The study was conducted in the northern sector of Virunga National Park (PNVi), located in North Kivu Province, in the eastern Democratic Republic of the Congo (Figure 1). Covering an area of approximately 7900 km², this park stretches from the Virunga volcanic massifs to the Ruwenzori Mountains, encompassing the Rwindi plains and part of Lake Edward, all of which are recognized for their high biological diversity [4]-[6].

Listed as a UNESCO World Heritage Site since 1979, the PNVi presents a unique ecological mosaic that combines three major ecoregions: The Afro-alpine zone, the Afro-montane forest, and the forest-savanna mosaic, to which are added aquatic systems of high ecological and socio-economic value [1]-[3]. Geographically, it is located between 01˚35'S latitude and 29˚01'E-30˚01'E longitude, giving it a strategic position in the Albertine Rift, one of the world’s major biodiversity hotspots [4]-[6].

The data used in this study were derived from both secondary and primary

Figure 1. Location of the study area.

sources. Secondary data were obtained through an extensive review of scientific publications, institutional reports, and technical documents related to Muridae and anthropogenic pressures within Virunga National Park.

Primary data were collected between April and June 2025, following a preliminary exploratory survey conducted in March 2025 after obtaining the necessary research permits. This reconnaissance phase enabled the georeferencing of study sites, the characterization of anthropogenic disturbance gradients, and the adaptation of the sampling protocol to local environmental conditions.

Field sampling was carried out along eleven 2-km recce transects (Figure 2), randomly distributed within a 2 × 2 km grid system and representing three levels of anthropogenic disturbance (low, high, and very high). Each transect was surveyed for approximately two hours by a team of four observers. Geographic coordinates were recorded using a Garmin eTrex 32x GPS receiver.

Evidence of Muridae occurrence was systematically recorded through both direct and indirect observations. Direct observations consisted of visual sightings of individuals, whereas indirect evidence included footprints, feces, burrows, hair, remains, and carcasses. Species identification was based on the field identification guides developed by [12]. In addition, signs of anthropogenic activities, including traps, fire traces, hunting camps, and cleared areas, were systematically documented along each transect. The collected data were organized and managed using Microsoft Excel 2021 and subsequently analyzed with R version 4.1.0 for statistical analyses, QGIS version 3.44.0 and ArcGIS version 10.8 for spatial analyses and cartographic production, and Google Earth Pro for geographic data visualization and validation.

In the northern sector of the PNVi, sampling was carried out along reconnaissance transects established in the three anthropogenic disturbance gradients. The

Figure 2. Location of Recce transects in the study area.

operation involved a team of four observers with complementary roles, ensuring the standardization of the surveys. A total of eleven 2 km recce transects were covered at a constant speed (≈two hours per transect) to ensure uniform coverage of the different habitats. The geographical coordinates of the transects and observations were recorded using a Garmin eTrex 32x GPS and systematically recorded on standardized field sheets.

Evidence of Muridae presence was collected in two categories: 1) direct evidence (visual observations of individuals) and 2) indirect evidence (feces, footprints, burrows, hair, food remains, carcasses). Identification was carried out using field guides developed by Kingdon [12]. Simultaneously, signs of human activity (traps, bush fires, temporary huts, and agricultural clearings) were recorded in accordance with the ICCN monitoring protocol [7].

Data from Muridae presence indices were used to estimate their abundance, specific diversity, and spatial distribution according to disturbance gradients. Information on anthropogenic pressures was used to develop a map of human activities, which was integrated and superimposed on the official PNVi Map using Google Earth and ArcGIS, in order to analyze the spatial correspondence between disturbances and the dynamics of Muridae communities.

2.3.1. Specific Abundance and Presence Indices of Muridae

The frequency of occurrence was estimated using the encounter rate (ER), defined by the following formula:

ER = (n/L) × 100,

where n is the total number of indices observed and L is the total length of the transects (in km).

