METSÄNTUTKIMUSLAITOKSEN TIEDONANTOJA 864, 2002 THE FINNISH FOREST RESEARCH INSTITUTE, RESEARCH PAPERS 864,2002 Interactions between host trees and fungi associated with the spruce bark beetle (Ips typographus) Heli Viiri SUONENJOEN TUTKIMUSASEMA SUONENJOKI RESEARCH STATION METSÄNTUTKIMUSLAITOKSEN TIEDONANTOJA 864, 2002 The Finnish Forest Research Institute, Research Papers 864,2002 Interactions between host trees and fungi associated with the spruce bark beetle (Ips typographus) Heli Viiri ACADEMIC DISSERTATION IN FOREST PROTECTION To be presented, with the permission of the Faculty of Forestry of the University of Joensuu, for public criticism in Auditorium 81, of the Natura Building,Yliopistokatu 7, Joensuu, on November I st 2002, at 12 o'clock noon. SUONENJOEN TUTKIMUSASEMA SUONENJOKI RESEARCH STATION Supervisor: Professor Pekka Niemelä Faculty of Forestry University of Joensuu Joensuu, Finland Reviewers: Professor Riitta Julkunen-Tiitto Department of Biology University of Joensuu Joensuu, Finland Professor Michael Wingfield Faculty of Biological and Agricultural Sciences University of Pretoria Pretoria, Republic of South Africa Opponent: Professor Bo Längström Department of Entomology Swedish University of Agricultural Sciences Uppsala, Sweden Available at: The Finnish Forest Research Institute Library P.0.80x 18 FIN-01301 Vantaa Phone: +358 9 8570 5580 Fax: +358 9 8570 5582 Email: kirjasto@metla.fi Publisher: The Finnish Forest Research Institute Suonenjoki Research Station Lay-out: Reija Viinanen Cover photos: Ips typographies infested Norway spruce, Heli Viiri Accepted by: Kari Mielikäinen, Research Director 10.10.2002 ISBN 951 -40-1846-X ISSN 0358-4283 Joensuun yliopistopaino Joensuu 2002 Abstract Viiri, H. 2002. Interactions between host trees and fungi associated with the spruce bark beetle (Ips typographus ). The Finnish Forest Research Institute, Research Papers 864. ISBN 951-40-1 846-X. ISSN 0358-4283. Among the European scolytids the spruce bark beetle, Ips typographus, is the most important species in conifers. Some of the fungi associated with spruce bark beetles have a role in overwhelming the resistance of vigorous Norway spruce, Picea abies. The species composition of ophiostomatoid fungi associated with I. Typographus was investigated at different population levels of beetle in Finland and France. The fungal flora varied at different study areas. In Finland the most frequent ophiostomatoid species were Ophiostoma penicillatum, O. piceaperdum and O. bicolor. In France, in addition of O. bicolor and O. penicillatum, Ceratocystis polonica and Ceratocystiopsis minuta were frequent species. The frequency of highly pathogenic C. polonica was lower in Finland than in post-epidemic areas of spruce bark beetle in France. Interactions between the host and the pathogen were studied after artificial inoculation of Norway spruce with C. polonica, the most pathogenic fungus associated with I. typographus. In experiment, plots of mature Norway spruce were fertilized with nitrogen, phosphorus or combination of nitrogen, phosphorus and potassium. One year after fertilization the trees were artificially infected with C. polonica. Changes in the main secondary compounds of Norway spruce, the stilbenes and terpenes, and soluble carbohydrate concentrations of phloem were studied in relation to the nutrient status of trees. The response of stilbenes to fungal inoculation was qualitative. The concentration of stilbene glycosides in the phloem decreased. Corresponding stilbene aglycones were more frequent inside the reaction lesion. Fungal inoculation caused a strong quantitative response in terpenes. The total terpene concentration of the phloem increased to almost 100 times greater near the inoculation site compared to the constitutive values. N fertilization significantly reduced the total terpene and total stilbene aglycone concentrations near the inoculation sites. Thus, N fertilization may reduce the ability of Norway spruce to defend itself against fungal pathogens. The concentration of total soluble carbohydrates in the outer border of the lesion was significantly decreased in P-fertilized trees compared to corresponding unfertilized trees. However, changes in soluble carbohydrate concentration caused by fungal inoculation were more pronounced than changes caused by fertilization. The main soluble carbohydrate was sucrose. Near the site of fungal inoculation the concentration of total soluble carbohydrates decreased significantly compared to corresponding values in unwounded phloem. N and NPK fertilization treatments increased radial growth of the stem and the vigour indices. Despite the increased radial growth of the stem, the only indication that enhanced growth might reduce the level of resistance was the modest positive correlation between lesion length and radial growth of the stem. Key words: Ceratocystis, induced defense, monoterpenes, Ophiostoma, phenolics, Picea abies, soluble carbohydrates, stilbenes Author's address: Heli Viiri, The Finnish Forest Research Institute, Suonenjoki Research Station, Juntintie 40, FIN-77600 Suonenjoki, Finland. Tel. +358 17 5138 219, Fax: +358 17 513 068. E-mail: heli.viiri@metla.fi. Tiivistelmä Viiri, H. 2002. Kirjanpainajan (Ips typographus) seuralaissienten ja kuusen väliset vuorovaikutussuhteet. Metsäntutkimuslaitoksen tiedonantoja 864. ISBN 951-40- 1846-X. ISSN 0358-4283. Kirjanpainaja (Ips typographus) on kuusella esiintyvä tuhohyönteinen, joka kuljettaa mukanaan puuhun patogeenisiä värivikaa aiheuttavia sieniä. Tässä työssä tutkittiin kirjanpainajan pyydystysmenetelmien vaikutusta seuralaissienilajiston rakenteeseen (osajulkaisu I), seuralaissienilajistoa Suomessa ja Ranskassa (osajulkaisut I ja II), sekä patogeenisen seuralaissienen (Ceratocystis polonica) aiheuttamia muutoksia kuusen terpenoidien, stilbeenien ja liukoisten hiilihydraattien määrissä erilaisilla lannoitetasoilla (osajulkaisut 111 ja IV). Sienilajisto vaihteli eri tutkimusalueilla. Suomessa kirjanpainajan endeemisellä esiintymisalueella yleisiä seuralaissieniä olivat Ophiostoma penicillatum, O. piceaperdum and O. bicolor. Ranskassa kirjanpainajaepidemian jälkeen O. bicolor, O. penicillatum, Ceratocystis polonica ja Ceratocystiopsis minuta olivat yleisiä seuralaissieniä. Erittäin patogeeniseksi todettu C. polonica oli harvinaisempi Suomessa kuin Ranskassa. Isäntäpuun ja patogeenisen sienen välisiä vuorovaikutussuhteita tutkittiin ymppäämällä keinotekoisesti C. polonica sientä eläviin terveisiin kuusiin. Koepuita oli lannoitettu vuosi ennen sieni-infektiota typellä, fosforilla tai NPK-lannoitteella. Rungon sekundääriaineista stilbeenit reagoivat sieniymppäykseen kvalitatiivisesti, mutta terpeenit kvantitatiivisesti. Terpeenien määrä lisääntyi yli 100-kertaiseksi lähellä sieni-infektion kohtaa vioittamattomiin puihin verrattuna. Stilbeenien glykosidien määrä aleni lähellä sienen ymppäyskohtaa merkittävästi. Lisäksi stilbeenien aglykoneita tavattiin merkittävästi useammin lähellä sienen ymppäyskohtaa kuin vioittamattomassa nilassa. Typpilannoitus alensi kokonaisterpeenien ja stilbeenien aglykonien kokonaismäärää infektiokohdan läheisyydessä, mikä osoittaa, että typpilannoitus voi heikentää kuusen kestävyyttä tuhonaiheuttajia vastaan. Sieniymppäyksen aiheuttamat muutokset liukoisissa hiilihydraateissa olivat merkittävimpiä kuin lannoituksen aiheuttamat muutokset. Lähellä sienen ymppäyskohtaa liukoisten hiilihydraattien kokonaismäärä aleni merkittävästi verrattuna vioittamattomien puiden nilaan. Yleisin liukoinen hiilihydraatti oli sakkaroosi. Typpi- ja NPK-lannoitukset lisäsivät rungon paksuuskasvua ja puiden elinvoimaisuusindeksiä. Avainsanat: Ceratocystis, fenoliset yhdisteet, indusoitunut puolustus, kuusi, liukoiset hiilihydraatit, monoterpeenit, Ophiostoma, stilbeenit Tekijän osoite: Heli Viiri, Metsäntutkimuslaitos, Suonenjoen tutkimusasema, Juntintie 40, 77600 Suonenjoki. Puh. 017-5138 219, Fax 017-513 068. Sähköposti: Heli.Viiri@metla.fi. Preface Numerous people in several institutions have been helpful during the course of this thesis. I acknowledge greatly the help of following persons and institutions. Kim von Weissenberg (currently University of Helsinki), Erik Christiansen and Halvor Solheim (Norwegian Forest Research Institute, As) and Susanne Harding (Royal Veterinary and Agricultural University, Copenhagen) guided me years ago as under-graduated student to the fascinating world of the bark beetles associated fungi. They have given valuable advice in early phase and your work has given inspiration during the whole project. This work has been financially supported mainly (1995-1999) by the Graduate School of Forest Sciences, Ministry of Education. Further funds have been provided: Finnish Forest Research Institute; Faculty of Forestry and Centre of Excellence Programme, University of Joensuu; The Finnish Society of Forest Science and Niemi Foundation. I express my thanks for the financial support. Pekka Niemelä has been supervisor of this thesis. I thank you from encouragement and support during these years. You have been a good company. Esko Valtonen gave valuable advice for statistical analysis for Paper I. Jacques Garcia, Paul Romary, Annie Yart (Institut National de la Recherche Agronomique, Orleans) helped in practical issues during my stay in Orleans. Konkordia Foundation and Emil Aaltonen Foundation supported my visit to France. Co-authors Erkki Annila, Veikko Kitunen and Frarujois Lieutier widened my original aims to current form of this thesis. I am also grateful to Erik Christiansen, Markku Keinänen, Maarit Kytö, William Mattson, Vladimir Ossipov, Halvor Solheim and Elina Vapaavuori for improving suggestions concerning manuscripts. Joann von Weissenberg helped to correct the English language and Reija Viinanen made the text lay-out. In last phase, since May 2000 my current immediate superiors at Suonenjoki Research Station, Heikki Smolander and Marja Poteri have been cheering superiors. Thank you from pushing me to write the final dot. Marja also give constructive criticism about the whole bunch. Riitta Julkunen-Tiitto and Michael Wingfield are acknowledged for their work in reviewing the manuscript. Finally I want to express my gratitude to my family. My parents and parents-in-law have been helpful during these years. My husband Mika Räty, helped with some statistical issues, but mainly I appreciate your efforts to keep my mind and home clean about scientific papers. This thesis is dedicated to our daughter, Vilma, the sparkling light of our lives. Suonenjoki, 17 th September 2002 Heli Viiri Abbreviations List of abbreviations used in the text, presented in alphabetical order. Journal names are mainly abbreviated according to instructions of the ISI Journal Abbreviation Index, http://www.nal.usda.gov/indexing/ lji99/ and the International Organization for Standardization, International Serials Data System, CIEPS Paris 1985, ISBN 2-904938- 02-8. Symbol: Description: BA! cross-sectional area of the current annual ring BAySA vigour index, the ratio of the cross-sectional area of the current annual ring and the sapwood basal area at breast height C-based carbon-based CHN analyser carbon-hydrogen-nitrogen elemental analyser CNB hypothesis carbon/nutrient balance hypothesis C/N ratio carbon/nitrogen ratio CBSCs carbon-based secondary compounds DBH diameter at breast height (1.3 m) de novo novus = new, de novo synthesis of organic compound in biochemical reaction DF, df degree of freedom e.g. exempli gratia, for example GC gas chromatography GDB hypothesis growth/differentiation balance hypothesis ICP-AES inductively coupled plasma emission spectrometer in vitro outside the living organism MS mass spectrometry MT-type Myrtillus-type in the Finnish forest soil-type classification N nitrogen n sample size N-based nitrogen-based NPK nitrogen-phosphorous-potassium P phosphorous PAL phenylalanine ammonia lyase PP cells polyphenolic parenchyma cells SA sapwood basal area at breast height (1.3 m) SD standard deviation of series s.s. sensu stricto List of original articles This thesis is based on the following articles, which are referred to in the text by Roman numerals I-IV. The published articles are reprinted with kind permission of the publishers. I Viiri, H. 1997. Fungal associates of the spruce bark beetle Ips typographus L. (Col. Scolytidae) in relation to different trapping methods. Journal of Applied Entomology 121: 529- 533. II Viiri, H. and Lieutier, F. Ophiostomatoid fungi associated with the spruce bark beetle, Ips typographus, in post epidemic areas in France. (Submitted). 111 Viiri, H., Annila, E., Kitunen, V. & Niemelä, P. 2001. Induced responses in stilbenes and terpenes in fertilized Norway spruce after inoculation with blue-stain fungus, Ceratocystis polonica. Trees - Structure and Function 15: 112-122. IV Viiri, H., Niemelä, P., Kitunen, V. & Annila, E. 2001. Soluble carbohydrates, radial growth and vigour of fertilized Norway spruce after inoculation with blue-stain fungus, Ceratocystis polonica. Trees - Structure and Function 15: 327-334. In Study II all stages were planned and carried out by H. Viiri; F. Lieutier assisted in selecting study areas and collecting the beetles. For Papers 111-IV E. Annila and P. Niemelä were in charge of organising the field experiment; V. Kitunen was in charge of the chemical analyses and H. Viiri analysed the results and submitted the papers. Contents Introduction 9 Aims of the study 11 1. Spruce bark beetle-fungi-host tree interaction - a review 12 1.1 Spruce bark beetle 12 1.1.1 Damage 12 1.1.2 Ecology 13 1.2 Associated fungi 15 1.2.1 Ecology 15 1.2.2 Taxonomy 17 1.3 Host tree 18 1.3.1 Vigour 18 1.3.2 Constitutive defense 19 1.3.3 Induced wound response 20 2. Methods 23 2.1 Isolation and identification of fungi 23 2.2 Fungal inoculations in living trees 24 2.3 Chemical analyses 24 3. Results and discussion 26 3.1 Fungal flora 26 3.2 Induced responses and allocation of resources to defense 29 3.3 Beetle-fungi-host tree interaction 34 Conclusions 38 References 39 Original articles I-IV Heli Viiri 9 Introduction Phloem-feeding bark beetles are destructive pests in coniferous forests. They breed sub-cortically in trees, thus utilising a wide range of dead and dying trees to healthy trees. There are about 2.000 beetle (Coleoptera) species in Finnish forests; and about 800 of these species are saproxylic, depending upon dead trees (Siitonen 1998). Only a few bark beetles do attack healthy trees. The Eurasian spruce bark beetle, Ips typographic L. (Coleoptera, Scolytidae), is one of these and is a serious and widely distributed pest on Norway spruce, Picea abies (L.) Karsten. Adults and larvae live under the thick bark, usually in dying and weakened trees. While constructing breeding chambers and galleries in the phloem, beetles distribute spores of fungi to their hosts. As they grow, fungal hyphae suppress water transportation, cause discolouration of wood and help the beetles to kill trees. Each new generation of beetles transports the fungi to new host trees. All aspects of symbiotic relationship are not well understood. Among these features is that the fungi seems to make host carbohydrates available to the beetles as nutrition. Many species of fungi have been reported to occur in association with Scolytidae. Most of them belong to the genus Ceratocystis sensit stricto ( s.s.) Ellis & Halstedt, Ophiostoma H. & P. Sydow, Ceratocystiopsis Upadhyay & Kendrick and anamorph genera such as Pesotum Crane & Schoknecht and Leptographium Lagerberg & Melin (Rumbold 1931, Mathiesen- Käärik 1960, Upadhyay 1981, Whitney 1982, Solheim 1986, Schowalter and Filip 1993, Okada et al. 1998, 2000). This economically important but taxonomically controversial group has been referred to as the "ophiostomatoid fungi" (Wingfield et al. 1993). These fungi are adapted for dispersal by insects: elongated ascocarps bear ascospores at the apices of their necks, which may be protected by a gelatinous matrix. The bark beetles transport spores both laterally and in the digestive tract. A main aspect of the symbiosis between bark beetles and their associated fungi is their joint action to overcome the resistance mechanisms of their host trees. The symbiosis between bark beetles and their associated fungi has been reviewed extensively (Mathiesen-Käärik 1960, Francke-Grosmann 1963, Whitney 1982, Beaver 1989, Paine et al. 1997). In several investigations Ophiostoma bicolor Davidson & Wells, Ophiostoma penicillatum 10 Introduction (Grosm.) Siemaszko, Ophiostoma piceae (Munch) H. & P. Sydow, Ophiostoma piceaperdum (Rumbold) Arx and Ceratocystis polonica Siemaszko have been found to be associates of Eurasian spruce bark beetles (Siemaszko 1939, Solheim 1986, Harding 1989, Krokene and Solheim 1996). C. polonica is the most pathogenic of these fungi and when mass inoculated into Norway spruce, is able to kill even healthy trees (Horntvedt et al. 1983, Christiansen 1985b). Although O. piceaperdum can also invade the phloem and sapwood and disrupt the water-conducting system (Harding 1989), mostly C. polonica is thought to play a special role in the bark beetle -host tree interaction (Solheim 1993, Krokene and Solheim 1998). Exploitation of healthy trees as breeding material and a nutrition base causes inevitable difficulties to bark beetles. Conifers have several constitutive mechanisms for protecting themselves against pests and pathogens, e.g. thick bark, needle waxes and sticky primary resin. Conifers also have induced defensive mechanisms, e.g. the hypersensitive reaction against infection by pathogens (Berryman 1969). The pathogen is sealed off from living tissues of the host by a rapidly expanding, controlled necrosis (Berryman 1969, 1972). Several secondary metabolites are antifungal and antiherbivory, thus preventing the growth and reproduction of fungi and insects. After wounding, the content and quantity of the terpenoid compounds, resin acids, monoterpenes and sesquiterpenes in resin change and become more toxic. Similar changes occur in phenolic compounds like stilbene derivatives. In spite of the constitutive and induced responses that work together to protect the tree, in some circumstances fungi and beetles can escape from the entrapping lesion (Berryman 1982). Variation in the qualitative and quantitative production of secondary metabolites is an essential factor in modifying plant-herbivore interactions. Heli Viiri 11 Aims of the study The main aim of this work was to study interactions between the host tree, the pathogen and its bark beetle vector as well as to identify factors that influence the resistance of Norway spruce to infection by C. polonica. The first section is methodological and considers whether the spruce bark beetles collected (i) individually, (ii) with pheromone traps using an insulating material in the collecting bottles and (iii) in pheromone traps without insulating material differ in frequency and abundance of associated fungi. The second section includes identification of the ophiostomatoid species associated with I. typographus at different populations. The hypothesis was that there are no differences in frequencies of fungi collected at different geographic locations in low and higher population level of beetles. The third section covers interactions between the host and the pathogen. Changes in the main secondary compounds of Norway spruce, terpenoids and stilbenes, were studied in relation to carbon allocation and to the nutrient status of trees after artificial inoculation with C. polonica. 