Although evidence for secular variation in the major-element chemistry of seawater has accumulated in recent years (e.g., Holland, 1984; Wilkinson and Algeo, 1989; Hardie, 1996; Kovalevich et al., 1998; Lowenstein et al., 2003), the possibility of secular variation in the trace-element inventory of seawater has received less attention.  One of the most important influences on trace elements (TEs) in seawater is uptake of redox-sensitive species by black shales.  At present, only ~0.3% of seafloor globally is subject to benthic anoxia, limiting the flux of redox-sensitive TEs to the sediment.  For example, the Black Sea accounts for <10% of the burial fluxes of Mo, U, and V.  Ancient episodes of widespread marine anoxia, such as during the Late Devonian, may have resulted in a much larger flux of redox-sensitive TEs to organic-rich sediments.  Reservoir modeling of changes in the seawater TE inventory as a function of the Late Devonian burial flux represented by black shales suggests that certain trace elements (e.g., Mo) might have been drawn down to <50% of their present-day concentrations.  Potential empirical evidence of seawater TE depletion comes from the Upper Devonian black shale succession of the Central Appalachian Basin.  There, most redox-sensitive TEs show strong positive covariation with TOC and DOP (degree of pyritization, a paleoredox proxy) throughout the 50-m-thick Ohio Shale.  A major exception is Mo, which declines by 70-80% on a TOC-normalized basis in the uppermost 8 m of the Cleveland Member of the Ohio Shale (representing the last <1 m.y. of the Famennian Stage), despite correlative increases in TOC and other redox-sensitive trace elements that are consistent with a decrease in redox potential (as proxied by DOP).  The most likely explanation for this anomaly is a marked reduction in the concentration of Mo in Late Devonian seawater owing to massive uptake by black shales over several million years.