Located off the south-eastern tip of New Guinea, the D'Entrecasteaux archipelago in Milne Bay Province, Papua New Guinea, forms part of the D'Entrecasteaux and Trobriand Islands Endemic Bird Area (Stattersfield et al. 1998, BirdLife International 2024a). The archipelago comprises three main islands—Goodenough, Fergusson and Normanby—with the middle island in the chain, Fergusson, being the largest of the three (1,437 km²). High levels of endemism in this archipelago are the product of these being true oceanic islands that were never connected to the New Guinea mainland (except potentially part of Normanby, see Hill et al. 2023) and are separated by 18 km at their closest point from mainland New Guinea. Their avian communities overlap substantially with mainland New Guinea (Mayr & Van Deusen 1956) but are less diverse due to the dispersal barrier presented by the open-water channel between the mainland and the islands (Diamond 1972). The EBA is home to four endemic species: Long-billed Myzomela Myzomela longirostris, Curl-crested Manucode Manucodia comrii, Goldie's Bird of Paradise Paradisea decora and Oya Tabu White-eye Zosterops crookshanki, as well numerous endemic subspecies including Otidiphaps nobilis insularis, which is often treated as a Critically Endangered endemic species under the vernacular name Black-naped Pheasant-Pigeon (Beehler & Pratt 2016, del Hoyo & Collar 2016, Gregory 2017).

Despite the ornithological significance of the D'Entrecasteaux and Trobriand Islands, research on the regional avifauna has been limited and scientific knowledge of the EBA remains poor (BirdLife International 2024a). The first ornithological collections from the islands were made in the late 1800s, notably in 1882 by Andrew Goldie (for whom the endemic bird of paradise is named), with multiple visits by Albert S. Meek between 1894 and 1913 (see Frith & Beehler 1998, Gregg et al. 2020, for summaries of historical visits). The most recent avifaunal study on Fergusson was a two-week survey of the south-western slope of the Oya Tabu Massif in August 2019 by JG, DN & JB (Gregg et al. 2020). That this study contributed five new records for the island, including four breeding species, evidenced that even basic information such as overall breeding bird diversity remains incomplete for the D'Entrecasteaux Archipelago.

Even less is known about the natural history, vocal behaviour and conservation threats for taxa endemic to the EBA. As one notable example, there is apparently no published description of the nest of Goldie's Bird of Paradise despite this being a charismatic and relatively common species on Fergusson and Normanby. Equally, recordings or descriptions of the vocalisations of several endemic taxa, such as the local subspecies of Yellow-billed Kingfisher Syma torotoro ochracea, Little Shrikethrush Colluricincla fortis fortis and Red-bellied Pitta Erythropitta erythrogaster finschii, are limited. In addition to providing basic ecological information, these data are relevant for taxonomic evaluations and understanding potential threats to the region's birds.

We surveyed areas in the eastern half of Fergusson Island, with a specific focus on the Oya Tabu Massif, during 5–30 September 2022. This work built on the 2019 survey by JG, DN & JB, and formed part of a larger project to assess the conservation status of Otidiphaps nobilis insularis. Here we report the results of those field surveys and provide comments on our records of Pheasant Pigeon (see also Gregg et al. 2024), new distributional records for Fergusson Island, and observations of breeding behaviour and the vocalisations of some of the endemic taxa.

Methods

We visited eastern Fergusson Island during 5–30 September 2022. Our survey sites included three communities in the environs of the Oya Tabu Massif (see Fig. 1)—Basima (09^o46.60's, 150^o83.31'E), Sion (09^o44.80's, 150^o80.95'E) and Bosalewa (09^o44.60's, 150^o72.05'E)—and four communities in the lowlands, Salamo (09^o66.62's, 150^o79.39'E), Galubwa (09^o61.25's, 150^o78.16'E), Upper Momoawa (09^o55.98's, 150^o81.48'E) and Bibio on Sebutuia Bay (09^o57.54, 150^o86.35'E). In addition to visiting habitats around these communities, we conducted surveys at three camps in primary forest, each c.1-day's walk from the nearest community: on Mt Oya Tabu above Sion village (09^o45.54's, 150^o78.88'E; 1,110 m, 7–11 September); Kalatupe near Bosalewa (09^o45.35's, 150^o74.51'E; 525 m, 23–25 September); and on the Kwama River south of Bosalewa (09^o51.28's, 150^o75.84'E; 860 m, 26–28 September). These camps were on the northern, western and southern slopes of the Oya Tabu Massif.

