The study area comprised nine sampling sites located within Grand Libreville, including the municipalities of Libreville, Owendo, Akanda, and Ntoum (Figure 1). Four sites were located in Libreville (Ondongo, Ambowe, Lowe (Mindoube), and Ozoughe). Additional sites included Alenakiri and Rouger (in Owendo), Okala (Avolé-Zamé village) and Angondje (in Akanda), and Bizango (in Ntoum). At each site, five sampling points were selected. This design resulted in a total of 45 water samples collected for analysis.

Figure 1. Location of the study area (adapted from [17]).

Among all these sites, those of Lowe, Ambowe, and Ozoughe were particularly affected by anthropogenic activities. The Lowe site is characterized by unregulated stilt constructions within the mangrove ecosystem, established by foreign communities engaged in fishing activities in these waters. As a result, traditional latrines and bathing facilities have been built, leading to the direct discharge of domestic wastewater into the mangrove environment. A similar situation is observed at the Ozoughe and Ambowe sites, where the distinguishing feature is the disposal of all types of waste into the ecosystem, resulting in multiple forms of pollution [18][19].

In contrast, sites such as Okala and Ondongo benefit from significant conservation efforts aimed at protecting the ecosystem. For instance, at Okala, a non-governmental organization actively works toward ecosystem preservation, ensuring that the area remains clean.

The sampling period extended from July to September 2025. During this period, surface water samples were collected at ebb tide from five points at each site. For each of the 45 water samples analyzed, the sampling unit consisted of 500 mL of surface water. Samples were collected using sterile polyethylene containers. These containers were immediately stored in a light-protected isothermal bag containing ice packs (−4˚C) and transported to the laboratory for analysis.

Upon arrival at the laboratory, each water sample underwent four successive serial dilutions up to 10⁻⁴. Subsequently, 1 mL of each original sample and its dilutions were spread onto D-Cococcel agar plates (bioMérieux, France) using a sterile 10 µL inoculation loop, and incubated at 37.0 ± 0.5˚C for 48 hours. After incubation, all bacterial colonies exhibiting distinct morphology and coloration were isolated on the same medium under identical conditions.

From each plate, only grey colonies on a black background suspected to belong to the genus Aerococcus were selected. Culturing of the 45 wastewater samples resulted in a total of 51 colonies. Each colony was subjected to Gram staining, as well as oxidase and catalase tests. Final identification was performed using API 20 Strep galleries (bioMérieux, France). Results were interpreted using the ApiWeb software.

Antibiotic susceptibility testing was carried out using the disk diffusion method on Mueller-Hinton agar (bioMérieux, France), as described by [20]. To ensure a comprehensive assessment for this taxon, CLSI standards (2024) were utilized to supplement EUCAST (2023) criteria where data were limited. This approach precluded the use of surrogate criteria, thereby ensuring a rigorous interpretation of inhibition zone diameters. A total of 14 antibiotics were tested to determine the susceptibility profiles of the isolates, including: oxacillin (5 µg), penicillin (10 µg), erythromycin (10 µg), chloramphenicol (30 µg), azithromycin (30 µg), vancomycin (30 µg), teicoplanin (30 µg), streptomycin (10 µg), gentamicin (30 µg), amikacin (30 µg), tetracycline (30 µg), linezolid (30 µg), cefotaxime (30 µg), and ciprofloxacin (30 µg).

Overall and site-specific resistance/susceptibility proportions were compared us-ing Pearson’s Chi-square test or Fisher’s exact test when expected cell counts were below 5. Multidrug resistance (MDR) score, defined as the number of distinct an-tibiotic classes to which an isolate exhibited resistance, was compared across sites using the Kruskal-Wallis test. All analyses were performed with statistical signifi-cance set at α = 0.05.

The identification of the 51 colonies selected on D-coccocel agar resulted in the detection of Aerococcusviridans with excellent and very good confidence levels. The proportion of the 51 selected suspect colonies identified as A.viridans was 100%. Only a single isolate per positive sample was selected for antimicrobial susceptibility testing to ensure the statistical independence of observations.

