Molecular Phylogenetics and Evolution 109 (2017) 246–258 Contents lists available at ScienceDirect Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev Bioacoustic and multi-locus DNA data of Ninox owls support high incidence of extinction and recolonisation on small, low-lying islands across Wallacea Chyi Yin Gwee a, Les Christidis b,c, James A. Eaton d, Janette A. Norman b, Colin R. Trainor e, f a, Philippe Verbelen , Frank E. Rheindt ⇑ a Department of Biological Sciences, National University of Singapore, 14 Science Drive 4, 117543 Singapore, Singapore b National Marine Science Centre, Southern Cross University, Coffs Harbour, New South Wales 2450, Australia c Department of Genetics, University of Melbourne, Australia d Birdtour Asia Ltd, 17 Keats Avenue, Littleover, Derby DE23 4EE, United Kingdom e Faculty of Science and Technology, Federation University Australia, Mt Helen, Victoria 3350, Australia f Krijgsgasthuisstraat 89, 9000 Ghent, Belgium article info abstract Article history: Known for their rich biodiversity and high level of endemism, the islands of Wallacea serve as natural Received 22 August 2016 laboratories for the study of spatio-temporal evolution and patterns of species diversification. Our study Revised 23 November 2016 focuses on the owl genus Ninox, particularly the Southern Boobook (N. novaeseelandiae) and Moluccan Accepted 20 December 2016 Boobook (N. squamipila) complexes, which are widely distributed across Australasia. We conducted bioa- Available online 23 December 2016 coustic and multi-locus DNA analyses of 24 Ninox owl taxa to evaluate relationships and levels of diver- gence within the two complexes and ultimately assess the relationship between patterns of taxonomic Keywords: differentiation and bioclimatic factors. We found that taxa that are vocally and/or genetically distinct Island biogeography from populations on the Australian mainland are found on islands that are significantly larger and higher Evolution Molecular phylogeny in altitude than taxa that are vocally and/or genetically indistinct from populations on the Australian Archipelago mainland. This pattern suggests that taxa occurring on small, low-lying Wallacean islands are likely to Faunal diversification be recent colonisers that have dispersed from Australia. Overall, our observations demonstrate that the genus Ninox is likely to have colonised the Wallacean region multiple times as small, low-lying islands undergo frequent extinction, whereas populations on large and high-altitude islands are more resilient. Ó 2017 Elsevier Inc. All rights reserved. 1. Introduction and disappeared in the course of glacial-interglacial cycles (Bird et al., 2005; Caputo, 2007; Sathiamurthy and Voris, 2006; Siddall Australasia is a biogeographically distinct region comprising the et al., 2003; Voris, 2000). archipelagos of Wallacea, New Guinea and satellites, Australia, The islands of Wallacea constitute a natural laboratory for bet- New Zealand and several Pacific islands. It is a region of rich biodi- ter understanding faunal diversification and evolutionary pro- versity and high levels of endemism (Marchese, 2015; Mittermeier cesses across Australasia (Filardi and Moyle, 2005; Lohman et al., et al., 1999, 2011; Myers et al., 2000), characterised by exception- 2011; Wallace, 2009; Whittaker and Fernández-Palacios, 2007). ally rapid tectonic movements at a much larger scale than typical Various studies have shown that the diversification of faunas for most other regions in the world (Hall, 2002, 2012; Hall and across archipelagos is not always driven by a simple one-way, Holloway, 1998). Additionally, Australasia has gone through a per- downstream flow of colonists from continent to islands (Filardi iod of complex changes in its land distribution during the Pleis- and Moyle, 2005; Gohli et al., 2015; Jønsson et al., 2014; Päckert tocene epoch across the last 2 mya as a result of sea level et al., 2013; Stervander et al., 2015). Evolutionary processes are fluctuations. Australia and New Guinea have periodically formed sometimes more complex, involving multiple events of recolonisa- a continuous landmass, while land bridges connecting some of tion and even backward colonisation from islands to continents. A the Wallacean islands, but not others, have repeatedly emerged study by Jønsson et al. (2014) revealed that Pachycephala colonised Melanesia and Wallacea multiple times, resulting in old and young Corresponding author. lineages occurring in the same region today. The young lineages ⇑ E-mail address: [email protected] (F.E. Rheindt). are found to have distributions of variable extent, while the old http://dx.doi.org/10.1016/j.ympev.2016.12.024 1055-7903/Ó 2017 Elsevier Inc. All rights reserved. C.Y. Gwee et al. / Molecular