Scale morphometry, geometry and ultrastructure of three Nemipterus species from the Egyptian part of the Red Sea

Fish scales, composed of ossified platelets, function as rigid external structures exhibiting complex morphological and structural characteristics. These features include scale type, shape, size, and specific surface ornamentations such as circuli, radii, lepidonts, and granules¹. Because scales grow proportionally with the fish, new circuli are added at the periphery and remain preserved throughout the lifespan of the scale^2,3. Scale shape is generally species-specific and has therefore been widely applied in elucidating fish systematics, determining stock structure, and assessing evolutionary relationships^4,5. Considerable variability in scale morphology has been documented among different fish species, reflecting both intraspecific morphological plasticity and regional variation within individual specimens^3,6. Such variation in scale shape may also contribute to reducing frictional drag during locomotion¹.

In addition to their protective role against external injuries, certain diseases and predation, scales can influence swimming performance. Surface features such as ctenii, circuli, radii, and scale curvature affect the flexibility and rigidity of the scales, thereby influencing the magnitude of force required for body bending during locomotion and ultimately shaping overall swimming efficiency^2,5,7.

Scale extraction is a non-lethal procedure that enables easy acquisition, preservation, and preparation for image-based analyses¹. Owing to their distinctive morphological and structural characteristics, scales provide valuable information across a wide range of scientific applications, including fisheries assessment, age and growth determination, taxonomy, evolutionary and phylogenetic studies, archaeology and environmental monitoring^{1,2,3,5,8,9,10,11,12,13,14}.

Numerous studies have applied a variety of methods to examine the morphometrics, geometric properties, morphological characteristics, and surface ornamentations of fish scales in both marine and freshwater species^{1,3,4,6,7,8,11,15,16,17,18,19,20}.

The family Nemipteridae (threadfin breams) comprises approximately 77 species across five genera²¹ and is characterized by small to moderate body size and vibrant coloration²². Members of this family are primarily distributed throughout the tropical and subtropical Indo-West Pacific but are absent from the eastern Pacific and Atlantic Oceans²³. The largest genus, Nemipterus²³, includes many species of economic importance as a food source²⁴. Extensive research on Nemipterus has addressed its biology^{24,25,26,27,28,29}, population dynamics and fisheries^30,31,32,33 as well as its biochemistry³⁴ species identification and genetic diversity^23,35,36,37.

Since the opening of the Suez Canal, the Mediterranean Sea has become connected to the Red Sea, facilitating a substantial influx of tropical fauna into the predominantly eastern Mediterranean basin. Numerous Indo-Pacific species (Lessepsian migrants) have consequently established populations and expanded their ranges within the region³⁸. Nemipterus japonicus and N. randalli have been documented in the Mediterranean Sea^39,40.

Threadfin breams constitute an important component of coastal demersal fish communities and regional fisheries^41,42. In recent years, they have emerged as a significant fishery resource in Egyptian waters, particularly in the Suez Gulf trawl fishery, where their catch now accounts for approximately 7% of the total trawl yield⁴³.

Despite the economic importance of Nemipterus species in Egypt, their scale characteristics have been relatively understudied. The present study aimed to investigate scale traits that can differentiate three nemipterid species; Nemipterus zysron, N. randalli, and N. japonicus from the Red Sea near Hurghada. Scales were examined using light and scanning electron microscopy, and univariate and multivariate analyses were applied to assess morphometric and geometric variations in size and shape.

Specimen collection

A total of 601 scales were collected from three Nemipterus species: N. zysron (12.2–18.5 cm standard length), N. randalli (13–18 cm standard length), and N. japonicus (8.5–17.2 cm standard length). Specimens were obtained from fish markets in Hurghada (27° 15′ N, 33° 48′ E; Red Sea, Egypt) between September 2021 and August 2022.

Preparation and measuring of scales

Scales were carefully extracted from the left side of the body, targeting seven specific anatomical regions as shown in Fig. 1a: (A) beneath the anterior portion of the dorsal fin (BDFS), (B) post-operculum (POS), (C) beneath the lateral line, between the pectoral and pelvic fins (BLLS), (D) caudal peduncle (CPS), (E) anterior lateral line (ALLS), (F) middle lateral line (MLLS), and (G) posterior lateral line (PLLS). Extracted scales were thoroughly cleaned in a 10% ammonia solution for 24–36 h to remove adhering tissue without compromising surface microstructure. Following cleaning, scales were dried on filter paper and mounted between two glass slides for storage.

