Yue Pan · Jun Lu · Peng Chen · Zefen Yu ·Huihong Zhang · Hui Ye · Tao Zhao

Abstract Ophiostomatalean fungi may facilitate bark beetle colonization and reproduction. In the present study of the fungal community associated with bark beetle species belonging to Tomicus in Yunnan, China, six ophiostomatalean fungi ( Ophiostoma canum , O. ips, O. tingens,Leptographium yunnanense, Leptographium sp. 1 and Leptographium sp. 2) were isolated from the beetles or their galleries; O. canum was the most common fungal species.The distribution of O. canum was associated with stands heavily damaged by Tomicus species and a higher percentage of valid galleries of Tomicus yunnanensis and T. minor in Yunnan pine ( Pinus yunnanensis). After inoculation of Yunnan pine with the fungus, a phloem reaction zone formed and monoterpenes accumulated in the phloem. These results suggested that O. canum was pathogenic to Yunnan pine and that the wide distribution of the fungus might be benef icial to reproduction of pine shoot beetles in Yunnan pine. However, because the reaction zone and monoterpene accumulation were mild, fungal damage of Yunnan pine might be limited. A more integrated study considering all the fungal species should be done to better understand the interactions among bark beetles, blue-stain fungi, and the tree hosts in the region.

Keywords Ophiostomatalean f ungi · Fungal i solation a nd identif ication · Ophiostoma canum · Pinus yunnanensis ·Pathogenicity t est

Introduction

A bark beetle—fungus symbiosis evolved (Sapp 1994; Six 2012) that benef its both organisms, enabling them excel in marginal habitats, occupy new niches, avoid competition and adapt to survive in resistant trees (Bruno et al. 2003;Mueller et al. 2005; Six and Wingf ield 2011). These fungal associates, some of which are plant pathogens, may facilitate bark beetle colonization and reproduction by lowering tree resistance and efficiently capturing resources (Six 2012). In addition, a recent study has suggested that fungal symbionts of the spruce bark beetle affect the synthesis of beetle aggregation pheromone (Zhao et al. 2015, 2019). Fungal symbionts, thus, seem to be closely related to the colonization of bark beetles in host trees.

Yunnan pine (Pinus yunnanensisFranch.), economically and ecologically the most important tree species in southwestern China, has suffered a recordTomicusinfestation since the 1980s. In Yunnan province, 93,000 ha ofP. yunnanensistrees have been killed by the collective action of fourTomicusbark beetles (Tomicus yunnanensisKirkendall& Faccoli,T. minor(Hartig),T. brevipilosus(Eggers), andT. armandiiLi & Zhang) (Lu 2011), which makes up 36.6%,38.1%, 19.7% and 5.6%, respectively, of the beetle population in the region (Wang et al. 2018). Pine shoot beetles usually attack pine shoots and reduce tree growth (Lanne et al.1987). The mass tree-killing by pine shoot beetles seems only to be reported in south-western China (Chen 2011).Fungal associates and warm, dry weather have been thought to contribute to severe damage ofTomicusin the area (Li et al. 1997; Zhou et al. 2000, 2013; Chang et al. 2017; Pan et al. 2017, 2018a).

Fungi in the order Ophiostomatales (Ascomycota) are frequent associates of hardwood- and conifer-infesting bark beetles. Currently, the order Ophiostomatales includes nine phylogenetic lineages represented by the following genera:Aureovirgo,Ceratocystiopsis,Fragosphaeria,Graphilbum,Hawksworthiomyces,Leptographium s.l.,Ophiostoma s.l.,Raffaelea s.s.andSporothrix(De Beer and Wingf ield 2013;De Beer et al. 2016). Most ophiostomatalean species are the causal agent of blue-stain, which causes economic loss in the forest industry (Seifert 1993; Jacobs et al. 2006). So far, of the 27 species of ophiostomatalean fungi reported in Yunnan, China, nine are associated withTomicusspp.(Zhou et al. 2000, 2013; Chang et al. 2017; Pan et al. 2017,2018b). One species,Leptographium yunnaneseX.D.Zhou,K.Jacobs, M.J.Wingf. & M.Morelet is reported to be highly virulent toP. yunnannesis(Liao and Ye 2002, 2004). However, knowledge about ophiostomatalean species in the area is still very limited.

