THE END-DEVONIAN HANGENBERG EVENT: CAUSES AND CONSEQUENCES OF A MAJOR BOTTLENECK IN VERTEBRATE EVOLUTION

The end-Devonian Hangenberg mass extinction (359 Ma) was long considered a minor black shale bioevent, one among many in the “Late Devonian Biotic Crisis.” However, recent work has linked the Hangenberg to the most severe extinction of vertebrates in the Paleozoic (Sallan and Coates 2010, Sallan et al. 2011; Friedman and Sallan 2012). This involved losses of 44% of higher-level clades (Fig. 1) and restructuring of both marine and freshwater ecosystems worldwide (Sallan and Coates 2010). Modern dominant groups (e.g., ray-finned fishes, sharks and tetrapods) experienced a severe bottleneck, with a handful of lineages radiating in a 10-20 My recovery interval associated with “Romer’s Gap” (a lull in the tetrapod record; Sallan and Coates 2010; Sallan et al. 2011; Sallan and Friedman 2012). Vertebrate selectivity at the Hangenberg is difficult to parse, as taxa of every lineage, size and ecology were lost. At higher levels, the Hangenberg event seems to have impacted pelagic and terrestrial faunas while sparing benthic invertebrates, the reverse of the more famous Frasnian-Famennian event that left vertebrates unscathed (Sallan and Coates 2010; Sallan et al. 2011; Friedman and Sallan 2012). The causes of the Hangenberg extinction need further examination; it might be a double event. Open marine fishes are lost at the start of black shale deposition associated with transgression, euxinia and volcanism. A subsequent regressive stage, linked to glaciation and either deposition of the Hangenberg sandstone or absence of the boundary, marks nearshore and freshwater losses. The Hangenberg extinction might coincide with an opening of the Siberian traps and/or the Woodleigh bolide impact in Australia, yet no attempt has been made to link these phenomena. Since similar drivers have also been implicated in other Devonian events, the “biotic crisis” interval illustrate how invertebrate and vertebrate extinction can be disjunct, and all attempts to determine causes must take such selectivity into account.