We conducted an ornithological survey of the Kilkerran massif on Fergusson Island, Milne Bay Province, Papua New Guinea. Our records of Torrent Flycatcher Monachella muelleriana represent the first confirmation of the family Petroicidae in the D'Entrecasteaux Archipelago. Additionally, Rainbow Bee-eater Merops ornatus, Peregrine Falcon Falco peregrinus, Red-collared Myzomela Myzomela rosenbergii longirostris and Eurasian Tree Sparrow Passer montanus were all additions to the avifauna of Fergusson Island. We also report the first visual record of Papuan Mountain Pigeon Gymnophaps albertisii. In total, we recorded 70 species. Notably, we did not observe Black-naped Pheasant-Pigeon Otidiphaps insularis, a Fergusson Island endemic, and express concern as to the vulnerability of this virtually unknown species to extinction.
The D'Entrecasteaux Archipelago in Milne Bay Province, Papua New Guinea is an area of pronounced avian endemism (Stattersfield et al. 1998). The largest islands are Goodenough (687 km2; max. elevation 2,536 m), Fergusson (1,437 km2; max. elevation 2,073 m) and Normanby (1,000 km2; max. elevation 1,158 m) (Fig. 1). The avifauna of Milne Bay Province has been studied by a number of collectors and ornithologists over the last 150 years, however, the outlying islands have received less attention than the mainland (Frith & Beehler 1998). Past ornithological visits to Fergusson Island include those by C. Hunstein and A. Goldie (1882), O. Finsch (1885), Rickard (1891), A. Meek (1894, 1895, 1896, 1897, 1899, 1901, 1912–13), H. Hamlin (1928), F. Mayer (1935), the Fourth Archbold Expedition (1953), the Fifth Archbold Expedition (1956), M. LeCroy (1978), T. Pratt and M. Moore (2003), P. Gregory (2004), and T. Laman and E. Scholes (2011) (Frith & Beehler 1998).
At their closest point, the D'Entrecasteaux Islands are c.18 km from mainland New Guinea, with the Goschen Strait separating East Cape (on the mainland) and Normanby Island (Fig. 1). During past ice ages, sea levels were up to 120 m lower than at present, and the D'Entrecasteaux Archipelago formed a contiguous landmass termed the D'Entrecasteaux Shelf (Diamond 1972, Bintanja et al. 2005). Consequently, the gap between the D'Entrecasteaux Shelf and mainland New Guinea was narrower during these periods, both between East Cape and Nuakata Island and Normanby Island in the south, as well as between the Mount Trafalgar peninsula, and the Lusancay Islands and Trobriand Islands in the north (Bintanja et al. 2005; D. Mitchell pers. comm.).
The D'Entrecasteaux Islands share many bird families with mainland New Guinea, including birds of paradise, honeyeaters, and gerygones. However, the archipelago lacks genera characteristic of those New Guinean satellite islands that were formerly connected by Pleistocene land bridges, such as the Aru Islands, Salawati and Misool (Diamond 1972). The absence of avian genera in the D'Entrecasteaux group including Talegalla (Megapodiidae), Goura (Columbidae), Crateroscelis and Sericornis (Acanthizidae), Ptilorrhoa (Cinclosomatidae), Arses (Monarchidae) and Pitohui (Oriolidae) indicates that these are true oceanic islands (Diamond 1972). High rates of avian endemism in the D'Entrecasteaux Archipelago reinforce this conclusion. The current species-level ranking of Oya Tabu White-eye Zosterops crookshanki, and the recommendations that Black-naped Pheasant-Pigeon Otidiphaps nobilis insularis and Red-collared Myzomela Myzomela rosenbergii longirostris merit species status results in a heightened level of importance for the D'Entrecasteaux Islands, which are already recognised as an Endemic Bird Area (Stattersfield et al. 1998, del Hoyo & Collar 2014, 2016, Beehler & Pratt 2016). Additional, morphologically distinct subspecies including Papuan Myzomela Myzomela nigrita forbesi and Little Shrikethrush Colluricincla megarhyncha fortis deserve further taxonomic study.
Mayr & Van Deusen (1956) noted high levels of similarity between the avifaunas of Goodenough, Fergusson and Normanby Islands. Exceptions do occur, such as Goldie's Bird of Paradise Paradisaea decora, which is restricted to Fergusson and Normanby, and Fantail Monarch Symposiachrus axillaris, which to date has been recorded only on Goodenough. Further surveys will permit a fuller comparison of these islands' avifauna, as many species records are based on observations made during a single visit.
