Fungi as Ecosystem Engineer Species of the Pannonian Grasslands: The Effect of Fungal Fairy Rings on Grassland Vegetation

Abstract

Fungal fairy rings (FFRs) are circular patterns primarily formed by basidiomycete fungi. These structures significantly influence grassland ecosystems by mediating nutrient cycling, altering soil microbial communities, and driving shifts in plant community composition. The present study investigates FFR formed by Agaricus xanthodermus in a Pannonian sandy grassland, with a focus on vegetation structure, productivity, and diversity. Field surveys conducted along transects across FFR quantified plant species cover, height, and additional ecological parameters. The findings demonstrate that FFR alters species dominance, reduces diversity at the ring edge, and based on ecological indicator values of plant species it increases soil nitrogen, and modify the movement of water and nutrients within the soil. Collectively, these results suggest that FFRs function as ecosystem engineers, shaping ecological processes and affecting the agricultural potential of semi-natural grasslands.

1. Introduction

Among global ecosystems, temperate grasslands play a prominent role in preserving biodiversity [1,2,3]. In Europe most natural and semi-natural grasslands have been used for agricultural production for centuries [4]. As a result, especially since the early 20th century, the area of semi-natural grasslands has decreased significantly across Europe. This decline has been reinforced by land-use change [5,6,7], and also led to a decline in biodiversity [8]. The land-use change is particularly pronounced in sandy habitats, rendering sandy vegetation and sand grasslands among the most endangered habitats in Europe. This presents a significant challenge for nature conservation, as sandy soils support numerous specialised and endangered grassland species. These species are adapted to dry, nutrient-poor conditions [9] and benefit from disturbances such as sheep grazing [10] and soil disruption. Their decline has been extensively documented across Europe, often associated with eutrophication following the abandonment of traditional agricultural practices, which accelerates natural succession [11,12].

Grasslands deliver a range of ecosystem services [13,14], which are influenced by land use practices [15,16]. Fungal fairy rings (FFRs) produced by various species of fungi (e.g., Agaricus spp.) represent one such service [17]. Getzin et al. [18] refined the terminology of circular plant structures by distinguishing FFRs, which are formed by fungal taxa, from non-fungal Fairy Rings. The latter occur in desert grasslands in regions such as Namibia, South Africa, Angola, and parts of Western Australia, and their formation is primarily associated with low-precipitation environments [18].

Within basidiomycetes, the most extensively studied grassland fairy ring species include Marasmius oreades (Bolton) Fr., Agaricus campestris L., and A. arvensis Schaeff. In forest ecosystems, the ectomycorrhizal fungus Tricholoma matsutake (S.Ito & S.Imai) Singer is the most researched species and is primarily distributed in Asia, including Japan, the Korean Peninsula, China, and Russia, as well as in certain Northern European countries [19,20,21]. Advances in genomic research on native fungal communities have facilitated the identification of new species with strong indicator potential [22].

Multiple studies have examined the soil microbiome, including that of forest soils, and have detailed the impacts on fungal communities [23,24]. Fungal species maintain close associations with the soil microbiota they colonize and the plant communities present. Their influence on soil chemical parameters is well documented [25,26]. The balance among fungal symbionts, pathogens, and saprotrophs in the soil [27,28,29], as well as the degree to which plant matter decomposition affects the microbial community, is critical for understanding plant community patterns [30,31]. However, limited information exists regarding the impact and persistence of changes in soil-dwelling fungal and bacterial communities [32].

The effects of these fungi on vegetation are species- and habitat-dependent [33], resulting in considerable variability. The first systematic classification was conducted by Shantz and Piemeisel [34] in their foundational work, which remains widely used for classifying FFRs:

Type 1: has two vegetation zones: a narrow outer necrotic belt with infertile soil and sparse, yellowed or absent vegetation, and an inner belt with more vigorous, dark green vegetation.

Type 2: A single belt of dark green vegetation, no necrotic outer belt.

Type 3: No visible vegetation change, only seasonal fruiting bodies.

Salvatori et al. [35] divided type 1 rings into three further subtypes:

Type 1A: The outer vegetation belt is lower than the inner. A necrotic belt separates them. Type 1B: The outer stimulation zone is higher than the inner. Type 1C: No stimulation is present; identification relies on the necrotic zone.

