Morphological studies help in understanding the relationship between physiological and biochemical functions and molecular mechanisms. The gastrointestinal tract (GIT) of fish has shown a remarkable diversity of morphological and functional traits, particularly in connection with unusual nutritional behaviors, body shape and sex (Díaz et al. 2008b, Kapoor et al. 1975). There have been many contributions on the morphology of the GIT structure of fish, including marine, freshwater and estuarine species (Murray et al. 1994, Park and Kim 2001). Furthermore, due to various structural differences in the GIT of fish, nutritional studies have been the focus of most research in the field. For instance, past research has tried to correlate morpho functional characters of the GIT of fish with their nutritional habits and behaviors, the results of which are important to commercial fish management.

The large yellow croaker, Larimichthys crocea (Richardson, 1846), known as drums or croakers in reference to the repetitive drumming echoes they make, belong to Sciaenidae (Ramcharitar et al. 2006). This important carnivorous marine fish species is cultivated in many Asian countries, particularly China (Feng and Cao 1979). The ontogenetic development of the digestive tract of L. crocea was briefly described by Mai et al. (2005). Young fish start out with a simple, indistinguishable GIT that changes during development according to the environmental conditions and prey availability (Govoni et al. 1986). Functional differences in the GIT of fish are reflected in their capacity to digest, absorb and assimilate chemical compounds and are also associated with the developmental situation (Dabrowski 1984). In aquaculture, studying the anatomical and histological characteristics of the GIT of fish may help to understand the development of pathological conditions, to devise nutritional formulations and promote the management of L. crocea stocks in aquaculture. The primary goal of the present study was to describe the macroscopic and histological organization of the GIT of L. crocea.

Twelve mature L. crocea were obtained from the Marine Fisheries Research Institute of Zhejiang Province, Zhoushan, China. After being anesthetized with tricaine methanesulfonate, MS 222; Sigma Chemical Co. solution by using a dosage of 100 mg L⁻¹, the fish were euthanized via a longitudinal incision through their ventral side. The length of the intestine was measured to the nearest millimeter, using a tape measure and the intestinal coefficient (IC) was calculated using the equation IC = IL/SL (Angelescu and Gneri 1949). The segmental tissue fragments were removed after dissection and immediately processed for histological, histochemical and scanning electron microscopic studies.

Sliced tissue fragments from the separated GIT were fixed in 10% formalin (for 24 hours). After fixation, tissues were repeatedly washed in 70% ethanol and dehydrated via ethanol series. Then tissues were cleared in xylene and embedded using paraffin wax at 56–58 °C. Sections were sliced at 5 µm using a 820 HistoStat Reichert Rotatory Microtome (Reichert Technologies, NY, USA). After routine histological procedure, dewaxed sections were stained with haematoxylin – eosin (HE) solutions. Stained slides were examined and photographed under Olympus B202 microscope (Olympus Optical, Ltd, Tokyo, Japan).

The fragments were preserved and processed similarly to the procedure described above. Five to seven µm sections were prepared and stained in Periodic acid-Schiff (PAS), Alcian blue (AB pH 2.5) and PAS+AB pH 2.5 to identify the neutral and acidic mucins respectively.

The sample processing method for scanning electron microscopy is described elsewhere in detail (Kalhoro et al. 2017). In brief, tissues were fixed in 4% paraformaldehyde and 2.5% glutaraldehyde in phosphate buffer (pH 7.0) (for 24 hour at 4 °C). After fixation, tissues were post-fixed for 1–2 hours in 1% osmium tetroxide (pH 7.0) and desiccated in a graded ethanol sequence placed in a mixture of alcohol and isoamyl acetate (v: v = 1:1) about 30 minutes, then shifted to pure isoamyl acetate for overnight. Finally, tissues were washed and desiccated in Hitachi Model HCP – C critical-point dryer through liquid CO₂. Desiccated samples were coated with gold-palladium in Hitachi Model E-1010 ion sputter for 4–5 minutes and observed in SEM (TM-1000, Hitachi, Japan).

