Molt pattern identification through gastrolith examination on Nephrops norvegicus ( L . ) in the Mediterranean Sea *

Samples of Nephrops norvegicus were collected on a monthly basis from October 1993 to September 1995 in different areas of the Mediterranean and the Atlantic, both to investigate molt periodicity and to verify the existence of a common molt pattern for this species in the Mediterranean. The presence of well-developed gastroliths in the wall of the stomach was used as an indicator of molting periods. A well defined molting periodicity was not found among juveniles, which seem to molt all year round. In each studied area molt synchronism was evidenced among both males and females after the onset of sexual maturity, even if the seasonal molting pattern differs between the two sexes. Adult females have only one molting period per year (December-March), immediately after the hatching of eggs. Adult males present a molting period in late summer-autumn (August-October). In some areas a second molting period is evident and co-occurs with that of females (late winter-early spring). This molt synchrony among adults is common to all the Nephrops populations of the Mediterranean Sea. The different results obtained for the Atlantic population could be due to the influence of environmental factors on molt cycle regulation.


INTRODUCTION
In Crustaceans, growth is a discontinuous process with a succession of molts separated by intermolt periods.During each molt the old exoskeleton is shed and the animal grows very quickly, before the new exoskeleton hardens.Two components of growth have been recognized (Hartnoll, 1982): the increase in size at each molt (growth factor) and the time between two successive molts (intermolt period).To identify molting periods and periodicity is of great importance in growth studies; in some decapods (i.e.lobsters) this is facilitated by the presence of gastroliths.In species with a very hard exoskeleton and to which calcium is not readily available in the environment, calcium is stored in these transitory structures.Gastroliths are situated in the wall of the cardiac chamber of the stomach, between the epithelium and the cuticularized layer.Here, during pre-molt phases, calcium, mobilized from the old exoskeleton to the blood by the molt hormone, is stored as calcium carbonate crystals (calcite) (Sardà, 1981;Skinner, 1985).After ecdysis, gastroliths are quickly dissolved and the calcium is used to harden the new exoskeleton.
The duration of the molt cycle (from proecdysis to anecdysis) differs among species and is highly influenced by environmental factors (in particular, temperature and light) which affect molt hormone secretion (Skinner, 1985).As a consequence, the period in which gastroliths are present is variable.To define the duration of molt phases, identified in accordance with the Drach (1939) scale, the structural changes in pleopods of lobsters were studied by Aiken (1973) on Homarus americanus, and by Sardà (1983) on Nephrops norvegicus.
Experiments on the variation in calcium concentration in different body structures indicate that in N. norvegicus the maximum increment in gastrolith volume takes place during last phases of premolt (D2 to D4).Gastroliths reach full development about 15 days before molt, and are reabsorbed very quickly (less than 24 hours) after ecdysis (Sardà and Cros, 1984).
Therefore, the presence in this species of welldeveloped gastroliths indicates that the animal is ready to molt within a few days.The use of gastroliths as indicators of molting periods can also support results obtained with other techniques such as percentage of 'soft' specimens in size frequency distributions (Conan, 1975;Sardà, 1991), or studies in aquaria (Thomas, 1965;Conan, 1978;Sardà, 1985).
The aim of this paper, based on samples from different areas of the Mediterranean and the Atlantic coast, was to improve knowledge on molting periodicity and to compare the molting pattern among different Nephrops populations.

