Comparative study on the diversity and abundance of bacterial composition in four non-biting synanthropic flies from The Gambia and China

Abstract

Background

Non-biting synanthropic flies from the order Diptera have long been implicated as mechanical vectors to several human and animal pathogens, yet their microbiome variations across different regions remain poorly understood. Here, we compared the abundance of gut microbial composition in four species (Chrysomya megacephala, Lucilia cuprina, Musca domestica, and Physiphora clausa of non-biting synanthropic flies from two countries (The Gambia and China), providing insights into the potential pathogenic bacterial taxa they harbor. This study was conducted in Kanifing Municipal Council, The Gambia, and Changsha city, China, with sampling occurring between August and November 2023. Illumina NovaSeq6000 sequencing was used to amplify the V3-V4 region of the 16S rDNA from the midgut of these flies, which were pooled (n = 20 per sample) into 21 samples comprising Chrysomya megacephala, Lucilia cuprina, Musca domestica (from both cities), and Physiphora clausa (from KMC). Alpha and beta diversity indices were used to compare bacterial composition in the midguts of these flies. All bioinformatics and statistical analyses were performed using the BMKCloud online platform (http://www.biocloud.net).

Results

Taxonomic classification was annotated into 2 kingdoms, 44 phyla, and 2,379 species. Proteobacteria (41.14%) and Firmicutes (39.51%) dominated across all the samples. Significant geographic differences were observed. Shannon Alpha diversity analyses differed significantly between countries (P = 0.023), and Bray-Curtis-based PERMANOVA confirmed distinct bacterial compositions (R²=0.201, P = 0.001). Gambian flies harbored more Wohlfahrtiimonas chitiniclastica (13.79%), while Changsha samples contained higher Pseudomonas (3.99%). To the best of our knowledge, this is the first description of the gut microbiome of P. clausa, which was dominated by Corynebacterium (15.82%). These geographic and species-specific patterns highlight flies as reservoirs for regionally relevant pathogens.

Conclusion

This study highlights the geographic variability in gut microbiota of non-biting synanthropic flies as potential carriers of regionally relevant bacterial taxa and motivates further investigation into fly-borne pathogen transmission in both regions. This study also provides, to the best of our knowledge, the first description of the gut microbiome of P. clausa.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12917-026-05620-7.

Background

Synanthropic flies are insects that live in close association with humans, animals, and their environments, where they play important ecological and public health roles [1, 2]. These flies may be biting or non-biting, with the latter predominantly belonging to the families Calliphoridae, Muscidae, and Sarcophagidae [3]. More than 300 species of synanthropic flies have been documented worldwide, and many of which harbor diverse microorganisms on their body surfaces and within their digestive tract, making them potential mechanical vectors to pathogens [4]. Among the most common non-biting synanthropic flies reported are the Musca domestica, Chrysomya megacephala, and Lucilia cuprina, which possess sponging mouthparts adapted for feeding on a wide range of substrates, including decaying organic matter and human food [5, 6].

Non-biting synanthropic flies have long been implicated as mechanical vectors of microorganisms of medical, forensic, and veterinary significance, owing to their close contact with microbe-rich substrates and their frequent movement between contaminated environments and human or animal food [7, 8]. Larval stages of these flies have been found to cause myiasis in humans [9, 10]. Their life cycles and feeding behaviors frequently bring them into contact with microbe-rich environments such as animal manure, carrion, refuse, and decomposing organic matter, facilitating the dissemination of diverse microorganisms [3, 11–13]. During feeding and movement, flies can transfer pathogens via contaminated legs, mouthparts, body hairs, regurgitated gut contents, or feces, thereby contaminating food, surfaces, and wounds [14]. Consequently, a broad range of pathogens, including bacteria [3], parasites [15], fungi [16], and viruses [17] are known to be mechanically transmitted by non-biting synanthropic flies [3, 13, 18–21].

Increasing evidence suggests that the gut, particularly the midgut, serves as a more stable and diverse reservoir of microbial communities than the external body surfaces of flies [20]. While surface-associated microbes may be transient and influenced by environmental contact, midgut-associated bacteria are shaped by host physiology, diet, and microbial interactions, allowing for potential proliferation [21, 22]. Several studies have reported the gut microbial loads of these flies, suggesting a more reliable compartment for assessing fly-borne associated pathogens for a more contaminant-free bacterial microbiome [23–25]. Importantly, differences in gut microbiota composition have been observed among fly species, reflecting variations in feeding habits, habitat preferences, and ecological niches [23–25].

Despite growing interest in fly-associated microbiomes, most studies have focused on single fly species [26] or restricted geographic locations, limiting our understanding of how microbial communities vary across species and regions. Comparative studies examining multiple non-biting synanthropic fly species across distinct geographic and environmental settings remain scarce, particularly across countries with contrasting environmental and sanitation conditions. This study aimed to compare the midgut bacterial composition of four non-biting synanthropic fly species collected from KMC and Changsha, two cities that differ in sanitation infrastructure and waste management practices, providing insights into species- and geography-associated variation in microbial composition.

