Status of Charcoal Canker on Oak Trees at a Site of Community Importance: Case Study of the Relict Castelﬁdardo Forest (SIC Area IT520008, Castelﬁdardo, AN, Italy)

: Oaks are dominant and key tree species in Mediterranean forest ecosystems. However, in recent decades, oak forests have been heavily impacted by oak decline, a worldwide phenomenon exacerbated by climate change. The charcoal disease agent Biscogniauxia mediterranea is involved in the decline of Mediterranean oak formations in a variety of contexts. Here, we investigated the impact and role of B. mediterranea in the decline of oaks in Castelﬁdardo Forest, a relict wood of the late Holocene and a Site of Community Importance. We established ﬁve plots within which we recorded tree positions, any symptoms and signs of decline, association of B. mediterranea to declining trees, and deadwood and associated mycota. Of 471 oaks inspected, 7.0% showed brownish exudates on the stems, 46.9% showed epicormic shoots along the main trunk, and 24.4% showed black carbonaceous stromata on diseased branches and trunks. The decline was most severe for Quercus cerris , which comprised plots #4 and #5, at 50.0% (81/162 trees) and 29.0% (33/114), respectively; then for Quercus robur for plot #3, at 40.0% (38/95); and ﬁnally for Quercus pubescens for plots #1 and #2, at 13.7% (7/51) and 12.3% (6/49), respectively. Bark tissue was collected from trees with charcoal cankers via microscopy examination and identiﬁed by mycological and molecular methods. This investigation revealed a close association between oaks with pronounced reduction of vitality and incidence of B. mediterranea . Deadwood was equally distributed among the ﬁve plots, and was heavily colonized by Basidiomycota . The high incidence of the charcoal canker pathogen B. mediterranea appeared to be related to environmental stresses. However, the absence of silvicultural management, high competition among physiologically mature trees, and the geographic isolation of this residual forest may have predisposed oaks to decline.


Introduction
Forest decline may be considered the most worrying and complex syndrome that affects wide forested areas across a range of ecosystems throughout the world [1][2][3][4][5]. When dealing with forest decline in the Mediterranean area, special attention should be focused on oak forests, due to the importance and widespread presence of oaks in Mediterranean forest ecosystems and the severity and extent of the decline in oak formations [6].
Oak decline involves various members of the genus Quercus, and it is seen as a general and progressive decline that is accompanied by a range of symptoms. The most common of these are: leaf chlorosis and yellowing; microphyllia; premature leaf drop and crown transparency; branch dieback; emission of dark exudates through the trunk; and production The aims of this study were to: (i) determine the health conditions of the oak trees in the Castelfidardo Forest; (ii) describe the incidence of the charcoal canker pathogen Biscogniauxia mediterranea on the declining oaks; (iii) determine the amounts of deadwood and take a census of the associated Mycota; and (iii) determine whether there is a causeeffect relationship between the water stress and drought and the decline of the oak species.

Study Site, Plot Design, and Tree Health Assessment
An examination of the phytosanitary status of the oak trees was carried out in the Castelfidardo Forest (43°27′59″ N, 13°35′30″ E) from May to July 2015. For this purpose, five plots were located in the upper, middle, and lower areas that were representative of the soil types and vegetation that characterize the forest (Table 1). Each plot was circular, with an area ranging from 0.6 ha to 0.9 ha. The plots were of variable sizes and were designed to include all of the heterogeneity conditions in terms of soil type, vegetation, and intensity of decline.  The aims of this study were to: (i) determine the health conditions of the oak trees in the Castelfidardo Forest; (ii) describe the incidence of the charcoal canker pathogen Biscogniauxia mediterranea on the declining oaks; (iii) determine the amounts of deadwood and take a census of the associated Mycota; and (iv) determine whether there is a causeeffect relationship between the water stress and drought and the decline of the oak species.

