What happened to these fish, as they adapted to living exclusively in freshwater

Sticklebacks

In what will prove to be an exciting project, David Kingsley's laboratory developed a genome-wide linkage map for the threespine stickleback, Gasterosteus aculeatus, which they used to study the development of body armour (dermal skeleton) and feeding morphology in benthic (deep) and limnetic (lake-dwelling) pairs of sympatric species in six lakes in British Columbia, Canada. Benthic species have reduced body armour, increased body depth and decreased numbers of gill rakers, in contrast to limnetic species with their increased body armour, more slender bodies and increased numbers of gill rakers.

Their initial finding (Peichel et al., 2001) is that independent chromosome regions control plate number, spine length and gill raker number. No major QTL contributes significantly to the number of long gill rakers, leading to the conclusion that many genes, each of small effect, control the number of rakers, à la the murine mandible as modeled by Bailey (1986). Two QTLs contribute significantly to the number of short gill rakers, explaining two-thirds of the variance in number. QTLs belonging to several linkage groups also influence the length of dorsal spines one and two, the pelvic spines, and the number of lateral plates, explaining 17–20 per cent of the variance in these characters.

Threespine sticklebacks are covered with dermal plates. Plate number is polymorphic, the number of plates varying between populations in a predictable manner. Different variants – we are dealing with a single species – are known as lateral-plate morphs and result from genetic polymorphism for plate number in combination with paedomorphosis, which influences the duration of ontogeny when the plates are laid down. Using sticklebacks from an insular lake in the Queen Charlotte Islands, British Columbia, individual fish (especially large adults) with high or low plate numbers were shown to have increased asymmetry of plate numbers, a phenomenon these researchers linked to parasite burden. Recent QTL analysis shows that a single major locus contributes to most of the variation in both number and patterning of lateral plates. Additional QTL analysis would allow genetic and environmental influences to be partitioned and/or integrated further.e

Different populations of threespine sticklebacks also show various stages of reduction of their pelvic girdles. Again, the mechanism is paedomorphosis, as truncation of development removes individual elements one at a time, or removes entire pelvic girdles. Pelvic reduction, as analyzed in populations in 179 lakes, requires the combination of absence of predatory pike from the lake and low Ca++ concentrations in the water, indicating how unrelated environmental variables can interact and influence a part of the skeletal system.f This system is ripe for a QTL analysis.

The combination of QTL analysis and identification of genes of major or minor effect is the most exciting genetic approach to skeletal growth and development available today.

Add in environmental and life history components and the prospects for understanding skeletal development in the real world looms on the horizon.

These are exciting times to be a skeletal biologist.

II WHY STUDY STICKLEBACK AGGRESSION?

Among the stickleback species, male three-spined sticklebacks (Gasterosteus aculeatus) have the highest levels of aggression, the most pronounced breeding coloration, and the best-developed morphological defense mechanism against vertebrate predators (Bell and Foster, 1993). It has been suggested that the relative freedom from predators has facilitated the change from breeding in areas of dense vegetation to the open (Morris, 1958; Wilz, 1971; Wootton, 1976, 1984), where competition for females would be more intense. This habitat shift may have permitted the evolution of male traits that enhance competitive abilities (e.g., high aggression levels: see the following) and attractiveness toward females (e.g., red breeding coloration: Milinski and Bakker, 1990). Furthermore, this fish species is remarkably variable for a wide array of features, including the aforementioned traits, and is actually a large complex of differentiated allopatric populations and biological species (Bell, 1984; Bell and Foster, 1993). Variation among three-spined stickleback populations is (like the exaggeration of the aforementioned male traits) often interpreted in terms of adaptation (Bell and Foster, 1993), but the genetics of most traits has not been studied.

In studying the evolution of stickleback aggression, the assessment of heritable variation in male territorial aggression would be a necessary first step, but would be of limited value in understanding its evolution because natural and sexual selection do not act on single traits (e.g., Lande, 1988). Our understanding of the evolution of territorial aggression would gain substantially by knowing the important genetic relationships between territorial aggressiveness and other traits. Through reproductive physiology, territorial aggression has obvious links with other aspects of reproductive biology such as male courtship and coloration (e.g., Munro and Pitcher, 1983; Villars, 1983). Laboratory and field research on stickleback aggression and the situations in which it may occur has been strongly biased toward territorial aggression of reproductively active males for reasons of both conspicuousness and interest. Its occurrence among juveniles or subadults and among adult females is less well known. These other forms of aggression can be very pronounced (Bakker, 1986, 1993a; Bakker and Feuth de Bruijn, 1988) and cannot be neglected when studying the evolution of aggression in this species.