12 1 Spruce bark beetle-fungi host tree interaction - a review 1.1 Spruce bark beetle 1.1.1 Damage Large-scale outbreaks of spruce bark beetles can cause severe damage throughout spruce forests over areas covering tens of thousands of square kilometres. In central Europe severe damage has been recorded since the 1700's. The damage was most serious immediately after the Second World War, when throughout Central Europe about 30 million m 3 of spruce died. During recent years the spruce bark beetle has killed millions of cubic metres of spruce annually e.g. in Germany, France, Austria and Japan (Boutte 1993, Klimetzek and Yue 1997). In Japan, the related species Ips typographies japonicus (Niijima) is a destructive pest on Picea jezoensis (Siebhold and Zucc.) (Yamaoka et al. 1997). In Fennoscandia in 1971-1982 there was an extensive epidemic, which followed windfall and dry summer periods (Austarä etal. 1983, Bakke 1983, Worrell 1983, Risberg 1985). In Norway, the damage reached its peak in 1978-1980, when about one million cubic metres of spruce were killed annually (Christiansen and Bakke 1988). In Sweden 1.7-2.8 million m ? of spruce died during this epidemic. In Denmark, the damage was slight up until the 1960's but drought and windfalls favoured bark beetles' reproduction in the 1970-1980's (Ravn 1985). Also in Finland, the spruce bark beetle is the most damaging scolytid on Norway spruce. However, during the recent period of intensive forest management, bark beetle populations have not risen to high levels. In Finland, compared with the other Nordic countries, both damage and population levels of the spruce bark beetle have been modest (Saalas 1919, 1949, Löyttyniemi and Uusvaara 1977, Löyttyniemi etal. 1979, Weslien etal. 1989, Valkama etal. 1997). Heli Viiri 13 1.1.2 Ecology As breeding material, spruce bark beetles usually favour weak and dying trees, windfalls and snow breaks. When more suitable breeding material is available, the population size starts to increase rapidly. If the population level is high, the spruce bark beetle can successfully attack even healthy trees (Bakke 1983, Mulock and Christiansen 1986). Damage is considerable because these beetles favour large spruce trees; and once started, an outbreak can continue for several years. In Fennoscandia the beetles disperse in May-June, when the beetles emerge from their hibernation sites in the forest litter. Small numbers of beetles also overwinter under the bark of felled trees or in the lower trunk of standing trees. Timing of the flight period has been investigated in several studies (Annila 1969, 1977, Bakke et al. 1977 a). The most significant regulators of flight have been found to be an air temperature and thermal sum. Most individuals become sexually mature and swarming starts when the air temperature reaches +2O °C. In southern and central Finland the dispersal period normally occurs between the end of May and early June, and in northern Finland in the middle of June (Annila 1969, 1977). New generations emerge from trees into the forest in late July or early August. In Fennoscandia and at high altitudes, only one generation is produced per year, while in more favourable areas in southern Europe normally two, or even three, generations occur each year. Spruce bark beetles have a chemical system of communication that helps them to select suitable host trees and allows the beetles to colonize them effectively (Bakke 1983). The beetles' pheromone system co-ordinates the intensity of the attack, causing rapid aggregation of large numbers of beetles (Bakke et al. 1977 a, Birgersson et al. 1984, Schlyter and Birgersson 1989). In Ips species, the male beetles initiate construction of an entrance tunnel and gallery by producing mainly the components of aggregation pheromone (Bakke et al. 1977b). In I. typographus , the aggregation pheromone is composed of two primary components; methylbutenol is a short range attractant that promotes landing and entering of holes, while the heavier and less volatile c/s-verbenol also promotes landing but acts over a longer distance (Bakke et al. 1 977 b, Schlyter et al. 1987). Conspecifics of both sexes are attracted to the site; males initiate their own nuptial chambers, and one to four females arrive to the completed chambers to mate with the resident male (Christiansen 1988). After mating, the female starts to excavate 14 Spruce bark beetle-fungi-host tree interaction - a review longitudinal egg galleries in the phloem and the male helps remove from the gallery the dust produced by boring. The pheromone production of bark beetles is linked to the secondary metabolism of the host tree. The resin monoterpenes can have contrasting effects as bark beetle repellents and as pheromone precursors. C/s-verbenol is synthesized by hydroxylation of the exogenous monoterpene of the tree, (-)-a-pinene. The synthesis of this compound is regulated by the availability of the precursor (-)-a-pinene (Ivarsson 1995). Since spruce produces (-)-cx-pinene to defend itself, the beetles can produce m-verbenol and attract more individuals to the tree (Birgersson 1989, Ivarsson 1995). The spruce bark beetle can synthesize methylbutenol de novo via mevalonate, independently of precursors from the host tree (Lanne et al. 1989). In addition, yeasts associated with spruce bark beetles can convert cis verbenol to verbenone (Leufven et al. 1984). Furthermore, verbenone and ipsenol act as antiaggregation pheromones regulating breeding density in trees (Bakke 1981, Birgersson et al. 1984). During gallery construction, many bark beetles transport various micro-organisms to the phloem and cambium: yeasts, bacteria and fungi that help the beetles establish galleries and begin to oviposit as well as aiding digestion (Whitney 1982, Leufven and Nehls 1986, Furniss et al. 1990). In some bark beetles slimy secretions may be produced to preserve ascospores and conidia from desiccation and UV-light (Dowding 1969). Many species of bark beetles have special organs, known as mycangia, in which spores are transported. In general, the mycangia have a similar basic structure: tubes, pouches, cavities or pits associated with glandular cells (Beaver 1989). The spruce bark beetle, however, does not have mycangia and it transports fungal spores mainly laterally on the posterior half of the pronota and in pits scattered over most of the surface of the elytra. Small numbers of spores are also transported in the digestive tract and on phoretic mites associated with the beetles (Moser et al. 1989, Furniss et al. 1990). Either I. typographus has slimy excretions in the pits, where the spores occur or the spores may be covered with wax (pers. comm. E. Christiansen and H. Solheim). Thousands of ascospores and conidia may be found on the surface of each beetle, but the number of spores transported by different individuals varies greatly. Neither ultrasonic cleaning nor chemical sterilizing agents have successfully eliminated fungal spores from adult beetles (Harding 1989, Furniss et al. 1990). Heli Viiri 15 1.2 Associated fungi 1.2.1 Ecology Ophiostoma and Ceratocystis fungi are well adapted for dispersal by insects: elongated ascocarps bear ascospores on the apices of their necks; these ascospores may be released in a slimy matrix. The fruiting structures of the fungi are formed in beetle galleries and under bark flaps on the phloem. Since spores lose their vitality and dry in sunshine, rapid insect dispersal guarantees reliable inoculation of the fungi to a suitable habitat (Dowding 1969). In addition, the conidia of ophiostomatoid fungi may accumulate in sticky drops at the apex of conidiophores. The cirri of some ophiostomatoid species disperse in conifer resin rather than only in water (Whitney and Blauel 1972). According to Wingfield et al. (1993), species with short ascomatal necks tend to have long ascospores and vice versa. In spite of that, variation in the length of necks or the presence of ostiolar hyphae does not correspond to the probability of successful dispersal at the species level. Many ophiostomatoid fungi associated with bark beetles are highly pathogenic and also cause sap-stain, which is a grey, black or bluish discoloration of sapwood caused by the optical effect of pigmented fungal hyphae (Wingfield et al. 1993). It is quite often referred to as blue-stain; thus the word "blue" for describing the discoloration can in many cases be misleading. There is a continuum from truly pathogenic sap-staining fungi that occur in living trees to pathogenic fungi that grow on weakened trees to truly saprobic fungi that utilise dead trees (Wingfield et al. 1993). Typically, discoloration is spread over a wide area of the stem, (more rapidly longitudinally than radially) causing deep staining throughout the sapwood. Hyphae grow mainly in tracheids and ray parenchyma cells, preventing water transport, and can kill living tissues far from the beetle galleries (Wong and Berryman 1977, Horntvedt et al. 1983). No actual staining of the cell walls occurs (Wingfield et al. 1993). Despite the fact that tangential growth and radial growth in the xylem ray system are minor, fungi can reduce the commercial value of the lumber considerably. However, the main harmful effects of pathogenic ophiostomatoid fungi are visible in the forest when trees are dying. Common associates of the spruce bark beetle during a beetle epidemic and immediately following are C. polonica, O. bicolor, O. penicillatum and Pesotum species (Table 1). These species are frequent in the area of visible staining (Solheim 1986, 1992a,b). When the population level of beetles has been low, the same species have occurred; but the frequency of C. polonica has been 16 Spruce bark beetle-fungi-host tree interaction - a review lower than during epidemics (Harding 1989, Solheim 1993). It has been proposed that C. polonica may be replaced by other species during periods when the population level is low and the beetles infest dead trees and timber (Solheim 1993). During epidemics, however, living trees are attacked and the frequency of C. polonica may increase (Solheim 1993, Wingfield et al. 1993). C. polonica is the most aggressive species, being able in inoculation experiments to invade sapwood and kill healthy trees (Horntvedt et al. 1983, Christiansen 1985b, Solheim 1988, Kirisits 1998, Krokene and Solheim 1998). It tolerates low oxygen levels and grows extensively in the phloem and sapwood (Solheim 1991), while other ophiostomatoid fungi cannot grow well in such low levels of oxygen. Table 1. Ophiostomatoid species associated with the spruce bark beetle I. typographus with their anamorphs. *Note: Dubious validity of species according to Jacobs and Wingfield (2001) References: I=Davidson et al. 1967, 2=Grosmann 1931, 3=Harding 1989, 4=Kirisits 1996, s=Krokene and Solheim 1996, 6=Mathiesen 1951, 7-B=Mathiesen-Käärik 1953,1960, 9=Savonmäki 1990, 10=Siemaszko 1939, 11-14=Solheim 1986, 1992a,b, 1993 and 15=Yamaoka et al. 1997. Species Anamorph Reference Ceratocystiopsis minuta (Siemaszko) Upadhyay & Kendrick Hyalorhinocladiella 3,4,6,9,11-15 Ceratocystis polonica Siemaszko Thielaviopsis 1,3-7,11-15 Ophiostoma ainoae Solheim Pesotum 3,11-15 0. bicolor Davidson & Wells Hyalorhinocladiella 1,3-5,9,11-15 0. cainii (Olchowecki & Reid) Harrington Pesotum 3 0. cucullatum Solheim Pesotum 3,4,11,15 0. japonicum Yamaoka & M.J. Wingfield Pesotum 15 0. flexuosum Solheim Sporothrix 3,11 0. penicillatum (Grosm.) Siemaszko Leptographium 1-9,11-15 0. piceae (Munch) H. & P. Sydow Pesotum 2-5,7-9,11-15 0. piceaperdum (Rumbold) Arx = 0. europhioides (Wright & Cain) Solheim Leptographium 1,3,4,11-15 0. piliferum (Fries) H. & R Sydow Sporothrix 9 0. pluriannulata (Hedgcock) H. & P. Sydow Sporothrix 7 0. stenoceras (Robak) Melin & Nannfelt = O. albidum Mathiesen-Käärik Sporothrix 7 0. tetropii * Mathiesen Sporothrix 6,9,11,13 Heli Viiri 17 1.2.2 Taxonomy Most of the fungi associated with bark beetles that cause discoloration belong to the ascomycetes or the fungi imperfecti. Important and economically significant associates belong to the family Ophiostomataceae and in particular to the genera Ceratocystis and Ophiostoma (Table 1). Ceratocystis , Ceratocystiopsis and Ophiostoma are differentiated according to properties of ascospore morphology, development of the ascomatal centrum, carbohydrate content of the cell walls, conidial stages and conidia. Order, family and several species have been a source of taxonomic controversy. The families Ceratocystis s.s., Ceratocystiopsis and Ophiostoma have generally been accepted. Most species in the genus Ceratocystiopsis share common characteristics with species in Ophiostoma, and it has been proposed that the species form a monophyletic group (Wingfield et al. 1993, Viljoen et al. 2000). Many bark beetle associates formerly considered to be in the genus Ceratocystis are now placed in the genus Ophiostoma (Wingfield et al. 1993, Hawksworth et al. 1995, Viljoen et al. 2000). The main species in this thesis, C. polonica, was originally described as O. polonicum and has lately been transferred to Ceratocystis (Visser et al. 1995, Harrington et al. 1996). Furthermore, based on the morphology of the perithecia, Jacobs et al. (2000) concluded that O. europhioides and O. piceaperdum are indistinguishable thus supporting the synonymy O. piceaperdum proposed by Upadhyay (1981). In the interest of consistency, throughout this review O. europhioides is called O. piceaperdum. The members of the genus Ceratocystis s.s. contain no cellulose or rhamnose in the cell walls, and traditionally the asexual stage has been Chalara Corda (Weijman and De Hoog 1975, Upadhyay 1981, De Hoog and Scheffer 1984). Recently Paulin-Mahady et al. (2002) transferred anamorphic Chalara species to the genus Thielaviopsis Went. Within these Ceratocystis species, antibiotic cycloheximide in the growth media prevents growth, which has been used as a taxonomic characteristic of the genus (Harrington 1981). Other characteristics are darkly pigmented (nearly black) perithecia and hyaline ascospores that vary in shape but are not falcate (De Hoog and Scheffer 1984). Ophiostoma and Ceratocystiopsis species have cellulose and rhamnose in the cell walls, which may be the reason for their tolerance to cycloheximide (Harrington 1981). The perithecia are almost glutinous, and the ascospores vary in shape. Ceratocystiopsis species are otherwise similar to Ophiostoma, but the ascospores are elongated or falcate and are 18 Spruce bark beetle-fungi-host tree interaction - a review surrounded by a hyaline, gelatinous ascospore wall (previously defined as a sheath) (Upadhyay and Kendrick 1975, Upadhyay 1981, De Hoog and Scheffer 1984). Ophiostomatoid fungi can have several asexual forms or, in some species, only the asexual form is known. The family Ophiostomataceae has contained as many as 16 asexual forms (Wright and Cain 1961, Upadhyay and Kendrick 1975, Upadhyay 1981, Wingfield et al. 1993). Asexual forms are classified according to the morphology of the conidiophores or by the form of the conidia. The conidiophores can be mononematous like Leptographium or form synnemata like Pesotum. Synnematous anamorphs of Ophiostoma species were quite recently placed in the genus Graphium , although Graphium species are considered to be anamorphs of the Microascales (Okada et al. 1998). However, the synnematous anamorphs of Ophiostoma species are phylogenetically unrelated to Graphium s.s., and should be currently referred to the anamorph genus Pesotum (Okada et al. 2000). The genus Verticicladiella is considered to be a synonym for Leptographium (Wingfield 1985, Schowalter and Filip 1993). Most of the asexual forms of the Ophiostoma species can be classified to the genera Pesotum (Okada et al. 1998, 2000), Leptographium, Hyalorhinocladiella or Sporothrix (Harrington 1987, Hausner et al. 2000). The taxonomy of the family Ophiostomataceae and its genera still needs clarification. In recent years, several new fungi that cause wilting and discolouration on trees have been described. Unknown species complexes probably exist even in fungi that have worldwide distribution and a wide host range, for example, C. fimbriata, for which related species, C. albofundus Wingfield, De Beer and Morris have lately been found (Wingfield et al. 1996). Morphological similarities in teleomorph and anamorph structures of phylogenetically distinct Ceratocystis and Ophiostoma have apparently developed as an adaptation to an insect-associated habitat (Visser et al. 1995, Hausner et al. 2000). 1.3 Host tree 1.3.1 Vigour Measurements of tree vigour have been used to determine the risk of attack by bark beetles and as a substitute for the term "resistance" (Waring and Pitman 1983, Mulock and Christiansen 1986). The tree vigour index is defined as the ratio of the cross-sectional area of the current annual ring (BAj) to the Heli Viiri 19 sapwood basal area (SA) at breast height (Waring et al. 1980, Munster-Swendsen 1987). The basic assumption is that good growth generates high vigour. The index is based on the pipe model theory and on the assumption that the relationship between the cross-sectional sapwood area and the weight or area of leaves supported by the conducting sapwood within a tree species is linear. The number of beetle attacks, their distribution on the trunk and their timing determine the effectiveness of the bark beetle - fungi association complex in overcoming the resistance of a tree (Berryman 1972, 1982, Raffa and Berryman 1983, Christiansen 1985a,b). The spruce bark beetle can colonise standing healthy Norway spruce trees if the number of attacking beetles is large enough to overcome the resistance of the trees. In experiments, attacks of 150-200 beetles or artificial inoculations of C. polonica have killed the trees (Christiansen and Horntvedt 1983, Christiansen 1985b). During an epidemic in nature, hundreds or even thousands of spruce bark beetles can attack a single mature tree. Bark beetles both co-operate in overwhelming the host defense due to pheromone mediated mass-attack and simultaneously compete for the available resources. Stand and climatic conditions often reduce plant assimilation and may consequently lower a tree's ability to resist the attack of bark beetles and associated fungi. Changes in herbivore abundance have often been correlated positively with unfavourable environmental factors. On the other hand, the ability of a tree to defend itself is linked to its overall vigour and to the amount of carbohydrates that can be used as a source of energy for synthesising defensive compounds (see reviews: Berryman 1972, Christiansen et al. 1987, Paine et al. 1997). This is especially important in mature conifers, which have reduced ability to replace damaged structures. 1.3.2 Constitutive defense Conifers protect themselves against attack by bark beetles and associated fungi with constitutive defense, based mainly on pre formed primary resin, and by production of secondary resin in the induced wound response. Constitutive defense is found especially in trees with well-developed pre-formed resin duct systems, such as Pinus. Primary resinosis is a continuous defense system against generalists. It is particularly important during the early phases of the beetle-fungus colonization as an agent that cleanses the wound (Berryman 1972) and enables activation of the induced defense. 20 Spruce bark beetle-fungi-host tree interaction - a review Primary resin is under turgor pressure in the vertical and horizontal resin ducts and blisters and depends on the water potential of the tree. The resin starts to exude when the tree is wounded, e.g. by the boring activity of beetles. The flow and the crystallization of the primary resin are sometimes sufficient to deter beetles soon after they initiate construction of galleries (Reid et al. 1967, Hodges et al. 1979). In addition, pre-formed primary resin inhibits the growth of ophiostomatoid fungi (Cobb et al. 1968). During a mass attack by bark beetles the resin system is often exhausted due to simultaneous exudation of resin in separate wounds that lead to successfully constructed galleries. In spruce, constitutive resin is carried primarily in vertical resin ducts in the bark, and the amount of resin content depends mainly on the storage capacity of the duct system (Christiansen and Bakke 1988). Thus, in Norway spruce exudation of primary resin varies considerably between individual trees (Christiansen and Horntvedt 1983). In addition to resin, lignified stone cell masses (lignin) provide an important pre-formed system of defense in living trees by preventing both construction of bark beetle galleries and oviposition (Wainhouse et al. 1998). 