Study Site, Plot Design, and Tree Health Assessment
An examination of the phytosanitary status of the oak trees was carried out in the Castelfidardo Forest (43 • 27 59 N, 13 • 35 30 E) from May to July 2015. For this purpose, five plots were located in the upper, middle, and lower areas that were representative of the soil types and vegetation that characterize the forest (Table 1). Each plot was circular, with an area ranging from 0.6 ha to 0.9 ha. The plots were of variable sizes and were designed to include all of the heterogeneity conditions in terms of soil type, vegetation, and intensity of decline. In the upper layer, plots #1 and #2 were characterized by the arenaceous substrate where Q. pubescens was predominant, and was associated with sporadic presence of Fraxinus ornus trees. In plot #3, in the middle layer, the dominant species was Q. robur, with rare Carpinus betulus found. In the lower layer, where plots #4 and #5 were located, the arenaceous component definitively disappeared, and the dominant forest vegetation was Q. cerris, with Carpinus orientalis occurring sporadically. In all of these plots, Ruscus aculeatus was the predominant understory species, and this covered large proportions of the plots, strongly limiting the natural regeneration of the oaks.
As summarized in Table 1, the main parameters recorded in each plot were: surface area; tree species; tree densities; and total numbers of trees inspected. A number of trees were selected according to the expression of symptoms (i.e., production of tarry exudates, epicormic shoot formation, and eruption of charcoal cankers through the bark), from 49 to 162 trees in each plot. For each tree selected, the diameter at chest height (i.e., 1.3 m above the ground) and its sanitary status were assessed and recorded, noting in particular the presence and frequency of: early symptoms of decline (e.g., emission of exudates, production of epicormic branches); advanced symptoms and/or signs of decline (e.g., dieback, black carbonaceous stromata indicative of charcoal cankers); and tree mortality. The position of each tree was recorded using the Global Positioning System (GPS; e-Trex 30, Garmin), and metallic labels with progressive identification numbers were affixed to the trunks. The data obtained relating to individual trees were managed through the Quantum Geographic Information System (GIS) open source software, and displayed using the Google Earth software.

Sample Collection, Mycological Analysis, and Morphological Identification
Bark tissue with charcoal canker (mean dimension, 4 × 4 cm) was collected from 20 Q. pubescens trees in plots #1 and #2, from 10 Q. robur trees in plot #3, and from 20 Q. cerris trees in plots #4 and #5. These samples were incubated in a humidity chamber with five layers of tissue paper that had previously been soaked in sterile distilled water, and placed in glass Petri dishes (diameter, 14 cm), at 25 • C, in the dark, for 10 days. The samples were then observed under a stereo microscope (MZ125; Leica) to document the sporulation. Various samples were photographed using a digital camera (FinePix S1 Pro; Fujifilm).
The fungal agents were isolated from the bark after the incubation in the humidity chamber, from differentiated fruiting structures. Under aseptic conditions, portions of the mycelia were transferred onto Petri dishes with potato dextrose agar (PDA) with added antibiotics (150 mg/L streptomycin, 150 mg/L ampicillin; to prevent bacterial contamination). Five original mycelial samples were put into each plate, and five replicates were prepared from each sample. Each plate was labeled, sealed with parafilm, and incubated at 25 • C in the dark. After 5 days, the developed fungal colonies were subcultured onto fresh PDA plates, then transferred into glass test tubes, and finally stored at 4 • C. For 50 isolates (10 for each plot), identification was carried out by recording the characteristics of 50-100 units of spores for each isolate, by light microscopy (Eclipse E600; Nikon).

