Introduction

Unable to carry out their entire lifecycle in lagoons and estuaries, as sedentary species do, so-called "migratory" species make regular movements, at relatively stable dates, which are predictable from one year to the next, between the sea and the lagoon, and vice versa, or between lagoons and freshwater (Hervé and Bruslé, 1979; Lasserre, 1989; Quignard and Zaouali, 1980, 1981; Quignard, 1984, 1994). Egg laying takes place either in the sea or in freshwater, but never in lagoons or estuaries. In this introduction, we will sketch the main lines that characterize the ecology, biology and the exploitation of this guild, before providing a detailed description, species by species. The list of species retained in this context is justified by their frequency in these environments and/or due to their confirmed abundance in the few sites they occupy.

It is important to know that many erratic occupants of lagoons and estuaries, still referred to as "casual migrants", generally make short-lived incursions into these environments. These "episodic visitors" represent about 70% of the species we have mentioned (sedentary, migratory and occasional). Their frequency in the environments studied does not exceed 9% and is generally poorly represented (sometimes only one individual) (see Volume 1). Often, they are carnivorous ichthyophagous, as the Belone belone garfish, the Scombrus mackerel, etc., which chase the schools of small clupeidae Sardina pilchardus, of engraulids Engraulis sp., but also the atherine Atherina hepsetus, the red mullet Mullus sp., etc. In addition to these species, there are some exotic lessepsiens and herculeans whose presence is linked to more or less important geoecological manipulations (modifications of the channels connecting the sea to the lagoons, immersion of artificial reefs, opening of the Suez Canal, etc.) and general hydroclimatic upheaval.

From 97 fish censuses carried out in 45 Mediterranean lagoons (see Volume 1), the best-represented families of migrating fish are listed in descending order according to their occurrence in the censuses: Anguillidae, Moronidae, Mugilidae, Sparidae, Soleidae:

• - eight species are present in at least 75% of the environments considered (Anguilla anguilla, Dicentrarchus labrax, Liza aurata, Sparus aurata, Mugil cephalus, Solea solea, Chelon labrosus, Liza ramada);
• - seven species in at least 50% (Liza saliens, Diplodus annularis, D. sargus, Engraulis encrasicolus, Sarpa salpa, Belone belone, Mullus surmuletus);
• - eight species in at least 25% (Pomatoschistus minutus, Platichthys flesus, Diplodus vulgaris, Lithognathus mormyrus, Diplodus puntazzo, Mullus barbatus, Boops boops, Scomber scombrus).

It is generally accepted that migration from the sea to the lagoon is induced by trophic and "anti-predator" needs, insofar as shallow lagoons are not favorable for the penetration of large predators, and intralagoonal algal and phanerogamic vegetation make "sight hunting" difficult. Lagoon-sea "outmigration" is, in turn, induced by reproductive needs and/or thermal constraints, and probably by other unidentified causes.

The specific richness of lagoon-migrating fish varies from one environment to another and depends on:

1. 1) the species richness of the adjacent littoral zone and their morphological and behavioral type;
2. 2) the structure of the channels connecting the lagoon to the sea;
3. 3) the physical, chemical and other characteristics of lagoons;
4. 4) the hydrodynamics of sea-lagoon and lagoon-sea exchanges.

Occupation of lagoons and migration phenology

In shallow lagoons (approximately 1 m), migrants are generally 0+ young individuals, larvae and juveniles depending on the species, whereas in deep lagoons, individuals of all ages make such displacements, but 0+ are still usually the most abundant age group. Apart from intrinsic lagoon factors, the success of the recruitment of migratory species depends on the success of breeding at sea, hydrographic conditions (survival and dispersal of larvae), the topographic and architectural qualities of the communication channels between the sea and the lagoon and the extent of the volume of water coming out from the lagoons to the sea.

