Introduction

Otoliths (ear stones) are calcium carbonate accretions in the inner ear of fishes, and have proven useful in a variety of scientific contexts. Most otolith studies have been devoted to the saccular otolith (sagitta) because it is the largest of the three otolith pairs and displays traits that enable the identification of species and populations (Fig. 1). Saccular otoliths (termed otolith in the following) have served as informative characters in systematic studies of both fossil and extant fishes (e.g., Nolf 1985, 2013; Tuset et al. 2008, 2012; Lord et al. 2012; Gierl et al. 2013), and permitted inferences relating to physiological and ecological traits, such as habitat use (e.g., Lychakov and Rebane 2000; Jaramillo et al. 2014), sound reception capabilities (e.g., Popper and Lu 2000), age and growth rate (e.g., Campana and Neilson 1985; Green et al. 2009), as well as food preferences and feeding habits (e.g., Chancollon et al. 2006). They have also shed light on the timing of critical life history events such as settlement (e.g., Rehberg-Haas et al. 2012) and on migration strategies (e.g., Hermann et al. 2016). Furthermore, they have facilitated the interpretation of genetic hybrids, as in the case of cyprinodontiform fishes of the genus Aphanius (Masoudi et al. 2016).

Fig. 1
figure 1

a SEM image of the left otolith (sagitta) of a female A. farsicus (TL 32.6 mm) reared under laboratory conditions showing the otolith terminology used in this study. b Close-up of a showing the subdivision of the sulcus into ostium and cauda. c Close-up of b clearly reveals the sulcus pit marks

Full size image

Evidence of ontogenetic otolith variation has been provided in several studies (e.g., Lombarte and Lleonart 1993; Volpedo and Echeverría 1999; Morales-Nin 2000; Lombarte et al. 2003; Hüssy 2008; Capoccioni et al. 2011). According to these works it appears that both genetic and environmental factors impact the ontogenetic change of otolith size and shape. However, the ontogeny of the detailed otolith morphology, i.e. the ontogeny of rostrum, antirostrum or sulcus (see Fig. 1) has been the subject of very few investigations (see Volpedo and Echeverría 1999; Reichenbacher et al. 2009a; Kumar et al. 2012; de Carvalho et al. 2015). The results of these studies indicate that species-diagnostic otolith traits are fully developed only in sexually mature and adult individuals, and that otoliths from juvenile individuals are only of limited use or even not useful for taxonomy.

The genus Aphanius Nardo, 1827, to which our study refers, has been the subject of morphological and osteological studies, as well as molecular and genetic analyses of species and hybridization (e.g., Doadrio et al. 2002; Hrbek et al. 2006; Tigano et al. 2006; Ferrito et al. 2013; Esmaeili et al. 2014a, b; Masoudi et al. 2016). Otolith-based research on Aphanius has been used to identify species and populations, and to infer both population connectivity and past vicariance events (e.g., Reichenbacher et al. 2007, 2009b; Teimori et al. 2012a, b; Annabi et al. 2013; Gholami et al. 2015a, b). However, little is known about the ontogenetics of otolith variation within the genus, although such knowledge is critical for the identification of its constituent species and populations (see Reichenbacher et al. 2009b). The purpose of the present work was to elucidate the ontogenetic development of otolith morphology during the early life stages of a critically endangered endemic species, Aphanius farsicus (Teimori et al. 2011), which is restricted to a small locality in the endorheic Maharlu Lake Basin in southern Iran.

Material and methods

Sampling and preparation

The 34 specimens of A. farsicus used for this study (Table 1) were obtained from the Aquatic Animal Breeding Center of Shiraz University (ABCSU). They represent the F2 generation of a wild sample collected from the Barm-e Babonak site in the Maharlu Lake Basin (29°33′17.04“ N, 52°44’23.70” E) in South-western Iran in 2013 (see Sanjarani Vahed et al. 2017).

Table 1 Numbers of specimens of Aphanius farsicus studied at different development stages (indicated in days after hatching). Standard lengths of larvae (ranges) and otolith lengths (ranges) are also given
Full size table

Sampling and extraction of otoliths

Freshly hatched individuals were grouped into nine batches, and otoliths were prepared every 15 days from the time of hatching until maturity (120 days post hatching). Fish were anesthetized with an overdose of clove oil. After placing the samples in 70% ethanol for 5 min, each specimen was photographed, and its total length (TL) and standard length (SL) were measured using Image J Tools software (Rasband 1997–2016). For otolith analysis, fish skulls were opened dorsally, and right and left otoliths were removed. Otoliths were cleaned of organic remains by immersion in a 5% KOH solution for 5 min, and rinsed in distilled water for 10 min. Fishes were then fixed and preserved in 99.9% ethanol. SEM images of the saccular otoliths (sagitta) were prepared using a TESCAN vega3 instrument (Shiraz University). The general terminology used for otolith morphology in species of Aphanius follows previous work (e.g., Reichenbacher et al. 2007; Teimori et al. 2012a), except for the introduction of the new term ‘sulcus pit marks’ here (see Fig. 1c and below).

