Larval feeding and survival
A defining event in a larva’s life is when yolk supplies are exhausted and the fish becomes dependent on exogenous food sources, usually in the form of small planktonic organisms such as diatoms, larval copepods and mollusks, and
Even fishes that are herbivorous as juveniles and adults are usually carnivorous as larvae, as in rabbitfishes (Bryan & Madraisau 1977). This generalization may reflect the difficulty with which energy and nutrients are extracted from plants.
The potential importance of food availability at the onset of exogenous feeding has greatly influenced our thinking about sources of larval mortality and the subsequent strength of year classes of fishes. Several infl uential hypotheses address the relationship between early larval biology, food availability, and adult population size in marine fishes. Hjort (1914) proposed the Critical Period Hypothesis, which stated that starvation at a critical period, perhaps the onset of exogenous feeding, was a strong determinant of later year class strength. Blaxter and Hempel (1963) coined the phrase point-of-no-return to describe when larvae, as a result of starvation, are too weakened to take advantage of food even if it were available. Such irreversible starvation depends on larval condition and age: well-fed, young anchovy may last only 1–2 days before irreversible starvation sets in, whereas healthy, older flatfish larvae may be able to go 2–3 weeks without food. Cushing (1975) proposed the Match– Mismatch Hypothesis, which suggests that the timing of reproduction in many marine fishes has evolved to place larvae in locales where food will be available, i.e., that fish reproduction and oceanic production are synchronized. Since the cues of photoperiod and temperature that fish use to initiate spawning (see Reproductive seasonality) are not necessarily the same ones that determine plankton production, mismatches can occur and result inhigh larval mortality (May 1974; Russell 1976; Hunter 1981; Blaxter 1984; Houde 1987).
Because of the relationship between larval survival and later population size, the actual causes and patterns of larval mortality are of considerable theoretical and obvious practical interest. Literally billions of larvae are produced by most populations of marine fishes annually. In most species, >99% of these larvae die in their first year from the combined effects of starvation and predation; the average fish probably dies in less than a week (Miller 1988). Hence very minor shifts in mortality rates can have major implications for later year class strength and for recruitment into older, catchable size classes (Hobson et al. 2001).
The temperatures at which larvae develop affect individual growth and development rates, metabolic rates, and energy requirements, all of which can influence mortality (Houde 1989). Across a 25°C temperature range characteristic of the difference between tropical and temperate conditions (5–30°C), mortality rates of marine larvae can vary four-fold, being highest at the higher temperatures. Growth rates at these higher temperatures at any given size are six times faster. Larval duration at the lower end of the temperature range typically exceeds 100 days, whereas at 25– 30°C, metamorphosis occurs in 1 month or less.
One might expect that spending the least amount of time as a small, vulnerable larva would lessen the chances of both predation and mortality. However, the metabolic requirements of small ectothermic animals such as fish larvae increase in direct relationship to temperature. Gross growth effi ciency (weight increase/weight of food consumed) is constant despite temperature. A larva, because of its higher metabolic rate at higher temperatures, must consume more food to achieve the growth rate of a larva at lower temperatures. This is additionally compromised because gross growth efficieny declines with increased ingestion, and assimilation effi ciency, which is how much of the food is actually useful to the larva, declines with increased temperature. To maintain the same average growth rate, a tropical larva has to eat three times more food than a temperate larva of the same species. Mismatching larvae with food availability becomes more critical at higher temperatures.
Spawning patterns among species appear to represent adaptations to these temperature relationships. Tropical fishes typically spawn over an extended period, producing multiple batches of young, rather than releasing all their eggs in one large spawning session. This kind of “bet hedging” strategy increases the probability that some larvae will encounter the kind of conditions necessary for successful growth, whereas a single spawning might lead to complete reproductive failure if food abundance were low, as it usually is in tropical pelagic areas. Temperate marine fishes that spawn in the summer, such as Atlantic mackerel and white hake, tend to spread their reproduction out over a longer period than do winter spawners such as herring and capelin at the same latitude (Houde 1989).
Pelagic larvae are particularly common among coral reef species. Whereas many nearshore temperate species have short larval periods or retain their larvae near the adult habitat, dispersal via a pelagic stage is almost universal among coral reef fishes. Of the 100 or so families that commonly inhabit coral reefs, 97 have pelagic larvae. The exceptions are instructive in that it is easy to postulate historical constraints or adaptive disadvantages to dispersal. Marine plotosid catfishes are a freshwater-derivative family with highly venomous spines. The brightly colored young form dense, ball-shaped shoals and probably gain predator protection from this behavior, a lack of dispersal helping keep siblings together and facilitating the formation of monospecific shoals. Many of the viviparous brotulas live in fresh or brackish water caves near coral reefs, a habitat that could be difficult to relocate by a settling larva. One species of damselfish, Acanthochromis polyacanthus, lacks a dispersed larval stage. It is also the only damselfish worldwide that continues to care for its young after they hatch; interestingly, Acanthochromis larvae develop more slowly than most other damselfishes (Kavanagh & Alford 2003). Other non-dispersers include batrachoidid toadfishes, the monotypic convict blenny (Pholidichthyidae), and apparently reef species of the croaker family.
The adaptiveness of a floating larva for the other 97 families probably results in part from selection for avoidance of abundant predatory fishes and invertebrates in benthic habitats, larvae generally not settling until they have developed avoidance capabilities. Dispersal may also reflect: (i) the possibility that successfully reproducing adults live in saturated habitats that offer few opportunities for settlement for their young; and (ii) the widespread and spotty distribution of coral reefs relative to immense oceanic expanses, necessitating the dispersal of offspring over as wide an area as possible. However, tagging, genetic, behavioral, and otolith chemistry methodologies are increasingly indicating that larvae are retained in nearshore gyres and currents and may actively return to parental regions. For example, marking and DNA genotyping studies of anemonefish on the Great Barrier Reef showed 15–60% of juveniles recruited back to their natal population, many settling <100 m from the anemone where they hatched (Jones et al. 1999b, 2005; see also below, Getting from here to there: larval transport mechanisms).
Pelagic existence is about the only thing these larvae have in common. Reproductive strategies include viviparity and oviparity, mouth-brooded eggs, nest builders, and demersal and floating eggs, some of the latter attached to seaweed. Larval periods range from 9 to >100 days, sizes at settlement from 8 to 200 mm. Some settle prior to metamorphosis to a juvenile stage, some after, and one family (Schindleriidae) is even mature at the time of settlement (Leis 1991).