1.3.3 Induced wound response The induced wound response plays an essential protective role in conifers that lack a well-developed pre-formed system of resin ducts, e.g. species in the genus Picea. The induced wound response has been thought to protect trees against host-adapted pests and pathogens. Fungal cell-wall components and bark beetle feeding elicit a metabolically active induced wound response in the phloem and sapwood (Miller et al. 1986, Lieutier and Berryman 1988). Living cells in the reaction zone die and form a necrotic area, thus preventing fungal nutrition. Berryman (1969) described this controlled necrosis as the hypersensitivity reaction which is an active metabolic process affected by the physiological vigour of the tree. Soluble nutrients and carbohydrates are mobilized from tree storage reserves to produce toxins in tissues near the point of infection; the amount of soluble carbohydrates decreases and the content of secondary metabolites increases and becomes more toxic (Reid et al. 1967, Wright et al. 1979, Cook and Hain 1985, 1987, Delorme and Lieutier 1990, Raffa and Smalley 1995). Thus the reaction zone fills with resinous and C based compounds. Monoterpenes, sesquiterpenes, resin acids and phenolic compounds repel beetles and inhibit the growth of fungi or bark beetle larvae (Shrimpton and Whitney 1968, De Groot Heli Viiri 21 1972, Bordasch and Berryman 1977, Raffa et al. 1985, Bridges 1987, Woodward and Pearce 1988, Delorme and Lieutier 1990, Solheim 1991, Klepzig et al. 1996, Lindgren et al. 1996, Evensen et ai. 2000). Secondary resin, as well other secondary metabolites, is produced by ray and phloem parenchyma cells. In addition, callus and wound periderm develop rapidly in tissues surrounding the wounded area (Reid et al. 1967, Berryman 1969, Wong and Berryman 1977). A reaction zone is formed to separate wounded tissues from healthy wood. Fungal inoculation tests and resin toxicity tests with bark beetles have shown that the rate of accumulation and the concentration of secondary metabolites in the reaction zone are crucial for successful protection. The components of primary and secondary resin vary qualitatively and quantitatively between individual trees, species, age, season and type of resin (Russell and Berryman 1976, Raffa and Berryman 1982, 1983, Toscano Underwood and Pearce 1991, Lindberg et al. 1992). In general, non-host resins are more toxic to insects and fungi than resins from the host species; and the induced resins are more toxic than the pre-formed resins. Formation of the reaction zone and terpene synthesis represents two independent activities during the wound response, necrosis proceeding more rapidly than terpene synthesis (Raffa and Berryman 1982). Thus, the fungus is first contained by the removal of essential nutrients from the entry site, and only secondarily by resinosis (Wong and Berryman 1977, Raffa and Berryman 1982, Christiansen and Ericsson 1986). In woody plants the induced wound response is considered to be a non-specific response to wounding. However, it has been detected that concentrations of monoterpenes increase as the virulence of the fungal species increases (Cook and Hain 1985, Poppet al. 1995). On the other hand, it is a competitive advantage for many pathogenic fungi to tolerate stilbenes better than non pathogenic species do (Hart 1981). When inoculated into a tree, a pathogenic fungus causes a typical reaction zone in the phloem, the size of which depends on variation in fungal growth, virulence or elicitor production. The phloem of Norway spruce contains so-called polyphenolic parenchyma cells (PP cells) and traumatic resin ducts. These cells and ducts react to fungal inoculation and wounding, which indicate involvement of both constitutive and inducible defense responses (Franceschi et al. 1998, 2000). Pre formed and induced wound responses are separate, but overlapping, dynamic phases of wound reaction; the pre-formed reaction consists mainly of processes ending up to visual symptoms, and the induced wound response consists more of 22 Spruce bark beetle-fungi-host tree interaction - a review biochemical changes at the cell and tissue level. Induced defense, in particular, is energy-demanding response involving de novo synthesis of secondary compounds and new tissue. Despite the fact that the constitutive and induced wound responses work together to protect the tree, in favourable circumstances fungi can escape from the entrapping reaction zone. Heli Viiri 23 2 Methods (For details, please refer to the original articles I-IV) 2.1 Isolation and identification of fungi Living beetles were inoculated into fresh logs in order to isolate the associated ophiostomatoid fungi (I, II). The inoculation method was originally described by Wright (1933) and has been used widely in corresponding studies. Beetles were placed with forceps into the hole made by a cork-borer, and the bark plug was then reinserted. Empty holes made by a cork-borer but without beetles were used as control inoculations. The aim of the control inoculations was to detect fungi that had become established in the phloem or bark before inoculation or invaded trees during the inoculation phase as contamination. To kill air-dispersed spores, the surface of the logs was sterilized with ethanol before inoculations. After 3 and 4 weeks incubation, samples were taken from bark, phloem or sapwood. Malt-agar media suitable for most ascomycetes were used in culturing. To promote fungal sporulation, wooden chips were occasionally added to the growth media. The growth media were not optimised for the growth demands of any special fungi, because the aim was to isolate all possible ophiostomatoid fungi. Identification concentrated on ophiostomatoid species, and thus other species that did not produce teleomorph were mostly ignored. The only anamorph of Ophiostoma and Ceratocystis species identified to species level was Leptographium penicillatum Grosmann. There are several discrepancies among the anamorph descriptions, so in many cases identification would have been uncertain without a teleomorph (Wingfield et al. 1993). Following cultures from the collections of the Norwegian Forest Research Institute were used as reference material in identification: O. penicillatum (80-91/54), C. polonica (80-53/7), O. piceae (80-92/34), 0. europhioides (80-91/9), (). ainoae (80-85/37), O. tetropii (80-113/9) and Graphium sp. (80-52/24). 24 Methods 2.2 Fungal inoculations in living trees Low-density inoculations were made in a Myrtillus-lype mature spruce forest in Vesijako, southern Finland (111, IV). Four plots were marked, and in May 1993 fertilization treatments (N, P and NPK) were applied to the plots to manipulate the defensive potential of the trees. The control treatment was a non-fertilized plot. From each plot, 30 trees that were free of visible wounds, a total of 120 trees, were selected. In June 1994 a Finnish culture (origin Tuusula) of C. polonica , on malt-agar in petri dishes was inoculated with a cork-borer into ten trees per fertilization treatment. Each wounded tree received four inoculations, one at each of the cardinal points of the compass, at 1.3 m above ground level. For mechanical wounding, the trees were injured with a cork borer in the same way as inoculation, but without the fungus. Ten trees per fertilization treatment were wounded mechanically. After a two-month incubation period around each fungal inoculation site, phloem samples were taken with the borer to the level of the cambium. One sample was collected from the distal ends of the visible reaction lesion (later called the "far" samples) and one each from the areas immediately above and below the site of inoculation (later called the "near" samples). Two samples were taken from near the site of mechanical wounding, one above the inoculation site and another below the inoculation site. Unwounded phloem from non-bored trees was used as the unwounded control. 2.3 Chemical analyses Chemical analyses were conducted in the Central Laboratory of the Finnish Forest Research Institute, Vantaa Research Centre. Phloem samples were ground in liquid nitrogen and analysed for their terpene, stilbene and carbohydrate composition. To obtain sufficient amounts of compounds for gas chromatography (GC) analysis, organic solvents were used for extraction. Stilbenes, carbohydrates, mono- and sesquiterpenes were analysed in a GC mass spectrometry (MS) system with a capillary column. The degradation of stilbene compounds was reduced by silylating the samples. Stilbene glycosides were quantified according to the response of rhapontin, and stilbene aglycones were quantified by using the response factors of diethyl stilbestrol and resveratrol. Monoterpenes were identified using the retention and mass spectral data of authentic model compounds. Sesquiterpenes were identified according to the method of Pohjola (1993) and Heli Viiri 25 quantified according to the response factor for caryophyllene. Due to the complexity of the compounds analysed, both retention and MS data were used for identification. In identification of compounds previously published data were also used (Mannila 1993, Pohjola 1993). The enantiomers of chiral monoterpene hydrocarbons were not separated. For quantification of compounds, calibration with both internal and external standards was used. When possible, commercial substances were used as reference compounds. The needles of the same trees were used for analysing nutrient concentrations with an inductively coupled plasma emission spectrometer (ICP-AES). The total C and N in needles were determined by dry combustion with a CHN analyser. 26 3 Results and discussion 3.1 Fungal flora Species Ceratocystis polonica, O. ainoae, O. bicolor, O. piceaperdum, O. penicillatum, O. piceae and O. tetropii and asexual forms of Leptographium spp. and Graphium spp. were found to be associated with spruce bark beetles in Finland (Paper I). These species correspond to the fungal flora previously found to be associated with galleries of spruce bark beetle in southern Finland (Savonmäki 1990). In addition, C. polonica , O. ainoae and O. piceaperdum were found as new species. Furthermore, in France were found C. minuta and O. cucullatum Solheim. The frequency of C. polonica was lower in Finland and France than in Norway (Krokene and Solheim 1996). Whether the frequency of pathogenic C. polonica as an associate of spruce bark beetle would be more common during an epidemic, could not be clearly answered here. Nevertheless, low frequency of associated pathogenic species is not contradictory to low level of damage in Finland. In French study areas, the spruce bark beetle population were in the post-epidemic phase. Also elsewhere, the high frequency of C. polonica has been sporadic and irregular (Harding 1989, Kirisits 1996, Yamaoka et al. 1997, Kirisits et al. 2000). In French isolations C. minuta was frequent, but it was not detected in Finnish isolations. Lack of C. minuta in the Finnish samples may be partly due to difficulties identifying the mixed unsporulated strains (I). On the other hand, C. minuta perithecia were abundant and easy to identify from the French primary isolations containing even several species (II). Also elsewhere the association of C. minuta with I. typographus has been inconstant (Solheim 1986, Kirisits 1996, Kirisits et al. 2000). Previously no specific importance has been ascribed to O. piceaperdum, but my results suggest that the role of this species may vary (I, II). O. piceaperdum is frequently found in Denmark and Austria, invading the sapwood of Norway spruce and disrupting the water-conducting system (Harding 1989, Kirisits 1996). It is not able to grow as rapidly as C. polonica (Harding 1989), but it causes long broad lesions in the phloem (Kirisits 1996, 1998). Results concerning the ability of O. piceaperdum to cause sapwood discoloration are conflicting; in Denmark the Heli Viiri 27 species has caused as extensive desiccation as C. polonica , but in Austria the desiccation observed after artificial inoculation was slight (Kirisits 1998). O. piceaperdum has been found in several conifers in Europe (Davidson et al. 1967) and in North America (Wright and Cain 1961, Davidson and Robinson-Jeffrey 1965). Recently the species was found to be a constant associate of the Douglas-fir beetle, Dendroctonus pseudotsugae Hopkins (Solheim and Krokene 1998). Because of the revised status of O. europhioides and O. piceaperdum, the pathogenicity and distribution of the O. piceaperdum complex needs further investigation in different host trees throughout its broad geographical range. O. penicillatum is frequently associated with I. typographus (I, 11, Grosmann 1931, Siemaszko 1939, Mathiesen-Käärik 1953, 1960, Harding 1989) and has been isolated from the phloem near visible stains (Solheim 1986, 1992 a). O. penicillatum does not grow deep into the sapwood, and thus it has been speculated that this species is not able to kill a tree without the beetle (Horntvedt et al. 1983, Solheim 1988, Harding 1989). The large internal variation in morphological characters supports the expectation that the pathogenicity of this species is variable (Mathiesen- Käärik 1953, 1960). O. bicolor occurs with C. polonica in the beetle galleries, especially in the early phase of attack (Solheim 1986, Harding 1989). It is a fast growing species (Solheim 1986,1991); but not, according to the inoculation experiments, as aggressive as C. polonica (Solheim 1988, Kirisits 1998). Most likely O. bicolor can overcome the resistance of a tree with other fungi, but alone it is not able to kill a healthy tree. O. bicolor has been isolated from //«-infested spruce both in Europe and in western parts of North America. Among the ophiostomatoid species, O. bicolor , O. piceaperdum and C. minuta were easy to identify. In the Finnish isolates of O. piceaperdum , perithecia formation was very abundant and stable. There were no signs of degeneration that might have influenced its frequency of occurrence. French isolates produced noticeably less perithecia, thus suggesting intra specific variation. However, the ability of O. piceaperdum to produce perithecia seemed to be more stable than that of e.g. O. bicolor (I, II). C. polonica, O. ainoae and O. bicolor degenerated rapidly after two transfers, producing only sterile mycelia. Degeneration of axenic cultures and loss of pathogenicity has also been noticed by Kirisits (1996) and Krokene and Solheim (2001). In old cultures, after several months incubation and storage at +4 °C, O. ainoae produced few perithecia (I). O. piceae colonies 28 Results and discussion can also be hard to maintain in normal condition on artificial media, and some colonies do not form perithecia (Davidson 1953). Here on malt-agar some O. piceae colonies formed only light brown sectors or grey dots with few coremia and no perithecia. It has lately been reported that the O. piceae complex forms a monophyletic group of nine recognized insect-dispersed species, delimited by synnemata morphology, growth rate, mating reactions and sequences of the internal transcribed spacer (ITS) region of the rDNA operon (Harrington et al. 2001). The large numbers of Leptographium sp. can be partly ascribed to the ability of O. penicillatum to degenerate on artificial growth media when retained for long periods (I, II). After several transfers the species produces few perithecia and only occasionally (Davidson et al. 1967). According to Kendrick's (1962) extensive review of Leptographium species, the Leptographium sp. found in study (I) appears to be very similar to L. penicillatum. Leptographium spp. may contain asexual stages of both O. penicillatum and O. piceaperdum, so asexual stages were not used as the main characteristic in identification. All synnematous anamorphs were included in the group of Graphium spp. in Paper I and Pesotum spp. in Paper 11, which group probably includes several different species since there was great variation in width, length and colour of the synnemata. Some species produce few or no perithecia on artificial growth media, and some reproduce more abundantly in wood (Furniss et al. 1990). Most species of Ceratocystis and Ophiostoma vary widely in colony growth, formation of asexual and sexual structures and sporulation, which makes it difficult to produce sexual stages on artificial media. Ophiostomatoid fungi need adequate nutrients and high C/N ratios to produce perithecia on artificial growth media (Mathiesen-Käärik 1960). Progressive sub-culturing favours vegetative growth at the expense of reproductive structures. Moreover, after two or three transfers (I, II) some species produce only sterile mycelia. When fungi have been isolated from attacked trees, inoculated logs or directly from beetles, significant differences have been detected in the fungal flora (Yamaoka et al. 1996, Solheim and Krokene 1998). The most reliable method for identifying ophiostomatoid fungi is to isolate them from recently hatched insects. However, after hibernating in the ground, dispersing beetles are covered with soil microbes. Here inoculations of insects (I, II) into wooden logs prevented mostly other fungi and microbes from affecting isolation of associated fungi. On the other hand, the isolation technique used here favour pathogenic fungi. Fungi isolated from control inoculations had Heli Viiri 29 become established in the phloem or bark before the inoculation or invaded trees during the inoculation phase as contaminants. Some of the same species, such as Nectria sp„ were isolated from the true inoculation. In the control inoculation there might have been fungi belonging to the genus Ophiostoma, e.g. O. piceae or O. piliferum, which can also disperse by air. But surface sterilization of logs (I, II) was enough to kill air-dispersed spores. Furthermore, there was no evidence of cross-contamination from other beetles due either to the presence or to the abundance of fungal species when fungi were isolated from beetles collected with different methods (I). Morphological plasticity, conidia production for partitioning conidial stages, and degeneration of conidiophores on artificial growth media during storage and subculturing have led to large numbers of conidial stages in ophiostomatoid fungi. Ultrastructural studies and ribosomal DNA sequencing have been used to explore species characteristics and taxonomic relationships within the genus and between genera (Van Wyk and Wingfield 1990, 1993, Jacobs et al. 1996, Hausner et al. 2000, Okada et al. 2000, Harrington et al. 2001, Paulin-Mahady et al. 2002). Genera that are indistinguishable with light microscopy may differ in ultrastructural morphology, and phylogenetic and some ascosporic characteristics observed in light microscopy studies can even be misleading. These aspects must be taken into account when these results are considered (I, II). To clarify the discrepancy between the results of different investigations, the pathogenicity of different strains of C. polonica, O. penicillatum and O. piceaperdum needs to be tested. 