3
Biology and Genetics

As we have seen above, lagoons can be regarded as “extreme environments”. A limited number of species live there permanently (sedentaries), but a greater number come and stay for varying lengths of time, often on an annual basis (migrators). Whether they are lagoon-dwelling sedentaries, sea-lagoon migrators or river-lagoon migrators, these fishes generally represent only a proportion of specific marine or freshwater stocks. There are only five species that can be considered endemic or subendemic to lagoons: Aphanius iberus, Aphanius dispar, Aphanius fasciatus, Symphodus cinereus staitii and Atherina lagunae.

All of the sedentaries come from very coastal demersal marine species, and thus are preadapted to life in challenging environments. They are all euryvalent and they protect their spawn (non-planktonic eggs) from the negative contingencies in lagoons. All of the migrators (with the exception of Pomatoschistus minutus) are nectonic coastal fishes, relatively euryvalent, but not capable of taking care of their eggs, which are planktonic. Their spawning takes place in the sea where hydroclimatic conditions are less restrictive, and where they come together on the spawning grounds with conspecific marine sedentary spawning adults. From these factors, we can deduce that long-term or brief stays in lagoons are not essential for the survival of populations and species, and that the latter present no morphoanatomical and/or biophysical-ethological characteristics compared to individuals from the marine populations of origin. What, then, are the factors that encourage or discourage certain individuals from coming to live in a lagoon area, and what advantages do they derive from it? The biological traits reviewed below may perhaps give some clues that will enable us to better understand the originality of the fishes in the Mediterranean lagoons. It should be borne in mind that genetic scientists are still trying to find the “candidate gene” that governs migratory behavior.

3.1. Sexuality

Mediterranean marine sedentary fishes are all gonochoric (Atherinidae, Blenniidae, Cyprinodontidae, Gasterosteidae, Gobiidae, Labridae, Poeciliidae, Syngnathidae) and reach their first sexual maturity around the age of 12 months at the latest (Tsikliras and Stergiou, 2015). Sexual dimorphism (color, size) is particularly marked in the polygynandric nest-building species, gynovivipares and androvivipares. It should be remembered that in the Gobiidae family, some non-Mediterranean species are “bi-directional” hermaphrodites and that in Labridae, cases of protogynous hermaphrodites occur frequently in every ocean, including the Mediterranean. Sea-lagoon migrators are also gonochoric, with the exception of the gilthead bream Sparus aurata, which is a protrandric hermaphrodite (Figure 3.1), as are many other Sparidae that visit lagoons on a more or less regular basis (Diplodus sargus, Lithognathus mormyrus, etc.) and their sexual dimorphism, which is barely noticeable, is limited to differences in the maximum sizes attained.

images — **Figure 3.1.** *Genital gland of the hermaphrodite fish Sparus aurata (J.P. Quignard). For a color version of this figure, see www.iste.co.uk/kara/fishes1.zip*

At the family level, in Atherinidae, fishes with marine affinities regarded as seeking out brackish and fresh water, all individuals of the same sex present the same morphoanatomy and the same behavior. Secondary sexual dimorphism in adults has implications in terms of maximum size, which is slightly larger in females, and, during the reproductive period, the more rounded appearance of the female’s abdomen due to ovary development.

In Cyprinodontidae of the genus Aphanius and the Gasterosteidae Gasterosteus aculeatus (stickleback), which are typically potamo-limnic fishes with affinities to brackish waters, the females are very noticeably larger than the males. However, these fishes do not have the same reproductive behavior. Stickleback males build and guard very elaborate nests, whereas Aphanius males do not build nests, but pair briefly for the duration of the spawning act on a spawning site among vegetation. In both cases, the males, especially during the reproductive period, are more brightly colored than the females. Pairing takes place after a sexual parade.

In Gobiidae, Labridae and Blenniidae, things are completely different. In these fishes, which have clear marine affinities, the nest-building males are larger than the females, but other types of non-nest-building males (sneakers, satellites), smaller than nest-building males, can also exist (Taborsky, 2001; Mazzoldi et al., 2002b; Ruchon et al., 1995, 1999).

Certain large males, which we term “type 1”, prepare and guard a spawning nest and keep watch over a certain area (territory) around this nest, within which, usually, several females will come to lay their eggs. These dominant or “bourgeois” males attain a larger size than the females and present more radiant colors than the females (during the reproductive season at least). In addition, they have a number of attributes of their own: a cephalic crest, anal glands in the peacock blenny Salaria pavo (Figure 3.2), trapezoidal genital papilla in all gobies, and also a well-developed, very tall dorsal fin, in the black goby Gobius niger. The testicles of these males are little to moderately developed.

The males we term “type 2” are not territorial and do not own nests. These subordinate males are smaller in size than bourgeois males, have a duller hue, of the “female” type and present no visible masculine attributes. In their small stature (size), their morphology (rounded belly like mature females, due to very pronounced testicular development) and their coloring, they resemble females (mimicry). There can even be “chemical silence” (no sexual pheromones released) in these males. With their female appearance, these males can thus “fool” dominant males, approach closely and even enter nests and ejaculate onto the ovocytes that a female, attracted by the dominant male, has just laid. These males are regarded as “fertilization thieves”, orkleptogamic (they steal from the dominant male the possibility of fertilizing the entire clutch of ovocytes), and in view of their behavior, they are generally called “sneakers”. The greater the shortage of suitable locations and/or material for nest building (lack of lamellibranch shells, rocky sites, etc.), the more secondary males there are.

In the peacock blenny (Salaria pavo), a third type of male has been described, the “satellite”; this type is larger than the sneaker and its presence is tolerated by nest-owning males, whereas the sneaker is sometimes driven out and chased away by the dominant male. The status of “sneakers” and “satellites” is not stable. These males can acquire the status of nest-owning “dominant male”, socioenvironmental conditions permitting. “Satellite” males and “sneakers” undermine the reproductive success of the bourgeois (dominant) male, but at population and even species level, fertilization thievery is a way of mitigating the consequences of potential sterility in a nest-owner and introducing mixing, which is beneficial to genetic diversity. Satellites can be found in Labridae of the genus Symphodus (Taborsky, 2001).

In Syngnathidae, all species are gonochoric. The males are morphoanatomically very clearly differentiated from the females by the presence in Nerophis sp. males of an abdominal incubation area or “open brooding area” (Gastrophori), or a brood pouch located in the tail region (Urophori) in Syngnathus sp. (Figure 3.3) and Hippocampus sp. It should be noted that it was not until the 19th Century (Ekström, 1813) that the male Syngnathidae was recognized as being the “incubatory” sex. Prior to this time, individuals with rounded bellies were cataloged as females. Sexual behavior is generally reversed (Jones et al., 1999), with both courting behavior and coloring being adjusted according to the sex ratio, i.e. the shortage of one sex in relation to the other.

Visual signals play an important role in sexual recognition and selection; scent, at least in the case of Syngnathus typhle, appears not to be involved in the seeking out and thus the choosing of females by the male (Linddqvist et al., 2011), but hearing may play a part. With regard to the maximum sizes of males and females, the available data are often contradictory. In seahorses and the syngnathid Sygnathus abaster, the maximum sizes usually appear to be very similar. In Syngnathus thyphle, the females are larger than the males and in Syngnathus acus both scenarios have been recorded.

Monogamy (social and genetic), which is usually linked to low individual density and low mobility of individuals, has been reported in some seahorses and a few syngnathids (Vincent et al., 1992; Sogabe et al., 2008; Mobley et al., 2011a; Rosenqviste and Berglund, 2011), but has sometimes been disputed (Whiteman and Côté, 2004; Weiss, 2010). Be that as it may, in Syngnathidae, and also in certain nest-building fishes, monogamy (for the male) can last for a spawning season or be limited to one mating (in cases where the male only receives, or only accepts the eggs of one female in his brood pouch or on the open brooding area on his body strictly biparental pregnancy), ready to mate with another female after parturition. This type of monogamy seems to be “enforced” in Nerophis ophidion (which has no brood pouch) and certain seahorses. Until recently, temporary monogyny (when a clutch consists uniquely of the ovocytes of one female, transferred into the male’s brood pouch during one or several copulatory acts) used to be a frequent occurrence, even the norm in Syngnathidae, but genetic research has revealed that polygyny often dominates in the filling of brood pouches (poly-parenting). In the long-snouted seahorse Hippocampus guttulatus, Naud et al. (2009) state that couples are ephemeral, since the male is constantly seeking to mate with the largest possible female to maximize his chances of filling his brood pouch (Weiss, 2010).

In the Poeciliidae Gambusia holbrooki, a viviparous gonochoric fish native to America, sexual dimorphism is marked. The male has a gonopod, an intromittent organ deriving from the anal fin (Figure 3.4) that, via copulation, facilitates internal fertilization of the female. This male attains a maximum size of approximately half that of the female and is more brightly colored than she is, especially during the reproductive period.

The age of first sexual maturity in migrators is very variable from one species to another and can differ according to sex and environmental conditions (Tsikliras and Stergiou, 2015). The sea bream Sparus aurata is the most remarkable among the migratory species from this point of view, since it is a hermaphrodite. From the age of 2–3 years, the age of first sexual maturity, up to the age of 4–5 years, all specimens function as males (protandrous hermaphroditism), and then gradually a certain proportion of these males undergoes a fairly rapid sex reversal (male sexual death) between two spawning seasons, and remain female until death (somatic death). According to Chauvet (1978, 1986), the ratio of sex reversal may be regulated by social pressure exercised by females upon males, this pressure being proportional to the numerical density of females; high female mortality induces a compensatory reversal of males to females. The disadvantages to the population of this type of hermaphroditism should be the imbalance, caused by natural mortality, between the number of males (all young) and the number of females (older and therefore fewer in number). However, in terms of reproduction, there is a compensation. Female fertility is proportional to size, and their size is large to ensure good reproductive success. Furthermore, since ovocyte size is also proportional to size of female, a great many large larvae are produced with the capacity to quickly become competitive. As for the males, their small size is not a handicap. Since they are numerous, natural mortality has little effect on their stock, and spermatozoa production per individual is sufficiently abundant to not be a limiting factor in fertilization.

3.2. Reproduction

Lagoon and estuarine fishes reproduce from the end of winter to the beginning of fall, usually quite early so that the neonates are large enough and have sufficient reserves to withstand winter constraints. In most cases, for sedentaries, first sexual maturity is reached after the first winter following their birth, but in some species such as Syngnathus abaster, Gambusia sp. and Aphanius sp., it can be earlier, occurring a few weeks or months after hatching; this means that at least two generations can be born during a single spawning season. Sedentary females have relatively low fecundity: several dozen or occasionally several hundred ovocytes per egg-laying act, each of them large in size (>1 mm in diameter), and several hundred or even, exceptionally, several thousand per spawning season. Migratory females are very fecund (with the exception of females of the goby Pomatoschistus minutus, a nest-building fish): from several thousand up to millions of small ovocytes (≈1 mm in diameter). In all these species, but especially in the sedentaries, intrasexual competition and male versus female, as well as female versus male, sexual selection are very intense and involve all of the sensory functions (sight, hearing, smell, etc.) (see section 3.5).

3.2.1. Nest building, gestation and fecundity

Except in the case of species of the geni Aphanius and Atherina which abandon their eggs after fertilization and females of the genus Gambusia which are viviparous, Labridae, Blenniidae, Gobiidae, Gasterosteidae and Syngnathidae have nests in the broader sense of the word (spawning nest, brood pouch or abdominal open brooding area on the outer surface of the body) and provide “parental care” of varying degrees of sophistication to their eggs. Only the stickleback Gasterosteus aculeatus (Gasterosteus gymnurus) provides a degree of protection to the neonates for a few days after hatching, before their “emancipation”. Nesting in its broad sense is viewed by bioethologists as a function of gestation. Several types can be identified, namely “open” external gestation (nest of the Labridae Symphodus staitii, abdominal open brooding area of Nerophis ophidion), semiclosed external gestation (nests of Gobius sp., Pomatoschistus sp., Zosterisessor ophiocephalus, Knipowitschia sp. and Gasterosteus sp.), and internal gestation (brood pouches of Syngnathus sp. and Hippocampus sp. males, genital equipment of Gambusia sp. females, Figure 3.5).

All these types of “gestation” have implications for the “operational” fecundity of males and sometimes that of females (Avise and Liu, 2010). In fact, “nest-building/gestation” brings physical constraints in terms of the space (surface area or volume) provided to receive the egg clutches (the size of the nest, of the surface area on the body and of the incubatory organs is always limited), and this drastically reduces the chances of realizing potential fecundity. In addition, for females as well as for nest-building males, there are two other physical constraints: one linked to the volume of the mature, thus relatively large eggs (over 1 mm in diameter), which can increase with the size of the female but decrease between the beginning and end of the spawning season; the other linked to their shape. In fact, among species where eggs are attached to substrates (gobies, blennies), an oblong goby egg attached to a substrate takes up less surface area than would a spherical egg of the same volume. Often, the number of ovocytes a female can attach is lower than the number she can produce, because her size increases at the point of release (final hydration) and sometimes during embrogenesis (Avise and Liu, 2010). To these constraints, we must of course add those caused by a possible imbalance of the operational sex ratio and the shortage of space and material needed for nest building.

Species that practice external, semiexternal or internal nest building will often, in order to partially compensate for the disadvantages of nest building, be polygynandrous: one male mates with several females and one female spawns with several males in the course of one reproductive episode during the reproductive season. In such cases, no long-lasting pairs are formed except, perhaps, in a few seahorses. Usually, among nest-building males in the broad sense of the term, the number of times a male mates with different females (polygyny) is necessarily limited by the size of the nest or brood pouch and by the degree of immediate fertility (the quantity of mature ovocytes likely to be laid at the moment of mating) of the females admitted successively to lay their eggs. Moreover, during the external guarding or internal incubation phase following egg laying, the nest or brood pouch being “full”, the male is no longer able to perform “legitimate” fertilizations (long incubation period, sometimes several weeks). However, males with open or semi-closed nests may be “guilty” of acts of sexual piracy toward neighboring nests (fertilization thievery), although that is impossible in Syngnathidae (Kvarnemo and Simmons, 2004) which are all androviviparous, and likewise in Nerophis ophidion (Ah-King et al., 2006) even though the ovocytes are directly accessible since they are not contained in a brood pouch (only one case, reported by McCoy et al., 2001, of a clutch fertilized in part by a male other than the pregnant father).

Seasonal fertility of nest-building males (sometimes called “nest fecundity” in Gobiidae, Blenniidae, etc.) or gestating males (Syngnathidae) is, moreover, limited by the duration of the incubation period (from several days to several weeks) which, it should be noted, varies according to egg size but above all according to temperature, which is a very variable factor in lagoons. From this point of view, the impact of nest building on the potential fecundity of females is less drastic.

In fact, no matter what the type and size of the nest or brood pouch of the male, or how full it is, a female can always expect to find a nest or brood pouch ready to receive her eggs, except of course in the eventuality of a severe shortage of operational males. It should be noted that in all cases of nest building and androviviparity, particularly in syngnathids, this shortage sets in rapidly from the beginning of the spawning season. The number of receptive males diminishes and “vacant” nests or brood pouches soon become increasingly rare. The operational sex ratio, though it may have been in balance outside of the spawning season, thus becomes biased toward females, and this leads to competition between them for access to males. The males, confronted by an overdensity of females seeking a spawning male, employ drastic selection toward them (Vincent et al., 1994). In the gynoviviparous Gambusia sp., the individual fecundity of each female is determined by her morphoanatomical characteristics, principally her size (ovarian fecundity and the abdominal space reserved for gestation being proportional to the female’s size) and, exceptionally, by the fertilizing activity of the males. Moreover, she will mate, voluntarily or involuntarily (sexual harassment), with more than one male. As well as the size of the spawning female, the length of the gestation period, which correlates to ambient temperature, is a factor that encourages or limits the realization of potential fecundity of mosquitofish during the spawning season. In the case of mosquitofish, male fecundity may be limited even though they do not take charge of their offspring, as a consequence of a lack of receptive females (gestating or immature females). This shortage is the factor that gives rise to the aforementioned sexual harassment.

