Fungal Communities Associated with Bark Beetles in Pinus radiata Plantations in Northern Spain Affected by Pine Pitch Canker , with Special Focus on Fusarium Species

Fusarium spp., as well as other endophytic or pathogenic fungi that form communities, have been reported to be phoretically associated with bark beetles (Coleoptera; Scolytinae) worldwide. This applies to Fusarium circinatum Nirenberg and O’Donnell, the causal agent of pine pitch canker (PPC), which threatens Pinus radiata D. Don plantations in northern Spain. The main objective of this study was to study the fungal communities associated with bark beetles and their galleries in stands affected by PPC, with special attention given to Fusarium species. Funnel traps and logs were placed in a P. radiata plot known to be affected by F. circinatum. The traps were baited with different attractants: four with (E)-pityol and six with ethanol and α-Pinene. In addition, fresh green shoots with Tomicus piniperda L. feeding galleries were collected from the ground in 25 P. radiata plots affected by PPC. Extracts of whole insects and gallery tissues were plated on agar medium to isolate and identify the associated fungi. A total of 24 different fungal species were isolated from the bark beetle galleries constructed in logs and shoots, while 18 were isolated from the insect exoskeletons. Ten different Fusarium species were isolated from tissue and insects. Fusarium circinatum was isolated from bark beetle exoskeletons (1.05% of the Pityophthorus pubescens Marsham specimens harboured F. circinatum) and from the galleries (3.5% of the T. piniperda feeding galleries harboured the pathogen). The findings provide information about the fungal communities associated with bark beetles in P. radiata stands in northern Spain.


Introduction
Fungal endophyte species form fungal communities, together with saprotrophic and pathogenic species, in forests.Knowledge about the species composition and the factors influencing the presence of Forests 2018, 9, 698 2 of 20 different fungal communities is important for understanding the role that fungi play in regulating other organisms [1].Bark beetles are known to be closely associated with fungi worldwide.They are well known for the associations that they form with endophytic, plant-pathogenic and entomopathogenic fungi, such as Fusarium species, which are widespread and abundant in living and dead plants [2].
The genus Fusarium includes important plant pathogens that affect both forest and agricultural species [3] by producing different types of wall-degrading enzymes (e.g., cellulases, glucanases and glucosidases) and mycotoxins such as beauvericin and fumonisins [4,5].Some of the species in this genus, such as the Fusarium oxysporum spp.complex, cause disease symptoms in a large number of vegetable crops [6] and forest trees.The pathogenic fungi Fusarium circinatum Nirenberg and O'Donnell (teleomorph = Gibberella circinata) is an ascomycete fungus that causes pine pitch canker (PPC) [7].The main symptom of this disease is the presence of pitch-soaked cankers which can girdle both trunks and large branches in adult trees [8].Fusarium circinatum is currently threatening pine plantations and natural stands throughout the world [9], especially Pinus radiata D. Don, a highly susceptible pine species [10].Up to 60 Pinus species and Pseudotsuga menziesii (Mirb.)Franco [11,12] are also susceptible to the pathogen.In Spain, the disease caused by F. circinatum was reported for the first time in 2005 [13] and has led to important ecological and economic losses in forest plantations and nurseries.Regarding the interaction between F. circinatum and other Fusarium spp., introduction of the non-pathogenic Fusarium lateritium Nees as a potential biocontrol agent was found to inhibit the pathogen growth when introduced as a pioneer [14].
The fungal communities that inhabit P. radiata trees may strongly influence the distribution of PPC in Spain.Endophytic species which do not cause damage to the host [1], such as Trichoderma viride Bissett, could be used for biological control of Fusarium spp.[15][16][17].The antagonistic properties of species of the genus Trichoderma are mediated in five different ways: antibiotic production, competition, production of enzymes (chitinases and/or glucanases), induction of host plant defence mechanisms and parasitism [18].The last of these occurs when the hyphae of Trichoderma coil round the hyphae of other fungi and eventually penetrate the tissues [19].Other fungal species like Diplodia sapinea (Desm.)Kickx, which is a latent pathogen in pine trees, have been associated with PPC in P. radiata trees, causing shoot dieback and the presence of resin drops and necrotic stem lesions [20].Some bark beetles have been reported to be phoretically associated with F. circinatum in P. radiata plantations in northern Spain: Pityophthorus pubescens (Marsh.),Ips sexdentatus (Börner) and Tomicus piniperda L. [21][22][23].In California, the importance of Pityophthorus spp. as the main vectors of F. circinatum has also been demonstrated [24], although in northern Spain P. pubescens was found to be only weakly associated with the pathogen [23].Differences in bioecology may determine spread of fungal infections; for example, Hylastes spp.feed on roots or trunks of declining trees whereas T. piniperda feeds on the shoots of healthy crowns [25].Moreover, there is a risk that bark beetle populations may increase from endemic to epidemic levels [26].For example, I. sexdentatus populations may increase as a result of forest disturbances, leading to subsequent attacks on healthy trees [25].
We hypothesize that different bark beetle species living in these plantations play specific roles in the spread of Fusarium spp.Moreover, the fungal composition of the associated mycobiota may determine the distribution of the pathogens.The main objective of this study was to characterize the fungal communities associated with bark beetles and their galleries in stands affected by PPC, with special focus on Fusarium species.

