SOME EMERGING PRINCIPLES OF CARBONATE-SHELL ACCUMULATION RELEVANT TO CONSERVATION PALEOBIOLOGY – WHAT CLAMS SAY

Molluscan death assemblages (DAs), sieved from the uppermost 5-20 cm of the seabed, have received much taphonomic study focusing on ‘direct-dating’ to establish scales of time-averaging and ‘live-dead comparison’ to evaluate bias in species preservation. Meta-analysis and dynamic modeling of these data are providing new insights into DAs as sources of historical information on benthic communities, and many of these will have parallels for the microfossil record. For example, 14C-calibrated amino-acid racemization reveals that, although DAs can contain shells that are several ka in inner-shelf and ~20 ka in outer-shelf settings, the frequency-distribution is very hollow to 'L-shaped', with most specimens dating to the last few decades or century. Shells that survive aggressive taphonomic conditions in the initial postmortem period apparently shift to a significantly lower loss-rate via burial to a sequestration zone and/or diagenetic stabilization. Such a two-phase model of fossilization reconciles the high loss-rates observed in short-term experiments and the very old shells present in DAs. Live-dead studies in diverse coastal systems reveal that DAs differ from censused ‘living assemblages’ primarily in being temporally coarse samples, contrary to concerns that bias from transport and destruction dominates. DAs differ from local living assemblages – in richness, species identities, etc – because they pool past live variability; the strongest correlate of *poor live-dead agreement is human activity, which can shift today’s community from its historic composition, which the DA remembers. Temporal pooling thus damps the ability of DAs to detect short-term (e.g. seasonal) variation but promotes their ability to detect significant increases and decreases in the populations of key species or functional groups at the habitat scale, estimate the abundance structure of the regional metacommunity, inventory rare species, and identify now-absent species, community states, and anthropogenically shifted baselines. We are now working to develop practical protocols for managers, and are evaluating the effects of burial on temporal resolution and biological fidelity, using cores from an urban shelf where a well-known history of 20^th Century pollution provides a benchmark for ecological history.