Molecular Identification of Fungal Colonies
Mycelia were collected from 25 representative fungal colonies grown on PDA, transferred into 1.5 mL microcentrifuge tubes and pulverized, with the addition of 600 µL extraction buffer (20 mM EDTA, 0.1 M Tris-HCl, pH 8.0, 1.4 M NaCl, 2% cetyltrimethylammonium bromide, 4% polyvinylpyrrolidone, 0.1% sodium metabisulfite [added just before use]). The extraction of total DNA was carried out using the CTAB protocol [31]. The quality and quantity of the extracted DNA were determined directly on 1% agarose gels, with evaluation using a biophotometer (Eppendorf, Hamburg, Germany). The DNA was finally diluted to 20 ng/µL for further amplification. Nucleic acids were stored at −20 • C, and later analyzed by polymerase chain reaction (PCR). Amplification of the 5.8S ribosomal DNA (rDNA) and flanking internal transcribed spacer (ITS) region was performed in 20 µL PCR reactions that contained 2 µL genomic DNA (at~20 ng/µL) from the fungal isolate, 10 µL EmeraldAmp GT PCR Master mix 2× (Takara, Madison, WI, USA), and 0.5 µL of each primer (10 µM), in a thermal cycler (MyCycler; Bio-Rad Laboratories, Hercules, CA, USA). MED1/MED2 primers were used, which are specific for B. mediterranea [32]. The PCR products (9 µL per sample) were separated by electrophoresis in 1.5% agarose gels stained with Red Gel (Biotium, Hayward, CA, USA), then visualized and captured using an imagining system (Gel Doc XR; BioRad, Hercules, CA, USA). The same ITS region was also amplified from fungal isolates with universal primers ITS1/ITS4 [33]. Amplicons from three isolates (WP101, WP20, WP376), were purified using Wizard SV gels and PCR clean-up kits (Promega Corporation, Madison, WI, USA), and quantified using the biophotometer (Eppendorf, Hamburg, Germany). Sequencing was carried out by Genewiz UK. The Bioedit software, v. 7.0.0 (http://www.mbio.ncsu.edu/Bioedit/bioedit.html, accessed on 12 April 2021) was used to cut off 20 bp to 30 bp of the terminal end sequence. All of the sequences were used as the query sequences in BLAST searches (http://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 26 April 2021), to determine the nucleotide identities available in GeneBank NCBI. The nucleotide sequences were deposited in the NCBI database (MZ407593-MZ407595).

Deadwood and Associated Lignicolous Macrofungi
As deadwood and deadwood-dwelling macrofungi are fundamental for the maintenance of the biodiversity of forest ecosystems, the quantity and quality of the deadwood served as a proxy indicator for the level of biodiversity of Castelfidardo Forest. In particular, the deadwood was classified into standing dead trees (snags) and lying dead woody debris (logs). The quantities were expressed as percentages, considering the snags and logs with respect to the total number of oak trees in each plot.
The associated lignicolous macrofungal assemblage was analyzed concurrently. The macrofungi were collected gently and taken to the laboratory for detailed mycological examination.

Climatic Data
Data for the minimum, maximum, and average temperatures, and for the average annual and seasonal rainfall for the period 1951-2015 were obtained from the Regional Informative System located at Ancona Torrette 2944 (RT 701) (13 • 26 , 43 • 36 ).