Sample Collection
This study was carried out in a P. radiata plot affected by F. circinatum, located in Vejorís (Cantabria, Spain).Two types of traps were set up in the plots: piles of bait logs and funnel traps.The bait logs consisted of 6 piles of branches (diameter: 6.4-16.9cm) and 6 piles of thick logs from trunks (diameter: 16-31 cm), all obtained from healthy P. radiata trees collected from a F. circinatum-free plot.Plant tissues (xylem and phloem) and insects were collected from the breeding galleries weekly between June and October 2010.
In addition, four funnel traps baited with (E)-pityol and six traps baited with ethanol-α-Pinene (Econex) were placed within the plantation.(E)-pityol, an aggregation pheromone produced by P. pubescens, attracts both males and female beetles [27].Insects were collected weekly between June and October 2010.
To examine the association between T. piniperda and F. circinatum, fresh fallen green shoots with T. piniperda feeding galleries were collected from 25 P. radiata plots affected by F. circinatum.A total of 285 fallen P. radiata shoots with a T. piniperda entrance hole and feeding gallery were collected from the ground between June and October 2010.Twenty-seven shoots were collected during the summer (from June to August) and 258 were collected in autumn (September and October).

Molecular and Morphological Identification of Fungi
Shoots and xylem and phloem from beetle galleries in logs were plated on PDAS (potato dextrose agar with 0.3 g/L of streptomycin sulfate) culture medium previously treated by surface sterilization (1 min tap water, 1 min ethanol 70%, 1 min sodium hypochlorite 20% and 1 min distilled sterilized water).In addition, a total of 438 bark beetles collected from logs and funnel traps belonging to 11 different species (Table 1) were plated by pressing the whole bodies onto Fusarium selective media (FSM: 15 g bactone peptone, 1 g KH 2 PO 4 monobasic, 0.5 g MgSO 4 -7H 2 O, 20 g agar, 0.2 g of pentachloronitrobenzene (PCNB) and 0.3 g streptomycin sulfate per liter) to avoid bacterial and soil fungi contamination.All specimens were processed in this way, except those collected from the shoots, which were used for molecular identification.Fungi isolated from plant material and insects were classified into morphological units, i.e., colonial morphotypes (CMs), on the basis of cultural characteristics [28].Thus, fungi obtained from plant tissues were grouped into 30 CMs according to macromorphological features of the colony growing on the PDAS.Fungi derived from insects and isolated on FSM were grouped into 17 different CMs.One isolate from each CM was selected for molecular identification.However, 29 isolates of reddish, orange, yellowish, violet or pinkish colonies, which probably belonged to the genus Fusarium, were selected.Regarding the identification of F. circinatum, 16 isolates were selected on the basis of their macroscopic features for specific molecular identification.Single hyphae cultures were grown prior to molecular identification.
DNA extraction was carried out on the fungal culture following the protocol described by Vainio et al. [29].Once the DNA was extracted, the internal transcribed spacer (ITS) region of the rDNA was amplified by polymerase chain reaction (PCR), with primers ITS-1F (5 -CTTGGTCATTTAGAGGAAGTAA-3 ) and ITS-4 (5 -TCCTCCGCTTATTGATATGC-3 ) [30].The As the ITS region is not a suitable molecular marker for identifying Fusarium spp. to species level, microscopic morphological and morphometric identification was carried out together with molecular identification using specific molecular markers.Thus, 29 isolated fungi belonging to the genus Fusarium, according to the ITS region, were plated on Spezieller Nahrstoffarmer Agar (SNA) and Carnation Leaf Agar (CLA), both specifically used for identification of Fusarium species [31].After 10 to 20 days, SNA diagnostic characters including the shape of the macroconidia, the presence or absence of microconidia, the