To better understand the diversity of fungi associated withTomicusand evaluate the pathogenicity of the most common species associated with Yunnan pine, we isolated and identif ied ophiostomatalean species associated withTomicusspp. in Yunnan Province. We also evaluated the effect ofOphiostoma canum(the most common fungal associate) onTomicuscolonization in the gallery system in the f ield, tested its pathogenicity and measured tree resistance responses to the fungus inoculation.

O. canumis weak pathogen of Scots pine (P. sylvestrisL.) in Europe and Japanese red pine (P. densif loraSieb. et Zucc.) in Japan. It colonizes tree stems and causes phloem necrosis and xylem blue-stain in both hosts (Solheim et al.2001; Masuya et al. 2003; Jankowiak 2008). Inoculations with a low density ofO. canuminduced a smaller reaction zone than didL. wingfi eldiiM.Morelet andO. minus(Hedgc.) Syd. & P.Syd. (Solheim et al. 2001) and a weaker monoterpene response thanL. wingf ieldiiin the phloem of Scots pine (F?ldt et al. 2006).

Materials a nd methods

Field s ampling

To gain a better knowledge of ophiostomatalean fungi associated withTomicusspecies, we monitored 22 forest areas withTomicusdamage; 19 were severely damaged and 3 lightly damaged as def ined by Lu ( 2011), Zhang ( 2001),He and Zhang ( 2004) (for details, see Table 1 and Fig. 6),which covered the edge of theTomicusdistribution in the province during 2014 and 2016. In each area of about 5 hectares, trees (30—40 years) with clear beetle damage such as withered shoots, entrance holes, oleoresin pitch tubes or frass, were selected for sampling.Tomicusgalleries in the tree stems were dissected and the adults removed with sterilized forceps and placed in separate 1.5-mL Eppendorf(EP) tubes. The phloem around the bark beetle gallery was excised using sterile scalpels and stored in separate disposable plastic bags. Altogether, 630 adults and 1450 phloem samples were collected (Table 1).

Fungal is olation an d identif ication

The species of pine shoot beetles were identif ied based on morphology using a previously described method (Lu 2011). After species identif ication, the beetle was individually squashed with sterilized forceps and placed individually onto the surface of 2% malt extract agar (MEA; 20 g Biolab malt extract, 20 g Biolab agar, 1000 mL deionised water) amended with 0.05% cycloheximide and 0.04% streptomycin. The phloem samples were individually incubated in sterilized Petri dishes with moist f ilter paper at room temperature until spores formed. Single spores were transferred using a sterile needle and a dissecting microscope to 2%MEA and incubated at 20 °C in the dark for 10 days. All the pure cultures isolated from the adult beetles and phloem samples are maintained at the Herbarium of the Laboratory for Conservation and Utilization of Bio-resources, Yunnan University, Kunming, Yunnan, P. R. China (YMF; formerly Key Laboratory of Industrial Microbiology and Fermentation Technology of Yunnan).

