Comparative analysis of gut microbiota between common (Macaca fascicularis fascicularis) and Burmese (M. f. aurea) long-tailed macaques in different habitats

The environment has an important effect on the gut microbiota—an essential part of the host’s health—and is strongly influenced by the dietary pattern of the host as these together shape the composition and functionality of the gut microbiota in humans and other animals. This study compared the gut microbiota of Macaca fascicularis fascicularis and M. f. aurea in mangrove and island populations using 16S rRNA gene sequencing on a nanopore platform to investigate the effect of the environment and/or diet. The results revealed that the M. f. fascicularis populations that received anthropogenic food exhibited a higher richness and evenness of gut microbiota than the M. f. aurea populations in different habitats. Firmicutes and Bacteroidetes were the two most abundant bacterial phyla in the gut microbiota of both these subspecies; however, the relative abundance of these phyla was significantly higher in M. f. aurea than in M. f. fascicularis. This variation in the gut microbiota between the two subspecies in different habitats mostly resulted from the differences in their diets. Moreover, the specific adaptation of M. f. aurea to different environments with a different food availability had a significant effect on their microbial composition.

Nanopore sequencing of bacterial 16S rRNA gene.The full-length bacterial 16S rRNA gene from 120 fecal samples of long-tailed macaques was successfully sequenced using high throughput nanopore sequencing.In total, 2,444,551 sequencing reads were obtained from 120 samples with an average read per sample of 20,371 (Table 1).The average classified reads were 18,091 per sample.According to the rarefaction analysis, all the samples had sufficient sequencing depth for estimation of the bacterial diversity (Fig. 2).
Bacterial diversity in the gut microbiome in Mff and Mfa.Bacterial alpha diversity (level of diversity within individual samples) comparisons between Mff and Mfa in the respective mangrove and island populations were evaluated based on the Chao1 index (Fig. 3a), while the richness and evenness of bacterial operational taxonomic units (OTUs) were determined using the Shannon diversity index (Fig. 3b).Statistical comparisons of indices between groups were carried out using a Kruskal-Wallis test, accepting significance at the P < 0.05 level.
The Chao1 index and Shannon's diversity between different habitat types of the M. fascicularis subspecies were compared.The Chao1 index of the Mff-KPE population on the island had a significantly higher OTU richness (P = 0.0021) than the Mff-BTB population in the mangrove forest.Likewise, the Shannon's diversity was noticeably and significantly higher (P = 0.0002) for the Mff-KPE island population than the Mff-BTB mangrove population.Similarly, the Mfa-PNY population living on the island showed a significantly higher OTU richness (P = 0.0021) and Shannon's diversity index (P = 0.0332) than the Mfa-MFRC mangrove population.Overall, the alpha diversity of Mff was significantly higher than that for the Mfa populations in both habitat types.
To further examine the differences between the samples, beta diversity (level of diversity or dissimilarity between samples) analyses was performed using the Bray-Curtis cluster analysis index to compare the microbial community compositions between Mff and Mfa in mangrove and island populations.The beta diversity (Fig. 3c) between Mff and Mfa in different habitat types (mangrove and island) were significantly different (P = 0.001, Figure 1.The total number of food items (per visit) consumed by Mff and Mfa living in a mangrove forest or on an island.The black, grey, and white columns indicate marine invertebrates, plants, and anthropogenic foods, respectively.
Table 1.Summary of the sequencing and reads classification (mean ± SD) in each population of Macaca fascicularis fascicularis (Mff) and M. f. aurea (Mfa) in the two respective habitat types.permutational multivariate analysis of variance [PERMANOVA]).However, the Mfa-MFRC mangrove population had a significant divergence from the other populations.

