Influence of Golden Moles on Nematode Diversity in Kweek Grassland, Sovenga Hills, Limpopo Province, South Africa

Abstract

This study investigates the impact of golden moles (Amblysomus sp.) on the abundance, diversity, and community structure of nematodes in kweek grass (Cynodon dactylon) within the Sovenga Hills of Limpopo Province, South Africa. Eight sites were sampled: four with active moles (sites: M1–M4), and four without (sites: T1–T4). Eighty soil samples were collected, and nematodes were extracted. A total of 23 nematode genera were identified, including 3 plant-parasitic and 20 free-living genera. The frequency of occurrence (FO) data showed that Aphelenchus sp. and Acrobeles sp. were the most prevalent nematodes, each occurring in 87.5% of the samples. In contrast, Eucephalobus sp., Tripylina sp., Discolaimus sp., and Tylenchus sp. had the lowest FO, appearing in only 12.5% of samples. The diversity indices (the Shannon index, the maturity index, and the plant-parasitic index) showed significant differences between the two environments. The Shannon index (H′) and maturity index were the most effective indicators of ecosystem disturbance. The lowest H′ was found at T4 (1.7 ± 0.2), compared with a higher value at M1 (2.4 ± 0.1). The principal component analysis (PCA) results revealed a positive correlation between Ditylenchus and the clay in the soil. In addition, Cervidellus was associated with soil pH. Network analysis revealed increased complexity in the nematode community structure at mole-affected sites. These findings suggest that mole activity alters soil properties and indirectly affects nematode diversity and trophic structure.

1. Introduction

Soil serves as a rich habitat for a diverse range of microorganisms, including essential bacteria, fungi, and nematodes, all of which play essential roles in maintaining ecosystem health. Among them, nematodes are particularly important for regulating pest populations, decomposing organic matter, and facilitating nutrient cycling. Because of these functions, nematodes are widely used as bioindicators of soil health due to their sensitivity to physical and chemical soil disturbances and their roles in nutrient cycling [1,2], offering insights into the effects of pollution, land use, and farming practices [3,4,5]. While many nematodes contribute to ecosystem function, some plant-parasitic species may hinder crop growth and yield [6].

Moles, small burrowing mammals, also influence soil ecosystems in complex ways. Burrowing mammals like kangaroo rats (Dipodomys spp.) and mole rats are known to bioturbate soil, influencing hydrology and plant–soil interactions [7,8]. Their tunneling may damage plant roots, but it also improves aeration and helps control insect populations. In South Africa, golden moles, such as the golden mole Neamblysomus julianae Thomas and Schwann, 1906, are common in Limpopo Province and may play a significant ecological role [9,10,11].

Golden moles, which are native to Sub-Saharan Africa, are insectivorous mammals known for their burrowing behavior, which significantly modifies soil structure. They are commonly associated with sandy soils throughout South Africa [12]. Their activities could potentially influence soil fauna, including nematode communities, but this relationship remains poorly studied.

Molehills are often viewed as unsightly disturbances in the landscape, detracting from the overall aesthetic of gardens and lawns. These mounds of soil not only create an unpleasant appearance but also serve as breeding grounds for invasive weeds, including species like Cirsium, which can quickly overtake desirable vegetation [13]. Furthermore, the tunneling activities of moles may lead to the mixing of subsoil, potentially affecting forage quality. This contamination can result from the mixing of soil with organic matter and debris, potentially affecting crop quality and yield [14,15]. Moreover, research findings revealed that both community diversity and species richness in grassland ecosystems in the Swabian Alb, Baden-Württemberg, Germany, were markedly diminished on molehills compared with areas of undisturbed vegetation [16]. This reduction in biodiversity highlights the ecological impact of molehill formation, which disrupts the native plant communities and alters the habitat conditions essential for various species, including nematodes. Certain plants, including the delicate German knotweed (Scleranthus annuus L.) and the vibrant yellow bedstraw (Galium verum L.), have been noted to thrive in the vicinity of molehills. Their presence suggests a relationship between these resilient species and the unique soil dynamics created by moles [17]. However, the underlying ecological reasons for this association have not been extensively explored. Additionally, factors such as the timing of mole activity and prevailing climatic conditions could significantly influence the types of vegetation that flourish atop these earthy mounds.