We visited Fergusson Island on 6–22 August 2019, focusing our efforts on a ten-day survey of the eastern mountain range, the Kilkerran Massif (known locally as Oya Tabu). Specifically, we ascended Mount Othona (Oya Nai), the massif's second tallest peak, which to our knowledge has not been surveyed previously for birds. Vegetation on Fergusson Island is characterised by lowland and montane rainforest. Areas around villages present extensive mosaics of subsistence gardens and secondary forest. On the east slope of Mount Othona, forests above c.600 m appeared undisturbed, with substantial undergrowth and no established trails.
We report observations of Torrent Flycatcher Monachella muelleriana on Fergusson (the first confirmed record of the Petroicidae in the D'Entrecasteaux Archipelago), a range extension for the endemic Red-collared Myzomela Myzomela rosenbergii longirostris, as well as four additional first records (three of them documented). We also discuss our failure to find Otidiphaps nobilis insularis, which is listed by some authorities as a species (del Hoyo & Collar 2014) and is treated as Endangered (IUCN 2020). This taxon has not been recorded by biologists since it was collected in 1882 (Godman & Salvin 1883, Frith & Beehler 1998, Beehler & Pratt 2016, IUCN 2020, del Hoyo et al. 2020a).
On 7–10 August 2019, we made opportunistic observations between sea level and c.300 m in and around Sebutuia Bay (09o56.962'S, 150o86.927'E) and the south-east slopes of the Kilkerran Massif, using active logging roads from Sebutuia to access higher elevation areas. Thereafter, on 11–20 August, we surveyed the avifauna of the eastern slope of Mount Othona. On 11 August, we hiked from sea level at the village of Basima to 920 m, where we camped on a forested ridge (09o49.466'S, 150o79.464'E). On 12 August, we cut a trail from this camp to c.1,245 m through primary montane forest. We conducted observations opportunistically between 11 and 17 August from this trail. On 17 August, we descended to 388 m and established a second camp (09o49.362'S, 150o81.018'E). Until 20 August we made observations of birds in a 1 km-stretch of riparian and adjacent forests defined by a large, fast-flowing river (known locally as Awalota).
We follow the taxonomy of Beehler & Pratt (2016) but also acknowledge species splits made in del Hoyo & Collar (2014, 2016). Birds were identified to species visually using Pratt & Beehler (2014) and Gregory (2017), and by ear with the aid of recordings archived at Xeno-canto ( https://www.xeno-canto.org). No systematic or quantitative surveys were attempted. Birds were photographed and sound-recorded opportunistically by JG using a Sony RX10 IV, a Røde videomic pro shotgun microphone, or an Apple iPhone. Images and sound-recordings have been deposited in the Macaulay Library ( https://www.macaulaylibrary.org/) and Xeno-canto, respectively. Reference numbers are provided in Tables 1–2 for all such media.
In December 2019, we studied specimens of Monachella muelleriana (AMNH 341290–294, 420600–603, 426812–816, 608387–405 and 70592–599) and M. m. coultasi (AMNH 334453–456, 334568–570, 417432) at the American Museum of Natural History, New York, comparing them with photographs of M. muelleriana taken by JG on Fergusson.
We identified 70 species on Fergusson Island, including four migrants and 66 residents. Our new records are presented below, followed by a discussion of Otidiphaps nobilis insularis and the general conservation of Fergusson's avifauna. Tables 1–2 present our overall species lists, divided into lowland observations made on 7–10 August, during which 37 species were identified (Table 1), and our more thorough survey (45 species) of the montane avifauna up to 1,245 m on 11–20 August (Table 2).
All bird species observed 7–10 August 2019, at 0–300 m. * = new island record. Reference numbers correspond to photographs uploaded to the Macaulay Library (ML).
All bird species observed on 11–20 August 2019, at 300–1,245 m. * = new island record. Reference numbers correspond to sound-recordings uploaded to Xeno-canto (XC) and photographs uploaded to the Macaulay Library (ML).