Taxonomic composition of the gut microbiota in
Mff and Mfa at different habitats.Firmicutes was the most dominant bacterial phylum among the mangrove and island Mff populations (Fig. 4, upper panel) with a mean ± SD proportion of 57.6 ± 14.6% and 57.3 ± 5.9%, respectively.The Bacteroidetes accounted for 24.0 ± 10.2% and 28.9 ± 6.8% in the Mff mangrove and island populations, respectively, making it the second most abundant phylum.However, the relative abundance of Firmicutes and Bacteroidetes was not significantly different between the Mff mangrove and island populations (4.7 ± 8.1 and 2.1 ± 1.9 for the mangrove and island populations, respectively; Mann-Whitney U test; P < 0.05).Proteobacteria, which made up 8.7 ± 18.0% and 4.4 ± 2.0% of the bacteriome in the Mff mangrove and island populations, respectively, was the third most Similarly, Firmicutes was the most dominant phylum in the Mfa mangrove (74.7 ± 27.2%) and island (64.9 ± 17.9%) populations, while Bacteroidetes was the second most abundant phylum in both Mfa populations (5.4 ± 10.7% and 20.6 ± 13.2% for the mangrove and island populations, respectively).In contrast to the Mff populations, the relative abundance of Firmicutes to Bacteroidetes ratio in the Mfa-PNY island population (8.9 ± 12.4) was significantly lower than the Mfa-MFRC-mangrove population (68.3 ± 82.5; Mann-Whitney U test; P < 0.001), which was aligned with the higher abundance level of Bacteroidetes in the Mfa-PNY island population.The Proteobacteria (14.0 ± 25.8%) and Verrucomicrobia (6.7 ± 6.1%) were the third most abundant phyla in the Mfa mangrove and island populations, respectively.
The top 10 most dominant genera in the bacterial communities of both macaque subspecies in the mangrove and island populations were identified and are shown in Fig. 4, middle panel.The most dominant bacterial genus in Mff was Oscillibacter at 13.3 ± 6.5% and 12.2 ± 3.3% in the mangrove and island populations, respectively.The other predominant bacteria in the bacterial microbiome of Mff were Prevotella, Clostridium sensu stricto, Clostridium XlVa, Faecalibacterium, and Intestinimonas.The proportions of these bacteria varied across samples and were different between the Mff mangrove and island populations.Oscillibacter was also the most dominant www.nature.com/scientificreports/bacterial genus in the fecal microbiome of the Mfa island population (17.8 ± 7.6%) and was higher than that in the Mfa mangrove population (4.9 ± 4.5%).In contrast, Clostridium sensu stricto was the most predominant bacterial genus in the Mfa mangrove population (23.1 ± 22.9%) and with a significantly higher abundance (P < 0.0001) than in the Mfa island population (4.9 ± 7.4%).
At the bacterial species level, Oscillibacter valericigenes was the most dominant species in Mff (13.38 ± 6.56% and 12.20 ± 3.36% in the mangrove and island populations, respectively) (Fig. 4, lower panel).The other less dominant bacterial species were Prevotella copri, Intestinimonas butyriciproducens, and Faecalibacterium prausnitzii; however, their abundance varied between populations.In contrast, Clostridium sardiniense was the most predominant bacterial species in the Mfa mangrove population (14.0 ± 15.5%) with a significantly higher abundance (P < 0.00001) than in the Mfa island population (0.03 ± 0.1%).
Comparison of the bacterial species between the different macaque subspecies (Mff and Mfa) in the same habitat types (island or mangrove) was examined by Mann-Whitney U tests (P < 0.05).The results revealed that the Firmicutes and Bacteroidetes were the two most abundant phyla in the Mfa and Mff island populations; however, the proportion of Firmicutes was not significantly different between them.In contrast, Bacteroidetes were significantly higher in the Mff (28.9 ± 6.8%) than in the Mfa (20.6 ± 13.2%) island populations.Similarly, the mangrove population of Mfa showed a significantly higher abundance of Firmicutes (74.7 ± 27.2%) and a lower abundance of Bacteroidetes (5.4 ± 10.7%) than the Mff mangrove population.
