Jukuri, open repository of the Natural Resources Institute Finland (Luke) All material supplied via Jukuri is protected by copyright and other intellectual property rights. Duplication or sale, in electronic or print form, of any part of the repository collections is prohibited. Making electronic or print copies of the material is permitted only for your own personal use or for educational purposes. For other purposes, this article may be used in accordance with the publisher’s terms. There may be differences between this version and the publisher’s version. You are advised to cite the publisher’s version. This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): S.M. Velmala, I. Vuorinen, A. Uimari, T. Piri & T. Pennanen Title: Ectomycorrhizal fungi increase the vitality of Norway spruce seedlings under the pressure of Heterobasidion root rot in vitro but may increase susceptibility to foliar necrotrophs Year: 2018 Version: Accepted manuscript Copyright: The Authors 2018 Rights: CC BY-NC-ND 4.0 Rights url: http://creativecommons.org/licenses/by-nc-nd/4.0/ Please cite the original version: Velmala, S. M., Vuorinen, I., Uimari, A., Piri, T., & Pennanen, T. (2018). Ectomycorrhizal fungi increase the vitality of Norway spruce seedlings under the pressure of Heterobasidion root rot in vitro but may increase susceptibility to foliar necrotrophs. Fungal Biology, 122(2–3), 101–109. https://doi.org/10.1016/j.funbio.2017.11.001 1 Ectomycorrhizal fungi increase the vitality of Norway spruce seedlings 1 under the pressure of Heterobasidion root rot in vitro but may increase 2 susceptibility to foliar necrotrophs 3 Velmala SM1, Vuorinen I1, Uimari A2, Piri T1, Pennanen T1. 4 1Natural Resources Institute Finland (Luke), Latokartanonkaari 9, 00790 Helsinki, Finland; 5 sannakajsa.velmala@luke.fi, irmeli.vuorinen@luke.fi, tuula.piri@luke.fi, 6 taina.pennanen@luke.fi; 2Natural Resources Institute Finland (Luke), Suonenjoki, Finland; 7 anne.uimari@luke.fi 8 Author for correspondence: Sannakajsa M. Velmala, Tel: +358 29 532 2580, e-fax: +358 295 9 326 100, Email: sannakajsa.velmala@luke.fi 10 Abstract 11 We tested if root colonisation by ectomycorrhizal fungi (EMF) could alter the susceptibility 12 of Norway spruce (Picea abies) seedlings to root rot infection or necrotic foliar pathogens. 13 Firstly, spruce seedlings were inoculated by various EMF and challenged with 14 Heterobasidion isolates in triaxenix tubes. The ascomycete EMF Meliniomyces bicolor, that 15 had showed strong antagonistic properties towards root rot causing Heterobasidion in vitro, 16 protected spruce seedlings effectively against root rot. Secondly, spruce seedlings, inoculated 17 with M. bicolor or the forest humus, were subjected to necrotrophic foliar pathogens in 18 conventional forest nursery conditions on peat substrates. Botrytis cinerea infection after 19 winter was mild and the level of needle damage was independent of substrate and EMF 20 colonisation. Needle damage severity caused by Gremminiella abietina was high in seedlings 21 grown in substrates with high nutrient availability as well as in seedlings with well-22 established EMF communities. These results show that albeit M. bicolor is able to protect 23 spruce seedlings against Heterobasidion root rot in axenic cultures it fails to induce systemic 24 protection against foliar pathogens. We also point out that unsterile inoculum sources, such as 25 the forest humus, should not be considered for use in greenhouse conditions as they might 26 predispose seedlings to unintended needle damages. 27 Keywords: Meliniomyces bicolor, Botrytis cinerea, Gremmeniella abietina, forest nurseries, 28 integrated pest management IPM, Picea abies 29 2 1 Introduction 30 Norway spruce (Picea abies) is the most widely used tree species in Northern forest 31 nurseries; In Finland more than 100 million Norway spruce seedlings are produced annually 32 (Finnish Food safety Authority 2010–2016). During and after winter storage in forest 33 nurseries spruce seedlings are vulnerable to various microbial diseases. Grey mould Botrytis 34 cinerea commonly infects spruce seedlings by spreading from diseased seedlings to healthy 35 ones during winter storage or thawing resulting in mild discoloration spots or even total 36 decay of needles (Petäistö 2006). Another disease showing belated symptoms after winter 37 dormancy is the ascomycete Gremmeniella abietina that causes Brunchorstia disease in pine 38 and spruce. The fungal hypha of G. abietina spreads inside the aboveground plant tissues 39 during dormancy. The infection becomes visible gradually during the following growing 40 season as browning of needles in the mid-section of the shoot, discolouring of foliage, 41 presence of resin impregnated necrotic stem lesions and sometimes even dieback of the 42 whole shoot (Petäistö 2008, Børja et al.2006). 