All experimental procedures involving animals were reviewed and approved by the Animal Ethics Committee of the Indonesian Agency for Research and Innovation (BRIN), Bogor, Indonesia (Approval Number: 020/KE.02/SK/8/2022). The study was conducted in full compliance with national animal welfare regulations and internationally accepted ethical guidelines, including the ARRIVE guidelines 2.0 and the standards established by the World Organization for Animal Health (WOAH).

The study involved non-invasive sampling procedures using oral and cloacal swabs, which did not cause pain, injury, or long-term distress to the animals. Handling of T. haematodus was performed by trained personnel under the supervision of a licensed veterinarian to minimize stress and ensure animal welfare. No surgical procedures, anesthesia, or euthanasia were required during the study.

All birds were clinically healthy and sourced from a monitored captive population. Prior to inclusion, animals were assessed by a veterinarian to confirm the absence of disease or abnormal physiological conditions. Throughout the study period, birds were maintained under controlled environmental conditions and monitored daily for behavioral changes, physical abnormalities, and signs of distress. Any indication of compromised welfare would have resulted in immediate veterinary intervention; however, no such events occurred.

Housing, feeding, and environmental enrichment were provided in accordance with best practices for captive avian management. The study design adhered to the principles of the 3Rs (Replacement, Reduction, and Refinement), using the minimum number of animals required to achieve scientific objectives while ensuring high standards of care and minimizing animal use.

No endangered or protected individuals were harmed, and the study did not involve wild capture or invasive manipulation. All procedures were conducted in a manner that ensured the highest level of ethical responsibility, animal welfare, and scientific integrity.

The study was conducted from July to October 2024 at a controlled research facility under the Indonesian Agency for Research and Innovation (BRIN), Bogor, Indonesia. The experimental and sampling activities were performed during a continuous 14-day observation period under standardized environmental conditions.

This study employed a controlled experimental design involving six clinically healthy Trichoglossus haematodus, consisting of three adults and three juveniles. Birds were maintained under identical housing, environmental, and dietary conditions. Samples were collected from two anatomical sites (oral and cloacal) and categorized into four groups: adult oral (AO), adult cloaca (AC), juvenile oral (JO), and juvenile cloaca (JC) (Figure 1). The study focused on comparative microbiome characterization using full-length 16S rRNA sequencing and descriptive bioinformatic analysis.

A total of six T. haematodus (three adults and three juveniles) were housed individually in cages (70 cm × 43 cm × 52 cm) at a controlled temperature of 29.9 ± 1.1°C. During captivity, birds were provided with a diverse diet including fruits (apple, mango, papaya, banana, guava), seeds (fresh corn, sunflower seeds, cooked edamame), vegetables (bean sprouts, cucumber, long bean), and commercial baby biscuits soaked in brown sugar solution. Feeding was performed daily at 9:00 am, and water was provided ad libitum.

Feed intake for each bird was recorded daily over 14 days. The quantities of feed offered and remaining were weighed, and intake was calculated as the difference between these values. All birds were sourced from a monitored population and showed no clinical signs of disease. Additional diagnostic screening was deemed unnecessary by the attending veterinarian. Health monitoring included daily observation of morphology, behavior, and clinical signs of stress or disease. All animals remained under continuous veterinary supervision throughout the study.

Six T. haematodus (three adults and three juveniles) were maintained under identical environmental and dietary conditions. Oral and cloacal swab samples were collected from each bird using sterile cotton swabs approximately 2 h after feeding. A total of 12 samples were obtained, including three JO, three JC, three AO, and three AC samples.

Each swab was placed into a sterile tube containing 10% glycerol as a cryoprotectant and stored at −80°C for long-term preservation until DNA extraction.

Sex determination was performed using polymerase chain reaction (PCR) with universal primers 2550F (5’-GTTACTGATTCGTCTACGAGA-3’) and 2718R (5’-ATTGAAATGATCCAGTGCTTG-3’). These primers target the CHD gene, which exists in two forms: CHD-Z (Z chromosome) and CHD-W (W chromosome). Breast feather samples (plumae) were used for DNA extraction.

DNA extraction was performed individually for each sample before pooling. Extraction was carried out using the QIAamp Fast DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol with modifications, including treatment with lysozyme (25 mg/mL) (Calbiochem, Millipore, USA and Canada), RBC lysis buffer, and RNase (10 µL/mL) (Geneaid Biotech Ltd., Taipei, Taiwan).