2.3.2. Mapping the Spatial Distribution of Muridae

The GPS coordinates of Muridae presence indices and signs of human activity were integrated into QGIS 3.44.0, ArcGIS 10.8, and Google Earth (2025). These data were then overlaid on the official map of Virunga National Park to generate accurate georeferenced maps illustrating the location of recce transects, the spatial distribution of Muridae species, and the distribution of anthropogenic disturbance gradients in the northern sector of the park.

2.3.3. Diversity Indices

Specific diversity, following the approaches proposed by Jost [13] and Piélou [14], was measured using three ecological indices:

Shannon-Wiener index (H′): H′ = −∑ (pi × log₂ pi), where pi = ni/N;Simpson index (IS): D = ∑ [ni (ni − 1)]/[N (N − 1)]; IS = 1 − D;Piélou’s evenness index (J′): J′ = H′/log S.

2.3.4. Correspondence Factor Analysis (CFA)

A correspondence factor analysis (CFA) was performed to examine the relationships between the relative abundance of Muridae and anthropogenic disturbance gradients. This multivariate method allows for the visualization of associations and dissociations between species and habitats, while quantifying the contribution of each species and each habitat to the total inertia of the data set [15].

2.3.5. Statistical Tests

Two statistical tests were used to examine the relationships between anthropogenic disturbance gradients and Muridae presence indices. Kendall’s correlation test, which is non-parametric and robust to extreme values, was chosen because of the non-normality of the variable distributions. At the same time, the Chi-square (χ²) test was used to assess the association between the distribution of Muridae presence indices and the intensity of disturbances, as defined according to anthropogenic disturbance gradients.

3.1.1. Encounter Rate of Muridae Based on Presence Indices

In the northern sector of Virunga National Park, the presence of Muridae was confirmed using ten types of indirect indices (Figure 3). Among these, the most frequently observed were droppings (21.7%) and footprints (17.4%), followed by carcasses (13.8%), gnaw marks (12.2%), burrows (11.7%), and food remains and hair (8.3% each). These results highlight the diversity of evidence that can be used to assess the presence and activity of Muridae in different habitats and disturbance gradients.

Figure 3. Proportion of indirect signs of rodent presence in the PNVi (MG = gnaw marks; RA = food remains; VO = vocalization).

3.1.2. Relative Abundance of Muridae in the Study Area

Table 1 shows the encounter rates of Muridae along 11 survey transects distributed according to the three anthropogenic disturbance gradients, based on 253 direct indices recorded. The most frequently observed species are Arvicanthis niloticus (22.1%; 2.54 signs/km) and Lemniscomys striatus (17.8%; 2.04 signs/km), indicating high abundance or activity in the study area. Conversely, Dasymys montanus (5.9%; 0.68 signs/km), Oenomys hypoxanthus (5.1%; 0.59 signs/km) and Praomys jacksoni (4.7%; 0.54 signs/km) had lower encounter rates, suggesting either a low presence or a reduced probability of detection in these habitats.

Table 1. Encounter rate (number/km) and relative abundance (%) of Muridae.

3.1.3. Grouping of Muridae Species According to Relative Abundance

In Virunga National Park (PNVi), correspondence analysis (CA) based on the relative abundance of Muridae species revealed three groups structured along gradients of anthropogenic disturbance (Figure 4). The first group, associated with habitats subject to moderate disturbance, includes five species typical of relatively preserved environments: Hybomys lunaris, Dasymysmontanus, Praomys jacksoni, Oenomys hypoxanthus, and Grammomys kuru. The second group, present in all habitats but strongly linked to areas affected by marked disturbance, includes generalist and tolerant species such as Lemniscomys striatus, Otomys tropicalis, and Mastomys natalensis. Finally, the third group, consisting exclusively of Arvicanthis niloticus, is concentrated in environments characterized by very high anthropization, particularly fallow land, wasteland, and peri-urban areas.