Phylogenetics and Evolution 109 (2017) 246–258 247 lineages are more narrowly endemic to islands, supporting the morphology (Johnstone and Darnell, 1997; Mayr, 1943), the spe- taxon cycles hypothesis where a taxon goes through sequential cies limits of the Southern Boobook complex remain controversial. phases of expansion and contraction (Wilson, 1959, 1961). Additionally, discrepancies in the taxonomy of the Moluccan Boo- Here, we focus on the owl genus Ninox, which has more than book (N. squamipila) complex, consisting of one to five species half of its species members distributed across Australasia (Debus, (Christidis and Boles, 2008), also contribute to the vastly compli- 2002). Owing to their cryptic plumage, many Ninox species have cated taxonomic classification of the genus Ninox. complex patterns of taxonomic diversity. During the early taxo- We analysed a total of 24 Ninox taxa, focusing primarily on the nomic exploration of the region’s avifauna, even the famous Southern Boobook and Moluccan Boobook complexes (Fig. 1), ornithologist Ernst Mayr once remarked on the extreme challenges using bioacoustic and multi-locus DNA data. Ultimately, we aimed posed by the taxonomic classification of the Southern Boobook to (i) examine the complementarity of bioacoustic and genetic (N. novaeseelandiae) complex (Mayr, 1943), which has been treated results, (ii) evaluate the species limits within the Southern Boo- as one (Christidis and Boles, 2008; Norman et al., 1998a), two book complex and Moluccan Boobook complex, (iii) investigate (Schodde and Mason, 1997), three (del Hoyo and Collar, 2014) or the evolutionary history of the genus Ninox, and (iv) assess the even four species (Dickinson and Remsen, 2013; Mikkola, 2014). relationship between patterns of taxonomic distribution and bio- Despite various independent studies on bioacoustics (Brighten, climatic conditions. In order to assess this relationship, we tested 2015; Olsen and Debus, 2013; Olsen et al., 2010, 2002b, 2002c), the hypothesis that deeply diverged taxa, supported by genetics (Norman et al., 1998a; Wink et al., 2009, 2008) and strongly distinct vocal and/or genetic data, occur on islands that Fig. 1. Geographical distribution of the Southern Boobook (N. novaeseelandiae) complex and Moluccan Boobook (N. squamipila) complex. The inset on right shows a detailed view of the Wallacean region. Abbreviations used in legend: SBC, Southern Boobook complex, and MBC, Moluccan Boobook Complex. Ninox novaeseelandiae undulata was endemic to Norfolk Island before it went extinct. Other taxa analysed in the study but not illustrated on the map include the Australian species N. rufa, N. strenua and N. connivens, the Christmas Island endemic N. natalis, and the two outgroups, N. jacquinoti and Athene superciliaris. 248 C.Y. Gwee et al. / Molecular Phylogenetics and Evolution 109 (2017) 246–258 are significantly larger and higher in altitude than islands inhab- we assessed the vocal diagnosability of variables using the crite- ited by weakly diverged taxa. Our null hypothesis was that size rion outlined by Isler et al. (1998), henceforth referred to as the and maximum altitude of islands where distinct taxa occur are Isler criterion. The Isler criterion is based on two conditions: (i) not significantly greater than islands where indistinct taxa occur. there must be no overlap between the ranges of measurements Assessing the extinction potential of landmasses according to their between the two taxa being compared, and (ii) the means x ð Þ geographic and climatic attributes is an important step forward in and standard deviations (SD) of the taxon with the smaller unraveling the mechanisms that have led to faunal diversification set of measurements (a) and the taxon with the larger set of in biologically complex regions such as Australasia. measurements (b) have to meet the following requirement: x t SD x t SD , where t refers to the t-score at the 95th a þ a a 6 b À b b i percentile of the t distribution for n – 1 degrees of freedom. Since 2. Materials and methods there must be no overlap between the two sets of measurements, measurements of one taxon are uniformly higher than the other, 2.1. Collection and measurements of Ninox sound recordings and a one-tailed test with a significance level of 5% was used. The Isler criterion is substantially more discriminating than the We collected a total of 207 sound recordings from various t-test and Mann-Whitney U test because it uses the standard devi- sources, including our private collections, McPherson ations of the sample points, and not the standard deviation of the National History Unit Sound Archive (http://www.archivebirdsnz. taxon mean, which is much smaller (Isler et al., 1998; Rheindt com), Professional