The overall resistance observed across the entire sample was 38.40%, whereas the overall susceptibility reached 61.60%. The overall resistance and susceptibility rates were calculated using the total number of individual susceptibility tests as the explicit denominator. Accordingly, the overall resistance percentage was defined as the total number of results classified as ‘Resistant’ (R) divided by the total count of isolates tested.

Table 1. Overall antibiotic susceptibility.

The results (Table 1) highlight two major trends among Aerococcusviridans strains isolated from mangrove waters. Overall susceptibility (61.6%): the majority of the isolates remained susceptible to the antibiotics tested, and overall resistance (38.4%), where more than one-third of the isolates exhibited resistance. In microbiology, resistance levels approaching 40% are generally considered high.

The presence of approximately 38% resistance in A.viridans within this specific environment suggests two main hypotheses. First, it may reflect strong selective pressure, likely due to anthropogenic pollution such as domestic, agricultural, or hospital wastewater discharge. Second, the presence of antibiotics in aquatic ecosystems may drive bacterial adaptation. Thus, proximity to human activities is likely a key factor in this observed resistance. Such resistance levels in A.viridans could potentially impact the survival of certain mangrove species if infections occur.

Resistance levels varied markedly between sampling sites. The highest resistance rates were recorded at Lowe (76.19%) and Ambowe (61.90%), whereas the lowest were observed at Alenakiri (10.71%) and Okala (16.67%) (see Figure 2). This spatial analysis provides critical geographical insight, demonstrating that Aerococcusviridans resistance is not uniformly distributed across mangrove ecosystems in Grand Libreville, but instead varies significantly depending on the sampling location.

Two sites exhibited particularly concerning resistance profiles, in which resistance exceeded susceptibility.

Lowe (76.2% resistance) represents the most critical point in the study. The inversion of the susceptibility/resistance ratio (only 23.8% susceptible) suggests a strong and continuous contamination source, potentially linked to nearby hospital, industrial, or dense urban discharge. Ambowe (61.9% resistance), where resistance is also predominant at this site. Together, Lowe and Ambowe may act as hotspots for the dissemination of resistance genes within the mangrove ecosystem. These sites are the ‘‘High-Risk Areas’’ (Resistance-Dominated Zones).

Additionally, there are transition areas (moderate risk): Bizango, Ondongo, Ozoughe, and Rouger. These areas showed intermediate profiles, with susceptibility remaining dominant (57% - 75%) but with resistance already well established (approximately 30% - 40%). These findings raise the possibility of a correlation with human activity. For instance, are Lowe and Ambowe located near residential areas or waste disposal sites? Alternatively, could this reflect localized pollution events rather than intrinsic resistance patterns of the species?

Finally, there are “high-sensitivity” areas. In contrast, some sites show a much healthier microbiological profile. Alenakiri (89.3% susceptibility) appears to be the least impacted, with a low resistance rate (10.7%), and could serve as a reference for physicochemical comparisons. Okala (83.3%) and Agondje (76.8%) seem relatively less affected by antibiotic selective pressure.

Figure 2.Resistance and sensivity of A. viridans by location.

Resistance prevalence to penicillin (P) reached 100% at all nine sites, con-firming this antibiotic as ineffective against A. viridansacross the entire study area, while linezolid (LZ) remained fully effective (0% resistance) at every site. Beyond these two extremes, resistance profiles diverged sharply by location: Lowe and Ozoughe exhibited near-complete resistance (67-100%) to most antibiotic classes tested, including vancomycin, teicoplanin, azithromycin and erythromycin, whereas Alenakiri, Okala and Ondongo showed resistance limited to one or two antibiotics (mainly erythromycin or streptomycin), with 0% resistance to the remaining twelve (see Figure 3).

Figure 3. Prevalence of antibiotic resistance (%) of A. viridans by location. Legend: Vancomycin (VA); Teicoplanin (TEC); Azithromycin (AZM); Erythromycin (E); Tetracycline (TE); Amikacin (AK); Streptomycin (S); Gentamicin (GEN); Chloram-phenicol (C); Linezolid (LZ); Oxacillin (Ox); Penicillin (P); cefotaxime (CTX), and ciprofloxacin (CIP).