PAPUAN MOUNTAIN-PIGEON Gymnophaps albertisii
On 11 August, one was seen perched in the understorey at c.900 m. Its overall drab, non-green plumage quickly differentiated it from Ptilinopus fruit doves. The large patch of red orbital skin and dark wings contrasting with a paler breast differed from Pinon's Ducula pinon and Zoe's Imperial Pigeons D. zoeae. On the same date, we also heard the distinct mechanical wing sounds produced by a flock of Gymnophaps in flight. This is the first sight record of G. albertisii on Fergusson, although it was previously heard by T. K. Pratt (pers. comm.) during a 2003 visit but has not been otherwise reported. It is endemic to mainland New Guinea, the Bismarck Archipelago, Bougainville, and nearby Goodenough Island (Beehler & Pratt 2016).
RAINBOW BEE-EATER Merops ornatus
On 7 August, we identified one at c.250 m by its brightly coloured orange, green and blue plumage, and black upper breast-band. On 8 August, we observed four, two of them perched directly by the shore in Sebutuia Bay. On both days, individuals perched on prominent branches providing clear views. This is the first record of M. ornatus on Fergusson, although the species has previously been recorded in the D'Entrecasteaux Archipelago on Goodenough (LeCroy & Peckover 2000). M. ornatus primarily breeds in northern Australia, but has also been reported nesting at a few localities in New Guinea. It occurs throughout New Guinea, the Bismarck Archipelago, Sulawesi and the Lesser Sundas in the non-breeding season, and has been reported on several New Guinea satellite islands, including Woodlark and Karkar, as an austral migrant (Diamond & LeCroy 1979, Vang 1991, Fry et al. 1992). We suspect previous ornithological surveys did not encounter this easily detected species on Fergusson due to its presumed absence outside of the austral winter.
PEREGRINE FALCON Falco peregrinus
On 12 August, we observed a vocalising individual, subspecies ernesti, on high branches. We subsequently found an active nest in a large emergent Araucariaceae tree above the forest canopy (ML 238386491). Two adults were seen and the nest was observed daily on 12–17 August, during which at least one individual was apparently incubating. F. peregrinus has been previously reported on Normanby Island in the D'Entrecasteaux Archipelago (Beehler & Pratt 2014), but this is the first record on Fergusson and the first breeding record in the archipelago. Throughout mainland New Guinea, F. p. ernesti is a widespread, uncommon resident that has been seen on other satellites such as Woodlark Island, making its presence on Fergusson expected (Rothschild & Hartert 1896, Gregory 2017).
RED-COLLARED MYZOMELA Myzomela rosenbergii longirostris
On 15 August, at 1,055 m, we observed a Myzomela that we identified as this taxon foraging in the forest canopy. Its black head and wings, and contrasting bright red neck and breast, immediately differentiated it from Papuan Black Myzomela M. nigrita forbesi and any other passerine on Fergusson. On 16 August, we saw another within 20 m of our initial observation. This bird was foraging near a M. nigrita, offering a direct comparison. In addition to the obvious plumage differences, its bill appeared longer and less steeply decurved. The same day, we observed what we suspect to have been an immature male, based on its mottled black-and-brown head, foraging at c.1,245 m. Finally, we observed a female foraging within a mixed-species flock. Identified by its olive-green body and red collar, female plumage is the most distinctive feature of the D'Entrecasteaux taxon vs. the mainland form. Camera failure due to high humidity and limited survey time meant that we failed to photograph M. r. longirostris.
M. r. longirostris was described from Goodenough, where it has been reported to be common to very common at 1,000–2,750 m (Mayr & Van Deusen 1956, Gregory 2017). Recently, del Hoyo & Collar (2016) treated it specifically (Long-billed Myzomela) due to its morphology, and the taxon is listed by the IUCN as Near Threatened. Our observations of M. r. longirostris on Fergusson are the first on the island of this little-known taxon, but are not unexpected given Pleistocene land connections with Goodenough and because Fergusson's mountains cover its elevational range.
We observed M. r. longirostris just four times, all at 900–1,245 m. Our survey only reached the lower elevational range reported for this species on Goodenough and it is possibly commoner at higher elevations on Fergusson. Future work throughout the archipelago is required to better understand this taxon's distribution and its basic biology. However, we suspect it is unlikely that M. r. longirostris is present on nearby Normanby, whose mountains lie below this taxon's reported elevational range.
TORRENT FLYCATCHER Monachella muelleriana
On 10 August, at least two M. muelleriana (Fig. 2) were seen along a fast-flowing large river (the Awalota) with massive boulders and deep pools, at c.300 m. They were perched side by side on branches above the channel and alighted on boulders within it, vocalising frequently. On 17–20 August, we returned to the same general area and made sound-recordings of, and photographed, individuals and pairs. None was observed at smaller, higher elevation streams.
