Mediterranean forests are characteristic ecosystems of the Mediterranean Basin, but also present in California, South Africa and Chile. These forests are a hot spot of genetic diversity in the European continent, as they harbor 290 tree species, compared to 135 reported in the rest of Europe (Gauquelin et al., 2018). The organisms that make up the Mediterranean forests are usually adapted to very specific climatic factors, such as the existence of a marked seasonality characterized by long periods of summer drought, and great variability in the total annual rainfall. In addition, Mediterranean forests have been subjected for millennia to anthropogenic impacts, mainly fragmentation due to land use change toward agriculture and grazing, overexploitation to obtain goods and services and, more recently, abandonment of their use (Fady et al., 2022). These characteristics have confirmed that Mediterranean forests are as vulnerable to management policies as they are to climate change (Morán-Ordóñez et al., 2020).

Some of the main tree species found in Mediterranean forests are chestnut, cork oaks, holm oaks and Pyrenean oaks. Chestnut (Castanea sativa Mill.) covers more than 2.5 million hectares in the Mediterranean Basin, being the only species of the genus native to this region. Chestnut has been widely cultivated as a monoculture for centuries and as a source of timber and fruit, although in recent decades it has been progressively felled (Garfì et al., 2022). Cork oak (Quercus suber L.) spreads over between 1.7 and 2.7 million ha of the Mediterranean Basin and has been cultivated for 3,000 years for the production of bark, representing the second most important marketable non-timber forest product (Gauquelin et al., 2018). Holm oak (Quercus ilex L.) forests comprise 4% of the forests of the Mediterranean Basin, being fully adapted to a wide range of soil and climatic conditions. In spite of the continuous anthropic exploitation of holm oaks during several centuries, the reduction in the use of their wood as fuel in the last 50 years is turning holm oak forests into high forests (Camponi et al., 2023). Pyrenean oak (Quercus pyrenaica Willd.) is found only in France, Spain, Portugal and Morocco. Pyrenean oak forests have historically been exploited by man for timber, and still are to this day, and constitute the main ecosystem of the Mediterranean Basin in diversity of hoverfly species (Perez-Luque et al., 2020; Martín-Pinto et al., 2023).

At present, the Mediterranean forest is in a situation of regression and dieback in several areas of the Mediterranean Basin, due to increased periods of drought, overexploitation, fire, soil degradation and the spread of pests and diseases (Peñuelas and Sardans, 2021). The decline of Mediterranean forests is a complex phenomenon due to multiple and very different causes that can occur slowly and over several years, ending with tree death (Gentilesca et al., 2017). In the case of chestnuts, drought (Waldboth and Oberhuber, 2009), or pathogens such as Cryphonectria parasitica (Elliott and Swank, 2008; Waldboth and Oberhuber, 2009), or Phytophthora sp. (Vettraino et al., 2005; Akıllı et al., 2012) have been reported as causes related to decline. In cork oak, causes include drought (Zribi et al., 2016; Camilo-Alves et al., 2017; El Ahmadi et al., 2022) and fungi such as Phytophthora sp. (Smahi et al., 2017). Or in the case of holm oak, drought events (Limousin et al., 2009) or pathogens such as Phytophthora cinnamomi (Frisullo et al., 2018) have been associated with decline.

Soil fungi play fundamental roles in the functioning of ecosystems, as decomposers of organic matter and as beneficial and detrimental organisms in interaction with plants (Wardle and Lindahl, 2014; Fernandes et al., 2022). In the study of soil mycobiome diversity, the DNA metabarcoding methodology represents one of the most widely used tools in the last decade (Wardle and Lindahl, 2014; Tedersoo et al., 2021). Precisely, metabarcoding has identified how disturbance events of natural and anthropogenic origin have a significant effect on the soil mycobiome of forests (Bowd et al., 2022; Martín-Pinto et al., 2023). This methodology of mycobiota diversity analysis has already been used in Mediterranean forest, being described as an effective way for the real knowledge of microbial diversity and functionality (Adamo et al., 2021a,b), also in situations of disturbance, such as fire (García-Carmona et al., 2021) or climate change (Diez-Hermano et al., 2022).