Thus, sticklebacks of both sexes show aggressive behavior in a variety of contexts. Consequently, aggression is subject to diverse selective forces. This diversity can be expected to be reflected in the underlying causal mechanisms and genetic bases of different forms of aggression, making stickleback aggressiveness a suitable example with which to study multivariate evolution. I have used multiple artificial selection experiments to evaluate the extent of common genetic control of different forms of aggression and to examine the underlying hormonal influences on aggressive behavior.

Evolutionary and Natural History of the Threespine Stickleback

The threespine stickleback, Gasterosteus aculeatus, is a small fish (3–8 cm in length at breeding) that is widespread in holarctic oceanic and coastal freshwater habitats. It is a member of the family Gasterosteidae (the sticklebacks), a monophyletic group most closely related to the marine family Aulorhynchidae, which is paraphyletic with respect to the sticklebacks. The Gasterosteidae includes five genera, only one of which, Gasterosteus, includes more than one extant, widespread species (G. aculeatus and G. wheatlandi). Ancestrally oceanic, this family includes one marine genus (Spinachia), three genera that are found in both marine and freshwater (Apeltes, Gasterosteus, Pungitius), and one that is found exclusively in freshwater (Culea). Because the family includes mostly monotypic genera, it is not likely to provide great insight into the nature and pattern of evolutionary transitions in a standard phylogenetic framework, but phylogenetic comparative studies have identified behavioral phenotypes likely to be evolutionarily ancient in the family.

The best-studied species in the Gasterosteidae is the threespine stickleback, a species with characteristics that have afforded it model system status (indeed, supermodel status) in evolutionary developmental genetics, including behavior (Wootton, 1976; Cresko et al., 2007; Östlund-Nilsson et al., 2007). This small fish exhibits extensive population-level adaptive differentiation in freshwater, much of which must have occurred in the last 15,000 yr as glacial ice retreated. With the withdrawal of ice from coastal regions, oceanic threespine stickleback invaded newly created freshwater habitats (Fig. 1) radiating into uncountable freshwater populations.

This radiation possesses two features that make it especially valuable for the evolutionary study of behavior (Bell and Foster, 1994). The first is that living oceanic fish closely resemble those which gave rise to the recent freshwater radiation, making it possible to understand a trait’s ancestral condition relative to the postglacial freshwater radiation. Second, upon invasion of freshwater habitats, the oceanic colonists gave rise repeatedly, and independently, to similar populations in similar habitats that differed predictably from those found in divergent habitats (parallelism). The consistency of phenotype–environment associations among freshwater populations independently derived from oceanic ancestors suggests that some aspects of genetically based divergence in phenotypes, including elements of behavior, are adaptive. Hence the term ‘adaptive radiation’, in which the first term reflects the role of natural selection, and the second, the diversification in fresh water.

Parallelism is apparent in the loss of armor in freshwater populations. This is likely caused by reduced availability of calcium relative to oceanic environments, decreased predation by vertebrates, and increased predation by invertebrates. Most freshwater populations have lost the heavy posterior flanking plates characteristic of oceanic fish (Fig. 1, center image) after invasion of fresh water. In some freshwater lakes free of piscine predators, stickleback have also lost the forward-most plates that support the first dorsal spine and paired pelvic spines on the underside of the fish. The ancestral complex provides a spiny, robust defense against some vertebrate predators. In addition to the loss of armor in predator-free environments, stickleback can exhibit different behavioral responses to attack than do those from populations where predatory fish are abundant.

A second axis of ecotypic divergence in the stickleback radiation in northwestern North America is the benthic-limnetic divergence associated with foraging habit (Fig. 2). Extreme and limnetic forms are best known in the small number of cases in which they are found as lake-dwelling pairs of benthic and limnetic species that have not been given named species status for a number of reasons. Benthic and limnetic ecotypes are also found alone in lakes, and the resident ecotype is predictably associated with lake characteristics. The correlations between ecotype and lake characteristics offer insight into the ecological causes of the differences in morphology and behavior of the two forms. Stickleback in small, shallow lakes have evolved a deep-bodied (benthic) form specialized for feeding on benthic invertebrates, while those in deep, oligotrophic lakes evolved a slender form (limnetic) adapted for feeding on plankton (Fig. 2). Stickleback of these two ecotypes exhibit predictable differences in foraging and reproductive behavior that are linked to the habitats in which they live, but that can only be fully interpreted in an evolutionary context with reference to the ancestral condition.