Oya Tabu, also known as Mt. Kilkerran, is the largest and highest mountain on Fergusson Island (max. elevation 2,073 m) and forms part of a massif with several lower peaks such as Mt. Olaba on its southern edge. Vegetation on the massif constitutes a mosaic of forest fragmented by subsistence agriculture around villages and logging, to an elevation of c.300 m (see Fig. 2 for habitat photographs). Forest is mostly undisturbed above 600 m, and montane cloud forest is characterised by epiphytes, epiphylls and stands of bamboo at elevations in excess of 1,300 m.

To evaluate which species had previously been reported on Fergusson, we reviewed published records from the D'Entrecasteaux Archipelago (e.g., Beehler & Pratt 2016), unpublished resources (e.g., M. K. Tarburton's lists) and specimens in the American Museum of Natural History, New York.

Bird surveys.—Surveys were conducted on all 26 days we were present on Fergusson. At our forest camps, surveys commenced at dawn and continued until late morning, with additional outings in the late afternoon and evening. At sites around villages, we performed morning walks accompanied by local community members. Surveys were supplemented by opportunistic observations whenever possible. We recorded species using the eBird mobile app following the stationary (n = 26), travelling (n = 37) and incidental (n = 7) protocols (Sullivan et al. 2009). Locations and survey distances were recorded using the GAIA GPS application ( www.gaiagps.com) or a Garmin 60CSX GPS. We used a Sennheiser ME66 microphone and Marantz Solid State Recorder (PMD661) to record vocalisations, and DSLR cameras with telephoto lenses to photograph birds, when possible.

Figure 1.

Map of survey and camera trap locations and major villages where we based our operations on Fergusson Island, Milne Bay Province, Papua New Guinea. Some camera locations are jittered from their actual location to reduce overlap with bird survey points.

Camera traps.—We deployed eight camera traps along an elevational transect from 360 to 1,360 m on the northern slope of the Oya Tabu Massif above Sion village on 7–11 September, and 12 cameras along a similar elevational transect between 360 and 970 m above Basima village on 13–30 September. Cameras were placed a minimum of 200 m apart, alternating between Browning Strikeforce Pro BTC-5DCL (Birmingham, AL) and Reconyx Hyperfire H2X (Holmen, WI) models along the transect. During 26–28 September we deployed an additional eight cameras in primary forest between 550 and 980 m around our Kwama River camp. In all areas, camera traps were set in sites with little or no human disturbance, prioritising game trails and places with recently fallen fruit when possible. Cameras were mounted vertically on PVC poles, placed c.10 cm above ground, and set to high sensitivity and rapid-fire photo bursts. We also opportunistically deployed cameras near nests for up to 24 hours to confirm species identifications and record behaviour. For full details on camera trap methods, see Gregg et al. (2024).

Figure 2.

Representative images of habitats surveyed on Fergusson Island. (A) Fergusson Island from sea level to higher mountains, showing gradient of human-altered and primary forest ecosystems. (B) Human-modified habitat around villages with betel nut palms and ornamental plants. (C) Garden on the slopes of Mt. Oya Tabu with subsistence crops including banana, tapioca and taro. (D) Upper montane habitat on Mt. Oya Tabu, showing understory of ferns and trees blanketed in moss and epiphytes (John C. Mittermeier)

Results

We recorded 97 bird species during our surveys on Fergusson Island and estimated abundance for each based on the number of detections during audiovisual surveys and camera-trap hours in suitable habitat (see Appendix). Audiovisual surveys detected 95 species compared to eight species photographed by camera traps (Fig. 4). Pheasant Pigeon and Red-necked Crake Rallina tricolor were the only species recorded by the camera traps but not observed during the rest of the survey. Orange-footed Scrubfowl Megapodius reinwardt and Red-bellied Pitta accounted for most bird detections by the camera traps overall (36% and 28% of bird detections, respectively).

The majority (n = 72) of the 97 species we recorded are documented by media vouchers (photographs or sound recordings), all of which are archived and publicly available via our eBird trip report ( https://ebird.org/tripreport/103629). Comments on the endemic subspecies of Pheasant Pigeon, new island records for Fergusson, noteworthy breeding observations and vocalisations of endemic taxa are presented below. Taxonomy and nomenclature follow Beehler & Pratt (2016), which differs from taxonomy used by Macaulay Library where our supporting media is archived (Clements et al. 2024).

PHEASANT PIGEON Otidiphaps nobilis insularis

O. n. insularis is known only from Fergusson Island and is treated as a Critically Endangered species, Black-naped Pheasant-Pigeon, on the IUCN Red List and HBW/BirdLife Taxonomic Checklist (del Hoyo & Collar 2016, BirdLife International 2024b). Prior to our field work, this taxon was known from just three specimens: two collected by Goldie in 1882, which formed the type series, and a single collected by Meek in 1896 (Kirwan & van Grouw 2023, Kirwan et al. 2023). With no confirmed documentation in 126 years, this taxon was identified as a ‘lost’ bird species by the Search for Lost Birds project (Rutt et al. 2024) and provided the impetus for much of our survey efforts (see Gregg et al. 2024).