Phytosanitary Monitoring
A total of 471 oak trees were assessed from May to July 2015 for their health conditions in the five plots ( Table 1). Many of the trees showed an advanced decline. The patterns of symptom appearance showed a succession of symptoms of increasing severity. Oaks in the initial stages of decline showed some nonspecific symptoms, such as leaf yellowing, early leaf browning, and thinning of canopy cover. Those in advanced states of decline showed foliar rolling and leaf wilting, production of epicormic shoots on the main branches and stems, and loss of branches, which eventually lead to tree mortality.
The emission of dark-brownish mucilaginous exudates through the bark (Figure 2A,B) occurred mainly during the growing season, and this was recorded for 7.2% (34/471) of the trees. Debarking the trees at the points of emission of exudates showed clearly visible necrotic areas of the underbark tissue. Furthermore, the trees with these symptoms often showed marked transparency of the canopy. Trees with production of exudates through the bark were more frequently recorded for plots #5 and #2, at 9.7% (11/114 trees) and 8.2% (4/49), respectively ( Figure 3A). Epicormic shoot formation along the main trunk was recorded for 42.7% (201/471) of the trees ( Figure 2C,D), and it occurred equally in all  Figure 2E). In cross-sections of the trunks, necrosis of the xylem and wood decay were observed in correspondence with the black carbonaceous stromata that had developed externally ( Figure 2F). The charcoal-black cankers were more frequently recorded for plot #3, at 40.0% (38/95), followed by plots #4 and #5, at 30.3% (49/162) and 25.4% (29/114), respectively; in plots #1 and #2, occurrence of the typical charcoalblack surface was lower ( Figure 3C). Declining trees generally died in the growing season following the appearance of these erumpent, coriaceous black stromata. In plot #4, there were sporadic fallen trees that showed root rot and thick, black, string-like structures, which were usually seen under the bark or in the soil near the roots, as typical structures of Armillaria sp. (Figure 2G,H).
visible necrotic areas of the underbark tissue. Furthermore, the trees with these symptoms often showed marked transparency of the canopy. Trees with production of exudates through the bark were more frequently recorded for plots #5 and #2, at 9.7% (11/114 trees) and 8.2% (4/49), respectively ( Figure 3A). Epicormic shoot formation along the main trunk was recorded for 42.7% (201/471) of the trees ( Figure 2C,D), and it occurred equally in all of the five plots ( Figure 3B). Charcoal cankers erupted through the bark in about 27.6% (130/471) of the trees ( Figure 2E). In cross-sections of the trunks, necrosis of the xylem and wood decay were observed in correspondence with the black carbonaceous stromata that had developed externally ( Figure 2F). The charcoal-black cankers were more frequently recorded for plot #3, at 40.0% (38/95), followed by plots #4 and #5, at 30.3% (49/162) and 25.4% (29/114), respectively; in plots #1 and #2, occurrence of the typical charcoal-black surface was lower ( Figure 3C). Declining trees generally died in the growing season following the appearance of these erumpent, coriaceous black stromata. In plot #4, there were sporadic fallen trees that showed root rot and thick, black, string-like structures, which were usually seen under the bark or in the soil near the roots, as typical structures of Armillaria sp. (Figure 2G,H).

Mycological and Molecular Identification
We sporadically isolated from the bark tissue of oaks some fungi belonging to the genera Botryosphaeria, Diplodia, Neofusicoccum, Cytospora, Discula (Apiognomonia). However, we focused our attention on the agent of charcoal canker B. mediterranea as the incidence of this fungus was, in terms of isolation frequencies, extraordinarily higher.
Stereoscopic observations revealed carbonaceous, perithecial stromata in all of the 50 samples. The stromata appeared slightly convex, ellipsoid, and elongated, 7.2 to 20.5 × 3.5 to 4.2 cm. Perithecia were ovoid to tubular, 0.74 to 0.80 × 0.12 to 0.15 mm. A total of 152 fungal colonies were obtained in purity, most of which (94.1%; 143) were identified as B. mediterranea by mycological examination of their distinguishing micromorphological traits ( Figure 4). The colonies completely filled the agar plates (diameter, 90 mm) after 7 days of incubation at 25 °C. The colonies were white-grey in color and velvety (viewed from the top). Short-stipitate, amyloid asci that were dark brown were observed; these produced ellipsoid ascospores, 14 μm to 19 × 7 to 9 μm, which perfectly matched the description by Mirabolfathy et al. [34]. No reproductive structures were observed in cultures that could be used to identify the microorganisms on a micromorphological basis. The reverse of the colonies were darker, tending to dark grey-black. Colonies of Trichoderma sp. were also sporadically isolated from the black carbonaceous stroma (four isolates).
The morphological identification of the fungi was supported by molecular identification. The DNA extraction protocol was very efficient, and provided ~210 ng/µ L, with high quality (A260/280 > 1.87; A260/230 > 1.92). Only a few of the extracted samples (n =