Group 1 exhibited a mean resistance of 64.30%, compared to 28.60% in Group 2. Conversely, mean susceptibility was 35.70% in Group 1 versus 71.40% in Group 2 (as shown in Figure 4). Overall, isolates from Lowe, Ambowe, and Ozoughe were more resistant than those from Okala, Ondongo, Rouger, Bizango, Alenakiri, and Agondje.

High anthropogenic influence (Group 1), comprising the Lowe, Ambowe, and Ozoughe sites, exhibits critical levels of antibiotic resistance. This group shows a complete reversal of the standard microbiological profile of Aerococcus viridans, with resistance predominating (64.3%) and susceptibility becoming a minority trait (35.7%). By combining Lowe, Ambowe, and Ozoughe, a distinct “red zone” of the mangrove ecosystem can be identified. The fact that resistance reaches nearly two-thirds of the isolates suggests that these three sites may share highly pronounced ecological or geographical characteristics. This group most likely corresponds to mangrove sectors located closest to urban discharge outlets and wastewater drainage channels, where antibiotic residues accumulate and exert strong selective pressure favoring resistant strains.

Figure 4. Resistance and sensitivity of A.viridans by location.

Lowanthropogenicinfluence(Group2) includes the Okala, Ondongo, Rouger, Bizango, Alenakiri, and Agondje sites. The microbial dynamics observed in these sites are markedly reversed compared with Group 1 and more closely resemble a natural environmental profile, characterized by a clear predominance of susceptible isolates (71.4%) and a lower residual level of resistance (28.6%). A susceptibility rate exceeding 70% indicates that the ecological resilience of the mangrove ecosystem remains largely functional in these areas. These sites are likely situated in more remote zones, distant from direct anthropogenic pressures. Nevertheless, the presence of 28.6% resistant isolates indicates that chemical contamination and/or the transport of resistance genes through water circulation are not entirely absent, even in the least impacted areas.

These findings demonstrate that antibiotic resistance in the mangrove ecosystem is not randomly distributed, but rather follows a clear pattern of spatial compartmentalization. Comparison of the mean resistance rates per isolate between Groups 1 and 2 revealed a highly significant difference according to both the Chi-square and Fisher’s exact tests (p < 0.001). This result confirms that the observed difference in resistance levels between the two groups cannot be attributed to simple random variation.

Figure 5 shows the distribution of multidrug resistance scores among Aerococcusviridans isolates across the nine sampling sites. The MDR score represents the number of antibiotic classes to which each isolate exhibited resistance. It varied significantly between sites (Kruskal-Wallis test, p = 1.19 × 10⁻⁵), indicating substantial spatial heterogeneity in antimicrobial resistance profiles. While some sites exhibited relatively low and homogeneous MDR scores (Group 1), others showed higher median values and a broader dispersion of scores (Group 2). The observed variability suggests that local environmental and anthropogenic factors may influence the selection and dissemination of antimicrobial resistance. Sites characterized by higher MDR scores may be exposed to stronger selective pressures, potentially associated with antimicrobial use, wastewater contamination, livestock activities, or other human-related inputs. Conversely, sites with lower MDR scores may represent environments subjected to reduced antibiotic pressure.

Figure 5. Distribution of MDR scores by sampling location (Kruskal-Wallis test, p = 1.19 × 10⁻⁵).

The occurrence of A.viridans in the environment constitutes a significant indicator of contamination of human origin, although its precise sources may sometimes be complex to identify [16][21]. Moore etal., [22] reported that the isolation of A.viridans at high frequencies exclusively in polluted marine water samples (Avalon Bay, USA), but not in wastewater effluents or urban runoff, could be attributed to the existence of alternative sources of this bacterium in marine environments, even though such sources remain poorly documented to date. Nevertheless, its presence in the oral cavity [21] or among domestic fauna [23]-[25] may contribute to indirect environmental contamination.