3.2 Induced responses and allocation of resources to defense Inoculation of C. polonica caused extensive lesions around the inoculation site varying in length from 0.5 to 38 cm. Changes in CBSCs were more pronounced with fungal inoculation than with fertilization treatments (III). The further lesion formation and induced response had progressed (wounding, fungus inoculation), more total soluble carbohydrates were utilized to prevent fungal invasion (Figures 1-3). It has been proposed that the induced phenolic response of Norway spruce phloem consists of activation of the phenolic pathway, finally leading to production of tannins and insoluble polymers (Brignolas et al. 1995, 1998). Present results agreed with these previous findings; detection frequencies of stilbene glycosides decreased in phloem inoculated with fungi 30 Results and discussion Figure 1. Relationship between total soluble carbohydrates and total stilbene glycosides (µg mg - 1 fresh phloem) in Norway spruce. For sampling and compounds see Papers III, IV. (Near, Wounded, n=38; Far, Unwounded, n=39). Figure 2. Relationship between total soluble carbohydrates and total stilbene aglycones (µg mg- 1 fresh phloem) in Norway spruce. For sampling and compounds see Papers 111, IV. (Near, Wounded, n=38; Far, Unwounded, n=39). Heli Viiri 31 but remained high or at initial concentrations in unwounded phloem (see Fig. 3 in Paper III). However, in this study the corresponding aglycones were also detected in small amounts near the point where the fungus had been inoculated. As suggested by the results of (III) and those of Woodward and Pearce (1988 a), the stilbene aglycones might play a more central role in fungus challenged tissues than the corresponding glycosides do. Thus the newly synthesized stilbenes might be incorporated into stilbene aglycones and partly into tannins (Figures 1-2). This could explain why in the studies of Brignolas etal. (1995,1998) stilbene glycosides accumulated at higher levels in the susceptible clone than in the resistant clone. However, the results of Evensen et al. (2000) contradict those of Brignolas etal. (1995,1998) and do not support stronger induction of the flavonoid pathway than of the stilbene pathway. Changes in glycosylation of the stilbenes may be located in the PP cells on the basis of their high phenolic and PAL (phenylalanine ammonia lyase) content and to dynamic changes in the cells after wounding (Franceschi et al. 1998, 2000, Nagy et al. 2000). PAL, a key enzyme in phenolic synthesis, is also present in ray parenchyma cells. The PP cells are the most abundant living cells in the secondary phloem and are thus the most probable site of PAL synthesis. In Norway spruce the terpenoids are stored mainly in the resin ducts, whereas phenolic compounds and PAL are stored in the vacuoles of PP cells (Franceschi et al. 1998, Krekling et al. 2000). Release of phenolics or metabolites from the vacuole phenolics and traumatic formation of resin ducts provides inducible and sustained release of defensive compounds away from the initial site of wounding or invasion (Franceschi et al. 1998, 2000, Krekling et al. 2000, Nagy et al. 2000). In this study, the accumulation of terpenes near the inoculation point was extensive; the total concentration of terpenes was almost 100 times higher than that in unwounded trees (III). Due to the large volume of accumulated terpenoids near the inoculation point (Figure 3), in addition to de novo energy-demanding biosynthesis at the site, it is possible that photosynthates were also translocated. The relationship between host resistance and carbohydrate status in the phloem may be more complex than a simple source and sink relationship. CBSCs cannot be synthesized without substrate, but the presence of substrate does not necessarily lead to high synthesis level of CBSCs. Variation in the responses of CBSCs to fertilization might also be caused by differences in their biosynthesis (Haukioja et al. 1998, Koricheva et al. 1998) or the storage, transport or maintenance of defenses (Gershenzon 1994). Sesquiterpenes and triterpenes are biosynthesised via the 32 Results and discussion Figure 3. Relationship between total soluble carbohydrates and total terpenes µg - 1 fresh phloem) in Norway spruce. For sampling and compounds see Papers III, IV. (Near, Wounded, n=38; Far, Unwounded, n=39). mevalonic acid pathway and monoterpenes, diterpenes and tetraterpenes are biosynthesised via the pyruvate-glyceraldehyde 3-phosphate pathway. Meanwhile, phenylpropanoids and defense proteins are synthesised via the same pathway and share a common precursor, PAL. Protein-pathway may compete for excess carbon out of the reach of phenolics, while the mevalonic acid and pyruvate-originated pathways guarantees more "a private" synthesis routes for terpenoids. A recurring theme in defense allocation theories is the assumption that there is a trade-off between growth and defense (see Herms and Mattson 1992). This has led to numerous attempts to find a trade-off between plant fitness and the level of constitutive and induced defense (Haukioja et ai 1998, Koricheva et ai. 1998, Warren et ai. 1999). The CNB hypothesis predicts that fertilization will increase growth at the expense of C-based (e.g. phenolics and terpenoids) production of secondary metabolites (Bryant et al. 1983, Herms and Mattson 1992). In this study, the results support the CNB theory in induced defense reactions of N fertilized trees, but not in unwounded trees and with other fertilizations. Estimation of the cost of the defense reaction Heli Viiri 33 involves several unknown aspects; in many organisms the costs can be partly the costs for traits that are strongly correlated with defense but are not themselves defensive. The currency used, individual carbohydrates or total soluble carbohydrates, may not be representative either (IV). In mature trees, which have already passed the most active growth phase, it is difficult to relate minor changes in current growth rate to defense reactions. Difficulties in quantifying and detecting the costs of constitutive defense from induced defense are also relevant. The phloem probably represents the accumulation of photosynthates over a longer period of time, about 15-30 years of growth. Hamilton et al. (2001) have argued that CNB theory should be replaced with a new theory based on hypotheses that have an evolutionary underpinning that presupposed an adaptive value of any trait. In the long run we will see whether limitations in the predictive tool of CNB hypothesis will lead to the final formulation of the new theory that successfully predicts concentrations of defense-related secondary compounds. The question of whether N fertilization decreases defense due to increased growth was partly answered. Vigour indices and diameter growth were higher in N-fertilized treatments than in the control and with P fertilization (IV). Diameter growth and the vigour index correlated positively with the length of the lesion caused by C. polonica inoculation. Because fertilization increased the diameter of the experimental trees, density of the resin duct may have decreased as a result of increased cell division. This may partly explain why the total terpene concentration of N-fertilized trees was lower than in the control (III). Interestingly, fertilization did not affect total concentration of soluble carbohydrates, although with fertilization the growth of experimental trees improved significantly (IV). Thus the differences in carbohydrate composition after wounding are supported by the histological observations of Nagy et al. (2000). When the two-month incubation time and the low density of the inoculations are taker into account, short lesions probably indicate successful defease with minimum loss of energy and resources. Overall, when tested on artificial media, many Ceratocystis species are able to utilise various nitrogen sources (Mathiesen-Käärik 1960). Accordingly in the present study, carbon resources probably were not the restricting factor in defense reactions. 34 Results and discussion 3.3 Beetle-fungi-host tree interaction Various morphological, chemical and behavioural modifications in bark beetles and the ophiostomatoid fungi ensure a close association. Spruce bark beetles are among the phloem-feeding insects that benefit from association with fungi mainly because fungi reduce tree resistance, thus increasing the breeding success of the beetles. Although the first beetles in an attack may be overwhelmed and killed by the secondary compounds of the host tree, they may inoculate the tree with fungi that may linger for a long time. The fungi are dependent on the bark beetles for transport and for inoculation into suitable habitats. The potential of the fungi to weaken and predispose trees to subsequent attacks has been considered important, especially when the density of the beetle population has been low and there have not been enough beetles to mount successful mass attacks (Berryman 1982). As the vigour of Norway spruce increases, experimental trees tolerate an increasing number of spruce bark beetle attacks without being killed (Mulock and Christiansen 1986). According to Reid and Robb (1999), trees killed while growing vigorously, such as those felled by windthrow, may contribute significantly to initial increases in population, leading from endemic to epidemic populations of bark beetles. They suggest that the importance of phloem quality might be the key determinant of bark beetle performance. Bark beetle-fungal infection is only slightly inhibited by constitutive compounds; thus the induced response of the host is elicited and is essential in stopping invasion (Klepzig et al. 1996). In an elegant series of bioassays, Klepzig et al. (1996) showed that beetle responses to induced reactions could be a component of the host selection process. The beetles probably discriminate between susceptible host trees and vigorous trees capable of mounting an effective defense. The above-mentioned ideas have been united by characterization of the important role of fungal infection in activating PP cells quickly throughout several cell layers and launching a systemic induced defence reaction far away from the original site of infection (Franceschi et al. 2000). Some recent results have also supported the idea that mechanical stress, rather than associated fungi, plays an essential role in induction and development of the response of trees to bark beetles (Lieutier et al. 1995). Mechanically made control wounds were here cleaned effectively and sealed with constitutive resin, whereas only fungal inoculations caused induced accumulation of secondary compounds (III). However, both mechanical wounding and fungal infection of Norway spruce have resulted in enhanced Heli Viiri 35 resistance to subsequent mass inoculation with C. polonicci (Christiansen et al. 1999). One artificial inoculation cannot be considered to correspond to one bark beetle attack because artificial inoculations lead more surely to fungal infection of the phloem than to bark beetle attacks (Christiansen 1985b). The number of spores in one inoculum, 5 mm diameter slant of the colony of Ophiostoma fungus, can contain 1.5 x 106 spores (Lieutier etal. 1989). In natural infection, gallery construction by the beetles will spread the fungal spores more efficiently over a larger area than artificial point inoculations do. In addition, in natural bark beetle attacks, growth conditions are probably more favourable for fungi. Occasionally, the attacking beetles may not carry enough fungal spores for infection (Bridges and Moser 1983). Thus, the inoculation density might be more important than the fungal load per inoculation. An increase in the density of inoculations above a certain level has been reported to result in decreased resinosis (Raffa and Berryman 1983, Christiansen 1985 a). In low-density inoculations the ultimate reason for the long reaction zones may be the aggressiveness of fungus or the weakness of the host. Lesion length may vary depending on the season when trees are inoculated and on the length of the incubation period (Raffa and Smalley 1988, Parmeter et al. 1992). As used here, an incubation period of 8-10 weeks probably provides data from near the endpoint, i.e. when all defense reactions elicited by wounding are completed (Parmeter et al. 1992, 111, IV). According to Solheim (1988), C. polonica produced slightly shorter lesions than the less aggressive O. penicillatum did. According to Poppet al. (1995), large lesions with a high monoterpene concentration in secondary resin also reflect greater fungal virulence rather than greater host resistance. Thus, lesion length alone may not be a good measure of the aggressiveness of fungi or the resistance of trees. Lesion length should be considered a supporting and descriptive parameter, not a main parameter judging solely the resistance (111, IV). With low inoculation level, a tree saves energy and allocates resources to defense by forming a reaction zone that is just large enough to prevent the growth of fungi. However, with low inoculation density, resin exudation is expected to be powerful, since the tree response is not exhausted immediately. Different bark beetles can have specific fungal associates, which occur only with their host species (Jacobs and Wingfield 2001). Both ophiostomatoid fungi and spruce bark beetles can exist and develop without their associates, but in nature they are commonly found together; the relationship is nearly symbiotic. 36 Results and discussion C. polonica, C. minuta, O. bicolor and O. penicillatum were constant associates of I. typographus (I, II). C. polonica has been isolated mainly from spruce bark beetles (Solheim 1986, Kirisits 1996, Krokene and Solheim 1996, I, II) whereas the rest of the ophiostomatoid species mentioned here occur together with several bark beetles and some of them also spread by air. Species were classified here according to their frequency of occurrence (Table 2). Some fungi can be classified as constant associates of I. typographus, while others are common or casual associates. Nevertheless, it is difficult to decide whether the symbiosis between beetles and specific fungi is obligate. The success of rearing sterile adult beetles alone does not rule out the presence of a symbiotic association. It is possible to rear generations of sterile bark beetles in vitro with species that normally are associated with fungi (Grosmann 1931, Harding 1989). Thus, the presence of fungi is not a prerequisite for larval nutrients and establishment of the spruce bark beetle, in spite of all the evidence pointing to a symbiotic relationship between the beetle and fungi. Also in natural conditions successful bark beetle infestations have been found without a constant fungal associate, like Dendroctonus frontalis Zimmermann infestations without Ceratocystis minor (Hedgcock) Hunt (Bridges et al. 1985). In addition, different associated fungi can have a variable role in different interactions and under different environmental conditions (Harding 1989, Lieutier et al. 1995). According to Paine et al. (1988), mycangial fungi did not trigger the induced wound response in host trees compared with sterile wounding, whereas the less adapted or less specialized non-mycangial C. minor did induce lesion formation. The virulence of ophiostomatoid fungi may differ geographically, and the resistance of host trees may vary in different environments. Recent investigations (Stout et al. 1998) point out that induced resistance responses show variation in elicitation specificity and the organism involved. To understand the interaction between variable defensive reactions against stress factors and how physiological processes affect these responses, studies that take into account the specificity of induced responses are needed. Investigations are also needed to evaluate different strains of ophiostomatoid species; thus variation in sporulation ability and frequency (I, II) and virulence (Kirisits 1996, 1998) indicates variation within the current species concepts. 37 Table 2. Comparison of basic characteristics of I. typographus associated ophiostomatoid fungi. Intimacy of association refers the relationship between spruce bark beetle and associated fungi. Constant = fungus present > 90 % of studied beetles; common = fungus present 89-11 % of studied beetles; casual = fungus present < 10 % of studied beetles. Abbreviations of bark beetle names: D.ruf- Dendroctonus rufipennis, D.val = Dendroctonus valens, Dryo = Dryocoetes sp., H.gla = Hylorgops glabratus, H.pal = Hylorgops palliatus, I.ami = Ips amitinus, l.cem = Ips cembrae, l.dup = Ips duplicatus, I.typo = I. typographus, P.cha = Pityogenes chalcographus, P.pol = Polygraphus poligraphus, T.pini = Tomicus piniperda. h- O) CD > II 00 CO CD CD .d CN CD CD CD 00 CD E fc CD Ö "2 hj- Q) IT) CO . - CD O CD TO 2 E TO I» £ T o -Q CO c ■O C T— TO CD 0 2 c 1 1 5 § Ä .<£ C —1 Cl) T3 O C C TO 0 £ I S. o> Intimacy of association 6 UOIUIUOO casual 12 constant 46 common 579 casual 8 common 26 casual 58 common 3 9 casual 125 casual 15 constant 128 O) 00 o m LO (N CD O 00 h~ CT) CT> O CD O O lO O h- LO r-- CT) "3- O) O LO 00 O O O O "3" CT) o LO CD CD o CD LO Oi O CT) CM O o -51- CD CD 00 CD CM 00 CM O O CD O 00 CM CD 00 CO CM h- O o CM o CD O O CT) O) O CO LO o CD O h- b 0) CL o 0) c CM "tf- C\J CD CO r^- CD 00 LO CD to CO O CD CO o o CD O «O CO CM CM ö ■sl- CD 4 CM Lf> LO o h- CD CO CM CO O CD CM O o CO ■<3- O LO CD CD CT) CD CD O 00 CD LO CM CD CM CM CD Q. Z> "Sf LO h- LO Q. CO Z3 Tfr E 3 E zL O N «S 0 V) ro -Q CJ) O lO O a> O CO O CO O O x: "iZ Q) CL d) O) O 00 Ö CD 48-87 11 186-392 : CD Tf CO 1 CM C\J O CD CO O o CNJ 90-140 5 CM N- CT) 206-333 : CM CM CO lö CM h- co CO CD 76-110 5 o h- o CD CD CD CO CD 125-160 ! 186-303 : 250-327 f 200-310 ! O CD CO ö 00 196-362 : o h- CM CO o CM o CD CM O CM CM 121-197- O CO CT) O CM O O CM O CO LO I s - "sj- CO h- CO Ascospore shape orange section-shape, falcate ellipsoid, cucullate cylindrical cylindrical, oblong mandarin slice-shape cylidrical, ossiform hat-shape allantoid, cylindrical orange section-shape allantoid (/> Q) O 0 ■2 s c § 03 O c o o Q. CL> Ct3 O C CD .O O .o S 3 S "S o s o E CO o s X .s/Cerafoc>'.sris-infection of Nor way spruce: What is deadly dosage? Z. ang. Ent. 99, 6- 11. Dowding, P., 1969: The dispersal and survival of spores of fungi causing bluestain in pine. Trans. Br. mycol. Soc. 52, 125-137. Francke-Grosmann, H„ 1963: Some new aspects in forest entomology. Ann. rev. Ent. 8, 415-438. Furniss, M. M.; Solheim, H; Christiansen, E„ 1990: Transmission of Blue-Stain Fungi by Ips typographus (Coleoptera: Scolytidae) in Norway Spruce. Ann. Ento mol. Soc. Am. 83, 712-716. Harding, S.. 1989 a: Blue stain fungi associated with Ips typo graphy L. (Coleoptera: Scolytidae) in host trees of differ ent vitality and at different beetle population levels. In: The influence of mutualistic blue stain fungi on bark beetle population dynamics. PhD thesis. Department of Zoology, Royal Veterinary and Agricultural University, Copenhagen II; 1-30. —, 1989b: The pathogenicity of the blue stain fungi Ophios toma polonicum, Ophiostoma europhioides, and Ophios toma penicillatum, associated with Ips typographus L. (Coleoptera: Scolytidae), to Norway spruce. In: The influence of mutualistic blue stain fungi on bark beetle population dynamics. PhD thesis. Department of Zoology, Royal Veterinary and Agricultural University, Copenhagen III: 1-34. —, 1989 c: Rearing of spruce bark beetles, Ips typographus L. (Coleoptera: Scolytidae), without symbiotic blue stain fungi. In: The influence of mutualistic blue stain fungi on bark beetle population dynamics. PhD thesis. Depart ment of Zoology, Royal Veterinary and Agricultural Uni versity, Copenhagen IV; 1-34. Harrington, T. C.. 1993: Diseases of conifers caused by species of Ophiostoma and Leptographium. In: Cera tocystis and Ophiostoma, Taxonomy, Ecology and Patho genicity. Ed. by Wingfield, M. J., Seifert, K. A., Webber, J. F. St. Paul, Minnesota: APS Press, 161-172. Horntvedt, R.; Christiansen, E.; Solheim, H.; Wang, S., 1983: Artificial inoculation with Ips typographus-nssoci ated blue-stain fungi can kill healthy Norway spruce trees. Medd. Nor. inst. skogforskning 38, 1-20. Krokene, P.; Solheim, H„ 1996: Fungal associates of five bark beetle species colonizing Norway spruce. Can. J. For. Res. 26,2115-2122. Malloch, D.; Blackwell, M„ 1993: Dispersal biology of the Ophiostomatoid Fungi. In: Ceratocystis and Ophio stoma, Taxonomy, Ecology and Pathogenicity. Ed. by WINGFIELD, M. J., SEIFERT, K. A., WEBBER. J. F. St. Paul, Minnesota: APS Press, 195-206. Mathiesen-KääRIK, A., 1960: Studies on the ecology, taxo nomy and physiology of Swedish insect-associated blue stain fungi, especially the genus Ceratocystis. Oikos. 11, 1-25. M ehta, C. R.; Patel, n. R., 1989, 1991: StatXact™. Sta tistical Software for Exact Nonparametric Inference. User Manual. Version 2. CYTEL Software Corporation. Cam bridge. Moser, J. C.; Perry, T. J.; Solheim, H., 1989: Ascospores hyperphoretic on mites associated with Ips typographus. Mycol. Res. 93, 513-517. SOLHEIM, H., 1986: Species of Ophiostomataceae isolated from Picea abies infested by the bark beetle Ips typo graphus. Nord. J. Bot. 6, 199-207. —, 1992 a: The early stages of fungal invasion in Norway spruce infested by the bark beetle Ips typographus. Can. J. Bot. 70, 1-5. —, 1992b: Fungal succession in sapwood of Norway spruce infested by the bark beetle Ips typographus. Eur. J. For. Path. 22, 136-148. —. 1993: Fungi associated with the spruce bark beetle Ips typographus in an endemic area in Norway. Scand. J. For. Res. 8, 118-122. UPADHYAY, H. P., 1981: A Monograph of Ceratocystis and Ceratocystiopsis. Athens: The University of Georgia Press. 