On the other hand, the fecundity of some males can be limited by postcopulatory selection practiced by the females of certain species among the spermatozoa they have received from the various males they have copulated with. This postcopulatory selection applies not only to spermatozoa but also, in some species, to ovocytes. In fact, nest-building males of Gobiidae, Labridae, Gasterosteidae, etc., eliminate unhealthy eggs and often prevent egg-overload in their nest by consuming part of the clutch (filial cannibalism), and likewise when they have a requirement for high energy levels. In syngnathids, cannibalism, be it filial or not, is little practiced and only neonates are targeted (Franzoi et al., 1993; Cunha et al., 2015–2016), but, perhaps uniquely in the animal world, males “consume” some eggs in their marsupium without taking them out. This intramarsupial digestion of eggs might be viewed as an “original” form of cannibalism. In fact, the male may be able to identify, among the eggs in his brood pouch, those that came from an undesirable female (usually a small female) with whom he mated “accidentally” (Paczolt and Jones, 2010; Fang, 2010; Mobley et al., 2011a). The fact that a male is able to choose, after having consented to copulation, not to fully commit himself to ensuring the survival of all the ovocytes entrusted to him is an “intriguing” fact (Fang, 2010). It has been shown that the nutrients released by lysis of some of the eggs inside the marsupium were used during development by the surviving eggs (embryos), and also by the pregnant male (Ahnesjö, 1996; Ripley and Foran, 2006, 2009; Sagebakken et al., 2010). Regardless of the nest-building species and type of nesting (Labridae, Gobiidae, Bleniidae, Syngnathidae), the male monitors and even controls, more often than not, the present and future reproductive act and success of spawning adults. This control is always exercised to the detriment of the reproductive success of one or more females during the precopulatory (rejection of females) and postcopulatory (removal of certain unwanted eggs) phases. In many instances, the male's oophagous behavior, representing filial cannibalism, is effected primarily for his own benefit, since it keeps him in good physical condition and enables him to successfully achieve at least one further “gestation” of eggs from a female who is (perhaps) more attractive and of better quality. It thus favors its own reproductive success, in the short, medium and longer term, over that of one or more females, since it exploits and hijacks part or all of the reproductive efforts of these females for its own benefit. Remarkably, this hijacking can have the knock-on effect of benefiting the reproductive success of one or more future partners of higher quality than the undesirable partners whose eggs he has consumed. Overall, the male’s behavior can on the whole be regarded as beneficial to the population’s reproductive success.

In all instances of “guarded” gestation, we can accept that the reproductive success of the individual and population is based more on seeking to maintain genetic diversity through multiple matings than on parental fecundity (Avise and Liu, 2010); environment-dependent parental care (oxygenation, cleaning the brooding area, etc.) is, furthermore, equivalent. As has already been mentioned, long-lasting pairings may form in Syngnathidae, lasting for one spawning season or for life. Pairing seems to be largely a result of an acute shortage of one or the other sex, but above all a severe shortage of females (Whiteman and Côté, 2004). Permanent pairing has no advantages in terms of individual or populational reproductive success, because it limits the fecundity of one of the partners or of both sexes as the case may be, and limits the potential for gene mixing and thus genetic variability within the population, but it avoids the need for spawning adults to look for a partner after each spawning, which is a waste of time and energy.

Nest quality is evaluated on its general appearance, and also its sanitary qualities. To maintain it in good condition, the males secrete antibiotic and antifungal substances that are contained in the adhesive protein (spiggin) secreted by the kidneys and used to stick together the nest-making materials in the case of the stickleback, in secretions from the anal glands of the peacock blenny and in the sperm trails of gobies.

3.2.2. Reproductive success and gamete management

As mentioned above, reproductive success in nest-building lagoondwelling fishes is dependent on a set of very diverse factors. While it is generally accepted that the development of secondary sexual traits is linked to intersexual selection (Anderson, 1994), the correlation between their quality and the success of reproductive mating opportunities does not always follow a clear pattern (Alonzo, 2008). The examples we have considered above suggest that reproductive success is also subject to many other environmental, ecological, social and behavioral factors. Among these factors, in species where one of the spawning adults practices “parental care” (guarding), the quality of this is a determining criterion in the choice of partner. Quality is often perceived as being linked to the size and color of the guardian, the intensity of the courtship behavior and a factor that is more complex to define, his/her “experience”.

According to some authors (Bruslé and Quignard, 2004, 2012), females can tell whether they are in the presence of an “experienced” male by the appearance of his nest, and also by the presence of eggs in it. Experiments have shown that females prefer to lay eggs in a nest that already contains eggs. This preference may be induced by a basic “copycat” tendency (if one or more females have already chosen to mate with this male, it shows that the individual and his nest are of good quality), but the female’s choice may have been determined by other objectives. In laying her eggs in a nest already containing eggs, the female can expect a good survival rate of her progeny by diluting the predation risks from external egg predators or from the guardian male (filial cannibalism). The female may in fact consider that this male has already “had his fill” of eggs from previous clutches (Forsgren et al., 1996). While empty or almost empty nests are “suspect”, over-full nests are also avoided by the females; the risk of infection and anoxia increasing with egg density (confinement, poor water circulation). Also, the risk of being unable to deposit a sufficient number of ovocytes there, so as not to have to go and find another nest to finish laying (minimizing energy expenditure), is doubtless also taken into consideration. Adding ovocytes to a nest that already contains many eggs can lead to the male eating eggs to reduce the density (filial cannibalism) in order to avoid the aforementioned issues. The survival of clutches of eggs laid in nests is to a large extent “density-dependent” (Klug et al., 2006) and cannibalism can save lives. A very special case of “intra marsupial egg cannibalism” to assure certain reproductive success has been described in syngnathids (see sections 3.2.1 and 3.2.4). However, nests containing “old eggs” close to hatching are equally shunned by females (Sikkel, 1989; Matsumoto et al., 2011). In fact, the presence of old eggs can be interpreted by the female thus: that she has found an “old nest”, potentially in poor condition, and also that the male “approaching the end of guarding”, who is himself probably in a poor physiological state, may be attracted by fresh eggs (filial cannibalism) with high nutritional value (Manica, 2002a). There is the additional risk that the nest may be abandoned after the hatching of the “old eggs”. It should be noted that, in experiments, Matsumoto et al. (2011) were not able to show that Rhabdoblennius ellipes (Blenniidae) females were capable of distinguishing a “young egg” 0–2 days old, from an “old egg” 3–5 days old.

Reproductive success is linked to the quality of the spawning adults and nests, to the female’s egg management (usually she does not put them all in the same nest or brood pouch) and to “management by selection” of the spermatozoa stored in her genital passages (postcopulatory selection), for instance in Poeciliidae (Chapman et al., 2003; Chapman, 2009; Bruslé and Quignard, 2012). Sperm management by the male is of equal importance, since the quantity of sperm emitted depends on the female’s “appearance” or on the level of competition between males. Data concerning these aspects of reproduction are non-existent for the non-nest-building sedentary species (Aphanius sp., Atherina sp.,) and sparse for Syngnathidae. However, gobies have been the subject of much research. To ensure a maximum fertilization rate, while at the same time ensuring quasi-permanent guarding of his nest against egg predators and “fertilization thieves”, i.e. sneakers (also called cuckolders or kleptogamic males), the male who owns a spawning nest releases a layer of viscous sperm (sperm trail) on the substrate, before (or sometimes at the same time as) the females come to attach their eggs (Mazzoldi et al., 2005). Since spermatozoa are released progressively from this coating, over several hours, the females’ ovocytes (egg laying by the female can last for several hours according to size and species) can be fertilized even during the male’s absence. The male can, with his reproductive success virtually guaranteed, attend to other activities that are beneficial to nest quality, and also to staying in good physical shape (finding food). These activities all have favorable consequences for the overall reproductive success of the pair(s). The sperm of the peacock blenny (Salaria pavo) is also viscous, but we have no data on how it is used. In sticklebacks, sperm management may also influence the quality of sperm emitted dependent on the presence of rival males, and the timing of sperm emission (Zbinden et al., 2003). If there is a lot of competition, the male stickleback ejaculates in his nest before the eggs are laid (the precautionary principle). The ovocytes of sticklebacks appear to be covered in a mucus that prolongs the life of spermatozoa.

In all sedentary lagoon-dwelling species, egg laying is fragmented (a number of egg-laying acts) throughout the reproductive season, regardless of whether the female spawns annually (semelparity with fragmented egg laying) or multiannually (iteroparity with fragmented seasonal egg laying). Thus, sedentary lagoon-dwelling females do not put all their eggs (ovocytes) “in the same basket”; they spread them out in time and space by entrusting them to a number of males. They adopt the precautionary principle, a principle that is essential in order to ensure good reproductive success in the lagoon environment, which is subject to severe and sudden hydroclimatic changes and, whatever might be said on the subject, significant predation. The splitting of clutches of eggs also has advantages for the males, since any quality defects in the ovocytes of one female can be compensated for (only partially, because poor quality ovocytes take up space, unless the male eliminates them) by the quality of those from another female. Fecundity per clutch of eggs laid is low: from a few dozen to a few hundred, rarely more than 1,000 ovocytes, relatively large in size (generally more than 1–1.5 mm in diameter). We have little information on absolute fecundity (number of ovocytes released by a female over a laying season). Overall, we can accept that it is low, but when taken in conjunction with body mass as per the gonadosomatic index (ratio of the mass of mature ovaries in relation to somatic mass), energy use is considerable. For example, in the annually spawning goby Pomatoschistus microps a female can lay an egg mass equivalent to twice (possibly more) her body mass.

3.2.3. Reproductive particularities in Blenniidae, Gobiidae and Labridae

In lagoon-dwelling Blenniidae (Ruchon et al., 1995), Gobiidae (Mazzoldi and Rasotto, 2002a) and Labridae, reproductive success is not confined solely to the meeting of a male and a female, but can involve several types of males whose appearance and behavior differ from one to another, and whose functions are more or less complementary. There are generally two types of male, sometimes three:

– type 1: the bourgeois male, a dominant, nest-building and nest-owning male. He is larger than the females and, during the spawning season, his coloring is brighter than theirs. His testicles are relatively undeveloped, compared to those of a sneaker;
– type 2: the sneaker, a male dominated by the bourgeois male, who has no nest and who seeks, by trickery, to fertilize the eggs deposited by females in the nest of a bourgeois male. The “kleptogamic” sneaker has the morphological and chromatic appearance (mimicry) of a female (small in size; rounded stomach like a female ready to spawn, due to very well-developed testicles; a relatively drab color like that of females) and produces altered pheromones (chemical silence). Certain glands that are active in the bourgeois male (for example the anal glands in the peacock blenny Salaria pavo, the seminal vesicles and mesorchial glands annexed to the testicles of Gobiidae that secrete pheromonal steroids emitted into the water in urine and ejaculates) are not functional or not fully functional in sneakers. Furthermore, according to Locatello et al. (2002), the chemical structure of these glucronic steroids is certainly altered, making them undetectable to nest-building males. These authors have shown, through experiments, that nest-building Gobius niger males do not “respond” if they are brought into contact with the ejaculates of sneakers, whereas they do react to the presence of the ejaculates of other nest-building males;

– type 3: the satellite, a larger male than the sneaker, with the coloring but not the morphology of a female. The satellite, who sometimes seems to be socially castrated (although cases of parasitic fertilization by these satellites have been reported), is accepted in the male bourgeois domain, unlike the sneaker who is always equipped with hypertrophic testicles. The latter is a potentially formidable competitor (sperm competition) in terms of fertilization.

Type 2 and 3 males are not genetically determined. A sneaker, depending on socio-environmental context, can become a bourgeois male. Moreover, they are relatively more numerous when there is a significant shortage of nests, which highlights that sneaker status is linked to the impossibility of acquiring a nest and that the only other way to reproduce is to opt for “sneakage”. In terms of the population, the value of “sneakage” is limited to cases of sexual inadequacy of the bourgeois male, but we should not underestimate the fact that sneakers increase the genetic variability of the neonates produced from one clutch of eggs.

3.2.4. Reproductive particularities in Syngnathidae

Syngnathidae are remarkable in that the males incubate the eggs on or in a corporeal structure of greater or lesser complexity (marsupium, brooding area), but they are also unique in terms of their genital glands and more particularly their ovaries, whose morphohistological structure and mode of egg production, which varies from one genus to another, impact on the reproductive behavior of spawning adults (Sogabe et al., 2008; Sogabe and Annesjö, 2011).

With regard to the testicles, Syngnathidae present a very particular gonadic structure and method of spermatozoa production: each testicle is formed of a single hollow lobule and spermiation occurs early (semicystic). Spermatides are released into the tubular lumen from giant multinucleoid cells, which each yield four spermatozoa (Carcupino et al., 1999; Dzyuba et al., 2008; Biagi et al., 2014; Piras et al., 2015a, 2015b). The ovary of species of the genus Syngnathus is a tube that has a longitudinal dorsal germinal bulge (germinal crest) with oogonial then oocytic follicular cells, each forming a layer stretching from the dorsal end of the bulge to the opposite ventral end where ovulation takes place (Begovac and Wallace, 1987; Wallace and Selman, 1990; Sogabe et al., 2008 Sogabe and Ahnesjö, 2011). In species of the geni Hippocampus and Nerophis, there are two longitudinal germinal bulges, and reproductive behavior is connected to the ovarian type (Selman et al., 1991; Sogabe et al., 2008; Sogabe and Ahnesjö, 2011. In fact, seahorses like pipefish which have ovaries equipped with two germinal bulges produce successive “waves” of ripe ovocytes at well-separated time intervals (sychronic production). The frequency distribution of the intraovarian ovocytes, depending on their size, is at most bimodal: one group of ripe or maturing ovocytes and one group of small, immature cells. In the latter case, after depositing her ovocytes in her partner’s pouch (seahorse and sygnathid-pipefish) or on the abdominal brooding area (nerophis-pipefish), the female must wait until the end of gestation (around 15 days) to spawn again with the same male in the case of certain monogamous seahorses, but even if she finds another available male, whether she is monogamous or not, she cannot mate with him immediately or soon after the previous mating, because it takes her between 10 and 19 days to mature a new wave of ovocytes. In both seahorses and nerophispipefish, the brood pouch and abdominal open brooding area are filled at mating by one female (“enforced” intermittent masculine monogamy). The length of the ovocyte maturation cycle in females and the gestation cycle in males are, in these two Syngnathidae, limiting factors in the reproductive success of females and males alike. This is compounded for the females by a potential shortage of males and vice versa. In nerophis-pipefish, females are larger than males and their fecundity is sufficient for one female to provide a male with enough ovocytes to cover his brooding area.

It is not known what happens to any surplus ovocytes that may be produced by the female. Three hypotheses may be envisaged:

– they may be released into the open water at the end of copulation and thus lost;

– they may remain in the ovarian lumen where they will gradually be joined by new ovulations;
– or they may be lysed and recycled by the female to meet her vital energy requirements.

The positive aspect of monogamy is the time and energy saved by not looking for a partner. According to Sogabe and Yanagisawa (2007), while temperature is the main factor governing the length of the gestation period and of ovocyte maturation, the two remain well synchronized even if the temperature varies. Syngnathidae of the genus Syngnathus are polygamous (polygynandry). In their case, the ovaries have only one germinal bulge (Sogabe et al., 2008; Sogabe and Ahnesjö, 2011) and production of ripe ovocytes is continuous, “asynchronous production” (multimodal distribution of intraovarian ovocytes according to size), which allows one female to mate successively and with virtually no time lapse with a number of males and to transfer a very variable number of ovocytes to each of them (she does not put all her ovocytes “in the same basket”, which limits failures) according to the number of ripe eggs she has available and the “requirements” of the male (full or less full pouch). In this case, only the number of available males and the size of their brood pouch limit the females’ operational fecundity (Berglund et al., 1988; Sogabe et al., 2008; Sogabe and Ahnesjö, 2011); the functional fecundity of the males is limited by the size of the brood pouch and length of the gestation period, and rarely by a shortage of females (females essentially the same size as males and sex ratio balanced or biased in favor of females). It should be noted that in monogamous fishes, as for polygamous fishes, the intergestation rest period is very short, sometimes less than 1 h (Silva et al., 2006) or up to 1 or 2 days at most (Matsumoto and Yanagisawa, 2001).

