The experimental protocol was reviewed and approved by the Walailak University Institutional Animal Care and Use Committee (WU-IACUC), Walailak University, Nakhon Si Thammarat, Thailand (Approval No. WU-ACUC-68002). All procedures involving NT were conducted in accordance with institutional guidelines for the care and use of animals in research and were designed to minimize pain, distress, and unnecessary handling. Fish used in this study were obtained from commercial farms during routine production and were handled only for blood collection. Before sampling, fish were anesthetized by immersion in a clove oil solution until loss of equilibrium and cessation of opercular movement, and sampling was performed promptly under humane conditions by trained personnel. During the procedure, fish were monitored for adequate anesthetic depth based on absence of response to external stimuli. No surgical intervention or experimental disease challenge was performed, and no fish were subjected to prolonged restraint or intentionally induced suffering. Blood collection was limited to the minimum volume required for hematological analysis, and all efforts were made to reduce stress associated with capture, handling, transport, and sampling. Therefore, the study complied with accepted ethical standards for the humane use of aquatic animals in research.

This study was conducted from June to August 2025 in Southern Thailand. Blood samples were collected from farmed sex-reversed NT raised in Nakhon Si Thammarat (8°37′ N, 99°53′ E; n = 31) and Surat Thani (9°5′ N, 99°12′ E; n = 15). Blood collection and anesthesia were performed at the Center for Scientific and Technological Equipment (B3), Walailak University (8.64′ N, 99.89′ E). Hematological analyses were conducted at the Laboratory of Veterinary Clinical Pathology, Akkhraratchakumari Veterinary College, Walailak University (8.64′ N, 99.89′ E).

Sex-reversed male NT with body weights of 500–1,000 g were obtained from commercial farms in Nakhon Si Thammarat (8°37′ N, 99°53′ E; n = 31) and Surat Thani (9°5′ N, 99°12′ E; n = 15), Southern Thailand, between June 2025 and August 2025 using convenience sampling. Farms used semi-intensive earthen pond systems with commercial pellet feed. Data on water quality parameters (temperature, pH, dissolved oxygen, ammonia, and nitrite) were unavailable. The fish were approximately 7 months old and considered market-ready adults.

Fish were anesthetized by immersion in a 30 ppm clove oil solution (prepared by diluting 6 mL of 10% clove oil stock in 20 L aerated water) until loss of equilibrium and cessation of opercular movement (typically 3–5 min). Blood (approximately 0.5–1.0 mL) was collected from the caudal vasculature using a 24-gauge needle attached to a 1 mL syringe. Blood was immediately transferred into tubes containing ethylenediaminetetraacetic acid (EDTA; Quetainer™, Cangzhou Fukang Medical Supplies, Hebei, China) at a final concentration of approximately 1.8 mg/mL.

EDTA was selected as the anticoagulant based on its common use in piscine hematology for cell counting; however, the literature indicates potential erythrocyte swelling with EDTA in some teleosts [32], and heparin may preserve morphology better in certain species.

Samples were grouped as: market-transported (Group I, n = 14), fresh-caught with 24-h iced storage (Group II, n = 15), and fresh-caught with immediate analysis (Group III, n = 17).

Blood smear preparation and staining: Three blood smears were prepared per fish using the push-pull technique on clean glass slides. Smears were air-dried and stained with Wright’s stain (Wright eosin methylene blue solution, Merck KGaA, Darmstadt, Germany) as follows: the dried smears were immersed with Wright’s stain for 6 min (this step also served as the fixation step [33]; no prior methanol fixation was required), after which buffer was added and allowed to stand for 8 min. Then, the stained blood smears were washed in running tap water.

Hb concentration was measured using an automated analyzer (URIT-300Plus, URIT Medical Electronic, Dhaka, Bangladesh) on whole blood. Other hematologic parameters were manually evaluated by veterinary technicians following standard methods [34–36]. PCV was determined by microhematocrit centrifugation (14,000 × g for 5 min).