175 pp. —, 1993: Classification of the Ophiostomatoid Fungi. In: Cera tocystis and Ophiostoma, Taxonomy, Ecology and Patho genicity. Ed. by WiNGFIELD, M. J., SEIFERT, K. A., Webber, J. F. St. Paul, Minnesota: APS Press, 7-13. Whitney, H. S., 1982: Relationships between Bark Beetles and Symbiotic Organisms. In: Bark beetles in North American Conifers. A System for the Study of Evo lutionary Biology. Ed. by Mitton, J. 8., Sturgeon, K. B. Austin: University ofTexas Press, 183-211. WHITNEY, h. s.; Blauel, R. A., 1972: Ascospore dispersion in Ceratocystis spp. and Europhium clavigerum in conifer resin. Mycologia 64, 410-414. Wright, E., 1935: Trichosporium symbioticum, n.sp., a wood staining fungus associated with Scolytus ventralis. J. Agric. Res. 50, 525-538. Author's address: Heli Vli'Rl, University of Joensuu, Faculty ofForestry, PO Box 111, FIN-80101 Joensuu, Finland II 1 Ophiostomatoid fungi associated with the spruce bark beetle, Ips typographus, in post-epidemic areas in France HELI VIIRI u " 'Faculty of Forestry, University of Joensuu, PO.Box 111, FIN-80101 Joensuu, Finland 2Present address: Finnish Forest Research Institute, Suonenjoki Research Station, Juntintie 40, FIN-77600 Suonenjoki, Finland; heli.viiri @metla.fi. Tel. +358 17 513 8219, Fax+3sB 17 513 068 Corresponding author FRANCOIS LIEUTIER 3 4 Tnstitut National de la Recherche Agronomique, Station de Zoologie Forestiere, Ardon, F-45160 Olivet, France 4 Laboratoire de biologie des ligneux et des grandes cultures, Universite d'Orleans, B.R 6759, 45067 Orleans Cedex 02, France; francois.lieutier@univ-orleans.fr. Tel. +33 2 38 49 48 07, Fax +33 2 38 41 78 79 Abstract The species composition of ophiostomatoid fungi associated with Ips typographus was studied in the Vosges, Alps and Massif Central regions of France. In each region, damage caused by bark beetles has increased during recent years. For this study, beetles were collected individually by hand from freshly attacked trees and crushed in healthy Picea abies logs. Fungi were isolated from log phloem and sapwood, and identified. The most frequently found species were Ophiostoma bicolor, O. penicillatum , Ceratocystiopsis minuta and Ceratocystis polonica. Results are discussed in terms of differences between locations and in relation to previous investigations in which populations of spruce bark beetle have been sparse. The potential role of associated fungi in the population dynamics of the spruce bark beetle is discussed. Key words: associated fungi / Ceratocystis / Ips typographus / Ophiostoma 2 1 INTRODUCTION European spruce (Picea spp.) forests suffer regularly from extensive outbreaks of the spruce bark beetle Ips typographus L. (Coleoptera: Scolytidae). Eurasian spruce bark beetles together with associated pathogenic fungi have killed millions of cubic metres of spruce in western and central Europe during recent years. In north-eastern France alone, damage has been as high as 100,000 m 3 in 1991, 212,500 m 3 in 1992 and 113,000 m 3 in 1993 (Boutte 1993, Nageleisen 1994, 1995). Severe beetle damage often follows heavy storm damage and windfall, e.g. as a result of the severe windstorm in December 1999. Adults of the spruce bark beetle transport spores of blue-staining fungi both in the pronota and elytra and in the digestive tract (Furniss et al. 1990). When building breeding chambers and galleries, spruce bark beetles introduce the spores of Ophiostoma and Ceratocystis species into the phloem and cambium of Norway spruce, Picea abies L. Karsten. Together with associated fungi, spruce bark beetles can overcome the resistance of vigorous spruce trees. In the most harmful species, Ceratocystis polonica (Siemaszko) Moreau, pathogenicity is based on its ability to grow rapidly in moist wood through the tracheids and to disrupt water transport in the tree, finally leading to high mortality (Horntvedt et al. 1983, Christiansen 1985, Solheim 1988, Krokene and Solheim 1998, Kirisits 1998). The main aim of this investigation was to describe the ophiostomatoid fungi associated with /. typographies in a locally high population level of spruce bark beetles. As further aim was to compare the fungal flora associated with spruce bark beetles collected from different regions. This information will provide us with useful elements for understanding the role of associated fungi as possible regulators of bark beetle epidemics. 2 MATERIAL AND METHODS 2.1 Study areas Beetles were collected at the beginning of the main swarming period of the first generation, in late May and early June 1996, from three regions in France: Vosges, Alps and Massif Central (Figure 1, Table I). Two locations in each region were selected on the basis of previous large populations of beetles, and 50 beetles were collected at each location. At all locations, extensive spruce bark beetle damage has occurred in 1990-1995 (Boutte 1993, 1994). In Vosges, where two generations occur each year, the volume of dead Norway spruce varied between 1,200 and 5,900 m 3 in 1991-1995. In 1995, beetles were collected in pheromone traps and the total catch for three pheromone traps was 2,219 spruce bark beetles, thus indicating a declining trend (Office National des Forets, Raon l'Etape). In Massif Central, at the Mezenc collecting site, the 3 high altitude reduces reproduction and only one generation of spruce bark beetles occurs annually. In Meygal, depending on weather conditions, 1-2 generations occur per year. Fig. 1. Location of the I. typographus collecting areas. 1 = Vai de Senones, 2 = Vologne, 3 = St Pierre de Belleville, 4 = St Michel de Maurienne, 5 = Meygal, 6 = Mezenc. Table 1. Study areas used for collection of 1. typographus. Location Forest Elevation (m.s.l.) Stand age (yrs) I.Col de Praye, Vosges Val de Senones 910 130 2.Tete de Nayemont, Vosges Vologne 840 130 3. St Pierre de Belleville, Savoie St Pierre de Belleville 1350 150 4. St Michel de Maurienne, Savoie piles 450 150 5. Boussoulet, Haute Loire Meygal 1300 120 6. Mont d'Alambre, Haute Loire Mezenc 1480 120 4 2.2 Collecting beetles At all locations, except St Michel de Maurienne, beetles were collected individually by digging out adult females and males with a knife and forceps from windblown Norway spruce trunks lying in the forest. In St Michel de Maurienne, beetles were collected in the Norway spruce trunks lying in a timber yard. Beetles were placed individually into sterile Eppendorf-test tubes. The equipment used for collection was sterilized after extraction of each individual. The logs had fallen during the previous winter and the beetles had just started to build galleries in them. Mostly the construction of nuptial chambers was completed and the mother galleries had been initiated. However, the mother galleries were less than 4 cm long. The collected beetles were stored individually at +4 °C in Eppendorf-test tubes for a maximum of three days before they were introduced into logs. 2.3 Inoculation, isolation and identification of fungi Fungi were pre-cultivated in fresh uninfected Norway spruce bolts (one metre long, diameter 15 cm) according to the method described previously by Furniss et al. (1990). The bolts were brushed and the surfaces wiped with 70 % alcohol. To prevent drying, the ends of the bolts were dipped into melted paraffin. Then 25 beetles were introduced individually into each log through holes (5 mm diameter) bored previously with a cork-borer at the level of the cambium. After the beetle was introduced, the bark plugs were replaced and the beetles were crushed gently. In each log two control holes without beetles were made and treated similarly. After 21 days of incubation at room temperature (+2O °C) reaction zones formed with phloem around each inoculation point. These reaction zones were then cut from the logs, wrapped in foil and stored at +4 °C for two weeks until used for isolations. Two phloem samples (50-60 mm 3 ) were taken from inside each necrotic zone, one at the border of the visible reaction and one at a distance of 15 mm. Two samples were also taken from a depth of Imm in the sapwood. When the reaction zones were less than 20 mm long, all four samples were taken from the edge of the visible reaction zone. When reaction zones were more than 150 mm long, six samples were taken, four from the phloem and two from the sapwood. A total of 1,221 primary samples were taken around the inoculation points. Samples were cultivated in petri dishes (2 % malt and 1.4 % agar medium) at room temperature. Occasionally, pieces of fresh autoclaved phloem or sapwood of Norway spruce were added to the dishes to promote formation of sexual stages. A type specimen of each ophiostomatoid fungi was purified by transferring mycelium or spores from individual colonies to fresh malt agar medium. 5 Reproduction structures of the fungi were prepared with lacto-fuchsin, lactic acid or cotton blue for identification. Fungal structures were compared with the species descriptions given in the literature (Siemaszko 1939. Davidson 1953, Mathiesen-Käärik 1953, Wright and Cain 1961, Kendrick 1962, Davidson et ai 1967, Upadhyay 1981, De Hoog and Scheffer 1984, Solheim 1986, Kirisits 1996, Yamaoka et al. 1997). 2.4. Statistics Frequencies of ophiostomatoid species were analysed with the Kruskal-Wallis test. Due to the sparseness of the observed frequencies, the data were analysed with StatXact™ Version 2.11 software, a statistical package for exact nonparametric inference (Mehta and Patel 1989, 1991). Since the data sets were too large for exact calculation of p-value, the Monte-Carlo estimates of the p-value were computed by generating 100,000 tables. The level of significance in the tests was p < 0.01. 3 RESULTS The most common and consistently occurring species were Ophiostoma bicolor Davidson & Wells, O. penicillatum (Grosm. Siemaszko), Ceratocystiopsis minuta (Siemazko) Upadhyay & Kendrick and C. polonica. Other frequently isolated species were O. piceaperdum (Rumbold) Arx and O. ainoae Solheim (Table II). O. piceae (Miinch) H. & P. Sydow and O. cucullatum Solheim were isolated occasionally. There was no visible staining on any of the control inoculations, and no ophiostomatoid fungi were detected in the control inoculations. Furthermore, an unidentified strain of ophiostomatoid species was isolated which differed from known species by having perithecia formed in mixed cultures within the malt extract substratum, perithecia black, 160-230 |im in diameter, necks 3,840-4,800 pm long, tapered, dark brown to black, 40-50 (I m wide at the base, 20 (xm wide at the apex, and ostiolar hyphae absent. Ascospores were cucullate with a hyaline wall, released in a slimy mass. No conidia or conidiophores were seen. When the frequencies of nine ophiostomatoid species were compared simultaneously at six beetle-collection locations, the Kruskal-Wallis analysis of variance indicated a highly significant difference between locations (% 2 = 29.04, df = 8, asymptotic p-value = 0.0003). When the five most frequent species (C. minuta , C. polonica, O. bicolor , O. piceaperdum and (). penicillatum) were compared, there was also a significant difference between locations (locations (% 2 = 16.86, df= 4, asymptotic p-value = 0.0021). 6 Table 2. Frequencies of occurrence of ophiostomatoid fungi associated with I. typographus collected at six locations in France. Locations presented in Table 1. n =5O beetles per location. 4 DISCUSSION The blue-stain fungi Ceratocystis polonica, Ophiostoma bicolor, O. europhioides Wright & Cain (Solheim) and O. penicillatum have been reported to be associated with I. Typographus, occurring with varying frequency in different environmental conditions and investigations (Solheim 1986, 1992, 1993, Harding 1989, Krokene and Solheim 1996, Viiri 1997). O. europhioides was recently synonomised with O. piceaperdum, since they cannot be distinguished on the basis of morphology (Jacobs et al. 2000). Some characteristics of the long-necked ophiostomatoid species isolated here are the same as the species characteristics of O. piceaperdum, but possible synonymy needs to be clarified in more detailed studies of teleomorph and anamorph morphology. The most common and consistently occurring fungus in this study was O. bicolor, which was recovered in Vologne from 74 % of the bark beetles examined. At nearly all locations, C. minuta, O. ainoae, O. bicolor, O. penicillatum and O. piceaperdum occurred at higher frequencies than recorded Vosges Alps Massif Central Senonne Vologne Belleville Maurienne Meygal Mezenc C. minuta 62 36 36 30 28 24 C. polonica 40 32 22 50 42 30 0. ainoae 2 10 28 12 24 10 0. bicolor 54 74 26 44 66 42 0. cucullatum 0 0 0 0 2 0 0. piceaperdum 20 34 30 10 16 28 0. penicillatum 40 40 24 26 60 40 0. piceae 8 12 10 12 8 2 "Long-necked" 2 0 6 4 8 0 Pesotum spp. 36 28 46 46 50 58 Leptographium spp. 2 2 0 0 0 0 Primary isolations 204 204 202 202 205 204 7 from the low population density areas of I. typographus (Solheim 1993, Viiri 1997). The following ophiostomatoid species have been reported to be associated with Ips sexdentatus Boern in France: C. minuta, O. bicolor, O. brunneo-ciliatum Mathiesen-Käärik, O. europhioides, O. Ips (Rumbold) Nannf., O. piceae and O. minus (Hedgcock) H. & P. Sydow (Levieux et al. 1989, Lieutier et al. 1989, 1991). Correspondingly, the species found to be associated with Ips acuminatus Gyll are O. brunneo-ciliatum, O. Ips, O. minus and Ceratocystiopsis minima (Olchow. and Reid) (Lieutier et al. 1991). According to surveys made in previous years, in all sampling areas, especially in Vosges and Massif Central, the population levels of the spruce bark beetle were high. This resulted in numerous spontaneous attacks on spruce trees in these areas. Pheromone trapping, although done only in Vosges, showed declining population size already during the year of beetle sampling. Thus the isolated fungal flora constantly corresponded to a beetle population in the post-epidemic phase. According to Weslien et al. (1989), fewer than 15,000 spruce bark beetles in a group of three traps corresponds to a low population level. Hiibertz et al. (1991) caught 3,400-12,000 individuals and Valkama et al. (1997) at highest 14,000 individuals per season with a group of three traps during a period when the beetle population was low. According to Yamaoka et al. (1997), the technique used to isolate ophiostomatoid fungi from various niches can greatly affect the frequencies of occurrence. Thus when results are compared to those of other authors, discrepancies in fungal frequencies may be partly due to differences in methods of sampling and isolation. In this study, however, the fungal flora differed significantly between locations. Both C. polonica and O. piceaperdum have been suggested to play a special role in the population dynamics of the spruce bark beetle (Solheim 1993, Harding 1989). It has been proposed that during endemic periods when beetles utilise dead trees and timber for breeding, pathogenic species can be replaced by less harmful ones. In Norway, the frequency of C. polonica has been low during periods of low population level when beetles use dead trees and timber, whereas the frequency has been higher during the epidemic phase when living trees are attacked (Solheim 1992, 1993, Krokene and Solheim 1996). The previous finding that the frequency of the pathogenic species, C. polonica (Viiri 1997), in the endemic population of spruce bark beetle is low does not conflict with the fact that associated pathogenic fungi can regulate the damage by spruce bark beetles. Our results are thus in agreement with those suggesting that pathogenic species can be replaced by other species during endemic periods. Furthermore, they support the idea that the role of the associated fungi may vary under different environmental conditions. The success or failure of bark beetle attacks on living trees is ultimately determined by the beetle-fungus-host tree interaction. Owing to conflicting results concerning frequency and pathogenicity, genetic variation within the 8 species O. piceaperdum and C. polonica needs to be clarified (Harding 1989, Kirisits and Angelberger 1999, Krokene and Solheim 2001). The pathogenicity of geographically different strains of O. piceaperdum and C. polonica should be tested. Acknowledgements This work was supported by the Graduate School of Forest Sciences, Ministry of Education, Finland and the Institut National de la Recherche Agronomique, France. Collection of samples in France was supported by grants from Konkordialiitto and the Halonen Foundation. We thank Jacques Garcia and Eeva Vehviläinen for technical assistance, Marja Poteri for comments on the manuscript and Joann von Weissenberg checking the English language. REFERENCES Boutte B„ Le typographe de I'epicea, in: Departement de la Sante de Forets, La Sante des Forets [France] en 1992, Ministere de l'agriculture, de la peche et de ralimentation, (1993) 12-13. Boutte 8., Le typographe de I'epicea, in: Departement de la Sante de Forets, La Sante des Forets [France] en 1993, Ministere de l'agriculture, de la peche et de l'alimentation, (1994) 14-17. Christiansen, E., lps/Ceratocystis- i n fee ti on of Norway spruce: what is deadly dosage? Z. ang. Ent. 99 (1985) 6-11. Davidson R.W., Two common lumber-staining fungi in the western United States, Mycologia 45 (1953) 579-586. Davidson R.W, Francke-Grosmann H., Käärik A., A Restudy of Ceratocystis penicillata and report of two American species of this genus from Europe, Mycologia 59 (1967) 928-932. Furniss M.M., Solheim H., Christiansen E., Transmission of blue-stain fungi by Ips typographus (Coleoptera: Scolytidae) in Norway spruce, Ann. Entomol. Soc. Am. 83 (1990) 712-716. Harding S., The influence of mutualistic blue stain fungi on bark beetle population dynamics, Ph.D. thesis, Department of Zoology, Royal Veterinary and Agricultural University, Copenhagen, 1989. De Hoog G.S., Scheffer R.J., Ceratocystis versus Ophiostoma: A Reappraisal, Mycologia 76 (1984) 292-299. Horntvedt R., Christiansen E., Solheim H., Wang S., Artificial inoculation with Ips typographus-associated blue-stain fungi can kill healthy Norway spruce trees, Rep. of Norwegian For. Res. Inst. 38 (1983) 1-20. Hiibertz, H., Larsen, J.R., Bejer, 8., Monitoring spruce bark beetle (Ips typographus ) populations under non-epidemic conditions, Scand. J. For. 9 Res. 6(1991)217-226. Jacobs, K., Wingfield, M.J., Crous, P.W., Ophiostoma europhioides and Ceratocystispseudoeurophioides, synonyms of O. piceaperdum, Mycol.Res. 104 (2000) 238-243. Kendrick W. 8., The Leptographium complex. Verticicladiella Hughes, Can. J. Bot. 40(1962) 629-770. Kirisits, T., Untersuchungen iiber die Vergesellschaftung von Bläuepilzen (Ophiostoma/Ceratocystis spp.) mit den rindenbriitenden Fichtenborkenkäfern Ips typographic L., Pityogenes chalcographus L. und Hylurgops glabratus Zett. in Österreich. Master thesis, Insitut fur Forstentomologie, Forstpathologie und Forstschutz, Universität fiir Bodenkultur, Wien. 175 p. 1996. Kirisits, T., Pathogenicity of three blue-stain fungi associated with the bark beetle Ips typographus to Norway spruce in Austria, Österr. Z. Pilzk. 7 (1998) 191-201. Kirisits, T. & Anglberger, H., Report on a strain of the pathogenic blue-stain fungus Ceratocystispolonica with low virulence, Österr. Z. Pilzk. 8 (1999) 157-166. Krokene P., Solheim H., Fungal associates of five bark beetle species colonizing Norway spruce, Can. J. For. Res. 26 (1996) 2115-2122. Krokene P., Solheim H., Pathogenicity of four blue-stain fungi associated with aggressive and nonaggressive bark beetles. Phytopathology 88 (1998) 39- 44. Krokene P., Solheim H., Loss of pathogenicity in the blue-stain fungus Ceratocystis polonica , Plant Pathology 50 (2001) 497-502. Levieux, J., Lieutier, F., Moser, J.C., Perry, T.J., Transportation of phytopathogenic fungi by the bark beetle Ips sexdentatus Boerner and associated mites, J. App. Ent. 108 (1989) 1-11. Lieutier, F., Garcia, J., Yart, A., Vouland, G., Pettinetti, M. & Morelet, M., Ophiostomatales (Ascomycetes) associees ä Ips acuminatus Gyll (Coleoptera: Scolytidae) sur le pin sylvestre (Pimts sylvestris L) dans le Sud-Est de la France et comparaison avec Ips sexdentatus Boern, Agronomie 11 (1991) 807-817. Lieutier, F., Yart, A., Garcia, J., Ham, M.C., Morelet, M. & Levieux, J., Champignons phytopathogenes associes ä deux coleopteres scolytidae du pin sylvestre (Pinus sylvestris L.) et etude preliminaire de leur agressivite envers l'hote, Ann. Sci. For. 46 (1989) 201-216. Mathiesen-Käärik A., Eine Übersicht liber die gewohnlichsten mit Borkenkäfern assoziierten Bläuepilze in Schweden und einige fiir Schweden neue Bläuepilze, Medd. Stat, skogsforskningsinst. 43 (1953) 1-74. Mehta, C.R., Patel, N.R., StatXact™. Statistical Software for Exact Nonparametric Inference. User Manual. Version 2. CYTEL Software Corporation. Cambridge, MA. 1989. 1991. Nageleisen L-M., Lutte Contre les Scolytes en 1993, in: Information technique 10 N°2l, Departement de la Sante des Forets Nord-Est, Ministere de P agriculture, de la peche et de Palimentation, 1994. Nageleisen L.