The modalities concerning the meeting of gametes in Syngnathidae are still inadequately explained or even not known (Ah-King et al., 2006; Goncalves et al., 2011). With regard to Syngnathidae (genus Syngnathus), there is still a great deal of uncertainty despite Lafont having described, in 1869, the mating of syngnathids in the Arcachon aquarium (France). According to Paczolt (see Fang, 2010), the female usually courts the male. A couple forms and a dance begins: the partners give each other little taps and wind themselves into a spiral (double helix). While they are thus intertwined, the female transfers her ovocytes in two rows (single layer) along the ventral surface of the male. The male then fertilizes these ovocytes and the two lateral flaps of flesh, located one each side of the layer of eggs, come together and “seal” along the median line of the body, thus forming a brood pouch that will open only to expel the neonates. This description, suggesting that the male’s brood pouch closes after the ovocytes have been deposited, is not generally accepted. According to Sagebakken et al. (2011), in Syngnathus typhle the female transfers her ovocytes into the male’s pouch in successive waves. This brood pouch, open only via a pore situated in the fore section, fills up progressively from the rear section to the fore section. This means that when the male mates with a number of females, the embryos from the first female are located in the posterior section and those from the last female to have deposited ovocytes in the brood pouch are in the anterior section. From this description, although the authors do not mention it, it would seem that the females deposit their eggs into a closed pouch, apart from the anterior section (pore), and there is no indication of the point at which the spermatozoa are released. Silva et al. (2006) suggest that after the “courtship” phase, the Syngnathus abaster female places her dilated urogenital papilla in an anterior opening in the marsupial pouch (pore), created by the separation of the two lateral folds of the tail forming the brood pouch (urophoric syngnathid). The sides of this opening are “visibly” tumefied. When contact is firmly established, the ovocyte transfer phase begins and lasts for between six and 25 seconds, and then the female moves away and the male continues swimming slowly, his body shaken by violent contractions, before returning to the substrate to rest. After a 15–20 min “recuperation period”, a new “courtship-coupling-spawning cycle” begins, usually with another female. Despite never having seen the male ejaculate, the authors believe, like Kvarnemo and Simmons (2004), that fertilization occurs in the marsupium after the transfer of ovocytes – but how does that happen given that there is no “visible” junction between the testicles and the marsupial pouch (Van Look et al., 2007; Dzyuba et al., 2008)? Another possibility: after the female's withdrawal (coitus interruptus), the pore of the marsupial pouch remains open for a brief moment during which the spermatozoa, released into the water by the male at that point, are “pumped” into the pouch as a result of the violent contractions that agitate him. With regard to species of the genus Nerophis, fertilization appears to take place in a cloud of sperm emitted by the male just after receiving ovocytes from a female; or else in the genital tracts of the female; or else at the exact moment when the female begins to deposit her ovocytes (Monteiro et al., 2002; Kvarnemo et al., 2004). External fertilization, while the ovocytes are being deposited, appears to be corroborated by the fact that the spermatozoa are activated not by ovarian fluid nor by sea water, but by a delicate mixture of the two (Ah-King et al., 2006) – no “sperm cloud” has been seen.

Newborn syngnathids and seahorses, which resemble adults, can be regarded as juveniles. In experiments, by inducing ingestion, via gastric intubation, of radioactive amino acids and glucose by gestating males, Kvarnemo et al. (2011) demonstrated that males supply the embryos with amino acids and probably glucose. In species of the genus Syngnathus and more particularly in Syngnathus typhle, Ahnesjö (1996) and Ripley and Foran (2006, 2009) suggest that the eggs and embryos that are “destroyed” in the brood pouch are broken down and used as nutrients for the living embryos. Sagebakken et al. (2010) show that the destroyed eggs in the marsupial pouch are also beneficial to the pregnant father, helping him keep himself in good physical condition through to the end of gestation. Thus, the brood pouch is not simply a sac containing eggs and embryos, but an actual “organ” with multiple functions. It provides the eggs and embryos with protection, oxygenation, nutrition, elimination of undesirable metabolites, and osmoregulation. It should be noted that neonates from eggs from species of the genus Nerophis, which are in direct contact with the environment and which do not appear to benefit from paternal nutrient input, are, upon hatching, at a less advanced stage of development than those of the geni Hippocampus and Syngnathus and close to those of Atherinidae, Gobiidae and Blenniidae. Another point of divergence is that in this family, “sneaker” status does not exist. It is excluded by the mating system; another male cannot introduce spermatozoa into the pouch of a pregnant male (to do that, the male would have to be equipped with a penis, as an equivalent of the female’s ovipositers). The possibility of this kind of “sneakage” could be envisaged in Nerophis ophidion, since the ovocytes are not protected after being attached to the body of the male, but visually and genetically the presence of a sneaker has never (except just once, McCoy et al., 2001) been identified (McCoy et al., 2001; Avise et al., 2002; Ah-King et al., 2006). For syngnathid males, the sexual rest period between giving birth and the start of a new gestation is often a few hours, sometimes even less than 1 h and more infrequently 1 or 2 days.

Most often, in Syngnathidae of the geni Syngnathus such as Syngnathus abaster and Hippocampus sp., it is the female who courts the male with whom she wishes to mate. This behavior is reinforced if males are the major limiting factor in the reproductive success of females. This can happen as a result of a numerical shortage of males, or by an excessive presence of small and therefore not very “marsupium-fecund” males; the size of their brood pouch being in proportion to body size. To this, we can add the length of the incubation period, often almost two weeks, although this can vary depending on temperature. The courtship consists of dance movements and a series of upward and downward movements in the water column, above the phanerogame meadows (posidonia, zosteria) that provide their habitats. “Role inversion” can occur (Berglund et al., 1986). Thus, the courtship behavior of females of the Syngnathus Syngnathus abaster depends on the number of accessible males in the immediate environment. There is a return to a “conventional” pattern of behavior in cases where there is a surplus of males of a “good size”. When they find themselves in keen competition for access to the females who have become relatively rare, the males take on ornate liveries and court them in order to attract them with the goal of copulation.

It is widely accepted that copulating couples are not different in size, and that the male receives the eggs of only one female at each spawning. This appears to be the case in Nerophis, but for the other Syngnathidae the available data are sometimes conflicting. In experiments, Paczolt and Jones (2010) demonstrated that in Sygnathus scovelli at least, the male prefers to pair and copulate with the largest female from among those available for him to choose from (precopulatory sexual selection). He therefore chooses the female who is potentially the most fertile, with the largest mature ovocytes likely to produce the finest neonates. Gonçalves et al. (2011) suggest that in Syngnathus typhle, both sexes prefer to copulate with the largest possible partner. As we have mentioned above, in Syngnathus typhle (Ahnesjö, 1996; Ripley and Foran, 2006, 2009), the number of neonates is often lower than the number of ovocytes introduced into the male’s pouch. The missing, “destroyed” eggs have been used as energy resources, either for the surviving eggs (a form of adelphophagy), or for the incubating father (a form of infanticide, filial cannibalism) or both. Paczolt and Jones (2010) identified that the number of “destroyed” eggs or embryos is higher if the copulating female is less desirable, because she is small, and thus her fecundity is low and she produces small eggs. Consequently there is, in Syngnathi, a “postcopulatory sexual selection” favoring large females, via a possible process of elimination by the male of the propalgules from unwanted matings. The mechanisms for elimination, and also those that favor survival, are unknown. It seems most probable that postcopulatory selection occurs along the same lines as precopulatory selection (Paczolt and Jones, 2010) and that the pregnant male’s use of energy resources from embryos coming from undesirable females enables him to keep himself in good physical condition, in the hopes of a union with a better quality, larger female.

Moreover, according to the same authors, success in terms of embryo survival is inversely proportional to the success of the previous clutch. In other words, a “high success rate” exhausts the male incubator who then limits his efforts toward embryo survival, in order to derive as much energy as possible from them, after their death, with a view to future reproductive acts. The Syngnathus male is thus able to “manipulate” the broods entrusted to him for his own benefit, in a rather similar way to the viviparous Poeciliidae female who selects, to fertilize her ovocytes, the spermatozoa of males with whom she agrees, voluntarily or involuntarily, to copulate, or the spermatozoa of the last male with whom she had relations (Bruslé and Quignard, 2012). In terms of the genetic homogeneity or heterogeneity of the brood (monogyny or polygyny) inside the marsupial pouch of a Sygnathus male, it appears that heterogeneity resulting from a number of copulatory matings is frequent. Sagebakken et al. (2011) show in experiments that, in this case, the intrapouch survival of the embryos is higher than if the “brood” came from a single female. Moreover, they report that survival is inversely proportional to the number of embryos, and this holds true irrespective of the number of matings.

In Syngnathidae, the reproduction mode is very unusual when compared to that of migrators and that of the sedentary nest-building species (Gobius sp., Pomatoschistus sp., Gasterosteus sp., Salaria pavo). However, there is one point of convergence with these latter in that Syngnathidae males, like those of Gasterosteidae, Blenniidae and Gobiidae, are in charge of the females’ ovocytes and thus the eggs up to the release of the neonates in the aquatic environment. The divergences come from the fact that the males carry (phoresis) the eggs attached by the female to the ventral surface of their body, on the abdomen (gastrophores), with no other protection (Nerophis sp.), or deposited by one or more females (Avise and Liu, 2010; Sagebakken et al., 2011) into a pouch located on the ventral surface of the tail (urophores) known as a “brood pouch, marsupial or marsupium” (Syngnathus sp., Hippocampus sp.). In the latter case, there are metabolic exchanges between the “pregnant” father and the eggs, embryos and larvae (androviviparity, paraviviparity) contained within his marsupial pouch until the release of the subjuvenile neonates in the open water, which does not occur in nest-building species or in Nerophis. From this point of view, there is a very close convergence between the androviviparity of Syngnathus sp. and Hippocampus sp. and the gynoviviparity developed in Gambusia sp. It should be stressed here that there are at least 24 separate “forms” of viviparity in fishes, and that the most unusual is that (androviviparity) adopted by Syngnathidae (Whittington et al., 2015), which share some genes associated with s.l. viviparity with other vertebrates.

3.2.5. Other aspects of nest building and parental care

In the Aphanius, the male courts the female and draws her a little aside toward an area rich in vegetation. The female releases her ovocytes, which are fertilized immediately. The relatively large benthic eggs (≈ 2 mm in diameter), hidden in the meadow, are then abandoned. This behavior, like that of the atherine, foreshadows nest building with the practice of parental care. Atherina lagunae (Atherina boyeri), a small fish that moves in shoals, is regarded as a “group spawner” but it is possible that transient couples form and last for the time it takes for the female to attach her eggs, equipped with filaments, to erect algae (Gaillardia sp., Solieria sp., Laurencia sp., Cystoceira sp., etc.) or to phanerogames (Ruppia sp., Zostera sp.). After fertilization, the clutches are immediately abandoned (Figure 3.6). This tactic prevents the eggs from coming into contact with sediment, which is often muddy and thus rich in organic matter in the early or more advanced stages of decomposition. The quasi-permanent swaying movement of this vegetation prevents them from being covered by deposits of sediments in suspension in the water, and thus being asphyxiated. Moreover, being in contact with plants means the eggs, like those of Aphanius, can take advantage of oxygen production by photosynthesis during the day, which is not without its importance in lagoon environments that frequently present a deficiency in this gas (heat, fermentation of the bed rich in organic matter), more so by day than by night. In all other lagoon-spawning species, eggs are attached to a structure that can either be living (algae, phanerogames) or inert (mineral, metal, synthetic), and are prepared by the males (nest building) and then guarded by the males at least until hatching (paternal or parental care). Syngnathi, seahorses and mosquitofish are special cases in their practice of “parental care”.

Labridae are only present in deep and relatively marinized lagoons, such as the Thau lagoon (France). The wrasse Symphodus (Crenilabrus) cinereus staitii lives in this lagoon. The mature adult male is larger than the females and sneakers. To reproduce, he builds a nest of vegetation in a territory he defends. To do this, he picks sprigs of erect “branched” algae, a few centimeters long, and arranges them on a bed of sand and mud, interweaving them to form a cup-shaped nest of the “classic bird’s nest” type, measuring around 20–30 cm in diameter and around 15 cm in height. The nest is weighted down with sand and gravel. The male, in relatively sober “wedding” dress, waits in the vicinity of the nest but usually in the middle. He attracts females (polygyny) who take turns to come and attach their ovocytes to the algae, leaving the nest immediately after laying. The male fertilizes the eggs and is responsible for guarding them. He cleans the nest regularly, removing “by mouth” the sediments deposited on it, dead eggs, fungal mycelium growth, etc. He also “ventilates” the nest, accelerating water renewal by body and fin movements, primarily by sustained fanning of the pectoral fins. This ventilation limits fine sediment deposit and, all being well, supplies the eggs with oxygenated water. As in the case of Atherinidae, the eggs are not in contact with the floor and benefit from the oxygen released by photosynthetic activity of the algae. In this polygynous fish, spawning is performed as a couple and lasts for as long as the act takes, but since the nest is “open”, the nest-owning male may not be able to prevent some of the ovocytes, deposited by the female he selected, being fertilized by a male without a nest, a sneaker. This latter approaches the spawning couple and ejaculates an abundance of sperm into the open water (sperm cloud). The nest-owning males try to drive these sneakers away, but how can they ejaculate while simultaneously defending their nest against intruders? Moreover, sneakers present all of the characteristics of mature females: small in size, rounded stomach due to strong testicular development (abundant sperm), drab gray-brown coloring like that of females and (although it has not been proven) “chemical silence”. Sneakers are thus difficult to recognize.

In the stickleback Gasterosteus aculeatus (Gasterosteus gymnurus), the male, unlike in the other nest-building sedentary lagoon-dwelling species, is smaller than the female, but like them he is very brightly colored during the spawning season. In spite of his small size, he builds a very elaborate nest from vegetation, as does the wrasse Symphodus staitii, but the construction technique and shape are very different. The male tidies and makes a little hollow in the sediment, and fetches vegetal materials and also various bits of rubbish including plastic items (lagoons are often polluted), which he assembles by gluing them together using a glycoprotein, spiggine (Jakobsson et al., 1999), secreted by the posterior part of the kidneys and stored in the urinary bladder (Wootton, 1976). This glue possesses properties that protect the eggs from bacterial and fungal infections. The nest, the appearance of which can vary, is generally shaped like a sleeve with one opening. The male guards and maintains the nest, and looks after the eggs and even, for a time, the neonates. A number of females attracted by the male can enter the nest one after another to lay eggs there. Only after the female has departed does the male fertilize the ovocytes. These are covered with a “mucus”, produced by the female, that prolongs the life of the spermatozoa. A different male strategy has been described in regard to sticklebacks. The quantity of sperm ejaculated by the nest-owning male varies according to how many rival males there are. When competition is keen, the bourgeois male cuts short his courtship activities and proceeds to early ejaculations in his nest, even before the female is in position to lay her eggs; since his gametes are fertile for 15–20 minutes, this boosts his chances of paternity (the same process can be found in Gobiidae). When the neonates are ready to “leave the nest” (emancipation), the male “pierces” it with a number of openings, thus facilitating their departure. In sticklebacks, sneaker status exists. Nest-owning males sometimes abandon their nest and transform themselves into “nest thieves” if there is a nearby nest that they consider to be better quality than their own. Compared to other lagoon-dwelling fishes, the stickleback seems to be more opportunistic.

Gobiidae and Blenniidae males also use tactics that enable them to better “monitor” the various phases of reproduction and ensure the brood is protected effectively until the eggs hatch. To this end, they endeavor to insulate the brood as well as possible from the abiotic (water, sediment) and biotic (sneakers, predators, parasites, pathogens) environments by building more “enclosed” nests than those of the aforementioned species. Consistently larger than the females and equipped with features and coloring that make them distinguishable from females, some sexually mature males prepare a nest under a shelter or in a cavity of natural or anthropic origin (a tin can, a pipe, a hollow brick, an empty lamellibranch shell, various rockfills, etc.). The males of some species “self-build” their nest (Pomatoschistus microps, Pomatoschistus marmoratus, Pomatoschistus minutus, Zosterisessor ophiocephalus, sometimes Gobius niger, etc.). Male sand gobies of the genus Pomatoschistus all have the same building tactic. They choose a lamellibranch mollusc’s valve (shell) according to their body size. If the inside of the valve is not facing the ground, the “building” male turns it over and excavates a hollow beneath it, creating an entrance, and more or less covers the shell with sediment. From the entrance, where he often waits with his head facing out (Figure 3.8), he can manage the “ventilation” (water current) of the nest cavity and watch for females ready to spawn, as well as for sneakers, potential fertilization thieves and egg predators. The coveted female, seduced by his courtship, consents to follow the male into his nest where, “belly up”, she attaches her ovocytes, for around 1 h, to the inner surface of the shell that forms the nest’s ceiling or “sky”. Affixing them is done using adhesive filaments located at the “animal pole” of the ovocytes (Figure 3.9) where they form a kind of net around the micropyle (Figure 3.10). The mesh of this net is loose enough to allow the spermatozoa to reach the micropyle and fertilize the egg once it is attached to the substrate (Giulianini and Ferrero, 2001a; Giulianini et al., 2001b). Positioning the eggs thus insulates them from the mud bed and limits surface deposits of sediments that can be carried in by the water current circulating inside the nest.

