The presumed geometry of clam and brachiopod clades (brachiopod declines matched closely by clam increases) has long served as primary data for the classic case of gradual replacement by competition in geological time. Agassiz invoked the geometric argument to assert the general superiority of clams, and it remains the standard textbook illustration today. Yet, like so many classic stories, it is not true. The supposed replacement of brachiopods by clams is not gradual and sequential. It is a product of one event: the Permian extinction (which affected brachiopods profoundly and clams relatively little). When Paleozoic and post-Paleozoic times are plotted separately, numbers of clam and brachiopod genera are positively correlated in each phase. Each group pursues its characteristic and different history in each phase—clams increasing, brachiopods holding their own. The Permian extinction simply reset the initial diversities. The two groups seem to track each other in each phase and a plot of brachiopod vs. clam residuals (each from their own within-phase regressions against time) yields significantly positive association. Some of this tracking may be an artifact of available rock volumes; we could, however, detect no effect of stage lengths. Passive extrapolation of microevolutionary theory into the vastness of geological time has often led paleontologists astray. Competitive interaction may rule in local populations, but differential response to mass extinctions (surely not a matter of conventional competition) may set the relative histories of large groups through geological time. Similarly, adaptive superiority in design cannot, in the usual sense of optimal engineering, have much to do with the macroevolutionary success of clams. The interesting question lies one step further back: what in the inherited Bauplan of a clam permits flexibility in design and why are other groups, however successful in their own domain, unable to alter their basic design.
Constrained inefficiency in GEI: A geometric argument
