,78 a unfertil. 6,10 a 22,42 a 28,45 a 22,85 a 4,48 a 12,2 a 26,29 a 157 Funk, A. 1972. Sirococcus shoot-blight of Western Hemlock in British Columbia and Alaska. Plant Dis. Rep. 56: 645-647. Halmschlager, E., Anglberger, H. and Neumiiller, A. 2000 a. Die Bedeutung des Sirococcus- Triebsterbens in sekundären Fichtenvväldern. In: Umbau sekundärer Nadehvälder. Ed. by Miiller, F.FBVABerichte 111,95-100.(Schriftenreihe der Forstlichen Bundesversuchsanstalt Wien), Federal Forest Research Centre Vienna, Austria. Halmschlager, E., Gabler, A. and Andrae, F. 2000b. The impact of Sirococcus shoot blight on radial and height growth of Norway spruce (Picea abies [L. ] Karst.) in young plantations. For. Path. 30: 127-133. Horsfall, J.G. and Dimond, A.E. 1957. Interaction of tissue sugar, growth substances and disease susceptibility. Z. Pflanzenkrankh. Pflanzenschutz 64: 415-421. Illingworth, K. 1973. Variation in the Susceptibility of Lodgepole Pine provenances to Sirococcus Shoot Blight. Can. J. For. Res. 3: 585-589. Jandl, R., Anglberger, H., Reh, M. and Halmschlager, E. 2000. Auswirkung von Dungemaßnahmen auf einen sekundären Fichtenbestand im Kobernaußerwald mit Symptomen des Fichten- Triebsterbens. Die Bodenkultur 51: 247-258. Magasi, L.P., Manley, S.A. and Wall, R.E. 1975. Sirococcus strobilinus, a new disease of spruce seedlings in maritime nurseries. Plant Dis. Rep. 59: 664. Maresi, G. and Capretti, P. 1995. Sirococcus blight on Picea abies in northern Italy. In: Shoot and Foliage Diseases in Forest Trees. Ed. by Capretti, P., Heiniger, U. and Stephan, R. Proc. Joint Meeting lUFRO Working Parties 52.06.02 and 52.06.04, Vallombrosa, Italy, pp. 287- 288. Meister, R. 1998. Flug-, Brut- und Entwicklungszyklen des Kupferstechers, Pityogenes chalcographus L. (Col. Scolytidae), inAbhängigkeit vom Überwinterungsstadium. Diploma thesis, Univ. f. Bodenkultur, Wien. Motta, E., Annesi, T. and Balmas, V., 1996. Seedborne fungi in Norway spruce: testing methods and pathogen control by seed dressing. Eur. J. For. Path. 26: 307-314. Neumiiller, A. 1994. Beteiligung von Pilzen am Zweig- und Aststerben der Fichte im Revier Sonnenwald (Böhmerwald). In: Zustandsdiagnose und Sanierungskonzepte fur belastete Waldstandorte in der Böhmischen Masse. Ed. by Fuhrer, E. and Neuhuber F. Forstl. Schriftenreihe Univ. Bodenkultur Wien 7. pp. 171-190. O'Brien, J.T. 1973. Sirococcus shoot blight of Red pine. Plant Dis. Rep. 57: 246-247. Peterson, G.L., 1977: A simplification of the protein assay method of Lowry et al. which is more generally applicable. Analyt. Biochem. 83: 346-356. Smith, R.S. 1973. Sirococcus Tip Dieback of Pinus spp. in California forest nurseries. Plant Dis. Rep. 57(1): 69-73. Redfern, D.B. 1973. Sirococcus strobilinus. Forest Commition London, Rep. Forest Research: 100-101. Schafellner, C., Berger, R., Mattanovich, J. and Fuhrer, E. 1996. Variations in spruce needle chem istry and implications for the Little spruce sawfly, Pristiphora abietina. In: Dynamics of Forest Herbivory: Quest for Pattern and Principle. Ed. by Mattson, W.J., Niemelä, P. and Rousi, M. USDAFor. Serv. Gen. Tech. Rep. NC-183. pp. 248-256. Sutherland, J.R., Lock, W. and Farris, S.H. 1981 .Sirococcus blight: a seed-borne disease of con tainer-grown spruce seedlings in Coastal British Columbia forest nurseries. Can. J. Bot. 59: 559-562. Sutherland, J.R. 1987. Sirococcus blight. In: Cone and Seed Diseases of North American Coni fers. Ed. by Sutherland, J.R., Miller, T. and Quinard, R.S. NAFC Pubi. 1. pp. 34-41. Wall, R.E. and Magasi, L.P. 1976. Environmental factors affecting Sirococcus shoot blight of black spruce. Can. J. For. Res. 6: 448-452. Winter, K., Richter, A., Engelbrecht, 8., Posada, J., Virgo, A. and Popp, M. 1997. Effect of elevated C0 2 on growth and crassulacean-acid-metabolism activity of Kalanchoepinnata under tropical conditions. Planta2ol: 389-396. 158 Nutritional status of mature Norway spruce related to infection by Sirococcus shoot blight Hans Anglberger, Monika Sieghardt, Klaus Katzensteiner and Erhard Halmschlager H. Anglberger, E. Halmschlager, Institute of Forest Entomology, Forest Pathology and Forest Protection, University of Agricultural Sciences, Billrothstraße 53/1/4, A-1190 Vienna, Austria. E-mail: halmi@mail.boku.ac.at M. Sieghardt, K. Katzensteiner, Institute of Forest Ecology, University of Agricultural Sciences, Peter-Jordanstraße 82, A-l 190 Vienna, Austria. Abstract An inventory on the nutritional status of 72 mature Norway spruce trees (Picea abies) revealed significant differences in element contents of the current year and 3-year-old needles between healthy and Sirococcus-diseased trees. Contents of P, Ca and Mg were significantly lower in the