Coffee (Coffea arabica L.—Rubiaceae) is the second most valuable commodity in the world after oil. It is a major crop in many countries in Latin America, Africa and Asia. Colombia is the third-largest global coffee producer after Brazil and Vietnam (de Oliveira Junqueira et al., 2019). Colombian coffee cultivation produces income that accounts for 15% of the agricultural GDP and employs 2.5 million people annually (Cano-Sánz et al., 2013). The success of coffee cultivation in Colombia is mainly due to rich volcanic-ash soil, abundant annual rainfall (200 mm per month on average), and high altitudes (1,800–2,000 m) where the Arabic coffee grows (Lara-Estrada and Vaast, 2005). However, such conditions overlap with areas of enormously rich biodiversity (Komar, 2006). Consequently, coffee represents a promising crop in the described environment but entails an undeniable impact on local and global biodiversity (Jha et al., 2014).

Different cropping strategies for coffee cultivation have a particular impact on biodiversity (Moguel and Toledo, 1999). Agroforestry, also known as shade-grown management, supports coffee plantations in the understory of native and exotic trees. Diversified shade management has been the basis of traditional coffee plantations and offers a refuge for biodiversity, such as arthropods, mammals, amphibians and birds (De Beenhouwer et al., 2013; Perfecto et al., 2014), with tree leaf litter being an important source of carbon that modulates soil biota structure and functions (Rao et al., 2020). Coffee farm management is gradually shifting away from shade-grown management in favor of sun-intensive, higher-yield farms (DaMatta, 2004; Bravo-Monroy et al., 2016; Velmourougane and Bhat, 2017; Harvey et al., 2021). Compared to the available information about the impact of coffee cultivation on aboveground diversity, relatively little is known about belowground microbial diversity.

Among soil-borne microorganisms, soil fungi regulate key ecosystem processes, including plant productivity and carbon mineralization and sequestration. They are essential decomposers, mutualists, and pathogens (Webster and Weber, 2006). Certain pathogenic and mutualistic fungi have been extensively studied in coffee (Lovera et al., 2022; Lu et al., 2022b). For example, in the period from 2008 to 2011, until the appearance of resistant cultivars, Hemileia vastatrix caused a massive outbreak of coffee leaf rust in C. arabica plantations with significant economic losses (Cristancho et al., 2012; Talhinhas et al., 2017; Gichuru et al., 2021). On the other hand, the mutualistic association between coffee and arbuscular mycorrhizal fungi (AMF) has been well documented (Urgiles-Gómez et al., 2021; Lovera et al., 2022). Nevertheless, there is scarce information about overall soil fungal communities and their ecological functions, which are of utmost importance for sustainable crop management (Duong et al., 2020; Rao et al., 2020).

Soil fungal responses are context-dependent, and local conditions determine the resultant assemblage and functions of fungal communities (Lekberg et al., 2021; Wang et al., 2021; Tedersoo et al., 2022). For example, fungal diversity showed high sensitivity to the legacy effects of land uses in the past (Turley et al., 2020; Correia et al., 2021) and along altitudinal gradients (Ogwu et al., 2019). In coffee plantations, earlier studies revealed that the soil fungal community is strongly influenced by agronomical management (i.e., conventional vs. organic) and the canopy composition in agroforestry systems on Nicaraguan farms (Jurburg et al., 2020), while altitude, the regional floristic domains and Coffea species were the main drivers of fungal diversity on Brazilian farms (Veloso et al., 2020, 2023). However, in situ surveys of soil fungi are missing in important and traditional coffee-growing regions of Colombia. Notably, unsustainable farm practices can favor pathogenic species (Jurburg et al., 2020; Rao et al., 2020), and impoverish the pool of mutualistic fungi from which plant roots can be colonized (Brinkmann et al., 2019; Prates Júnior et al., 2019; Lovera et al., 2022). In addition, valuable agents against pathogens and pests in coffee plantations, such as entomopathogenic and endophyte fungal species, are also sensitive to management practices (Vega et al., 2008; Duong et al., 2020; Bayman et al., 2021).