43 During the first years after out planting, Norway spruce seedlings commonly suffer from pine 44 weevil damage (Hylobius abietis) (Luoranen et al.2017) but in longer term, also from root rot 45 caused by Heterobasidion annosum sensu lato, the group of most severe fungal coniferous 46 tree pathogens in northern temperate regions, particularly in Europe. In Finland, two 47 Heterobasidion species, H. annosum sensu stricto and H. parviporum (Niemelä and 48 Korhonen 1998), infect spruces of all ages causing root and butt rot. Heterobasidion root rot 49 becomes a risk especially at previously infected regeneration sites, where old conifer stumps 50 as well as decaying root pieces in soil colonised by Heterobasidion spp. act as infection 51 sources for the next tree generation (Stenlid 1987, Piri 1996, Piri 2003). On infested sites, the 52 first spruce seedlings can be infected by Heterobasidion root rot already four to five years 53 after planting (Piri and Hamberg 2015). H. annosum and H. parviporum inhabit tree roots and 54 spread via belowground root contacts (Garbelotto and Gonthier 2013) thus overlapping with 55 habitats of symbiotic ectomycorrhizal fungi (EMF) in forest soils. 56 Formerly fungal diseases in seedling production have been tackled with chemical fungicides, 57 but due to the restricting legislation and banning of several chemicals (European Parliament 58 2009) there is an urgent need of integrated pest management that can be cost efficiently 59 utilized in forest nurseries. Mycorrhizal fungi have suggested triggering signals that move 60 through plants causing biochemical changes affecting growth allocation, nutrition levels, 61 inducible defence mechanisms and resistance as well as host vitality (Whipps 2004). 62 3 In arbuscular mycorrhizal angiosperm plants both positive and negative feedbacks have been 63 observed between aboveground and belowground microbial communities; Priming and 64 protective systemic effects against microbial and invertebrate pathogens have been shown to 65 move to distal plant organs (Pozo and Azcón-Aguilar 2007, Pineda et al.2010) but the 66 outcome seems to be context dependant (Jung et al.2012). Arbuscular mycorrhiza-induced 67 jasmonic acid-regulated resistance boosts host aboveground durability towards pathogens 68 with a necrotrophic lifestyle but in contrast may make the host more susceptible to biotrophic 69 and viral pathogens (Pozo and Azcón-Aguilar 2007, Jung et al.2012). 70 Induced resistance has shown to move from roots to shoots also in Norway spruce, as root 71 inoculation with the mycorrhiza helper bacterium Streptomyces sp. increased needle 72 resistance against B. cinerea (Lehr et al.2007). Furthermore, Rajala et al.(2013) found a 73 negative relationship between the abundance of fungal needle endophytes and the EMF 74 richness on Norway spruce. Some earlier studies have tested the biocontrol potential of EMF 75 against soil and air derived plant pathogens in conifers, but again the effects seem to be 76 fungal strain specific and vary according to the physiological state of the host (Zhang et 77 al.2011, Farquhar and Peterson 1991, Hwang et al.1995, Martín-Pinto et al.2006). In the 78 case of unspecific general responses of Norway spruce to fungi, the effects of symbiotic EMF 79 Pisolithus tinctorus on peroxidase activity and on the amount of free salicylic acid in roots 80 seem to be transient in comparison to the more permanent effects caused by H. annosum 81 (Likar and Regvar 2008). EMF may still permanently modify the chemical content of distal 82 organs as the colonisation by EMF Hebeloma bryogenes and Cadophora finlandica reduced 83 seedling growth but increased nutrient content and photosynthetic pigment levels in needles 84 of Norway spruce (Mrnka et al. 2009). Thus there is an increasing interest in the long-term 85 effects of these interactions on distant plant parts. Furthermore, it is important to address the 86 effects of EMF on growth allocation, access to nutrients, and the formation of induced and 87 constitutive defence (Bonello et al. 2006). 88 Paxillus involutus (Basidiomycota) and Meliniomyces bicolor (Ascomycota) form 89 ectomycorrhizal associations with Norway spruce roots in boreal forests. M. bicolor is part of 90 the Pezoloma ericae (formely Rhizoscyphus ericae) aggregate including phylogenetically 91 related fungi (M. variabilis, R. ericae and C. finlandica) inhabiting roots of both hardwood 92 and herbaceous plants in various habitats, both in the Northern and Southern hemisphere 93 (Vrålstad et al. 2000, Hambleton and Sigler 2005, Bruzone et al. 2017). Individual genets of 94 fungal species belonging to the P. ericae aggregate can simultaneously form both ericoid and 95 4 ectomycorrhizal associations (Grelet et al. 2010). Previously it has been found that certain 96 isolates of P. involutus and M. bicolor were able to restrict or even halt the growth of H. 97 parviporum and H. annosum strains in vitro on nutrient agar plates in laboratory (Hyder et 98 al.2013). Hence, we aim to test in triaxenic cultures how certain EMF isolates, that have 99 shown to be antagonistic against Heterobasidion species in vitro, affect the viability of 100 Norway spruce seedlings. In addition, we study the effects of the antagonistic EMF M. 101 bicolor on spruce growth allocation, and susceptibility of distant organs towards necrotic 102 foliar pathogens B. cinerea and G. abietina in conventional forest nursery conditions. As a 103 result, we gather elementary knowledge prior to compile recommendations for the use of 104 certain EMF strains in nursery. 105 2 Material and Methods 106 2.1 Triaxenic testing of protective effects of EMF against Heterobasidion root rot 107 2.1.1 Fungal strains and preparation of vegetative inoculum 108 The ectomycorrhizal strains that outperformed standard antagonism against H. parviporum 109 and H. annosum (Hyder et al. 2013) were selected to be used in the triaxenic test (Table 1). 