MARINE ECOLOGY PROGRESS SERIES Published January 31 Mar Ecol Prog Ser ' Daily variations of the saccular endolymph and plasma compositions in the turbot Psetta maxima: relationship with the diurnal rhythm in otolith formation 'Laboratoire de Sclerochronologie des Animaux Aquatiques, IFREMER. DRVIRHILASAA, and ZLaboratoire de Physiologie des poissons, IFREMER, DRVIRA, BP 70,29280 Plouzane, France 3Laboratoire de Physiologie et Toxicologie Environnementale EA 2138, and "Laboratoire de Physiologie Cellulaire et Moleculaire UMR CNRS 6548. Universite de Nice-Sophia Antipolis, Faculte des Sciences, BP 71, 06108 Nice Cedex 2, France ABSTRACT We studied the circadian variations of the ionic composition ([Na']. [K'], [Cl-]),total CO2 ([totCO*]), and proteln concentration ([protein])in the plasma and endolymph in the distal and proxi- mal sides of the sacculus in the turbot Psetta maxima Daily variations in total plasma calcium ([tot Ca]) were also recorded. Significant diurnal fluctuations occurred In most parameters and the spati.al heterogeneity of the endolymph composition was maintained throughout a 24 h period. The main parameters implicated in otolith calcification, endolymph [tot CO,] and endolymph [protein],fluctuated diurnally. [tot CO,] was highest and [protein] lowest at night; the relative concentrations were reversed during the day. The inverse, alternating relationship of maximal and minimal [tot CO2]and [protein] implies that calcium carbonate and organic matrix rates of deposition vary in antiphase. We therefore propose that in the turbot P. maxima, in autumn, the incremental zone (L-zone) is formed at night whereas the discontinuous zone (D-zone) is formed during the daytime. KEY WORDS: Daily cycle . Endolymph . Otolith . Chemical composition . Teleost INTRODUCTION daily increment (Pannella 1980). Since the structure and chemical composition of otoliths is scarcely modi- Teleost otoliths are calcareous structures involved in fied after deposition, they can be cons~deredas biolog- fish balancing and hearing processes (Fay 1980, ical archives and are commonly used for age and Gauldie & Radtke 1990). They are composed of cal- growth estimations (Campana & Neilson 1985). They cium carbonate, mainly in the mineral form of arago- also allow studies on stock discrimination of exploited nite (Carlstrom 1963) and an organic matrix in which populations (Carnpana et al. 1995, Severin et al. 1995) acidic amino acids dominate (Degens et al. 1969). and permit the characterisation of certain events in the Accretion occurs by successive depositions of incre- life cycle (Kalish 1992, Otake et al. 1994, Secor et al. mental (calcium dominant) and discontinuous (matrix 1995). dominant) zones (Watabe et al. 1982, Morales-Nin In relation to the cyclic formation of otolith incre- 1987). In many fish the deposition of these 2 layers ments, several studies have suggested the importance occurs throughout the day, producing a recognisable of photoperiodicity (Taubert & Coble 1977, Tanaka et al. 1981, Radtke & Dean 1982). Tanaka et al. (1981) demonstrated that in Tilapia nilotica (L.), the order of 'Corresponding author. E-mail: [email protected] formation of the incremental zone (L-zone) and discon- O Inter-Research 2000 Resale of full art~clenot permitted Mar Ecol Prog Ser 192: 287-294, 2000 tinuous zone (D-zone) was photoperiod dependent, MATERIALS AND METHODS since a reversal of the light-dark cycle induced a rever- sal in the order of the 2 layers. It has also been shown The experiments were carried out over the 2 days of that the daily rate of increment formation is maintained 22 to 23 October 1997, on turbot Psetta maxima reared in juvenile fish after transfer to constant light (Cam- in running seawater at 18"C, and subjected to a pho- pana 1984, Wright et al. 1991) or darkness (Radtke & toperiodicity of 13 h light to 11 h darkness (the light Dean 1982). Mugiya (1987) and Wright et al. (1992) period starting at 07:OO h). Turbot weighing about 400 suggested that the rhythm in otolith calcification is to 500 g were fed once a day in the morning (around controlled by a circadian variation in plasma chem- 10:OO h) but food was withheld 24 h prior to the begin- istry, since they found that during the dark period ning of the experiments. The fish were acclimated to there is a parallel decline in otolith calcification and the experimental conditions 3 wk before sampling. total and free plasma calcium concentratlons. Further- Batches of 6 fish netted from the tank were taken ar more, Mugiya (1984) found a seasonal reversal in the 3 h intervals. Blood was rapidly collected from the cau- rhythm of otolith calcification associated with a rever- dal vessels and, centrifuged, the separated plasma sal in the plasma calcium cycle. being kept in a refrigerator until analysis. The tech- Fish have, in each inner ear, 3 otollths bathed in niques of endolymph and plasma sampling have been endolymph within a sac. The largest otolith, called the described previously (Payan