Scales obtained from the first four regions were used to count primary (R1) and secondary (R2) radii and to obtain morphometric measurements, including scale width (W), caudal field length (L2), rostral field length (L1), and scale length (L) (Fig. 1b), for assessing stock-related differences among species. The following morphometric indices were subsequently calculated and evaluated: R1/W, W/L, W/L1, W/L2, L1/L, L2/L, and L1/L2. To assess variation associated with the lateral line canal, scales from the remaining regions were analyzed.

Scale landmark coordinations

In order to distinguish between distinct fish stocks, scales collected from the region beneath the pectoral fin of the Nemipterus species under study were examined using geomorphometric techniques. Eight homologous landmarks were digitized on consistent anatomical positions of each scale using tpsUtil and tpsDig2 software⁴⁴. These landmarks included the ventro- and dorso-lateral tips of the rostral margin (landmarks 1 and 2), the dorso- and ventro-rostral centers (landmarks 3 and 7), the dorsal and ventral tips of separation line (landmarks 4 and 6), the posterior tip of the scale (landmark 5) and the focal point of the scale (landmark 8) (Fig. 1b).

Microscopic study

Scales from the first four body regions were photographed using a stereomicroscope equipped with an AxioCam ERc 5s camera (Carl Zeiss–Promenade 10; 07745 Jena, Germany) and operated with Zeiss imaging software at the National Institute of Oceanography and Fisheries. Scales from all regions were subsequently examined using scanning electron microscopy (SEM) to assess their microstructural characteristics. For SEM preparation, cleaned and dried scales were mounted on specimen stubs using adhesive tape and coated with a 30-nm layer of gold. Electron micrographs were obtained using a JEOL JSM-5400LV SEM operated in backscattered electron mode at an accelerating voltage of 15 kV at the Assiut University Electron Microscope Center, Assiut, Egypt. Interpretation of SEM features followed the descriptions and criteria outlined in Hassanien⁴⁵, Sadeghi et al.¹³, Al Jufaili et al.¹, and Mekkawy, et al.³, the SEM results were interpreted and described.

Statistical analysis

IBM SPSS version 26⁴⁶ was used to compute descriptive statistics of the scale morphometric indices. A two-way analysis of variance (ANOVA; design: species + scale location + species × scale location) was performed to evaluate the effects of species and scale location. Post hoc comparisons were conducted using Tukey’s HSD test to assess interspecific variation among Nemipterus species. A multivariate analysis of variance (MANOVA) was also applied. In terms of homogeneity of variances, Levene’s test revealed significant findings for every variable.

A two-way permutational multivariate analysis of variance (PERMANOVA) was conducted using the PAST version 4.11 software⁴⁷ to further examine interspecific variation based on multivariate distances (Euclidean Index, 9,999 permutations). Furthermore, principal component analysis (PCA), discriminant function analysis (DFA), and cluster analysis based on Mahalanobis distances were performed Classification accuracy in DFA was assessed using leave-one-out cross-validation. Frequency distributions of radii counts across species and sampling locations were compared using Chi-square tests (Pearson Chi-square and likelihood ratio) to detect significant intra- and interspecific differences among Nemipterus stocks.

Geometric morphometric analyses were conducted using MORPHOJ version 1.07c⁴⁸. Generalized Procrustes Analysis (GPA) was used to standardize landmark configurations by removing variation due to size, position, and orientation, generating a set of shape variables⁴⁹. Multivariate analyses, including PCA and Canonical Variate Analysis (CVA), were performed to examine interspecific differences in shape. Deformed outline drawings were generated to depict the shape variations identified by the CVA, using the mean shape as a reference framework for comparative analysis. The Mahalanobis distance (P < 0.0001) were used to evaluate the significance of mean shape differences among species. To assess the effects of size and shape variation in fish scales, distinct multivariate regression analyses of Procrustes coordinates on centroid size were conducted following Cavalcanti, et al.⁵⁰.

The landmark coordinates for the three Nemipterus species were transformed into distance measurements using PAST software. Stock discrimination was further investigated using 28 landmark-transformed distances analyzed through DFA and Multivariate Analysis of Covariance (MANCOVA). These analyses were conducted using raw data, distance indices standardized by standard length (SL), and measurements corrected for Burnaby’s allometry and isometry, employing both PAST and SPSS packages.