Total RBC and WBC counts were performed manually after 1:200 dilution in Natt and Herrick’s solution (in-house preparation: 0.10 g methyl violet 2B, 3.88 g sodium chloride, 2.50 g sodium sulfate, 2.91 g disodium hydrogen phosphate dodecahydrate, 0.25 g potassium phosphate, 7.50 mL formalin 37%, distilled water to 1,000 mL; pH adjusted to 7.3) [37]. Counts were made in a Neubauer hemocytometer chamber (BOECO GermanyTM, Boeckel + CO, Hamburg, Germany) under 400× magnification.

WBC counts were corrected for thrombocytes using the formula:

Corrected WBC count (×10⁹/L) = Uncorrected WBC (×10⁹/L) × 100 / (Thrombocytes per 100 WBC + 100)

Differential leukocyte counts and thrombocyte enumeration were performed on 100 leukocytes per smear at 1,000× magnification. All counts were conducted by two trained veterinary technicians blinded to group assignment.

Absolute counts of leukocyte types were calculated as:

Absolute number (×10⁹/L) = Relative percentage × Corrected WBC count (×10⁹/L) / 100

RBC indices were computed as follows:

MCV (fL) = PCV (%) × 10 / RBC (×10⁶/µL)

MCH (pg) = Hb (g/dL) × 10 / RBC (×10⁶/µL)

MCHC (g/dL) = Hb (g/dL) × 100 / PCV (%)

Smears were screened at 400× magnification for 100 fields to detect morphological abnormalities, parasites, or artifacts, followed by detailed examination at 1,000× magnification for another 100 fields. Only smears from Group III fish without blast cells, apoptotic cells, or significant karyorrhexis were included for morphology and morphometric analysis, and for calculation of preliminary hematological values. Images were captured using an Olympus BX43 microscope equipped with a DP27 camera and CellSens software (version 1.18, Olympus, Tokyo, Japan). Morphometry was performed on 100 randomly selected cells per type (except basophils, which were limited to 10 cells due to rarity) in the monolayer area. Cell length, width, area (erythrocytes), or diameter (leukocytes) were measured using calibrated software tools.

Data were analyzed using descriptive statistics (mean ± standard deviation, minimum–maximum, and median). Histograms were generated using the ggplot2 package in R version 4.5.2 [38] to visualize distributions. Given the small final sample size (n = 14), formal reference intervals were not calculated according to the American Society for Veterinary Clinical Pathology recommendations (ideally n ≥ 20–40 for non-parametric methods; n ≥ 120 preferred for robust estimation) [39]. Outliers were screened visually using the boxplot package in R version 4.5.2 [38] but were retained unless biologically implausible. No inferential statistics were applied between groups due to exclusion criteria and sample size limitations.

Blood smear examination of 46 sex-reversed male NT revealed blast cells/transformed mature cells (Figure 1A) and apoptotic cells (Figure 1B) in all samples from Group I (market-transported fishes) and Group II (fresh-caught with prolonged-storage), respectively. Additionally, all blood smears from Group II contained several basket cells (Figure 1C). Thus, these 29 samples were excluded from hematologic analysis and morphological studies.

The main findings of this study were observed in Group III (fresh-caught with immediate hematological analysis). In this group, two samples showed hemolytic plasma and one sample revealed an intracytoplasmic inclusion body in a monocyte (Figure 1D). This structure morphologically resembled morulae of bacteria belonging to the order Rickettsiales, family Anaplasmataceae. However, diagnostic confirmation was not performed.

Within the remaining 14 fish from Group III, without clinical signs or external lesions, examination of blood smears revealed the morphology of RBCs, WBCs, and thrombocytes. Notably, some fish contained a few karyorrhectic cells and immature cells.

The types of WBCs identified in NT raised in Southern Thailand included neutrophils, basophils, eosinophils, lymphocytes, and monocytes (Figures 2A–M).