et al. 1997, 1999). Briefly, sagitta, is located with~nthe sacculus. Although the after decapitating the fish and exposing the inner ear, main characteristics of the endolymph have been de- 2 samples of endolymph of about 5 p1 each were taken scribed earher (Mugiya & Takahashi 1985), it is only in the left sacculus using calibrated capillaries. The recently that endolymph samples have been analysed right sacculus was sampled in the same way immedi- individually (Payan et al. 1997).It has been suggested ately after The sample withdrawn near the macula will that otolith biocalcification is related to endolymph be designed as 'proximal' while that from the opposite composition (Romanek & Gauldie 1996). In a recent side will be designed as 'distal' The endolymph con- work Payan et al. (1999) showed that inside the saccu- tained in capillaries was stored in a refrigerator and lus the endolymph has a lack of spatial uniform~tyof its anal>,sed shortly after The right sacculus samples chemical composition, probably due to the presence of werc? used to measure [protein], [Na'], [Cl-] and [Kt], several types of ionocyte population on the 2 sides of while [tot CO2] was measured on samples of the left the saccular epithelium which are probably concerned sacculus. with active ion transport (Mayer-Gostan et al. 1997). Plasma and endolymph were analysed as described Payan et al. (1999) proposed that the spatial hetero- in Payan et al. (1997). [Na+]and [K'] were measured geneity of some param.eters in the endo1ymp.h is neces- by flaine photometry. [tot Ca], [Cl-], [tot CO?] and sary for the mineralization of the otolith. They sug- [protein] were measured by spectrophotometry using gested that CaCO, deposition is initiated on the Signla kits for the first 3 parameters and Commassie proximal side of the otolith When CaC03is generated, blue for proteins with BSA (bovin serum albumin) as a H+is produced which must be removed. As the sac- standard. As fish were not anaesthetised in these cular epithelium is relatively far from the mineraliza- experiments the level of stress and its variability were tion slte, the proximo-distal bicarbonate gradient estimated by measuring the plasma cortisol concentra- would locally buffer the H' and help calcification to tions by radioimmunoassay. proceed all around the otolith. These authors also Data are expressed as means and standard errors. demonstrated a close relationship between the spatial The diurnal cycles are analysed using univariate heterogeneity in the K concentration of the endolymph ANOVA after having checked variable normality. In and that of the otolith. Very little research has been case of deviation, a KruskalI dnd Wallis test was em- carried out on daily variations in endolymph chemistry. ployed. A contrast analysis was used for the comparison Mugiya & Takahashi (1985) found simultaneous diur- of individual peaks or troughs wlth the initial value at nal variations in plasma and endolymph [tot CO2]. 14:00 h. Signijicance was attributed at p < 0.05. The present study was undertaken to determine, by sampling both distaP and proximal zones, whether this eiiJvlj-r,ph hE:e;ogccci:). ir, rLcin!einc2 tIyCnnhrn>t3----- RFSI TI.TS the d~urnalcycle. Rhythms of otolith calcification re- flect the chemical compositions oI plasma and endo- Plasma cortisol concentrations (mean 42.8 i 5.32 ng lynlph during the daily cycle. Diurnal profiles of plasma ml-') did not vary significantly during the diurnal and endolymph concentrations of protein, carbon di- cycle. Values at the beginning (14:OO h) and at the end oxide, sodiunl, polassium and chlorine, and also plasma of experiments (14:00h following day) were not sig- calcium concentration were therefore made. nihcantly dfferent. Edeyer et al.: Daily variations; of endolymph composition 289 To clarify the variations of endolymph composition proximal and distal sides, within the sacculus, was and their relationships to those of plasma, the data for maintained throughout the experiment. Plasma [pro- the 2 fluids are presented on the same graphs. First, tein] appeared to decline slightly during the 24 h the initial and final concentrations measured at the period and were characterised by 2 significant minima, same hour of the day were compared, and although at 23:OO h (p < 0.001) and 11:OO h (p < 0.001).Endo- some variation was observed, none of the differences lymph [protein] clearly decreased during the dark were significant. Statistical comparisons of the data period particularly in the proximal side. Between 20:OO against the reference initial value (corresponding to and 02 00 h the magnitude of variation was 65 and the midpoint of the daylight period) were made, except 58 %, for proximal and distal sides, respectively. Never- where mentioned. theless, a large scatter of the values obscured this phe- Fig. 1 shows the daily variations of the concentra- nomenon and none of the mean values was signifi- tions of monovalent ions in plasma and endolymph (distal and proximal).