current year and 3-year-old needles of diseased trees, and supply of Ca and Mg was insufficient, when compared with threshold values. Diseased trees further showed higher N/Mg and N/Ca ratios compared with healthy trees, indicating unbalanced nutritional status. Keywords: Sirococcus conigenus, shoot blight, Norway spruce, Picea abies, nutrition Introduction The mitosporic fungus Sirococcus conigenus (DC.) P. Cannon & Minter is constantly associated with shoot blight and branch dieback in secondary spruce forests on poor soils in Upper Austria, leading to top and branch mortality of Norway spruce. Symp toms have been observed since the early eighties (Neumiiller 1992, 1994, Halmschlager et al. 1998, 2000 a, 2000b), however, differences in disease severity often occur within a single stand, probably due to different nutritional status of indi vidual trees (Anglberger and Halmschlager 2000, Jandl et al. 2000). To verify this hypothesis, the nutritional status of healthy and diseased mature Norway spruce trees was assessed on a homogenous site. Material and methods Investigations were carried out in a 90-year-old pure Norway spruce stand in the Kobernausser Wald (Fig. 1). Half of the 72 sample trees showed severe crown thin ning due to shoot, twig and branch dieback, caused by S. conigenus whereas the other trees were apparently healthy and vigorous. In December 2000 one branch at 159 Figure 1. Location and site characteristics of the experimental site. the 7th whorl was cut from each sample tree. Current year needles and 3-year-old needles were individually analyzed after oven drying and acidic extraction with HN0 3 - HCI0 4 for K, Ca, Mg, P and S using inductively coupled plasma emission spectrometry (ICP-OES). N-content was determined by semi-micro-Kjeldahl method. Only living needles were used as analytic samples and current year needles used for analyses were taken from shoots expressing no symptoms of Sirococcus shoot blight. Differences between needle element contents of severely affected and healthy trees were tested by Student's t-test using SPSS statistical software, release SPSS 8.0, SPSS Inc. (Kähler 1998). Results and discussion Table 1 shows the macronutrient contents in the current year and 3-year-old needles of P. Abies related to infection. Results obtained from needle analysis revealed sig nificant lower contents of P, Ca and Mg in the current year and 3-year-old needles of diseased trees. Furthermore the nutrient contents of needles from trees severely af fected by S. conigenus reflected insufficient Mg and Ca supply and enhanced N/Mg and N/Ca-ratios, when compared to threshold values given by Htittl (1986) and Stefan (1995), respectively (Tables 1 and 2). The content of P in the current year needles of diseased trees was just at the limit to deficient supply (Table 1). Contents of N, K and S showed no relationship to tree health status, in both current year and 3-year old needles. Increased host susceptibility due to unbalanced nutrition has already been dem onstrated from other fungi, causing shoot blight and canker in conifers. Ylimartimo (1990/1991) reports about increased disease severity on Scots pine seedlings with latent infections by Gremmeniella abietina (Lagerb.) Morelet after increasing N availability. Increased susceptibility of Scots pine to G. abietina related to nutri tional imbalances also has been demonstrated by Barklund (1993). As well the out 160 Table 1. Means of macronutrient contents (mg/g dry weight) in the current year and 3-year-old needles of Picea abies related to infection by S. conigenus. Means followed by different letters are significantly separated at p<0.05 by t-test. 1 Threshold values for current year needles from Hiittl (1986) Table 2. N/Ca and N/Mg ratios in current year needles of Picea abies related to infection by S. conigenus. Means followed by different letters are significantly separated at p0.05) concerning the infection width. There was a large variation between years, which means that this experiment must be repeated some more winters before any conclu sions can be drawn. Snow blight infection vs stand residuals During the winters 1998/99, 99/00, 00/01, totally 45 pair-wise controlled compari sons situated at Svartberget Experimental Station, Vindeln (64°14'N, 19°46'E) have been set up and analysed. The experiment was arranged so that one inoculated 60 cm high Scots pine seedling was placed together with one "secondary seedling" within 30 cm (Fig. 1). This arrangement was repeated both with and without low shelter, consisting of 2 m Birch (Betula sp.) saplings. The result varied highly between the years but as an average the mycelial growth within these "inoculation units" was 31 cm wider (horizontally) in the shelter (pO.001). "Secondary seedlings" with shelter 186 were infected to 40% higher proportion than those standing without shelter (p<0.001). There was now significant difference in snow depth. Figure 1. The "inoculation unit" used by the author in most of the experiment concerning Phacidium infestans in Northern Sweden. As inoculumn 2 g of dry infected needles enclosed in a polyester bag with 1 mm masks is used. The pines are approx. 