Results

Taxonomic Annotation

The bacterial 16S rDNA gene was analyzed from samples of Chrysomya megacephala, Musca domestica, Lucilia cuprina, and P. clausa. As shown in supplementary file_1_Table 1, C. megacephala, M. domestica, and L. cuprina each constituted six samples (3 samples from each city), amounting to 18 out of the 21 samples. Each of these three fly species had three replicates, three months apart, in both cities. Each fly species had one sample each month (fly midgut). Three P. clausa samples, NDS1, NDS2, and NDS3 from the KMC, were also included in the study, bringing the total to 21 samples.

A total of 2,593,693 high-quality reads were obtained from the 21 samples after quality checks and were then grouped into operational taxonomic units (OTUs). The OTUs were annotated into 2 kingdoms, 44 phyla, 120 classes, 326 orders, 706 families, 1,681 genera, and 2,379 species. The species distribution histogram shows the proportional representation of the top 10 organisms from each phylum, class, genus, and species across the four species. The other extremely low (< 1%) bacterial genera or species were consolidated into a collective category labeled as “OTHERS”.

After fly species identification, three non-biting flies were selected for midgut bacterial isolation analysis due to their high abundance in both cities, while the P. clausa was included because it was found only in the KMC (Gambia) samples, and no previous studies have reported on the bacterial composition of this fly. In this study, four non-biting synanthropic fly species were assessed, and 12,529 OTUs were detected. The bacterial taxa from the Changsha samples from all three species (C. megacephala, L. cuprina, M. domestica) showed a notably higher OTU abundance compared to the KMC samples. The Changsha samples collectively represented 63.93% of the total detected OTUs, in contrast to the lower OTU representation from the KMC samples. L. cuprina had the highest percentage of OTUs at 46.88%, followed by M. domestica at 43.5%, and C. megacephala at 23.4%. Bacterial taxa from the P. clausa accounted for 12.27% of the detected OTUs.

Additionally, Changsha samples recorded 7,556 unique OTUs while KMC samples recorded 4,399 unique OTUs. Samples from both cities recorded 574 shared OTUs (Figure 1). At the phylum level, Changsha recorded 6 unique OTUs, while KMC samples had no unique OTUs, and 38 shared OTUs existed between the two cities (Fig. 2). Similarly, at the genus level, Changsha samples recorded 545 unique OTUs, while KMC samples recorded 265 unique OTUs and shared 772 OTUs (Fig. 3).

Fig. 1.

Venn diagram of OTU-level classification between all Changsha and KMC samples

Fig. 2.

Venn diagram of phylum-level classification between all Changsha and KMC samples

Fig. 3.

Venn diagram of genus-level classification between all Changsha samples and KMC samples

The overall most abundant bacterial taxa at the phylum level across all samples (both cities) were Proteobacteria (40.11%) and Firmicutes (37.27%), followed by Actinobacteroidota (8.46%) and Bacteroidota (6.67%) (Fig. 4A, B). Extremely low percentage bacterial phyla were also detected. At the family level, Wohlfahrtiimonadaceae (12.15%) was the most represented across all samples, followed by Vagococcaceae (10.85%) and Streptococcaceae (6.82%). Of the identified genera, Vagococcus (10.85%) and Wohlfahrtiimonas (8.48%) were the most abundant bacterial taxa. This was followed by Lactococcus (6.03%) and Corynebacterium (4.24%). Low-abundant genera included Acinetobacter (2.94%), Wolbachia (2.85%), unclassified Enterobacteriaceae (2.60%), Dysgonomonas (2.34%), Ignatzschineria (2.30%), and Escherichia-Shigella (2.24%). Most of these bacterial genera are known to harbor potential bacterial pathogens of humans and animals. Extremely low-abundant taxa “Others” accounted for 54.83% and each reported < 1% (Fig. 5A, B) (Supplementary file 2).

Fig. 4.

AB: Relative abundance (%) of bacterial taxa at phylum levels across all 21 samples from KMC and Changsha. A Phylum-level relative abundance (%) of the top ten bacterial taxa across all samples. B Relative abundances (%) of the top five phylum-level level bacterial taxa across all samples

Fig. 5.

AB: Relative abundance (%) of bacterial taxa at genus-levels across all 21 fly midgut samples from KMC and Changsha. A Genus-level relative abundance (%) of the top ten bacterial taxa across individual samples. B Relative abundances (%) of the top five genus-level bacterial taxa across all samples

Species-specific bacterial composition and abundance

C. megacephala samples

Significant differences in bacterial composition at the phylum and genus levels were observed within C. megacephala samples from Changsha and KMC. The most abundant bacterial phyla were Proteobacteria (41.26%) which were higher in KMC samples, and Firmicutes (41.25%) which were higher in Changsha samples. Other low-abundant bacterial taxa encountered were Fusobacteria (6.73%), Bacteroidota (6.25%), and Actinobacteriota (3.18%) (Fig. 6A, B) which were reported in samples from both cities.

Fig. 6.