This study aimed to identify and describe the community composition and ecological functions of soil fungi in coffee farms in three traditional coffee-growing regions of Colombia. We conducted our study in six farms with divergent management practices (one conventional and five organic), which also varied in shade management, the plantation age and environmental properties. Certainly, we acknowledge that the multiplicity of conditions that characterize the farms across the studied regions can determine the existing fungal community (Lekberg et al., 2021; Wang et al., 2021; Tedersoo et al., 2022). For this reason, we focused on the identification of abundant genera, indicator species, and functional guilds. Despite the aforementioned sampling limitations, this investigation offers insight into the soil fungal taxonomy and ecological functions and provides an empirical baseline supporting the implementation of sustainable agricultural practices in Colombian coffee plantations.

Agriculture generates artificial systems through multiple human activities, which greatly alter the resources and environment, and affects the biodiversity patterns of pre-existing biota (Sun et al., 2016; García-Delgado et al., 2019; Vukicevich et al., 2019). Here, we evaluated the composition of soil fungal communities in six coffee farms across three main coffee-growing regions in Colombia (Cauca, Risaralda, and Magdalena). These regions exhibited particular climatic and edaphic conditions, typical of soils intended for coffee production (Vaast et al., 2006). We found differences in community composition and ecological functions of soil fungi, which could be expected given the distinct geographical locations (Tedersoo et al., 2014; Nilsson et al., 2019), altitudinal gradients (Ogwu et al., 2019), topology (Veloso et al., 2020), floristic domains (Veloso et al., 2023), and inherent soil properties (Sun et al., 2016; Jurburg et al., 2020; Ning et al., 2021). However, to the best of our knowledge, the great predominance of the saprotroph-symbiotroph Mortierella and the genus Saitozyma is an unprecedented result that highlights the need for further investigation to disentangle the potential benefits as well as the consequences of this finding on the maintenance of sustainable coffee plantations (Duong et al., 2020).

We found that the conventional long-term shaded farm showed a diverse fungal community and high number of indicator species. Ascomycota and Basidiomycota exhibited the highest abundances (33.7 and 44%, respectively) in Mag_1, the farm under conventional management, and were non-dominant on the remaining farms (2–22% and 1–8%, respectively). On the other hand, Mortierellomycota dominated fungal abundance, with up to 88% of the total number of reads in a single sample belonging to the organic farms. Such differences in phyla distribution are consistent with further taxonomical and functional indicators evaluated in this study (i.e., indicator species, abundant genera, and abundant trophic modes).

Ascomycota, Basidiomycota, and Mortierellomycota are considered ubiquitous in natural and anthropogenic soils (Yang et al., 2019; Chen et al., 2020; Wang et al., 2021). Ascomycota and Basidiomycota comprise a great variety of species displaying a wide range of lifestyles (Webster and Weber, 2006). In a previous report on soil fungi in orchards in Boyacá Department, Colombia, Landinez-Torres et al. (2019) found that ~60% of the abundance belonged to Ascomycota followed by ~20% Basidiomycota and ~ 15% Zygomycota (formerly the phylum that contained Mortierellomycota). This study observed a similar pattern of relative abundance across the seven sites assessed in their survey (Landinez-Torres et al., 2019). Guevara (2005) found that conversion of natural forest to coffee plantation greatly affected the abundance of Basidiomycota, particularly mycelial cord-forming fungi, due to changes in soil micro-environmental conditions.

Mortierellomycota is frequently represented in investigations of soil fungi. It has been identified among generalist fungi of cropping systems (Toju et al., 2018; Grzadziel and Gałazka, 2019; Wang et al., 2021), and among the most characteristic micro-fungal species in montane cloud forest (Velez et al., 2021). Mortierellomycota became highly dominant in soils after the addition of organic amendments (Li et al., 2018, 2021), and high soil P availability negatively affected Mortierellomycota abundance (Detheridge et al., 2016; Orrù et al., 2021). Hence, considering that our study was conducted mostly in P-impoverished soils, we infer that such conditions might have created a favorable trophic niche for Mortierellomycota development (Sun et al., 2016; Li et al., 2021). Indeed, Mortierella was the predominant genus in five out of the six studied coffee farms, and it was identified as indicator species of organic farms. This result can be associated with the application of organic amendments on the organic farms (Li et al., 2018, 2021), which are not generally used on conventional farms as these rely on mineral fertilization (Perfecto et al., 2019).