Currently, no published research directly examines the relationship between nematode diversity and mole activity in South African grasslands. Sovenga Hills in Limpopo Province provides a suitable setting to explore this relationship, as it supports native grasses, particularly kweek grass (Cynodon dactylon (L.) Pers., and is home to golden moles that frequently burrow in the root zone.

The objectives of this study were to (1) characterize the composition of soil nematode communities in kweek grasslands of Sovenga Hills and (2) evaluate how golden mole activity and soil physicochemical properties influence nematode diversity and structure.

2. Materials and Methods

2.1. Soil Sampling

Soil samples were collected from kweek grass (C. dactylon) sites located in Sovenga Hills, Limpopo Province, South Africa (GPS coordinates: 23°53′28.3″ S, 29°44′31.3″ E). Eight sampling sites were selected: four with visible signs of golden mole activity (designated as M-sites: M1–M4) and four without mole presence, representing undisturbed grass growth (T-sites: T1–T4) (Figure 1).

A total of 80 soil samples were obtained, with 10 subsamples per site. Each sample was collected from a depth of 10–30 cm, targeting the root zone after removing surface plant debris. This depth was selected to capture nematodes occupying active root regions. The samples were stored in cooler boxes at 4 °C and transported to the Nematology Laboratory at the Aquaculture Research Unit, the University of Limpopo, for extraction and identification of nematodes.

2.2. Nematode Identification

Nematodes were extracted from 200 g of soil per site using a modified tray method [18,19]. They were initially counted with a stereomicroscope (Zeiss^® Discovery V8, Carl Zeiss AG, Oberkochen, Germany) and subsequently identified to the genus using a compound light microscope (VWR^® BL384, Milan, Italy) [20]. After counting, the nematodes were fixed in hot 4% formaldehyde solution and then transferred to anhydrous glycerine for species identification [21]. Genus-level identification was based on morphological keys and diagnostic features described by Andrássy [22], Siddiqi [23], Geraert [24], and Shokoohi and Abolafia [25].

2.3. Soil Property Analysis

Soil chemical properties, including ammonia, nitrate, and phosphate, were analyzed at the Aquaculture Research Unit Laboratory using a Hach^® spectrophotometer (Loveland, CO, USA), following the manufacturer’s protocols. Soil pH was measured using a Thermo Scientific^® Orion 3 Star pH Benchtop Meter (Waltham, MA, USA). Soil texture was determined according to the method described by van Capelle et al. [26].

2.4. Statistical Analysis

To identify dominant genera in kweek grass soils of Sovenga Hills, Limpopo Province, the relationship between the nematode population density (MPD) and the frequency of occurrence (FO) was used to calculate the prominence value (PV) for each genus, following Norton and Schmitt [27].

$$\mathrm{PV}=\mathrm{Population} \mathrm{density} \times \sqrt{\mathrm{frequency} \mathrm{of} \mathrm{occurrence}}$$

The frequency of occurrence (FO%) was calculated as follows:

FO% = (Number of samples containing a genus/Total number of samples) × 100

The relative abundance (RA) was determined as follows:

RA = Total number of individuals of a genus per gram of soil and root/Total number of samples (including those with zero counts) [28].

Nematode biodiversity was assessed using the Shannon index (H′) [29], calculated in the XLSTAT software v. 2025 [30]. The richness was calculated based on the species at each site [29]. Community structure analysis was further explored using NINJA (Nematode Indicator Joint Analysis; Wageningen University & Research (WUR), Wageningen, The Netherlands) [31].

The sigma maturity index (MI) or plant-parasitic index (PPI) was calculated through the NINJA online software v. 1.0 (Sieriebriennikov et al., 2014) [31] using the equation below:

MI or PPI = ∑ (v_i × f_i)/n

where v_i = the colonizer–persister (c-p) value assigned to the family, f_i = the frequency of family i in the sample, and n = the total number of individuals in a sample.

The nematode biomass was calculated using the equation below:

W = (L³/a²)/(1.6 × 10⁶)

where W is the fresh weight (µg) per individual, L is the nematode length (µm), and a is the ratio of body length/greatest body diameter (µm).