However, this latter classification omits three-part witch’s circles, such as A. arvensis [36,37] and M. oreades [25,38]. These have a type 1 zone on the outer edge and a type 2 zone on the inner edge. Zotti et al. [39] summarised a comprehensive classification of FFRs.

Type 1 FFRs exert the most pronounced effects on vegetation. Similar to other disturbance factors such as drought, grazing, and fire, these fairy rings create conditions that favour taxa reliant on disruption by the advancing fungal front [32,34,36]. The hyphal progression in type 1 circles induces soil hydrophobicity [25], nutrient immobilization [40], pathogenic activity [26,41,42], production of phytotoxic compounds such as cyanides [43,44], and microbial imbalance [17], all of which negatively impact grassland plants. In most Agaricus species, the fungal front supports low species richness and diversity. Richness and diversity are higher at the outer and inner edges of the circle and lower in the center, though not as low as at the front [45]. A similar pattern is observed for fungi and bacteria [32]. In type 1 circles, plant growth increases in the belt as the front recedes, a pattern also observed in type 2 fairy rings. This effect may be attributed to plant-growth-promoting secondary metabolites, referred to as “fairy chemicals” [46,47,48], or to a beneficial microbial community that includes arbuscular mycorrhizal (AM) fungi.

The distinctive spatial patterns of FFRs influence landscape structure by redistributing ecological niches and affecting nutrient cycles [49,50]. Fairy rings may facilitate the colonization and spread of small, nitrogen-demanding species that are prone to invasion during mid-successional stages, such as Bothriochloa ischaemum (L.) Keng, Calamagrostis epigeios (L.) Roth, and Solidago spp. [51,52]. Higher nitrogen availability influences above-ground production in grasslands, which is significant not only for agriculture but also for various ecological processes and for maintaining the diversity of Pannonian grasslands, thereby supporting grassland persistence [53,54,55]. Fungal species that form fairy rings can enhance grassland yield [56,57,58], and their capacity to modify soil flora is also important [57,59,60]. Notably, these fungi increase the biological availability of nitrogen, other nutrients, and microelements [50,58], resulting in greater quantity and improved quality of pasture grass for grazing animals [61,62].

This study contributes to the development of a comprehensive database aimed at elucidating the multifaceted ecological roles of FRRs. Fairy rings in grasslands were examined from multiple perspectives and with various methods, including remote sensing techniques [63], which complement field studies and facilitate more detailed research. Remote sensing is particularly effective for detecting fairy rings [64,65]. These approaches are well-suited for investigating mosaic-like patterns in vegetation, aquatic systems [66], and terrestrial habitats such as FFRs. In agricultural contexts, remote sensing is widely used for crop estimation [67,68,69] and may have economic implications in the context of FFRs.

In depth, this case study investigates the influence of Agaricus xanthodermus Genev., a potential ecosystem engineer, on vegetation. The following hypotheses are proposed: (H1) A. xanthodermus significantly alters the abundance of sandy grassland species; (H2) the magnitude of this effect varies across concentric FFR bands; (H3) during fungal engineering, species with high nitrogen demand temporarily proliferate in the inner band due to nutrient release, resulting in a non-linear spatial pattern in plant community diversity; (H4) this pattern represents the spatial analogue of the Intermediate Disturbance Hypothesis (IDH).

2. Materials and Methods

2.1. Sample Area

The sand steppe was selected in the Hungarian Great Plain, nearer on the Pest Alluvial Plain [70], on the outskirts of Fót (latitude: 47°35′41.75″; Latitude: 19°10′52.26″; altitude: 136 m) (Figure 1). The sampling area is a Potentillo arenariae-Festucetum pseudovinae Soó (1938) 1940 grassland community of a closed sand steppe habitat, which was previously used as pasture, but since 2000 has been the site of a waterworks, with four disused water wells. The degraded, closing sandy grassland is mowed in July each year and is grazed by sheep, helping maintain the grassland’s biodiversity.

2.2. Method of Botanical Recording

Based on the aerial photographs and Google Earth images examined on 5 June 2025, the fungal colony forming the fairy ring is 5–6 years old (first visible in a Google Earth image taken on 28 July 2020). FFR is 14 m in diameter, measured using a measuring tape. The fungi consists of Agaricus xanthodermus and was determined based on morphological features. The FRR could be classified [35] as Type 1A. The inference of this study is based on one ring in one grassland only and that the replication across rings, sites and multiple years is needed. We recorded 0.5 × 0.5 m squares that touched along the six diagonals of the FFR (Figure 2), estimating species coverage as a percentage. The transects were 2 m longer than the diameter of the circle in both directions, so we also sampled the control grassland outside the FFR. This resulted in 216 squares in total. The coenological survey was conducted on 26 May 2025, at the beginning of the flowering season of the dominant grass species.