Overall structural characteristics of scales

The scales of the studied species are predominantly ctenoid with a sectioned structure, characterized by well-defined radii. Simple scales, characterized by the absence or faint development of radii, were not observed. The scales exhibited variation in shape they were predominantly polygonal (pentagonal or hexagonal) or rectangular in form. Among the examined body regions, the largest scales were observed in the post-opercular (POS) region. Scales collected from the first three body regions were generally broad along the dorsoventral axis (Fig. 2a), whereas those from the caudal peduncle (CPS) were broader along the anteroposterior axis (Fig. 2b). Each scale comprised four distinct fields: dorsal, ventral, rostral (anterior), and caudal (posterior).

Surface ornamentation across the scale fields was primarily characterized by circular patterns created by alternating ridges (circuli) intersected by distinct, deep, and narrow grooves known as radii, that extended radially from the focus toward the anterior margin. Depending on their position and point of termination, the radii were classified into two categories: primary and secondary (Fig. 1b). This configuration was clearly visible in most fields, except in the caudal field, where circuli were absent and replaced by granulated structures (Fig. 2). The scale focus was situated posteriorly. The rostral margin of the scales exhibited a wavy and striated appearance, contrasting with the crescent-shaped contour of the caudal margin. The dorsal and ventral fields consistently displayed a convex form across all examined body regions. Regenerated scales were also observed in various body regions, distinguished by the lack of surface ornamentation, especially in the central portion of the scale (Fig. 2c).

Traditional morphometrics and meristics of scale

The number of R1 varied across different body regions, ranging from 4 to 10 in N. zysron and from 4 to 9 in both N. randalli and N. japonicus. In contrast, R2 counts ranged from 0 to 3 in N. zysron, 0 to 4 in N. randalli, and 0 to 1 in N. japonicus (Tables 1 and 2). A statistically significant correlation was observed between the number of R1 and body size in N. zysron (P < 0.01), suggesting that R1 development in this species is size-dependent. However, no significant correlation was observed between R1 count and body size in N. randalli and N. japonicus, indicating that this trait is not influenced by body size in these species. Conversely, a significant correlation between R2 count and fish size was found in N. japonicus, whereas no such relationship was detected in N. zysron or N. randalli (P > 0.01), implying that R2 count is independent of body size in the latter two species.

Descriptive statistics (mean ± SD) and occurrence percentages of R1 and R2 counts across the first four body regions of the examined Nemipterus species are summarized in Tables 1 and 2. The distribution of R1 differed significantly among body regions in all species (P < 0.001) and exhibited interspecific variation within each region (P < 0.035). The distribution of R2 varied significantly among body regions in N. zysron and N. randalli (P < 0.001), but not in N. japonicus (P > 0.206). Interspecific differences in R2 distribution were also detected across all examined regions (P < 0.033).

The basic statistical values of scale morphometric indices obtained from the first four body regions are shown in Table 3. ANOVA revealed that species, body region, and their interaction significantly influenced all scale morphometric indices (P < 0.001). The parameters W/L2, L1/L and L1/L2, in region BDFS, L2/L in region POS, L1/L, L1/L2, W/L, W/L1, and W/L2 in region BLLS, as well as W/L, W/L1, and W/L2 in region CPS exhibited the highest values for N. zysron. The greatest values for N. randalli were observed in region BDFS for W/L, W/L1 and L2/L; in region POS for W/L, W/L1, W/L2, L1/L and L1/L2; and in regions BLLS and CPS for L2/L. The highest values for N. japonicus were observed in W/L in region BDFS, L1/L and L1/L2 in region CPS, and R1/W across all regions.

All body regions under examination showed interspecific differences in scale indices, with the exception of the BDFS region with regard to the W/L2 index. Moreover, intraspecific variations in scale indices were observed in all species examined. The PERMANOVA results on the traditional scale indices across different body regions of the three Nemipterus species indicated no significant effect of species (P > 0.01), whereas scale location and the interaction between species and scale location exhibited highly significant effects (P < 0.0001).

Figure 3 illustrates the clustering of scale locations derived from the significant Mahalanobis distance matrix (P < 0.0001) of the examined scale indices. No distinct clustering pattern was observed in relation to species or body regions, as there was considerable overlap between the body regions of different species. Interspecific differences were evident, irrespective of scale location, as demonstrated by the Mahalanobis distances (P < 0.0001). N. japonicus formed a distinct cluster, whereas the remaining species showed no clear separation on the CVI (82.49%), with 50.9% of cross-validated grouped cases accurately classified.