The birds had a pale grey back, white rump, white throat, breast and belly, black bill and legs, large white forehead spot, and dark brown eyes. The head, wings and tail were black (ML 238404731). No substantial individual variation was observed, even among pairs of suspected males and females, and no individuals matching the description of a juvenile M. muelleriana were observed.
Two primary vocalisations were recorded. The most commonly heard was a frequently given call of 1–6 piping, bell-like notes on the same pitch (XC 567490). Presumed to be a contact call, it was heard near-constantly while observing closely associating pair members. Less frequently heard was a rapidly delivered, repeated, 2–3-second descending staccato series of notes, followed by several longer, more emphatic, higher pitched ones (XC 567488). We believe this to be the species' song given that it was made during what appeared to be courtship behaviour, as well as when a lone individual moved quickly along the river, as if patrolling or defending a territory.
Our records of Torrent Flycatcher are the first confirmation of the family Petroicidae in the D'Entrecasteaux Archipelago. A report of M. muelleriana from Goodenough Island by De Vis (Mayr 1941) was doubted by subsequent authors due to the lack of further observations, but the species' presence on Fergusson now offers support for the earlier record (J. M. Diamond pers. comm). An eBird record, verified by us, also reports the species for Normanby (E. Enbody pers. comm.). We consider ours the first confirmed records of M. muelleriana for the D'Entrecasteaux Archipelago.
This New Guinea endemic is widespread and uncommon on fast-flowing hill and mountain streams to 2,130 m (Boles & Christie 2019). A second population, M. m. coultasi, suggested for species status by del Hoyo & Collar (2016) based on plumage differences, inhabits a limited number of streams on New Britain (del Hoyo et al. 2020b). The apparently small population of M. muelleriana in the D'Entrecasteaux Archipelago is presumably confined to the few larger streams on the islands.
The individuals we observed did not appear different from M. muelleriana of mainland New Guinea, based on our comparison of photographs with specimens at AMNH. However, those on Fergusson presumably represent an isolated population due to the persistent sea barrier between the mainland and D'Entrecasteaux Archipelago since at least the Pleistocene (Bintanja et al. 2005). Given the high degree of endemism in the archipelago, specimens as well as a comparison of the vocalisations of mainland, New Britain and Fergusson birds is necessary to determine their taxonomic placement, but is likely to require additional sound-recordings from all three areas.
EURASIAN TREE SPARROW Passer montanus
On 7 August, we observed a flock of five in Sebutuia, an active logging camp and timber shipping port. They were identified by their chestnut cap, black cheek mark, and stout bill, and the species is well known to us. Now breeding in New Guinea, P. montanus is introduced in the region, and was first reported in Port Moresby in 2006 (Gregory 2009). Its range is rapidly expanding across mainland Milne Bay Province, where it has invaded several villages outside Alotau over the last five years (JB pers. obs.). This is the first record for the D'Entrecasteaux Archipelago. Its presence in the port of Sebutuia on Fergusson is not unexpected, given the frequent arrival of large logging ships there, which could easily transport this and other invasive species.
Conservation of Black-naped Pheasant-Pigeon.—Otidiphaps insularis is a large, terrestrial, ground-nesting bird endemic to Fergusson Island, known to science only from the type specimen, collected in 1882 (Godman & Salvin 1883) and now at the Natural History Museum, Tring. For the purposes of this review we follow del Hoyo & Collar (2016) in treating it as a species, which is listed as Endangered by IUCN (2020).
We failed to observe O. insularis on Fergusson despite visiting appropriate habitat and being familiar with the loud, far-carrying, whistled songs of the genus. On 17–20 August we surveyed an area identified by our local guides as a place where O. insularis is hunted. We placed four camera traps at ground level at locations suggested by our guides as favoured by the bird. This area was also reported by our guides as a breeding location for the species, where they had found an active nest. We failed to detect O. insularis there.
Several ornithological surveys and birding tourist groups that have visited Fergusson during the last two decades have also failed to find O. insularis, and indeed there is no modern-day record of it (Gregory 2007, Beehler & Pratt 2016, del Hoyo et al. 2020a; G. Dutson pers. comm., D. Mitchell pers. comm., T. K. Pratt pers. comm.). We urgently recommend an island-wide survey to determine the status of this highly unique and apparently imperilled island endemic.
