Otolith morphology and body size relationships for juvenile Sciaenidae in the Río de la Plata estuary ( 35-36 ° S ) *

Seven species of the family Sciaenidae have been reported inhabiting bottom waters of the Río de la Plata estuary. Of these juvenile stripped weakfish (Cynoscion guatucupa), king weakfish (Macrodon ancylodon), whitemouth croaker (Micropogonias furnieri) and banded ground drum (Paralonchurus brasiliensis) are by far the most abundant. These species are also important prey of several top-predator species among large fishes, seabirds and mammals. In this study we provide new information regarding otolith morphology and body size relationships of juvenile sciaenid fish. The sagittae morphology of juvenile sciaenids showed strong changes in the course of their development (e.g. in overall shape and development of concrescences on the outer surface). Analysing the morphometric relationships, we concluded that otolith length and otolith mass are good indicators of fish standard length and fish weight in all species. If otolith length or mass is used to estimate fish length, the regression explained more than 97% of the variation in all species. All equations relating fish standard length and fish weight for juvenile stages of the species studied explained a very large proportion of the variance in the data. These data will help researchers studying food habits of top predators to determine size and weight of usually juvenile fish-prey from length, broad and/or weight of recovered otoliths.

The identification and quantification of fish species in the diet of top predators (marine mammals, fish and seabirds, among others) usually require the analysis and identification of diagnostic hard remains found in food samples.Particularly, sagittae otoliths recovered from digestive tracts, faeces and/or regurgitated scats of piscivorous animals have been extensively used for identification of prey (Granadeiro and Silva, 2000;Everson et al., 1999;Tollit et al., 1997;Corrêa and Vianna, 1993;White, 1991;Jobling and Breiby, 1986).The adequate identification of fish prey from otoliths is essential in feeding studies, where the relationship between otolith length and fish length or weight given by specific regressions can also be used to estimate the size and/or mass of prey (Harvey et al., 2000;Everson et al., 1999;Baldás et al., 1997;Reid, 1996;Corrêa and Vianna, 1993;Gamboa, 1991).Furthermore, fishery biologists have been using this information as an indirect indicator of fish distribution and also for quantifying the potential effects of natural predation (Corrêa and Vianna, 1993;Mauco et al., 2001).
The sagittae otoliths of each sciaenid species show typical morphology features, most of which have been studied by several authors in adult fishes (Volpedo and Echeverría, 2000;Baldás et al., 1997;Corrêa and Vianna, 1993;Torno, 1976).Nevertheless, information about the morphology and morphometry of otoliths in juvenile individuals is scarce, and only partial information is available on whitemouth croakers (Volpedo and Echeverría, 1999).The aim of this study was to provide new information regarding otolith morphology and body size relationships of juveniles of the four most abundant sciaenids in the Río de la Plata estuary: C. guatucupa, M. ancylodon, M. furnieri and P. brasiliensis.These data might be used by researchers studying food habits of top predators to determine the size and weight of juvenile fish-prey from length, width and/or weight of recovered otoliths.We also provide new otolith images from juvenile fish, usually away from adult morphology, to be used as and aid for identifying fish species.

MATERIALS AND METHODS
Fishes were collected from December 1999 to December 2000 using a beach seine designed for capture of post-larval and juveniles (mesh = 5 mm) during fishing research cruises in the estuary of the Río de la Plata.Samples were preserved in 98% ethanol until examination.Fishes were identified following keys provided in Menni et al. (1984).Standard length (SL, from the most anterior point to the base of hypural plate at caudal flexion) was measured to the nearest mm and fish were weighed to 0.01 g using a digital scale (Mettler PC 2000).
Sagittae were removed, cleaned, and stored dry for subsequent examination.Maximum otolith length (OL) and width (OW) were measured using a binocular magnifier (OLYMPUS 52 H10, 10X) and RATOC otolith image analyser program version 3.0.Total otolith mass (OM) was weighed using a standard analytical scale to the nearest 0.1 mg.Sagittae were described following Volpedo and Echeverría (2000) and Baldás et al. (1997).The following characters were analysed: overall morphology and dif-ferent types of marginal sculpture, mainly dorsal and ventral margins, differences in the levels reached between the ostium and the cauda, presence and absence of a dorsal areal depression, calcareous concrescence and curvature of the otolith (Fig. 1).
Differences between right and left sagittae were tested using a paired t-test.The following morphometric relationships were analysed: fish weight (W) vs. SL, OL vs. SL, OW vs. SL, otolith total mass (OM) vs. SL, OL vs. W, OW vs. W, and OM vs. W. Morphometric relationships were built using potential regression models which best fit the data distribution.