Geometric morphometric analysis

Principal components analysis (PCA) of the Procrustes coordinates of scales located beneth the pectoral fin revealed interspecific variations among the Nemipterus species studied (Fig. 4a). Notably, N. zysron was positioned within the range of the other two overlapping species on PCI (29.5%). Canonical variate analysis (CVA) of these coordinates further demonstrated clearer separation of species with minimal interspecific overlap on CVI (65.08%) (Fig. 4b). The predicted group membership accuracy for the original cases of N. zysron, N. randalli, and N. japonicus was 96.7%, 93.3%, and 90.0%, respectively, with an overall mean accuracy of 93.3% for correctly classified original cases. In contrast, the average cross-validation accuracy for grouped cases was 80.0%, with correct classification rates of 76.7% for N. zysron, 83.3% for N. randalli, and 80.0% for N. japonicus. Furthermore, a statistically significant difference was observed among species based on the combined Procrustes coordinates after controlling for standard length (SL) (Wilks’ λ = 0.172, p < 0.0001, partial η² = 0.586).

The DFA of the regression residuals of the Procrustes coordinates revealed an interspecific pattern of variation consistent with the previous analysis (Fig. 4c). The predicted group membership accuracy for the original cases of N. zysron, N. randalli, and N. japonicus was 96.7%, 93.3%, and 90.0%, respectively, resulting in an average accuracy of 93.3% for correctly classified original cases. In contrast, the average accuracy for cross-validated grouped cases was also 93.3%, with correct classification rates of 96.7% for N. zysron, 93.3% for N. randalli, and 90.0% for N. japonicus. Furthermore, a statistically significant difference was found among species based on the combined regression residuals after controlling for centroid size (Wilks’ λ = 0.174, p < 0.0001, partial η² = 0.583).

The DFA of the principal components of regression residuals derived from the Procrustes coordinates exhibited an interspecific variation pattern that substantially coincided with the aforementioned patterns (Fig. 4d). The predicted group membership for the original cases of N. zysron, N. randalli, and N. japonicus is 96.7%, 93.3%, and 90.0%, respectively, resulting in a mean classification accuracy of 93.3% for the original grouped cases. An average of 81.1% of cross-validated grouped cases were correctly classified for N. zysron (80.0%), N. randalli (80.0%), and N. japonicus (83.3%). A statistically significant difference among species was observed in the combined principal components of regression residuals after adjusting for centroid size (Wilks’ A = 0.166, p < 0.0001, partial η2 = 0.592).

The DFA of landmark-transformed distances revealed interspecific variation among the studied Nemipterus species. Raw landmark distances (without normalization) showed species separation with some overlap (CVI = 64.84%) (Fig. 5a). When indices were scaled relative to standard length (SL), species clusters separate more distinctly, especially between N. japonicus and the other two species (CVI = 67.81%) (Fig. 5b). After removing allometric effects, the species clusters remained partially overlapping; however, N. japonicus tended to form a distinct group. In contrast, controlling for isometric size scaling resulted in a clearer separation of N. japonicus, while N. randalli and N. zysron remained partially overlapping (Fig. 5c, d). Removing allometric effects further enhanced separation among species (Fig. 5c). In contrast, controlling for isometric size scaling (i.e., analyzing shape variation independent of overall size) resulted in minor overlap between N. randalli and N. zysron, while N. japonicus remained distinct from N. zysron (Fig. 5d). Nenipterus species exhibited average correct classification rates of 91.1%, 91.1%, 90%, and 92.2% for raw distance, indices, isometry, and allometry transformed distances, respectively. Conversely, the species exhibited means of 76.7%, 73.3%, 75.6%, and 76.7% for correctly classified cross-validated grouped cases based on raw distance, indices, isometry, and allometry transformed distances, respectively. Clustering utilizing Mahalanobis distance revealed the subsequent relationship patterns among Nemipterus species: N. zysron + N. randalli vs. N. japonicus for raw distance, distance indices and isometry and allometry transformed distances.

In conclusion, significant interspecific variations in size, shape, and shape-size among the studied Nemipterus species were observed through both univariate and multivariate analyses of their traditional and landmark-based scale morphometric traits. Isometry and allometry transformed distances, revealed unique patterns of shape variation.