50 cm high detached Scots pine (Pinus sylvestris) shoots from an infection free area. Figure 2a. Mean snow blight infection height versus clear-cut area in Swedish Lapland the winters 1998/99-2000/01. 187 Figure 2b. Mean snow blight infection width versus clear-cut area in Swedish Lapland the winters 1998/99-2000/01. Figure 2c. Mean snow depth 1998/99-2000/01 at the investigated clear-cuts in Swedish Lapland. Effect of fresh logging slash In an inoculation experiment near Fredrika in Swedish Lapland, (64°03'N, 18°28'E) the mycelial growth with or without fresh logging slash from Scots pine was inves tigated. The slash was fresh in the way that the needles still were green and could serve as nutrient source for the pathogen. Already after one winter the infection width was 17 cm (p<0.001) wider within the slash than without, and this difference 188 was at the same level after next winter although the width now was 62 cm within the slash compared to 47 cm without slash (p<0.001). Effect of high shelter trees At Svartberget Experimental Station, Vindeln (64°14'N, 19°46'E) the mycelial growth of snow blight was tested within a seed tree stand (160 stems/ha) and on an open 1 ha clear-cut during the winter 1998/99-2000/01. In each environment 18 replications of the "inoculation unit" (Fig. 1) were set out each late autumn just before the first snow. In the seed tree stand, the "inoculation units" were placed at three different distances from the seed trees (0.5, 1 and 3 m). The snow depth (mean of 120 meas urements during three winters) was 72 cm on the clear-cut compared to 53 cm in the seed tree stand (pO.001). As an average over all three winters, there was signifi cantly higher (28 cm) snow blight infection on the clear-cut area compared to 18 cm in the shelter (pO.001) and significantly (p<0.001) higher proportion infected sec ondary saplings (0.5) on the clear-cut area compared to in the shelter (0.3). Infection potential through infected needles As a graduate student project, we have studied the ability for the snow blight fungus to infect through ascospores in a environment with or without low shelter. Prelimi nary data indicate that ascospore infection occur at least 40 m from the source. In a wind thunnel experiment we have collected ascospores at a distance of 100 m from the source. 189 Phylogenetic analysis of Guignardia cryptomeriae based on rDNA-ITS sequences Shun-Ichiro Miyashita and Toshihiro Yamada S. Miyashita, Kansai Research Center, Forestry and Forest Products Research Institute, Momoyama, Fushimi, Kyoto 612-0855, Japan. E-mail: miyaty@affrc.go.jp T. Yamada, Graduate School of Agricultural and Life Sciences, The University of Tokyo, Tokyo 113-8657, Japan. Abstract In order to infer the taxonomic placement of G. cryptomeriae, we determined the sequences of the rDNA-ITS region for isolates of G. cryptomeriae and some related fungi. Phylogenetic analysis among these sequences revealed that G cryptomeriae was not distinct from Botryosphaeria dothidea and B. berengeriana. Phylogenetic analysis among these sequences and Botryosphaeria sequences obtained from the DNA databank showed that all isolates of G cryptomeriae were included in a Botryosphaeria main clade. These results strongly suggested that G cryptomeriae should be trans ferred to the genus Botryosphaeria. Introduction Shoot blight is one of the most important diseases of the Japanese cedar (Cryptomeria japonica). The causal agent, Guignardia cryptomeriae, was first described in 1950 (Sawada 1950). Some taxonomists assumed that this fungus might be placed in the genus Botryosphaeria, but further studies have not been conducted. Consequently, the taxonomy of this fungus remains unclear. The purpose of this study is to infer the taxonomic placement of G. cryptomeriae based on molecular phylogenetic analysis. We determined sequences of the rDNA ITS region for isolates of G. cryptomeriae, B. laricina, which is a causal agent of shoot blight in the Japanese larch (Larix kaempferi), an undescribed species of Botryosphaeria (tentatively nicknamed the "Botryosphaeria large-spore species"), which is a causal agent of shoot blight in the Japanese cypress (Chamaecyparis obtusa), and two species of Botryosphaeria (B. dothidea and B. berengeriana), which were isolated from several broadleaved trees. Phylogenetic analyses were made among these sequences and additional sequence data of Botryosphaeria spp. taken from the DNA databank. 