From the species coverage (c) and plant height (m) data (10 measurements/species, uppermost leaf tip), we calculated their relative yield (t) from Szentes et al. [71].

$$c\times m=t$$

The average grassland height (M) was calculated, the relative green mass of the grassland (T), which is the sum of the species relative green mass meaning yield, divided the total coverage.

M = T/∑c

The species names follow the nomenclature of https://europlusmed.org/ [72] (Supplementary Table S1).

2.3. Data Processing and Statistics

We evaluated the quadrats based on the relative nitrogen demand (NB) and relative water demand (WB) values from Borhidi’s relative ecological indicators [73] fitted to the Hungarian flora. Their naturalness status and trends were evaluated based on Borhidi’s social behaviour types (SBT) and their scores named naturalness values (VAL) [73], using a species-weighted average.

All analyses were performed in R (v4.5.2) [74] using the packages vegan [75] and ggplot2 [76]. The Shannon diversity index was used for diversity analyses. Within-subject diversity (α-diversity) was evaluated using observed species counts (richness). For this, species abundance profiles were examined using summed abundances per species and quadrats. To account for spatial dependence among contiguous quadrats, samples were aggregated into biologically meaningful zones within each diameter: Outside circular line (quadrats 1–4 and 33–36), Circular line as the fungal front (quadrats 5 and 32), and Inner ring (all remaining quadrats within each diameter). Species abundances were summed per Diameter × Zone × Species to generate the community matrix used for all inferential analyses.

Differences in community composition among zones were tested using PERMANOVA (adonis2) based on Bray–Curtis dissimilarities [77]. Permutations were constrained within diameters (strata = Diameter) to account for spatial autocorrelation along transects. Homogeneity of multivariate dispersion among zones was assessed using betadisper.

Community patterns were visualized using non-metric multidimensional scaling (NMDS; metaMDS, k = 2) based on the same Diameter × Zone matrix. Two NMDS ordinations were generated from this matrix: (1) points colored by dominant species per Diameter × Zone combination (Figure 3), and (2) points colored by zone with diameters distinguished by shape (Supplementary Figure S1). Stress values were reported to assess ordination goodness-of-fit.

Exploratory NMDS ordinations based on a pooled distance matrix (i.e., abundances summed across diameters for each distance position) were presented separately for descriptive purposes (Supplementary Figures S2 and S3). All figures were produced using ggplot2, with 95% confidence ellipses included where appropriate to illustrate group-level clustering.

3. Results

3.1. Species Composition and Dominant Species

Community composition differed significantly among zones (PERMANOVA: F = 36.92, R² = 0.83, p = 0.001), with no significant differences in multivariate dispersion (betadisper: F = 1.04, p = 0.38), confirming that PERMANOVA assumptions were met. NMDS ordinations revealed clear patterns: quadrats were separated by zones in multivariate space, with a stress value of 0.15 indicating good fit. The overall separation of zones and the spatial structure of diameters within zones are presented in Figure 3 and Figure 4 respectively. Confidence ellipses around zones illustrate distinct clustering, consistent with the PERMANOVA results. Festuca valesiaca subsp. parviflora (Hack.) Tracey was the dominant species outside the FFR (Figure 3, Supplementary Figure S2). On the fungal front (quadrat 5), Elytrigia repens (L.) Nevski became dominant. On the other side of the transect (quadrat 32), Poa angustifolia L. reached slightly higher coverage (34%) than E. repens (29%). Moving towards the centre of the circle, P. angustifolia gradually became dominant (Figure 4). This species was the only one present in all quadrats. Its average coverage exceeded 14.5% in each quadrat. Therefore, it can be considered the dominant species in the FFR.

Examining the more common species (Figure 4), Dactylis glomerata L. and Vicia sativa subsp. nigra (L.) Ehrh., Vicia cracca L. and Vicia hirsuta (L.) Gray had significant coverage. They thrived in the belt zone and in the outer part of this area.