Mycological and Molecular Identification
We sporadically isolated from the bark tissue of oaks some fungi belonging to the genera Botryosphaeria, Diplodia, Neofusicoccum, Cytospora, Discula (Apiognomonia). However, we focused our attention on the agent of charcoal canker B. mediterranea as the incidence of this fungus was, in terms of isolation frequencies, extraordinarily higher.
Stereoscopic observations revealed carbonaceous, perithecial stromata in all of the 50 samples. The stromata appeared slightly convex, ellipsoid, and elongated, 7.2 to 20.5 × 3.5 to 4.2 cm. Perithecia were ovoid to tubular, 0.74 to 0.80 × 0.12 to 0.15 mm. A total of 152 fungal colonies were obtained in purity, most of which (94.1%; 143) were identified as B. mediterranea by mycological examination of their distinguishing micromorphological traits (Figure 4). The colonies completely filled the agar plates (diameter, 90 mm) after 7 days of incubation at 25 • C. The colonies were white-grey in color and velvety (viewed from the top). Short-stipitate, amyloid asci that were dark brown were observed; these produced ellipsoid ascospores, 14 µm to 19 × 7 to 9 µm, which perfectly matched the description by Mirabolfathy et al. [34].

Mycological and Molecular Identification
We sporadically isolated from the bark tissue of oaks some fungi belonging to the genera Botryosphaeria, Diplodia, Neofusicoccum, Cytospora, Discula (Apiognomonia). However, we focused our attention on the agent of charcoal canker B. mediterranea as the incidence of this fungus was, in terms of isolation frequencies, extraordinarily higher.
Stereoscopic observations revealed carbonaceous, perithecial stromata in all of the 50 samples. The stromata appeared slightly convex, ellipsoid, and elongated, 7.2 to 20.5 × 3.5 to 4.2 cm. Perithecia were ovoid to tubular, 0.74 to 0.80 × 0.12 to 0.15 mm. A total of 152 fungal colonies were obtained in purity, most of which (94.1%; 143) were identified as B. mediterranea by mycological examination of their distinguishing micromorphological traits (Figure 4). The colonies completely filled the agar plates (diameter, 90 mm) after 7 days of incubation at 25 °C. The colonies were white-grey in color and velvety (viewed from the top). Short-stipitate, amyloid asci that were dark brown were observed; these produced ellipsoid ascospores, 14 μm to 19 × 7 to 9 μm, which perfectly matched the description by Mirabolfathy et al. [34]. No reproductive structures were observed in cultures that could be used to identify the microorganisms on a micromorphological basis. The reverse of the colonies were darker, tending to dark grey-black. Colonies of Trichoderma sp. were also sporadically isolated from the black carbonaceous stroma (four isolates).
The morphological identification of the fungi was supported by molecular identification. The DNA extraction protocol was very efficient, and provided ~210 ng/µ L, with high quality (A260/280 > 1.87; A260/230 > 1.92). Only a few of the extracted samples (n = No reproductive structures were observed in cultures that could be used to identify the microorganisms on a micromorphological basis. The reverse of the colonies were darker, tending to dark grey-black. Colonies of Trichoderma sp. were also sporadically isolated from the black carbonaceous stroma (four isolates).
Amplification with the specific MED1/MED2 primers gave fragments of the expected length (377 bp) in all 25 of the samples analyzed. No ambiguous bands were generated, and the lack of bands in the water control test confirmed the absence of contamination. A portion of the ITS region (approximately 580 bp) was sequenced for additional confirmation of the identity of the fungus. Three nucleotide sequences selected revealed 100% identity with reference sequence MT819849 in the NCBI database, as B. mediterranea isolated from Quercus suber in Portugal.

Deadwood and Deadwood-Dwelling Mycobiota
The ground within Castelfidardo Forest was characterized by high amounts of deadwood, for all of the plots investigated. Plot #2 showed the lowest abundance of snags (12.2%, versus total number of oak trees) and logs (4.0%), while the highest abundance was for plot #5 (39.5%, 18.4%, respectively) ( Figure 5).  6) showed low quality DNA values (A260/280 < 1.80; A260/230 < 1.80), and these were not processed further. Amplification with the specific MED1/MED2 primers gave fragments of the expected length (377 bp) in all 25 of the samples analyzed. No ambiguous bands were generated, and the lack of bands in the water control test confirmed the absence of contamination. A portion of the ITS region (approximately 580 bp) was sequenced for additional confirmation of the identity of the fungus. Three nucleotide sequences selected revealed 100% identity with reference sequence MT819849 in the NCBI database, as B. mediterranea isolated from Quercus suber in Portugal.

