Sagitta otolith of three demersal species in a tropical environment

The objective of this study was to describe the morphology and morphometry of saggitta otoliths of Polydactylus virginicus, Menticirrhus cuiaranensis and Conodon nobilis in a tropical environment. Fishes were caught with rod and reel in competitive fishing events promoted in 2014-2015 along the coast of Sergipe. A total of 174 pairs of sagitta otoliths of P. virginicus, 181 of M. cuiaranensis and 77 of C. nobilis was extracted. In general, the sagitta otoliths of all three species analyzed here presented different morphology and shape indices. The permutational multivariate analysis of variance (PERMANOVA) demonstrated significant differences among species and ontogenetic phases within each species using morphometry and shape indices. The linear discriminant analysis (LDA) presented a 98.3% correct reclassification of the otoliths by species.


INTRODUCTION
Otoliths are calcium carbonate precipitated primarily as aragonite structures, present in the inner ear of fishes in three pairs: sagittae, asterisci and lapilli (Popper et al., 2005). They are metabolically inert, i.e., there is no chemical alteration or reabsorption after their formation. These structures assist the hearing and perception of the external environment (Ladich & Schulz-Mirbach, 2016). Otoliths are excellent tools for various studies of fish populations such as: habitat use and connectivity between populations using the chemistry of the otolith Maciel et al., 2020), environmental stress indicated by the deposition of vaterita (Carvalho et al., 2019a;Holmberg et al., 2019), and age and growth (Soeth et al., 2018;Maciel et al., 2018).
Many studies correlate the ontogenetic developments of otoliths to life history and changes in habitat use as distribution in the water column (Cruz & Lombarte, 2004;Assis et al., 2020). Taylor et al. (2020) correlated changes in the morphology and morphometry of the otoliths of Argyrosomus japonicus with the segregation of habitat use of juveniles and adults. The morphology and morphometry of otoliths may present sexual variation, as observed in Atherinella brasiliensis (Quoy & Gaimard, 1825), Micromesistius australis (Norman, 1937) and Porichthys notatus (Girard, 1854) (Carvalho & Correa, 2014;Leguá et al., 2013;Bose et al., 2016), as well as before and after the first maturity within the same species . The intraspecific morphologies and morphometric patterns of otoliths make them an important tool for studies of trophic ecology of ichthyophagous species (Miotto et al., 2017;Carvalho et al., 2019b). Some of the species identified as preys are also exploited in artisanal and recreational fisheries (Cattani et al., 2011;Freire et al., 2017;Passarone et al., 2019). Polydactylus virginicus (Linnaeus, 1758), Menticirrhus cuiaranensis (Marceniuk et al., 2020) and Conodon nobilis (Linnaeus, 1758) are examples of species exploited by recreational fishers (Freire et al., 2017) and are usually described as food items in the diet of ichthyophagous species (Tavares & Di Beneditto, 2017). Those species are demersal and distributed in the Southwestern Atlantic Ocean (Menezes & Figueiredo, 1980;Froese & Pauly, 2021 Menticirrhus cuiaranensis and C. nobilis are benthophagous  whereas P. virginicus is zooplanktivore (Medeiros et al., 2017).
Several studies have described the morphology and morphometry of otoliths of species important for artisanal and recreational fisheries in southeastern-southern Brazil (Siliprandi et al., 2014;Brenha-Nunes et al., 2016;Maciel et al., 2019;Dal Negro et al., 2021). Conversely, few studies have been conducted in northeastern Brazil, making it impossible future studies of connectivity or fishing stocks for species with wide distribution along the Brazilian coast (Freire et al., 2017;Assis et al., 2020;Bot et al., 2020). Thus, this study was conducted with the objective of describing the morphology and morphometry of sagitta otoliths of P. virginicus, M. cuiaranensis and C. nobilis, and assessing the occurrence of ontogenetic changes to serve as basis for future studies related to diet composition and connectivity between populations.

MATERIAL AND METHODS
Samplings were performed between March 2014 and August 2015 in distinct beaches along the coast of the state of Sergipe, in northeastern Brazil (Fig. 1). Recreational fishers used rod, reel, and hook with natural bait during these events (Freire et al., 2017). The specimens caught were identified, measured for total length and total weight (TL, in centimeters and W, in grams, respectively). Power relations were also estimated between total length and total weight (W = a•TL b ) (positive allometry, b > 3; negative allometry, b < 3; and isometry, b = 3; Froese, 2006). The specimens were separated into six length classes (5.1-10; 10. 1-15; 15.1-20; 20.1-25; 25.1-30; 30.1-35 cm) to test for ontogenetic variation. Sagitta otoliths were washed, dried, and stored separately in small paper envelopes. Voucher specimens are deposited in the collection of the "Acervo Zoológico da Universidade Santa Cecília" under the numbers AZUSC-UNISANTA n. 5155, 5169, and 5935.

