DEAD OR ALIVE: USING SCLEROBIONT DENSITY DISTRIBUTION AND HOST GROWTH MODELS TO HELP DISTINGUISH BETWEEN LIVE-HOST AND POST-MORTEM COLONIZATION

Sclerobiont fossil assemblages potentially record ecological interactions between host taxa and their hard-substrate colonists with excellent spatial and temporal resolution, providing a wealth of paleoecological data. Many of these interactions require the demonstration of live-host colonization, limited to exposed hard surfaces of the host, such as the external shell surfaces of epifaunal bivalves or brachiopods. Site-selective behaviors and other evidence of biotic interaction between host and colonist, such as parallelism in growth to maintain feeding position at the mantle edge, are commonly used to demonstrate live-host relationships, but in the absence of these interactions it may be difficult to distinguish between live-host and post-mortem colonization. Host growth models may provide evidence to discriminate between post-mortem colonization and the gradual accumulation of encrusting and bioeroding organisms during the life of their host in the absence of site selective behavior and growth. For live epifaunal host taxa, exposed surface area increases as a function of age. It is possible to model the relationship between surface area and age using sclerochronologic evidence (such as age/length curves or von Bertalanffy growth models) if the relationship between shell length and area is determined. A continuous live-host accumulation model (CLAM) accounting for the increase in surface area through the life of the host results in greater abundance of colonists towards older parts of the shell than a null model based on delayed encrustation after death (DEAD), in which colonization is proportional to shell area. This trend is most apparent when sclerobiont density is determined as a function of position relative to the growth axis of the shell, in which live-host colonization, with or without subsequent post-mortem colonization, can be identified by increasing sclerobiont density towards the umbo. These models can be used in conjunction with digital mapping and image-stacking techniques to pool evidence from multiple host specimens in order to better elucidate the presence or absence of live-host colonization in a host population in the absence of site-selective behavior.