Otolith Fingerprints and Tissue Stable Isotope Information Enable Allocation of Juvenile Fishes to Different Nursery Areas

: Integrated otolith chemistry and muscle tissue stable isotope analyses were performed to allocate juvenile Diplodus puntazzo and Diplodus vulgaris to nurseries in the Adriatic Sea. Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS) was used to quantify the concentrations of chemical elements in the otoliths. Fish muscle samples were analysed for δ 13 C and δ 15 N. In general, Ba/Ca and Sr/Ca ratios and isotopes varied between sites and species. Values of δ 13 C and δ 15 N were signiﬁcantly different between species and sites. Multivariate analysis detected a signiﬁcant difference in the element signature between species while there was no evidence for a signiﬁcant interaction for sites. A clear pattern across the four groups of interest, D. puntazzo _Estuary > D. vulgaris _Estuary > D. puntazzo _Coastal > D. vulgaris _Coastal, following decreases in δ 13 C, and increases in δ 15 N were found. It seems that these species are feeding on the same local food web within more productive estuarine site while at costal site, feeding segregation among investigated species is evident. Both species were re-allocated correctly to the estuarine waters based on the otolith chemistry and stable isotopes information and higher value of δ 15 N. Combining otolith chemistry with tissue isotope ratios of juvenile ﬁsh provided complementary information on nursery habitat use at different spatial scales and elucidated ecological and environmental linkages.


Introduction
Elucidating movement and life-history characteristics of marine organism is of crucial importance for their management and conservation [1][2][3] and the knowledge gap still represents a challenge to scientists working on this issue. Nearshore estuarine and marine ecosystems such as seagrass meadows, marshes and mangrove forests are often referred to as nursery grounds [4] due their positive effects on the diversity and productivity of fish and invertebrates in coastal waters. The greater food abundance and lower predation risk of these shallow habitats support high juvenile densities and may contribute juveniles or sub-adults to adult populations [5]. Coastal ecosystems are highly structured and fragile environments, and many valuable coastal systems are under high anthropogenic pressures, resulting in species loss and habitat degradation [6][7][8]. In particular, the highly populated ever, these species are contemporaneous in nurseries [53], thus confirming the successful temporal partitioning of habitat use between different Diplodus species [51].
Juvenile fish from the genus Diplodus have been previously investigated in three studies. Correira et al. [33] applied solution-based analyses on whole otoliths and laser ablation analysis of otolith cores to obtain insight into the population structure of D. vulgaris. Di Franco et al. [54] investigated within-otolith variability in chemical fingerprints and found that individuals at the same site can show significant variability in elemental uptake. The possible use of otolith fingerprints as natural tags for the identification of juvenile D. sargus and D. vulgaris in ports were studied by Bouchoucha et al. [34]. However, there are no reports of any otolith chemistry studies using D. puntazzo. Other authors have recently conducted chemical analyses of juvenile fish otoliths [12,20,[55][56][57][58][59].
The aim of the present study was to use both otolith chemistry and muscle stable isotope composition to allocate two closely related juveniles of D. vulgaris and D. puntazzo (age-zero) to two different nursery sites: an estuarine and a coastal (marine) nursery. We hypothesized that these closely related fish species, simultaneously present in the same nursery areas, exhibit different chemical signatures in estuarine and coastal waters as a reflection of their different behavior in foraging prey in specific nursery, which should consequently allow for the proper allocation of juveniles to a specific nursery. Such knowledge can help to accurately identify nursery origin and determine the relative contributions of individual nurseries to the coastal population of these species.