This process yielded three JO, three JC, three AO, and three AC DNA extracts. DNA concentration and purity were measured using a nanophotometer (Implen, Munich, Germany). After quality assessment, equal amounts of DNA from individuals within each group were pooled, resulting in four composite samples: JO, JC, AO, and AC.

Sequencing libraries were prepared following standard protocols. Sequencing was performed using ONT with the PromethION platform. Raw signal data were acquired using MinKNOW software version 24.02.16 in FAST5 format. Basecalling was conducted using Dorado version 7.3.11 [18], generating FASTQ files under a high-accuracy model (Table 1).

Quality control was performed using NanoPlot [19, 20] to evaluate read length and quality distribution. Reads were filtered using NanoFilt version 1.8.0 (https://github.com/wdecoster/nanoflt), with a minimum read length of 1000 bp and minimum quality score threshold. High-quality reads were taxonomically classified using Centrifuge against the SILVA 16S rRNA reference database with a confidence threshold of 0.7 [21].

Taxonomic abundance was visualized using Pavian, and radial community structures were generated using KronaTools. Downstream analyses included alpha diversity (Observed operational taxonomic unit (OTUs), abundance-based coverage estimator (ACE), Simpson, and Fisher) and beta diversity (Venn diagram and principal coordinate analysis (PCoA)) using R software version 4.2.3 (https://www.R- project.org/). No inferential statistical tests were performed; all analyses were descriptive.

PCR products were analyzed using gel electrophoresis on 1% TBE agarose. Migration patterns of samples (lanes 1–4) were compared with a 1 Kb DNA ladder (lane M) to determine fragment size. The non-template control confirmed the absence of contamination (Figure 2).

This study used a bioinformatics-based metagenomic approach. Sequencing data were processed through ONT pipelines including MinKNOW, Dorado, NanoPlot, and NanoFilt version 1.8.0 (https://github.com/ wdecoster/nanoflt). Taxonomic classification was performed using Centrifuge with the SILVA 16S rRNA database, followed by downstream analysis using Pavian https://github.com/fbreitwieser/pavian), KronaTools (https:// github.com/marbl/Krona), and R software version 4.2.3 (https://www.R- project.org/).

Alpha diversity (Observed OTUs, ACE, Simpson, and Fisher) and beta diversity (Venn diagram and PCoA) were calculated using R Studio (version 4.2.3). No inferential statistical analysis was conducted, and all results are presented descriptively.

The biological data of the experimental subjects are presented in Table 2. The data include sex, age, body weight, and rearing period for both juvenile and adult groups.

The average daily feed intake data for juvenile and adult T. haematodus for all feed types are presented in Table 3. The results show variations in consumption depending on feed type and age group during the 14-day observation period.

The nutritional content of T. haematodus feed, including baby biscuit in nectar solution, is presented in Table 4. The table summarizes macronutrients, vitamins, and mineral contents per 100 g fresh weight.

The gut microbiota of T. haematodus was evaluated using cloacal and oral samples to assess age-related differences. A total of 1859 bacterial species were identified across all groups, indicating substantial microbial diversity influenced by both age and sampling site.

Alpha diversity of the gut microbiome in T. haematodus was measured using observed OTUs, ACE, Simpson, and Fisher indices (Figure 3). The results demonstrate variations in bacterial diversity across the four sample groups.

Observed OTUs showed that JC and AO had similar values (~980) and the highest richness, whereas AC and JO exhibited lower values (~800). The ACE index indicated higher richness in JC (1829) and AO (2131), while AC and JO showed lower richness (1283 and 1358, respectively). The Simpson index demonstrated that JC had the highest diversity value, whereas JO and AO had similar values (0.91), and AC showed the lowest diversity. The Fisher index indicated higher diversity in AO (153.00) and JC (1522.70), with AC and JO showing lower values (119.35 and 113.21, respectively).

Beta diversity of microbiota samples in T. haematodus is illustrated using Venn diagrams (Figure 4). A total of 106 OTUs were shared across all samples. Unique OTUs were observed as follows: AC (238), JC (351), JO (299), and AO (499).

JC and JO shared 246 core OTUs, whereas AC and AO shared 220 OTUs. JC exhibited higher unique OTUs than JO (734 vs 516), while AC had fewer unique OTUs than AO (580 vs 762). Furthermore, AC and JC shared 493 OTUs, and AO and JO shared 365 OTUs. AC had fewer unique OTUs than JC (307 vs 487), whereas AO had more unique OTUs than JO (617 vs 397).

The relative abundance of bacterial communities was analyzed at genus and species levels (Figures 5a and b). Microbial composition differed across AO, AC, JO, and JC groups.

