Live, dead, and fossil mollusks in Florida freshwater springs and spring-fed rivers; taphonomic pathways and the formation of multisourced, time-averaged death assemblages

Taphonomic processes are informative about the magnitude and timing of paleoecological changes but remain poorly understood with respect to freshwater invertebrates in spring-fed rivers and streams. We compared taphonomic alteration among freshwater gastropods in live, dead (surficial shell accumulations), and fossil (late Pleistocene-early Holocene in situ sediments) assemblages from two Florida spring-fed systems, the Wakulla and Silver/Ocklawaha Rivers. We assessed taphonomy of two gastropod species: the native Elimia floridensis (n=2504) and introduced Melanoides tuberculata (n=168). We quantified seven taphonomic attributes (aperture condition, color, fragmentation, abrasion, juvenile spire condition, dissolution, and exterior luster) and combined those attributes into a total taphonomic score (TT). Fossil E. floridensis specimens exhibited the greatest degradation (highest TT scores), whereas live specimens of both species were least degraded. Specimens of E. floridensis from death assemblages were less altered than fossil specimens of the same species. Within death assemblages, specimens of M. tuberculata were significantly less altered than specimens of E. floridensis, but highly degraded specimens dominated in both species. Radiocarbon dates on fossils clustered between 9792 and 7087 cal BP, whereas death assemblage ages ranged from 10,692 to 1173 cal BP. Possible explanations for the observed taphonomic patterns include: (1) rapid taphonomic shell alteration, (2) prolonged near-surface exposure to moderate alteration rates, and/or (3) introduction of reworked fossil shells into surficial assemblages. Combined radiocarbon dates and taphonomic analyses suggest that all these processes may have played a role in death assemblage formation. In these fluvial settings, shell accumulations develop as a complex mixture of specimens derived from multiple sources and characterized by multimillennial time-averaging. These findings suggest that, when available, fossil assemblages may be more appropriate than death assemblages for assessing preindustrial faunal associations and recent anthropogenic changes in freshwater ecosystems.