3.2. Height of the Grassland

The species change resulting from the expansion of the fungal colony can also be clearly observed in changes in grass height (Figure 5). Outside the circle, F. valesiaca subsp. parviflora forms an average grass height of 32.7 cm. On the fungal front (quadrate 5), E. repens becomes dominant with an average height of 70.9 cm. In contrast, on the other side of the fungal front, in quadrate 32, P. angustifolia is also present with greater coverage alongside E. repens, which is why the grass is slightly lower here, with an average height of 58.8 cm. As we move towards the centre of the circle, becomes increasingly abundant, and with it, the average height gradually decreases. The lowest height is in the centre of the circle, where the P. angustifolia has formed a monodominant patch, but at 36.9 cm, it still exceeds the average height outside the circle.

3.3. Relative Ecological Values

Based on the species’ relative water demand, there is less variation between the individual bands (Figure 6). An increase in values can also be observed in the belt ring of the FFR and on its inner side (mainly in quadrats 25 and 29). This is primarily due to the presence of Centaurea jacea subsp. angustifolia (DC.) Gremli, D. glomerata and Medicago lupulina L.

Based on the relative nitrogen demand of the species, a sudden spike in the values can be seen on the fungal front (squares 5 and 32), from 3.5 to 5.8, and from 3.6 to 4.7, respectively. As we move towards the centre of the circle, the value falls below the values outside the circle due to the dominance of P. angustifolia (Figure 7). The larger jump at the beginning of the transect is due to the homogeneous closed stand of E. repens, while on the other side of the transect, species valuable from a forage perspective, such as D. glomerata, Arrhenatherum elatius (L.) J. Presl & C. Presl, V. sativa subsp. and V. hirsuta raised the value.

3.4. Social Behaviour Types

The boxplots depicting the naturalness value show a trend that is precisely the opposite of that for relative nitrogen demand (Figure 8). In the control area, the dominance of F. valesiaca subsp. parviflora as a natural competitor resulted in the most natural quadrats. Subsequently, a sudden and very strong degradation can be observed on the fungal front, which is due to the dominance of the ruderal competitor E. repens. The reason for this is that the fungus suddenly made a relatively large amount of nitrogen available to the plants. As the plants utilise this, the value increases continuously towards the centre of the circle, but at this point, it has not yet reached the values of the control area.

3.5. Diversity

Figure 9 shows the change in Shannon’s diversity index in transects taken along the diagonals. The first and last four points show the control grassland quadrats. It can be seen that diversity decreases significantly on one side of the circle (quadrat 5) at the fungal front. Here, the dominant species is E. repens, whereas on the other side, only a slight decrease is observed (square 32). Here, the co-dominant species are P. angustifolia and, in places, D. glomerata. Then, moving towards the centre of the circle, diversity increases to 7 and 5.5 m, respectively (1.7 and 1.9), then drops sharply to a minimum of less than 0.4 in the centre, which is formed by a homogeneous patch of P. angustifolia.

4. Discussion

Despite their ecological importance and agricultural potential, little scientific research has been conducted on FFRs, even though several authors emphasise the indispensable role of interdisciplinary research in the complex exploration of the diverse ecological processes associated with them [15,17,39,63]. Despite their role as ecosystem engineers, fungi have only recently been included in nature conservation programmes, following the decline of fungal species in grassland habitats, which justified the launch of conservation measures [78].

4.1. Confirmation/Rejection of Hypotheses

H1. 

A. xanthodermus significantly alters the abundance of sandy grassland species. The results show an apparent change in the dominant species. F. valesiaca subsp. parviflora dominates the control area; meanwhile, E. repens takes over on the fungal front. Subsequently, inside the ring, P. angustifolia becomes dominant (Figure 4 and Figure 5). This replacement, in turn, affects the frequency of grassland species and changes dominance relationships.

H2. 

The results show clear belt zones (outer, front, interior) with different species composition, plant height, nitrogen demand (the amount of nitrogen required by plants to grow), and diversity (the variety of species present) (Figure 3, Figure 5, Figure 7 and Figure 9). The hypothesis is true.

H3. 

Due to its nutrient-releasing effect, species with higher nitrogen demand temporarily proliferate in the inner zone. This pattern is further illustrated by Figure 7, which clearly shows that the relative nitrogen demand of grassland species increases significantly at the fungal front (squares 5 and 32). At this location, the fast-growing E. repens, with a high nitrogen demand, becomes dominant, thereby supporting the hypothesis perfectly.