T ot al n o. i s o l a t e s D am ag e l ev el b********************************922 26 60 4 22 13 5 22 9 15 26 28 129 22 6 65 7 3 4 N o. is ol at es fr om g a l l e r i e s a Ta a a Tb Ty a Tm 38 31 4 42 8 45 1 9 34 3 4 2 7 14 9 37 57 42 9 3 2 14 12 4 7 17 14 11 64 5 6 rovince 9 6 unnan P. g a l l e r y les No samp 16 56 96 15 78 54 52 52 52 43 96 42 36 45 54 43 g sites in Y a Tb a Ta a 7 3 5 6 4 2 1 2 1 3 32 plin ith Tomicus species at 22 sam 9 3 11 2 1 2 2 2 1 N o. is ol at es fr om b e e t l e Ty a Tm 14 66 No. b e e t l e s a m p l e s 22 42 63 34 33 22 24 20 19 33 18 13 22 27 20 gi associated w ngal species ma. canum hiosto um um L ep to gr ap hi um yu n- n a n e n s e u m O. can O. can um O. ips O. can nanense um um O. can um L. yun O. can um O. can um O. can um O. can O. ips um um um um O. can um O. tingens O. can Fu O. ips O. can O. can O. can O. can O. can atalean fun nanensis P. yunnanensis nanensis Op s yun nanensis nanensis nanensis L) 26°5′48″N, 100°2′12″E P. yun P. yunnanensis P. yunnanensis nanensis P. yunnanensis nanensis nanensis P. yunnanensis P. yunnanensis lates of ophiostom P. yunnanensis Plant host P. yun P. yun P. yun nanensis Pinu P. yun P. yun P. yun P. yun n for iso Latitude, longitude, 99°58′16″E atio, 1 0 3°4 8′1 0″E , 1 0 4°1 5′0 3″E, 1 0 3°2 2′1 6″E , 1 0 3°4 9′1 7″E, 1 0 3°5 9′6 8″E , 1 0 2°6 5′7 7″E , 1 0 3°7 0′8 3″E , 1 0 3°0 7′2 4″E, 1 0 3°3 4′4 5″E , 1 0 4°2 1′3 5″E, 1 0 2°4 6′4 3″E , 1 0 2°5 0′2 5″E , 1 0 0°5 7′3 4″E , 1 0 0°4 6′1 9″E inform) 25°35′32″N 25°29′53″N 23°24′13″N 24°55′16″N 25°05′80″N 25°08′37″N 25°14′62″N 26°13′13″N 25°15′44″N 24°92′58″N 25°16′42″N 26°27′52″N 27°44′40″N XY) 24°43′06″N CH) 24°29′52″N ling ML JY Samp)ongh uang uozh on ile (HH(L long nn shen Table 1 iangyun (DL hanghu (SL g, L g, S g, F Dali, X Z)Qujing (QJ)Wenshan (WS Shilin, C ing, D a i (L L L H)lian lianLS Lijiang,Yu ing, T g, B Site Lu Lu(Llianeishan (L L B S)Lu liananghua (L L F H)Kunmgchuan (K M D C)Honghe, M i ub ei (W S Q B)Wenshan, Q ing, A Lu Kunming (K M A N)Kunmuanjie (K M T J)g (L J Y S)Lijiang,Yong i ng la ng (L J N L)Lijiang, N

T ot al n o. i s o l a t e s D am ag e l ev el b******16 8 10 134 5 1 69 7 N o. is ol at es fr om g a l l e r i e s a Ta a 9 a Tb 13 4 Ty a Tm 14 4 78 5 249 17 9 19. g a l l e r y les No samp 36 67 48 45 88 42 1450 aged a Tb a Ta a 4 N o. is ol at es fr om b e e t l e 4 am Ty a Tm 3 1 12 50% of trees d 39 47 7, ＜aged No. b e e t l e s a m p l e s 16 29 25 17 36 29 630 andii lightly dam ngal species rm aged; *um micus a am f trees d Fu O. can L. sp. 1 O. ips L. sp. 1 O. ips L. sp. 2 s; Ta, To% o 50, ＞aged Plant host P. yunnanensis P. yunnanensis P. yunnanensis andii P. arm P. kesiya P. kesiya micus brevipilosu Latitude, longitude, 99°24′55″E, 99°04′50″E 25°17′30″N 24°57′49″N 01°24′01″E, 98°31′54″E 6″N, 1, 1 0 0°5 3′1 9″E age at site: ** severely dam 23°03′01″N, 1 0 1°0 3′0 6″E 23°11′18″N 26°27′25″N) 25°0′2(continued)ZX)ixi (CXNR MJ)a np in g (N J L P)inor; Ty, Tomicus yunnanensi; Tb, To n of Tomicus dam s m)Table 1 g, L ojiang (PR Chuxiong, Z Nujian Baoshan (BS ong, Y Tengchongle (T C Y L)inger (PR b Visu Puer, N Puer, M Total a Tm Site, Tomicu al evaluatio

