Biology Reference
In-Depth Information
(see
Kimmel et al., 1995
). Whether it actually is so conservative is controversial; several
studies have challenged the idea that embryos initially diverge and then converge, an
idea that has been termed the “hour-glass” model (e.g.
Bininda-Emonds et al., 2003
)on
the grounds that the phylotypic period is not as conservative as generally thought. But
even if distantly related species do resemble each other through somitogenesis, it
would not be surprising to find that larvae or neonates differ substantially in shape.
Even fairly modest differences in development are likely to have large impacts on mor-
phology when rates of development are high. What is surprising is not that divergence
occurs during larval or fetal development but rather than divergence occurs only
then
overtheentireobservedphaseofdevelopment, the ontogenetic trajectories are
parallel.
One hypothesis that could explain parallel trajectories is that later stages of
ontogeny are subject to more severe functional constraints. This has been postulated for
thecaseofsigmodontinerodents;thereason why post-weaning growth allometries
might be conservative is the biomechanical constraints of masticatory function; that
earlier developmental stages are less conserved could result from the absence of
such constraints in the fetus or suckling pup (
Voss and Marcus, 1992
). Another
explanation is serial correlations between developmental stages; just as modifications
early in development will have cascading effects on later stages, selection on adult
morphology will lead to divergence at earlier stages. This hypothesis presupposes
that individuals who deviate from the mean as adults also do so (and in the same
direction) much earlier in ontogeny. Although there is some evidence for that correla-
tion between developmental stages, one recent study of human craniofacial shape
found a weak correlation between newborns and adults, although correlations are high
between three year olds and adults (
Bulygina et al., 2006
). An alternative hypothesis,
which also assumes a high correlation between early and late developmental stages, is
that selection for changes of early ontogeny leads to modifications of adult mor-
phology; the shape of the adult may even beconstrainedbytheshapeoftheinfant
(
Young, 2008
).
Divergent Ontogenies of Shape
The remaining hypotheses predict that ontogenies diverge in ontogenetic trajectories for
shape. What differentiates the hypotheses is what else is expected to differ and whether
morphologies are expected to diverge or converge over the course of development. The
reason for expecting one of these patterns, as opposed to heterochrony or parallel trajecto-
ries, is that there is no good a priori reason to expect that the spatiotemporal patterning of
morphogenesis, or any other aspect of morphogenesis, growth or maturation is conserved.
Rather, ontogenies can evolve by modifications of allometry, by accelerating or retarding
growth, or by increasing or decreasing the durations of growth or by increasing or
decreasing either the rates or durations of morphogenesis. Thus, in the absence of any rea-
son for expecting that ontogenies are constrained in their evolutionary possibilities, to
determine which modifications occur we need to consider all the possibilities shown in
Figure 11.10
as well as all variants upon them.