110 Additionally, Thelephora terrestris, a common nursery EMF, and a common root endophyte 111 Phialocephala fortinii were included as non-antagonistic controls. H. parviporum isolates 112 were collected from Norway spruce (isolate A) thinning stump and from living trees (B,C) 113 and H. annosum isolates from final cutting stumps (D, E) of Scots pine in southern Finland. 114 All fungal isolates were collected from different stands and represented different genotypes 115 based on pairing tests (Stenlid 1985). Vegetative mycelia of H. parviporum (A, B, C) and H. 116 annosum (D, E) were grown on Brown & Wilkins (1985) agar on Petri dishes. Twice 117 autoclaved wooden sticks (121 ˚C 60 min; Ø 3mm x 20mm) were placed on top of the 118 cultures for three weeks to allow mycelia to grow into the sticks. The ectomycorrhizal, root 119 endophytic and Heterobasidion fungal isolates used in this study were obtained from the 120 culture collection of the Natural Resources Institute Finland, except P. involutus BOUX 121 which originates from France, INRA-Nancy. T. terrestris strain was initially isolated from a 122 mycorrhizal root tip of a Norway spruce seedling (Flykt et al. 2008). M.bicolor and P. fortinii 123 were isolated similarly from spruce mycorrhizas and the Finnish P. involutus strains were 124 originally isolated from fruiting bodies (details in Vuorinen et al. 2015). 125 2.1.2 Preparation of triaxenic synthesis tubes 126 5 The potential of EMF isolates to protect Norway spruce seedlings from H. annosum and H. 127 parviporum was tested using a modification of the synthesis tube system described previously 128 by Timonen et al. (1993). Briefly, surface sterilized (30% H2O2 15 min) seed-orchard-seeds 129 of Norway spruce (ROI-89-1002 SV. 109) of southern Finnish origin were germinated on 130 glucose (5%) agar plates. After germination seedlings were transferred into glass tubes (Ø 25 131 mm x 200 mm) on 9 ml slants of Brown & Wilkins agar covered with Leca-clay balls (acid 132 washed and twice autoclaved for 20 min 121°C) moistened with 2,5 ml of Modified Melin-133 Norkrans (Marx 1969) nutrient solution with reduced sugar and malt content (½ MMN). 134 Tubes were closed with cotton wool plugs. 135 2.1.3 Tube experiment 136 Triaxenic synthesis tubes containing spruce seedlings were placed in Conviron PGR15 137 growth chambers (Controlled environments Ltd., Manitoba, Canada) under 280 µM light 138 intensity at 23 °C and 65% humidity with a 16 h day length, and 8 hours of darkness. Lights 139 were composed of high pressure sodium (250W, Lumalux®, Sylvania) and metal halide 140 (250W, Britelux®, Sylvania) light bulbs (Feilo Sylvania Finland Oy). 141 At the age of four weeks, 30 replicate seedlings per each root associated fungi (Table 1) were 142 inoculated with two agar plugs taken from the peripheral growth zone of an actively growing 143 fungal colony in ½ MMN agar. 30 seedlings were left as uninoculated controls. Tubes were 144 incubated for six weeks and ectomycorrhiza formation was assessed by dissecting 145 microscope prior to further inoculations with pathogens. Heterobasidion isolates (A–E) were 146 introduced into the system by placing wood sticks colonised with each isolate close to the 147 base of the seedlings. This resulted in total 48 treatments with five replicate seedlings each; 148 seedlings pre-infected with six EMF fungi, one endophyte and no EMF-control, and all these 149 challenged with five Heterobasidion isolates and the pathogen free control treatment. 150 Seedlings were grown in growth chambers for four months and irrigated with 2 ml of sterile 151 water once a month. 152 After altogether six months, as soon as the seedlings seemed to reach stationary stage, the 153 vitality of the seedlings was assessed visually to a five-step discrete variable according to the 154 following classification: i) dead seedlings with totally brown needles (B), ii) chlorotic 155 needles with brown lesions (CB), iii) chlorotic needles (C), iv) light green needles (LG), and 156 v) healthy seedlings possessing only green needles (G). Seedlings were removed from the 157 6 tubes, washed, dried in +70˚C and weighed for their biomass. The roots of seedlings with 158 severe Heterobasidion infection had already started to decay before the sampling. 159 2.2 Impact of antagonistic EMF on susceptibility to foliar pathogens 160 2.2.1 Preparation of plant and EMF material for the nursery experiment 161 In late May seed-orchard-seeds of Norway spruce (ROI-89-1002 SV. 109) were sown in six 162 Plantek-81F containers (vol 85 cm3, BCC) on top of unfertilized blonde sphagnum peat 163 (PP03, Kekkilä Group, Vantaa, Finland) supplemented with five different inoculums (Table 164 2) from which one contained 15% (vol:vol) vegetative fungal hypha of M. bicolor grown in 165 solid silica-based malt extract medium as in Vuorinen et al.