Soil properties were used as supplementary variables to identify relationships with the abundances of the main nematode genera. The scores’ values were determined for each variable based on each of the principal components, and the scores for the first two components were used to form a two-dimensional plot (PC1 and PC2) based on eigenvalues given with the software XLSTAT software v. 2025 [30].

2.5. Visual Analytics

The distribution of nematode genera across the sampling sites was visualized using the Gephi 0.10.1 software [32], a network-based tool increasingly used in ecological research to reveal trophic complexity and community robustness [33]. Each sampling location was first added as a fixed “site node” and spatially positioned. Nematode genera were then added as additional nodes, with their placement determined by the ForceAtlas2 layout algorithm, which groups nodes based on connection strength. Genera with stronger associations to a site appeared closer to it. The thickness and direction of the connecting lines (edges) represented the strength and directionality of these associations. The ForceAtlas2 layout also highlighted weak or isolated associations, helping to identify less-connected taxa. Additionally, shared nematode genera among sites were compared using Venny ver. 2.1 [34], a tool for visualizing overlap in community composition.

3. Results

3.1. Analysis of Nematode Communities

Twenty-three genera were identified (

Table 1. Mean abundance and trophic groups of nematodes found in soil associated with kweek grass (C. dactylon (L.) Pers. in Sovenga Hills, Limpopo Province, South Africa. M-sites: Soil from areas with golden mole activity (N. julianae). T-sites: Soil from areas with kweek grass soil growth, without mole activity. [C-p = colonizer–persister; P-p = plant-parasitic nematodes].

Genus	C-p Class	P-p Class	Feeding Type	Mass, μg	M	T
Rotylenchulus	0	3	Herbivorous sedentary parasites	1.77	4.1	0
Rotylenchus	0	3	Herbivorous semi-endoparasites	0.859	4.9	0
Tylenchorhynchus	0	3	Herbivorous ectoparasites	0.231	0	2.7
Tylenchus	0	2	Fungivorous	0.36	2.7	0
Aphelenchoides	2	0	Fungivorous	0.151	10	13.3
Aphelenchus	2	0	Fungivorous	0.218	14.9	10.8
Ditylenchus	2	0	Fungivorous	0.451	21	9.6
Acrobeles	2	0	Bacterivorous	0.64	13.1	6.8
Acrobeloides	2	0	Bacterivorous	1.263	8.8	3.6
Cervidellus	2	0	Bacterivorous	0.174	2.7	4
Eucephalobus	2	0	Bacterivorous	0.236	0	1.3
Geomonhystera	2	0	Bacterivorous	0.283	1.3	2.6
Panagrolaimus	1	0	Bacterivorous	0.66	2.6	12.8
Plectus	2	0	Bacterivorous	0.858	2.9	4.4
Prismatolaimus	3	0	Bacterivorous	0.357	3.7	6.2
Pseudacrobeles	2	0	Bacterivorous	0.217	2.9	3.5
Wilsonema	2	0	Bacterivorous	0.061	1.3	4.8
Zeldia	2	0	Bacterivorous	0.717	19.7	1.4
Tripylina	3	0	Predators	1.422	0	1.3
Mylonchulus	4	0	Predators	1.723	2.6	0
Discolaimus	5	0	Omnivorous	2.652	0	1.3
Aporcelaimellus	5	0	Omnivorous	9.54	9.5	6.2
Aporcella	5	0	Omnivorous	6.546	0	2.6

). Eighteen genera were collected in kweek grass with golden moles (M-sites; M1–M4), and nineteen genera were associated with normal kweek grass without golden moles (T-sites; T1–T4) (Table 1).

From the overall genera collected, Rotylenchulus, Mylonchulus, Tylenchus, and Rotylenchus were detected only in M-sites. In contrast, Tylenchorhynchus, Aporcelaimellus, Aporcella, Discolaimus, Eucephalobus, and Tripylina were only in T-sites (Figure 2; Table 1). The findings revealed that 60.9% of the nematode genera were shared between the soil samples of M- and T-sites. In contrast, 17.4% of the nematode genera were found exclusively in the soils of the M-sites, while 21.7% were solely present in the soils of the T-sites (Figure 2).