Morphological studies on the GIT of fish are useful when developing strategies for the management and conservation of fish species (Germano et al. 2014, Xiong et al. 2011). Our results revealed some peculiarities of the GIT of L. crocea, including the rather uniform structure of a shorter esophagus, well-developed stomach, pyloric appendages, and intestine.

In teleosts, the length of the intestine is correlated with nutritional habits (Canan et al. 2012, Machado et al. 2013). On the basis of the IC obtained (0.80), L. crocea can be categorized as omnivorous or carnivorous (0.2–2.5) (Bertin 1958), whereas the intestinal coefficient assortment for omnivorous (0.6–8.0) and herbivorous (0.8–15.0) teleosts is far greater (Al-Hussaini 1949, Bertin 1958, Xiong et al. 2011). However, some herbivorous teleosts have short intestines with less intestinal coefficients (Takesue 1954).Thus the intestinal coefficient might only be a criterion-referenced factor for classification of fish feeding habits.

Similar to other teleosts, the esophageal mucosa of L. crocea has squamous epithelium (frontward zone) and columnar epithelium (hindmost zone) (Carrassón et al. 2006, Cataldi et al. 1987). Frontward, the mucous cells of the epithelium contain sialosulphon glycoproteins. This secretion increases in the frontward zone and most likely protect the anterior region of the esophagus. The histochemical staining reacted positively, confirming the function of mucus in regulating lubrication (Oliveira Ribeiro and Fanta 2000). The lack of taste buds and lamina muscularis mucosae indicates that this organ has the function of pushing the swallowed micronutrients towards the gut (Ezeasor 1984). The submucosal part is a thick layer of connective tissue associated with a thick muscle layer, which might be compensate for the absence of a mucosal muscular layer (Albrecht et al. 2001). In fish, abundant mucous cells in the esophagus help food pass to the stomach (Cataldi et al. 1987, Grau et al. 1992); those cells are equivalent to esophageal glands found in the lamina propria or in the submucosa of higher vertebrates (Andrew and Hickman 1974). The formation of striated skeletal muscle fibers present in the esophageal wall found in L. crocea has also been observed in Stegastes fuscus (Cuvier, 1830), Pterodoras granulosus (Valenciennes, 1821) and Oligosarcus hepsetus (Cuvier, 1829) (Canan et al. 2012, Germano et al. 2014, Vieira-Lopes et al. 2013). It provides expeditious mobility with powerful contractions to clear the esophageal lumen (Chaves and Vazzoler 1984). As observed in our study, the presence of muscular layers in the esophagus (longitudinal inner sublayer and circular outer sub layer) and striated muscles that may possibly expand the luminal surface area of the esophagus, enable food intake and have been similarly observed in other Sciaenidae species such as Micropogonias furnieri (Desmarest, 1823) (Diaz et al. 2008a, b) and other teleosts; Hemisorubim platyrhynchos (Valenciennes, 1840) (Faccioli et al. 2014) and Odontesthes bonariensis (Valenciennes, 1835) (Diaz et al. 2006). In many teleostean families, the stomach is absent (Genten et al. 2009). However, it is developed in Sciaenidae (Andrade et al. 2017). The stomach of L. crocea has a U-shaped saccular structure, as reported by Pessoa et al. (2013) for Hypostomus pusarum (Starks, 1913) and by Andrade et al. (2017) in M. furnieri. A U-shaped stomach fundus characterizes a debris-eater (Moraes et al. 1997), while a Y-shaped, caecum sort of stomach is adapted to the assimilation of the entire prey (Rodrigues et al. 2008). The gastric glands are the main part of the gastric mucosa in vertebrates and teleost species (Al Abdulhadi 2005). The epithelium of L. crocea contains tubular gastric glands in cardiac and fundic regions, as reported in Sparus aurata Linnaeus, 1758, Pelteobagrus fulvidraco (Richardson, 1846), Glyptosternum maculatum (Regan, 1905) and Hemisorubim platyrhynchos (Valenciennes, 1840) (Cao and Wang 2009, Cataldi et al. 1987, Faccioli et al. 2014, Xiong et al. 2011). Our results showing an absence of gastric glands within the pyloric part of the stomach confirm the results reported by (Clarke and Witcomb 1980, Purushothaman et al. 2016). On the other hand, Arellano et al. (2001) reported that gastric glands were absent from the cardiac part of the stomach, while tubular glands within the pyloric part of the stomach were reported for a few teleosts by Al-Hussaini (1949). The gastric mucosa of L. crocea consists epithelial cells rich in neutral mucins similarly as reported for other teleosts (Cao and Wang 2009, Cataldi et al. 1987, Hernández García et al. 2001), formed by microvilli that may possibly serve in nutritive absorption (Grau et al. 1992, Ortiz-Delgado et al. 2003). On the other hand, this secretion from neutral mucous may protect epithelia from auto-digestion by hydrochloric acid and gastric glands secretion (Ferraris et al. 1987).