A-B: Relative abundance (%) of bacterial taxa at the phylum levels in Chrysomya megacephala midgut samples from KMC and Changsha. A Phylum-level relative abundance (%) of the top ten bacterial taxa across individual of C. Megacephala samples. B phylum-level relative abundances (%) of the top five bacterial taxa across individual C. megacephala samples

At the family level, Wohlfahrtiimonadaceae (13.27%) and Vagococcaceae (12.66%) were the most abundant families, followed by Anaplasmataceae (9.76%), Streptococcaceae (8.78%), Fusobacteriaceae (6.70%), and Lactobacillaceae (6.38%). Family Wohlfahrtiimonadaceae, Vagococcaceae, and Anaplasmataceae were relatively predominant in KMC samples compared with Changsha replicates. Conversely, Streptococcaceae was prominently higher in the latter. Fusobacteriaceae, Lactobacillaceae, and Enterobacteriaceae were detected in all replicates in varying low abundances.

Most bacterial taxa were classified up to the genus level across the C. megacephala samples. Vagococcus (12.66%) was found in all samples; Wolbachia (9.77%) was found only in KMC samples; Wohlfahrtiimonas (8.28%) was found only in KMC samples; and Lactococcus (8.09%) was found only in Changsha samples. These bacterial taxa recorded the highest abundances in C. megacephala. Lactilactobacillus bacteria were present in all Changsha replicates but absent from all three KMC samples. Fusobacterium was detected in all samples except the ChOC replicate, suggesting that the month of collection influenced the bacterial composition of the samples. Dysgonomonas and other bacterial genera were present in all six samples in varying abundances (Fig. 7A, B) (Supplementary file 3).

Fig. 7.

A-B: Relative abundance (%) of bacterial taxa at the phylum and genus levels in Chrysomya megacephala midgut samples from KMC and Changsha. A Relative abundances (%) of the top ten phylum-level bacterial taxa in C. megacephala samples from both cities. B Relative abundance (%) of the top five genus-level bacterial taxa in C. megacephala samples from both cities

L. cuprina samples

Phylum Firmicutes and Proteobacteria accounted for approximately 74% of the total bacterial OTU of the L. cuprina samples from both cities. However, we observed significant variations among the six samples across the phyla. For example, LOC, LSG, and LNC recorded the highest abundances in the phylum Firmicutes compared to the other three samples. Unlike the C. megacephala samples, the L. cuprina samples show no significant variation between the two cities. Fusobacteriota was < 1% in all three Changsha samples, absent in the LAG sample, and relatively higher in LOG and LSG. Probacterium phyla were higher in the LSC sample (9.23%) and were < 1% in all the other samples (Fig. 8A, B).

Fig. 8.

A-B: Relative abundance (%) of bacterial taxa at the phylum levels in Lucilia cuprina midgut samples from KMC and Changsha. A Phylum-level relative abundance (%) of the top ten bacterial taxa across individual L. cuprina samples. B Phylum-level relative abundance (%) of the top five bacterial taxa across individual L. cuprina samples

At the family level, Streptococcaceae, Lactobacillaceae, Vagococcaceae, and Wohlfahrtiimonadaceae (respectively) recorded the highest abundances across the L. cuprina samples. The Vagococcaceae and Wohlfahrtiimonadaceae families were slightly more abundant in the KMC samples.

Lactococcus (8.19%) and Vagococcus (7.65%) were the most abundant bacterial genera. This was followed by Bacteroides (4.34%), Pseudomonas (3.99%), Weissella (3.84%), Paracoccus (3.36%), and Wohlfahrtiimonas (3.26%). Lactococcus bacterial taxa were present only in the Changsha samples. Vagococcus and Bacteroides genera recorded higher abundances in the KMC samples. Pseudomonas bacterial taxa were present in all samples, but showed higher abundances in the Changsha samples. Weissella and Paracoccus were present only in the Changsha samples. On the contrary, Wohlfahrtiimonas taxa were highly prevalent in the KMC samples (high abundances). Enterococcus was detected in all samples except the LAG replicate. Luteimonas bacterial taxa were present in only the Changsha samples (Fig. 9A, B) (Supplementary file 4).

Fig. 9.

A-B: Relative abundance (%) of bacterial taxa at the phylum levels in Lucilia cuprina midgut samples from KMC and Changsha. A Relative abundances (%) of the top ten phylum-level bacterial taxa in L. cuprina samples from both cities. B Relative abundances (%) of the top five genus-level bacterial taxa in L. cuprina samples from both cities

M. domestica

Phylum Proteobacteria (45.22%) and Firmicutes (34.54%) recorded the highest relative abundances from the M. domestica samples from both cities. This was followed by low proportions of bacterial phyla, Bacteroidota (9.37%) and Actinobacteriota (6.07%). All the remaining phyla, Chloroflexi, Patescibacteria, Fusobacteriota, and Cyanobacteria, recorded extremely low abundances (< 2% each) (Fig. 10A, B) from samples in both cities.

Fig. 10.

