Otol ith preparation and analysis
After extraction, the sagittal otoliths were washed three times in an ultra-pure water bath (milliQ 0.0055 μS). After the remaining tissues were removed under a binocular, otoliths were dried and stored in 1.5-mL plastic Eppendorf tubes. The left otolith extracted from each fish was embedded in araldite resin 2020 (Huntsman) with the sulcus acusticus downward.
They were grounded in the sagittal plane up to the core with ultra-pure water and sandpaper with grains gradually decreasing from 2400 μm to 1200 μm, 9 μm, and 3 μm. Finally, the otoliths were rinsed with ultra-pure water and air-dried.
Otolith microchemical composition was assessed using 257 nm femtosecond laser ablation (Lambda 3, Nexeya, France) inductively coupled with plasma mass spectrometry (Elan DRCII, Perkin Elmer) (LA-ICPMS). This delivers 360 fs pulses at wavelengths of 1030 nm and can be operated at high repetition rates (up to 100 kHz). A 2D galvanometric scanner allows the fast movement of the laser beam (10 μm) at the surface of the sample to simulate virtual beam shaping when the laser is operated at a high repetition rate. Considering the otolith growth ring pattern, an elongated laser beam (10 × 50 μm2) was simulated in order to preserve the high spatial resolution while keeping the highest signal sensitivity. The laser was operated at 300 Hz with a pulse energy of 35 μJ while the scanner was doing a permanent 35- μm-wide, back-and-forth movement at a speed of 2 mm/s, resulting in a 20 × 50 μm laser beam. Combined with this back-and-forth movement, the sample was continuously moved along the posterior axis from the nucleus to the edge of the otolith at a speed of 5 μm/s, resulting in an uninterrupted ablation on the grounded surface. In order to prevent a blast effect on the nucleus, the ablation was started 200 μm before the nucleus. The ablation depth was evaluated at 10 μm.
At the beginning and end of each session, careful calibrations were carried out using NIST 610, 612, and 614 (National Institute of Standards and Technology). Quality control was systematically evaluated using pelletized CRM NIES 22 otolith powder (Certified Reference Material produced by the National Institute for Environmental Studies). 43Ca was used as an internal standard for each ablation to correct for instrumental error in terms of ablation yield, sample transport and detection. Analysed isotopes were 86Sr, 135Ba, 138Ba, 24Mg, 26Mg, 55Mn, 63Cu, 65Cu, 66Zn, 68Zn, 57Fe, 232Th, and 238U, which are frequently used in microchemistry studies (Vasconcelos et al., 2011). Isotopes for which 75% of the measurements were above the limit of detection for at least one individual were retained. Furthermore, for elements with two isotopes (e.g. 63Cu, 65Cu), only the isotope with the highest natural abundance was kept after checking that no spectral interference was affecting the reliability of the result. After standardization by calcium (Campana, 1999), the remaining element ratios were Ba/Ca, Sr/Ca, Mn/Ca, Zn/Ca, Cu/Ca, Fe/Ca, and Mg/Ca.
Growth est imat ion: val idation from cohorts and otol ith microst ructures
Four species were caught at Lancieux Bay: Hyperoplus lancelolatus, Hyperoplus. immaculatus, Gymnammodytes semisquamatus and Ammodytes tobianus, the latter accounted for 73.04% (n = 642) of the catches.
At Lancieux Bay, fish sizes ranged from 38 to 175 mm (Figure 3). A maximum age of 5 years was observed, but most fishes were one or two years old. Whatever the sampling period, individuals from the G2 (131.62 ± 7.53 mm) and G3 (146.47 ± 8.34 mm) age classes were detected. At the beginning of the survey, G1 individuals born in spring 2011 (96.71 ± 7.83 mm) were still very abundant in February and April 2012, with an estimated age of around one year old (i.e. G1, Figure 3, black stars).
Ontogenic and temporal var iat ion in otol i th fingerprints
Mg/Ca varied significantly both according to season and fish age (Df = 3, Deviance explained per factor = 77%, p = 1.45e-06***) and exhibited higher ratios in juvenile otoliths and during the summer (Figure 6). The microchemical fingerprints of the other elements ratios were not different (mean ± sd of all individuals: Ba/Ca = 2.12e-06 ± 6.24e-07, Sr/Ca = 4.40e-03 ± 1.02e-07, Mn/Ca = 1.19e-05 ± 6.51e-06, Zn/Ca = 8.63e-06 ± 7.38e-06, Cu/Ca = 8.12e-06 ± 7.95e-06, and Fe/Ca = 2.02e-03 ± 2.40e-04). Meaning the variation in Mg/Ca ratio in otolith fingerprint is more due to seasonal effects and physiological changes during the fish’s development than to ontogenic habitat changes (see Discussion). Therefore to avoid biaises, Mg/Ca was removed in further analysis.
Discriminat ion of si tes and opt imal element combinat ion
The three classification methods performed (LDA, RF, and ANN) provided good maximal accuracy of prediction, falling between 78.44% and 83.79% (Table 1). LDA had the best maximal prediction accuracy (83.79%), and the best element combination was composed of Cu/Ca and Mn/Ca (Table 1). Capture signature in otoliths from Lancieux was significantly different from those of Chausey (Manova, p = 0.0070**) and Rotheneuf (Manova, p = 0.032**) but Chausey and Rotheneuf were not globally different (Manova, p = 0.27). Mn/Ca ratios were significantly higher in otoliths from Rotheneuf and Chausey than those from Lancieux (Figure 7). Cu/Ca and Zn/Ca ratios were significantly higher in otoliths from Rotheneuf than the two other sites (Figure 7).