The fungal isolates were f irst classif ied into different morphological groups based on culture characteristics and morphology (Paciura et al. 2010) and then fungi were identif ied mainly based on DNA sequencing. ForOphiostomaspecies,ITS rDNA regions (ITS1-5.8S-ITS2) were amplif ied with primers ITS1F/ITS4 (White et al. 1990; Gardes and Bruns 1993), partial β-tubulin gene region (TUB) was amplif ied using primers Bt2a/Bt2b (Glass and Donaldson 1995) and elongation factor 1 α (EF-1α) was amplif ied with primers EF1F/EF2R (Jacobs et al. 2004), exceptO. tingensusing primers NS1/NS4, ITS3/LR3 and BT2E/BT12 to amplify the nuclear small subunit (SSU), nuclear large subunit(LSU) andTUB, respectively (Alamouti et al. 2009; Pan et al. 2017). ForLeptographiumspecies, part of the nuclear ribosomal DNA operon including the internal transcribed spacer 2 region and part of the large subunit (ITS2-LSU)was amplif ied using primers ITS3/LR3 (White et al. 1990).βT and EF-1α were amplif ied as done for theOphiostomaspecies. PCR reactions conditions were according to Grobbelaar et al. ( 2009).

One isolate from each sampling site for each taxonomic group was selected for phylogenetic analyses. The obtained sequences were aligned with reference sequences from Gen-Bankusing ClustalX1.81 (Jeanmougin et al. 1998). Phylogenetic relationships were examined based on two phylogenetic analyses. Bayesian (BI) inference was done using MrBayes 3.1.2 (Ronquist and Huelsenbeck 2003). The appropriate models of sequence substitution were tested in the jModelTest v. 2.1.4 software package (Posada 2008).Four Markov chain Monte Carlo (MCMC) iterations were run from random trees for 5,000,000 generations, and trees were sampled every 1000 generations. The f irst 25% of the trees were discarded as the burn-in phase for each analysis, and the remaining trees were used to calculate posterior probabilities. Maximum likelihood (ML) analysis was conducted using the program RAxML, and the RAxML-HPC BlackBox was selected with default parameters (Stamatakis 2014) (for details on representative isolates, see Tables S1,S2 and Figs. 1, 2, 3, 4, 5). For distinguishing betweenO.canumandO.cf.canum, BI and ML analysis were conducted forO. piceaecomplex (Fig. 2).

Effect of fungal associates on Tomicus colonization in the gallery system in the f ield

To evaluate the possible effect of fungal associates on the colonization of pine shoot beetles, we analyzed the correlation of the isolated fungi inT. minorandT. yunnanensisgalleries with the development of the gallery system in Yunnan pine stands in Suozhuang (SZ, 25°05′N, 103°59′E), Beishan (BS, 25°08′N, 102°65′E) and Fanghua (FH, 25°14′N,103°70′E) in Luliang County. In late April 2016, the peak period of beetle pupation, f ive trees (mean diameter 15 ± 2 cm, height 15 ± 2 m) with beetle entrance holes were felled at each stand (about 80—100 entrance holes per tree).Fifty entrance holes were randomly selected from each tree.Bark around the entrance hole was removed with a knife to completely expose the gallery. The beetles were taken fromeach gallery for species identif ication and fungal associate isolation. The gallery with larvae were recorded as valid and those without larvae as invalid using the method described by Duan ( 2011). As described above, the phloem around each gallery was collected, and fungi were isolated and identif ied. All isolated strains were identif ied asO. canum.

Fig. 1 Bayesian consensus tree for Ophiostoma s. s. species based on ITS1-5.8S-ITS2 data.Support values are indicated at nodes. Bayesian posterior probabilities ＜ 75% and maximum likelihood bootstrap percentages ＜ 75% are indicated by a hyphen (-). Sequences generated in the present study are in bold.T: ex-type

Fig. 2 Bayesian consensus tree of the Ophiostoma piceae complex based on β-tubulin (left) and EF1-α data (right), respectively. Support values are indicated at nodes. Bayesian posterior probabilities ＜ 75% and maximum likelihood bootstrap percentages ＜ 75% are indicated by a hyphen (-). Sequences generated in the present study are in bold.T: ex-type