Scanning electron microscope studies (SEM)

Rostral field

The inter-radial tongues and the first circulus

Within the inter-radial space, the rostral margins of the scales exhibit tongue-like projections that lack circuli near the rim (Fig. 6). Two distinct forms of these projections, along with the first inter-radial circulus, are observed in the scales of the examined nemipterid species. The first form is characterized by conical tongue-like projections and a crescent-shaped first inter-radial circulus, which was identified in N. zysron (BDFS, POS, CPS, MLLS, and PLLS), N. randalli (BDFS), and N. japonicus (BDFS) (Fig. 6a). The second form exhibits conical tongue-like projections with a straight first inter-radial circulus (Fig. 6b), this form was recorded in N. zysron (BLLS and ALLS), N. randalli and N. japonicus (all regions except BDFS).

Radial grooves

The radial grooves of the scales of the nemipterid species under study are categorized into two distinct forms (Fig. 7): Form 1: wide and deep grooves with a thin membrane-like structure (Fig. 7a). This form was recorded in N. zysron (BDFS, BLLS & PLLS), N. randalli (BDFS, BLLS, ALLS, MLLS & PLLS) and N. japonicus (all regions). Form 2: narrow and deep grooves with a thin membrane-like structure (Fig. 7b). This form was recorded in N. zysron (POS, CPS, ALLS & MLLS) and N. randalli (POS & CPS). In both types, the ridges extend almost to the edge of the groove.

The inter-radial circuli, grooves and denticles

The inter-circular grooves in the rostral field were narrow and deep. The circuli exhibited small denticles within the inter-radial space. Five distinct forms of denticles were recorded on the dorsal surfaces of the inter-radial circuli (Fig. 8): Form 1: The denticles are conical in shape, featuring recurved, hook-like teeth (Fig. 8a) recorded in N. zysron (BDFS, BLLS & CPS). Form 2: blunt denticles are (Fig. 8b) recorded in N. zysron (POS) and N. randalli (BDFS). Form 3: conical denticles are with unicuspid ends (Fig. 8c) recorded in N. zysron (ALLS, MLLS & PLLS), N. randalli (POS, ALLS, MLLS & PLLS) and N. japonicus (POS, CPS ALLS, MLLS & PLLS). Form 4: the denticles are irregular shape (Fig. 8d) recorded in N. randalli (CPS & BLLS). Form 5: the denticles are conical with bicuspid and unicuspid ends (Fig. 8e) recorded in N. japonicus (BDFS & BLLS).

Outer lateral circuli, grooves and denticles

In the studied Nemipterus species, the outermost lateral circuli display thin, flat, and wide grooves. These circuli either bear denticles (Fig. 9a), as observed in N. zysron (BDFS, POS, CPS, MLLS & PLLS) and N. randalli (BDFS, POS & BLLS), or are free of denticles (Fig. 9b) as documented in N. zysron (BLLS & ALLS), N. randalli (CPS, ALLS, MLLS & PLLS), and N. japonicus (all regions).

Inner lateral circuli, grooves and denticles

In the three studied Nemipterus species, the inner lateral circuli display two types: either thin, bearing numerous denticles with flat, shallow, and wide grooves (Fig. 10a), observed in N. zysron and N. randalli (all regions) and N. japonicus (BDFS, BLLS, CPS, ALLS, MLLS & PLLS), or thick, bearing few denticles with flat, deeper, and narrower grooves (Fig. 10b), documented in N. japonicus (POS).

The focus region

The focus region in the three Nemipterus species appears in two forms: it is either oval-shaped and surrounded by oval ridges (circuli), as observed in N. randalli (BDFS, POS & BLLS) and N. japonicus (POS & BLLS) (Fig. 11a) or round-shaped and surrounded by round ridges, as documented in N. zysron (all regions), N. randalli (CPS) and N. japonicus (BDFS & CPS) (Fig. 11b).

Caudal field

In all the studied species, the caudal field area is weakly delineated from the anterior rostral field by the separation line, and the posterior rim exhibits a curved shape, featuring ctenii (Fig. 12a). The granulation area in the caudal field is devoid of circuli; yet, it features fine segments and is permeated by many pores throughout all regions (Fig. 12b).

Based on the separation line, the shape of the caudal region and the posterior rim five distinct scale forms were identified in the three Nemipterus species studied (Fig. 13): Form 1: The separation line is almost straight and the posterior rim is irregular crescent resulting in an oval caudal area. Form 2: The separation line is convex and the posterior rim is V-shaped resulting in a nearly conical caudal area. Form 3: The separation line and the posterior rim are irregular resulting in an elongated irregular oval caudal area. Form 4: The separation line is irregular crescent and the posterior rim is rounded resulting in almost a parallelogram caudal area. Form 5: The separation line is almost straight and the posterior rim is rounded resulting in an oval caudal area.