Climatic Data
Analysis of climatic data related to Castelfidardo Forest revealed that the average annual temperature (17 °C) recorded over the period from 2001 to 2015 has increased by 2 °C with respect to the historical series   (Figure 7). For rainfall, there were no evident differences, with two negative peaks recorded in 1983 (376 mm) and 2014 (488 mm). A more detailed analysis of the climatic data relating to the growing seasons from 2001 to 2015 revealed that in the spring and summer periods for that time interval, the temperatures increased by 2 °C and 2.5 °C respectively, compared to the former period of 1957 to 2000.

Discussion
Oak forest ecosystems are of paramount importance in the Mediterranean basin due to the multiplicity of functions and services they provide in this region. Indeed, this region is at risk of desertification due to climatic and socio-economic driving forces, including rainfall variability, droughts, changes in land-use patterns, overgrazing, and demographic saturation of rural areas [35]. Oak formations have crucial roles here for the prevention of land degradation and for climate mitigation, carbon storage, and the safeguarding of the environment; they also generate economic, esthetic and landscaping benefits [25].
Unfortunately, Mediterranean oak forests have been plagued by oak decline for decades now, which is a phenomenon that appears strongly related to climate anomalies, as mainly heat stress, but also water scarcity and droughts [18,36]. Indeed, the Mediterranean region is a hotspot for climate change, as it is characterized by a negative water balance and pronounced warming [37,38]. This region is currently experiencing high water stress conditions and prolonged droughts, and these are worsening the effects of climate change, with a strongly negative impact on plant biomass production. Our analysis of climatic data showed increased temperatures during spring and summer for the period from 2001 to 2015 in the Castelfidardo area, which confirms the general warming trend that is constraining the whole Mediterranean region.
The stressful conditions determined by the water deficit, high temperatures, and droughts have created conditions conducive to the pervasive spread of the thermophilic fungus B. mediterranea in oak forests [39]. Several fungal species have also been isolated from trunk and branch cankers as well as from twigs and foliage of declining oaks, including Botryosphaeria dothidea, Diplodia corticola, Diplodia seriata, Neofusicoccum parvum, Cytospora spp., Discula quercina, and Neocucurbitaria cava [22,23,40,41]. The occurrence of these fungi was relatively erratic in the plots investigated, with lower isolation frequency compared to B. mediterranea. However, in particular here, we focused our attention on trees with evident symptoms of decline that were accompanied by the specific presence of B. mediterranea stromata. The relationship between oak decline, climate change, and the proliferation of B. mediterranea has been well documented in Mediterranean countries [12,[42][43][44][45][46]. Indeed, recent outbreaks of B. mediterranea in countries where it had not been reported to cause harm, like in Turkey, Iran, Slovenia [47,48], and Croatia [49], confirm that this fungus is expanding its range and is accentuating its role as a contributing factor to the decline of oak formations.
Members of the Xylariaceae fungi are widely known for surviving across a range of lifestyles. They live as saprotrophs, as they can decompose many natural substrates, and