A-B: Relative abundance (%) of bacterial taxa at the phylum levels in Musca domestica midgut samples from KMC and Changsha. A Phylum-level relative abundance (%) of the top ten bacterial taxa across individual M. domestica samples. B Phylum-level relative abundance (%) of the top five bacterial taxa across individual M. domestica samples

Wohlfahrtiimonadaceae, Vagococcaceae, Streptococcaceae, Lactobacillaceae, Enterobacteriaceae, Moraxellaceae, and Dysgonomonadaceae were the seven most abundant bacterial families recorded from M. domestica across the six samples. Wohlfahrtiimonadaceae and Vagococcaceae were more abundant in the KMC samples. On the other hand, Streptococcaceae and Lactobacillaceae families were most predominant in the Changsha samples.

Vagococcus (13.80%) and Wohlfahrtiimonas (13.79%) were the most predominant bacterial genera and were higher in KMC samples than in Changsha samples. Conversely, the Changsha samples showed the highest bacterial dominance in the Dysgonomonas, Lactococcus, and Acinetobacter bacterial genera. Weissella and Fructilactobacillus bacterial taxa were higher in the Changsha samples. Corynebacterium and Providencia bacterial genera were present in all samples in similar proportions. Ignatzschineria bacterial taxa were present in only the KMC samples (Fig. 11A, B) (Supplementary file 5).

Fig. 11.

A-B: Relative abundance (%) of bacterial taxa at the genus levels in Musca domestica midgut samples from KMC and Changsha. A Relative abundances (%) of the top ten genus-level bacterial taxa in M. domestica samples from both cities. B Relative abundances (%) of the top five genus-level bacterial taxa in M. domestica samples from both cities

P. clausa

Proteobacteria (41.96%), Firmicutes (27.57%), and Actinobacteria (22.97%) bacterial phyla recorded the highest relative abundances from P. clausa samples. This was followed by an abundant bacterial phylum, Fusobacteriota (5.82%) (Fig. 12A, B). Unlike the other three fly species, all these families—Morganellaceae, Wohlfahrtiimonadaceae, Enterobacteriaceae, Vagococcaceae, Fusobacteriaceae, Lactobacillaceae, Coriobacteriaceae, Bacillaceae, Peptostreptococcaceae from P. clausa were detected in very high abundance across the sample.

Fig. 12.

A-B: Relative abundance (%) of bacterial taxa at the phylum levels in Physiphora clausa from KMC. A Phylum-level relative abundance (%) of the top ten bacterial taxa across individual P. clausa samples. B Phylum-level relative abundance (%) of the top five bacterial taxa across individual P. clausa samples

Corynebacterium (15.82%) recorded the highest bacterial genus from the P. clausa samples. This was followed by low-abundant bacterial genera such as Wohlfahrtiimonas (8.70%), Vagococcus (7.74%), Providencia (5.36%), Fusobacterium (5.01%), Escherichia-Shigella (4.84%), Proteus (4.45%), Collinsella (3.79%), Bacillus (3.67%), and Ignatzschineria (2.71%) (Fig. 13A, B) (Supplementary file 6).

Fig. 13.

A-B: Relative abundance (%) of bacterial taxa at the phylum and genus levels in Physiphora clausa from KMC. A Relative abundances (%) of the top ten phylum-level bacterial taxa in P. clausa samples. B Relative abundances (%) of the top five genus-level bacterial taxa in P. clausa samples

Alpha and Beta diversity metrics

The richness and bacterial diversity among the midgut of four non-biting synanthropic fly samples were estimated using the Shannon and Simpson alpha diversity indices. Wilcoxon rank-sum was used to analyze the statistical significance of the data, with a significance level set at P < 0.05. Shannon indices of the bacterial composition of samples from Changsha were significantly different (P = 0.023), and the Simpson index showed the same pattern with a significant difference (P = 0.049) (Fig. 14). However, samples from both cities recorded a high Simpson index exceeding 0.9, indicating generally diverse and evenly distributed bacterial communities.

Fig. 14.

Alpha diversity indices of the midgut of synanthropic flies from both cities (p < 0.05). Shannon index; Wilcoxon rank-sum test (p < 0.023) from KMC and Changsha. Simpson index (Wilcoxon rank-sum test, p < 0.049) from KMC and Changsha. Beta Diversity of the midgut of non-biting synanthropic flies; Principal Component Analysis (PCoA) and PERMANOVA analysis of the midgut of four synanthropic flies from both cities, Wilcoxon rank-sum test p < 0.05

The limited variance reported by PCoA (21.57%) explained, and the dispersed clustering patterns suggest that additional environmental and ecological variables, including sanitation conditions, seasonal effects, and feeding substrates, may likely contribute to microbiome variation. Samples from both countries were dispersedly clustered. However, samples from Changsha were less clustered, indicating higher bacterial diversity, especially in the Changsha samples. The PCoA analysis using the Bray-Curtis dissimilarity metric, followed by PERMANOVA, reveals a statistically significant difference in bacterial composition between samples from the two countries, supporting the influence of geographic factors on microbial structuring. The findings of this study demonstrate that the microbial composition of Changsha city and KMC samples differed considerably across all groups (R2 = 0.201, P = 0.001*) (Fig. 15).

Fig. 15.