Certain aspects of the behavior of stickleback seem to be characteristic of oceanic populations and also appear in most freshwater populations. In early spring, males move into shallow littoral regions of marine, brackish, or freshwater habitats where they establish territories and build tubular nests of organic material that is collected from the substratum and glued together with kidney secretions. During this period, males may develop the red coloration usually considered typical of a courting male (Fig. 3), although in oceanic and benthic populations, drab coloration can be retained well into the courtship phase of reproduction, or even until the onset of parental care. Once the nest is completed, males court receptive females, which are distinguishable by their distended abdomens, solitary behavior, and characteristic head-up postures when soliciting males. If the female responds positively to the male, he leads her to his nest, inserts his nose in the entrance, and turns on his side, ‘showing’ the nest entrance. If the female enters the nest, the male presses his snout against her tail, which protrudes from the nest, ‘quivering’ until the female leaves the nest. If the male is not distracted from quivering, the female usually spawns in the nest, releasing all of her eggs. She then resumes foraging and clutch production. Females can mature clutches at approximately weekly intervals.

Courtship behavior of male threespine stickleback is usually considered to be the prominent, ‘species-typical’ zig-zag dance, in which the male swims rapidly toward an approaching female, with forward progress interrupted by a pronounced ‘jumps’ from side to side. A second courtship behavior, originally interpreted as ‘interrupting courtship’ and deterring a female from approaching the male’s nest, is termed ‘dorsal pricking’. When performing this behavior, females are positioned above the male, and both meander erratically, with the male occasionally pressing backward and up, pushing his erect dorsal spines into the belly of the female. In early laboratory observations, males were described as initiators of the behavior. However, in many stickleback populations (primarily cannibalistic populations), females initiate this behavior, and it is the primary form of courtship. The overall pattern of courtship in cannibalistic populations is less conspicuous than in noncannibalistic populations, presumably because cannibalistic foraging groups detect nests as a consequence of activity at the nest.

Males spawn multiple times, usually making the transition to parental behavior in 24–36 h. They provide all subsequent care of young for a 7–14 day period (temperature dependent), modifying the nest to provide for changing needs of the young, removing damaged embryos, fanning the nest to avoid sedimentation and to provide well-oxygenated water, and defending the young from predators. Predators on young can include leeches, sculpin, and other fish. Conspecifics – most often large groups of females between reproductive bouts – may also attack nests and consume young. Cannibalistic conspecifics overwhelm the defenses of the males through sheer numbers, with the result that direct attack by males is unlikely to deter them. Instead, males in populations with abundant cannibals perform diversionary displays – conspicuous displays that, if effective, attract approaching groups to the male and away from their vulnerable nests. Like the diversionary displays of birds, these conspicuous displays appear to include elements of behavior co-opted from other contexts, and subsequently ritualized or made more conspicuous in the new context of the diversionary display (Foster, 1995).

As the fry hatch, males cease fanning, and enter a guarding phase in which they protect the nest, and, in some populations, retrieve fry that stray from the nest area. Free swimming fry leave the territory and males may initiate a new nesting cycle. Fry may feed in shallow water for a month or so before moving into deeper water, or before migrating from fresh or brackish breeding areas back into salt water. First reproduction typically occurs at 1 or 2 years of age, rarely lasting beyond year 3.

The rise of threespine stickleback as a model species for behavioral studies lies in their small size, relative ease of rearing, and willingness to complete nesting cycles and respond to environmental challenges such as predators in arenas ranging in size from 7-l aquaria to experimental ponds. They respond readily to appropriate dummies and to video images, and in most populations are sufficiently bold to behave normally after a short period of adjustment to a novel environment, and when under scrutiny by human observers in both laboratory and field. The fact that stickleback populations comprise an adaptive radiation with a known, extant ancestral type offers an unusual opportunity to address questions concerning the evolution of behavioral phenotypes, and the recent development of molecular tools offers unparalleled opportunities to meld the four levels of ethological study, described later, that were identified by Niko Tinbergen.