Egg density (number per unit of surface area) varies from one species to another depending on their shape (subspherical, oval, piriform etc.; see Figure 3.11) and their size, which is itself variable within the same species depending on the size of the egg-laying female. Also, the size of the ovocytes released by the same female can diminish over the spawning season in Pomatoschistus (Bouchereau et al., 1991) and Salaria pavo (Ruchon et al., 1993). This reduction in egg volume shortens the incubation period, but leads to a reduction in larvae size, which seems counter productive in the lead-up to bad fall and winter conditions. It should be remembered that size and shape are factors in the egg surface/volume relationship, and thus in exchanges with the aquatic environment. Egg shape also has some impact on water flow within the surface area of the nest they are in. The environment, too, can be a determining factor in the females’ egg-laying tactic, and thus in how they use the available nest surface to attach their ovocytes. Thus, Knipowitscha panizzae females who frequently lay their eggs in nests built on anoxic beds, space out their ovocytes sufficiently to permit good water circulation at their level (Massironi et al., 2005). Species of the genus Gobius are more saxicolous; the males, always larger than the females and sneakers, seek out natural cavities, overhangs formed by heaps of large stones or rocks, to build their nests. As in the case of the aforementioned small gobies, the eggs are attached to the ceiling of the nest.

The black goby Gobius niger ejaculates thick, viscous sperm that he spreads out in trails called “sperm trails”, mainly on the ceiling of the nest, before or while the females lay their eggs, which can take longer than in the small species. The viscosity comes from secretions of sialomucus and proteins produced, along with steroid pheromones, by the two seminal vesicles (sperm duct gland), annexes of the male reproductive system. The seminal vesicles are typical of polygynous nest-building species; they are not present in monogamous species (Mazzoldi et al., 2005). The spermatozoa contained in this “glue” are inactive and become “active” as they are released progressively by the dilution of the gangue, which can take several hours.

Sneakers of this species produce sperm in “trails” that they deposit, using surprise tactics, on the walls of the nest with only one well-guarded entrance. Their sperm differs from that of the aforementioned; it is around 10 times richer in spermatozoa (Mazzoldi and Rasssotto, 2002a; Mazzoldi et al., 2011) and more fluid, the seminal vesicles in which it is stored producing less sialomucus. This mucus becomes dilute more rapidly than that of the nest-building males, thus releasing a large quantity of spermatozoa in a short space of time (Mazzoldi et al., 2011). These spermatozoa, which possess twice as much ATP, live for longer and are faster moving than those of the nest-building males (Locatello et al., 2007). The spermatozoa of the latter are thus swamped by those of the sneakers, which significantly outperform them. The grass goby Zosterisessor ophiocephalus digs a burrow with a number of entrances, usually under meadows of zostera or posidonia, i.e. in relatively fine sediments that are rich in organic matter and bacteria. The internal environment of this type of very confined nest is favorable to the proliferation of pathogenic agents (bacteria, fungus) that the male controls, together with “ventilation”, using a “biotechnique” that involves depositing sperm trails of mucus and proteins, produced by the seminal vesicles, onto the walls. These trails, containing spermatozoa, consolidate the walls, partially insulate the eggs from sediment and, as we will see, have major antibacterial properties. Contrary to what we have said about Gobius niger, the sperm of Zosterisessor ophiocephalus sneakers is only about five times richer in spermatozoa, and these are no more active than those of the nest-building males (Mazzoldi et al., 2011). The differences in the sperm of the sneakers in these two species can be correlated to nest structure (Mazzoldi et al., 2000, 2011). In the case of Gobius niger, the nest, which has only one opening, often fairly wide, and “healthy”, loosely constructed walls, is easy to guard and the water circulation inside is relatively intense; by contrast, the nest of Zosterisessor ophiocephalus, sunk into the sediment, is more tortuous, more confined and more difficult to guard since it has several entrances (Figure 3.12). In the latter scenario, a sneaker can easily enter the nest, since the owner can only guard one entrance at a time. Moreover, in view of the time the owner spends in going round the various entrances to defend them, and also in feeding and attracting spawning females, a sneaker has a good chance of having enough time to deposit his sperm trail in contact with an egg-laying female who spends several hours attaching her ovocytes.

Finally, since water circulation is weak, the loss of spermatozoa through dilution and leaching is low. The male’s reproductive success is easily assured and he can avoid expending significant energy in high sperm production. Conversely, it is more difficult for a sneaker of Gobius niger to gain access to an egg-laying female and leave his sperm trail; the Gobius niger nest-owning male can guard his nest more easily than the Zosterisessor ophiocephalus male, due to its structure. If, by cunning, the sneaker manages to get close to a female who is in the process of laying her eggs, he must quickly produce a sperm trail or “cloud” before being driven away by the nest owner. His ejaculate is rich in highly active spermatozoa that are good swimmers and long-living, which are both qualities that give hope for some degree of reproductive success, even though he emits his sperm some distance away from the spawning female, he does not have much time to do this, and there is significant dilution effect. As we have seen, all of these conditions are satisfied insofar as the sperm of Gobius niger is concerned, but the cost in terms of energy expenditure is higher than in the case of Zosterisessor ophiocephalus. Unfortunately, we have no genetic data to enable a comparison between the relative reproductive success of sneakers and nest owners in the two species.

The behavior of the lagoon-dwelling blenny Salaria pavo is close to that of gobies. The male seeks out cavities and crevices in rocky biotopes (rare in lagoons), empty mollusc shells and “cavernous” areas that may exist within reefs of the serpulid Ficopomatus enigmaticus (Mercierella enigmatica), which occur frequently in a good many Mediterranean lagoons. He also uses tin cans, hollow bricks, etc., thus any “cavernous” object that humans have thrown into the water (Figure 3.13).

Adult nest-owning males (bourgeois males) have a “reproductive apparatus” that presents strong morphological, histological and functional analogies with that of gobies. In fact, we can note the presence of a testicular gland on the ventral surface of the two testicles (Eggert, 1931; Pern and Laumen, 1977; Seiwald and Patzner, 1987; Lanhsteiner and Patzner, 1990a; Lahnsteiner et al., 1990c), which produces sialomucus that contribute to increasing the viscosity of the sperm together with the sulfomucus secreted in the sperm ducts (Lanhsteiner and Patzner, 1990c), in the part that we sometimes call the “seminal vesicle”. These two processes play a role in spermatogenesis, which is “semicystic” (early spermiation). The testicles actually release spermatides (Lanhsteiner and Patzner, 1990c). Spermiogenesis, up to the “mature spermatozoa“ stage, takes place in these two annexes (Seiwald and Patzner, 1987; Lahnsteiner and Patzner, 1990a). Mature spermatozoa are stored, remaining inactive, in the vas deferens (sperm ducts). This type of semicycstic spermatozoa production (early spermiation) can be found in various gobies and in Syngnathidae. With this type of production, a large quantity of spermatozoa can be achieved on an ongoing basis: production being shared, before the final storage, between the testicles and their annexes. Moreover, two evaginations from the sperm ducts, the “blind pouches” (Lahnsteiner et al., 1993c), produce sialomucus, but most importantly these pouches are where the synthesis of glucorinic steroids takes place (Barata et al., 2008; Barata and Gonçalves, 2011), and these act as sexual pheromones. These water-soluble pheromones are emitted with sperm and urine. The blind pouches do not store spermatozoa. “Anal glands“, located on the first two radii of the anal fins, secrete mucus (glycoproteins) (Barata et al., 2008; Serrano et al., 2008a, 2008b, 2008c), odiferous compounds (amino acids and peptides), pheromonal compounds (Laumen et al., 1974; Barata and Gonçalves, 2011) and antibacterial compounds with which the eggs can be coated, thus protecting them from any pathogenic germs. The activity of the testicular glands, sperm ducts, blind pouches and anal glands is linked to the blenny’s sexual cycle.

Eggs are attached inside the nest by a wreath of filaments, 1.5–3 μm in diameter, close to their base, creating what is generally called the “adhesive disk”, a kind of “mesh” located at the egg’s animal pole. The micropyle is, as in gobies, located in the center of the adhesive disk, with its opening facing toward the substrate, probably covered in viscous sperm trails deposited by the male before the ovocytes are laid (Patzner, 1984; Patzner and Lahnsteiner, 2011). According to Patzner (1984), the ovocytes of Salaria pavo present no change in volume (diameter after preservation = 1,200 μm), either after laying or after fertilization (eggs), whereas the ovocytes (thus the eggs) of gobies can treble in volume (Tavolga, 1950).

A number of field studies (Ruchon et al., 1999) have demonstrated that the quality of the nest, and also the quality of the nest-owning male (termed the bourgeois male), play a determining role in the choice the females can make from a number of proposals, not forgetting that the nest must also have certain qualities to enable the male to ensure his own success: to achieve as many fertilizations and as many hatchings as possible. The surface area to which the female can “stick” her ovocytes and the size of the nest’s opening are determining factors in this respect. The characteristics of the opening dictate whether or not the male will be able to effectively manage water circulation in the nest in relation to its volume, manage the coming and going of females and limit incursions by predators and sneakers, and thus the number of “parasitic” fertilizations. Moreover, this type of nest maximizes his chances of fertilizing all the ovocytes entrusted to him by females; his spermatozoa, after ejaculation, remain “confined” within the reproduction chamber, and this enables him to achieve a high fertilization success rate with lower spermatozoa production than in the open water spawning strategy. At least the black goby Gobius niger, the grass goby Zosterisessor ophiocephalus and the peacock blenny Salaria pavo maximize their chances of fertilizing the ovocytes attached to the walls of their nest, by depositing viscous sperm trails (a mixture of mucous substances secreted by the seminal vesicles annexed to the sperm duct) onto the walls of their nest before the females deposit their eggs, which they attach with the micropyle almost in contact with the sperm-coated wall, not necessarily touching the trails, but close. Gradually (sometimes over the course of several hours, 20 h in Zosterisessor ophiocephalus), the spermatozoa detach themselves from the mucous gangue, and this allows the ovocytes to be fertilized by the spermatozoa of the bourgeois male even during his absence (asynchronic emission of male and female gametes). The bourgeois male is thus able to devote more time to defending his nest against sneakers and predators, and feeding. Moreover, the permanent presence of spermatozoa plays an important role in sperm competition by limiting the effect of ejaculates from sneakers, which are more fluid (low in mucus) and very rich in spermatozoa, which disperse rapidly and easily in the water inside the nest. Secretions from the seminal vesicles of gobies that build semiclosed nests have antibacterial capabilities (Giacomello et al., 2008), guaranteeing them a nest with a relatively antiseptic environment, favorable for eggs. Moreover, in Zosterisessor ophiocephalus where confinement and the sedimentary nature of the nest walls are favorable to the development of potentially pathogenic bacteria (Santos et al., 1999), the “sperm trail” strategy, alongside its antibiotic capabilities, acts as a coating that consolidates the nest and creates a brooding area insulating the eggs from direct contact with sediment, as is done, among other strategies, by Atherina lagunae, Gasterosteus gymnurus, Pomatoschistus microps, Symphodus cinereus staitii, etc. In lagoon-dwelling nest builders (gobies, blennies, sticklebacks), the production of antibiotic substances deposited by the male on the inside walls of the nest, provides some moderation of the development of pathogenic bacteria. This seems widespread and can, according to Little et al. (2008), be viewed as a novel form of parental protection.

With regard to the wrasse Symphodus cinereus staitii, the male, who builds and then protects a basin-shaped nest made of vegetation (like a bird’s nest), attracts (courtship behavior) a number of females in succession. His choice appears to favor those who present the largest urogenital papillae (turgescence) tinged with black, a sign of maturity. The ovocytes are attached to the algae of which the nest is made, but we have no data on the structure used to attach them, nor whether there are, apart from paternal guarding (chasing away nest scavengers, mechanical elimination of dead eggs and parasites), any other strategies in operation to protect the eggs from pathogenic agents. In this wrasse, as in other species of the genus Symphodus, the presence of sneakers is common.

3.2.6. Other aspects of reproduction in migrators

All migrators spawn in the sea, but in general the precise geographical location of the reproduction sites is rarely known exactly, even for the eel Anguilla anguilla. Apart from eels and the goby Pomatoschistus minutus who only engage in one spawning season in their lives (semelparous parents), all migrators are iteroparous, thus engaging in several spawning seasons during their lifetime. They spawn in open water; their eggs are pelagic, small and relatively numerous or even very numerous; and their development is rapid. Gametes are released very promiscuously within groups or shoals, and it is not possible to identify whether couples form, which does not exclude the possibility, as has been demonstrated in cod. It is a difficult issue to address, because group spawning occurs between parents whose sexual dimorphism is limited, more often than not, to a difference in size. In mullets, we have a description of the formation of microgroups composed of one female accompanied by several males (Breder, see Bertin, 1957), but these observations are little documented. In Pomatoschistus minutus, a nest-building goby, one female lays eggs successively in the nests of several males (polyandry); therefore transient couples are formed, limited to the time taken for the egg-laying act. With regard to eels, while more and more is known about their migratory journey (Amilhat et al., 2016) and spawning grounds, their spawning behavior remains a mystery.

In these fishes, whether they are semelparous or iteroparous, there may be only one spawning act during the season or several (fragmented spawning). The distinction between these two “types” of ovocyte emission is usually deduced from the shape of the frequency polygons of size of ovocytes contained in the ovaries. If this shape is bimodal (one mode for the small cells and one for the large ovocytes), it is accepted that spawning occurs only once or closely grouped together in time. This interpretation is not necessarily true. In fact, this configuration can indicate that there is a relatively long time lapse between a succession of several spawnings that cannot be seen from the degree of precision of the measurements of small cells and the graphic representation. With the exception of Pomatoschistus minutus, a nest-building fish that practices parental care of the eggs entrusted to him, all migrators present ovarian fecundity considerably greater than the sedentary lagoon-dwelling species. For instance, in the flounder Platichthys flesus, absolute fecundity is from 325,800 to 1,450,000 (25-45 cm TL) (Vianet, 1985); from 50,000 to 272,000 (36-56 cm TL) in the sea bass Dicentrarchus labrax (Kara, 1997); in farmed gilthead sea bream Sparus aurata, relative fecundity is from 1,000,000 to 2,000,000/kg (Zohar et al., 1984).

3.3. Feeding and energy transfer

In general, food is not a limiting factor for fishes, due to the great biomass of prey in lagoons, even though they generally present low diversity. Moreover, the wide trophic flexibility of lagoon-dwelling species enables them to easily satisfy their nutritional requirements. It should be stressed that lagoons have very different interlagoon and intralagoon nutritional potential (see section 3.4) according to their interactions with the sea and the continent (Isnard et al., 2015; Escalas et al., 2015).

3.3.1. Alimentary guilds and competition

Two major guilds should be considered in the feeding of lagoon-dwelling fishes:

1) the sedentary guild, characterized by their small maximum size. These, regardless of the size of the constituent individuals, feed on small planktonic, nectonic and benthic prey (endogean and epigean);
2) the migratory guild, the individuals of which present a very wide range of sizes and because of this, mostly have very different diets and feeding behavior from each other, ranging from microphagy to macrophagy of invertebrates and vertebrates.

The major difference between these two guilds is that the sedentaries contribute to the lagoon’s self-sufficiency and thus form part of the lagoon’s ongoing “tropho-energetic riches”, while the migrators are users, exploiters, exporting the lagoon's energy resources to the sea and to fresh water. Above a certain size, migrators are rarely the object of significant intralagoon predation. Schultz and Kruschel (2010) suggest that the predation risk, particularly for juvenile fishes, is higher in the meadows and algae beds than on the adjoining naked substrates. Bird predation is limited by the shallowness of lagoons and only fishing, sometimes to excess, has an impact on the populations. It should be noted that the only typically herbivore autochtone fish that visits certain lagoons is the Mediterranean bream Sarpa salpa, but the Siganidae lessepian fishes Siganus luridus and S. rivulatus, both herbivores, are new lagoon residents in the Eastern Mediterranean. The goby Gobius cobitis is sometimes regarded as omnivorous, and juveniles of the white seabream Diplodus sp. are temporarily herbivorous.