This structuring along the disturbance gradient illustrates a clear differentiation between sensitive species, dependent on relatively intact habitats, and opportunistic species capable of colonizing environments that have been heavily transformed by humans. It highlights the role of Muridae as relevant bioindicators of the impact of anthropogenic pressures on the dynamics and resilience of ecosystems in the PNVi.

Figure 4. Biplot ordination by habitat gradients and anthropogenic disturbance in the study area.

3.1.4. Diversity of Muridae in the Study Area

According to Table 2, nine species of Muridae have been recorded in the northern sector of the PNVi, two of which are threatened according to the IUCN: Dasymys montanus (Endangered) and Hybomys lunaris (Vulnerable). The other species are classified as “Least Concern.” Local knowledge indicates a high perceived abundance for Arvicanthis niloticus, Lemniscomysstriatus, and Otomys tropicalis, while D. montanus and H. lunaris are considered less abundant, corroborating the low encounter rates observed in the field.

3.1.5. Specific Diversity Indices

The accumulation curves (Figure 5) and diversity indices (Table 3) indicate that the specific diversity of Muridae is significantly higher in the PNVi, particularly

Table 2. Specific richness and conservation status of Muridae.

(−): Very rare, (+++): Very abundant, (++): Abundant, (+): Rare.

Table 3. Shannon, Simpson, and Piélou diversity of Muridae in the study area.

Figure 5. Accumulation curve for Muridae diversity (1 = Shannon diversity; 2 = Simpson diversity).

in lightly disturbed habitats (H′ = 1.936; D = 0.825; J′ = 0.881), compared to heavily disturbed habitats (H′ = 1.182; D = 0.581) and very heavily disturbed habitats (H′ = 1.014; D = 0.545). These results highlight the fundamental role of the PNVi in maintaining and conserving small mammal biodiversity. Across all habitats studied, the average diversity of Muridae was H′ = 1.457 and D = 0.684, reflecting a significant reduction in diversity in environments heavily transformed by human activities.

3.1.6. Similarity of Specific Composition

The specific composition of Muridae is strictly identical between lightly disturbed habitats and heavily disturbed habitats (r = 1.00), reflecting relatively similar ecological conditions in these two types of habitat. In contrast, highly disturbed habitats have an impoverished composition (r = 0.714), marked by the absence of four species sensitive to disturbance: Dasymys montanus, Grammomys kuru, Hybomys lunaris, and Oenomyshypoxanthus (Figure 6). These results highlight the negative impact of intense anthropogenic disturbances on the species diversity of Muridae and the potential loss of certain species that are indicators of habitat quality.

Figure 6. Clustering dendrogram showing the similarity of the specific composition of Muridae in the study area (slightly disturbed habitats, heavily disturbed habitats, and very heavily disturbed habitats).

3.1.7. Spatial Distribution of Muridae in the Study Area

Mapping of Muridae presence indices (Figure 7) highlights a heterogeneous distribution of Muridae in the northern sector of the PNVi, with relative abundance varying across transects: high (red dots) in heavily disturbed habitats, moderate (black dots) and low (blue dots) in lightly and heavily disturbed habitats. This spatial distribution illustrates the combined influence of habitat characteristics and local anthropogenic pressures on the dynamics and density of Muridae populations.

3.1.8. Frequency of Anthropogenic Activities in the Study Area

Table 4 shows that hunting is the main pressure on Muridae populations (6.81 signs/km), followed by inappropriate agricultural practices (2.27 signs/km) and deforestation (1.5 signs/km). Although less frequent, the settlement of displaced populations (1.13 signs/km), as well as other anthropogenic activities such as fishing, charcoal production, grazing, and urbanization, also contribute to habitat degradation and disturbance of Muridae communities, increasing the vulnerability of certain species sensitive to human pressures.

Figure 7. Location of evidence of rodent presence in the study area.

Table 4. Frequency of anthropogenic activities in the study area.