4 Sticklebacks from populations experiencing higher levels of predation engage in more risk-taking behaviors than their counterparts from safer environments

As mentioned above, sticklebacks are widely distributed throughout the northern hemisphere and have a penchant for rapidly adapting to their local environments. Populations that occur in similar environments have evolved similar behaviors. One of the most important selective forces shaping stickleback populations is predation pressure. That is, some lakes contain many predators which prey on sticklebacks (“high predation”) while other lakes are relatively predator-free (“low predation”). We have been comparing the risk-taking behaviors of sticklebacks in a set of populations in Scotland that vary in predation pressure (Bell et al., 2009). Scotland is especially well suited for studying variation in risk-taking behavior in response to predation pressure because the country is teeming with postglacial waterbodies. Many lochs and ponds have been isolated for long enough (up to 15,000 generations) to independently evolve adaptations to high or low predation pressure and to become genetically differentiated from each other, but still capable of interbreeding (Malhi et al., 2006).

Interestingly, sticklebacks from areas where there are high levels of predation tend to be more risk-prone (i.e., they show higher levels of risk-taking behaviors) than their counterparts in safer environments (Huntingford and Coulter, 1989; Huntingford et al., 1994; Walling et al., 2003, 2004). This pattern has been documented in other small fish species as well, such as guppies (Magurran, 1986). While risky behavior in a dangerous environment might seem nonintuitive, this result is predicted by life history theory. The reason is that small individuals are especially vulnerable to predation, so when predation pressure is high, individuals that grow quickly will be favored because they are not small and vulnerable for long. Therefore, risk-taking behaviors that improve growth rate such as active foraging and aggression that results in access to resources should be favored when predation pressure is high (Mangel and Stamps, 2001). It is also worth noting that increased levels of risk-taking behaviors in humans have been documented in harsh or impoverished environments (Farrington, 2005; Kendler et al., 1995).

Defense

Male caring fishes, such as sticklebacks, damselfish, girabali, bluegill sunfish, plainfin midshipman, and the black-chinned tilapia, have high plasma androgen levels (testosterone and 11-ketotestosterone) during pre-spawning, when males compete for territories, construct nests, and court females. The androgen levels then gradually drop following spawning while males provide care. This was believed to indicate an androgen-mediated trade-off between aggression and parental care and a minimal role of androgens during parental care.

However, recent studies show that androgen levels often rise again to pre-spawning levels once eggs have hatched. Other studies even show that androgens remaining high in the early stages of care, correlated with the frequency of parental care/defense of young in the biparental cichlid N. pulcher. In addition, experimental elevation of androgen levels does not inhibit paternal behavior in a number of fish species, suggesting that an elevated androgen level is not necessarily incompatible with the expression of paternal behavior. This may be true of fish in general, but not other vertebrates. Male fish often continue to court and attract females even after they have begun to provide care, but in other taxa, the mating/courtship phase of reproduction is commonly temporally separated from the parental phase. Fish do not typically feed young, but instead defend young against predators (and sometimes fan the eggs); so, high androgen levels might, in fact, be beneficial and necessary for the aggression needed during parental care. More research is needed to clarify the role of androgens in mediating parental behavior with the importance of estrogen and mechanisms that modulate female care remaining particularly understudied.

Pitx1 enhancers in sticklebacks and the evolution of the pelvic region

The modification of the locus Pitx1 in sticklebacksGasterosteus aculeatus is also a case study of extreme morphological evolution. Three-spine sticklebacks have the ability to live both in freshwater and the ocean even if most of the population lives in the latter. Freshwater populations are supposed to have been isolated during the last glacial retreat in newly formed isolated lakes, thus becoming adapted to their novel environment (Bell and Foster, 1994). While ocean stickleback harbor full skeletal pelvic structures, some populations of freshwater stickleback exhibit a reduction or loss of skeletal armor (dorsal spine and pelvic girdle), a trait associated with reduced calcium and fewer large gap predators (Shapiro et al., 2004; Figure 5(a)). To decipher the changes at the origin of these variations, the authors crossed marine and freshwater populations and mapped the responsible genetic locus. The Pitx1 locus was shown to have the major effect of such variation. When looking at the sequence of Pitx1 in each population, no protein coding changes were found in the freshwater one compared to the marine one showing that the casual mutation must be regulatory. Moreover, mRNA localization experiments by in situ hybridization (Figure 5(b)) showed that freshwater populations lack Pitx1 expression in the pelvic region. Yet, the region was not fully characterized in this study, and the exact nature of the regulatory changes remained unclear. A few years later, another study carried on these experiments (Chan et al., 2010). First, to avoid changes linking to a different transcription factor context, the authors crossed two freshwater populations that partially or totally lack the pelvic part of the sticklebacks and confirmed that the loss of expression is due to a modification of a CRE at the Pitx1 region. Then, a high resolution mapping between marine and pelvic-reduced sticklebacks identified a 124 kb region of interest containing Pitx1 and another gene. The study of the correlation of microsatellites with the absence or presence of pelvic phenotypes in natural population reduced this interval to 23 kb in the intergenic region of Pitx1, a region which is conserved in teleosts and may thus contain ancestral enhancers. Finally, a smaller putative 2.5 kb enhancer, Pel (Figures 5(c) and 5(d)), cloned upstream Pitx1 driven by a drhsp70 promoter was used in transgenic experiments to successfully restore pelvic expression in reduced pelvic fishes. The sequencing and analysis of this region in various fishes identified a 1868 bp deletion in this region, as well as other independent deletions of this locus in various independent freshwater populations of sticklebacks.