Pathogenicity o f O. canum on Yunnan pine

Trunks of 10 healthy Yunnan pines (20—30 years old, diameter: 13 ± 3 cm, height: 15 ± 2 m) in a Yunnan pine experimental stand (25°12′N, 102°77′E) in Kunming city in July and August 2016 were inoculated as described earlier (e.g.,Solheim et al. 2001; Viiri et al. 2001; Masuya et al. 2003).Three sterilized agar plugs and three mycelial plugs (O.canum) were placed into separate holes in each trunk made with a 1-cm iron borers in a spiral on the trunk at a height between 0.80 m and 1.60 m. A paired set of holes (with agar plug [control], with mycelial plug) was sampled 10, 20 and 30 days after inoculation, respectively. The vertical length of each reaction zone was measured. A phloem sample within the reaction zone was excised for chemical analysis using a sterilized scalpel and forceps, placed in a 1.5-mL Eppendorf tube and stored in a liquid nitrogen tank. A few extra phloem pieces (12 ± 3 mm × 1.5 ± 0.5 mm) were removed and cultured on 2% MEA to reisolate the fungal strain and conf irm its identity and pathogenicity.

Monoterpene s eparation, i dentif ication and quantif ication

Fig. 3 Bayesian consensus tree for the Ophiostoma ips complex based on β-tubulin (left) and EF1-α data (right), respectively. Support values are indicated at nodes. Bayesian posterior probabilities ＜ 75% and maximum likelihood bootstrap percentages ＜ 75% are indicated by a hyphen (-). Sequences generated in the present study are in bold.T: ex-type

To investigate the chemical response of Yunnan pine toO. canum, we quantif ied the monoterpenes in the phloem reaction zones for the inoculated and the control sites at three times after inoculation. The phloem sample was extracted in 1.5 mL hexane at 20 °C for 48 h. The extract was centrifuged at 5000 rpm for 10 min at 8 °C, then 1 mL supernatant was transferred to a 2-mL sample vial (Agilent, USA). Nonane (99%, Dr. Ehrenstorfer) was used as an internal standard at 0.3 mg/mL hexane. The extracts were stored at - 20 °C until GC—MS analysis. The phloem was dried at 80 °C for 6 h and weighed using an electronic balance (Sartorius) to calculate monoterpene concentration.

An Agilent 6890 gas chromatography combined with a 5973 mass spectrometer and a HP-5MS fused silica capillary column (30 m length, 0.25 mm id, and 0.25 μm f ilm thickness, J&W Scientif ic, USA) was used for the analysis.The temperature program was set at 40 °C, increased to 80 °C at a rate of 3 °C/min, to 280 °C at a rate of 5 °C/min, and then held at 280 °C for 20 min. One microliter of the extract was injected into a split/splitless injector with a 30-s splitless injection at 250 °C. The monoterpene hydrocarbons were identif ied using the Wiley 7n.l and NIST 98.L reference libraries and by comparing retention times and mass spectra with available authenticated standards(Pan 2018).

The absolute amount of each monoterpene hydrocarbon (mx ) was calculated using the equation (Pan 2018):n=k(Ax /As )(ms )/Dw , (wherekis the relative response factor,Ax is the peak area of a monoterpene hydrocarbon in the sample,As is the peak area of nonane,ms is the mass of nonane andDw is the dry mass of the sample. The relative response factor (k) was calculated using the equationk=Asmi /Aims , where A s is the peak area of nonane,Ai is the peak area of a monoterpene standard,ms is the mass of nonane,mi is the mass of the monoterpene standard).

Statistical anal yses

The proportion of valid galleries with fungus versus the proportion without the fungus and absolute amount of monoterpene hydrocarbons after fungal inoculation versus control treatment at each time were compared using an independentsamplettest. The lesion length in phloem, absolute amounts of monoterpene hydrocarbons among different times after fungal inoculation versus control treatment were analyzed using one-way analysis of variance (ANOVA) test followed by Bonferroni test. All analyses were conducted with SPSS 17.0 (SPSS Inc., Chicago, IL, USA), and graphs were generated using Origin 8.5 (Origin Lab Corp., Northampton,MA, USA).

