Abstract The Late Pennsylvanian was a time of ice ages and climate dynamics that drove biotic changes in the marine and non-marine realms. The apex of late Paleozoic glaciation in southern Gondwana was during the Late Pennsylvanian, rather than the early Permian as inferred from more equatorial Pangaea. Waxing and waning of ice sheets drove cyclothemic sedimentation in the Pangaean tropics, providing an astrochronology tuned to Earth-orbital cycles, tied to climatic changes, reflected in aeolian loess and palaeosol archives. Vegetation change across the Middle–Late Pennsylvanian boundary was not a ‘Carboniferous rainforest collapse’, but instead a complex and drawn out step-wise change from one kind of rainforest to another. Changes in marine invertebrate and terrestrial vertebrate animals occurred across the Middle–Late Pennsylvanian boundary, but these did not lead to substantive changes in the organization of those communities. The base of the Upper Pennsylvanian is the base of the Kasimovian Stage, and this boundary needs a GSSP to standardize and stabilize chronostratigraphic usage. To avoid further chronostratigraphic confusion, the Cantabrian Substage should be abandoned, and the traditional Westphalian–Stephanian boundary should be returned to and recognized as the time of major floristic change, the lycospore extinction event.

Abstract The Late Pennsylvanian was a critical juncture in tetrapod evolution when many terrestrially adapted taxa first appeared. The Middle Pennsylvanian (Moscovian) tetrapod record reflects a taphonomic megabias that favoured preservation, discovery and collection of aquatic tetrapods that lived in wetland palaeoenvironments (‘coal swamps’). The Kasimovian tetrapod record is limited to seven localities, all but one in the USA, and two of which are singleton records, so it is less abundant, diverse or widespread than earlier Moscovian and later Gzhelian tetrapod records. This ‘Kasimovian bottleneck’ hinders interpretation of tetrapod evolutionary events across the Middle–Late Pennsylvanian boundary. Significant changes did take place across that boundary, but they were spread out over Moscovian through Gzhelian time. Many of the perceived changes in tetrapods across the Middle–Late Pennsylvanian boundary are largely artefacts of facies changes and the Moscovian tetrapod taphonomic megabias and of the limited fossil record of Kasimovian tetrapods. Therefore, there is no simple link between Late Pennsylvanian tetrapod evolutionary events and changes in climate and vegetation.

Abstract Tetrapod (amphibian and amniote) fossils of Carboniferous age are known almost exclusively from the southern part of a palaeoequatorial Euramerican province. The stratigraphic distribution of Carboniferous tetrapod fossils is used to identify five land-vertebrate faunachrons: (1) Hortonbluffian (Givetian–early Visean), the time between the first appearance datum (FAD) of tetrapods to the beginning of the Doran; (2) Doran (late Visean–early Bashkirian), the time between the FAD of the baphetid Loxomma and the beginning of the Nyranyan; (3) Nyranyan (late Bashkirian–Moscovian), the time between the FAD of the eureptile Hylonomus and the beginning of the Cobrean; (4) Cobrean (Kasimovian–late Gzhelian), the time between the FAD of the eupelycosaur Ianthasaurus and the beginning of the Coyotean; and (5) Coyotean (late Gzhelian–early Permian), the time between the FAD of the eupelycosaur Sphenacodon and the beginning of the Seymouran. This biochronology provides insight into some important evolutionary events in Carboniferous tetrapod evolution.