shape and mode of aggregation of microconidia, the shape of conidiogenous cells, and the presence or absence of coiled sterile hyphae and chlamydospores were observed.The colour and size of the sporodochia were observed on CLA.These samples were amplified with the primers EF1 (forward primer; 5 -ATGGGTAAGGA(A/G)GACAAGAC-3 ) and EF2 (reverse primer; 5 -GGA(G/A)GTACCAGT(G/C)ATCATGTT-3 ) [32] for sequencing the translation elongation factor (α-TEF) region, which encodes an essential part of the protein translation machinery and is highly informative at the species level in Fusarium [33].PCR reactions were conducted in 50 µL volumes containing 1 µL template DNA, 1× reaction buffer, 200 mM dNTPs, 0.4 µM forward and reverse primers, 1U of Kapa Taq DNA polymerase.The thermal profile of PCR was one cycle of 10 min at 94 • C followed by 36 cycles of 30 s at 94 • C, 55 s at 62 • C and 1 min at 72 • C and a final 10 min extension at 72 • C (modified from Pérez-Sierra et al. [34]).The α-TEF region was sequenced and the sequences were corrected with Geneious Pro Software (V 5.6.5) and blasted in the GenBank and FUSARIUM ID sequence databases.
Multiple sequence alignment (ClustalW) of the α-TEF region sequences and evolutionary analyses were conducted with MEGA 6 [35].The phylogenetic tree was constructed according to the neighbor-joining statistical method [36].The bootstrap method (1000 replicates) was used to represent the phylogenetic history of the taxa analyzed [37].The evolutionary distances were computed using the Kimura 2-parameter method [38].The rate of variation between sites was modeled with the gamma distribution (Shape parameter = 1).All positions containing gaps and missing data were eliminated.Fusarium equiseti (Corda) was used as a outgroup (GenBank accession number: AJ543571.1).

Molecular Identification of Insects
Tomicus destruens (Woll.)and T. piniperda are morphologically difficult to distinguish, although there are some differences between their life cycles [33,34].In order to confirm the identification, 17 specimens randomly collected from the feeding galleries were sent to the Department of Animal Biology (University of Murcia) for molecular identification following the protocol described by Gallego and Galian [40].

Statistical Analysis
Analysis of variance (ANOVA) and multiple comparison procedures were used to test the effect of insect species from the insect sampling and to examine how fungal species richness was affected by the season of shoot sampling.Robust methods were applied as the data did not fulfil two of the ANOVA assumptions (normality and homoscedasticity) [41].Specifically, one-way fixed-effects ANOVAs were performed under the assumption of non-normality and inequality of variances using the generalized Welch procedure and a 0.2 trimmed mean transformation.The ANOVAs were carried out using the 'Wilcox' Robust Statistics (WRS)' package (Version 2014 [42]) implemented in the R software environment (V 3.4.3),R Foundation for Statistical Computing, Vienna, Austria).A paired-sample Wilcoxon test was carried out to determine whether fungal species richness varied according to the tissue surveyed, particularly xylem and phloem.Pearson's correlation coefficient was calculated to examine the association between the frequency of appearance of T. harzianum and F. circinatum, in those samples that were positive for the pathogen.
Non-metric multidimensional scaling (NMDS) and the multiple response permutation procedure (MRPP) were conducted with the VEGAN package (Version 2015) [43] implemented in the R software environment in order to analyze the fungal communities associated with (1) the insect bodies, (2) the logs, depending on the tissue and the insect species, and (3) the shoots, depending on the season.NMDS was conducted using Bray-Curtis as the distance metric and the multivariate ordination was created using the metaMDS results.MRPP was also performed using Bray-Curtis dissimilarity with 1000 permutations.