Study Locations and Fish Collection
Newly settled juveniles of sharpsnout seabream, Diplodus puntazzo and common two-banded sea bream Diplodus vulgaris were collected from two sites along the eastern Adriatic (Figure 1a,b): the estuarine site Pantan and coastal site Sovlja (Figure 1c), as sites known to be essential nursery areas for these species [60][61][62]. They are separated by a distance of 200 km and hydrologically represent different water types in the Adriatic Sea. The Pantan estuary is near Split, and receives the waters of the Pantan River, exhibiting variable salinity gradients during the year (transitional waters), with a muddy-sandy bottom partially overgrown with Zostera marina. Sovlja Cove is near Šibenik and is a typical coastal site, with a partially rocky-sandy bed with patches of Cymodocea nodosa meadows, and less influence of freshwater springs (Table 1). Samples of juvenile fish specimens were collected using a special constructed small shore seine net (L = 25 m; mesh size 4 mm) in June 2018. Three hauls for each site were performed to collect an adequate number of specimens. To avoid temporal variation in otolith chemistry and stable isotope analysis, sampling was carried out in the shortest possible time. At both sites, Pantan and Sovlja, both species, Diplodus vulgaris and Diplodus puntazzo, were present with similar abundance (up to 7 specimens in each haul) and similar sizes (from 38 to 71 mm and 31 to 72 mm, respectively). Additionally, 5 individuals per site of blue mussel, Mytilus galloprovincialis were sampled. Upon collection, specimens were transported to the laboratory and frozen until analysis. For the analysis, total length (TL; Water 2021, 13, 1293 4 of 19 cm) and weight (TW; g) were recorded and specimens were dissected to extract white fish muscle tissue and otoliths for stable isotope analyses and otolith chemistry, respectively. Samples of juvenile fish specimens were collected using a special constructed small shore seine net (L = 25 m; mesh size 4 mm) in June 2018. Three hauls for each site were performed to collect an adequate number of specimens. To avoid temporal variation in otolith chemistry and stable isotope analysis, sampling was carried out in the shortest possible time. At both sites, Pantan and Sovlja, both species, Diplodus vulgaris and Diplodus puntazzo, were present with similar abundance (up to 7 specimens in each haul) and similar sizes (from 38 to 71 mm and 31 to 72 mm, respectively). Additionally, 5 individuals per site of blue mussel, Mytilus galloprovincialis were sampled. Upon collection, specimens were transported to the laboratory and frozen until analysis. For the analysis, total length (TL; cm) and weight (TW; g) were recorded and specimens were dissected to extract white fish muscle tissue and otoliths for stable isotope analyses and otolith chemistry, respectively.

Sample Preparation
Sagittal otoliths (hereafter: otoliths) were removed, rinsed with water, cleaned of soft tissue with plastic dissecting pins, washed with Milli-Q water, air dried, and stored in labelled plastic vials. The otoliths were embedded in epoxy resin (Buehler EpoThin 2) and sectioned transversely through the core using a low-speed precision saw (Buehler Isomet 1000) equipped with a 0.4 mm thick diamond-coated blade. Otoliths sections were affixed to glass slides using clear Crystalbond and subsequently ground (F800 and F1200 grit SiC powder) and polished using a soft cloth impregnated with diamond paste (3 µm). After polishing, otoliths were rinsed and cleaned ultrasonically (2 min).
Muscle tissue of M. galloprovincialis was used as appropriate baseline since sedentary bivalves can be useful indicators of isotopic baseline [63] in the coastal ecosystem. That is needed to integrate the variation in isotope values at the base of food webs [64] when trophic status of specific marine organisms is requested while data of prey spectra trophic status is unknown.
Standard preparation for stable isotope analysis consisted of oven drying samples at 60 • C until constant weight. Tissues were then ground to a fine powder with a mortar and pestle and approximately 1 mg of sample was weighed into tin cups. Lipid extraction of fish muscle samples was not performed as individuals were juveniles and body lipid was uniformly low (<5%) and insufficient to bias carbon stable isotope analysis or require corrections as suggested by Post et al. [65].

LA-ICP-MS Analysis of Otoliths
The concentrations of Li, Na, Ca, Mg, Mn, Zn, Sr, Mo, Ba, Pb, and U were determined using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) in line scan mode, through the otolith core from edge to edge ( Figure 2). Each point on the otolith corresponds to a specific point on laser trajectory enabling selection of the otolith part to be analyzed.