The taxonomic bacterial composition at a genera level showed that the Mfa island population (17.8 ± 12.2%) had a significantly higher abundance of Oscillibacter than the Mff island population (12.2 ± 3.3%), while Clostridium sensu stricto was significantly higher in the Mfa mangrove population (23.1 ± 22.9%) than in the Mff mangrove population (5.8 ± 11.1%).Further comparison at the bacterial species level reveled that Oscillibacter valericigenes was significantly more abundant in the Mfa island population (17.8 ± 7.6%) than in the Mff island population (12.2 ± 3.3%), while the Mfa mangrove population (14.0 ± 15.5%) had a significantly higher abundance of Clostridium sardiniense than the Mff mangrove population (0.4 ± 1.6%).

Differential abundance of gut bacteria between Mff and Mfa in different habitat types.
The taxonomic abundance of the gut bacterial microbiota of Mff and Mfa living in the mangrove forest and on the island were compared further using LEfSe analysis (LDA score > 2, P < 0.05) 39 , as shown in Fig. 5. Differences in the gut bacterial microbiota between the Mff and Mfa populations in the different habitat types of mangrove forest and island were identified.Porphyromonadaceae, Phascolarctobacterium succinatutens, Acidaminococcaceae, and Prevotella fusca were the most enriched taxa in the Mff-BTB mangrove population, while the Mff-KPE island population had a greater number of significantly enriched taxa, including Tannerella forsythia, Bdellovibrionaceae, Rikenella microfusus, Barnesiella viscericola, Ethanoligenens harbinense, Olivibacter sitiensis, and Fibrobacter intestinalis.In contrast, the Mfa-MFRC mangrove population was enriched in Lachnospiraceae incertae sedis, Clostridium saccharolyticum, and Eubacterium hallii.Moreover, Haloferula helveola and Bacteroides fluxus were abundant in the Mfa-PNY island population.These results indicate the significant differences in the compositional abundance of gut microbiota between Mff and Mfa in the mangrove and island populations.

Discussion
Comparative analysis of the gut bacterial microbiota between Mff and Mfa living in different habitat types (mangrove forest and island) revealed the potential influence of different environments and diets on their gut bacterial composition.Overall, the gut bacteria's alpha diversity in Mff was significantly higher than in the Mfa populations.This difference likely resulted from the increased enrichment of bacterial species in Mff populations (BTB and KPE), such as Oscillibacter valericigenes, Prevotella copri, Faecalibacterium prausnitzii, and Intestinimonas butyriciproducens, which was primarily influenced by the consumption of anthropogenic foods.A previous study in rhesus macaques (M.mulatta) also indicated that the population which consumed anthropogenic foods exhibited a higher microbial richness compared to the wild population that freely foraged for natural foods 40 .Generally, the gut microbial diversity tends to be higher in wild animals compared to captive animals, which is mainly attributed to the complexity of their diet in their natural habitats 41,42 .One possible explanation for the higher bacterial richness in the Mff KPE and BTB populations is that, apart from the natural foods in their natural habitats, these Mff could access anthropogenic foods regularly.These findings suggest that the gut microbiome's bacterial composition in Mff is primarily influenced by types of food they consume rather than the habitat types they inhabit.Nevertheless, the effect of host (macaque) genetics, which differ between Mff and Mfa, cannot be ruled out.