190 Material and methods Fungal strains and DNA isolation The cultures that were used in this research are listed in Table 1. Total genomic DNA was extracted from mycelium grown in PD broth by the method of Lee and Taylor (Lee and Taylor 1990). Table 1. Isolates used in this study. PCR amplification and sequencing The nuclear ribosomal ITSI-5.85-ITS2 region was amplified by polymerase chain reaction (PCR) using conditions and primers (ITSS and ITS 4) described in White et al. (1990). Nucleotide sequences of the PCR products were obtained by an Applied Biosystems 377 DNA sequencer, using the Big Dye Terminator Reaction Kit (Ap plied Biosystems). Sequence analysis Sequences were aligned using the Clustal W (Thompson et al. 1994). The alignment was checked visually, and minor adjustments were made manually. Phylogenetic analysis was performed with PA UP 4.0 (Swofford 2000) for maximum parsimony. Isolate Host G. cryptomeriae MA3 Japanese cedar MA7 Japanese cedar MA12 Japanese cedar MA 18 Japanese cedar MA19 Japanese cedar MA8 Japanese cypress MA 10 Japanese cypress MA21 Japanese cypress M-T-l Japanese thuja B. laricina GC56 Japanese larch GC59 Japanese larch GC78 Japanese larch GC79 Japanese larch Botryosphaeria sp. B-H-la Japanese cypress (--Botryosphaeria large-spore species "). B-H-3 Japanese cypress B-H-5 Japanese cypress B. dothidea M-C-l Cherry tree M-C-6 Cherry tree M-C-l 1 Cherry tree B-MU-1 Japanese apricot B-JP-1 Japanese pear B. berengeriana B-P-l Peach B-P-8 Peach 191 Results PCR amplification of genomic DNA using the primers ITSS and ITS 4 generated a product of approximately 500 bp for all the isolates. The phylogenetic relationships among G. cryptomeriae and Botryosphaeria spp. were inferred from heuristic parsi mony analysis of the aligned ITS sequences. The analysis showed that tested isolates were divided into four major groups. The first group consisted of 4 isolates of G. cryptomeriae, 4 isolates of B. dothidea, and all isolates of B. berengeriana. The second group consisted of 5 isolates of G. cryptomeriae and 1 isolate of B. dothidea. The third group consisted of all isolates of B. laricina. The fourth group consisted of all isolates of the "Botryosphaeria large-spore species". The sequence data of G. cryptomeriae were aligned and analyzed with Botryosphaeria spp. sequences obtained from the DNA databank. The results showed that all isolates of G. cryptomeriae were included in a Botryosphaeria main clade. Discussion Although isolates of G. cryptomeriae were divided into two groups, both groups contained isolates of Botryosphaeria. Furthermore, phylogenetic analysis with Botryosphaeria sequences obtained from the DNA databank showed that all isolates of G. cryptomeriae were included in a Botryosphaeria main clade. These results strongly suggest that G. cryptomeriae should be transferred to the genus Botryosphaeria. Acknowledgments We thank Dr. S. Ito of Mie University and Dr. S. Kanematsu of the National Institute of Fruit Tree Science for providing the fungal isolates. References Lee, S. B. and Taylor, J. W. 1990. Isolation of DNA from fungal mycelia and single spores. In PCR: Protocols: A Guide to Methods and Applications. Eds. N. Innis, D. Gelfand, J. Sninsky, and T. White. Academic Press: New York, NY, U.S.A. pp. 282-287. Sawada, K. 1950. Fungi inhabiting on conifers in the Tohoku district. I. Fungi on "sugi" (■Cryptomeria japonica D. Don). Bull. Gov. For. Exp. Stn. (Tokyo), 45: 27-53. Swofford, D. L. 2000. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4. Sinauer Associates, Sunderland, Massachusetts. Thompson, J. D., Higgins, D. G., and Gibson, T. J. 1994. CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Research, 22: 4673-4680. White, T. J., Bruns, T., Lee, S., and Taylor, J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In PCR: Protocols: A Guide to Methods and Applications. Eds. N. Innis, D. Gelfand, J. Sninsky, and T. White. Academic Press: New York, NY, U.S.A. pp. 315-322. 