Results

Fungal ide ntif ication and DNA sequence comparisons

A total of 567 fungal isolates were collected. Most were from the beetles and galleries ofT. yunnannesisandT. minorinP. yunnannesis, and a small number of isolates were from those ofT. brevipilosusandT. armandiiinP. kesiyaandP.armandii(Table 1).

Fig. 4 Bayesian consensus tree of Ophiostoma tingens and its related species generated using the combined sequences for nSSU, nITS2-LSU, and β-tubulin. Support values are indicated at nodes. Bayesian posterior probabilities and maximum likelihood bootstrap percent- ages ＜ 75% are indicated by a hyphen (-). Sequences generated in the present study are in bold. T: ex-type. A. = Ambrosiella, R. = Raffaelea, G. = Grosmannia, D. = Dryadomyces, C. = Ceratocystiopsis,C. = Ceratocystis, M = Microascus

The DNA sequence analysis revealed that the isolates belonged toOphiostoma sl.orLeptographium s.l.(Figs. 1, 2, 3, 4, 5), and six species were identified (Tables S1, S2). WithinOphiostoma s.l., the isolates belonged toO. piceaeorO. ipsspecies complexes (Fig. 1); 17 isolates were identified asO. canum(taxon 1) withinO.piceaecomplex, and four were identified asO. ips(taxon 2) (Fig. 1). The TUB and EF1-α sequences analysis confirmed that all the obtained isolates in these complexes belonged to two known species (Figs. 2, 3). Additionally, based on the phylogenetic analysis of SSU + ITS2-LSU + βT, one isolate was identified asO. tingens(taxon 3) (Fig. 4).

The comparisons of the ITS2-LSU + βT + EF-1α sequences obtained for theLeptographiumisolates showed that our isolates belonged to theL. lundbergiicomplex (Fig. 5). Taxon 4 grouped in the same clade withL. yunnanense, while taxa 5 and 6 represented different clades most closely related toL. yunnanense,L. conjunctumPaciura, Z.W. de Beer & M.J.Wingf. andL. wushanenseY.Pan, J.Lu, & H.Ye (Fig. 5).

Ophiostomatalean f ungi i n relation t o tree s pecies and damage level

Six ophiostomatalean species (O. canum,O. ips,O.tingens,L. yunnanense,Leptographiumsp. 1 andLeptographiumsp. 2) were identif ied from 22 sampling sites infested byTomicus(Table 1).O. canumwas the species most commonly associated with pine shoot beetles in Yunnan, which was isolated from 17 of the 19 investigated Yunnan pine stands (Fig. 6, Table 1).O. canumandL.yunnanensiewere mainly obtained fromP. yunnanensis.In contrast,O. ipsand two unknownLeptographiumspecies (taxon 5 and 6) were isolated fromPinus kesiyaandPinus armandii. More ophiostomatalean isolates were obtained from severely damaged areas.O. canumwas isolated mainly from 16 Yunnan pine stands heavily attacked byTomicus, but not from the two stands with low beetle damage (Table 1).

Fig. 5 Bayesian consensus tree of the Leptographium lundbergii complex generated from the DNA sequences of the ribosomal ITS2-LSU regions, combined with β-tubulin and EF 1-α. Support values are indicated at nodes. Bayesian posterior probabilities ≤ 75% are indicated by a hyphen (-). Sequences generated in the present study are in bold

Effect of O. canum on Tomicus colonization in the gallery system in the f ield

The percentage of valid galleries was signif icantly higher for galleries withO. canumthan those without the fungus, for bothTomicusspecies (Table 2). ForT. minor, the percentages of valid gallery withO. canumwere 85.71%,77.78% and 77.27%, respectively, at the three experimental stands, and were signif icantly higher than in the galleries without the fungus (64.29%, 42.11% and 27.27%, respectively;t= 3.216,p＜ 0.05). A similar pattern was discovered forT. yunnanensis(t= 3.317,p＜ 0.05).