At the genus level, Rosenbergiella was the dominant genus in AC (56.3%), followed by Klebsiella (15.1%) and Acinetobacter (15.6%). AO showed a more balanced profile, dominated by Alcaligenes (46.5%), followed by Neisseria (21.3%), Psittacicella (20.5%), and Uruburuella (10.8%).

JC exhibited a more complex microbial structure, dominated by Periweissella (27.3%), Rosenbergiella (24.8%), and Weissella (22.3%), with smaller contributions from Klebsiella (8.9%) and Acinetobacter (8.8%). In JO, Psittacicella (56.4%) was dominant, followed by Alcaligenes (23.8%), Neisseria (12.8%), and Uruburuella (6.3%).

At the species level, AC was dominated by Rosenbergiella epipactidis (73.0%), followed by Rosenbergiella spp. (23.1%) and Weissella confusa (3.2%). AO showed higher diversity, with Alcaligenes faecalis (40.3%) as dominant, followed by Psittacicella meloposticti (15.4%), Uruburuella suis (12.8%), Neisseria weaveri (11.5%), and P. hinzii (7.9%).

JC was dominated by W. confusa (31.7%), Periweissella fabaria (30.1%), and R. epipactidis (28.2%). JO showed higher complexity, with P. meloposticti (32.2%), P. hinzii (26.5%), A. faecalis (20.4%), and U. suis (7.5%).

The top 50 OTUs were visualized using a heatmap (Figure 6). Color intensity reflects microbial abundance, with lighter colors indicating higher abundance. Distinct microbial patterns were observed across age groups and sampling sites. R. epipactidis was most abundant in cloacal samples, particularly AC, whereas P. meloposticti dominated oral samples. W. confusa and P. fabaria were consistently observed across all groups.

The specific pathogenic bacteria taxa identified across all samples, along with their associated references are summarized in Table 5 [22, 24–30]. This table provides a detailed overview of these potential pathogens, highlighting their relevance to the health monitoring of captive birds.

PCoA analysis demonstrated distinct clustering patterns among sample groups (Figure 7). The first axis explained 79.64% of variance, while the second and third axes explained 11.21% and 9.16%, respectively. These results indicate clear separation among microbial communities, primarily driven by sampling site and age.

Based on the present results, the oral microbiota of adult T. haematodus (AO) exhibited higher species richness, as indicated by the ACE index, but showed lower evenness. This pattern may reflect the dominance of a few bacterial taxa, resulting in a less balanced community structure. This finding is consistent with observations in other vertebrates, in which older individuals often maintain a more stable but less diverse microbiota, possibly due to established ecological niches and age-associated phenotypes, including oxidative stress and proinflammatory responses [31].

Conversely, the oral microbiota of juvenile T. haematodus (JO) showed a more balanced microbial distribution, characterized by greater taxonomic diversity. Younger animals tend to harbor more varied microbiota and often show higher alpha diversity indices than adults [32, 33]. Studies on captive zebra finches and Bengalese finches have also indicated that juveniles have higher alpha diversity than adults. Bacteria in juveniles may represent early microbial colonizers adapting to the host because the immune system is not yet fully developed [34]. The presence of diverse bacteria in juvenile T. haematodus may reflect developmental processes in which a rich microbial community provides essential metabolic functions and improves resilience against pathogens [35].

The cloacal microbiota of adult (AC) and juvenile (JC) T. haematodus showed similar community structures, characterized by high evenness and a more balanced distribution of taxa. Compared with the oral microbiota, the cloacal microbiota appeared more stable across age groups. Microbiota diversity in the cloaca of parrots was higher than that in the oral cavity [36]. The higher diversity observed in the cloacal microbiota may be associated with feed breakdown into nutrients, which can influence the gut environment and provide substrates for bacterial growth.

The present results also indicated that the gut microbiome of juvenile birds remains strongly influenced by environmental factors and feeding behavior. The high alpha diversity observed in juvenile cloacal samples of T. haematodus suggests that juvenile birds have not yet achieved microbial stability in the digestive tract. Lower alpha diversity in adult cloacal samples indicates that microbial stability develops as birds mature.

Gastrointestinal tracts of juvenile birds are initially colonized by numerous transient bacterial species, which gradually develop into a more stable adult-like community [36]. In addition to bacteria, other studies have shown that juvenile birds harbor more transient viral species, probably acquired early in life, whereas adults exhibit a more stable viral community. These changes in microbial diversity reflect the influence of age on both bacterial and viral communities within avian hosts [37]. Meanwhile, the high alpha diversity observed in the oral microbiome of adult birds may indicate more varied feeding behavior in adults than in juveniles [38].