192 Diversity of endophytic fungi of single Norway spruce needles and their role as pioneer decomposers Jarkko Hantula and Michael Mtiller The Finnish Forest Research Institute, P.O. Box 18, 01301 Vantaa, Finland. E-mails: jarkko.hantula@metla.fi, michael.mueller@metla.fi Abstract The diversity of endophytic fungi within single symptomless Norway spruce needles was found to be low at species level as 90% of isolates were identified as Lophodermium piceae. The number of individual infections was, however, high. The diversity within detached needles that retained their green color had similar diversity even after a long incubation, but after the loss of colour the needles went through a considerable population change and became inhabited by only one or few individuals belonging to species other than L. piceae. The ability of needle endophytes to act as pioneer decomposers was found to be considerable, which, however, does not concern L. piceae as this fungus was not found to persist in needles during decomposition. Keywords: endophytic needle fungi, decomposition, Lophodermium piceae, Norway spruce, ITS RFLP, RAMS Introduction In Norway spruce (Picea abies (L.) Karst.) the highest diversity of fungi appears to be found in needles as epiphytes and endophytes (Miiller and Hallaksela 2000). Sieber (1988) detected one hundred endophytes from Norway spruce needles in Switzer land. The most frequently occurring species is Lophodermium piceae (Fuckel) Höhn. inhabiting often more than 50% of the needles (Barklund 1987, Sieber 1988, Solheim 1994, Miiller and Hallaksela 1998 a). Other less known endophytes occurring in nee dles of Norway spruce are Tiarosporella parca (Berk. & Broome), Sclerophoma pythiophila (Corda) Höhn., Rhizosphaera kalkhoffii Bubäk, Thysanophora penicillioides (Roum.) W. B. Kendr. and Lirula macrospora (R. Hartig) Darker (Sieber 1988, Solheim 1994, Miiller and Hallaksela 1998 a). The significance of L. piceae and other Norway spruce needle endophytes to their host has not been solved, but it is clearly not an opportunistic pathogen as it occurs more commonly in young, healthy stands than in stressed, diseased ones (Barklund 1987, Sieber 1988). So far it is not known whether these endophytes are activated from their dormant state within the needle by unusual events like herbivory, attacks by other microbes or by needle senescence. Neither do we know whether these endophytes protect needles against insects or needle pathogens, allow redistri 193 bution of nutrients prior to needle cast or are just "waiting" for needle senescence to be the first pioneers during needle decomposition, the last hypothesis of which is tested in a study published recently (Miiller et al. 2001). Here we review shortly the results and provide an additional illustration of that study. Results Fresh green needles In the analysis of Norway spruce endophyte diversity Miiller et al. (2001) isolated fungi from thin (thickness 0.2 (am) slices of surface sterilized healthy needles. 52% of fresh green needles were infected with endophytic fungi. The fungi were identi fied to species level by comparing their restriction fragment length polymorphism (RFLP) patterns of PCR-amplified Internal Transcribed Spacer (ITS) sequences to known reference isolates. Using restriction enzymes Mspl and Cfol it was possible to identify the majority, i.e. 147 out of 182, isolates obtained from fresh green nee dles. In single densely infected needles up to eight different ITS-types could be found. The majority of isolated endophytes had similar ITS-patterns as found among the L. piceae reference isolates (Miiller et al. 2001). The within species diversity of fungi was analyzed using random amplified microsatellite (RAMS) analysis (Zietkiewicz et al. 1994, Hantula et al. 1996) with two primers (Miiller et al. 2001). In this analysis even isolates from adjacent needle slices were almost without exception of different haplotypes (see Fig. 1). Detached needles Changes in endophyte communities of detached needles were analyzed from sam ples incubated on sterilized soil. L. piceae remained dominant after needle detach ment as long as the needles retained their green color. When the needles became Figure 1. Examples of endophytic infections within single healthy green needles. 194 brown L. piceae made way for T. parca, S. pythiophila and other (unidentified) endophytes, most of which were not found in fresh green needles (Miiller et al. 2001). In the analysis of Miiller et al. (2001) the fungal diversity within needles re mained extremely high (i.e. only genetically dissimilar isolates were found) during incubation as long as the needles were green or partly green. After turning brown or black, single fungal haplotypes overgrew the needle and all isolates had the same or only few haplotypes. Endophytes as primary decomposers The role of endophytes in primary decomposition was measured as a difference be tween weight loss of needles incubated on sterilized soil and needles incubated on live soil. After twenty weeks incubation at 15° C a dry weight loss of 13-17% was found in initially green symptomless needles. The dry weight loss of needles on the live soil was somewhat higher