The replacement of Paleozoic brachiopod-dominated marine benthic communities by post-Paleozoic assemblages dominated by molluscs was one of the most significant ecological transitions in the Phanerozoic, completely restructuring benthic ecosystems and paving the way for modern marine communities. The timing of the abrupt diversity switch has been tightly constrained to the catastrophic mass extinction at the Permian-Triassic boundary. In contrast, the shift in ecological dominance, as measured by relative abundance in marine communities, has only been assumed to be synchronous with the taxonomic change. This assumption ignores environmental changes throughout the Permian as well as potential effects of the earlier end-Guadalupian extinction (at the end of the Middle Permian). In order to test whether the ecological transition was contemporaneous with the end-Permian taxonomic shift, I quantified Permian community change based on fossil assemblages collected from the western United States (Early and Middle Permian, 15 samples), Greece (Late Permian, 6 samples), and China (Late Permian, 6 samples). All assemblages were derived from offshore carbonate deposits formed in tropical environments.

During the Early Permian, paleoenvironmental trends in community composition parallel those documented from biodiversity studies. Molluscs were extremely abundant in nearshore assemblages (comprising close to 100% of the assemblage), while inner shelf settings contained a mosaic of communities with co-dominant brachiopods and molluscs. However, Early and Middle Permian offshore fossil communities were overwhelmingly dominated by rhynchonelliform brachiopods, with a mean abundance of 98.9%. Brachiopods were approximately evenly split between pedically-attached and reclining forms, with cementing genera rare in most samples. Bivalves only accounted for 0.7% and were strongly dominated by epibyssate suspension-feeding forms (>90% of the bivalve population).

In contrast, Late Permian offshore assemblages contained a mixture of brachiopods and molluscs. Brachiopods only comprised 34.6%, with bivalves accounting for 17.9% and gastropods the most abundant group at 47.5%. Bivalve life habits were also more evenly distributed between epifaunal suspension feeders (52%) and infaunal suspension feeders (42%). In addition, bivalves were comparable in size to co-existing brachiopods. These results demonstrate that a substantial portion of the ecological transition from brachiopods to bivalves---in terms of relative abundance, ecological dominance of infaunal forms, and size distributions---had occurred prior to the end-Permian biotic crisis and was apparently synchronous with the end-Guadalupian extinction.

Despite this large ecological change, a re-evaluation of the severity and selectivity of the end-Guadalupian extinction reveals that it was only a minor event, with an overall extinction rate of 33.8%. Some groups, such as corals and bryozoans, suffered more than others, but the selectivity between rhynchonelliform brachiopods (33.8%) and bivalves (32.7%) was minimal. Paleobiogeographic patterns of extinction show that elevated extinction intensities only occurred in western North America (among brachiopods) and eastern Australia (among bivalves). Both of those regions had a strongly endemic fauna during the Capitanian Stage and lacked marine deposition during the Late Permian because of tectonic activity. Extinction rates throughout most of the Tethyan region were low and not significantly different from either preceding or succeeding stages, implying that this tectonically-induced loss of biotic provincialism was the primary cause of the apparent extinction.

The minimal selectivity and severity in all regions except for North America and eastern Australia (and to some extent, south China) therefore suggests that the ecological change was not triggered by severe taxonomic effects, as has been inferred for the end-Permian crisis. It is not clear what may have triggered the substantial decoupling of global diversity (minor biotic crisis) and local community ecology (major shift in relative abundance). Given the minimal severity of the end- Guadalupian crisis, potential causes of this dramatic increase in the relative abundance of molluscs include: (1) acquisition of evolutionary innovations that conferred a competitive advantage to molluscs; or (2) a change in environmental conditions that favored molluscs over rhynchonelliform brachiopods. As nearly all of the abundant bivalve and gastropod genera in Late Permian assemblages were also present during the Middle Permian, it does not seem likely that a new evolutionary innovation was responsible for their increased Late Permian abundance. Instead, the increase in molluscan abundance may have resulted from environmental changes, possibly precursors to the end-Permian mass extinction, during the Guadalupian- Lopingian interval.