Results
A total of 24 different fungal species were obtained from the bark beetle galleries constructed in logs and shoots, and 18 fungal species were obtained from the insect exoskeletons.The bark beetle species collected during the sampling are listed in Table 1.
As F. circinatum and T. harzianum were negatively correlated (r = −0.8143875,p < 0.1), higher frequencies of T. harzianum corresponded to the lowest frequencies of F. circinatum.
The prefix M indicates unidentified species.

Fungal Communities from Insect Exoskeletons
Eighteen fungal species were obtained from the insect exoskeletons.Six Fusarium spp., Candida fructus Nakase Neonectria radicicola (Gerlach and L. Nilsson) Mantiri and Samuels, Penicillium sp., T. atroviride and G. roseum were obtained, together with seven other species that remained unidentified (Table 3).

Fungal Communities from Insect Exoskeletons
Eighteen fungal species were obtained from the insect exoskeletons.Six Fusarium spp., Candida fructus Nakase Neonectria radicicola (Gerlach and L. Nilsson) Mantiri and Samuels, Penicillium sp., T. atroviride and G. roseum were obtained, together with seven other species that remained unidentified (Table 3).

Fungal Communities from Insect Exoskeletons
Eighteen fungal species were obtained from the insect exoskeletons.Six Fusarium spp., Candida fructus Nakase Neonectria radicicola (Gerlach and L. Nilsson) Mantiri and Samuels, Penicillium sp., T. atroviride and G. roseum were obtained, together with seven other species that remained unidentified (Table 3).The species richness on the insect exoskeletons differed significantly depending on the insect species (F We = 4.8, p < 0.001) (Figure 3).Species richness was highest for X. saxaseni Ratzenburg and P. pubescens and was significantly different from the values associated with I. sexdenatus and C. mediterraneus.The fungal communities present on the insect bodies were clustered in two different groups depending on the insect species (Figure 4): one group was associated with Hylastes spp.and another one with P. pubescens.The species richness on the insect exoskeletons differed significantly depending on the insect species (FWe = 4.8, p < 0.001) (Figure 3).Species richness was highest for X. saxaseni Ratzenburg and P. pubescens and was significantly different from the values associated with I. sexdenatus and C. mediterraneus.The fungal communities present on the insect bodies were clustered in two different groups depending on the insect species (Figure 4): one group was associated with Hylastes spp.and another one with P. pubescens.Fungal communities from the insects collected from logs differed significantly (A = 0.3538, p < 0.001) from those associated with log tissues.Thus, two distinct groups of fungi were observed: one  The species richness on the insect exoskeletons differed significantly depending on the insect species (FWe = 4.8, p < 0.001) (Figure 3).Species richness was highest for X. saxaseni Ratzenburg and P. pubescens and was significantly different from the values associated with I. sexdenatus and C. mediterraneus.The fungal communities present on the insect bodies were clustered in two different groups depending on the insect species (Figure 4): one group was associated with Hylastes spp.and another one with P. pubescens.Fungal communities from the insects collected from logs differed significantly (A = 0.3538, p < 0.001) from those associated with log tissues.Thus, two distinct groups of fungi were observed: one Fungal communities from the insects collected from logs differed significantly (A = 0.3538, p < 0.001) from those associated with log tissues.Thus, two distinct groups of fungi were observed: one associated with the xylem and phloem and the other related to the insect exoskeletons.Fusarium spp.were associated with both types of samples although they appeared to be more closely related to the insect exoskeletons (Figure 5).were associated with both types of samples although they appeared to be more closely related to the insect exoskeletons (Figure 5).

Fusarium spp. from Insect Galleries
Fusarium species were isolated from both xylem and phloem from logs.A total of 11 isolates were identified: two belonged to the F. oxysporum species complex, one was identified as Fusarium sporotrichioides Sherbakoff, two were identified as Fusarium beomiforme (Nelson, Toussoun and Burgess), three belonged to Fusarium avenaceum (Fries) Saccardo, and finally three were identified as to F. circinatum (Tables 2 and 4).Fusarium circinatum appeared in 0.85% and 0.43% of the phloem and xylem samples, respectively.In the Hylastes angustatus (Herbst) galleries, F. circinatum appeared in 5.2% and 5.5% of the xylem and phloem samples respectively, whereas F. circinatum was found in only 1.36% of the phloem samples from the H. attenuatus galleries (Table 2).