Discussion
Oak forest ecosystems are of paramount importance in the Mediterranean basin due to the multiplicity of functions and services they provide in this region. Indeed, this region is at risk of desertification due to climatic and socio-economic driving forces, including rainfall variability, droughts, changes in land-use patterns, overgrazing, and demographic saturation of rural areas [35]. Oak formations have crucial roles here for the prevention of land degradation and for climate mitigation, carbon storage, and the safeguarding of the environment; they also generate economic, esthetic and landscaping benefits [25].
Unfortunately, Mediterranean oak forests have been plagued by oak decline for decades now, which is a phenomenon that appears strongly related to climate anomalies, as mainly heat stress, but also water scarcity and droughts [18,36]. Indeed, the Mediterranean region is a hotspot for climate change, as it is characterized by a negative water balance and pronounced warming [37,38]. This region is currently experiencing high water stress conditions and prolonged droughts, and these are worsening the effects of climate change, with a strongly negative impact on plant biomass production. Our analysis of climatic data showed increased temperatures during spring and summer for the period from 2001 to 2015 in the Castelfidardo area, which confirms the general warming trend that is constraining the whole Mediterranean region.
The stressful conditions determined by the water deficit, high temperatures, and droughts have created conditions conducive to the pervasive spread of the thermophilic fungus B. mediterranea in oak forests [39]. Several fungal species have also been isolated from trunk and branch cankers as well as from twigs and foliage of declining oaks, including Botryosphaeria dothidea, Diplodia corticola, Diplodia seriata, Neofusicoccum parvum, Cytospora spp., Discula quercina, and Neocucurbitaria cava [22,23,40,41]. The occurrence of these fungi was relatively erratic in the plots investigated, with lower isolation frequency compared to B. mediterranea. However, in particular here, we focused our attention on trees with evident symptoms of decline that were accompanied by the specific presence of B. mediterranea stromata. The relationship between oak decline, climate change, and the proliferation of B. mediterranea has been well documented in Mediterranean countries [12,[42][43][44][45][46]. Indeed, recent outbreaks of B. mediterranea in countries where it had not been reported to cause harm, like in Turkey, Iran, Slovenia [47,48], and Croatia [49], confirm that this fungus is expanding its range and is accentuating its role as a contributing factor to the decline of oak formations.
Members of the Xylariaceae fungi are widely known for surviving across a range of lifestyles. They live as saprotrophs, as they can decompose many natural substrates, and they can also be pathogenic, promoting disease on living hosts, and especially those that are water-stressed. The Xylariaceae fungi can persist within trees in a latent stage, due to their endophytic aptitude [50]. B. mediterranea is a weakness pathogen that can live as an endophyte inside healthy plant tissues, thus causing extensive, although symptomless, infections. Such a cryptic existence can last for undefined periods, as long as the host does not become weakened by adverse environmental conditions. When stressful conditions that alter the carbohydrate physiology in trees occur, B. mediterranea switches to a pathogenic behavior to extensively colonize its debilitated host. This pathogenic endophyte then moves externally to sporulate profusely over the plant surfaces, a mechanism used to escape from a dying tree and colonize other trees that are already severely weakened, and possibly injured [11,23].
The aggressive colonization of oaks by B. mediterranea in the Castelfidardo Forest must thus be mainly ascribed to the chronic debilitation of these trees by hot temperatures, recurrent water deficit and droughts. Exposure to these environmental adversities will have triggered tree infection by this opportunistic fungus whereby its pathogenic action will also be facilitated by the intrinsic, low resilience of this residual forest. As a relict population, Castelfidardo Forest is more prone to any disturbance. Indeed, the extent of oak decline in Castelfidardo Forest appears greater than for other oak stands of the Italian Peninsula. Millennia of anthropogenic pressure, which was mainly represented by agricultural encroachment and overgrazing, led to the isolation of this residual portion of primeval forest. Such forest fragmentation can result in long-lasting and complex changes in biodiversity that can go beyond the loss of species, and thus include alterations to the functional diversity of the remaining communities, with isolated forests being more fragile and vulnerable [51].