(2015) (Table 2). The vitality of 166 EMF vegetative inoculum was confirmed on ½ MMN agar. 167 In each of the six containers (9 × 9 plants per container) two rows (9 plants in a row) were 168 randomly filled with one of the four treatment substrates (18 seedlings/substrate/container), 169 and one row of each container was filled with conventional fertilized blonde Sphagnum-peat 170 (Ctrl treatment, Table 2). The seedlings were germinated in a glasshouse, and transferred 171 outside to the nursery field in the end of June. Between June and August each seedling was 172 fertilized weekly with approximately 5 ml 0.1% w:w liquid fertilizer Turve Superex® N-P-K 173 (12–5–27) (Kekkilä Group, Vantaa, Finland) to keep the peat conductivity as low as 0.5–1 174 mS cm−1. The experiment was established at the Haapastensyrjä nursery, Southern Finland 175 and plants were grown in the Suonenjoki nursery (62.625N, 27.122E) in Eastern Finland. 176 2.2.2 Preparation of spore suspensions and foliar infection 177 The foliar inoculations with the two fungal pathogens Botrytis cinerea and Gremmeniella 178 abietina, and the uninfected water only- treatment were performed in early June and August. 179 Two out of six Plantek-81F containers were randomly chosen to each foliar treatment and 180 labelled accordingly, totalling in 36 replicate seedlings per treatment and 18 replicate Ctrl 181 seedlings in fertilized peat (Table 2). B. cinerea was grown on potato dextrose agar (PDA) at 182 +17 °C for three days in darkness followed by three to seven days under photoperiodic 183 ambient lighting (16 hours of light:8 hours of dark). The B. cinerea spores from the fungal 184 cultures were liberated into 10 ml of sterile water by rubbing the plates with a sterile glass 185 rod. Hyphae from the spore suspension were removed by running the suspension through 186 sterile cotton gauze. G. abietina was grown on autoclaved wheat substrate (10 g wheat and 25 187 ml distilled water in 100 ml growth bottle) for one to two months at +17 °C under artificial 188 7 light with the photoperiod of 16 hours of light and 8 hours of dark. The spore mass on the 189 culture was collected by a sterile loop and suspended into 20 ml of sterile water. The number 190 of spores was counted by Neubauer-improved chamber (Wertheim, Germany). Before 191 inoculation, the shoots were sprayed with tap water to moisten the foliage. The seedlings 192 were spray-inoculated with 2 ml of spore suspension of either B. cinerea isolate BcSjk1.1 193 (Petäistö et al. 2004) or G. abietina containing ca. 300 000 spores/ml. The vitality of spore 194 suspensions was confirmed by monitoring the spore germination on PDA plates. The water 195 only-seedlings were sprayed with 2 ml of autoclaved tap water. After inoculation, the 196 seedlings were maintained in greenhouse under plastic cover for three days and the shoots 197 were repeatedly sprayed with tap water to keep the foliage humid. Thereafter the seedlings 198 were grown in a plastic greenhouse ventilated by opening side panels and doors. The 199 seedlings were monitored daily for two weeks after the inoculation. A throughout inventory 200 of needle damage was done twice, in October after the growing season and in May after 201 wintering in the open field nursery under natural snow cover. 202 2.2.3 Sampling and determination of fungal colonisation 203 In early June seedling height was measured and roots were carefully washed with tap water. 204 From each combination of foliar and growth substrate treatments ten replicate root systems 205 were randomly chosen for EMF colonisation assessment. 206 EMF colonisation and number of fungal morphotypes were determined under dissecting 207 microscope. Representative root tips from each morphotype were amplified with the Thermo 208 Scientific Phire Plant Direct PCR Kit® as described in detail in Velmala et al.2013. Root tips 209 were crushed in 20 μl Dilution Buffer and 1 μl was used as template in a 25 μl reaction 210 volume, and amplified with primers ITS1F (Gardes and Burns 1993) and ITS4 (White et 211 al.1990). PCR products were separated in 1% Synergel (Diversified Biotech, MA, USA) for 212 3h 120V. Distinguishable bands were cut and dissolved in water, and used as template for re-213 amplification with DreamTaq DNA Polymerase (Thermo Scientific, Waltham, MA, USA) 214 with the same primers for 25 cycles. End products were sequenced by the Macrogen 215 Sequencing Service (Macrogen Europe, The Netherlands) using the ITS4-primer. The 216 sequences were revised in Geneious R6.1.8 (Biomatters, Auckland, New Zealand, available 217 from http://www.geneious.com/). The OTUs were identified by comparing the sequences 218 against sequences in UNITE (http://unite.ut.ee/) and the INSD database. Sequences are 219 deposited in GenBank under the accession numbers MF947534–MF947542. 220 8 2.2.4 Determination of biomass and nitrogen content 221 Roots and shoots were separately oven dried for 48h in 50°C and weighted for biomass 222 determination. Needle nitrogen concentration was determined with a CHN-100 analyser 223 (Leco, analysis based on ISO 10694 and ISO 13878 standards) in Natural Resources Institute 224 Finland. Before analysis 200 mg powdery needle sample (from 0.5 g of dry needles 225 homogenised in 2 ml screw cap tubes with FastPrep® (MP Biomedicals, CA, USA) for 2 x 226 30s with 4.5 m/s speed) from three replicate seedlings from each row of the water only- 227 treatment were pooled resulting in altogether four samples per growth substrate. 228 2.3 Statistical analyses 229 The data from both experiments was analysed as linear models with R software version 3.2.3. 