after 8 and 20 weeks of incubation but the difference was not statistically significant (Miiller et al. 2001). On average a proportion of 13% of slices of fresh green needles were infected at the start of the incubation experiment. The infected proportion did not change significantly in those needles which remained green during the 20 weeks incubation but in decoloured needles the infected proportion increased considerably (ending from 36% to 93%; Miiller et al. 2001). Conclusions In the analysis of Norway spruce endophytes (Miiller et al. 2001) the two hypotheses suggested by Carroll and Petrini (1983) for endophytes proved true regarding to the dominant Norway spruce needle endophyte L. piceae. Norway spruce needles com monly inhabit numerous genetically different endophytic individuals, the majority of which do not contribute significantly to the decomposition of the needles after their abscission. These results are in agreement with earlier hypotheses made by Carroll and Petrini (1983) based on variation in substrate utilization capacity among isolates of single species in single leaves. However, the study produced another as important but highly unexpected result. This was the drastic decrease of fungal geno type diversity in needles during the early phase of the decomposition process. This suggests that the endophytic mycota of plants may in fact be more diverse in terms of the numbers of different genotypes than the saprophytic mycota of the same tissue. It would be important to test whether a similar result would also be obtained from other plant systems. The results of Miiller et al. (2001) confirmed that L. piceae is the most fre quent endophyte in needles of P. Abies and that it is a highly diverse species (Miiller and Hallaksela 1998 a). Further information was obtained from individual level di versity within single needles, which shows that endophytic infections occupy only a small spatial domain of the needle and the needles may thus accommodate a large 195 number of genetically different fungal individuals. This infection pattern is essen tially similar to that of R. parkeri in needles of Douglas fir (McCutcheon et al. 1993, Carroll 1995). As L. piceae disappear from decaying needles it clearly has no marked role in the decomposition of spruce needles and is likely to become overgrown by other fungi after needle cast. This is supported by observations that fruiting of L. piceae on fallen brown needles is less common than could be expected from the infection frequency of L. piceae on green intact needles (Solheim 1994). Thus, it can be con cluded that other less frequent endophytes of Norway spruce needles like S. pythiophila and T. parca seem to be responsible for the initial decomposition proc ess. The ecological function of L. piceae in needles of Abies alba may be different from that in P. Abies as microscopic investigations of decomposing fir needles sug gest L. piceae to be a primary saprophyte (Gourbiere et al. 1999). The result of Miiller et al. (2001) on the significance of endophytic needle fungi during the initial needle decomposition is in accordance with earlier results obtained by Mitchell and Millar (1978) on the decomposition of Corsican pine nee dles. However, the saprophytic ability of P. Abies needle endophytes varies largely (Miiller et al. 2001). This is in good agreement with earlier findings on needle endophytes of A. alba (Canavesi 1987). The overall results of Miiller et al (2001) strengthen the hypothesis that L. piceae occupies a similar ecological niche in Norway spruce needles as R. parkeri in needles of Douglas fir because L. piceae (1) occupies small spatial domains in the needles, (2) is genetically highly diverse and (3) seems not to have an essential role during needle decomposition (Carroll 1995, McCutcheon et al. 1993). Taking this in consideration and remembering that R. parkeri has been shown to act antagonisti cally toward gall midges (Contarinia spp.) in Douglas fir (Pseudotsuga menziesii) needles (Carroll 1986), it would be intriguing to speculate whether L. piceae would have a similar role in Norway spruce. Evidence for a mutualistic relationship be tween L. piceae and Norway spruce, however, is not available. References Barklund, P. 1987. Occurrence and pathogenicity of Lophodermium piceae appearing as an endophyte in needles of Picea abies. Transactions of the British mycological Society 89: 307-313. Canavesi, F. 1987. Beziehungen zwischen endophytischen Pilzen von Abies alba Mill, und den Pilzen der Nadelstreue. PhD thesis. Swiss Federal Institute of Technology, ETH, No. 8325, Ztirich, Switzerland. Carroll, G 1986. The biology of endophytism in plants with particular reference to woody perennials. In: Microbiology ofthe phylloplane. Eds. Fokkema, N.J. and van den Heuvel, J. Cambridge University Press, Cambridge, England, pp. 205-222. Carroll, G. 1995. Forest endophytes: pattern and process. Canadian Journal of Botany 73: 51316- 51324. Carroll, G and Petrini, O. 1983. Patterns of substrate utilization by some fungal endophytes from coniferous foliage. Mycologia 75: 53-63. Gourbiere, F., Pöpin, R. and Bernillon, D. 1999. Microscopie de la mycoflore des aiguilles de sapin (Abies alba). 