Population-level variation in senescence suggests an important role for temperature in an endangered mollusc

Age-related declines in survival and function (senescence) were thought not to exist in wild populations as organisms, and particularly in invertebrates, do not live long enough. While, recent evidence has demonstrated that senescence is both common and measurable even in wild populations under field conditions, there are still organisms that are thought to exhibit “negligible senescence”. We explore variation in rates and patterns of senescence in the biogerontological model organism Margaritifera margaritifera across five populations, which differ in their age profile. In particular, we tested the theory of negligible senescence using time-at-death records for 1091 specimens of M. margaritifera. There is clear evidence of senescence in all populations, as indicated by an increase in mortality with age, but the nature of the relationship varies subtly between populations. We find strong evidence of a mortality plateau at later ages in some populations but this is unequivocally absent from others. We then demonstrate that the temporal scaling of the rates of senescence between five populations of M. margaritifera can be explained by the variation in the thermal environment of the population. Hence climate change may pose a threat to the demography of this long-lived, endangered species, and a greater understanding of the relationship between river temperature and population structure will be essential to secure the species against global temperature increases. Our findings demonstrate that useful insights can be drawn from a non-invasive monitoring method to derive demographic data, and we suggest a wide-scale application of this method to monitor populations across the whole latitudinal (and, hence, thermal) range of the species.