Sample Preparation
Sagittal otoliths (hereafter: otoliths) were removed, rinsed with water, cleaned of soft tissue with plastic dissecting pins, washed with Milli-Q water, air dried, and stored in labelled plastic vials. The otoliths were embedded in epoxy resin (Buehler EpoThin 2) and sectioned transversely through the core using a low-speed precision saw (Buehler Isomet 1000) equipped with a 0.4 mm thick diamond-coated blade. Otoliths sections were affixed to glass slides using clear Crystalbond and subsequently ground (F800 and F1200 grit SiC powder) and polished using a soft cloth impregnated with diamond paste (3 μm). After polishing, otoliths were rinsed and cleaned ultrasonically (2 min).
Muscle tissue of M. galloprovincialis was used as appropriate baseline since sedentary bivalves can be useful indicators of isotopic baseline [63] in the coastal ecosystem. That is needed to integrate the variation in isotope values at the base of food webs [64] when trophic status of specific marine organisms is requested while data of prey spectra trophic status is unknown.
Standard preparation for stable isotope analysis consisted of oven drying samples at 60 °C until constant weight. Tissues were then ground to a fine powder with a mortar and pestle and approximately 1 mg of sample was weighed into tin cups. Lipid extraction of fish muscle samples was not performed as individuals were juveniles and body lipid was uniformly low (<5%) and insufficient to bias carbon stable isotope analysis or require corrections as suggested by Post et al. [65].

LA-ICP-MS Analysis of Otoliths
The concentrations of Li, Na, Ca, Mg, Mn, Zn, Sr, Mo, Ba, Pb, and U were determined using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) in line scan mode, through the otolith core from edge to edge ( Figure 2). Each point on the otolith corresponds to a specific point on laser trajectory enabling selection of the otolith part to be analyzed. Analyses were performed at the Institute of Geosciences, JGU, Mainz, Germany, using an ESI NWR193 ArF excimer laser ablation system equipped with the TwoVol2 ablation cell, operating at 193 nm wavelength, coupled to an Agilent 7500ce quadrupole ICP-MS. Sample surfaces were preablated prior to each line scan to prevent potential surface Analyses were performed at the Institute of Geosciences, JGU, Mainz, Germany, using an ESI NWR193 ArF excimer laser ablation system equipped with the TwoVol2 ablation cell, operating at 193 nm wavelength, coupled to an Agilent 7500ce quadrupole ICP-MS. Sample surfaces were preablated prior to each line scan to prevent potential surface contamination. The laser repetition rate was 7 Hz and laser energy on samples was about 3 J/cm 2 . Background intensities were measured for 15 s. Line scans were carried out at a scan speed of 5 µm/s, using a rectangular beam of 50 x 40 µm (preablation beam 80 × 40 µm). Synthetic glass NIST SRM 612 (National Institute of Standards and Technology; Gaithersburg, Maryland, United States ) was used to calibrate element concentrations of otolith samples and quality control materials (QCMs) (USGS MACS-3, USGS BCR-2G, NIST SRM 610) ( Table 2) were used to monitor accuracy and precision of the LA-ICP-MS analysis applying the preferred values available from the GeoReM database ([66], application version 26; compared with [67][68][69]). Signals were monitored in time-resolved mode and processed using an in-house Excel spreadsheet [70]. Details of the calculations are given in Mischel et al. [71]. The concentration of 43 Ca as an internal standard in otoliths was taken as 38.8% by weight or 388,000 ppm following the determination of otolith Ca concentration [72]. Concentrations determined on the otoliths were converted to molar concentrations and standardized to calcium.

Stable Isotope Analyses of Muscle Tissue
Muscle samples were analyzed for δ 13 C and δ 15 N using a PDZ Europa ANCA-GSL elemental analyser interfaced to a PDZ Europa 20-20 isotope ratio mass spectrometer (Sercon Ltd., Cheshire, UK) at the UC Davis Stable Isotope Facility. Samples were combusted at 1000 • C in a reactor packed with chromium oxide and silvered copper oxide. Following combustion, oxides were removed in a reduction reactor (reduced copper at 650 • C) and the helium carrier was released through a water trap (magnesium perchlorate and phosphorous pentoxide). N 2 and CO 2 were separated on a Carbosieve GC column (65 • C, 65 mL/min) before entering the Isotope-ratio mass spectrometry (IRMS). Stable isotopes were expressed in standard delta (δ) notation as parts per thousand (‰).