230 (R Core Team 2015). We used glht function from ‘multcomp’ (Hothorn et al. 2008) package 231 to make Dunnett’s multiple comparison of means and to calculate the confidence intervals for 232 general linear hypotheses. Model assumptions for homogeneity of variances, and normality 233 of residuals were checked by plotting residuals against fitted values, and creating a Q-Q plot 234 of residuals. Bar plot figures were plotted with ‘ggplot2’ (Wickham 2009) and arranged in 235 grids with ‘gridExtra’ (Auguie 2016) packages. 236 2.3.1 Triaxenic tube experiment 237 As the explanatory variable we used a 24 level variable that was coined to cover all eight 238 combinations of root associated fungi with either no pathogen (=control), H. parviporum (A-239 C) or H. annosum (D, E). All treatments were compared to the reference, 0-level, seedlings; 240 No inoculation with root associated fungi and no pathogens. Shoot and root biomass were 241 fitted as linear models with lm function. For needle vitality class response variable we fitted 242 a generalised linear model glm assuming Poisson error distribution and log link function (no 243 over-dispersion was observed). The relationship between vitality and root and shoot growth 244 were calculated as Pearson’s product-moment correlations. 245 2.3.2 Foliar pathogen experiment in the nursery 246 As the explanatory variable we used a 15 level variable that was coined to cover all foliar and 247 growth substrate combinations. All treatments were compared to the reference level 248 seedlings; water only- foliar treatment grown in Ctrl fertilized peat (Table 2). Root biomass, 249 shoot:root- ratio, and shoot height were fitted as linear models with lm function. For EMF 250 9 colonisation and needle damage percentage response variables we fitted a generalised linear 251 model glm with binomial error distribution and logit link function. EMF richness count was 252 analysed as a generalized model assuming Poisson error distribution and log link function (no 253 over-dispersion was observed). N content and concentration was analysed as linear models 254 with growth substrate as the explanatory variable, and all substrates were compared to Ctrl 255 fertilized peat substrate. The relationship between EMF, Meliniomyces sp., T. terrestris, C. 256 geophilum and Amphinema sp. colonisation, and species richness with needle damage %, N 257 content, shoot and root growth were calculated as Pearson’s product-moment correlations. 258 3 Results 259 3.1 Triaxenic testing of spruce, EMF and root rot pathogen Heterobasidion spp. 260 The vitality of seedlings varied between green healthy seedlings and slightly chlorotic yellow 261 phenotypes when no pathogens were present (Fig. 1a). When seedlings without any root 262 associated fungi were subjected to either H. parviporum or H. annosum isolates they turned 263 brown and died. In general, it seemed that all EMF provided some protection against root 264 pathogens, and seedling needle vitality was only slightly lowered in most treatments (Fig. 265 1a). H. parviporum or H. annosum isolates did not show within species differences and thus 266 the results are bulked together within Heterobasidion species. 267 Through the treatments, shoot biomass was the highest in T. terrestis and P. involutus 2 268 inoculated seedlings. M. bicolor inoculated seedlings had the most even shoot growth when 269 infected with Heterobasidion isolates (Fig. 1b), and showed least disease symptoms under H. 270 annosum infection (Fig. 1a). The correlation between vitality and shoot biomass was 271 moderate (0.53, P<0.001). 272 Root growth responded differently to various root inoculations; when Heterobasidion was 273 present root growth was significantly reduced in seedlings inoculated with P. fortinii and 274 P.involutus 1. P. fortinii and P.involutus 1 also resulted in the lowest vitality classes even 275 without pathogens present (Fig. 1a). P. involutus strains 2, 3, and 4 were able to efficiently 276 protect seedlings against Heterobasidion infection, even though some discoloration and 277 redness of needles was more common than with other EMF. Roots inoculated with M. bicolor 278 and T. terrestris were least affected by Heterobasidion isolates (Fig. 1c). Root biomass had a 279 strong positive correlation with vitality class (r=0.7, P<0.001) indicating that root growth in 280 in vitro set ups can be used to extrapolate the vitality of seedlings. 281 10 3.2 Impact of M. bicolor on susceptibility of spruce seedlings against foliar pathogens 282 3.2.1 Needle damage became visible after winter dormancy 283 Needle damage in autumn after the first growing season in nursery was low (1.35±0.1%) and 284 no statistically significant differences could be observed between the treatments. After winter 285 dormancy, needle damage caused by necrotrophic pathogens became visible, and in May 286 clear symptoms with almost half of the needles injured could be seen in G. abietina infected 287 seedlings (Fig. 2a). Needle damage in spring was statistically significantly higher in G. 288 abietina infected seedlings than in the Ctrl seedlings treated with water only. The foliage of 289 water only- treated seedlings suffered some damage especially when growing in Humus 290 substrate, which possessed even higher amount of needle damage than B. cinerea infected 291 seedlings (Fig. 2a). 