Figure 3.

Sonogram of recording (ML 610283282; Jason Gregg) of King Quail Synoicus chinensis at Bosalewa village, 22 September 2022; the first confirmed record of this species on Fergusson Island. Audacity v.3.4.2 was used to create the sonogram.

We detected the species using camera traps on three occasions in September 2022: a single individual photographed on the north slope of Oya Tabu above Basima (746 m) on 22 September, and what was presumably the same individual near our Kwama River camp on 27 and 28 September (971 m, Fig. 4H). The fact that we detected this species just twice despite significant search effort indicates that the population is both rare and extremely elusive, and suggests its Red List status of Critically Endangered is warranted.

The two locations where the bird was found were both in relatively undisturbed primary hill forest without any logging activity. Conversations with local residents, however, suggested that at least one of these areas could be designated for logging in the near future.

New distributional records for Fergusson Island

We recorded eight species that lack previously published records from Fergusson Island, including five without published records anywhere in the D'Entrecasteaux Archipelago (Fig. 5). These included two austral migrants to New Guinea (Little Egret Egretta garzetta and Australian Hobby Falco longipennis), three species that are probably resident in open and human-modified habitats (King Quail Synoicus chinensis, Intermediate Egret Ardea plumifera, Singing Starling Aplonis cantoroides) and one breeding forest-dependent species (Rallina tricolor).

(Online)

Figure 4.

Camera-trap images for: (A) Orange-footed Scrubfowl Megapodius reinwardt macgillivrayi, (B) Grey-headed Goshawk Accipiter poliocephalus, (C) Papuan Boobook Ninox theomacha goldii, (D) Red-bellied Pitta Erythropitta erythrogaster finschii, (E) Stephan's Emerald Dove Chalcophaps stephani, (F) Red-necked Crake Rallina tricolor, (G) Little Shrikethrush Colluricincla megarhyncha fortis, and (H) Pheasant Pigeon Otidiphaps nobilis insularis.

KING QUAIL Synoicus chinensis

On 22 September, JG flushed two quail from open grassland at sea level near Idubada station, Bosalewa community, and made an audio recording of the birds calling at close range (ML 610283282; see Fig. 3). Based on the recorded vocalisations, we identified the birds as King Quail. This species occurs in lowland habitats over most of New Guinea and in the Bismarck archipelago, but there were no previously published records from the D'Entrecasteaux Islands.

Figure 5.

Nest-building behaviour and nests of selected species: (A) Curl-crested Manucode Manucodia comrii comrii visiting nest and inset image of nest in building stage, (B–C) Oya Tabu White-eye Zosterops crookshanki collecting nesting material, (D) female Goldie's Bird of Paradise Paradisea decora carrying nesting material, and (E) female Goldie's Bird of Paradise on a nest in an epiphyte (arrow points to tail) (A: Doka Nason and B–E: John C. Mittermeier)

YELLOW-BILLED KINGFISHER Syma torotoro ochracea

We found Yellow-billed Kingfisher to be common at most of our survey sites in forest habitats from sea level near Salamo, Bibio and Bosalewa to the highest elevations we visited on Oya Tabu (c.1,800 m). S. t. ochracea has been suggested as a potential split from mainland populations based on differences in morphology, vocalisations and genetic divergence (Linck et al. 2020, Berryman in press). We recorded the calls of three individuals, on 9 September (ML 548009611 and ML 548010441) and 15 September (ML 513164643), nearly doubling known audio recordings of the taxon. Both the mainland (S. t. meeki) and D'Entrecasteaux taxa possess short and long versions of their main call (see Fig. 6): the former's call is a fast trill, whose pitch rises and falls once in the short version and twice in the long version. S. t. ochracea possesses a considerably slower trill, with a slight vibrato in each note that steadily decreases in pitch and accelerates during the final notes. Long and short versions differ only in that the former consists of a long repetition of one of the final notes in the sequence.

Figure 6.

Sonograms of Yellow-billed Kingfisher Syma torotoro vocalisations: (A) long version of mainland S. t. meeki call (XC 24796; Niels Krabbe), (B) short version of S. t. meeki call (XC 38235; Frank Lambert), (C) long version of D'Entrecasteaux endemic S. t. ochracea (ML 548009611; Jordan Boersma), a n d (D) short version of S. t. ochracea (ML 613164643; Jordan Boersma). Audacity v.3.4.2 was used to create the sonograms.