Lateral line scales

In the studied species, the lateral line scales generally feature a wide tube (canal) that runs parallel to the anteroposterior axis of the scale in a straight manner, while some other tubes are oriented obliquely to this axis (Fig. 14a). This canal exhibits a posterior opening positioned medially within the focus region on the inner surface of the scale and located at a considerable distance from the posterior margin. The anterior opening is similarly situated away from the anterior margin. Additionally, lateral line canal pores are present on both sides of the canal (Fig. 14b).

Scale morphometrics, geomorphometrics, and microstructural analyses have been widely utilized in comparative, phylogenetic, and stock identification studies of fish species^{2,3,13,15,18,20,25,51,52}. The methodologies utilized by these authors often incorporate univariate and multivariate analyses^3,5,8,20, scanning electron microscopy (SEM)^1,3,8,13,18 and geomorphological analyses^2,16,17,20.

To date, only a limited number of studies have investigated the scale morphology of nemipterid species, collectively demonstrating that scale characteristics can serve as reliable indicators for distinguishing among Nemipterus stocks. Renjith, et al.⁵¹ analyzed morphological variation in the tenth lateral line scale of N. japonicus, N. bipunctatus and N. randalli using elliptical Fourier analysis, revealing distinct interspecific differences in scale shape and highlighting the taxonomic value of scale morphology within the genus. Devi, et al.⁷ further described the scales of N. japonicus as exhibiting clear morphological variations across three different body regions, while Ujjania and Jaiswar⁵³ compared variations in scale size between N. japonicus, N. bipunctatus and N. randalli, reinforcing the diagnostic potential of scale-based traits.

The predominant scale type among marine fishes is the elasmoid type, encompassing both cycloid and ctenoid scales⁷. In the present study, the scales of Nemipterus species were identified as ctenoid and sectioned in type, exhibiting distinct surface ornamentation. These observations corroborate the findings of Devi, et al.⁷, who characterized the scale morphology of several species including N. japonicus, and those of Lelli, et al.⁴¹, who examined N. randalli from the eastern Mediterranean Sea. In addition, our results are consistent with the general morphological patterns reported in various marine taxa^1,3,13,18.

According to Spinner, et al.⁵⁴, the evolution of ctenoid scales was likely driven by mechanical interactions with the surrounding environment, offering an adaptive advantage by reducing physical damage. This adaptation is particularly relevant for demersal species such as Nemipterus, which inhabit areas where contact with the sea bottom, rocks, and corals is frequent. Thus, the presence of ctenoid scales in Nemipterus may represent an evolutionary response to the mechanical constraints of a demersal lifestyle. Moreover, beyond their protective function, ctenoid scales also contribute to hydrodynamic efficiency. Complementary to this, Harabawy, et al.¹¹ reported that the epidermal surface of fishes, especially those inhabiting coral reefs, is exposed to frictional forces generated by water movement. By regulating turbulence within the boundary layer, ctenoid scales help maintain smooth water flow along the body surface, thereby reducing drag and enhancing swimming performance². Furthermore, some fish species exhibit both cycloid and ctenoid scales, reflecting variations in scale morphology and functional adaptations^9,13,15.

Scale shape is also thought to influence swimming performance. For instance, Ibañez, et al.⁵⁵ suggested that scales elongated along the anteroposterior axis may reduce thrust and water pressure generated during swimming. In contrast, scales that are comparatively wide along the dorsoventral axis and shorter in the anteroposterior axis are thought to be advantageous for a subcarangiform swimming mode. In the present study, scales of the latter form were predominantly observed in the first three body regions. However, due to the lack of detailed information regarding the swimming behavior and performance of nemipterid species, a potential association between scale shape and swimming mode cannot be formally evaluated. The caudal peduncle scales are clearly distinguishable from those of the other body regions with respect to their size (as indicated by the W/L ratio) and overall shape. The elongation of these scales along the anteroposterior axis suggests a functional role in enhancing swimming efficiency by modulating thrust and water pressure⁵⁵. These findings are consistent with observations reported for cichlid species from Lake Tanganyika².

The current study, along with various other investigations^3,12,45,56 supports the validity of scale radii-meristics distribution and morphometrics in the identification of fish species and stocks. The number and morphology of radii serve as valuable taxonomic characters for species identification¹. In the present study, both primary and secondary radii exhibited interspecific differences in the studied species. The primary radii varied intraspecifically among body regions in all species and was size-dependent only in N. zysron. Whereas, the secondary radii showed intraspecific variations in N. zysron and N. randalli, with size-dependence observed solely in N. japonicus. Few studies have investigated such variations^1,3,12,56,57. In addition, primary radii in Nemipterus species exhibited higher counts than secondary radii, a situation also recorded previously by different authors^3,45,58. According to Raffealla and Nath⁵⁹ the formation of primary, secondary, and tertiary radii is generally considered a growth-related phenomenon.