The fungal functional guild analysis revealed a strong association between predicted functions and taxonomic composition. The great predominance of Mortierellomycota is consistent with the prevalence of the saprotroph-symbiotroph trophic mode and the representative species detected in this study (i.e., Mortierella alpina, M. elongata, M. exigua, and M. minutissima). It is noteworthy that the genus Mortierella comprises several phosphate-solubilizing species (Osorio Vega et al., 2015; Ceci et al., 2018), as well as species able to assist roots and mycorrhizal fungi in P acquisition (Tamayo-Velez and Osorio, 2017). Our results highlight the potential role of Mortierella in carbon turnover and P acquisition on the surveyed farms, particularly in recently established coffee plantations suggesting this genus can be considered a beneficial component of the soil fungal community.

Potentially pathogenic fungi were abundant in the studied soils and included species of crucial agronomical importance. Animal, plant, and insect pathogens are the main functional guilds of pathotrophs (Nguyen et al., 2016). Plant pathogens were found in very low abundance across the surveyed farms, and only Thanatephorus (teleomorph of Rhizoctonia solani) was found to be a potential coffee disease (Priyatmojo et al., 2001). This pathogen is not a menace in productive plantations and has rarely been reported in coffee plantations (Duong et al., 2020; Lu et al., 2022b). Apiotrichum and Fusarium were among the dominant genera, and this finding was previously reported in coffee plantations and montane forest (Arias and Abarca, 2014; Velez et al., 2021). Apiotrichum and Fusarium have been described as devastating pests affecting coffee production in Africa, where they usually infect mature trees and affect coffee quality by the premature ripening of coffee beans (Rutherford and Phiri, 2006). Although farmers did not report damage caused by the fungal species in this study, they are a potential menace that deserves further consideration.

Conversely, insect pathogen species belonging to Metarhizium, in particular M. anisopliae and M. acridum, were characteristic of the conventional long-term shaded farm (Mag_1) (Figures 4, 6). In addition to their well-documented entomopathogenic function, Metarhizium can also colonize plant root tissues as an endophyte, resulting in increased tolerance against pests and diseases (Behie and Bidochka, 2014; Altinok et al., 2019). In coffee plantations, M. anisopliae is used as a biopesticide to control nematodes (Tylenchida: Heteroderidae) and the coffee berry borer (Coleoptera: Cucurlionidae), which cause significant economic losses in coffee crops (Escobar-Ramírez et al., 2019; Cure et al., 2020; de Oliveira et al., 2021). It remains to be established whether the sequences identified in this study are vestiges of the commercial products applied on the farm in recent years or the native inhabitants of the soil microbiome.

Saprotrophs were among the abundant trophic modes with a high number of species with low prevalence (i.e., <25%) and low abundance per site (i.e., <10%) that were grouped as “other.” Dominance of saprotroph was reported earlier in tropical forests and coffee plantations suggesting that saprotrophs are key players in nutrient cycling (Looby and Treseder, 2018; Velez et al., 2021). Among the surveyed farms, we found that Ascomycotan Arachnotheca glomerata (Onygenaceae) and Basidiomycotan Geastrum spp. (Geastraceae) were prevalent with low abundance. A. glomerata possesses cellulose degradation abilities (Cano et al., 1987), while Geastrum has been reported to be an ectomycorrhizal fungus (Hibbett et al., 2000). The Cauca farm exhibited a distinct abundance of Pseudocoleophoma sp. that accounted for >75% of reads within the saprotrophic group and was identified as an indicator species of this site. Pseudocoleophoma is a genus belonging to Dictyosporiaceae and was recently isolated from decayed wood of Coffea arabica plantations in China (Lu et al., 2022a). According to the authors, their finding represents the first report of coffee-associated fungi in Dictyosporiaceae, and our results support this novelty. Probably, Pseudocoleophoma is a common saprotroph soil fungus without any record in coffee plantations due to the limited surveys of fungi on coffee debris.

