WHATS GROWTH GOT TO DO WITH IT - PROCESS AND PRODUCT IN THE EVOLUTIONOF ONTOGENY

Growth allometries are regularly used to diagnose evolutionary perturb ations in the rate and timing of development, and to study their effec ts on morphological change. Metric variables are measured for ancestor and descendant, plotted on logarithmic scales, and the two estimated linear relationships are compared. The relationship between these two lines has been a tool of choice for inferring heterochronic history, p articularly when age data are not available. Size, plotted on the absc issa, is often used as a proxy for age; the variable plotted along the ordinate typically measures some aspect of ontogenetic change. Under the rubric of ''allometric heterochrony'', changes in the slopes and i ntercepts of ancestral and descendant growth allometries are taken to reveal heterochronic processes (changes in the rates of growth, offset timing, and onset timing of shape relative to size) and their two pro ducts-paedomorphosis (the retention in mature descendants of ancestral juvenile characteristics) and peramorphosis (the acquisition in matur e descendants of features that transcend ancestral adult characteristi cs). This framework is said to be grounded in the pioneering work of G ould (1977) and Alberch et al. (1979). In this paper, we examine the m eanings of ''shape'' in heterochronic analysis. We argue that the curr ent diagnostic toolkit departs significantly from those constructed by Gould (1977) and Alberch et al. (1979), and bears little relevance to the problems they addressed. We show why growth allometries (as typic ally read) do not capture evolutionary relationships between growth (G ould's ''size'') and development (Gould's ''shape''). (C) 1995 Academi c Press Limited