As a conclusion, this example clearly shows how a deletion of a CRE can lead to morphological changes and can also be selected repeatedly. It also perfectly shows the challenges involved in fully characterizing a CRE and the need to perform multiple experiments to decipher such changes in a natural population. Finally, this case highlights how the fine-tuning of CRE can produce different levels of expression leading to a variation in phenotypes. Another interesting case of extreme morphological change was also observed in sticklebacks with the EDA pathway (see coding sequence, Colosimo et al., 2005) that also exhibit a reduction of armor plate in a specific environment. The mutation is not yet identified but is thought to be linked to cis-regulatory mutation.

b Learning in the Absence of a Predator

As reported above (Section III,A,1) sticklebacks develop their normal antipredator behavior in the absence of any predator only when raised by their father. Significantly, only young from an environment with the predator under study develop the response when thus raised. The mechanism by which this education is achieved is presently unknown, yet work on guppies hints at a possible reason: fish that had been chased by adult conspecifics when newborn escaped capture more successfully (Goodey and Liley, 1986; Tulley and Huntingford, 1987a).

Little attention has been devoted to the role of imprinting-like phenomena in the development of vital survival behavior in birds. Csermely et al. (1983–1984) claim that red-legged partridge chicks develop no fear of humans if they happen to see them within 48 hr after hatching (Table I). It is unclear to what extent maternal care contributes to the full level of fear. According to T. Silva (personal communication), young parrots of various species develop a fear of humans only when they are parent-raised, regardless of whether their parents are themselves shy of humans or tame. Whether this observation stands quantitative scrutiny remains to be seen. If true, an explanation based on habituation resulting from affiliation with the human caretaker, which would prevent an IRM from developing properly, might be most parsimonious at present.

There is a dearth of information on how the various risk-assessment mechanisms come to operate. An exception to this is a noteworthy observation by Trost (see Moholt, 1989) on how magpies develop recognition of the dead conspecific and, thus, risk assessment based on the “sign-of-predator.” A hand-raised magpie failed to mob a dead conspecific. By contrast, another magpie, raised later in the presence of that older individual, readily mobbed a dead one. It appears that experience with a live companion is necessary for the response to a dead one to develop. Interestingly, mobbing episodes by the second, socially less deprived bird in encounters with a deceased magpie increasingly induced the response in the deficient bird. These observations indicate that the possibility of social effects bringing about the typical corvid response to dead conspecifics through an IRM cannot be ruled out; the potential shortcomings of any experimental deprivation are always a problem (Section III,A,1).

4 Other Examples of Mate Preferences for Indicator Traits that may be Derived from Foraging Responses

Three-spined stickleback (Gasterosteus aculeatus) and nine-spined stickleback (Pungitius pungitius) of both sexes are attracted to red objects (plastic strips) in a foraging context but only three-spined stickleback display red (carotenoid-based) breeding coloration (Smith et al., 2004). In the clade that includes both of these species, red breeding coloration is a derived condition unique to three-spined stickleback. Nine-spined stickleback and some populations of three-spined stickleback have black nuptial coloration, which is ancestral for the clade. In most populations of three-spined stickleback, females prefer males with red coloration, and this is a condition-dependent trait that reflects variation among males in parental care and parasite resistance (reviewed in Smith et al., 2004). There is no direct evidence, as yet, that the mate preference for red males and the foraging preference for red prey are linked, but the strength of the mate preference has been correlated with variation among populations in sensitivity to red light (Boughman, 2001). The sensitivity of female three-spined stickleback to red light varies seasonally and peaks during the spawning season when males express maximal red coloration (Cronly-Dillon and Sharma, 1968). Male three-spined stickleback are also sensitive to red light and preferentially attracted to red food items, but males do not show seasonal variation in sensitivity to red. This sex difference may be a product of the indicator process, although whether females benefit in other ways from exhibiting seasonal variation in sensitivity to red light has not been investigated. Another piece of evidence that the mate preference for red coloration in three-spined stickleback has been shaped by the indicator process is that the strength of this preference correlates positively with the degree of condition-dependence of red coloration among populations of three-spined stickleback in the lakes of British Columbia (Boughman, 2007). The direction of causality is unclear, however, because directional sexual selection may cause secondary sexual characters to become condition-dependent (Iwasa and Pomiankowski, 1999; Kodric-Brown and Brown, 1984). If it could be shown that the variation in condition-dependence was caused by an environmental factor (e.g., carotenoid availability), then the correlation between condition-dependence and the mate preference for red coloration would constitute strong support for the indicator hypothesis.