Moreover, it was suggested that the radii exhibited a weak influence from genetic factors, suggesting that other environmental or physiological factors may play a more significant role. In this regard, the enhanced nutritional conditions of the fish may correlate with an increased number of radii⁶⁰. Given that radii represent gaps in scale ossification that contribute to scale flexibility⁶¹ while maintaining protective functionality⁶², their variation could reflect adaptive responses to environmental conditions rather than strictly genetic determinants. However, due to the absence of data on body flexibility of Nemipterus, it is currently impossible to comment on such a relationship. In addition, Esmaeili and Gholami⁶¹ reported that the number of radii depends on the scale location on fish body. Our results revealed that the BDFS, POS, and CPS regions showed the greatest number of radii counts among the examined species. According to Wainwright and Lauder⁶², such elevated radii counts typically occur in body areas that are inherently curved, such as the dorsal region, or in those subjected to greater lateral bending during locomotion, such as the caudal peduncle and tail.

In the present study, both intraspecific and interspecific relationships were reflected through quantitative, size-independent morphometric indices of the scales. The obtained results are consistent with those of Ujjania and Jaiswar⁵³, who investigated N. japonicus and N. randalli. Furthermore, several studies have employed comparable morphometric indices of fish scales in other teleost species^{1,3,5,8,11,63,64}. The observed morphological differentiation among the studied species may be attributed to environmental, geographical, and biological variations influencing phenotypic expression⁵³. In contrast to the present findings, previous authors did not address the variability of primary radii in relation to scale width across different species and sampling locations.

Numerous investigations have focused on the analysis of fish populations and species identification through the examination of scale geometry utilizing different scale landmarks in multivariate size and shape assessments^{2,3,4,17,19,20}. Also, the geometric analysis is harmless, rapid, and cost-effective than genetic analysis¹⁶.

Our results revealed clear species-specific scale shape differences. This was supported by high classification accuracies and statistical significance, demonstrating that the methods effectively discriminate species by scale shape variation. In addition, landmarks transformed distances, substantially improved species discrimination. N. japonicus consistently clusters apart, whereas N. zysron and N. randalli, though more similar, are better separated once size-related shape variation is accounted for. This highlights the essential role of morphometric data transformations for accurate species classification, especially in closely related taxa with subtle morphological differences. These results align with recent studies employing geometric morphometrics for species discrimination in other fish taxa^2,17,20,65.​.

Various studies have investigated the ultrastructure and surface ornamentation of scales in the rostral and caudal regions to distinguish between species^1,3,8,13,15. In the present study, the initial inter-radial circulus exhibited either a crescentic or straight configuration, a feature that may represent species-specific variation. Similar observations have been documented in various fish taxa^3,11,45,56. Correspondingly, Lippitsch⁶⁶ noted that the morphology of the first inter-radial circuli is characteristic within cichlid species, although it may occasionally be influenced by environmental conditions. Furthermore, the shape of the first circuli has been recognized as a potentially valuable taxonomic character in certain groups, such as goatfishes (Mullidae)¹⁵.

Previous studies have reported that denticles exhibit considerable variation in shape and size across different fish species^1,3,8,15,18. Variations in denticle morphology have been shown to depend on species identity, body size, and the specific location of the scales on the body^6,15,18. The present investigation did not reveal significant interspecific differences in denticle morphology among the three Nemipterus species examined. This finding aligns with the observations of Al Jufaili, et al.¹, who indicated that the taxonomic significance of denticles in Garra shamal remains uncertain, and with Gholami, et al.⁶⁷, who reported that the microstructure of scales, particularly with respect to denticles, is insufficient for distinguishing among aphaniid species but may assist in differentiating certain populations. Conversely, Ferrito, et al.⁶⁸ demonstrated that denticle morphology remains constant throughout the life of the fish and represents a reliable taxonomic feature for species identification. This is further supported by Lippitsch⁶⁶ who reported that denticle morphology may serve as a useful character in phylogenetic analyses.