The gut microbiota of Mff and Mfa in this study were mainly composed of two phyla, the Firmicutes and Bacteroidetes, which were most likely similar to that of humans and other NHPs, including other wild and captive Thai Mff [34][35][36][37][43][44][45][46] . Note tat the composition of Firmicutes in the Mfa-MFRC mangrove population was highest among the four examined populations of long-tailed macaques.Thus, the relative abundance of Firmicutes exhibited variations among different subspecies and habitat types.Specifically, the Mfa-MFRC mangrove population showed a higher relative abundance compared to the Mfa-PNY island population, and the Mfa populations displayed a higher relative abundance compared to the other two Mff populations.Firmicute species contained numerous genes encoding enzymes related to energy metabolism, and these bacteria can produce a wide variety of digestive enzymes to decompose various substances, assisting the host in the digestion and absorption of nutrients 47 .According to previous studies, a higher ratio of Firmicutes to Bacteroidetes is associated with a higher absorption of dietary energy 48,49 .Bacteroidetes species helped the host in metabolizing the proteins and carbohydrates in the diet 50,51 .Taken together, it can suggest that the abundance of Firmicutes and the ratio of Firmicutes to Bacteroidetes are related to the genetic characteristics (leading to a different subspecies of Mff and Mfa), habitat type (mangrove forests or island), and anthropogenic foods (only in Mff populations).The higher www.nature.com/scientificreports/abundance of Firmicutes to Bacteroidetes may partially be related to the consumption of the high-energy mollusk foods that were observed to be heavily consumed in the Mfa populations in this study.These Mfa populations were observed to primarily rely on the natural food sources available in their respective habitats subject to their specific foraging techniques to acquire these foods.In addition, the abundance of bacteria belonging to the phylum Proteobacteria in the Mfa-MFRC mangrove population was significantly higher than in the Mfa-PNY island population, which could reflect the effects of the habitat type and food items.
During fecal specimen collections, we discovered that the Mff-KPE populations, especially adults, sporadically used percussive stone tools for opening oysters.Thus, the higher bacterial species richness observed in the Mff-KPE population can be attributed to their consumption of anthropogenic foods and their stone-tool use behavior, which allows them to access more food items requiring foraging techniques.This indicates that while diet plays a significant role in bacterial diversity, stone-tool use behavior also contributes to the bacterial diversity.However, due to the short-time stay and lack of individual animal identification and stone-tool use in this study, we were unable to collect data on the proportion of food types consumed by the monkeys on a daily basis.Besides, the data on stone-tool use by each population was obtained from previous studies 24,27,38 , and were also confirmed during the field observations.This limitation hinders our ability to analyze the microbiome composition at an individual level based on the proportion of food consumption.To address this limitation in future research, it would be beneficial to identify each animal individually, collect data on the proportion of their daily www.nature.com/scientificreports/food consumption with or without stone-tool use, and then analyze the microbiome composition at the individual level.Thus, collecting data on the proportion of food items acquired through stone-tool use and without the use of stone-tools for each individual animal would allow for a more detailed analysis of the relationship between stone-tool use, dietary habits, and the gut microbial profiles.Such an individual-level analysis would provide valuable insights into how specific dietary behaviors shape the gut microbiota within each population of M. fascicularis, contributing to a deeper understanding of the factors driving gut microbiome variation in these macaque populations.At the genus level, our results indicated that the microbiome of long-tailed macaque populations was enriched with Prevotella, which is one of the most predominant genera in the human microbiome.In line with these findings, a previous study also reported that the macaque microbiome exhibited a higher abundance of Prevotella than the human microbiome 52 .The predominance of Prevotella was associated with a diet high in carbohydrate and fiber from plant sources 53 .Similarly, western lowland gorillas (Gorilla gorilla gorilla) that consumed a high number of fruits had a high relative abundance of Prevotellaceae 54 .These findings suggest that populations with a higher abundance of Prevotella possess the capacity to effectively break down and utilize the natural plantbased diet.