Three genera of plant-parasitic nematodes (PPNs), Tylenchorhynchus, Rotylenchulus, and Rotylenchus, were identified in the kweek grass soil samples (Table 1). In contrast, 20 genera of free-living nematodes representing various trophic groups (bacterivorous, fungivorous, omnivorous, and predators) were also recorded (Figure 3 and Figure 4; Table 1).

Frequency of occurrence (FO) data (Figure 5;

Table 2. Community analysis of nematode genera associated with kweek grass (C. dactylon (L.) Pers.) in Sovenga Hills, Limpopo Province, South Africa.

Species	Average	Positive Samples%	FO%	RA	PV
Aporcelaimellus	7.8125	63	50	0.78	1.38
Aphelenchus	12.825	88	87.5	1.28	1.71
Plectus	3.7	50	50	0.37	0.65
Acrobeles	9.95	88	87.5	1.00	1.33
Acrobeloides	6.1625	38	37.5	0.62	1.26
Geomonhystera	1.9375	38	37.5	0.19	0.40
Zeldia	10.525	50	50	1.05	1.86
Rotylenchulus	2.0625	25	25	0.21	0.52
Cervidellus	3.325	50	50	0.33	0.59
Tylenchorhynchus	1.325	25	25	0.13	0.33
Mylonchulus	1.3	25	25	0.13	0.33
Aporcella	1.275	25	25	0.13	0.32
Ditylenchus	15.3	75	75	1.53	2.21
Tylenchus	1.3375	13	12.5	0.13	0.47
Rotylenchus	2.425	25	25	0.24	0.61
Pseudacrobeles	3.2	38	37.5	0.32	0.65
Prismatolaimus	4.925	63	62.5	0.49	0.78
Aphelenchoides	11.6375	75	75	1.16	1.68
Discolaimus	0.6625	13	12.5	0.07	0.23
Wilsonema	3.075	38	37.5	0.31	0.63
Panagrolaimus	7.7	38	37.5	0.77	1.57
Eucephalobus	0.6375	13	12.5	0.06	0.23
Tripylina	0.65	13	12.5	0.07	0.23

) showed that Aphelenchus sp. and Acrobeles sp. were the most prevalent nematodes, each occurring in 87.5% of the samples. In contrast, Eucephalobus sp., Tripylina sp., Discolaimus sp., and Tylenchus sp. had the lowest FO, appearing in only 12.5% of samples. The highest prominence value (PV) was recorded for Ditylenchus sp. (PV = 2.2).

3.2. Indices of Nematode Communities

Community analysis using NINJA revealed significant differences (p < 0.001) in several indices, including the maturity index (MI), plant-parasitic index (PPI), enrichment index (EI), channel index (CI), and structural index (SI) (

Table 3. Nematode community indices, biomass, and metabolic footprints associated with kweek grass (C. dactylon (L.) Pers. in Sovenga Hills, Limpopo Province, South Africa. Values represent the mean ± standard deviation (SD) for each sampling site. M-sites: Soil from areas with golden mole (N. julianae) activity. T-sites: Soil from areas with normal kweek grass growth and no mole activity. [* = p < 0.001].