-M., Epicea: le typographe, in: Bilan Departement de la Sante des Forets Nord-Est 1994, Ministere de I' agriculture, de la peche et de Palimentation, 1995. Siemaszko W., Zespoly grzybow kornikom Polskim, Planta Polonica7 (1939) 1-54. Solheim H., Species of Ophiostomataceae isolated from Picea abies infested by the bark beetle Ips typogrciphus, Nord. J. Bot. 6 (1986) 199-207. Solheim H., Pathogenicity of some Ips typographus-associated blue stain fungi to Norway spruce, Rep. of the Norwegian For. Res. Inst. 40 (1988) 1-11. Solheim H., The early stages of fungal invasion in Norway spruce infested by the bark beetle Ips typographus, Can. J. Bot. 70 (1992) 1-5. Solheim H., Fungi associated with the spruce bark beetle Ips typographus in an endemic area in Norway, Scand. J. For. Res. 8 (1993) 118-122. Upadhyay H.P., A Monograph of Ceratocystis and Ceratocystiopsis, The University of Georgia Press, Athens, Georgia, 1981. Valkama, H., Räty, M., Niemelä, P., Catches of Ips duplicatus and other non target Coleoptera by Ips typographus pheromone trapping, Entomol. Fenn. 8 (1997) 153-159. Viiri H., Fungal associates of the spruce bark beetle Ips typographus L. (Col. Scolytidae) in relation to different trapping methods, J. Appl. Ent.l2l (1997) 529-533. Weslien, J., Annila, E., Bakke, A., Bejer, 8., Eidmann, H.H., Narvestad, K., Nikula, A. & Ravn, H.P., Estimating risks for spruce bark beetle (Ips typographus ) damage using pheromone baited traps and trees. Scand. J. For. Res. 4 (1989) 87-98. Wright E.F., Cain R.F., New species of the genus Ceratocystis, Can. J. Bot. 39 (1961) 1215-1230. Yamaoka Y., Wingfield M.J., Takahashi 1., Solheim H„ Ophiostomatoid fungi associated with the spruce bark beetle Ips typographus f.japonicus in Japan, Mycol. Res. 101 (1997) 1215-1227. III Trees (2001)15:112-122 DOI 10.1007/sOO4-680000082 Heli Viiri • Erkki Annila • Veikko Kitunen Pekka Niemelä Induced responses in stilbenes and terpenes in fertilized Norway spruce after inoculation with blue-stain fungus, Ceratocystis polonica Received: 15 May 2000 / Accepted: 25 October 2000 / Published online: 16 January 2001 © Springer-Verlag 2001 Abstract In conifers, attacks by bark beetles and associ ated pathogenic fungi cause an induced wound response, which is characterized by accumulation of antifungal compounds and morphological changes that aid wound healing. In this article the stilbene and terpene concentra tions of Norway spruce phloem were monitored as symptoms of induced wound responses in relation to changed nutrient conditions caused by fertilization. Plots of mature Norway spruce were fertilized with N, P or NPK. One year after fertilization the trees were artificial ly infected with Ceratocystis polonica , a pathogenic fun gus associated with the bark beetle Ips typographus. The response of stilbenes to fungal inoculation was mainly qualitative. The concentration of stilbene glycosides in the phloem decreased, and in the immediate vicinity of the site of fungal inoculation, stilbene glycosides were less frequent than in mechanically wounded or un wounded phloem. Corresponding stilbene aglycones were most frequent inside the reaction lesion. The con centration of total stilbene aglycones near the inoculation site was significantly lower in N-fertilized trees than in unfertilized trees. Fungal inoculation caused a strong quantitative response in terpenes. The total terpene con centration of the phloem increased significantly, to al most 100 times greater near the inoculation site com pared to the constitutive values. N fertilization signifi cantly reduced the total terpene and total stilbene agly cone concentrations near the inoculation sites. Thus, N fertilization may reduce the ability of Norway spruce to defend itself against fungal pathogens. H. Viiri (S3) • R Niemelä Faculty of Forestry, University of Joensuu. PO Box 111. 80101 Joensuu, Finland E. Annila • V. Kitunen Finnish Forest Research Institute, Vantaa Research Centre, PO Box 18. 01301 Vantaa. Finland Present address: H. Viiri. Finnish Forest Research Institute, Suonenjoki Research Station, Juntintie 40, 77600 Suonenjoki. Finland, e-mail: heli.viiri@metla.fi, Tel.: +358-17-5138219, Fax: +358-17-513068 Keywords Carbon/nutrient balance hypothesis • Fertilization • Induced defence • Monoterpenes • Phenolics Introduction Many bark beetles distribute spores of fungi laterally or internally in the digestive tract while constructing breed ing chambers and galleries in the phloem of their host trees. As they grow, fungal hyphae suppress water trans portation in the host, causing discolouration of wood and helping the beetles to kill the trees. New generations of beetles transport fungi to new host trees. For many spe cies of fungi, transmission by insects is vital, and special associations have arisen between insects and fungi. Most fungi associated with the Scolytidae belong to the genus Ceratocystis sensu stricto De Hoog and Scheffer, Ophio stoma H. and P. Sydow and their anamorph genera (Wingfield et al. 1993; Krokene and Solheim 1996). These fungi are adapted to dispersal by insects: elongat ed ascocarps bear ascospores at the apices of their necks, which may be protected by a gelatinous matrix (Wingfield et al. 1993). From the human point of view, a major part of the symbiosis between bark beetles and associated fungi is their joint action to overcome the resistance mechanisms of their host trees. Some fungi are pathogenic and, when mass inoculated into trees, are able to kill healthy trees (Christiansen 1985). When phloem is infected by bark beetles and associated pathogenic fungi, a resistance re action may be initiated that is characterized by rapid des iccation of cells and necrosis around the site of the wound (Berryman 1972). The lesion surrounding the at tacked site is impregnated with secondary metabolites, i.e. phenolic and terpenoid compounds, and isolated by the formation of wound periderm, callus tissue and trau matic resin cavities at the cambium-sapwood interface (Reid et al. 1967; Berryman 1969; Woodward and Pearce 1988 a, 1988b: Delorme and Lieutier 1990). The co determinants of conifer resistance to bark beetle-fungus attack have been tentatively identified as the primary, or 113 resident, resin canal system and the hypersensitive reac tion, which functions as wound cleansing, containment of infection and wound healing (Berryman 1972). In duced wound responses can start as a result of mechani cal wounding, but formation of traumatic secondary compounds and extension of the lesion require fungal in fection or boring activity by beetles (Lieutier et al. 1989; Franceschi et al. 1998). Accumulation of monoterpenes in high concentrations is characteristic of the defensive response in several coni fers; a-pinene. (3-pinene, myrcene, terpinolene, sabinene, A-3-carene or limonene, in particular, have been shown to increase in response to inoculation with fungi (Russell and Berryman 1976; Raffa and Berryman 1982 a; Gershenzon 1994). This resinous material is stored in specialized resin ducts or granular cells (Berryman 1969; Franceschi et al. 2000; Nagy et al. 2000). Monoterpenes prevent growth of pathogenic fungi (Cobb et al. 1968; Delorme and Lieutier 1990) and repel or kill bark beetles (Raffa et al. 1985; Bridges 1987; Raffa and Smalley 1995). Thus, a change in any of these compounds may be important for the defence of the tree against beetles and fungi. The bark of many conifers also contains antifungal phe nolic compounds, i.e. stilbenes and flavonoids. In healthy phloem of Norway spruce, Picea abies (L.) Karst., stil benes typically occur as glycosides (Brignolas et al. 1995). The main constitutive stilbene glycosides in the bark of Picea species are astringin and isorhapontin (Woodward and Pearce 1988 a; Solhaug 1990; Toscano Underwood and Pearce 1991 a, 1991b; Lindberg et al. 1992). In vitro these compounds have antifungal properties (Woodward and Pearce 1988 a). Rapid accumulation of stilbenes in response to injury or fungal infection is considered to be an active defence response (Nicholson and Hammerschmidl 1992; Schultz et al. 1992). Phenolics are formed in situ from carbohydrates in phloem parenchyma cells (Hart 1981; Franceschi et al. 1998) via the shikimate/phenylpro panoid-acetate pathways. The proportion and type of syn thesized stilbenes (Brignolas et al. 1995, 1998) and other secondary compounds are controlled by the plant geno type. However, physiological conditions also affect the ability of trees to produce secondary compounds (Raffa and Berryman 1982b; Klepzig et al. 1995). Brignolas et al. (1998) proposed that the induced re sponse in spruce phloem after fungal inoculation involves two phases. The first of these is the tree's response to wounding, which is characterized mainly by both an in crease in the (-t-)catechin concentration and a slow de crease in the protein-precipitation capacity of wound ex tract (or tannins). Thereafter, a violent fungus-dependent reaction occurs, which is characterized by a strong de crease in the stilbene glycoside concentration, delayed appearance of the corresponding aglycones and a strong decrease in both the (+)-catechin concentration and the capacity to precipitate protein. This agrees with the non specific healing model proposed by Woodward and Pearce (1988b). In their model, stilbene glycosides, coupled with terpenes, provide the first line of defence af ter the bark surface is breached. The action of p-glycosi dases produced by the fungi close to the wound releases stilbene aglycones, thus inhibiting the growth of most pathogens. The stilbenes and terpenes provide the first bar rier to invasion, allowing time for the formation of necro phylactic periderm and isolating the necrotic tissues. Balanced availability of nutrients is assumed to affect the vigour of trees and their ability to resist attack by bark beetles and associated fungi. As hypersensitivity is an active metabolic process, the reaction capacity and speed must be affected by the physiological vigour of the tree (Berryman 1972). The number of bark beetle attacks on a tree may also modify the resistant response. Each attack severs resin ducts, causes the expenditure of ener gy in hypersensitive reaction, and reduces the area of functional phloem and sapwood (Berryman 1969, 1972). The growth/differentiation balance hypothesis predicts a physiological trade-off between plant growth and differ entiation processes. When environmental conditions are favourable, resources are generally allotted by giving priority to vegetative growth over secondary metabolism and storage (Herms and Mattson 1992; Lerdau et al. 1994). Even moderate shortages of nutrients or water slow growth processes considerably. Net photosynthesis, however, is not as sensitive to limitations in resources (Chapin 1980). Thus, when a moderate nutrient deficien cy imposes sink limitations upon growth, the growth is limited more by nutrient availability than by photosyn thate production. Therefore, carbohydrates accumulate in "excess" of growth requirements and are allocated to the production of C-based secondary metabolites, e.g. phe nolics and terpenoids (Gershenzon 1994). If growth is stimulated more strongly than photosynthesis, then con centrations of carbohydrates and C-based secondary compounds decline and the production of N-based com pounds increases. These predictions are known as the carbon/nutrient balance hypothesis (CNB). More specifi cally, concentrations of C-based secondary metabolites are positively correlated with the C/N ratio of the plant (Bryant et al. 1983; Herms and Mattson 1992). The aim of the present experiment was to test for changes in the C-based secondary compounds, such as stilbenes and terpenes, in Norway spruce after mechani cal wounding and artificial inoculation with Ceratocystis polonica Siem. Moreau, a pathogenic fungus transmitted by the spruce bark beetle Ips typographus L. (Solheim 1988; Krokene and Solheim 1996). In addition, the ex tent and formation of induced wound responses were studied in relation to increased nutrient concentration and growth caused by fertilization treatments. Results concerning carbohydrates and growth of experimental trees are reported in a companion paper. Materials and methods Experimental area and trees The study area was located in a Myrtillus-type spruce forest in Karttavuori, Vesijako. southern Finland. The stand was of natural origin and the selected trees were mature (age ca. 80 years) Nor 114 way spruce. Four plots of 0.5 ha were marked, and fertilization treatments were randomly applied to the plots. There was a 10- to 15-m-wide buffer zone between plots. Fertilizers N (173 kg ha -1 year 1; NH4-N 13.5%; NOr N 13.5%; Ca 3%; Mg 1%; S 3%; B 0.02%). P (41 kg ha -1 year l; total P 9%; water-soluble P 8%; Ca 20%; S 11%) and NPK (800 kg ha- 1 year l; NH4 -N 11%; NO rN 7%; total P 5%; water-soluble P 3.6%; K 10%; Ca 3%; Mg 0.5%; S 3%; B 0.02%; Se 0.001%) were spread by hand at the beginning of the growing season in May 1993. The control treatment was an unfertilized plot. Due to space limitations, we used a single plot for each fertilization treatment and one control. Thus, our experi mental design was pseudoreplicated (Hurlbert 1984); this is taken into account in the interpretation of results, where fertilization treatments are compared with the control, not with each other. From each plot 30 trees that were free of visible wounds, a total of 120 trees, were selected. Trees were chosen so that there were 97 dominants, 22 co-dominants and one intermediate tree. Two trees that were infected with root-rot and two that had been attacked by 1. typographus were omitted from the analysis. At the end of the experiment, the sample trees averaged 34±0.5 cm in diameter at breast height and 28±0.2 m in total height. Fungal inoculations In June 1994 a culture of blue-stain fungus, C. polonica, on 2% malt extract agar in Petri dishes was inoculated into ten trees per fertilization treatment. The fungal strain had been isolated from Norway spruce that had been infected naturally with I. typo graphus in Tuusula, southern Finland. The fungus was inoculated into the phloem with a cork-borer (diameter 5 mm), and the bark plug was re-inserted into the hole. Each tree received four evenly spaced inoculations made at the cardinal points of the compass 1.3 m above ground. For mechanical wounding, the trees were in jured with a cork borer in the same way as for inoculation, but without the fungus and malt agar. Ten trees per fertilization treat ment were wounded mechanically. In the middle of August, after a 2-month incubation period, the dead rhytidome was removed carefully with a sharp knife around each inoculation point so that the reaction lesion was visible. The vertical length of the lesion and the bark, sapwood and phloem thicknesses, were measured. Lesion length includes only the area outside the bark plug. Around each fungal inoculation site, four phloem samples were taken with a cork-borer (diameter 15 mm) to the level of the cambium: two from the distal ends of the visible reaction lesion (henceforth called "far" samples) and two immedi ately above and below the site of inoculation (henceforth called "near" samples). Two samples were taken from around the site of mechanical wounding, one above the wounding site and the other below the wounding site. In cases where there was visible lesion formation around the mechanical wounding (necrotic area more than 3 mm at the upper and lower side of the inoculation site, «=11), these necrotic tissues were included as part of the samples. Two samples per tree of phloem from intact trees were used as un wounded controls. All samples were immediately placed on dry ice and stored at -20° C until analysed. The fungus was not re-iso lated from inoculation sites, but all reaction lesions were visually distinct from each other. Extraction and analysis of stilbenes Phloem samples were ground in liquid N. Then 100 mg ground phloem powder was extracted with 2 ml of 80% v/v EtOH. As in ternal reference compounds, 50 |il (10,000 ppm) rhapontin and 50 fil (5,000 ppm) diethyl stilbestrol (Sigma) were added to the extracts. The extraction liquid was placed in an ultrasonic bath for 45 min, after which the liquid was extracted overnight. The extract was shaken on an orbital shaker and then centrifuged (4,332 g; 20 min -1 ); 500 (il of supernatant was evaporated to dryness in a stream of N. Dried samples were silylated for 2 h at room temper ature with 400 |jl silylating reagent prepared from 100 ml dry pyri din and 21 ml trimethylsilyl-imidazole (Fluka) and preserved in complete dryness over silica gel. The silylated samples were shak en and preserved at -20° C until analysed. Stilbenes were analysed with a gas chromatography (Hewlett- Packard 5890) mass spectrometry (Hewlett-Packard 5988 A) system with a HP-5 (Hewlett-Packard) capillary column (25 mxo.2 mmxo.3 pm). Split injection was used. The carrier gas was helium; the column head pressure was 100 kPa and the split flow was 25 ml min -1 . The initial temperature of the oven was 110° C, increasing at 10° C min -1 to a final temperature of 300° C. Stilbene glycosides were quantified by the response of rhapontin. and stilbene aglycones were quantified by using the response fac tor of diethyl stilbestrol and resveratrol. Identifications of com pounds were confirmed by mass spectra. Results are expressed as micrograms of stilbenes per milligram of fresh phloem. Extraction and analysis of mono- and sesquiterpenes One hundred milligrams of phloem powder obtained by grinding the phloem in liquid N was extracted with 2 ml hexane. The hex ane contained 200 ppm each of the following three internal stan dards: ClO H 21C1, C l 4 H29C1 and C lBH37C1 (Fluka). The samples were extracted in an ultrasonic bath for 30 min and left in test tubes overnight. The hexane extracts were shaken and centrifuged (4,332 g; 20 min -1 ); and the extracted samples were stored at -20° C until analysed. Mono- and sesquiterpenes were analysed with the previously de scribed gas chromatography system and the mass spectrometry system using an NB-351 (HNU-Nordion, Helsinki, Finland) capil lary column (25 mxo.2 mmxo.3 (jm). The programme started at 60° C (splitless injection, 0.5 min), rising by 5°C min -1 to a final temperature of 230° C. The analysis time was 40 min, the carrier gas was helium, and the split flow rate was 20 ml min- 1 . Monoterpenes were identified and quantified using the retention and mass spectral data of authentic model compounds. Sesquiterpenes were identified according to the method of Pohjola (1993) and quantified according to the response factor of caryophyllene. Model compounds were ob tained from the Pharmacognosy Division, Department of Pharmacy. University of Helsinki, Finland. The results are expressed as micro grams of terpenes per milligram of fresh phloem. Statistical analysis Univariate ANOVA was used to examine overall differences be tween inoculation methods and between inoculation directions. The means of chemical compounds calculated from opposite sides of the inoculation site were used in the analysis. For mechanically wounded trees and unwounded controls, the means of two sam pling sites were used. Detection frequency is the mean (%) of all Fig. 1 Within fertilization treatments, lesion lengths in fungus-in oculated and mechanically wounded trees differed significantly from each other at the 5% level; these differences are indicated by different letters 115 Fig. 2 Stilbene concentrations (pg mg- 1 fresh phloem tissue. mean±SD) in Picea abies phloem inoculated with Ceratocystis polonica. mechan ically wounded and unwounded phloem (see Materials and methods regarding sampling), grouped according to fertiliza tion. Values indicated with an asterisk at the end of a bar differ significantly from corresponding unfertilized phloem at the 5% level. Within fertilization treatments, bars with different letters differ significantly from each other at the 5% level samples in which the substance in question was detected. The Dunnett-test (Myers and Well 1995) was used to examine differ ences between fertilized plots and the unfertilized control and within fertilization treatments. Homogeneity of the variances was tested with Levene's test (Snedecor and Cochran 1980). If the as sumption of homogeneity was not shown to be true (Jeffers 1960), nonparametric Dunnett-C tests were used. The values presented in the tables and figures are untransformed means, and if not men tioned otherwise, the significance level is 5%. Results Lesion length Fungal inoculation produced a necrotic lesion around the inoculation site, the vertical length of which was 7.1 ±O.B cm (mean of all fertilizations±SE; n=39). In me- 116 Fig. 3 Detection frequency of stilbene compounds (%) in P. abies phloem after inoculation