T. haematodus primarily consumes nectar and pollen from flowers. Based on Figure 5b, Rosenbergiella species, such as R. epipactidis, are Gram-negative, rod-shaped, and facultatively anaerobic bacteria. They have specialized growth characteristics, including temperature and salinity tolerance, that allow them to thrive in nectar habitats [39]. The microbial community of R. epipactidis and related species has evolved to survive in the unique chemical environment of floral nectar. Several factors affect their growth, including sugar concentration and nitrogen availability in nectar [40]. Although R. epipactidis is most commonly associated with floral nectar, bacteria perform critical functions in multiple ecosystems, including soil and aquatic environments. Their metabolic flexibility allows them to participate in biogeochemical cycles and adapt to diverse environments [41].

Rosenbergiella belongs to the family Enterobacteriaceae, which is commonly isolated from floral nectar. The high abundance of Rosenbergiella in cloacal samples of lorikeets may be associated with their role as pollinators or nectarivores [25]. In addition, the abundance of Enterobacteriaceae in lorikeets may indicate susceptibility to enteritis. Captive lorikeets are commonly fed commercial nectar powder diets that contain trypsin inhibitors. High levels of trypsin inhibitors in lorikeet diets may play a major role in susceptibility to enteritis [25]. Fortunately, in the present study, some taxa predictive of enteritis susceptibility, including Rhodococcus fascians, Kocuria spp., Pseudomonas umsongensis, and Aeromonas spp., were present at low abundance in cloacal samples of lorikeets.

Rosenbergiella is commonly detected in nectarivorous birds, such as hummingbirds [42]. However, it has rarely been identified as a specific bacterium in parrots. One study detected Rosenbergiella in a captive Alexandrine parakeet (Palaeornis eupatria), but did not specifically discuss Rosenbergiella [43].

Microbiome diversity is strongly influenced by environmental factors, including habitat and lifestyle. Birds with broader habitat ranges and more diverse diets typically have more diverse microbiomes [44]. The diversity of microbiota in birds differs substantially between the digestive tract and oral cavity because of differences in pH, enzymes, and organic substrates available in each habitat [45]. Dominant bacteria in the oral cavity of vertebrates often belong to Firmicutes and Bacteroidetes, whereas Proteobacteria are usually more common in the cloaca or digestive tract. These phyla are frequently found in avian gut ecosystems and play important roles in digestion and immune function. Differences in the dominance of these genera or species can be influenced by diet, lifestyle, and environmental conditions [46, 47].

The presence of Weissella in juvenile cloacal samples highlights the importance of microbial diversity in maintaining gut health and preventing pathogen colonization. Most Weissella species meet the prerequisites required to be categorized as probiotics [48]. In parrots, high abundance of Leuconostoc and Weissella and low levels of Clostridium and Escherichia/Shigella have been correlated with weight gain. Moreover, anorexia has been associated with higher abundance of Kocuria and Streptococcus [49].

The gut microbiome is crucial for avian health because it influences disease resistance and overall well-being. In chickens, modulation of the gut microbiota has been shown to alleviate infections, highlighting the potential role of microbiome management in disease control [22]. Microbiomes also contribute to ecological interactions, including pathogen transmission. Migratory birds can act as reservoirs for pathogens, posing risks to other species and humans [50, 51]. Similar to mammalian systems, the gut microbiota supports immune system function and nutrient absorption. It contributes to general physiological health and assists in the processing of pollutants [52]. The microbiome also protects against pathogens, thereby supporting disease resistance. However, wild birds can harbor antibiotic-resistant bacteria, creating a risk of disease transmission and resistance spread [53].

The gut microbiome of wild birds is highly diverse and dynamic, playing a crucial role in digestion, immune function, and overall health. Multiple factors, including diet, age, environmental conditions, antibiotic use, and pathogen exposure, can shape the composition of avian gut microbiota [54]. However, among the many microorganisms present in the intestine, some may act as enteropathogens and potentially cause gastrointestinal infections and other health issues [55]. These pathogenic bacteria include Salmonella, Escherichia, Clostridium, Pseudomonas, Staphylococcus, Klebsiella, and Campylobacter (Table 5).

Among these taxa, Klebsiella was the most abundant in the cloaca of both adult and juvenile birds. Wild birds are known carriers of Klebsiella, and microbial studies have confirmed its presence in their intestines. The prevalence of these bacteria in wild birds is influenced by diet, environmental exposure, and interactions with contaminated water, soil, or human-associated waste. In addition, Klebsiella forms part of the intestinal microbiota of humans and animals and has been found in several domestic and wild animal species, including mammals, fish, birds, mollusks, insects, and earthworms [27]. In the present study, the adult cloacal sample showed high prevalence of Klebsiella. In young birds, bacterial flora is still developing, and microbial diversity tends to increase as certain avian species mature [56].