LITTLE SHRIKETHRUSH Colluricincla megarhyncha fortis

Like Erythropitta erythrogaster, this complex has been split into multiple species by most authorities, with the taxon we observed representing one of five subspecies under the name Variable Shrikethrush (Marki et al. 2018, Boles 2021). A recording made by JB on 9 September (ML 548004101) is the first published recording of D'Entrecasteaux C. f. fortis and differs considerably from the few known recordings of mainland C. f. despecta, the only one of the five C. fortis taxa that had previously been sound-recorded. Song of C. f. despecta differs qualitatively across the five publicly available recordings, but always includes at least one monotone introductory note, followed often by a downslurred or upslurred whistle (see Fig. 7). C. f. fortis song typically consists of a sequence of three distinct downslurred notes, the first two on approximately the same pitch and the last at higher pitch. Often the song alternates between the three-note sequence being of overall higher pitch followed by a sequence at lower pitch. Occasionally, raspy notes of rising and falling pitch precede the standard song.

Figure 7.

Comparison of sonograms of Little Shrikethrush Colluricincla megarhyncha taxa: (A) typical song (XC 279259; John V. Moore) and (B) adult plumage of mainland C. m. despecta (ML 575874601; © Dubi Shapiro), versus (C) representative song (ML 548004101) and (D) adult plumage of the D'Entrecasteaux endemic C. m. fortis (ML533344161; John C. Mittermeier). Audacity v.3.4.2 was used to create the sonograms.

Discussion

We found published records for 143 species on Fergusson Island and 160 species in the D'Entrecasteaux archipelago as a whole (Table 1). Our field surveys added another six species with no previously published records from Fergusson, increasing the total on the island to 149. Four of these species had no records from the D'Entrecasteaux group, increasing the total for the archipelago to 164 species. Combined with the study by Gregg et al. (2020), recent surveys on Fergusson have increased documented avian diversity on Fergusson by 8% and for the archipelago by 4%.

Among the six new island records, two are migrants (Australian Hobby and Little Egret), three are species of savanna, second growth near habitation and gardens near villages (King Quail, Intermediate Egret and Singing Starling), and one is an easily overlooked species found in lowland forest that we documented by camera trap (Red-necked Crake). Together these results highlight the value of surveying across seasons and habitats, as well as the benefit of employing multiple survey methods and landowner engagement. Two of the species we recorded (Intermediate Egret and Singing Starling), occurred only in the vicinity of villages and may constitute recent arrivals to Fergusson following human-led habitat change. The recent record of Eurasian Tree Sparrow by Gregg et al. (2020) also fits that pattern.

TABLE 1

Species list for the D'Entrecasteaux Archipelago, with presence/absence data indicated for each major island. Species names in bold involve new records for Fergusson Island and highlighted rows new records for the D'Entrecasteaux archipelago made during our 2022 survey. Local subspecies are indicated, where appropriate.

Continued

Continued

Continued

Continued

Camera traps permitted us to confirm the presence of two inconspicuous terrestrial species that had either not been recorded on the island previously or had not been documented for more than a century: Red-necked Crake and Pheasant Pigeon. This demonstrates that while camera traps may not add significantly to overall numbers of species detected during ornithological surveys, they can increase the detection of inconspicuous terrestrial birds, even when deployed for relatively brief periods of time. Both species we documented only by camera traps were also reported by local people who described the birds and suggested locations to set the cameras.

We regularly observed three of the four species endemic to the D'Entrecasteaux and Trobriand Islands EBA during our surveys: Curl-crested Manucode, Goldie's Bird of Paradise and Oya Tabu White-eye, but failed to find the fourth, Long-billed Myzomela. This poorly known species was originally described as endemic to montane forest on Goodenough Island and was first documented on Fergusson in 2019 (Gregg et al. 2020). While it appears to be uncommon on Fergusson, our failure to detect the species was probably due to the limited time we spent in forest above 1,000 m (four days total, including one with prohibitively poor weather conditions).

The avifauna of the D'Entrecasteaux archipelago faces threats from climate change and habitat destruction. Two of the regionally endemic taxa—Oya Tabu White-eye and Long-billed Myzomela—are restricted to the highest elevations and could possess reduced suitable habitat as global climate continues to warm (Freeman et al. 2018). Logging has already degraded much of the lowland and hill forest, and conversations with local landowners indicated that new operations are now targeting higher elevations not subject to prior disturbance. The most threatened bird on Fergusson is almost certainly the local subspecies of Pheasant Pigeon, which despite weeks of focused effort was only detected by two of our camera traps (see Gregg et al. 2024, for more details). This taxon is probably at risk from habitat degradation due to logging, but does not seem to face current pressure from local hunters due to its scarcity. Partnerships with local people were vital to the success of our field work, as they guided us and shared their immense knowledge of local fauna. These local natural historians are ideal partners for developing conservation priorities before species decline or disappear.