292 Mortality of the nursery experiment seedlings after winter was low, less than 0.7%. The 293 highest mortality rates were observed when seedlings were exposed to G. abietina and B. 294 cinerea foliar treatments; 15 out of 162 seedlings died in both G. abietina- (evenly in all 295 growth substrates) and in B. cinerea- treatments (half growing in sieved humus and the rest in 296 Me, Ctrl-M, and Ctrl-H substrates). Only four seedlings died during the experiment from the 297 water only- treatment and again mostly from the humus substrate. 298 3.2.2 B. cinerea and G. abietina caused needle damage despite EMF colonization 299 EMF colonisation in either M. bicolor or Humus treatments did not protect seedlings against 300 foliar pathogens: B. cinerea infection severity seemed to be independent of growth substrate 301 and EMF status. On the contrary, the severity of needle damage caused by G. abietina was 302 high in growth substrates containing EMF inoculum (Me, Humus) or well-balanced nutrients 303 (Ctrl) (Fig. 2a, Table 2). 304 3.2.3Needle N content was lowest in seedling growing in EMF inoculated substrates 305 Shoot height and biomass as well as root biomass were affected by both foliar pathogen 306 treatments and growing substrate (Figs 2b and 2c). Fertilized peat (Ctrl) and sterile humus 307 (Ctrl-H) inoculations ensured the best shoot growth. These seedlings were on average 20% 308 taller than in other treatments (Fig. 2b). Furthermore, the good nutrient status in Ctrl and Ctrl-309 H treatments was also reflected as the highest total N contents of needles (3.40±0.11 mg and 310 3.65±0.30 mg, ns, respectively). The lowest N content was found in M. bicolor inoculated 311 seedlings (2.09±0.35 mg, P<0.001), and Ctrl-M (2.48±0.13 mg, P<0.01) and Humus 312 11 treatments (2.55±0.07 mg, P<0.05) when compared to the reference level (Ctrl). Needle N 313 content correlated negatively with Meliniomyces sp. colonisation (r=-0.49, P<0.01), and 314 positively with colonisation percent of T. terrestris (r=0.45, P<0.01) and Amphinema sp. 315 (0.36, P<0.05) and EMF species richness (r=0.40, P<0.02). Needle N concentration was on 316 average 1.14±0.03% in all growth substrates. 317 Roots had significantly lower biomass in Ctrl-M and M. bicolor inoculated substrates 318 compared to uninoculated Ctrl substrate, especially when exposed to B. cinerea. The 319 infection by G. abietina reduced root growth statistically significantly in all substrates except 320 Ctrl-H (Fig. 2d), and shoot growth was restricted even more than root growth as the 321 shoot:root- ratios were statistically significantly lower after G. abietina infection than in the 322 water only- foliar treatment (Fig. 2c). EMF colonisation % had a weak positive relationship 323 with root growth (r=0.29, P<0.001). Furthermore, EMF species richness had a weak positive 324 relationship with both shoot and root biomass (r=0.24and r=0.26, P<0.001). 325 3.2.4 Roots were well colonized with M. bicolor 326 As expected EMF colonisation was highest in M. bicolor inoculated substrate, almost 87%, 327 but there was a lot of variation in colonisation levels between seedlings in different 328 treatments (Table 3). In all the other substrates, roots were abundantly colonised (from one 329 fifth to almost 60%) with Thelephora terrestris. Also other root associated fungi were found 330 in lower quantities on the roots; forest humus inoculated seedlings, Humus, hosted the most 331 diverse fungal community associated to their roots comprising Amphinema, Tylospora, 332 Cenococcum, Meliniomyces, Varicosporium, Phialocephala, Psilocybe and 333 Archaeorhizomyces fungal genera (Table 3). 334 4 Discussion and conclusions 335 In the triaxenic tube experiment EMF colonisation of spruce roots by M. bicolor, T.terrestris 336 and three out of four P. involutus isolates provided protection against Heterobasidion root 337 rot, and as in the study by Hyder et al. (2013), the strongest protective effect was provided by 338 the ascomycetous EMF M. bicolor. In general the same strains, that were found to be 339 antagonistic to each other when subjected to mycelial confrontation in vitro (Hyder et 340 al.2013), had positive impacts on the viability of the seedlings also in present triaxenic 341 system which contained a tree seedling as the tested subject. However, a few exceptions were 342 found, most probably due to differences in experimental systems; the variable protective 343 12 mechanisms of EMF against root pathogens may include production of antifungal substances, 344 induction of inhibitory compounds and root exudates from the host plant, and even microbial 345 competitive potential amongst other things (Marx 1969, Marx 1973, Duchesne et al.1989, 346 Chakravarty and Hwang 1991, Buscot et al.1992). For example the T. terrestris strain, that 347 did not show any antagonism against Heterobasidion isolates in vitro plate tests (Hyder et 348 al.2013), provided protection against the less aggressive root rot species H. parviporum 349 (Swedjemark and Stenlid 1995) in the present study. T. terrestris grew thick external mycelia 350 in the tubes (data not shown) and thus most likely formed also a physical barrier around the 351 roots. Furthermore, T. terrestris seemed to provide efficient nutrient allocation to support 352 good growth. Similarly Buscot et al. (1992) reported greater vigour of mycorrhizal Norway 353 spruce seedlings when inoculated with P. involutus and Laccaria laccata since EMF could 354 have increased host protection trough reinforcement of plant resistance and production of 355 