There is evidence that large, terrestrial, ground-nesting Columbidae in Australasia are vulnerable to extinction, as evidenced by the Choiseul Pigeon Microgoura meeki, which occurred on Choiseul in the Solomon Islands. M. meeki was last seen in 1904 and is considered to be extinct. Like Black-naped Pheasant-Pigeon, it was known from a single island and believed to lay just one egg. M. meeki is thought to have been driven to extinction by the introduction of cats to the island (Tennent 2009, Baptista et al. 2020).
Bird conservation on Fergusson Island.—Fergusson, along with Goodenough, Normanby, and the Trobriand Islands form an Endemic Bird Area (EBA 196). BirdLife International lists the priority for this EBA as ‘high' and considers knowledge of the area ‘poor’ (Stattersfield et al. 1998). The archipelago hosts the endemic Goldie's Bird of Paradise (Vulnerable), another endemic bird of paradise, the Curl-crested Manucode, and many other poorly studied endemic avian taxa. Our observations of five previously unrecorded species on Fergusson demonstrate that significant gaps remain in our knowledge of the D'Entrecasteaux Archipelago avifauna, which clearly warrants additional field work.
Our results support the assertion made by Mayr & Van Deusen (1956) that Fergusson, Goodenough and Normanby Islands share a very similar avifauna based on their Pleistocene connections as a contiguous landmass, the D'Entrecasteaux Shelf (Diamond 1972). Aside from P. montanus, a recently introduced species in New Guinea, our first records on Fergusson are of species that have previously been observed on Goodenough and / or Normanby. Based on this pattern, we suspect that additional species most likely to be present on Fergusson have already been reported elsewhere in the archipelago. Nevertheless, the possibility of undescribed taxa also remains, given the few ornithological surveys of Fergusson, especially above 1,500 m.
Habitat destruction represents a significant threat to the avifauna of the D'Entrecasteaux Archipelago and is currently occurring at a vast scale throughout New Guinea, including other comparably sized satellites such as Woodlark and Manus Islands (D. Mitchell pers. comm). Shearman & Bryan (2011) reported that between 1972 and 2002, the D'Entrecasteaux group suffered the highest level of deforestation relative to forest cover throughout Papua New Guinea, with most loss and degradation in the lowlands. Also, Shearman & Bryan (2011) classified the D'Entrecasteaux Islands as representing a distinct bioregion (one of 11 recognised by their study) based on a ‘radically’ distinct freshwater and amphibian fauna, and further supported by their high degree of avian endemism (Allison 1993, Polhemus et al. 2004, Shearman & Bryan 2011).
Selective logging of primary rainforest has been ongoing on Fergusson for decades. A large network of maintained logging roads originates in Sebutuia and extends throughout central and eastern parts of the island, and we observed the accumulation of many logs bound for foreign markets. The D'Entrecasteaux Archipelago, including Fergusson, also has a large and increasing human population, with a resulting increase in forest clearance for subsistence agriculture. We suspect effective conservation of the islands' birds requires locally driven strategies and sustained investment in local communities by NGOs and the Papua New Guinea government. Notably, escalating rates of sea piracy have made essential travel for local people difficult, while also presenting a barrier to research and conservation. Finally, given the frequent arrival of large ships at Sebutuia Bay, it is a potential location for the introduction of species such as cats and rodents which could severely impact Fergusson's ecosystems.
Foremost, we thank the people of Sebutuia, Basima and upper Basima for their hospitality and trust, without whom our study of the island's avifauna would have been impossible. Their generosity and knowledge of the terrain and fauna were essential to the success of our research. Eli Malesa provided vital logistical support and Aisodi Madiu, Brandon Sam, Rio Balaya, and Teliwa Sedebo helped with navigation, cutting trails and establishing camps. We also thank Serena Ketaloya, Captain John Sarusaruna, and the rest of our crew for transporting us safely. David Mitchell of Eco-custodian Advocates provided indispensable advice during this manuscript's preparation, and Jared Diamond and Thane Pratt greatly improved it with their detailed reviews and enthusiasm for the topic. Jack Dumbacher and David Bishop refereed the submitted manuscript. Guy Kirwan proffered generous assistance and patience as editor. David Bishop and Guy Dutson provided their notes and support, and Thane Pratt and Michael Moore generously shared their unpublished trip reports and bird lists.