At the bacterial species level, Oscillibacter valericigenes was the most dominant species in the Mfa and Mff populations with the exception in the Mfa-MFRC mangrove population.Oscillibacter valericigenes is a representative bacterium in the Oscillibacter group that can produce valerate 55 , a short-chain fatty acid that can replace butyrate as an energy source for colonocytes.This bacterium's abundance showed its potential relevance to the macaque's health.These results are also consistent with a previous study reporting a significant abundance of O. valericigenes in healthy humans 56 .Similarly, Faecalibacterium prausnitzii was present in all four populations of long-tailed macaques examined in this study, which is supported by previous studies that F. prausnitzii was the dominant butyrate producer of Clostridium cluster IV, the most common bacteria in the microbiome of humans, and which exhibited anti-inflammatory effects 57 and enhanced the gut barrier functions 58 .The depletion of F. prausnitzii is associated with Chron's disease 59 .Note that the microbiome of the Mfa-MFRC mangrove population was enriched with Clostridium sardiniense and less diversified.These findings are significant for the health of long-tailed macaques because a reduced diversity in the gut microbiota results in fewer microbial metabolic pathways interacting with food items and providing fewer nutritional benefits to the hosts.Similarly, in other mammalian species, a low gut microbial diversity has also been associated with heightened vulnerability to opportunistic pathogens 60 .Clostridium sardiniense is a glycolytic cluster I species that uses anaerobic carbohydrate fermentation to produce butyrate 61 .This species can also promote a more severe infection of Clostridioides difficile in mice by modulating the virulence, growth, and colonization of the pathogen 62 .Also, the reduced Chao 1 and Shannon alpha diversity of the microbiome in the Mfa-MFRC mangrove population could potentially be attributed to the higher abundance of C. sardiniense.It is essential to highlight that the Mfa-MFRC mangrove population in this study are wild animals and are not habituated to human presence.Due to COVID-19 restrictions, human activities were limited during the field observations of Mfa-MFRC mangrove population.As a result, these animals predominantly relied on natural food sources, leading to a less diverse range of microbial species compared to the Mff populations, which had access to both natural and anthropogenic foods.
The LEfSe-based differential species abundance analysis of the Mff-BTB mangrove population revealed that Porphyromonadaceae and Phascolarctobacterium succinatutens were the most enriched taxa.The Porphyromonadaceae have a potential role as adiposity modulators by producing two short-chain fatty acids: acetate and propionate 63,64 .Phascolarctobacterium succinatutens is known for its utilization of succinate and has been previously identified in the gut of healthy humans 64 .These results suggest that the Mff-BTB mangrove population have a specific diet that promotes the growth and proliferation of Porphyromonadaceae and P. succinatutens.These bacteria are known to thrive on certain dietary components, such as complex carbohydrates and fibers, which are abundant in the macaques' food sources in the mangrove habitat.
Tannerella forsythia, a well-known oral human pathogen 65 , was found to be more abundant in the Mff-KPE island population, as indicated by the LEfSe analysis.Periodontitis in humans is strongly associated with the presence of T. forsythia and this species has a significant role in the pathogenicity of the microbiota in subgingival plaques 66 .In the short-time observations during fecal specimen collection and our previous observations before the COVID-19 episode, the Mff-KPE island population was seen to be heavily provided with fresh and leftover foods by humans compared to the other three macaque populations.Thus, it is possible that their diet, which includes anthropogenic food, might have contributed to their higher abundance of T. forsythia in the gut microbiota.
Following the LEfSe analysis, the Mfa-PNY island population, which did not receive anthropogenic foods, showed an enrichment of Haloferula helveola and Bacteroides fluxus.Haloferula helveola is commonly associated with marine environments 67 , and is not known to inhabit the human gut in any marked abundance according to the data from the U.S. NIH Human Microbiome Project 68 and the search engine of EZBioCloud 69 .This is in accord with a previous report that indicated that marine invertebrates were the main food source of the Mfa-PNY island population 38 .Bacteroides fluxus has been isolated from the feces of healthy human individuals 70 .Nevertheless, one case of its presence in an abdominal infection has been reported 71 .Overall, the higher abundance of these bacterial species in the Mfa-PNY island population can be attributed to their specific diet, which includes marine-based foods, and their adaptation to a distinct island habitat, which likely influenced the composition of their gut microbiome.