Index Name	M1	M2	M3	M4	T1	T2	T3	T4
Maturity index *	2.2 ± 0.04	2.2 ± 0.06	2.6 ± 0.07	2.2 ± 0.08	2.4 ± 0.04	2.9 ± 0.05	1.9 ± 0.04	2.2 ± 0.02
Maturity index 2–5 *	2.3 ± 0.04	2.2 ± 0.06	2.6 ± 0.07	2.2 ± 0.08	2.4 ± 0.04	3.3 ± 0.07	2.2 ± 0.05	2.2 ± 0.02
Sigma maturity index *	2.3 ± 0.04	2.3 ± 0.06	2.7 ± 0.06	2.2 ± 0.07	2.4 ± 0.04	2.9 ± 0.05	1.9 ± 0.04	2.2 ± 0.02
Plant-parasitic index *	3.0 ± 0.0	3.0 ± 0.0	3.0 ± 0.0	2.0 ± 0.0	3.0 ± 0.0	3.0 ± 0.0	NA	NA
Shannon index (H′) *	2.4 ± 0.1	2.1 ± 0.2	2.2 ± 0.1	2.0 ± 0.1	2.2 ± 0.2	1.9 ± 0.1	2.1 ± 0.2	1.7 ± 0.2
Channel index *	61.2 ± 2.58	100 ± 0.0	100 ± 0.0	100 ± 0.0	100 ± 0.0	11.8 ± 3.63	33.2 ± 1.93	100 ± 0.0
Enrichment index *	44.7 ± 1.1	26.1 ± 2.0	38.5 ± 1.59	31.2 ± 1.1	24.5 ± 1.83	61.2 ± 4.51	67.3 ± 1.27	34.1 ± 2.18
Structure index *	43 ± 3.8	34.6 ± 5.89	63.1 ± 3.38	28.6 ± 9.6	46.5 ± 3.56	86.0 ± 1.36	26.6 ± 6.54	30.1 ± 3.59
Total biomass, mg *	0.2 ± 0.02	0.2 ± 0.02	0.2 ± 0.02	0.1 ± 0.02	0.1 ± 0.01	0.3 ± 0.01	0.1 ± 0.01	0.1 ± 0.01
Abundance	1783	1373	1091	892	762	849	1686	664
Richness	14	9	10	9	10	7	10	6
Evenness	0.81	0.88	0.88	0.83	0.90	0.85	0.74	0.93
Herbivorous, % of total	4.5	5.8	13.2	0	5.2	5.4	0	0
Fungivorous, % of total	55.5	36.6	41.3	36.7	17.0	5.6	58.7	22.0
Bacterivorous, % of total	75.7	78.2	35	34.6	40.9	49.3	78	22.6
Predators, % of total	1.7	0	2.3	0	0	0	0	0
Omnivorous, % of total	11.4	10.4	11.1	5.0	7.8	24.6	1.8	0

). Site-level differences were also observed in the relative abundance of nematode trophic groups: herbivorous, fungivorous, bacterivorous, and omnivorous. Both the sigma maturity index and plant-parasitic index differed significantly between M-sites and T-sites. The highest MI was recorded at T2 (2.9 ± 0.05), while the lowest was observed at T3 (1.9 ± 0.04) (Table 3). The PPI was the highest (ranging from 2 to 3) in M-sites but could not be calculated for T3 and T4 due to the absence of plant-parasitic nematodes. The Shannon index (H′) and maturity index were the most effective indicators of ecosystem disturbance (Table 3). The lowest H′ was found at T4 (1.7 ± 0.2), compared with a higher value at M1 (2.4 ± 0.1). Again, T2 showed the highest MI (2.9 ± 0.05) among all sites (Table 3). For total nematode biomass (p < 0.001), the highest biomass was recorded at T2 (0.3 ± 0.01 mg). Furthermore, the basal index (BI), enrichment index (EI), and structure index (SI) effectively distinguished between the M-sites and T-sites.

Food web analysis (Figure 6) was conducted for all nematode taxa identified in the soil samples associated with kweek grass. This analysis assessed the soil health status of each site based on the structure and composition of the nematode communities. The results of the food web are displayed across four quadrants, labeled A to D. Soil samples from T3 were placed in quadrant A, indicating high nitrogen levels, a low carbon-to-nitrogen (C–N) ratio, and significant disturbance. In T3, the colonizer–persister nematode (cp-1; Table 1), Panagrolaimus, was observed more frequently. This bacterivorous nematode has a short lifespan and is sensitive to soil disturbances. The soil samples of T2 were placed in quadrat B, indicating high nitrogen, a low C–N ratio, and less disturbance than T3. In T2, the colonizer–persister nematode (cp-5; Table 1), Aporcelaimellus, was observed more than in other feeding groups. This omnivorous nematode has a long lifespan and is highly sensitive to soil disturbances. The majority of the samples, M1, M2, M4, T1, and T4, were placed in quadrat C, which they classified as degraded, with conductivity and a high C–N ratio, representing 63% of all soil samples. In the mentioned soil sites, cp-2 (Table 1) was more frequent and better adapted to soil stress due to the decrease in food resources caused by competition from fungivorous and bacterivorous nematodes. The soil samples from M3, located in quadrat D, indicated fertile and well-structured soil. In M3, the colonizer–persister nematode (cp-4), specifically Mylonchulus, was observed more frequently than other feeding groups. This predator nematode has a long lifespan and thrives in soil with high organic matter decomposition. It is also highly sensitive to soil disturbances.