Morphology and morphometry of otolith
The identification of otolith shape, position, type and shape of the sulcus acusticus, shape of the cauda, and ostium type was performed according to Corrêa & Vianna (1992), Tuset et al. (2008), and Volpedo & Vaz-dos-Santos (2015). The otolith length (OL, greater longitudinal distance in mm) and height (OH, greater perpendicular distance in mm) were measured (Fig. 2). Power relations were also estimated between otolith length (OL) and height (OH) and fish total length (TL): OL = a•TL b and OH = a•TL b . The following otolith shape indices were used to verify the ontogenetic variation as well as variation among species: OL/TL, OH/OL × 100 (Volpedo & Echeverria, 2003), and ellipticity [E = (OL -OH)/(OL + OH)] (Tuset et al., 2003).

Statistical analysis
The morphometry and shape indices did not meet the assumptions required for parametric tests in all three species analyzed here (Shapiro-Wilk: p < 0.05; Bartlett's test: p < 0.05). Hence, a Permutational Analysis of Variance (PERMANOVA) was applied to verify ontogenetic differences within species and shape variation among species using the morphometry and shape indices for all three species. A linear discriminant analysis (LDA) was performed to verify the percentage of correct   Oliveira, C.S. et al.: Otolith of three demersal species Pap. Avulsos Zool., 2021; v.61: e20216154 2/7 reclassification of otoliths among the three species analyzed (Linde et al., 2004). All statistical analyses were performed using the R software (R Core Team, 2020) and Past (Hammer et al., 2001).

RESULTS
A total of 487 individuals of P. virginicus (TL = 7.2 to 27.5 cm; n = 229), M. cuiaranensis (TL = 12.5 to 35.5 cm; n = 182), and C. nobilis (TL = 7.4 to 33.5 cm, n = 76) were analyzed (Fig. 3). The morphological analysis of the otoliths indicated interspecific variation (Table 1), except for the shape of the sulcus acusticus and cauda, which are similar for all three species. The position and type of the sulcus acusticus are similar for P. virginicus and C. nobilis, i.e., median and ostial. Menticirrhus cuiaranensis showed on rostrum and P. virginicus and C. nobilis showed a small and underdeveloped rostrum.  Morphometry and indices demonstrated ontogenetic variation within the three species analyzed here (Fig. 4). The otoliths of C. nobilis tend to be more rounded along the ontogeny and the otoliths of M. cuiaranensis and P. virginicus tend to be more elongated as the ontogenetic development occurs (Fig. 4D). The otoliths of the three species show greater growth in the anteroposterior axis along the ontogeny (Fig. 4E). The results of the PERMANOVA (F = 807.1; p < 0.0001) indicated significant differences in the morphometry of otolith and shape indices among species.
Both P. virginicus and M. cuiaranensis presented positive body allometry when analyzing the weight-length relationship (b = 3.233 and 3.159, respectively; Fig. 5). On the other hand, C. nobilis showed body isometry (b = 3.022), with an increase in its body weight proportional to the cube of its length. At the same time, C. nobilis was the only species, among all analyzed, with otoliths presenting a strong positive allometry (b = 3.744; Fig. 5).
The LDA showed a variation in the shape of the sagitta otoliths: M. cuiaranensis presents otoliths distributed along axis 1 for presenting more elongated otoliths; and P. virginicus and C. nobilis are distributed along axis 2, which better explains the variability of the more rounded otoliths (Fig. 6). The LDA presented a 98.3% correct reclassification of the otoliths by species (Table 2).