Moreover, the presence of denticles on the inter-radial circuli and rostrolateral regions of scales in Nemipterus species suggests a functional role in reducing frictional forces through mechanical anchoring, a mechanism previously described in several other fish taxa^12,69,70. In addition, the circuli of the rostral field bear a higher density of denticles compared to those of the lateral fields, which likely enhances the attachment of scales to the underlying skin⁹. Yang, et al.⁷¹ proposed that the circuli and other microroughness features on the surface of fish scales serve to concentrate tensile stresses within the depressions between circuli during scale bending. By restricting tensile stress to these regions, the overall stress experienced during bending is reduced, thereby influencing the magnitude of force required to achieve a given level of deformation⁶².

The present study observed that the outer lateral circuli of N. japonicus, as well as the inner and outer lateral circuli of scales in specific locations of the other two Nemipterus species examined, lacked denticles. The absence of denticles in certain regions of the scales suggest that anchoring is not crucial in these regions, particularly in newly formed circuli³. These findings are consistent with the observations of Mekkawy, et al.¹², Mekkawy, et al.³ and Mahmoud, et al.⁷².

In the studied nemipterid species, the focus exhibits an oval or circular shape, encircled by elongated ridges that gradually diminish toward the caudal field, and is typically positioned in the posterior region of the scale. This observation aligns with the findings of Devi, et al.⁷ for N. japonicus. Although the position of the scale focus can vary among fish species¹⁸, it is generally considered stable once formed and does not shift throughout the lifespan of the individual⁷³. Teimori, et al.⁷⁴ further indicated that the shape and size of the focus are influenced by a combination of genetic factors, environmental conditions, and developmental processes. Variation in focus morphology has been widely reported among different fish taxa, including triangular, rectangular, circular, and oval forms^1,3,56,75.

The caudal region of Nemipterus scales lacks conventional microstructural features such as circuli and grooves; instead, it is distinguished by ornamentations composed of granulations (tubercles) and ctenii. Similar structural patterns have been documented in previous studies^3,7,15,45,51. Additionally, the caudal field of the scales in the studied nemipterid species contained multiple pores, which likely function as minute canaliculi facilitating mucus secretion. This mucus layer may play an essential role in wound healing and in reducing hydrodynamic drag at the fish water interface, thereby enhancing locomotory efficiency⁷⁶. Also, Teimori, et al.⁷⁷ suggested that modification in the ornamentation of the posterior region may have hydrodynamic significance, and such features are subject to modification throughout ontogenetic development.

Previous studies have demonstrated the taxonomic significance of lateral line scales in fish classification^51,78,79,80. Khalil, et al.⁸¹ reported substantial genus-level variability in the anterior–posterior placement of the lateral line tube in Brycinus nurse and Alestes baremose. In contrast, the Nemipterus species examined in the present study showed no distinct species-specific pattern in lateral line scale morphology.

In this investigation, the lateral line canal ran parallel to the anteroposterior axis of the scale, with posterior and anterior openings positioned medially within the focus region and separated from their respective scale margins. These observations correspond with the findings of Voronina and Hughes⁷⁸, who examined more than 1,000 teleost species, including N. zysron, and classified such scales as Unmodified Tubular-Scalar I. In this type, the canal tube forms as an elevated structure on a fully developed elasmoid scale, with the tube and scale following independent developmental pathways^82,83. The position of canal openings is considered an important diagnostic feature for differentiating taxa^6,11,79.

In general, Fish scale morphology has been shown to be most effective for distinguishing species within the same family that exhibit differing body morphologies, moderately effective for differentiation within fish families, and least effective for separating distantly related taxa¹⁷. Variations in scale shape and structure are influenced by multiple environmental factors, including temperature, water currents, population density, food availability, and other ecological conditions^59,84. In addition to environmental influences, genetic factors have also been highlighted as critical determinants of the taxonomic utility of scale characteristics⁸⁵.

This study demonstrates that scale morphology, together with traditional morphometric and geometric shape analyses, provides powerful tools for interspecific discrimination among Nemipterus species. The observed differences in scale characteristics, including surface ornamentation and structural features, underscore the value of scale-based assessments for species differentiation and stock identification. Collectively, these findings highlight the promising utility of scale morphology as a cost-effective approach for refining taxonomic classification and supporting conservation efforts. Nevertheless, classification accuracy may be enhanced by incorporating additional scale measurements and traits, such as scale thickness, flexibility, circuli counts, and lateral line canal patterns. Further comparative studies are required to validate the potential use of scale morphology and architectural design in species identification and phylogenetic applications within Nemipterus and related taxa.