In chestnut, the study of soil fungal diversity by metabarcoding has been confirmed as a more comprehensive strategy than the study of fruiting bodies (Baptista et al., 2015). In the presence of the pathogenic oomycete Phytophthora cambivora, it has been reported by metabarcoding how soil mycobiota show resilience to the pathogen, and may play a key role in the forest’s resistance to the disease (Venice et al., 2021). The use of metabarcoding in the study of rhizospheric mycobiota of cork oaks has revealed the influence of abiotic factors, such as climate, on its composition (Costa et al., 2022). In addition, the diversity of soil mycobiota is closely related to the decline of cork oaks (Maghnia et al., 2019). In this sense, the fungal diversity most studied by metabarcoding in relation to decline is that of the holm oaks of Spanish “dehesas.” In this way, it has been possible to identify how pathogens such as Phytophthora cinnamomi in combination with others (such as Fusarium spp., Alternaria spp. or Pythium spiculum) and not in isolation are involved in the development of decline, as well as how fungi such as Trichoderma spp. or ectomycorrhizae improve the response of trees affected by these pathogens (Ruiz-Gómez et al., 2019a,b). With respect to Pyrenean oak, to our knowledge, the study of its rhizospheric/soil mycobiota by metabarcoding has been little addressed so far. For example, it has been possible to relate how different management strategies have an impact in soil fungal diversity (Martín-Pinto et al., 2023).

Therefore, the aim of this work is to relate soil mycobiota with the health status of Mediterranean forests. To this end, chestnut, cork oak, holm oak and Pyrenean oak have been selected as key forest species, obtaining rhizospheric soil samples from healthy and declining trees in forests with known presence of pests and pathogens. Soil mycobiota was analyzed by metabarcoding using the ITS region of ribosomal DNA.

Soil mycobiota play a key role in forest functioning and health (Fernandes et al., 2022). Therefore, knowing the fungal diversity of forest soils under different situations and health status is fundamental to understand the role that each component of this diversity plays in the sustainability of the ecosystem, being metabarcoding a proven tool for this purpose (Bowd et al., 2022). In our study, this methodology has allowed us to identify 674 different fungal genera as inhabitants of the rhizosphere of chestnuts, cork oaks, holm oaks and Pyrenean oaks.

Some fungal genera were only present in soils of declining trees. However, none of these genera has been previously described as a forest pathogen, which might be indicative of the soil conditions playing an important role as a predisposing or inciting factor of decline per se. Humicola is a genus of endophytic fungi of great biotechnological interest in recent years, mainly related to halophytic plants and production of chemical compounds for industrial use (Hosseyni-Moghaddam et al., 2020). Meliniomyces is a genus of endophytic fungi and ericoid mycorrhiza (Ohtaka and Narisawa, 2008; Vohnik et al., 2013). Pleorotus is a genus of widely known edible saprophytic fungus (Suwannarach et al., 2020; Doroški et al., 2022) which also includes biological control agents against nematodes (Singh et al., 2019), and its presence is most likely tied to clearings in declining areas. Polyphilus is a recently created fungal genus, which includes nematophagous fungi of eggs and cysts of plant-parasitic nematodes (Ashrafi et al., 2018). However, other important microbial communities with putative pathogenic species and capable of exerting direct and indirect effects on these forest species, such as bacteria, oomycetes or nematodes, have not been addressed here. Future work would be interesting in this regard, and could provide important information on their relationship with the health status of the trees as well as their impact on population dynamics of non-pathogenic species.

Similarly, we also found fungal genera exclusive to healthy forest soils, which could be related to control activity of different pathogens. Varicellaria is a fungal genus whose life form is algal symbiosis to form lichens (Schmitt et al., 2012) and could be related to the presence of more humidity due to better vegetative condition of the trees. A few species of Podila have been described as mycophilic fungi or mycoparasites (Telagathoti et al., 2022), but the absence of specific studies on their potential biocontrol effect prevents associating their presence with the absence of decline. Some genera found can act as animals’ pathogens, such as the genus Geomyces (Warnecke et al., 2012). The genus Sagenomella is widely known to cause systemic illness in animals (Gené et al., 2003), whereas the genus Scytalidium includes the causal agent of scytalidiosis and onychomycosis, human skin diseases (Elewski, 1996; Machouart et al., 2013). Interestingly, the genus Coleophoma, also absent from declining samples, has been widely described as a forest pathogen (Pehl and Wulf, 2003; Bianchinotti and Rajchenberg, 2004). Grazing activities that take place regularly in the sampling sites could offer an explanation to the presence of fungi pathogenic to animals, although their association with healthy trees soils would point to a preference of cattle for such areas that has not been explored.

