Geological context
The geological history of the IAA is complex. Here we focus on summarising geological inference on the history of three broad clusters of putative island systems that may have formed to the north of the Australian continent, and which have been hypothesized to have a role in regional biotic diversification and dispersal through the Oligo-Miocene. Respectively these are: island arc and continental ribbon terranes that may underpin contemporary archipelagos spanning from the Philippines and to Melanesia (Oliver et al. , 2018a): the proto-Papuan Archipelago (Jønsson et al. , 2010); and the islands now clustered into the contemporary region of Wallacea (Moyle et al. , 2016). Due to the complex history of the IAA, some components of these clusters are not mutually exclusive. For example the southward migrating Caroline Arc has accreted onto the northern edge of New Guinea (e.g. Zahirovic et al. 2016). Similarly, the Halmahera Arc has migrated west from the New Guinea region into the contemporary Wallacean region (e.g. Hill & Hall 2003).
In this paper, when we refer to the region of Melanesia we include the island of New Guinea - reflecting the broader contemporary conceptualisation of the region across biotic and cultural spheres. We note that this broad definition differs from some key literature on birds which uses a more restrictive definition for Melanesia that excludes New Guinea (e.g. Mayr & Diamond, 2001),
For the broad history of island terranes in the IAA we here focus on Zahirovic et al. (2016) plate reconstructions but recast into a pure paleo-magnetic reference frame from Merdith et al. (2021) using GPlates (www.gplates.org). This allows us to preserve relative plate motions but remain true to “true” geographic paleo-latitudes that are relevant to climatically-sensitive analyses, such as those required for paleobiogeography. The distribution of shallow marine, emergent land, and elevated topographic regions are modified from paleogeographic reconstructions presented in Cao et al. (2017). Geological reconstructions will be uploaded to Dryad, and can also be accessed here - https://www.dropbox.com/t/nqL4mrvMdPlCxjQX
Island arc terranes and continental “ribbon terranes ”: Continental ribbon terranes probably had their origin in the convergence zone between the Eurasian, Indo-Australian, and Pacific tectonic plates. The oldest parts of the Philippine Archipelago likely formed 160 million years ago along the northern margin of Australia (Deng et al. , 2015; Zahirovic et al. , 2016). In addition to these Australia-derived terranes, younger volcanic arcs formed from ongoing subduction further to the north, including the Halmahera-Caroline, the Izu-Bonin-Mariana, and the Melanesian arc systems. For a broad period spanning from around 45 to 25 Ma tectonic reconstructions suggest that the Philippine, Caroline, and Melanesian arc systems may have broadly aligned to form a chain extending thousands of kilometres across the tropical south-west Pacific (Fig. 1). Within this chain, the Philippine arc terrane was proximal to Sundaland, and potentially provided a pathway for colonisation of the more distant islands of the Vitiaz Arc (itself comprised of the eastward continuation of the Melanesian and Caroline arc systems).
The proto-Papuan Archipelago : Within the IAA a persistent area of geological uncertainty in biogeographic analyses has been around the extent of islands in the present-day region of New Guinea, and whether these now-accreted fragments were contiguous or separate from the main Australian continent. This is further complicated by geologically long- and medium-term sea level fluctuations of up to ~250 m since the Cretaceous (Haq, Hardenbol, & Vail, 1987) that may periodically isolate or connect these landmasses. Geological constructions suggest that the Sepik terrane likely collided with the Australian continental margin ~50–30 million years ago, and generated uplift and topographic relief over the collisional timeframe (Schellart & Spakman, 2015; Zahirovic et al. , 2016; Mahoney et al. , 2019). The Caroline Arc likely collided between 15 and 5 million years ago (Zahirovic et al. , 2016; Mahoneyet al. , 2019). Progenitor islands uplifted by these collisions were likely either partially or entirely isolated from the Australian mainland by a shallow sea over most of the last ~50 million years (Norvick, 2003; Golonka et al. , 2006; Harringtonet al. , 2017). This flooding of the northern Australian continental margin has been ascribed to the role of sinking tectonic plates (Harrington et al. , 2017), causing “dynamic subsidence” of the margin and north-eastward tilting of the Australian continent. The proto-Papuan archipelago itself was emergent only due to the collisions (e.g. the Sepik and Caroline arc accretions) occurring on this continental margin (Mahoney et al. , 2019), leading to crustal shortening and uplift on the leading edge of the advancing Australian continent (Fig. 1).
Wallacea : This region denotes the collection of islands bound to the west by Wallace’s Line and to the east by Lydekker’s Line, including the Lesser Sundas, Sulawesi and Maluku (Fig. 1). It forms the contemporary geological and biotic bridge between the Sundaland (Eurasian) and Sahul (Gondwanan) bioregions. The islands of Wallacea are a collage of continental fragments, volcanic island arcs related to subduction, and possibly also accreted volcanic plateaus from mantle hotspots (Zahirovic et al. , 2016). Their complex tectonic evolution is evident in the network of suture zones (Zahirovic et al. , 2016), indicating the consumption of ancient ocean gateways during the main phase of equatorial ocean gateway closure since Australia’s northward drift from Antarctica (Whittaker, J. M., Williams, S. E., and Müller, 2013). The islands between the Bird’s Head Peninsula in north-western New Guinea and the east/southeast arms of Sulawesi represent the ‘Sula Spur’, which is a continental promontory attached to the Australian continent. The arrival of these blocks in the near-equatorial region by 20 million years ago marks the final closure of the Indonesian oceanic gateway, and the onset of the main phase of biological exchange between Australia and Asia. Most of the concentration and uplift of islands in Wallacea has likely occurred subsequent to this closure (in the last 20 million years) (Zahirovicet al. , 2016). The Lesser Sunda Islands are thought to have become emergent in the last ~10 million years, due to the interplay of Australia-Sundaland collision and the opening of the Banda Sea (Hinschberger et al. , 2005). Parts of Maluku and Sulawesi are also built on older Australian continental crust, forming the Sula Spur (Audley-Charles et al. , 1979), but have experienced extensive uplift and reconfiguration in the last 20 million years.