There is significant trophic competition within each of the two guilds, and also between the sedentaries (regardless of their ontogenetic phase) and migrators in the larval, post larval and juvenile phases. Gisbert et al. (1996) studied trophic resource sharing between the larvae and fry of five migrators (Chelon labrosus, Liza aurata, L. ramada, L. saliens and Mugil cephalus) and three residents (Cyprinus carpio, Gasterosteus holbrooki and Atherina boyeri) in the Ebro Delta (Canal Vell lagoon). They noted that there was no competition between the following pairs: Gasterosteus holbrooki–Cyprinus carpio, Gasterosteus holbrooki–Chelon labrosus, Atherina boyeri–Cyprinus carpio and Atherina boyeri–C. labrosus. By contrast, there was keen competition between the various coexisting species of Mugilidae, as well as also between G. holbrooki and Atherina boyeri. The five mullet species actually have a very similar diet composition in terms of prey diversity and size. These mullets feed principally on planktonic invertebrates measuring around 800 μm, such as the cyclopoid Acanthocyclops robustus, but also feed on prey of more than 1,500 μm, particularly adult Chironomus salinarius and Daphnia curvirostris. The diet composition of Cyprinus carpio is very similar to that of the mullets, but includes fewer copepods than cladocera. Gambusia holbrooki and Atherina boyeri consume small prey, especially rotifers, nauplii and copepodites. They never take prey measuring over 1,200 μm. Adult and older juvenile migrators, which are often larger than sedentaries, compete a little with the latter. This competition, linked to size, is of limited duration in view of the often rapid growth of the young migrators. Also, having, like the adults, entered the lagoons in late winter and early spring, they return to the sea during late summer and fall. The sedentary's small stature is therefore an advantage. It enables them to fully exploit, all year round and throughout their lives, small prey (invertebrates) living in the shallow areas that are inaccessible to large migrators, with their only competition, for a few months, coming from young migrators of small stature. Shaiek et al. (2015) studied the diet of 16 species in Ichkeul Lake: eight species forming a homogeneous trophic group, and the other eight each having their own diet.

3.3.2. Cannibalism

Sedentary lagoon-dwelling fishes are not fundamentally “piscivorous”, but on some occasions they can become so. In most cases, this ichthyophagy manifests itself as “egg cannibalism” (Barras, 1998), particularly “filial”, as in the gobies Pomatoschistus minutus and Pomatoschistus microps (Lindstrom, 1998; Kvarnemo et al., 1998; Svensson and Kvarnemo, 2007) and the stickleback Gasterosteus aculeatus (Sparkes et al., 2008; Mehlis and Bakker, 2009). To meet their energy requirements so as to maximize their reproductive success, nest-guarding males sometimes feed on the eggs they are guarding. These eggs are rich in nutrients, are easily accessible, and do not necessitate a period of absence that would be detrimental to guarding. In experiments, Marzano and Gandolfi (2001) show that Knipowitschia panizzae females, in the event of a food shortage, are likely to become oophagic. Cannibalism is little practiced among syngnathids (Franzoi et al., 1993; Cunha et al., 2015–2016; see also section 3.2.4), but a very unusual case of nutrient input is that found in the incubating male Syngathus typhle. Sagebakken et al. (2010) demonstrate, by marking the females’ ovocytes at ¹⁴C, that the nutrients resulting from the destruction of dead eggs in the marsupial pouch are metabolized by the pregnant father (a form of necrophagy), allowing him to keep himself in good physical condition to complete gestation and ready himself for a new reproductive cycle (the embryos can also benefit from these nutrients).

According to Kvarnemo (1995, 1997), lack of food does not in itself fully explain cannibalism. In fact, paternal cannibalism can occur, without there being a famine, to the benefit of the parents’ reproductive success, maintaining an egg density compatible with favorable environmental conditions (water circulation, elimination of damaged eggs, etc.) for the propitious development of the broods in the male’s care (Klug et al., 2006). It can be that only the father benefits (Klug and Lindströme, 2007); in fact, when he eats the largest eggs, or the most recently laid, he reduces the length of the guarding period (incubation time being proportional to size of egg), and this enables him to keep himself in good physical condition (guarding being arduous) and increase the number of broods he is able to achieve over the duration of the spawning season. In experiments, Marzano and Gandolfi (2001) demonstrate that cannibalism between adult Knipowitschia panizzae females in the Po Delta can be induced after three days of fasting, and that males are not the object of predation if the sex ratio is out of balance (2.5 females/1 male during the period of the experiment). The female predator is always larger than her prey: attacks by the tail are most frequent, but frontal attacks present the highest percentage success rate. The prey is swallowed whole.

3.3.3. Feeding behavior

Migrators, just like sedentaries, whether in deep lagoons, laminar lagoons or estuaries, are extremely efficient users of the food riches of these environments. They exploit every level, from the bacterial film on the substrate (certain mullets), to molluscs (sea bream), crustaceans and fishes (sea bass), etc. They also take an interest in all sizes and all forms of prey (planktonic, endo and epibenthic, sessile and vagile, nectonic), according to their size and requirements. Competition for access to prey between individuals within the migrator guild is relatively low, given the existence of a relatively high degree of specialization, especially in the large specimens, be they juveniles or adults.

Prey capture strategies are varied and evolve during the course of ontogenesis (Yaniv et al., 2014). The search for food is based simultaneously on visual, olfactory and gustative stimuli. These three sensory systems often act in conjunction in prey detection and in governing feeding behavior. Mecanoreception and/or electrolocalization are also involved at a secondary level. Fishes develop tactics to catch their food (surface feeding or open-water feeding, grazing, scraping, digging, sit-and-wait or open-water hunting, etc.). It should be noted that, in most cases, each species and thus each individual has a “repertoire” of a number of tactics applicable to seeking, detecting, pursuing, attacking, capturing and manipulating their prey, and selects the one best suited to the diverse circumstances of the “predator–prey” interaction. This degree of flexibility enables the individual, group or shoal to adapt themselves to environmental conditions (light, temperature etc.) and biotic conditions (prey density, presence of competitors or predators, etc.). Thus, the sea bass Dicentrarchus labrax is able to hunt nectonic prey using “sit-and-wait” tactics, or as a lone explorer, or in groups of two to five juvenile individuals to attack shoals of various small fishes or mysidacean crustaceans. Adult sea bass often attack groups of sardines, anchovies or mullets (Barnabé, 1978). In certain circumstances, they can become benthophagic. “Cleaning behavior” in fishes that visit lagoons is unusual. Only young Diplodus sargus (110 mm maximum TL) sometimes feed on copepod ectoparasites (Caligidae, Caligus pageti) that are present on the bodies of mullets (Fogliano and Caprolace lagoons, Italy; Mariani, 2001) and sometimes on other sparid fishes (Thau lagoon, France; Rosecchi, 1985 and 1987).

Predator–prey interactions are thus flexible and rarely give rise to rigid, stereotypical behavior. Despite this, in terms of capture tactics, fishes can be classified according to five categories.

3.3.3.1. Sit-and-wait hunters

In lagoons, Syngnathidae and especially the seahorses Hippocampus sp. practice the sit-and-wait type of hunting. Seahorses are unusual fish, most often associated with a vegetational environment. They can remain relatively immobile there, lying in ambush while they watch for small crustaceans (amphipods, copepods, shrimps) to come within reach, less than 4 cm away. They have the gift of excellent sight and independent movement of the eyes and head (30% greater movement than syngnathids), enabling them to cover a wide visual field (Van Wassenbergh et al., 2011). Prey capture is effected by suction when the prey comes within 1 cm of their mouth (thereby creating a current with a speed of 1 cm per second). The success of sit-and-wait hunting in seahorses is partly due to the fact that they are “chemically silent” (no kairomone emissions) (Bruslé and Quignard, 2004).

Gobies, which are benthic carnivores (Pomatoschistus sp., Gobius sp., etc.), can be included in this category, but their capture strategy is more “active”. In fact, this entails the fish making a sudden leap forward to seize the prey as it comes within reach (small invertebrates, essentially vagile crustaceans belonging to the epibenthic meiofauna) (Zander, 2011a). According to Zander (2011b), these gobies can sometimes go into open water to supplement their food rations. They may also practice a certain degree of “trophic exploration” in search of prey visiting the algae beds. This is the case for Gobius cobitis which sometimes presents stomach contents relatively rich in algae (it is sometimes considered to be omnivorous).

The flounder Platichthys flesus, which lives more or less hidden in the lagoon’s mud and sand sediment, has prey capture behavior similar to that of gobies. This flatfish feeds primarily on itinerant polychete annelids (Gammarus sp.) and also, in the case of large individual, amphipod crustaceans (Gammarus sp.) and even benthic shrimps (Crangon crangon) that pass within reach, with molluscs being secondary prey (Gandolfi and Giannini, 1977; Bekhti et al., 1985). The soles Solea solea and Solea senegalensis, whose feeding requirements and “bioecological” behavior are similar to those of the flounder, belong to the same guild. However, the data available on the diet composition of these fishes in lagoons do not make it possible to assess the degree of interspecific competition, especially with Solea senegalensis, an exotic species that recently arrived in the Mediterranean.

3.3.3.2. Active solitary hunters

The syngnathids Syngnathus sp. are equipped with a long tubular snout and a bucco-pharangeal system (Leysen et al., 2011; Van Wassenbergh et al., 2011) that permits them to create a strong suction current of the adjacent water and the prey (shrimps, various larvae) contained in it, within a radius of around 10 cm. They therefore practice the same capture method as seahorses, but the suction is most often performed as the final phase of an active search (hunt) for prey (Van Wassenbergh et al., 2011). The sea bass Dicentrarchus labrax, which feeds on shrimps and small fishes, hunts in open water but, in shallow lagoons, it may also feed on benthonectonic animals (scud, idotea) and also molluscs. The sticklebacks Gasterosteus sp. mainly hunt alone, but in cramped environments they may, involuntarily, be in small groups. Moreover, depending on the type of lagoon and certainly on the quantity of available prey (opportunism), they can be planktonophagic or benthophagic; the dominant prey are crustaceans, and also fish larvae including elvers (Daniel, 1965). Opportunism can lead them to practice “filial cannibalism” (Sparkes et al., 2008). According to Hagen (1967), sticklebacks of the low-plated type (Gasterosteus aculeatus gymnurus) are basically benthophagic. But in the Carmargue, Pont et al. (1991) indicate that these sticklebacks (Gasterosteus aculeatus gymnurus) are very voraciously zooplanktonophagic.

3.3.3.3. Active hunters in groups or shoals

The atherines Atherina sp. travel in shoals and usually hunt together. In deep lagoons, they hunt in open water for planktonic creatures, mainly crustaceans, while in laminar lagoons they catch benthic and benthonectic creatures (small molluscs, amphipod crustaceans: Gammarus sp., isopods: Idothea sp., copepods: Corophium sp., etc.). This guild also includes the white anchovy Engraulis russoi and the blue anchovy E. encrasicholus, planktonophagic hunters. The sea bass Dicentrarchus labrax sometimes hunts in small groups of three to five individuals (Barnabé, 1978).

3.3.3.4. Pickers, gatherers, grazers and wanderers

The gilthead sea bream Sparus aurata can be regarded as this type. Basically conchyliphagous, it tears with its canines and crushes, using its molars, lamellibranch molluscs, particularly mussels. What is more, the damage that large groups of bream can cause to lagoon and marine mussel and oyster farms demonstrate this beyond dispute. In lagoons, usually shallow ones, where the serpulid tubeworm Ficopomatus enigmaticus thrives, bream graze on the upright tips of the tubes that actually form reefs touching the surface. The wrasse Symphodus cinereus staitii (Labridae), less specialized than the gilthead sea bream, has the same morphoanatomical capabilities: jaws equipped with strong canines, enabling it to seize and ingest prey either unattached or attached to the substrate; presence of pharyngeal teeth, more or less molariform, capable of crushing prey with a shell or carapace (molluscs and crustaceans). However, no descriptions exist of the prey approach and capture tactics used by this species. Its diet consists of, in roughly equal proportions, crustaceans (shrimps: Crangon crangon, Leander sp.; amphipods: Gammarus sp.) and molluscs (gasteropods: Bittium sp., Rissoa sp., Amycla sp.; small lamellibranches, Mytilus galloprovincialis and Modiolaria sp. spats) (Quignard, 1966; Quignard and Pras, 1986b).

The blenny Salaria pavo is strictly carnivorous in some lagoons, such as for example the Mauguio lagoon in France (Ruchon et al., 1998), where it feeds on amphipod crustaceans and grazes on reefs constructed by the exotic serpulid F. enigmaticus. In other areas, such as the North Adriatic, it is omnivorous (Patzner, 1983; Santic et al., 2007). The search for food occurs during more or less random journeys over short distances, especially in the case of nest-building males who continue feeding during the reproductive period (Ruchon et al., 1998). The young of Diplodus sargus occasionally adopt the behavior of cleaners–grazers–pickers. Mariani (2001) and Rosecchi (1985 and 1987) actually found the copepod Caligus pageti, a mullet parasite, in the stomach of this sea bream.

3.3.3.5. Limivore scrapers

Mullets belonging to the geni Mugil, Liza and Chelon are often included in the “limivore” or “detritus-eater” guild. However, their diet compositions are relatively varied, including benthic as well as planktonic species. These mullets form itinerant groups (shoals), easily identifiable when they swim close to the surface and perform leaps out of the water. They have a tendency to become benthic to search for food (Olla and Samet, 1974) which consists of small living organisms in the upper layers of the sediment: microfauna of protozoaires, copepods, microflora of unicellular algae (diatoms etc.), protobacterial film. Picking up food is facilitated by the position of the mouth, surrounded by relatively well-developed lips, and the position of the body at an angle of 15–30° when picking up food, thus allowing them to scrape the superficial layer of sediment rich in organic matter. This mixture is filtered in their pharyngobranchial organ (branchiospines). Some sandy particles are ingested and assist with mechanical trituration of the alimentary bolus in the extremely muscular stomach, in the same way as the gizzard of a fowl.

In lagoons and estuaries, mullets have access to considerable trophic reserves: the organic benthic biomass (dead or living) is virtually unlimited, which reduces intra- and interspecific competition. Each mullet species tends to select, according to its morphoanatomy, benthic material with distinctive physical characteristics (particle size) and perhaps chemical characteristics (vegetation debris, microalgae, type of proteobacterial film, zoobenthos) (Marais, 1980). In some regions, there have been descriptions of other mullet species with different feeding rhythm periodicities (day or night), which also limits interspecific competition (Marais, 1980). In one lagoon in the Balearic Islands, Cardona (2001) found no competition between the five mullet species that, in additional, are capable of expanding their trophic niche as necessary.

3.4. Age and growth

The hydroclimate in lagoons presents far more marked seasonal variations than in the marine environment to which they are connected. Consequently, the feeding rhythm and growth of their resident fish show seasonal fluctuations, whether or not they undertake periodic sea-lagoon migrations. For this reason, a number of authors have modified the von Bertalanffy (1938) equation generally used to express fish growth, so as to include the “seasonal variations” factor (Longhurst and Pauly, 1987; Hoenig and Chaudhury, 1982; Somers, 1988; Soriano and Jarre, 1988). One example of the application of this model is given in Figure 3.14 concerning the sea bass Dicentrarchus labrax in the Etang de l'Or (Mauguio, France). As shown, the model predicts a non-growth period of around three months (January to March), a characteristic that previous growth models were not able to take into consideration (Pauly and Yañez-Arancibia, 1994).