A.viridans, a fastidious Gram-positive coccus associated with a wide range of diseases in both humans and animals, has been isolated from environmental samples, aquatic organisms, wildlife, livestock, and animal-derived products [21][26]. To our knowledge, this is the first study investigating A.viridans in wastewater samples from urban mangrove aquatic ecosystems. Our results indicate that the prevalence of antibiotic resistance is higher in the Lowe and Ambowe sites, which correspond to areas with greater anthropogenic pressure, consistent with findings from previous studies [27]. Urbanization and industrialization contribute to increased loads and adverse effects associated with antibiotic discharge into aquatic environments [28]. As reported by several authors, a high prevalence of antibiotic-resistant bacteria in the environment often serves as an indicator of anthropogenic pressure, such as antibiotic pollution or urban discharge [29][30]. Several studies corroborate our findings. For instance, in India, the highest abundance of antibiotic-resistant bacteria was recorded in mangrove ecosystems receiving human waste discharge [31], suggesting that these sites were the most polluted by anthropogenic activities, including untreated domestic, commercial, industrial, and agricultural effluents. This supports the hypothesis that the relatively high levels of antibiotic resistance may result from the exchange of resistant bacteria between humans, aquatic fauna, and the environment [28][32]. Overall, these findings suggest that antibiotic-resistant A.viridans should be considered a zoonotic bacterial pathogen [21][33]. Furthermore, its presence may also be regarded as an indicator of human-derived contamination [4].

The presence of antibiotic-resistant A.viridans strains in mangrove ecotones may pose significant public health concerns [5]. This bacterium is frequently isolated in hospital settings and has also been identified as a marine pathogen responsible for fatal diseases in several aquatic species, including sea turtles, fish, and lobsters [34][35]. Available data indicate that A.viridans can be highly pathogenic to certain aquatic animals, representing both ecological and economic risks in coastal ecosystems [35][36]. This ecological risk is compounded by a direct threat to human populations, particularly those relying on subsistence fishing in these waters. Comparison with existing literature highlights that seafood consumption constitutes a well-documented route of human infection by A.viridans [12]. Mangroves serve as nurseries for crabs and shrimp, which, once contaminated, may transmit the pathogen to humans [37]. A clinical case in India linked A.viridans infection to lobster consumption two weeks prior to symptom onset [38].

In our study, resistance rates to penicillin and erythromycin were 100% and 60%, respectively. These values are comparable to those reported in Burkina Faso, China, and Algeria, where penicillin resistance reached 83.3% [14][39][40], oxacillin 83.3%, chloramphenicol 80%, and erythromycin 66.7% [21][40]. Contrary to some earlier findings [40], A.viridans was previously considered susceptible to commonly used hospital antibiotics. However, recent studies indicate that it often remains susceptible to β-lactam antibiotics, particularly penicillin, which is regarded as the treatment of choice for infections caused by this bacterium [21]. The observed resistance to penicillin and erythromycin may be explained by their frequent use in treating Group B Streptococcus (GBS) infections, with erythromycin serving as an alternative to penicillin [41]. Variations in resistance and MDR across study sites may be attributed to differences in antibiotic usage patterns between geographical regions and other anthropogenic pressures [30]. Indeed, several reports indicate considerable diversity in antibiotic resistance profiles among isolates from different locations [10]. Moreover, published data on the antibiotic susceptibility of environmental A.viridans isolates remain limited. The presence of this rare environmental pathogen is likely to pose public health challenges. Transmission to humans may occur through foodborne or environmental exposure [21][38]. Numerous studies have reported infections in fish and crustaceans caused by this bacterium [36][42], which may subsequently lead to human infection following seafood consumption.

Mangrove ecosystems provide suitable habitats for many of these seafood species [43][44], including mud crabs and shrimp [45][46]. Notably, A.viridans has been detected in mangrove mud crabs [45]. These organisms represent an essential source of nutrients, such as proteins, lipids, and minerals, for human populations in many regions [44][47].

Consequently, this situation may lead to significant health risks, particularly in developing countries where resources for managing public health crises are often limited.