Results

Although all individuals were reared under the same conditions, the larvae and early juveniles showed considerable variation in growth rate, resulting in size (SL) overlaps between almost all age groups (Table 1). Furthermore, a significant positive correlation was found between the overall size of the larvae/early juveniles and the size of their otoliths (Fig. 2).

Fig. 2
figure 2

Linear relationship between length of larvae/early juveniles and otolith length (r2 = 0.907, p < 0.001)

Full size image

Representative examples of the specimens of A. farsicus studied and their sagittae (hereafter ‘otolith’) are presented in Figs. 3 and 4. The newly hatched embryos are 3.5–4.9 mm SL. At the time of hatching, the paired fin buds are still rudimentary; there are no rays in either the ventral- or dorsal-fin folds, while rays are present in the caudal fin (Fig. 3a1). The yolk sac has not yet been fully absorbed. During the first few hours after hatching, larvae remain on the substrate and do not swim. The otoliths at this point are almost round, with the ventral contour being more elongate than the dorsal (Fig. 3a2, b). A shallow depression is discernible on the inner surface, but no other indications of structure are yet visible.

Fig. 3
figure 3

Larvae (a1, c1), an early juvenile of uncertain gender (e1) and early juvenile females (g1, i1) of A. farsicus, their corresponding otoliths (left sagittae, a2, c2, e2, g2, i2) and additional otoliths (left sagittae) of early juveniles (b, f, uncertain gender; d, h, j, females). The different development stages are indicated in days after hatching. Otolith characters (arrows) appear successively over the course of early development. SL refers to the standard length of each individual depicted

Full size image
Fig. 4
figure 4

Early juvenile females of A. farsicus (a1, c1, e1, g1) and their otoliths (left sagittae a2, c2, e2, g2), together with additional otolith samples of early juvenile females (b, d, h) and an early juvenile male (f). The different development stages are indicated in days after hatching. An adult female individual (i1) and its otolith (left sagitta, i2) are also shown. SL refers to the standard length of each individual depicted

Full size image

At 15 days after hatching the larvae are 4.2–4.6 mm in SL. As yet, no rays are present in the paired and unpaired fins; only the caudal-fin rays are well developed (Fig. 3c1). The yolk sac is smaller than in the previous stage. In addition, sexual dimorphism is now recognisable: females already show the characteristic dense dark spot on the caudal fin base which is lacking in males. The otoliths are now ovate or round, and the first overt signs of structure can be seen, in the form of small sulcus pit marks (Fig. 3c2, d). The bulge at the anterior rim of the otolith shown in Fig. 3c2 is probably an aberrant feature, because we observed similar protuberances in some otoliths of mature individuals.

By 30 days after hatching, the sizes of the early juveniles range from 5.6 to 7.3 mm SL, the yolk sac has been completely resorbed and the mouth has opened. Paired and unpaired fins and fin rays have developed (Fig. 3e1), and the fish are now constantly active. In this growth stage, the sulcus first appears as a localized shallow depression in the middle of the otolith, but does not reveal a division into ostium and cauda (Fig. 3e2, f). Clear pit marks can be observed in the region which will later become the anterior part of the sulcus (see below). In addition, the otoliths still lack any indications of a rostrum, antirostrum or excisura. Only the otolith depicted in Figure 3e2 displays a slight projection at the ventral edge of the anterior rim of the future sulcus, which could be the first sign of a rostrum.

Between post-hatching days 45 and 60, early juveniles grow from 5.9–9.2 mm (SL) to 6.2–12.2 mm (SL). The paired and unpaired fins have increased in size, but rudiments of fin folds are still present (Fig. 3 g1, i1). These early juveniles swim actively and continuously. Otolith morphology develops markedly between the ages of 45 and 60 days. Otoliths from 45-day-old specimens exhibit an almost complete sulcus, but there is still no clear division into ostium and cauda (see Fig. 3 g2, h). The sulcus pit marks seem to be complete, as they cover almost its entire surface. Furthermore, a crista inferior, but only a very weak crista superior, is now recognizable. In addition, these otoliths show a broadly rounded, very short rostrum, an antirostrum of almost equal length, and a shallow excisura. In the 60-day-old, early juvenile otolith, its basic structural features are clearly discernible, but the rostrum and antirostrum are still very short (see Fig. 3i2, j). The sulcus is now divided into ostium and cauda, the crista superior is well developed and the deepening of the dorsal area is also recognizable. Of the four larvae studied at 60 days, three had otoliths that differed from those of the younger stages in possessing a straight ventral margin, while one otolith (Fig. 3j) still retained a rounded ventral contour.