11. Lophodermium piceae. Canadian Journal of Botany 64: 102-107. 196 Hantula, J., Dusabenyagasani, M. and Hamelin, R 1996. Random amplified microsatellites (RAMS) - a novel method for characterizing genetic variation within fungi. European Journal of Forest Pathology 12: 167-174. McCutcheon, T.L., Carroll, G.C. and Schwab, S. 1993. Genotypic diversity in populations of a fungal endophyte from Douglas fir. Mycologia 85. 180-186. Mitchell, C.R and Millar, C.S. 1978. Mycofloral successions on Corsican pine needles colonized on the tree by three different fungi. Transactions of the British mycological Society 71: 303- 317. Mttller, M.M. and Hallaksela, A.-M. 1998 a. Diversity of Norway spruce needle endophytes in various mixed and pure Norway spruce stands. Mycological Research 102: 1163-1168. Mtiller, M.M. and Hallaksela, A.-M. 2000. Fungal diversity in Norway spruce - a case study. Mycological Research 104: 1139-1145. Mtiller, M.M., Valjakka, R., Suokko, A. and Hantula, J. 2001. Diversity of endophytic fungi of single Norway spruce needles and their role as pioneer decomposers. Molecular Ecology 10: 1801-1810. Sieber, T. 1988. Endophytische Pilze in Nadeln von gesunden und geschädigten Fichten (Picea abies (L.) Karsten). European Journal of Forest Pathology 18: 321-342. Solheim, H. 1994. Occurrence of the Norway spruce needle endophytes Lophodermium piceae and Tiarosporella parca at the permanent plots of the Norwegian monitoring programme for forest damage. In: Shoot and Foliage Diseases in Forest Trees. Proceedings of a Joint Meeting of the Working Parties Canker and Shoot Blight of Conifers and Foliage Diseases. Eds. Capretti, P., Heiniger, U. and Stephan, R. Vallombrosa, Firenze, Italy June 6-11, 1994. pp. 17-21 Zietkiewicz, E., Rafalski, A. and Labuda, D. 1994. Genome fingerprinting by simple sequence repeat (SSR)-anchored polymerase chain reaction amplification. Genomics 20: 176-183. 197 Black yeast-like fungi of Norway spruce Michael M. Miiller, Jarkko Hantula and Anna-Maija Hallaksela Finnish Forest Research Institute, Vantaa Research Centre, RO.Box 18, 01301 Vantaa, Finland. E-mail: michael.mueller@metla.fi Abstract During two studies on the diversity of endophytic fungi in Norway spruce (Picea abies) we ob tained 66 dark yeast-like fungal isolates from wood and needles. Some of these isolates could be identified by their morphology as Exophiala sp. and Hortaea werneckii. Representatives of these genera are known to be human pathogens causing skin mycoses. We characterized ourisolates and several culture collection (CBS) strains of these genera using restriction fragment length polymor phism (RFLP) analysis of their 18S rDNA. According to their RFLP-patterns the isolates showed to be diverse. None of them are closely related to culture collection strains of H. Werneckii or of five Exophiala species used for comparison in this study. Black yeast-like fungi could not be cultured from spruce timber avail able in timber yards. 198 Needle disease on Taxus baccata caused by Cryptocline taxicola Alfred Wulf and Leo Pehl Institute for Plant Protection in Forests, Federal Biological Research Centre for Agriculture and Forestry, Messeweg 11/12, D-38104 Braunschweig. E-mail: a.wulf@bba.de Abstract The paper reports a conspicuous appearance of Cryptocline taxicola at several locations in Ger many in the year 2000. The fungus is one of the few parasitic fungi on English Yew and its charac teristics and the symptoms of the needle disease are described. Keywords: Cryptocline taxicola, Taxus baccata, English Yew, needle disease Introduction In contrast to other European tree species, English Yew (Taxus baccata) has a con spicuously low affinity for fungi (Schiitt et al. 1994). Not only does this species apparently grow entirely without the presence of mycorrhizal fungi, but it also ex hibits little tolerance for substrate-specific saprophytic or parasitic fungi (De Vries et ai. 1996). Therefore it is all the more remarkable that a characteristic needle fun gus with obvious parasitic potential has occurred more often in recent years. The appearance of a needle disease on English Yew, which was found at several places in Germany and caused