H4. 

Grassland plant diversity temporarily decreases and then increases in the inner strip. Figure 9 (Shannon diversity) shows this pattern: first, a sharp decrease at the fungal front; next, an increase as disturbance lessens (inside the ring, around squares 8–9); and finally, another decrease in the monodominant centre.

4.2. Agaricus xanthodermus as an Ecosystem Engineer

Our results clearly confirm that A. xanthodermus is an ecosystem engineer species, as it causes profound changes in the functioning, species composition and spatio-temporal diversity of grassland communities [79], and is also a special example of plant–soil interactions in grassland ecosystems [17,36,80]. The plant bands they form (dominance shifts, height and diversity gradients) are direct consequences of their ecosystem engineering, which results from their multidimensional activities. Their physical activity involves altering the soil structure (hydrophobicity, aggregation) and the flow of water and nutrients through the hyphal network [25,26]. Their chemical activity involves nitrogen mobilisation [40,57], the production of “fairy chemicals” [48,81,82], and changes in soil pH and redox status [80]. Their biological activity involves transforming the composition and diversity of the soil microbial community (bacteria and AM fungi) [59,83,84,85], thereby affecting the plant community. Thus, a direct outcome of their ecosystem engineering activity is increased landscape heterogeneity in the form of habitat mosaics. The different conditions prevailing in different bands (high N, low N, allelopathy, etc.) allow different plant communities to coexist, thereby increasing the overall landscape (γ) diversity. However, their characteristic spatial patterns not only reshape the landscape but also significantly impact niche redistribution and nutrient cycles [49,50]. This can reduce the likelihood of colonisation by small nitrogen-demanding species prone to invasion in the middle stage of succession, such as Bothriochloa ischaemum [86,87].

A. xanthodermus radically changes the frequency of sandy grassland species. Instead of F. valesiaca subsp. parviflora, E. repens became the dominant species in the belt. This process had a significant impact on many ecological parameters of the grassland. As the former species is short grass and the latter is tall grass, the change in dominance has significantly increased the average height of the grassland. The height of grassland species is a highly informative indicator of the vertical structure of grassland vegetation and a vital determinant of above-ground biomass [88,89]. It can also be a good indicator for monitoring changes in biodiversity [90]. The phytomass of grasslands plays an essential role in preserving diversity and maintaining the natural state of grasslands, not only from a management perspective but also for ecological and conservation reasons [91]. In Pannonian sandy grasslands located in dry, nutrient-poor habitats, the amount of above-ground phytomass has a positive effect on species in the event of slight disturbance [92]. In these habitats, therefore, increased production may contribute to species richness [93,94]. However, increased nutrient availability promotes succession and may lead to the decline of stress-tolerant specialised species due to the overgrowth of stronger competitors [95,96,97].

4.3. Implications for the Intermediate Disturbance Hypothesis (IDH) Theory

Species richness depends on numerous factors, such as geographical location [98,99], habitat extent [100,101], fragmentation (breaking up of habitats into smaller patches) [98,102,103], community type [104], habitat succession status (stage of ecological development) [100], and grassland production [93]. The highly competitive E. repens limited diversity in the belt, as its extensive tillering system (production of side shoots) enabled it to rapidly take up and utilise the nutrients that became suddenly available through the fungus. The tall, dense stand it formed, the light limitation it exerted on the soil surface, and the allelochemicals (chemicals released by plants that affect other organisms) it produced [105] prevented Shannon diversity (a measure of community diversity) from increasing. One of the main factors contributing to the increased growth of the inner vegetation may be the proliferation of AM (arbuscular mycorrhizal) fungi in the belt ring [32], which promotes the species richness of higher plants, soil microbiota (microorganisms), and mycobiota (fungal community) [17]. In other words, they can increase plant gamma diversity (overall diversity across habitats) in grassland ecosystems by creating new niches, even when the dominant species limits vegetation diversity within the strip. No such effect has been observed in type 3 circles to date.

Thus, diversity reached its maximum value in both directions around the 8th-9th quadrant, i.e., at half the radius of the ring (between 4 and 5 metres). The diversity pattern found—a sharp decline at the front, a maximum at half the radius of the ring, and then a repeated decline at the monodominant centre—an excellent spatial example of the diversity pattern that has developed is the IDH [106], for which the study must be repeated between rings and seasons.