antifungal phenolics even though no direct antagonism was present. Likar and Regvar (2008) 356 noted that induction of defence cascade after inoculation with the EMF P. tinctorius was only 357 transient, and thus they claimed that Norway spruce could recognise the infecting fungus and 358 activate and adjust appropriate defence mechanisms. Then again EMF colonisation have 359 shown explicit protective effects against Heterobasidion spp., and also other root pathogenic 360 fungi (Rhizoctonia solani, Fusarium damping-off, Ilyonectria destructans (in both pine and 361 spruce (Farquhar and Peterson 1991, Hwang et al.1995, Buscot et al.1992, Zhang et al.2011, 362 Martín-Pinto et al. 2006). Thus also in root tips that were highly covered with dense M. 363 bicolor hypha, fungal colonization might have provided a physical protective barrier around 364 roots in addition to direct antagonism of M. bicolor against Heterobasidion. 365 The most evident exceptions in the tube experiment were P.involutus isolate 1 and the root 366 endophyte P. fortinii, which lacked all antagonism, and tended to reduce the fitness of the 367 seedlings even in the absence of root pathogens. Similar within species variability of 368 antagonistic abilities of certain EMF (P. involutus) against Heterobasidion isolates has 369 previously been reported amongst others by Červinková (1989). The root endophyte 370 Phialocephala sphareoides was shown to protect spruce seedlings from H. parviporum root 371 infections in vitro by means of antifungal metabolites (Terhonen et al. 2016). Yet the 372 antagonistic abilities of root endophytes are variable; Tellenbach et al. (2013) found only one 373 antagonistic root endophytic fungal strain from over 80 tested isolates. Furthermore, also 374 growth reduction in spruce, caused by the root endophyte P. fortinii, has been previously 375 reported by Reininger et al. (2012), although they emphasize that plant growth responses are 376 13 both fungal strain and host species dependent. Despite possible negative correlations between 377 host biomass and endophytic biomass it has also been speculated that spruce might even 378 actively attract root endophytic fungi to provide protection against more serious pathogens 379 (Tellenbach et al. 2011). Nevertheless, the present results suggest that this hypothesis may 380 not hold with the necrotrophic Heterobasidion sp. pathogens. 381 M. bicolor was selected from the tube experiment for further testing in nursery conditions as 382 it showed good antagonistic and protective properties in vitro. It has also shown to be a 383 suitable EMF species for large scale inoculum production (Vuorinen et al.2015), and to 384 abundantly colonize the roots of spruce. However typically the exploration of the extraradical 385 mycelium of ericoid mycorrhizal fungal species of the P. ericae aggregate is narrow (Read 386 1984) and there is no evidence for the formation of large mycelial networks (Grelet et al. 387 2010). 388 In the nursery experiment, the infection of foliar pathogen G. abietina caused severe damage 389 after winter in particular in seedlings growing in conventional fertilized peat (Ctrl), in contact 390 with M. bicolor or in forest humus containing natural microbial fauna and flora. Hence, on 391 the contrary to its effectiveness towards root pathogens M. bicolor was not able to provide 392 protection from the foliar G. abietina infection. G. abietina changed resource allocation 393 towards roots, as the decrease in growth was stronger in shoots than in roots. The addition of 394 forest soil microbes along with humus increased the EMF diversity, N contents and biomass 395 of the seedlings but did not provide benefits against foliar pathogens. Instead, the addition of 396 fresh unheated humus seemed to be a slight risk factor probably due to the exposure of 397 seedlings to wild needle damage caused by the microbes of the soil. 398 The only soil treatment where shoot and especially root growth seemed to be unaffected by 399 G. abietina was the heat-treated humus. These seedlings also had the highest N storage in 400 their needles and the best root and shoot growth within all the substrates. This was probably 401 due to the autoclaving of humus that released microbe-bound nutrients or other beneficial 402 compounds into the growing media without exposing seedlings to potential pathogens. 403 Heating of humus up to less than 200 degrees Celsius has shown to increase the quantity of 404 water soluble potassium (K) and phosphorus (P) without yet reducing the levels of N (White 405 et al. 1973). Seedlings growing in the heat-treated humus were also the ones most abundantly 406 colonised by EMF T. terrestris. T. terrestris has been found to secrete both N and P 407 solubilizing exoenzymes in moderate quantities (Velmala et al.2014a), and the seedlings with 408 high T. terrestris colonisation had high N content in our study. 409 14 Good nutritional status increases seedling survival (van der Driessche 1992). Seedlings 410 grown in conventional fertilized peat invested more resources belowground when exposed to 411 foliar stress by B. cinerea compared to the situation with no foliar stress, as has been reported 412 previously (Velmala et al.2014b). Lack of nutrients seemed to reduce shoot and root growth 413 significantly in Ctrl-M and M. bicolor inoculated treatments. Reduced growth was especially 414 clear under foliar stress caused by B. cinerea and G. abietina when compared to the 415 conventional (Ctrl) substrate. The frequently reported short term slowdown effects of EMF 416 on shoot growth in early developmental stages of seedlings (Corrêa et al.2006, Vaario et al. 417 2009) could clearly be seen also in the present study, even though the amplitude seems to be 418 milder under high N availability. 