several requests for determination especially in the year 2000, seems to be worth a short communication. Information on the disease symptoms and on the causal fungus are presented. Material and methods The material for examination was taken from an infected tree in Hamburg, which is described in detail by the portrayal of a botanical garden (Biederstedt 2001). Inves tigations of the symptoms was carried out macroscopically by hand lens while the examination of the fungal structures required histological preparation. After embed ding needle pieces in glycol-methacrylate (GMA, Kulzer method) needle cross sec tions of 8 jim were prepared and stained with a 1 % aqueous solution of thionine. For cultivation of the fungus malt-extract-agar (MEA) containing 2% malt and 2% agar was used, and cultures were stored at 21 °C in 7 cm Petri-dishes. 199 Results Symptoms The first disease symptoms consist of single necrotic spots mainly on the current year's needles. Not all needles are damaged to the same extent, heavily damaged needles being found in the direct vicinity of green, apparently healthy needles (Fig. 1). However, as the disease progresses, the affected shoots turn completely brown and both the upper and under side of needles is covered by numerous black fruit bodies of the fungus (Fig. 2). In the presence of sufficient moisture, the conidia of the fungus become visible as a white to cream-colored pustule on the ripe acervuli which rupture the epidermis (Figs. 3,4). During dry conditions, the fruit bodies have a black, shrunken appearance (Fig. 5). Figure 1. Figure 2. Figure 3. Figure 4. 200 Causal fungus The dark gray to black, spherical fruit bodies, measuring 250-450 in diameter and up to 150 |im in height, develop beneath the epidermis, as shown in the cross section (Fig. 6). The needle epidermis is first pushed upwards in a pustular manner and then ruptures irregularly when the fruit bodies become ripe (Figs. 4, 5). During spells of wet weather the egg-shaped to elliptical, smooth-walled and hyaline co nidia are produced, measuring 12-18 x 5-8 (Fig. 7). The determination of the fungus based on its demonstrated characteristics leads to Cryptocline taxicola (All.) Petr (Petrak 1925). On malt extract agar, the fungus exhibits felty to woolly, olive to gray-brown aerial mycelium (Fig. 8) and shows linear growth of 1,4 mm per day at 21° C on MEA. Figure 5. Figure 6. Figure 7. Figure 8. Discussion 201 The spectrum of fungi reported specifically for English Yew in literature is small and limited mainly to the following species: Botryosphaeriafoliorum, Chaenothecopsis caespitosa, Diplodia taxi, Dothiora taxicola, Guignardia philoprina, Phacidium taxicola, Physalospora gregaria, Phytophthora cinnamomi (Brandenburger 1985, De Vries et ai. 1996, Ellis and Ellis 1997, Peace, 1962, Schutt et al. 1994, Von Arx and Miiller 1954). Few of these species seem to be able to threaten their host tree seriously. Among the needle fungi only Diplodia taxi is reported to possibly have lethal effects (Schutt et al. 1994). Cryptocline taxicola (All.) Petr. was first described in 1896 as Gloeosporium taxicola All., and PETRAK placed it in the genus Cryptocline in 1925 (Petrak 1925). Among the needle diseases it was judged to be of less importance. Recent observa tions show that it obviously possesses a larger degree of parasitic ability. Further investigations will show if this disease can be more harmful than previously thought. References Biederstedt, K. 2001. Parkpflegewerk: Arboretum Marienhof Poppenbiittel - Maßnahmenkatalog, Bepflanzungskonzepte, Konstruktionsdetails. Diploma thesis, Fachbereich Landschaftsarchitektur, Fachhochschule Osnabruck. pp. 53-54. Brandenburger, W. 1985. Parasitische Pize an Gefäßpflanzen in Europa. Gustav Fischer Verlag, Stuttgart, New York, pp. 32-33. De Vries, 8.W.L., Kuyper, T.W. and Schmid, H. 1996. Eibenbegleitende Pilze. In: Kölbel, M. and Schmidt, O. (Hrsg.) 1996. Beiträge zur Eibe. Berichte aus der LWF, Nr. 10. Bayerische Landesanstalt ftir Wald und Forstwirtschaft, Freising. Ellis, M.B. and Ellis J.P. 1997. Microfungi on land plants. The Richmond Publishing Co. Ltd., Slough, England, pp. 258-259. Peace, T. R. 1962. Pathology of trees and shrubs. Oxford at the clarendon press, 367-368. Petrak, F. 1925. Mykologische Notizen VIII. In: Annales Mycologici. Eds. H. Sydow. pp. 23-24. SchUtt, P., Schuck, H.J., Aas, G., Lang, U.M. 1994. Enzyklopädie der Holzgewächse: Handbuch und Atlas der Dendrologie. Ecomed Verlagsgesellschaft, Landsberg. pp. 1-11. Von Arx, J.A., Muller E. 1954. Die Gattungen der amerosporen Pyrenomyceten. Beiträge zur Kryptogamenflora der Schweiz, 11, pp. 42. ISBN 951-40-1809-5 ISSN 0358-4283 Hakapaino 2002