Regarding biodiversity, a higher alpha diversity was observed in soils of healthy chestnuts than in declining trees. In this sense, other authors have previously described that the rhizosphere of chestnuts has a high diversity, even much higher than that of other forest ecosystems (Kelly et al., 2021). Moreover, this fungal diversity plays a key role in the resistance of trees against pathogens such as the oomycete Phytophthora cambivora (Venice et al., 2021). On the other hand, in both cork oak and holm oak, we found a higher fungal diversity in the rhizosphere of declining trees than in healthy ones. Therefore, a higher fungal diversity in these trees may not be related to the biological control of the decline, as has been described by other authors (Gómez-Aparicio et al., 2022). Finally, no differences in alpha diversity were reported between soils of healthy and declining Pyrenean oaks, a novel aspect that requires further studies. Differential abundance analysis indicated that the genus Metarhizium is more present in healthy chestnut soils than in declining soils. Metarhizium has been widely described as a biological control agent against plant pests and pathogens (Stone and Bidochka, 2020), which might be indicative of this genus playing a significant role in the absence of declining in the selected healthy chestnut stands. Influence of environmental drivers on biodiversity such as slope or elevation has not been evaluated and would be an important confusion factor to be considered in future works.

With respect to the assignment of primary lifestyles, most common functional niches found were EcM and soil saprophytes, very widespread and important functions in fungi that make up the mycobiota of all forest soils (Li et al., 2022). EcM fungi were dominated by some ubiquitous genera such as Cenococcum, Cortinarius, Inocybe, Laccaria, Russula and Tomenteria, with Russula being particularly more abundant in holm oak and Pyrenean oak soils regardless of health status (see complete list in Figure S5). In general, all oak species shared a predominance of EcM over other functional guilds. Inverse relationship between abundance of EcM and abundance of other guilds has been found before in oaks’ soils (Ruiz Gómez et al., 2019b). In contrast, chestnut showed evenness of EcM, saprotrophs and pathogens. We hypothesize that these differences in balance of functional guilds between oaks and chestnuts could be related to oaks’ soils being generally poorer and dryer and thus more prone to prevalence of EcM than chestnuts’ soils, where water is more readily available to the plant. The aforementioned guilds were also found to be more abundant in declining chestnut soils. Higher presence of opportunistic pathogens is to be expected in the rhizosphere of weakened trees, whereas saprobes might benefit from the excess of dead matter and leaf litter. Declining chestnuts could also have increased generation of roots as an attempt to compensate for compromised water transport, which in turn would allow for more colonization of EcM, however, this phenomenon was not observed in the oak species. Regarding other lifestyles, a significant proportion of AM belonging to 15 genera was found in chestnut and declining cork oak soils, whereas holm oak, Pyrenean oak and healthy cork oak soils showed a remarkable absence of AM (Supplementary Figure S5). In particular, presence of some genera such as Archaeospora, Entrophospora and Dominikia was tied to health status in cork oak soils, favoring declining samples. The persistence of AM has been reported in dying succulent trees in correlation with NO3⁻ levels in soil (Vivas et al., 2018), although we found no differences in nitrogen compounds between healthy and declining cork oak soils in our samples.

This work describes the fungal soil mycobiota of four Mediterranean forest species and explores differences between healthy and declining stands. Healthy chestnut soils were characterized by higher alpha diversity and higher abundance of the genus Metarhizium, but such relationships were not found in the rest of the forest species included in the study. Finally, with respect to the functional profile, declining chestnut soils harbored more abundance of EcM, saprobic and pathogenic guilds than healthy samples. EcM were the dominant lifestyle in all oak species regardless of health status, whereas AM were preferentially found in healthy cork oak soils. Further experimental validations would be necessary to assess whether functional shifts in soil mycobiota are causally related to trees’ health status.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://www.ncbi.nlm.nih.gov/genbank/, PRJNA953856 https://github.com/serbiodh/2023_FrontiersForests_Soil, GitHub.

IB-A collected the data. SD-H analyzed the data. JP, SD-H, and JD wrote the manuscript. IB-A, JN-S, PM-P, AP, and JD designed the sampling scheme. All authors contributed to the article and approved the submitted version.