3.1.9. Spatial Distribution of Anthropogenic Activities in the Study Area

Figure 8 illustrates a contrasting spatial distribution of anthropogenic activities in the northern sector of the PNVi, with high intensity (blue dots) localized in highly disturbed habitats, moderate intensity (black dots) and low intensity (red dots) in slightly and strongly disturbed habitats, depending on the areas studied. This mapping is based on indices recorded in the field (tracks, human installations, signs of disturbance) and allows the spatial impact of anthropogenic pressures on Muridae habitats to be visualized.

Figure 8. Location of signs of human activity in the study area.

3.2.1. Influence of Anthropogenic Activities on Signs of Rodent Presence in the Study Area

Kendall’s correlation test reveals a significant negative correlation between the intensity of anthropogenic activities and indirect signs of Muridae presence (τ = −0.54; p = 0.04), indicating an adverse impact of human disturbances on these populations. A moderate but non-significant negative association was also observed with direct indices (τ = −0.44; p = 0.12) (Figure 9), suggesting that direct detection of individuals is less sensitive to local variations in anthropogenic pressures.

Figure 9. Relationship between the presence indices (indirect and direct) of Muridae and the presence of entropic indices.

3.2.2. Spatial Variation in Muridae Presence Indices in Relation to Anthropogenic Activity Gradients in the Study Area

The analysis revealed a significant variation in Muridae occurrence indices across different types of anthropogenic disturbances (χ² = 439.8, df = 8, p < 0.001). The proportion of occurrence indices was highest in urbanized areas (21%) and lowest in hunting zones (2.4%) (Figure 10). These findings indicate that the distribution

Figure 10.Proportion of Muridae presence indices according to type of anthropogenic disturbance.

and detectability of Muridae are significantly influenced by the nature and intensity of human-induced habitat modifications.

The high diversity of indirect signs recorded in the northern sector of Virunga National Park indicates substantial Muridae activity within this protected area. The predominance of feces (21.7%) and footprints (17.4%) suggests not only a regular presence of these small mammals but also their active use of the various microhabitats available. Indirect signs constitute particularly effective tools for assessing the presence and activity of small mammals in forest habitats, where direct observations are often difficult. Similar findings have highlighted the importance of indirect evidence in evaluating species richness and ecological integrity in forest ecosystems [16].

The high relative abundances of Arvicanthis niloticus (22.1%) and Lemniscomys striatus (17.8%) reflect the ability of these species to exploit open or disturbed habitats. These rodents are known for their ecological plasticity, opportunistic feeding behavior, and high reproductive capacity, characteristics that enable them to rapidly colonize environments modified by human activities. The predominance of these generalist species in disturbed habitats is consistent with observations from Nairobi National Park, where habitat degradation favored disturbance-tolerant species at the expense of specialized forest taxa [11].

The grouping of species into three functional assemblages, as revealed by Correspondence Factor Analysis, clearly reflects the influence of the anthropogenic disturbance gradient on Muridae communities. The forest-associated group, including Hybomys lunaris, Dasymys montanus, Praomys jacksoni, Oenomys hypoxanthus, and Grammomys kuru, remained strongly linked to relatively intact forest habitats. These species generally exhibit strict ecological requirements related to vegetation cover, humidity, and habitat structural complexity. In contrast, the anthropophilic group, dominated by Arvicanthis niloticus, was primarily associated with open and degraded environments, whereas ubiquitous species occupied a broader range of habitats owing to their high adaptive capacity. Similar studies have emphasized the role of closed-canopy forests as refugia for specialized species [17].

Furthermore, the results suggest that moderate levels of disturbance may temporarily increase environmental heterogeneity and promote the coexistence of species with different ecological requirements. According to the Intermediate Disturbance Hypothesis, the creation of new microhabitats and increased structural diversity may locally enhance biological communities [18]. However, when disturbances become more intense or frequent, they generally lead to habitat simplification, reduced resource availability, and the gradual disappearance of highly specialized species. Similar patterns have been reported in heavily disturbed habitats, where a decline in small mammal diversity was observed [19].