Only a fraction of the marine population migrates to the lagoons (Quignard, 1984; Quignard and Zaouali, 1980, 1981; Lasserre, 1989; Mercier et al., 2012), but we have no estimate of the relative importance of this "migratory phase" in relation to the original marine stock, nor do we have knowledge about the determinism and the "laws" governing fish migrations between sea and lagoons. For example, what is the share that depends on the fish and which part is related to the conditions at sea and/or the lagoon? In other words, within the same species, why do some individuals move, while others remain at their original marine territory? Within a marine or a freshwater population, are there any sedentary or nomadic "genetic lineages" that regularly extend their distribution area to lagoons? Over the past 15 years, genetic, molecular and mineralogical studies (microchemistry of otoliths) have begun to provide interesting information about Mediterranean lagoon-thalassic migrants: the visit of different nurseries by the gilt-head sea bream S. aurata (Mercier et al., 2012; Tournois et al., 2013, 2017) and by the eel A. anguilla (Panfili et al., 2012); the return periodicity to the lagoon for the same gilt-head sea bream population (Mercier et al., 2012); the independence of lagoon recruitment from nesting sites by the sole S. solea (Morat et al., 2009); differentiation (the existence of exclusive alleles in lagoon migrants) and the genetic adaptation of the gilt-head sea bream (Chaoui et al., 2012; Guinand et al., 2016) and the sea bass (Lemaire et al., 2000, 2004-2005; Guinand et al., 2015) to local lagoon conditions; etc. At the genetic level, the question is whether these divergences are premigratory or whether they result from "ongoing" natural selection processes acting on the new recruits that colonize the lagoons, causing an increase in the frequency of the alleles that allows migrants to become adapted to lagoon systems.

The return to sea (outmigration) of marine migrants is made against the current during the entry of marine waters by the inlet. These trips are made by fish of all ages having stayed for a few weeks, a few months or a few years (eel) in the lagoons. For each species, they occur in several waves of groups of individuals that are often of the same size and sometimes the same sex, during a fairly constant period, from one year to another. However, we should emphasize that returning to the sea does not always have a "reproductive" purpose, since it involves juveniles that are far from sexual maturity. Perhaps, this could be induced by autumn lagoon hydroclimatic conditions being more unfavorable than those at sea (Hotos et al., 2000; Katselis et al., 2007) or by other yet unidentified factors.

Lifespan and growth

The lifespan and size of migratory species are much higher than those of sedentaries, most of whom are annual or subannual. For example, 12-year-old gilt-head sea breams visit the Mirna estuary in Croatia (Kraljevic and Dulcic 1997), 11-year-old eels are found in the Commacchio lagoon in Italy (Rossi and Colombo, 1976) and 10-year-old sea basses live on the Tunisian coasts (Bouain, 1977). Migrants with a relatively long lagoon stay in comparison to their lifespan, such as the A. anguilla, the common goby P. minutus, two semelparous species, reach their maximum size in lagoons, before moving to lay at sea. As for other migratory species, likely to carry out several relatively short lagoon stays throughout their life, the comparative approach regarding the share of their growth at sea and in the lagoon, for the 0+ and especially for adults, is harder to pin down. Microchemical techniques applied to otoliths could teach us more.

The growth performance of migratory species is often judged to be better in lagoons than at sea, an opinion which is not shared by all researchers, some of whom describe these environments as "deadly traps" (Boutière, 1974) and even as "places of death" (Chauvet, 1986). Differences in growth also exist between neighboring lagoons (Bruslé and Cambrony 1992, Cambrony 1983, Quignard et al., 1984; Mosconi and Chauvet, 1990; Isnard, 2015), and even within the same lagoon among different biogeographical sectors (Escalas et al., 2015). Taking into account the "adaptive strategies" deployed by the species occupying lagoon ecosystems (Amanieu and Lasserre, 1982), the chances for better survival and for better growth vary considerably depending on lagoons, intra-lagoonal sites (marine sectors and continental sectors) and on the years (Amanieu, 1973). As a result, no "uniform" scenario can be drawn and the "lagoon advantage" regarding individual growth cannot always be retained. Comparative data between "sea-lagoons", carried out at the same time (same period, same year), at marine and lagoon sites close to each other, are very little documented or entirely missing. Nevertheless, we can observe that certain lagoons or intralagoonal zones, especially those rich in...