During analysis, samples were interspersed with several replicates of at least four different laboratory reference materials, previously calibrated against international reference materials, including: IAEA-600, USGS-40, USGS-41, USGS-42, USGS-43, USGS-61, USGS-64, and USGS-65 reference materials. A sample's provisional isotope ratio was measured relative to the reference gas peak analyzed against each sample. These provisional values were finalized by correcting the values for the entire batch based on the known values of the included laboratory reference materials. The long term standard deviation is 0.2 ‰ for 13 C and 0.3 ‰ for 15 N [73].

Data Analysis
Element-to-Ca data for Li, Na, Mg, Ba, Sr, Mn, Zn, Mo, Pb, and U were determined for all specimens. Most of these element-to-Ca data were below quantification and detection limits. Some ratios including Na/Ca, Mg/Ca, Zn/Ca, Mn/Ca, and Li/Ca exceeded the detection limit in several otoliths, although they were below the quantification limit in most samples. Ba/Ca and Sr/Ca ratios were above the detection and quantification limits [74] and thus subjected to further analysis. Element concentration data Ba/Ca and Sr/Ca ratios for D. vulgaris and D. puntazzo samples exceeding 31-point (31-pt) running averages by 5σ were considered outliers and excluded from further analysis (see [75,76]). For data visualization, element linear raster was smoothed using a 31-pt arithmetic running average. Differences in otolith chemistry composition were evaluated via the permutational analysis of variance (PERMANOVA) using Manhattan distance dissimilarity matrices [77], since both elements were on very comparable measurement scales. The metric Multi-dimensional Scaling (mMDS) ordination were used for showing the patterns across the four groups of interest and the contribution of each element isotope composition to the obtained distance. Starting point for data selection on linear raster was 200 µm which corresponds approximately to the third month of fish juvenile life according to settlement mark [34,51,78,79]. We calculated the Manhattan measure separately for each of the barium and strontium variable sets and then averaged the resulting Manhattan distance matrices to get a single overall matrix that measures the differences between fish species for the overall otolith signatures for both elements. Differences in muscle δ 13 C and δ 15 N isotope ratios were normalized and evaluated via PERMANOVA using Euclidean distance dissimilarity matrices.
Canonical analysis of principal coordinates (CAP) was used to estimate the accuracy of otolith element signatures and muscle stable isotopes in classifying fish to their collection site. CAP is a routine for performing canonical analysis by calculating principal coordinates from the resemblance matrix among groups of samples to predict group membership, positions of samples along another single continuous variable or finding axes having maximum correlations with some other set of variables [77].
CAP analyses were run separately for each of the two factors: "Site" and "Species". The CAP routine output scores were then merged for both factors. Finally, we relate the distance matrix based on otoliths to the distance matrix based on isotopes and performed CAP as a canonical correlation analysis of the otolith distance matrix on the isotope (continuous quantitative) values [77].
Univariate permutational analysis of variance (PERMANOVA) was used to test the difference of site or species effects on elemental data obtained from otoliths and stable isotope data obtained from white muscle. Statistical analysis was done using PRIMER (V. 7.0.13; Auckland, NZ) and graphs were prepared using SigmaPlot (v. 13.0; Systat Software Inc, San Jose, CA, USA).