419 Seedlings inoculated with forest humus had the highest EMF richness, and established widely 420 associations with both basidio- and ascomycoteus fungi that are recognized as EMF and 421 endophytic fungi commonly found on spruce (Vohnik et al. 2013, Rosling et al. 2011). 422 However proportions of other EMF than T. terrestris were minor, underlining the strong 423 competitive strength of T. terrestris in nursery conditions. Yet, foliar infection of G. abietina 424 seemed to reflect in ectomycorrhizal formation as the colonisation degree of these seedlings 425 was the lowest in all substrates except for the heat-treated humus substrate. Similarly, we 426 have found that severe exposure to the needle pathogen Gibberella avenacea has a slight 427 negative effect on the EMF richness on Norway spruce seedlings showing slow long-term 428 growth performance (Velmala et al. 2014b). Regardless of these indications of basipetal, top-429 down, movement of pathogen induced systemic signals we found no signs of EMF induced 430 acropetal systemic effects. Moreover, in the present study the positive effect of high EMF 431 colonization and diversity seems to be due to either direct antagonism or root mediated 432 improved nutritional status. Thus, it seems that the host-microbe interactions are highly 433 species and even strain-specific, and no general responses can be expected without 434 throughout knowledge on the identity of the organisms. It appears very characteristic to 435 EMF-host interactions that the within-species variations of effects on host performance are as 436 high as the variation among different EMF species (Pennanen et al., unpublished). Our results 437 also contradict the findings on the effects of C. finlandica, the close relative of M. bicolor, on 438 spruce needle chemistry (Mrnka et al. 2009) which further supports our claim that even 439 phylogenetically closely related EMF species may induce very different effects in trees. 440 In conclusion, our study showed that there are several potential EMF isolates that protect 441 Norway spruce seedlings towards Heterobasidion root rot in vitro. Nevertheless, seedlings 442 15 that were inoculated with the antagonistic EMF isolate M. bicolor or were naturally colonised 443 by T. terrestris in the nursery showed variable aboveground susceptibility towards foliar 444 pathogens in nursery. Before introducing EMF inoculations into forest nursery practice 445 further studies of possible effects of the tripartite interactions should be considered. 446 5 Acknowledgements 447 Funded by Finnish Cultural Foundation for Art and Science, and Academy of Finland project 448 292967 and Natural Resources Institute Finland. We thank Lemström E, Petäistö RL, 449 Jalkanen ML, Oksanen M, Tiikkainen S, Ruhanen H, Hytönen T and Vanhanen R for 450 assistance in inoculations, sampling and laboratory work. We are grateful for Jean Garbaye 451 for the isolate Paxillus involutus BOUX. 452 6 Figure captions 453 Figure 1 Seedling vitality classes and growth of six month-old Norway spruce seedlings 454 grown in glass tubes on Brown and Wilkins -media and inoculated with seven root associated 455 fungi and subjected to Heterobasidion spp. pathogens. Seedling vitality is scored based on 456 visual symptoms: the highest class includes only green needles (G), the middle classes light 457 green (LG) and chlorotic needles (C), and the lowest classes either chlorotic needles with 458 brown lesions or totally brown dead needles (B). Figure panels group H. parviporum and H. 459 annosum infected seedlings visually together. The mycorrhizal inoculum is visible on the X-460 axis; (details in Table 1). Grey bars indicate that there is no statistically significant difference 461 between the treatments and white bars show that the treatment in question differ statistically 462 significantly (P<0.05) from the leftmost base line 0 treatment (no EMF and no 463 Heterobasidion). The black error bars show the 95% confidence interval of the mean (n=5). 464 Figure 2 Needle damage %, and shoot and root growth of 1-yr-old Norway spruce seedlings 465 in spring after wintering under natural snow cover grown in different substrates and subjected 466 to two foliar fungal diseases Botrytis cinerea and Gremmeniella abietina. Figure panels 467 group seedling subjected to same foliar treatments visually together (Water only, B. cinerea 468 and G. abietina). The growth substrate is visible on the X-axis: Ctrl conventional fertilized 469 peat, Me inoculated with EMF M. bicolor, Humus inoculated with natural forest humus layer 470 and their sterilized controls Ctrl-M (no EMF) and Ctrl-H (sterilized humus), respectively 471 (details in Table 2). All treatments are compared to the leftmost reference level Ctrl-472 16 treatment that is fertilized peat and water only- foliar treatment; Grey bars indicate that there 473 is no statistically significant difference, and white (P<0.05) and light grey (0.05