Alpha diversity indices of the midgut of synanthropic flies from both cities (p < 0.05). PCoA1-2 samples from the KMC and Changsha. PERMANOVA analysis of samples from KMC and Changsha (R2 = 0.201, p = 0.001*)

Abundance and prevalence of potential pathogens

Potential pathogenic bacterial taxa of human and animal relevance were detected in all four fly species from both cities. Variations in bacterial abundance were largely species-dependent; however, geographic location and local environmental factors also influenced bacterial diversity and composition, although these factors were not the focus of this study.

Most detected potential pathogenic bacterial taxa were country-, month-, and species-specific. The potential bacterial pathogens recorded from KMC samples included Wohlfahrtiimonas chitiniclastica, Ignatzschineria ureiclastica, Ignatzschineria indica, and Vagococcus teuberi. The abundance of Weissella ceti was highly affected by the fly species (L. cuprina). Erysipelothrix (Erysipelothrix rhusiopathiae), Flavobacteriaceae (Myroides phaeus), Fusobacterium sp. RMA1065, Acinetobacter sp. PP1 bacterial species were detected only in C. megacephala samples. Similarly, the prevalence of Fusobacterium sp. Marseille P3599 was also detected in both C. megacephala and L. cuprina samples. Corynebacterium (Corynebacterium ulceribovis, Corynebacterium humireducens) was detected only in the P. clausa from the Gambia. Other important unclassified potential pathogens detected included unclassified Enterobacter, unclassified Klebsiella, and unclassified Enterococcus across the 21 samples, in varying proportions and at low abundances.

Discussions

Several studies have implicated non-biting synanthropic flies as mechanical vectors to a variety of human and animal pathogens [26]. The house fly (M. domestica), the oriental fly (C. megacephala), and the Australian blowfly (L. cuprina) are among the most widely studied insects that live in close association with humans and have been found to carry several bacteria on their midgut [27, 28]. Their feeding and environmental settings affect the gut microbial composition. Despite these flies sharing a similar environmental niche, they have been found to differ in the bacterial composition of their gut contents [29–31]. We equally investigate the gut bacterial content from P. clausa (Diptera: Ulidiidae), and to the best of our knowledge, this is the first study to isolate bacteria from P. clausa. This is also the first bacterial isolation study to compare the bacterial composition of four different non-biting synanthropic flies from KMC and Changsha over a three-month study period, with midgut bacterial content as the main comparison factor [21]. This study was a 3-month replicate study in both countries, conducted to monitor the consistency of bacterial composition over the study duration. However, we found that bacterial composition of the same fly species shows considerable variations in the different replicate months, suggesting the month of fly species collection largely influences the bacterial composition of a fly species. While this study did not primarily focus on the seasonal variation of bacterial composition, as reported elsewhere, it was not the focus of this study [36], the pattern discussed aims to shed light on the differences in fly midgut bacterial composition associated with species and geography.

Bacterial composition analysis

Proteobacteria and Firmicutes recorded the highest relative abundances in samples from both cities. Bacteroidota, Actinobacteriota, and Fusobacteriota phyla were present in all samples; however, they were recorded in extremely low abundances. These phyla were also dominant phyla in previous studies of the gut bacterial composition in C. megacephala, which reported Proteobacteria, Firmicutes, and Bacteroidetes to be the most predominant phyla, and this is similar to the results from this study [25, 32]. This could be due to their similar ecological niches. Even though the bacterial composition of P. clausa is unknown, results from our study revealed a bacterial composition at the phylum level similar to that reported in other synanthropic fly studies, for example, the housefly [33]. The high abundance of the phyla Proteobacteria and Firmicutes in flies could be associated with their roles in decomposing organic matter [34].

Despite all the studied samples having similar bacterial composition at the phylum taxonomic level, interesting variations were detected at other taxonomic levels, such as family, genus, and species, with varying relative abundances, which is crucial to our findings. For example, the most prevalent genus from all the samples in this study is Corynebacterium, and this genus was detected in very high abundances from the P. clausa samples from KMC, as reported in houseflies elsewhere [33]. Vagococcus and Wohlfahrtiimonas recorded the highest abundances from C. megacephala and M. domestica samples in this present study, and they were most predominantly detected from the KMC samples. These results might be associated with the widely shared microbiome that exists between C. megacephala and M. domestica due to shared feeding sources [31], and may also reflect the higher availability of decomposing organic substrates at KMC collection site where these bacterial taxa are highly enriched.