Fusarium spp. from Insect Galleries
Fusarium species were isolated from both xylem and phloem from logs.A total of 11 isolates were identified: two belonged to the F. oxysporum species complex, one was identified as Fusarium sporotrichioides Sherbakoff, two were identified as Fusarium beomiforme (Nelson, Toussoun and Burgess), three belonged to Fusarium avenaceum (Fries) Saccardo, and finally three were identified as to F. circinatum (Tables 2 and 4).Fusarium circinatum appeared in 0.85% and 0.43% of the phloem and xylem samples, respectively.In the Hylastes angustatus (Herbst) galleries, F. circinatum appeared in 5.2% and 5.5% of the xylem and phloem samples respectively, whereas F. circinatum was found in only 1.36% of the phloem samples from the H. attenuatus galleries (Table 2).In the T. piniperda shoot feeding galleries, Fusarium species were isolated from 21.05% of the shoots collected (Table 2).Eighteen isolates were found to correspond to five different Fusarium species: one was identified as Fusarium cortaderiae (=Fusarium graminearum clade), one as F. sporotrichioides, three as F. avenaceum, three as Fusarium tricinctum (Corda) Saccardo (Table 4) and 10 as F. circinatum, corresponding to 3.5% of the feeding galleries in shoots (Tables 2 and 4).

Fusarium spp. on Insect Exoskeletons
A high percentage of Fusarium spp.appeared on the insect exoskeletons, showing that the bark beetle species usually carried at least one Fusarium species (Table 3).In some cases, one insect species was associated with more than one Fusarium species (Table 4).A total of seven Fusarium species were identified: four isolates were identified as F. oxysporum spp.complex, three isolates as F. circinatum, one as Fusarium anthophilum (A.Braun) Wollenw, six as F. avenaceum, one as Fusarium sambucinum Fuckel, one as F. tricinctum and one as Fusarium konzum.Of those isolates identified as F. circinatum, one was isolated from H. attenuatus collected from logs, one from P. pubescens, and one from I. sexdentatus.
Phylogenetic analysis of the α-TEF region indicated that all Fusarium species identified were clustered together (Figure 6) regardless of the type of sample (insect, logs or shoots).Moreover, the results of the sequence analysis support the molecular identification as the isolates that were identified as the same species clustered together.In the T. piniperda shoot feeding galleries, Fusarium species were isolated from 21.05% of the shoots collected (Table 2).Eighteen isolates were found to correspond to five different Fusarium species: one was identified as Fusarium cortaderiae (= Fusarium graminearum clade), one as F. sporotrichioides, three as F. avenaceum, three as Fusarium tricinctum (Corda) Saccardo (Table 4) and 10 as F. circinatum, corresponding to 3.5% of the feeding galleries in shoots (Tables 2 and 4).

Fusarium spp. on Insect Exoskeletons
A high percentage of Fusarium spp.appeared on the insect exoskeletons, showing that the bark beetle species usually carried at least one Fusarium species (Table 3).In some cases, one insect species was associated with more than one Fusarium species (Table 4).A total of seven Fusarium species were identified: four isolates were identified as F. oxysporum spp.complex, three isolates as F. circinatum, one as Fusarium anthophilum (A.Braun) Wollenw, six as F. avenaceum, one as Fusarium sambucinum Fuckel, one as F. tricinctum and one as Fusarium konzum.Of those isolates identified as F. circinatum, one was isolated from H. attenuatus collected from logs, one from P. pubescens, and one from I. sexdentatus.
Phylogenetic analysis of the α-TEF region indicated that all Fusarium species identified were clustered together (Figure 6) regardless of the type of sample (insect, logs or shoots).Moreover, the results of the sequence analysis support the molecular identification as the isolates that were identified as the same species clustered together.