3.3. Principal Component Analysis (PCA)

Principal component analysis (PCA) was conducted to explore the relationship between nematode genera and soil physicochemical properties across all sampling sites (M1–M4 and T1–T4) (Figure 7). The PCA results revealed distinct associations between specific nematode genera and soil parameters: the ordination plot presents the spatial distribution of sampling sites and nematode taxa in relation to environmental vectors (arrows). Each arrow represents a soil parameter gradient, with the direction and length indicating the strength and direction of its influence. Ditylenchus is closely aligned with the clay content vector, suggesting a positive correlation between this genus and clay-rich microhabitats. Clay likely supports moisture retention, favoring the survival of this genus. Cervidellus appears to be associated with soil pH. Aporcelaimellus, Prismatolaimus, and Wilsonema clustered near the phosphate vector, suggesting their abundance increases with higher phosphorus availability. Tylenchorhynchus shows an affinity with ammonia, potentially reflecting the influence of nitrogen forms on plant-parasitic nematode dynamics. Pseudacrobeles is positioned along the silt gradient, suggesting a niche preference for fine-textured soils.

The mole-active sites (M1–M4) and non-mole sites (T1–T4) are spatially separated in the ordination space, indicating distinct nematode community structures shaped by golden mole activity. The M-sites are generally associated with a higher abundance of certain genera and correlate with particular soil parameters, suggesting that bioturbation by moles indirectly modifies soil microenvironments, thereby altering nematode assemblages (Figure 7).

3.4. Data Visualization

A network projection was generated using Gephi to visualize the distribution of nematode genera across the soil samples (Figure 8). The co-occurrence network revealed 31 nodes (various trophic groups of nematodes and 8 sites of this study) and 77 edges (connections between the nodes). The average degree of the network was 4.9, indicating that each node was, on average, connected to approximately 5 other nodes. Moreover, the results showed M1, M2, M4, and T3 with stronger edge connectivity (Figure 8). The clustering coefficient was calculated as 0.0, indicating no connection between the nodes. In addition, the clustering coefficient indicated that there was no internal clustering of genera, suggesting clear habitat preferences and a low level of taxonomic redundancy.

The co-occurrence network showed bacterivorous (38.7%) was the most dominant trophic group connected to the sites of this study (M1–M4; T1–T4), followed by fungivorous, with 12.9%. The analysis revealed several site-specific associations: T3 was strongly associated with Ditylenchus. M3 was linked to both Aphelenchoides and Tylenchus. M1 showed a notable association with Zeldia, while M2 was strongly connected to Aphelenchus. Aporcelaimellus was unique to M4, Aporcella was associated with T3, and Discolaimus appeared only in T1 (Figure 8). Among herbivorous nematodes, Rotylenchus was present in M1 and was also shared between M1 and M3. Tylenchorhynchus was shared between T1 and T2. The only predatory genus identified, Mylonchulus, was found at M1 and M3. The network analysis complements Table 3 by shifting the focus from site-level averages to community-wide structure and interactions, which is essential for understanding the ecological implications of mole-driven soil disturbance on nematode biodiversity.

The PCA plot (Figure 7) is a constrained ordination based on the abundance of data and environmental gradients. It reflects continuous relationships among all taxa and environmental variables, even when associations are weak. In contrast, the network graph (Figure 8) is a qualitative co-occurrence diagram that emphasizes strong and exclusive site-level associations. To prevent overcrowding, only nematode genera with stronger frequency or prominence at specific sites were visualized.

4. Discussion

The grassland biome of South Africa, known locally as grassveld, occupies the elevated interior plateau and is defined by rolling hills and extensive plains [35]. In Limpopo Province, this biome transitions into the savanna, creating a mosaic of open grasslands and scattered trees [36]. Despite its ecological importance, this biome remains underexplored regarding its belowground biodiversity, particularly nematodes [37]. South Africa’s grasslands support half the country’s endemic mammal species and deliver critical ecosystem services, including nutrient cycling, soil stabilization, and forage provision [38,39].