Chauvet (1988) suggests that shallow, eutrophic lagoons (Tunis) permit better growth compared to deep lagoons that are strongly influenced by the sea (Salses-Leucate, El Biben). He also finds, especially in the sea bass Dicentrarchus labrax and the gilthead sea bream Sparus aurata, that the lagoon habitat appears to permit better growth of juveniles and young adults (Figure 3.15) than the sea. In terms of size (growth) and condition corpulence) of 0+ sea bream (juveniles), Isnard et al. (2015) highlight interlagoon differences (Thau, Mauguio, Bages, Salses Leucate, France) and Escalas et al. (2015) intralagoon differences (Mauguio). In both cases, areas under continental influence, rich in particulate organic matter (CPOM), are the most favorable for the development of gilthead sea bream fry. For older individuals, growth in lagoons is at best equal to, but generally less than, growth in the sea (Ravagnan, 1978; Lemoalle et al., 1984b; Chauvet, 1986). Sudry (1910) writes: “according to fishermen, their size (that of the fish) increases more rapidly in Thau than in the sea”. In fact, the growth of these relatively long-living migrators is linked to a number of factors, including their marine past, and then, in the lagoon, the richness in accessible food and its nutritional quality, intraspecific and interspecific competition that can differ from one year to another according to reproductive success and conditions that are more or less favorable to migrations (settlement density), and finally how close the coincidence is between the time of the fish entering the lagoon, its morphophysiological stage of development and the state of the lagoon in terms of both water quality and trophic resources.

However, since methods of addressing age (scalimetry, otolithometry, frequency polygons of size, tagging, etc.) are very diverse, comparisons between years and populations are overshadowed by significant doubts. Up to now, there have been no reliable techniques making it possible to distinguish individuals that have visited lagoons from those who have remained in the sea. It was therefore objectively difficult to assess, after migration, the advantage gained from spending time in a lagoon and to estimate its consequences on the later life of the fish. The advance of research into the microchemistry of otoliths and the analysis of stable isotopes, making it possible to identify individuals who have spentt time in lagoons, opens up interesting perspectives in this domain (Dierking et al., 2010, 2012; Morat et al., 2009). Finally, the “migratory phase” of marine stock may perhaps be composed of individuals selected on a genetic basis from the start or within the lagoons, which could interfere with their biotic and abiotic conditions (Lemaire et al., 2000, 2004–2005; Gonzalez-Wangümert et al., 2004, 2006; Chaoui et al., 2012; Guinand et al., 2016). Other studies have demonstrated that while lagoon environments are “attractive” to fry and juveniles of euryhaline species, they do not always facilitate growth rates higher than those recorded in the sea. Growth rates can even be lower in some mullets. For example, 0⁺ juveniles of Liza aurata present a mass of 0.5 to 1 g in Salses-Leucate lagoon and 2.37 g in the Gulf of Marseilles (Albertini-Berhaut, 1980). Marfin (1981) highlighted a weaker growth rate in Atherina boyeri (Atherina lagunae) in lakes than in the adjacent coastal fringe and underlined the unfavorable character of Canet-Saint-Nazaire lagoon (France). Likewise, Quignard et al. (1984b) recorded a summer growth halt in a number of species of lagoon-dwelling fish, as well as a slowing or halt in growth over the winter; this halt is also visible in eel otoliths (Lecomte-Finiger, 1983). The goby Pomatoschistus minutus in the Mauguio lagoon presents not only weak growth in summer but also summer weight loss (Bouchereau et al., 1989). However, we must be careful because, with very few exceptions (Mathias and Jalvy, 1958; Lasserre, 1989), we do not have any studies carried out simultaneously in the sea and lagoons. In general, we have to make do with statements indicating that weight gain and growth are (or may be) favorable at certain growth periods but not all (Amanieu and Lasserre, 1973b, 1974), and in certain lagoons (or parts of lagoons) but not all (Roblin, 1980).

The inequality in the growth of juveniles is apparent not only between lagoons and the sea, but also between lagoons. Chauvet (1988) suggests that shallow, eutrophic lagoons (Tunis) allow better growth in comparison with deep lagoons that are strongly influenced by the sea (Salses-Leucate, El Biben). Isnard et al. (2015) demonstrate that shallow, eutrophic lagoons with significant input of continental particulate organic material (CPOM) (Mauguio, Bages, France) are, for “0⁺” juveniles of the gilthead sea bream Sparus aurata, nurseries that prove more favorable to their growth (size and condition upon leaving the lagoons in fall) than the large, deep, marinized, mesotrophic lagoons (Thau, Salses-Leucate), which are relatively low in continental OM. The content of stable ¹³C and ¹⁵N isotopes shows that OM of continental origin represents more than 33% of total organic matter in Mauguio as against 15% in Thau and less than 5% in Salses-Leucate. Within the same lagoon basin, we can also note differences such as disharmonies in growth between micro-cohorts of Liza aurata on the fringes of the Salses-Leucate lagoon: the La Coudalère basin and the Les Dindilles inlet (Bruslé and Cambrony, 1992).

Thus, newly recruited fry very quickly localize themselves in certain areas of the lagoon, adopting territorial behavior that is at odds with the reputation generally attributed to them as wanderers. Their growth will therefore reflect strictly local ecobiological conditions. Consequently, microcohorts of juveniles develop separately from each other in the marginal lagoon areas, all of which constitute “autonomous” microenvironments that are more or less juxtaposed. Similarly, for the gilthead sea bream, Quignard et al. (1984b) have demonstrated that the size distribution of individuals in the “0⁺” cohort upon entering the Mauguio lagoon, is unimodal and becomes bimodal. By monitoring these two modes it has been possible to give growth curves (size and body mass) for the sub-cohort composed of slow-growing individuals and for that composed of fast-growing individuals. To explain this bimodal structure, these authors put forward a number of hypotheses: intralagoon intraspecific competition or colonization by two successive waves (various causes). In this case, it is difficult to speak of “slow and fast growth”, since the duration of life in the lagoon is not the same. The possibility of new arrivals, from the same wave, being divided between the two parts of the lagoon described by these authors (marine in the south-west and limnic in the north-east, with different trophic potentialities), as Bruslé and Cambrony (1992) suggest for the Salses-Leucate lagoon, is not considered. This is the hypothesis that Escalas et al. (2015) confirmed. In fact, in the Mauguio lagoon (France), these authors demonstrated that the young “0⁺” gilthead sea bream living in the north-eastern part of the Mauguio lagoon, which is influenced by continental waters, were, when they left the lagoon (at the age of a few months old), significantly superior in condition and size (therefore in growth) to those that had remained in the relatively marinized western part. Analysis of the ¹³C and ¹⁵N isotopes shows that 62% of white muscle of the individuals in the north-eastern section is composed of organic matter of continental origin. These examples of differential growth, inter- and intralagoon, clearly highlight the role of the lagoons’ enrichment in continental particulate organic matter (CPOM) in the production of juveniles (aged around eight months) of good quality (size, condition) who will go back to the sea after spending a few months in a lagoon. Yet, it is still difficult to evaluate what consequences the advantages gained by these juveniles will have on the course of their adult lives. However, Lasserre and Labourg (1974b) accept that the size attained as a consequence of spending time in lagoons has an impact on marine “stock” dynamics. These insights are all the more important since, according to Mercier et al. (2012), all of the gilthead sea bream in the Gulf of Lion bear the chemical signature of having spent time in a lagoon on their otoliths.

Mediterranean lagoons do not hold the record for the smallest fish, nor the fish with the shortest life expectancy, but they have a strong representation of small, sedentary, lagoon-dwelling annual species.

Some Syngnathus abaster, Gambusia sp. individuals, sexually mature a few months after hatching and shortly before the first winter of their lives, may have a lifespan of less than 1 year.

Only one species, Pomatoschistus minutus, is currently regarded as an annual species. Gobius cobitis and Gobius paganellus are the most long-living species; they can live for around 10 years. The annual or subannual species are Pomatoschistus microps (possibly 2 years), Pomatoschistus marmoratus (possibly 2 years), Pomatoschistus canestrini, Pomatoschistus tortonesei, Gasterosteus aculeatus (Gasterosteus gymnurus), Syngnathus abaster (17 months).

Others live for at least 2 years but not more than 5 years (Gambusia sp.: 2 years; Syngnathus taenionotus: 2 years; Aphanius iberus: 2 years; Aphanius fasciatus: 3–4 years; Salaria pavo: 2–3 years; Atherina lagunae: 3–4 years; Gobius niger: 4–5 years; Zosterisessor ophiocephalus: 5 years).

Migratory species have a considerably longer lifespan: 10 years for the sea bass Dicentrarchus labrax on the Tunisian coast (Bouain, 1977); 8 years or even 12 years for the gilthead seabream Sparus aurata in the Mirna estuary in Croatia (Kraljević and Dulčić, 1997); 7 years for the flounder in the Mauguio lagoon (Vianet et al., 1989); 8 years for Chelon labrosus in the Pantan estuary in Croatia (Morovic, 1960) and up to 9 years in the Dardanelles (Erman, 1961); 9 years for Liza ramada in the Porto-Lagos lagoon in Greece (Koutrakis and Sinis, 1994); 7 years for Mugil cephalus in the Venetian lagoon in Italy (Morovic, 1954, 1957) and up to 9 years in the Bosphorus (Erman, 1959); 11 years for the eel in the Commacchio lagoon (Rossi and Colombo, 1976).

In the sedentary species, 50–100% of their maximum size is reached during the first year, and sometimes even by the beginning of the first winter of their life. Without exception (Quignard et al., 1984b), the summer reproduction period does not cause a slowdown or halt in growth. The growth rate and maximum size of nest-building males are superior to those of females (Labridae, Blenniidae, Gobiidae); Gasterosteidae are a notable exception. In the ovuliparous and viviparous non-nest-building species (Atherinidae, Cyprinodontidae and Poeciliidae), females all attain a size superior to that of males.

Being a small fish presents an obvious advantage when it comes to conquering lagoons which are generally shallow. This trait allows them to occupy a wide variety of micro and mesohabitats, for instance reefs built by the worm Ficopomatus enigmaticus, from which large predators are excluded. Moreover, their role in energy transfer (food chain) between meio and mesofauna, thus between lagoon and sea via the lagoons’ large migrators, are determining factors in lagoon-sea energy dynamics.

Migrators whose stay in the lagoon is relatively long in relation to their lifespan, such as the eel Anguilla anguilla and the sand goby Pomatoschistus minutus (two semelparous species), attain their maximum size in the lagoons before going to spawn in the sea. As is the case for sedentaries, the males of Pomatoschistus minutus, a migratory species that builds nests in the sea, are superior in size to the females (better growth?); the opposite is true for Anguilla anguilla. With regard to the other migratory species, which may potentially effect a number of relatively short stays in lagoons over the course of their life, the comparative approach to proportional growth in the sea and/or lagoon, for those aged 0⁺ and especially for adults, is more difficult to gauge, but it is generally accepted that growth in lagoons is superior to the growth of the individuals who remained in the sea (Lasserre, 1989).

3.5. Intra- and interspecific communication

Emission and reception of “signals” by fishes are key elements governing intraspecific interactions (gathering in shoals, attracting spawning adults, etc.) and interspecific relationships (prey–predator and host–parasite relationships, commensalism, symbiosis, etc.). Intra- and interspecific communication uses visual, chemical, acoustic and electrical channels. Depending on the environment and the fishes’ potential for emission and perception, hierarchization in the use of these signals can vary considerably. The reasons for these variations can be geoclimatic, on either a grand or a small scale, seasonal, etc. They can also be biophysiological (ontogenetic stage, sexual cycle, etc.). Moreover, as we have already mentioned, lagoons can be or can rapidly change from a monotonous sand-mud system to a highly structured, vegetation-rich system, which limits visibility and free circulation of water and vagile species. This system also provides shelter from predators and meets certain vital needs, such as rest, sit-and-wait hunting and reproduction (only in the case of sedentaries). The physical properties of the waterbody can also be modified by an excessive proliferation of plankton, especially phytoplankton (green water), and also by suspension of sediments (turbidity), which changes light transmission both quantitatively and qualitatively, and consequently limits visibility. The dissolved gas content and composition can sometimes also be altered.

3.5.1. Visual functions

The ability to distinguish shapes, contrasts and colors is essential in establishing various vital strategies for reproduction, feeding, protection, movement, etc. The morphologies (shapes and sizes) presented by fishes differ greatly from one fish to another. This visual signal is theoretically the first “warning” signal of conformity or non-conformity with the expectation of an observer encountering an object (alive or inert). What the fish sees is a determining factor in its choice of reaction in response to this object. In this domain, sedentary and migratory fish present notable divergences, and within the sedentary species, size differences between the sexes are almost universal.

In sedentaries, visual signals have a very important original role in social organization and sex life. In fact, while there is a very low degree of morphological sexual dimorphism in the nectonic non-nest-building species Atherina lagunae (Atherina boyeri), the nectobenthic non-nest-building species of the genus Aphanius, in the stickleback Gasterosteus gymnurus which is a nest-builder and the mosquitofish Gambusia holbrooki which is viviparous, the females are larger than the males. Conversely, in the nectonic Labridae Symphodus (Cr.) cinereus staitii, Gobiidae (Pomatoschistus sp. and Gobius niger) and the blenny Salaria pavo, the nest-building males are larger than females, while the so-called sneaker and satellite males are comparable in size to females. In the latter three families, the visual signal of “size” plays a major role in social hierarchization. In fact, when a sneaker or a satellite reaches a size near to that of the nest-owner, the latter drives them away vigorously no matter what their color. As well as size, some adult males develop striking characteristics that play a role in sexual attraction: the adult male Gobius niger presents a very tall first dorsal fin; Salaria pavo has a cephalic crest and anal glands (swellings) on the first two radii of its anal fin, developed to a greater or lesser degree according to its physiological state; male mosquitofish are equipped with a intromittent organ (gonopod) formed by the third, fourth and fifth radii of the anal fin. These signals allow females to evaluate the quality of a male and respond or not to his advances. In addition to size, mature female Gobiidae also send specific visual morphosignals. In fact, they can be identified by the rounded profile of the abdomen and their trapezoid urogenital papilla, with a greater or lesser degree of turgescence – that of the male is slender and less conspicuous. Similarly, in the wrasse S. (Cr.) cinereus staitii, besides the shape of the abdomen, the urogenital papilla of mature females, conical in shape and tinged with black, is very well developed, whereas that of the male is small and white-ish. Morphovisual signs carrying intra- and interspecific significance are those based on modifications to the outline and stature of the fish: dilation of the gills and use of the fins with the objective of intimidating an “unwelcome intruder”.

Even when the ambient light conditions and optical qualities of the water and substrate in lagoons are far from optimal, the colors sported by fishes often resemble a mosaic composed of pigment spots of shape, size, color and radiance that can change according to the interactions between the organism and its abiotic and biotic environment. These are long-lasting or short-lasting visual signals that are complementary to morphovisual signals. The pigment spots, given their plasticity, have a more important role, because they are more “nuanced” than the morphological signals in social communication: coordinating shoals, dominance interactions in territorial behavior, sexual selection between spawning adults. Color dynamics also have a functional meaning in terms of cynegetic and anti-predator camouflage. In lagoon-sea migratory fishes, the dominant colors are quite muted and fairly monochrome. Apart from a few darker-colored spots and the occasional permanent golden spot or stripe, such as on certain mullets and the gilthead sea bream, coloring is of a fairly uniform brownish gray. In stress situations, dark-colored transversal stripes can appear. By contrast, with the exception of the silvercolored atherines, sedentary lagoon-dwelling species present more diverse and brighter coloring in the males than in the females, especially during the spawning period; blues and reds often tend to dominate.

Color is sign of good health, as evidenced by the example of the stickleback Gasterosteus aculeatus. In this fish, the female seeks as reproduction partners males with blue irises, whose chin and throat are of the brightest red. This nuptial coloration comes from an accumulation of caroteoid pigments (astaxanthin) whose quantity correlates to the condition of the male and his reproductive capacity. The intensity of coloration thus constitutes an indication of the “quality” of the spawning male and assures her of high “mating-reproduction” success. Conversely, males with parasites, infested with the protozoaire Ichthyophthirius, present a drab-colored livery and are shunned by the females (Barber et al., 2000). In fish, it is not only colors visible to the human eye that play a role in communication.

In the stickleback, besides the red coloration of their throat and the blue color of their eyes, the males have another method of seducing females: their ability to reflect solar UV radiation (300–400 nm) due to the silver pigments on their flanks. In reciprocation, at the moment of reproduction the females present a ventral nuptial livery of a silver color, reflecting UV radiation and increasing their chances of being clearly seen by the males and then courted by them (Rick et al., 2006). Gambusia holbrooki males seek to mate by preference with large females who have a well-developed and very black abdominal spot near the anus.