Meanwhile, the prevalence of Salmonella, Escherichia, and Clostridium was higher in the cloaca of juvenile birds than in adults. The higher prevalence of Clostridium, Escherichia, and Salmonella in juvenile cloacal samples may be linked to environmental conditions, developmental processes, and ecological interactions. Juvenile birds have less developed immune systems than adults, making them more susceptible to pathogen colonization [57]. This immune immaturity can result in higher bacterial loads in the cloaca because juvenile birds are less capable of clearing infections.

Identification of Salmonella, Escherichia, and Klebsiella through complete 16S rRNA sequencing provides an important basis for health monitoring in captive populations. Although these taxonomic profiles detect bacterial DNA, they do not distinguish between commensal strains and strains carrying active virulence factors. However, the high abundance of Klebsiella in adult cloacal samples and the presence of Salmonella in juveniles are important ecological indicators of potential zoonotic risk and susceptibility to enteritis. This finding highlights the need for biosecurity and hygiene protocols for breeders in captive breeding facilities.

Furthermore, dietary changes that cause loss of host flexibility and metabolic shifts are frequently observed in captive breeding programs [58]. Captive birds can also spread antibiotic-resistant bacteria and zoonotic pathogens; therefore, mapping the gut microbiome of birds intended for reintroduction has become a useful tool for wildlife conservation and the One Health concept [59].

Several recent studies have demonstrated that captivity and environmental transition can substantially influence the prevalence and distribution of antibiotic resistance in wildlife. One study identified that rehabilitation and environmental changes in migratory birds were associated with increased antibiotic resistance, including the emergence of multidrug resistance in E. coli. Similarly, captive musk deer (Moschus berezovskii) showed higher diversity and abundance of antibiotic resistance genes than wild deer based on metagenomic and culture-based methods [60, 61].

Young birds may have different diets and foraging behaviors than adults, which can increase their exposure to contaminated feed sources. The cloacal microbiota of juvenile birds is often more diverse and less stable than that of adults. This diversity may include a higher prevalence of pathogenic bacteria, such as Escherichia and Salmonella, which may not persist in the microbiota of older birds because of the establishment of a more stable microbial community [23, 57].

Staphylococcus species are frequently present on the skin, feathers, and mucosal surfaces of wild birds, including the beak and oral cavity. These bacteria are natural components of the avian microbiota and can be isolated from various body regions. In a previous study, Staphylococcus aureus was recovered from the oral cavity of racing pigeons showing lesions characteristic of pigeon pox virus infection [26].

Commensal bacteria are integral components of the gut microbiome in wild birds and contribute to physiological functions such as digestion and immune modulation. Among these, Enterococcus species are frequently detected. A study analyzing fecal samples from wild birds reported a high occurrence of Enterococcus species, with Enterococcus faecalis being the most predominant, followed by E. hirae and E. faecium [30]. Lactococcus species have been shown to modulate the gut microbiota and produce beneficial effects for the host. In rats, Lactococcus petauri substantially modified the gut microbial community by enhancing short-chain fatty acid producers, which increased the production of total short-chain fatty acids, acetic acid, and propionic acid [29].

The separation along PCoA1 suggests that this axis represents a key ecological gradient influencing microbial community composition. Clustering of samples within the same group indicates that community structures are more similar within groups than between groups, suggesting that specific environmental or biological factors shape microbial composition. PCoA results in migratory birds indicated that the microbial structure of migratory birds was more heterogeneous than that of environmental samples [62]. Overall, this pattern suggests that the main factor distinguishing the groups plays a significant role in shaping community structure. Samples within the same group tended to cluster together because of shared characteristics, whereas clear separation between groups reflected distinct compositional differences.

There was clear separation of sample points, suggesting that the community composition of AC, AO, JC, and JO differed distinctly. Clustering of points within the same group indicated compositional similarity, whereas larger distances between groups suggested substantial differences. Separation along PCoA1 indicates that this axis represents the most influential factor differentiating the groups, which may include environmental conditions, species interactions, or external influences. Gut microbes can influence phenotypic traits of animals, whereas metabolite concentrations in the gut regulate microbial functions [43, 63]. In this study, dietary changes through pellet feeding and age directly affected the intestinal microbial community and metabolites in Alexandrine parrots [43].