Contrary to C. megacephala and M. domestica, the Lactococcus genus was the most prevalent bacterial taxon among L. cuprina samples, in contrast to other fly samples from this study. Ignatzschineria genus was detected only in M. domestica and P. clausa samples and was most dominant in the KMC samples. The Wolbachia genus was present on C. megacephala samples and in very high abundance (9.77%), especially from the Changsha samples, and this bacterial taxon has been previously reported from pupal stages of C. megacephala in a much higher percentage [29]. Results from this analysis have shown that the bacterial composition is related to the type of fly species. For example, Pseudomonas, Paracoccus, Enterococcus, and Luteimonas bacterial taxa were detected only from L. cuprina samples with high abundances. While Providencia, Fusobacterium, Escherichia-Shigella, Proteus, Collinsella, and Bacillus bacterial taxa were recorded in high abundances from only the P. clausa samples from KMC, Gambia. Most bacterial taxa found in P. clausa are similar to bacterial taxa found in M. domestica of this study. Other M. domestica studies have found a similar bacterial composition [33], with the most predominant bacterial taxa; therefore, a thorough comparative analysis of the gut contents of these two flies is needed. These bacterial taxa were isolated from urban flies, and flies from this study were collected from urban settings in both cities, which can partly explain the similarity in bacterial composition between the two studies. Some of the variation observed in the presence of certain bacterial taxa at the genus level across the different fly species in this study could also be due to the fact that synanthropic flies share food sources. The observed differences in bacterial abundance between KMC and Changsha samples may reflect differences in sanitation and waste-management practices, available decomposing substrate, and dominant feeding sources in both cities [24].

Most of the reported bacterial taxa observed in the samples from our study were classified at the genus level. Several unclassified bacterial species were reported. This could indicate new species from these genera, and the fact that the Silva reference database has no reference sequences that are similar to those from this study [24, 35].

Our study also revealed that despite some samples being from the same species and collected from the same country and location, the month the samples were collected (sampling collection time) shows overwhelming variations in the bacterial composition of these flies, just as suggested in a previous study [33].

Potential bacterial pathogens

Potential pathogenic bacterial taxa have been isolated from the midgut of these flies from both cities. These pathogenic taxa are implicated as disease-causing agents in humans and animals and therefore warrant appropriate attention for public health concerns. Empedobacter falsenii, isolated from M. domestica, is an opportunistic Gram-negative pathogen that causes several infections in humans, especially immunocompromised individuals [36]. This bacterium has been isolated from various clinical samples (stool, ear pus, urine) and has also been found to cause bacteremia and wound infections [37–39]. In this study, the environmental conditions, such as the sample collection sites in both cities, were approximately 300–400 m near public hospitals, which could explain why this bacterium was detected, as previous studies have implicated it in carrying multi-resistance genes and in causing nosocomial infections [40]. The bacterial composition due to fly collection-site variation is not the purpose of this study; however, it has been reported to cause significant variation in the bacterial composition of non-biting synanthropic flies elsewhere [41].

Wohlfahrtiimonas chitiniclastica, Ignatzschineria indica, and Ignatzschineria ureiclastica are fly-borne bacteria found in close association with humans [42]. These bacteria were found in all samples of our study, from both cities, but were most predominantly found in the KMC samples. The parasitic genera Ignatzschineria and Wohlfahrtiimonas have been isolated from the Wohlfahrtia magnifica fly and ever since have been found to cause human and animal myiasis— infection of open wounds (septicemia, and bacteremia) [43] which could lead to septic shock and eventually death [44]. The bacteria have been isolated from several clinical cases [45–48]. Snyder and colleagues, for example, recovered these bacteria from barrowing larvae in the toes of an 82-year-old patient [49]. Our findings further confirm that these bacteria are fly-associated and underscore the need for proper fly-control strategies, particularly around healthcare facilities. These bacteria are associated mainly with sarcophagid flies [50], which contrasts with our findings, where they were detected in calliphorids, muscids, and the picture-winged Ulidiidae. In our study, these bacterial species were detected in higher relative abundance in KMC than in Changsha samples, although environmental parameters were not recorded in either city [51]. Wohlfahrtiimonas chitiniclastica may, in addition to its pathogenic role, also degrade chitin, a major component of the insect skeleton [52]. This could explain its detection in all the samples from both cities at relatively high abundance.

We equally detected the pathogenic bacteria (Weissella ceti) that cause infections in rainbow trout, fish, and whale, as previously reported [53–55]. The bacterium was found mainly in the Changsha samples. The high abundance of this bacterium in the Changsha sample may reflect differences in feeding substrates between the two cities, which needs further monitoring.

The presence of Corynebacterium variable in all the flies from both countries suggested its association with the digestion and degradation of lipids and the nutrient cycle in flies [56], and its ability to utilize proteins and lipids in microbe-rich surface habitats is consistent with a role for Corynebacterium in processing the decomposing organic substrates that flies feed on, and could explain its detection in all the flies sampled in this study. However, it is worth noting that some species of the genus Corynebacterium are pathogenic to both humans and animals. Corynebacterium ulceribovis and Corynebacterium humireducens are Gram-positive bacteria from a genus recognized for animal-associated zoonotic pathogens. C. ulceribovis was first isolated from the ulcerated skin of a cow’s udder in Germany [57, 58]. C. variabile was originally isolated from the surface of smear-ripened cheeses, where it contributes to flavor and texture development through casein hydrolysis, fatty-acid degradation, and the catabolism of organic acids such as lactate and propionate, and C. humireducens was originally recovered from compost. Bacteria of the genus Corynebacterium were detected in all four fly species in this study, but these two species in particular were detected exclusively in P. clausa from KMC, Gambia. Their presence in this fly is notable because non-biting synanthropic flies are recognized as mechanical carriers of pathogens of veterinary relevance [59–61].