It should be noted that vigorous growth of meadows and algae beds (eutrophication), and also an abundance of boulders, allows hunters to hide in ambush to wait for a prey, and conversely, allows a prey to escape a potential predator. On another note, limiting the “field of vision” leads to a reduction in opportunities for females to encounter males, and thus a reduction in reciprocal choices. In sticklebacks, it has been shown that the females’ “motivation” is seriously affected by this, resulting in fewer eggs being deposited inside nests.

3.5.2. The olfactory functions

Sensitivity to chemical messages in the form of “smells” and “tastes”, termed “chemoreception”, corresponds to the olfactory and gustative senses. Olfactory and gustative sensitivities are both based on the common solubility characteristics of chemical messages, as opposed to organisms that live in an aerial environment in which olfaction is linked to the perception of volatile molecules. Chemical messages rely on molecules emitted by organisms, facilitating various types of communication between individuals or populations of the same or different species. They are generally contained in the epidermic mucus that is rich in amino acids, in the bile, in the urine, in the sexual fluids (spermatic or ovarian fluid), or produced by specialized exocrine glands (external anal gland of the male peacock blenny). These messages, which can be attractive or repulsive, are used in the same domains as visual signals: detecting food or an enemy, choosing and recognizing habitats (homing), establishing a social hierarchy, identifying and choosing a sexual partner (sexual selection), the gathering together of individuals (shoal, lek, etc.). Chemical signals have an advantage in some lagoons, in that they are operational regardless of the light conditions, but as in the case of visual signals, certain types of pollution can disrupt chemical communication by limiting the molecule’s effectiveness by modification or substitution; the proliferation of obstacles, by slowing and changing the flow of the currents, can limit their range (as it limits the area of vision). The olfactory sensitivity of fish is very keen. The eel responds to a concentration of 3.5 × 10^–18 mol·L^–1 of β-phenylethyl alcohol, and according to recorded electroolfactograms, to a wide variety of amino acids: glycine, L-alanine, L-valine, L-leucine, L-asparagine (Sola et al., 1993).

Three major categories of messengers are distinguished:

– at the interspecific level, two types of “pheromones” have been identified: “inducer” pheromones which induce an immediate behavioral reaction and “behavior-modifying” pheromones which induce physiological changes in the receiver. According to their function, they are classed as sexual, social, alarm or defense pheromones;
– at the interspecific level, allelochemical compounds, known as “allelomones”, which primarily contribute to the localization and identification of prey or predators, and in the latter case, to the organism’s active (chemical weapon) or passive defense. A number of types are distinguished:

- “allomones” which act to the advantage of the individual emitting them (repulsive and defensive effects);
- “kairomones” which benefit the receiver (localization of a prey by a predator);
- “synomones” whose effects are mutually beneficial (mutualism);
- “apneumones”: molecules emitted by non-living organic matter.

In sedentary lagoon-dwelling fishes, the role of the sexual pheromones is very important (sexual selection). They are emitted by the testicles and ovaries, and also by the annex glands of the genito-urinary system and “external” glands (anal gland of Salaria pavo males), the development and activity of which are linked to the sexual, hence hormonal, cycle of the fish. Steroidal sex pheromones, unlike amino acids which do not trigger specific responses, present great specificity. In this domain, it should be remembered that “major histocomptability complex” genes (MHC, a group of coding genes for various proteins: antigenic peptide presentation molecules, etc.), in addition to their role in immune functions, have an influence on reproductive behavior (Penn and Potts, 1998; Milinski, 2006) and sexual selection via mechanisms involving olfaction (Milinski, 2003; Milinski et al., 2005, 2009). In Gobius niger and Gobius paganellus, there is a gland in the testicles known as the “mesochorial gland” whose steroidal secretions are eliminated in the urine and ejaculates. These secretions are pheromones which, when released in the nest, attract females and trigger egg-laying (Colombo et al., 1980; Marconato, 1980). The seminal vesicles (sperm glands) are two evaginations from the sperm duct, present on all Gobiidae (Miller, 1984, 1992; Mazzoldi et al., 2005). The lining of these vesicles produces sialomucines, various proteins, and steroids that have a pheromonal function related to reproduction (Jasra et al., 2007; Mazzoldi et al., 2011). In Salaria pavo, there are under-developed testicular glands (15% of the volume of the testicles) and a pair of “blind pouches”, distal evaginations from the sperm ducts. These pouches present a “development/regression” cycle linked to the sexual cycle (Lahnsteiner et al., 1990c; Lahnsteiner and Patzner, 1993). They are the site of the synthesis of sex pheromones. Moreover, in this species, pheromonal substances released by the anal glands, located on the first two radii of the male’s anal fin, play a crucial role in attracting females.

Chemical signals produced by specialized epidermal cells are “alarm/stress/alert pheromones”, which can also be found in the urine, and these act as warnings to the community. These emissions are particularly important when cutaneous lesions occur as a result of an attack, and can induce anti-predator flight behavior in their companions. A clear collective advantage is gained by these emissions, especially when visual signals are disrupted by water turbidity. It has also been demonstrated that these signals can attract other predators and that competition for access to the prey increases its chances of survival by 40% (Lonnstedt, 2015).

With regard to lagoon-sea migratory fishes, chemical messages carried by the currents certainly play a role in localizing the lagoon entrance and govern the fishes’ journey upstream to enter the lagoon. This is called “the call of the lake” (Heldt, 1929). Exiting the lagoon may perhaps also be partly induced and guided by chemotaxic stimuli corresponding to “the call of the sea”. In both cases, in addition to chemotaxy, rheotaxic and thermotaxic responses are certainly to be envisaged and may perhaps take priority.

3.5.3. Auditive and mechanoreceptive functions

Fish live in a complex acoustic environment. Their living environment is not homogeneous from the three-dimensional point of view: it is characterized by varied hydrodynamic movements (currents, swells, waves, etc.) generating “physical noise” which is complemented by the noise of rowing boats, motor boats and the various “biological noise“ emitted by all living beings (plants as well as animals) including the fishes themselves, their prey and their predators. The physical characteristics of the aquatic environment which is dense, viscous and very incompressible, are favorable to the acoustic transmission of sound vibrations, greater in the water than aerial transmission. Propagation speeds are approximately 1,500 m/s in water and 330 m/s in air, or a speed in water 4.5 times higher than in air – the loss of energy with distance being low (speed and absorption-extinction variables depend on salinity, turbidity and temperature).

Sensorial perception (internal ear) of sounds is an important function for fishes, especially in frequently turbid environments. Their auditive system varies according to species, from infrasonic to ultrasonic detection. Quite a few fishes produce sonic emissions (Tavolga, 1967), generally in the medium frequencies (between 100 Hz and 4 kHz). These emissions are caused by friction of anatomical parts (teeth, pharyngian plates, fins, etc.) or can result from muscular contractions (sonic muscles). Sound emission and perception are often amplified by the swim bladder which acts as a “sound box”, in conjunction with the tympanic bulla and the ear labyrinth. These emissions are thus involved in sexual encounters and courtship behavior, and in the defense of territory or a nest during spawning. They can thus have an “attractive” or “repulsive” function, but they can also have no social significance at all. A fish can sing “for its own pleasure”, but no instance of this has yet been identified in Mediterranean lagoons.

In Mediterranean lagoons, some representatives of Anguillidae and Gobiidae emit sounds, namely: Anguilla anguilla, Gobius niger, Gobius cobitis, Gobius paganellus, Knipowitschia panizzae, Pomatoschistus canestrini, Pomatoschistus marmoratus, Pomatoschistus minutus, Pomatoschistus microps, Zosterisessor ophiocephalus, etc. (Parmentier, 2013; Bolgan et al., 2015). However, it has been demonstrated that some species of Syngnathidae belonging to the geni Syngnathus (Ripley and Foran, 2007) and Hippocampus (Colson et al., 1998), which are present in other geographical sectors, can emit sounds in a variety of circumstances. The same applies to Blenniidae (De Jong et al., 2007; Coers et al., 2008).

In gobies, sound production can be caused by a sudden ejection through the gills (opercules) of the water content of the mouth. In Gobius niger, nest-owning males emit sonic signals like the “croaking of a frog” directed at a female ready to lay eggs: 11 emissions/min, for a duration of 347 ms and a maximum frequency of 107 Hz (Malavasi et al., 2008). Females can be attracted to a greater or lesser degree by the quality of the sonic message emitted by the “singer”, as well as by his size, his color, the height of his first dorsal fin and the appearance of his nest. Furthermore, in an aggression situation, Gobius niger emits “groans” (Malavasi et al., 2005, 2008; Scaggiante et al., 2005). Zosterisessor ophiocephalus emits sounds during aggressive confrontations (Ota et al., 1997, 1998, 1999b), at around 7 emissions/mn, with a duration of 210 ms and a frequency of 213 Hz (Malavasi et al., 2008a). In this fish, the frequency of the sounds emitted decreases with size by around 23 Hz per centimeter between 10 and 20 cm TL (Malavasi et al., 2003). The vocal function thus “matures” with age. One might wonder whether there is a “juvenile” vocabulary and a different “adult” vocabulary, and also whether there is a learning process from contact with elders. In Knipowitschia panizzae, Pomatoschistus canestrini, Pomatoschistus marmoratus and Pomatoschistus minutus, males courting females emit a wide variety of sounds, quite powerful (12–15 dB), lowpitched or strident, varying in length, and even some vocalized sounds (Lugli and Torricelli, 1999; Malavasi et al., 2008, 2009). These emissions continue in the nest until the spawning act (prespawning sounds). The soundproducing mechanisms in gobies are as yet unknown and suggestions in this regard are debatable (Kasumyan, 2008). It has been demonstrated that depositing silt and sand on a lamellibranch valve, used as a nest by certain Pomatoschisus sp., amplifies the sounds emitted by nest-building males, thus facilitating communication between “male and female” and between “male and malicious intruder”. Malavasi et al. (2008) investigated whether there is a relationship between the molecular phylogeny of seven species of goby, their habitat and the type of sounds they emit.

Another issue that is sometimes raised concerns the existence of local dialects in gobies, as found in birds and mammals, given the isolation of sedentarized lagoon-dwelling populations. This phenomenon has been identified in Gobius paganellus, with individuals from Brittany not having the same “language” as those from the Venetian lagoon (Parmentier et al., 2013).

In two families that are present in Mediterranean lagoons and are considered “mute”, the potential to emit sounds has been reported in two exotic Syngnathidae, Hippocampus zostera and H. erectus (Colson et al., 1998), by “displacing” the “coronet” bones, the unpaired bones located at the back of the cranium. The Blenniidae Chasmodes bosquianus (Tavolga, 1958), Parablennius sanguinolentus and P. parvicornis (De Jong et al., 2007; Coers et al., 2008) also emit sounds. Finally, the eel, a physotomic fish, can produce sounds via its digestive tracts, by expelling the air present in its swim bladder (Kasumyan, 2008), but the meaning of these noises is unknown.

Mechanoperception presents no particularity. Its seat is the lateral line, a specialized system for the perception of low to very low frequencies. This system makes it possible to detect every living or non-living object that generates waves by agitating itself or moving from one place to another. It also enables the fish, through perception of the waves it produces, reflected from a foreign body, to situate itself in relation to that foreign body. All or part of the body of the fish, usually the fins, vibrates during courtship or in situations of aggression. “Vibration is communication”; these vibrations carry specific messages that inform a partner or an enemy of their intentions.

3.6. Ecological genetics

Sedentary lagoon-dwelling species, due to their ecobiology and behavior, are good subjects for research on mechanisms that have created genetic divergences between different lagoon populations, and also between them and the marine populations from which they are presumed to have originated; in certain cases, the existence of lagoon endemism can be envisaged. Moreover, they provide useful models for understanding sea/lagoon and interlagoon genetic divergences (Lemaire et al., 2000, 2004–2005; Chaoui et al., 2009; Guinand et al., 2016; Boudinar et al., 2016), and also changes to the diversity of gene pools of local populations under the influence of changes that have an impact on the environments (Cimmaruta et al., 2003; Angeletti et al., 2010). It should be borne in mind that genetic scientists are still trying to find the “candidate gene” related to migratory behavior.

3.6.1. Sedentary species

Sedentary lagoon-dwelling populations are generally fairly well isolated from each other (natural geographical and anthropic barriers) and from the marine world (canal links to the sea are often narrow and shallow). Nevertheless, the sea can provide interlagoon continuity and we should not ignore the fact that coastal currents, depending on their direction, are likely to promote, or conversely to impede, exchanges between certain lagoons. Moreover, in addition to “geo-geographical” and hydrological isolation, there is also “ecobiological and behavioral” isolation which limits the potential for active (swimming) or passive (floating) emigration and immigration at every stage in the life of sedentary lagoon-dwelling fishes. In fact, these are benthic or nectobenthic and present mediocre swimming abilities. Their eggs are either free benthic (Aphanius sp.); or attached to a variety of supports such as rocks, mollusc shells, macroalgae, etc. (atherines, gobies, blennies, sticklebacks); or incubated by one of the parents (syngnathids, seahorses, mosquitofish). Moreover, the neonates (larvae, postlarvae, prejuveniles) are not easily “dispersible”; their ability to float is mediocre and their life as planktonic larvae is relatively short, and even virtually non-existent for syngnathids, seahorses and mosquitofish – neonates of these species are however recorded in the neuston (Vandendriessche et al., 2005; Pérez-Ruzafa et al., 2004). It is therefore difficult for these species to escape and for their stocks to be renewed by inputs external to their host lagoon. From a temporal point of view, this isolation is very relative, because most lagoons are ephemeral areas (becoming infilled, or sand spit destruction) and a good number of them, such as the lagoons in Languedoc (Gulf of Lion, France) which first began to form in Roman times with acceleration since the 16–17th centuries, communicate with each other via the Rhône-Sète Canal (sometimes called the “Canal des Étangs” or Canal of the Lakes) built between 1700 and 1879. Finally, there is a rapid turnover in their population, all species being short-living. Some semelparous annual species (Pomatoschistus sp., Gambusia sp., Gasterosteus aculeatus (Gasterosteus gymnurus), Syngnathus abaster) can beget two generations, and perhaps more, during their single spawning season.

Sedentary lagoon-dwelling fishes, while extremely euryvalent, are subjected, alongside relatively stable and long-lasting lagoon factors, to the destabilizing, stressful hazards of certain ephemeral environmental conditions (hydroclimatic changes: temperature, salinity, eutrophication, sedimentation (turbidity), overfishing, etc.). These changes, of varying degrees of intensity and persistence, induce or exacerbate certain selection factors, causing genetic modifications, often rapid, through the erosion of allele frequencies (Hoffmann and Willi, 2008). This is the case with Aphanius fasciatus in the Tarquinia lagoon in Italy where, following a change in the environment, allele diversity of the population fell from 1.88 in 1998 to 1.54 in 2003, in the same period (five years) as a reduction in heterozygosity from 0.169 to 0.126 (Angeletti et al., 2010). This level of reduction in genetic diversity can lead to a loss of overall “adaptive plasticity” in the population, which is detrimental to its survival. However, it is often difficult to clearly identify the mechanisms that cause a loss of genetic biodiversity (allele variability) because they are numerous. In addition to those mentioned above, we could add:

1) fluctuations in genetic flux, linked to geographical and behavioral factors;
2) demographic changes, such as a reduction in numbers due to pollution, overfishing or limited reproductive success (Leonardo and Sinis, 1998);
3) genetic drift;
4) selection (Angeletti et al., 2010; Willi et al., 2007; Frankham, 1996); etc.

Finally it should be noted that, in extreme lagoon conditions, within a population, “genetically close” individuals tend to form groups according to their “resistance” to adverse conditions (family groups), and interbreed, leading to deleterious consanguinity (Eanes, 2002; Gysels et al., 2004a).