Compar ison of l ife stage signatures f rom Lancieux sandeels wi th signatures of si tes of capture (Chausey, Lancieux, and Rotheneuf )
The large majority of the microchemical fingerprints found for the different life stages (i.e. macrostructural zones of otoliths) of Small Sandeels captured at Lancieux always appeared different from the Chausey and Rotheneuf capture signatures (Table 2). Among the elements, Cu/Ca, Mn/Ca, and Zn/Ca ratios were always significantly different from the Rotheneuf capture signature, and Mn/Ca and Fe/Ca ratios differed significantly and the most frequently from Chausey capture fingerprint (Table 2).
Site fidelity of the Small Sandeel according to age: insights from microchemical fingerprint s of otol iths
Firstly, the microchemical signatures of the Chausey and Rotheneuf sites were never similar to those of Lancieux fishes (40 and 20 km away, respectively) suggesting that (i) environmental imprints may occur among nearby marine areas, and (ii) there was a priori no population connectivity between these intertidal sites. This is in accordance with Jensen et al. (2011), who found high fidelity of A. marinus to their nighttime burrowing sites, with a range that did not exceed 5km. Nevertheless, it appeared that diurnal movements could extend about 15 kilometres away from nighttime burrowing sites (Engelhard et al., 2008 for A. marinus).
Finally, the swimming capacity of the Small Sandeel (1 to 1.5 km/h) (Kühlmannand Karst, 1967), suggested that fish caught in Lancieux were unable to reach either the Chausey archipelago or Rotheneuf Beach through daily movement, which does not exclude potential migrations over several days. Therefore, as telemetry is difficult for small fishes and traditional surveys and acoustic methods do not permit tracking individual behaviour, otolith microchemistry appears to be a potentially useful tool to determine the connectivity between sandeel stocks.
Secondly, the signatures of the larval stage (L) and the early juvenile growth stage (i.e. S0b, the beginning of the juvenile zone), including the metamorphosis stage (Wright, 1993), of fish caught in Lancieux, presented significantly distinct microchemical fingerprints from older stages. This first fingerprint change over the Small Sandeel lifespan could be explained by a change inhabitat, probably when the metamorphosis occurred. Indeed, the size estimated from otoliths, at the beginning of the juvenile zone (S0b), just after the larval stage, ranged between 24.96 ± 0.96 mm and 34.04 ± 0.96 mm. Even if no data were found for Small Sandeel, Wright described A. marinus larvae as undergoing metamorphosis over the length range 35 to 55 mm TL, leading to a change from a pelagic to a semi-demersal habitat (Wright, 1993). This early life stage seems to occur for Small Sandeel in coastal waters according to Langham (1971), who never found larvae and post-larvae of the species in the Scottish offshore waters. Our sampling tended to confirm this result, since the smallest size of Small Sandeel settled (metamorphosed) detected at Lancieux was 38 mm. Contrary to the following stages, larvae are notably not in contact with sand during the night, which could explain the change in the microchemical fingerprint. Finally, the change in behaviour and habitat (pelagic to semi-demersal) and the influence of particular ecophysiological characteristics (growth and feeding) of these early stages (larval and during metamorphosis) on the microchemical composition of the otoliths cannot be excluded (Otake et al., 1997; Arai et al., 2000; Chittaro et al., 2006; Tanner et al., 2011).
Study area and fish sampl ing
Three Sandeels species H. lanceolatus, H. immaculatus and G. semisquamatus were collected during May and September 2012 with a four panel fishing trawl (width: 15.90 meters, length: 22.80 meters, 60 mm mesh size in the wings and 3 mm at the bottom) in Lannion Bay, south-west Channel along the coast of the Norman-Breton Gulf (Figure 1). We did not catch any larvae and small juveniles (< 6 cm TL) due to the selectivity of the nets. The total length (TL) and weight (W) of 1037 fishes were measured to nearest g and mm, respectively. Subsamples of 55 fishes were stored at -20°C within one hour of capture for further identification and measurement at the laboratory. To test the spatial variations of microchemical signature in H. lanceolatus, specimens were provided from another subtidal site at Hébihens Isles (n = 7) and one coastal site at Lancieux Bay (n = 15), both located in Saint Malo Bay.
Table of contents :
1. Problématique et objectif principal
2. Les modèles biologiques
3. Objectifs scientifiques et organisation de la thèse
PARTIE 1: LE CAS DE LA COMMUNAUTE DE LANÇON
I. Contexte de l’étude
II. Article 1 : Life history of the Small Sandeel, A. tobianus, infered from otolith microchemistry. A methodological approach.
2. Materials and methods
Conclusions and perspectives
III. Article 2 : Contrasted life histories of three sympatric sandeels cross validated by otolith microchemistry, stable isotopes and functional traits
2. Material and methods
PARTIE 2 : LE CAS DU BAR EUROPEEN, DICENTRARCHUS LABRAX
I. Contexte de l’étude
II. Article 3 : Diversity of life histories of juvenile European sea bass, Dicentrarchus labrax, revealed by the microchemistry of their otolith.
2. Materials and methods
III. Article 4 : Variability of feeding and life traits to young European sea bass in contrasted nursery habitats of the western Channel
2. Material and methods
Histoire de vie inter-espèce à l’échelle des communautés de lançon
Variabilité d’histoire de vie intra-espèce au sein du bar européen
Relations entre histoires de vie et connectivité des HEE