Neutrophils were round cells with abundant clear cytoplasm (with some slightly light basophilic areas) and an eccentric round to oval nucleus, occasionally exhibiting a bi-lobed shape (Figure 2C). Granulation of neutrophils was not visible. Basophils were small, round cells with densely packed basophilic granules. The nucleus was round (Figures 2D and F), bi-lobed (Figure 2E), and usually obscured by these granules (Figure 2F), although it was occasionally clearly visible.

Lymphocytes were small, round cells with a narrow rim of light-blue cytoplasm. The nucleus was round (Figure 2G) to indented (Figure 2H) and eccentrically positioned. Occasionally, cytoplasmic blebs were observed (Figure 2I). Lymphocytes were the smallest WBCs, with a mean diameter of 6.89 ± 0.99 µm.

Monocytes were the largest cells, with a mean diameter of 15.56 ± 1.8 µm. They were round, oval, or irregular in shape and had abundant deep gray-blue cytoplasm, often containing vacuoles. The nucleus was round to oval (Figure 2J), kidney-shaped (Figure 2K), or irregularly shaped (Figure 2L).

Eosinophils, rarely found in NT, were round cells with a round to oval nucleus. Their cytoplasm contained large, round to oval eosinophilic granules (Figure 2M). Eosinophils were not found in the monolayer area and were detected only after examination of the entire smear.

Mature erythrocytes of varying sizes (Figure 2N) were ellipsoidal, nucleated cells with smooth eosinophilic cytoplasm and a centrally positioned oval nucleus. The mean cell length was 10.35 ± 0.85 µm, and the mean width was 7.03 ± 0.60 µm (Table 1). Immature erythrocytes with polychromatophilic cytoplasm were also found (Figure 2O).

Thrombocytes were small, round (Figure 2P) or oval-shaped cells (Figure 2Q) with a condensed nucleus and clear cytoplasm that sometimes contained vacuoles. Occasionally, deep gray-blue cytoplasm was observed (Figure 2R).

Table 1 presents the morphometric measurements of blood cells in farmed sex-reversed male NT from Southern Thailand.

Individual hematologic values and their ranges are presented in Table 2, with distributions illustrated by histograms (Figure 3).

Farmed NT in Southern Thailand had PCV ranging from 0.23 to 0.45 L/L. Hb concentration ranged from 76 to 119 g/L. The total RBC count ranged from 0.94 to 2.72 × 10¹²/L.

Leucogram analysis showed that total WBC counts ranged from 4.17 to 23.01 × 10⁹/L. Lymphocytes were the predominant WBC type, with absolute numbers ranging from 3.00 to 10.35 × 10⁹/L, followed by neutrophils ranging from 0.71 to 9.66 × 10⁹/L. Monocytes ranged from 0.30 to 2.99 × 10⁹/L, whereas eosinophils and basophils were rare.

Blood cells identified in farmed NT included RBCs, WBCs, and thrombocytes, and were similar to those reported in other teleost species [27, 36]. WBCs were further classified into neutrophils, basophils, eosinophils, lymphocytes, and monocytes. Although heterophils have been reported in some teleost species, such as the armored catfish (Hoplosternum littorale) [40], these cells were not observed in NT from Southern Thailand. The authors suggested that the images of all blood cell types obtained in this study could be used to train artificial intelligence models, which might be integrated into automated hematology analyzers to improve the accuracy and precision of piscine hematology analysis, reduce time consumption, and support the implementation of CBC in the aquaculture industry.

Neutrophils in teleost fish act as primary phagocytic cells, migrate to inflammatory sites, and are able to eliminate pathogens through complementary mechanisms [41, 42]. In NT, neutrophils were positive for periodic acid–Schiff (PAS) reaction [43], suggesting that glycogen serves as an important energy source for the phagocytic process. Additionally, NT neutrophils were positive for myeloperoxidase (MPO) staining [43], indicating that this enzyme is likely involved in neutrophil bactericidal activity. Lymphocytes were the most common blood cells in NT, consistent with findings in other teleost species [27, 36, 44].