We found a low abundance of Pathotroph-Saprotroph-Symbiotroph, Pathotroph-Saprotroph, and Pathotroph-Symbiotroph. Among these groups, we would like to highlight Trichoderma, which soared up to 2.5% of relative abundance in Mag_1 and was among the abundant genera. Trichoderma’s antimicrobial action has been used as a biocontrol agent against diverse phytopathogens in agroecosystems (Láng, 1936; Zamanizadeh et al., 2011). Mulaw et al. (2010) and Hoyos-Carvajal and Bissett (2011) isolated Trichoderma species inhabiting the rhizosphere of coffee plants. Arias and Abarca (2014) found Trichoderma as one of the most abundant and frequent genera among the saprotrophic fungi in coffee plantations. These authors suggest that its potential to produce large quantity of spores that are able to colonize a wide variety of substrates coupled with the ability to produce mycotoxins might constitute a competitive advantage of Trichoderma over other soil organisms in coffee plantations.

Among the fungi without a functional annotation, Papiliotrema laurentii (Basidiomycota), Cladosporium sphaerospermum, and Saitozyma podzolica were detected as indicators species. P. laurentii is an oleaginous yeast frequently found in soil, characterized by its ability to produce enzymes that degrade a wide diversity of carbon sources (de Almeida et al., 2022). In addition, P. laurentii can enhance mycorrhizal colonization of roots (Sampedro et al., 2004). Cladosporium sphaerospermum is usually found in soil air and has been reported to be a plant growth-promoting fungus (Hamayun et al., 2009). Dietzel et al. (2019) reported that C. sphaerospermum is a pathogenic fungus highly found in dust. Hamdouche et al. (2016) identified C. sphaerospermum among post-harvest fungi in Coffea arabica beans. Despite the ecological and agronomical role of P. laurentii and C. sphaerospermum in this survey, both were detected in limited abundance (i.e., <1%). In contrast, Saitozyma (Basidiomycota) was the second most dominant genus in our study (i.e., up to 50%). It has been recently cited as a common member of soils associated with Coffea arabica plantations in Brazil (Veloso et al., 2023). Although the FUNGuild database did not assign any ecological function to this genus, Veloso et al. (2023) propose that Saitozyma can be used as a starter culture in coffee fermentation to improve the sensory perception of the coffee beverage due to its ability to degrade particular lignocellulosic compounds. In addition, according to Starmer and Lachance (2011), Saitozyma species are key soil yeasts that can act as saprotrophs, mutualists and parasites. Yurkov et al. (2012) and Yurkov (2018) proposed this species as a potential indicator of acidic, well-drained, soils. Future investigations are necessary to disentangle the ecological role of Saitozyma in tropical and subtropical soils under anthropogenic use as well as the potential implications in further coffee beverage processing.

Admittedly, evident design constraints affected the generalizability of the findings of this study. One of the main limitations of this research was the restricted sampling design, associated with the absence of comparable farms in the sampled regions and scarce funding. We are aware that our results might not be representative of heterogeneous regional agroecosystems. Studying farms with comparable land practices is pivotal to disentangling agronomical-based causal effects, and here we attempted to explore the broadest area in order to include more sites from unexplored regions. Given the persistent lack of data on coffee soil microbiomes, our study makes a valuable contribution not only to local coffee farmers but also to global datasets of biodiversity from these poorly studied soils.

This exploratory survey sheds light on the composition of soil fungi in a poorly studied area with the goal of providing important information for ensuring a sustainable food production. Different fungal communities prevailed across the sampled regions. The saprotroph Mortierella was broadly frequent and dominant. Saitozyma was prevalent in the long-term agroforestry farm, and even though we could not assign an ecological function, this yeast may hold a valuable potential to improve the beverage quality of coffee. The identification of beneficial fungi such as Metarhizium and Trichoderma, widely used as biological control agents in Colombia’s coffee production, is an indication of the health of coffee soils in the country. In addition, no major pathogens of coffee roots, such as Rosellinia, were identified in any of the samples.

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 in the article/Supplementary material.

VO-H: Investigation, Methodology, Writing – original draft, Writing – review & editing. VF: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Supervision, Writing – original draft, Writing – review & editing. LG-G: Investigation, Methodology, Writing – original draft. MR: Funding acquisition, Investigation, Project administration, Resources, Supervision, Writing – original draft. MC: Conceptualization, Funding acquisition, Investigation, Methodology, Resources, Supervision, Writing – original draft, Writing – review & editing.