Morphological analysis
Based on changes observed in otolith morphology and following the criteria given by Volpedo and Echeverría (1999), we distinguished three otolith categories according to otolith growth: category I includes fishes from 30 to 60 mm SL, category II includes those from 70 to 110 mm SL, and category III includes fishes from 120 to 170 mm. to ventral margin in all categories; ostium and cauda were at the same level in category I but the latter was much deeper than the former in categories II and III.
Only category III showed calcareous concrescence at the back of the outer face.The inner face was convex, mainly in otoliths of categories II and III.All categories presented dorsal areal depression.

Morphometric analysis
Most parameters measured did not show significant morphometric differences between left and right otoliths: only OW in whitemouth croaker and 236 J.A. WAESSLE et al.OL in king weakfish showed significant differences between left and right otoliths.However, these differences were on average smaller than 3% in whitemouth croaker and 1% in king weakfish (Table 1).The relationship between fish SL and W was described by an exponencial equation, the coefficient of determination being higher than 0.98 in all species (Table 2).
Analysis of otolith morphometric parameters vs. SL showed that the potential regression model explained more than 95% of the variance in all species (Table 3, Fig. 3), except for the relationship between OW and SL in banded ground drum.In all species the relationships between morphometric parameters and W were described by the potential regression model, which explained 94% of the variance.Only banded ground drum showed no significant accuracy in the OW vs. W relationship (Table 4, Fig. 4).