The saprophytic stage is of fundamental importance in the epidemiology of B. mediterranea in Mediterranean oak stands. Through its extensive production of stromatal charcoal cankers, this xylariaceous fungus can introduce massive amounts of spores into the oak forest ecosystem, enormously increasing its biomass. The number of individuals introduced at a given location is a primary determinant of the successful establishment of an invasive organism into new environments [52]. On the other hand, the importance of propagule pressure in the spread of fungal pathogens and in the colonization of new hosts in oak stands has already been established [53].
Recent studies have shown how some functional traits can exacerbate the negative effects of drought, such as stand density and intraspecific and interspecific competition, which can then predispose trees to decline [54,55]. The Castelfidardo Forest was subjected to intense cutting in 1970, which has not been followed by any other interventions to date. This lack of management for half a century has had repercussions on the tree physiology and vigor; in particular, the high density of the stands causes excessive competition among the trees for the already limited resources (i.e., water, nutrients, light). This has led to depletion of the plant carbon reserves, and therefore to less resistance to colonization by the weakness pathogen B. mediterranea. Indeed, higher incidence of charcoal canker and higher tree mortality were recorded for Q. robur in plot #3 and for Q. cerris in plots #4 and #5. In these plots, tree densities were much higher than in plots #1 and #2. Differential susceptibility to decline across oak species was shown in previous studies. In particular, in France, Q. robur was more susceptible than Quercus petraea [56]. In North America, species of the red oak group were more susceptible than those of the white oak group [57]. In central and southern Italy, Q. cerris is more susceptible to decline than Q. pubescens [58].
In the present study, the disease caused by B. mediterranea was also tree-density dependent, and it represented only the terminal stage of infection. In this context, the fungus would function as an ecological factor that regulates the structure and composition of forest stands, to make them better suited to changing environmental conditions [59].
We recorded an abundance of deadwood that was equally distributed across all of the plots. The role of deadwood in forest ecosystems is linked to various aspects, which include improvement of natural diversity [60], storage and slow release of carbon and nutrients [61,62], maintenance of soil fertility, creation of pedoclimatic conditions conducive to natural regeneration of the forest, conservation of the soil, and improvement of the stability of slopes from hydrogeological risk [63]. The deadwood richness of the Castelfidardo Forest suggests that the carbon reserves in the stand are substantial and can therefore be used by the trees. However, it must be considered that when trees are chronically impaired by abiotic (i.e., environmental) and biotic (mainly anthropogenic) stressors beyond a threshold of tolerance, their energies become so depleted that they can no longer benefit from the availability of substrate. This happens especially to physiologically overmature and veteran trees, like the majority of oaks in Castelfidardo Forest [64,65]. Armillaria mellea and A. ostoyae were also associated with the deadwood. These two Agaricales deserve special attention because they play a central role in the dynamics of numerous woody ecosystems [66]. They are ubiquitous species, more commonly found in mature stands with an abundance of old trees, where they play a valuable ecological role in the recycling of dead wood. However, A. ostoyae and A. mellea are also polyphagous, facultative parasites with parasitic ability, causing white rot to a number of host species [67]. In Castelfidardo Forest these two fungi live preferentially as saprotrophs on dead wood and only sporadically are they found to attack pathogenically live but visibly weakened trees [65].

Conclusions
The present study focused on the decline of oaks in Castelfidardo Forest, which is a forest of great ecological importance. The attacks of B. mediterranea on oaks in this forest have increased dramatically in recent years and must be related to hot temperatures and recurrent droughts which chronically weaken trees (Eugenio Paoloni, personal communication). The life history strategy of B. mediterranea has been investigated in other oak woods of central Italy where it emerged, as in this study, that the agent of charcoal canker takes advantage of environmental stresses to aggressively colonize physiologically impaired oaks [68]. Since we cannot interfere with climatic conditions, to mitigate the impact of the decline, it would be advisable to reduce the source of inoculum and tree density. This opportunistic fungus sporulates abundantly on the trunks and branches of dying or standing dead trees, as well as on deadwood on the ground. Therefore, among the possible control strategies, it would be appropriate to: (i) eliminate all dead trees with sporulating cankers; (ii) carry out thinning, to reduce tree densities, particularly in plots #3, #4, and #5; and (iii) explore the possibility of using antagonistic microorganisms (e.g., Trichoderma spp.) for repressing this stress-mediated disease [69].