Discussion
Amongst the fungal species isolated from the shoots and log tissues, we can highlight the presence of D. sapinea, Penicillium spp., Pestalotiopsis spp., Trichoderma spp.and Fusarium spp.for their pathogenicity, parasitism and potential as biocontrol agents.Diplodia sapinea was the species most frequently isolated in the plant tissue samples, although it did not appear on the insect bodies.This

Discussion
Amongst the fungal species isolated from the shoots and log tissues, we can highlight the presence of D. sapinea, Penicillium spp., Pestalotiopsis spp., Trichoderma spp.and Fusarium spp.for their pathogenicity, parasitism and potential as biocontrol agents.Diplodia sapinea was the species most frequently isolated in the plant tissue samples, although it did not appear on the insect bodies.This fungus, which is responsible for shoot blight and dieback on pine trees, has previously been found in association with bark beetles [44].Penicillium spp.have previously been isolated as saprotrophs in pine species, although they may also occur as endophytes in healthy tissues [45].Four Penicillium species (P.glabrum, P. Penicillium polonicum K.M. Zalessk, Penicillium minioluteum Dierckx and Penicillium melinii Thom) were found to be associated with healthy Pinus sylvestriss L. twigs in Spain [46].However in the present study, only one species of this genus was found in plant tissues.Pestalotiopsis is a ubiquitous genus that acts as an endophyte, saprotroph and pathogen in different hosts worldwide [47].It has also been found in healthy tissues of pine species in Spain [45], although some species such as Pestalotiopsis funerea can cause damping off in other conifers [48].Four CMs were identified as Trichoderma spp. in this study, but two of them could not be identified to species level from the ITS region.Trichoderma harzianum, T. atroviride and T. viride have been proposed as effective biological control agents for pathogenic Fusarium spp.[3,49], indicating the importance of determining the endophytic species that inhabit P. radiata trees.Our findings seem to indicate a negative correlation between T. harzianum and F. circinatum.Thus, T. harzianum reached a maximum of 52.7% of the phloem samples from I. sexdentatus galleries, while F. circinatum occurred in only 0.9% of such samples (Table 2).In addition, the higher rates of occurrence of F. circinatum in plant tissue (3.5% in shoots and 5.2%-5.5% in H. angustatus galleries) were associated with lower rates of occurrence of T. harzianum (11.9%, 0% and 10.5%, respectively).Although in the present study T. atroviride did not appear in high proportions, this fungus has been shown to parasitize a large variety of phytopathogenic fungi due to the production of hydrolytic enzymes, which has led to its use as a biological control agent [50].
Regarding the fungal communities associated with T. piniperda-colonized shoots, D. sapinea was the species most frequently isolated, as it appeared in 88.4% of the samples.Seasonal variation in the fungal species isolated from shoots was observed, and D. sapinea was more frequently found in samples collected during autumn.Fusarium spp.also appeared more frequently in those plots sampled during autumn.This is consistent with findings of a previous study [22], in which we isolated higher percentages of F. circinatum during autumn and winter than during the rest of the year.
Bark beetles are known to be associated with endophytic, phytopathogenic and entomopathogenic fungi.In this study, 18 species were isolated from insect bodies.Among these, the ascomycete yeast Candida fructus was identified in nine insect species.Candida species have been reported to be associated with several bark beetle species as entomopathogens, although many yeasts appear in symbiosis with bark beetles [51].Trichoderma atroviride appeared on the bodies of six different insect species, mainly on C. mediterraneus (8% of the samples), coinciding with the lower rate of appearance of Fusarium spp.(12% of the samples).In order to evaluate the importance of Trichoderma spp. as an antagonist of Fusarium spp., more detailed studies of the role of bark beetles are required.Other fungi, such as Neonectria radicicola, were isolated from four bark beetle species.In Norway, this fungal species has been observed as an endophyte in pine roots without causing any damage [52].However, other species of the genus Neonectria are known to cause neonectria canker disease on subalpine fir in Denmark [53] and stem cankers on P. radiata in Chile [54].The species richness and fungal communities differed according to the bark beetle species.This difference may be due to the presence of specialized structures for carrying spores such as mycangia on some species (e.g., H. ater, I. sexdentatus and Xyleborinus saxesenii).The fungal communities associated with the insect exoskeletons and their galleries also differed significantly, although Fusarium spp.appeared to be present in both types of samples.This may indicate the importance of bark beetles in spreading Fusarium species.
Ten species of Fusarium were identified in this study by both molecular and morphological methods.These species were associated with the insect exoskeletons and galleries.The rate of occurrence of Fusarium spp. in plant tissue samples reached maximum levels as the fungi appeared in all of the xylem samples from galleries constructed by O. erosus, and this insect species also showed a high rate of phoresy for these fungal species (56.6% of the specimens).Fusarium spp.comprise a polyphyletic group that can act as endophytes or as a plant pathogens depending on the species and on the plant host, according to host specificity for plant-pathogen interactions.Moreover, these species are widely distributed, infecting a wide range of organisms worldwide.Several Fusarium spp.have mutualistic associations with insects, e.g., Fusarium solani (Martius) which has a symbiotic relationship with Hypothenemus hampei (Ferrari) while colonizing coffee