mammalian rate), but the sample size (12 genera) is, again, much too small for statistical significance.
Selectivity for Specific Traits. Somewhat more success has been achieved by focusing directly on aspects of physiology, behavior, habitat, and biogeography. Jablonski (1986a) showed, for example, that marine mollusks with planktonic larvae survive longer during "background" times than those which develop directly from the egg. This result is reasonable because species with a planktonic stage have greater dispersal capabilities and can attain wider geographic distributions, and thus are more likely to survive stresses that eliminate species in small areas. Whereas Jablonski has confirmed that widespread species are significantly less likely to go extinct during most geologic intervals, he has also shown that this protection breaks down at times of severe mass extinction, when broad geographic range at the genus level becomes important (Jablonski, 1986b).
It is commonly thought that tropical organisms suffer more extinction than those in higher latitudes. Although this is supported by anecdotal data for several extinction events, a recent study based on a global data base (3514 occurrences of 340 genera) found no recognizable geographic pattern in extinction of bivalve mollusks at the end of the Cretaceous, once reef-dwelling rudists were omitted (Raup and Jablonski, 1993). Approximately 50% of all genera died out, regardless of geographic position. Unpublished follow-up studies show a lack of habitat selectivity for bivalves and gastropods of the Gulf Coast during the same interval (D. Jablonski, University of Chicago, pers. comm., 1993). It may be that extinction is selective when overall extinction rates are low (so-called background extinction), but not when rates are high.
Large body size is thought to increase the risk of extinction. Indeed, many apparently good examples exist, including the now-extinct dinosaurs, ammonites, eurypterids, mammoths and mastodons, and rudist clams (LaBarbera, 1986). In the terrestrial realm, at least, decreased survival can be related easily to body size through demographic considerations (small populations, large home range, low birth rates, etc.). But the issue is clouded by the lack of rigorously controlled statistical analysis and by the fact that in several cases (e.g., eurypterids and ammonites), the largest species did not exist late in the group's range.
Species Selection. A special case of selective extinction involves differential origination and extinction of species in an evolving clade. In theory, species carrying a favored trait should survive longer and thus have greater opportunity for speciation than less well-adapted species.