The mate preference for another carotenoid-based trait, the terminal yellow band (TYB) of some Goodeinae fishes, also appears to be derived from a foraging adaptation, but has become partially decoupled from foraging behavior in species that possess a TYB (Garcia and Ramirez, 2005). Species without TYBs, which may represent the ancestral condition, respond to TYBs as though they were food items, while such responses are reduced in species in which males have evolved TYBs (Garcia and Ramirez, 2005).

The pheromones that male lizards use to attract females contain chemical compounds that are also found in prey and evoke feeding responses in the lizards (reviewed in Martin and Lopez, 2008). In Iberian rock lizards (Iberolacerta cyreni), the provitamin D content of a male's pheromones reflects the quality of his diet, and females show stronger sexual responses to the secretions of males with higher provitamin D content. Experimentally, increasing female hunger levels through food deprivation increases the rate at which females tongue-flick (a feeding response) to cotton swabs scented with mealworm prey, male pheromones, and provitamin A, but not to unscented control swabs (Martin and Lopez, 2008). Provitamin D has health benefits in lizards, and thus the provitamin D content of a male's pheromones may be an indicator of his health and quality as a mate. If so, this may be another example of a indicator trait that evolved to exploit a sensory bias that originally evolved as a foraging adaptation and which now serves the dual function of enabling females to find high-quality food and high-quality mates.

Sudden Evolution of Morphology in the Threespine Stickleback, Gasterosteus aculeatus

When female (with riped ovaries) and male (with breeding colorations) threespine stickleback fish of the species Gasterosteus aculeatus, at the beginning of their breeding season, were transferred from marine tide pools to freshwater ponds, they produced offspring of different shape and with less armor plates than their parents (Kristjansson, 2005).

In 1987, the Hraunsfjordur, a fjord in north-west Iceland, was dammed to form a freshwater lagoon for cultivating salmon (Salmo salar). Ever since, within 12 years, the stickleback (G. aculeatus) population of the freshwater lagoon experienced a rapid morphological evolution that led to a remarkable divergence in morphology from its marine ancestral population. The rates of evolution for dorsal spines and for keeled armor plates were comparable only to the exceptionally fast rates of evolution of coloration and decoration of Trinidad guppies, Poecilia reticulata, and of the beak in Darwin's finches (Kristjansson, 2005).

In the Queen Charlotte Islands (British Columbia, Canada) as well, populations of the stickleback G. aculeatus, show remarkable differences in morphology. These differences seem to have arisen as adaptive responses to the local habitat and fish predators (Moodie and Reimchen, 1976; Fig. 10.25).

Neither changes in gene functions nor selection on the existing genetic variability have been proposed for explaining the exceptionally rapid morphological evolution of this fish species. Hence, epigenetic changes in the regulation of expression of genes are the only alternative explanation of the phenomenon.

Publisher Summary

This chapter describes the role of the intestine in maintenance of hydromineral balance of sexually mature male sticklebacks in fresh water. The kidneys are the main pathway by which osmotically accumulated water is eliminated in freshwater fish. As their kidneys are characterized by a high glomerular filtration rate and the epithelial cells of their nephronic tubules have a high ion-reabsorbtive capacity, fish are able to excrete large volumes of dilute urine. During the reproductive period the second proximal and collecting tubules of the kidneys of male three-spined sticklebacks, Gasterosteus aculeatus, are transformed into mucus secreting cells indicating a considerable loss of ion reabsorbtive capacity. This glandular transformation is induced by testosterone. Testosterone dependent structural changes also take place in the glomeruli of mature males. Physiological experiments showed that the intestine of the three-spined stickleback may play an important role in hydromineral regulation because sexually mature males excrete a slightly hypotonic fluid in amounts that are about 3 to 4 times higher than in immature males, which produce small amounts of isotonic intestinal fluid.