This study provided the first description of the characteristics of granules in both basophils and eosinophils in Thai NT using Wright’s staining. Despite the use of different staining methods, the eosinophil granule outline observed in this study was similar to that reported previously [43].

Monocytes were present in low numbers in the circulation, accounting for 2%–13% of leukocytes (Table 2). These cells were phagocytic and were considered transient blood cells, as they migrate into connective tissues during inflammation and differentiate into macrophages [45].

Thrombocytes in fishes show a variety of appearances, including round, oval, spike-shaped, fusiform, or elongated forms [10]. Piscine thrombocytes possess both coagulation functions and phagocytic activity [45].

The PCV of Thai NT ranged from 23% to 45%, which is higher than the threshold indicating anemia in fish (20%) [34]. Campbell and Grant [34] explained that the normal PCV of many fish species ranges from 20% to 45%, that PCV values greater than 45% are considered indicative of dehydration, and that PCV values lower than 20% are considered indicative of anemia. The PCV may be influenced by intrinsic and extrinsic factors, and normal PCV values in some fish species may be as low as 20%.

The PCV in this study was slightly higher than that reported in Egypt [46], where NT were raised on various dietary ingredients. Another study from Egypt reported PCV values in NT fed diets supplemented with coconut oil, which fell within our preliminary hematologic values [47]. However, both studies [46, 47] reported higher RBC counts than those observed in this study (1.37 ± 0.45 × 10¹²/L).

Total WBC counts in NT (12,860 ± 4,400 cells/µL) were lower than the reference interval for tilapia (21,600–154,700 cells/µL) [31], which was established from mixed groups of fish with different blood collection methods, seasons, ages, and sexes. It is noted that the preliminary hematological values in this study may be significantly limited by the small sample size. Thus, further studies with larger sample sizes are required.

Furthermore, recent studies indicate that hematological parameters in NT are influenced by nutritional and immune-stimulatory factors [48], suggesting that variations may reflect dietary effects rather than solely pre-analytical or technical factors, highlighting the importance of feeding regimes in hematological assessment of farmed fish.

The diluent, Natt and Herrick’s solution, used in this study, is commonly used in piscine hematology [31, 36] as well as in avian hematology [49, 50]. Thus, the lower values observed may have been caused by differences in breed, age, nutrition, or environmental factors.

Corrected WBC counts were performed in this study because the appearance of small WBCs (lymphocytes) and thrombocytes on the hemocytometer was difficult to distinguish. Therefore, all non-RBC cells were counted, and subsequently, corrected WBC counts were calculated using the formula described for species with non-nucleated RBCs [51], with the nucleated RBC number replaced by the thrombocyte number.

Blood smears of fish in Group II had several karyorrhectic (apoptotic) cells and smudge (basket) cells (Figures 1B and 1C), suggesting blood cell damage. These artifacts may have been induced by prolonged-storage in EDTA, as described previously [52, 53]. Previous studies have also indicated erythrocyte swelling associated with EDTA in some teleost species [32], whereas heparin may better preserve cellular morphology in certain species. Thus, heparin should be considered the anticoagulant of choice for piscine blood analysis. To standardize and improve laboratory management in piscine hematology, further investigations into the effects of storage conditions are recommended.

Furthermore, several blast cells (Figure 1A) were found in the market-transported fish (Group I). The authors assumed that the increased number of blast cells or transformed mature cells may have been related to physiological changes occurring during transportation, during which several fish were kept in the same tank. Although NT subjected to hypoxic stress have been reported to show no significant changes in blood biochemistry [54], the lineage and underlying cause of these blast cells require confirmation.

Although blood parasites were not within the scope of this study, the authors observed an Anaplasmataceae-like organism. A previous study reported Anaplasmataceae infection in NT from Brazil [17]. This study may represent the first potential report of an emerging Anaplasmataceae-like organism in Thai NT, providing baseline information for future studies. The authors suggested that further investigations using both molecular techniques and transmission electron microscopy are required to confirm Anaplasmataceae infection in NT in Thailand.