These findings should be interpreted in light of the limitations of this study. Biological replication was modest (three pooled samples per fly species per site, with each pool comprising 20 individuals), which is limited within-group statistical power and masks individual-level variation. The sampling months did not fully overlap between countries (August-October in KMC and September-November in Changsha); therefore, the inter-city signal was partly confounded by temporal variation. Environmental parameters (ambient temperature, substrate type, and sanitation indices) were not taken into account during sampling.

Conclusion

This study explored the bacterial composition of the midgut of non-biting synanthropic flies from two countries. This comparative study not only addresses critical gap in understanding geographic microbial variation in these flies but also reported several bacterial genera with recognized veterinary importance, including Wohlfahrtiimonas, Ignatzschineria, Vagococcus, Weissella, and Corynebacterium, with city- and species-specific abundance patterns providing insights on species and geography-associated variation in microbial composition with observed city-specific abundance patterns at the genus level. Results from this study also provide the first reference data on the bacterial composition of P. clausa in the literature. Extensive pathogen isolation studies over a longer study duration are needed to confirm the viability and transmission potential of bacterial taxa identified in this study. We call for an in-depth characterization of the vector potential of P. clausa from the KMC, Gambia.

Methods

Fly sampling

Adult non-biting synanthropic flies were sampled between August and November 2023 in Kanifing Municipal Council (KMC) (Latitude: 13.45°N, Longitude: 16.68°W), Gambia, and Changsha City (Latitude: 28.23°N, Longitude: 112.94°E), China. We chose these two cities for two main reasons. The first is that they have very different sanitation and waste-management systems. Changsha operates a regulated, source-segregated municipal waste system, while KMC relies on a largely uncontrolled and informal one. This, in turn, affects the kind of decomposing material flies can feed on and how easily they can access it. Secondly, similar fly species were present in both cities, making cross-species comparison possible because different fly species have been reported to differ in gut bacterial composition. These factors have been shown to shape the gut microbiome of synanthropic flies, which enables a useful contrast in which a species-by-geography comparison was made.

Flies were collected at both cities around the main local market and 300–400 m away from the public hospital using commercial ranch-baited fly traps (manufactured in China). On each sampling day, one trap was deployed at each location and retrieved after a 9-hour deployment period. In each month’s collection, a new trap was used on each sampling day to avoid cross-contamination across sampling visits. Sampling was conducted 3 times per month for three consecutive months. After sorting and morphological identification, the three most abundant flies were selected for bacterial composition analysis in addition to the newly collected P. clausa flies from KMC, and 20 individuals of the same species were pooled into a single sample per month. This design yielded 21 pooled samples in total, with 6 traps in each month, amounting to 18 traps per country (18 trap-days in total). Nine from Changsha (three monthly pools each of C. megacephala, L. cuprina, and M. domestica) and 12 from KMC (three monthly pools each of C. megacephala, L. cuprina, M. domestica, and P. clausa. A total of 420 individual flies were analyzed. Sample identifiers and the corresponding species, city, and month for each pool are listed in Supplementary File 1, Table 1. Photographs of the trap setup at the representative sampling site in each city are provided as Supplementary Fig. 1.

A stereomicroscope (Motic SMZ-171) based on morphological diagnostic characteristics and region-wide taxonomic keys was used for species identification [62–65] and was confirmed by Sanger sequencing to ensure accuracy for taxonomic purposes. Convergence between morphological diagnostic features and molecular sequence measurements ensured reliable assignment of the specimens at the species level. Subspecies-level differentiation was not performed, as molecular confirmation provides sufficient resolution for species delimitation in the tested taxa.

Upon collection, flies from each trap were transferred into sterile plastic specimen bags at the sampling site and promptly placed in field coolers containing dry ice to ensure low temperatures during transport to the laboratory. The samples from Changsha were directly transported to the Meng Medical Entomology Laboratory, Department of Medical Parasitology, School of Basic Medical Sciences, Central South University, Changsha, China. Conversely, the KMC samples were initially transported and stored at the Medical Research Council Unit, The Gambia (MRCG) and subsequently shipped to the Meng Medical Entomology Laboratory in Falcon tubes, which were tightly sealed with parafilm and packed with silica gel desiccants to maintain dryness during international transport. All specimens from both cities were stored in − 80 °C until DNA extraction and subsequent molecular analyses of all samples were conducted at the Meng Medical Entomology Laboratory. All sequencing data generated in this study, including metabarcoding datasets, are publicly available in GenBank under the SRA accession number PRJNA1077112.

