Bivalve death assemblages from subtidal environments within the tropical Bocas del Toro embayment of Caribbean Panama permit a test of the extent to which levels of damage are determined by the intrinsic nature of shell supply (proportion of epifaunal species, thick shells, calcitic shells, low-organic microstructures), as opposed to the extrinsic postmortem environment that shells experience. Only damage to interior surfaces of shells was used, to ensure that damage was unambiguously postmortem in origin. We find that facies-level differences in patterns of damage (the rank order importance of postmortem encrustation, boring, edge-rounding, fine-scale surface degradation) are overwhelmingly controlled by environmental conditions: in each environment, all subsets of the death assemblage present the same damage profile. The composition of shell supply affects only the intensity of the taphonomic signature (i.e., percentage of shells affected) only in environments containing hard substrata (patch reefs, Halimeda gravelly sand, mud among patch reefs). In these environments, epifauna, whether aragonitic or calcitic and whether thin or thick, exhibit significantly higher damage than co-occurring infauna, probably due to the initial period of seafloor exposure they typically experience after death. Thick shells (>0.5 mm), regardless of life habit or mineralogy, are damaged more frequently than thin shells, probably because of selective colonization by fouling organisms. Calcitic shells show no consistently greater frequency of damage than aragonitic shells high-organic microstructures yield mixed patterns. Taphofacies surveys in such depositional systems could thus be confidently based on any subset of the fauna, including diagenetically residual assemblages of calcitic shells and thick-shelled molds. Further tests are needed to determine whether the higher levels of damage observed on some subsets of shells are a consequence of greater time-averaging (thus lower temporal resolution), greater exposure time, preferential attack (potential bias in relative abundance), or some combination of these. Paleobiologically, however, the implication is that ecological subsets of bivalve assemblages are not isotaphonomic, either in tangible damage or in probable bias, within hard-substrate environments, although they may be within soft-sediment environments. In actualistic studies, targeting broad classes of taxa for comparison across environments maximizes our ability to extrapolate taphonomic guidelines into the fossil record, where life habits, skeletal types shallow subtidal habitats have dramatically different patterns of abundance and deployment.