Multi-parameter Comparison
When the otolith chemistry data were combined into a single matrix, PERMANOVA analysis detected that "Species" differed significantly in their element signatures, although significant level is not high (P = 0.049) while "Site" did not (Table 3). There was also no evidence for a significant interaction, as PERMANOVA analysis conducted after pooling the Site x Species interaction term did not change this result. Table 3. Summary of PERMANOVA results for the multivariate analysis of overall elemental composition of strontium (Sr) and barium (Ba) in otoliths (a) and overall carbon (δ 13 C) and nitrogen (δ 15 N) stable isotope values in muscle tissue (b) for juvenile Diplodus puntazzo and Diplodus vulgaris collected at different sites. After pooling the isotope data, the plot clearly showed effects for each of the four "Species x Site" groups and for each of the stable isotopes ( Figure 5). PERMANOVA

Multi-parameter Comparison
When the otolith chemistry data were combined into a single matrix, PERMANOVA analysis detected that "Species" differed significantly in their element signatures, although significant level is not high (P = 0.049) while "Site" did not (Table 3). There was also no evidence for a significant interaction, as PERMANOVA analysis conducted after pooling the Site x Species interaction term did not change this result. Table 3. Summary of PERMANOVA results for the multivariate analysis of overall elemental composition of strontium (Sr) and barium (Ba) in otoliths (a) and overall carbon (δ 13 C) and nitrogen (δ 15 N) stable isotope values in muscle tissue (b) for juvenile Diplodus puntazzo and Diplodus vulgaris collected at different sites. After pooling the isotope data, the plot clearly showed effects for each of the four "Species x Site" groups and for each of the stable isotopes ( Figure 5). PERMANOVA showed that both factors ("Species" and "Site") had main effects and a significant interaction term ( Table 3). The metric MDS of the bivariate isotope data showed patterns across the four groups, with an evident pattern with a decrease in δ 13 C ( Figure 5A showed that both factors ("Species" and "Site") had main effects and a significant interaction term ( Table 3). The metric MDS of the bivariate isotope data showed patterns across the four groups, with an evident pattern with a decrease in δ 13 C ( Figure 5A) and increase in δ 15 N ( Figure 5B) (going from left to right). The four groups ordered along this axis as follows: D. puntazzo_Estuary > D. vulgaris_Estuary > D. puntazzo_Coastal > D. vulgaris_ Coastal. Separate CAP analysis for each of the two factors ("Site" and "Species") gave successful discrimination for species but not for sites. In particular, 80% D. puntazzo specimens were correctly allocated based on the otolith chemistry information, as opposed to 77.8% of D. vulgaris specimens. The two-way CAP plot obtained by merging output scores for the CAP analysis of "Site" and "Species" showed separation of the two species (Figure Separate CAP analysis for each of the two factors ("Site" and "Species") gave successful discrimination for species but not for sites. In particular, 80% D. puntazzo specimens were correctly allocated based on the otolith chemistry information, as opposed to 77.8% of D. vulgaris specimens. The two-way CAP plot obtained by merging output scores for the CAP analysis of "Site" and "Species" showed separation of the two species ( Figure 6). It is apparent that the site differences ("E" estuary vs. "C" coastal) were able to distinguish for D. vulgaris. In contrast, the D. puntazzo samples from the estuary were consistently clustered, while coastal samples were more variable, making them difficult to classify.  . It is apparent that the site differences ("E" estuary vs. "C" coastal) were able to distinguish for D. vulgaris. In contrast, the D. puntazzo samples from the estuary were consistently clustered, while coastal samples were more variable, making them difficult to classify. The mean isotope values for the four groups of factors (Species x Site) were plotted as distances among centroids based on otolith data (Figure 7), which showed a clear separation of the coastal and estuarine sites. This was confirmed by CAP as a canonical correlation analysis of the otolith distance matrix on the isotope (continuous quantitative) values ( Figure 8). According to our results, based on the otolith chemistry and stable isotope information, correct re-allocation of D. vulgaris individuals to the estuarine waters were confirmed. Samples of D. puntazzo were correctly re-allocated due to the higher value of δ 15 N to estuarine waters. The mean isotope values for the four groups of factors (Species x Site) were plotted as distances among centroids based on otolith data (Figure 7), which showed a clear separation of the coastal and estuarine sites. This was confirmed by CAP as a canonical correlation analysis of the otolith distance matrix on the isotope (continuous quantitative) values ( Figure 8). According to our results, based on the otolith chemistry and stable isotope information, correct re-allocation of D. vulgaris individuals to the estuarine waters were confirmed. Samples of D. puntazzo were correctly re-allocated due to the higher value of δ 15 N to estuarine waters. Water 2021, 13, x 12 of

Discussion
This study investigated the potential of otolith chemistry and tissue stable isotope analyses to distinguish between two different nursery areas of two closely related fish species of the genus Diplodus. Juveniles of D. puntazzo and D. vulgaris from the Pantan and Sovlja sites have similar reproductive and early life characteristics [52], inhabiting nursery habitats and leaving them in early summer [53,60]. The larvae of D. puntazzo settle in these shallow sites earlier as they hatch several weeks before D. vulgaris, so their juveniles are larger at both sites [51].