MATERIAL AND METHODS
Samples of Nephrops norvegicus were collected monthly from October 1993 to October 1995 in some areas of the Mediterranean Sea and in the Eastern Atlantic.The areas sampled were the Aegean Sea (Euboikos Gulf), the Adriatic Sea (off Ancona), the Tyrrhenian Sea (off Elba island), the Ligurian Sea (off Genoa), the Catalan Sea (off Barcelona), the Alboran Sea (off Malaga) and the eastern Atlantic (off Faro, southern coast of Portugal).
The samples were collected by commercial bottom trawlers in all the areas, except in the Adriatic where a research vessel was used.The gears used had meshes of 40 mm (stretched) in the codend except for Greece (32 mm) and Portugal (55 mm).
The sampling techniques were different among the areas.In Portugal, Euboikos Gulf and Adriatic Sea, Nephrops catches were systematically sampled with the aim to obtain at least 10 specimens per sex and length class (each length class had a range of 5 mm CL).In the other areas, the whole catch of N. norvegicus (or a random sample if the catch was too large) was examined.Unfortunately in some areas and in some seasons it was impossible to have regular samples for rough sea, trawler unavailability, etc.The number of specimens examined (by area and sex) is reported in Table 1.
All the specimens in the sample were sexed and measured (CL) with a dial calliper to the mm below.Then the carapace was opened at stomach level to check for gastroliths; only gastroliths whose crystals were noticeable by touch were considered as 'present'.
From growth studies on the species (Conan, 1985;Sardà, 1985;Froglia and Gramitto, 1988) it is known that molt frequency is higher in juveniles than in adults, and differs between adult males and females, because females do not undergo molting during the period they are carrying external eggs.For these reasons, in data processing each sex was treated separately and for each sex three groups were formed based on information on size at sexual maturity (Orsi Relini et al., 1998).The same class limits were chosen for all the areas, except for the Alboran Sea.The first group (CL<25 mm) included young specimens, while the third one (CL≥35 mm) included the largest animals, all adults.The intermediate group (25≤CL<35 mm) could include both juveniles and adults, so that it had been excluded from further analysis.The Nephrops population sampled in the Alboran Sea seemed to be composed of larger specimens.The smallest berried female had 30 mm CL and the 100% of maturity was over 40 mm CL.Only for this area the upper limit of the first group and the lower limit of the third one were therefore moved to 30 and 40 mm CL respectively.
The rather small size of the samples examined for the Atlantic population and the unusually low gastrolith frequency make the identification of molting periods difficult.In an attempt to increase sample size, 'soft' specimens were included in gastrolith frequency calculations.'Soft' specimens are animals in the first post-ecdysis stage, whose exoskeleton is not yet completely hardened (parchment-like).
For each sex and month, the frequency of occurrence of gastroliths (F%) was calculated as the percentage of specimens with gastroliths on the total of specimens in the size range considered.For each percentage, the 95% confidence interval was also calculated.To avoid large confidence intervals due to paucity of data, only length classes with 10 or more specimens were considered.Negative values of the lower confidence limit were considered unrealistic and discarded.

Gulf of Euboikos
In juveniles, specially in females (Fig. 1F), a molting period in winter-early spring was well evidenced in both the sampling years.Other molting peaks were present in summer and early autumn, but their existence can only be supposed, because the confidence intervals are quite large.Adult males (Fig. 2A) showed a preferential molting period at the end of summer with gastrolith frequency higher than 20%.Adult females (Fig. 2G) clearly underwent only one molt per year, in winter (January-March).This molting period is not well evidenced in the second sampling year, but supported by data of the first year.Trends were very muddled in both sexes for juveniles (Fig. 1B, G), with very wide fluctuations in gastrolith frequency between months, evidenced by large confidence intervals.Only in the first sampling year, there was a molting peak in young males between June and September.The data for adult males showed a main molt peak between June and September in both years (Fig. 2B).In adult females, the frequency of gastroliths was always very low (Fig. 2H), so that little can be said on their molt cycle, other than adult females do not molt between August and January, when they carry external eggs (Froglia and Gramitto, 1981)

Frequency (%)
Nothing can be said about juveniles.The paucity of specimens sampled (Fig. 1C, H) was unsuitable for an analysis (less than 10 specimens) or resulted in very wide confidence intervals of the estimated gastrolith frequencies.Adult males showed two molting periods: spring (March-April) and late summer (August-October) (Fig. 2C).These peaks were more evident in 1995 than in 1994, when the frequency of gastroliths was quite low (generally less than 10%) in spite of the high number of specimens examined.Adult females confirmed the existence of only one molt period per year (winterearly spring) (Fig. 2I).