Pyloric caeca or tubular pouches vary in size and quantity according to the nutritional habits of teleosts (Canan et al. 2012, Purushothaman et al. 2016). Sixteen tubular pouches of almost similar sizes were observed in L. crocea. A different amount has been found in different teleost such as 10 in Leporinus friderici (Bloch, 1794), 4 tubular pouches in Acanthopagrus schlegelii (Bleeker, 1854) and 5 in Lates calcarifer (Bloch, 1790) respectively (Albrecht et al. 2001, Kalhoro et al. 2017, Purushothaman et al. 2016). It is predicted that the number of intestinal caeca and the length of the intestine are greater in herbivorous species, intermediate in omnivorous ones, and minimized in carnivorous fish, but is not always consistent (Alonso et al. 2015). From a histological categorization, the tunica mucosa has simple prismatic epithelia with mucus-secreting cells and longitudinal folded structure. The chief purpose of the tubular pouches is explained as “absorption” in contrast with mammals that function as fermentation. These are filled up as well as collapsed together by the anterior intestine, expanding the entire absorptive surface.

The intestinal mucosa is covered by columnar epithelia (enterocytes) and dominant mucus cell type called the goblet cell, like in other teleosts. Carrassón et al. (2006) stated that gel-forming mucin secretion produced by these goblet cells is involved in epithelial protection as well as in lubrication for nutrient pass way. The increased concentration of mucus-secreting cells appearing from the anterior to the posterior region followed by the rectum indicate that mucous production help to defend the lining of the intestine and support waste exclusion (Machado et al. 2013). The mucus-secreting goblet cells of L. crocea reacted positively to PAS (neutral mucins), specifying that these mucins might be involved in providing co-factors that are essential to nutrients enzymatic breakdown (Anderson 1986), and AB-positive reaction (acidic mucins) specified that these mucins shield the intestinal surface from glycosidase deprivation (Carrassón et al. 2006). As defined by Pelaseyed et al. (2014), mucus-secreting goblet cells are frequently dispersed among enterocytes. They have glycoproteins and are released by exocytosis of the intestinal epithelium, as observed in this study. Their abundance and secretive material are dependent on environmental factors (Rogers 1994). Various glycosaminoglycan or mucopolysaccharides eventually interact in the digestion process. In our study, the secretion of acid mucopolysaccharides suggest a secretory role of the intestinal epithelia. Moreover, these acid mucopolysaccharides are known to be sulfated peptic protease inhibitors that are useful in preventing pathogenic infections and providing protection from mechanical actions (Campbell 1999). Neutral mucopolysaccharides are associated with the function of digesting and emulsifying nutrients (Clarke and Witcomb 1980). Besides, it can be assumed that the chemical nature and the amount of mucus may influence the nature of the nutrients swallowed (Pelaseyed et al. 2014). This statement might support the increased concentration of goblet cells found from the anterior to the posterior intestinal regions in our study. Moreover, finding the intraepithelial lymphocytes (IEL) with the epithelial cells and mucus-producing cells indicate an active role in the teleost gut immune response. Particularly, they are involved in a regulatory function increasing the lymphocytic number in the intestinal mucosa, indicating the presence of a mucosal immune system (Miura et al. 2012).

On the basis of our results, the morphology of the GIT of L. crocea is consistent with omnivorous and carnivores feeding habits. Further detailed studies should be conducted on the ultrastructure level of the gastrointestinal tract of L. crocea.