According to the LEfSe analysis, the Mfa-MFRC mangrove population was enriched with bacterial species from the family Lachnospiraceae.These bacterial species are known to degrade complex polysaccharides, producing butyrate that can be utilized for energy 72 .This finding aligns with the dietary habits of herbivores, which are known to have a higher abundance of Lachnospiraceae compared to omnivores 73 .The results may reflect the plant-based dietary sources available to the Mfa-MFRC mangrove population in their habitat.In conclusion, this is the first report to compare the gut microbiomes of different subspecies of M. fascicularis (Mff and Mfa) living in two different habitat types (mangrove forest and island).The results revealed a significant difference in the gut microbiome associated with the different genetic background of the animals (between the two subspecies of M. fascicularis) and their diverse dietary habits (comparing between mangrove forest and island habitats, as well as anthropogenic foods).The latter factor could be associated with the use of stone-tools in foraging for foods.It was previously reported that the Mfa-PNY island population used percussive stone tools daily 24,32,38 , while the Mfa-MFRC mangrove population performed only food-pounding behaviors 30 .The food-pounding behavior is when the animals used the food (i.e., shell) to pound the food or to pound the stone, while the stone-tool use behavior is using the stone to pound the food, as seen in the Mfa-PNY macaques 30 .Furthermore, the study offered intriguing insights into the potential influences of stone-tool use and anthropogenic foods on the macaque's health, as evidenced through their gut microbiome.The higher gut bacterial diversity observed in the Mff populations, especially in Mff-KPE island population with access to anthropogenic foods and stone-tool use behavior, suggested that both diet and stone-tool use play significant roles in shaping the gut microbiome.In contrast, the reduced diversity in the Mfa-MFRC mangrove population that relies solely on natural food sources may reflect limitations in accessing a diverse range of microbial species.However, to comprehensively elucidate the influences of stone-tool use and diet acquisition on macaque health through the microbiome, further research is needed to investigate the individual-level relationship between stone-tool use, dietary habits, and gut microbial profiles.This is the next question for us to explore further.

Methods
Permit and ethical note.The permits for research and sample collection in the four populations of freeranging long-tailed macaques sampled in this study in Thailand were approved by the Department of National Parks, Wildlife, and Plant Conservation of Thailand.The Institutional Animal Care and Use Committee (IACUC) of the National Primate Research Center of Thailand-Chulalongkorn University approved the study's experimental protocols (Protocol Review no.2075007).The research adhered to the American Society of Primatologists (ASP) Principles for the Ethical Treatment of Non-Human Primates.All methods were performed in accordance with the relevant guidelines and regulations.
Study sites and consumed food items.Two subspecies of free-ranging Macaca fascicularis (Mff and Mfa) at two habitat types (island and mangrove forest), giving a total of four populations, in Thailand were selected for this study (Table 2).The subspecies were identified based on their geographical distribution and morphological characteristics 23,27,30 .The information regarding the food consumed by the monkeys was gathered through direct observation of foraging animals and their consumed foods, or by observing the remaining food item(s) after the animals had finished eating.Food items were identified and photographed using a Nikon COOLPIX W300 (Nikon, Japan).

Fecal specimen collection.
A total of 120 freshly defecated specimens (n = 30 for each population) were non-invasively collected using the fecal swab method in their natural habitats.In each location, the survey was conducted over at least five consecutive days, at 7:00-16:00 h (see Table 2).To avoid contamination with the soil microbiome, the fecal samples were collected from the inner part using cotton swabs (Citoswab, China).Samples were preserved in 2 mL of DNA/RNA shield (Zymo Research, USA) for viral inactivation and nucleic acid stabilization.To avoid double collection, the physical characteristics (i.e., color, texture, and shape) of each fecal specimen were recorded.DNA extraction.DNA was extracted using ZymoBIOMICS™ DNA Miniprep kit (Zymo Research, USA).
Briefly, 750 µL of fecal suspension were lysed in a ZR BashingBead™ lysis tube using TissueLyser LT (Qiagen, Germany) at 50 Hz for 3 min.The cell lysate was then extracted following the manufacturer's instruction.The concentration of DNA was determined using A 260 / 280 nm by NanoPhotometer® C40 (Implen, Germany).