DISCUSSION
In agreement with Volpedo and Echeverría (1999) for whitemouth croaker, the sagittae morphology of other juvenile sciaenid species major strong changes during their development.Overall morphology changed from polygonal to oblong in stripped weakfish, oblong to oblong with angular front zone in banded ground drum, oblong to polygonal in whitemouth croaker, but remained with a generally oblong shape in king weakfish.According to Volpedo and Echeverría (1999), the variations may be due to the way in which the CaCO 3 is deposited during the sagittae development.While in small otoliths CaCO 3 is uniformly deposited, as the otolith grows the deposition of this salt occurs mainly in the front-back axis, resulting in a large elongation of these otoliths.
All species showed ostium and cauda at the same level in category I, the latter being much deeper than the former in categories II and III.This agreed with the observations of Volpedo and Echeverría (1999) for juvenile and adult whitemouth croakers, and of Baldás et al. (1997) in adult whitemouth croakers, stripped weakfish, king weakfish, banded ground drums and Argentine croakers (Umbrina canosai Berg, 1895).The presence of a dorsal depression area on the inner face, dorsally accompanying the sulcus, was observed in all analysed sagittae of stripped weakfish, king weakfish and whitemouth croaker.These results were in agreement with those reported by Volpedo and Echeverría (1999) for whitemouth croaker and Chao (1978) for Cynoscion spp.We also agree with the position of a dorsal depression area in stripped weakfish reported by Corrêa and Vianna (1993).
A calcareous concrescence on the outer face of category III sagittae was strongly developed in stripped weakfish, whitemouth croaker and banded ground drum.Saenz Echeverría (1950) reported pearls or little smooth pebbles as calcareous concrescences in juvenile fishes, changing to prominent protuberances in otoliths of adult individuals in sciaenids from the Sahara Coast.The presence of crestlike, bloblike projections or granulations in Western Atlantic sciaenids was reported for Chao (1978).Baldás et al. (1997) reported the occurrence of these calcareous concrescences in adult whitemouth croakers, banded ground drums and Argentine croakers.Volpedo and Echeverría (1999) argued that the differential pattern of calcium carbonate deposition and the presence of calcareous concrescence in juvenile (120 to 300 mm total length [TL]) and adult whitemouth croakers could be due to changes in physical and chemical parameters in different habitats.Therefore, the distinctive characteristics of the otoliths in the Sciaenidae is the absence of the rostrum, excisura and antirostrum (Torno, 1976), the presence of cauda and ostium with different textures, cauda curved to ventral margin, and much deeper than the ostium (Corrêa and Vianna, 1993), the presence of calcareous concrescence in the inner face (Baldás et al., 1997) and a dorsal depression on the inner face of the former otolith stages (this study).
All equations relating otolith variables to fish standard length and fish weight for juvenile sciaenid specimens studied explained a very large proportion of the variance in the data (see Table 2); these results are in agreement with those published by Haimovici and Velasco (2000) for the same sciaenid species from southern Brazil.Also, our data on OM vs. SL of banded ground drum fall in the range of the South Brazilian population of this species (Vargas, 1976).Analysing the morphometric relationships, we concluded that otolith length and otolith mass are good indicators of fish standard length and fish weight in all species.If otolith length is used to estimate fish length, the potential regression explained more than 97% of data variation in all species.If otolith mass is used, potential regression explained more than 97% of the variation in banded ground drum, king weakfish and whitemouth croaker and more than 95% in stripped weakfish.Baldás et al. (1997) described the relationship between otolith length and fish total length by using potential models in stripped weakfish (50-600 mm TL) and linear models in whitemouth croaker (140-370 mm TL).Also, Torno (1976) developed linear models to describe the same relationship in king weakfish (230-350 mm TL).It was impossible to compare our model with those from the literature due to differences in the size distribution of the fish sampled and the different morphometric measurements taken.When otolith length and otolith mass were used to estimate fish weight, the coefficient of determination (R 2 ) in all species was slightly smaller than the respective values of fish weight estimated from fish length (see Tables 2 and 4).However, the straight estimation of fish weight by using otolith length or otolith mass could avoid the introduction of additional errors due to the use of two regressions instead of only one.
FIG. 1. -Terminology used throughout the text in the descriptions of otoliths of four sciaenid species from the Río de la Plata estuary.Left: view from inner side, right: dorsal view, anterior is up.
): overall morphology changed from oblong in category I to oblong with angular front zone in the other categories.Sinuate margins in category I presented entire margins after category II.Cauda was curved to ventral margin in all categories; ostium and cauda were at the same level in category I but the latter was much deeper than the former in categories II and III.Otoliths from category III presented calcareous concrescence distributed mainly on the back of the outer face.The inner face was most convex after category II.No categories presented dorsal areal depression.Whitemouth croaker (Fig.2C): overall morphology changed from oblong in category I to oblong with angular zone in the frontal margin in category II, and polygonal in category III.Margins sinuate in category I became irregular in category II, while in category III the ventral margin became sinuate and entire.Cauda curved to ventral border; ostium and cauda were at the same level in category I but the latter was much deeper than the former in categories II and III.The outer face presented slight undulations in category I, undulations and calcareous concrescences distributed in the centre of the outer face were present from category II onwards.The inner faces were most convex in otoliths of categories II and III.All categories presented a dorsal depression.King Weakfish (Fig.2D): overall morphology was oblong in all categories.Margins sinuate in category I, changed to entire dorsal margin and sinuate ventral margin in category II.Category III presented entire margins.Cauda curved to ventral margin; ostium and cauda were at the same level in category I but the latter is much deeper than the former in categories II and III.No categories presented calcareous concrescence.The inner face was rather convex from otoliths of category II onwards.All categories showed a narrow dorsal depression and a mortise in the dorsal margin, located in the back of otolith, next to the end of the cauda.

TABLE 1 .
-Differences between right and left otoliths of four sciaenid fish from the Río de la Plata estuary tested by paired t-test.

TABLE 4 .
-Relationships between otolith morphometric parameters and Weight.Coefficient of determination (R 2 ) and sample sizes (n) are indicated.OL: otolith length, OW: otolith width, OM: otolith mass.All regressions were significant at P< 0.05