Pathogenicity t est a nd monoterpene l evels

The lesion length in phloem from fungal inoculated trunks was signif icantly longer than in controls treatment at all three sampling times (10 days:t= 3.171,p＜ 0.01; 20 days:t= 6.513,p＜ 0.01; and 30 days:t= 7.156,p＜ 0.01), suggesting thatO. canumstimulated tree defense response and might be pathogenic to Yunnan pine. Interestingly, the lesion length increased gradually over time after fungal inoculation(F2,27 = 8.916,p＜ 0.01) but not in the control (F2,27 = 0.536,p＞ 0.05), suggesting gradual growth of the fungus in the tree phloem during the sampling period (Fig. 7).

Fig. 6 Geographic distribution of sampling sites in Yunnan Province,China. Solid circles: Ophiostoma canum isolated; open circles: O.canum not isolated

Table 2 Impact of blue-stain fungus Ophiostoma canum on number of galleries with larvae (valid) formed by Tomicus yunnanensis and T.minor at three Pinus yunnanensis stands in Yunnan Province

Six monoterpenes were detected: α-pinene, camphene,β-pinene, β-myrcene, β-phellandrene and α-terpinolene(Fig. 8). The absolute amount of α-pinene, camphene and β-myrcene in the phloem reaction zone afterO. canuminoculation was significantly higher than after mechanical wounding 30 days after treatment (α-pinene:t= 3.004,p＜ 0.05; camphene:t= 2.629,p＜ 0.05; β-myrcene:t= 2.589,p＜ 0.05). However, no signif icant difference for any individual monoterpene hydrocarbon was observed among sampling times in any treatments.

Fig. 7 Mean (± SE) length of phloem reaction zone in Pinus yunnanensis trunk at three times after inoculation with a mycelial plug of Ophiostoma canum (black bars) or sterilized agar (grey bars).**Lesion lengths differed signif icantly between O. canum and control for a specif ic time ( n = 10); different capitals and lowercase letters indicate signif icant difference among time points within the same treatment at p = 0.01 and 0.05 level, respectively

Discussion

In this study, we isolated and identif ied six fungal associates ofTomicus:O. canum,O. ips,O. tingens,L. yunnanensis,Leptographiumsp. 1 andLeptographiumsp. 2,representing an increase in the diversity of fungal species associated withTomicusin southwestern China. Interestingly, our study showed thatO. canumwas the main fungal species associated withTomicusin Yunnan, whereas Chang et al. ( 2017) did not isolate this fungus. The difference is probably due to the investigation time: their major investigation onP. yunnanensis(the main host of

O. canum) was done in 2001 and 2002, 13—14 years earlier than ours. Therefore, we speculate thatO. canummight be a new invasive species. Our population genetic data about the fungus seemed to support this speculation (Pan 2018).Ophiostoma ipshas been reported to be associated with the bark beetlesIps grandicollis(Eichhoff),Hylurgus ligniperda(Fabricius),Ips sexdentatusBoerner,Orthotomicus erosus(Wollaston) and others that infest pines such asP. radiataD. Don,P. caribaeaMorelet,P. taedaL.,

P. elliottiiEngelm.,P. sylvestrisin Asia, Europe, North America, South America and Africa (Zhou et al. 2007,2015; Jankowiak 2012).Ophiostoma tingenshas been reported as associated with bark beetles in genera ofIps

andTomicus(Batra 1967, Harrington et al. 2010) and isolated from the galleries ofT. minorinP. yunnanensis(Pan et al. 2017). Our f inding that more ophiostomatalean fungiwere isolated from severely damaged areas agrees with the results of Jankowiak and Bilanski ( 2007).