The occurrence of Dasymys montanus, classified as Endangered (EN), and Hybomys lunaris, classified as Vulnerable (VU), highlights the conservation importance of the study area. Combined with the relatively high Shannon diversity index (H′ = 1.936), this finding underscores the role of Virunga National Park as a refuge for species sensitive to anthropogenic disturbances. The high similarity observed between the park and fallow lands (r = 1.00) may indicate ecological connectivity that facilitates movement and recolonization of surrounding habitats. Conversely, the lower similarity observed in urbanized areas (r = 0.714) suggests community simplification resulting from urbanization, characterized by the dominance of a limited number of generalist species. Comparable trends have demonstrated that increasing human activities progressively promote the biotic homogenization of small mammal communities [16][18].

Overall, these findings highlight the critical role of habitat quality and anthropogenic disturbance intensity in shaping Muridae communities. They confirm the importance of the forest habitats of Virunga National Park for maintaining specialized and threatened species while emphasizing the need to reduce anthropogenic pressures to preserve the diversity and ecological functionality of small mammal communities.

The significant negative correlation observed between anthropogenic activities and indirect Muridae occurrence indices (τ = −0.54, p = 0.04) highlights the overall detrimental influence of human disturbances on these communities. This relationship suggests that as the intensity of anthropogenic activities increases, Muridae occurrence tends to decline, particularly among the most specialized and disturbance-sensitive species. The similar trend observed for direct occurrence indices (τ = −0.44) further supports this interpretation.

However, the significant variation in occurrence indices among disturbance types (χ² = 439.8, df = 8, p < 0.001) indicates that Muridae species do not respond uniformly to different forms of anthropogenic disturbance. Urbanized environments exhibited the highest proportion of occurrence indices (21%), suggesting that certain opportunistic species, particularly Arvicanthis niloticus and Lemniscomys striatus, benefit from food resources, artificial shelters, and ecological conditions generated by human activities. Similar observations have been reported in Gabon and Kenya, where generalist species dominate highly modified habitats [11][16].

In contrast, hunting areas exhibited the lowest proportion of occurrence indices (2.4%). This pattern may result from disturbances associated with human presence, hunting dogs, trapping activities, and progressive habitat degradation. Although Muridae are not always the primary target species, traditional hunting practices exert non-selective pressure on wildlife and alter the ecological conditions required for the persistence of numerous species. Comparable findings have been reported in West and Central Africa [20]-[22].

The variation in Muridae distribution across habitats further confirms the importance of habitat structural characteristics. Forest habitats, characterized by closed canopies, dense understories, and extensive vegetation cover, provide greater availability of shelter, breeding sites, and food resources, thereby supporting more diverse communities. In contrast, fallow lands and urbanized areas, which are characterized by canopy opening and habitat simplification, are primarily inhabited by disturbance-tolerant species.

These results are consistent with findings from several African ecosystems, including the savannas of Gabon [16], the Mabira Forest Reserve in Uganda [9], the Mukwe villages in Namibia [23], and the Mau Forest Complex in Kenya [10]. In these ecosystems, anthropogenic disturbances generally favor generalist species at the expense of forest specialists. Similarly, declines in specialized forest species within agricultural landscapes of southern Africa have been documented [24].

Taken together, these findings indicate that anthropogenic activities promote the progressive homogenization of Muridae communities, characterized by the increasing dominance of opportunistic species capable of adapting to disturbed habitats. The conservation of continuous forest blocks, the reduction of habitat fragmentation, and the sustainable management of human activities therefore appear essential for maintaining species diversity and the ecological functioning of small mammal communities [25].

It should be noted, however, that species identification in this study relied primarily on indirect evidence (footprints, feces, burrows, hair, and other traces), the interpretation of which can be challenging for cryptic Muridae species exhibiting similar morphological or ecological characteristics. Consequently, this approach may have led to occasional species misidentifications and influenced relative abundance estimates for certain taxa. Nevertheless, the combined use of multiple evidence types and specialized identification guides helped minimize this potential source of bias as much as possible.