with C. polonica, in mechanically wounded and unwounded phloem (see Materials and methods for sampling information). Each fertilization treatment and the control were combined. Values indicated with an asterisk at the end of a bar differ significantly from unwounded phloem at the 5% level. For the same compound, bars labelled with a different letter differ significantly from each other at the 5% level chanically wounded trees the vertical length of the lesion around the inoculation sites was 0.2±0.3 cm («=3B), and horizontal growth was <0.3 cm. The difference in lesion length between fungal inoculation and mechanical wounding was highly significant (Fig. 1). The lesions appeared to be longest in P-fertilized trees and shortest in N-fertilized trees. After fungal inoculation the differ ence between fertilization treatments was nearly signifi cant (F= 2.47, df= 3, P=0.064). Stilbenes Concentrations of the stilbene glycosides (piceid, astrin gin, isorhapontin) and their aglycones (resveratrol. as tringenin and isohapontigenin) in relation to inoculation and fertilization are shown in Fig. 2. Near the inocula tion site, concentrations of individual glycosides and to tal glycosides were lower than in unwounded phloem re gardless of treatment (Fig. 2). The total stilbene glyco side concentration showed a negative and significant correlation with the total terpene concentration near the inoculation site (r=-0.493, «=33, P<0.01). and at the outer border of the lesion «=39, P<0.01). The Fig. 4 Composition of the major monoterpenes and the sesquiterpene germacrene (%, mean±SD) grouped according to fertilization. Labels as in Fig. 2 117 Table 1 Quantitative changes in the terpene fraction (µg mg _I fresh phloem) of Picea abies phloem inoculated with Ceratocystis polonica, mechanically wounded and unwounded phloem. All fertilization treatments were pooled. Samples were taken around the fungal inoculation sites. Near Mean of two samples from the areas immedi ately above and below the site of inoculation, Far mean of two samples from the distal ends of the visible reaction lesion *P0.0 1) total stilbene glycoside concentration did not correlate with the lesion length near the inoculation site (r=-0.338, «=33, P>0.05) or at the outer border of the reaction lesion (MXO66, «=39, P>0.05). On the other hand, the stilbene aglycone isohaponti genin and total aglycones were detected in higher con centrations near the fungal inoculation site than at the other sampling points (Fig. 2). Total aglycone concentra tion did not correlate with lesion length (data not shown). It correlated positively with the total terpene concentration at the far end of the lesion n=39, P<0.05) but not with that near the inoculation site (r— 0.231, n= 3B, P>0.05). In N-fertilized trees, near the inoculation site the concentration of resveratrol was sig nificantly lower than in the phloem of unfertilized trees. In addition, the concentration of total stilbene aglycones decreased significantly after N fertilization compared to values for the corresponding unfertilized phloem (Fig. 2). The detection frequencies of stilbene glycosides near the inoculation site were significantly lower than in me chanically wounded or unwounded phloem (Fig. 3). Cor respondingly, near the inoculation site the frequencies of stilbene aglycones were constant. When the detection frequencies of glycosides and aglycones in different fer tilization treatments were compared to those of the cor responding unfertilized control, the differences were usually not significant. Terpenes The main monoterpenes both in the reaction lesion and in the constitutive phloem tissue were a-pinene, |3- pinene, (3-phellandrene, A-3-carene and a-phellandrene, which made up 82% of the total terpene fraction (Fig. 4). Near Far Wounded Unwounded Tricyclene 0.172*** 0.022*** 0.004 0.004 a-Pinene 26.469*** 3.407*** 0.326 0.351 Fenchene 0.012*** 0.009** 0.003 0.003 Camphene 0.560*** 0.080*** 0.007 0.008 (3-Pinene 27.878*** 2 0.278 0.278 Sabinene 1 228*** 0 ]4^*** 0.012 0.012 A-3-Carene 7^812*** 0.854*** 0.038 0.049 a-Phellandrene 5.550*** 0.646*** 0.054 0.063 Limonene 3.531*** 0 437*** 0.028 0.029 p-Phellandrene 13.121*** 1.376*** 0.142 0.163 y-Terpinene 0.128*** 0.024*** 0.011 0.012 Terpinolene 1.578*** 0.201*** 0.012 0.015 Bornylacetate 0.414*** 0.065*** 0.010 0.010 Total monoterpenes 88.446*** 10.149*** 0.903 0.975 p-Caryophyllene 0.200*** 0.031*** 0.007 0.009 a-Humulene 0.144*** 0.022** 0.004 0.004 y-Muurolene 0.422*** 0.048** 0.008 0.010 Germacrene 2.898*** 0.410*** 0.072 0.068 B icyclogermacrene 0.354*** 0.027* 0.005 0.007 A-Cadinene 1.181*** 0.138*** 0.026 0.037 Total sesquiterpenes 5.193*** 0.648* 0.110 0.119 Total terpenes 93.639*** 10.797*** 1.019 1.094 118 Fig. 6 Total terpene concentrations ( µg mg -1 fresh phloem. mean+SD) in fungus-inoculated, mechanically wounded and unwounded phloem, grouped according to fertilization. Labels as in Fig. 2 Minor amounts of the monoterpenes limonene, terpinol ene and sabinene occurred in all trees. The sesquiterpene fraction was found to consist mainly of germacrene and A-cadinene. Each of the other mono- and sesquiterpenes represented lOO times higher than the constitutive values (Table 1). After mechanical wounding, terpene quantity remained stable compared to that in unwounded phloem. Near the inoculation site, the total terpene concentra tion of N-fertilized trees was significantly lower than the values for unfertilized control trees: according to the Dunnett-test, the mean difference was -0.59 and />=o.ooB. The total terpene concentration was about half that found in the control (Fig. 6). Other sampling sites or fertilization treatments did not differ significantly from the corresponding unfertilized control. On the other hand, in some cases the percentage composition of the terpene fraction differed significantly after fertilization treatment (Fig. 4: Table 2). Of the total terpene fraction, the proportion of a-pinene increased after N and P fertil izations, and the proportion of germacrene increased af ter NPK fertilization. The proportions of (3-pinene and (3- phellandrene decreased after N and P fertilization. N P NPK Control Tricyclene 0.34 0.36 0.18 0.27 a-Pinene 34. IS*** 34 32*** 29.48 28.29 Fenchene 0.13 0.30*** 0 72*** 0.10 Camphene 0.80*** 0.98*** 0.64 0.59 (3-Pinene 24.71*** 24.04*** 28.40 27.87 Sabinene 1.12 1.18 1.56 1.26 A-3-Carene 5.38 5.90 6.47 6.04 ct-Phellandrene 5.54 5.52 4.98 5.42 Limonene 4.31 3.54 2 72*** 3.69 {3-Phellandrene 12.46*** 11 3S*** 14.36 15.13 y-Terpinene 1.26* 1.23* 0.66* 0.96 Terpinolene 1.53 1.75 1.57 1.56 Bornylacetate 1 1 18*** 0.68 0.70 Monoterpenes 92.94 91.31 88.82* 91.76 p-Caryophyllene Q 93*** 0.59 0.76 0.59 a-Humulene 0.23 0.27 0.40 0.27 y-Muurolene 0.63 0.510* 0.96 0.86 Gennacrene 2.980* 4.75 8.25*** 3.78 Bicyclogermacrene 0.35 0.37 0.68 0.57 A-Cadinene 2.23 2.60 2.27 2.28 Sesquiterpenes 7.06 8.69 12.07* 8.24 119 Discussion A mere change in the concentration of allelochemical com pounds is not sufficient evidence of changes in resistance to bark beetle attack. Such a change is a general response to stress, wounding or reduced availability of nutrients (Wright et al. 1979; Delorme and Lieutier 1990). Neverthe less, aseptic mechanical wounding generally yields a mod est response (Johansson and Stenlid 1985; Lieutier et al. 1989; Raffa and Smalley 1995). In this experiment me chanical wounding did not change the amount of stilbenes and terpenes compared to unwounded phloem. In most cases the bark plugs healed and there were no signs of ac cumulation of secondary compounds. Mechanical wound ing combined with inoculation of sterile malt agar media or autoclaved fungus might have given more variable re sults, as reported e.g. hy Raffa and Smalley (1995). A low-density method of inoculation was used since the aim was to study phloem chemistry after nutrient levels changed. With this method the tree's defence response is the most efficient mobilizing of all resources against a cor responding attack (Raffa and Berryman 1983). After low density inoculation, however, lesion length alone is not a good measure of the resistance of the tree or of fungal viru lence. Here lesion length was about the same as in a low density study in Norway spruce after a corresponding incu bation time (Krokene and Solheim 1997). In general, small lesions indicate either the weakness of the aggressor or the resistance of the host. The efficient response of a resistant tree might produce a small lesion, while a longer lesion might imply either a physiologically weaker host tree or a more virulent fungus (Krokene and Solheim 1999). The positive relationship between lesion length and total ter pene concentration near the inoculation site supports the latter idea. For determining the virulence of fungi, the depth of desiccated sapwood would be a more useful vari able than lesion length (Krokene and Solheim 1997, 1999). In stilbenes. the qualitative changes after inoculation with fungus were more pronounced than the quantitative responses. Concentrations of all stilbene glycosides de creased considerably after inoculation, while the concen tration of the aglycone isorhapontigenin increased signif icantly (Fig. 2). These results agree with those of Wood ward and Pearce (1988 a), Lindberg et al. (1992) and Brignolas et al. (1995), all of whom observed similar changes a few days after fungal infection. It is known that the fi-glycosidase enzymes produced by pathogens, rather than the constitutive host enzymes, are able to me tabolize stilbene glycosides to the corresponding agly cones (Johansson and Stenlid 1985; Woodward and Pearce 1988 a; Woodward 1992). In this study the increase in (3-glycosidase activities in fungus-inoculated phloem might also have been instrumental in the release of agly cones from glycosides (Fig. 2); the longer the reaction lesions were, the less glycosides were present, suggest ing that they were actually metabolized. Since (3-glycosi dase activities were not measured here, we can only speculate about the accumulation of stilbene aglycones after the activities of the fi-glycosidase enzymes. Nevertheless, aglycones cause higher levels of anti fungal activity in wounded tissues than glycosides do. For instance, in vitro the antifungal activity of the agly cone isorhapontigenin against the decay fungus Phaeolus schweinitzii (Fr.) Pat. was 6 times greater than that of the corresponding glycoside (Woodward and Pearce 1988 a). Stilbene aglycones were not observed in previous studies on fungus-inoculated or unwounded phloem of Norway spruce, despite increased stilbene synthase activity (Brignolas et al. 1995, 1998). This may be due to differ ent methods of sample preparation and differences in as say methods, as in a corresponding case in Sitka spruce [Picea sitchensis (Bong.) Carr.] (Toscano Underwood and Pearce 1991 a). In the present study the accumulation of aglycones was an actual response to fungal inocula tion (as shown in Figs. 1. 2) and not an artefact caused by sample preparation. In healthy spruce, the concentrations of stilbene glyco sides are known to differ between trees but may also vary according to age, season, provenance and site (Solhaug 1990; Toscano Underwood and Pearce 1991 a. 1991b; Lindberg et al. 1992). With regard to the stilbene glycoside fraction, our results agree with those of Solhaug (1990) and of Lindberg et al. (1992), who concluded that the bark of Norway spruce contains more isorhapontin than astringin. The concentration of stilbene increases towards the outer bark, but in Sitka spruce the procedure for bark plug sampling still provides an adequate measure of the stilbenes (Toscano Underwood and Pearce 1991 a). The terpene responses of Norway spruce to fungal inoc ulation were strong and quantitative rather than qualitative. The major terpenes in the reaction lesion did not differ from the constitutive compounds. The enantiomeric com position of monoterpenes, which differs in the different tis sues of Norway spruce, and the volatile compounds emit ted by the host tree are important in the chemical commu nication system of the spruce bark beetles and host selec tion (Ivarsson 1995; Persson et al. 1996). The host tree monoterpene, (-)-a-pinene, acts as a precursor for synthe sis of the aggregation pheromone, ci.v-verbenol (Ivarsson 1995). On the other hand, the fungistatic effect of terpenes in conifers seems to be quantitative rather than qualitative; the response of a tree to fungal inoculation is characterized by considerable increases in the concentrations of all the phloem terpenes (Russell and Berryman 1976; Raffa and Berryman 1982 a, 1982b; Lieutier et al. 1989; Delorme and Lieutier 1990; Raffa and Smalley 1995). On the basis of our sampling procedure, it is difficult to determine whether terpenoids accumulated here in large quantities after fungal inoculation originated from translocation or from on-site de novo biosynthesis (Lewinsohn et al. 1993). In Norway spruce, parenchyma cells near the fungal inoculation site differentiate into resin ducts (Franceshi et al. 2000; Nagy et al. 2000), which makes long-range translocation of ter penes less obvious. In addition, changes in terpene compo sition after inoculation support the idea that the neosynthe sis occurred around the inoculation site. Sesquiterpenes and resin acids (Björkman et al. 1991), which occur in small quantities, may not be representa 120 tive indicators for testing the CNB hypothesis. Further more, resins that require not only the synthesis but also the production of specialized storage structures or com partments may begin to decline sooner than compounds, such as phenolics, that can be sequestered simply by the surrounding cell walls or vacuoles (Lerdau et al. 1994). Previous investigations have also indicated the opposite; C may not be the limiting factor for resin accumulation, but rather physiological and anatomical constraints (Björkman et al. 1991: Muzika 1993: Kytö et al. 1999). Stilbene compounds that are present in low concentra tions after dynamic reactions are difficult to detect, and therefore quantitative changes following changes in re source availability are difficult to observe and document reliably. Trade-off between growth and defence seems to be feasible only when there is actual biosynthetic compe tition for the same precursor, as in the case of the phe nylpropanoids. The CNB hypothesis is not applicable in all cases because not all C-based secondary compounds seem to respond to fertilization in the same way (Lawler et al. 1997; Koricheva et al. 1998, and references there in; Keinänen et al. 1999). For example, total phenolics, which was previously a widely used variable, may in clude ecologically interesting but opposite responses of secondary compounds, since tannins and non-tannin phe nolics have not been separated and identified. Here, for stilbene aglycones and stilbene glycosides, the trends de tected in response to fertilization were opposite. In this study, the accumulation of stilbene aglycones and terpenes was significantly reduced near the inocula tion site only after N fertilization, but not by P or NPK fertilizations. Thus, N fertilization might affect the ability of spruce to defend itself against aggression by C. polo nica. Responses after N fertilization were consistent with the CNB hypothesis. Even so, according to a wide meta analysis (Koricheva et al. 1998, and references therein), the ability of the CNB hypothesis to predict changes in plant terpenoids in response to experimental manipula tions seems weak. That analysis, however, contained many studies in which constitutively occurring secondary compounds were included or changes outside the active growing phase were analysed. The predictions of plant de fence theories might best be fulfilled during those times of the year when growth and defence are competing for re sources, so that an inverse relationship between N avail ability and allocation to mobile C-based defences might be expected (Lerdau et al. 1995). Minor changes in growth and defence caused by P or NPK fertilization or by other stress treatments can easily be hidden within natural variation, especially if these changes occur outside the ac tive growing season. Since here sampling was done only once, the lack of reference samples at the beginning of the experiment is a weakness and should be taken into ac count when results are interpreted. The shift from stilbene glycosides to aglycones in the lesion (Fig. 2) indicates, however, the dynamics involved in the wound reaction. Here, NPK fertilization increased the total amount of terpenes but decreased the total concentration of stilbene aglycones near the inoculation site. The N-fertilizer in our NPK regime was moderate (144 kg ha-1 year 1 ), since the goal was to imitate actual fertilization in practical for estry and not to create artificial responses with overdoses. These results agree with a those of Muzika et al. (1989) and Lerdau et al. (1995). who found that treatment with a high level of N often produced less terpenes, whereas an intermediate level of N apparently had little influence on terpene production. Most likely, our NPK fertilization was too modest to cause proper responses in allocation of secondary metabolites. It might be that NPK fertilization has a more balanced effect on secondary metabolite pro duction than pure N fertilization does. In boreal conifer ous forests, N is commonly a growth-limiting factor, whereas P seldom reduces growth; when it does, this typ ically occurs on peat lands. Thus it is not surprising that P fertilization had no significant effect on secondary metab olite concentrations in this study or in the meta-analysis (Koricheva et al. 1998). This may also be the reason why we found a significant decrease in total terpenes only in the induced response of N-fertilized trees. In summary, N fertilization resulted in lower concen trations of total terpenes and stilbene aglycones, but P fer tilization led to only a minor decrease in the concentration of total terpenes. The CNB hypothesis predicts changes in these groups of secondary metabolites in response to the above-mentioned nutrient changes. On the other hand, the concentrations of stilbene glycosides did not correspond to the predictions of the CNB hypothesis; their role as in duced defensive compounds may be less important. How ever, the effect of fertilization was seen only at the point where the induced wound response was the most exten sive, near the inoculation site. Results indicate that further studies of plant-herbivore interactions are needed, in par ticular, studies using a model in which allocation varies in response to phenological and herbivory demands. Acknowledgements The authors thank Mr Pekka Helminen and Mr Jukka Lehtonen for field work and Mr Pauli Karppinen for conducting most of the chemical analyses. Drs Markku Keinänen, Maarit Kytö, William Mattson and Vladimir Ossipov provided helpful comments on the manuscript, and Joann von Weissenberg corrected the English. This study was supported by and conducted at the Finnish Forest Research Institute, Vantaa Research Centre. The Graduate School of Forest Sciences. Ministry of Education and Faculty of Forestry, University of Joensuu, provided grants to H. V. The Pharmacognosy Division. Department of Pharmacy, University of Helsinki, provided model compounds of terpenes. The above-mentioned individuals and institutions are gratefully acknowledged. References Berryman AA (1969) Responses of Abies grandis to attack by Scolxtus ventralis (Coleoptera: Scolytidae). Can Entomol 101: 1033-1041 Berryman AA (1972) Resistance of conifers to invasion by bark beetle-fungus associations. Bioscience 22:598-602 Björkman C, Larsson S. Gref R (1991) Effects of nitrogen fertili zation on pine needle chemistry and sawfly performance. Oecologia 86:202-209 Bridges JR (1987) Effects of terpenoid compounds on growth of symbiotic fungi associated with the southern pine beetle. Phytopathology 77:83-85 121 Brignolas F, Lacroix B. Lieutier F. Sauvard D. Drouet A, Claudot A-C, Yart A. Berryman AA. Christiansen E (1995) Induced re sponses in phenolic metabolism in two Norway spruce clones after wounding and inoculations with Ophiostoma polonicum, a bark beetle-associated fungus. Plant Physiol 109:821-827 Brignolas F, Lieutier F. Sauvard D. Christiansen E, Berryman AA (1998) Phenolic predictors for Norway spruce resistance to the bark beetle Ips typographic (Coleoptera: Scolytidae) and an associated fungus, Ceratocxstis polonica. Can J For Res 28: 720-728 Bryant JP, Chapin FS 111. Klein DR (1983) Carbon/nutrient bal ance of boreal plants in relation to vertebrate herbivory. Oikos 40:357-368 Chapin FS 111 (1980) The mineral nutrition of wild plants. Annu Rev Ecol Syst 11:233-260 Christiansen E (1985) Ips/Ceratocystis-infection of Norway spruce: what is a deadly dosage? Z Angew Entomol 99:6-11 Cobb FW Jr. Krstic M. Zavarin E. Barber HW Jr (1968) Inhibitory effects of volatile oleoresin components on Fomes annosus and four Ceratocxstis species. Phytopathology 58:1327-1335 Delorme L. Lieutier F (1990) Monoterpene composition of the preformed and induced resins of Scots pine, and their effect on bark beetles and associated fungi. Eur J For Pathol 20: 304-316 Franceschi R. Krekling T. Berryman AA. Christiansen E (1998) Specialized phloem parenchyma cells in Norway spruce (Pinaceae) bark are an important site of defense reactions. Am J Bot 85:601-615 Franceschi VR. Krokene P. Krekling T. Christiansen E (2000) Phloem parenchyma cells are involved in local and distant de fense responses to fungal inoculation or bark-beetle attack in Norway spruce (Pinaceae). Am J Bot 87:314-326 Gershenzon J (1994) Metabolic costs of terpenoid accumulation in higher plants. J Chem Ecol 20:1281-1328 Hart JH (1981) Role of phytostilbenes in decay and disease resis tance. Annu Rev Phytopathol 19:437-458 Herms DA, Mattson WJ (1992) The dilemma of plants: to grow or defend. Quart Rev Biol 67:283-335 Hurlbert S H (1984) Pseudoreplication and the design of ecological field experiments. Ecol Monogr 54:187-211 Ivarsson P (1995) The pheromone systems of the spruce bark bee tles, Ips duplicatus and I. typographus: biosynthesis and regu lation. Dissertation. Faculty of Natural Sciences, Göteborg University, Göteborg Jeffers JNR (1960) Experimental design and analysis in forest re search. Almqvist and Wiksell, Stockholm Johansson M. Stenlid J (1985) Infection of roots of Norway spruce (Picea abies) by Heterobasidion annoswn. 