| The most current thought on the ecological causes of heterochrony postulates that neoteny (retardation of the growth rate) and hypermorphosis (delay in the time of maturation) will be associated with more stable (K-selecting) environments while acceleration (increase of the growth rate) and progenesis (earlier maturation) will be associated with less stable (r-selecting) environments. Most of the evidence so far for this 'Gould hypothesis' is anecdotal. Here I undertake a systematic, quantitative study with a very large data base to test this hypothesis. I chose to work with Cenozoic irregular echinoids from the Gulf Coastal Plain because they are abundant, well-studied, and well-preserved as fossils. In addition, the strata where they occur are also well-studied and paleoenvironmental information is available. The data base consists of biometric measurements of 11 traits on 2,333 individuals. These belong to nine genera including 32 species from a variety of paleoenvironments.; Computer-assisted bivariate analyses of the ontogenetic changes in the traits yielded many cases of heterochronic change. Two types of patterns were observed. One was that while different traits often show different types of heterochrony (only 2 of 14 cases show the same type throughout), there are some clear associations among the types: neoteny and hypermorphosis often occur together and almost always lead to larger sizes. Changes which are not global (the same throughout) or of this neotenic/hypermorphic association seem to be due to (1) heterochronic 'drift', where selectively neutral traits are affected, and (2) 'fine-tuning' such as changes in feeding ability to permit larger sizes.; The second pattern is that between the environment and the types of heterochrony occurring there. In 13 of 15 cases, the hypermorphic/neotenic member of a pair of related species occurs in the more stable environment. A smaller sample with more reliable data shows 7 of 8 cases where the hypermorphic/neotenic member lived in the more stable environment. Statistically, neither of these results is easily attributable to 'chance'. Along with other, less clear-cut, data showing the same general pattern, these results provide empirical substantiation in favor of the Gould hypothesis. |