A commonly used method is laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS), which produces an elemental fingerprint at a discrete time-point in the life of a fish [80]. Trace elements (e.g., Ba, Li, Mg, Mn, and Sr) and heavy metals (e.g., Pb, Cu, and Zn) are acquired by fish during the life history and preserved within the otolith structure [19,[80][81][82]. In addition to these typically analyzed elements, we also examined Na, Mo, and U in line with the protocol of the Institute of Geosciences, JGU [83]. Unfortunately, as most of the analysed element/Ca ratios were below the quantification and detection limits, only Ba/Ca and Sr/Ca were analysed in this study. A number of factors, such as salinity, temperature, water chemistry, age and growth, physiology, and metabolism may be responsible for the incorporation of trace elements into otoliths, though this is a complex process and remains poorly understood for most elements (with the exception of Ba and Sr) [27,33,[84][85][86][87][88][89][90][91][92].
Data for Ba/Ca elemental composition were not significant between species and sites, although more prominent differences were obtained between species. Generally, Ba incorporation into otoliths appears largely determined by ambient concentrations, which are spatially variable and typically higher in inshore waters, estuaries, and upwelling zones [1,[91][92][93][94][95][96]. Although, both sites are inshore, Pantan is estuarine and Sovlja is coastal, and therefore the hydrological conditions differ. Though not substantial, there is some enrichment of Ba in the coastal Sovlja site, likely influenced by local fluvial runoff and groundwater input, as suggested by Correira et al. [33] which consequently raise this concentration of above expected. The Ba/Ca concentration ratios were different in both species at both sites, confirming variability in element uptake of different species at same site [34,74]. Further on, Bouchoucha et al. [34] studying life of juvenile D. vulgaris and D. sargus reported that Ba was systematically the most discriminating element, since its concentrations in otoliths were generally higher outside ports than inside, probably due to river runoff. The Sr/Ca ratio was also more variable between species and sites but this difference was not significant for site neither for species. Sr incorporation is also influenced by ambient concentration, and has been linked to salinity, though temperature, ontogeny and growth rate may also influence patterns of Sr incorporation into otoliths [1,16,33,87,92,97,98]. The higher Sr levels from Sovlja are likely related to exogenous factors (marine site with higher salinity and temperature), though there may also be certain endogenous causes since D. vulgaris incorporated more Sr at both sites but this influence is too weak to make a significant difference. However, the variability with at each otolith concentrations have to be discussed with attention due different sampling size and site.