DISCUSSION
In general, the sagitta otoliths of all three species analyzed here presented different morphology. However,   Avulsos Zool., 2021;v.61: e20216154 4/7 similarities were observed in the shape of the sulcus acusticus and cauda for all of them, and also in the position and type of sulcus acusticus for P. virginicus and C. nobilis. Even though P. virginicus, M. cuiaranensis and C. nobilis are demersal species, the otolith shape is different among them (rectangular, bullet shaped, and oval, respectively). For Gauldie (1988), it is the shape of the sulcus, and not the otolith shape, that is associated with acoustic sensitivity, which in turn depends on the habitat. However, this author did not present details on how this association works. In the present study, the shape of the sulcus acusticus was heterosulcoid for all three demersal species analyzed. The exact mechanism that relates the sulcus shape with higher acoustic sensitivity would have to be better investigated in future studies. Additionally, Cruz & Lombarte (2004) pointed out that Sciaenidae and Haemulidae families, which include M. cuiaranensis and C. nobilis, respectively, belong to groups specialized in sound production. This could be associated with relatively large otoliths for better acoustic communication.
Our study was the first to evaluate the ontogenetic variation of otoliths for P. virginicus, M. cuiaranensis and C. nobilis using shape indices. Only the OH/OL x100% demonstrated a change in trend associated with ontogenetic variation mainly for P. virginicus and C. nobilis. According to Volpedo & Echeverria (2003), values of aspect ratio between 21% and 96% and the absence of rostrum indicate species associated with a soft bottom. Conversely, species associated with a consolidated substrate show values between 41% and 67% and a small rostrum. The aspect ratio and the presence of a small rostrum indicate that P. virginicus and C. nobilis inhabit consolidated and soft substrates, respectively (Brenha-Nunes et al., 2016;Santificetur et al., 2017). The absence of rostrum indicates that M. cuiaranensis lives in environments of unconsolidated sediment as mentioned for other species of the same genus Menticirrhus . Polydactylus virginicus showed a reduction in the aspect ratio in specimens larger than 15 cm TL, when this species starts the maturation process (Freire et al., 2020). This may indicate a change in habitat after maturity between juveniles and adults, or in swimming capacity associated with feeding change. On the other hand, the aspect ratio of the otoliths from C. nobilis increased until specimens were 20 cm TL long and kept stable onwards. Note that the length at first maturity for this species is about 21 cm TL in a neighboring area (da Silva et al., 2019). Here, again, a change in habitat for juveniles and adults may be occurring. The otoliths of Menticirrhus are elongated (Siliprandi et al., 2014;Carvalho et al., 2020), as also observed for M. cuiaranensis in the present study, but no change was observed in the aspect ratio for the length range analyzed here.
According to Iizuka & Katayama (2008), demersal species present otoliths within two groups defined by their external shape: elongate (or elliptical) and orbicular (or ellipsoidal). This difference was evident for the three species analyzed here, with elliptical otoliths found in P. virginicus and M. cuiaranensis, and ellipsoidal in C. nobilis. Moreover, the former two species presented a positive allometry in the relationship between body weight and length, but negative when the same relationship was estimated for the otolith. On the other hand, C. nobilis presented body isometry, but a positive otolith allometry, which is related to the ellipsoidal shape of its sagitta otolith. Hence, otoliths of C. nobilis are disproportionately heavier than the other two species. One could hypothesize that C. nobilis presents higher acoustic sensitivity based on Lychakov & Rebane (2000), who stated that the heavier the otolith, the higher its acoustic sensitivity in lower frequency. This higher acoustic sensitivity should be further investigated in future studies involving C. nobilis.
The otolith shape is usually species-specific, but intraspecific geographic variation may occur due to environmental factors Maciel et al., 2020). The morphology and morphometry of the otoliths from P. virginicus and C. nobilis were analyzed in previous studies carried on southeastern-southern Brazil  Oliveira, C.S. et al.: Otolith of three demersal species Pap. Avulsos Zool., 2021;v.61: e20216154 5/7 acteristics to specimens inhabiting northeastern Brazil. The applied methodology was not the most sensitive to determine contour variations that may show latitudinal influence. Several external factors can influence the otolith shape (Torres et al., 2000;Capoccioni et al., 2011;Avigliano et al., 2012), including beach hydrodynamics. However, no study has been conducted in Brazil correlating these two variables. Could the differences observed in the otoliths of the three species represent a response to the beach dynamics? Future studies of the morphology and morphometry of the otoliths of such demersal coastal species should try to correlate the beach hydrodynamics with the otolith shape.

ACKNOWLEDGMENTS
All authors would like to thank the Universidade Federal de Sergipe for the scholarship provided to the first author (COPES/UFS), to all recreational fishers from the Associação Sergipana de Pesca Amadora Bons Ventos (ASPA-BV) for donating their catch from fishing competitive events to be analyzed in this study, to Rodrigo Melins, Rafael Barbosa, Ana Cláudia dos Santos, and Osmir Fabiano for helping with sample processing, and to Daniel Assis and Roberto Schwarz for their comments on an earlier version of this manuscript. B.M.C. is thankful to the National Council for Scientific and Technological Development (CNPq 153090/2019-7).

AUTHORS' CONTRIBUTIONS
C.S.O.: Collection, data collection, data analysis and writing. K.M.F.F.: Collection, data collection, data analysis and writing. L.C.R.: Discussion of data and assistance in the preparation of the final version of the manuscript. B.M.C.: Data analysis and writing.