Processes of hardpart breakdown and models of stratigraphic disorder in shallow marine environments.

Taphonomy is the study of the fate of information in the fossil record. Information can be lost through the partial or complete destruction of fossils, or through the disruption of their original spatial relationships. Information can be "gained" if the alteration of fossils allows environmental information to be retrieved. In Bahia la Choya, northern Gulf of California, bioerosion, dissolution/maceration, and abrasion produce distinctive textures on the surfaces of shells in intertidal and shallow subtidal environments. Shells from different environments possess different surface textures, suggesting that textures on fossil shells could serve as paleoenvironmental indicators. Algal bioerosion is the chief mode of shell alteration and destruction in Bahia la Choya, though dissolution/maceration and abrasion are locally important. Algal bioerosion of shell surfaces is accelerated by the grazing activity of snails, and is most intense where snails are abundant. Microstratigraphic resolution is limited by vertical mixing of fossils and by the reworking of older fossils into younger deposits. Stratigraphic disorder is the departure from perfect chronological order of fossils in a stratigraphic sequence. I simulated mixing and reworking of fossils by simple computer models, and measured stratigraphic disorder using rank correlation statistics. As modeled, mixing produces disorder slowly, and its effects can be minimized by increasing sample size at each horizon and by increasing the vertical spacing between sampled horizons (though this reduces vertical resolution). Reworking generates disorder more efficiently, and its effects are not reduced by increasing sample size or spacing. The generation of stratigraphic disorder in fossiliferous sediments can also be modeled using M (depth of vertical mixing), I (thickness of sedimentary increments), and L (taphonomic loss rate) as parameters. Increasing M increases the disorder generated, and increasing I and L decreases disorder. For a worst case--high M and low I and L--the vertical spacing between samples must at least 3 times M to ensure a 5% temporal overlap between adjacent samples. A 1% temporal overlap requires a vertical spacing of 4.6 times M