Catalan Sea
In juveniles of both sexes (Fig. 1D, I) gastrolith frequency fluctuated widely from month to month over the two years, without any evidence of molt periodicity.In adult males (Fig. 2D) the gastrolith frequency was always very low (less than 8%) and the confidence limits very large.It can be argued that molt occurred the year round with a possible maximum peak in September (10% in both years).Adult females (Fig. 2J) were the less abundant, but the percentage of specimens with gastroliths reached peak values in February andMarch (14-15% in 1994, 13-24% in 1995).Data collected in the first sampling year can be sustained by those of the second one.This means that, as in the other areas examined, adult females molt once per year, in late winter.The lack of data between September 1994 and January 1995 is due to paucity of adult females in the samples.It may be the result of the lower vulnerability of berried females.

Alboran Sea
Young specimens were almost completely absent in the first sampling year, and the gastrolith frequencies obtained in the second year are meaningless, due to the paucity of specimens in the samples (Fig. 1E, J).In adult males only a small peak was detectable in spring 1994 (Fig. 2E).In adult females (Fig. 2K), the trend is the classical one, with a well defined molting period, from January to April (F max > 25%, in March).Atlantic Samples obtained off Faro, were characterized by the almost complete absence of young specimens (always less than 10 specimens in each length class).Available data for adults (Fig. 2F, L) suggest a molting period extending from March to May for females and a slightly shorter one for males (from May to June).

DISCUSSION AND CONCLUSIONS
To define molting periods for juveniles is quite difficult for the wide fluctuations in gastrolith frequency (Fig. 1).Also from data presented in this paper, sex does not seem to influence molt periodicity in young specimens, confirming a common growth pattern for males and females in the juvenile phase (Hillis, 1979;Sardà, 1985;Froglia and Gramitto, 1988).Results obtained in areas where the number of specimens caught is sufficiently high (Catalan Sea and Adriatic Sea), suggest that molt synchrony does not exist among juveniles which can molt all year round (see Sardà, 1991 for the Catalan Sea).In other areas (Tyrrhenian and Alboran Sea, Atlantic) juveniles are scarce or almost absent from samples.Probably the main cause of this low catchability is the behaviour of young specimens, which spend most of their time within burrows, where they also feed.They either build their own burrows (Crnkovic, 1968) and find food by digging tunnels in the sediment (Chapman, 1980), or they can live in small tunnels linked to the burrows of adults (Chapman, 1980), feeding on remains of food caught by adults inhabiting the same burrow system (Baden et al., 1990).
Within each area, slight shifts in molt peaks of adult Nephrops are evident between the two years studied (Fig. 2).This may be due to the influence of environmental factors on molt cycle regulation (Conan, 1985;Skinner, 1985).Even so, the synchrony in seasonal molt pattern appears quite evident, particularly in adult females.Differences in molting periodicity between males and females after the onset of sexual maturity (Charuau and Conan, 1977;Froglia and Gramitto, 1988;Sardà, 1995) is also evidenced in gastrolith frequencies (Fig. 2).
Adult females molt before mating, immediately after the hatching of eggs (Conan, 1985;Sardà, 1991).In the Mediterranean they undergo molt once per year between December and March, their molting period off Faro (Atlantic) seems to shift towards spring (March-May).
The molting period of adult males is situated in late summer-autumn.In some areas (Tyrrhenian and Alboran Sea) there is also evidence for a second molting period, co-occurring with that of females in late winter-early spring.Probably this period is present also in other Mediterranean areas, but it is not so evident as the first one.In the Atlantic area, only a single period in late spring can be noted.
Data from the Ligurian Sea were too scattered to allow any analysis on molting periodicity.Even so, the available values for gastrolith frequency suggest a molting period in late spring both for males and females, as in the Atlantic area.
The existence of a common seasonal molting pattern among Nephrops populations in the Mediterranean Sea suggested by Sardà (1995) was only partially confirmed by present work for adult specimens.There is no such evidence among juveniles.
The low frequencies of gastrolith occurrence observed in most of the samples may be due to extended molting periods or to inappropriate sampling periodicity, as well-developed gastroliths are present for about 15 days before molt (Sardà and Cros, 1984).Probably, better results could be obtained with a more intense sampling frequency.
FIG. 1. -Frequency of occurrence of gastroliths (in percentage) in young males (A-E) and females (F-J) in the studied areas.Vertical bars indicate the 95% confidence interval

TABLE 1 .
-Number of specimens of Nephrops norvegicus per sex and area collected monthly for gastrolith examination.M, males; F, fema- .