PCR amplification and sequencing on MinION™.
The full length bacterial 16S small subunit ribosomal RNA (16S rRNA) gene, ca.1,500-bp size, was amplified based on PCR with the specific primers; 16S-V1F 5′-TTT CTG TTG GTG CTG ATA TTG CAG RGT TYG ATYMTGG CTC AG-3′ and 16S-V9R 5′-ACT TGC CTG TCG CTC TAT CTT CCG GYT ACC TTG TTA CGA CTT -3′ 74 .The 10 µL PCR reaction mixture consisted of 5 µL of 2 × UltraHiFi mix (Tiangen, China), 2 µL of PCR Enhancer (Tiangen, China), 0.25 µM each of forward and reverse primers, 1.5 µL of ddH 2 O, and 1 µL of the nucleic acid template.The PCR was thermal cycled at 94 °C www.nature.com/scientificreports/for 2 min, followed by 25 cycles of 98 °C for 10 s, 60 °C for 30 s, and 68 °C for 45 s, and then a final 68 °C for 5 min.The amplicons were barcoded by a five-cycle PCR using the barcode primers based on the PCR Barcoding Expansion 1-96 kit (EXP-PBC096; Oxford Nanopore Technologies, UK).The barcoded libraries were enriched using a QIAquick® PCR Purification kit (QIAGEN, Germany) following the manufacturer's instructions.The enriched libraries were quantified by Quant-iT™ dsDNA HS Assay kit using Qubit 4 fluorometer (Invitrogen, USA), and then equimolarly pooled for multiplexing.The pooled library was enriched using 0.5 × Agencourt AMPure XP beads (Beckman Coulter, USA).Afterwards, the library was subjected to end repair and adaptor ligation steps using Ligation Sequencing Kit (SQK-LSK114).Finally, the library was loaded onto the R10.4.1 flow cell and sequenced on a MinION™ Mk1C sequencer (Oxford Nanopore Technologies, UK).

Data analysis.
The FASTQ files were generated from the FAST5 data based on a super-accuracy model with a minimum acceptability quality score (Q > 10) using the Guppy basecaller software v6.0.7 (Oxford Nanopore Technologies, UK) 75 , while MinIONQC was used for the evaluation of the quality of the reads 76 .Porechop v0.2.4 was used for adaptor-trimming and demultiplexing of FASTQ sequences 77 .NanoCLUST was used for clustering, polishing, and taxonomically classifying the filtered reads, based on the size of the sequences for the V1-V9 region of 16S rRNA gene sequences from the Ribosomal Database Project (RDP) database 78,79 .The files were converted into QIIME (Quantitative insight into microbial ecology) format, and the QIIME2 toolkit v2021.2 was used for calculation of the alpha diversity using Chao1 and Shannon indices, and the beta diversity by Bray-Curtis cluster analysis 80 .The MicrobiomeAnalyst was used for the visualization of normalized data 81 .Finally, the Galaxy server was used for the differential abundance analysis of gut microbiota using linear discriminant analysis Effect Size (LEfSe) with P < 0.05 and a linear discriminant analysis (LDA) score > 2 39 .

Figure 2 .
Figure 2. Rarefaction analysis showing that an adequate sequencing depth was obtained for estimating the diversity of all the samples.

Figure 5 .
Figure 5. Differential abundance analysis by Linear discriminant analysis Effect Size (LEfSe) of the gut bacterial microbiome of Mff and Mfa living in a mangrove forest and on the island.The bar plots indicated the differentially abundant bacterial microbiota at different taxonomic ranks.The LDA score shows the effect size and ranking of each differentially abundant taxon (LDA score > 2, P < 0.05). https://doi.org/10.1038/s41598-023-42220-zwww.nature.com/scientificreports/

Table 2 .
Subspecies, code, location, geographical coordinate, habitat types, and date of specimen collection of the wild Mff and Mfa populations in this study.