Genetic research (electrophoresis of enzymatic systems, mtDNA, cytochrome b, rRNA, microsatellites) has shown that major intraspecific divergences sometimes exist between sedentary lagoon-dwelling populations (individuals who reproduce in lagoons) and those residing permanently in the sea, so that lagoons are regarded as “melting pots” where species are “in the making” (in statu nascendi). This is certainly what Tougard et al., (2014) meant, in relation to Pomatoschistus microps, when they said that, according to their initial genetic research, the populational structuration, and thus differentiation within the species, is “an ongoing process”. Likewise, Focant et al. (1991, 1992, 1993), Cammarata et al. (1996), Trabelsi et al. (2002a, 2002b) and Boudinar et al. (2016) demonstrated the existence of a set of lagoon-estuarine species and a set of marine species within the Atherina boyeri complex. These findings, obtained in the western Mediterranean basin, were corroborated by studies in the eastern basin and the Adriatic Sea (Congiu et al., 2002; Klossa-Kilia et al., 2002, 2007; Astolfi et al., 2005; Mauro et al., 2007; Milana et al., 2008; Kraitsek et al., 2008; Francisco et al., 2008, 2011).

It is now accepted that there is one lagoon species, on the Mediterranean coasts Atherina lagunae (Trabelsi et al., 2002), and two marine species, Atherina boyeri (Risso, 1810) and Atherina punctata (Trabelsi, 2002). Although we have no genetic data, we can also accept that the species Symphodus (Crenilabrus) cinereus (Bonnaterre, 1788) is represented in deep lagoons by a subspecies S. (Cr.) cinereus staitii according to Quignard (1966), to which we can assign the species rank of S (Cr.) staitii (Nordmann, 1840). The same pattern seems be repeated for anchovies, in that there is one coastal-lagoon species (white anchovy), Engraulis russoi (Duzetto, 1947) syn. Engraulis albidus (Borsa et al., 2004) and one species found in the open sea (blue anchovy), Engraulis encrasicolus (Linnæus, 1758).

According to these examples, lagoons seem to be sites that promote speciation in spite of their relative “geo-sedimentary” youth. This is true of species with relatively little mobility, and also for mobile species with high passive dispersal potential (pelagic eggs and larvae) such as in anchovies. The white anchovy is less closely linked to lagoons than the lagoon-dwelling atherine or the corkwing wrasse and Tortonese’s goby, whose eggs are attached to the substrate.

Focusing particularly on benthic organisms, Riggio and Chemello (1992) consider, for instance, that the Stagnone di Marsala lagoon is sufficiently separated from the sea to give rise to microevolutionary phenomena. On the other hand, Guelorget and Perthuisot (1983) believe that the paralic domain is “a kind of potential reservoir of lineages that could replace marine or continental lineages in the event of a crisis leading to the disappearance of the latter”.

3.6.2. Migratory species

The fact that lagoons provide a temporary refuge for many species and function as nurseries and feeding grounds for marine species, means that these organisms have to develop adaptation strategies, as do sedentary lagoon-dwellers, in response to the multiple threats and constraints of the lagoon environment in order to temporarily survive in sometimes extreme conditions. These strategies are very costly in energy, which should have repercussions on mortality and growth rates and on reproductive performance, but apparently there are none (see section 3.4). It should be stressed that migratory individuals, who live in lagoons at one period in their lives, represent a greater or lesser proportion of the population from which they originated and there are notable divergences between deep lagoons and laminar lagoons in hosting potential. The determinism of the migratory behavior of a fraction of a given marine population is still imperfectly understood. Furthermore, a major problem in stock management is knowing whether individuals from this “migratory phase” are in some way different from those who do not undertake such migrations. From the point of view of genetics, the existence of genetic differentiation or greater heterozygosity has been recognized, as has the existence of exclusive alleles in lagoon migrating individuals compared to strictly marine individuals of the population (Gonzalez-Wangüemert et al., 2004, 2006; Chaoui et al., 2012; Guinand et al., 2016). The presence of lagoon-sea migratory coastal subpopulations originating from strictly “marine” populations has been envisaged, for example in the gilthead sea bream Sparus aurata (Chaoui et al., 2012; Guinand et al., 2016) and the sea bass Dicentrarchus labrax (Lemaire et al., 2000, 2004–2005; Guinand et al., 2015). Are these genetic divergences premigratory? Or are they a result of the “ongoing” process of natural selection making an impact new recruits colonizing the lagoons, which causes an increase in allele frequency, enabling the migrants to adapt to lagoon systems? It should be noted that some of these alleles are rare in individuals who do not undertake lagoon migrations with regard to the gilthead sea bream populations of the Gulf of Lion and the Algerian coast (Chaoui et al., 2009). On the subject of allele frequencies, it would make sense to compare gilthead sea bream visiting the “dilution”-type lagoons such as Mauguio (France) to those visiting “concentration”-type lagoons such as Bardawil (Egypt). The knock-on effect on the species’ genetic pool, when the adults return to the sea to reproduce, may play a major role in the adaptation of these species to climate change.

3.6.3. Other structural factors

In the Mediterranean, phylogeographic studies of a number of migratory and sedentary lagoon-dwelling species have highlighted the role of straits (Gibraltar, Siculo-Tunisian, Messina, Otranto) as “bottlenecks” that limit migratory and thus genetic flux (Naciri et al., 1999; Astraldi et al., 1999; Bianchi and Mori, 2000; Lemaire et al., 2004–2005; Patarnello et al., 2007; Souche et al., 2015). These “bottlenecks”, often described as “breakpoints” in the flux, have promoted the emergence of genetic and morphoanatomical divergences (Bahri-Sfar and Ben Hassine, 2009) between the Atlantic and the Mediterranean (Naciri et al., 1999; Lemaire et al., 2004–2005, Patarnello et al., 2007; Souche et al., 2015), between the east and west Mediterranean (Bahri-Sfar et al.; 2000, Souche et al., 2015) (Figure 3.16), and between the Mediterranean and the Atlantic (Stefanni and Thorley, 2003; Gysels et al., 2004b; Souche et al., 2015).

Thus, the genus Aphanius is represented on either side of Gibraltar, on the Spanish coast, by two distinct species (Aphanius iberus in the east and A. baeticus in the west) and not by only one, as had been accepted before the genetic research of Perdices et al. (2001) and Doadrio et al. (2002).

The Siculo-Tunisian straits, between the east and west Mediterranean, represent a genetic breakline (Cape Bon-Mazara del Vallo). This has been demonstrated in the case of sedentary lagoon-dwelling fishes, such as the gobies Pomatoschistus tortonesei and P. marmaratus (Mejri et al., 2009, 2011), and also migrators such as the flounder Platichthys flesus (Borsa et al., 1997) and the sea bass Dicentrarchus labrax (Bahri-Sfar et al., 2000; Souche et al., 2015). Between the Mediterranean and the Adriatic, divergences have been recorded for the goby Pomatoschistus minutus (Stefanni and Thorley, 2003; Gysels et al., 2004b), Pomatoschistus marmoratus (Mejri et al., 2011) and the sea bass Dicentrarchus labrax (Souche et al., 2015). The structure of these bottlenecks (canal and threshold effect) and their hydrodynamics, together with the hydroclimatic differences that prevail on each side of the Siculo-Tunisian Strait and the Strait of Otranto, are determining factors in the genetic structuration of populations. Sometimes, however, only the presence of hydroclimatic fronts and currents (Astraldi et al., 1999) can explain a certain “isolation” (at least partial autonomy) of populations. This certainly seems to be the case for the populations of Pomatoschistus marmoratus in the north of the Aegean Sea, which diverge genetically from those of the south-east coast (Gulf of Gabes) in Tunisia (Mejri et al., 2011). In this case, the Aegean thermohaline front, resulting from the meeting of waters from the Black Sea with those from Levant, can limit exchanges between populations. Conversely, Tougard and Berrebi (2009), based on a comparative study of the mtDNA of populations of Pomatoschistus microps and Pomatoschistus marmoratus, conclude “that the populations of sedentary gobies are not so strictly isolated” and add that “no clear structure can be identified among populations of Mediterranean lagoons”.

To explain the genetic structure, we should not underestimate, even slightly, the more global events, i.e. movements of the continental plates, the Messinian crisis or crises, the alternation of cold and hot periods, accompanied by variations in water level, bringing about changes in the relationship between territories (Penzo et al., 1998; Huyse et al., 2004). Only changes in the relationship that used to exist between Sicily, Sardinia and Tunisia can, for instance, account for the genetic divergences and convergences between the Sicilian, Sardianian and North Tunisian populations of Aphanius fasciatus. In the same way, Tigano et al. (2006) identified genetic diversions between the south-eastern and western locations in Sicily. The latter have affinities with those of Malta and Tunisia. Findings obtained by Pappalardo et al. (2008) in this sector show genetic divergences between the east Sicilian populations (Longarini and Marcellino) and the west Sicilian populations (Trapani and Marsala). The latter have affinities with the Sardinian populations. For the lagoon-dwelling atherine, Milana et al. (2008) demonstrate that the Sicilian population in Marsala differs genetically from those in the neighboring lagoons. The latter findings corroborate those of Congiu et al. (2002), Astolfi et al. (2005) and Francisco et al. (2008) concerning the atherine, and those of Rossi et al. (2006) with regard to the gilthead sea bream. These examples show the complexity of the relationships between lagoons in the key Sardinia-Sicily-Tunisia sector. According to these authors, affinities detected should be seen in relation to paleogeographic and paleoclimatic events (Miocene and Pleistocene) in this region. Research by Astolfi et al. (2005) on the atherine (Atherina sp. = Atherina lagunae) clearly highlights the complexity of the factors involved in the determinism of the genetic architecture of Mediterranean lagoon populations. Based on a study focusing on the genetic variability (mtDNA) of this species within and between seven lagoons in the western Mediterranean, three in the Adriatic, one in the Tage estuary and one in the Danube, these authors identify strong structuration and clear interlagoon fragmentation. They recognize five major sets:

1) the Siculo-Tunisian canal with Sicily and Tunis;
2) the Adriatic;
3) the Tyrrhenian Sea;
4) the north-west Mediterranean with the Etang de l’Or (Mauguio lagoon, France) and Tage (Portugal);
5) the Black Sea (Danube).

This structuration can only be explained by making reference to geological history, hydrology and climatology.

Regional, and especially local, factors (lagoon confinement, hydroclimatic and hydrodynamic characteristics, competition, predation, etc.), play, in synergy with the biobehavioral properties of sedentary lagoon-dwelling fishes which limit their dispersal potential (mediocre swimmers, non-pelatic eggs, moderately pelagic larvae), a role in the genetic structuration of lagoon populations, and thus in their genetic diversity, with these factors being very major and specific to each lagoon (Frisoni et al., 1983). For this reason, genetic architecture can differ between lagoons that are very close geographically. Thus, Maltagliati et al. (2003) identified morphological (development of the caudal fin) and genetic differences between two contiguous “populations” of Aphanius fasciatus (Pilo lagoon and an adjacent lake in Sardinia), which they believed were the result of difference in predatory pressure. Based on an electrophoretic study (allozymes), Doadrio et al. (1996) identified a significant correlation between the salinity of the environment and the genetic variability (heterozygoty and polymorphism) of Aphanius iberus on the east coast of Spain; the lowest genetic variability values were attained by the populations living in the most saline lagoons. However, we should not underestimate historical geological factors, for example the possibility that during the Messinian crisis, hypo- and hyperhaline “lagoon refuges” may have hosted euryvalent species and, in doing so, initiated genetic divergences that we term Messinian (relict species) before their redeployment when the waters returned to the Mediterranean. The same scenario of lagoon isolation/genetic divergences in these “Messinian relict” populations may have occurred during the ice ages. It should be remembered that the Mediterranean sector, and thus the lagoons, has been the scene of violent crises. Based on a study of sediments of the Mar Menor (Spain), Dezileau et al. (2016) recognize eight “extreme” lagoon events on the following dates: 5,250, 4,000, 3,600, 3,010, 2,300, 1,350, 650 and 80 years BP (C¹⁴ dating). The present-day genetic structure of lagoon populations is thus a result of this long and tumultuous history inscribed in the ADN of individuals, still preserved and adapted in the original areas that are the lagoons of today. Tougard et al. (2014) found that the formation of lagoons during the Holocene has had a certain degree of impact on the structuration of populations of Pomatoschistus microps. These authors confirm that Gibraltar is not, as far as this species is concerned, a significant “phylogenetic barrier”, but they identify, in the Atlanto-Mediterranean area, four major lineages, each characterized by a large number of “private” haplotypes including a prelagoon “Mediterranean” haplotype. Subsequently, the formation of lagoons permitting small “subpopulations” to be isolated led to the appearance of new haplotypes that could be characteristic of each of them. The divergences found in lagoons these days are thus recent and have been rapid, and the differentiation process is still ongoing. It should be stressed that it is impossible to generalize the role of straits, fronts and lagoon conditions in the structuration of intralagoon and interlagoon genetic biodiversity. Counterexamples are not uncommon. Thus, for species that can be considered to be ecobiologically close, such as seahorses and syngnathids, significant divergences can be noted on the phylogeographical level. Thus, Ben Alaya et al. (2011) identify a significant divergence (mtDNA) between the Syngnathus abaster population of the Tunis North lagoon and the Mauguio lagoon (Languedoc, France). However, in the case of the seahorse Hippocampus hippocampus, Woodall et al. (2011), using two mitochondrial DNA markers, found no divergence within the Mediterranean, or between the Mediterranean and the Iberian Atlantic: “the absence of significative genetic differentiation among populations of Hippocampus hippocampus… is surprising”. According to these authors, the same schema is found in Hippocampus guttulatus (H. ramulosus). Likewise, Mejri et al. (2011) find no genetic divergence (mtDNA) between three populations of Pomatoschistus marmoratus from the western Mediterranean basin (Thau and Vaccarès lagoons in France, Bizerte lagoon in north Tunisia), while in the eastern basin they record, for the same species, significant interpopulational divergences between the Gulf of Gabes and the north of the Aegean Sea. Congiu et al. (2002), based on a study focusing on individuals of Atherina sp. = Atherina lagunae, from 11 populations in Adriatic and Tyrrhenian lagoons and two freshwater lakes (Bolsena and Trasimeno, Italy), show that there are no specific markers for these lagoons (no strait effect?), but that the Marsala lagoon (Sicily) indicates clear differentiation, a quirk linked to this sector’s “geo-geographic” past.

There are many applications for an understanding of the genetic affinities between populations of the same species: fish farm stock management, management aimed at the conservation of populations of threatened or endangered species, management of repopulation programs, etc. Genetic research carried out on the Spanish coast focusing on Aphanius iberus is evidence of this. Doadrio et al. (1996), based on an electrophoretic study (allozymes), recognize three OCUs (Operational Conservation Units = Distinct Conservation Units) for Aphanius iberus, on the Spanish Mediterranean coast (OCU 1: Catalonia; OCU 2: Levantine; OCU 3: Murcia). From 16 samples taken from six localities (Emporda, Salou, Ebro Delta, Albuixec, Santa Pola, Mar Menor) and six populations reintroduced following local extinction, Araguas et al. (2007) define on genetic bases (allozyme electrophoresis) six operational management units (OCUs) for this species, which is considered to be “under threat”. In the Adriatic, Cano et al. (2008) recognize two “distinct conservation units” for the stickleback, one linked to Neretva and the other linked to the Zeta.

We have previously considered a number of factors structuring lagoon settlements such as water quality (salinity, temperature, etc.), substrate type (sandy, rocky, vegetation cover, etc.), the steepness of physicochemical gradients, confinement, etc., but until recently hydrodynamic factors have rarely been taken into account in the models put forward. However, there is no doubt that the more or less passive transportation of certain planktonic stages of the lifecycle of animals between sea and lagoon and then interlagoon, plays a major role in their procurement and dispersal in lagoons, and thus in the spatial organization of specific “assemblages”. Digital models are currently being developed, based on the concept of “connectivity” between the significant points of a lagoon-sea/intralagoon hydrographic system. Some of these models consider only the passive physical dispersal of elements (larvae), while others take into account potential interaction of planktons with their ecophysicological needs, thus habitats favorable to their occupation. Taking this viewpoint requires a good understanding of planktonic type, duration of stages, etc. Approaches such as this have been implemented focusing on the Venetian lagoon (Umgiesser et al., 2004; Ghezzo et al., 2010, 2015), the Mar Menor (De Pascalis et al., 2012; Ghezzo et al., 2015) and other lagoons (Umgiesser et al., 2014).

Unquestionably, on the basis of these data, lagoons form, from an ecological point of view, a system that has definite consequences for the structure of the populations living there, and this has given rise to the interlagoon diversity observed as much from the morphoanatomical as from the genetic point of view. These lagoons also have repercussions on the genetic structure of the species who visit them temporarily (migratory species).