1. Initial reactions in sapwood by wounding and infection. Eur J For Pathol 15: 32-45 Keinänen M, Julkunen-Tiitto R. Mutikainen P, Walls M. Ovaska J, Vapaavuori E (1999) Trade-offs in phenolic metabolism of sil ver birch: effects of fertilization, defoliation, and genotype. Ecology 80:1970-1986 Klepzig KD. Kruger EL, Smalley EB. Raffa KF (1995) Effects of biotic and abiotic stress on induced accumulation of terpenes and phenolics in red pines inoculated with bark beetle-vec tored fungus. J Chem Ecol 21:601-626 Koricheva J. Larsson S. Haukioja E. Keinänen M (1998) Regula tion of woody plant secondary metabolism by resource avail ability: hypothesis testing by means of meta-analysis. Oikos 83:212-226 Krokene P. Solheim H (1996) Fungal associates of five bark beetle species colonizing Norway spruce. Can J For Res 26:2115-2122 Krokene P. Solheim H (1997) Growth of four bark-beetle-associat ed blue-stain fungi in relation to the induced wound response in Norway spruce. Can J Bot 75:618-625 Krokene P. Solheim H (1999) What do low-density inoculations with fungus tell us about fungal virulence and tree resistance? In: Lieutier F. Mattson WJ. Wagner MR (eds) Physiology and genetics of tree-phytophage interactions. INRA Editions. Paris, pp 353-362 Kytö M, Niemelä P. Annila E, Varama M (1999) Effects of forest fertilization on the radial growth and resin exudation of insect defoliated Scots pines. J Appl Ecol 36:763-769 Lawler IR, Foley WJ, Woodrow lE. Cork SJ (1997) The effects of elevated C0 2 atmospheres on the nutritional quality of Euca lyptus foliage and its interaction with soil nutrient and light availability. Oecologia 109:59-68 Lerdau M, Litvak M. Monson R (1994) Plant chemical defense: monoterpenes and the growth-differentiation balance hypothe sis. Trends Ecol Evol 9:58-61 Lerdau M. Matson P, Fall R. Monson R (1995) Ecological controls over monoterpene emissions from Douglas-fir (Pseudotsuga menziesii). Ecology 76:2640-2647 Lewinsohn E, Gijzen M, Muzika RM. Barton K, Croteau R (1993) Oleoresinosis in grand fir (Abies grandis) saplings and mature trees. Plant Physiol 101:1021-1028 Lieutier F, Cheniclet C, Garcia J (1989) Comparison of the de fense reactions of Pinus pinaster and Pinus sylvestris to at tacks by two bark beetles (Coleoptera: Scolytidae) and their associated fungi. Environ Entomol 18:228-234 Lindberg M. Lundgren L. Gref R. Johansson M (1992) Stilbenes and resin acids in relation to the penetration of Heterobasidion annosum through the bark of Picea abies. Eur J For Pathol 22:95-106 Muzika RM (1993) Terpenes and phenolics in response to nitro gen fertilization: a test of the carbon/nutrient balance hypothe sis. Chemoecology 4:3-7 Muzika RM. Pregitzer KS. Hanover JW (1989) Changes in terpene production following nitrogen fertilization of grand fir {Abies grandis (Dougl.) Lindl.) seedlings. Oecologia 80:485-489 Myers JL, Well AD (1995) Research design and statistical analy sis. Erlbaum. Hillsdale, N.Y. Nagy NE. Franceschi VR, Solheim H. Krekling T, Christiansen E (2000) Wound-induced traumatic resin duct development in stems of Norway spruce (Pinaceae): anatomy and cytochemi cal traits. Am J Bot 87:302-313 Nicholson RL. Hammerschmidt R (1992) Phenolic compounds and their role in disease resistance. Annu Rev Phytopathol 30: 369-389 Persson M. Sjödin K, Borg-Karlson A-K. Norin T. Ekberg I (1996) Relative amounts and enantiomeric compositions of monoterpene hydrocarbons in xylem and needles of Picea ab ies. Phytochemistry 42:1289-1297 Pohjola J (1993) A Headspace gas chromatographic study on the variation of needle volatile terpenes in Scots pine (Pinus syl vestris L.). Dissertation. Pharmacognosy Division. Department of Pharmacy, University of Helsinki. Helsinki Raffa KF, Berryman AA (1982 a) Accumulation of monoterpenes and associated volatiles following inoculation of grand fir with a fungus transmitted by the fir engraver, Scolytus ventralis (Coleoptera: Scolytidae). Can Entomol 114:797-810 Raffa KF, Berryman AA (1982b) Physiological differences be tween lodgepole pines resistant and susceptible to the moun tain pine beetle and associated microorganisms. Environ Ento mol 11:486-492 Raffa KF. Berryman AA (1983) Physiological aspects of lodge pole pine wound responses to a fungal symbiont of the moun tain pine beetle. Denroctonus ponderosae (Coleoptera: Scolyt idae). Can Entomol 115:723-734 Raffa KF, Smalley EB (1995) Interaction of pre-attack and in duced monoterpene concentrations in host conifer defense against bark beetle-fungal complexes. Oecologia 102:285-295 Raffa KF. Berryman AA, Simasko J. Teal W. Wong BL (1985) Ef fects of grand fir monoterpenes on the fir engraver. Scolytus ventralis (Coleoptera: Scolytidae). and its symbiotic fungus. Environ Entomol 14:552-556 Reid RW. Whitney HS. Watson JA (1967) Reactions of lodgepole pine to attack by Dendroctonus ponderosae Hopkins and blue stain fungi. Can J Bot 45:1115-1126 Russell CE. Berryman A A (1976) Host resistance to the fir en graver beetle. 1. Monoterpene composition of Abies grandis pitch blisters and fungus-infected wounds. Can J Bot 54:14-18 122 Schultz TP, Boldin WD, Fisher TH, Nicholas DD, McMurtrey KD. Pobanz K (1992) Structure-fungicidal properties of some 3- and 4-hydroxylated stilbenes and bibenzyl analogues. Phy tochemistry 31:3801—3806 Snedecor GW, Cochran WG (1980) Statistical methods. lowa State University Press, Ames, lowa Solhaug KA (1990) Stilbene glucosides in bark and needles from Picea species. Scand J For Res 5:59-67 Solheim H (1988) Pathogenicity of some lps typographus-associ ated blue-stain fungi to Norway spruce. Commun Norw For Res Inst 40:1-11 Toscano Underwood CD, Pearce RB (1991 a) Astringin and isorha pontin distribution in Sitka spruce trees. Phytochemistry 30: 2183-2189 Toscano Underwood CD, Pearce RB (1991b) Variation in the lev els of the antifungal stilbene glucosides astringin and isorha pontin in the bark of Sitka spruce [Picea sitchensis (Bong.) Carr.]. Eur J For Pathol 21:279-289 Wingfield MJ, Seifert KA, Webber JF (1993) Ceratocystis and Ophiostoma. Taxonomy, ecology and pathogenicity. APS, St. Paul. Minn. Woodward S (1992) Responses of gymnosperm bark tissues to fungal infections. In: Blanchette RA, Biggs AR (eds) Defense mechanisms of woody plants against fungi. Springer, Berlin Heidelberg New York, pp 62-75 Woodward S, Pearce RB (1988 a) The role of stilbenes in resis tance of Sitka spruce [Picea sitchensis (Bong.) Carr.] to entry of fungal pathogens. Physiol Mol Plant Pathol 33:127-149 Woodward S, Pearce RB (1988b) Wound-associated responses in Sitka spruce root bark challenged with Phaeolus schweinitzii. Physiol Mol Plant Pathol 33:151-162 Wright LC, Berryman AA, Gurusiddaiah S (1979) Host resistance to the fir engraver beetle, Scolytus ventralis (Coleoptera: Scolytidae). 4. Effect of defoliation on wound monoterpene and inner bark carbohydrate concentrations. Can Entomol 111: 1255-1262 IV Trees (2ool) 15:327-334 DOI 10. 1007/5004680100117 Heli Viiri • Pekka Niemelä • Veikko Kitunen Erkki Annila Soluble carbohydrates, radial growth and vigour of fertilized Norway spruce after inoculation with blue-stain fungus, Ceratocystis polonica Received: 10 March 2001 / Accepted: 5 July 2001 / Published online: 15 August 2001 © Springer-Verlag 2001 Abstract The aim of this study was to determine whether fertilization and the consequent increase in growth reduce the allocation of soluble carbohydrates in response to an induced wound. Norway spruce trees fer tilized with N, P or NPK were artificially infected with Ceratocystis polonica, a blue-stain fungus associated with the spruce bark beetle Ips typographus. N and NPK fertilization treatments increased radial growth of the stem and the vigour indices. The concentration of total soluble carbohydrates in the outer border of the lesion was significantly decreased in P-fertilized trees com pared to corresponding unfertilized trees. However, changes in the soluble carbohydrate concentration caused by fungal inoculation were more pronounced than changes caused by fertilization. The main soluble carbohydrate was sucrose, and after fungal inoculation its concentration decreased considerably near the site of inoculation. Thus, near the site of fungal inoculation the concentration of total soluble carbohydrates also de creased significantly compared to corresponding values in unwounded phloem. Despite the fact that in all fertil ized trees the radial growth of the stem increased, the on ly indication that enhanced growth might reduce the lev el of resistance was the modest positive correlation be tween lesion length and radial growth of the stem. Keywords Soluble sugars • Sugar alcohols Carbon/nutrient balance hypothesis • Resource allocation Vigour index H. Viiri (CS) • P. Niemelä Faculty of Forestry, University of Joensuu, P.O. Box 111. 80101 Joensuu, Finland V. Kitunen • E. Annila Finnish Forest Research Institute. Vantaa Research Centre. P.O. Box 18, 01301 Vantaa. Finland Present address: H. Viiri. Finnish Forest Research Institute. Suonenjoki Research Station. Juntintie 40. 77600 Suonenjoki. Finland e-mail: heli.viiri@metla.fi Tel.: +358-17-5138219, Fax: +358-17-513068 Introduction Bark beetle-fungus attack causes accumulation of sec ondary metabolites in the phloem tissues of conifers. In response to attack, the parenchyma cells in the phloem produce large quantities of C-based secondary com pounds (CBSCs) that, compared to primary metabolites, are energetically costly to synthesize (Lorio 1986; Gershenzon 1994 a, 1994b). The attack leads to a de crease in the amount of soluble carbohydrates in the in ner bark and sapwood (Shrimpton 1973; Wright et al. 1979; Christiansen and Ericsson 1986; Cook and Hain 1987). Furthermore, primary fungal invaders of the phloem utilize sugars as nutrients. Growth of the fungus is first delayed by the removal of essential nutrients, like glucose and fructose, and later by secondary metabolites, which reduce the availability of water and oxygen in the lesion (Wong and Berry man 1977; Raffa and Berry man 1982). Finally, traumatic formation of resin ducts and wound periderm seal the lesion (Reid et al. 1967; Franceschi et al. 2000; Nagy et al. 2000). Carbohydrates are especially important because they are direct products of photosynthesis and are therefore the primary energy-storage compounds and the basic or ganic substances from which most other organic com pounds found in plants are synthesized (Kozlowski and Pallardy 1997). The ability of trees to withstand at tack by bark beetles and their associated fungi is thought to be linked to the amount of carbohydrates that can be utilized for defence reactions. Thus, the ability of a tree to resist attacks by bark beetles is directly related to its physiological status and vigour (Mulock and Christiansen 1986; Christiansen et al. 1987; Paine et al. 1997). Therefore, any environmental factor that restricts the size of the canopy or its photosynthetic efficiency can weaken the resistance of the tree. Production of car bohydrates in plant tissues, if this exceeds the amount required for growth, favours the synthesis of C-based secondary metabolites, e.g. phenolics and terpenes (Bryant et al. 1983; Herms and Mattson 1992). The quantitative ability of a tree to respond to infection in 328 creases until the tree reaches maturity, and then declines with age (Raffa and Berryman 1982). In plants, in gen eral, the biosynthesis of terpenoids proceeds via two in dependent pathways: first, sesquiterpenes, triterpenes and sterols are biosynthesized via the cytosolic, classic acetate/mevalonate pathway or via the alternative, non mevalonate I -deoxy-D-xylulose-5-phosphate pathway for the biosynthesis of plastidic terpenoids, such as mono- and diterpenes (Lichtenthaler 1999). Both path ways form the active C 5-unit isopentenyl diphosphate as the precursor from which all other terpenoids are formed via head-to-tail addition (Lichtenthaler 1999). Phenolics are derived from carbohydrates via the shiki mate pathway through which about one fifth of all the C fixed by plants flows (Matsuki 1996). According to the C/nutrient balance (CNB) hypothesis, fertilization with a growth-limiting nutrient or release of some other nu trient will stimulate growth more strongly than photo synthesis (Chapin 1980), so that concentrations of car bohydrates and C-based secondary metabolites decline (Bryant et al. 1983; Herms and Mattson 1992). The main aim of the present study was to investigate interactions between soluble carbohydrates, defence mechanisms and radial growth of the stem of Norway spruce. The specific objectives were: 1. To examine the effect of artificial inoculation with the blue-stain fungus Ceratocystis polonica Siem. Moreau, transmitted by the spruce bark beetle, Ips typographus L., on the composition of soluble carbo hydrates in the phloem. 2. To determine whether fertilization affects the relative amounts of soluble carbohydrates. 3. To examine the interactions between soluble carbohy drates, terpenes, stilbenes, radial growth of the stem in Norway spruce and the length of lesions caused by C. polonica. A companion paper (Viiri et al. 2001) indicated that after fungal inoculation in Norway spruce the total amount of terpenes increased and the total concentrations of stil bene glycoside decreased. On the other hand, N fertiliza tion decreased the concentrations of both total terpenes and total stilbene aglycones (Viiri et al. 2001). Material and methods Experimental area and fungal inoculations of trees The study area was located in a mature natural stand of Norway spruce [Picea abies (L.) Karst.] in Karttavuori, Vesijako. southern Finland (see Viiri et al. 2001 for details). At the end of the experi ment. the experimental trees averaged 34±0.5 cm in diameter at breast height and 28±0.2 m in total height. Four plots of 0.5 ha were marked, and fertilization treatments [N (173 kg ha -1 year 1; NH 4 -N 13.5%: NOrN 13.5%: Ca 3%: Mg 1%: S 3%: B 0.02%), P (41 kg ha -1 yean 1: total P 9%: P water soluble 8%: Ca 20%: S 11%) and NPK (800 kg ha- 1 year 1 : NH4 -N 11%: NO r N 7%; to tal P 5%: P water soluble 3.6%: K 10%: Ca 3%: Mg 0.5%; S 3%; B 0.02%; Se 0.001%)] were spread by hand on 4 May 1993. The control treatment was a non-fertilized plot. Due to limited space, we used a single plot for each fertilization treatment and one con trol plot. Thus, our experimental design was pseudoreplicated (Hurlbert 1984); this is taken into account in the interpretation of results, where fertilization treatments are compared with the con trol, not with each other. From each plot, 30 trees that were free of visible wounds, a total of 120 trees, were selected. Two trees that were infected with root-rot and two attacked by I. typographus were omitted from the analysis. In June 1994 a culture of blue-stain fungus. C. polonica, on malt extract agar in Petri dishes was inoculated into ten trees per fertilization treatment. The fungus was inoculated into the phloem with a cork borer (diameter 5 mm), and the bark plug was re inserted into the hole according to the method originally described by Wright (1933). Each tree received four inoculations, one at each of the cardinal points of the compass, 1.3 m above ground level. For mechanical wounding, ten trees per fertilization treat ment were injured with a cork-borer in the same way as for inocu lation, but without the fungus and malt agar. In the middle of August, after a 2-month incubation period, phloem samples were taken and the vertical length of each lesion was measured. Around each site of fungal inoculation, four phloem samples were taken with the cork-borer (diameter 15 mm) to the level of the cambium: two from the distal ends of the visible reaction lesion (henceforth called "far" samples) and two immediately above and below the site of inoculation (henceforth called "near" samples). Two sam ples were taken from around the site of mechanical wounding, one above the wounding site and the other below the wounding site. In cases where there was visible lesion formation around the me chanical wounding (necrotic area more than 3 mm at the upper and lower side of the inoculation site, «=11), these necrotic tissues were included as part of the samples. Two samples of phloem per tree from intact trees were used as unwounded controls. Phloem samples of intact trees were taken at the same height and location as in mechanically wounded and fungus-inoculated trees. All sam ples were immediately placed on dry ice and stored at -20° C until analysed. Tree characteristics At the end of the growing season, on 5 October 1994, the experi mental trees were felled and their main characteristics were mea sured: age, height, diameter at breast height, crown length, bark and phloem thicknesses (Table 1). Stem discs were cut from each tree at a height of 6.1 m for assessment of radial growth of the stem during the last 5 years (1990-1994) and determination of the tree-vigour index (Waring et al. 1980; Munster-Swendsen 1987). The cross-sectional area of the sapwood is rather constant from breast height to the base of the live crown. The vigour indices were calculated as the ratio of the cross-sectional area of the annu al ring and the basal area of the sapwood. Diameters were mea sured twice in opposite directions, and all other measurements of thickness (annual rings, bark, phloem and sapwood) were means of four measurements made at four positions along the radii 90° apart. In some cases the heartwood-sapwood border was impossi ble to distinguish precisely due to the changing direction of the tracheids at the base of the branch whorls or similarity in the natu ral colours of the sap and heartwood. In these cases (n= 3o), the thickness of the sapwood was estimated from the regression model v= 13.111+0.655 x (r2=0.183, /;=9O) based on the actual diameter (x) and sapwood (y) measurements in the same experiment. Analysis of chemical elements When trees were felled at the end of the experiment, needle sam ples for analyses of chemical elements were taken from the mid crown on the southern side of the trees. From all experimental trees, the needles of the current year class, 1994. and the previous year class, 1993. were sampled separately. After all sampled nee dles had been dried for 24 h at 60° C and ground, the total concen trations of C and N were determined with a CHN elemental ana lyser (CHN-1000: Leco). The concentrations of other elements 329 Table 1 Characteristics (mean±SD) of the experimental trees at the end of the experi ment *P<0.05, **/>