Since the investigated species are closely related and show no temporal segregation in nursery areas, we hypothesized that foraging behavior and diet composition may have contributed to the observed differences in the element incorporation between species and sites. Both, δ 13 C and δ 15 N differed significantly between sites and species. The median of δ 15 N was higher for D. puntazzo while D. vulgaris had higher values of δ 13 C at both sites. For soft tissue stable isotopes, lower δ 13 C values were found at Pantan, which agrees with the expected natural patterns of δ 13 C variation and displays an enrichment trend along the terrestrial-estuarine-marine gradient [99]. In addition, the overall richer δ 15 N values at Pantan than at the coastal Sovlja site were likely due to anthropogenic nitrogen inputs in the estuary (e.g., wastewater, and fertilizers) [20,36,46,100]. The observed intra-species differences in fish muscle stable isotopes reflected the isotope composition of local food webs and available prey [20,101]. It seems that in estuarine Pantan, both species feed on the same local food web for a longer period and do not disperse widely around the sampling site. Since targeted specimens in this study are juveniles representing similar growing stage and values obtained for stable isotopes were adjusted to blue mussel baseline, one should consider that in general for fish muscle the turnover rate is around months [102,103], while short-living consumers, such as zooplankton, have high tissue turnover rates, similar to that of phytoplankton [104]. Abecasis et al. [42] reported that in estuarine waters, juvenile D. vulgaris make only short movements and typically remain in the same areas for extended periods, and this is likely also the case for D. puntazzo. Higher isotope values in D. puntazzo may reflect that these specimens are possible several weeks older and, thus, larger and are likely to forage on bigger prey. Estuarine areas are often highly productive with a narrow prey spectrum, but with high prey availability and abundance [105]. The marked differences in isotope concentration of muscle tissue in specimens from the coastal site Sovlja suggest that these two species feed on different local food webs, with D. vulgaris foraging at a higher trophic level [106]. In coastal areas, the availability and abundance of prey are usually lower though the prey spectrum is wider [107].
PERMANOVA clearly confirmed the different element signatures of D. vulgaris and D. puntazzo. Although the incorporation of Ba and Sr is largely influenced by environmental factors (temperature and salinity), these differences in the otolith fingerprints likely resulted from the homeostatic apparatus of the individual fish, i.e., its physiology and ultimately its genetic makeup [98]. The fact that PERMANOVA did not reveal significant difference between sites raises the question of how these sites, defined as estuarine and coastal, really differ in the study area due to the specific oceanographic properties of the eastern Adriatic Sea, with many freshwater grounds in the coastal area [108]. Unfortunately, lack of water sample from both habitats disable relevant comparison and establishment of the relationship between Ba and Sr concentration and otolith microchemistry in this study. For sure, such limitations have to be consider in future sampling designs.
The metric MDS of the bivariate isotope data clearly shows patterns that can be interpreted as decreases in δ 13 C and increases in δ 15 N (D. puntazzo_Estuary > D. vulgaris_Estuary > D. puntazzo_Coastal > D. vulgaris_Coastal). Both species exhibited different behaviours in estuarine and coastal waters, which is likely related to foraging and feeding. D. puntazzo is more efficient in feeding in estuarine waters than D. vulgaris, and it grows faster, incorporating more δ 15 N in the more productive estuarine waters [105]. Moreover, this greater efficiency of D. puntazzo over D. vulgaris is even more prominent in coastal waters, where prey is generally less available and foraging time is longer [106,107].
Furthermore, we attempted to correctly allocate these species to the estuarine or coastal environments through CAP analyses. 80% D. puntazzo and 77.8% of the D. vulgaris specimens were allocated correctly based on the otolith chemistry information. However, the results suggested that over time, the otolith fingerprint differences observed in D. vulgaris in different waters will become more significant and thus it can be allocated correctly in estuarine water using otolith chemistry and stable isotope information. D. puntazzo incorporates elements into otoliths in different environments in a similar way and therefore can be allocated according to the higher value of δ 15 N in estuarine waters.
The present study provides preliminary insight into juvenile fish nursery use at different spatial scales in the Adriatic Sea by combining otolith chemistry with tissue isotope ratios of the same individuals to determine distinct ecological and environmental linkages [20]. Although, conducted on relatively small sampling size, otolith chemistry results reflected the environmental characteristics of the juvenile Diplodus nursery areas, while muscle stable isotope analysis indicated the isotope differences between species and between sites, accentuating the need to consider both environmental gradients and species behaviour in movement and connectivity studies based on otolith fingerprints. Such knowledge can help to accurately identify nursery origin and determining the relative contributions of individual nursery areas to the adult coastal populations of species [18,39,46]. Moreover, better understanding of settlement and recruitment processes, and nursery habitat use and movement patterns between juveniles and adults enables more sustainable management of fishery resources and essential habitat conservation based on ecological principles.