The Predictable Complexity of Evolutionary Allometry

Allometry has been a paradigm of constraints, including intrinsic constraints on the evolvability of allometry, as a source of developmental and genetic constraints on the evolution of form, and of functional constraints, maintaining functional equivalence as body size evolves. Yet, allometry may be the simplest case of varied constraints, and of morphological integration, even though allometry itself is not simple. Evolutionary allometry may be especially complex because it depends not only on developmental origins of allometry and determinants of allometric variation but also on the evolutionary dynamics of size and shape. It should also depend on the ecological opportunity for size-dependent ecomorphological specialization. We predict that lineages that converge in those would exhibit similar evolutionary allometries but otherwise, evolutionary allometries would be heterogeneous. Countering this expectation are familiar craniofacial evolutionary allometries, often ascribed to developmental bias. To test both those hypotheses, we compare evolutionary allometries of mandibles across lineages of squirrels and evolutionary to growth allometries. As expected, lineages that converge on size-dependent specializations exhibit similar evolutionary allometries, but otherwise their allometries are no more similar than expected by chance. Growth allometries of squirrels (and a cricetid rodent) slightly resemble the evolutionary allometry of one lineage, but growth allometries of species from other lineages are orthogonal to their own lineages' evolutionary allometry. We would expect that craniofacial allometries that are not brain-driven would, like mandibular evolutionary allometries, be predictable only from size-dependent ecological specializations.