beans [55] and with Xyleborus ferrugineus (Fabricius) when colonizing dead insects as saprophytes or entomopathogens.In general, the Fusarium species associated with beetles are weakly entomopathogenic, although infection of the pine beetle Dendroctonus frontalis (Zimmerman) with F. solani resulted in the death of 90% the insects in 5 days [2].In the present study, Fusarium avenaceum was the most commonly identified species and was directly isolated from I. sexdentatus, H. attenuatus and H. ater specimens and their galleries.Moreover, it also appeared on T. piniperda-infested shoots.Fusarium avenaceum has been previously isolated from P. radiata in New Zealand, where it was associated with dieback caused by physical injury [56].However, the F. graminearum clade and F. avenaceum are commonly associated with crops [57] and cause Fusarium head blight (FHB) on wheat [58].Molecular identification of Fusarium species was necessary to distinguish isolate number 21 (Table 4) and the DNA sequence corresponded to F. avenaceum, although the presence of napiform microconidia in this isolate had identified it morphologically as Fusarium anthophilum.Fusarium tricinctum was isolated from I. sexdentatus and from T. piniperda feeding galleries.This fungus usually acts as a saprophyte or a weak parasite in Europe and North America [31].The Fusarium oxysporum spp.complex was found to be associated with O. erosus, H. attenuatus and detected in the I. sexdentatus galleries in logs (both, xylem and phloem).This is a saprophyte and soil pathogen species complex with a wide range of plant hosts divided into many formae specialis depending on the host specificity.Fusarium sambucinum, which was isolated from P. pubescens, causes Fusarium dry rot in potatoes [59] but has also been isolated from P. radiata in New Zealand and found to be associated with dieback and root rot [56].Other species isolated in this study, such as Fusarium sporotrichioides, F. konzum and F. beomiforme, have not been previously described as plant pathogens or are known only as very weak pathogens [26].
Fusarium circinatum, the causal agent of PPC, appeared in 3.5% of the fallen shoots occupied by T. piniperda.This insect species has been proposed as a plausible vector of the disease in the study region, where 12% of the shoots collected from the ground were found to be associated with the pathogen in plantations affected by PPC [22].However, the pathogen was not found in association with specimens of T. piniperda and T. destruens analyzed by Muñoz-Adalia et al. [60] in the same study area, although other Fusarium spp.were detected.The singularity of this association lies in the shoot-feeding maturation behavior of T. piniperda in the crowns of healthy pine trees [61], which may enable transmission of the fungal spores into healthy crowns.This insect has also been associated with some other highly phytopathogenic fungi such as Leptographium wingfieldii Morelet in Europe and North America [62,63].These species have caused significant economic losses due to blue staining of infected wood.Brood galleries of Hylastes spp.also appeared to harbour to F. circinatum in this study.These insects have a similar way of interacting with sapwood fungi in several pine species [64], including P. radiata, although no ophiostomatoid fungi were found in the present study.Hylastes angustatus galleries frequently harboured F. circinatum (5.2% of xylem samples and 5.5% of phloem samples).However, Hylastes spp.are secondary pests, attacking weakened trees and roots, which suggests that they are not able to inoculate healthy trees with the pathogen as in the case of T. piniperda.In addition, some of the specimens of the bark beetle P. pubescens (1.05%) captured carried F. circinatum.The importance of other species of this genus has been noted in several countries.For example, in California the potential role of Pityophthorus carmeli Swaine and Pityophthorus setosus Blackman in spreading F. circinatum was shown by the wounding behaviour of these twig beetles while they tested the suitability of P. radiata branches as hosts [24].Bonello et al. [65] reported that Pityophthorus spp.discriminated between healthy and pitch canker-diseased branches, preferring symptomatic branches due to the increased emission of ethylene.However, P. pubescens is a secondary species in the present study area, affecting small branches from weakened trees [23], and it does not seem to have an important role as a vector.On the other hand, F. circinatum was present in 0.9% of the I. sexdentatus specimens.Although Ips sexdentatus is a secondary bark beetle, it can act as a primary parasite when the population reaches epidemic levels [66] and could thus inoculate healthy trees with the pathogen.In the Basque Country (Spain), 8.57% of the I. sexdentatus analyzed [21] carried the pathogen.Likewise, in California (USA) other species such as Ips mexicanus (Hopkins) and Ips paraconfusus Lanier, have been reported to be vectors of the pitch canker fungus [67].Fusarium circinatum spores may be harboured on the exoskeletons of bark beetles when they feed on or breed in pitch canker diseased trees.Insects carrying the spores could then inoculate healthy pines.Some authors report that the association between insects and pathogenic fungi leads to stimulation of tree resistance [68].However, as epidemic levels of bark beetles can kill healthy trees without carrying any phytopathogenic fungi, other authors state that insect-pathogen interactions could only benefit the fungal species, thus helping them expand their range to trees that they would otherwise not reach without this association [69].Bark beetles are also principal wounding agents, contributing to the infections caused by airborne spores, although the insects themselves do not carry the pathogen [70].The presence of the airborne inoculum of the causal agent of PPC has been recently studied by Dvorak et al. [71], who demonstrated that spores are present in the air throughout the year.

