Mid-Cenozoic insular radiation in the IAA
Temporal and spatial patterns of insular diversification within the IAA are poorly understood, and have at times been contentious (Jønssonet al. , 2010; Moyle et al. , 2016). A key challenge is that while geological reconstructions suggest the position of key geological terranes, it is often unclear if these were above or below water (Caoet al. , 2017). A further issue for biogeographic analysis is that the contemporary proximity of geological features differs greatly from paleogeographic configurations. Differing geological, biogeographic and cultural delineations of key regions have potentially further confounded analyses and comparison, most importantly New Guinea is not always considered part of Melanesia (Mayr & Diamond, 2001) and sometimes even lumped together with Australia as a single landmass of Sahul.
Our analyses emphasise that islands in the east of the IAA, here considered to be part of the broader Melanesian region, have been a hotspot of pigeon evolution and speciation since the Oligocene (Fig. S2, S7). Deep (mid-Cenozoic) insular origins have now been inferred for several radiations centred on Melanesia, strongly indicating that key terrane complexes (Vitiaz Arc and Sepik Arc) and/or the proto-Papuan region have likely been shaping diversity across the IAA since the mid-Cenozoic (Aggerbeck et al. , 2014; Oliver et al. , 2018a; Bank et al. , 2021) (Table S8). Wallacea to the west also appears to be a potentially important source of upstream colonists in pigeons, although this smaller region shows younger and weaker signals of lineage accumulation and endemic cladogenesis (Fig 2B,C) and geological evidence suggest more recent arrival and uplift of key geological features (Zahirovic et al. , 2016). To the north-west the Philippine pigeon fauna (Fig. S7B) appears to be relatively young and derived when compared to that of Melanesia or Wallacea. However, in light of the extreme geological dynamism of the region and the mobility of pigeons, we suggest that further discussion of the extent to which Oligo-Miocene insular diversification of pigeons may be linked to specific geological features within Melanesia - such as the Caroline, Sepik or Melanesian Arcs - remains speculative.
While the importance of Melanesian islands in the early diversification of pigeons is strongly supported, the ultimate source from which this region was colonised is obscured by the long stem lineage, the apparent rapid radiation at the base of the Columbiformes and the sparse fossil record. This contrasts with most other old insular radiations in the IAA, for which biogeographic analyses typically point clearly to either Australian or Asian origins (Table S8). One intriguing potential scenario for pigeons is South America; with westward dispersal to island arcs, leading to populations on quasi-continental fragments (Melanesia, Australia/Zealandia and Philippines) and later dispersal to Old World. This is speculative, but we suggest is as consistent with the current information on distribution, diversity, phylogeny, dispersal ability, biogeography and fossil record of pigeons as any other inference. Other Pacific lineages also show evidence of deep ‘out-of-South America’ origins (Malone, Reynoso, & Buckley, 2017), showing that eastward migration across the Pacific is possible, if not common.
The rapid early radiation of pigeons around the mid-Cenozoic also mirrors the estimated timing of initial diversification in two lizard radiations similarly centred on island arcs in the IAA (Oliver et al. , 2018a; Slavenko et al. , 2022). This concordance may reflect the timeframe when island arcs were especially conducive to biotic colonisation and diversification – for example the putative Philippine-Caroline-Vitiaz arcs (Zahirovic et al. , 2016). An alternative or complementary explanation is a shared legacy of global climatic shifts linked to changes in southern ocean circulation patterns (Zachos et al. , 2001). Analyses of plant (Nge et al. , 2020) and reptile (Oliver & Hugall, 2017) lineage diversification patterns in Australia, patterns of lineage diversity in mammal faunas globally (Stadler, 2011), and substantial turnover in marine and terrestrial fossil records (McGowran et al. , 2004; Sun et al. , 2014) all suggest the early Oligocene as a time of profound biotic turnover.