Our study of C. dactylon (kweek grass) in Sovenga Hills revealed diverse nematode communities consisting of both plant-parasitic and free-living taxa. These organisms serve as bioindicators of soil health and contribute to key ecological processes [40,41,42]. Plant-parasitic nematodes such as Rotylenchus, Rotylenchulus, and Tylenchorhynchus were consistent with previous reports in South African pastures [43]. Tylenchorhynchus is known to cause significant root and shoot damage [44], which may impact the productivity of native grasses.

Golden moles, fossorial mammals that prefer loose, sandy soils, have been shown to modify subsurface environments [45,46]. Their burrowing activities improve aeration and moisture retention, creating conditions favorable for the movement and survival of plant-parasitic nematodes. This may explain the elevated abundance of parasitic species in mole-active soils. Additionally, reduced populations of free-living nematodes in these soils—likely due to lower organic matter—may reduce competition, facilitating parasitic nematode dominance.

In contrast, mole-free sites support greater numbers of fungivorous, bacterivorous, and omnivorous nematodes. These groups thrive under minimal disturbance, benefiting from more stable microbial food sources [47,48].

Environmental variables such as pH were positively associated with genera like Cervidellus and Eucephalobus, supporting previous observations of pH sensitivity in bacterivorous nematodes [49,50]. Seasonal rainfall in Sovenga Hills likely contributes to the natural input of organic matter, enhancing microbial and nematode activity in summer months.

Soil texture significantly influences nematode community composition. Sandy soils favor both nematode activity and golden mole tunneling [51,52]. Our results showed that mole-active soils had lower sand and higher silt contents, conditions less favorable to free-living nematodes but more suitable for certain plant-parasitic taxa. High silt content has been negatively correlated with nematode abundance in other ecosystems [6,53].

The low clay content (2–5%) may have limited overall nematode populations, though we observed a positive correlation between clay and Ditylenchus abundance, consistent with other studies [6,52]. Clay-rich sub-surfaces help retain moisture and may create microhabitats favorable for this genus.

Nematode indices provided further insights into soil condition. The Sigma maturity index (SMI) is a reliable indicator of ecological disturbance in non-agricultural soils [54]. Our findings showed that all sites except T3 exceeded the SMI threshold of 2, suggesting generally mature and undisturbed soil food webs. The low SMI at T3 points to disturbance or stress, despite the absence of mole activity.

The channel index (CI), which reflects decomposition pathways, was below 50 at T2 and T3, indicating fungal-dominated breakdown and slower organic matter turnover. In contrast, higher CI values at other sites suggested bacterial dominance and more rapid decomposition [47].

Nematode diversity, as measured by the Shannon Index, was highest at most sites except T2 and T4, suggesting that golden mole activity may contribute to higher nematode diversity, possibly through increased soil heterogeneity [6,55].

The enrichment index (EI) and structure index (SI) provide insights into nutrient availability and food web complexity, respectively. Sites with high C–N ratios were dominated by fungivorous nematodes, suggesting slow decomposition and nutrient depletion. Over time, microbial buildup can shift communities toward bacterivorous and fungivorous dominance, accelerating nutrient release [56,57]. The analysis of the grassland food web indicated that the majority of soil samples were categorized in quadrat C, which is characterized by relatively structured soils with moderate enrichment, characterized by the cp-2 trophic group of nematodes [47]. Furthermore, some grassland samples from quadrats A and C displayed a higher presence of cp-1 and cp-2 nematodes, suggesting that these areas represent managed land with soils ranging from disturbed to structured [47]. Consistent with these findings, the current study also found that most soil samples were classified in quadrat C.

In addition to golden mole activity, other environmental and biological factors may also influence the structure of the nematode community. For example, vegetation heterogeneity in the Sovenga Hills, including variability in root exudates and canopy cover, could affect organic inputs and microclimate, indirectly shaping nematode assemblages [45,46,55]. Similarly, nematode-antagonistic microflora, such as predatory fungi (Arthrobotrys spp.) and parasitic bacteria (Pasteuria penetrans), can naturally regulate nematode populations, particularly in undisturbed soils with high microbial activity [58,59]. Moreover, soil compaction, which is typically reduced in mole-active sites due to burrowing, can influence aeration and moisture retention, thereby altering nematode mobility and survival, as indicated by earworm activity [60]. These confounding factors underscore the complexity of belowground ecological interactions and suggest that mole-related effects should be interpreted within the broader context of interacting biotic and abiotic soil parameters.