Fig. 8 Mean (± SE) absolute amounts of monoterpenes in the phloem reaction zones of Pinus yunnanensis at three sampling times after inoculation with a mycelial plug of Ophiostoma canum (solid line) or sterilized agar (dotted line). n = 10

O. canumwas the species most frequently associated with pine shoot beetles in Yunnan pine stands, suggesting it may have a close relationship withTomicusinfestingP. yunnanensis.O. canumhas been reported from many countries in Europe, Australia and Asia (Münch 1907;Mathiesen 1950; Rennerfelt 1950; Masuya et al. 1999;Zipfel et al. 2006; Jankowiak 2008; Linnakoski et al.2010; Goldazarena et al. 2012; Davydenko et al. 2014;Wang et al. 2014). This study shows that the fungus is also widely distributed in Yunnan, China and thus represents an increase in the known geographical distribution of this fungus. Many studies have demonstrated thatO.canumis closely associated withT. minorandT. piniperda(Francke-Grosmann 1952; Mathiesen-K??rik 1953; Masuya et al. 1999; Jankowiak 2008; Linnakoski et al. 2010).Our results showed that this fungus was associated with three pine shoot beetles in the province. Except for pine shoot beetles,O. canumis also associated with other bark beetles, such asHylurgops palliatusGyll. andTrypodendron lineatum(Oliv.) (Linnakoski et al. 2010). Thus,O.canumseems to be a common fungal associate of many bark beetle species. Pinus sylvestrisandP. densif loraare previously reported as the major hosts ofO. canum(Solheim et al. 2001; Masuya et al. 2003). Here we isolated the fungus from beetles and galleries in Yunnan pine. In an inoculation experiment,O. canumcolonizedP. kesiyaandP. armandiiseedling (data not shown), suggesting that the fungi may be a pathogen of many pine species.

For bothTomicusspecies, our f ield investigation clearly showed that more galleries withO. canumhad larvae than when the fungus was absent from the galleries, suggesting the presence ofO. canumenhance the formation of valid galleries forT. yunnanensisandT. minor. The formation of valid galleries is an important embody of the successful invasion of the bark beetles (Duan 2011). Our observation suggests thatO. canummay promote the successful colonization of the beetles in Yunnan pine stands. The observation that this fungi is a food source forT. minorlarvae which is benef icial to the brood development of the beetle (Francke-Grosmann 1952) further supports our argument. However,more study is needed to reveal the effect ofO. canumonTomicusperformance.

The reaction length in tree phloem after fungal inoculation has long been used as a measurement for fungal virulence or tree resistance (Krokene and Solheim 1999; Zhao et al. 2011). In our study,O. canuminoculation induced signif icantly longer lesion length compared mechanical wounding, indicating that the fungus might be pathogenic to Yunnan pine. This observation is consistent with previous results from Scots pine in Europe and Japanese red pine in Japan (Solheim et al. 2001; Masuya et al. 2003; Jankowiak 2008). The pathogenicity of blue-stain fungi such as ofO.canumis thought to help the bark beetle to overcome tree resistance (Lieutier et al. 2009) or mediate competitive interactions among fungi (Six and Wingf ield 2011).

Monoterpenes play a critical role in conifer resistance against bark beetles and their associated fungi (Gershenzon and Dudareva 2007). Some of these monoterpenes are toxic to insects and fungal pathogens and can repel beetle invasion, increase mortality, and inhibit fungal germination and mycelia growth (Bordasch and Berryman 1977; Delorme and Lieutier 1990; Raffa and Smalley 1995; Joseph et al.1998). Some studies have shown that the concentration of monoterpenes such as camphene, β-pinene and myrcene increase after fungal inoculation (Raffa and Berryman 1982;Gershenzon and Croteau 1991). Accumulation of monoterpenes and other defense chemicals has long been considered as an indicator of tree defense responses (Krokene and Solheim 1999; F?ldt et al. 2006). In this study, the levels of α-pinene, camphene and β-myrcene 30 days after the fungal treatment were signif icantly higher than after the control treatment, suggesting thatO. canuminoculation induced monoterpene accumulation in trees.

Based on the present results and previous reports, at least 10 fungi are associated withTomicusin southwestern China. The most common fungal species in the present study,O. canum, induced a defense response in trees and enhanced the formation of valid galleries and thus might be benef icial to tree colonization by the beetles. However, since inoculation ofO. canumonly induced a mild reaction zone and monoterpene accumulation, the damage ofO. canumto Yunnan might be limited. Integrated studies with all the fungal associates are needed to understand the potential role of blue-stain fungi in bark beetle damage toTomicus.