Conclusions
In conclusion, the species richness and also the fungal communities associated with the bodies of the bark beetles and their galleries varied depending on different factors such as insect species, type of plant tissue and season.The interactions between different fungal species associated with insects galleries has been shown to be a relevant factor determining the composition of the mycobiota, as the presence of T. harzianum was negatively correlated with a the frequency of F. circinatum.Moreover insect-fungi interactions are important for the distribution of fungal species.In this study, an association was found between the Fusarium species and bark beetles and their galleries, reaching the maximum rate in H. ater samples, 64% of which harboured Fusarium species.The importance of bark beetles in the distribution of Fusarium spp.may be related to the season (as observed in this study) as well as to the population levels, as bark beetles can reach epidemic levels after forest disturbances as fires, storms or outbreaks of pathogens.Specifically, our results showed that bark beetles appeared at a low frequency as phoretic agents of the PPC pathogen, suggesting that they may not be the main means of spreading of the disease in the study area.This should be considered together with the fact that the bark beetles were found at endemic levels, and if the population increased, this association would become more important.However, as F. circinatum is a quarantine pathogen, all pathways by which it can be introduced to new areas represent a potential risk to forest health.Further studies are required to identify the specific associations between different bark beetle species and fungal communities, in particular to clarify the role of the different Scolytinae species in the distribution of Fusarium spp.

Forests
for amplification was as follows: 10 min denaturation at 95 • C followed by 13 cycles of 35 s at 95 • C, 55 s at 55 • C and 45 s at 72 • C; 13 cycles of 35 s at 95 • C, 55 s at 55 • C and 2 min at 72 • C; 9 cycles of 35 s at 95 • C, 55 s at 55• C and 3 min at 72 • C; and a final elongation 7 min at 72 • C. The purified PCR product (NucleoSpin Gel and PCR Clean up, Macherey Nagel) was sent to a commercial sequencing service (SecuGen, Madrid, Spain).The ITS region sequences were revised with the Geneious Pro 5.6.5 software package for Blast search in the GenBank database.

Forests 2018, 9 ,
x FOR PEER REVIEW 2 of 20 associated with the xylem and phloem and the other related to the insect exoskeletons.Fusarium spp.

Figure 6 .
Figure 6.Phylogenetic tree inferred using the neighbor-joining method.The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches.The evolutionary distances are in the units of the number of base substitutions per site.

Figure 6 .
Figure 6.Phylogenetic tree inferred using the neighbor-joining method.The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches.The evolutionary distances are in the units of the number of base substitutions per site.

Table 1 .
Insect species collected from logs, funnel traps and shoots.

Table 2 .
Percentage of plant tissue samples containing each fungal species.

Table 3 .
Percentage of bark beetles harbouring each fungal species.