Late Miocene drying of central Australia

Highlights

• Calcareous paleosols reveal late Miocene drying of central Australia.

• Late Miocene vegetation was open woodland and gallery woodland, not rainforest.

• Late Miocene drying of central Australia is globally correlated.

Abstract

A back to the future approach to climate change refines Neogene records of paleoclimatic cooling and drying as future climate states for a warming world. Near Alcoota station in central Australia is a late Miocene fossil mammal site for Alcoota (10 Ma) and Ongeva (8 Ma) local faunas. The Alcoota fauna includes the largest known land bird (Dromornis stirtoni) and early diprotodontid (Kolopsis torus), but the geologically younger Ongeva fauna has a larger diprotodontid (Zygomaturus gilli). These faunas are evolutionary predecessors of Australian Pleistocene megafauna. Calcareous paleosols recorded a marked turn toward aridity compared with underlying thick lateritic duricrusts which lack fossils. Laterites of likely middle Miocene age are evidence of wet (>1100 mm mean annual precipitation) tropical (>17 °C mean annual temperature) paleoclimates, but these were deeply eroded and redeposited by the time of the late Miocene mammals. A succession of 19 paleosols of 5 pedotypes were recognized in the 19 m thick section, with shallow calcareous nodule (Bk) horizons as evidence of mean annual precipitation ranges from 100 to 400 mm, and mean annual range of precipitation of 50 mm, very similar to the modern climate of Alcoota (296 mm precipitation with 56 mm difference between wet and dry month). This was neither a rainforest nor monsoonal paleoclimate, and vegetation inferred from paleosols was open woodland and gallery woodland similar to that of today. Central Australia was not covered by monsoon rainforest during the late Miocene and its paleoclimatic history of Neogene desertification matches well that of other parts of the world.

Introduction

Neogene paleoclimate records may be guides to climate change expected with increased carbon dioxide in the atmosphere, because the future will reverse Neogene cooling and drying now known from a variety of detailed proxies (Breecker et al., 2010; Beerling and Royer, 2011). Thus climate and vegetation of the future can be predicted by reconstructing plants, animals and soils of the past formed at times of known atmospheric carbon dioxide concentration (Retallack et al., 2018a). Estimates of 612 ± 24 ppm atmospheric carbon dioxide for the middle Miocene are of special interest because comparable with predicted carbon dioxide by the year 2100 with current emission scenarios (Retallack et al., 2016). One difficulty for predicting Australian future climate in this way is debate whether Miocene monsoon rain forest was widespread in central Australia (Archer et al., 1994; Travouillon et al., 2009) or limited to northern Australia (Megirian et al., 2004; Herold et al., 2011). Paleosols are useful to such a task because rainforest soils are thick lateritic or bauxitic duricrusts (Retallack, 2010), and aridity and monsoonal seasonality is reflected in patterns of pedogenic carbonate nodules (Retallack, 2005). This study addresses paleoclimate and vegetation changes for two successive mammal assemblages in the late Miocene, Waite Formation near Alcoota, Northern Territory, Australia (Woodburne, 1967; Megirian et al., 1996; Murray and Vickers-Rich, 2004). These can be added to other records of Australian Neogene paleoclimate (Metzger and Retallack, 2010) for comparison with paleoclimatic records from paleosols in other parts of the world (Retallack, 1991a, Retallack, 2007; Retallack et al., 2016, Retallack et al., 2018a).