Network analysis provided an additional layer of ecological interpretation by visually representing the co-occurrence patterns between nematode genera and sampling sites. While diversity indices and ordination methods (e.g., PCA) quantify compositional and environmental variation, the network diagram (Figure 8) allowed us to evaluate the trophic structure, site-specific associations, and connectivity within the nematode communities under different disturbance regimes.

The analysis revealed that bacterivorous nematodes were the most dominant and widely distributed trophic group, comprising 38.7% of all genera in the network, and were particularly enriched in mole-affected sites (M1–M4). This aligns with previous reports that bacterivorous are often the first colonizers in disturbed or enriched environments due to their fast reproductive strategies and strong association with microbial activity [1]. Mole activity likely enhances these conditions by loosening the soil and facilitating microbial proliferation through increased aeration and organic matter turnover [7,8].

Distinct associations were observed between specific nematode taxa and sampling sites. For instance, Zeldia and Rotylenchus were tightly linked to M1, while Aphelenchus dominated in T3, and Aporcelaimellus in M2. Aphelenchus is a genus known for antifungal potential under high-density conditions [61]. In non-mole sites, Ditylenchus and Aporcella were prominent in T3, indicating that fungivorous and omnivorous taxa may act as indicators of habitat filtering in less-disturbed soils. These relationships may reflect microhabitat specialization or the influence of localized soil physicochemical gradients, such as pH, nutrient availability, or texture [2,6].

The low clustering coefficient (0.0) of the network suggests a lack of tight co-occurrence clusters, indicating that most genera were functionally or spatially separated. This is consistent with high niche differentiation and selective pressures acting on nematode communities, possibly driven by the presence or absence of golden mole bioturbation. Sites such as M1, M2, and M4 demonstrated higher connectivity (average degree ~5), suggesting more diverse and ecologically complex soil food webs compared with the more fragmented networks in T-sites.

Moreover, network analysis enables the identification of indicator species that may not be statistically dominant but are structurally significant in the ecological web (e.g., Mylonchulus, a predatory nematode at M1). These taxa are often underrepresented in abundance tables but play pivotal roles in regulating trophic interactions and maintaining food web stability [33,40].

Overall, network analysis served as a complementary tool to PCA and diversity metrics by enabling a visual synthesis of community structure, identifying unique taxa–site interactions, and revealing the functional organization of nematode communities in disturbed vs. undisturbed soils. This integrative approach underscores the ecological impact of fossorial mammals like golden moles in shaping belowground biodiversity and supports the utility of nematode networks as indicators of soil ecosystem health.

5. Conclusions

This study provides the first detailed account of nematode community structure associated with kweek grass in the Sovenga Hills region of Limpopo Province, South Africa, and highlights the ecological influence of golden mole activity on soil biotic composition. Sites with mole activity supported a distinct nematode assemblage, characterized by increased abundance of plant-parasitic genera and reduced representation of free-living trophic groups. These shifts may be driven by changes in soil structure, moisture, and organic matter availability resulting from mole burrowing.

Kweek grass ecosystems associated with golden mole activity exhibited greater nematode diversity, including an increased presence of plant-parasitic species. These findings highlight the need for targeted management practices to minimize potential damage to grasslands. Strategies such as regular nematode monitoring, soil health improvement, and rotational planting may help maintain ecological balance and promote sustainable grass growth.

Nematode-based ecological indices, including the maturity index, plant-parasitic index, and channel index, proved valuable in distinguishing disturbed from stable soil food webs. Network and redundancy analyses further demonstrated clear associations between specific nematode taxa and soil physicochemical properties, particularly clay, phosphate, and pH. The findings underscore the utility of nematodes as bioindicators for assessing belowground biodiversity and ecosystem function in grassland environments. Moreover, they reveal a previously underappreciated interaction between fossorial mammals and soil microfauna.

To strengthen future research, broader sampling across multiple sites in Limpopo Province is recommended. This will enable a more comprehensive understanding of the ecological role of golden moles in shaping nematode communities and their effects on grassland health. A systematic investigation across varying habitats and soil conditions will provide deeper insights into the complex interactions between burrowing mammals and soil biodiversity.