Fish length continues to increase up to 120 days after hatching, but there is considerable variation in their growth rates and sizes (Table 1). They show almost the same morphology, with all fins (paired and unpaired) and fin rays well developed (Fig. 4a1, c1, e1, g1, i1). The pigmentation pattern becomes clearer as overall size increases. Females not only have a dense dark spot at the base of the caudal fin, but display small vertical bands on the body, which are the specific pattern elements characteristic of the females in this species. The otoliths of the 75-, 90-, 105- and 120-day-old specimens show considerable variation in the curvature of the ventral rim, as some have a straight ventral rim, while others have a curved ventral rim (e.g., compare Figs. 4e2 and 4f). In addition, some are relatively longer in the dorsal-ventral axis than others (e.g., Fig. 4c2 vs. Fig. 4d). All otoliths still have a very short, rounded rostrum, an antirostrum of almost equal length, and a shallow, wide excisura (see Fig. 4a–h). In this regard, all otoliths from the various developmental stages studied are clearly differentiated from otoliths of mature adults (1 year old or more) that were reared under the same laboratory conditions as the larvae and early juveniles (see Fig. 4i2).

Discussion

The natural habitats of Aphanius farsicus are springs and streams in the Maharlu Lake Basin (see Teimori et al. 2011). Individuals collected from their natural habitats have yielded otoliths with a sub-triangular to trapezoid shape, a relatively extended rostrum that is clearly longer than the antirostrum, a thickened and rounded to slightly pointed antirostrum and a deeply incised V- or U-shaped excisura (Gholami et al. 2015a). The same characters are present in otoliths of adult (mature) specimens of A. farsicus (1 year and older) that were reared under laboratory conditions (shown here in Figs. 1a, 4i2). As a result, the differences in otolith morphology between larvae/early juveniles and adults are unlikely to be attributable to environmental factors, but rather document a clear pattern of ontogenetic variation in the otoliths of A. farsicus.

Similar differences in otolith morphology between juvenile and adult individuals of Aphanius have previously been reported for A. ginaonis and A. dispar (Reichenbacher et al. 2009a). It is also known for other groups of fishes, i.e. Merluccius and Sciaenidae, and a relationship with habitat differences and/or behavioral disparities between young and adult individuals has been suggested (e.g., Lombarte and Lleonart 1993; Lombarte et al. 2003; Monteiro et al. 2005). In the case of Aphanius, a reduced rostrum length characterizes the larvae/early juveniles of A. farsicus (this study) and also those of A. dispar and A. ginaonis (Reichenbacher et al. 2009a). Furthermore, it has been observed that larvae/early juveniles of Aphanius thrive exclusively on the substrate or near the bottom of the water body, whereas larger specimens swim actively near the surface (Reichenbacher et al. 2009a; this study). Generally, an underdeveloped or lacking rostrum is characteristic for bottom-dwelling fishes, while a long rostrum is typical for pelagic fishes (e.g., Nolf 1993; Volpedo and Echeverría 2003). We therefore follow Reichenbacher et al. (2009a) in concluding that the short rostrum length and probably also the other ontogenetic differences in the otoliths of larvae/early juveniles of A. farsicus mainly result from differences in lifestyles, i.e. demersal in larvae and early juveniles, and pelagic in adults. Interestingly, de Carvalho et al. (2015) presented very similar results for the early ontogenetic development stage (up to 40 mm TL) of the anchovy Anchoa tricolor (Spix and Agassiz, 1829) from a subtropical estuary in Brazil. The dorsal and anal fins of the small anchovies were not yet completely developed (pointing to low swimming capability), and their otoliths were rounder and the rostrum smaller than in the later development stages (see de Carvalho et al. 2015).

A further interesting outcome of this study is that the appearance of the sulcus at the otolith’s medial surface and the emergence of the dorsal and anal fins are temporally and perhaps causally linked. The sulcus is the area where the sensory tissue comes into contact with the otolith, and it plays a role in both hearing and posture (e.g., Aguirre 2003; Schulz-Mirbach et al. 2011). It can thus be assumed that A. farsicus larvae do not hear well and their control of posture may be less well developed in the early post-hatching period, but that their sense of balance improves as soon as the dorsal and anal fins have formed and the larvae start to swim.

Moreover, our new data together with the results provided in Reichenbacher et al. (2009a) and de Carvalho et al. (2015) suggest that otoliths of larvae or early juvenile fish can be recognized as such not only because of their small size, but also based on their short and rounded rostrum and antirostrum and shallow, wide excisura. Such small otoliths are not diagnostic at the species level in A. farsicus, and this probably holds for other species of tooth-carps also. Furthermore, the same conclusion has been drawn for species of the Sciaenidae (Volpedo and Echeverría 1999; Kumar et al. 2012) and also for the above-mentioned anchovy (de Carvalho et al. 2015). In addition, the small otoliths studied here showed considerable variation in the curvature of the ventral rim and the ratio of length to height (see above and Figs. 3 and 4). This outcome of our study is also important for palaeontologists working with fossil killifish otoliths (or with otoliths of other fish groups), because it will facilitate the recognition of fossil otoliths from adult fish (which have species-diagnostic otolith morphology) and thus minimize the risk of overestimating species diversity.