Live (biocoenoses) and dead (taphocoenoses) populations of regular and irregular echinoids inhabiting shallow water environments of San Salvador and Leestocking Island, Bahamas (and elsewhere in the Caribbean) have been censused over the last five years to test Kier's (1977) hypothesis that the relatively poor fossil record of the regular echinoid is the result of taphonomic bias. In general, results reveal that distributions of living regular echinoids are not reflected by accumulations of their carcasses, while the reverse is true for irregular taxa: subfossil material is more often associated with living populations suggesting that irregulars may have relatively greater likelihood of preservation. Population distributions of four regular echinoid taxa (Diadema antillarum, Eucidaris tribuloides, Echinometra lucunter and Tripneustes ventricosus) were systematically studied in reef and near-reef environments. Although these taxa are common components of the benthic fauna, the distributions of the biocoenoses were patchy. High abundance of living individuals (densities approaching 60/m2) of Echinometra was not correlated with an increase in subfossil material even though previous work has indicated the remarkable durability of the corona relative to other regular echinoid taxa. Tripneustes, Diadema and Eucidaris occurred in lower abundances than Echinometra and were associated with limited amounts of skeletal remains >2 mm in size. Shallow-burrowing irregular echinoids selected for study included two mellitid sand dollars (Mellita quinquiesperforata, Leodia sexiesperforata), and the spatangoid Meoma ventricosa. In contrast to the results obtained for regular echinoids, more subfossil material comprised the taphocoenoses, the majority of it occurring on the sediment surface: the latter observation suggests that the shallow infaunal life mode does not guarantee that biostratinomic processes will not affect skeletal material. Multivariate analyses of census-data reveal that live echinoid distribution is likely related to substrate: the regular echinoids under study occurred on hard or rubbly substrates that inhibit burrowing, while the irregular echinoids occurred in soft substrates of clean carbonate sand. In Fernandez Bay, Echinometra and Tripneustes were most common closest to shore (between 0 m and 18 m from the shoreline), while Diadema and Eucidaris occurred farther out (between 28 m and 138 m). Mellita and Leodia occurred farthest from shore (between 46 m and 156 m). However, echinoid carcasses do not reflect the same pattern: regular taxa, though rare, are restricted to the same areas in which living counterparts occur. Conversely, the irregular echinoid taphocoenosis has undergone some transport: carcasses possess a wider distribution than their living counterparts. Results of this study suggest that Kier's (1977) hypothesis, while essentially correct, requires modification: 1) Skeletal durability may be more important than life habit in determining preservation potential; 2) Regular echinoid fragments, although rare in the fossil record, are useful tools for paleoenvironmental reconstruction based on the fidelity with which they reflect the distribution of a once-living fauna; 3) The poor fossil record of the regular echinoid may also be an artifact of taxonomic bias; finally, 4) As is the case with regular echinoids, exceptionally well-preserved assemblages of abundant mellitid and shallow-burrowing spatangoids must be viewed as the result of extrinsic processes rather than intrinsic traits. Analysis of Pleistocene analogues to the Recent environments studied generally corroborates these conclusions.