The fungal front is a strong disruptive factor that reduces α-diversity, thereby allowing for a rearrangement of the competition hierarchy. As a result, only a few fast-reacting species, such as E. repens and P. angustifolia, can dominate. As the effect of disturbance decreases (towards the inside of the ring), diversity increases as competition becomes more equal and more niches become available. In the centre of the circle, the monodominance of P. angustifolia reduces diversity again. In addition, outside the circle, low production due to nutrient deficiency and the dominance of F. valesiaca subsp. parviflora; in the inner zone, the dominance of highly competitive species capable of high above-ground production, such as E. repens, limited diversity. This species quickly absorbed the nutrients made available by the fungus through its extensive tillering system. Thus, it quickly formed a tall, dense stand, which prevented increases in Shannon diversity due to the light limitation it imposed on the soil surface and the allelochemicals it produced [105].

The reason why diversity reaches its maximum at half the radius of the ring may be that at this distance, a significant portion of the nutrients made available by the fungus has already been utilised, which is supported by the relative nitrogen demand of the species (Figure 7). This result confirms the findings of Fraser et al. [93]. According to this, the highest diversity is expected in medium production, since in low production, site stress limits diversity, whereas in excessive production, dominant or codominant species limit diversity. Diversity showed a decreasing trend from half the ring radius towards the centre, due to the formation of a homogeneous P. angustifolia stand. The relative nitrogen demand of this species is similar to that of F. valesiaca subsp. parviflora.

By mobilising nutrients and modifying soil properties, they affect the entire grassland community, from soil bacteria [32,60,107,108] to grazing animals [61,62]. Their impact on biomass production is particularly noteworthy [56,57,58], in which the altered bacterial flora also plays a significant role [59,94], particularly through increased availability of biologically important nitrogen and other nutrients and microelements [5,58,109].

4.4. Implications for Farming and Nature Conservation

Fairy chemicals can enhance plant resistance to environmental stresses such as cold, heat, and salt, and increase crop yields; they may also have biological activities, such as anti-ageing and anti-tumour effects [47,81,110]. Their yield-enhancing effects [111] have already been demonstrated in the case of wheat [48,82,111], which is of outstanding economic importance.

The results of Friesová et al. [112] indicate relatively small changes in sand vegetation over a two-year study period. This is consistent with studies reporting the long-term stability of sand vegetation under appropriate disturbance regimes [113,114,115]. Thus, in these grasslands, FFRs play a crucial role in increasing landscape diversity by facilitating vegetation dynamics processes [32], especially in habitats under anthropogenic influence [17,59,84].

4.5. Limitations and Future Research

To understand their multifaceted ecological role as described above, the first step is to map FFRs and create databases, which require the use of remote sensing techniques [73]. This supports more detailed research and enables field studies. Evaluating data requires a holistic approach. The results obtained in this way are not only beneficial for ecology. Still, they can also be of great help in understanding the resistance of diverse grassland communities to climate change, in grassland restoration work, and in agricultural production.

5. Conclusions

This study demonstrates that FFR formed by Agaricus xanthodermus acts as an ecosystem engineer in Pannonian sandy grasslands. The phenomenon fundamentally alters the composition, structure, and functioning of plant communities. The rings create concentric zones with distinct shifts in dominant species and ecological indicators. The original grassland dominated by F. valesiaca subsp. parviflora, is replaced at the fungal front by fast-growing, high-nitrogen-demanding species (E. repens, D. glomerata), while towards the ring centre, P. angustifolia becomes monodominant. This species succession causes a significant increase in average sward height and aboveground biomass, primarily due to the release of plant-available nitrogen mobilised by the fungus.

The dynamics of diversity (Shannon index) follow a pattern consistent with the Intermediate Disturbance Hypothesis, showing how the fungal engineer creates a spatial disturbance gradient that drives succession and competitive release. At the landscape scale, the distinct plant communities in each ring belt significantly enhance beta-diversity, underscoring the role of FFRs as architects of ecological heterogeneity.

The FFRs act as natural, small-scale disturbance agents, enhancing landscape heterogeneity (beta-diversity) and fostering complex plant–soil–fungus interactions. Our findings emphasise that the bottom-up ecological effects of fungi must be integrated into grassland management and conservation strategies, particularly for maintaining the biodiversity of semi-natural grasslands and their resilience to climate change. Further research should combine remote sensing, molecular techniques, and long-term monitoring to understand this phenomenon fully.