DNA extraction and PCR

Pooled flies were surface sterilized with 75% ethanol and 10% bleach (0.8% sodium hypochlorite) for 60 s. This was followed by two rinses in autoclaved distilled water, each lasting 5 min. The samples were subsequently rinsed in sterile phosphate-buffered saline (PBS) for a duration of 15 min to avoid cross-contamination [66]. The midgut samples were crushed using a motorized hand homogenizer to form fly homogenates, and total genomic DNA was extracted from adult flies using the TIANamp Genomic DNA Kit (Cat. no. 4992199/4992254; TIANGEN, China) according to the manufacturer’s instructions. To recover the maximum amount of DNA, each sample was subjected to elution twice using 50 µL of elution buffer (TE buffer). DNA concentration was quantified using a spectrophotometer, and integrity was verified by 1% agarose gel electrophoresis. The midgut of 420 flies was pooled (n = 20/sample) by species (C. megacephala, L. cuprina, M. domestica from both cities and P. clausa from KMC, Gambia).

The V3-V4 hyper-variable region of the bacterial 16S rDNA gene of adult non-biting synanthropic flies was sequenced on the Illumina NovaSeq6000 platform using the primer 341F: 5’-ACTCCTACGGGAGGCAGCA-3’; 805R: 5’-GGACTACHVGGGTWTCTAAT-3’ [67, 68]. PCR was performed in triplicate in a 25 µL reaction mixture following the instructions of KFX-101 (TOBOYO, Tokyo). Purified amplicons from different samples were pooled in equimolar amounts and sequenced by BioMarker Technologies Co., Ltd. (Beijing, China) (www.biomarker.com.cn).

Bioinformatics and statistical analysis

Raw data were filtered and processed by Trimmomatic (version 0.33) [69] to generate high-quality reads. The primer sequences were identified and removed by cutadapt (version 1.9.1) [70]. High-quality reads were clustered into OTUs by USEARCH (version 10.0) [71] and followed by chimera removal using UCHIME (version 8.1) [72]. Operational taxonomic units (OTUs) were defined as features in the analysis. Taxonomic annotation was performed at the OTU level. Microbial community composition at the phylum, class, order, family, genus, and species levels was revealed using taxonomic annotation in the Silva 138 Database [73]. Relative abundance was used to describe how much each bacterial taxon contributed to the total bacterial community at a given taxonomic level. It was calculated from the proportion of sequencing reads or OTUs assigned to each taxon within a sample or sample group and then expressed as a percentage for visualization and comparison.

Alpha and Beta diversities were evaluated by QIIME2 (version 2020.06) [74], combined with the algorithms R package, all statistical analyses were performed in R using the Vegan package [75] and ggplot2 was used for data visualization [76]. A bacterial diversity analysis of samples from the two cities of the two countries was compared using the Wilcoxon rank-sum test. The differences were then statistically quantified (P ≤ 0.05) using PERMANOVA (permutational multivariate analysis of variance) [77]. A spreadsheet of all analyzed taxonomic OTU counts per sample was also generated through Microsoft Excel to allow further comparison and is attached as supplementary file 2–6.

Abbreviations

rDNA

Ribosomal DNA

PCA

Principal Component Analysis

PCoA

Principal Coordinate Analysis

KMC

Kanifing Municipal Council (in this study, KMC is used to refer to the area from Kanifing Municipal Council to Brikama Marakissa)

PERMANOVA

permutational multivariate analysis of variance

ChSC

Chrysomya megacephala September Changsha

ChOC

Chrysomya megacephala October Changsha

ChNC

Chrysomya megacephala November Changsha

ChAG

Chrysomya megacephala August Gambia

ChSG

Chrysomya megacephala September Gambia

ChOG

Chrysomya megacephala October Gambia

LSC

Lucilia cuprina September Changsha

LOC

Lucilia cuprina October Changsha

LNC

Lucilia cuprina November Changsha

LAG

Lucilia cuprina August Gambia

LSG

Lucilia cuprina September Gambia

LOG

Lucilia cuprina October Gambia

MSC

Musca domestica September Changsha

MOC

Musca domestica October Changsha

MNC

Musca domestica November Changsha

MAG

Musca domestica August Gambia

MSG

Musca domestica September Gambia

MOG

Musca domestica October Gambia

NDS1

New Species - Physiphora species (P. clausa) August

NDS2

New Species - Physiphora species (P. clausa) September

NDS3

New Species - Physiphora species (P. clausa) October

Authors’ contributions

**BJJJ** collected all fly samples and performed all molecular analyses. **BJJJ** and **FM** designed the study. **BJJJ, ML, YM, YH, ST, AL, JH, JC, ** and **FM** did the conceptualization. **BJJJ** performed all molecular and statistical analyses. **BJJJ, ML, YM, YH, ST, AL, JH, ** and **JC** were responsible for imaging software. **F M** supervised, acquired the funding, and handled project administration. **BJJJ** wrote the original draft of the manuscript. **BJJJ** and **AL** edited the manuscript. All authors revised and approved the final version of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (grant number 32370554, 32460248), Science Foundation of Hunan Province (2022JJ30693), Research and Innovation Team Project, Xinjiang Medical University (XYD2024C05).

Data availability

The datasets generated and/or analyzed during the current study are available in GenBank under SRA accession: **PRJNA1077112**.

Declarations

Ethics approval and consent to participate

No formal permits and ethical approval are required for this study, as the sampled flies; C. megechephala, L cuprina, M domestica, and P clausa, are not protected under any national law.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.