The proposal (Balon 1975a) to define the ontogeny of fishes by uniform,
objectively-recognized intervals in an hierarchical system, has lead to increas-
ingly widespread reassessments (Balon 1978, 1980; Bond 1979; Noakes
1981), although the proposal itself has been considered by some as little
more than a semantic exercise (Balon 1976; Richards 1976).
The life history (ontogeny) of any fish is composed of a series of intervals.
By now most would accept this concept, but disagreements continue as to
the definition and nomenclature for these intervals. The proposal of a
uniform terminology for these intervals, based on a hierarchical system from
the smallest recognizable stages in development to the broadest periods in
the life history has significant and far-reaching implications (the fundamental
implications of hierarchical systems in biology are well-established, Koestler
1967). These intervals are arranged, in sequence from broadest to smallest, as
periods, phases, steps and stages. Each interval is composed of one or more
at the next lower level in the hierarchy.
These terms overlap somewhat with previous usage, especially as they have
been applied by some developmental biologists and ecologists, but they
impose an order on otherwise ambiguous usage (e.g. Keenleyside 1979) and
align with current theories of ontogeny.
The life history of a fish can consist of a maximum of five periods; em-
bryonic, larval, juvenile, adult and senescent. These encompass the entire life
of an individual, from beginning to the end of its life. The embryonic period
begins with activation (insemination) of the egg, and terminates with the
transition to exogenous feeding. The larval period begins with the transition
to exogenous feeding and terminates with metamorphosis, when most of the
temporary embryonic structures are replaced by permanent "adult" organs
(typically including differentiation of the median fin fold, and ossification of
vertebral centra). The juvenile period extends from this point until the first
maturation of the gonads. The first gonadal maturation marks the beginning
of the adult period, which continues until the last reproduction. The
senescent period includes that time after the fish has ceased reproduction (or
gametes are very few and of inferior quality), and ends with death.
At least some of these intervals may be greatly extended (e.g., larval period
of several years in anguillid eels), or be entirely eliminated (as is the larval
period in mouthbrooding cichlids (Fig. I), the coelacanth, Latimeria cha-
lumnae, and many cyprinodontids). Recognition of the alterations of life
histories by the relative timing and duration of these intervals (usually
referred to as heterochrony; Jones 1972; Gould 1979) leads to productive
insights (Balon 1980). We must again emphasize that hatching (or parturition)
plays no role in defining these intervals. This is perhaps the most common
and persistent misunderstanding in considerations of ontogeny and life
history features (Balon 1980, 1981a; Noakes 1981)., We tend to date most
things from birth (parturition) or hatching, and assume that the event has
general significance in ontogeny, likely as a result of peculiarities of our own
development (Noakes 1981). It does not.
Our terminology provides a uniform scheme to apply to any organism for
the sake of comparisons. Typically, authors attempting such comparisons
have assumed (usually implicitly) that hatching (or birth) should be the
"zero point" at which to adjust the scales of measurement for the species in